Risk Factors for Melanoma in Renal Transplant Recipients

Key Points

Question

What are the risk factors for developing melanoma in renal transplant recipients?

Findings

In this cohort study, renal transplant recipients had greater risk of developing melanoma compared with their non–renal transplant counterparts. Risk factors for developing melanoma included older age, male sex, recipient white race, living donors, sirolimus therapy, and cyclosporine therapy.

Meaning

Renal transplant recipients should be monitored for risk factors for developing melanoma.

This cohort study uses data from the United States Renal Data System to investigate risk factors and characteristics of renal transplant recipients who develop melanoma.

Abstract

Importance

Melanoma risk factors and incidence in renal transplant recipients can inform decision making for both patients and clinicians.

Objective

To determine risk factors and characteristics of renal transplant recipients who develop melanoma.

Design, Setting, and Participants

This cohort study of a large national data registry used a cohort of renal transplant recipients from the United States Renal Data System (USRDS) database from the years 2004 through 2012. Differences in baseline characteristics between those who did and did not develop melanoma were examined, and a survival analysis was performed. Patients with renal transplants who received a diagnosis of melanoma according to any inpatient or outpatient claim associated with a billing code for melanoma were included. A history of pretransplant melanoma, previous kidney transplantation, or transplantation after 2012 or before 2004 were exclusion criteria. The data analysis was conducted from 2015 to 2016.

Exposure

Receipt of a renal transplant.

Main Outcomes and Measures

Incidence and risk factors for melanoma.

Results

Of 105 174 patients (64 151 [60.7%] male; mean [SD] age, 49.6 [15.3] years) who received kidney transplants between 2004 and 2012, 488 (0.4%) had a record of melanoma after transplantation. Significant risk factors for developing melanoma vs not developing melanoma included older age among recipients (mean [SD] age, 60.5 [10.2] vs 49.7 [15.3] years; P < .001) and donors (42.6 [15.0] vs 39.2 [15.1] years; P < .001), male sex (71.5% vs 60.7%; P < .001), recipient (96.1% vs 66.5%; P < .001) and donor (92.4% vs 82.9%; P < .001) white race, less than 4 HLA mismatches (44.9% vs 37.1%; P = .001), living donors (44.7% vs 33.7%; P < .001), and sirolimus (22.3% vs 13.2%; P < .001) and cyclosporine (4.9% vs 3.2%; P = .04) therapy. Risk factors significant on survival analysis included older recipient age (hazard ratio [HR] per year, 1.06; 95% CI, 1.05-1.06; P < .001), recipient male sex (HR, 1.53; 95% CI, 1.25-1.88; P < .001), recipient white race, living donors (HR, 1.35; 95% CI, 1.11-1.64; P = .002), and sirolimus (HR, 1.54; 95% CI, 1.22-1.94; P < .001) and cyclosporine (HR, 1.93; 95% CI, 1.24-2.99; P = .004) therapy. The age-standardized relative rate of melanoma in USRDS patients compared with Surveillance, Epidemiology, and End Results patients across all years was 4.9. A Kaplan-Meier estimate of the median time to melanoma among those patients who did develop melanoma was 1.45 years (95% CI, 1.31-1.70 years).

Conclusions and Relevance

Renal transplant recipients had greater risk of developing melanoma than the general population. We believe that the risk factors we identified can guide clinicians in providing adequate care for patients in this vulnerable group.

Introduction

Whereas renal transplantation is considered the only long-term curative treatment for end-stage renal disease, patients must cope with complex lifestyle changes. Adequate graft function necessitates lifelong immunosuppressive treatment, rendering the recipient susceptible to a multitude of disease processes. Immunosuppression increases the risk of various cancers, most commonly cancer of the skin, which leads to substantial morbidity and mortality in this patient cohort. In 1971, Walder et al published the first study to identify the increased risk of skin cancer in a small cohort of Australian renal transplant recipients receiving immunosuppressive therapy. Therefore, annual dermatological assessment is recommended for all renal transplant patients.

