BACKGROUND:Lung cancer is the leading cause of cancer mortality worldwide despite the availability of low-dose computed tomography (LDCT) for screening in high-risk populations. METHODS:To develop an approach and identify blood-based protein signatures for lung cancer that can be deployed across platforms, we combined data-independent acquisition mass-spectrometry (DIA-MS) and proximity extension assay (PEA) with explainable artificial intelligence (XAI)-led machine learning (ML) for plasma-based biomarker discovery. Using a cohort of 490 lung cancer patients and 124 matched controls, ML models were trained to predict lung cancer and XAI was used to characterise networks of model-consistent features. We then introduced a DNA-aptamer based proteomic approach to assess cross-platform concordance and define a cross-platform signature. This signature was subsequently evaluated using an external cohort. RESULTS:Here we show that ML models achieve an AUROC of 0.91 [95% CI: 0.88-0.93] and 0.97 [95% CI: 0.92-0.98] in DIA-MS and PEA, respectively, using a 80/20% train/holdout split. XAI further characterises networks of model-consistent features related to chemotaxis, cell adhesion, wound healing and immune response. Introduction of the DNA-aptamer proteomic approach identifies a cross-platform signature, with performances of 0.88 [95% CI: 0.80-0.90] and 0.88 [95% CI: 0.81-0.95] in DIA-MS and PEA, respectively. Assessment of this signature in an external cohort separates lung cancer from control cases. CONCLUSIONS:This study develops an approach combining multi-dimensional proteomics with XAI-ML and demonstrates the characterisation of cross-platform biomarker signatures for lung cancer.

BACKGROUND:Anaplastic lymphoma kinase inhibitors (ALKi) are a mainstay of therapy for patients with advanced non-small cell lung cancers (NSCLC). ALKi are associated with increased serum creatinine, which may represent reduced renal tubular creatinine secretion and/or true acute kidney injury (AKI). METHODS:We performed a retrospective study of patients who received ALKi for NSCLC (2013-2022). The primary outcomes were incidence of AKI within 90 days of ALKi start and chronic kidney disease (CKD) within 1-year. AKI and CKD were defined via KDIGO criteria, using the CKD-EPI equation with creatinine (and cystatin C in a subcohort). We performed logistic regression for AKI risk factors and Kaplan-Meier analysis for overall survival (OS). Spline curves were generated for eGFR means over time. RESULTS:Among 114 NSCLC patients, median age 61 years; 53% were female; 191 ALKi treatments were initiated: Alectinib (n = 91), Crizotinib (n = 33), Ceritinib (n = 27), Brigatinib (n = 21) and Lorlatinib (n = 19). There were 20 (10%) AKI events within 90 days after ALKi initiation, with 4 treatment changes attributed to AKI. Twenty-eight (14%) patients developed CKD, with 10 treatment changes; none requiring dialysis. In multivariate analysis, hypertension and male sex were associated with AKI. OS did not differ by AKI status. CONCLUSION:AKI/CKD events were frequent post-ALKi initiation (using creatinine-based eGFR), with a minority resulting in treatment change. Mean eGFR declined in the 90-days post-ALKi start. Most patients had mild CKD, with eGFR recovering postdrug cessation. AKI did not impact OS. Our findings suggest that most patients may continue ALKi therapy despite creatinine-based eGFR changes.

Introduction:The prognostic significance of baseline and on-treatment brain and liver metastasis in ALK+ or ROS1+ metastatic NSCLC (mNSCLC) remains unclear. As we consider intensification strategies, it is critical to identify factors that predict high-risk disease. Methods:Clinical characteristics and outcomes were abstracted from the electronic medical records of patients with ALK+ or ROS1+ mNSCLC. Baseline characteristics and the cumulative incidence (CI) of brain and liver metastases were compared (≥2-year survivors versus <2-year; pre-2017 versus post-2017). Multivariable Cox proportional hazard models were used to evaluate the association between factors and overall survival, and multivariable logistic regression models were used for the odds of death within 2 years. Results:A total of 310 patients with ALK+ mNSCLC were identified (≥2-y: 229, <2-y: 81). There was no difference in cumulative incidence of brain metastases between survival groups (29% at 21 mo). However, the cumulative incidence of liver metastasis was higher in those who survived less than 2 years (20.9% versus 5.4% at 21 mo). The cumulative incidence of brain but not liver metastases has improved post-2017 with the newer generation of ALK tyrosine kinase inhibitors. There were 69 patients with ROS1+ mNSCLC who were identified (≥2-y: 46, < 2-y: 23). There was no significant difference in the cumulative incidence of brain or liver metastases between less-than-2-year and greater-than-or-equal-to-2-year survivor cohorts (p = 0.664, p = 0.201). Conclusions:Among patients with ALK+ but not ROS1+ mNSCLC, the presence of liver metastases at baseline and on-treatment was associated with worse survival. In the ALK+ population, the cumulative incidence of brain but not liver metastases is improving, highlighting a need for therapies effective at the treatment and prevention of liver metastases.

