Institutional and Patient Criteria for Heart-Lung Transplantation: Health Technology Assessment Number 1

Patient Selection Criteria

Selected published studies of HLT are summarized in Table 1. (5,12-22). Compilations of such medical conditions as well as contraindications to HLT have been extracted from these and other studies (3,5,12-36). and from a personal communication (Starnes VA, Stanford University, August 1990). They are listed in outline form in Tables 2 and 3, respectively.

Several of the authors cited in Table 1 and in References 3, 5, and 23-32 and in the personal communication from Starnes indicated that suitable candidates should have the following general characteristics: They should have a life expectancy of only 6 to 12 months with a deteriorating course, should be unresponsive to medical therapy, and should be severely disabled, e.g., the New York Heart Association's functional classification III or IV (class III = comfort is present at rest, but less than ordinary physical activity results in dyspnea or fatigue; class IV = dyspnea or fatigue may be present at rest and is made worse by any physical effort). Patients should have no other serious medical problems, and they should be psychologically willing to undergo the stressful surgery and postoperative care necessary for a successful

Few of the reports shown in Table 1 (5,12-26). provide cardiac or pulmonary functional criteria in the form of actual numerical values that quantitate the degree of disability. Rather, terms such as "increased pulmonary artery pressure," "cardiomegaly," and "corpulmonale" are used. In addition, appropriate patient ages have been specified as follows: 5 to 45 years of age, younger than 50, and no older than 60 years of age. Preoperative prognosis has been specified as expected survival of less than 6 months, "severely disabled," "poor prognosis," and "deteriorating clinical condition." Thus, the patient selection criteria for HLT are not defined in any quantitative, systematic, or uniform manner in the published literature.

Another source of information relevant to criteria for patient selection was a report in 1991 by Evans et al in the National Cooperative Transplantation (37). This was an opinion survey carried out at 366 transplant programs in the United States. The survey indicated that the respondents believed that among the more severe contraindications for HLT were human immunodeficiency virus (HIV) antibody positivity, chronic renal failure, a recent history of intravenous drug abuse, a history of cerebrovascular accidents resulting in significant impairment, treatment with 50 mg or more of prednisone per day, Down's syndrome, active smoking, primary heart or lung malignancy, history of noncompliance with medical regimens, recurrent pulmonary emboli, and an age over 60.

It should be noted that the patient selection criteria for HLT are in flux, partly because a number of diseases previously treated with HLT can now be treated by LT.(7,38)

Patient Factors

Comprehensive registries contain far more data on patient survival following HLT, diagnosis, and age than do published case studies. Data from registries are often more current than those from the published literature, and the data set from the international registry is larger than that from the United States alone. This type of information cannot be obtained easily from reports such as those cited in Table 1, because each case series contains only a relatively small number of patients with a particular diagnosis, and information on long-term outcome is limited by the duration of patient followup prior to publication.

Tables 4and 5Table 4

These data also are plotted in Figure 1, together with the survival data from studies of more than 20 patients cited in Table 1. The survival curve from the registry data is comparable to the individual case series data cited in Table 1. Data from the United States on 374 patients are shown in Table 5. The survival rates are quite similar to the international data.

Among the factors examined that affect patient survival rates are the diagnosis and age at the time of surgery. Information addressing this point from RISHLT is shown in Tables 6 and 7 for four of the diseases treated with HLT. As can be seen, patients with congenital heart disease and primary pulmonary hypertension have a worse survival rate at all time periods than do patients with Eisenmenger's syndrome or cystic fibrosis.

Tables 8 and 9 show data from two sources that demonstrate the relationship between the age of the patient at the time of HLT and the survival rates al 1, 3, and 5 years. The data indicate that age has very little influence on survival rates.

Other patient characteristics may affect survival rates in patients undergoing HLT. Because sufficient data are not yet available for HLT, the information regarding HT may be the best available surrogate data by which to examine such patient characteristics; HT and HLT share many medical, technical, and logistic features.