The type, intensity, and duration of immunosuppressive therapy contribute to the risk of developing skin cancer, such as melanoma, after transplantation. Patients receive induction therapy and maintenance therapy to minimize graft rejection. Induction therapy includes treatment with a biologic agent such as a lymphocyte-depleting agent or an interleukin-2 receptor antagonist, with the goal of limiting the T-cell response and reducing the possibility of acute rejection. Maintenance immunosuppression provides long-term therapy to prevent acute rejection and preserve graft viability. Commonly used agents include tacrolimus, cyclosporine, mycophenolate mofetil, and azathioprine. Corticosteroids have fallen out favor as a result of their adverse effects.

Immunosuppressive agents can accelerate the development of melanoma in transplant recipients via 2 mechanisms. First, immunosuppressive agents are inherently carcinogenic, and second, chronic immunosuppression inhibits the body’s self-defense against malignant changes. Several prospective studies have highlighted the carcinogenic effect of calcineurin inhibitors, which are widely used in immunosuppressive regimens for transplant patients. An increase in the appearance of melanocytic nevi has been observed in immunosuppressed patients, potentially indicating its role as a risk factor for melanoma.

Solid-organ transplant recipients bear a 3- to 5-fold increased risk for melanoma compared with the general population. A study by Hollenbeak et al in 2005 examined melanoma incidence among 89 786 renal transplant recipients registered in the United States Renal Data System (USRDS) database from 1988 to 1998. They found melanoma to be 3.6 times more common in the USRDS population compared with populations from the Surveillance, Epidemiology, and End Results (SEER) database. In 2006, Le Mire et al reported an 8-fold increased melanoma incidence in a cohort of renal transplant recipients compared with the general population for the region, one of the highest reported incidences in the literature.

Given changes in immunosuppression regimens and the advent of novel immunosuppressive agents in the past decade, a large-scale current report of melanoma in renal transplant recipients is necessary and informative. The present study aims to provide an up-to-date review of melanoma incidence in renal transplant recipients. Using a cohort from the USRDS database for the years 2004 through 2012, we analyze risk factors for development of melanoma in renal transplant recipients to enhance knowledge of this topic.

Methods

Study Population

The USRDS combines data from the United Network for Organ Sharing with Medicare and Medicaid payment data, which represents a preponderance of payment source for patients with end-stage renal disease. USRDS Medicare 5% sample claims data from the years 2004 through 2012 were used to identify a cohort of patients with a diagnosis of melanoma according to any inpatient or outpatient claim associated with an International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM), code for melanoma (172.0-172.9). A history of previous kidney transplantation, transplantation after the year 2012 or before the year 2004, and a history of melanoma prior to transplantation were exclusion criteria. Any patient missing a date of last follow-up in their Medicare claims was excluded to allow for a survival analysis.

This study was exempted from institutional review board approval by University Hospitals Case Medical Center due to its retrospective nature; informed consent was waived.

Statistical Analysis

Descriptive statistics for variables present in Table 1 were calculated, and statistically significant differences across the melanoma vs the nonmelanoma group were determined by 2-sided χ² and t tests. Differences were considered statistically significant if P < .05. The cumulative prevalence of melanoma among kidney transplantation recipients in the USRDS database was calculated for the years 2004 through 2012.

Table 1. Baseline Characteristics of the United States Renal Data System Population.