BACKGROUND:Data are limited on the efficacy of different TKIs for patients with atypical EGFR-mutated (AM) mNSCLC, a heterogeneous group excluding classical mutations (CM) L858R and exon19del. In our previous single-institution analysis, AM patients had longer survival with osimertinib than afatinib, but outcomes for patients with specific mutations could not be compared due to sample size. METHODS:We performed a multi-institution, retrospective survival analysis of atypical EGFR mutated (AM) mNSCLC patients treated with 1L osimertinib or afatinib between 2015-2021 at 12 US institutions. Time to discontinuation (TTD) and overall survival (OS) were estimated using Kaplan-Meier curves and compared using log rank tests between treatment or mutation groups. RESULTS:Among 52 patients identified, 32 (62 %) were treated with osimertinib and 20 (38 %) with afatinib. 20 had mutations in G719X (38 %), 12 in L861Q (23 %), and 5 in S768I (10 %). 34(65 %) had compound mutations: 20(62 %) had AM + CM, and 14(38 %) had ≥ 2 AMs. Among G719X (n = 20), afatinib was associated with longer time to discontinuation (TTD) (log-rank: p = 0.047) and longer OS (p = 0.043) vs. osimertinib. Median TTD (mTTD) was 20.3 m[95 %CI 7.3-24.2] and 9.4[1.7-14.0], respectively. For L861Q (n = 12), osimertinib was associated with longer TTD vs. afatinib (p = 0.004), with no statistical difference in OS (p = 0.215). mTTD was 7.2 m[2.2-12.3] and 1.3[0-3.1], respectively. In AM + CM (n = 20), osimertinib was associated with longer TTD and OS compared to those receiving afatinib (p = 0.037, p = 0.042, respectively). CONCLUSIONS:Patients with G719X alterations experienced longer TTD and OS with afatinib than osimertinib. In contrast, patients with L861Q alterations had longer TTD with osimertinib. In AM + CM pts, TTD and OS with osimertinib were longer than afatinib, suggesting that classical mutations may be driving the outcomes. Atypical EGFR mutations may warrant distinct treatment recommendations based on the specific mutation(s) identified, but more studies are needed.

Immune checkpoint inhibitors (ICIs) have transformed cancer treatment, providing survival benefits across multiple cancer types. Despite this remarkable advancement, ∼15-30% of patients develop immune-related adverse events (irAEs). One of the common irAEs is immune-mediated colitis, and the incidence ranges from 1% to 25% by type of therapy. Previously, we have shown that polygenic risk scores of ulcerative colitis predict colitis irAE. However, the role of rare coding variants on the development of colitis irAE remains largely unknown. We conducted whole exome sequencing for colitis irAE in 1, 215 cancer patients treated with ICIs. Association analyses were carried out for each gene separately for protein-truncating missense variants, predicted splice (Sp) variants, and non-synonymous predicted deleterious (deN; CADD≥20 or VEST≥0.8) or any non-synonymous (N) variants. The main association analyses were burden tests in which genotypes were collapsed to a 0/1 variable based on whether samples carried a variant of the given class. All models were adjusted for age at diagnosis, sex, type of ICI therapy, recruitment site, and five ancestry-informative principal components. The exome-wide threshold for significance was set at 2.5 × 10-6, with genes beyond this considered significant. Among 1, 215 patients treated with immune checkpoint inhibitors (ICIs), 60 (4.9%) developed colitis. We identified rare coding variants in two genes associated with an increased risk of colitis at the exome-wide significance level: PT+Sp+N variants in LRRC47 (OR = 5.5, P = 2.4 × 10-6) and PT+S+deN variants in PFKP (OR = 65.9, P = 1.2 × 10-6). Additionally, PT+S+deN variants in PTPRH showed a suggestive association with colitis irAEs (OR = 8.8, P = 7.1 × 10-5). To our knowledge, this is the first study to assess the role of rare coding variants in colitis immune-related adverse events (irAEs) among cancer patients treated with ICIs. Our findings highlight two exome-wide significant associations, implicating variants in LRRC47 and PFKP in the development of colitis irAEs. LRRC47 has previously been linked to inflammatory bowel disease through differential methylation patterns. Furthermore, genes involved in glycolysis, including PFKFB3, PKM, and PFKP, have been shown to be upregulated in patients with ulcerative colitis. Abnormal activity of PTPRH has also been reported to disrupt the regulatory mechanisms of the intestinal mucosa, potentially contributing to inflammation. These findings provide novel insights into the genetic underpinnings of colitis irAEs in cancer immunotherapy. Pooja Middha, 1 Zoe Quandt, 2 Karmugi Balaratnam, 3 Eduardo Cardenas, 1 Christina J. Falcon, 4 Princess Margaret Lung Group, Matthew A. Gubens, 1 Scott Huntsman, 1 Khaleeq Khan, 3 Min Li, 1 Christine M. Lovly, 5 Devalben Patel, 3 Luna Jia Zhan, 3 Melinda C. Aldrich, 6 Geoffrey Liu, 7 Adam J. Schoenfeld, 4 Elad Ziv1. Whole exome sequencing of immune-mediated colitis in cancer patients treated with immune checkpoint inhibitors [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 1 (Regular Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_1):Abstract nr 2276.

