Association of germline variants in telomere maintenance genes (POT1, TERF2IP, ACD, and TERT) with spitzoid morphology in familial melanoma: A multi-center case series

Background Spitzoid morphology in familial melanoma has been associated with germline variants in POT1, a telomere maintenance gene (TMG), suggesting a link between telomere biology and spitzoid differentiation. Objective To assess if familial melanoma cases associated with germline variants in TMG (POT1, ACD, TERF2IP, and TERT) commonly exhibit spitzoid morphology. Methods In this case series, melanomas were classified as having spitzoid morphology if at least 3 of 4 dermatopathologists reported this finding in ≥25% of tumor cells. Logistic regression was used to calculate odds ratios (OR) of spitzoid morphology compared to familial melanomas from unmatched noncarriers that were previously reviewed by a National Cancer Institute dermatopathologist. Results Spitzoid morphology was observed in 77% (23 of 30), 75% (3 of 4), 50% (2 of 4), and 50% (1 of 2) of melanomas from individuals with germline variants in POT1, TERF2IP, ACD, and TERT, respectively. Compared to noncarriers (n = 139 melanomas), POT1 carriers (OR = 225.1, 95% confidence interval: 51.7-980.5; P < .001) and individuals with TERF2IP, ACD, and TERT variants (OR = 82.4, 95% confidence interval: 21.3-494.6; P < .001) had increased odds of spitzoid morphology. Limitations Findings may not be generalizable to nonfamilial melanoma cases. Conclusion Spitzoid morphology in familial melanoma could suggest germline alteration of TMG.


INTRODUCTION
Approximately 10% of cutaneous melanoma patients have a family history of the disease (familial melanoma). 1 Germline pathogenic variants (GPVs) in melanoma susceptibility genes CDKN2A, CDK4, and BAP1 are identified in up to 40%, 0.7%, and 1.0% of familial melanoma cases, respectively. 1,2 More recently, GPV in telomere maintenance genes (TMG) POT1, TERF2IP, ACD, and TERT have been observed in up to 6% of melanoma-prone families. [3][4][5][6][7] POT1 GPV also predispose to lymphoid and myeloid malignancies, angiosarcoma, and glioma. 8,9 BAP1 GPV carriers commonly develop spitzoid melanocytic neoplasms, characterized by large epithelioid melanocytes with abundant cytoplasm. The biological mechanisms underlying this genotypehistotype association are unknown. 10,11 Spitzoid subtype melanomas with complete spitzoid morphology (100% of tumor) and cutaneous melanomas with focal spitzoid morphology ($25% of tumor) have also been Koplowitz Center, Barcelona. The work at Fundaci on Instituto Valenciano de Oncolog ıa was partially funded by the Instituto de Salud Carlos III (PI19/00667), the Generalitat Valenciana (Prometeo 21/067), and the EADV (PPRC,-2018-36). The funders had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication. IRB  Spitzoid morphology is common in familial melanoma cases from individuals with POT1, TERF2IP, ACD, and TERT germline variants. Dermatologists and dermatopathologists should be aware of these associations when ordering familial melanoma gene testing panels.
observed in POT1 GPV carriers, suggesting dysfunctional telomere maintenance as a possible mechanism for spitzoid histology. 12 Here, we further investigate this hypothesis by evaluating the pathology of familial melanoma cases from individuals with germline variants in TMG.

METHODS
Whole slide images of cutaneous melanomas from individuals with any germline variant in a TMG (POT1, TERF2IP, ACD, and TERT ) and at least one first, second, or third degree relative with melanoma, were requested during the virtual Melanoma Genetics Consortium (GenoMEL) annual meeting on July 27, 2020. Images were provided to the National Cancer Institute by research groups conducting familial melanoma studies in the United States (1 group), Europe (3 groups), and Australia (2 groups). Collaborators submitted all melanoma cases they had access to from variant carriers, which were independently reviewed by 4 dermatopathologists (MRS, DEE, EYC, and DM) with expertise in melanocytic lesions, who were blinded to clinical information and the specific gene altered. Each dermatopathologist reported histologic subtype and the percentage of tumor cells exhibiting spitzoid morphology, defined as large epithelioid or spindled melanocytes with a cytoplasm to nuclear ratio of greater than or equal to 1. To be consistent with our previous study on this topic, melanomas were classified as spitzoid if at least 3 of 4 dermatopathologists reported spitzoid morphology in 25% or more of the tumor cells. 12 Logistic regression was used to estimate odds ratios (ORs) and 95% confidence intervals (CIs) for spitzoid morphology and subtype in variant carriers compared to unmatched noncarriers (United States, Europe). All noncarriers had whole exome sequencing performed to exclude GPV in melanoma predisposition genes (CDKN2A, CDK4, BAP1, POT1, TERT, ACD, and TERF2IP). Noncarrier cases (all cutaneous melanomas) were previously reviewed by a single dermatopathologist (MRS), 12 and were from individuals with at least one first, second, or third degree relative with melanoma (same as variant carriers). CIs accommodated within person and within family correlations by using a generalized estimating equation approach with a working independence covariance matrix. Analyses were performed in R Studio and SAS 9.4 (SAS Institute Inc).
Familial melanoma studies submitting whole slide images to the National Cancer Institute had local institutional review board or ethics review committee approval.
Compared to noncarriers, individuals with germline POT1 variants had a 10.5-fold increased odds (95% CI: 2.1-51.7; P = .004) of developing a spitzoid subtype melanoma (complete spitzoid morphology) and a 225.1-fold increased odds (95% CI: 51.7-980.5; P \ .001) of having a melanoma with spitzoid morphology involving at least 25% of the tumor. Restricting the analysis to cases from individuals with likely pathogenic or pathogenic variants in POT1 based on the American College of Medical Genetics and Genomics and Association for Molecular Pathology guidelines for the interpretation of sequence variants also yielded similar results (OR [spitzoid subtype] = 12.1, 95% CI: 2.2-65.7, P = .004; OR [spitzoid morphology] = 205.5, 95% CI: 8.7-1089.9, P \ .001). BAP1 germline status was Abbreviations used:

