Single-guide RNA (sgRNA) lentiviral infection with Cas9 protein electroporation (SLICE) enables CRISPR screening in primary cell types that require transient Cas9 expression, yet is limited by scalability and robustness. Here, we introduce dual guide RNA infection with Cas9 electroporation (DICE), which expresses two guides from the same lentiviral construct that target the same gene. In genome-wide screens, DICE outperformed SLICE in defining essential genes and modulators of PD-L1 expression in Interferon-gamma-activated THP1 cells. Collectively, these data demonstrate that DICE can be utilized for reduced-scale CRISPR screens in cell types with transient Cas9 protein expression without sacrificing screening quality.

Abstract Characterizing immune-related adverse events (irAEs) among patients with cancer receiving immunotherapy remains challenging. Traditional ICD-based identification often misses or misclassifies irAEs, and manual chart review is labor-intensive, error-prone, and not scalable. Existing large language model (LLM)-based approaches often do not fully capture irAE onset dates and severity information, frequently missing mild events and limiting their utility for time-sensitive and severity-stratified analyses.We developed a two-stage prompting strategy to extract irAE type, severity, and onset date from large volumes of unstructured clinical notes for patients receiving immunotherapy. In the first stage, the model summarizes irAE-related information from long and heterogeneous clinical documentation. In the second stage, it extracts structured irAE details. This approach improves extraction accuracy and supports more reliable identification of irAE types and onset timing. Using expert-curated ground truth, we evaluated multiple reliable and increasingly used large language models, including ChatGPT 4o, Llama 3.1 with 70B and 450B, Llama 3.3 with 70B, and DeepSeek R1. Across models, LLM-based methods achieved higher sensitivity (0.95 vs. 0.57), higher precision (0.69 vs. 0.64), and higher F1 scores (0.77 vs. 0.55) compared with ICD-based extraction. For cases in which both LLM and ICD identified an irAE, LLMs more accurately captured event types (83% vs. 38%) and consistently identified earlier onset dates (on average 27 days earlier). This strategy is model-agnostic and will continue to improve extraction accuracy as the underlying LLM models improve. We applied our pipeline, which incorporates ChatGPT 4o, to 340,277 notes from 8,768 patients treated with immunotherapy in the Dana-Farber Cancer Institute Profile cohort, covering 21 cancer types. 87% of patients developed at least mild irAEs within two years of treatment initiation. Consistent with prior work, CTLA-4 inhibitors were associated with higher irAE incidence (HR = 3.93; P = 1.94e-19), using Cox models that incorporate irAE onset timing, information typically unavailable in earlier studies. Survival analyses stratified by time-dependent irAE severity showed that mild irAEs were associated with reduced hazard of death (HR = 0.86; P = 5.64e-92), whereas severe irAEs were associated with an increased hazard of death (HR = 1.04; P = 1.15e-3). We also observed system-specific severe irAEs that aligned with underlying cancer type; for example, patients with non-small cell lung cancer had a higher instantaneous risk of severe respiratory irAEs (HR = 2.37; P = 1.29e-9) compared with other cancers. These findings demonstrate that LLM-based clinical text extraction enables scalable and accurate characterization of irAEs, providing deeper insights into their prognostic implications. Citation Format: Zeyun Lu, Mustafa Saleh, Charles Lu, Intae Moon, Razane El Hajj Chehade, Elio Ibrahim, Yevgeniy R. Semenov, Toni K. Choueiri, Alexander Gusev. LLM-based extraction of immunotherapy toxicities reveals severity-dependent effects on overall survival [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 2762.

