SEARCH

SEARCH BY CITATION

Keywords:

  • renal cell carcinoma;
  • melanoma;
  • genetic predisposition;
  • multiple primary neoplasms;
  • familial melanoma

Abstract

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. CONFLICT OF INTEREST DISCLOSURES
  8. REFERENCES

BACKGROUND:

Patients with melanoma (MM) have an increased risk of kidney cancer, and there is an excess risk of MM among patients with renal cell carcinoma (RCC). The objective of the current study was to analyze a series of 42 patients with both MM and RCC to identify clinical and pathologic features as well as risk factors of this association.

METHODS:

Clinical and pathologic characteristics of 42 patients who developed both MM and RCC (the MM + RCC series) were compared with 2 published series in each cancer alone: a series of 293 patients with MM (MM series) and a series of 1527 patients with RCC (RCC series).

RESULTS:

RCC was diagnosed concomitantly or after MM in 83% of patients in the MM + RCC series. Those patients displayed a high proportion of asymptomatic RCC at diagnosis (70%) and a higher frequency of stage I tumors (61%) than patients in the RCC series. Compared with the MM series, patients in the MM + RCC series more often were men, had a higher frequency of blond/red hair, had poor tanning ability, and had a higher number of nevi. In addition, patients in the MM + RCC series had a high aggregation of other malignancies (mainly skin cancers) and a significantly higher frequency of family history of MM (P = .005). Only 2 cyclin-dependent kinase 2A gene (CDKN2A) germline mutations were identified among patients in the MM + RCC series who also were members of MM-prone families.

CONCLUSIONS:

The high aggregation of cancers among patients in the MM + RCC series and the familial clustering of MM argued for a genetic predisposition that may be partly independent of CDKN2A. Cancer 2010. © 2010 American Cancer Society.

There has been a dramatic rise in incidence of melanoma (MM) over the last decades. In France, the annual age-standardized MM incidence rates increased from 2.4 to 7.6 per 100, 000 person-years in men and from 3.9 to 9.5 per 100,000 person-years in women between 1980 and 2000.1 The median age at diagnosis of MM has increased from 45 years to 55 years over the last 3 decades in France. Cutaneous MM is a complex disease that results from genetic and other risk factors. Epidemiologic studies have demonstrated that exposure to sunlight is the major environmental risk factor, whereas high numbers of melanocytic nevi (banal nevi and/or atypical nevi), hair color, eye color, skin color, extent of freckling, and skin reactions to sun exposure are the major host factors.2 Approximately 10% of MMs occur in a familial setting. It has been demonstrated that rare germline mutations in 2 genes, the tumor suppressor gene cyclin-dependent kinase 2A (CDKN2A) and the oncogene cyclin-dependent kinase 4 (CDK4), confer a high risk of MM and segregate in MM-prone families.3 Common genetic variants that confer a low-to-moderate risk of MM also play a role and have been characterized in candidate gene studies4 and, more recently, in genome-wide association studies.5, 6

Like MM, the incidence of renal cell carcinoma (RCC) has increased in the last decades and is approximately twice as high in men as it is in women. In France, the age-standardized incidence rate (per 100, 000 person-years) has risen from 7.3 to 11.4 in men and from 3 to 4.5 in women between 1980 and 2005.7 RCC accounts for 85% of renal cancers and usually is discovered between ages 60 years and 70 years. RCCs arise from the renal epithelium, and the main histologic types are the clear cell type (85% of cases) and the papillary type (10% of cases). A few RCC risk factors have been identified consistently, including smoking, obesity, and hypertension.8 Although most RCC cases are sporadic, evidence for an inherited predisposition also has been reported.9, 10 Approximately 3% of RCCs are related to heritable syndromes, notably the von Hippel-Lindau syndrome, which is caused by mutations in the von Hippel-Lindau (VHL) suppressor gene.11

The development of multiple cancers in the same patient has been widely documented. It is well established that patients with cutaneous MM are at higher risk than individuals from the general population for developing a second primary tumor, including a second primary MM, nonmelanoma skin cancers, and noncutaneous cancers.12-14 Second primary malignancies of various types also are associated with RCC.15-17 A significant excess of MM after kidney cancer and an excess of kidney cancer after MM have been reported.14, 16-21 Although the clinical characteristics and risk factors of both of these cancers have been investigated separately, the characteristics of patients who present with both MM and RCC remain unknown.

