Comprehensive genomic profiling (CGP) of clinically advanced and metastatic cutaneous adnexal carcinomas (CAs; MCADCA): A genomic landscape study.

Abstract

9587 Background: Cutaneous adnexal tumors are a large group of malignant neoplasms that exhibit morphologic differentiation towards one of the four primary adnexal structures present in normal skin: hair follicles, sebaceous glands, sweat-apocrine glands, and sweat-eccrine glands. Methods: 276 cases of MCADCA underwent hybrid capture-based CGP to evaluate all classes of genomic alterations (GA). Microsatellite instability (MSI) status, tumor mutational burden (TMB), genomic loss of heterozygosity (gLOH), genomic ancestry prediction and COSMIC genomic signature were determined from sequencing results. PD-L1 was measured by IHC (TPS; Dako 22C3). Fisher exact method was used for statistical analysis with the false discovery rate corrected using Benjamini-Hochberg adjustment. Results: Sequencing was performed on the primary cutaneous tumor in 131 (47.4%) and on a local recurrence or metastatic site biopsy in 145 (52.5%) MCADCA. Male preponderance (64-81%) and age distributions mean (59-63 yrs) were similar in all groups with apocrine (APO) older than eccrine (ECC) (72 vs 62 yrs; p = .001). There were 173 (62.7%) sweat gland (SWT) derived, 55 (19.9%) were sebaceous gland (SEB) derived, 14 (5.1%) were hair follicle (HRF) derived and 34 (12.3%) were unclassified (UNK). 150/173 (86.7%) of SWT were ECC and 23 (13.3%) were APO. The SWT cases included 12 (6.9%) digital papillary ca’s (DPA), 11 (6.3%) mucinous ca’s (MAC), 19 (11.0%) poroca’s (POR), 14 (8.1%) spiradenoca’s (SPR), 10 (5.8%) syringoadenoca’s (SRNG) and 77 (44.5%) unclassified ca’s (UNC). GA/tumor were highest in SEB vs SWT (7.9 vs 4.9; p = .004) and lowest in DPA (2.1 vs 5.0 non-DPA; p = .03). There we no ancestry distribution differences. When compared with the SWT group, the SEB group had higher frequencies of RB1 GA (38.2% vs 8.1%; p < .0001) and TP53 GA (76.4% vs 43.4%; p = .0002) indicating likely neuroendocrine differentiation. The MAC group had significantly higher PTCH1 GA than the non-MAC group (36.4% vs 1.8%; p = .044). MSI-High status was highest in the SEB group vs all other groups (SEB 15.7% vs SWT 1.2%; p = .005). gLOH > 16% was most frequent in SEB when compared with SWT (19.6% vs 7.2%; p = .081). The MMR signature was more frequent in SEB than SWT (32.0% vs 2.1%; p = .005). The mean TMB was high in all MCADCA ranging from 10.4 mut/Mb in HRF to 38.8 mut/Mb in MAC with the exceptions of APO (2.7 mut/Mb; p = .001 and DPA 1.4 mut/Mb; p = .003). PD-L1 expression was not significantly different in all the MCADCA with low positive scores SWT vs SEB (37.0% vs 33.3%; NS). Conclusions: As there is limited evidence on the treatment of MCADCA, we provide a repertoire of common GA found. An increased frequency of GA is observed in SEB MCADCA. Clinical trials are needed to evaluate MCADCA-based biomarkers prognostic of ICI response. Only then can we successfully incorporate novel therapeutic strategies in the evolving landscape of treatment.