Importance:Enfortumab vedotin (EV) is an antibody-drug conjugate approved for the treatment of locally advanced or metastatic urothelial cancer (la/mUC). Cutaneous adverse events (cAEs) are common during EV therapy, with prior studies suggesting an association between EV-related cAEs and improved survival; however, there is insufficient data to delineate the survival benefit of EV-induced cAEs from those associated with concurrent immune checkpoint inhibitors (ICIs). Objective:This study aims to evaluate the association of EV-induced cAEs and survival, and to characterize the timing and morphology of EV-induced cAEs. Design:We conducted a multi-institutional retrospective study of patients with la/mUC treated with EV between 2020 and 2025. Setting:Multicenter academic referral center. Participants:A total of 449 EV-treated patients were included. Patient characteristics were extracted manually, and likelihood scoring was used to attribute cAEs to either EV or other etiologies. Exposure:EV treatment. Main Outcomes and Measures:We estimated progression-free (PFS) and overall (OS) survival using Kaplan-Meier method. Multivariable time-varying and landmark Cox regression models were used to evaluate associations between EV-induced cAE and survival. Sensitivity analyses were performed at landmarks from 15 to 105 days. Results:Of 449 patients, 206 (45.9%) developed a cAE; 39 (18.9%) were high-grade and 127 (61.7%) were attributed to EV. The most common cAEs were pruritus (41.3%), unspecified and desquamating dermatitis (37.3%), and morbilliform dermatitis (27.7%). Across all treatment groups, survival was longer in patients with EV-induced cAEs. Developing an EV-induced cAE was protective across all examined landmark times, with hazard ratio (HR) 0.60 (95% CI: 0.43-0.82, p<0.001) for PFS and HR 0.46 (95% CI: 0.31-0.67, p<0.001) for OS at primary landmark time of 30 days. Early-onset EV-induced cAEs were protective at all landmark times and high-grade EV-induced cAEs were not associated with worse survival. Conclusions and Relevance:EV-induced cAEs were independently associated with improved PFS and OS in patients with la/mUC, even after accounting for immortal time bias and ICI exposure. Distinguishing EV-induced cAEs from other etiologies in timeline and morphology may help guide oncology and dermatology management.

9555 Background: Melanoma is the 4th most common cancer among US Veterans. Non-acral cutaneous melanoma has well-established risk factors including history of sunburn, ultraviolet (UV) exposure, fair skin and features. Risk factors for acral (AM), mucosal (MM), or uveal melanoma (UM), are less understood. Methods: We conducted a nested case-control study (2000-2024) identifying cases of acral (N=1,292, Veterans only), mucosal (N=520, combined Veteran and civilian cohort), and uveal (N=2,010, combined Veteran and civilian cohort) melanoma using the Veterans Affairs (VA) Cancer Registry, natural language processing applied to pathology reports, and Mass General Brigham Research Patient Data Repository. Results: For AM, current smoking was associated with reduced odds, while Agent Orange Exposure (AOE) was associated with increased odds compared to both control populations. Compared to non-melanoma controls, prior nevus, keratinocyte carcinoma (KC), and actinic keratosis (AK) were associated with increased odds of AM. For MM, prior KC was associated with higher odds of MM compared to non-melanoma controls, consistent across anogenital, sinonasal, and upper aerodigestive tract sites. Smoking status was associated with reduced MM odds in the sinonasal cohort. For UM, we conducted a case-control study using data from 3 academic medical centers and the VA. Prior cutaneous and systemic immune disease were protective of UM development compared to CM. Smoking status and prior nevi had higher odds of UM compared to CM and non-cancer controls. Conclusions: To our knowledge, this is the largest cohort of AM, MM, and UM cases to date. Our study identified unique factors associated with risk of, or protection from, AM, MM, or UM, compared to CM or non-cancer controls. These findings provide novel insights into potential risks for rare melanomas and can help guide future biologically driven studies. Risk and protective factor associations across rare melanomas. Melanoma type Factor type Both cohorts CM only Non-cancer only Acral High risk association factors Younger age, female sex, Black race, Hispanic ethnicity, AOE Asian race Prior nevus, prior KC, prior AK Protective factors Current smoking ≥2 comorbidities, AUDIT-C positive, prior KC, prior AK Former smoking No associations BMI Mucosal High risk association factors Younger age, female sex Black race, Hispanic ethnicity Prior KC Protective factors Former smoking Current smoking, ≥2 comorbidities No associations Prior nevus, BMI Uveal High risk association factors Older age, former smoking, current smoking, prior nevus Black race, American Indian/Alaska Native race, obesity Protective factors ≥1 comorbidities, prior photosensitizing medication use, prior cutaneous autoimmune disease, prior systemic autoimmune disease Prior non-melanoma skin cancer Hispanic ethnicity, normal or underweight BMI No associations Sex

