Editas Medicine (United States)

PIK3CA and KRAS are among the most frequently mutated oncogenes and often co-mutated in colorectal cancers. Understanding the impact of KRAS codon-specific mutations on cross-talks between the PI3K and MAPK pathways and response to targeted therapies, such as the p110α-specific inhibitor inavolisib (GDC-0077), is critical for advancing precision oncology. Focusing on colorectal PIK3CA + KRAS co-mutated models, we found that KRASG12D-mutated cells were more sensitive to inavolisib than models with KRASG13D, or other MAPK pathway mutations, even though the PI3K and MAPK pathways were active in both genotypes. In most co-mutated models, regardless of the type of KRAS alteration, the combination of inavolisib with MAPK pathway inhibitors showed synergy in vitro and in vivo. Our work highlights how specific codon substitutions in KRAS differentially toggle pathway activity and alter sensitivity to inavolisib, which could inform whether patients would benefit more from single-agent inavolisib or combination with MAPK pathway inhibitors.

Purpose:Many preclinical hypotheses, including reciprocal feedback activation between the androgen receptor (AR)–PI3K pathway in PTEN loss and AR signaling inhibitor–induced “BRCAness” regardless of BRCA1/2 status, have struggled to translate into clinical benefit for patients with metastatic prostate cancer. This gap in translatability, particularly in metastatic castration-resistant prostate cancer (mCRPC), may stem from a limited understanding of prostate cancer evolution. A key challenge is the correlation between early-stage tumor genetics and mCRPC. By examining clinical, genomic, and molecular changes over time, we aimed to refine clinical trial design.Experimental Design:Using a comprehensive dataset from electronic health records with genomic profiling, a shift in the prognostic value of biomarkers from metastatic hormone-sensitive prostate cancer (mHSPC) to mCRPC was observed. Additionally, genomic and transcriptomic analyses of primary tumors from the IPATential150 trial and mCRPC samples from the Stand Up To Cancer cohort examined the changing AR–PI3K signaling correlation.Results:PI3K–AKT pathway alterations lost their prognostic significance in later stages. Although AR and PI3K–AKT signaling were inversely correlated in primary tumors, this relationship was disrupted in mCRPC, independent of PIK3CA/AKT1/PTEN status. We identified broad transcriptional rewiring associated with AR signaling, increasing tumor heterogeneity. Improving the understanding of early-stage disease, we identified a high-risk mHSPC subset enriched for AR alterations. Additionally, in a subset of patients, AR ligand-binding domain mutations preceded amplification, potentially leading to preferential amplification of the mutant AR form.Conclusions:Our findings underscore the dynamic nature of prostate tumor biology and emphasize the need for translational research to validate preclinical hypotheses in clinically relevant settings, ultimately improving trial design and therapeutic strategies.

Not available.

BRAF mutations are rare in biliary tract cancers (BTC), but are of interest given the recent developments in targeted therapy for BTC. We investigated the clinical outcomes in a cohort of BRAF-mutant advanced BTC treated with first-line chemotherapy. Furthermore, we investigated the genomic landscape of BRAF class I, II, and III mutations in the intrahepatic cholangiocarcinoma (iCCA) subgroup of BTC. We analyzed two nonoverlapping cohorts. We examined the genomic landscape of BRAF-mutated iCCA in a "genomic cohort" [187 class I, 82 class II, 113 class III BRAF mutants and 8,026 wildtype (WT)]. We also analyzed median progression-free survival (PFS) and overall survival (OS) on first-line chemotherapy in a separate multi-institutional "clinical cohort" of patients with BTC (including iCCA and extrahepatic cholangiocarcinoma (eCCA) and gallbladder cancer; 41 class I, 32 class II+III BRAF mutants and 1,042 WT). In the entire BTC clinical cohort, the median PFS was shorter for class I [HR, 2.11 (P < 0.001)] and class II+III [HR, 1.72 (P = 0.007)] as compared with BRAF WT. OS was also shorter in class I [HR, 2.04 (P = 0.011)] and class II+III [HR, 1.86 (P = 0.002)] as compared with BRAF WT. In the iCCA subgroup, class I alterations were mutually exclusive with FGFR2, IDH1/2, ERBB2, and KRAS mutations. Class II+III mutations appear to be mutually exclusive with FGFR2 and KRAS. In BTC, all classes of BRAF mutations are associated with a worse prognosis. BRAF mutations occur in 5% of iCCA subgroup and may be mutually exclusive with other targetable mutations.

