• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of annsurgLink to Publisher's site
Ann Surg. Oct 2000; 232(4): 549–556.
PMCID: PMC1421187

Gastrointestinal Carcinoids: Characterization by Site of Origin and Hormone Production

Abstract

Objective

To describe a large series of patients with carcinoid tumors in terms of presenting symptoms, hormonal data, stage at diagnosis, pathologic features, and survival.

Summary Background Data

Published series have described significant prognostic features of carcinoid tumors as site of origin, age, sex, stage at diagnosis, presence of high hormone levels, and increased T stage. Of these, stage at diagnosis and T stage seem to emerge most often as independent predictors of survival in multivariate analyses. Of carcinoid tumors, those arising from a midgut location have higher levels of serotonin and serotonin breakdown products, as well as more frequent metastatic disease at presentation, than those arising from either foregut or hindgut locations.

Methods

A prospective database of carcinoid patients seen at Duke University Medical Center was kept from 1970 to the present. Retrospective medical record review was performed on this database to record presenting symptoms, hormonal data, pathologic features, and survival. Statistical methods included analysis of variance, Kaplan-Meier analysis, and Mantel-Cox proportional hazard survival analysis, with P < .05 considered significant for all tests.

Results

Carcinoids arising in different locations had different presentations: rectal carcinoids presented significantly more often with gastrointestinal bleeding, and midgut carcinoids presented significantly more often with flushing, diarrhea, and the carcinoid syndrome. Patients with midgut tumors had significantly higher levels of serotonin and serotonin breakdown products, corresponding to higher metastatic tumor burdens. Although age, stage, region of origin, and urinary level of 5-hydroxyindoleacetic acid predicted survival by univariate analysis, only the latter three were independent predictors of survival by multivariate analysis. Of the patients with metastatic disease at diagnosis, those with midgut tumors had better survival than those with foregut or hindgut tumors.

Conclusions

Although region of origin is certainly an important factor in determination of prognosis, stage of disease at presentation is more predictive of survival. Pancreatic and midgut carcinoids are metastatic at diagnosis more often than those arising in other locations, leading to a worse overall prognosis. Among patients with distant metastases, patients with midgut primary tumors have improved survival despite increased hormone production compared with patients with tumors arising in other primary sites.

In 1907, Obendorfer coined the term “karzinoide” to describe a set of ileal tumors that behaved in a more benign manner than carcinomas. 1 Since then, carcinoid tumors have been found to be relatively uncommon neuroendocrine tumors arising from amine precursor uptake and decarboxylation cells. These tumors are usually classified according to their site of origin as foregut, midgut, or hindgut carcinoids. 2,3

One of the unique features of carcinoid tumors is that they produce a variety of protein and peptide products, the most characteristic of which is serotonin. Hormonal measurements reveal that foregut and midgut carcinoids produce the highest levels of serotonin. 4,5 Systemic serotonin release is purported to cause most of the symptoms of the carcinoid syndrome, including diarrhea, flushing, wheezing, and right-sided heart disease. However, many have reported variance within these broad embryologic subgroups as to clinical behavior. 6,7 The goal of this study was to define the relations between presenting variables and prognosis among carcinoids arising in different gastrointestinal (GI) sites, with particular attention to stage and hormonal production.

METHODS

Since 1970, one author (J.M.F.) has prospectively maintained a database of patients with carcinoid tumors at Duke University Medical Center and the Durham Veterans Affairs hospital. Carcinoid tumors were defined as tumors with neuroendocrine histologic features and evidence of serotonin production. All patients had 24-hour urine samples and blood samples collected at the time of initial evaluation. The urine samples were stored at 4°C and assayed for 5-hydroxyindoleacetic acid (5-HIAA) and serotonin using spectrophotometric and radioenzymatic methods, respectively. Serotonin levels were also measured in both serum and platelets. High-pressure liquid chromatography was used to measure serotonin levels after 1988. In addition to hormonal data, information on site of origin and patient demographics was collected.

