Malignant Melanoma in Solid Transplant Recipients:  Collection of Database Cases and Comparison With Surveillance, Epidemiology, and End Results Data for Outcome Analysis FREE

Malignant melanoma (MM), an immune-responsive tumor,^1- 5 poses concern for adverse outcomes when encountered in immunosuppressed solid organ transplant recipients. The outcomes of MM in the general population have been well documented and depend greatly on the stage of disease.⁶ Key prognostic factors that determine the stage of disease include depth of invasion (reported as Breslow thickness), Clark level, presence of ulceration, lymph node involvement, mitoses, and distant metastasis.⁷ Outcomes in immunosuppressed solid organ transplant recipients are not as well documented, with data contained in a few case reports and case series.^8- 18 Furthermore, even less published evidence lists prognostic information to assist in the interpretation of outcomes among patients with MM and organ transplants.^19- 24 Some of us previously reported a series of cases from 1 institution,¹⁹ and although prognostic information was more complete than in prior reports, the number of patients studied was too small to make outcome conclusions based on statistically valid comparisons. The present study is a continuation of the previous work, with the addition of cases collected from national databases, and uses individuals with cases of MM reported to the Surveillance, Epidemiology, and End Results (SEER) Program of the National Cancer Institute to derive expected survival rates for comparison.

Three unique subgroups exist within the population of solid organ transplant patients with MM, each with its own set of predisposing risk factors, potential outcomes, and management considerations: patients with MM diagnosed before transplantation, patients with MM diagnosed after solid organ transplantation, and patients with MM that developed through transmission from the organ donor (our study does not include patients in this third subgroup).

The primary objective of this study was to assess MM cause-specific and overall survival rates for patients with MM diagnosed after transplantation and to determine the risk of recurrence of MM after transplantation in those with a history of MM before transplantation. Survival data for the study cohort were compared with expected survival data derived from a national sample with MM.

Cases were identified from surgical and medical index databases (Mayo Clinic in Rochester, Minnesota, 1978-2007), physician recall (ie, recall from memory of physicians caring for patients; Mayo Clinic in Arizona and Mayo Clinic in Florida, 1999-2007), and national databases (the Organ Procurement and Transplantation Network/United Network for Organ Sharing [UNOS] database, 1999-2006, and the Israel Penn International Transplant Tumor Registry [IPITTR], 1967-2007). Data from the different sources were compared, and duplicate information was removed.

Maintenance, validation, and updating were attempted for cases of reported MM within the IPITTR. Original medical records and pathology reports were requested for all possible cases to clarify stage of disease and outcomes.

Abstracted data for all patients identified at Mayo Clinic sites and in the national databases included the following: date of birth, sex, race, Fitzpatrick skin type, family history of MM, personal history of nonmelanoma skin cancer, history of dysplastic nevi, history of more than 50 nevi, number of blistering sunburns, number of tanning bed exposures, date of latest follow-up, date and cause of death, date of first and any subsequent solid organ transplantation, type of transplant received, disease that led to end-organ failure, type of immunosuppressive agents at induction and for maintenance, number of short-term rejection episodes, presence of long-term rejection at latest follow-up, human leukocyte antigen type, date of diagnosis of MM, location of MM, Breslow thickness, Clark level, clinical or histologic ulceration, histologic regression, histiogenic type, radial or vertical growth, mitoses per 10 high-power fields, presence of tumor-infiltrating lymphocytes, neurotropism, angiolymphatic invasion, associated nevi, sentinel lymph node status, serum lactate dehydrogenase level, elective lymph node dissection status, imaging (ie, chest radiography, positron emission tomography, and computed tomography), American Joint Committee on Cancer (AJCC) 2001 pathologic stage, treatment of MM, presence of recurrence (defined as local recurrence within the scar of the original MM treatment site), presence and date of metastasis, and site of metastasis.

Cases were excluded if the patients did not have a documented transplantation history or a pathologic report of MM. Inclusion criteria included 1 or more melanomas documented by a pathology report or a clinical note and a known transplantation history also documented by a clinical note. Because many of the diagnoses were made before the establishment of the published 2001 AJCC staging criteria,⁷ much of the analysis was based on Breslow thickness as the primary prognostic indicator of disease because this was the most reported and consistent prognostic factor available during the time frame of the cases. Breslow thickness was categorized into 5 groups (0-0.75 mm, 0.76-1.50 mm, 1.51-3.00 mm, >3.00 mm, and unknown). The UNOS database recorded cases only according to these groupings and not with specific individual Breslow thickness for each case, thereby limiting the analysis to these groupings. Separate analysis of cases also was performed on the basis of Clark level and AJCC stage when this information was available.