The objective of the current study was to determine the clinical and pathologic characteristics of the coexistence of RCC and MM by comparing a series of 42 patients who developed both RCC and MM with 2 series of patients who were affected either by primary MM or RCC. Our goal was also to make comparisons of MM risk factors that were collected in both an MM + RCC series and an MM series.

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. CONFLICT OF INTEREST DISCLOSURES
  8. REFERENCES

Recruitment of Patients and Data Collection

A series of 61 consecutive patients who were diagnosed with both MM and RCC was recruited over of a 34-year period (from 1975 to 2009). These patients were referred for MM care to M.-F. Avril's dermatology clinic at the Gustave Roussy Institute (Villejuif) from 1975 through 2005 and then at Cochin Hospital (Paris) from 2005 to 2009. Most patients were referred by private dermatologists, and 6 patients were treated at other French clinical centers (Civil Hospices, Lyon; St. Louis Hospital, Paris Public Assistance Hospitals, Paris; Foch Hospital, Suresnes; and C. Nicolle Hospital, Rouen). Patients with MM and renal cancer represented 0.49% (39 of 7867) of all patients with MM and 0.93% (39 of 4187) of all patients with renal cancer who attended the Gustave Roussy Institute from 1975 to 2005.

Diagnoses of MM and renal cancer were confirmed pathologically in 53 patients. We analyzed only the 42 patients with renal cancer who had confirmed RCC. Among these patients with both MM and RCC (the MM + RCC series), 35 patients had clear cell carcinoma, 6 patients had papillary carcinoma, and 1 patient had chromophobe carcinoma according to the Heidelberg classification.22

The data collected for each patient included date of birth, sex, date of diagnosis for each cancer, treatment, histologic type, pathologic stage, and (for patients who died) age at death and cause of death. Pathologic stage was assigned according to the 2009 International Union Against Cancer TNM classification of renal tumors23 and according to the 2009 American Joint Committee on Cancer (AJCC) for MM.24 Additional variables included Breslow thickness and body site for MM and modes of presentation for RCC. The initial RCC clinical presentation was defined as asymptomatic for tumors that were discovered at any radiologic examination and symptomatic for tumors that were discovered by local or systemic symptoms. In addition, information about other primary tumors was retrieved from medical records and/or pathology reports for all 42 patients who had both confirmed MM and RCC.

The majority of patients in the MM + RCC series (78%) agreed to complete a standardized questionnaire that recorded MM risk factors that was used in our previous MM studies.25, 26 These factors included skin color (pale or dark), eye color (pale or dark), hair color (red, blond, light brown, dark brown, or dark), the presence or absence of freckles, the total number of moles (<10, 10-50, or >50 nevi), and the degree of exposure to sunlight (low, medium, or high) during holidays (intermittent sun exposure). The skin reaction to ultraviolet light was evaluated by sensitivity to sunburn (low, medium, or high) and tanning ability (low, medium, or high). Written informed consent was obtained from all patients before participation under an Institutional Review Board-approved protocol (Committee for the Protection of Patients, Necker Hospital, Paris).

Clinical and pathologic characteristics of the 42 patients in the MM + RCC series were compared with 2 hospital-based series that were ascertained systematically: a series of 293 French patients with cutaneous MM (the MM series) who were recruited by us between 1986 and 198926 and a published series of 1527 French patients with RCC who were diagnosed between 1982 and 2003 (the RCC series).27 The variables that were available for these 2 series were mean age at diagnosis, sex, stage at diagnosis, location and histologic type of the primary tumor, survival status, and (for MM) Breslow thickness. We compared the frequency of other tumors that occurred among patients in the MM + RCC series at least 3 months after both diagnoses of MM and RCC with the frequency of tumors that occurred after the first MM among patients in the MM series. The same type of comparison was made for tumors that occurred after a diagnosis of renal cancer using data from the literature, because such tumors were not documented in the French RCC series. A family history of either MM or RCC also was compared between our MM + RCC series and either the MM series or published data for RCC.9, 10 Finally, the distribution of MM risk factors (pigmentary traits, nevus phenotypes, and sun-related covariates) was assessed in the samples from patients in the MM + RCC series and the MM series.

