Michael J. Demeure

Purpose: To estimate the relative and absolute risks (ARs) for ovarian, prostate, male and female breast, pancreatic, biliary tract, and colorectal cancers associated with germline BARD1 pathogenic and likely pathogenic variants (PVs/LPVs) because these risks have not previously been reported from a genotype-first perspective. Methods: We performed a secondary retrospective analysis of multicenter data comprising 445 three-generation pedigrees with confirmed BARD1 PVs/LPVs from 25 institutions nationwide. Data were aggregated at the family-relatedness level. The risks were estimated using a Bayesian survival model that incorporated information from the Surveillance, Epidemiology, and End Results (SEER) Program. Results: We confirmed the previously established increased risk for female breast cancer (relative risk [RR] = 3.53; 95% CI, 1.95 to 5.81; posterior probability [PP] = 0.0040) and found positive associations with female pancreatic cancer (RR = 3.76 [95% CI, 2.16-7.25]; PP < .0001), male pancreatic cancer (RR = 2.89 [95% CI, 1.66 to 5.61]; PP < .0001), and prostate cancer (RR = 3.02 [95% CI, 1.95 to 4.71]; PP < .0001). The predictive ARs to age 85 were 41.84% (95% CI, 23.09 to 68.91) for female breast cancer, 4.80% (95% CI, 2.76 to 9.25) for female pancreatic cancer, 3.98% (95% CI, 2.29 to 7.73) for male pancreatic cancer, and 36.11% (95% CI, 23.29 to 56.29) for prostate cancer. Risk estimates suggest an association for male breast, ovarian, and female colorectal cancers; however, the clinical significance is unclear, and these findings warrant cautious interpretation. In contrast, we found no evidence of association for biliary tract cancer or male colorectal cancer. Conclusion: These results support BARD1 's association with female breast cancer and suggest associations with pancreatic and prostate cancers. These findings have potential implications for genetic counseling and personalized risk management for male and female BARD1 PV/LPV heterozygotes.

BRAF inhibitors have a 50–70% response rate in melanoma but are less effective for thyroid cancer. Differential response may be from activation or expression of downstream mitogen-activated protein kinase (MAPK) pathway genes. Retrospective analysis compared whole exome and transcriptome sequencing in melanoma and thyroid cancers from April 2019 to October 2023. The MAPK Activation Score (MPAS) was calculated using Z-score normalized/log-transformed values indicating expression across 10 MAPK-associated genes. Our tumor registry provided outcome data. BRAF V600E mutations were identified in 33 of 200 (17%) melanomas and 14 (7%) had other BRAF mutations (V600K/R). Of 49 thyroid tumor samples, BRAF V600E mutations were found in 19 (39%). RNA expression of BRAF and the 10 MAPK-associated genes were increased in melanomas with V600E compared to wild-type BRAF (p = 0.02). Conversely, BRAF V600E mutation in thyroid cancer was not associated with increased expression nor MAPK pathway activation. No significant difference in overall survival based on BRAF mutation was observed in the subset of patients where data was available. The MAPK pathway is differentially affected by the different cancers, with increased MAPK activation observed in melanoma and not in thyroid cancer. This may account in part for the observed differential response to BRAF inhibitors.

Differential expression of HIF1A and ESR1 transcripts across LMS subtypes

Prevalence of prior treatment and biomarkers among patients stratified by ICI time-on-treatment

Proportion of samples by PD-L1 (IHC) tumor cell staining percentage.

Purpose:Comprehensive molecular profiling was used to define the genomic and immune landscapes of leiomyosarcomas (LMS) by anatomic subtypes, which have not been completely characterized.Experimental Design:A total of 1,115 LMS samples, categorized into uterine LMS (uLMS), retroperitoneal LMS, or other LMS (oLMS), underwent DNA/RNA sequencing (Caris Life Sciences). Genomic/transcriptomic profiles were compared across subtypes. Immune profiling was compared with melanoma (n = 1,255), an immunogenic tumor. Insurance claims data were used to infer real-world outcomes with immune checkpoint inhibitors (ICI) in LMS.Results:uLMSs (n = 701) were molecularly distinct from retroperitoneal LMSs (n = 166) and oLMSs (n = 248). RB1 mutations and MAP2K4 copy-number amplification were more common in non-uLMS. MED12 mutations were almost exclusive to uLMS. Traditional ICI response biomarkers (i.e., PD-L1) did not vary by anatomic site. Non-uLMS demonstrated upregulated immune-related gene sets, including IFN and inflammatory response pathways, and higher immune cell infiltration, especially CD8+ T cells and B cells (>2-fold increase, P < 0.0001). LMS had lower immune cell abundance and T cell–inflamed scores (TIS) compared with melanoma, though 11% of oLMS samples had high TIS scores. In a real-world cohort (n = 138), 29% of patients with LMS receiving ICI were treated >6 months, indicating potential clinical benefit.Conclusions:Comprehensive profiling suggested that uLMS represents a molecularly distinct disease from non-uLMS. Although traditional ICI response biomarkers were similar across anatomic subtypes, uLMSs were immune cold compared with non-uLMS. Signals for ICI responsiveness, such as high TIS and immune cell abundance, in some tumors suggest that further research into immunotherapies for LMS is warranted.

Overall survival and ICI time-on-treatment (TOT) for LMS by anatomic subtype and mutational status for TP53, ATRX, and RB1

Additional methods for DNA NGS and RNA WTS and fusion detection

Purpose: Comprehensive molecular profiling was used to define the genomic and immune landscapes of leiomyosarcomas (LMS) by anatomic subtypes, which have not been completely characterized. Experimental Design: A total of 1,115 LMS samples, categorized into uterine LMS (uLMS), retroperitoneal LMS, or other LMS (oLMS), underwent DNA/RNA sequencing (Caris Life Sciences). Genomic/transcriptomic profiles were compared across subtypes. Immune profiling was compared with melanoma (n = 1,255), an immunogenic tumor. Insurance claims data were used to infer real-world outcomes with immune checkpoint inhibitors (ICI) in LMS. Results: uLMSs (n = 701) were molecularly distinct from retroperitoneal LMSs (n = 166) and oLMSs (n = 248). RB1 mutations and MAP2K4 copy-number amplification were more common in non-uLMS. MED12 mutations were almost exclusive to uLMS. Traditional ICI response biomarkers (i.e., PD-L1) did not vary by anatomic site. Non-uLMS demonstrated upregulated immune-related gene sets, including IFN and inflammatory response pathways, and higher immune cell infiltration, especially CD8+ T cells and B cells (>2-fold increase, P < 0.0001). LMS had lower immune cell abundance and T cell–inflamed scores (TIS) compared with melanoma, though 11% of oLMS samples had high TIS scores. In a real-world cohort (n = 138), 29% of patients with LMS receiving ICI were treated >6 months, indicating potential clinical benefit. Conclusions: Comprehensive profiling suggested that uLMS represents a molecularly distinct disease from non-uLMS. Although traditional ICI response biomarkers were similar across anatomic subtypes, uLMSs were immune cold compared with non-uLMS. Signals for ICI responsiveness, such as high TIS and immune cell abundance, in some tumors suggest that further research into immunotherapies for LMS is warranted.