The 336 patients with carcinoids of GI origin who were referred to us for consultation underwent retrospective medical record review to determine the presenting symptoms, pathologic features, and survival. Our experience with foregut carcinoids has been reported previously 8 but has been updated and included in this manuscript for reference. The previous foregut manuscript focused on the differences in clinical behavior and biochemical production of gastric, pancreatic, and duodenal carcinoids. The current manuscript examines how foregut, midgut, and hindgut tumors differ. No patients with appendiceal carcinoids were referred to our tertiary referral center, and none were included in the database. Stage at diagnosis was also recorded, broken down into local, regional, and distant disease.

One-way analysis of variance was used to compare groups, with the Newman-Keuls method applied for pairwise comparison. Kaplan-Meier analysis was used to estimate survival from diagnosis to death. The Mantel-Cox log-rank test was used to compare survival curves between groups. P < .05 was considered significant for all tests. The computer program Statistica for Windows version 5 (Statsoft, Tulsa, OK) was used for statistical calculations and survival curve generation.

RESULTS

The demographics and presenting symptoms of carcinoids arising at different GI sites are presented in Table 1. The most common site of disease was the ileum. The most common presenting symptoms were abdominal pain, frequently accompanied by obstructive symptoms, diarrhea, flushing, and GI bleeding. In 28 patients (8%), carcinoid disease was found incidentally at laparotomy for other purposes or at endoscopy. Thirteen percent of the patients had carcinoid syndrome, mostly with ileal sites of origin (77% of patients with carcinoid syndrome).

Table thumbnail
Table 1. DEMOGRAPHICS AND PRESENTING SYMPTOMS

When analyzing the presenting data by analysis of variance, several significant associations become clear. Rectal carcinoids manifested less often with pain or diarrhea than did foregut or midgut carcinoids. In contrast, these hindgut carcinoids manifested more often with bleeding symptoms or without specific symptoms. Midgut tumors (jejunal, ileal, and cecal) manifested with flushing and frank carcinoid syndrome more often than the other subtypes.

The extent of disease at presentation is presented in Table 2. Most patients had metastatic disease at presentation. Broken down by site of origin, duodenal and rectal carcinoid tumors were most likely to be localized at presentation; pancreatic tumors were universally metastatic at diagnosis. The wide spectrum of survival based on site of origin is also demonstrated in the survival curves shown in Figure 1. Hindgut and foregut tumors other than pancreatic carcinoid had the most favorable prognoses (46% 10-year survival for rectum, 62% 10-year survival for duodenum, 56% 10-year survival for stomach). Pancreatic primaries had the poorest prognosis, with 10% alive at 10 years. Midgut tumors had an intermediate prognosis, with 10-year survival rates in the range of 40% to 45%. Although these numbers were close to those of foregut tumors at 10 years, midgut survival continued to decline after this point.

figure 10FF1
Figure 1. Survival by site (P < .01).
Table thumbnail
Table 2. EXTENT OF DISEASE AT DIAGNOSIS

To account for the fact that more midgut and pancreatic tumors were metastatic at presentation, we analyzed the survival data after classifying patients based on the extent of disease at presentation (Fig. 2). As Figure 2A reveals, patients with localized disease had statistically similar survival curves, irrespective of the site of origin. Analysis of survival of patients with regional or distant metastases at presentation (see Fig. 2B) revealed that patients with tumors of midgut origin did better than those with tumors of foregut or hindgut origin. However, these data may be biased in that wide mesenteric resection of midgut tumors provided more lymph nodes for analysis. Thus, the same analysis, excluding the patients with regional metastases, was performed to evaluate survival in patients with distant metastatic disease. As Figure 2C shows, the better survival of patients with metastatic midgut tumors persisted. Given that patients with the carcinoid syndrome are supposed to have a dismal prognosis, we wondered whether the midgut curve would portray even better survival if we excluded patients with the carcinoid syndrome. Figure 2D indicates that the curve did not change, demonstrating that the presence of the carcinoid syndrome does not portend a worse prognosis for metastatic midgut tumors. Looked at another way (see Fig. 2E), patients with metastatic carcinoid of midgut origin had the same prognosis with or without the carcinoid syndrome.

figure 10FF2
Figure 2. (A) Survival by region of patients with local disease at diagnosis (P = .92). (B) Survival by region of patients with metastatic disease (regional lymph node/distant metastasis) at diagnosis (P < .01). (C) Survival by region ...

