Figure 1. Possible presentations of incidentally discovered adrenal masses during imaging studies. The size of the ellipse is not representative of the likelihood of findings
 |
|
|
The Agency for Healthcare Research and Quality (AHRQ), through its Evidence-Based Practice Centers (EPCs), sponsors the development of evidence reports and technology assessments to assist public- and private-sector organizations in their efforts to improve the quality of health care in the United States. The reports and assessments provide organizations with comprehensive, science-based information on common, costly medical conditions and new health care technologies. The EPCs systematically review the relevant scientific literature on topics assigned to them by AHRQ and conduct additional analyses when appropriate prior to developing their reports and assessments.
To bring the broadest range of experts into the development of evidence reports and health technology assessments, AHRQ encourages the EPCs to form partnerships and enter into collaborations with other medical and research organizations. The EPCs work with these partner organizations to ensure that the evidence reports and technology assessments they produce will become building blocks for health care quality improvement projects throughout the Nation. The reports undergo peer review prior to their release.
AHRQ expects that the EPC evidence reports and technology assessments will inform individual health plans, providers, and purchasers as well as the health care system as a whole by providing important information to help improve health care quality.
We welcome written comments on this evidence report. They may be sent to: Director, Center for Practice and Technology Assessment, Agency for Healthcare Research and Quality, 6010 Executive Blvd., Suite 300, Rockville, MD 20852.
Carolyn M. Clancy, M.D.
Acting Director
Agency for Healthcare Research and Quality
Robert Graham, M.D., Director
Center for Practice and Technology Assessment
Agency for Healthcare Research and Quality
The authors of this report are responsible for its content. Statements in the report should not be construed as endorsement by the Agency for Healthcare Research and Quality or the U.S. Department of Health and Human Services of a particular drug, device, test, treatment, or other clinical service.
Objective. This evidence report on the Management of the Clinically Inapparent Adrenal Mass was requested by the Office of Medical Application and Research (OMAR) at the National Institutes of Health for a State-of-the-Science Conference. The widespread use of computed tomography (CT), magnetic resonance imaging (MRI), and ultrasound imaging (US) has resulted in the incidental discovery of asymptomatic adrenal masses (also referred to as incidentaloma) that has created a management dilemma.
Search Strategy. Studies for the review of the primary literature were primarily identified through searches of the English language literature published between 1966 and March 2001 in Medline, BIOSIS, and Embase.
Selection Criteria. Adrenal incidentaloma can occur in a variety of situations and authors have used a wide range of definitions for this condition. Because incidentaloma is not a specific disease entity, we accepted all studies that used their own definitions. Articles were selected for their relevance to answering five key questions.
The literature search yielded over 5,000 independent citations. We evaluated over 600 articles and included about 200 English language articles in the evidence report. Specific inclusion criteria and methods of synthesis were developed for each of the key questions. Relevant data from each article were abstracted into evidence tables. Information from the evidence tables was synthesized into summary tables describing the findings of each study.
Forty-five studies provided data about the prevalence of incidentaloma or the distribution of adrenal pathologies. Thirty-two studies evaluated diagnostic tests to differentiate adrenal masses. Over 80 studies provided outcome information on adrenal surgery techniques. Thirty-two studies reported prognostic information on patients with adrenal carcinoma after surgical excision, and nine articles reported follow-up results of untreated incidentaloma. Most of these studies were retrospective and the overall methodological quality was low.
Because incidentaloma is not a disease entity, the prevalence of incidentaloma will vary with the circumstances. One study used transabdominal ultrasound for general health examination reported 11 verified adrenal masses out of 41,357 subjects (prevalence 0.027 percent). Combining individual cases from several studies, the prevalence of adenoma among incidentaloma was 41 percent, metastases 19 percent, adrenocortical carcinoma 10 percent, myelolipoma 9 percent, and pheochromocytoma 8 percent, and other mostly benign lesions comprised the remainder of the lesions. The distribution of pathology varies with the definition of incidentaloma applied.
Most of the diagnostic studies were conducted for the purpose of diagnosing adrenal malignancy. In general, unenhanced or immediate enhanced CT had fair test characteristics. Delayed enhanced CT or MRI using the mass to spleen ratio had excellent test performance. Combined unenhanced CT with MRI improved the accuracy in one study. Scintigraphy had fair to excellent sensitivity and specificity. Fine needle aspiration had good to excellent test performance; however, inconclusive biopsies were common. In one small study, DHEAS had perfect sensitivity, but poor specificity.
The clinical outcomes of various surgical adrenalectomy approaches were reassessed. Evidence suggests that laparoscopic adrenalectomy in general results in less blood loss and fewer major complications. The optimal surgical approach may depend on the size and the type of tumor.
Limited follow-up data on patients with untreated incidentaloma found most tumors remain unchanged in size, some tumors disappeared or decreased in size, and about one-sixth of the tumors increased in size but none of these were adrenocortical carcinoma.
With few exceptions, the overall methodological quality of the studies we examined was low. Future studies of incidentaloma need to broadly cover diverse manifestations of this condition but individual studies should apply rigorous inclusion criteria for each of the manifestations or provide thorough descriptions and careful analyses of well-defined subgroups. Future studies should incorporate rigorous methodologies to properly assess the clinical usefulness of various diagnostic tests and follow-up strategies for adrenal incidentalomas. An international registry of patients with well-documented adrenal incidentaloma may provide the best means of collecting these data.
This evidence report on the topic of Management of the Clinically Inapparent Adrenal Mass was produced on request from the Office of Medical Application and Research (OMAR) at the National Institutes of Health for a State-of-the-Science Conference.
The widespread use of computed tomography (CT), diagnostic ultrasound imaging (US), and magnetic resonance imaging (MRI) has resulted in the incidental discovery of asymptomatic adrenal masses (also referred to as incidentaloma). While there is little debate about the management of large adrenal masses, there is disagreement over the proper management of adrenal masses smaller than 6 cm.
Adrenal incidentaloma is not a single pathological entity. Adenoma, adrenocortical carcinoma, pheochromocytoma, metastases, myelolipoma, and cysts may all present as incidentalomas. The likelihood of any specific pathology depends both on the circumstances of discovery and the definition of incidentaloma. Depending on the presence or absence of known cancer, patient age, indication for imaging, presenting signs and symptoms, and the performance characteristics of the initial test, different pathologies will be more or less likely. Mass size and appearance on imaging further modify this likelihood. Appropriate management depends on understanding the influence of these factors on disease prevalence.
Modalities for diagnosing adrenal masses include CT, MRI, transcutaneous needle biopsy (TNB), radioisotope imaging, and biochemical tests. Each of these has an associated degree of diagnostic accuracy as well as risk.
Management of adrenal incidentaloma is a complex decision making process that involves considering a range of possible diagnoses and their natural history, diagnostic testing and interpretation, and weighing the risks and benefits of interventions in light of the patient's age and the tumor size. While some authors have advocated surgical removal of all incidentally discovered adrenal masses, because complete and early removal of small, potentially cancerous lesions could be life saving, removal of all such lesions would result in unnecessary surgery for most patients. This is further complicated by recent studies reporting that some adenomas are subclinical biochemically active and lead to long-time morbidities.
The following key questions were formulated by the State-of-the-Science conference planning committee, composed of staff from OMAR, National Institute of Child Health and Human Development, National Cancer Institute, national and international experts on this topic, and Evidence-based Practice Center (EPC) staff.
Question 1. What are the causes and prevalence of clinically inapparent adrenal masses identified by CT, MRI, or ultrasound that are confirmed by histology? What are the prevalence rates for the various causes of inapparent adrenal masses? Are there differences in the rates among the initial diagnostic tests used? What is the relationship of the age and sex of the patient to the likelihood of having a particular pathology? What is the relationship of the size of the mass with the likelihood of having a particular pathology?
Question 2. What is the diagnostic accuracy (sensitivity, specificity) of evaluation modalities (fine needle aspiration/biopsy, CT, MRI, US, biochemical tests) used to differentiate adrenal masses (adrenal carcinoma, pheochromocytoma, adenoma, adrenal hyperplasia, etc.)? What is the risk of metastatic spread of adrenal carcinoma by fine needle aspiration (FNA)?
Question 3. What are the surgical complication rates for various approaches used to excise adrenal masses; specifically laparoscopic, transabdominal, and retroperitoneal approaches?
Question 4. What are the patient outcomes after surgical excision of adrenocortical carcinoma (morbidity and mortality)? Are there data on the influence of age and tumor size on the outcomes?
Question 5. What evidence is there to support the use of periodic biochemical and imaging studies to follow untreated adrenal masses?
Authors have used a wide range of definitions for incidentaloma. Rather than choosing a narrow definition that may exclude potentially relevant situations and many relevant studies, we accepted all studies that used their own definitions of incidentaloma.
Studies were identified primarily through a Medline search of the English language literature published between 1966 and October 2000. We also consulted technical experts and examined references of selected review articles to identify additional studies. Articles that met the inclusion criteria were incorporated into our evidence report. The staff at the National Library of Medicine conducted a search in October 2000 on the following databases: Medline, PreMedline, BIOSIS, and Embase. The content of the original key questions was later expanded, and we conducted an updated search in March 2001 in the Medline database. Additional subject headings were included to address questions on diagnostic accuracy, surgical complication rates, morbidity and mortality outcomes for adrenal masses, and monitoring strategies for untreated adrenal masses.
The literature searches yielded a total of 5,386 independent citations. The abstracts were screened and reports published only as letters or abstracts in proceedings were rejected from further consideration. We retrieved 602 articles. Specific inclusion criteria and methods of synthesis were developed for each of the key questions. In general, we included all English language studies with at least 10 human subjects. There were no age limitations.
About 194 articles met our inclusion criteria for one of the key questions and were included in the evidence report. Forty-five studies provided data about the prevalence of incidentaloma or the distribution of adrenal pathologies. Thirty-one studies evaluated various diagnostic tests to differentiate adrenal masses. Over 80 studies provided outcome information on various adrenal surgical techniques. Thirty-two studies reported prognostic information on patients with adrenal carcinoma after surgical excision, and nine articles reported results of follow-up strategies.
From these articles, we abstracted detailed information into evidence tables and summary tables. We also assessed the methodological quality of studies that evaluated diagnostic performance and adrenal surgery. Using a three-category scale, we graded these articles based on study design, conduct, and reporting. We also assessed the applicability of studies to the population of interest as determined by study location, tumor size and tumor type.
There are over 44 reports from various countries describing the causes and prevalence of various pathologies found in incidentaloma. Because incidentaloma is not a disease entity, the prevalence of incidentaloma will vary with the circumstances. We found one study that used transabdominal ultrasound for general health examination reported 11 adrenal masses (verified clinically or with pathology) out of 41,357 subjects (prevalence 0.027 percent). The prevalence of incidentaloma has been reported to be 0.6 percent in a study of 2200 patients undergoing upper abdominal CT. Combining the studies that used the broadest definitions and that reported individual cases, the prevalence among incidentaloma was as follows: adenoma 41 percent, metastases 19 percent, adrenocortical carcinoma 10 percent, myelolipoma 9 percent, pheochromocytoma 8 percent, with other, mostly benign lesions comprising the remainder. This distribution is similar to that reported in a single large study. Sixty percent of the incidentalomas occurred between the ages of 41 and 60, and 90 percent between the ages of 31 and 70.
For tumors 4 cm or less, the data are limited. Sixty-five percent were adenomas and approximately 21 percent were metastases. For tumors above 6 cm, adrenal carcinomas comprised to 25 percent and adenomas only 18 percent. The percentage of metastases was 18 percent. Approximately 64 percent of the adenomas and 70 percent of the adrenal carcinomas were found in females, whereas 60 percent of metastases were in males.
We identified 32 studies that met the search criteria for this report and reported diagnostic performance of tests to evaluate adrenal masses. Four studies included only subjects with adrenal incidentalomas. Overall, about one-third of the subjects were reported to have incidentalomas. Study quality was as follows: one good, 16 fair, and 15 poor.
Ten studies (N = 467) evaluated CT. Four studies of unenhanced CT to diagnose adrenal cancer found high sensitivity and moderate to high specificity at density thresholds below 18 Hounsfield units. Three studies of enhanced CT found that high sensitivity could be achieved only with low specificity, and vice versa. Two studies of delayed enhanced CT found high sensitivity (83 to 100 percent) and high specificity (91 to 100 percent). One of three studies of combined unenhanced and enhanced CT found excellent test performance; the other two found poor test performance. Nine studies of size and other morphological features on various types of CT found overall poor test performance. Three studies concluded that on unenhanced CT, attenuation performs better than morphological parameters. Two other studies of enhanced CT concluded that attenuation and size perform similarly, and both were worse than delayed enhanced CT.
Eight studies (N = 302) evaluated unenhanced MRI in diagnosing adrenal cancer. They found sensitivity of 50 to 100 percent and specificity of 24 to 100 percent. Four studies found the best test performance using adrenal mass to spleen signal intensity ratio. Sensitivity was 95 to 100 percent and specificity was 78 to 100 percent.
One study (N = 33), which combined unenhanced CT and MRI to diagnose adrenal cancer, reported sensitivity of 100 percent and specificity of 94 percent. Six studies (N = 387) using different parameters to categorize scintigraphy to diagnose adrenal cancer, reported sensitivity of 71 to 100 percent and specificity of 50 to 100 percent. One study (N = 54) of ultrasound to diagnose adrenal cancer found sensitivity of 67 percent and specificity of 52 percent. One study (N = 20) of positron emission tomography to diagnose adrenal cancer found sensitivity and specificity of 100 percent.
Nine studies (N = 515) of FNA to diagnose adrenal cancer found sensitivity of 81 to 100 percent and specificity of 83 to 100 percent. Inconclusive biopsies occurred in 6 to 50 percent of samples.
Three studies evaluated biochemical tests to diagnose adrenal cancer in patients with adrenal incidentaloma. One study (N = 84) found that elevated or normal morning dihydro-epiandrostenedione sulfate level had sensitivity of 100 percent and specificity of 38 percent.
Twelve studies reported on 888 biopsies of adrenal masses in 866 patients. Only two studies explicitly reported data on the risk of metastatic spread due to biopsy. One patient (0.3 percent of subjects with long term follow-up) with bronchogenic carcinoma had needle tract metastasis after adrenal biopsy. Thirty-six complications (4 percent) were reported, including 26 that were potentially serious. Because of the wide variety of biopsy techniques, unclear or incomplete reporting, and the small study sizes, no reliable estimates can be made about the relative safety of the different biopsy techniques.
We evaluated nine case series of open adrenalectomy, four studies comparing open adrenalectomy techniques, 20 series of transperitoneal laparoscopic adrenalectomy, 10 series of retroperitoneal laparoscopic adrenalectomy, 28 studies comparing laparoscopic adrenalectomy to open surgery, and nine studies comparing transperitoneal with retroperitoneal laparoscopy. Overall study quality was poor.
The open series included 525 patients. Major complications occurred in 2 to 24 percent of patients and minor complications in 0 to 14 percent. Mortality occurred in 0 to 3 percent of patients. The most common complications included pleural tear, wound infection, bleeding, splenectomy and urinary tract infection. Four studies comparing the open transabdominal and posterior approaches in 566 patients found similar complication rates with both approaches, but average length of stay was shorter with the posterior approach.
The transperitoneal laparoscopic series included 1,189 patients. Major complications occurred in 0 to 25 percent of patients and minor complications occurred in 0 to 72 percent. Two deaths were reported. The most common complications were bleeding, wound infection and deep vein thrombosis. The retroperitoneal laparoscopic series included 537 patients. Major complications occurred in 0 to 10 percent of patients and minor complications occurred in 0 to 63 percent. One patient died. The most common complications included retroperitoneal hematoma, subcutaneous emphysema, and pancreatic or splenic injury. Complications were less common in larger studies.
Studies comparing open and laparoscopic surgery included 1388 patients. Most studies did not apply statistical methods to analyze complication rates. Five studies found laparoscopy to have fewer complications, while one found no difference. Studies comparing different laparoscopic approaches did not find any statistical differences in complications, operating time or blood loss.
Thirty-two studies with a total of 1,684 patients were examined to gather information about patient outcomes after surgical excision of adrenocortical carcinoma. Most of these data came from retrospective studies. There were wide variations in the quality and quantity of reported information about the tumor size, patient characteristics, surgical approaches, and outcomes. Nine studies reported patients who had surgery in the 1940s and 1950s. One study went back to 1929. Fifteen of the 32 studies reported perioperative mortality data. Twenty perioperative deaths were reported out of a total of 625 patients, a rate of 4.6 percent.
There were diverse methods of reporting the long-term survival. Seventeen studies reported 5-year survival data that ranged from 19 to 62 percent with a median of 34 percent (weighted average = 34.6 percent). There does not appear to be any important difference in the overall survival rates between the earlier and the more recent series. Most of the studies included patients over a wide range of years, making it difficult to discern any trend over time. Age and tumor size did not appear to influence prognosis after surgery. Men had worse outcomes: in the three studies that reported individual patient data, out of 18 men, only 3 survived more than 24 months, and none survived to 5 years. However, men tended to have nonfunctional tumors, which are typically asymptomatic until a late stage.
There were four prospective studies of incidentaloma with established follow-up protocols for untreated patients. Two studies were clearly prospective, one was ambiguous, and one study identified patients retrospectively from medical records but prospectively followed the patients who gave consent. Two studies did not report exclusion criteria. The other two excluded patients with hormonally active and/or malignant tumors, and one excluded cysts as well. Overall, of 121 patients followed, 78 had no changes, 20 had increases in tumor size, and 11 had either decreases in tumor size or disappearance of the tumor. There were two cancer deaths unrelated to adrenal pathology. Adenoma was confirmed at autopsy.
Adrenal incidentaloma is neither a single pathological entity nor a disease. A strict definition of incidentaloma would aid in the interpretation of results from clinical studies; however, it will not be sufficient to address the diverse manifestations of this condition that are also clinically relevant. Future studies of incidentaloma need to broadly cover all possible scenarios but individual studies should apply rigorous inclusion criteria for each of these scenarios or provide careful analyses of well defined subgroups.
Many studies assumed that the major reason for the further evaluation of adrenal incidentalomas is for the purpose of detecting adrenal carcinoma. Given the rarity of this tumor and the lack of effective therapy in the later stages, the overall benefit of detection is small. On the other hand, subclinical, biochemically active adrenal adenomas are common and need to be better understood regarding their prevalence and long-term clinical outcomes. Given their prevalence and the significance of hypertension and diabetes, these tumors may become the reason for aggressive intervention in adrenal incidentaloma.
With few exceptions, the overall methodological quality of the studies we examined was poor. Higher quality studies are needed to properly assess the clinical usefulness of various diagnostic tests for adrenal incidentalomas. In contrast to published studies, future studies should make a rigorous attempt to clearly define and report eligibility criteria, sample characteristics, test methodology, and the definitions of positive and negative tests. Reference standards should be clearly delineated, completely independent of the tests being investigated, and as much as possible, include well defined outcomes, such as surgical or autopsy diagnoses.
Future research should concentrate on defining the best procedure for each indication, as well as identifying other factors, such as the number of operations performed by the surgeon, which may contribute to operating complications. In order to compare across series, it would be useful for authors to agree on standard definitions of such seemingly simple outcomes as operating time or post-operative length of stay. It would also be particularly useful to standardize the measurement and reporting of complications.
There is very sparse data to guide the management of untreated, incidentally discovered adrenal masses. Most of the available studies are either too small to provide meaningful results or suffer from methodological problems. Future studies should include prospective application of pre-specified protocols to clearly defined populations. Given the rarity of adrenocortical carcinoma, assessing the usefulness of follow-up protocols will require a large number of subjects. If recent studies are confirmed, follow-up strategies that include biochemical evaluations for subclinical biochemically active adrenal masses will be useful.
Well-designed clinical trials will provide the most reliable evidence regarding the management of these patients, but these trials will take many years and may not always be feasible. An alternative approach would be to create an international registry of patients with well-documented adrenal incidentaloma. This registry should be established using standardized definitions and inclusion criteria. Investigators could then analyze individual patient data to understand the influence of specific factors such as tumor size, age, and inclusion criteria on the development of subsequent adrenal pathologies.
The Office of Medical Applications of Research (OMAR) at the National Institutes of Health requested that the Agency for Healthcare Research and Quality (AHRQ), through its Evidence-based Practice Center (EPC) program, produce an evidence report for a State-of-the-Science conference on the topic of Management of the Clinically Inapparent Adrenal Mass.
EPCs generally review relevant scientific literature on assigned clinical care topics and produce evidence reports and technology assessments, conduct research on methodologies and the effectiveness of their implementation, and participate in technical assistance activities. The purpose of evidence reports is to search for and summarize evidence on several key questions on a specific topic. It is usually beyond the scope of an evidence report to comprehensively cover all relevant questions on a single topic. EPCs collaborate with science partners to formulate specific key questions. As specified by AHRQ in the EPC program, evidence reports do not make specific clinical recommendations, however, recommendations for future research are typically provided. Public and private sector organizations may use the reports and assessments as the basis for their own clinical guidelines and other quality improvement activities.
This evidence report summarizes the evidence on several key questions on the management of the clinically inapparent adrenal mass. These masses are also commonly known as “incidentaloma” in the medical literature. These two terms will be used interchangeably in this report. The key questions were formulated by the State-of-the-Science Conference planning committee composed of staff from OMAR, National Institute of Child Health and Human Development, National Cancer Institute, national and international experts on this topic, as well as the EPC staff. Anticipating that most of the data available to address the key questions will come from case series and retrospective cohorts, the EPC was also asked to develop a discussion paper on the biases of case series and the potential hazards in making recommendations from case series reports (Appendix C).
The widespread use of computed tomography (CT), diagnostic ultrasound imaging (US), and magnetic resonance imaging (MRI) has resulted in the incidental discovery of asymptomatic adrenal masses. While there is little debate about the management of large adrenal masses, there are disagreements in the recommendations for the management of small adrenal masses 1 to 6 cm in size.
A major concern for investigating small adrenal masses is the detection and treatment of early adrenal carcinoma. A highly lethal disease, adrenal carcinoma is nevertheless rare, and the search for it must balance the morbidity and mortality resulting from the search against the possible improvement in the outcome of proven carcinoma.
Most reported adrenal carcinomas are large and patients are often symptomatic. The overall prognosis is poor even with surgical removal (Pommier and Brennan, 1992). There is no current effective chemotherapy. Some practitioners have advocated surgical removal of all incidentally discovered adrenal masses, the rationale being that complete and early removal of small potentially cancerous lesions could be life saving (Seddon, Baranetsky, and Van Boxel, 1985). However, removal of all such lesions would result in unnecessary surgery for most patients.
Many approaches have been proposed to manage patients with adrenal incidentalomas. These proposals include repeated CTs at intervals of 2, 4, 6, 12 and 18 months to detect changes in lesion size; needle biopsy, MRI, and radioisotope imaging have been proposed to distinguish adenoma and pheochromocytomas (Abecassis, McLoughlin, Langer, et al., 1985; Belldegrun, Hussain, Seltzer, et al., 1986; Bernardino, 1988; Copeland, 1983; Geelhoed and Druy, 1982; Glazer, Weyman, Sagel, et al., 1982; Guerrero, 1985; Katz and Shirkhoda, 1985; Mitnick, Bosniak, Megibow, et al., 1983; O'Leary and Ooi, 1986; Prinz, Brooks, Churchill, et al., 1982; Seddon, Baranetsky, and Van Boxel, 1985; Thompson and Cheung, 1987; Wood, Delbridge, and Reeve, 1987). Many of these recommendations appear to reflect subspecialty selection bias and the personal experiences of the authors. Several recent studies suggest that removal of small incidentalomas is unnecessary and that repeat CT scanning is of limited benefit (Barry, van Heerden, Farley, et al., 1998; Barzon, Scaroni, Sonino, et al., 1999).