The Battelle-Seattle Research HT study, The Center Effect in Heart Transplantations, (40). was based on data collected by UNOS, the International Society for Heart and Lung Transplantation, the Battelle-Seattle Research Center, HCFA, and the American Hospital Association. This analysis, based on transplants performed in 1988, was from data received from 106 of the 117 transplant programs contacted (in which 1,602 transplants were performed). This figure represented 97.3 of all transplants performed worldwide in 1988. The effects of certain patient variables are shown in Table 10. The use of life-support systems, artificial hearts, or other assist devices, as well as retransplantation, had significant adverse effects on 1-year survival rates. Survival was positively affected by the use of triple or quadruple immunosuppressive protocols. Differences in age, sex, and diagnosis had no effect on 1-year survival rates.

Institutional Factors

At the present time, there are few data available that specifically address guidelines and criteria for institutions performing HLT. However, HT was performed earlier than HLT, and guidelines and criteria for institutions performing HT began to evolve in the early to mid-1980s. Since virtually the same institutions performing HT also have been leaders in the development of HLT, these standards and guidelines are useful as an initial model. In 1986, the Task Force on Organ Transplantation of the US Department of Health and Human Services developed criteria for cardiac transplantation centers which included many aspects of institutional organization and function. These include:

• Methods and criteria for the selection of patients.

• Patient management plans.

• Commitment of resources, including those in the areas of cardiology, surgery, anesthesiology, immunology, infectious disease, pulmonary disease, pathology, blood banking, radiology, nursing, and social services. Individuals must be identified in these areas, and the responsible medical/surgical members must be board-certified or eligible or must have demonstrated transplantation competence. The various members of the team must be organized and integrated into a stable, functioning unit.

• Facility plans that include commitment of resources that will ensure a reasonable concentration of experience.

These requirements then were used to determine which institutions were to be certified for Medicare coverage of HT, and the resulting rules were published in April 1987 in the Federal Register.(11). In brief, they comprise several categories that must be addressed by each center submitting an application. These include patient selection and management, commitment, facility plans (plans to perform 12 or more cardiac transplantations per year), experience, and survival rates. A facility must have an established program with documented evidence of 12 or more transplants in each of the two preceding 12-month periods and 12 patients prior to that, since January 1, 1982. Thus, for approval, an institution must have performed 36 HTs. The facility also must demonstrate actuarial survival rates of 73 for 1 year and 65 for 2 years for transplants performed since January 1, 1982. The number of cases required were chosen, in part, because they were large enough to determine statistically meaningful survival outcomes.

Before considering the number of HLTs to be performed as an institutional criterion, it is pertinent to consider the general subject of the number of surgical procedures performed and patient outcomes. A number of studies support the notion that there is a positive relationship between the number of a particular type of surgical procedure performed at an institution and better patient outcomes.(41,42). This effect was seen in some types of surgical procedures but not in others, and, in general, this positive relationship was more often observed in more complex surgical procedures. There are several alternative explanations for such a relationship. First, it is possible that patient selection factors and different mixes of patients in different types of institutions are the cause of the observed effect. The second explanation is that, indeed, "practice makes perfect," and institution with higher volumes and greater experience with particular procedures learn by doing and thus have better results. The third explanation is that the observed positive relationship is due to the fact that hospitals with more skillful surgeons and operating and recovery room teams are known to provide superior outcomes and thus may attract more patients suffering from a particular condition. In this context, larger surgical volumes in particular institutions are due to referral patterns that lead to more patients being sent to institutions that have better results. An analysis of the second and third hypotheses concluded that, for different surgical procedures and medical conditions, one or the other of these explanations may be more important.(43) The studies, however, did not directly examine these relationships with regard to HT or HLT.

With regard to organ transplantation in general, the Battelle-Seattle Research Center recently published a comprehensive report, The Concept of Center Effect in Organ Transplantation.(44) This study concluded, in part, that there is little evidence, based on published reports, that the volume of procedures performed has a beneficial impact on the outcomes of transplantation. However, this report did not specifically address HT.