CDKN2A and CDK4 Mutation Testing

Although the coexistence of MM and RCC in the same patients is not an indication for CDKN2A or CDK4 testing, we ran these tests because we observed that a high proportion of these patients had a family history of MM, as indicated below (see Results). Among the 42 patients in the MM + RCC series, 34 patients had their blood drawn, and 28 of those blood samples were tested for CDKN2A and CDK4 mutations. We screened for point mutations in CDKN2A (exons 1α, 2, and 3), adenosine diphosphate-ribosylation factor (ARF) (exon 1β), and CDK4 (exon 2) using denaturing high-performance liquid chromatography or direct sequencing as described previously.28, 29 Genomic deletion screening of the CDKN2A locus also was carried out as described previously.30

Statistical Analysis

Clinical and pathologic features were compared between the MM + RCC series and either the MM series or the RCC series using chi-square or Fisher exact tests for qualitative variables and Student t tests or nonparametric Wilcoxon-Mann-Whitney tests for quantitative variables. Similar tests were used to compare MM risk factors (pigmentary traits, nevus phenotypes, and sun-related covariates) between the MM + RCC series and the MM series. All computations were done using SAS software (version 9.1; SAS Institute, Inc., Cary, NC).

RESULTS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. CONFLICT OF INTEREST DISCLOSURES
  8. REFERENCES

The MM + RCC series of 42 patients included 23 men (55%). The mean time between the diagnosis of MM and the diagnosis of RCC was 8.3 years (range, 1-20 years) when MM was diagnosed first and 10.3 years (range, from 5 months to 19 years) when RCC was diagnosed first (P = .50). The mean age at diagnosis of first cancer, either MM or RCC, did not differ significantly (55.8 years for MM vs 58.5 years for RCC; P = .32).

The RCC diagnosis was subsequent to the first primary MM diagnosis in 22 patients (52%), whereas it was synchronous to the MM diagnosis in 13 patients (31%) and occurred before the first MM in 7 patients (17%). RCC was either was discovered because of clinical symptoms and confirmed by imaging or, in the absence of clinical symptoms, was discovered on imaging studies that were obtained in the context of a workup. Overall, 70% of RCCs were diagnosed on imaging studies without previous clinical symptoms. For these patients, imaging evaluation was performed at the initial staging of MM (37.5%), during a surveillance evaluation for MM (54.2%), or during an evaluation of other disease (8.3%). Among all 42 patients in the MM + RCC series, 17 patients had other primary tumors, including 5 patients (12%) with from 1 to 4 other primary MMs, 5 patients (11.9%) with basal cell carcinomas, and 9 patients (21.4%) with various types of noncutaneous cancers, as presented in Table 1. These tumors were subsequent to primary MM and RCC tumors in 11 patients (Table 1).

Table 1. Other Associated Cancers in 42 Patients With Melanoma Plus Renal Cell Carcinoma
PatientSexType of MMNo. of MMs (No. of Subsequent to MM+RCC)Other Cancer(s)
  • MM indicates melanoma; RCC, renal cell carcinoma.

  • a

    These tumors were diagnosed subsequent to both melanoma and renal cell carcinoma.

1WomanCutaneous Meningiomaa
2ManCutaneous B-cell follicular lymphomaa
3WomanOcular Uterine cervical carcinoma
4WomanCutaneous3 (1) 
5WomanCutaneous Meningioma
6WomanCutaneous Contralateral urothelial carcinoma
7WomanCutaneous5 (2) 
8ManCutaneous Prostate adenocarcinoma
9WomanCutaneous2 (1) 
10ManCutaneous Chronic lymphocytic leukemia; colorectal carcinomaa
11ManCutaneous2 (1)Colorectal carcinomaa
12ManCutaneous Bladder carcinoma
13ManCutaneous2 (1)Basal cell carcinomaa
14WomanCutaneous Two basal cell carcinomasa
15WomanCutaneous Basal cell carcinoma
16ManCutaneous Basal cell carcinomaa