Recent studies have reported that some of these adrenal adenomas may have subtle cortisol production and abnormalities in the hypothalamus-pituitary-adrenal axis. The patients with these tumors have subclinical Cushing's Syndrome and are associated with hypertension and other metabolic derangements (Rossi, Tauchmanova, Luciano, et al., 2000). Estimates of the prevalence of this condition has ranged from 5 to 16 percent. Verification of these findings in future studies will add another layer of complexity to the management of this condition.
Adrenal incidentaloma is generally defined as unexpected adrenal masses seen on diagnostic imaging studies (Reinig, Doppman, Dwyer, et al., 1986) performed for purpose other than imaging the adrenal glands, in patients without signs and symptoms of adrenal pathologies. However, varying definitions in the literature could lead to different meaning and interpretations of the data. For example, Aso and Homma (1992) used the following definition: “Incidental adrenal tumors were defined as adrenal tumors that were detected incidentally primarily as adrenal lesions on occasions unrelated to tumor signs and symptoms. Under the definition, previously unsuspected adrenal masses found at a routine health examination, or during diagnostic or therapeutic procedures for unrelated diseases were considered to be incidental adrenal tumors even if the patient retrospectively proved to have had symptoms or signs attributable to an adrenal tumor.”
Including patients with signs and symptoms attributable to an adrenal tumor will increase the proportion of large masses, which are more likely to be cancerous. Including symptomatic, biochemically active tumors will increase the proportion of pheochromocytomas. Conversely, studies that exclude patients with signs or symptoms will find a greater proportion of small masses and biochemically silent tumors. Some studies categorize adrenal tumors discovered as part of the workup for abdominal pain as incidentalomas. While adrenocortical carcinoma may not have been suspected initially in patients with abdominal pain, clearly this symptom is a potential manifestation of this tumor, especially when it is large.
Adrenal incidentaloma is not a single pathological entity. Adrenal adenoma, adrenocortical carcinoma, pheochromocytoma, cyst, metastasis from primary cancers, and various benign tumors may all present as incidentalomas. The underlying distribution of adrenal pathology in incidentaloma is influenced by a number of factors. These include the history or the presence of known cancer, age, reason for performing the test that led to discovery of incidentaloma, presenting signs and symptoms, and the performance characteristics of the test itself. The likelihood of any particular pathology is further modified by the findings on the imaging study (mass size, appearance) and biochemical activity of the tumor. Appropriate management depends on understanding these factors and how they influence the probability that the incidentaloma is a particular tumor. Unfortunately, published reports are inconsistent in applying inclusion and exclusion criteria for these various factors, making the results difficult to interpret. For example, authors have both included and excluded biochemically active tumors in their published reports. Obviously, those who exclude biochemically active tumors report very few of these tumors as a cause of incidentaloma.
Figure 1
illustrates the effect of
applying various commonly considered criteria to define adrenal
incidentaloma. By categorizing patients according to recent history of
cancers known to have the potential of metastasizing to the adrenals, signs
and symptoms of adrenal biochemical activity, and signs and symptoms
attributable to adrenal diseases (abdominal enlargement or pain), we form
eight different categories of patients with a different distribution of
adrenal pathologies. Category H has the strictest inclusion criteria and
represents the purest definition of incidentaloma. An example from this
category might be an incidentaloma found in a patient who underwent a chest
CT for solitary pulmonary nodule that turns out to be a benign granuloma.
While the probability of the adrenal mass being a metastasis is relatively low in patients without a known cancer, the probability could be quite high in breast cancer patients (Abrams, Spiro, and Goldstein, 1950). This group of patients belongs to category D.
would generate other distributions of
adrenal pathologies. Published clinical studies often used variations of
inclusion and exclusion criteria that result in mixtures of different
categories, further confounding the interpretation of prevalence data. The
retrospective nature of most of the published studies makes interpretation
even more difficult.Adenomas comprise the vast majority of incidental asymptomatic adrenal masses. Pheochromocytomas, metastases to the adrenal gland, adrenocortical carcinomas, and lesions such as myelolipomas and adrenal cysts make up the rest. Adenomas are benign, and there is no evidence that they degenerate into malignant lesions. The adrenal glands are frequent sites for metastases from many cancers, including lung and breast cancer. Most of the time either a primary site is obvious, or widespread disease is apparent. Occasionally, an adrenal mass may present as a metastatic cancer of unknown primary. These tumors will not respond to surgical removal and should be treated with systemic therapy based on the origin of the primary. Myelolipomas and cysts are benign lesions that usually require no therapy.
Adrenocortical carcinoma is rare. The incidence ranges from 0.6 to 2 cases per 1 million people (Nader, Hickey, Sellin, et al., 1983). A review published in 1954 of autopsies from 1918 to 1947 at the Los Angeles County Hospital found adrenocortical carcinoma to account for 0.2 percent of all known malignancies (MacFarlane, 1958; Steiner, 1954). Adrenocortical carcinoma can be functional or nonfunctional and the definition varies in the literature. Some authors require a clinically apparent endocrine syndrome, while others will classify a tumor as functional if biochemical activity can be demonstrated in serum or urinary assays. Using the clinical definition, functional tumors accounted for 24 to 46 percent of adrenocortical carcinoma. One series of 34 cases found 53 percent of them to be biochemically functional (Huvos, Hajdu, Brasfield, et al., 1970). When data from six different series are combined, 95 of 237 patients (Ribeiro, Sandrini Neto, Schell, et al., 1990) have functional tumors. Several authors have noted that functional tumors occur more frequently in women (Didolkar, Bescher, Elias, et al., 1981; Hajjar, Hickey, and Samaan, 1975; Hogan, Gilchrist, Westring, et al., 1980; Huvos, Hajdu, Brasfield, et al., 1970; MacFarlane, 1958; Sullivan, Boileau, and Hodges, 1978). One series of twenty-one cases of adrenocortical carcinoma reported that anaplastic adrenocortical carcinomas were nonfunctional, while many of the differentiated tumors were functional. The nonfunctional undifferentiated tumors had a poor prognosis with survival time less than one year in all cases. The only effective treatment is en bloc surgical resection of the cancer before it metastasizes. In addition, nonfunctional tumors did not respond to chemotherapy with op'DDD (Hogan, Gilchrist, Westring, et al., 1980)
While adrenocortical carcinomas are rare, benign adenomas are common, and their prevalence increases with age (Commons and Callaway, 1948). The true incidence of adrenal adenomas is difficult to determine, but several large autopsy series have found adrenal adenomas larger than 2 to 5 mm in 1.5 to 8.7 percent of the population. In the only prospective study, it was found that adenomas with a size greater than 3 mm occur in 8.7 percent of 739 autopsies (Hedeland, Ostberg, and Hokfelt, 1968). Most of these masses are small, and in a large series of 12,000 autopsies only 3 were 6 cm or larger (Copeland, 1983). In a review of 98 patients with suspected adrenal disease, CT scans showed that adenomas ranged in size from 1.4 to 9 cm, with a mean of 3.3 cm (Adams, Johnson, Rickards, et al., 1983).
Pheochromocytomas occur much less frequently than adenomas. This type of tumor secretes vasoactive substances and can lead to significant morbidity and mortality. Between 10 and 13 percent of pheochromocytomas are malignant (ReMine, Chong, van Heerden, et al., 1974; Sutton, Sheps, and Lie, 1981). In a review of 40,078 autopsies at the Mayo Clinic between 1928 and 1977, pheochromocytoma was found in 54 patients (0.13 percent) (Sutton, Sheps, and Lie, 1981). Ninety percent of these tumors are located in the adrenal glands, and the remaining 10 percent are located in the para-aortic sympathetic chain, aortic bifurcation, and wall of the urinary bladder. Diagnosis is made by biochemical testing, which will identify most pheochromocytomas. The false negative rate of urinary vanillylmandelic acid (VMA) is 30 percent, metanephrines 10 percent to 20 percent, and catecholamines 5 percent (ReMine, Chong, van Heerden, et al., 1974).
Adrenal glands are frequent sites for metastases. Carcinoma of the lung, breast, and lymphoma account for a large portion of adrenal metastases. In a review of 1,000 consecutive autopsies of patients with carcinoma, the adrenal glands were involved in 27 percent of the cases (Abrams, Spiro, and Goldstein, 1950). The incidence of adrenal metastases in patients with breast and lung cancer is approximately 39 and 35 percent, respectively (Abrams, Spiro, and Goldstein, 1950; Lumb and Mackenzie, 1959).
CT is an accurate tool for detecting the presence of adrenal masses. Asymptomatic adrenal masses were identified in 14 of 2,200 upper abdominal CT scans (0.6 percent) (Glazer, Weyman, Sagel, et al., 1982). Using a fast scanner and 1 cm scanning intervals, both adrenals can be identified in 97 to 99 percent of patients. Size has been reported to be the most reliable way to distinguish benign adenomas from adrenocortical carcinomas. Adrenal adenomas are often small, well defined, homogeneous lesions that do not enhance on CT, and are believed to remain constant in size on serial CT scans (Hussain, Belldegrun, Seltzer, et al., 1986; Mitnick, Bosniak, Megibow, et al., 1983). Adrenocortical carcinomas are usually large, dense, irregular, heterogeneous, enhancing lesions that invade other structures (Hussain, Belldegrun, Seltzer, et al., 1986). However, small masses, in the range of 1 to 6 cm, may be difficult to discriminate (Fishman, Deutch, Hartman, et al., 1987; Hussain, Belldegrun, Seltzer, et al., 1985; Mitnick, Bosniak, Megibow, et al., 1983)
MRI has been used to differentiate among adenomas, metastases, and pheochromocytomas. Both T1 and T2 relaxation times have been studied. Signal intensity ratios between the adrenal mass and various organs, including spleen, fat, liver, and muscle, have been tested to discriminate adrenal masses. In general, malignant masses are denser than benign masses, though various benign lesions can mimic malignancies (Baker, Spritzer, Blinder, et al., 1987; Chezmar, Robbins, Nelson, et al., 1988; Reinig, Doppman, Dwyer, et al., 1986). There is still debate over the best method to discriminate adrenal masses using MRI.
Transcutaneous needle biopsy, or fine needle aspiration (FNA) of adrenal mass has been advocated by some for the investigation of incidentally discovered adrenal masses (Bernardino, 1988). Needle biopsy is an accepted procedure in evaluating adrenal masses in patients with known carcinoma. The biopsy is generally performed under either CT or US guidance. While accuracy appears to be high, up to 15 percent of biopsies are inconclusive (Karstrup, Torp-Pedersen, Nolsoe, et al., 1991). Complications of adrenal mass needle biopsy include pneumothorax, bleeding, and bacteremia (Bernardino, 1988; Pagani and Bernardino, 1982). Rare instances of metastatic seeding of the cancer along the needle track have been reported (Mody, Kazerooni, and Korobkin, 1995).
Three radioisotopes, 131I-iodomethyl-norcholesterol (NP-59), 6-methyl-75-selenomethyl-19-norcholesterol, and 123I-meta-iodobenzylguanidine (MIBG) scintigraphy have been used for imaging of adrenal morphology and function. Various methods of analyzing scintigraphy have been used to differentiate adrenal masses, including relative uptake of marker, concordance with CT, and imaging patterns (Dominguez-Gadea, Diez, Bas, et al., 1994; Gross, Shapiro, Francis, et al., 1994; Nakajo, Nakabeppu, Yonekura, et al., 1993). In addition to its potential role in diagnosing malignancy, scintigraphy may also be capable of differentiating autonomously secreting adenomas from non-functioning adenomas, adrenal hyperplasia and other adrenal diseases (Gross, Shapiro, Francis, et al., 1994).
Biochemical tests, including dehydroepiandrosterone sulfate (DHEAS), urine cortisol, plasma catecholamines, and others are used to differentiate metabolically active from inactive adrenal disease (Bardet, Rohmer, Murat, et al., 1996; Bondanelli, Campo, Trasforini, et al., 1997). Biochemical tests to diagnose malignant incidentalomas are rare (Terzolo, Ali, Osella, et al., 2000).
Surgery with en bloc resection of the tumor is the only chance for a cure. Reported mortality of adrenal carcinoma surgery ranges from 0.8 to 9 percent (Didolkar, Bescher, Elias, et al., 1981; Mitnick, Bosniak, Megibow, et al., 1983; Wood, Delbridge, and Reeve, 1987), and may depend on the stage of disease (Henley, van Heerden, Grant, et al., 1983). In a review of 315 adrenal surgery patients from 1970 to 1979 at the Mayo Clinic, an overall operative morbidity was 14 percent. Morbidity includes wound infections, pneumothorax, pneumonia, hemorrhage, thromboembolic events, gastrointestinal hemorrhage, sepsis, small bowel obstruction, and wound dehiscence.
It has been suggested that tumor size may not correlate with survival, and may actually be inversely related to survival. However, these patients all presented symptomatically. It is not surprising that large low grade malignancies that present as masses can at times be surgically resected, and that smaller high grade symptomatic lesions may be unresectable. Since size, stage, and resectability are often related, extrapolation of this data from symptomatic masses to incidentalomas may not be valid. One study reported that when patients were not candidates for surgery, more than 70 percent of them died within 2 years (Lipsett, Hertz, and Ross, 1963). On the other hand, significant survival benefit is obtained when total excision is achieved. Survival decreases as the stage of the tumor increases (Sullivan, Boileau, and Hodges, 1978). In a review of 28 cases, one stage I patient had no evidence of disease 12.5 years after surgery. Four of five stage II patients were alive at 5 years. Of nine patients at stage III, only one was alive at 5 years. Finally, of the 13 stage IV patients, only one survived for 1 year and none for 2 years (Sullivan, Boileau, and Hodges, 1978).
Management of adrenal incidentaloma is a complex decision-making process that involves considering a range of possible diagnoses, choosing additional diagnostic tests and interpreting their results, understanding the natural history of various adrenal pathologies, estimating the benefits and risks of interventions, and customizing therapy based on patient age and the size of the adrenal mass. Ideally one would like to have randomized trials of a sufficiently large number of patients to address this problem, but this might never be feasible given the heterogeneity of incidentaloma.
Our evidence report on the management of clinically inapparent adrenal mass is based on a systematic review of the literature. Meetings and teleconferences of the EPC staff with technical experts were held to identify specific issues central to this report. A comprehensive search of the medical literature was conducted to identify studies addressing several key questions. We compiled evidence tables of study characteristics and results, appraised the methodological quality of the studies, and summarized their results.
The purpose of an evidence report is to summarize information from relevant studies addressing specific key questions. It is beyond the scope of an evidence report to cover all possible related issues for a topic. Speakers at the State-of-the-Science Conference covered other relevant issues within this topic. The following key questions were addressed in this report:
Question 1. What are the causes and prevalence of clinically inapparent adrenal masses?
1.1 What are the causes of clinically inapparent adrenal masses identified by CT, MRI, or US that are confirmed by histology?
1.2 What are the prevalence rates for the various causes of inapparent adrenal masses?Are there differences in the rates among the initial diagnostic tests used?
1.3 What is the relationship of the age and sex of the patient to the likelihood of having a particular pathology?
1.4 What is the relationship of the size of the mass with the likelihood of having a particular pathology?
Question 2. What is the diagnostic accuracy (sensitivity, specificity) of evaluation modalities (FNA/biopsy, CT, MRI, US, biochemical tests) used to differentiate adrenal masses (adrenal carcinoma, pheochromocytoma, adenoma, adrenal hyperplasia, etc.)?
2.1 What is the risk of the metastatic spread of adrenal carcinoma by FNA?
Question 3. What are the surgical complication rates for various approaches used to excise adrenal masses; specifically laparoscopic, transabdominal, and retroperitoneal approaches?
Question 4. What are the patient outcomes (morbidity and mortality) after surgical excision of adrenocortical carcinoma?
4.1 Are there data on the influence of age and tumor size on the outcomes?
Question 5. What evidence exists to support the use of periodic biochemical and imaging studies to follow untreated adrenal masses?
) would likely result in very few studies qualifying for
evaluation. Furthermore, because most of the published studies on this topic are
retrospective and have methodological problems, it is unclear whether this
strict approach would be useful. Therefore, rather than choosing a single
definition of incidentaloma that may exclude many potentially relevant studies,
we accepted all studies that used their own definitions of incidentaloma.Studies were identified primarily through a Medline search of English language literature published between 1966 and October 2000. We also consulted technical experts and examined references of published meta-analyses and selected review articles to identify additional studies.
Using the PubMed search engine, the staff at the National Library of Medicine conducted a search on the databases Medline, PreMedline, BIOSIS, and Embase on September 19–20, 2000. A combination of search terms was used to map to the subject heading, publication title, or publication abstract. Additional filters or limitations were not used. The searches resulted in 2,636 citations, of which 2,323 came from Medline, 21 from PreMedline, 85 from BIOSIS, and 207 from Embase.
The content of the original key questions was later expanded, and the EPC conducted an updated Medline search on March 21, 2001, using the OVID search engine. Additional subject headings were included to address questions on diagnostic accuracy, surgical complication rates as well as morbidity and mortality outcomes for adrenal masses, and monitoring technologies for untreated adrenal masses. The search yielded an additional 2,750 citations. A final Medline update of surgical series was conducted on October 10, 2001, yielding an additional 41 citations.
The literature searches yielded a total of 5,386 independent citations. The abstracts were screened manually and either categorized to one or more of the key questions or rejected. After screening, we retrieved 602 articles for further examination. Reports published only as letters or abstracts in proceedings were excluded. Specific inclusion criteria and methods of synthesis, developed for each of the key questions, are discussed below. In general, the EPC included all English language studies with at least 10 human subjects. There were no age limitations.
About 194 articles met our inclusion criteria for one of the key questions and were included in the evidence report. Forty-five studies provided data about the prevalence of incidentaloma or the distribution of adrenal pathologies. Thirty-one studies evaluated various diagnostic tests to differentiate adrenal masses. Over 80 studies provided outcome information on various adrenal surgical techniques. Thirty-two studies reported prognostic information on patients with adrenal carcinoma after surgical excision, and nine articles reported results of follow-up strategies.
We report the evidence in three forms. The evidence tables offer a detailed description of the studies that we identified addressing each of the key questions. Summary tables report on each study in an abbreviated form using summary measures of the main outcomes. Finally, for several key questions, we provide an overall summary of information presented in various related summary tables.
Evidence tables provide detailed information about the study design, patient characteristics, inclusion and exclusion criteria, intervention or test evaluated, and the outcomes. Where appropriate, we graded the studies according to the methodological quality, applicability, size, and the effect or test performance. The specific methodologies and the results for each key question are presented in the respective sections in the evidence report.
The evidence tables are condensed into summary tables to provide a more succinct impression of the study quality and results. Summary tables include the important variables regarding study size, patient population (location of study, tumor size and type), outcome measures, and methodological quality. Summarizing the data in such a way allows for ease of comparison among studies. The study (sample) size offers a measure of the weight of the evidence. In surgical series, it also indicates the experience of the surgeons with that particular technique. In general, larger studies provide a more precise estimate of the effect in question, though patient population governs more the applicability of any given study
We used autopsy series in order to estimate the rates of adrenal adenoma. Both prospective and retrospective series were searched to gather data about the prevalence of adrenal incidentaloma and the likelihood of various adrenal pathologies among incidentalomas.
We searched for articles on any diagnostic test of adrenal masses. We specifically included articles on CT, MRI, US, scintigraphy, biochemical tests, and FNA; however, studies of other diagnostic tests were also included. To qualify for inclusion, articles must have reported sufficient data to comment on the performance of a diagnostic test. Our initial screening of abstracts and articles included only studies of subjects with incidentalomas. However, because very few studies met these criteria, and many studies did not clearly define their eligibility criteria, we decided to include all studies of subjects with incidentalomas, adrenal masses found during extra-adrenal cancer evaluation (i.e., staging or search for a primary tumor), or “adrenal masses.” We excluded studies whose subjects primarily or exclusively had subjects with clinically suspected adrenal masses. Because the initial screening of abstracts excluded articles without incidentalomas, the subsequent set of included studies of cancer patients or those with adrenal masses is not comprehensive and can only be considered as representative of all the studies of these populations.
Due to the different prevalence of disease in different populations of patients, we categorized the studies into five groups according to their eligibility criteria. These included 1) incidentally-discovered adrenal masses (incidentalomas); 2) adrenal masses found during extra-adrenal cancer evaluation; 3) adrenal masses (no further description); 4) incidentally-discovered adrenal adenomas only; and 5) clinically-suspected adrenal masses.
Outcomes Considered. The principle measures of test performance were sensitivity and specificity. Whenever possible, we extracted the number of subjects with true positive, true negative, false positive, and false negative tests. We then calculated the sensitivity (percentage of subjects with the outcome of interest who have a (true) positive test) and specificity (percentage of subjects without the outcome of interest who have a (true) negative test). If the data could not be extracted, we relied on reported values of sensitivity, specificity, or, if necessary, other measures of test performance (e.g., accuracy -- total percentage of subjects with either true positive or true negative tests). If authors reported test performance at multiple thresholds (e.g., different density values on CT) or provided raw data, we included the performance at each threshold in the summary tables. Some studies (especially those studies of scintigraphy) did not explicitly calculate test performance. For these studies we calculated test performance to maximize sensitivity.
Of note, tests with multiple thresholds demonstrate a trade-off between sensitivity and specificity. By lowering the threshold for a positive test, high sensitivity can be achieved at the expense of specificity, while raising the threshold has the opposite effect. As a result, studies that include only subjects with disease are of limited value, because only sensitivity, but not the associated specificity can be calculated. If the sensitivity is high, it may be simply that the threshold is set too low and specificity is close to zero. It is the combination of sensitivity and specificity that defines test performance.
Several studies evaluated test performance at multiple thresholds and used receiver operating characteristics (ROC) curve analysis to compare different tests. ROC curves offer a convenient means of graphically displaying the trade-off between sensitivity and specificity. In this method, sensitivity and specificity pairs are plotted for multiple thresholds. The area under the resulting ROC curve (AUC) can then be calculated. By definition, tests that have perfect test performance (100 percent sensitivity and specificity) have an AUC of 1.0; tests that provide no diagnostic information (50 percent sensitivity and specificity) have an AUC of 0.5. Different tests for the same diagnosis can thus be compared by measuring their respective AUCs. Those with a higher AUC can be considered superior to those with a lower AUC. This method is most valid when comparing different tests performed on the same subjects.
Study quality. The internal validity of each study was graded based on study design, conduct, and reporting of the clinical study. For studies of diagnostic test performance, we used a three-category scale to provide some indication of the methodological quality of each study summarized:
Grade A (least bias) - a study that mostly adheres to the traditionally held concepts of high quality diagnostic evaluation, including: prospective design, clear description of the population and setting, the reference standard, the test under investigation, and the diagnostic criteria; blinded interpretation of the reference tests and the test under investigation; verification of the diagnoses in all or most of the patients with negative results; no reporting errors that might hide substantial bias.