Characteristic	No Melanoma (n = 105 174)	Melanoma (n = 488)	P Value
Age, mean (SD), y	49.7 (15.3)	60.5 (10.2)	<.001
BMI, mean (SD)	27.68 (5.79)	28.13 (5.39)	.10
Most recent creatinine level, mean (SD), mg/dL	7.77 (3.48)	6.64 (2.99)	<.001
Recipient race, No. (%)a
White	69 405 (66.5)	468 (96.1)	<.001
Native American	1025 (1.0)	3 (0.6)
Asian	5228 (5.0)	5 (1.0)
Black	28 768 (27.5)	11 (2.3)
Hispanic	14 920 (14.2)	27 (5.5)	<.001
Recipient male sex, No. (%)	63 802 (60.7)	349 (71.5)	<.001
≥4 HLA mismatches, No. (%)	66 159 (62.9)	269 (55.1)	<.001
Episode of acute rejection, No. (%)
Yes, treated	2635 (2.5)	9 (1.8)	.57
Yes, not treated	484 (0.5)	3 (0.6)
No rejection	102 047 (97.0)	476 (97.5)
Recipient positive test result, No. (%)
Hepatitis B antigen	2268 (2.3)	7 (1.6)	.39
Hepatitis C	5354 (5.5)	27 (6.1)	.61
Peak panel reactive antibody, mean (SD), %	42.6 (34.4)	36.1 (32.8)	.055
Multiple organ transplantation, No. (%)	2701 (2.6)	16 (3.3)	.40
Donor positive test result, No. (%)
Cytomegalovirus	44 061 (63.5)	164 (61.2)	.47
Hepatitis B antigen	9 (0.0)	0	>.99
Hepatitis C	1603 (4.7)	8 (3.8)	.67
Donor age, mean (SD), y	39.2 (15.1)	42.6 (15.0)	<.001
Donor male sex, No. (%)	41 897 (60.1)	166 (61.5)	.69
Donor race, No. (%)
White	86 764 (82.9)	449 (92.4)	<.001
Native American	533 (0.5)	2 (0.4)
Asian	2623 (2.5)	5 (1.0)
Black	14 718 (14.1)	30 (6.2)
Donor skin cancer, No. (%)	139 (0.2)	0	.96
Donor type, No. (%)
Living	35 483 (33.7)	218 (44.7)	<.001
Not blood relative	7117 (25.8)	46 (28.9)	.42
Cancer spread to lymph nodes, No. (%)	456 (0.4)	53 (10.9)	<.001
Immunosuppressive drug therapy, No. (%)
Cyclosporine	3317 (3.2)	24 (4.9)	.04
Tacrolimus	92 066 (87.5)	411 (84.2)	.03
Sirolimus	13 864 (13.2)	109 (22.3)	<.001

Abbreviation: BMI, body mass index, calculated as weight in kilograms divided by height in meters squared.

SI conversion factor: to convert creatinine to micromoles per liter, multiply by 88.4.

^aHispanic ethnicity was counted as a separate predictor such that recipient race and Hispanic ethnicity are not mutually exclusive.

For the purpose of comparison, cases of melanoma were also identified within SEER data from the National Cancer Institute for the years 2005 through 2012, and demographic proportions of SEER patients with melanoma are presented alongside USRDS demographic data for the same categories. The age-standardized relative rates of developing melanoma were calculated between the USRDS cohort and the SEER cohort.

Survival Analysis

Cox proportional hazards models were calculated, where the event of interest was melanoma diagnosis. Date of last follow-up, as indicated by USRDS Medicare claims, was used to identify right-censorship among the population. Variables were selected by stepwise Akaike information criterion minimization procedures, across both covariates of interest and immunosuppressants for adjustment. Results of the disease-free survival model are presented in Table 2. Kaplan-Meier survival curves were calculated to determine median time of survival and for the purposes of graphical illustration. The median time to diagnosis is presented for those patients who did develop melanoma.

Table 2. Multivariable Disease-Free Cox Proportional Hazards Survival Analysis.

Variable	Hazard (95% CI)	P Value
Age, y
25-49	3.78 (1.37-10.40)	.01
50-74	14.40 (5.33-38.80)	<.001
75-100	16.40 (5.10-52.50)	<.001
Body mass index	1.00 (0.98-1.02)	.91
Recipient race
Native American	0.28 (0.07-1.14)	.08
Asian	0.15 (0.06-0.36)	<.001
Black	0.05 (0.03-0.10)	<.001
Hispanic	0.28 (0.19-0.42)	<.001
Recipient male sex	1.53 (1.25-1.88)	<.001
≥4 HLA mismatches	0.96 (0.80-1.17)	.72
Living donor	1.28 (1.05-1.55)	.01
Immunosuppressive drug therapy
Cyclosporine	1.69 (1.09-2.62)	.02
Tacrolimus	0.94 (0.72-1.24)	.68
Sirolimus	1.46 (1.16-1.84)	.001

All analyses were performed using R Statistical Computing Language, version 3.3.0.