Other variables from the database were examined and found to correlate with survival. Table 3 shows the serotonin measurements of patients at presentation with a carcinoid tumor. The midgut tumors had the highest levels of urinary 5-HIAA and serum/platelet serotonin. Patients with gastric and pancreatic carcinoid tumors had significantly higher levels of urinary serotonin. Patients with duodenal and rectal carcinoid tumors had low serotonin levels across the tests. High levels of serotonin breakdown products correlated strongly with both liver metastases and death.

Table thumbnail
Table 3. HORMONAL MEASUREMENTS AT DIAGNOSIS

Size has also been traditionally correlated with survival in patients with carcinoid tumors. As Table 4 shows, we found that the size of the primary tumor could be used to predict the extent of disease and prognosis. Of the 183 patients for whom we had accurate size data, only 2 patients of 21 with tumors less than 1 cm had evidence of metastatic disease at diagnosis, with no others developing distant metastases during the follow-up period. Of the 75 patients with 1- to 2-cm tumors, 51 had metastatic disease at diagnosis. Of the remaining 24 patients, metastatic disease developed in 3 during the follow-up period. Of the 87 patients with tumors larger than 2 cm, 81 had distant metastatic disease at diagnosis. Of the six thought to have local disease at diagnosis, all six were ileal in origin. Distant metastases developed in three of these patients.

Table thumbnail
Table 4. SIZE/STAGE RELATIONSHIP

Although size correlates strongly with the presence of metastases, it does not correlate with survival. As Table 5 shows, on univariate analysis, age (P = .04), stage (P < .01), region of origin (P < .01), and 5-HIAA (P < .01) were significantly associated with survival. On multivariate analysis, only the latter three were independent predictors of survival.

Table thumbnail
Table 5. SURVIVAL ANALYSIS RESULTS

DISCUSSION

One of the problems in effective management of GI carcinoid disease is the difficulty in diagnosis. Part of this difficulty is undoubtedly attributable to the nonspecific nature of many of the symptoms patients report. This frequently results in series of unfruitful diagnostic tests that often culminate in exploratory laparotomy. In the present series, 58% of patients reported abdominal pain; only 24% presented with obstruction, and 11% with evidence of GI bleeding. Twenty-one percent of patients reported diarrhea; half of these cases occurred in patients with pancreatic carcinoid tumors, highlighting the ability of these tumors to secrete a variety of hormones. Another factor confounding diagnosis is the rarity of carcinoid disease, with an incidence of 1 to 2.5 per 100,000 population. 6,7 This rarity, combined with the nonspecific GI symptoms, makes diagnosis a challenge.

Carcinoid syndrome is frequently discussed in relation to carcinoid tumors. However, the complex of flushing, diarrhea, abdominal pain, and occasional asthma or right-sided valvular problems is actually uncommon: only 10% of carcinoids exhibit some of these symptoms. 9,10 In this series, 13% of patients had carcinoid syndrome, with the vast majority having ileal tumors. This slightly high number may be explained by the exclusion of appendiceal carcinoids, which rarely have liver metastases. No patient in this series had the carcinoid syndrome in the absence of liver metastases, as has been rarely described in the literature in tumors draining into the systemic circulation. 11

In contrast to many other studies, 9,12,13 in which 40% to 60% of tumors are found incidentally, carcinoid tumors detected incidentally represented only 7% of the tumors in our analysis. This most likely can be attributed to two factors: the omission of appendiceal carcinoids, which are found incidentally most of the time, and the tertiary nature of our medical center, where patients have often been referred for definitive diagnosis or treatment. However, our study does corroborate the results of others in that patients with incidental tumors have favorable prognoses, with only 6 of 25 exhibiting liver metastases during the follow-up period. This observation is probably attributable to the relatively early diagnosis of these tumors, compared with patients in whom clinical symptoms have developed.

Size, 9,12–16 depth of bowel wall invasion, 9,12,13 site of origin, 6,9,12,13,17 age, 9,18 and histologic growth pattern 19 have all been shown to be predictive of survival in carcinoid disease. Of the above, size, depth of wall invasion, and histologic type were not recorded prospectively in this database. Because most patients in the series were referred to Duke after surgical management at outside hospitals, accurate assessment of the latter two has been difficult because of our inability to review pathology specimens for these variables. Regarding size, most series have agreed that size greater than 2 cm predicts metastasis and disease-specific death. 9,13–16 However, some authors have challenged the notion that small size accurately predicts long-term survival, reporting that 18% to 29% of ileal tumors less than 1 cm may develop liver metastases. 12,20 Our series corroborates this: almost 10% of patients with tumors less than 1 cm had metastatic disease.

