Germ Line BAP1 Mutation in Patients with Uveal Melanoma and Renal Cell Carcinoma

Abstract

Uveal melanoma (UM) and renal cell carcinoma (RCC) can occur sporadically and as a manifestation of BAP1 tumor predisposition syndrome. We aimed to understand the prevalence of germ line BAP1 pathogenic variants in patients with UM and RCC. We reviewed patients managed at Cleveland Clinic between November 2003 and November 2019 who were diagnosed with UM and RCC. Charts were reviewed for demographic and cancer-related characteristics. RCC samples were tested for BAP1 protein expression using immunohistochemical (IHC) staining, and testing for germ line BAP1 pathogenic variants was performed as part of routine clinical care. Thirteen patients were included in the study. The average age at diagnosis of UM was 61.3 years. Seven patients underwent fine-needle aspiration biopsy for prognostic testing of UM (low risk =5, high risk =2). Twelve patients were treated with plaque radiation therapy, and 3 patients developed metastatic disease requiring systemic therapy. The median time to diagnosis of RCC from time of diagnosis of UM was 0 months. RCC samples were available for 7 patients for BAP1 IHC staining (intact =6, loss =1). All patients underwent nephrectomy (total = 3, partial = 8, unknown =2), and 1 received systemic therapy for metastatic RCC. Six patients underwent germ line BAP1 genetic testing. Of these, 1 patient was heterozygous for a pathogenic variant of BAP1 gene: c.1781-1782delGG, p.Gly594Valfs*48. The overall prevalence of germ line BAP1 pathogenic variants in our study was high (1/6; 17%; 95% CI 0–46%). Patients with UM and RCC should be referred for genetic counseling to discuss genetic testing.

Introduction

Patients with uveal melanoma (UM) often have a history of another cancer [1, 2, 3]. The prevalence of a second malignancy in patients with UM may be similar to that observed in the age-matched general population (9.6 vs. 2.8–17%) [1, 2, 3, 4]. However, the occurrence of a second malignancy may be due to cancer predisposition syndromes, environmental exposures, and late side effects of treatment exposure [4].

In the case of UM, BAP1 tumor predisposition syndrome (BAP1-TPDS) may play a role and increase the risk of developing BAP1-TPDS defining cancers which include cutaneous melanoma (CM), malignant mesothelioma (MMe), renal cell carcinoma (RCC), and basal cell carcinoma (BCC) [5, 6, 7, 8]. The development of RCC has been described in 0.6–4.8% of all patients with UM, and such an occurrence may be sporadic, related to BAP1-TPDS, or some other genetic predisposition not yet discovered [5, 6, 7].

Loss-of-function variants of BAP1 gene (BRCA-associated protein 1) in UM [9] and RCC [10, 11] can be inherited or occur sporadically. BAP1 is a deubiquitinating protein involved in regulation of gene expression [9]. Loss-of-function variants of the germ line BAP1 gene result in BAP1-TPDS, which includes the previously mentioned neoplasms along with nonmalignant BAP1-inactivated melanocytic tumor (BIMT) [6, 7, 12, 13]. RCC in BAP1-TPDS is usually of a clear cell subtype [10]. The rate of the germ line BAP1 pathogenic variant in UM and RCC is similar, 1.6 and 1.2%, respectively [10, 14]. On the contrary, somatic mutations of the BAP1 gene and loss of function of BAP1 protein are more frequent in UM (42%) than in RCC (15%) [15]. Other genes such as PBRM1 are more commonly mutated than BAP1 in RCC, which are not observed in UM [11]. Given that synchronous and metachronous UM and RCC may occur sporadically or be related to BAP1-TPDS, we sought to assess the prevalence of germ line BAP1 pathogenic variants in patients with UM and RCC.

Case Presentation

This study was approved by the Cleveland Clinic Institutional Review Board. Patients seen at Cleveland Clinic between November 2005 and November 2019 who were diagnosed with UM and RCC were included. Charts were reviewed for demographic and cancer-related characteristics. Personal and family history was noted for BAP1-TPDS-defining cancers. These were defined as the following malignancies: CM, MMe, BCC, UM, and RCC. BIMT was also considered a feature of BAP-TPDS.