Introduction Immune checkpoint inhibitors (ICIs) can yield remarkable clinical responses in subsets of patients with solid tumors, but they also commonly cause immune-related adverse events (irAEs). The predictive features of clinically severe irAEs leading to cessation of ICIs have yet to be established. Given the similarities between irAEs and autoimmune diseases, we sought to investigate the association of a germline polygenic risk score for autoimmune disease and discontinuation of ICIs due to irAEs.Methods The Genetics of immune-related adverse events and Response to Immunotherapy (GeRI) cohort comprises 1302 patients with non-small cell lung cancer (NSCLC) who received ICI therapy between 2009 and 2022 at four academic medical centers. We used a published polygenic risk score for autoimmune diseases (PRSAD) in the general population and validated it in the All of Us. We then assessed the association between PRSAD and cessation of ICI therapy due to irAEs in the GeRI cohort, using cause-specific and Fine-Gray subdistribution hazard models. To further understand the differential effects of type of therapy on the association between PRSAD and cessation of ICI due to irAEs, we conducted a stratified analysis by type of ICI therapy.Results Using a competing risk model, we found an association between PRSAD and ICI cessation due to irAEs (HR per SD=1.24, p=0.004). This association was particularly strong in patients who had ICI cessation due to irAEs within 3 months of therapy initiation (HR per SD=1.40, p=0.005). Individuals in the top quintile of PRSAD had 4.8% ICI discontinuation for irAEs by 3 months, compared with 2% discontinuation by 3 months among patients in the bottom quintile (log-rank p=0.03). In addition, among patients who received combination programmed cell death protein-1 (PD-1)/programmed death-ligand 1 (PD-L1) inhibitors and cytotoxic T-lymphocyte associated protein 4 (CTLA4) inhibitors, ICI discontinuation for irAEs by 3 months occurred in 4 of the 13 patients (30.8%) with high PRSAD genetic risk (top quintile) versus 3 of 21 patients (14.3%) with low PRSAD genetic risk (bottom quintile).Conclusions We demonstrate an association between a polygenic risk score for autoimmune disease and early ICI discontinuation for irAEs. Our results suggest that germline genetics may be used as an adjunctive tool for risk stratification around ICI clinical decision-making in solid tumor oncology.