Site of origin has universally been found to predict survival, with the appendix 6,9,15,17 having the best survival and either the midgut or hindgut having the worst. 9,13,18 Our data are similar in many respects. First, considering survival regardless of stage categorized by site of origin, a pattern emerges. Pancreatic carcinoid tumors have a dismal prognosis, which has been shown by others, 15 and which is undoubtedly related to almost universal existence of metastatic disease at diagnosis. Excepting this unfortunate subset, patients with midgut tumors in our series did significantly worse than others in terms of survival.

However, in a previous analysis of foregut carcinoids, we found by multivariate analysis that site of origin was not predictive of survival when stage was taken into account. 8 Because of this, we examined the effect of tumor stage on survival in the larger series. Across various sites of origin, the 10-year survival curves for local disease at diagnosis were not significantly different. However, when only patients with metastatic disease to the lymph nodes or to distant sites were considered, hindgut and foregut carcinoid Kaplan-Meier curves could be virtually superimposed. Patients with metastatic midgut tumors had significantly better survival than those with tumors arising from other regions.

Because ileal carcinoids are resected with wide mesenteric resection, which has been shown to improve survival, 21 the possibility arises that the improved survival in midgut tumors may be influenced by therapy of the primary tumor. However, performing the same analysis with the exclusion of patients found to have regional disease continued to yield a significant survival advantage with midgut carcinoid tumors. These observations support the hypothesis that carcinoids arising from different embryologic sites are not only histologically and functionally different but also exhibit different clinical behaviors.

In addition to staging criteria, we examined the neurohormonal profiles of the patients in the database. Differences exist between sites of origin as to levels of serotonin and serotonin breakdown products. Midgut carcinoids had high levels of urinary 5-HIAA as well as serum and platelet serotonin. This is consistent with the results of other studies, which biochemically profile midgut carcinoids as producing high levels of serotonin and tachykinins. 22,23 Although the serum and platelet serotonin levels are detectable earlier than urinary 5-HIAA, they have not changed the management of patients with these tumors and are not directly correlated with disease burden, as 5-HIAA has been. With the advent of octreotide treatment for symptomatic patients, which markedly blunts serotonin production, 5-HIAA may also not correlate well with disease burden and may not be a reliable marker of recurrent disease.

Foregut carcinoids tend to have low levels of urinary 5-HIAA and serum serotonin. However, many foregut tumors do not contain DOPA decarboxylase and therefore cannot convert 5-hydroxytryptophan to serotonin. As previously reported, 24 when 5-hydroxytryptophan is released into the systemic circulation, DOPA decarboxylase in the renal parenchyma converts it to serotonin, leading to high levels of urinary serotonin. Thus, urinary serotonin may be a more reliable marker of liver metastases in these foregut patients. Although carcinoids of pancreatic origin are classified as foregut carcinoids, their biochemical profile falls in between that of midgut tumors and the other foregut carcinoids. They have relatively high levels of both urinary 5-HIAA and urinary serotonin. These values correlate with each other in these patients, suggesting that, in patients with pancreatic carcinoids that produce serotonin, some cells have active DOPA decarboxylase and others do not.

Serotonin levels strongly correlated with both liver metastases and disease-specific survival across sites, but they do not provide evidence of more aggressive tumor biology. In fact, the patients with the highest levels of serotonin production, the midgut group, did better than the metastatic foregut and hindgut groups in terms of survival. This raises the question as to what makes these latter tumors act more aggressively. Perhaps the foregut and hindgut tumors produce factors such as tumor necrosis factor-alpha, which have been shown to lead to cachexia and death. Monitoring the production of other hormones by these tumors may provide clues as to the factors involved in the worse prognosis of these subtypes once metastasis occurs.