Cases were separated into 2 subgroups: those in patients with MM diagnosed before their first solid organ transplantation and those in patients with MM diagnosed after their first solid organ transplantation. Within each subgroup, patient characteristics were compared between patients with and without Breslow thickness available using the χ² test for categorical variables and the Wilcoxon rank sum test for continuous variables. Duration of patient follow-up was calculated from the date of the patient's first MM diagnosis to the date of death or latest reported follow-up. The Kaplan-Meier method was used to estimate rates of overall survival, MM cause-specific survival, and survival free of metastasis after the initial MM diagnosis. The 95% confidence intervals (CIs) for the survival estimates were calculated using the log e transformation. Univariate associations between patient characteristics and each of the time-to-event outcomes were evaluated by fitting Cox proportional hazard models. Associations were summarized with hazard ratios (HRs) and corresponding 95% CIs. All calculated P values were 2-sided, and P   <  .05 was considered statistically significant. Analyses were performed with the SAS statistical software package (SAS Institute Inc, Cary, North Carolina).

For comparison, data regarding patients with a diagnosis of MM of the skin reported in the SEER Program were accessed using SEER*Stat software, version 6.6.2 (National Cancer Institute, Bethesda, Maryland). Patients with a diagnosis of MM of the skin between 1973 and 2006 from the 17 SEER regions were identified. The first primary diagnosis of MM was used, and patients were excluded if the report was obtained solely from a death certificate or from a report at autopsy or if no microscopic confirmation of diagnosis was available. To obtain complete data regarding survival time, patients were excluded if active follow-up continued but survival time was not available (ie, if the follow-up date was the same as the diagnosis date and no further follow-up data had been obtained). Patients were followed up for vital status until December 31, 2006. Data regarding the thickness of melanomas was available for SEER cases after 1988 and was recorded as the depth in millimeters of the lesion. For comparative purposes, the thicknesses were categorized into 5 groups (0-0.75 mm, 0.76-1.50 mm, 1.51-3.00 mm, >3.00 mm, and unknown). Tumor extension was categorized as Clark level I, II, III, IV, V, or unknown.

Overall and MM cause-specific expected survival rates were estimated by applying the age-, sex-, and calendar year –specific mortality rates from the SEER population to the age- and sex-specific person-years of follow-up in the study cohort.²⁵ For each Breslow thickness category and Clark level, the expected survival rates were derived from SEER patients with MM in the same category. A 1-sample log-rank test was used to determine whether the observed survival of the study cohort differed from the expected survival. Approval for this retrospective review and analysis of combined data from national databases was obtained from the institutional review boards of Mayo Clinic, the University of Cincinnati, and the Organ Procurement and Transplantation Network.

Malignant melanoma was diagnosed in 61 cases among 59 transplantation patients before receiving a solid organ transplant and in 724 cases among 638 transplantation patients after receiving a solid organ transplant (Table 1). Breslow thickness was available for 123 patients and Clark level for 175 in those with MM after transplantation. Among these 638 transplantation patients, 570 had 1 MM, 54 had 2 MMs, 12 had 3 MMs, and 2 had 4 MMs. All MM cases reported from UNOS were diagnosed after transplantation owing to the nature of this database. In the cases collected from the IPITTR, the response rate to the request to access medical records and pathology reports was 42%. This modest response rate most likely resulted from most of the cases occurring in patients who were aged 30 to 40 years and our difficulty locating those patients retrospectively. Table 1 summarizes sex, age at diagnosis of first MM, age at first transplantation, and type of first transplantation, according to the timing of the MM and whether the Breslow thickness was available. Among the patients with MM diagnosed before receiving a transplantation, the patients with Breslow thickness available were more likely to be older at the time of the first MM diagnosis (P    =  .03) and tended to have a shorter interval between MM and transplantation (P  =  .05) compared with patients without Breslow thickness. No statistically significant differences were observed between patients with and without Breslow thickness among the subset with MM diagnosed after receiving a transplant. Of all the 697 patients with MM, 57.1% had a kidney transplant, 25.7% had a heart transplant, and 9.9% had a liver transplant.

Table 2 summarizes data regarding Breslow thickness, Clark level, AJCC pathologic stage, and ulceration for MMs diagnosed before and after transplantation from all individual data sources and combined sources. Breslow thickness was available for 17 MM cases (27.9%) diagnosed before transplantation and 133 MM cases (18.4%) diagnosed after transplantation. Cases of MM diagnosed before transplantation, compared with cases diagnosed after transplantation, were more likely to have more shallow Breslow thickness ( ≤1.50 mm: 88.2% vs 60.2%, P  = .02). Cases of MM diagnosed after transplantation were fairly evenly distributed among the Breslow thickness and Clark level categories. However, the greatest number of cases of MM diagnosed after transplantation were categorized as Clark level IV disease. The AJCC pathologic stage was known for 80 cases.