Introduction Precision oncology aims to provide individualised treatment recommendations based on patient-specific characteristics. In this rapidly evolving field with increasing numbers of biomarkers and potential therapeutic targets, there is a growing unmet need for evidence guiding these individualised treatment recommendations. The Precision Oncology Program (POP) harnesses real-world data (RWD) and imaging mass cytometry (IMC) to evaluate the feasibility and utility of integrating different data modalities to inform personalised treatment recommendations. This program uses patient-matched clinicogenomic data and spatial single-cell proteomics analysis to support profiling-driven decision-making for patients with cancer at the Molecular Tumor Board.Methods and analysis The collaborative POP project recruits patients across all tumour entities and stages at the Comprehensive Cancer Center Zurich (CCCZ). For patients in the POP, a clinically and molecularly matched cohort is identified within the nationwide (US-based) de-identified Flatiron Health-Foundation Medicine clinicogenomic database (CGDB). It assesses whether clinical, genomic and outcome data of the CGDB cohort can inform treatment recommendations. In addition, multiplexed imaging mass cytometry (IMC) is performed in formalin-fixed paraffin-embedded tissue to assess the potential impact of spatial proteomics on personalised treatment decisions. RWD and IMC information is reviewed in the Molecular Tumor Board to assess the potential impact of this information on therapy decisions. However, since this is an observational study, these additional recommendations remain nonprescriptive and will not be forwarded to the treating physician.Ethics and dissemination The study is registered at ClinicalTrials.gov (NCT06680726) and approved by the Canton of Zurich Ethics Committee (Project ID: 2022-02289). Project-specific informed consent is obtained from all participants. Deceased patients may also be included. In this case, a signed general consent form must be available. Data privacy is ensured by unique patient numbers for pseudo-anonymised data. Study findings will be disseminated through international peer-reviewed journals, conferences, and direct communication with participants and relevant organisations.Trial registration number NCT06680726.

Synthetic lethality exploits the genetic vulnerabilities of cancer cells to enable a targeted, precision approach to treat cancer1. Over the past 15 years, synthetic lethal cancer target discovery approaches have led to clinical successes of PARP inhibitors2 and ushered several next-generation therapeutic targets such as WRN3, USP14, PKMYT15, POLQ6 and PRMT57 into the clinic. Here we identify, in human cancer, a novel synthetic lethal interaction between the PELO-HBS1L and SKI complexes of the mRNA quality control pathway. In distinct genetic contexts, including 9p21.3-deleted and high microsatellite instability (MSI-H) tumours, we found that phenotypically destabilized SKI complex leads to dependence on the PELO-HBS1L ribosomal rescue complex. PELO-HBS1L and SKI complex synthetic lethality alters the normal cell cycle and drives the unfolded protein response through the activation of IRE1, as well as robust tumour growth inhibition. Our results indicate that PELO and HBS1L represent novel therapeutic targets whose dependence converges upon SKI complex destabilization, a common phenotypic biomarker in diverse genetic contexts representing a significant population of patients with cancer.

Interleukin-1β (IL-1β) is dysregulated in chronic inflammatory diseases, yet the genetic factors influencing IL-1β production remain largely unknown. Myeloid-derived cells are the primary producers of IL-1β, which prompted a genome-wide CRISPR knockout screen in the human myeloid-derived U937 cells treated with lipopolysaccharide (LPS) to mimic inflammatory conditions and sorted for high and low intracellular IL-1β levels. A total of 295 genes are identified as regulators of IL-1β production, with 57 overlapping loci associated with inflammatory diseases, including the TNRC18 gene locus associated with multiple diseases in the Finnish population. U937 cells engineered with the Finnish-enriched rs748670681 risk allele demonstrate decreased expression of TNRC18 and an adjacent gene WIPI2, reduction in LPS-dependent gene activation and cytokine production, but elevation of interferon-responsive gene programs. Transcriptomic profiles for individual knockouts of TNRC18 and WIPI2 attribute the loss of LPS-dependent signaling primarily to TNRC18, which occurs through the modulation of H3K27 acetylation around inflammatory regulatory regions via TNRC18 and its protein interaction network. In contrast, the loss of WIPI2 is characterized by an exacerbation of interferon signaling. These findings delineate the global regulatory mechanisms of IL-1β production and provide molecular insights to the role of the rs748670681 variant in inflammatory diseases.