Prognostic testing was performed on UM samples when available. Classes 1A and 1B by gene expression profiling and disomy 3 by multiplex ligation-dependent probe amplification were considered low risk for the recurrence, while class 2 by gene expression profiling and monosomy 3 by multiplex ligation-dependent probe amplification were considered high risk. RCC samples were tested for nuclear BAP1 protein expression using immunohistochemical staining. BAP1 mouse monoclonal antibody was used on formalin-fixed, paraffin-embedded tissue sections. An antigen-antibody complex was localized using the OptiView detection Kit® (Roche Diagnostics, Indianapolis, IN, USA). BAP1 protein expression was scored as intact when nuclear BAP1 staining was present and as loss when nuclear BAP1 staining was absent [16]. Normal cells with intact staining in the background of each sample served as the internal control. Testing for germ line BAP1 pathogenic variants was performed as part of routine clinical care after referral to genetic counseling at a Clinical Laboratory Improvement Amendments (CLIA)-certified laboratory using next-generation sequencing of DNA obtained from salivary or blood samples.

Thirteen patients were included in the study. The average age at diagnosis of UM was 61.3 years. Four patients had a known history of RCC, while remaining 9 patients were incidentally diagnosed with RCC on staging imaging for UM (Fig. 1). Patient 8 had a diagnosis of bilateral metachronous RCC. Median time to diagnosis of RCC from diagnosis of UM was 0 months. Most patients were male (8/13), with low-risk UM (5/13) and RCC of clear cell subtype (9/13) (Table 1). Remaining RCC subtypes included papillary (1/13) and unclassified (1/13), and the subtype was not known for 2 patients. RCC samples were available for 7 patients for immunohistochemical staining. Of these, only 1 patient showed loss of expression of nuclear BAP1 protein. Twelve patients were treated with plaque radiation therapy, and 3 patients developed metastatic disease requiring systemic therapy. All patients underwent nephrectomy (total = 3, partial = 8, unknown = 2). Six patients underwent testing for the germ line BAP1 gene, and only 1 carried a pathogenic variant. The overall prevalence of germ line BAP1 pathogenic variants in this population was 17% (1/6; 95% CI 0–46%).

Fig. 1.

Time to diagnosis of RCC in relation to UM. RCC, renal cell carcinoma; UM, uveal melanoma.

Table 1.

Results of germ line BAP1 genetic testing

Group	N = 13	Patients with germ line BAP1 pathogenic variants, n (%)
Gender
Male	8	1 (12.5)
Female	5	0 (0)
UM prognostic class
Unavailable	6	1 (17)
Low risk	5	0 (0)
High risk	2	0 (0)
RCC subtype
Clear cell	9	1 (11)
Papillary	1	0 (0)
Unclassified	1	0 (0)
Unknown	2	0 (0)
RCC BAP1 stain status
Unavailable	6	0 (0)
Intact	6	0 (0)
Loss	1	1 (100)
Personal history of cancers*
Present	5	1 (20)
Absent	8	0 (0)
Family history of BAP1-TPDS-defining cancers
Present	1	1 (17)
Absent	12	0 (0)
Family history of any cancer
Present	5	1 (20)
Absent	8	0 (0)

UM was considered low risk for class 1a, 1b, and disomy 3 groups and was considered high risk for class 2 groups. Following cancers were considered BAP1-TPDS-defining cancers: UM, RCC, CM, MMe, and BCC. UM, uveal melanoma; RCC, renal cell carcinoma; BAP1-TPDS, BAP1-tumor predisposition syndrome; BCC, basal cell carcinoma; CM, cutaneous melanoma; MMe, malignant mesothelioma. * Other than UM, and RCC.

Patient 6 was heterozygous for a pathogenic variant of the germ line BAP1 gene (Fig. 2). This was a 50-year-old male patient with incidentally diagnosed RCC on staging imaging for UM. The patient had multiple nevi on examination and was later diagnosed with BIMT and CM. He had a family history of BCC in his mother and maternal grandmother, eye cancer of unknown type in his paternal grandmother, ductal carcinoma in situ in his sister, and abdominal cancer of an unknown type in his paternal aunt. The patient had a low-risk UM, RCC of clear cell subtype, and loss of BAP1 protein expression on the RCC stain.