Results

Baseline Characteristics of the USRDS Cohort

Between the years 2004 and 2012, 105 174 people were first-time kidney transplantation recipients who had record of a Medicare claim (Table 1). Of these, 488 (0.46%) patients had record of melanoma. Table 1 presents baseline characteristics of those who developed melanoma and those who did not develop melanoma.

Among patients with melanoma, 349 (71.5%) were male, compared with 63 802 (60.7%) individuals without a melanoma diagnosis (P < .001). Patients with melanoma were, on average, roughly 11 years older than those without melanoma (mean [SD] age, 60.4 [10.2] vs 49.7 [15.3] years; P < .001). Almost all of the patients with melanoma were white (96.1%), with smaller proportions of Native Americans, Asians, blacks, and Hispanics (0.6%, 1.0%, 2.3%, and 5.5%, respectively); among the nonmelanoma population, only 66.5% were white, with the other racial groups making up 1.0%, 5.0%, 27.5%, and 14.2%, respectively, of this group.

When stratified to 4 or more HLA mismatches vs less than 4 HLA mismatches, the proportion of patients with 4 or more HLA mismatches was greater among patients who did not develop melanoma (62.9% vs 55.1%; P = .001). Mean (SD) organ donor age was greater for those who developed melanoma (42.6 [15.0] vs 39.2 [15.1] years; P < .001).

The proportion of white kidney donors was greater for those who developed melanoma (92.4% vs 82.9%; P < .001). In addition, the group of patients who developed melanoma had a greater proportion of living donors than those who did not develop melanoma (44.7% vs 33.7%; P < .001).

Importantly, patients with melanoma were significantly more likely to be taking cyclosporine (4.9% vs 3.2%; P = .04) or sirolimus (22.3% vs 13.2%; P < .001)—2 common immunosuppressants—than those without melanoma.

Melanoma Prevalence

The overall crude prevalence of melanoma in the USRDS cohort was 0.47%, with 488 cases in 105 174 renal transplant recipients. The USRDS population prevalence and SEER population prevalence were standardized to the 2000 US standard population. Furthermore, the overall relative rate of melanoma for USRDS patients compared with SEER patients is 4.9, indicating that USRDS patients were 4.9 times more likely to develop melanoma than their SEER counterparts. The age-adjusted prevalence of melanoma in the kidney transplant population is 208 per 100 000 patients, while the crude prevalence is 462 per 100 000 patients. For the purposes of comparison, one study reports an incidence rate of 21.6 per 100 000 in the non-Hispanic white general American population.

Table 3 compares the demographic characteristics of SEER patients with melanoma with those of USRDS patients with melanoma. The USRDS population had a greater proportion of males than SEER, was similar to SEER in terms of race, and had a narrower age distribution than SEER.

Table 3. Demographic Characteristics of Patients With Melanoma in Surveillance Epidemiology and End Results (SEER) vs Patients With Melanoma in United States Renal Data System (USRDS)^a.

Variable	%
	SEER	USRDS
Sex
Female	41.7	28.5
Male	58.3	71.5
Race
White	94.4	96.1
Black	0.5	1.6
Other	5.1	2.3
Age, y
≤20	0.8	0
21-40	11.6	4.1
41-60	35.1	37.7
61-80	39.4	58.0
81-100	13.1	0.2

^aThis table juxtaposes crude proportions of patients across sex, race, and age quintile for each of the USRDS and SEER populations of patients with melanoma.

