Table S7. Clinical responses of advanced PDGFRA-mutant GIST patients treated with first-line imatinib.

Pathogenic POLE mutations (pPOLE) undermine mismatch error correction by polymerase ε during DNA replication, and the resulting somatic ultramutation predicts response to immunotherapy. Beyond frequently recurrent alleles, historical pPOLE classification has been largely based on exonuclease domain localization. A POLE-specific phenotypic classification model was developed, encompassing tumor mutational burden (TMB), mutational signatures, germline frequency, and consideration of comutation with other POLE mutations to identify pPOLE. This model was applied to >490,000 samples and identified 29 predicted pPOLE, including 16 not previously reported. A total of 748 tumors (0.2%) had one or more pPOLE, most commonly in endometrial and colorectal cancers, although pPOLE were observed in many additional cancer types. pPOLE were associated with ultramutation [median TMB, 186.3 mutations per megabase (mut/Mb)] across tumor types. Concurrent pPOLE and microsatellite instability were more common than previously appreciated and produced a synergistic TMB impact, with medians of 135.7 mut/Mb for pPOLE/microsatellite stable samples compared with 325.6 mut/Mb for pPOLE/microsatellite instability-high samples. Comutation analysis in endometrial and colorectal cancers highlighted associations with homologous recombination pathway gene mutations that were predominantly monoallelic passengers that are unlikely to predict response to therapies targeting DNA repair deficiencies. pPOLE have been incorporated into treatment guidelines for several malignancies and are an important predictor of immunotherapy response. This study provides biological insight to guide classification and clinical management of patients with tumors harboring pPOLE.

Mutation analysis of our PDGFRA-mutant GIST cohort reveals clustering around the exon 18, 842-codon position. A, The protein domains of PDGFRA (AL, activation loop; AP, ATP-binding domain; EC, extracellular domain) and the exon distribution of unique primary mutations seen in PDGFRA-mutant GIST (n = 1,379 cases). B, Breakdown of mutations seen in exon 18 PDGFRA-mutant GIST cases (n = 1,122/1,379). The gray-colored proportion in the chart indicates mutations with five or fewer reported cases (n = 90/1,122). C, Number of cases with exon 18 in/del mutations and the net number of deleted residues in the final protein sequence, along with the breakdown of the type of mutations observed with a 4-residue deletion. Red “X” denotes the number of amino acids inserted (X = 1, XX = 2). D, Proportion of cases with mutations that directly alter the 842 residue. E, Distribution of the amino acid occupying the 842-position in exon 18–mutant cases (n = 1,122). Colors in the legend correspond to the amino acid class of the 842-position residue (hydrophobic, polar uncharged, special case, positively charged, and negatively charged). [Portions of A were Created in BioRender. Khosroyani, H. (2026) https://BioRender.com/l88qy7p.]

Abstract The most common platelet-derived growth factor receptor α (PDGFRA) alteration in gastrointestinal stromal tumors (GIST) is the exon 18 activation loop mutation D842V, which is resistant to imatinib and other type II tyrosine kinase inhibitors (TKI) but sensitive to type I TKI avapritinib. Avapritinib is FDA-approved for first-line treatment of all PDGFRA exon 18–mutant GIST cases but is only available for D842V-mutant cases outside the United States. Non-D842V exon 18–mutant GIST cases are understudied and lack evidence-based treatment guidelines. However, there are a few previous reports describing non-D842V exon 18–mutant patients with GIST who responded well to imatinib therapy. Given that imatinib is more tolerable, globally accessible, and significantly less expensive than avapritinib, we sought to define which patients could be treated with imatinib rather than avapritinib. We assembled a cohort of more than 1,000 PDGFRA exon 18–mutant GIST cases and identified that 78% of these mutations involved a key autoinhibitory aspartic acid residue at position 842. Using cell-based models, we demonstrated that imatinib sensitivity was dependent on the amino acid class of the 842-position residue, with all hydrophobic amino acids except alanine conferring resistance. In contrast, all 842-position mutations were avapritinib-sensitive. Structural modeling supported our biochemical results and revealed how 842-position mutations induce changes that can interfere with imatinib binding. Lastly, our biochemical data were validated using imatinib response data for first-line metastatic disease; patients with predicted exon 18–sensitive mutations had longer progression-free survival than patients with predicted imatinib-resistant mutations. These results provide key evidence that should be used to guide therapy selection for PDGFRA-mutant GIST. Significance: Biochemical and structural modeling of PDGFRA exon 18 842-position mutations allows for the prediction of clinical TKI responses. These data can be used to generate new treatment guidelines for PDGFRA exon 18–mutant GISTs and select optimal therapies for patients.