Analysis of patient survival was focused on the 638 patients with MM after transplantation. No significant difference was observed in overall survival and MM cause-specific survival between the group of patients with (n  =  123) and those without (n  =  515) known Breslow thickness (Figure 1). Among the patients with known Breslow thickness, 39 deaths occurred (median time to death, 1.7 years), of which 17 were due to MM (median time to death, 1.4 years). Among the remaining 84 patients with known Breslow thickness who were alive at latest follow-up, the median duration of follow-up after the MM diagnosis was 3.1 years. Of the 47 patients with Breslow thickness of 0.75 mm or less (including in situ), 5 deaths (1 due to MM) occurred within 2 years and 1 additional death (not due to MM) occurred within 5 years of the MM diagnosis. Of the 26 patients with Breslow thickness of 0.76 to 1.50 mm, 4 deaths (2 due to MM) occurred within 2 years and 1 additional death (not due to MM) occurred within 5 years. Of the 26 patients with Breslow thickness of 1.51 to 3.00 mm, 7 deaths (3 due to MM) occurred within 2 years and 3 additional deaths (2 due to MM) occurred within 5 years. Of the 24 patients with Breslow thickness greater than 3.00 mm, 9 deaths (6 due to MM) occurred within 2 years and 2 additional deaths (1 due to MM) occurred within 5 years. Overall survival rates at 3 years among patients with MM after transplantation and with Breslow thickness of 0.75 mm or less (including in situ), 0.76 to 1.50 mm, 1.51 to 3.00 mm, and greater than 3.00 mm were 88.2%, 80.8%, 51.2%, and 55.3%, respectively (Figure 2A). The MM cause-specific survival rates among patients with MM after transplantation and with Breslow thickness of 0.75 mm or less (including in situ), 0.76 to 1.50 mm, 1.51 to 3.00 mm, and greater than 3.00 mm, respectively, were 97.8%, 89.4%, 73.2%, and 73.9% at 3 years (Figure 2B). Outcomes according to Clark level are shown in Figure 3.

Table 3 summarizes the 3-year overall and MM cause-specific survival rates for patients with MM after organ transplantation (ie, the study cohort) and the expected survival rates that were estimated by applying the age-, sex-, and calendar year –specific mortality rates from the MM patients in the SEER population to the study cohort. The MM cause-specific survival rate by Breslow thickness for patients in the study cohort was not significantly different from the expected estimates except for Breslow thickness of 1.51 to 3.00 mm (at 3 years, 73.2% vs 91.0% expected; P  =  .002). Also, the MM cause-specific survival rate for patients with Clark level III or IV was significantly different in the study cohort compared with that expected (Clark level III at 3 years, 82.8% vs 97.4% expected; Clark level IV at 3 years, 65.8% vs 92.4% expected; P   <  .001).

The effect of other variables on MM cause-specific survival was evaluated: sex, age, and type of organ transplanted were analyzed for 638 patients with MM diagnosed after transplantation. Sex was not associated with MM cause-specific survival (females vs males: HR, 1.01; 95% CI, 0.63-1.61; P  =  .98), and neither was age at MM diagnosis (HR per 10-year increase in age, 1.09; 0.91-1.31; P  =  .34). Type of transplantation was also not associated with MM cause-specific survival (cardiac vs noncardiac: HR, 0.86; 95% CI, 0.53-1.40; P  =  .54; and cardiac, pancreas, or double organ vs all other types: HR, 0.75; 0.47-1.22; P  =  .25). However, the combined subset of patients who had MM after cardiac transplantation with a Breslow thickness of 1.51 to 3.00 mm (n  =  10) or greater than 3.00 mm (n  =  9) had a significantly more adverse MM cause-specific survival compared with the expected survival derived from SEER (3-year MM cause-specific survival: observed, 45.2%; expected, 78.9%; P  =  .009).

Data regarding recurrences and subsequent metastases were available for patients in the Mayo Clinic and IPITTR cohorts. Among the 59 patients with MM before their first transplantation, 2 metastases and no documented recurrences were reported. Both metastases had occurred in kidney transplant patients; one of the metastases was noted in the lymph nodes 1 month after the MM diagnosis, and the other was noted in the lungs 14.8 years after the MM diagnosis.

Among the 244 patients from the Mayo Clinic and IPITTR cohorts with MM diagnosed after their first transplantation, 10 reported metastases and 4 recurrences have been observed. The 10 metastases occurred at a median of 10 months after the MM diagnosis (range, 2-60 months). The 5-year survival rate free of metastasis was 91.5%. The 4 recurrences occurred at 1, 9, 11, and 26 months, respectively, after the MM diagnosis. Of note, no cases of metastasis due to melanoma from an organ donor were observed in a transplant recipient.