The clinical and histologic characteristics of the 40 patients with cutaneous MM in our MM + RCC series and of the 293 patients with cutaneous MM in our systematically ascertained MM series26 are shown in Table 2. For that comparison, we excluded 2 patients in the MM + RCC series who had ocular MM. The mean age at MM diagnosis was significantly older in the MM + RCC series than in the MM series (aged 56.9 years vs 45.6 years; P < .001). Patients in the MM + RCC series were men significantly more often than patients in the MM series (57.5% vs 33.8%; P = .004). The distribution of body sites of primary MM, the frequency of superficial spreading MM, Breslow thickness, AJCC stage at diagnosis, and the frequency of death related to MM did not differ significantly between the 2 series (P > .9). Patients in the MM + RCC series developed other tumors after both MM and RCC at a significantly higher rate than patients in the MM series after their first MM (27.5% vs 15%; P = .05). The frequency of other primary MMs, nonmelanoma skin cancers, and noncutaneous cancers also was higher in the MM + RCC series than in the MM series, although the frequency was only marginally significant for primary MM (P = .08) (Table 2). No patient developed renal cancer subsequent to either MM + RCC or MM alone. A family history of MM was significantly more frequent in the MM + RCC series than in the MM series (22.5% vs 7.5%; P = .005) (Table 2). Most patients in the MM + RCC series who had a family history of MM (5 of 9 patients) belonged to families that had at least 3 MM cases.

Table 2. Characteristics of Melanoma in Patients With Melanoma Plus Renal Cell Carcinoma and in Melanoma Series
Clinical and Pathologic FeaturesMM+RCCMM SeriesaP
  • MM indicates melanoma; RCC, renal cell carcinoma; AJCC, American Joint Committee on Cancer; SD, standard deviation.

  • a

    The MM series was recruited by the current authors (see Grange et al26).

No. of patients40293 
Age at melanoma diagnosis: Mean±SD, y56.9±13.345.6±14.1<.001
Men, %57.533.8.004
Histologic type: Superficial spreading melanoma, %64.460.5.4
2001 AJCC stage, %   
 06.71.7 
 I53.348.6 
 II35.642.1.3
 III4.45.5 
 IV02.1 
Breslow thickness: Mean±SD1.7±1.31.9±1.6.3
Primary melanoma site, %   
Head and neck14.69.4 
Trunk37.527.4.1
Extremities47.963.2 
Death related to melanoma, %17.516.7.9
Other cancers and family history   
 Other cancers after MM and RCC, %   
  All cancers27.515.05
  Melanoma12.55.3.08
  Renal cancer00 
  Nonmelanoma skin cancers105.8.2
  Noncutaneous cancers105.5.2
 Family history of melanoma, %22.57.5.005

Regarding RCC, no significant difference was observed between the MM + RCC series and the RCC series27 with respect to sex or age at diagnosis (Table 3). The overall distribution of TNM stage did not differ significantly between the MM + RCC series and the RCC series. However, the frequency of TNM stage I was approximately 1.5-fold (significantly) higher in the MM + RCC series than in the RCC series (61.1% vs 40.5%; P = .02). A high proportion of patients in the MM + RCC series developed other tumors after the diagnosis of both MM and RCC: 27.5% of these patients developed other tumors of any type, 12.5% developed primary MM, 10% developed nonmelanoma skin cancers, and 10% developed noncutaneous cancers, whereas none developed a subsequent renal cancer. Compared with published data on other tumors after the diagnosis of renal cancer, these proportions were higher for all other tumors, for MM, and for nonmelanoma skin cancers, but the proportions were of the same order of magnitude for noncutaneous cancers and second renal cancers (Table 3). Familial clustering of RCC was identified in 2 of 40 patients (5%), which was similar to the frequency reported in published RCC series (Table 3).

Table 3. Characteristics of Renal Cell Carcinoma in Patients With Melanoma Plus Renal Cell Carcinoma and in Renal Cell Carcinoma Series
Clinical and Pathologic FeaturesMM+RCCRCC SeriesaP
  • MM indicates melanoma; RCC, renal cell carcinoma; SD, standard deviation.

  • a

    Clinical and pathologic characteristics of patients who had MM+RCC were compared with data from a French RCC series that was published by 1 of the current authors (see Patard et al.27).

