No abstract

No abstract

3125 Background: MAP2K1 (MEK1) mutations are potentially actionable driver mutations in cancer. MAP2K1 mutations can be classified into 3 classes according to molecular characteristics. The efficacy of MAPK inhibitors (MAPKi) for the treatment of MAP2K1 mutant tumors is not well understood. We sought to characterize the genomic and clinical landscape of MAP2K1 mutant tumors, and to evaluate the relationship between MAP2K1 mutation class and clinical activity of MAPKi in patients with MAP2K1 mutant metastatic solid tumors. Methods: We interrogated AACR GENIE (v13) to identify all tumors with Class 1/2/3 MAP2K1 mutant solid tumors. We performed a systematic review and meta-analysis of individual patient data from patients with MAP2K1 mutant cancer published between 2010-22. Key inclusion criteria were: MAP2K1 mutation, solid tumor, metastatic disease, treatment with MAPKi and available treatment response data. The primary endpoint was progression-free survival (PFS) and the secondary endpoints were overall response rate (RR) and duration of response (DOR). Chi-squared and Log-Rank tests were used to evaluate statistical significance of differences between groups. Results: MAP2K1 driver mutations were present in 917/167,423 (0.5%) tumors in the AACR GENIE dataset. MAP2K1 mutants were most commonly identified in melanoma, colorectal (CRC) and non-small cell lung cancer (NSCLC). In solid tumors, Class 2 mutations were the most prevalent (n=310, 63%) followed by Class 1 (n=119, 24%) and Class 3 (n=66, 13%). Co-occurring MAPK pathway activating mutations (KRAS, NRAS, HRAS, NF1, BRAF, RAF1, or EGFR) were significantly more likely (P<0.0001) to occur in Class 1 (82.3%), versus Class 2 (30.9%) or Class 3 (10.6%) MAP2K1 mutant tumors. We identified 55 patients with MAP2K1 mutant tumors who received MAPKi (n=16/30/6/3 for Class 1/2/3/unclassified, respectively). Of these, (n=22, 18, 12, 3) had melanoma, CRC, NSCLC, or other cancers, respectively. Patients were treated with BRAFi (n=12), MEKi (n=24), BRAF+MEKi (n=2), ERKi (n=1) or EGFRi (n=16). Co-occurring MAPK pathway mutations were present in 51% of tumors. In the entire cohort, the RR was 24% and median PFS was 3.3 months. The RR did not differ according to mutation class, cancer type or MAPKi regimen. However, patients with Class 2 mutations experienced longer PFS (4.0 months) and DOR (23.8 months) compared to patients with Class 1, 3 or unclassified MAP2K1 mutations (PFS 3.0 months, P=0.035; DOR 4.2 months, P=0.04). Conclusions: Class 2 MAP2K1 mutations are RAF-regulated oncogenic mutations with a relatively low incidence of co-occurring MAPK pathway activating mutations. Some patients with Class 2 MAP2K1 mutations may derive durable therapeutic benefit from MAPKi. Prospective clinical studies with MAPK inhibitors are warranted in patients with MAP2K1-mutated metastatic cancer.

Interstrand DNA crosslinks (ICLs) represent complex lesions that block essential biological processes, including DNA replication, recombination, and transcription. Several pathways have been involved in ICL repair, in particular nucleotide excision repair (NER), translesion DNA synthesis (TLS), Fanconi anemia (FA), and homologous recombination (HR). Still, the extent of factors involved in the resolution of ICL-induced DNA double-strand breaks (DSBs) remains poorly defined. Using CRISPR-based genome-wide screening, we identified the poorly characterized C1orf112 (also known as Apolo1) as a novel sensitizer to the clinically relevant ICL-inducing agent mafosfamide. Consistently, we noted that low expression of C1orf112 correlates with increased sensitivity to a series of ICL agents and PARP inhibitors in a panel of cell lines. We showed that lack of C1orf112 does not impact the initial recruitment and ubiquitylation of FANCD2 at the ICL site but rather impairs the resolution of RAD51 from ICL-induced DSBs, thereby compromising homology-directed DNA repair pathways. Our proximal mapping of C1orf112 protein neighbours coupled to structure-function analysis revealed that C1orf112, through its WCF motif, forms a complex with the N-terminal domain of the AAA+ ATPase FIGNL1 and regulates the interaction of FIGNL1 with RAD51. Our work establishes the C1orf112-FIGNL1 complex as an integral part of the HR-mediated response to ICLs by regulating the unloading of RAD51 during ICL repair.