ABSTRACT Reversion mutations (REVs) restore homologous recombination repair (HRR) and confer resistance to PARP inhibitors (PARPi) in HR-deficient cancers. Yet, their prevalence, mechanisms, and biological constraints remain undefined. We analyzed genomic profiling of 609,464 tissue and liquid biopsy samples across multiple cancer types to delineate the pan-cancer landscape of REVs. REVs were identified in eight HRR genes, most frequently BRCA2 and BRCA1 and notably never in ATM or CHEK2 . REVs exclusively impacted truncating pathogenic variants, predominantly through large in-frame and exon-level deletions, associated with repetitive sequences. Conserved functional domains are relatively depleted of REVs. Structural modeling and functional studies support that exon-level deletions preserve critical domain architecture and confer PARPi resistance. The study establishes HRR reversion as a structurally permissive, yet evolutionarily constrained, resistance mechanism with implications on response, monitoring, and therapeutic strategy. One Sentence Summary Homologous recombination reversion evolves through structurally tolerated, microhomology-driven alterations that restore DNA repair under therapeutic selection across cancers.

The mechanisms by which mutations of splicing factor gene U2AF1 contribute to lung adenocarcinoma pathogenesis are not well understood. Here we used prime editing to modify the endogenous U2AF1 gene in lung adenocarcinoma cells and assessed the impact on alternative splicing. One specific KRAS mutation, G12S, led to skipping of KRAS exon 2 and generation of a nonfunctional KRAS transcript. However, expression of the U2AF1S34F mutant reverted this exon skipping and restored KRAS function, leading to enrichment of U2AF1S34F mutations in KRASG12S-mutant lung adenocarcinomas. Comprehensive analysis of splicing factor-oncogene mutation co-occurrence in cancer genomes also revealed significant coenrichment of KRASQ61R and U2AF1I24T mutations. Experimentally, KRASQ61R mutation led to KRAS exon 3 skipping, which in turn could be rescued by expression of U2AF1I24T. Our findings provide evidence that splicing factor mutations can rescue splicing defects caused by oncogenic mutations in a dynamic process of cascading selection.

PURPOSE:Methylthioadenosine phosphorylase (MTAP) genomic loss is an emerging biomarker for PRMT5 and MAT2A inhibitors based on synthetic lethality. The MTAP gene is located on chromosome 9p21.3 near CDKN2A/B. MTAP homozygous loss across tumor types, specific exons lost, MTAP expression, and the landscape of coalterations were assessed. METHODS:409,755 tissue biopsies (TBx) and 85,801 liquid biopsies (LBx) were sequenced using hybrid capture-based NGS (FoundationOne CDx or FoundationOne Liquid CDx), evaluating all classes of genomic alterations and circulating tumor DNA (ctDNA) tumor fraction (TF). Additionally, 6,740 TBx were subjected to both RNA and DNA sequencing to compare MTAP gene expression with genomic MTAPloss. RESULTS:The prevalence of pan-tumor MTAPloss was 11.4% using TBx, with the highest prevalence observed in mesothelioma (32.8%), glioma (32.7%), pancreatic (28.9%), and bladder cancers (26.4%). Prevalence of MTAPloss on LBx approximated that of TBx when ctDNA TF ≥20% (positive percent agreement [PPA] = 86.2%). CDKN2Aloss was not a reliable proxy for MTAPloss; 33.7% of CDKN2Aloss cases did not have MTAPloss. MTAPloss was enriched with CDKN2A/B losses pan-tumor, EGFR in NSCLC, ERBB2 in colorectal cancer, and PTEN in prostate cancer, while MTAPno_loss was enriched for RB1 pan-tumor, APC in colorectal cancer, CCNE1 in breast and ovarian cancer, and SPOP in prostate cancer. Complete loss (exons 1-8) was observed in 82.9%, multiple exons in 16.5% and exon 8 alone in <1%. MTAPloss, specifically the number of exons lost, was correlated with reduced RNA expression. CONCLUSION:MTAPloss is a frequent pan-tumor alteration that predicts potential sensitivity to PRMT5 or MAT2A inhibitors. Although most tumors exhibit complete MTAPloss, 16.5% are characterized by partial MTAPloss where clinical benefit remains uncertain. The genomic coalteration landscape reported here may inform future PRMT5 or MAT2A clinical trials, potentially in combination with other agents.