Introduction:ALK tyrosine kinase inhibitors (ALK TKIs) are started at a standard dose regardless of patients' weight and body surface area (BSA). In this retrospective analysis, the authors explored whether body size variables were associated with toxicity and efficacy. Methods:Retrospective data for ALK-positive patients at the Princess Margaret Cancer Centre were extracted from electronic health records. Associations between BSA/weight quartiles and dose reductions (DRs), temporary interruptions (TIs), and permanent discontinuation due to toxicity were evaluated using generalized linear mixed modeling. Survival analysis was conducted using Kaplan-Meier curves and log-rank tests. Results:Among 142 patients with ALK-positive NSCLC treated between July 2012 and March 2024, the median age was 58 years, 54% were female, 78% never-smokers, 49% were Asian, and 37% were Caucasian. A significantly higher proportion of males were in the highest BSA Q4 (89%, p < 0.001), whereas more females (91%, p < 0.001) and Asians were in the lowest BSA Q1 (80%, p < 0.001).Higher weight at ALK TKI initiation was associated with an increased likelihood of DR at any time (adjusted odds ratio [aOR] = 1.20 per 10-kg increase, 95% confidence interval: 1.0-1.4, p = 0.04), as was having higher BSA (aOR = 2.00 per 0.5-U increase, 95% confidence interval: 1.0-3.9, p = 0.04). TIs were associated with higher weight (10-kg increase, aOR = 1.28, p = 0.006), BMI (5-U BMI increase, aOR = 1.40, p = 0.02), and BSA (0.5-U increase, aOR = 2.59, p = 0.005). These weight/BSA results were statistically significant in patients during alectinib or lorlatinib treatment. In contrast, permanent discontinuation was associated with higher weight/BSA in brigatinib/ceritinib/crizotinib-treated patients. In multivariable analysis, older age and male sex were independently associated with more DRs and TIs. Weight and BSA quartiles were not associated with progression-free survival (p = 0.4) or overall survival (p = 0.6). Conclusions:Higher weight and larger BSA at the start of ALK TKI treatment were associated with higher likelihood of toxicity, leading to more DRs and TIs-particularly in males and patients receiving alectinib and lorlatinib. However, weight and BSA were not associated with treatment outcomes.

Introduction: There is a paucity of real-world associations between EQ-5D-generated health utility scores (HUS), symptoms as measured by the Edmonton Symptom Assessment System (ESAS), and the patient-reported outcomes version of the common terminology criteria for adverse events (pro-CTCAE), and survival in patients with advanced Malignant Pleural Mesothelioma (aMPM). Methods: Clinico-demographic variables and treatment information were captured retrospectively in patients diagnosed with aMPM between January 2004 and February 2021 at Princess Margaret Cancer Centre. Quality of life outcomes were measured using HUS, ESAS, and pro-CTCAE scales, by stable versus progressive disease and line-of-treatment states. Survival by mean ESAS scores were analyzed using the Kaplan-Meier method. Results: Of the 262 patients, the median age was 69 years (interquartile range: 62–74), 77% were male individuals, 52% were ever-smokers, 67% were the epithelioid-subtype, and 62% received first-line systemic therapy for advanced disease. The mean baseline HUS at diagnosis was 0.68 (95% confidence interval: 0.62–0.74) with most symptoms consisting of pain, dyspnea, and fatigue. Pooled ESAS physical and psychological scores changed significantly with disease state: the mean scores were worst at baseline, improved with stable or responding disease (physical, p < 0.001; psychological, p < 0.001), and worsened at progressive disease (physical: p < 0.001; psychological, p < 0.001). Similar trends were seen in HUS and pro-CTCAE symptom severity/frequency. Patients with high baseline ESAS physical symptom burden had inferior overall survival (median = 8.9 [high] versus 12.6 months [low], p = 0.022). Weak-to-moderate correlations were observed between most ESAS domains and HU and between pro-CTCAE domains and HU. The strongest domain correlations were with well-being, shortness of breath, tiredness, and depression domains. Conclusions: Baseline quality of life burden is high in patients with aMPM and is well captured by both EQ-5D and ESAS Individual ESAS and pro-CTCAE domains reported low/moderate correlations with HUS, reflecting the inability of one symptom to predict the entire disease state, thus paving the way for future mapping studies. The baseline physical symptom burden (ESAS) was prognostic of survival.