Abstract Background:Third generation EGFR TKIs have emerged as a first line treatment of choice for aNSCLC patients with qualifying EGFR mutations due to high and durable response rates; however most patients will progress on therapy and may receive further treatment. Second line (2L) options include other EGFR TKIs, chemotherapy (C) and chemoimmunotherapy (C-IO). An optimal treatment strategy and utility of biomarkers in this 2L setting remain uncharacterized. In this study, we aim to describe the real-world treatment patterns and outcomes of patients with aNSCLC after treatment with first line (1L) EGFR TKIs to elucidate the potential for biomarker-guided approaches. Methods:We selected patients with non-squamous aNSCLC treated with 1L EGFR TKIs on or after April 1, 2018 (based on 1L approval of osimertinib) within the US-based de-identified Flatiron Health-Foundation Medicine (FMI) real world clinico-genomic database. Real world overall survival (rwOS) and progression free survival (rwPFS) curves and estimates of median survival time were generated using the Kaplan-Meier method. PD-L1 and tumor mutational burden (TMB) were summarized overall and by 2L treatment regimen. Results:Among 428 patients who received 1L EGFR TKI treatment, 386 (90%) received osimertinib, 30 (7%) received afatinib and 11 (3%) received erlotinib. For patients with known 2L treatment information (n=139, 32%) most received EGFR TKI alone (n=43, 31%), followed by C-IO (n=38, 27%), and C alone (n=14, 10%). Median rwOS following 2L treatment with EGFR TKI, C-IO, and C was 21.0 (95% CI 17.4-28.7) months, 9.9 (95% CI 7.7-19.8) months, and 17.4 (95% CI 4.3-NA) months respectively. Median rwPFS for the EGFR-TKI, C-IO, and C cohorts was 5.1 (95% CI 3.4-12.5) months, 5.5 (95% CI 3.3-7.0) months, and 3.5 (95% CI 1.9-NA) months. 416 (97%) of 428 patients harbored an EGFR mutation detected by FoundationOne®CDx, including 361 (84%) patients with at least one EGFR mutation in NCCN Guidelines. Among patients with available PD-L1 status (n=254, 59%), PD-L1 positivity (≥1% tumor proportion score) was more common among patients who received 2L C-IO (88%) compared to patients overall (59%). 30 (7%) of 428 patients had high TMB (≥10 mut/Mb); prevalence of high TMB was similar regardless of 2L therapy choice. Conclusions:This real-world data analysis confirmed that among patients with aNSCLC treated with EGFR TKIs, the vast majority received 1L osimertinib. Among 2L treatment options, additional EGFR TKI use was the most common and patients in this cohort had robust outcomes, with a median rwOS of 21 mos. Further analyses will describe the clinical characteristics of patients receiving different 2L treatment strategies and evaluate the association of PD-L1 status and select EGFR mutations with outcomes across therapeutic classes. Citation Format: Ericka Ebot, Jie He, David Fabrizio, Ivy Altomare, Neha Jain, Davey Daniel, Thomas Stricker, Edward Arrowsmith. Real world treatment patterns and outcomes in patients with advanced non-small cell lung cancer (aNSCLC) post-EGFR tyrosine kinase inhibitor (TKI) therapy [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 933.

Abstract Tumor mutational burden (TMB) derived from tissue biopsies (tTMB) has been associated with clinical efficacy in patients treated with immune checkpoint inhibitors; the clinical utility of TMB derived from blood samples (bTMB) has also been demonstrated but is less widely studied. Little is known about the concordance between tissue- and blood-derived genomic assessments and the factors that contribute to their discordance, underlying the need for further investigation of sequence alteration profiles for successful adoption of noninvasive tumor profiling. We explored the genomic landscape, including concordance between tTMB and bTMB, in samples from patients who were screened for enrollment into CheckMate 848, a prospective, phase 2 study of nivolumab plus ipilimumab and nivolumab monotherapy in patients with advanced or metastatic solid tumors with high TMB (NCT03668119). Of 1954 screened patients, 212 were randomized, with a cap of 15% per tumor type. Tissue- (FoundationOne® CDx—based Clinical Trial Assay) and blood-based (Foundation Medicine bTMB Clinical Trial Assay) genomic data were utilized for analysis of genomic variants, tTMB, and bTMB. In total, 1438 unique tissue and 1720 unique plasma samples were analyzed during trial screening (June 2021 database lock). Of over 100 screened disease ontologies (taken from the tissue diagnosis), pancreatic (9.7%), breast (8.8%), and ovarian (6.1%) cancers, as well as cholangiocarcinoma (5.2%), were the most common. A total of 1141 tissue and 1573 plasma samples passed established quality control criteria, resulting in ascertainment levels of 79.3% for tTMB and 91.5% for bTMB. A correlation between tTMB and bTMB scores was identified across 1017 tissue and plasma sample pairs (Spearman’s r, 0.48; P &amp;lt; 0.0001). Median (range) tTMB and bTMB were 3.8 (0—452.6) and 3.5 (0—1027.5) mutations per megabase (mut/Mb), respectively. High microsatellite instability (MSI) was detected in 25 (2.5%) MSI-evaluable tissue samples; in these patients, median tTMB was 25.2 mut/Mb. At the prespecified cutoff of 10 mut/Mb, 15.8% and 20.7% of samples had high tTMB and bTMB, respectively; the positive (PPA), negative, and overall percentage agreements between assays were 60%, 88%, and 84%, respectively. TMB correlation (Spearman’s r, 0.54; P &amp;lt; 0.0001) and PPA (66%) were improved among 806 (79.3%) sample pairs with plasma maximum somatic allele frequency ≥ 1%. In CheckMate 848, data from paired biopsies revealed the complementary nature of TMB assessments from tissue and blood, suggesting that both approaches may have the potential to identify high mutational burden in samples obtained from patients with advanced solid tumors. Further interrogation of the biological and analytical factors affecting tumor- and blood-derived genomic profiling is warranted to support their implementation in clinical settings. Citation Format: Jie He, Natallia Kalinava, Parul Doshi, Jie Ma, Dean C. Pavlick, Lee A. Albacker, Hanna Tukachinsky, Gina Fusaro, Geoffrey R. Oxnard, George Green, David Fabrizio, Jonathan Baden. Evaluation of tissue- and plasma-derived tumor mutational burden and genomic alterations of interest from the CheckMate 848 clinical trial [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr 2139.