Melanoma is considered an immunologic tumor owing to reports of its regression and its response to immune mechanisms.^1- 5 Therefore, a concern is that patients with induced immunosuppression, as required in solid organ transplantation, would have more adverse outcomes than the general population when affected by this potentially aggressive immune-responsive malignant tumor.

Compared with the expected survival rate derived from MM cases reported in SEER, overall survival rates in our retrospective study were significantly more adverse among patients with a history of transplantation regardless of Breslow thickness or Clark level. Cause-specific survival due to MM in solid organ transplant recipients with melanomas of 1.51 to 3.00 mm in Breslow thickness or Clark level III or IV also was significantly more adverse when compared with the expected survival rate. This finding of patients with thicker melanomas having more adverse outcomes in the setting of transplantation also has been shown in a previous study²³ of 100 melanomas in 95 organ transplant recipients. This finding suggests that immunosuppression plays a role in the aggressiveness of at least thicker melanomas. In particular, a subgroup analysis looking at supposedly more immunosuppressed patients (ie, cardiac transplant recipients) identified a statistically significant difference in MM cause-specific survival for patients with melanomas with a Breslow thickness greater than 1.51 mm.

The 59 patients in this study with MM before transplantation did not seem to demonstrate an increased risk of recurrence (no recurrences were observed during a median follow-up of 10.5 years after the MM diagnosis) or subsequent metastasis. Currently, no evidence exists in the literature suggesting that patients with MM before transplantation have more adverse outcomes. It is important, however, to keep in mind that a selection bias exists, with patients who have thinner MMs being more likely to be chosen as candidates for organ transplantation. With this selection bias in mind, it is difficult to make solid conclusions regarding the behavior of MM in patients who subsequently undergo organ transplantation. Furthermore, among the patients with MM before transplantation, Breslow thickness was available for only 17 MMs (Table 2).

The Skin Care in Organ Transplant Patients Europe (SCOPE) group showed that for posttransplantation cutaneous MM, 5-year survival rate for patients with tumors in stage T1 or T2 ( ≤2 mm) was similar to that of the AJCC control population.²³ This finding was similar to trends seen in the study by Dapprich et al,¹⁹ which contained a subset of the data in the present study. The SCOPE study, however, showed that the prognosis was significantly worse for posttransplantation MM of stage T3 or greater (P   <  .001).²³ Dapprich et al were concerned that the prognosis would be more adverse for patients with MM of stage T3 or more, but the number of patients with more advanced MM was too small to make conclusions. The SCOPE study was limited for a comparison of outcomes, with 8 deaths due to MM among 53 patients with complete AJCC staging. A limited number of sentinel events makes analysis difficult.

This study had numerous limitations. Databases were used to compile the largest number of cases of MM in solid organ transplant patients. Despite attempts to collect as many cases with Breslow thickness and outcome as possible, the number of cases was still too small for statistically meaningful analysis. The data were even more limited for the number of cases with data for AJCC staging, the most powerful prognostic indicator in the general population. In the attempt to collect numerous cases, no set systematic reporting of cases occurred; the use of databases with voluntary reporting from individual institutions may introduce a bias in the type of cases reported. If this were so, however, one would expect that more advanced cases would have been reported because these are often the cases remembered and reported. However, in the present report, most cases are thin MMs, with Breslow thickness of less than 1 mm. Thus, the possibility of report bias seems to be less of a possibility, but it cannot be ruled out.

We could not determine the type or degree of immunosuppression that the patients experienced, but these factors may have a clinically significant effect on outcomes. Follow-up information was lacking for many patients; therefore, outcomes were based on all-cause mortality instead of disease-specific (melanoma-specific) mortality, which would have been more meaningful. Knowing baseline risk factors that predispose individuals to melanoma also would have been helpful within this population, but we were unable to obtain those data with this study design because this information was not collected in the national databases.

Despite its limitations, the present study is the largest report of cases with primary prognostic assessment that uses Breslow thickness to assist in interpretation of outcomes. These data add to a growing understanding of the behavior of MM in the solid organ transplant population. Further work in this area is needed. We hope that with time, more extensive tumor databases (complete with staging information of the melanoma) can be linked with transplant databases (complete with outcomes and cause of death) to better answer the question of outcomes in this special population compared with the general population. Data collected prospectively would be even more meaningful. As the number of successful solid organ transplantations and the longevity of life increase, a growing population of patients will arise who are at risk of chronic complications from long-term immunosuppression. Malignant melanoma is a challenging complication that needs to be managed in the evaluation period before transplantation (if the patient has a history of MM before transplantation) or in the maintenance period after transplantation (if MM develops at that time). The importance of valid outcomes data to address this issue most likely will increase with time.