Grade B (susceptible to some bias) - a study that does not meet all the criteria of category A. It has some deficiencies, but none likely to cause major bias.
Grade C (likely to have significant bias) - a study with significant design or reporting errors that cannot preclude major bias. This category includes studies in which verification bias could be a large issue and studies that have large amounts of missing information or discrepancies in reporting.
Complications of fine needle aspiration. We searched for articles on FNA or adrenal biopsy. To qualify for inclusion, articles must have reported on short- or long-term procedure complications. Because of the almost complete lack of data specific to metastatic spread of adrenal carcinoma due to FNA, we broadened our eligibility criteria to include all complications. Studies that did not mention complications (either in the positive or the negative) were excluded. Similar to the search for diagnostic tests, our initial screening of abstracts and articles included only studies of subjects with incidentalomas. But since so few studies met these criteria we subsequently included any study of subjects with adrenal masses.
The internal validity of each study was graded based on study design, conduct, and reporting of relevant data. For studies of FNA complications, we used a three-category scale to provide some indication of the methodological quality of each study summarized:
Grade A (least bias) - a study that mostly adheres to the traditionally held concepts of high study quality, including: prospective design, clear description of the population and setting, the biopsy technique under investigation, and a priori definitions of complications; longitudinal follow-up of subjects; proper statistical analysis; complete reporting of complication data and no reporting errors that might hide substantial bias.
Grade B (susceptible to some bias) - study that does not meet all the criteria of category A. It has some deficiencies, but none likely to cause major bias.
Grade C (likely to have significant bias) - a study with significant design or reporting errors that cannot preclude major bias. This category includes studies with inadequate methodology and studies that have large amounts of missing information or discrepancies in reporting.
To address this question, we aimed to identify the various complications associated with different surgical approaches to adrenalectomy, in the context of the larger question of adrenal incidentaloma. Therefore, we concentrated on surgery for tumors that ostensibly could have represented incidentalomas, such as adrenal adenomas, but we did not address the issues of Cushing's disease or adrenal hyperplasia.
We included all studies that enrolled at least 10 patients, and excluded articles that did not explicitly state the surgical approach or did not list any surgical outcomes. When one author or institution published multiple articles containing the same patients, we included only the largest series. In October 2001 we conducted a supplementary search for articles published after our initial search. Due to time constraints, we included only those series with at least 50 patients. Overall, there were 81 articles that met our inclusion criteria.
Study population. Adrenal surgeries are performed only when the diagnoses of adrenal pathologies have been established. Thus, there are no surgical series on patients with adrenal incidentalomas. Adrenal pathologies necessitating surgery can occur at any age and in either sex, so we did not exclude any particular population based on age or sex. There were no studies that listed race as a variable, but most studies listed the country where the study was conducted. Tumor size and type perhaps best represent patient population. We excluded studies of adrenalectomy performed exclusively for indications such as Cushing's disease, which would not apply directly to incidentaloma. Studies of patients with hormone secreting tumors, including pheochromocytomas, were included, as these tumors may occasionally present as incidentalomas. We also included studies of aldosteronomas, which rarely present as incidentalomas, because we presumed that the complications from adrenalectomy for aldosteronomas would be similar to the complications from other similarly sized adenomas. When the data were available, the initial indication for surgery, as well as the final pathology, mean tumor size and range, were recorded in the evidence tables. Patient demographics and co-morbidities, location and nature of the study institution were recorded as well.
Outcomes considered. Although we were asked to consider the surgical complications of adrenalectomy, we found no standardized definition of complication. Most study authors collected operative time, blood loss, analgesic use, length of stay, mortality, and numbers of specific complications resulting from each procedure. Operative time also lacks a standardized definition and may represent total operating room time, skin-to-skin, or skin-to-gland-removal. Most authors did not specify their definition, but we noted it when they did. In addition, operative time could include a unilateral or bilateral adrenalectomy. When authors reported stratified results for unilateral and bilateral procedures, we included both times in the evidence table, but only the unilateral data in the summary table.
Blood loss is usually measured during the operation and recorded. Unfortunately, blood loss does not capture post-operative bleeding, which may be severe, nor does it address the more patient-centered outcome of transfusion. We recorded transfusions under the heading of major complications.
Analgesic use varied widely among institutions. There was no standard drug, method of administration, nor even a unit of measurement. Moreover, analgesic use was not always controlled by the patient, but rather by the attending surgeon.
Length of stay offers a potentially useful proxy for surgical complications. Unfortunately, length of stay may be confounded by secular changes in hospital practice over the span of many publication years, by social factors, by concurrent surgical or medical problems not directly related to the surgery, and by cultural expectations of the patients.
Most studies report some figure for mortality. While the definition of intraoperative mortality is clear, the definition of perioperative mortality is vague. The number of deaths in all series was small, and individual deaths were described in detail. There was no intra-operative mortality. We included all deaths within 30 days of surgery and noted when the author felt that the death was unrelated to surgery, which occurred in every case. Because complications were heterogeneous and often unique, we found it difficult to generate an overall rate. For the evidence table, we recorded the number and type of each individual complication in as much detail as possible.
Data extraction. Data extracted included study year, procedural approach, patient demographics, location and nature of the study institutions, inclusion and exclusion criteria, study design and enrollment years, surgical indication and pathology, patient co-morbidities and American Society of Anesthesiologists (ASA) score, tumor size, operative time, blood loss, mortality, post-operative analgesia requirements, length of stay, and complications.
Applicability (population and indication). Applicability in this series is governed by study location, tumor size and tumor type. Study location includes non-measurable variables such as surgical culture, institutional support, or patient expectations. Surgical culture (the local practices of surgeons) may influence certain outcomes, such as the placement of post-operative drains, or govern the use of intra-operative antibiotics or prophylaxis for deep vein thrombosis, which may affect complication rates, yet are unrelated to the specific procedure. Institutional support may determine how fast patients can be discharged after complicated operations such as pheochromocytoma. Patient expectations may affect different outcome variables such as length of stay or post-operative analgesic use. Furthermore, some subjective complications, such as local pain or paresthesia, may go unrecognized unless patients report them, or investigators ask specifically.
Tumor size and type are particularly useful in judging the applicability of a study to a particular patient known to have a tumor of that size or type. In particular, there is debate in the literature regarding the appropriate procedure to be used for large tumors, pheochromocytomas, adrenal carcinomas and metastatic disease. We captured the mean tumor size and range, as well as the percentage of patients with pheochromocytoma, adrenal carcinoma and metastatic disease. These features are also useful for comparing groups within the same study, when the two study arms are often drawn from different populations. These summary measures allow the reader to judge whether the historical controls chosen by the authors are indeed comparable to the study patients.
Estimating the complication rates. For each study we summarized the operative time, blood loss, length of stay and complication rate. We did not include post-operative analgesic use, because the measures were so diverse, and no comparable metrics could be devised. For operative time, blood loss, and length of stay, we reported the values for unilateral adrenalectomy, if results were stratified by operation. For studies that did not segregate their results, we reported the overall operative time, blood loss, and length of stay, all of which will be proportionately higher depending on the percentage of bilateral cases.
For complication rates, we classified complications as mortality, major or minor complications. We included 30-day post-operative mortality, as well as a comment when and if the author felt that death was not a direct outcome of surgery. Complications were classified according to the classification system developed by Clavien (Clavien, Sanabria, and Strasberg, 1992). We listed as minor all complications that were Clavien grade I. These include all events that, if left untreated, would have spontaneous resolution, or can be treated with a simple bedside procedure involving little or no anesthesia. They require no drugs other than analgesic, antipyretic, antiemetic, antidiarrheal, or drugs for treatment of urinary retention or lower urinary tract infection. They do not result in hospital stay greater than twice the median for the given procedure. All other complications, including Clavien grade II and III, were classified as major. These included potentially life-threatening complications, those requiring additional procedures, prolonged hospital stay, iatrogenic injury, residual disability and organ resection.
For ease of comparability, we graded the methodological quality (internal validity) of each of the studies using the following scheme:
Grade A (least bias) - a multicenter (or multi-surgeon) prospective series with matched controls, applying the same exclusion criteria to all study arms. Non-comparative case series of consecutive cases with data collected prospectively and no exclusions of difficult cases or bad outcomes (intention-to-treat analysis).
Grade B (susceptible to some bias) - a single surgeon retrospective case series with matched controls and exclusion criteria the same for all study arms. Retrospective non-comparative case series of consecutive cases without exclusions.
Grade C (likely to have significant bias) - a study with no matched controls or unequal treatment of the study arms. Non-consecutive cases due to exclusions.
Grade I (indeterminate) - a study with insufficient information to determine quality.
Summarizing the evidence for each surgical technique. We also constructed a summary table for each surgical technique and its comparisons. Summary tables for techniques include a single line summarizing the number of studies and patients, the range of tumor size and type, the ranges for each outcome, and for comparisons, the number of studies finding each technique to be advantageous. Finally, each table lists the number of studies for each of the methodological grades.
The summary table is intended to supply an overall picture for each surgical technique of the quantity of evidence, the quality of that evidence, and the main outcomes, or at least the ranges of evidence. In the comparison studies, it is also helpful to see the number of studies finding a statistical difference between the various approaches. Due to the wide variability of the study designs and definitions, we did not attempt to combine the data across studies.
This question is not specifically about adrenal mass incidentally found. A separate literature search was conducted to identify surgical studies of adrenocortical carcinoma that reported outcome data. Both retrospective and prospective studies qualify although mostly retrospective studies were found. We extracted information about the mean age of the study population, sex, study design, enrollment year of the study, procedures performed, mean tumor size, mean study follow-up duration, short term outcomes and long term survival information.
We looked for studies that prospectively applied pre-specified imaging or biochemical testing protocols to a population of patients with untreated incidentally discovered adrenal masses as part of their follow-up strategies. To supplement the few studies that we found with pre-specified protocols, we also accepted studies that reported analyses of patients followed with unspecified protocols. A total of nine studies were included in this report.
| Study | Sex | Age | Mass Size (cm a) | Reported Diagnosis | Revised Diagnoses b | Method of Diagnosis |
|---|---|---|---|---|---|---|
| Tütüncü 1999c | ND | ND | 7.7 | adrenocortical carcinoma | adrenal carcinoma | histology |
| ND | ND | 6.0 | adrenocortical adenoma: Cushing | adenoma | histology | |
| ND | ND | 5.9 | pheochromocytoma | pheochromocytoma | histology | |
| ND | ND | 4.5 | pheochromocytoma | pheochromocytoma | histology | |
| ND | ND | 4.0 | pheochromocytoma | pheochromocytoma | histology | |
| ND | ND | 5.6 | pheochromocytoma | pheochromocytoma | histology | |
| ND | ND | 5.0 | myelolipoma | myelolipoma | histology | |
| ND | ND | 9.2 | retroperitoneal fibrosis | retroperitoneal fibrosis | histology | |
| ND | ND | 4.0 | ganglioneuroma | ganglioneuroma | histology | |
| ND | ND | 7.0 | angiomyelolipoma | angiomyelolipoma | histology | |
| ND | ND | 11.5 | malignant epithelial carcinoma | malignant epithelial carcinoma | histology | |
| ND | ND | 15.0 | extraadrenal pheochromocytoma | pheochromocytoma | histology | |
| ND | ND | 4.2 | myelolipoma | myelolipoma | histology | |
| ND | ND | 7.0 | pheochromocytoma | pheochromocytoma | histology | |
| ND | ND | 4.3 | adrenocortical carcinoma | adrenal carcinoma | histology | |
| Corsello 1993 | F | 65 | 4.5 | adrenocortical adenoma | adenoma | histology |
| F | 20 | 6.0 | ganglioneuroma | ganglioneuroma | histology | |
| M | 52 | 3.0 | adrenocortical adenoma | adenoma | histology | |
| F | 52 | 5.0 | ganglioneuroma | ganglioneuroma | histology | |
| F | 46 | 5.0 | pheochromocytoma | pheochromocytoma | histology | |
| F | 35 | 12.0 | pheochromocytoma | pheochromocytoma | histology | |
| M | 66 | 5.0 | adrenocortical adenoma | adenoma | histology | |
| M | 52 | 16.0 | pheochromocytoma | pheochromocytoma | histology | |
| F | 30 | 6.0 | regenerative hepatic nodule | regenerative hepatic nodule | histology | |
| F | 62 | 3.0 | adrenocortical adenoma | adenoma | histology | |
| F | 48 | 4.5 | adrenocortical adenoma | adenoma | histology | |
| Guerrero 1985 | F | 44 | 2.8 | adrenocortical adenoma | adenoma | histology |
| F | 45 | 7.5 | adrenocortical adenoma | adenoma | histology | |
| F | 52 | 2.5 | adrenocortical adenoma | adenoma | histology | |
| M | 58 | 2.5 | adrenocortical adenoma | adenoma | histology | |
| F | 59 | 3.0 | adrenocortical adenoma | adenoma | histology | |
| M | 66 | 7.5 | myelolipoma | myelolipoma | histology | |
| M | 49 | 1.5 | adrenocortical adenoma | adenoma | histology | |
| F | 79 | 5.0 | adrenocortical adenoma | adenoma | histology | |
| F | 35 | 2.5 | adrenocortical adenoma | adenoma | histology | |
| F | 60 | 2.0 | metastatic small cell | metastasis | Histology | |
| Gaboardi 1991 | F | 55 | 3.0 | metastatic small cell | metastasis | histology |
| M | 56 | 11.0 | pheochromocytoma | pheochromocytoma | histology | |
| F | 49 | 5.0 | adenoma | adenoma | histology | |
| M | 52 | 9.0 | adenoma | adenoma | histology | |
| M | 45 | 8.0 | adrenal carcinoma | adrenal carcinoma | histology | |
| F | 45 | 4.0 | Lymphoma | lymphoma | histology | |
| F | 49 | 5.0 | pheochromocytoma | pheochromocytoma | histology | |
| M | 61 | 7.0 | myelolipoma | myelolipoma | histology | |
| M | 71 | 7.0 | myelolipoma | myelolipoma | histology | |
| M | 58 | 5.0 | adrenal carcinoma | adrenal carcinoma | histology | |
| F | 59 | 4.0 | metastatic small cell | metastasis | histology | |
| F | 55 | 6.0 | myelolipoma | myelolipoma | histology | |
| Virkkala 1989 | M | 52 | 1.3 | micro-macronodular hyperplasia | adenoma | histology |
| F | 66 | 2.0 | adenoma | adenoma | histology | |
| F | 75 | 2.5 | micro-macronodular hyperplasia | adenoma | histology | |
| M | 49 | 2.0 | adenoma | adenoma | histology | |
| F | 74 | 2.0 | renal angiomyolipoma | renal angiomyolipoma | histology | |
| F | 46 | 5.0 | atypical adenoma | adenoma | histology | |
| F | 57 | 6.0 | hematoma | hematoma | histology | |
| F | 60 | 4.0 | micro-macronodular hyperplasia | adenoma | histology | |
| Andreis 1985 | M | 32 | 7.0 | adenoma | adenoma | histology |
| F | 39 | 5.3 | adenoma | adenoma | histology | |
| F | 61 | 10.0 | carcinoma | adrenal carcinoma | histology | |
| F | 18 | 12.0 | carcinoma | adrenal carcinoma | histology | |
| F | 59 | 3.7 | carcinoma | adrenal carcinoma | histology | |
| M | 54 | 12.0 | carcinoma | adrenal carcinoma | histology | |
| M | 56 | 4.5 | mielolipoma | myelolipoma | histology | |
| M | 45 | 3.0 | fecromocitoma | pheochromocytoma | histology | |
| M | 62 | 10.0 | ganglioneuroma | ganglioneuroma | histology | |
| F | 60 | 9.0 | istiocitoma fibroso | istiocitoma fibroso | histology | |
| M | 47 | 5.5 | metastasis | metastasis | histology | |
| F | 64 | 2.0 | metastasis | metastasis | histology | |
| M | 48 | 7.0 | metastasis | metastasis | histology | |
| M | 74 | 13.0 | metastasis | metastasis | histology | |
| F | 40 | 4.0 | metastasis | metastasis | histology | |
| M | 51 | 5.5 | metastasis | metastasis | histology | |
| M | 44 | 8.3 | metastasis | metastasis | histology | |
| M | 45 | 5.0 | metastasis | metastasis | histology | |
| F | 50 | 8.0 | metastasis | metastasis | histology | |
| M | 50 | 6.0 | metastasis | metastasis | histology | |
| M | 42 | 3.5 | metastasis | metastasis | histology | |
| M | 72 | 6.6 | metastasis | metastasis | histology | |
| M | 59 | 2.0 | metastasis | metastasis | histology | |
| M | 56 | 4.4 | metastasis | metastasis | histology | |
| M | 45 | 4.2 | metastasis | metastasis | histology | |
| F | 48 | 4.0 | metastasis | metastasis | histology | |
| Pfister 1987 | F | 59 | 2.5 | benign adrenal cortical mass | adenoma | histology |
| F | 66 | 6.0 | myelolipoma | myelolipoma | histology | |
| F | 62 | 3.0 | cyst | adrenal cyst | histology | |
| F | 55 | 5.0 | benign adrenal cortical mass | adenoma | histology | |
| F | 51 | 5.0 | cyst | adrenal cyst | histology | |
| F | 61 | 2.5 | benign adrenal cortical mass | adenoma | histology | |
| F | 52 | 4.0 | benign adrenal cortical mass | adenoma | histology | |
| F | 59 | 5.0 | myelolipoma | myelolipoma | histology | |
| F | 42 | 2.0 | benign adrenal cortical mass | adenoma | histology | |
| F | 50 | 8.0 | benign adrenal cortical mass | adenoma | histology | |
| M | 70 | 6.5 | benign adrenal cortical mass | adenoma | histology | |
| F | 74 | 6.0 | ganglioneuroma | ganglioneuroma | histology | |
| Flecchia 1995d | F | 50 | 4.9 | adenoma | adenoma | histology |
| M | 51 | 2.9 | adenoma | adenoma | histology | |
| M | 30 | 5.2 | adenoma | adenoma | histology | |
| F | 68 | 4.1 | adenoma | adenoma | histology | |
| F | 69 | 3.5 | adenoma | adenoma | histology | |
| M | 74 | 2.8 | adenoma | adenoma | histology | |
| F | 44 | 5.5 | adenoma | adenoma | histology | |
| M | 78 | 2.3 | adenoma | adenoma | histology | |
| F | 44 | 9.6 | adenocarcinoma | adrenal carcinoma | histology | |
| F | 55 | 5.6 | adenocarcinoma | adrenal carcinoma | histology | |
| Ambrosi 1995 | F | 42 | 3.0 | adenoma | adenoma | histology |
| M | 66 | 4.0 | adenoma | adenoma | histology | |
| F | 39 | 6.0 | carcinoma | adrenal carcinoma | histology | |
| F | 68 | 2.8 | adenoma | adenoma | histology | |
| M | 72 | 6.3 | hematoma | hematoma | histology | |
| M | 68 | 4.6 | myelolipoma | myelolipoma | histology | |
| M | 64 | 2.0 | adenoma | adenoma | Histology | |
| Osella 1994 | M | 65 | 3.0 | adrenal adenoma | adenoma | histology |
| Terzolo 1995 | F | 61 | 5.0 | pheochromocytoma | pheochromocytoma | histology |
| F | 45 | 15.0 | adrenal carcinoma | adrenal carcinoma | histology | |
| M | 67 | 4.4 | pheochromocytoma | pheochromocytoma | histology | |
| F | 49 | 2.5 | adrenal adenoma | adenoma | histology | |
| F | 55 | 4.0 | adrenal adenoma | adenoma | histology | |
| M | 76 | 10.0 | ganglionic neoplasia | ganglioneuroma | histology | |
| M | 46 | 3.0 | adrenal adenoma | adenoma | histology | |
| F | 69 | 3.5 | adrenal adenoma | adenoma | histology | |
| F | 60 | 7.0 | adrenal myelolipoma | myelolipoma | histology | |
| F | 60 | 5.5 | adrenal adenoma | adenoma | histology | |
| F | 65 | 3.0 | metastasis | metastasis | histology | |
| F | 79 | 8.0 | adrenal carcinoma | adrenal carcinoma | histology | |
| M | 49 | 2.0 | adrenal adenoma | adenoma | histology | |
| M | 59 | 4.0 | adrenal adenoma | adenoma | histology | |
| F | 19 | 13.0 | adrenal carcinoma | adrenal carcinoma | histology | |
| F | 42 | 3.7 | adrenal cyst | adrenal cyst | histology | |
| Studies that excluded cancer cases | ||||||
| Terzolo 2000 | F | 61 | 5.0 | pheochromocytoma | pheochromocytoma | surgery/histology |
| F | 81 | 6.0 | myelolipoma | myelolipoma | surgery/histology | |
| F | 60 | 7.0 | myeliolipoma | myelolipoma | surgery/histology | |
| F | 71 | 3.0 | pheochromocytoma | pheochromocytoma | surgery/histology | |
| F | 64 | 3.5 | cortical adenoma | adenoma | surgery/histology | |
| F | 69 | 3.5 | cortical adenoma | adenoma | surgery/histology | |
| F | 52 | 4.2 | cortical adenoma | adenoma | surgery/histology | |
| M | 46 | 3.0 | cortical adenoma | adenoma | surgery/histology | |
| F | 49 | 2.5 | cortical adenoma | adenoma | surgery/histology | |
| F | 60 | 5.5 | cortical adenoma | adenoma | surgery/histology | |
| M | 61 | 2.0 | cortical adenoma | adenoma | surgery/histology | |
| M | 47 | 3.0 | hemorrhagic cyst | adrenal cyst | surgery/histology | |
| M | 66 | 7.0 | pheochromocytoma | pheochromocytoma | surgery/histology | |
| F | 55 | 4.0 | cortical adenoma | adenoma | surgery/histology | |
| M | 65 | 3.0 | cortical adenoma | adenoma | surgery/histology | |
| F | 42 | 3.7 | cyst | adrenal cyst | surgery/histology | |
| F | 59 | 4.0 | cortical adenoma | adenoma | surgery/histology | |
| F | 19 | 13.0 | adrenocortical cancer | adrenal carcinoma | surgery/histology | |
| F | 45 | 15.0 | adrenocortical cancer | adrenal carcinoma | surgery/histology | |
| F | 65 | 3.0 | metastasis | metastasis | surgery/histology | |
| Abecassis 1985 | M | 42 | 3.5 | black adenoma | adenoma | surgery |
| F | 56 | 4.0 | adrenocortical adenoma | adenoma | surgery | |
| M | 37 | 4.0 | ganglioneuroma | ganglioneuroma | Surgery | |
| Bondanelli 1997 | M | 55 | 4.2 | adrenal adenoma | adenoma | surgery/histology |
| F | 33 | 5.0 | lymphangiomatous cyst | adrenal cyst | surgery/histology | |
| M | 56 | 4.3 | adrenal adenoma | adenoma | surgery/histology | |
| M | 50 | 2.1 | pheochromocytoma | pheochromocytoma | surgery/histology | |
| M | 47 | 2.6 | ganglioneuroma | ganglioneuroma | surgery/histology | |
| F | 58 | 4.5 | adrenal adenoma | adenoma | surgery/histology | |
| M | 38 | 3.5 | ganglioneuroma | ganglioneuroma | surgery/histology | |
| M | 69 | 3.8 | adrenal hyperplasia | adrenal hyperplasia | surgery/histology | |
| F | 55 | 2.5 | hemorrhage | hemorrhage | surgery/histology | |
| F | 37 | 4.5 | adrenal adenoma | adenoma | surgery/histology | |
| F | 46 | 12.0 | adrenal carcinoma | adrenal carcinoma | surgery/histology | |
| M | 51 | 8.0 | myelolipoma | myelolipoma | surgery/histology | |
| F | 75 | 10.0 | pheochromocytoma | pheochromocytoma | surgery/histology | |
| F | 66 | 2.5 | adrenal adenoma | adenoma | surgery/histology | |
| Masumori 1998 | M | 47 | 2.7 | adrenocortical adenoma | adenoma | histology |
| M | 56 | 2.5 | adrenocortical adenoma | adenoma | histology | |
| M | 41 | 8.2 | ganglioneuroma | ganglioneuroma | histology | |
| M | 46 | 2.2 | pheochromocytoma | pheochromocytoma | histology | |
| F | 66 | 7.0 | neurinoma | neurinoma | histology | |
ND = no data; F, female; M, male
Includes the larger dimension for bilateral tumors; the smaller dimension if 2 dimensions used for singular tumor.