MET exon 14 skipping is a pathogenic event that results in decreased ubiquitin-mediated degradation of the MET receptor, sustained oncogenic signaling, and conferred sensitivity to MET tyrosine kinase inhibitors. While exon 14 skipping is most commonly caused by somatically acquired base substitutions and small indels near the exon 14 splice sites, here we report nine cases in which long interspersed element-1 (LINE-1, L1)-mediated insertions within or adjacent to MET exon 14, including one case of a LINE-1-mediated pseudogene insertion, appear to cause exon 14 skipping. These describe the first recurrent and clinically actionable mutations caused by LINE-1 retrotransposition in cancer.

Table S4. Imatinib IC50 values and 95% confidence intervals (CIs) of PDGFRA mutations modeled in Ba/F3 and CHO cells.

Abstract Background: Breast cancer is generally considered to be immune cold; however, emerging data indicate heterogeneity in its immune landscape. Triple negative and HER2-positive tumors typically show higher immune cell infiltration than hormone receptor-positive (HR+) tumors. Additionally, the local microenvironment at the biopsy site may influence the immune profile of the tumor, contributing to distinct immunophenotypes. In this study, we used gene expression profiling to evaluate biopsy site-specific patterns of immune activity in breast cancer. Methods: Tumor tissue from 1,009 patients with breast cancer (all-comers) underwent targeted RNA profiling using a laboratory-developed test, FoundationOne®RNA, as part of routine clinical care. Scores for immune and stromal gene sets were generated using a single-sample gene set enrichment analysis (ssGSEA) using a research use algorithm. K-means clustering of Z-normalized scores identified two immunophenotypes: immune-high and immune-low. Samples with a negative silhouette width were classified as unknown phenotype. Results: The most common sample sites were breast (n=363), liver (n=142), lymph node (n=120), bone (n=81), and lung (n=55). Immune ssGSEA scores varied significantly by sample site. Brain biopsies (n=23) exhibited the lowest immune scores, with only 5% (1/23) classified as immune-high, whereas lymph node and lung biopsies showed higher immune scores, with 42% (51/120) and 47% (26/55) classified as immune-high, respectively (p<0.05). Among local breast samples, 26% (96/363) were immune-high. Although liver metastases are typically considered to be immune cold, 12% (17/142) in this cohort were immune-high. Consistent with reports of lower immune activity in HR+ tumors, ESR1 expression was significantly higher in immune-low tumors compared to immune-high tumors (median expression 4,872 vs. 986 transcripts per million; p <10-4). Furthermore, distinct immune cell populations and gene signatures also showed site-specific patterns. IFN-Ɣ scores were above median in 60% of breast biopsies, compared with 17% of brain, 39% of liver, and 62% of lung metastases. T-cell and CD8+ T-cell scores were typically higher in lung (65%, 62% above median) and lymph node metastases (66%, 62%), intermediate in breast biopsies (53%, 56%), and lowest in liver (32%, 37%) and brain metastases (17%, 22%). In contrast, dendritic cell scores were highest in liver metastases (83% above median), compared with breast samples (48%). Conclusion: Immunophenotypes in breast cancer vary substantially by sample site, reflecting-heterogeneity of the local microenvironment. These site-specific immune signatures may have important implications for immunotherapy selection, prognostic assessment, and vaccine development strategies. Citation Format: Smruthy Sivakumar, Ericka Ebot, Douglas I. Lin, Meagan Montesion, Jeffrey S. Ross, Lee A. Albacker, Garrett M. Frampton, Michelle Marie Williams, Ethan S. Sokol. Biopsy site-specific variation in immune signatures identified by RNA profiling in breast cancer [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2026; Part 1 (Regular Abstracts); 2026 Apr 17-22; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2026;86(7 Suppl):Abstract nr 6468.