Fig. 2.

Pedigree of patient 6. The patient had a personal and family history of multiple BAP1-TPDS-defining cancers. Family history of other cancers included DCIS in his sister and abdominal cancer of unknown type in his paternal aunt and maternal grandmother. The patient was heterozygous for the pathogenic variant of the germ line BAP1 gene c.1781_1782delGG. UM, RCC, BIMT, and CM in self; BCC in his mother and maternal grandmother. RCC, renal cell carcinoma; UM, uveal melanoma; DCIS, ductal carcinoma in situ; TPDS, tumor predisposition syndrome; CM, cutaneous melanoma; BCC, basal cell carcinoma; BIMT, BAP1-inactivated melanocytic tumor.

The patient was heterozygous for the variant of the BAP1 gene: c.1781_1782delGG, p.Gly594Valfs*48. To our knowledge, this variant has not been previously reported and is expected to be pathogenic as it results in premature protein termination. The patient also demonstrated somatic loss of heterozygosity of the BAP1 gene on a metastatic UM sample from the liver, suggesting that the tumor was related to BAP1 loss of function. The parents were not tested for germ line mutation. The patient was alive at the end of the follow-up with metastatic UM undergoing localized radiation therapy to the liver and systemic therapy with immune checkpoint inhibitor.

Patients with a family history of non-BAP1-TPDS-defining cancers did not harbor germ line BAP1 pathogenic variants (Table 1). At the end of the follow-up, 4 patients had developed UM metastasis and 1 patient from RCC (Table 2).

Table 2.

Demographic and clinical characteristics of patients

Patient, N0./sex/race	Age at dx, years, UM/RCC	UM prognostic testing	Initial UM treatment	RCC subtype	RCC BAP1 IHC status	Presence of germ line BAP1 pathogenic variants	Follow-up since UM diagnosis, months	UM mets	RCC mets	Cutaneous nevi/melanoma	Other cancers
1/M/W	49/41	GEP 1B	Enucleation	CC	NK	Negative	1	NK	NK	Absent	Absent
2/M/W	66/65	NK	Plaque	NK	NK	NK	9	Present	Absent	Absent	Absent
3/F/W	70/70	NK	Plaque	UC	Intact	NK	12	Absent	Absent	Absent	Breast ca
4/F/W	68/69	NK	Plaque	CC	Intact	NK	157	Absent	Absent	Absent	BCC
5/M/W	68/68	MLPA disomy 3	Plaque	CC	Intact	Negative	49	Absent	Absent	Absent	Absent
6/M/W	50/51	NK	Plaque	CC	Loss	Positive	168	Present	Absent	Cutaneous nevi and melanoma present	Prostate ca BCC
7/F/W	76/76	GEP 2	Plaque	NK	NK	Negative	33	Present	Absent	Absent	Lung ca
8/F/W	60/51	GEP 1A	Plaque	CC	Intact	Negative	9	Absent	Present	Absent	Absent
9/M/W	64/64	GEP 2	Plaque	Papillary	NK	NK	34	Present	Absent	Absent	Absent
10/M/W	59/53	NK	Plaque	CC	Intact	NK	44	Absent	Absent	Absent	Absent
11/M/W	54/53	GEP 1B	Plaque	CC	Intact	NK	66	Absent	Absent	Absent	Absent
12/F/W	60/55	GEP 1B	Plaque	CC	NK	Negative	62	Absent	Absent	Absent	Absent
13/M/W	56/56	NK	Plaque	CC	NK	NK	5	Absent	Absent	Absent	Chronic myeloid leukemia

M, male; F, female; W, white; dx, diagnosis; UM, uveal melanoma; RCC, renal cell carcinoma; NK, not known; CC, clear cell; UC, unclassified; IHC, immunohistochemical; LOF, lost to follow-up; GEP, gene expression profiling; MLPA, multiplex ligation-dependent probe amplification; BCC, basal cell carcinoma; ca, cancer.