Survival Analysis

Table 2 presents the results of a multivariable Cox proportional hazards model of disease-free survival. Age, race, sex, donor type, cyclosporine therapy, and sirolimus therapy had hazard ratios that were statistically significant. Each extra year of age was associated with 1.06 times the hazard (95% CI, 1.05-1.06; P < .001); Asian, African American, and Hispanic transplantation recipients were found to have decreased hazard of melanoma compared with white recipients, with hazard ratios of 0.16 (95% CI, 0.06-0.38; P < .001), 0.06 (95% CI, 0.03-0.11; P < .001), and 0.30 (95% CI, 0.20-0.45; P < .001), respectively. Transplantation recipients with living donors had 1.35 times the hazard of those with deceased donors (95% CI, 1.11-1.64; P = .002). Cyclosporine and sirolimus use for long-term immunosuppression were associated with hazard ratios of 1.93 (95% CI, 1.24-2.99; P = .004) and 1.54 (95% CI, 1.22-1.94; P < .001), respectively. The Figure and eFigures 1 and 2 in the Supplement present Kaplan-Meier curves for renal transplantation recipients according to living vs deceased donors, sirolimus therapy, and cyclosporine therapy, respectively. eFigure 3 in the Supplement presents stratification of the model by age group. A Kaplan-Meier estimate of the median time to development of melanoma among those patients who did develop melanoma was 1.45 years (95% CI, 1.31-1.70 years).

Figure. Kaplan-Meier Fit of Melanoma-Free Survival Between Living and Deceased Donor Renal Transplantation Recipients.

Kaplan-Meier survival curve comparing renal transplant recipients who received a living-donor transplant and those who received a deceased-donor transplant. Patients receiving transplants from cadaveric donors had greater odds of melanoma-free survival. Shading represents 95% confidence intervals calculated based on cumulative hazard.

Discussion

Melanoma and nonmelanoma skin cancers after transplantation are common and are a cause of substantial mortality. Several studies have examined the risk of skin cancer after transplantation. Brewer et al reported 430 skin cancers per 1000 patient-years in a cohort of 312 heart transplant recipients. Modaresi Esfeh et al reported 84.8 skin cancer cases per 1000 person-years in a cohort of 998 liver transplant recipients. More specifically, Garrett et al reported 75 cases of melanoma per 100 000 person-years in a large retrospective study of 10 649 organ transplant recipients from the Organ Procurement and Transplantation Network database. Risk of cutaneous melanoma in solid-organ transplant (SOT) recipients has been reported to be 2- to 4-fold greater than in those who have not received SOT. Data are controversial regarding melanoma stage at the time of diagnosis; however, studies have found that melanoma in kidney transplant recipients had a higher pathologic stage, was more invasive, and conferred a higher risk of death.

Carcinogenesis following SOT is due to a number of factors: immunosuppressive therapy, UV exposure, genetic factors, viral infection, and psychogenic factors such as stress. Immunosuppression has been shown to be a prominent risk factor for melanoma after transplantation because it impairs immune surveillance and has direct oncogenic activity. We found that sirolimus, an immunosuppressant well known for its anti-oncogenic effects, was associated with a significantly increased risk of melanoma in the USRDS population (eFigure 1 in the Supplement). Cyclosporin, another immunosuppressive agent, was also associated with an increased risk of melanoma (eFigure 2 in the Supplement). Other risk factors related to posttransplant skin cancer include skin cancer prior to transplant, living vs deceased donor, history of malignant neoplasm prior to transplant, and age, sex, ethnicity, geographic location, and smoking. Melanoma-specific risk factors include older age at transplantation, skin type, sun exposure burden, and duration of immunosuppression.

Our analysis corroborates results from current literature. Increasing age, male sex, and white race were all associated with significantly increased risk of melanoma and have been demonstrated in past studies. A prior analysis of the USRDS data from 1988 to 1998 by Hollenbeak et al also found significant relationships between increased melanoma risk and male sex, increased age, and at least 1 episode of acute rejection. Age is a well-known risk factor, although specific age cutoffs at which cancer risk becomes significantly increased have yet to be elucidated. In our USRDS cohort, patients aged 75 years and older had the highest rates of melanoma (eFigure 3 in the Supplement). A single-center study of development of malignant neoplasm after renal transplantation identified age older than 45 years as the only risk factor associated with higher frequency of posttransplant malignant neoplasms. One study found a 6% increased relative rate for each 1-year increase in age.