Our analysis leads to the following conclusions. Pancreatic and midgut carcinoids are more often metastatic at diagnosis, which leads to a worse overall prognosis. Among patients with distant metastases, patients with midgut primary tumors have significantly improved survival compared with other primary sites, although they appear to be more hormonally active. The increased metastatic incidence of midgut carcinoids may be explained by diagnostic delay rather than biologic aggressiveness.

Discussion

Dr. Stanley R. Friesen (Prairie Village, Kansas): I compliment the authors for their emphasis on the function of carcinoid tumors as related to clinical survival and the aggressiveness and location of the tumors. They emphasized the neuroendocrine aspects of these tumors, which confirms the early studies of Paul Masson, who found a very strong neural component with these endocrine tumors way back in the 1920s.

These authors correlated the amine-secreting function of these tumors with their clinical picture. I want to point out, even though this may be beyond the scope of their study, that these kinds of tumors, being neural and submucosal in origin, will also secrete peptides. These peptides, such as ACTH and VIP and gastrin, produce clinical pictures which entirely overwhelm the picture of the carcinoid syndrome due to the amine secretion of these tumors. These being submucosal means that the tumors are sometimes very difficult to find; the neural origin in the submucosa of the intestine is therefore important.

The pancreatic carcinoids that you described are also due to neural tumors, probably not in the islets but among the islets and among the exocrine cells in the pancreas. This means that duodenal and pancreatic carcinoids are very difficult to diagnose, very hard to find, even by surgeons or pathologists.

For example, in 1968 and 1970, I presented to this Association a series of patients who had carcinoid tumors of the submucosa of the duodenum and the Zollinger-Ellison syndrome. They had total gastrectomy, with normal serum gastrin levels and negative secretin tests after the operation. I attributed the remarkable results of this to a so-called gastric factor.

Quite a few years later, we finally tumbled onto the idea that the gastric factor was mythical and that there must be something else. By that time, we recognized submucosal tumors, yellow tumors that have a carcinoid cell appearance. We recognized them by doing duodenotomies.

So my colleague, Dr. Romero Delcore, asked the pathologists to dig up the original gastrectomy specimens to look at these again. The pathologists studied them, and in four of the five specimens they were able to find, very small, minute, occult duodenal carcinoid gastrinomas. I think that paper, when it was reported in 1991, retracts the idea of the gastric factor and also points out that these tumors are very difficult to find. Now when we do duodenotomies we look at the duodenum, feel it, remove the tumor and its metastases under omeprazole protection. So my question is, in this large series that you report, did you find any clinical evidence of peptide secretions from these tumors?

Presenter Dr. Mark W. Onaitis (Durham, North Carolina): The only hormones and peptides that we looked at were the serotonin levels that I have stated here. Dr. Feldman, who is the endocrinologist who kept this database, has also looked at several other factors, including TNF-α, which he thinks may be responsible for the cachexia and more dismal prognosis of the patients with those tumors. So I think that TNF-α or some other factor may be responsible for the poor prognosis of those patients. But we did not look at any other peptide hormones.

Dr. Andrea Frilling (Essen, Germany): The definitions of carcinoids have changed over the years. As you have reviewed the data of patients treated over nearly 30 years, what were your histopathologic criteria for classification of a tumor as a carcinoid?

In addition to the question of Dr. Friesen, I would like to ask if you performed any immunohistologic studies. In particular, did you examine your tumors for chromogranin A, since this might be probably the most sensitive tumor marker for carcinoids?

Independent of the origin of the tumors, carcinoids can occur either as sporadic or familial tumors, as a later complement of MEN 1 syndrome. How many of your patients had sporadic and how many of your patients had hereditary disease? There might be a difference in the prognosis of these patients, as we saw in patients with medullary thyroid carcinoma.

Finally, as any endocrine tumors, carcinoids express somatostatin. In the presence of somatostatin receptor subtype 2, tumor growth and secretion can be influenced by octreotide. Did your group perform any studies on this topic?

Dr. Onaitis: The histologic diagnosis was performed by the pathologists looking for the characteristic vesicles. We went back and looked at histology and excluded some patients from the analysis so that everyone in the study had a carcinoid tumor. As far as immunohistochemical markers, we have used chromogranin A since 1987 and synaptophysin since the early 1990s.