New data concerning institutional criteria for HT now are available and can be used to evaluate the outcome probability relating to the volume and several other criteria of HT centers. Since too few HLT procedures have been performed at most institutions to allow for reliable statistical analyses, the information regarding HT may be the best available surrogate data by which to establish initial institutional criteria for HLT.

The Battelle-Seattle study of HT has been cited in regard to certain patient characteristics.(40) This analysis showed considerable variation of patient survival rates at different centers. For example, in the institutions examined, 10 centers had 1-year survival rates of 50 or less, 67 had 1-year survival rates greater than 80, and 32 reported 1-year survival rates greater than 95. A complex array of factors may contribute to such differences in survival rates among institutions. The study assessed the source of this variation both in terms of differences in the patient mix (age, race, sex, and diagnosis prior to transplant) across centers and differences among the centers themselves. More particularly, this study addressed such transplant center variables as cardiologist experience, surgeon experience, and program volume. When these factors were analyzed in four multivariate models, the following was observed: In the first model, a program volume of less than 12 had an adverse effect on patient survival. In the second model, when surgeon experience equaled or exceeded 50 transplants, there was a modest, positive effect on survival. In a third model, Medicare approval in 1988 or earlier had a positive effect on patient survival, suggesting that Medicare institutional criteria were valid.

It should be noted that some of these data were obtained after the HCFA institutional recommendations had already been put into effect, and it is possible that implementation of these recommendations may have had a positive effect on outcomes. In summary, the data in this report give some support to a hypothesis that there is a positive correlation between the number of HTs performed at an institution and surgeon experience and the 1-year survival rate of HT patients.(40) While this study included a large sample of more than 1,600 patients, these observations were limited to a single year (1988), and extrapolation beyond these data is problematic. Another such analysis of institutional influences on the outcomes of HT is now available.(45) Information for this study was obtained from RISHLT and a survey of 56 HT centers in the United States. The results were based on an analysis of 1,123 recipients who received HTs between 1984 and 1986.

Several institutional variables affecting the 90-day postoperative survival rate were considered. The correlation between outcome and previous HT experience by the members of the transplant team (i.e., cardiologist, transplant coordinator, and surgeon) was assessed, and it was found that previous experience of both the cardiologist and the coordinator played a major role. The risk of mortality was increased by factors of 2.7 and 2.0, respectively, when these individuals lacked prior experience. However, no correlation was found between surgeon HT experience and outcome.

Second, a number of variables related to the "learning curve" phenomenon were found to be significant. The mortality rate for the first five patients receiving transplants in an institution was significantly higher than that of subsequent patients (19.8 compared with 11). This finding not attributable to differences in patient selection. In addition, experience in HT on low-risk patients was found to be transferable to high-risk patients, since the mortality of a high-risk patient was inversely related to the number of low-risk patients previously operated on in that institution. A high transplant rate in startup centers was positively correlated with a favorable outcome but did not reach a level of statistical significance. The authors noted that there was an inverse relation between transplant program duration and mortality. Older transplant centers, without exception, have lower mortality rates than newer ones. This difference could not be explained by the differences in patient selection in the various centers.

However, despite the fact that the 56 HT centers exhibited marked heterogeneity in total volume (1 to 207) and in mortality (0 to 75), there was no correlation between these variables (r = -.03).

Finally, another factor that may have influenced HT outcomes during the time of the above study is the use of triple-drug therapy (cyclosporine, prednisone, and azathioprine). In the years between 1984 and 1986, use of triple-drug immunosuppressive therapy grew rapidly and could conceivably have been a factor resulting in the more favorable outcomes in the latter part of this study. Patients not receiving triple-drug therapy had a relative risk of 2.0 compared with patients receiving such therapy. This same beneficial effect of triple immunosuppressive therapy on survival also was seen in the Battelle-Seattle analysis.(40)

In summary, the Battelle-Seattle study(40). and the Laffel study(45). showed somewhat different results. In the Battelle-Seattle study, low program volumes had an adverse effect on patient survival, and surgeon experience had a positive effect on outcomes. However, in the Laffel study, this latter effect was not observed; rather, the degree of experience of the transplant coordinator and the cardiologist had a positive effect on patient outcomes, as did the duration of the transplant program. The Laffel study observed no correlation between overall program volume and patient outcomes. However, the authors did note an inverse relationship between cumulative experience with HT and postoperative mortality. It would, therefore, appear that a complex mix of factors affects survival outcomes.