Abstract Background: Genomic profiling with tissue sequencing is still considered as the gold standard despite several limitations including screening failures due to limited tissue availability, and inability to capture intratumor spatial and temporal heterogeneity, which may impair accurate treatment selection. Several studies have demonstrated the potential of circulating tumor DNA (ctDNA) to detect genomic alterations at high accuracy compared with tissue analysis. However, no studies have comprehensively evaluated differences between tissue and ctDNA by using a large panel in the same cohort. Methods: Genomic analysis was performed for each patient by using the Foundation One Liquid CDx Assay and the Foundation One CDx Assay (324 genes, tumor mutational burden [TMB], microsatellite instability). Each individual genomic report was reviewed and discussed weekly by a multidisciplinary tumor board (MTB). Actionable targets were defined by the MTB according to the existing level of evidence (ESCAT tiers) and molecular-based treatment suggestions were proposed where possible. Results: Between Dec 2020 and Nov 2021, 1021 patients (median age: 62 years) with advanced cancer underwent both tissue and ctDNA NGS. Five most frequent tumor types were colorectal (N=137,13%), NSCLC (N=130,13%), breast (N=120, 12%), prostate (N=82, 8%) and pancreas (N=65, 6%). Median time elapsed between request and assay results was 12 days for ctDNA and 46 days for tissue. Testing failure was 15% for tissue and 3.9% for ctDNA. Overall, 824 (81%) patients had evaluable results for both tissue and liquid. Total number of cancer-related alterations and variants of unknown significance were 4704 and 11673 vs 4645 and 7481 for ctDNA and tissue, respectively. Proportion of patients with a higher number of cancer alterations identified in ctDNA compared with tissue increased in parallel with the time elapsed between the tissue and ctDNA sampling (45% vs 33% for a delay &amp;gt; 26 months or &amp;lt; 8 months). MSI and TMB status were concordant for 71% and 64% of patients, respectively. MSI status was evaluable for 97% of patients through ctDNA vs 90% through tissue. Number of actionable alterations was similar in 346 (42%) of cases, whereas it was higher in tissue for 289 (35%) and in liquid for 189 (23%) patients. ctDNA profiling allowed the identification of an ESCAT I/II or III or IV alteration not present in tissue for 74 (9%), 113 (14%) and 52 (6%) patients, respectively. Overall, MTB recommended a matched therapy for 430 patients (52%). Such a recommendation would not have been made without the results of ctDNA for 120 patients (15%). Conclusion: This systematic comparison of ctDNA vs tissue sequencing demonstrates the capacity of ctDNA for capturing clinically relevant alterations to guide therapy in cancer patients with high accuracy and rapid turnover results. Citation Format: Arnaud Bayle, Florent Peyraud, Laila Belcaid, Maxime Brunet, Miha Aldea, Rebecca Clodion, Paul Dubos, Damien Vasseur, Claudio Nicottra, Santiago Ponce, Isabelle Soubeyran, Emmanuel Khalifa, Yohann Loriot, Benjamin Besse, Ludovic Lacroix, Etienne Rouleau, Geoffrey Oxnard, Fabrice Barlesi, Fabrice Andre, Antoine Italiano. Systematic comparison of ctDNA vs tissue sequencing with a large panel to guide therapy in patients with advanced cancer: A study from the French National Center for Precision Medicine (PRISM) [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr 3414.

Purpose:PARP inhibitors (PARPi) have demonstrated efficacy in tumors with germline breast cancer susceptibility genes (gBRCA) 1 and 2 mutations, but further factors influencing response to PARPi are poorly understood.Experimental Design:Breast cancer tumor tissue from patients with gBRCA1/2 mutations from the phase III EMBRACA trial of the PARPi talazoparib versus chemotherapy was sequenced using FoundationOne CDx.Results:In the evaluable intent-to-treat population, 96.1% (296/308) had ≥1 tumor BRCA (tBRCA) mutation and there was strong concordance (95.3%) between tBRCA and gBRCA mutational status. Genetic/genomic characteristics including BRCA loss of heterozygosity (LOH; identified in 82.6% of evaluable patients), DNA damage response (DDR) gene mutational burden, and tumor homologous recombination deficiency [assessed by genomic LOH (gLOH)] demonstrated no association with talazoparib efficacy.Conclusions:Overall, BRCA LOH status, DDR gene mutational burden, and gLOH were not associated with talazoparib efficacy; however, these conclusions are qualified by population heterogeneity and low patient numbers in some subgroups. Further investigation in larger patient populations is warranted.