CI:
confidence interval GPV: germline pathogenic variant OR: odds ratio TMG: telomere maintenance gene unknown in 10 cases from POT1 carriers and excluding these cases did not impact associations (Table II).
Spitzoid morphology was also more prevalent in melanomas from individuals with TERF2IP, ACD, and TERT variants compared to noncarriers

DISCUSSION
Identification of individuals with GPV in TMG is critically important for counseling and cancer screening. Our analysis identified spitzoid morphology ($25% of tumor) in 77% of cutaneous melanomas from POT1 carriers, consistent with a recent case series from our group where 60% of cutaneous melanomas from POT1 carriers exhibited this histology. 12 In the current study, all melanomas exhibiting spitzoid morphology from POT1 carriers occurred in adulthood (range: 35-77 years), although spitzoid subtype melanomas with complete spitzoid morphology (100% of tumor) have been reported in POT1 carriers as young as age 16 years. 12 We also identified spitzoid morphology in melanomas from individuals with variants in TMG other than POT1 (TERF2IP, ACD, and TERT ). All cases classified as having spitzoid morphology were independently observed to have this histologic finding by at least 3 of 4 dermatopathologists suggesting that spitzoid cells can be differentiated from other morphologies.
BAP1 variants have also been associated with spitzoid pathology. 10 Inactivation of BAP1 increases TERT expression and telomere elongation, suggesting that it may induce spitzoid differentiation via dysfunctional telomere maintenance. 13 In our study, BAP1 variants were not detected in any individuals with TERF2IP, ACD, or TERT variants. Furthermore, exclusion of cases from POT1 carriers with unknown BAP1 germline status did not significantly impact associations.
Our study findings support the inclusion of TMG (POT1, TERF2IP, ACD, and TERT ) on genetic testing panels when a familial melanoma case exhibits spitzoid morphology in at least 25% of  the tumor cells. Since genetic testing is most commonly ordered by dermatologists, oncologists, or geneticists who evaluate melanoma-prone families, dermatopathologists should report the presence ($25% of tumor) or absence of spitzoid morphology in the synoptic or microscopic description sections of all melanoma pathology reports regardless of the melanoma subtype (superficial spreading, nodular, lentigo maligna, spitzoid, acral, and desmoplastic) or family medical history, which could be incomplete or unavailable at the time of biopsy. Routine reporting of this information could assist with identifying individuals with GPVs in TMG, and detecting these variants will inform cancer surveillance for the patient, and possibly family members, based on the associated cancer risks.

LIMITATIONS
Although this case series is the largest histopathologic assessment of cutaneous melanomas from individuals with germline variants in TMG, our study ascertained cases from GenoMEL melanoma-prone families, and therefore, the study findings may not be  applicable to nonfamilial melanoma cases. Additionally, while cases were independently reviewed by multiple expert dermatopathologists, confirmation bias cannot be excluded. Therefore, case-control studies are needed to confirm the observed findings in this study and assess interobserver agreement for spitzoid classification. Noncarrier cases were reviewed by a single dermatopathologist and were not matched by country, which could have also impacted association estimates. However, previous GenoMEL studies of familial melanoma, involving histopathology review by multiple expert pathologists, did not identify spitzoid morphology in CDKN2A carrier and noncarrier familial melanomas. 14,15 Lastly, tissue was not available for cases with spitzoid morphology to evaluate for somatic alterations (ie, HRAS variants and gene fusions involving ALK, ROS1, NTRK1, NTRK3, MET, RET, BRAF, MAP3K8, and CRTC1::TRIM11) that have been associated with pediatric spitzoid tumors. 16,17 Therefore, tumor studies are needed to understand the biological mechanisms linking dysfunctional telomere maintenance to spitzoid differentiation in familial melanoma cases.

CONCLUSIONS
In conclusion, we identified that a high proportion of cutaneous melanomas from individuals with germline variants in TMG exhibit spitzoid morphology. Therefore, panel testing for POT1, TERF2IP, ACD, and TERT should be considered when familial melanoma cases exhibit spitzoid differentiation.