Information about HLT, although not as extensive as that about HT, is also available from the UNOS Report of the US Scientific Registry for Organ Transplantation.(6). A total of 150 HLTs were performed at 37 institutions in the United States between October 1, 1987, and December 31, 1989. Nineteen of these institutions performed only one or two Figure 2 shows the relationship between the number of HLTs performed and 1-year survival rates. As can be seen, programs performing a small number of transplants have greater variability in survival rates than do programs performing more transplants. Thus, survival rates in institutions with low patient volumes are not very meaningful. Twenty of these 37 institutions also performed HT and LT; the number of all three procedures performed per institution during that 27-month period ranged from 5 to 160 with a mean of 46. However, only three of the 37 institutions did more than 10 HLTs. As previously discussed, the number of particular procedures performed at a given center must be considered a significant factor in proposing institutional criteria. Thus, since the number of HLTs performed at individual institutions is often low, a combined figure including HLT, HT, and LT should logically be used as a basis for institutional review. The inclusion of HT and LT is based on the fact that many of the technical, surgical, patient-supportive, and logistic elements are shared among these three procedures.

Finally, limited evidence as to whether setting specific institutional standards can affect HT patient outcome is available from Medicare experience with HTs performed between August 1, 1987, and December 31, 1991. A total of 5,553 HTs were performed in Medicare-approved hospitals, and 2,332 Hts were performed in non-Medicare-approved hospitals. Life-table analysis of survival in Medicare-approved hospitals was 83. 8 +/- 0.7 at 1 year and 76.7 +/-0.7 at 3 years; the survival in non-Medicare-approved hospitals was 80.7 +/-0.6 at 1 year and 72.5 +/-0.7 at 3 years. The p value for this difference, .001, is statistically significant (Shekar S and Krakauer H, Health Care Financing Administration; personal communication, August 1992). Thus, Medicare approval is associated with a small but statistically significant improvement in 1-to 3-year survival after HT.

Patient Criteria

Lung and heart-lung transplantations have been done in patients with irreversible, progressively disabling end-stage lung disease (or heart and lung disease) with survival expected to be less than 12 to 18 months or less than 1 or 2 years. For each patient, this is the professional judgment of an experienced physician. It must be based on thorough clinical and physiologic assessment of the patients and observation of responses to medical therapy and attempts at rehabilitation.

In brief, both LT and HLT in carefully selected patients and by experienced teams yield significant increases in survival with reasonable quality of life and at costs that are not out of the range for other expensive therapies. However, at present, HLT is used less commonly and primarily in situations in which there is coexistence of end-stage pulmonary disease with advanced cardiac disease that is not secondary to pulmonary hypertension.

Affirmative criteria include:

• The patient must have irreversible, progressively disabling, end-stage pulmonary or cardiopulmonary disease (e.g., less than a 50 likelihood of survival for 8 months). Prognosis otherwise must be good both for survival and rehabilitation.

• All other medical and surgical therapies that might be expected to yield both short-and long-term survival comparable to that of transplantation must have been tried or considered.

• The patient must have a realistic understanding of the range of clinical outcomes that may be encountered.

• Plans for long-term adherence to a disciplined medical regimen must be feasible and realistic for the individual patient.

Many factors must be recognized as exerting an adverse influence on the outcome after transplantation. A patient who meets the above criteria and is free of the adverse factors cited on tne next page is considered a good candidate for transplantation.

Some of the factors cited below have important qualifications. It is recognized that some patients who may not be considered "good candidates" might also benefit from transplantation, but the likelihood or extent of benefit is thought to be significantly less.