INTRODUCTION:Germline mutations driving lung cancer have been infrequently reported in the literature, with EGFR T790M being a known germline mutation identified in 1% of NSCLCs. Typically, a somatic EGFR mutation is acquired to develop lung adenocarcinoma. Osimertinib has become a standard-of-care treatment for EGFR T790M-positive lung cancer. METHODS:We perform a retrospective analysis through the Lung Cancer Moon Shot GEMINI database at the University of Texas MD Anderson Cancer Center. Of the patients that underwent cell-free DNA analysis, germline mutations were identified by those with high variant allelic fraction approximating 50%, followed by further confirmation on genetic testing. RESULTS:We identified 22 patients with germline EGFR mutations, with the majority harboring an EGFR T790M mutation (95.5%) and an EGFR L858R somatic mutation (50%). Notably, most patients were female (86.4%), non-smokers (81.8%), white (86.4%), had a family history of lung cancer (59.1%), and stage IV at diagnosis (72.7%). A distinct radiographic pattern of small multifocal ground-glass pulmonary nodules was observed in the majority of our cohort (72.7%). Among the 18 with advanced-stage NSCLC, 12 patients (66.7%) were treated with first-line osimertinib, demonstrating a median progression-free survival (PFS) of 16.9 months (95% confidence interval [CI]: 6.3-not reached [NR]). Others were treated with first-line afatinib (11.1%) or chemotherapy (22.2%). Among the 17 patients treated with osimertinib (in first or second-line), median PFS was 20.4 months (95% CI: 6.3-NR) and median overall survival was 82.0 months (95% CI: 28.4-NR). CONCLUSIONS:Based on our institutional cohort, NSCLC driven by EGFR germline mutations occurs more frequently in non-smoking, white females with multi-focal pulmonary nodules radiographically. Osimertinib for advanced germline EGFR-mutated NSCLC renders similar PFS compared to somatic T790M EGFR-mutated NSCLC.

INTRODUCTION:Adjuvant immunotherapy (IO) following concurrent chemotherapy and photon radiation therapy confers an overall survival (OS) benefit for patients with inoperable locally advanced non-small cell lung carcinoma (LA-NSCLC); however, outcomes of adjuvant IO after concurrent chemotherapy with proton beam therapy (CPBT) are unknown. We investigated OS and toxicity after CPBT with adjuvant IO versus CPBT alone for inoperable LA-NSCLC. MATERIALS AND METHODS:We analyzed 354 patients with LA-NSCLC who were prospectively treated with CPBT with or without adjuvant IO from 2009 to 2021. Optimal variable ratio propensity score matching (PSM) matched CPBT with CPBT + IO patients. Survival was estimated with the Kaplan-Meier method and compared with log-rank tests. Multivariable Cox proportional hazards regression evaluated the effect of IO on disease outcomes. RESULTS:Median age was 70 years; 71 (20%) received CPBT + IO and 283 (80%) received CPBT only. After PSM, 71 CPBT patients were matched with 71 CPBT + IO patients. Three-year survival rates for CPBT + IO vs CPBT were: OS 67% vs 30% (P < 0.001) and PFS 59% vs 35% (P = 0.017). Three-year LRFS (P = 0.137) and DMFS (P = 0.086) did not differ. Receipt of adjuvant IO was a strong predictor of OS (HR 0.40, P = 0.001) and PFS (HR 0.56, P = 0.030), but not LRFS (HR 0.61, P = 0.121) or DMFS (HR 0.61, P = 0.136). There was an increased incidence of grade ≥3 esophagitis in the CPBT-only group (6% CPBT + IO vs 17% CPBT, P = 0.037). CONCLUSION:This study, one of the first to investigate CPBT followed by IO for inoperable LA-NSCLC, showed that IO conferred survival benefits with no increased rates of toxicity.

There is a need to define regions of gene activation or repression that control human kidney cells in states of health, injury, and repair to understand the molecular pathogenesis of kidney disease and design therapeutic strategies. Comprehensive integration of gene expression with epigenetic features that define regulatory elements remains a significant challenge. We measure dual single nucleus RNA expression and chromatin accessibility, DNA methylation, and H3K27ac, H3K4me1, H3K4me3, and H3K27me3 histone modifications to decipher the chromatin landscape and gene regulation of the kidney in reference and adaptive injury states. We establish a spatially-anchored epigenomic atlas to define the kidney's active, silent, and regulatory accessible chromatin regions across the genome. Using this atlas, we note distinct control of adaptive injury in different epithelial cell types. A proximal tubule cell transcription factor network of ELF3, KLF6, and KLF10 regulates the transition between health and injury, while in thick ascending limb cells this transition is regulated by NR2F1. Further, combined perturbation of ELF3, KLF6, and KLF10 distinguishes two adaptive proximal tubular cell subtypes, one of which manifested a repair trajectory after knockout. This atlas will serve as a foundation to facilitate targeted cell-specific therapeutics by reprogramming gene regulatory networks.