2034 Background: Patients with HER2+ breast cancer (BC) frequently develop leptomeningeal metastases (LM). While HER2-targeted therapies have demonstrated efficacy in the neoadjuvant, adjuvant, and metastatic settings, including for parenchymal brain metastases, their efficacy for patients with LM has not been studied in a randomized controlled trial. However, several single-armed prospective studies, case series and case reports have studied oral, intravenous (IV), or intrathecally (IT) administered HER2-targeted therapy regimens for patients with HER2+ BCLM. Methods: We conducted a systematic review and meta-analysis of individual patient data to evaluate the efficacy of HER2-targeted therapies in HER2+ BCLM in accordance with PRISMA guidelines. Targeted therapies evaluated were trastuzumab (IT or IV), pertuzumab, lapatinib, neratinib, tucatinib, trastuzumab-emtansine (T-DM1) and trastuzumab-deruxtecan (TDXd). The primary endpoint was overall survival (OS), with progression-free survival (PFS) as a secondary endpoint. To assess differences between groups, shared frailty Cox regression models were used to estimate the hazard ratio (HR), 95% confidence interval (CI) and p-value. Results: 7780 abstracts were screened, identifying 44 publications with 200 patients, corresponding to 257 lines of HER2-targeted therapy for BC LM which met inclusion criteria. In univariable (OS: HR=0.9, 95% CI: 0.64-1.4, P=0.76; PFS: HR=0.8, 95% CI: 0.57-1.2, P=0.35) and multivariable (OS: HR=1.4, 95% CI: 0.68-3.1, P=0.4; PFS: HR=0.8, 95% CI: 0.41-1.7, P=0.6) analyses, we observed no significant difference between IT, oral or IV administration of HER2-targeted therapy. Meanwhile, ECOG performance status remained independently associated with prolonged OS (HR=2.2, 95% CI: 1.5-3.2, P<0.001) and PFS (HR=2.2, 95% CI: 1.5-3.2, P<0.001 and HR=1.9, 95% CI: 1.4-2.8, P<0.001) in the final multivariable model. ECOG status was not associated with route of trastuzumab delivery (P>0.40). Anti-HER2 monoclonal antibody-based regimens did not demonstrate superiority over HER2 tyrosine kinase inhibitors (OS: P=0.647; PFS: P=0.983). In a cohort of 7 patients, TDXd demonstrated improved OS compared to other HER2-targeted therapies and compared to T-DM1 (P<0.05). Conclusions: The results of this meta-analysis suggest that IT administration of HER2-targeted therapy for patients with HER2+ BCLM confers no additional benefit over oral and/or IV treatment regimens. We also present the first evidence supporting the efficacy of TDXd compared to alternative strategies for this patient population. Although the number of patients receiving TDXd in this cohort is small, this novel agent offers promise for this patient population and requires further investigation in prospective studies.

Background: An estimated 20-40% of cancer patients develop brain metastases (BrM), mostly those affected by lung cancer, breast cancer or melanoma. Unfortunately, these patients suffer from poor outcomes and diminished quality of life. Few BrM treatment options beyond local therapy exist and this is often a short-term solution as 60% of resected BrM recur within 1 year. Our group discovered that patient prognosis is linked to BrM invasiveness, with highly invasive (HI) BrM more likely to recur, compared to minimally invasive (MI) BrM. This has presented us the opportunity to investigate exploitable biological mechanisms driving HI BrM. For instance, invasive signaling can be driven by microenvironmental proteins such as growth and immunological factors (cytokines) secreted by surrounding brain or cancer cells, through inter-cellular or self-feeding autocrine loops. Hypothesis: Considering the influence of secreted factors on cancer invasion and the brain microenvironment, I hypothesize that secretory profiling of BrM and brain parenchymal cells will be mechanistically insightful and help identify potential targetable drivers of BrM invasion. Results: To identify BrM invasion-related tumor- and brain-derived factors, I performed human- and mouse-specific high throughput Enzyme-Linked Immunosorbent Assay (ELISA)-based screens on conditioned media from mouse brain slices harboring intracranial MI or HI BrM patient-derived xenografts (PDX). This secretome screen reveals distinct MI and HI secretory profiles for melanoma, breast cancer and lung cancer BrMs and several HI BrM-derived factors of interest have been identified. Such factors are important to investigate as possible drivers of invasiveness in BrM through functional studies. Conclusion: BrM patients currently face a bleak prognosis, with few treatment options and a median survival of only 8-16 months. Considering the clinical availability of targeted therapies including inhibitors for growth factor-binding tyrosine kinases, antibody-drug conjugates, and immunotherapy, this project will help elucidate the cancer-brain crosstalk which may be exploited therapeutically with existing drugs in patients with frequently recurring HI BrM. ADDIN Citation Format: Caitlyn Mourcos, Sarah M. Maritan, Matthew G. Annis, Georgia Kruck, Alexander Nowakowski, Anna-Maria Lazaratos, Kevin Petrecca, Peter Siegel. Targeting tumor-brain crosstalk in invasive brain metastases. [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 3615.