Adverse criteria include:

• Acute illness (i.e., serious exacerbation of chronic end-stage disease or with nonchronic end-stage disease) or current requirement of mechanical ventilation for more than a very brief period (because there is difficulty in adequate assessment, a propensity for infection, and likelihood of poor results).

• Significant systemic or multisystem disease.

• Extrapulmonary site of infection (because of the probability of recrudescence once immunosuppression is instituted).

• Hepatic dysfunction, even secondary to right ventricular failure.

• Renal dysfunction, such as a preoperative serum creatinine greater than 1.5 mg-dL or a 24-hour creatinine level clearance less than 50 mL/min.

• Systemic hypertension that requires multidrug therapy for even moderate control.

• Cachexia, even in the absence of major end-organ failure (because of the significantly less favorable survival of these patients).

• Obesity, with weight being an increasingly severe adverse factor as the patient's weight exceeds 20 over the ideal weight for height and sex.

• A history of a behavior pattern or psychiatric illness considered likely to interfere significantly with compliance with a disciplined medical regimen.

• Continued cigarette smoking or failure to have abstained for a long enough time (at least 1 or 2 years) to indicate low likelihood of recidivism.

• Previous thoracic or cardiac surgery or other bases fro pleural adhesions that may be serious adverse factors depending on the site of thoracotomy/stermotomy.

• Age beyond 50 or 55 (an increasingly severe adverse factor).

• Recent or current history of gastrointestinal problems (because of common postoperative gastrointestinal problems and hemorrhage).

• Chronic corticosteroid therapy that cannot be tapered to a low dose and/or discontinued prior to transplantation. This has been considered a serious adverse factor by many because of the increased risk of tracheal or bronchial dehiscence in the early postoperative period.

Institutional and Team Experience

Institutional and team experience should be based on all LTs and HLTs performed since a given date that are at least 4 years prior to the date at which experience and survival are reported.

To be accepted for LT (i.e., both single and double lung), an institution and team should have:

Performed LT and/or HLT in at least 10 patients within the 12 months prior to application and in at least an additional 10 patients prior thereto, and achieved a documented Kaplan-Meier actuarial survival rate of no less than 65 at 1 year.

To be accepted for performance of HLT, an institution and team should:

Fulfill the above criteria for LT and the existing HCFA and Civilian Health and Medical Program of the Uniformed Services (CHAMPUS) criteria for HT. Thus, to perform HLTs, the institution and team would have to fulfill the criteria for both HT and LT.

Another NIH institute indicated that an important risk factor for HLT would be the presence of human leukocyte antigen (HLA) antibodies in the recipient. If these antibodies are present, it is likely that these transplanted organs might suffer hyper-acute rejection. They also stated that in HT recipients with HLA antibodies, there is a poorer 5-year survival rate and an increased incidence of coronary artery disease.

In response to the Federal Register notice of this assessment and solicitation of information and opinion from individuals and groups having experience with HLT, OHTA has received a number of comments (one of which will be cited in detail). The responses of these groups are summarized in Tables 11 and 12.

HRSA forwarded OHTA's request to the Thoracic Organ Transplantation Committee of OPTN. This committee provided the response cited below.

In general, the surgical principles for HLT are similar, although somewhat more complex, than those for HT. The basic immunosuppressive protocol and the approach to patient care should be essentially identical to that of cardiac transplant patients with the single exception that close surveillance of the lungs for possible rejection is critically important.

Medical indications for which HLT is performed include end-stage Eisenmenger's syndrome and primary pulmonary hypertension. Less common indications for HLT include cystic fibrosis with end-stage pulmonary failure and combined end-stage disease of both the heart and lungs.

The same general contraindications to HT also apply to HLT. Patients also are at particularly high risk during HLT if they have had previous major thoracic or mediastinal operations, due to the likelihood of excessive bleeding at the time of transplantation.

Institutional requirements for HLT are generally the same as those for HT. Specifically, a team that is qualified through experience and survival in HT is coupled with appropriate personnel. If the institution has been approved for CHAMPUS coverage of HT, then that institution also should be covered for HLT (if they have performed three or more HLTs with a 60 1-year survival). Given the appropriate degree of surgical expertise, HLT is a natural extension of HT in a highly successful and experienced HT program. HRSA stated that the information provided by this committee represents the opinion of its members and that OPTN and HRSA do not have an official opinion on this matter.