Accurate classification of BRCA1 and BRCA2 variants is essential for cancer risk assessment and therapy selection, yet over one-third remain variants of uncertain significance (VUS). Here, using 120,660 real-world cancer genomic profiles with BRCA1 or BRCA2 variants from a >800,000-sample cohort, we develop machine learning models that predict pathogenicity using clinical and tumor-derived features, including a pan-cancer homologous recombination deficiency signature, co-mutated genes, zygosity, and cancer type. Trained on classified variants from ClinVar, our models achieved near-perfect performance, with validation ROC-AUC of 1.000 for BRCA1 and 0.989 for BRCA2 variants with ≥5 observations, translating to strong benign or pathogenic evidence for VCEP classification. Applying these models to 1,073 BRCA1 and 1,639 BRCA2 VUS, we strengthened or enabled classification of 39.48% BRCA1 and 50.52% BRCA2 assessable variants. This approach transforms underutilized tumor profiling data into evidence that can be directly integrated into variant classification, providing a scalable framework for other tumor profiling datasets and cancer genes associated with defined tumor genomic features.

List of somatic SNV/Indel present in the concordance study samples and reported by at least one assay.

INTRODUCTION:MET amplification (METamp) can be a de novo or acquired resistance driver; however, the definition of METamp that best captures patients who may respond to targeted therapy remains debated. We explored the genomic landscape of METamp NSCLC including degree of amplification, co-drivers, amplicon size, and outcomes to MET inhibitors. METHODS:Hybrid-capture NGS-based genomic profiling from 88,547 tissue and 12,428 liquid NSCLC samples were queried for METamp (copy number (CN) ≥ ploidy + 4, or amplification ratio (AmpRatio; [CN/sample ploidy] ≥ 3). A nationwide de-identified real-world (rw) clinico-genomic database (CGDB) of NGS results linked to deidentified, electronic health record-derived clinical data was used to assess treatment and outcomes. RESULTS:Among 10,760 evaluable patients in CGDB, 362 (3.4%) had a METamp. In targeted therapy-naïve cases, MET AmpRatio negatively correlated with non-METex14 co-drivers (median 4.1 vs 2.9, p < 0.0001). MET AmpRatio was not significantly correlated with tumor mutational burden (p = 0.79) but was inversely correlated with amplicon size (p < 0.001). Among paired METamp tissue samples, 8/30 had METamp detected in liquid; higher tumor fraction and AmpRatio were associated with liquid detection. Among 39 METamp patients receiving MET inhibitors, longer median real-world progression free survival was observed with MET AmpRatio ≥ 3 vs < 3 (4.9 vs. 1.7mos, HR 0.53 [95 %CI:0.21-1.3]). CONCLUSIONS:MET AmpRatio positively correlated with focal amplification and absence of co-drivers and trended with increased benefit from MET inhibitors. Further studies evaluatingcombinatorial data including MET AmpRatio, amplicon size and presence of other potential drivers, as predictive biomarkers for therapies targeting MET amplification in NSCLC are warranted.

Summary and list of detected and reported alterations by alteration type (CNV, SNV/Indel) and assay (both assays, Central Lab (CL) or Assay H only, Pilot Lab only). For both detected and reported variants, positive percent agreement (PPA) with Pilot Lab and then CL as reference along with average percent agreement (APA), which is a weighted average of each PPA, were calculated.

Supplementary Table S1: First-, second-, and third-line systemic therapy in the SCLC clinical cohort. Supplementary Table S2: Genes analyzed as part of this study. Supplementary Table S3: Prevalence of alterations in the genes analyzed. Supplementary Table S4: Prevalence of gene alterations in 81 cases with liquid biopsies. Supplementary Table S5: Chromosomal losses and gains in SCLC. Supplementary Table S6: TET2 loss of function alterations detected in SCLCs. Supplementary Table S7: Manual review of specific mutation subgroups in SCLCs. Supplementary Table S8: Patterns of gene alterations in SMARCA4 altered and wild-type tumors. Supplementary Table S9: Prevalence of gene alterations in rearrangement-positive SCLCs. Supplementary Table S10: Univariate analysis of association between gene alterations and overall surival in the clinical cohort of stage III/IV SCLC tumors. Supplementary Table S11: Patterns of tumor mutational burden based on site of biopsy. Supplementary Table 12: Summary of tumor purity and tumor mutational burden by biopsy site. Supplementary Table S13: Aneuploidy patterns based on site of biopsy. Supplementary Table S14: Comparison of gene alterations in each biopsy site against lung biopsies. Supplementary Table S15: Comparison of gene alterations in liver and brain metastases. Supplementary Table S16: Summary of PTEN alterations in brain and other biopsy sites. Supplementary Table S17: Patterns of co-occurrence and mutual exclusivity between gene alterations in the overall SCLC cohort. Supplementary Table S18: Patterns of gene alterations in TP53 and/or RB1 mutant and wild-type cohorts. Supplementary Table S19: Patterns of gene alterations in young and older patients with SCLC. Supplementary Table S20: Patterns of gene alterations in HPV positive and HPV negative tumors. Supplementary Table S21: Manual review of specific HPV+ SCLCs. Supplementary Table S22: Patterns of gene alterations in STK11 mutant and wild-type cohorts. Supplementary Table S23: Patterns of gene alterations in EGFR mutant and wild-type cohorts. Supplementary Table S24: EGFR kinase domain alterations identified in the overall SCLC cohort.