  • b

    The frequency of asymptomatic initial presentation and of death related to RCC was reported in a subset of patients from the RCC series (see Patard et al27, 31).

  • c

    The presence of cancers after MM+RCC and a family history of renal cancer were compared with data from other published RCC series, because these variables were not recorded in the French RCC data. For these comparisons, the table provides the ranges of published frequency estimates rather than P values (see Gago-Dominguez et al,9 Clague et al,10 Rabbani et al,15 Beisland et al,16 Wilson et al,17 and Czene and Hemminki32).

No. of patients401527 
Age at RCC diagnosis: Mean±SD, y59.4±11.0461.9±12.8.2
Men, %57.565.3
Clear-cell RCC, %78.688.7.1
TNM stage I, %61.140.5.02
Asymptomatic initial presentation, %70.641.2b<.001
Death related to RCC, %1526.5b.1
Other Cancers and Family HistoryMM+RCC, %RCC Series: Range, %cP
Other cancers after MM and RCC   
 All cancers27.55.2-9.7
 Melanoma12.50.1-0.7
 Renal cancer00-0.7
 Nonmelanoma skin cancers100.1-0.4
 Noncutaneous cancer105.1-9.4
Family history of renal cancer53.8-6.5

The distribution of pigmentary traits, nevus phenotypes, and sun-related covariates in the MM + RCC series and the MM series is shown in Table 4. Compared with patients in the MM series, patients in the MM + RCC series had a 1.7-fold increased frequency of blond/red hair (P = .03), a 2-fold increased frequency of poor tanning ability (P = .04), and a 1.2-fold higher frequency of the propensity to sunburn (P = .02). The frequency of patients in the MM + RCC series with more numerous nevi (>50 nevi) was 1.7-fold higher than the frequency observed for patients in the MM series, but the difference was only marginally significant (P = .06). The distribution of other MM risk factors did not differ between the 2 series.

Table 4. Frequency Distribution of Melanoma Risk Factors (%) in Patients With Melanoma Plus Renal Cell Carcinoma and in Patients With Melanoma
Melanoma Risk FactorsMM+RCC (N=32), %aMM (N=293), %bP
  • MM indicates melanoma; RCC, renal cell carcinoma.

  • a

    The number of patients with MM+RCC who had information available on MM risk factors.

  • b

    The number of patients with MM who had information available on MM risk factors.

Pale skin color8079.2.92
Pale eye color6361.8.90
Blond/red hair color43.825.3.02
Freckles42.330.4.20
Poor tanning ability27.613.3.04
Propensity to sunburn96.880.8.02
High number of nevi (>50)38.723.2.06
Sun exposure during holidays75.974.7.89

Among the 28 patients in the MM + RCC series who were tested for CDKN2A and CDK4 mutations, only 2 patients (7%) who belonged to families with at least 3 MM cases were CDKN2A mutation carriers. These mutations, R.146U[RIGHTWARDS ARROW]G and R.159G[RIGHTWARDS ARROW]C, were reported previously in familial MM.33, 34 One patient who had 2 primary MMs carried 2 CDKN2A variants of unknown significance (R.−25C[RIGHTWARDS ARROW]U + C.−42U[RIGHTWARDS ARROW]A). Another patient who had a family history of MM without CDKN2A mutation had Cowden disease (a rare familial cancer syndrome) and was carrier of a germline phosphatase and tensin homologue gene (PTEN) mutation (L320X).35

DISCUSSION

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. CONFLICT OF INTEREST DISCLOSURES
  8. REFERENCES

The development of both MM and RCC in the same patient is a rare event. In the current study, patients who had both RCC and MM represented 0.5% of the patients with cutaneous MM and 1% of the patients with RCC, and these frequencies were similar to those reported in other studies to date.14-21 We identified the clinical and pathologic features of the association between these 2 cancers: Patients who had both MM and RCC more frequently were men, similar to patients with RCC alone; their mean age at MM diagnosis was older than that in patients who had MM only; and their mean age at RCC diagnosis was similar to that observed in patients who had RCC only. Most RCCs were asymptomatic and were diagnosed after MM on imaging studies. Patients in the MM + RCC series had a higher frequency of other tumors (mainly skin cancers) compared with patients in the MM series and the RCC series after their first cancer. They also more often had a family history of MM, whereas this was not the case in the RCC series. Patients in the MM + RCC series more frequently had phenotypic risk factors for MM compared with patients in the systematically ascertained MM series. These host factors included red/blond hair, poor tanning ability, high propensity to sunburn, and high numbers of nevi.