Inclusion criteria of masses ≥ 4 cm.
Study did not include individual patient data for other diagnoses (8) than adrenal adenomas and adrenal cortical carcinomas.
Because incidentaloma is not a disease entity, the prevalence of incidentaloma will vary according to the inclusion criteria of the study. In the general population setting, we found one study that used transabdominal US screening in general health examination and reported about 40 adrenal or retroperitoneal masses out of 41,357 subjects (prevalence of 0.1 percent) (Masumori, Adachi, Noda, et al., 1998). Eleven masses were confirmed to be adrenal origin with CT (prevalence of 0.027 percent). The prevalence of incidentaloma was reported to be 0.6 percent in a study of 2,200 patients undergoing upper abdominal CT for specific indications (Glazer, Weyman, Sagel, et al., 1982).
| Tumor Size | ||||||||
|---|---|---|---|---|---|---|---|---|
| < 2 – 4 cm | 4.1 – 6 cm | > 6 cm | Total | |||||
| Final Diagnosis | N | % | N | % | N | % | N | % |
| Adenoma | 45 | 62 | 20 | 38 | 5 | 11 | 70 | 41 |
| Adrenal carcinoma | 1 | 1.4 | 4 | 8 | 12 | 27 | 17 | 10 |
| Adrenal cyst | 4 | 5.5 | 2 | 4 | 6 | 3.5 | ||
| Adrenal hyperplasia | 1 | 1.4 | 1 | 0.6 | ||||
| Angiomyelolipoma | 1 | 2 | 1 | 0.6 | ||||
| Ganglioneuroma | 4 | 5.5 | 3 | 6 | 3 | 7 | 10 | 5.9 |
| Hematoma | 1 | 2 | 1 | 2 | 2 | 1.2 | ||
| Hemorrhage | 1 | 1.4 | 1 | 0.6 | ||||
| Istiocitoma fibroso (Italian) | 1 | 2 | 1 | 0.6 | ||||
| Lymphoma | 1 | 1.4 | 0 | 1 | 0.6 | |||
| Malignant epithelial carcinoma | 1 | 2 | 1 | 0.6 | ||||
| Metastases | 10 | 14 | 6 | 11 | 5 | 11 | 21 | 12 |
| Myelolipoma | 8 | 15 | 6 | 14 | 14 | 8.2 | ||
| Neurinoma | 1 | 2 | 1 | 0.6 | ||||
| Pheochromocytoma | 5 | 6.8 | 8 | 15 | 7 | 16 | 20 | 12 |
| Regenerative hepatic nodule | 1 | 2 | 1 | 0.6 | ||||
| Renal angiomyolipoma | 1 | 1.4 | 1 | 0.6 | ||||
| Retroperitoneal fibrosis | 1 | 2 | 1 | 0.6 | ||||
| Total | 73 | 43% | 53 | 31% | 44 | 26% | 170 | 100% |
N, number of subjects
| Tumor Size | ||||||||
|---|---|---|---|---|---|---|---|---|
| < 2 – 4 cm | 4.1 – 6 cm | > 6 cm | Total | |||||
| Final Diagnosis | N | % | N | % | N | % | N | % |
| Adenoma | 28 | 65 | 9 | 28 | 5 | 18 | 42 | 41 |
| Adrenal carcinoma | 1 | 2.3 | 2 | 6.3 | 7 | 25 | 10 | 9.7 |
| Adrenal cyst | 2 | 4.7 | 1 | 3.1 | 3 | 2.9 | ||
| Ganglioneuroma | 3 | 9.4 | 2 | 7.1 | 5 | 4.9 | ||
| Hematoma | 1 | 3.1 | 1 | 3.6 | 2 | 1.9 | ||
| Istiocitoma fibroso (Italian) | 1 | 3.6 | 1 | 1.0 | ||||
| Lymphoma | 1 | 2.3 | 1 | 1.0 | ||||
| Metastases | 9 | 21 | 6 | 19 | 5 | 18 | 20 | 19 |
| Myelolipoma | 5 | 16 | 4 | 14 | 9 | 8.7 | ||
| Pheochromocytoma | 1 | 2.3 | 4 | 13 | 3 | 11 | 8 | 7.8 |
| Regenerative hepatic nodule | 1 | 3.1 | 1 | 1.0 | ||||
| Renal angiomyolipoma | 1 | 2.3 | 1 | 1.0 | ||||
| Total | 43 | 42% | 32 | 31% | 28 | 27% | 103 | 100% |
N, number of subjects
Excluding studies with inclusion/exclusion criteria or missing data that limits generalixability: Tütüncü et. al., 1999 included mass ≥ 4 cm; Flecchia et. al., 1995 did not give individual patient data for cases (8) other than adenomas and adrenocortical carcinomas; four studies (Abecassis 1985, Bondanelli 1997, Masumori 1998, Terzolo 2000) with exclusion criteria of patients with known extra-adrenal malignancies.
The question about differences in the spectrum of pathologies identified by different imaging modalities cannot be answered with current data. In most of the studies, the adrenal incidentalomas were discovered with CTs. There were some studies that reported incidentalomas discovered with either US or MRI or CT, but most of the studies do not separately report the data that would allow meaningful subgroup analyses across studies.
Although the outcomes of interest for calculating sensitivity and specificity vary among studies, in this report sensitivity refers to the percentage of subjects with an adrenal malignancy (whether adrenal or metastatic in origin) with a positive test, and specificity refers to the percentage of subjects without adrenal malignancy with a negative test, except where noted.
There were ten studies that evaluated CT. Four of these evaluated unenhanced CT, three evaluated enhanced CT, two studies evaluated delayed enhanced CT, three studies evaluated combined enhanced and unenhanced CT, and six evaluated mass size on CT.
Three studies evaluated enhanced CT to differentiate malignant from benign adrenal masses, using various definitions of a positive CT. One study included only subjects with incidentalomas; two included all subjects with adrenal masses. Malignancy prevalence ranged from 20 to 50 percent. Two studies reported test performance based on total masses; the other reported test performance based on patients. Two studies found excellent sensitivity but poor specificity at low attenuation values. The study that used mass homogeneity had similar findings. Two studies compared the areas under ROC curves of attenuation values on enhanced CT to other parameters. Singer, Obuchowski, Einstein, et al. (1994) found that attenuation values on enhanced CT and mass size were both equally useful to diagnose adrenal malignancy. Korobkin, Brodeur, Francis, et al. (1996) found that attenuation values on enhanced CT were inferior to delayed enhanced CT.
Two studies evaluated delayed enhanced CT to differentiate either malignant masses from benign or adrenal adenomas from non-adenomas. Both studies included patients with adrenal masses. Twenty percent of masses were malignant in the first study; 47 percent were non-adenomas in the second. Both reported test performance based on total masses. Szolar and Kammerhuber (1997) found that enhanced CT delayed 180 seconds after IV contrast injection yielded good to excellent test performance using thresholds between 64 and 70 HU. Sensitivity ranged from 83 to 100 percent; specificity ranged from 100 to 91 percent, respectively, to differentiate non-adenomas from adenomas. Enhanced CT delayed 30 minutes after IV contrast injection resulted in perfect test performance. Korobkin, Brodeur, Francis, et al. (1996) also found excellent test performance when CT was delayed 45 to 75 minutes after IV contrast injection. Sensitivity was 100 percent; specificity was 95 percent to differentiate malignant from benign masses. As noted above, this study also found that using delayed enhanced CT was superior to immediate enhanced CT in order to diagnose adrenal malignancy.
Six studies evaluated adrenal mass size on CT to differentiate adrenal masses using various size thresholds. No study explicitly included only subjects with incidentalomas. Two included subjects with extraadrenal cancer; three included subjects with adrenal masses. One included subjects from both populations. Two studies tested mass size on CT to differentiate non-adenomas, including other benign lesions, from adenomas. Malignancy (or non-adenoma) prevalence ranged from 42 to 69 percent. Five studies reported test performance based on total adrenal masses; one reported test performance data based on both subjects and masses. Smaller size thresholds corresponded to higher sensitivity and lower specificity, and vice versa. No study found a size threshold that yielded both high sensitivity and specificity. As noted above, Singer, Obuchowski, Einstein, et al. (1994) found size threshold and attenuation value on enhanced CT to be equally good parameters to diagnose malignancy. Lee, Hahn, Papanicolaou, et al. (1991) and McNicholas, Lee, Mayo-Smith, et al. (1995) and van Erkel, van Gils, Lequin, et al. (1994) found size measurement to be an inferior parameter to diagnose malignancy or non-adenomas, respectively.
There were eight studies that evaluated MRI, all of them unenhanced MRI. One study also evaluated combined unenhanced and enhanced MRI.
Eight studies evaluated unenhanced MRI to differentiate adrenal masses. Bilbey, McLoughlin, Kurkjian, et al. (1995) included patients with incidentalomas, in addition to patients with cancer work ups and those with clinically suspected adrenal masses. Seven studies included patients with cancer work ups (Krestin, Freidmann, Fishbach, et al., 1991); one of these also included patients with incidentalomas and a few patients with clinically suspected adrenal masses; another included patients with other adrenal masses. One study included patients with adrenal masses. Two studies used MRI to differentiate adenomas from non-adenomas (including other benign lesions). The other studies used MRI to differentiate malignant masses from benign. Malignancy (or non-adenoma) prevalence ranged from 19 to 58 percent. Only two studies analyzed test performance based on patients instead of total masses.
Mass:spleen ratio. Four studies measured adrenal mass to spleen ratio. Two studies reported using T1 measurements; the others did not explicitly define which spin measurements were used. One study reported test performance at various thresholds. Bilbey, McLoughlin, Kurkjian, et al. (1995) found perfect test performance for MRI to differentiate non-adenomas (including other benign lesions) from adenomas; however, McNicholas, Lee, Mayo-Smith, et al. (1995) found much poorer test performance at the same threshold to diagnose adrenal malignancies. Overall, the studies found MRI had sensitivity of 84 to 100 percent with specificity of 78 to 94 percent to differentiate malignant from benign masses.
Mass:fat ratio. Three studies measured adrenal mass to fat ratio. One study evaluated T1 spin measurements, one evaluated both T1 and T2 spin measurements; the third did not explicitly report the type of spin measurement used. Using different thresholds and measurements, high sensitivity was achieved only with poorer specificity, and vice versa. To diagnose either adrenal malignancies or non-adenomas, sensitivity ranged from 50 to 100 percent and specificity ranged from 57 to 95 percent.
Mass:liver ratio. Three studies measured adrenal mass to liver ratio. One study evaluated MRI to diagnose adrenal malignancies; two diagnosed non-adenomas. Two studies used T1 spin measurements; one used T2 spin. Two studies found similar results with mass to liver ratios as with mass to fat ratios, where high sensitivity was achieved only with poorer specificity, and vice versa. Van Erkel, van Gils, Lequin, et al. (1994) found that T2 measurements achieved very poor test performance to diagnose non-adenomas.
Mass:muscle ratio. Two studies measured adrenal mass to muscle ratio to diagnose non-adenomas. Bilbey, McLoughlin, Kurkjian, et al. (1995) used T1 measurements and found that perfect sensitivity could be achieved with a specificity of 89 percent. Van Erkel, van Gils, Lequin, et al. (1994) found poorer test performance using T2 measurements.
Krestin, Freidmann, Fishbach, et al. (1991) compared MRI to CT qualitatively, and concluded that unenhanced MRI (mass:fat ratio) is somewhat superior to combined unenhanced and enhanced CT, based on overall test accuracy. McNicholas, Lee, Mayo-Smith, et al. (1995) performed a quantitative comparison of MRI to CT using area under the ROC curve; they concluded that unenhanced MRI (mass:spleen ratio) had similar test performance as unenhanced CT. Van Erkel, van Gils, Lequin, et al., (1994) found T2 spin measurements on MRI to be an inferior parameter to diagnose non-adenomas than CT.
Combination criteria. Burt, Heelan, Coit, et al. (1994) used both T1 and T2 weighted imaging. MRI positive for malignancy was defined as adrenal mass having a greater intensity than the opposite adrenal gland on both T1 and T2 weighted imaging, or having greater intensity than liver tissue on T2 weighted imaging. Sensitivity was 100 percent, however specificity was only 24 percent.
Six studies evaluated scintigraphy to differentiate adrenal malignancies from benign lesions. Two studies included patients with incidentalomas only; the other three included patients with either incidentalomas or cancer work ups. One included only patients with cancer work-ups. Gross, Shapiro, Francis, et al. (1994) included only subjects with unilateral masses; Gross, Shapiro, Francis, et al. (1995) included only subjects with bilateral masses. Malignancy prevalence ranged from 4 to 45 percent. Five studies evaluated 131I-6β-iodomethyl-19-norcholesterol (NP-59) scintigraphy and analyzed test performance for patients; Dominguez-Gadea, Diez, Bas, et al. (1994) evaluated 75-Se-cholesterol scintigraphy and analyzed test performance for total masses. No study of subjects with incidentaloma evaluated 123I-meta-iodobenzylguanidine (MIBG) scintigraphy in more than four subjects. Each study used different criteria to distinguish adrenal masses. One study categorized adrenal masses based on increased or decreased marker uptake. Another study categorized adrenal and juxtaadrenal masses based on nine different uptake patterns. Two studies categorized adrenal masses based on symmetry and uptake. Two studies compared concordance of scintigraphy to CT. Only largest study, comprising 229 subjects, explicitly calculated test performance results. Overall, the studies, using various definitions of positive scintigraphy, found that scintigraphy achieved high sensitivity (71 to 100 percent) with varying specificity (50 to 100 percent) to differentiate malignant from benign adrenal masses.
One study evaluated US to differentiate malignant from benign adrenal masses in patients with incidentalomas. Fontana, Porpiglia, Destefanis, et al. (1999) had a malignancy prevalence of 22 percent. Using a definition of positive test of inhomogeneous masses, this study found poor test performance for US to differentiate malignant from benign adrenal masses.
One small study of 20 subjects evaluated PET to differentiate malignant from benign adrenal masses in patients with either incidentalomas or cancer work ups. Boland, Goldberg, Lee, et al. (1995) had a malignancy prevalence of 55 percent. Using a definition of positive tests of increased uptake by the adrenal mass, this study found perfect test performance to differentiate malignant from benign adrenal masses.
Nine studies reported some measure of test performance for FNA to diagnose adrenal masses. There was no study which included patients with incidentalomas only, although five stated explicitly that these patients were included. Eight included patients with cancer work ups and one included patients with adrenal masses. Three studies also included some patients with symptoms suggestive of adrenal disease. Eight studies had a malignancy prevalence rate of 19 to 82 percent. One study did not report sufficient information to calculate a prevalence rate. This study and a second did not clearly define positive and negative biopsies. Six studies either did not clearly define the reference standard or, in part, used FNA as both test under investigation and reference standard (Berkman, Bernardino, Sewell, et al., 1984; Heaston, Handel, Ashton, et al., 1982; Katz, Patel, Mackay, et al., 1984; Montali, Solbiati, Bossi, et al., 1984; Welch, Sheedy, Stephens, et al., 1994). However, FNA was not always the primary test under investigation. Welch, Sheedy, Stephens, et al. (1994) provided test performance results but did not report complete data. Excluding biopsies that were inconclusive, eight studies showed that sensitivity for all patients (or masses) to diagnose malignancy ranged from 81 to 100 percent and specificity ranged from 83 to 100 percent. One study reported only that accuracy was 91 percent. In seven studies, between 6 and 50 percent of biopsies were inconclusive. One study also analyzed test performance based on mass size and needle size. The study found higher sensitivity and accuracy in masses larger than 3 cm than those that were smaller and also when 19 gauge or larger needles were used than smaller. Furthermore Welch, Sheedy, Stephens, et al. (1994) reported that “accuracy depended on the size of the needle used to perform biopsy but not on the size of the lesion.” The evidence is too sparse to draw conclusions about the test performance of different methods of adrenal biopsy (e.g., FNA versus coring biopsy).
Terzolo, Ali, Osella, et al. (2000) is the only study that evaluated the test performance of a biochemical test to differentiate malignant disease from benign. In a study of 84 patients, three of whom had malignancies (4 percent), high or normal 8 a.m. DHEAS levels were 100 percent sensitive but only 38 percent specific. Thresholds for normal DHEAS levels were age- and gender-specific based on a control group. Test performance values were reported, but complete data were not reported.
CT was evaluated in ten studies with 467 subjects, 138 of whom were reported to have incidentalomas. The rest had adrenal masses found during extraadrenal cancer work up and undefined adrenal masses; a small number had clinically suspected adrenal masses. Studies were generally small, ranging from 23 to 84 subjects, and were of fair to poor quality. In general, test performance was based on the total number of adrenal masses (as opposed to total number of subjects). The studies reported that CT has the potential to have good to excellent test performance to differentiate malignant from benign adrenal masses. Studies of unenhanced CT, immediate enhanced CT and combined unenhanced and enhanced CT generally found that high to excellent sensitivity can be achieved only with fair to good specificity. Two studies of delayed enhanced CT found excellent overall test performance. Size of adrenal mass on CT had generally poor test performance where high sensitivity was achieved only with low specificity, or vice versa. Four studies evaluated various morphological features of adrenal masses (other than size) using either unenhanced CT, enhanced CT, or a combination of the two. The morphological criteria were generally subjective, including such parameters as inhomogeneity, smoothness, shape, and extension of the mass; some objective parameters, such as size and attenuation, were used by some criteria. Only one study, using a variety of parameters on both unenhanced and enhanced CT to define a malignant pattern had good test performance (Terzolo, Ali, Osella, et al., 2000). The other studies found overall poor test performance for morphological or subjective CT interpretations. Four studies directly compared different CT measures to diagnose adrenal malignancy; one compared different CT measures to diagnose non-adenomas. Three concluded that attenuation on unenhanced CT was a superior measure to adrenal mass size; one of these also found that attenuation was superior to subjective criteria (Lee, Hahn, Papanicolaou, et al., 1991). One concluded that attenuation on enhanced CT and adrenal mass size were equivalent criteria (Singer, Obuchowski, Einstein, et al., 1994). One concluded that attenuation on delayed enhanced CT (approximately 1 hour after contrast injection) was superior to attenuation on immediate enhanced CT (Korobkin, Brodeur, Francis, et al., 1996).
MRI was evaluated in eight studies with 302 subjects, 38 of whom were reported to have incidentalomas. The rest had adrenal masses found during extraadrenal cancer work up, undefined adrenal masses, and a few with clinically suspected adrenal disease. The studies were small, ranging from 25 to 54 subjects and were generally of fair quality. The studies evaluated numerous methods of performing and analyzing MRIs. In general, test performance was based on the total number of adrenal masses (as opposed to total number of subjects). The studies reported that MRI has the potential to have excellent sensitivity, with fair to very good specificity, to diagnose adrenal malignancy. The best test performance was found using adrenal mass to spleen signal intensity ratio. One study found that MRI has excellent test performance to differentiate adrenal adenomas from non-adenomas. One study concluded that combined unenhanced and enhanced MRI was superior to both CT and unenhanced MRI alone. Two studies concluded that unenhanced MRI had similar, or somewhat superior, test performance as CT. One study concluded that unenhanced CT was superior to T2 measurements on MRI.
The only study that evaluated the combined use of CT and MRI to diagnose adrenal malignancy found near-perfect test performance in patients undergoing cancer work up, but the study was small, evaluating 37 masses in 33 subjects, and had fair quality.
The only study that evaluated US to diagnose adrenal malignancy found poor test performance in patients with incidentalomas. This study had 54 subjects and had poor quality.
The only study that evaluated PET to diagnose adrenal malignancy found perfect test performance. The study included only 20 subjects, some of whom had incidentalomas; others had adrenal masses found during cancer work ups. The study was of fair quality.
Scintigraphy was evaluated in six studies with 399 subjects, of whom about 210 had incidentalomas. The rest had adrenal masses found during cancer work ups. Studies were generally small, ranging from 28 to 229 subjects; five had fewer than 50 subjects. The studies were of variable quality and only one explicitly calculated test performance. The rest reported frequency of diagnoses with different scintigraphy patterns. Malignant masses generally had decreased or incomplete marker uptake, or had scintigraphy patterns that were discordant with CT. However, adenomas frequently had similar patterns. Sensitivity ranged from 71 to 100 percent; specificity ranged from 50 to 100 percent.
FNA was evaluated in nine studies with 515 subjects, of whom about 115 had incidentalomas. The rest had adrenal masses found during cancer work ups, undefined adrenal masses, and a few had clinically suspected adrenal disease. Studies were generally small, ranging from 15 to 270 subjects; six studies had fewer than 30 subjects. The studies all were of poor quality, some with major deficiencies. Some studies did not report sufficient data to confirm test performance or determine prevalence of malignancy. Some reported only test accuracy or sensitivity and accuracy. Five of the studies had either unclear reference standards or used FNA results also as a reference standard. Notwithstanding these caveats, sensitivity ranged from 80 to 100 percent and specificity ranged from 83 to 100 percent. Inconclusive biopsies occurred in six to 50 percent of samples, in both benign and malignant masses. One study found that FNA had higher sensitivity and accuracy in masses at least 3 cm in diameter and when 19 gauge or larger needles were used. There was insufficient evidence to draw conclusions about test performance of different needle biopsy techniques.
Three studies evaluated the test performance of biochemical tests in subjects with incidentalomas. One study included subjects with incidental adrenal adenomas only. The studies included 145 subjects and ranged in size from 31 to 84 subjects. The studies were of poor quality with regard to evaluating biochemical tests. One study found that DHEAS had excellent sensitivity but poor specificity to diagnose malignancy. However, only three subjects had malignancy. Two studies evaluated biochemical tests to predict unilateral uptake on scintigraphy in patients with incidental adenomas. It was implied that the scintigraphy was a marker for autonomous function of an adenoma, although neither study provided clinical results. One study tested six different biochemical tests and found poor sensitivity (0 to 50 percent) and fair to good specificity (62 to 94 percent). The other study found 100 percent sensitivity and 67 percent specificity for morning cortisol after an overnight suppression test. It is important to note that the second study used a lower threshold for lack of suppression.