Summary of detected and reported CNV alterations by assay (both assays, Central Lab (CL) or assay H only, Designated Lab (DL) assay only). For both detected and reported variants, positive percent agreement (PPA) with DL and then CL as reference along with average percent agreement (APA), which is a weighted average of each PPA, were calculated.

Scatter plots of continuous CN results obtained by NGS Central Lab (CL) or assay H, or NGS assay U and non-NGS techniques: A) ddPCR, B) microarray, C) FISH (Raw values) and D) FISH (Ratio). The solid line demonstrates a Deming regression line, while the dashed line demonstrates the identity line.

Genetic similarity of populations (or genetic ancestry) is associated with differences in somatic alterations in cancers. We meta-analyze two targeted panel sequencing cohorts with 275,605 samples from 14 cancer types. Here we find a recurrent depletion of TERT promoter mutations in patients of African and East Asian ancestry across multiple cancers. Several clinically actionable alterations, such as ERBB2 mutations in lung adenocarcinoma and MET mutations in papillary renal cell carcinoma, occur at a higher frequency in patients of non-European ancestry. Furthermore, in both cohorts, we show depletions in total driver alterations in non-European ancestries in multiple cancer types, potentially reflecting biases in current panel-based testing that prioritize established targets derived from predominantly patients of European ancestry. Our study highlights a need to increase population diversity in genomic studies to find new drivers and enhance precision oncology interventions for all populations.

Figure S1: Patterns of tumor mutational burden across different ancestry groups. Figure S2: Spectrum of TP53 and RB1 mutations in SCLC. Figure S3: Gene alteration landscape detected from targeted sequencing of liquid biopsies. Figure S4: Chromosomal loss events are frequent in SCLC and are preferentially enriched for tumor suppressor genes. Figure S5: Hematoxylin & Eosin (H&E) staining for some representative SCLC tumors with rare genomic alterations. Figure S6: Gene alterations detected in SMARCA4-altered SCLC. Figure S7: SCLC tumors exhibit recurrent gene rearrangements with therapeutic potential. Figure S8: Trends of overall survival across different ancestry groups and stage in the clinical cohort. Figure S9: Examination of tumor purity and tumor mutational burden (TMB). Figure S10: Liver and Brain SCLC metastases are enriched for chromosomal arm-level gains compared to lung-biopsied SCLC tumors. Figure S11: SCLC tumors present with unique co-occurring and mutual exclusive patterns of gene alterations. Figure S12: SCLC tumors with wild-type TP53 and/or RB1 have a low tumor mutational burden. Figure S13: SCLC tumors with wild-type TP53 and/or RB1 present with specific gene alterations. Figure S14: A smoking-associated mutational signature is rare in TP53/RB1 WT SCLC tumors. Figure S15: TP53 and RB1 alterations are less frequent in never smokers. Figure S16: Patterns of overall survival based on TP53 and RB1 alteration status. Figure S17: TP53 and RB1 mutations are more frequent in older patients with SCLC. Figure S18: Gene alterations detected in HPV-positive SCLC. Figure S19: Examination of HPV subtypes and prevalence in TP53/RB1 mutant and wild-type SCLC tumors. Figure S20: EGFR-mutant SCLC tumors exhibit unique mutational signature profiles. Figure S21: No overall survival differences based on EGFR mutation status in SCLC. Figure S22: Putative transformed SCLC tumors from EGFR-mutant NSCLC are enriched for EGFR exon 19 deletions.