N-linked glycosylation is a common post-translational modification that has various effects on multiple types of proteins. The extent to which an N-linked glycoprotein is modified and the identity of glycans species involved is of great interest to the biopharmaceutical industry, since glycosylation can impact the efficacy and safety of therapeutic monoclonal antibodies (mAbs). mAbs lacking core fucose, for example, display enhanced clinical efficacy through increased antibody-dependent cellular cytotoxicity. We performed a genome-wide CRISPR knockout screen in Chinese hamster ovary (CHO) cells, the workhorse cell culture system for industrial production of mAbs, aimed at identifying novel regulators of protein fucosylation. Using a lectin binding assay, we identified 224 gene perturbations that significantly alter protein fucosylation, including well-known glycosylation genes. This functional genomics framework could readily be extended and applied to study the genetic pathways involved in regulation of other glycoforms. We hope this resource will provide useful guidance toward the development of next generation CHO cell lines and mAb therapeutics.

Importance:Anaplastic thyroid carcinoma (ATC) is a rare and lethal cancer. Although progress has been made in recent years in patients with mutated BRAF tumors, those who respond initially eventually die of their disease; furthermore, there are no approved therapies for non-BRAF mutated tumors. Objective:To determine whether treatment with matched-targeted therapy plus immune checkpoint inhibitors were associated with improved overall survival (OS). Design, Setting, and Participants:A phase 2 trial at a single center, tertiary institution with parallel cohorts, assigning treatment with targeted therapy according to the tumor mutation status. Patients with mutated BRAF V600E tumors received vemurafenib/cobimetinib plus atezolizumab (cohort 1); those with mutated RAS (NRAS, KRAS, or HRAS) or NF1/2 tumors received cobimetinib plus atezolizumab (cohort 2). Patients without any of these variants were assigned to receive bevacizumab plus atezolizumab (cohort 3). Patients were enrolled from August 3, 2017, to July 7, 2021. All consecutive, systemic therapy-naive patients with ATC with active disease and who met eligibility criteria were considered for participation. The analysis was conducted in September 2023. Interventions:Patients were assigned to targeted therapy based on the driver mutation as follow: BRAF V600E (cohort 1, vemurafenib plus cobimetinib), RAS/NF (cohort 2, cobimetinib), or non-BRAF/RAS/NF (cohort 3, bevacizumab). All received atezolizumab. Main Outcomes and Measures:The primary outcome of the study was median OS of the entire targeted therapy cohort, compared with historical median OS of 5 months. Results:Forty-three patients with ATC were enrolled in the targeted therapy cohorts, of which 42 were included in the primary analysis. The median OS in patients across these 3 cohorts was 19 months (95% CI, 7.79-43.24). Median OS and progression-free survival per cohort were as follows: cohort 1: 43 months (95% CI, 16-not estimable [NE]), 13.9 months (6.6-64.1); cohort 2: 8.7 months (95% CI, 5.1-37.0) and 4.8 months (1.8-14.7); cohort 3 (vascular endothelial growth factor inhibitor group): 6.21 months (4.1-NE) and 1.3 months (1.3-NE), respectively. Conclusions and Relevance:In this nonrandomized clinical trial, atezolizumab combined with targeted therapy resulted in a longer median OS than historical landmark, achieving the study's primary end point, with cohort 1 achieving the longest OS. Trial Registration:ClinicalTrials.gov Identifier: NCT03181100.