Four patients had the MEN syndrome of patients with foregut tumors. These patients did not do statistically significantly worse than any of the other tumors. Interestingly, the patients who had chronic atrophic gastritis performed better than the rest of the foregut subgroup. We did not look at somatostatin levels as they relate to the biology of carcinoid tumors.

Dr. Keith A. Kelly (Scottsdale, Arizona): I have a question about the better survival in the patients with metastatic functioning midgut tumors and wondered if you had an explanation for that. It occurred to me that perhaps these tumors were better differentiated and hence less virulent tumors, as evidenced by the fact that they were making hormones and other substances. Or do you think that the substances they make discourage the growth of metastatic deposits in some way?

Dr. Onaitis: I am not sure how to explain the survival gap. I think it is possible that the foregut and hindgut tumors are producing factors that we are not measuring in addition to serotonin. These factors, like TNF-α, may have resulted in the worse prognosis for those patients. It very well may be that the midgut tumors are more well differentiated and therefore have a more indolent course even when they metastasize. It is definitely something worth looking into in the future.

Dr. Edwin L. Kaplan (Chicago, Illinois): I think it might be interesting, as the others stated, to look at your duodenal carcinoids and see if you included, for example, any gastrinomas or whatnot. I think you probably have not.

An interesting thing is that even though they look histologically like duodenal carcinoids, gastrinomas that have just lymph node metastasis without any other distant spread have a very good prognosis, far better than what is illustrated by your duodenal carcinoids. Do you wish to comment on that, please?

Dr. Onaitis: We did not include any patients who had a diagnosis of gastrinoma in the series. We tried to stick to the histologic and biochemical definition of carcinoid to exclude those patients. I think there were three of them in Dr. Feldman’s database that we did not include in this study.

Dr. James B. D. Mark (Stanford, California): Conventional wisdom has it that carcinoid tumors of the gut that metastasize to the liver are the ones that cause carcinoid syndrome. I have often wondered why pulmonary carcinoids, which are proximal to the liver, so rarely cause carcinoid syndrome. It may be, I thought, because they presented early with hemoptysis or a lesion on a chest x-ray or obstructive symptoms.

I discussed this at one time with the late Dr. Charles Carrington, a pathologist at our institution, and it was his opinion that it was the size of the tumor, not the fact that it metastasized to the liver, that was responsible for the carcinoid syndrome, since these tumors are very inefficient producers of hormone and it takes a lot of tumor size to produce enough hormone to cause the carcinoid syndrome. Do you agree or have any comments on that hypothesis?

Dr. Onaitis: I do agree that the higher the tumor burden, the more hormone production you are going to have that can be dumped into the systemic circulation. I think that the bronchial carcinoids are also foregut carcinoids which don’t contain DOPA decarboxylase and may not be able to produce some of the hormones that the midgut and hindgut tumors can. So that may be at the root of some of the difference as well.

Dr. Haile T. Debas (San Francisco, California): I don’t think the size necessarily relates to the symptoms these tumors produce. Some of the worst symptomatic cases we have seen, we can’t even find the tumors. I think it is the efficiency with which they produce the peptides and the lack of somatostatin production by those peptides that determine how symptomatic they will be.

Footnotes

Correspondence: Douglas S. Tyler, MD, Dept. of Surgery, Duke University Medical Center, DUMC Box 3118, Durham, NC 27710.

Presented at the 120th Annual Meeting of the American Surgical Association, April 6–8, 2000, The Marriott Hotel, Philadelphia, Pennsylvania.

E-mail: tyler002@acpub.duke.edu

Accepted for publication April 2000.