Discussion

Several studies have reported the occurrence of UM and RCC in the same patients [15, 17]. Of these, only one study reports the prevalence of germ line BAP1 pathogenic variants in this patient population [17]. The study by Boru et al. evaluated the prevalence of germ line BAP1 mutation in patients with UM who were at high risk for hereditary cancer syndrome (age of diagnosis of UM <35 years, familial UM, and personal and/or family history of other BAP1-TPDS-defining cancers). The overall prevalence of germ line BAP1 mutation in their study was 4.7% (8/172). The prevalence of germ line BAP1 mutation in patients with familial UM and a personal or family history of RCC was 33% (1/3) [17]. None of the 13 patients with UM in the absence of familial UM had a germ line BAP1 mutation, while one of the 16 patients with UM and a family history of RCC and CM carried a germ line BAP1 mutation [17].

The point prevalence of UM in patients with germ line BAP1 pathogenic variants is estimated to be 2.8% (95% CI, 0.88−4.81%) [18], similar to the frequency of BAP1 germ line pathogenic variants in patients with UM of 2% (0.80–3.08) [8, 14, 19, 20, 21, 22], with higher frequency in studies that have selected a population either with familial UM and personal or family history of MMe, and CM (4.7–8.0%) [6, 17, 23]. The prevalence of germ line BAP1 pathogenic variants in our study of 17% (1/6) is even higher as patients in our study represent a selected sample of a high-risk population.

The patient positive for a germ line BAP1 pathogenic variant in our study had a personal and family history of other BAP1-TPDS-defining cancers in addition to UM and RCC. The remaining 5 patients without a personal or family history of other BAP1-TPDS-defining cancers did not harbor a germ line mutation. Of note, the patient's parents were not tested for germ line mutation; therefore, it is unclear whether the mutation was inherited. His mother and maternal grandmother both had BCC which are also considered as BAP1-TPDS-defining cancers.

It is clear that patients with UM and a personal and/or family history of more than one BAP1-TPDS-defining cancer should be tested for germ line BAP1 mutation, given the high prevalence of pathogenic variants in this population (∼38%) [17]. However, based on the small number of patients in our study, we cannot conclude if patients with UM and a personal history of RCC alone should routinely be tested for germ line mutations. UM and RCC may occur sporadically in these patients since UM and RCC are both more frequent in men older than >60 years [6, 24]. BAP1 stain of RCC and UM is also likely not helpful in identifying patients needing BAP1 germ line testing as the prevalence of germ line BAP1 mutation in these populations is much lower (∼1.6% for UM and 1.2% for RCC, respectively) [10, 14].

Furthermore, it is possible that other unidentified germ line mutations may be at play in patients with UM and RCC. However, even if germ line testing is not routinely performed in this population, these patients should be referred for genetic counseling for a thorough assessment of family history and identification of high-risk patients who warrant germ line testing. The small size of our study population and the inability to test all patients for germ line BAP1 mutation limit the validity of our results. A larger study is needed to define the true prevalence of germ line BAP1 pathogenic variants in patients with UM and RCC.

Conclusion

Patients with UM and RCC should be referred for genetic counseling for thorough assessment of family history and identification of high-risk patients who will benefit from genetic testing.

Statement of Ethics

Our research complies with the guidelines for human studies, and the research was conducted ethically in accordance with the World Medical Association Declaration of Helsinki. The study protocol (17-1186) was approved by the Cleveland Clinic Institutional Review Board. Patients' consent was not required as the study was based on review of medical records.

Conflict of Interest Statement

Arun D Singh: one of the editors-in-chief of Ocular Oncology and Pathology. Other authors have no conflicts of interest to declare.

Funding Sources

No funding was obtained for this study.

Author Contributions

Yusra F. Shao: primary contribution to acquisition and analysis of data, writing and revising the manuscript for approval of final version, and agreement to be accountable for all aspects of the work.

Meghan DeBenedictis: contribution to acquisition of patient data, critically revising, final approval of the manuscript, and agreement to be accountable for accuracy and integrity of the work.

Gabriella Yeaney: contribution to acquisition of patient data, critically revising, final approval of the manuscript, and agreement to be accountable for accuracy and integrity of the work.

Arun D. Singh: primary contribution to the conception of study design, interpretation of data, revising the manuscript for intellectual content and accuracy, final approval of the published version of the manuscript, and agreement to be accountable for accuracy and integrity of the work.