BackgroundNeoadjuvant immune checkpoint inhibitors (ICIs) have improved survival outcomes compared with chemotherapy in resectable non-small cell lung cancer (NSCLC). However, the impact of actionable genomic alterations (AGAs) on the efficacy of neoadjuvant ICIs remains unclear. We report the influence of AGAs on treatment failure (TF) in patients with resectable NSCLC treated with neoadjuvant ICIs.MethodsTumor molecular profiles were obtained from patients with stage I–IIIA resectable NSCLC (American Joint Committee on Cancer seventh edition) treated with either neoadjuvant nivolumab (N, n=23) or nivolumab+ipilimumab (NI, n=21) followed by surgery in a previously reported phase-2 randomized study (NCT03158129). TF was defined as any progression of primary lung cancer after neoadjuvant ICI therapy in patients without surgery, radiographic and/or biopsy-proven primary lung cancer recurrence after surgery, or death from possibly treatment-related complications or from primary lung cancer since randomization. Tumors with AGAs (n=12) were compared with tumors without AGAs and non-profiled squamous cell carcinomas (non-AGAs+NP SCC, n=20).ResultsWith a median follow-up of 60.2 months, the overall TF rate was 34.1% (15/44). Tumor molecular profiling was retrospectively obtained in 47.7% (21/44) of patients and select AGAs were identified in 12 patients: 5 epidermal growth factor receptor(EGFR), 2KRAS, 1ERBB2, and 1BRAFmutations, 2 anaplastic lymphoma kinase(ALK)and 1RETfusions. The median time to TF in patients with AGAs was 24.7 months (95% CI: 12.6 to 40.4), compared with not reached (95% CI: not evaluable (NE)–NE) in the non-AGAs+NP SCC group. The TF risk was higher in AGAs (HR: 5.51, 95% CI: 1.68 to 18.1), and lower in former/current smokers (HR: 0.24, 95% CI: 0.08 to 0.75). The odds of major pathological response were 4.71 (95% CI: 0.49 to 45.2) times higher in the non-AGAs+NP SCC group, and the median percentage of residual viable tumor was 72.5% in AGAs compared with 33.0% in non-AGS+NP SCC tumors.ConclusionsPatients with NSCLC harboring select AGAs, includingEGFRandALKalterations, have a higher risk for TF, shorter median time to TF, and diminished pathological regression after neoadjuvant ICIs. The suboptimal efficacy of neoadjuvant chemotherapy-sparing, ICI-based regimens in this patient subset underscores the importance of tumor molecular testing prior to initiation of neoadjuvant ICI therapy in patients with resectable NSCLC.

6106 Background: Poorly differentiated thyroid carcinomas (PDTC) are a subset of thyroid cancers characterized by high-grade pathologic features. They hold an intermediate position on the spectrum of follicular-derived thyroid cancers. Due to their more aggressive clinical behavior, PDTCs typically exhibit shorter responses to the usual kinase inhibitors (KIs) used in advanced differentiated thyroid cancers and develop resistance earlier on. In fact, median progression free survival (PFS) with single-agent lenvatinib in patients with PDTC in the SELECT trial (n=28) was 14.8 months. Although there are increasing data to support the use of immunotherapy plus KI combinations in anaplastic thyroid cancer, evidence demonstrating the efficacy of this strategy in PDTC is sparse. We aimed to study the safety and efficacy of atezolizumab combined with KIs in PDTC. Methods: We enrolledpatientswith PDTC in a single-center phase II prospective trial of atezolizumab plus mutation-determined targeted therapy (NCT03181100). Patients with RAS or NF- mutated tumors were treated with atezolizumab plus the MEK inhibitor (MEKi) cobimetinib. Primary outcome was median overall survival (mOS). Best response to therapy was assessed per RECIST v1.1; survival by the Kaplan-Meier method. Results: Eight patients with RAS/NF-mutated PDTC were enrolled. Median age at treatment start was 68.5 years. All patients had distant metastatic disease at time of enrollment. Prior to study entrance, 7/8 patients (88%) had surgical resection of the primary tumor, 6/8 (75%) radioactive iodine, 2/8 (25%) external beam radiation to the neck, and 3/8 (38%) bridging cytotoxic chemotherapy. All patients were naïve to KIs. Six (75%) patients had RAS mutations (1 HRAS, 3 KRAS, 2 NRAS) and 3 (38%) had NF mutations (2 NF1, 1 NF2). One patient’s tumor harbored both KRAS and NF2 mutations at baseline. PD-L1 score was positive in 3/4 evaluable specimens. Median duration of follow-up was 65 months. Best response to therapy was stable disease in 6/8 (75%), partial response in 1/8 (12.5%) and progressive disease in 1/8 (12.5%). mOS was 23 months (95% CI, 12.8 – 33.2) and median PFS was 7 months (95% CI, 2.4 – 11.6). 4/8 patients received radioactive iodine (RAI) while on systemic therapy. Median dose of RAI was 149 millicuries (range, 108 – 153). mOS was significantly longer in patients who received RAI (32 vs 16 months; p= 0.034). The combination of cobimetinib + atezolizumab was overall well tolerated, with an expected adverse event profile. Conclusions: In patients with metastatic PDTC driven by RAS or NFmutations, combination of the anti-PD-L1 atezolizumab with MEKi cobimetinib showed some clinical efficacy, although PFS was shorter than with single-agent lenvatinib. However, SELECT trial did not delineate responses by driver mutations. This combination could thus be considered in selected patients at high risk of complications with antiangiogenic KIs. Clinical trial information: NCT03181100 .