The significantly higher frequency of men in the MM + RCC series compared with the MM series cannot be explained by differences in the period of recruitment for these patients. Indeed, we were able to verify that the frequency of men in the MM + RCC series always remained higher than that estimated from French MM cancer registries over the last 3 decades.7 The excess of men in both the MM + RCC series and the RCC series is in agreement with other studies of renal cancer.27, 36 In France, the frequency of renal cancer among men has been stable since 1980 and has been estimated as between 64% and 67% in RCC studies.27, 37

The older age at MM diagnosis in our MM + RCC series compared with our MM series may be explained in part by the later occurrence of RCC compared with MM in the general population. It also may have been caused in part by an increase in the mean age at MM diagnosis over the past decades and the by earlier period of recruitment for the MM series than the MM + RCC series. A variation in age at MM diagnosis over time also has been reported in other populations.38, 39

Regarding the clinical and pathologic features of the tumors, the 70% rate of RCCs that were asymptomatic at initial presentation in our MM + RCC series was much higher than what was reported in other RCC studies, in which the rate of asymptomatic RCCs varied from 8% to 55% over the last 2 decades.37, 40 RCCs were mainly of the clear cell type in patients with MM + RCC, like in patients with RCC alone, and they were diagnosed at an earlier TNM stage than in the current and other published RCC studies.27, 37 This is likely because of the high frequency of patients with MM + RCC (≥80%) who were diagnosed with RCC concomitantly or subsequently to MM. The contribution of a thorough initial evaluation of patients with MM and increased medical surveillance to an early diagnosis of RCC is in agreement with previous reports.21 With respect to MM, the predominant histologic type, as usually observed, was superficial spreading MM. We did not observe any significant difference in terms of AJCC stage or Breslow thickness between the MM + RCC series and the MM series. The distribution of body sites of primary MM did not differ significantly between these 2 series, although patients in the MM + RCC series had a higher frequency of MM in the trunk and a lower frequency in the extremities, in agreement with the reported over representation of men and the known preferential location of MM according to sex.14

Patients in our MM + RCC series developed more tumors subsequent to both MM and RCC diagnoses than patients from our MM series after a diagnosis of MM. This remained true when comparisons were made with other published MM series.12-14, 18, 19 This excess of other tumors appears to be caused mainly by an increased frequency of MMs and nonmelanoma skin cancers. Indeed, the rates of other MMs and nonmelanoma skin cancers in our MM series26 corresponded to the upper limits of the frequency of these cancers in published MM series12-14, 18, 19 and were more than 2 times lower than the rates observed in the MM + RCC series. An excess of skin cancers, as mentioned above (see Results), also was observed in the MM + RCC series with respect to other published RCC series.15, 32 Conversely, there was no excess of noncutaneous cancers in the MM + RCC series, because the frequency of these tumors was at the upper limit of published rates of noncutaneous cancers after MN12-14, 18, 19 or renal cancer.15-17, 32

Patients in our MM + RCC series had a higher degree of familial MM clustering than patients in our MM series26 and in other published series,41 whereas there was no excess of a family history of RCC.9, 10 Thus, both the high frequency of other tumors, especially MM, and the familial clustering of MM in patients with MM + RCC strongly suggest the involvement of genetic factors underlying the association of these 2 cancers, some of which also may be involved in MM alone. However, the role of environmental factors cannot be ruled out.

Patients in our MM + RCC series who displayed the phenotypic risk factors of MM (pigmentation and nevus phenotypes) at a higher frequency than patients in our MM series, whereas intermittent sun exposure was of similar magnitude in the both series. This may be explained in part by the higher frequency of a family history of MM in the MM + RCC series compared with the MM series, because it is known that MM-associated phenotypes cluster in families and that genetic variants influence both host factors and MM.5 However, we observed that the distribution of these risk factors did not differ significantly between patients in the MM + RCC series with and without a family history (results not shown).