Of note, patients with living donors had significantly higher incidence of melanoma compared with those who did not. Recipients of organs from living donors had 1.28 greater hazard of developing melanoma compared with deceased donor recipients (P = .01). This is discordant with prior studies, with patients receiving SOTs from deceased donors experiencing greater incidence of cancer and mortality. It is believed that greater doses of immunosuppressive agents in patients with SOTs from deceased donors result in greater cancer susceptibility. However, it is possible that patients with SOTs from living donors achieve a greater life expectancy, and older age increases melanoma risk.

The advent of novel immunosuppressive therapy in the past decade may have altered patterns of skin cancer development in the renal transplant population. Direct oncogenic effects of immunosuppressive agents are widely recognized in skin cancer development; cyclosporine and azathioprine are thought to have higher oncogenic potential. On the other hand, mammalian target of rapamycin (mTOR) inhibitors have been shown to protect against skin carcinogenesis. Interestingly, in our population, mTOR inhibitor therapy was associated with increased risk of melanoma. Optimal management of immunosuppression regimens to reduce the incidence of skin cancer risk after transplantation is currently being explored.

In our cohort, we also found that the median time to developing melanoma among renal transplant recipients was 1.45 years. Prior studies report a mean interval to melanoma diagnosis among SOT recipients of 5 years. Given our findings of a shorter duration to developing melanoma, adequate primary and secondary prevention of melanoma in renal transplant recipients is crucial. Patient counseling regarding melanoma risk and adequate skin checks are necessary. Intensive screening for skin cancer in renal transplant recipients has been recommended. Health care professionals should be aware of the necessity for regular skin screenings and early treatment of any suspicious pigmented lesions. Clinical practice guidelines from Kidney Disease Improving Global Outcomes include educating renal transplant recipients regarding the risk of skin and lip cancer, advising minimal sun exposure, use of UV-light blocking agents, self-examination of the skin and lips, and annual skin and lip examination by a dermatologist or experienced physician. Prior studies have shown that SOT recipients do not practice adequate sun protection behavior, with one-third not using sunscreen when in the sun and two-thirds not wearing protective clothing.

Limitations

This study is not without limitations. We used previously established methods of diagnosing melanoma from billing records. If a patient was not assigned the correct ICD-9-CM code, if his or her treatment was not covered by Medicare, and if melanoma was never detected and treated, then that patient would not have been included in the cohort. Consequently, it is possible that melanoma incidence was underreported. Furthermore, we used SEER as a comparison group; it is possible that the reported melanoma incidence in the SEER comparison group either underestimates or overestimates the true rate. Last, we could not assess certain risk factors for melanoma because they were not captured in the USRDS database. Information regarding lifetime sun exposure, Fitzpatrick skin type, stage of melanoma, and presence of melanocytic nevi was unavailable.

Longer life expectancy after renal transplantation permits greater duration of exposure to immunosuppressive therapy and increased age of organ recipients. Thus, clinicians should be aware of increased incidence and subsequent mortality from melanoma in this patient population.

Conclusions

Renal transplant recipients are at greater risk of developing melanoma than the general population. We provide detailed information and identified several important trends and risk factors in this vulnerable patient group that we believe can guide clinicians in providing adequate care.

Supplement.

eFigure 1. Kaplan Meier fit of Melanoma-Free Survival between Renal Transplantation Patients on Sirolimus and those not on Sirolimus

eFigure 2. Kaplan Meier fit of Melanoma-Free Survival between Renal Transplantation Patients on Cyclosporin and those not on Cyclosporin

eFigure 3. Kaplan Meier fit of Melanoma-Free Survival Stratified by Age