Search

PurposeLeiomyosarcomas (LMS) harbor frequent inactivation of TP53 and RB1, and extensive DNA copy number alterations. Here, we describe a distinct recurrent genomic signature in TP53/RB1-wildtype uterine LMS.MethodsTissues from 276,645 unique advanced cancers, including 2,570 uterine and soft tissue LMS were sequenced by hybrid-capture-based next-generation DNA and RNA sequencing/comprehensive genomic profiling of up to 406 genes. We characterized clinicopathologic features of relevant cases.ResultsWe identified 77 LMS with homozygous copy loss of CDKN2C at chromosome 1p32.3 (3.0% of LMS). Genomic alterations (GAs) in TP53, RB1, and ATRX were rare in comparison with the remainder of the LMS cohort (11.7% vs 73.4%, 0% vs 54.5%, 2.6% vs 24.5%, all pCDKN2C-null LMS cases were significantly enriched for GAs in CIC (40.3% vs 1.4%) at 19q13.2, CDKN2A (46.8% vs 7.0%), and RAD51B (16.9% vs 1.7%; all pCDKN2C-null LMS cases exhibited 1p/19q-codeletion, with significant enrichment in comparison to the remainder of the evaluated LMS cohort (85% vs. 5.1%, p99% of CDKN2C-null cases were in females; median age was 61 years at surgery (range, 36-81 years). 55 cases were of uterine primary, 4 cases non-uterine, and the remaining 18 of uncertain primary site. 6 patients had a prior history of leiomyomatosis or uterine smooth muscle tumor of uncertain malignant potential. 60% of cases showed at least focal epithelioid variant histology. Most cases were of known advanced stage, with 62% of confirmed uterine primary cases at FIGO stage IVB.ConclusionThe identification of this novel genomic subset may have prognostic and/or therapeutic clinical relevance, including use of specific cyclin-dependent kinase inhibitors.

Objectives: Among women with high-risk human papillomavirus (hrHPV) detected in cervical swabs, the distribution of hrHPV types is known to differ by racial/ethnic group, with East and West Africans and North Americans of African ancestry1 showing overrepresentation of hrHPV types that are not covered by the current 9-valent HPV vaccine. Possible downstream consequences in ethno-specific rates of aggressive hrHPV-related cancers, however, have not been systematically examined. Here, we surveyed a large, multiethnic database of aggressive cancers for non-9-valent-vaccine hrHPV types in association with individual ancestry data. Download : Download high-res image (398KB) Download : Download full-size image Methods: Our archive of 290,311 unique advanced cancer samples underwent comprehensive genomic profiling (CGP) using a hybrid-capture-based DNA sequencing platform. We detected hrHPV genomic sequences in 1,690 cervical, 1,196 anogenital, and 1,380 head and neck cancers by de novo assembly of nonhuman reads, followed by alignment to the RefSeq database, which distinguishes hrHPV types. Patient ancestry was determined by classifying specific SNPs by CGP based on their known variation among populations in the 1000 Genomes Project. Quantitative data were analyzed using the Fisher exact test. A two-tailed P value of Results: We identified four non-vaccine hrHPV types with >10 cancer cases each: HPV35 (n=71 cases), HPV51 (n=11), HPV59 (n=42), and HPV67 (n=12). Among patients whose cancers harbored hrHPV types that are covered by the 9-valent vaccine (here, HPV16, 18, 31, 33, 45, 52, and 58), 7.5% (308/4,131) were of African ancestry, lower than the overall representation of African ancestry in our database (9.1% [26,471/290,311]). In contrast, among cases with non-9-valent-vaccine hrHPV types, we identified several-fold enrichment for individuals of African ancestry (HPV35: 24% of African ancestry [17/71], HPV51: 27% [3/11], HPV59: 43% [18/42], HPV67: 25% [3/12]). In total, these four non-vaccine hrHPV types accounted for 11.4% (40/350) of all hrHPV-related advanced cancers in persons of African ancestry, but only 2.4% (95/3928) of hrHPV-related advanced cancers in individuals of other ancestries, a difference of nearly five-fold (p Conclusions: In our pan-hrHPV analysis of advanced cancers, hrHPV types not covered by the current 9-valent vaccine are strikingly increased in individuals of African ancestry. Our data strongly indicate a need to re-evaluate the choice of hrHPV types to include in the vaccine, as well as a need to re-evaluate the underlying process by which hrHPV types are chosen for inclusion.

Introduction: One major form of vulvar squamous cell carcinoma (vSCC) is associated with detectable high-risk strains of human papillomavirus (hrHPV) and is often accompanied by usual-type vulvar intraepithelial neoplasia (VIN). The second major form of vSCC is often associated with chronic dystrophic or inflammatory lesions in postmenopausal women, does not harbor detectable HPV infection, and is often preceded by p53-mutant differentiated VIN. While studies have examined the two subtypes, no large-scale genomic study has been performed to our knowledge. We sought to assess the genomics of a large cohort of aggressive vSCCs, with an aim to identify distinct mutational signatures based on the presence or absence of hrHPV genome reads. Methods: 280 vSCC were tested by hybridization capture of up to 406 cancer-related genes evaluated for base substitutions, small indels, amplification (amp), and rearrangements. HPV genome sequences were detected by de novo assembly of non-human sequencing reads and BLASTn comparison against all viral nucleotide sequences in the NCBI RefSeq database. Tumor mutational burden (TMB, mutations/Mb) was determined on ~1.1 Mbp of sequenced DNA. PD-L1 status was determined by IHC (Dako 22C3), with ≥50% tumor proportion score defined as high positive. Results: 102/280 vSCCs contained hrHPV sequences. Of these, 90 were HPV-16, 7 HPV-18, 1 HPV-31, 3 HPV-33, 1 HPV-58, and 1 HPV-67. Patients were significantly younger in the HPV(+) group (median 59 v. 64 years, p=0.001). Compared with the HPV(–) cohort, HPV(+) cases showed significantly more pathogenic genomic alterations (GA) in PIK3CA (31% vs. 17%, p=0.004), PTEN (14% vs. 2%, p10*6%24%3%19% Conclusions: vSCCs show significant differences in molecular profile based on HPV status. 63% of metastatic HPV(+) cases (54% overall) have a potentially actionable alteration in the PI3K/mTOR pathway, and 42% of metastatic HPV(–) cases (39% overall) have at least one potential predictive biomarker* for response to immunotherapy. Our findings provide compelling rationale for tandem comprehensive genomic profiling and HPV assessment of advanced vulvar SCCs to more fully inform therapeutic options and stratification in clinical trials. Citation Format: Erik A Williams, Adrienne J Werth, Meagan Montesion, Ethan S Sokol, Dean C Pavlick, Nikunj A Shah, Jo-Anne Vergilio, Natalie A Danziger, Jonathan K Killian, Douglas A Lin, Vincent A Miller, Jeffrey S Ross, Julia A Elvin. Vulvar squamous cell carcinoma: Comprehensive genomic profiling of HPV(+) versus HPV(–) forms reveals a different set of potentially actionable biomarkers [abstract]. In: Proceedings of the AACR-NCI-EORTC International Conference on Molecular Targets and Cancer Therapeutics; 2019 Oct 26-30; Boston, MA. Philadelphia (PA): AACR; Mol Cancer Ther 2019;18(12 Suppl):Abstract nr B091. doi:10.1158/1535-7163.TARG-19-B091