INTRODUCTION LM is a debilitating condition associated with metastatic cancers, including BC. When oncogenic drivers are identified, TT represents an appealing therapeutic strategy. However, the efficacy of TT for LM is unknown as LM patients are routinely omitted from clinical trials. METHODS We conducted a systematic review and meta-analysis of individual patient data to evaluate the effectiveness of HER2-TT in HER2+ BCLM in accordance with PRISMA guidelines. TTs evaluated included trastuzumab (intrathecal (IT) or intravenous (IV)), trastuzumab-emtansine, trastuzumab-deruxtecan, and lapatinib. Primary outcome was overall survival (OS). RESULTS Of 7780 abstracts screened, 91 publications and a total of 109 patients were included in the final analysis. Patients receiving chemotherapy (either IT, IV, or as part of an antibody-drug conjugate) concurrently with HER2-TT (N=57) exhibited a median OS (mOS) of 44.0 months, compared to patients treated with targeted anti-HER2 therapies alone (N=52), which exhibited a mOS of 14.5 months (P=0.009, hazard ratio (HR): 0.538, 95% confidence interval (CI): 0.328-0.883). Patients receiving IT trastuzumab (N=83) exhibited a median progression-free survival (mPFS) and mOS of 6.0 and 21.0 months, respectively, while patients receiving IV trastuzumab (N=14) exhibited a mPFS and mOS of 6.5 and 27.0 months, respectively (PFS: P=0.31, HR: 0.712, 95% CI: 0.331-1.531; OS: P=0.68, HR: 1.154, 95% CI: 0.587-2.266). In the subgroup of patients receiving IT trastuzumab (N=58), those who concurrently received IT chemotherapy (N=48) exhibited mPFS and mOS of 5.7 and 14.0 months, respectively, while patients concurrently receiving IV chemotherapy (N=10) exhibited a mPFS and mOS of 6.0 and 27.0 months, respectively (PFS: P=0.45, HR: 1.360, 95% CI: 0.602-3.073; OS P=0.29, HR: 1.821, 95% CI: 0.630-5.260). CONCLUSIONS HER2-TT is an effective therapeutic strategy for BCLM. Patients with BCLM receiving concurrent cytotoxic chemotherapy alongside HER2-TT experience prolonged mOS. IV and IT trastuzumab are similarly effective. Univariate and multivariate analyses will be presented.

e21118 Background: Leptomeningeal metastasis (LM) is a debilitating condition associated with advanced non-small cell lung cancer (NSCLC). When driver mutations in the epidermal growth factor receptor (EGFR) are identified, targeted therapies (TT) represent an appealing therapeutic strategy. However, the efficacy of TT for LM is unknown as LM patients are routinely omitted from clinical trials. Methods: We conducted a systematic review and meta-analysis of individual patient data to evaluate the efficacy of EGFR TT in EGFR-mutant NSCLC LM in accordance with PRISMA guidelines. EGFR-TTs evaluated were erlotinib, gefitinib, icotinib, afatinib, dacomitinib, osimertinib, zorifertinib, and furmonertinib. The co-primary endpoints were progression-free survival (PFS) and overall survival (OS). To assess differences between groups, shared frailty Cox regression models were used to estimate the hazard ratio (HR), 95% confidence interval (CI) and p-value. Results: 7780 abstracts were screened, identifying 126 publications with 232 patients, corresponding to 272 lines of EGFR-TT for NSCLC LM which met inclusion criteria. The overall median PFS (mPFS) in patients receiving any EGFR-TT was 9.15 months and the median overall survival was 14.5 months. In univariable analyses, osimertinib was associated prolonged PFS compared to other EGFR-TT, (mPFS of 11 versus 8 months, HR = 0.64; 95% CI: 0.44-0.93; P = 0.02). Osimertinib was numerically associated with prolonged OS compared to other EGFR-TT (mOS 16.8 months versus 13.6 months, HR = 0.72; 95% CI 0.44-1.17, P = 0.188). In multivariable analyses, ECOG performance status > 2 was independently associated with shortened PFS (mPFS 10 versus 8 months, HR = 1.54, 95% CI: 1.26-3.04; P < 0.05) and OS (mOS 16.5 versus 10 months, HR = 3.33, 95% CI: 1.80-6.18; P < 0.05). Conclusions: In the largest cohort of NSCLC LM treated with EGFR TT compiled to date, osimertinib is associated with marginally improved outcomes compared to other EGFR TTs. ECOG performance status is an independent predictor of prognosis in these patients. These results highlight the need for prospective studies for this difficult to treat patient population.