Background Functionally redundant paralogs in the human genome are the most common source of synthetic lethality (i.e., loss of one paralog conveys dependency to another). However, most paralogs have yet to be experimentally tested and the human paralogome remains largely uncharacterized. Results We performed the first pairwise genetic screen of all human paralogs using a multiplexed CRISPR-Cas12 library, which revealed that digenic synthetic lethalities are relatively rare (<0.5% of all paralog pairs) and varied in penetrance across different cancer models. We hypothesized that the variable penetrance of digenic synthetic lethalities was a result of complex polygenic interactions with endogenous factors. A multivariable regression analysis of 1,278 pairs and endogenous cellular features across 30 cancer models revealed that perturbations of related pathways were frequently predictive of paralog synthetic lethality. A machine learning classifier was also built to predict synthetic lethalities using a weighted set of true positives, accounting for the variable penetrance of synthetic lethal interactions. Intuitively, the predictive scores revealed that the penetrance of synthetic lethal interactions was driven by the overlap and essentiality of the protein-protein interactions for each paralog pair. Conclusions This study provided a comprehensive analysis of all digenic interactions in the human paralogome, as well as the key features that underlie the heterogeneity in synthetic lethalities that have been reported here and elsewhere. ### Competing Interest Statement The design, study conduct, and financial support for all other research were provided by AbbVie. AbbVie participated in the interpretation of data, review, and approval of the publication. All authors were employees of AbbVie at the time of the study. M.J.F. is currently a full time employee of Pfizer.

INTRODUCTION:Drug resistance remains a major clinical challenge in EGFR-mutant NSCLC tumors owing to pathway reactivation, pathway bypass, and pathway indifference resistance mechanisms to evade tyrosine kinase inhibitor (TKI) suppression. Fusion of receptor tyrosine kinases (RTKs), such as RET, ALK, and FGFR3, has been reported to mediate EGFR TKI resistance. Given the rarity of these fusions and the heterogeneous nature of the condition, no prospective clinical trials evaluated the incidence, safety, and therapeutic benefit of dual EGFR-RTK inhibition. METHODS:We queried clinical databases from multiple institutions to identify patients who had RTK fusions detected on next-generation sequencing testing results from tissue or blood at five institutions: the Second Affiliated Hospital Zhejiang University School of Medicine, Hunan Cancer Hospital, Prince of Wales Hospital Chinese University of Hong Kong, Chao Family Cancer Center, and the University of Texas MD Anderson Cancer Center from March 1, 2016, to September 30, 2023. The data analyzed included objective response rate (ORR) to treatment post RTK fusion detection, duration of treatment, and safety. A comprehensive literature search was conducted to identify patients with RTK fusion as the primary resistance mechanism in EGFR-mutated NSCLC patients. RESULTS:Twenty-seven patients were identified to be eligible in the analysis. ALK fusions were most reported (42.9%), followed by RET fusions (35.7%). Fifteen patients received dual TKI after fusion detection and nine received fusion targeting single TKIs. The median time on treatment was 169 days or 5.8 months (35-1050 d). ORR by the Response Evaluation Criteria in Solid Tumors in the evaluable 25 patients was 24% and the disease control rate was 80%. In 14 evaluable patients who received dual TKI therapy, ORR by the Response Evaluation Criteria in Solid Tumors was 21.4%, and the disease control rate was 78.6%. No new toxicities were observed with dual EGFR-RTK inhibition. In the literature review, after pooling 291 patients from 59 studies, RET fusions were the most common (50.0%), followed by BRAF (13.3%), ALK (13.3%), FGFR (10%), NTRK (5.3%), EGFR (1.7%), ROS1 (1.3%), MET (1%), and ERBB (0.7%). CONCLUSION:The emergence of RTK fusions is one of the mechanisms of bypass resistance of EGFR TKI. Dual inhibition of EGFR-RTK was safe and efficacious in patients with targetable RTK fusion after progression to EGFR TKIs.