Overall, CT and MRI have been shown to have the potential to be highly sensitive and specific to differentiate malignant disease from benign in patients with incidentally found adrenal masses. Scintigraphy may also show good test performance, although there was no consensus for what were appropriate definitions of positive and negative tests. FNA appears to be highly accurate to diagnose adrenal masses, although the rate of non-diagnostic biopsies was substantial. Furthermore, the poor quality of the studies and their small size make the conclusions dubious. US, PET and biochemical tests have been evaluated by too few studies to allow accurate conclusions.
With few exceptions, the overall methodological quality of the studies we examined was poor to fair. The heterogeneity of the studies limits the ability to estimate the overall diagnostic performance of each evaluated test. Future high quality studies are needed to properly assess the test performance of various diagnostic tests for adrenal incidentalomas. In contrast to published studies, future studies should clearly define and report eligibility criteria, sample characteristics, test methodology, and definitions of positive and negative tests. Reference standards should be clearly defined, should be completely independent of the tests being investigated, and should, as much as possible, include well defined outcomes, such as surgical or autopsy diagnoses. Given the heterogeneity of definitions of incidentally discovered adrenal masses, it is important to focus research on specific populations of patients. Many studies combined subjects with true incidentalomas (which have very low prevalence of malignancy and which tends to adrenal carcinoma) with subjects who had adrenal masses found during cancer work ups and who have a high prevalence of metastatic disease. Furthermore, many studies did not define how adrenal masses were initially discovered. It may also be of value for future studies to focus on test performance for specific outcomes, such as adrenal cortical cancer or metastasis, as opposed to all cancers combined. Studies of diagnostic test performance of clear, simple populations are much easier to interpret and should be more applicable to patients with the disease of interest. Finally, when possible, ROC curves should be drawn, and complete data reported, so that the reader can determine the proper threshold to use for a given test.
Twelve studies reported complications due to FNA of adrenal masses. Only two explicitly reported data on metastatic spread of cancer along the needle tract. In three studies patients were accrued prospectively, and the remaining nine were retrospective chart reviews, including the two that reported metastatic spread. Seven studies were judged to be of fair quality (quality grade B). These studies either did not fully describe their population, did not fully report biopsy techniques, did not fully report complication definitions or results, or were retrospective. The remaining five studies were judged to be of poor quality (grade C). These studies were too limited in their descriptions of biopsy techniques and complications.
Twelve studies reported a total of 36 complications (including the metastatic spread) in 888 biopsies (866 patients). The overall complication rate was 4.1 percent. No deaths were reported.
The most frequent complication was bleeding (15 events) which comprised adrenal bleeds, hematomas, and hematocrit drops. One patient required adrenalectomy; one patient received a 2 unit blood transfusion (in the early 1980s). One patient had a left-sided adrenal mass that was biopsied with four passes of a 22 gauge needle from a posterior approach. One patient had a right-sided adrenal mass that was biopsied with seven passes of a 22 gauge needle from a transhepatic approach. Five other patients had biopsies done with 18–20 gauge needles. No other data were provided about biopsy technique.
Seven patients developed pneumothoraces, and two of these required chest tube placement. One additional patient developed a hemothorax which also required chest tube placement. Four patients had biopsies with 22 gauge needles. One of these patients had a right-sided mass that was biopsied posteriorly with three passes. Another had a left-sided mass that was biopsied posteriorly with 5 passes. No data were provided about biopsy technique in four patients.
Two patients developed pancreatitis. One was hospitalized for 11 days, the other for 13. One received 2 units of blood. One patient had two passes with a 20 gauge needle and one pass with a 22 gauge needle. One patient had six passes with a 20 gauge needle.
Four patients experienced unusual pain; one required overnight hospitalization. Two patients had hypotension; one was asymptomatic, one received 2 units of blood (though there was no drop in hematocrit). One patient each suffered mild nausea, mild hematuria, and transient bacteremia.
One study which reviewed 277 biopsies performed statistical analysis but found no significant difference in complication rates by either needle or lesion size (Welch, Sheedy, Stephens, et al., 1994).
No study compared complication rates by biopsy technique, including aspiration versus core needle biopsy. The three studies with the highest complication rates either did not describe needle type or biopsy technique, or used multiple needle types and biopsy techniques, including coring (Bernardino, Walther, Phillips, et al., 1985; Mody, Kazerooni, and Korobkin, 1995; Silverman, Mueller, Pinkney, et al., 1993). However, since different definitions of complications were used across studies, one should be cautious of direct comparisons.
In summary, only two studies (of 10 or more subjects) evaluated the risk of metastatic spread due to biopsy of adrenal masses. In these two studies, one of 360 patients (0.3 percent) had needle tract metastases due to adrenal biopsy. This patient had bronchogenic carcinoma. Neither paper provides data on the number of subjects with adrenal carcinoma, so no conclusions can be offered about the risk of needle track metastases from FNA biopsy of adrenal carcinoma. Twelve studies reported on 888 biopsies in 866 patients. A total of 36 complications (4.1 percent) were reported. There were 26 potentially serious complications (bleeding, pneumothorax, hemothorax, pancreatitis, bacteremia). Nine patients (1 percent) received in-hospital treatment for complications due to adrenal biopsy. Because of the wide variety of biopsy techniques (including position, approach, needle size, aspiration, and coring), the commonly unclear or incomplete reporting, and the small study sizes, no reliable estimates can be made about the relative safety of the different biopsy techniques. The largest study found no difference in risk of complication based on needle or lesion size. Overall, study quality was fair to poor.
Further studies are required to determine the risks of both needle tract metastases and major complications. Ideally, future studies should be larger and clearly and completely report patient demographics (e.g., age, body mass index, gender, reason for initial discovery of mass, mass size and location), all relevant aspects of biopsy technique (e.g., position, approach, organs transected, needle size and type, number of passes, radiological tool), biopsy results and final diagnoses, short- and long-term complications, and patient disposition. Statistical analyses should also be performed to determine which patient characteristics and which biopsy techniques are predictors of increased risk of complications.
To date, few randomized trials have ever been conducted comparing the various surgical approaches to adrenalectomy. Most centers perform only one or two approaches. Published data consists primarily of case series, mostly retrospective, sometimes with historical controls, occasionally matched for demographic variables. We have organized the cohort/case series by surgical approach, and the comparative studies (when done) according to the comparison being made.
| Study Year | Tumor Type, % a | Total, N | Mean Tumor Size, cm (Range) | Operative Time, Min (Range) | Mean Blood Loss, mL (Range) | Length of Stay, Days (Range) | Complications | Quality b | ||
|---|---|---|---|---|---|---|---|---|---|---|
| Death | Major | Minor | ||||||||
| Favia 1992 | Pheo 0 | 52 | ND | ND | ND | ND | 0 | 2 | 0 | C |
| CA 0 | ||||||||||
| Met 0 | ||||||||||
| Proye 1993 | Pheo ND | 105 | 3.8 (0.5–9.6) | 132 (45–260) | ND | 7.6 (1–21) | 1 | 12 | 14 | C |
| CA ND | ||||||||||
| Met ND | ||||||||||
| Weigel 1994 | Pheo 0 | 40 | 1.48 | ND | ND | 6 | 3 c | 3 | 3 | B |
| CA 3 | ||||||||||
| Met 0 | ||||||||||
| Fahey 1994 | Pheo 18 | 51 | ND | 84 | ND | ND | 2 d | 14 | 6 | B |
| CA 8 | ||||||||||
| Met 6 | ||||||||||
| van Heerden, 1995 | Pheo 0 | 91 | 4.3 | ND | ND | 6 | 2 e | 3 | 0 | C |
| CA 4 | ||||||||||
| Met 0 | ||||||||||
| Nash 1995 | Pheo 0 | 40 | 1.8 (0.7–2.5) | 200 (130–355) | 232 (30–600) | 4.3 (2–11) | 0 | 5 | 10 | B |
| CA 0 | ||||||||||
| Met 0 | ||||||||||
| Sand 1997 | Pheo 12 | 59 | Median 2.5 (1–7) | ND | 237 (30–4500) | Median 8 (5–21) | 0 | 24 | 3 | C |
| CA 2 | ||||||||||
| Met 1 | ||||||||||
| Kolomecki 1999 | Pheo 41 | 32 | (1.5–14.0) | 90 | ND | 7 | 0 | 3 | 0 | B |
| CA 13 | ||||||||||
| Met 3 | ||||||||||
N, number of subjects; Min, minutes; ND, no data
Pheo, pheochromocytoma; CA, adrenal cortical cancer; Mets, metastatic cancer
See Methods, Question 3, Grading the methodological quality
Aneurysm rupture 5 days post-operative
Patient with severe coronary artery disease, considered high-risk
Deaths occurred within 30 days, determined not to be related to surgery
| Study Year | Procedure/approacha | Total, N b | Tumor Type, % c | Mean Tumor Size, cm (Range) | Operative Time, Min (Range) | Mean Blood Loss, mL (Range) | Length of Stay, Days (Range) | Complications | Quality c | ||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Pheo | CA | Met | Death | Major | Minor | ||||||||
| Russell 1982 | AA | 64 (94) | 0 | 0 | 0 | ND | 95 | ND | 9.3 e | 2 | 15 | 2 | C |
| PA | 39 (52) | 0 | 0 | 0 | 85 | 5.3 e | 0 | 4 | 19 | ||||
| Bruining 1984 | Laparotomy | 32 | ND | ND | ND | ND | 105 | 800 | ND | 0 | 38 | ND | C |
| Lumbotomy | 56 | 111 | 405 | 0 | 6 | ND | |||||||
| Dorsal | 12 | 98 | 288 | 0 | 25 | ND | |||||||
| Irvin 1989 | AA | 17 | 100 | 0 | 0 | 6.8 | ND | ND | 9.8 e | 0 | 47 | 24 | C |
| PA | 20 | 100 | 0 | 0 | 7.0 | 6.1 e | 0 | 5 | 15 | ||||
| Nagesser 2000 | AA | 59 | 24 | 48 | 0 | 160 g e | 160±54 f | 1050 f | 15.6±8.8 f | 6.8% g | 17 | 36 | C |
| PA | 267 | 17 | 6 | 0 | 12 g e | 101±44 f | 300 f (Median) | 12.8±6.7 f | 1.5% g | 10 | 20 | ||
N, number of subjects; Min, minutes; ND, no data
AA, anterior adrenalectomy; PA, posterior adrenalectomy
Number of procedures in parentheses.
Pheo, pheochromocytoma; CA, adrenal cortical cancer; Met, metastatic cancer
See Methods, Question 3, Grading the methodological quality
Statistically significant difference
Statistically significant difference in univariate analysis. In multivariate analysis, only blood loss was significantly correlated with operative approach. Operative time and length of stay were significantly correlated with tumor size only.
Not significantly different (p = 0.06)
In the early 1990s, Gagner, Lacroix, Prinz, et al. (1993) applied the laparoscopic technique to the transperitoneal approach. As with the posterior approach, initial indications were limited due to concerns about bleeding, the safety of removing pheochromocytomas, especially under carbon dioxide insufflation which theoretically might trigger a hypertensive crisis, the inability to do en bloc resections of invasive tumors, and the fear that removing cancers laparoscopically could result in metastatic seeding along the trocar port. As surgeons gained experience, indications for laparoscopic adrenalectomy expanded to include large tumors, pheochromocytomas, and metastases.
| Study Year | Tumor Type, % a | Total, N b | Mean Tumor Size, cm (Range) | Operative Time, Min (Range) | Mean Blood Loss, mL (Range) | Length of Stay, Days (Range) | Complications | Quality c | ||
|---|---|---|---|---|---|---|---|---|---|---|
| Death | Major | Minor | ||||||||
| Takeda 1994 | Pheo 0 | 17 | 7.9 g | 253 | 216 | 12 | 0 | 0 | 6 | C |
| CA 0 | ||||||||||
| Met 0 | ||||||||||
| Rutherford 1995 | Pheo 5 | 60 | ND | 124 | ND | 5.1±0.2 | 0 | 2 | 2 | C |
| CA 2 | 1993: 152 | |||||||||
| Met 0 | 1994: 109 | |||||||||
| Janetschek 1996 | Pheo 33 | 18 | 4.2 (1–8) | 160 | 76 | 4.6 | 0 | 0 | 72 | C |
| CA 0 | (2–7) | |||||||||
| Met 0 | ||||||||||
| Marescaux 1996 | Pheo 0 | 27 | Median 2.0 | 140 | 100–300, 1 with 1000 | 5.3 | 0 | 11 | 19 | B |
| CA 0 | (0.5 – 8.0) | (70–240) | ||||||||
| Met 0 | ||||||||||
| Walmsley 1996 | Pheo 16 | 12 | 4.6 | 120 | ND | 5.3 | 0 | 25 | 17 | C |
| CA 0 | (60–225) | |||||||||
| Met 0 | ||||||||||
| Fernandez-Cruz d 1996 | Pheo 38 | 31 | 3.8 | 137 | 217 | 3.5 | 0 | 0 | 6 | B |
| CA 0 | ||||||||||
| Met 0 | ||||||||||
| De Canniere 1996 | Pheo 25 | 16 | Median 3 | 132 | ND | 6 | 0 | 6 | 19 | B |
| CA 0 | (1.5–5) | (59–360) | (2–13) | |||||||
| Met 19 | ||||||||||
| Gagner e 1997 | Pheo 25 | 88 | 4.95 | 123 | 70 | 2.4 | 0 | 11 | 2 | B |
| CA 3 | (0.7–12) | (80–360) | (<20–1300) | (1–6) | ||||||
| Met 0 | ||||||||||
| Horgan 1997 | Pheo 25 | 16 | ND | 174 unilateral; 480 bilateral (Median) | 15 patients <100 | 3.9 f | 0 | 19 | 19 | C |
| CA 0 | (19) | |||||||||
| Met 0 | ||||||||||
| Terachi 1997 | Pheo 8 | 99 | 1.0–7.0 | 240±76 | 68±80 | 7.2 | 0 | 7 | 8 | C |
| CA 2 | (100) | |||||||||
| Met 0 | ||||||||||
| Shichman 1999 | Pheo 14 | 47 | 7.1 (adrenal size) | Right 210 | 142 | 3 | 0 | 4 | 6 | C |
| CA 0 | (50) | Left 227 | ||||||||
| Met 2 | ||||||||||
| Pujol 1999 | Pheo 20 | 27 | 3.1 | 156 g | ND | 3 | 0 | 7 | 7 | C |
| CA 0 | (30) | |||||||||
| Met 3 | ||||||||||
| Lucas 1999 | Pheo 7 | 36 | 2.5 | 262 overall; 194 unilateral | 116 overall; 108 unilateral | 2.0 overall; 1.1 unilateral; 3.0 bilateral | 0 | 10 | 10 | B |
| CA 0 | (42) | (1.0–8.0) | ||||||||
| Met 0 | ||||||||||
| Henry 2000 | Pheo 17 | 159 | 3.2 | 129 | ND | 5.4 | 0 | 2.5 | 5 | I |
| CA 2 | (169) | (0.7–11.0) | (48–300) | (3–15) | ||||||
| Met 3 | ||||||||||
| Lezoche h 2000 | Pheo 12 | 102 | 5.1 right | 85 right | ND | 2.5 | 0 | 5 | 0 | B |
| CA 0 | (108) | 5.8 left | 112 left | (2–3) | ||||||
| Met 3 | (3.5–12) | |||||||||
| Ishikawa 2000 | Pheo 5 | 55 | 3.0±1.6 | 143±42 | 49±50 | ND | 0 | 0 | 0 | I |
| CA 0 | (1.0–6.5) | 185 (first 14) | 92 (first 14) | |||||||
| Met 0 | 120 (last 41) | 32 (last 41) | ||||||||
| Guazzoni 2001 | Pheo 20 | 145 | 8 | 160 | ND | 2.8±0.9 | 0 | 5 | 0 | I |
| CA 0 | (161) | (7–10) | ||||||||
| Met 3 | ||||||||||
| Pisanu 2001 | Pheo 21 | 84 | 147 | ND | 5.0 | 0 | ND | ND | I | |
| CA 0 | (91) | 0.5–7.0 | (55–310) | (1–23) | ||||||
| Met 0 | ||||||||||
| Porpiglia 2001 | Pheo 11 | 72 | 5.0±2.3 | 135 | 80 | 4.3±2.0 | 1 | 17 | 3 | B |
| CA 1 | (76) | (1.8–11.0) | (85–200) | (50–700) | ||||||
| Met 0 | ||||||||||
| Valeri 2001 | Pheo 12 | 78 | 4.0 cm | 120 | 200 | 3.5 | 1 | 5 | 1 | B |
| CA 4 | (1–10) | (60–210) | ||||||||
| Met 6 | ||||||||||
N, number of subjects; Min, minutes; ND, no data
Pheo, pheochromocytoma; CA, adrenal cortical cancer; Mets, metastatic cancer
Number of procedures in parentheses.
See Methods, Question 3, Grading the methodological quality
Compared mixed laparoscopic techniques for pheo and non-pheo. No difference was found in operating time or blood loss. Patients with Cushing's disease had longer operating times and more blood loss than other patients.
Includes La Croix 1998, Weisnagel 1996 and Gagner 1993.
With the exception of one outlier, the mean was only 2.5
Includes time for other procedures.
Includes Filipponi 1998
| Study Year | Tumor Type, % a | Total, N b | Mean Tumor Size, cm (Range) | Operative Time, Min (Range) | Mean Blood Loss, mL (Range) | Length of Stay, Days (Range) | Complications | Quality c | ||
|---|---|---|---|---|---|---|---|---|---|---|
| Death | Major | Minor | ||||||||
| Takeda 1997 | Pheo 0 | 11 | 1.7 | 248 | 151 | ND | 0 | 0 | 63 | B |
| CA 0 | ||||||||||
| Met 0 | ||||||||||
| Gasman 1998 | Pheo 9 | 22 | 2.6 | 97 | 70 | 3.4 | 0 | 0 | 9 | B |
| CA 0 | (23) | (1.0–4.0) | (45–160) | (0–450) | ||||||
| Met 5 | ||||||||||
| Baba 1999 | Pheo 1 | 68 | ND | 144 | 43.5 | ND | 0 | ND | ND | C |
| CA 0 | ||||||||||
| Met 0 | ||||||||||
| Fazeli-Matin 1999 | Pheo 10 | 10 | ND | 180 | 50 | <1.0 | 0 | 10 | 0 | C |
| CA 0 | (150–270) | (20–390) | (<1–3) | |||||||
| Met 10 | ||||||||||
| Lee 2000 | Pheo 3 | 30 | 2.2 | 155 | <100 | 5.5 | 0 | 3 | 3 | C |
| CA 0 | (1–4) | (4 -8) | ||||||||
| Met 0 | ||||||||||
| Subramaniam 2000 | Pheo 100 | 11 | ND | 210–240 d | 240±200 | ND | 0 | ND | ND | B |
| CA 0 | (16) | |||||||||
| Met 0 | ||||||||||
| Tanaka 2000 | Pheo 19 | 54 | 2.65 | 227 | 176 | 10.6 | 0 | ND | ND | B |
| CA 0 | ||||||||||
| Met 0 | ||||||||||
| Bonjer 2000 | Pheo 20 | 95 (111) | 3.4 (0.2–7.0) | 118 unilateral; 214+26 bilateral | 66 unilateral; 121±39 bilateral | 2.0 (1.5–5.0) unilateral; 12.5 (7.3–16) conversion; 2.7 combined | 1 | 2 | 7 | B |
| CA 1 | ||||||||||
| Met 1 | ||||||||||
| Salomon e2001 | Pheo 18 | 106 (115) | 3.1 (1.0–6.2) | 118 (45–240) | 77 (0–550) | ND | 0 | 8 | 10 | A |
| CA 0 | ||||||||||
| Met 3 | ||||||||||
| Walz f 2001 | Pheo 22 | 130 (142) | 2.7±1.4 (0.5 – 7.0) | 101±44 g 78 (last 50) | 54±72 | Median 3 (1–21) | 0 | 0 | 19 | B |
| CA 0 | ||||||||||
| Met 2 | ||||||||||
N, number of subjects; Min, minutes; ND, no data
Pheo, pheochromocytoma; CA, adrenal cortical cancer; Mets, metastatic cancer
Number of procedures in parentheses.