8082 Background: Circulating tumor DNA (ctDNA) profiling in pleural mesothelioma (PM) is challenging due to its molecular heterogeneity and lack of mesothelioma-specific mutations. Diagnosis can be challenging and may require repeat biopsies. Cell-free methylated DNA immunoprecipitation sequencing (cfMeDIP-seq) of plasma cell-free DNA (cfDNA) offers a non-invasive approach to analyzing differentially methylated regions (DMRs), providing insights into epigenetic changes that could serve as potential biomarkers for diagnosis, histological differentiation, and prognosis in PM. Methods: cfMeDIP-seq was performed on plasma samples from 55 PM patients and 24 asbestos-exposed non-cancer controls (NCC). Libraries were sequenced to an average depth of 70 million reads, and chromosomes 1-22 were binned into 300 bp windows for read tallying. For NCCs, bins with a mean beta-value <0.3 and CG density >2 (n = 3,537,691 windows) were analyzed. DMR analysis and pathway enrichment were conducted using R packages (limma, clusterProfiler), and machine learning models were developed with Python modules (pandas, numpy, sklearn). Results: Among the 55 PM patients (72% epithelioid, 13% biphasic, 15% sarcomatoid), the median age was 70 years, 85% were male, and 78% had prior asbestos exposure. Using a stringent filter (mean beta-value <0.1; CG density >5), a random forest classifier was developed with 141 windows, distinguishing PM from NCC with 91% accuracy, 88% precision (or positive predictive value, PPV), and an area under the ROC curve (AUC) of 0.94 across 5-fold cross-validation cohorts. DMR analysis of epithelioid vs. sarcomatoid PM revealed 1,585 significantly different windows (adjusted p < 0.05), achieving 83% accuracy, 74% precision, and an AUC of 0.98. Gene ontology analysis indicated significant enrichment in RNA processing pathways. Among epithelioid PM patients, distinct DMRs were identified between those with overall survival (OS) ≤ 6 months and >6 months (n = 1,824 windows, adjusted p < 0.05). Patients with OS ≥ 36 months and <36 months showed 37 significantly differential windows (adjusted p < 0.05), though test performance assessment was limited by the small sample size. Conclusions: If validated, global methylome profiling of ctDNA via cfMeDIP-seq offers a novel, non-invasive method that may enhance accurate diagnosis and histological differentiation. Additionally, identifying epigenetic biomarkers could provide deeper insights into PM biology, paving the way for personalized medicine and improved patient outcomes.

Background: We evaluated outcomes in non-small cell lung cancer (NSCLC) patients who presented with brain-only metastatic (BOM) disease overall and by EGFR/ALK mutation status. Methods: We analyzed clinico-demographic, treatment and survival data for all NSCLC patients who presented to our center between 2014 and 2016 with BOM as their first presentation of metastatic disease. Differences in overall survival (OS) were evaluated using log-rank tests for NSCLC wildtype (NSCLCwt) versus NSCLC with an ALK-rearrangement/EGFR-mutation (NSCLCmut+). Results: Of 109 patients with BOM, median age was 68 years; 51 % were female; 69 % Caucasian; 76 % ever-smoker; 76 % adenocarcinoma; and 25 % NSCLCmut+. While 41 patients (38 %) had subsequent brain-only progressive disease (PD), 22 (20 %) developed extracranial metastases. A higher proportion of NSCLCmut+ (vs -wt) subsequently progressed outside the brain (37 % vs 15 %, p = 0.03). Median time-to-first-extracranial-metastases was 8.5 (NSCLCmut+) vs 21.0 months (NSCLCwt; p = 0.23). With 17.7 months median follow-up, median-OS was 15.9 months [95%CI: 11.5-21.3; all patients]; 12.3 [7.4-18.4; NSCLCwt] and 38.9 [21.3-not reached (NR); NSCLCmut+] (p = 0.09). In 33 of 80 patients with de novo BOM, the primary tumor was treated with surgery or radiotherapy. In patients with NSCLCwt, there was no OS benefit associated with local lung tumor treatment (p = 0.68), whereas in NSCLCmut + pts, local lung tumor treatment correlated with greater OS (median-OS NR vs 21.5 months; p = 0.05). Conclusion: In patients with NSCLCwt with BOM, we observed a -predominant pattern of brainonly secondary progression, however patients with NSCLCmut + more often progressed extracranially. In patients with NSCLCmut+ and BOM, definitive primary tumor treatment correlated with improved survival.

The field of precision oncology has witnessed several advances that stimulated the development of new clinical trial designs and the emergence of real-world data (RWD) as an important resource for evidence generation in healthcare decision-making. Here, we highlight our experience with an innovative approach to a set of Adaptive, Universal Principles for Real-world Observational Studies (AUPROS). To demonstrate the utility of these principles, we used a mixed-methods approach to assess three studies that follow AUPROS at Princess Margaret Cancer Centre: (1) Molecular Epidemiology of ThorAcic Lesions (METAL), (2) Translational Head And NecK Study (THANKS), and (3) CAnadian CAncers With Rare Molecular Alterations (CARMA; NCT04151342). We performed resource assessments, stakeholder-directed surveys and discussions, analysis of funding, research output, collaborations, and a Strengths-Weaknesses-Opportunities-Threats (SWOT) analysis. Based on these analyses, AUPROS is an approach that is applicable to a wide range of observational study designs. The universality of AUPROS allows for multi-purpose analyses of various RWD, and the adaptive nature creates opportunities for multi-source funding and collaborations. Following AUPROS can offer cost and logistical benefits and may lead to increased research productivity. Several challenges were identified pertinent to ethics approvals, sustainability, complex coordination, and data quality that require local adaptation of these principles.