The screening of patients in the MM + RCC series for mutations in the CDKN2A and CDK4 genes revealed that these genes accounted for a minor part of the genetic predisposition toward the association of these 2 cancers. Our main motivation for testing CDKN2A and CDK4 mutations, as mentioned above, was the high degree of familial clustering of MM among patients in the MM + RCC series. In addition, because the coexistence of these 2 cancers is a rare event, the underlying genetic factors are likely to be rare, and we believed that it was important to ensure that the rare genetic determinants of MM known to date did not account solely for the association of these 2 cancers. Because we observed an association between MM + RCC and pigmentation characteristics, pigmentation genes also may play a role in the coexistence of these 2 cancers. Recent genome-wide association studies have reported several pigmentation genes that are associated with both pigmentation phenotypes and MM, including the melanocortin 1 receptor (MC1R) and tyrosinase (TYR) gene regions6 among others.42 However, most of these associated genetic variants are relatively common; thus, a more thorough investigation of these genes, including resequencing, will be required to identify potentially rare variants that influence MM + RCC.

It is unlikely that our current results were caused by a bias in recruiting the patients in our MM + RCC series, because the frequency of the association of these 2 cancers among all patients with MM and RCC ascertained during the same period of time was similar to that reported in published registries.14, 16-18, 21 We are aware that 1 major limitation of our study is the small size of our MM + RCC series, which was mainly because of the rare occurrence of these 2 cancers in the same patient. Another limitation is the lack of information about known RCC risk factors (smoking, obesity, hypertension) in both the MM + RCC series and the RCC series, which precluded their investigation. There also was no information on pigmentation phenotypes in the RCC series; however, to our knowledge, no association of RCC with these phenotypes has been reported to date. Finally, we did not correct for multiple testing, because some studied factors were correlated, and the use of a Bonferroni correction would have been too conservative unless we applied it to independent tests that were difficult to assess. Therefore, further studies will be needed to confirm the current findings.

In conclusion, the most prominent features of the association between MM and RCC that were pointed out in the current study were a high aggregation of cancers (mainly skin cancers) in the same patient and a familial clustering of MM, which strongly argue for an inherited genetic predisposition. Given the marginal involvement of CDKN2A mutations, rare mutations in other high-risk genes and/or multiple rare variants with smaller effects, some of which may act through pigmentation and/or nevus formation, are yet to be discovered. However, the replication of our findings would be an important step before embarking on a thorough genetic investigation. This study also has 2 potential practical consequences: Because the symptomatic presentation of RCC is correlated with aggressive histology and advanced disease,43 abdominal ultrasonography or computed tomography scanning performed at the initial workup and during the follow-up of patients with MM may be of value for the early detection of renal cancer. Alternatively, a dermatologic screening visit may be recommended for patients with RCC who have blond/red hair and/or many pigmented nevi. Identification of the genetic determinants of the MM-RCC association will provide not only better insight into the mechanisms underlying these cancers but also guidance for targeting surveillance toward at-risk individuals.

Acknowledgements

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. CONFLICT OF INTEREST DISCLOSURES
  8. REFERENCES

We thank H. Baumert, M. Doré, N. Dupin, B. Escudier, D. Kerob, T. Lesimple, P. Joly, and C. Theodore for medical care of the patients with melanoma and renal carcinoma and for referring patients for this study. We are very grateful to F. Boitier for her support to this work. We also thank A. Carlotti, E. Marinho, A. Spatz, and M. J. Terrier-Lacombe for the pathologic analysis of tumor tissues.

CONFLICT OF INTEREST DISCLOSURES

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. CONFLICT OF INTEREST DISCLOSURES
  8. REFERENCES

Supported by Programme Hospitalier de Recherche Clinique (PHRC) 2001 (AOR-01-091; Drs. Avril and Bressac-de Paillerts), PHRC 2007 (AOM-07-195; Drs. Avril and Demenais), and the Ligue Nationale Contre le Cancer (PRE09/FD; Dr. Demenais).

REFERENCES

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. CONFLICT OF INTEREST DISCLOSURES
  8. REFERENCES