See Methods, Question 3, Grading the methodological quality
In successful laparotomies. In 3 cases converted to open procedure, blood loss averaged 2000±700
Includes Souliq 2000
Includes Walz 1996
Conversions excluded
| Study Year | Procedure/ approach a | Total, N b | Tumor Type % c | Mean Tumor Size, cm (Range) | Operative Time, Min (Range) | Mean Blood Loss, mL (Range) | Length of Stay, Days (Range) | Complications | Quality d | ||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Pheo | CA | Met | Death | Major | Minor | ||||||||
| Prinz e1995 | AA | 11 | 82 | 0 | 0 | 5.1 | 174±41 | 391±88 f | 6.4±1.5 f | 0 | ND | ND | C |
| PA | 13 | 0 | 0 | 8 | 3.1 | 139±36 f | 288±118 | 5.5±2.9 f | 0 | ||||
| LTLA | 10 | 40 | 0 | 0 | 4.8 | 212±77 f | 228±66 f | 2.1±0.9 f | 0 | ||||
| Guazzoni g1995 | AA +PA | 20 | 50 | 0 | 0 | 3.0 | 145 | 450 | 9.0 | 0 | 30 | 25 | C |
| LTLA | 20 | 35 | 0 | 0 | 3.0 | 170 | 100 | 3.4 | 0 | 5 | 0 | ||
| Brunt 1996 | AA | 25 | 72 | 0 | 0 | 3.4 h | 142 f | 408 f | 8.7 f | 4 | 28 | 44 | C |
| PA | 17 | 6 | 0 | 0 | 2.4 h | 136 f | 366 f | 6.2 | 0 | 12 | 41 | ||
| LTLA | 24 | 44 | 0 | 0 | 2.7 h | 183 f | 104 f | 3.2 f | 0 | 8 | 8 | ||
| MacGillivray 1996 | AA + PA | 9 | 11 | 0 | 0 | ND | 201±108 f | 500±574 h | 7.9±4.9 f | 0 | 60 | 40 | C |
| LTLA | 14 | 7 | 0 | 0 | ND | 289±86 f | 198±144 h | 3.0±1.6 f | 14 | 7 | |||
| Staren e1996 | Open mixed | 19 | 63 | 21 | 0 | 9.2 (2–25) | 177 | ND | 6.1 (4–8) f | 0 | ND | ND | C |
| Lap mixed | 20 | 35 | 0 | 0 | 3.8 (1–8) | 206 | 2.2 (1–7) f | 0 | |||||
| Mugiya 1996 | Open | 7 | 0 | 0 | 14 | 3.9±1.4 | 131±50 | 251±213 | ND | 0 | ND | ND | I |
| ATLA | 12 | 0 | 0 | 0 | 3.8±1.9 | 197±35 | 130±209 | ND | 0 | 8 | ND | ||
| Bonjer I1997 | PA | 12 | 0 | 0 | 0 | 3.0 | 60 e | 175 f | 6 f | 0 | 16 | 0 | B |
| ATLA | 9 | 0 | 0 | 0 | 3.0 | 150 e | 150 f | 6 f | 0 | 22 | 0 | ||
| RLA | 9 | 33 | 0 | 0 | 4.0 | 75 | 20 | 4 | 0 | 0 | 0 | ||
| (median) | (median) | (median) | (median) | ||||||||||
| Linos 1997 | AA | 86 | 28 | 12 | 1 | 8.07 | 155.3 f | 8.1% | 8.0 f | 0 | 14 | 0 | C |
| PA | 61 | 16 | 0 | 0 | 5.25 | 108.6 | 3.2% | 4.5 f | 0 | 8 | 2 | ||
| TLA | 18 | 17 | 0 | 11 | 4.03 | 116.1 | 5.5% | 2.3 | 0 | 6 | 0 | ||
| (Transfusion) | |||||||||||||
| Aldrighetti 1997 | AA | 12 | 0 | 0 | 0 | 4.6±0.3 f | 174±18 h | ND | 7.9±1.0 f | 0 | 16 | 0 | C |
| TLA | 8 | 0 | 0 | 0 | 3.4±0.4 f | 150±18 h | ND | 5.3±0.4 f | 0 | 12 | 12 | ||
| Vargas 1997 | Open | 20 | 30 | 3 | 0 | 11.5±4 h | 178±17 h | 283±62 h | 7.2±0.6 f | 0 | 25 h | 0 | B |
| LTLA | 20 | 30 | 3 | 0 | 13.9±6 h | 193±14 h | 245±52 h | 3.1±0.3 | 0 | 10 h | 0 | ||
| Jacobs 1997 | AA + PA | 19 | 26 | 0 | 10 | 3.3±1.9 f | 151±63 h | 263±242 f | 5.1±1.7 f | 0 | 32 f | 0 | B |
| LTLA | 19 | 15 | 0 | 0 | 2.5±1.2 f | 164±107 h | 109±75 f | 2.3±0.9 f | 0 | 5 f | 0 | ||
| Ishikawa 1997 | PA | 9 | 0 | 0 | 0 | 2.1±0.5 | 99±29 f | 126±54 | ND | 0 | ND | ND | C |
| AA | 10 | 60 | 0 | 0 | 4.2±1.5 f | 143±35 f | 407±285 f | ND | 0 | ND | ND | ||
| LTLA | 14 | 0 | 0 | 0 | 2.6±1.3 f | 185±19 f | 92±46 f | ND | 0 | ND | ND | ||
| Korman 1997 | Open | 10 | 20 | 10 | 10 | 6.1±2.8 | 124±29 f | 210±173 h | 5.9±1.1 f | 0 | 20 | 0 | B |
| ATLA | 10 | 50 | 0 | 10 | 2.9±2.0 | 164±47 f | 118±158 h | 4.1±2.5 f | 0 | 10 | 0 | ||
| Thompson 1997 | PA | 50 | 14 | 0 | 0 | 2.9 | 127 f | ND | 5.7 f | 0 | 18 h, j | 58 f, j | C |
| LTLA | 50 | 20 | 0 | 0 | 2.9 | 167 f | 3.1 f | 0 | 6 h, j | 0 f, j | |||
| Yoshimura k1998 | Open | 25 | 28 | 4 | 0 | 3.7 | 133±1 f | 345±84 h | 18.2±3.5 f | 0 | 20 | 8 | C |
| Lap | 28 | 4 | 0 | 0 | 1.9 | 376±30 f | 370±134 h | 12.0±0.7 f | 0 | 14 | 21 | ||
| Ting 1998 | PA | 56 | 0 | 0 | 0 | 1.6 a | 120 f | 150 f | 5 (2–17) f | 0 | 4 | 6 | C |
| TLA | 12 | 0 | 0 | 0 | 1.8 a | 160 f | 50f | 3 (2–5) f | 0 | 0 | 0 | ||
| Median | Median | Median | |||||||||||
| Winfield 1998 | Open | 17 | 41 | 0 | 0 | 2.5 h, L | 233 f | 266 h | 6.2 f | 0 | 35 | 35 | C |
| (165–310) | (100–600) | (4–10) | |||||||||||
| LTLA | 21 | 0 | 0 | 10 | 1.8 h, L | 309 f | 183 h | 2.7 h | 0 | 19 | 10 | ||
| (185–480) | (30–500) | (2–9) | |||||||||||
| Bonjer I1998 | PA | 30 | 19 | 0 | 0 | 3.0 | 60 f | 125 f | 7 f | 0 | 3 m | 7 m | C |
| RLA | 42 | 38 | 0 | 0 | 3.0 | 90 f | 20 f | 4 f | 0 | 12 m | 7 m | ||
| Soares 1999 | Open | 8 | 38 | 0 | 0 | 6.4 f | 166 f | 278 f | 5.4 f | 0 | 0 | 0 | C |
| LTLA | 11 | 9 | 0 | 0 | 2.0 f | 116 f | 132 f | 2.1 f | 0 | 9 | 0 | ||
| Dudley 1999 | LTLA | 31 | 42 | 0 | 0 | (1.6–8.2) | 158 | ND | 3.5 | 3 n | 6 | 0 | C |
| PA | 15 | 7 | 0 | 0 | (1.0–5.0) | 85 | 8.5 | 0 | 60 | 20 | |||
| Schell 1999 | LTLA | 22 | 18 | 0 | 0 | ND | ND | 1.7±1.0 f | 0 | 0 | 0 | C | |
| Open | 17 | 41 | 24 | 0 | 7.8±6.7 f | 0 | 12 | 6 | |||||
| Imai 1999 | LTLA | 40 | 20 | 0 | 0 | 2.8±1.7 | 180 f, o | 40 f | 12 f | 0 | 5 | 0 | B |
| PA | 40 | 18 | 0 | 0 | 2.7±1.4 | 127 f, o | 162 f | 18 f | 0 | 3 | 38 | ||
| Hobart 1999 | Needlescopic | 15 | 20 | 0 | 0 | 3.0 | 163±39 h | ND | 1.0±0.3 f | ND | ND | ND | C |
| AA | 15 | 20 | 0 | 0 | 3.7 | 136±24 h | 7.0±1.9 f | ND | ND | ND | |||
| Shen 1999 | Mixed open | 38 | 0 | 3 | 0 | ND | ND | ND | ND | 0 | 0 | 11 | C |
| Mixed LA | 42 | 0 | 0 | 0 | 0 | 0 | 0 | ||||||
| Inabnet 2000 | LTLA | 11 | 100 | 0 | 0 | 4.1±1.2 h | 146±36 h | ND | 5.5±2.2 h | 0 | 0 | 0 | C |
| AA | 11 | 100 | 0 | 0 | 4.6±1.2 h | 153±55 h | 6.1±1.6 h | 0 | 9 | 0 | |||
| Hobart 2000 | TLA or RLA | 14 | 36 | 29 | 7 | 8.0 f | 205 h | 400 f | 2.4 | 0 | 21 | 0 | C |
| Open | 14 | 58 | 0 | 0 | 7.8 | 216 h | 584 | 7.7 f | 0 | 57 | 21 | ||
| TLA or RLA | 45 | 36 | 0 | 4 | 2.2 f | 158 h | 113 f | 1.5 f | 0 | 7 | 2 | ||
| Rayan 2000 | LTLA | 19 | 5 | 0 | 0 | 3.3±1.9 f | 198 h | ND | 1.5 f | 0 | 5 | ND | C |
| Open | 48 | 40 | 0 | 0 | 5.2±2.9 f | 228 h | 6.3 f | 0 | 29 | ND | |||
| Sprung 2000 | LTLA | 14 | 100 | 0 | 0 | ND | 177±59 h | 100 f | 3.0 f | 0 | ND | ND | C |
| AA | 20 | 100 | 0 | 0 | ND | 196±69 h | 400 f | 7.5 f | 0 | ND | ND | ||
| (median) | |||||||||||||
N, number of subjects; Min, minutes; ND, no data
AA, anterior adrenalectomy; PA, posterior adrenalectomy; TLA, transperitoneal laparoscopic adrenalectomy (approach mixed or not specified); ATLA, anterior transperitoneal laparoscopic adrenalectomy; LTLA, lateral transperitoneal laparoscopic adrenalectomy; RLA, retroperitoneal laparoscopic adrenalectomy.
Number of procedures in parentheses.
Pheo, pheochromocytoma; CA, adrenal cortical cancer; Met, metastatic cancer
See Methods, Question 3, Grading the methodological quality
There is considerable overlap between Prinz, 1995 and Staren, 1996
Statistically significant
Includes Guazzoni 1994
Not significant
There is considerable overlap between Bonjer, 1997 and Bonjer 1998
Early and late complications, instead of major and minor
See Miyake
Mean weights (15.6 g vs. 45.5 g) were significantly different
Significant difference for all complications combined
Operation was uncomplicated. Patient had end-stage Cushing's disease and died at home on post-operative day 8. No cause of death was found.
Skin-to-skin time (34 patients)
| Study Year | Procedure/approach a | Total, N b | Tumor Type % c | Mean Tumor Size, cm (Range) | Operative Time, Min (Range) | Mean Blood Loss, mL (Range) | Length of Stay, Days (Range) | Complications | Quality d | ||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Pheo | CA | Met | Death | Major | Minor | ||||||||
| Fernandez-Cruz 1996 | LTLA | 7 (10) | 0 | 0 | 0 | 5.4 | 89±23 | 160±74 | 3.0±0.7 | 0 | 0 | 29 | A |
| RLA | 8 (11) | 0 | 0 | 0 | 5.3 | 105±19 | 180±84 | 2.8±0.8 | 0 | 0 | 0 | ||
| Duh 1996 | LTLA | 22(23) | 14 | 0 | 23 | 5.3 (1–13) | 226 | ND | 2.2 (1–5) | 3 e | 0 | 5 | B |
| RLA | 14 | 7 | 0 | 7 | 2.6 (1–6) | 202 | 1.5 (1–4) | 14 | 0 | ||||
| Miyake 1998 | ATLA | 17 | 6 | 0 | 0 | 1.92±1.0 | 423±184 f | 406 f | 12.8±4.0 | 0 | 18 | 24 | C |
| RLA | 12 | 0 | 0 | 0 | 1.96±0.6 | 300±73 f | 164 f | 10.3±3.7 | 0 | 8 | 8 | ||
| Gill 1998 | TNA | 15 | 13 | 0 | 7 | 41.6 g f | 169 f | 61 f | 1.1 f | 0 | 0 | 8 | C |
| LTLA | 21 | 0 | 0 | 10 | 15.7 g f | 220 f | 183 f | 2.7 f | 0 | 20 | 11 | ||
| Fernandez-Cruz 1999 | LTLA | 19 | 0 | 0 | 0 | ND | 89±23 | 160 | 3.0±0.7 | 0 | 0 | 11 | B |
| RLA | 17 | 0 | 0 | 0 | ND | 105±19 | 180 | 2.8±0.8 | 0 | 0 | 0 | ||
| Henry 1999 | LTLA g | 102 | 11 | 0 | 0 | 2.1 | 131 h | ND | 5.6 | 0 | 3 | 4 | B |
| LTLA g | 48 | 28 | 6 | 6 | 5.2 | 129 h | 5.1 | 0 | 0 | 8 | |||
| Terachi 2000 | TLA | 311 | 4 i | 0 i | 1 i | ND | ND | ND | ND | 0 | 6 | 9 | I |
| RLA | 59 | 0 | 5 | 10 | |||||||||
| Takeda 2000 | ATLA | 52 | 7 | 0 | 0 | ND | 206±81 h | 153±116 h | ND | 0 | 0 | ND | I |
| RLA | 24 | 0 | 0 | 0 | 260±79 h | 116±98 h | 0 | 0 | ND | ||||
| Suzuki 2001j | ATLA | 46 | 6 | 0 | 0 | 3.8±0.4 f | 201.1 f | 147.0 f | ND | 0 | 0 | 25 | B |
| LTLA | 32 | 0 | 0 | 0 | 3.5±0.4 f | 122.9 f | 68.1 f | 0 | 4 | 20 | |||
| RLA | 40 | 0 | 0 | 0 | 2.3±0.2 f | 174.6 f | 60.7 f | 0 | 0 | 11 | |||
N, number of subjects; Min, minutes; ND, no data
TLA, transperitoneal laparoscopic adrenalectomy (approach mixed or not specified); ATLA, anterior transperitoneal laparoscopic adrenalectomy; LTLA, lateral transperitoneal laparoscopic adrenalectomy; RLA, retroperitoneal laparoscopic adrenalectomy; TNA, transperitoneal needlescopic adrenalectomy.
Number of procedures in parentheses.
Pheo, pheochromocytoma; CA, adrenal cortical cancer; Met, metastatic cancer
See Methods, Question 3, Grading the methodological quality
Combined mortality for both groups
Statistically significant
Subset of Henry 2000. Compares subsets with small and large tumors.
Not statistically significant
Overall-no breakdown available by approach.
Includes Suzuki 1999, Suzuki 1995, and Suzuki 1993
Most studies did not include any information on patient co-morbidities. However, almost all studies list the indication for surgery. Because adrenal tumors are rare, most series include a heterogeneous mixture of indications. It is important to note that series are not randomized and rarely include all adrenalectomies at an institution. Therefore, for the sake of comparing studies or arms within the same study, we list the major indications for surgery, as well as the mean tumor size and range, if reported. We list the percentage of tumors found to be pheochromocytomas, carcinomas and metastases, as these tumors would be expected to have the longest operating time and highest complication rates, independent of operating approach.
One initial concern about laparoscopy was the prolonged operating time, especially in the earliest studies. Although most studies report operating time, comparison between studies is complicated by the lack of a standard definition of operating time. Some authors report total operating room time, including induction of anesthesia and patient positioning, which may be time consuming in certain approaches. Others report skin-to-skin, while others, in an attempt to negate the effects of multiple simultaneous operations (e.g. adrenalectomy plus cholecystectomy), report skin-to-gland-removal. Finally, some series include bilateral adrenalectomy, which always takes longer and may skew the average upward, if not reported separately. Unfortunately, the vast majority of studies do not specify how operating time is reported, making meaningful comparison impossible.
Most studies report mean blood loss during the operation, as well as the range. Although this is easy to measure, its clinical significance is unclear, especially as it does not take into account post-surgical bleeding which may be directly related to approach. A more meaningful measure might be post-op change in hematocrit or number of transfusions required. The former is never routinely reported. Number of transfusions is frequently reported; however, the decision to transfuse is not objective, but dependent on the surgeon. Moreover, the number of transfusions is so small that no study could demonstrate a significant difference between approaches.
Length of stay (LOS) offers a convenient and objective measure of overall morbidity associated with the operating technique. Other measures, such as time to ambulation or time to full recovery are less clearly defined and not universally reported. In collecting such data retrospectively, errors may occur. However, comparing LOS among studies poses its own set of problems. LOS varies greatly from country to country as different health care systems have different cost structures and patient expectations. Hospital stays in Japan, for example, are particularly long compared to the United States. Japanese authors explain this discrepancy due to the low cost of a hospital day and the patient's expectation that he will leave the hospital completely well (Imai T, Kikumori T, Ohiwa M, et al, 1999). LOS may also be confounded in patients with biochemically active tumors that frequently require post-op medical therapy related to hormone replacement which may prolong the hospital stay beyond the time required for surgical recovery. Finally, LOS has been decreasing globally throughout the 1980s and 1990s, the period encompassing all these studies. Comparative studies using historical controls may find that decreased LOS associated with newer procedures is exaggerated by secular trends in hospital reimbursement.
Most studies report rates of complications, though the amount of detail varies from study to study. Authors who simply report that “there were no major complications” may employ different standards from those who report the introduction of post-op drains as a complication. Unfortunately, there is no standard scale for reporting operating complications. We compared complication rates using a scale adapted from Clavien, Sanabria, and Strasberg (1992), even though only one author actually used that scale. Since most studies were retrospective, complications had to be discovered by chart review. We noted that different authors reporting on the same series of patients sometimes reported slightly different complication rates. (Gagner, Pomp, Heniford, et al., 1997; Weisnagel, Gagner, Breton, et al., 1996; Wells, Merke, Cutler, Jr., et al., 1998).
For each procedure there appears to be a learning curve. As the surgeon becomes more proficient with the procedure, there usually follows a decrease in operating time, blood loss, complication rates, and the number of conversions (midway through the procedure) to a more familiar approach. (Filipponi, Guerrieri, Arnaldi, et al., 1998; Henry, Defechereux, Raffaelli, et al., 2000; Ishikawa, Inaba, Nishiguchi, et al., 2000; Rutherford, Gordon, Stowasser, et al., 1995)
Malmaeus, Markaes, Oberg, et al. (1986) did a retrospective review of all the adrenalectomies at his institution between 1976 and 1984, all by the transabdominal approach. In addition, seven comparative studies included 289 patients operated on by the transabdominal route (Aldrighetti, Giacomelli, Calori, et al., 1997; Bruining, Lamberts, Ong, et al., 1984; Brunt, Doherty, Norton, et al., 1996; Inabnet, Pitre, Bernard, et al., 2000; Irvin, Fishman, Sher, et al., 1989; Linos, Stylopoulos, Boukis, et al., 1997; Russell, Hamberger, van Heerden, et al., 1982). Tumors tended to be large, and the proportion of patients with adrenal carcinoma was as high as 48 percent in one study (Nagesser, Kievit, Hermans, et al., 2000). Mean operating time ranged from 95 to 160 minutes, and mean blood loss ranged from 391 to 1,050 ml. Rates of major complications ranged from nine to 47 percent. Mortality ranged from zero to 6.8 percent. The most common complications were accidental injuries to the spleen (necessitating splenectomy), hemorrhage, wound infection, and pneumonia.
From 1992 to 1999 there were eight series, involving 470 patients, examining PA. None was prospective. Most studies had few or no cancers or pheochromcytomas, and mean tumor size ranged from 1.5–3.8 cm. Kolomecki, Pomorski, Kuzdak, et al. (1999), in a study of incidentalomas, included more difficult tumors such as adrenocortical carcinomas and tumors as large as 14 cm. Only four studies reported operating time, with a mean ranging from 84–200 minutes, and an overall range of 45–355 minutes. Only two studies reported mean blood loss—200 ml in Nash, Leibovitch, and Donohue (1995) and 237 ml in Sand, Saaristo, Nordback, et al. (1997). Major complications occurred in 2–24 percent of cases, minor complications in 0–14 percent. Mortality ranged from 0–3 percent, though all deaths occurred in high-risk cases, or were considered unrelated to surgery, according to the authors. The most common complications of PA included pleural tear, wound infection, chronic pain, and bleeding.
Four studies compare the transperitoneal or abdominal approach (AA) with PA in 564 patients. In general the quality of the studies was poor. All studies were retrospective and none matched cases for tumor size or type. All but one neglected to report on one or more of the outcome parameters. The three studies that measured LOS (Irvin, Fishman, Sher, et al., 1989; Nagesser, Kievit, Hermans, et al., 2000; Russell, Hamberger, van Heerden, et al., 1982) all found that mean LOS for PA was significantly shorter than for AA. The sole study to find a significant difference in the other parameters (Nagesser, Kievit, Hermans, et al. 2000) found PA quicker and with less blood loss as a result, but tumors in PA patients were less than one-tenth the size of those in AA patients (12 g versus 160 g). Indeed, in multivariate analysis for unilateral adrenalectomy, all advantages for PA disappeared except for blood loss. For this group of studies, mean operating time, blood loss, LOS, complication rates and types of complications were similar to those seen in the cohort series.
There are 20 cohort series of the transperitoneal approach, including both anterior and flank approaches, incorporating 1,189 patients. Most studies were retrospective with considerable selection bias and/or reporting bias. Half the studies included at least some cancers, and almost all included pheochromocytomas, with two studies (Janetschek, Altarac, Finkenstedt, et al., 1996; Wells, Merke, Cutler, Jr., et al., 1998) reporting that one-third of their operations were performed for pheochromocytoma. Few studies included adrenocortical carcinoma, and these tumors never accounted for more than 4 percent of the operations. Mean tumor size ranged from 2.0 to 5.1 cm and almost all series included tumors over 6 cm. Mean operating time ranged from 92–253 minutes, blood loss from 68–216 ml, and LOS from 2–12 days. There were two post-operative deaths. Major complications occurred in 0–25 percent and minor complications in 0–72 percent of cases. The most common complications were bleeding, conversion to open surgery, hypotension/ hypertension, and wound infection. A number of authors reported learning curves, whereby the operating time, blood loss, LOS and complication rates all decreased over time (Gagner, Pomp, Heniford, et al., 1997; Ishikawa, Inaba, Nishiguchi, et al., 2000; Lezoche, Guerrieri, Paganini, et al., 2000; Rutherford, Gordon, Stowasser, et al., 1995; Shichman, Herndon, Sosa, et al., 1999). In the four studies including at least 100 operations (Guazzoni, Cestari, Montorsi, et al., 2001; Henry, Defechereux, Raffaelli, et al., 2000; Lezoche, Guerrieri, Paganini, et al., 2000; Terachi, Matsuda, Terai, et al., 1997), major complications occurred in 2.5–7 percent and minor complications in 0–8 percent of cases.
There are 10 cohort series of the retroperitoneal approach, incorporating 537 patients. Most studies were retrospective, but overall quality was better than in previous series, including one prospective study (Salomon, Soule, Mouly, et al., 2001) which featured consecutive cases and reported outcome measures for both successful and failed laparoscopies. Almost half the studies included at least one case of metastasis, but there was only a single case of adrenocortical carcinoma in the entire series. Mean tumor size ranged from 1.7–3.4 cm and there were no tumors larger than 7.0 cm. Mean operating time ranged from 97–248 min, blood loss from 40–240 ml and LOS from <1–11 days. One patient died. Major complications occurred in 0–10 percent and minor complications in 0–63 percent of cases. The most common complications were retroperitoneal hematoma, subcutaneous emphysema, conversion to open or laparoscopic surgery, and pancreatic or splenic injury.
There are 28 studies, incorporating 1,388 patients, which compared open to laparoscopic approaches to adrenalectomy. No studies were entirely prospective, and none were randomized. All studies consisted of case series, collected prospectively or retrospectively, compared with historical controls, and occasionally matched for surgical indication and tumor size. Most studies suffer from considerable selection bias and/or reporting bias. Comparison groups vary among studies. Transperitoneal laparoscopic adrenalectomy (TLA) is compared to the anterior open approach (AA) in eight studies, posterior open approach (PA) in nine studies, or a mixture of the two in nine other studies. Retroperitoneal laparoscopic adrenalectomy (RLA) is compared to PA in two studies. A mixture of laparoscopic approaches is compared with a mixture of open approaches in four studies. Because most studies did not use matched controls, tumor sizes and types are often not comparable between study arms. In general, open approaches had a higher percentage of pheochromocytomas and adrenocortical carcinomas and a larger tumor size than laparoscopic approaches. In only half the studies that compared TLA to PA were there comparable tumor size and type in both arms. TLA and AA had similar operating times, but TLA resulted in less blood loss (four of five studies) and shorter LOS (six of seven studies). Most studies listed all complications but did not apply statistical methods to comparing the complication rates. Six studies did compare rates statistically. Sprung, O'Hara, Jr., Gill, et al., (2000) found a higher rate of hypotension for AA compared to TLA for pheochromocytoma. Jacobs, Goldstein, and Geer (1997) found fewer major complications from TLA when compared with a mixture of open approaches. Bonjer, van der Harst, Steyerberg, et al. (1998) found fewer overall complications from RLA than from PA. On the other hand, Linos, Stylopoulos, Boukis, et al. (1997) and Vargas, Kavoussi, Bartlett, et al. (1997) found no statistical difference in complications among AA, PA and TLA. Finally, Thompson, Grant, van Heerden, et al. (1997) found a comparable rate of early complications between TLA and PA, but an increased rate of late complications with PA, specifically muscle weakness and pain at the incision site. The authors, who sent questionnaires asking about these late complications, hypothesize that they are frequently underreported. PA was significantly quicker than TLA in eight of nine studies and RLA in one of two studies, but TLA resulted in less blood loss (three of six studies) and a shorter LOS (seven of eight studies). The remaining studies did not find any significant differences between specific techniques. Both AA and TLA each resulted in one death, neither of which was considered by the authors to be related to the surgery. There were no deaths from PA or RLA. Studies comparing mixed open approaches or mixed laparoscopic techniques had similar findings; laparoscopy usually took longer, but resulted in less blood loss and a shorter LOS.