References

1. Obendorfer S. Karzinoide tumoren des Dunndarms. Frankfurt Zeitschr Pathol 1907; 1: 426–432.
2. Stinner B, Kisker O, Zielke A, Rothmund M. Surgical management for carcinoid tumors of the small bowel, appendix, colon, and rectum. World J Surg 1996; 20: 183–188. [PubMed]
3. Williams E, Sandler M. The classification of carcinoid tumors. Lancet 1963; 1: 238–239. [PubMed]
4. Kema I, Vries E, Schellings A, et al. Improved diagnosis of carcinoid tumors by measurement of platelet serotonin. Clin Chem 1992; 38: 534–540. [PubMed]
5. Janson E, Holmberg L, Stridsberg M, et al. Carcinoid tumors: analysis of prognostic factors and survival in 301 patients from a referral center. Ann Oncol 1997; 8: 685–690. [PubMed]
6. Godwin J. Carcinoid tumors: an analysis of 2837 cases. Cancer 1975; 36: 560–569. [PubMed]
7. Modlin I, Sandor A. An analysis of 8305 cases of carcinoid tumors. Cancer 1997; 79: 813–829. [PubMed]
8. Kirshbom PM, Kherani AR, Onaitis MW, et al. Foregut carcinoids: a clinical and biochemical analysis. Surgery 1999; 126: 1105–1110. [PubMed]
9. Shebani K, Souba W, Finkelstein D, et al. Prognosis and survival in patients with gastrointestinal tract carcinoid tumors. Ann Surg 1999; 229: 815–827. [PMC free article] [PubMed]
10. Vinik A, Thompson N, Eckhauser F, Moattari A. Clinical features of carcinoid syndrome and the use of somatostatin analogue in its management. Acta Oncol 1989; 28: 389–402. [PubMed]
11. Feldman J, Jones R. Carcinoid syndrome from gastrointestinal carcinoids without liver metastases. Ann Surg 1982; 196: 33–37. [PMC free article] [PubMed]
12. Thompson G, van Heerden J, Martin JJ, et al. Carcinoid tumors of the gastrointestinal tract: presentation, management, and prognosis. Surgery 1985; 98: 1054–1063. [PubMed]
13. Zeitels J, Naunheim K, Kaplan E, Il F. Carcinoid tumors: a 37-year experience. Arch Surg 1982; 117: 732–737. [PubMed]
14. MacGillivray D, Snyder D, Drucker W, ReMine S. Carcinoid tumors: the relationship between clinical presentation and the extent of disease. Surgery 1991; 110: 68–72. [PubMed]
15. McDermott E, Guduric B, Brennan M. Prognostic variables in patients with gastrointestinal carcinoid tumors. Br J Surg 1994; 81: 1007–1009. [PubMed]
16. Moertel C, Saver W, Dockerty M, et al. Life history of the carcinoid tumor of the small intestine. Cancer 1961; 14: 901–912. [PubMed]
17. Morgan J, Marks C, Hearn D. Carcinoid tumors of the gastrointestinal tract. Ann Surg 1974; 180: 720–727. [PMC free article] [PubMed]
18. Greenberg R, Baumgarten D, Clark W, et al. Prognostic factors for gastrointestinal and bronchopulmonary carcinoid tumors. Cancer 1987; 60: 2476–2483. [PubMed]
19. Johnson L, Lavin P, Moertel C, et al. Carcinoids: the association of histologic growth pattern and survival. Cancer 1983; 51: 882. [PubMed]
20. Strodel W, Talpos G, Eckhauser F, Thompson N. Surgical therapy for small bowel carcinoid tumors. Arch Surg 1983; 118: 391–397. [PubMed]
21. Moertel C. An odyssey in the land of small tumors. J Clin Oncol 1987; 5: 1503–1522. [PubMed]
22. Chejfec G, Falkmer S, Askensten U, et al. Neuroendocrine tumors of the gastrointestinal tract. Pathol Res Pract 1988; 183: 143. [PubMed]
23. Falkmer S, Martensson H, Nobin A, Sundler F. Peptide hormones in various types of gastro-entero-pancreatic tumors: immunohistochemical patterns and evolutionary background. Prog Cancer Res Ther 1984; 31: 597.
24. Feldman J. Serotonin metabolism in patients with carcinoid tumors: incidence of 5-hydroxytryptophan-secreting tumors. Gastroenterology 1978; 75: 1109–1114. [PubMed]

Articles from Annals of Surgery are provided here courtesy of Lippincott, Williams, and Wilkins

Formats:

Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...

Links

  • Cited in Books
    Cited in Books
    PubMed Central articles cited in books
  • Compound
    Compound
    PubChem Compound links
  • MedGen
    MedGen
    Related information in MedGen
  • PubMed
    PubMed
    PubMed citations for these articles
  • Substance
    Substance
    PubChem Substance links

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...