Germline analysis in Latino young adults with early-onset cancer.

Abstract

e22639 Background: An estimated 5-10% of cancers are inherited, and 35% carry a variant of unknown significance (VUS). We conducted a retrospective study of young Hispanic adults to explore the burden of pathogenic germline variants (PGV), attempt to describe the significance of VUS, and explore whether cancer type or family history was associated with germline findings in this under-represented population. Methods: We conducted a retrospective chart review on 102 Hispanic patients diagnosed with malignancy at age ≤ 45 who underwent germline multi-gene panel testing between 2021 and 2024. PGV prevalence was estimated with exact binomial confidence intervals. Differences in VUS burden across cancer site and PGV status were evaluated using Fisher’s exact test. Nonparametric and exact methods were selected due to small cell counts across cancer subtypes and non-normal distribution of VUS counts. Associations between family history of cancer and germline findings were assessed using logistic regression with sensitivity analyses based on pattern-mixture modeling framework to evaluate robustness under different missing data assumptions. Results: The most common cancer diagnoses were breast (47%), gastrointestinal (24.5%), and hematologic malignancies (11.3%). Other cancer types each represented less than 5% of the cohort. PGVs were identified in 21 of the 102 patients, corresponding to a prevalence of 20.6% (95% CI 13.2- 29.7%). VUS were identified in 35 patients (34.3%). Among patients with VUS findings, most carried a single VUS (23.5%), while fewer patients carried two (8.8%), or three (2.0%). The distribution of VUS counts did not differ significantly across primary cancer sites (Kruskal-Wallis p = 0.886). Fisher’s exact test showed no evidence of an association between cancer type and PGV presence (p = 0.781). Of the 93 patients for whom family history was documented, 51 (54.8%) reported a family history of cancer. Positive family history was consistently associated with higher odds of both PGVs and VUS. Estimated odds ratios for PGV ranged from 2.8 to 3.0 and for VUS from 2.2 to 2.4 across sensitivity assumptions. Confidence intervals were wide and generally included the null value, but estimates were consistent in magnitude and direction. Conclusions: In our young Hispanic population, the PGV rate of 20.6% was significantly higher than the rate of about 8% in the general cancer patient population. Our VUS rate of 34.3% was similar to the VUS rate in the general cancer patient population. Family history of cancer was associated with higher odds of PGVs and twice the likelihood of VUS, suggesting that some of these VUS findings may actually be pathogenic. Our findings underscore the importance of further genetic data collection focused on Latino populations to better understand VUS and address potential disparities in cancer risk and outcomes.