The commentary received from several transplant centers (Table 11) regarding institutional criteria varied; e.g., the minimally suggested 1-year survival rates for HLT procedures ranged from 50 to 70. The number of HLT procedures considered necessary per institution ranged from 2 or 3 to 10 per year, or to a total of 25 HLTs. NIH, in this regard, recommended that an institution should have performed 20 HLTs or LTs and also fulfilled the existing HCFA and CHAMPUS criteria for HT. NIH, however, did not make a specific recommendation for the number of HLTs to be performed over a specific period of time. The total recommended NIH requirements therefore would be a total of 56 procedures (including HLT, HT, and LT) over a 3-year period. The ability of an institution to fulfill these requirements appears problematic, since only 15 of the 37 institutions performed more than 52 procedures (HLT, HT, and LT) in 37 months. Four institutions indicated that current CHAMPUS HT criteria also should apply to HLT; however, one institution indicated that this was inappropriate. Furthermore, another institution recommended that HLT should be performed only in a very limited number of institutions.

Recommendations regarding patient criteria also varied; the maximum age believed appropriate for HLT ranged from 45 to 60 years. There was general agreement that patients with irreversible cardiopulmonary disease are candidates for HLT. Among the diseases mentioned were pulmonary hypertension with congestive heart failure, cystic fibrosis, emphysema, bronchiectasis, and restrictive lung disease. Thus, there was a considerable range of opinion from responding institutions with regard to institutional criteria, and patient selection criteria were phrased only in general terms.

In response to the Federal Register notice of this assessment, the American College of Cardiology (ACC) also provided comment. The ACC indicated that many patients who were previously considered candidates for HLT are now more appropriately treated with LT. The advantages to individual patients, as well as advantages on a more general level due to the expanded supply of donor organs when hearts and lungs are used separately, are widely accepted concepts in the transplant community.

The ACC believes that absolute contraindications for HLT include:

• Active extrapulmonary infection.

• Other irreversible organ dysfunction, e.g., hepatic or renal failure.

• Malignancy (other than basal skin carcinoma) unless thought to be "cured. "

• HIV seropositivity.

• Previous thoracotomy involving pleurectomy or pleurodesis.

• Insulin-dependent diabetes with end-organ damage.

The ACC concluded that specific indications for appropriate cardiopulmonary (heart-lung or lung) transplantations are end-stage cardiopulmonary disease that cannot be treated by other types of medical or surgical management, and/or severe primary lung disease with only mild or absent right ventricular dysfunction. Patients to be considered for HLT should include:

• Patients with congenital cardiac abnormalities associated with severe pulmonary hypertension (Eisenmenger's complex).

• Patients with untreatable primary lung disease associated with severe secondary right-heart failure.

• Patients with cardiomyopathy associated with irreversible secondary pulmonary hypertension.

The ACC recommended that transplant team criteria should be similar to those already established for cardiac transplantation and should include the availability of a team of physicians, surgeons, nurses, social workers, and psychologists approved for cardiac transplantation. In addition, a separate group of pulmonary medicine specialists and surgeons with specialty training in thoracic surgery should be part of the team. A group of perfusionists capable of performing extracorporeal membrane oxygenation as well as prolonged left-and right-heart assist should be available as well. The physician team should include a minimum of two cardiothoracic surgeons, two transplant physicians, and one pulmonary medicine specialist. Institutional availability of expert consultation in the fields of infectious diseases, immunology, and nephrology should be the same as that required for cardiac transplantation. Such a center should be UNOS-and Medicare-approved for cardiac transplantation.

The ACC believes that existing CHAMPUS heart transplantation institutional and team criteria are appropriate for HLT and should be implemented as far as the team qualifications are concerned.

A brief summary of expert opinion taken from the published literature regarding contraindications to HLT is shown in Table 12.