9116 Background: Although osimertinib has become standard first-line (1L) therapy for patients (pts) with EGFRm mNSCLC, a subset treated with earlier-generation EGFR-targeted TKIs and chemotherapy have had long-term survival. We sought to characterize clinical features of long-term survival in pts with EGFRm mNSCLC treated prior to the osimertinib era. Methods: Data were abstracted from electronic medical records at 12 cancer centers participating in the Academic Thoracic Oncology Medical Investigator’s Consortium (ATOMIC). We included patients with mNSCLC with sensitizing mutations in EGFR who started 1L systemic therapy before 2015. Survival distributions and predictors were assessed using Kaplan-Meier estimates and a multivariable Cox proportional hazards model including time-dependent brain metastasis (met) development, age, 1L therapy (targeted vs. chemotherapy), sex, race, and smoking status. Multivariable logistic regression was used to evaluate baseline predictors associated with 5+ year survival vs. not for those with known survival status at 5 years after start of 1L. Results: We identified 304 patients (69% female, 56% White, 29% Asian, mean age 61.2). First-line targeted therapy was given in 70% of patients. With a median follow-up of 81.5 months, median overall survival was 63.5 months (95% CI 59.4-71.9). 135 (44%) pts survived 5+ years; among those with baseline next-generation sequencing (NGS), presence of a baseline pathogenic ERBB2 variant was higher in 5+ year survivors (4/51 [8%], 2x amplification, y772_A775dup, S310F) v others (1/65 [2%], amplification). Among 161 pts with baseline brain imaging, both baseline and on-treatment development of brain mets were associated with worse survival (HR 1.29, 95% CI 1.21-1.37; HR 1.59, 95% CI 1.47-1.72; both P < 0.001). Excluding 22 patients lost to follow-up before 5 years, a history of smoking (OR 2.99, 95% CI 1.35-6.90, P = 0.008) and baseline brain metastases (OR 3.57, 95% CI 1.64-8.13, P = 0.002) were associated with death before 5 years. Conclusions: Long-term survivors with EGFRm mNSCLC treated before the osimertinib era were more likely to be nonsmokers and have no baseline brain metastases. This highlights a subset of EGFRm patients who had excellent outcomes with older treatment strategies and may not benefit as much from intensification approaches. Additional baseline mutational data will be presented at the conference.

BACKGROUND:There is a paucity of information about the characteristics, treatment patterns, and outcomes of non-small cell lung cancer (NSCLC) patients with single organ metastasis (SOM). METHODS:This retrospective cohort study includes all patients with a diagnosis of stage IV NSCLC diagnosed from 2014 to 2016 and treated at Princess Margaret Cancer Centre. We compared baseline characteristics and patterns of metastatic sites between patients with SOM versus multiple (M)OM. Additionally, we identified treatment modalities and outcomes for patients with SOM. Cox multivariable models (MVA) were utilized to evaluate differences in overall survival (OS) between the SOM and MOM cohorts. RESULTS:Of 893 pts analyzed, 457 (51 %) had SOM, while 436 (49 %) had MOM at initial diagnosis. Demographics were comparable between the two groups. Brain was the most common site of metastasis for SOM patients. When compared to the MOM group, the SOM group had lower percentages of liver and adrenal metastases. Amongst SOM patients, 54 % received single modality treatment, and 20 % did not receive any treatment for their SOM. In MVA, patients with liver (HR 2.4), bone (HR 1.8), and pleural (HR 1.7) metastasis as their SOM site had the worst outcomes, with median OS of 6.8 months, 12.1 months, and 13.0 months respectively. Patients with SOM had a significantly improved median OS compared to those with MOM (15.9 months vs. 10.6 months; HR 0.56, 95 % CI 0.47-0.66, p < 0.001). CONCLUSION:In NSCLC patients who presented with SOM, survival correlated with the initial organ involved and was better overall compared to patients with MOM. SOM NSCLC may benefit from specific management strategies and SOM patients could be considered as a specific subgroup for survival analyses in observational and non-randomized interventional studies. In clinical trials, SOM can be considered as a stratification factor in the future.