e18070 Background: Anaplastic thyroid carcinoma (ATC) is an aggressive malignancy with a historical dismal prognosis. Although BRAF/MEK inhibitors have revolutionized the treatment of BRAF-mutant ATC, treatment options remain limited for those without a BRAF mutation. Single agent immunotherapy showed modest efficacy, while combination of immunotherapy with the anti-VEGF lenvatinib led to marked improvement in survival. Yet, given the locally advanced nature of this disease, many patients have contra-indications to antiangiogenics. We assessed the safety and efficacy of the anti-PD-L1 atezolizumab combined with taxane chemotherapy in patients with non- BRAF, non- RAS/NF mutated ATC with contra-indications to VEGF inhibitors. Methods: We enrolledpatientswith ATC in a single-center phase II prospective trial of atezolizumab plus mutation-determined targeted therapy (NCT03181100). Patients without a BRAF, RAS or NF mutation who had a contra-indication to anti-VEGF therapy (such as high risk for bleeding or fistula) were included in an exploratory cohort and were treated with nab-paclitaxel 100 mg/m2 on days 1, 8 and 15, combined with atezolizumab 1200 mg intravenously every 21 days. Paclitaxel could be substituted for nab-paclitaxel, but the latter was preferred as it does not require prophylactic steroids. Here we present results for this exploratory cohort. Best response to therapy was assessed per RECIST v1.1; survival by the Kaplan-Meier method. Results: Nine patients were enrolled in the exploratory cohort. All were naïve to systemic therapies, and 3/9 (33%) received prior bridging cytotoxic chemotherapy. Prior to study entrance, 6/9 (67%) had radiation therapy to the neck and 4/9 (44%) had surgery to resect the primary tumor. Seven (78%) had distant metastatic disease at time of enrollment. Most frequent baseline mutations were in TP53 (89%), APC (33%), PTEN (33%) and TERT promoter (33%). Median duration of follow-up was 4.4 months. Best response to therapy was partial response in 2 patients (1 confirmed and 1 unconfirmed), stable disease in 3 patients, and progressive disease in 4 patients. Overall response rate was 11 %, while disease control rate was 56%. Median OS was 4.44 months (95% CI, 1.12 –25.33) and median progression-free survival was 2.73 months (95% CI, 1.12 – 7.13). All patients had died at the time of data cut-off. Regarding treatment safety, the combination of atezolizumab plus taxane chemotherapy was overall well tolerated, with most adverse events being grade 2 or lower. Conclusions: Combination of atezolizumab plus taxane chemotherapy, although safe, offers limited additional survival benefit compared to single agent immunotherapy (median OS = 5.9 months) or the historical median OS of 5 months. Additional novel treatment strategies for patients with non- BRAF mutated ATC are warranted, especially when antiangiogenic drugs are contra-indicated. Clinical trial information: NCT03181100 .

Cancer dependency maps have accelerated the discovery of tumor vulnerabilities that can be exploited as drug targets when translatable to patients. The Cancer Genome Atlas (TCGA) is a compendium of ‘maps’ detailing the genetic, epigenetic and molecular changes that occur during the pathogenesis of cancer, yet it lacks a dependency map to translate gene essentiality in patient tumors. Here, we used machine learning to build translational dependency maps for patient tumors, which identified tumor vulnerabilities that predict drug responses and disease outcomes. A similar approach was used to map gene tolerability in healthy tissues to prioritize tumor vulnerabilities with the best therapeutic windows. A subset of patient-translatable synthetic lethalities were experimentally tested, including PAPSS1/PAPSS12 and CNOT7/CNOT78, which were validated in vitro and in vivo. Notably, PAPSS1 synthetic lethality was driven by collateral deletion of PAPSS2 with PTEN and was correlated with patient survival. Finally, the translational dependency map is provided as a web-based application for exploring tumor vulnerabilities.