There are eight studies comparing different laparoscopic approaches to adrenalectomy, incorporating 716 patients, as well as one study (Henry, Defechereux, Gramatica, et al., 1999) which compared surgery for large and small tumors using TLA in 150 patients. Overall quality of the studies was generally better than in the other series, and there was one prospective, randomized controlled trial (Fernandez-Cruz, Saenz, Benarroch, et al., 1996), albeit a small one. A few studies collected data prospectively, but because the surgeon still dictated the choice of approach, it is possible that considerable bias may have been introduced. One author (Suzuki, Kageyama, Hirano, et al., 2001) conducted a background-matched analysis on a subset of his patients.
In studies of TLA and lateral transperitoneal laparoscopic adrenalectomy (LTLA) vs. RLA, patients were generally well matched concerning tumor size and type. In each series, two studies found no difference in operating time or blood loss, while one study found RLA to be quicker than TLA (Miyake, Yoshimura, Yoshioka, et al., 1998), and another found that LTLA was quicker than either TLA or RLA (Suzuki, Kageyama, Hirano, et al., 2001). None of the studies demonstrated any difference in LOS, and none applied statistical methods to complication rates.
One study compared needlescopic surgery to traditional TLA. Although the tumors removed needlescopically were both larger and contained a higher percentage of pheochromocytomas, the needlescopic group had a shorter operating time, less blood loss and a shorter hospitalization. There were fewer complications in the needlescopic group, but due to the small sample size, the difference did not reach statistical significance.
Finally, Henry, Defechereux, Gramatica, et al. (1999) compared surgery for large (mean 5.2 cm) and small (mean 2.1 cm) tumors and found no difference in operating time, length of stay or complication rate. Moreover, the large tumor group contained 28 percent pheochromocytomas and 6 percent carcinomas that were subsequently converted to open procedures.
We evaluated nine case series of open adrenalectomies, four studies comparing open adrenalectomy techniques, 20 case series of TLA, 10 studies of RLA, 28 studies comparing laparoscopic adrenalectomy to open surgery, and nine studies comparing TLA with RLA.
Nine series of open adrenalectomy included 525 patients. Complications were common. Major complications occurred in 2 to 24 percent of patients and minor complications in 0 to 14 percent. Mortality occurred in 0 to 3 percent of patients. Although all deaths occurred in high-risk patients, they were not considered to be directly related to surgery, according to the authors. The most common complications included pleural tear, wound infection, bleeding and urinary tract infection. Four studies compared open abdominal with open retroperitoneal or posterior approaches in 566 patients. Average LOS was shorter for the posterior approach. In a multivariate analysis, only blood loss was lower for PA.
Twenty series of transperitoneal laparoscopic adrenalectomy included 1,189 patients. Major complications occurred in 0 to 25 percent of patients and minor complications occurred in 0 to 72 percent. Two deaths were reported. The most common complications were bleeding, conversion to open surgery, low or high blood pressure, and wound infection. Of note, in the larger studies both major and minor complications were less common.
Ten series of RLA included 537 patients. Major complications occurred in 0 to 10 percent of patients and minor complications occurred in 0 to 63 percent of patients. One patient died. The most common complications included retroperiotoneal hematoma, subcutaneous emphysema, conversion to a different surgery, pancreatic and splenic injury.
Twenty-eight series compared open and laparoscopic surgery in 1,388 patients. One study found a higher rate of hypotension with the abdominal approach to remove pheochromocytoma. Two studies found fewer complications with laparoscopy than with open approaches. Two studies found no difference in complication rates between different open and laparoscopic approaches. Another found a similar rate of early complications for different approaches, but a higher rate of late complications with posterior approach. The remaining studies did not find any significant differences between specific techniques.
Nine series compared different laparoscopic approaches in 866 patients. None applied statistical methods to complication rates. Two series found no difference in operating time or blood loss.
We found 32 studies with a total of 1,684 patients that met our inclusion criteria and reported data on at least 10 subjects. The number of patients in these studies ranged from 12 to 156. Almost all studies were retrospective. There were wide variations in the quality and amount of reported information about the tumor size, patient characteristics, surgical approaches, and outcomes. Nine studies reported patients who had surgery back in the 1940s and 1950s. One study went back to 1929. The age of the patients ranged from newborn to 85 years.
Fifteen of the 32 studies reported perioperative mortality data. Twenty deaths were reported out of a total of 625 patients for an overall perioperative mortality rate of 4.6 percent.
There were diverse methods of reporting the long-term survival. Seventeen studies reported 5-year survival data that ranged from 19 to 62 percent with a median of 34 percent (weighted average of 35 percent). There does not appear to be any important difference in the overall survival rates between the earlier and the more recent series. Most of the studies included patients from a wide range of years thus making it difficult to appreciate if there is any trend over time.
Eight studies with individual patient follow-up data were identified with a total of 134 cases of adrenocortical carcinoma determined by biopsy or at autopsy. Three studies (Lefevre, Gerard-Marchant, Gubler, et al., 1983; Michalkiewicz, Sandrini, Bugg, et al., 1997; Ribeiro, Sandrini Neto, Schell, et al., 1990) included pediatric cases only, and three studies (Borrelli, Ingenito, Cicchi, et al., 1989; Padberg, Lauritzen, Achilles, et al., 1991; Sullivan, Boileau, and Hodges, 1978) included adults and pediatric cases. The sample from Khorram-Manesh (1998) consisted of adults only, with ages ranging 36 to 76 years. Because of the lack of complete reporting of the data and the variation in the reporting, summary analyses of the influence of age and tumor size on the prognosis after surgery was not possible.
Individual patient data from three studies (Borrelli, Ingenito, Cicchi, et al., 1989; Khorram-Manesh, Ahlman, Jansson, et al., 1998; Padberg, Lauritzen, Achilles, et al., 1991), showed adults with large and small tumors surviving 2 to 5 years as well as having survival periods of a few months.
Three studies (Borrelli, Ingenito, Cicchi, et al., 1989; Padberg, Lauritzen, Achilles, et al., 1991; Sullivan, Boileau, and Hodges, 1978) offered individual patient data on age and survival with no discernable trends in survival. For each decade of age there were one or two individual patients who had 5 year survival. Short-term survivors of 1 to 6 months were also randomly spread across the age ranges.
Not surprisingly, the prognosis for the patient tended to be better for tumors discovered at earlier stages. In Sullivan (1978), all five of the 28 cases that survived were stage I or II with a follow-up from 86 to 188 months. Khorram-Manesh (1998) reported that eight of the ten stage II cases, as well as the single stage III case were alive at the end of the follow-up period. Of the seven patients with stage IV tumors, only one survived at 136 months, and of the other six cases, only one lived beyond 1 year. Using a histologic degree of malignancy to grade tumors, Borrelli (1989) reported five tumors graded as high had 100 percent mortality. None lived to 1 year. The other five tumors classified as moderate grade and two tumors as low grade were not a predictor of mortality.
Three studies (Borrelli, Ingenito, Cicchi, et al., 1989; Padberg, Lauritzen, Achilles, et al., 1991; Sullivan, Boileau, and Hodges, 1978) reported individual data on sex, functional status of the tumor, and mortality revealed a trend of shorter survival for adult males. Of the 18 adult males in three studies, only three survived beyond 24 months and were still alive at the end of the follow-up period. Two died at 51 months each, but none survived to 5 years. In addition the males also had a higher percentage of nonfunctioning tumors at 78 percent. This compares to 31 to 38 percent for adult females. Because males tended to have nonfunctioning tumors, it may have led them to seek out treatment when the tumor had progressed to a late stage.
Fourteen studies consistently reported nonfunctioning tumors to be predominant in males (Barzon, Fallo, Sonino, et al., 1997; Boscaro, Fallo, Barzon, et al., 1995; Didolkar, Bescher, Elias, et al., 1981; Greenberg and Marks, 1978; Grondal, Cedermark, Eriksson, et al., 1990; Henley, van Heerden, Grant, et al., 1983; Icard, Chapuis, Andreassian, et al., 1992; King and Lack, 1979; Luton, Cerdas, Billaud, et al., 1990; Nader, Hickey, Sellin, et al., 1983; Nakano, 1988; Soreide, Brabrand, and Thoresen, 1992; Tritos, Cushing, Heatley, et al., 2000; Venkatesh, Hickey, Sellin, et al., 1989), though there were conflicting results on sex or functional status as a predictor of survival rates.
None of the three pediatric studies that presented individual data could be generalized because of the lack of consistency in the reported data.
For subjects who presented with incidentally discovered adrenal masses, there are four prospective studies with established follow-up protocols for untreated patients. Two studies (Barzon, Scaroni, Sonino, et al., 1999; Jockenhovel, Kuck, Hauffa, et al., 1992) are clearly prospective, one is ambiguous (Kologlu, Akyar, Baskal, et al., 1988), and one study (Siren, Tervahartiala, Sivula, et al., 2000) identified patients retrospectively from medical records but prospectively follows the patients who gave consent. The studies by Kologlu, Akyar, Baskal, et al. (1988) and Jockenhovel, Kuck, Hauffa, et al. (1992) did not report any exclusion criteria whereas the study by Barzon, Scaroni, Sonino, et al. (1999) excluded patients with hormonally active tumors, cysts, and/or malignant tumors. The exclusion criteria in Siren, Tervahartiala, Sivula, et al. (2000) included the absence of malignancy or hormonally active tumors. The total number of patients in these four studies before follow up is 132. An age range of 19 to 80 is reported by Siren, Tervahartiala, Sivula, et al. (2000) for 27 patients but only 16 were followed. Complete follow-up data for the four studies is available for 109 subjects, the age range is 21–77 years, and the range of tumor size is 0.8–5.6 cm.
Three of the four studies utilized an imaging technology, CT or MRI, in combination with biochemical tests. Jockenhovel, Kuck, Hauffa, et al. (1992) conducted CT scans with biochemical tests with a median follow-up of 32 months (11–101) on 18 subjects. Only one tumor increased in size of 0.5 cm, two previously discovered tumors disappeared, and 15 remained unchanged. Barzon, Scaroni, Sonino, et al. (1999) followed 75 subjects at six-month intervals the first year and yearly thereafter for a median of 4.6 years (2–10). CT or MRI was used in combination with biochemical tests. Fifty-eight subjects had no changes, two had decreased tumor size, and 15 experienced hormonal changes or increase in tumor size. Siren, Tervahartiala, Sivula, et al. (2000) reported follow-up data on 16 subjects one year after initial CT with a mean of 98 months follow up (53–196). All hormonal tests proved normal. MRI imaging showed no changes in seven tumors (in five patients), absence of previously discovered tumors in four patients, increase in size in five tumors (four patients) and decrease in size in three tumors (three patients). One study (Kologlu, Akyar, Baskal, et al., 1988), utilized serial CT but stated that ‘several’ tests were conducted at three or six-month intervals but there were no follow-up data or outcomes.
Overall, 121 patients were followed, and of these, 78 subjects had no changes, 20 had increases in tumor size, and 11 had decreases in tumor size or absence of previously discovered tumor on imaging scans. There were two cancer deaths unrelated to adrenal pathology. Adenoma was confirmed at autopsy.
There are five studies that focused on the management of adrenal tumors but have no structured protocol for follow-up with regards to a predetermined time interval for testing. In addition, because some patients received treatment, a subset of the subjects was followed for each study. Exclusion criteria include adrenalectomy, diagnoses made at different institutions, and the presence of cysts, myelolipoma, and hormonally active tumors. One hundred forty-two of 341 participating patients were not treated and followed for 1 month to 12 years. Follow-up tests included CT in all studies--two studies (Barry, van Heerden, Farley, et al., 1998; Reincke, Winkelmann, Jaursch-Hancke, et al., 1989) utilized CT alone, one study (Mitnick, Bosniak, Megibow, et al., 1983) used serial CT but no information was available as to how many tests were conducted. Glazer, Weyman, Sagel, et al. (1982) used a combination of CT, clinical exam and biochemical tests, while Courtade, Carnaille, Ernst, et al. (1997) used CT, MRI, clinical exam, and telephone contact. Tumor changes were reported only as increases in size in four tumors in the large study of 91 subjects followed by Barry, van Heerden, Farley, et al. (1998).
Three studies had fewer than 20 participants each, showing no change, and the study by Courtade, Carnaille, Ernst, et al. (1997) with 44 patients did not report tumor change. In fact, due to confusing definitions and reporting, the follow up data in that particular paper cannot be reliably derived. Of the aggregate population of 341 in the five studies, the age and tumor size range from 5 to 86 years and 0.6 to 10 cm, respectively.
Overall, 124 patients were followed (rejecting the uncertain data in Courtade, Carnaille, Ernst, et al. (1997)), and of these, 120 had no changes and four had increases in tumor size. There were 81 cancer deaths unrelated to adrenal pathology.
Question 1. What are the causes and prevalence of clinically inapparent adrenal masses?
The pathologies found among adrenal incidentaloma are well known and include adrenal adenoma, pheochromocytomas, metastases of primary cancers to the adrenal gland, adrenocortical carcinomas, and benign lesions such as myelolipomas and adrenal cysts. Additional pathologies include: adrenal cyst, adrenal hyperplasia, angiomyelolipoma, ganglioneuroma, hematoma, hemorrhage, lymphoma, malignant epithelial carcinoma, metastases, myelolipoma, neurinoma, regenerative hepatic nodule, renal angiomyolipoma, and retroperitoneal fibrosis.
Because incidentaloma is not a disease entity, the prevalence of incidentaloma will vary with the definition applied. The prevalence of incidentaloma in the general healthy population is likely to be very low and higher when the imaging test is performed for specific indications. One study that used transabdominal US for general health examination reported 11 adrenal masses (verified clinically or with pathology) out of 41,357 subjects (prevalence 0.027 percent). The prevalence of incidentaloma has been reported to be 0.6 percent in a study of 2,200 patients undergoing upper abdominal CT for specific indications.
Combining retrospective patient series, the prevalence of adenoma among incidentalomas was 41 percent, metastases was 19 percent, adrenocortical carcinoma was 10 percent, myelolipoma was 9 percent, pheochromocytoma was 8 percent, and other mostly benign lesions comprised the remainder of the lesions. While adenoma is the most frequent cause, the likelihood of various adrenal pathologies is depending on the definition of incidentaloma and inclusion criteria of patients in the studies.
Sixty percent of the incidentalomas occur between the ages of 41 and 60 years. The majority of the incidentalomas, 90 percent, were found in patients between 31 to 70 years of age. The predominant diagnoses were adenomas and metastases at 45 and 20 percent, respectively. Overall, approximately 64 percent of adenomas cases and 70 percent of adrenal carcinomas cases were found in females whereas 60 percent of metastases cases were reported in males.
The limited data show that for small tumors, 4 cm or less, 65 percent were adenomas and approximately 21 percent were metastases. As tumors increased in size from 4 cm or less to greater than 6 cm, the number of adrenal carcinomas increased from 2.3 to 25 percent whereas, adenomas decreased from 65 to 18 percent. In addition, the percentage of metastases decreased slightly from 21 to 18 percent.
There is insufficient data to discern any differences in the rates of various pathologies among the initial diagnostic tests used. In most of the studies, the adrenal incidentalomas were discovered with CT. There were some studies that reported incidentalomas discovered with either US or MRI or CT, but most of the studies do not separately report the data that would allow meaningful subgroup analyses across studies.
The prevalence results we derived by combining data from case series should be interpreted carefully. As discussed in our paper, “Biases of Cases Series and Potential Hazards in Making Recommendations from Case Series Reports” (Appendix C), the lack of a uniform definition of incidentaloma (and because incidentaloma is not a single pathological entity), the selective sampling of patients and reporting of information, and retrospective nature of most of the studies may result in biased estimations of the prevalence of various pathologies.
Question 2. What is the diagnostic accuracy (sensitivity, specificity) of evaluation modalities (FNA/biopsy, CT, MRI, US, biochemical tests) used to differentiate adrenal masses (adrenal carcinoma, pheochromocytoma, adenoma, adrenal hyperplasia, etc.)?
With few exceptions, the overall methodological quality of the studies we examined was poor to fair. The evaluated studies examined multiple tests, used multiple variations of test (such as enhanced and unenhanced CT), used different definitions and thresholds for test results, and included a variety of different sample populations. Study size ranged from 16 to 270. The heterogeneity of the studies limits the ability to estimate the overall diagnostic performance of each of the evaluated tests. Although some studies reported that CT, MRI, scintigraphy, FNA and PET have good to excellent test performance for differentiating benign from malignant disease, others found only moderate to poor performance (except for PET, for which there was only one small study). US had poor performance in a single study. DHEAS had high sensitivity, but low specificity in detecting adrenal cancer, though the single study was small and of poor quality. High quality studies of well-defined diagnostic tests in well-defined populations (such as those with truly incidentally discovered adrenal masses) are required.
Twelve studies reported on complications due to FNA. Only two of these, involving 360 patients, explicitly reported the risk of metastatic spread along the needle tract. Only one patient was found to have metastatic spread. In a patient with metastatic lung cancer, metastatic seeds along the needle tract were found in the liver. No cases of metastatic spread of adrenal carcinoma were reported, although it is not clear how many of the 360 patients had adrenal carcinoma. While the evidence is limited, it appears that the risk of metastatic spread of adrenal cancers and metastases by FNA is very low. Further large, longitudinal studies of patients undergoing FNA for adrenal masses are required.
Question 3. What are the surgical complication rates for various approaches used to excise adrenal masses; specifically laparoscopic, transabdominal, and retroperitoneal approaches?
There are a number of surgical series, both reporting individual experience with a given adrenalectomy technique and comparing different techniques. Despite the large number of studies, involving thousands of patients, the quality of the evidence is poor. Randomized, controlled trials are lacking. Non-randomized series contain significant selection bias, as more difficult cases, larger tumors, and invasive cancers are routinely assigned to the control group as a result of surgeon preference. Nevertheless, the evidence consistently points in the same direction, at least for small, non-malignant tumors. The posterior approach appears to offer an advantage over the anterior approach in terms of surgical morbidity, as measured by post-operating hospital stay, and perhaps in terms of operating time and blood loss as well. Similarly, both RLA and TLA result in shorter hospital stays than PA or AA, and while PA is quicker, TLA and RLA result in less blood loss and perhaps fewer major complications. Although randomized, controlled trials would offer the best measure of the safety of laparoscopy versus open surgery, given the ostensible benefit seen in the non-randomized trials, and the current prevailing thought among surgeons, it is unlikely that such trials will ever be conducted.
When performing laparoscopic adrenalectomy, the lateral transperitoneal approach may be quicker and cause less blood loss than either RLA or anterior TLA, but in terms of hospital stay and complication rates, one approach does not appear to be superior to the others. In this area, more randomized, controlled trials are necessary.
Finally, TNA offers potentially the least morbid procedure, with the least blood loss, the shortest hospital stay, and a low complication rate. However, given that only 15 of these procedures have actually been reported, it is premature to assign needlescopic surgery a role in adrenalectomy. More trials are needed.
For pheochromocytomas, invasive carcinomas, and very large tumors, the best approach is still a matter of debate. There are few series looking at these indications alone, and many authors consider them contraindications to laparoscopy. However, others have challenged these limitations, operating on pheochromocytomas, large tumors, and potential carcinomas, though the latter are usually converted to open procedures once definitively identified. In these areas in particular, where there remains debate, randomized controlled trials are most needed and most appropriate.
Question 4. What are the patient outcomes after surgical excision of adrenocortical carcinoma (morbidity and mortality)?
There were 32 studies with a total of 1,684 patients that met the inclusion criteria. Fifteen of the 32 studies reported perioperative mortality data with an overall perioperative mortality rate of 4.6 percent. There were diverse methods of reporting the long-term survival. Seventeen studies reported 5-year survival data that ranged from 19 to 62 percent with a median of 34 percent (weighted average = 35 percent). There does not appear to be any important difference in the overall survival rates between the earlier and the more recent series. Most of the studies included patients over a wide range of years, making it difficult to discern any trend over time.
Question 5. What evidence is there to support the use of periodic biochemical and imaging studies to follow untreated adrenal masses?
Although many have proposed follow-up strategies for untreated adrenal incidentalomas, there are only a few studies that have evaluated these proposals. Only four studies used a pre-specified follow-up protocol in following up the patients. The studies were small and it is not surprising that few events were identified. If the goal of following small adrenal incidentalomas is the early detection of adrenal carcinomas and pheochromocytomas, prospective clinical trials designed to evaluate follow-up management strategies may not be practical. Given the rarity of these pathologies, large numbers of patients and a long duration of follow-up will be required.
Adrenal incidentaloma is neither a single pathological entity nor a disease. A strict definition of incidentaloma would aid in the interpretation of results from clinical studies; however, it will not be sufficient to address the diverse manifestations of this condition that are also clinically relevant. Future studies of incidentaloma need to broadly cover all possible scenarios but individual studies should apply rigorous inclusion criteria for each of these scenarios or provide careful analyses of well-defined subgroups.
Many studies assumed that the major reason for the further evaluation of adrenal incidentalomas is for the purpose of detecting adrenal carcinoma. Given the rarity of this tumor and the lack of effective therapy in the later stages, the overall benefit of detection is small. On the other hand, subclinical, biochemically active adrenal adenomas are common and need to be better understood regarding their prevalence and long-term clinical outcomes. Given their prevalence and the significance of hypertension and diabetes, these tumors may become the reason for aggressive intervention in adrenal incidentaloma.
Most of the studies we identified for this topic were retrospective. Like many retrospective studies, they suffered from inaccurate, incomplete, or inadequate definitions and descriptions of the population, reference standard for diagnosis, and follow-up. Future studies should provide more precise definitions of the conditions and standards of measurements, along with complete reporting of essential data.
With few exceptions, the overall methodological quality of the studies we examined was poor to fair. Higher quality studies are needed to properly assess the clinical usefulness of various diagnostic tests for adrenal incidentalomas. In contrast to published studies, future studies should make a rigorous attempt to clearly define and report eligibility criteria, sample characteristics, test methodology, and the definitions of positive and negative tests. Reference standards should be clearly delineated, completely independent of the tests being investigated, and as much as possible, include well defined outcomes, such as surgical or autopsy diagnoses.
Given the heterogeneity of definitions for incidentaloma, it is important to define specific populations of patients. Many studies combined subjects whom had true incidentalomas, which are rarely malignant, with subjects who had adrenal masses found during cancer work-ups and who frequently have metastatic disease. Other studies failed altogether to report how the adrenal masses were discovered. Without a clearly defined population, studies of diagnostic test performance are difficult to interpret and have limited applicability. In addition, ROC curves should be drawn, and complete data reported whenever possible, so that the reader can determine the proper threshold to use for a given test.
More studies of FNA are required to determine the risks of both needle-tract metastases and major complications. Ideally, future studies should be large, with thorough data collection. Authors should report the following minimum components: patient demographics (e.g., age, body mass index, gender, reason for initial discovery of mass, mass size and location), all relevant aspects of biopsy technique (e.g., position, approach, organs transected, needle size and type, number of passes, radiological tools), biopsy results and final diagnoses, short- and long-term complications, and patient disposition. When appropriate, authors should perform statistical analyses to identify patient characteristics and biopsy techniques associated with an increased risk of complications.
Future research should concentrate on defining the best procedure for each indication, as well as identifying other factors, such as number of operations performed, which may predict surgical complications. In order to compare across series, it would be useful for authors to agree on standard definitions of such seemingly simple outcomes as operating time or post-operative length of stay. It would also be particularly useful to standardize the measurement and reporting of complications.
With regard to complications, almost all studies to date on adrenal surgery complication rates have been retrospective case series. Many contain unrepresentative samples of patients, and data reporting was typically incomplete. Statistical analyses of complication rates are uncommon in the literature. To properly assess the risks and benefits of different surgical approaches for adrenal mass excision, high quality studies are needed in the future. Ideally, these should be comparative, prospective studies of clearly defined populations; if possible, randomized controlled trials would provide the best evidence. Although conducting randomized surgical trials is difficult, at least one small study suggests that it is possible (Fernandez-Cruz, Saenz, Benarroch, et al., 1996). Otherwise, well-matched, prospective, comparative studies would have considerable value. All subjects should be categorized by well-defined risk factors, including co-morbid conditions. Studies should define a priori major and minor complications of interest and should follow subjects for a sufficient time post-operatively to assess long-term complications. Statistical analyses should be performed, when possible, with multivariate analysis. Given the generally low risk of major complications, the value of multivariate analysis, and the inherent variability of surgical techniques among surgeons and facilities, meaningful results will require large, multicenter trials.
There is very sparse data to guide the management of untreated, incidentally discovered adrenal masses. Most of the available studies are either too small to provide meaningful results or suffer from methodological problems. Results from retrospective studies or studies without specific follow-up protocols are difficult to interpret. Variable definitions of incidentaloma further complicate the problem.
Future prospective studies should apply pre-specified protocols to clearly defined populations. Given the rarity of adrenocortical carcinoma and pheochromocytoma, assessing the usefulness of follow-up protocols will require a large number of subjects. If recent studies are confirmed, follow-up strategies that include biochemical evaluations for subclinical biochemically active adrenal masses will be useful.
Well-designed clinical trials will provide the most reliable evidence regarding the management of these patients, but these trials will take many years and may not always be feasible. An alternative approach would be to create an international registry of patients with well-documented adrenal incidentaloma. This registry should be established using standardized definitions and inclusion criteria. Investigators could then analyze individual patient data to understand the influence of specific factors such as tumor size, age, and inclusion criteria on the development of subsequent adrenal pathologies.
#1 Adrenaloma*
#2 Adrenal gland neoplasms OR adrenal glands OR adrenal gland diseases OR Adrenalectomy OR carcinoma, adrenal cortical OR Adrenomegaly [tw]
#3 (Adrenal OR Adrenocortical)
#4 (incidental* OR silent OR subclinical OR pre-clinical OR non-hormonal OR nonhormonal OR nonsecretory OR nonfunction* OR asymptomatic OR serendipitous* OR non-hypersecretory OR nonhypersecretory OR occasionally OR pre-cushing OR precushing OR inapparent OR unapparent OR nonapparent)
#5 (#2 AND #4) OR (#3 AND #4)
#6 #1 OR #5
s1 (adrenal OR adrenocortical) AND (adenoma OR pheochromocytoma OR mass OR Masses OR lesion?? OR tumor?? OR tumour?? OR cyst?? OR nodule?? OR nodular OR enlargement or hyperplasia)
s2 incidental? OR silent OR subclinical OR pre()clinical OR non()hormonal OR nonhormonal OR nonsecretory OR nonfunction? OR asymptomatic OR serendipitous?
s3 non()hypersecretory OR nonhypersecretory OR occasionally OR pre()cushing OR pseudo()cushing OR inapparent OR unapparent OR nonapparent)
s4 s1 AND (S2 OR S3)
s5 s4/CONF OR s4/MEETING
s1 (adrenal OR adrenocortical) AND (adenoma OR pheochromocytoma OR mass OR Masses OR lesion?? OR tumor?? OR tumour?? OR cyst?? OR nodule?? OR nodular OR enlargement or hyperplasia)
s2 incidental? OR silent OR subclinical OR pre()clinical OR non()hormonal OR nonhormonal OR nonsecretory OR nonfunction? OR asymptomatic OR serendipitous?
s3 non()hypersecretory OR nonhypersecretory OR occasionally OR pre()cushing OR pseudo()cushing OR inapparent OR unapparent OR nonapparent)
s4 s1 AND (S2 OR S3)
s5 (adrenal or adrenocortical) (5W) (carcinoma? OR neoplasm??)
s6 adrenalectomy
s7 s4 AND (s5 OR s6)
s8 s7/human
adrenaloma$.af.
exp adrenal gland neoplasms/ or exp adrenal glands/ or exp adrenal gland diseases/ or exp adrenalectomy/ or exp carcinoma, adrenal cortical/ or adrenomegaly.tw. or (adrenal gland neoplasms or adrenal glands or adrenal gland diseases or adrenalectomy or carcinoma, adrenal cortical).af.
(adrenal or adrenocortical).af.
(incidental$ or silent or subclinical or preclinical or pre-clinical or non-hormonal or nonhormonal or nonsecretory or nonfunction$ or asymptomatic or serendipitous$ or non-hypersecretory or nonhypersecretory or occasionally or pre-cushing or precushing or inapparent or unapparent or nonapparent or unsuspected).af.
(2 and 4) or (3 and 4)
1 or 5
exp pheochromocytoma/
exp adenoma/
pheochromocytoma.tw.
(adrenal adj6 adenoma).tw.
7 or 8 or 9 or 10
11 and 4
follow-up studies/
follow-up.tw.
exp Case-Control Studies/
case-control.tw.
exp Longitudinal Studies/
longitudinal.tw.
exp Cohort Studies/
cohort.tw.
(random$ or rct).tw.
exp Randomized Controlled Trials/
exp random allocation/
exp Double-Blind Method/
exp Single-Blind Method/
randomized controlled trial.pt.
clinical trial.pt.
(clin$ adj trial$).tw.
((singl$ or doubl$ or trebl$ or tripl$) adj (blind$ or mask$)).tw.
exp PLACEBOS/
placebo$.tw.
exp Research Design/
exp Evaluation Studies/
exp Prospective Studies/
exp Comparative Study/
13 or 14 or 15 or 16 or 17 or 18 or 19 or 20 or 21 or 22 or 23 or 24 or 25 or 26 or 27 or 28 or 29 or 30 or 31 or 32 or 33 or 34 or 35
exp Adrenal Gland Neoplasms/co, su [Complications, Surgery]
exp Adrenal Gland Diseases/co, su [Complications, Surgery]
exp Carcinoma, Adrenal Cortical/co, su [Complications, Surgery]
exp Pheochromocytoma/co, su [Complications, Surgery]
37 or 38 or 39 or 40
36 and 41
exp Postoperative Complications/
(post operative or postoperative).tw.
((surg$ or operat$) adj7 (outcome$ or complication$)).tw.
(laparoscopic or transabdominal or retroperitoneal).tw.
43 or 44 or 45 or 46
2 or 11
47 and 48 and 36
(FNA or needle).tw.
48 and 50
12 or 42 or 49 or 51
limit 52 to human
limit 53 to english language
limit 54 to (addresses or bibliography or biography or comment or dictionary or directory or editorial or festschrift or interview or lectures or legal cases or letter or news or periodical index)
case report.sh.
54 not (55 or 56)
57 not 6
limit 58 to (guideline or meta analysis or practice guideline or review or review literature or review, academic or review, multicase or review, tutorial)
58 not 59
The Evidence-based Practice Center staff acknowledges the collaboration as technical experts by the members of the Planning Committee of the National Institute of Health State-of-the-Science Conference on the Management of the Clinically Inapparent Adrenal Mass (“Incidentaloma”)
Dr. Robert C. McIntyre, Jr., MD, FACS
Associate Professor of Surgery
University of Colorado Health Sciences Center
Denver, Colorado
Quan-Yang Duh, MD, FACS
Professor of Surgery
University of San Francisco Medical Center and SF VA Medical Center
San Francisco, California
John Cronan, MD
Department of Diagnostic Imaging
Rhode Island Hospital
Providence, Rhode Island
Nick Fitterman, MD, FACP
North Shore Medical Group
Huntington, NewYork
Elise M Brett, MD, FACE
1192 Park Avenue
New York, New York
Joseph Lau, MD; EPC/Project Director
Ethan Balk, MD, MPH; Assistant Project Director
Michael Rothberg, MD
John P.A., Ioannidis, MD
Deirdre DeVine, M Litt; Project Manager
Priscilla Chew, MPH; Research Associate
Bruce Kupelnick, BA; Research Associate
Kimberly Miller, BA, Research Assistant
Case reports and case series have been the earliest method of accumulating evidence in medicine. The Hippocratic writings abound in descriptions of specific classical disease conditions. Most diseases in the past were defined on the basis of case reports or case series that brought new entities into focus. With the advent of controlled studies, the prominence of case series in medicine and in the hierarchy of evidence has diminished (Harbour and Miller, 2001; U.S. Preventive Services Task Force, 1996), but they continue to enjoy substantial popularity among clinicians (Jenicek, 2001). Clinicians may often feel intimidated by the “inroad of statistics” into their disciplines, as this is reflected in sophisticated controlled clinical research and the new methods of meta-analysis, decision-analysis and cost-effectiveness analysis that comprise the foundation of evidence-based medicine. However, clinicians still appreciate the interesting case report and find it easy to understand the findings of a case series. Moreover, the majority of studies published in several major clinical specialty peer-reviewed journals still pertain to uncontrolled data. In this section, we discuss some of the hazards that exist in the use of evidence from case series for making clinical decisions and generating recommendations for medical practice.
Many conditions may be rare enough to make it impractical to perform controlled research. Controlled trials are typically performed to test hypotheses, and power calculations are conducted to estimate the required sample sizes. For rare diseases, controlled trials and power calculations are often thought to be exercises in futility. Nevertheless, an underpowered controlled trial may still provide useful evidence to estimate the magnitude of a therapeutic effect. Such information would be less prone to bias compared with the evidence that may be obtained with the same subjects with an uncontrolled case series design. Bayesian inference may be used to interpret evidence from small trials. Unfortunately, most trial designs are still dependent on frequentist considerations. Finally, single case reports or small case series may be the only way to discover and study new or uncommon diseases, such as genetic syndromes (Simpson and Griggs, 1985).
In some situations, controlled designs may be difficult to implement because of logistic considerations. For example, controlled designs and in particular randomized trials are relatively rare in many surgical disciplines and therefore the case reports and case series have remained very popular in the surgical literature. The technical aspects of a new surgical procedure may present difficulties in creating an appropriate control arm. Parameters such as the experience of the surgeon and the point in the learning curve of each surgeon may require particular ingenuity to incorporate in randomized designs. Therefore, case series are very common for many surgical subjects. In medical domains, the adoption of controlled designs varies considerably. In some areas such as cardiology and oncology there is relatively wide appreciation of the merits of randomized controlled trials. In other specialties such as rheumatology, uncontrolled designs are still common research practice.
Health care practices sometimes become established on the basis of data accumulated before the availability of evidence from randomized trials. For example, the effectiveness of Pap tests in reducing cervical cancer mortality has never been tested with the rigorousness of randomized trials. However, it is difficult to question its use at the moment, despite the fact that the diagnostic characteristics (sensitivity and specificity) of this test are far from optimal (Fahey, Irwig, and Macaskill, 1995). In general, for interventions that once seemed to be extremely effective in uncontrolled research in the past, one may now perceive an ethical barrier to testing them under controlled conditions. While many interventions based on evidence from non-randomized data may truly be efficacious (Ioannidis, Haidich, Pappa, et al., 2001), unfortunately we cannot tell whether some of them may lead to practices that later will prove to be erroneous when randomized evidence become available (Antman, Lau, Kupelnick, et al., 1992).
For some research questions, such as estimation of incidence and prevalence or descriptive research, non-controlled case series may be the only legitimate mode for gathering information. Conducting clinical trials with rare events is often unfeasible. High costs may also make conducting a trial impractical. Actually, controlled studies may offer less reliable estimates of frequency and other descriptive parameters since they are usually focused on selected populations that have already satisfied the eligibility criteria of the controlled study.
One should caution that case series are sometimes used to make inferences about the prevalence or incidence of a disease. While, in some cases, this may be the only approach possible, the denominator (total population) is typically not well defined. For example, one cannot be certain that all cases in a certain “catchment” area during a specific period of time have been included in a given case series. Thus, whenever possible, case series should not replace more appropriate prevalence studies.
Most case series are based on retrospective examination of information, but sometimes information may be collected prospectively. Retrospective information is easier to collect than prospective information. Typical sources include medical records, electronic hospital and physician archives, and large-scale databases from insurance or other organizations (Kelsey, Whitemore, Evans, et al., 1996). Recording of information in retrospective case series may be more selective and subject to larger error than information that is being collected prospectively with a specific plan in mind. Retrospective data are also likely to be more heterogeneous both in quality and quantity. Since there is no a priori research plan when the information is being collected, the type and quality of data that are captured for each patient may vary enormously even within the same database. In general, retrospective data that pertain to parameters that are not routinely captured and unequivocally defined is likely to be incomplete. These biases do not exist with prospective collection of information where data may be captured according to a predefined plan and standardized definitions.
Measurement error is likely to be larger in uncontrolled designs than in controlled studies. The problem may be larger when it pertains to retrospective case series when there was no common a priori rule on how the pertinent measurements should be made and recorded. Repeatability conveys the concordance of successive measurements of the same parameter performed by the same evaluator (or laboratory) on different days or at different runs of the assay. Reproducibility refers to the concordance of successive measurements of the same parameter performed by different evaluators or different laboratories. The repeatability and reproducibility decrease when measurements are performed by different individuals or different methods, with different standards and different interpretation and recording procedures. Often, but not always, in controlled prospective research, there are assurances that both the repeatability and reproducibility of the measured parameters would be appropriately high. Quality control programs may exist in multicenter trials, for example, that assures that measurements of the parameters of interest deviate within only a reasonable extent between different centers, laboratories and evaluators. In case series, usually there are no such assurances.
A special consequence of measurement error is the phenomenon of regression-to-the-mean (Davis, 1976). Regression-to-the-mean occurs when the following conditions are satisfied: (a) subjects are selected on the basis of having extreme values (higher than a certain cutoff or lower than a certain cutoff) in a given parameter of interest; (b) only one measurement of this parameter is done at baseline; and (c) the change in this parameter is used to evaluate the outcome of a subject during follow-up. Regression-to-the-mean is one common explanation why case series report spurious therapeutic efficacy for ineffective interventions. Subjects with extreme characteristics are often the target of studying interventions. For example, it is common to select high-risk patients for a new intervention. Alternatively, one may select only low-risk patients if the intervention is conceived to be potentially toxic or its toxicity is unknown. Even in the absence of any true therapeutic effect, a group selected on the basis of an extreme value in a given parameter is likely to show less extreme values on average upon re-measurement. For example, let us assume that patients with rheumatoid arthritis are selected on the basis of having an elevated sedimentation rate >60 mm per hour. Upon re-measurement, the mean ESR of the group will be lower than the baseline measurement, even if there is no true change in the disease activity. The regression-to-the mean can be eliminated, if a separate measurement is performed after the screening measurement and then the second measurement is used as the baseline. Unfortunately, this simple correction is rarely performed in most case series even in studies published in prestigious journals.
Besides regression-to-the-mean, one needs to consider also the natural course of the disease both as short term and long-term assessment. Uncontrolled studies do not allow us to evaluate what would have happened to subjects if they had not received the intervention(s) that they actually received. For some diseases with continuously progressive disease, simple stabilization of a lesion or the patient's overall condition may be a success. For other diseases, where the disease severity may naturally improve over time, stabilization of the patient's condition may actually signal a detrimental, toxic intervention that should be avoided. For the same reasons, in treatment of malignant or pre-malignant lesions, uncontrolled studies may give us biased estimates about the effectiveness of specific interventions in delaying their development and consequences.
Subjects who are included in case series may suffer from various manifestations of selection biases. Selection bias occurs when the sample of a case series is not representative of the general population of the condition that it represents. In some situations, even self-selection may occur; for example, patients with some characteristics suggestive of worst prognosis may be selectively lost-to-follow-up, and the remaining group of subjects may subsequently and falsely give a picture of a group with more favorable prognosis than the truth would be. Another form of selection bias is diagnosis bias, where the attribution of a diagnosis may depend on the availability of additional information. For example, a lesion may be more likely to be screened for malignant features by a radiologist, if the radiologist knows that the patient has clinical signs that are suggestive of malignancy. More exhaustive imaging studies may be performed in such cases. Moreover, the radiologist may also be tempted to make a diagnosis of malignant lesion, if such clinical information is available.
Confounding is very difficult to handle in case series. The definition of a confounder (Rothman and Greenland, 1998) relies on the following criteria: (a) the confounder is a risk factor for the disease among those that do not have the studied parameter; (b) the confounder is associated with the studied parameter in the source population from which the subjects arise; and (c) the confounder is not affected by the disease or the studied parameter. In most case series research, information on confounders is not collected or may be collected erratically and selectively. Thus it is very difficult to evaluate how confounding impacts the results.
Case series that aim to differentiate between two or more diagnoses based on various parameters of interest may be conceived as applications of diagnostic test research. Typical biases that occur in this setting (Mulrow, Linn, Gaul, et al., 1989) include, among others: (1) lack of accurate definitions of the diagnostic categories; (2) verification bias (different diagnostic work-up depending on the features and categorization of the lesion); (3) improper handling of gray measurements and uninterpretable results; (4) lack of accurate and reproducible definitions for the putative discriminating parameters; and (5) information bias (misclassification) that may be non-differential or differential. The issues are similar to the biases that exist for other types of case series.
Many case series may report selected data, rather than all the data that has been accumulated from the investigators. In retrospective research especially, the eligibility criteria of what is to be included in a study report may be decided after the database has been collected. A special concern is publication bias, i.e. the tendency to publish more easily results that show statistically significant findings, while at the same time leaving small studies with non-significant results unpublished. Such a bias may influence the strength of associations observed in isolated case series as well as in their overall synthesis. For example, it is possible that such biases may tend to create a picture that some parameters are stronger predictors of an outcome or stronger discriminating factors for differentiating between two types of lesions than the truth would be. Publication bias is stronger in uncontrolled research (Easterbrook, Berlin, Gopalan, et al., 1991).
The synthesis of data from uncontrolled research is further confounded by the fact that it is unlikely that the different research reports use the same definitions for the various parameters of interest or for the study outcomes. Some heterogeneity is unavoidable and to some extent it is even useful in assessing the generalizability of the conclusions. This is especially true for differences in eligibility and selection criteria. However, significant differences in definitions may be problematic. Reports of case series may give suboptimal information and it may not be feasible to merge all data adjusting them to common definitions.
Lack of standardized definitions is likely to be prominent even for putative confounders. Information on confounders may have been collected with different means in different studies, or according to different definitions. In many cases, information on important confounders may have been collected only in some studies, but not in others. Synthesizing such disparate data may be very problematic.
The quality of the overall data is likely to be different in various studies. Most importantly, the quality of the data may be difficult or even impossible to discern simply from the written report of an uncontrolled study. Decisions on data synthesis for case series should always employ a priori the use of some rules that define the minimal prerequisites of quality that need to be met in order to include a case series into the final quantitative synthesis. While it is important to be able to integrate knowledge obtained from case reports and case series into the larger frame of evidence-based medicine (Vandenbroucke, 1999; Vandenbroucke, 2001), quality issues should be taken seriously.
Free Full text in PMC]| AA | anterior approach open adrenalectomy |
| ACC | adrenocortical carcinoma |
| ACTH | adrenocorticotropic hormone |
| AHRQ | Agency for Healthcare Research and Quality |
| ASA | American Society of Anesthesiologists |
| AUC | area under the ROC curve |
| BL | blood loss |
| BMI | body mass index |
| CA | cance |
| CAD | coronary heart disease |
| CHF | congestive heart failure |
| CRH | corticotropin-releasing hormone |
| CT | computed tomography |
| ctr | center |
| CVA | cerebrovascular accident |
| d | day |
| Dept | department |
| DEX | dexamethasone |
| DHEAS | dehydroepiandrosterone sulfate |
| DM | diabetes mellitus |
| DVT | deep vein thrombosis |
| dx | diagnosis |
| EPC | Evidence-based Practice Center |
| FN | false negative |
| FNA | fine needle aspiration |
| FNAB | fine needle aspiration biopsy |
| FP | false positive |
| FU | follow up |
| FUO | fever of unknown origin |
| HTN | hypertension |
| HU | Hounsfield units |
| IHD | ischemic heart disease |
| IV | intravenous |
| LFT | lost to follow up |
| LOS | length of (hospital) stay |
| LTLA | lateral transperitoneal laparoscopic adrenalectomy |
| MEN | multiple endocrine neoplasia |
| Mets | metastatic or metastases |
| MIBG | 123I-meta-iodobenzylguanidine |
| min | minute |
| mo | month |
| MRI | magnetic resonance imaging |
| N or n | number |
| NA | not available |
| ND | not documented |
| NP-59 | 131I-6β-iodomethyl-19-norcholesterol |
| ns | not significant |
| NSCLC | non-small cell lung cancer |
| OMAR | Office of Medical Applications of Research |
| OT | operative time |
| PA | posterior approach open adrenalectomy |
| Path | pathology |
| PE | pulmonary embolis |
| PET | positron emission tomography |
| Pheo | pheochromocytoma |
| PO | post-operative |
| PRA | plasma renin activity |
| prev | prevalence |
| radx | radiation therapy |
| RLA | retroperitoneal laparoscopic adrenalectomy |
| ROC | receiver operating characteristics (curve) |
| SCLC | small cell lung cancer |
| susp | suspected |
| SUV | standardized uptake value (in scintigraphy) |
| TB | tuberculosis |
| TLA | transperitoneal laparoscopic adrenalectomy |
| TN | true negative |
| TP | true positive |
| UFC | urinary free cortisone |
| US | ultrasonography |
| UTI | urinary tract infection |
| VMA | vanillylmandelic acid |
| Xr | X-Ray |
| Y | years |
Free Full text in PMC]
[PubMed]Free Full text in PMC]Free Full text in PMC]Free Full text in PMC]
[PubMed]Free Full text in PMC]Free Full text in PMC]Free Full text in PMC]Free Full text in PMC]
[PubMed]Free Full text in PMC]
[PubMed]Free Full text in PMC]Free Full text in PMC]
[PubMed]Free Full text in PMC]Search strategies #1–3 courtesy of the National Library of Medicine reference librarian, Mary Conway