INTRODUCTION:Determining lines of therapy (LOT) using real-world data is crucial to inform clinical decisions and support clinical research. Existing rules for determining LOT in patients with metastatic non-small cell lung cancer (mNSCLC) do not incorporate the growing number of targeted therapies used in treatment today. Therefore, we propose rules for determining LOT from real-world data of patients with mNSCLC treated with targeted therapies. METHODS:LOT rules were developed through expert consensus using a real-world cohort of 550 patients with ALK+ or ROS1+ mNSCLC in the multi-institutional, electronic medical record-based Academic Thoracic Oncology Medical Investigators Consortium's (ATOMIC) Driver Mutation Registry. Rules were subsequently modified based on a review of appropriate LOT determination. These resulting rules were then applied to an independent cohort of patients with EGFR+ mNSCLC to illustrate their use. RESULTS:Six rules for determining LOTs were developed. Among 1133 patients with EGFR mutations and mNSCLC, a total of 3168 regimens were recorded with a median of 2 regimens per patient (IQR, 1-4; range, 1-13). After applying our rules, there were 2834 total LOTs with a median of 2 LOTs per patient (IQR, 1-3; range, 1-11). Rules 1-3 kept 11% of regimen changes from advancing the LOT. When compared to previously published rules, LOT assignments differed 5.7% of the time, mostly in LOTs with targeted therapy. CONCLUSION:These rules provide an updated framework to evaluate current treatment patterns, accounting for the increased use of targeted therapies in patients with mNSCLC, and promote standardization of methods for determining LOT from real-world data.

Introduction: Osimertinib is now a standard first-line (1L) therapy for EGFR-mutated (EGFRm) advanced NSCLC. We aimed to characterize patterns of therapy and longitudinal risk of brain and liver metastasis in a cohort of EGFRm NSCLC. Methods: Patients with metastatic EGFRm NSCLC who received 1L systemic therapy at sites within the Academic Thoracic Medical Investigator's Consortium were included; demographic and clinical data including treatment patterns were described. Analyses of overall survival, time to next treatment, and incident brain and liver metastasis were performed using the Kaplan-Meier method, Cox regression, and cumulative incidence functions on patients who started 1L therapy in 2015 or later. Results: The full cohort included 1132 patients and the mean age of the participants was 63.4 years; among the participants, 53% were White individuals, 68% were female individuals, and 67% were nonsmokers. Among the participants, 830 patients received 1L systemic therapy in 2015 or later. The predominant first EGFR-tyrosine kinase inhibitor was erlotinib (65%) before 2018 and osimertinib (81%) after 2018. The median time to the next treatment after the start of 1L therapy was 13.9 months overall and the longest in patients receiving 1L osimertinib (28 months). In the post-2015 cohort, the baseline prevalence of brain metastasis (BM) was 54% and among patients without baseline brain metastasis, the probability of incident BM at 12, 24, and 48 months was 8%, 22%, and 44%, respectively. Development of an on-treatment brain metastasis among patients without baseline brain metastasis was associated with a 3.2 times higher risk of death. Conclusion: Even in a contemporary era with prevalent osimertinib use, the baseline and longitudinal risk of BM development was high. The ongoing risk of developing BM, together with the associated survival detriment, argues for routine surveillance of the brain through magnetic resonance imaging for patients with EGFRm NSCLC, which is not currently included in the guidelines. (c) 2024 The Authors. Published by Elsevier Inc. on behalf of the International Association for the Study of Lung Cancer. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/ 4.0/).

Immune checkpoint inhibitor-mediated colitis (IMC) is a common adverse event of treatment with immune checkpoint inhibitors (ICI). We hypothesize that genetic susceptibility to Crohn’s disease (CD) and ulcerative colitis (UC) predisposes to IMC. In this study, we first develop a polygenic risk scores for CD (PRS CD ) and UC (PRS UC ) in cancer-free individuals and then test these PRSs on IMC in a cohort of 1316 patients with ICI-treated non-small cell lung cancer and perform a replication in 873 ICI-treated pan-cancer patients. In a meta-analysis, the PRS UC predicts all-grade IMC (OR meta =1.35 per standard deviation [SD], 95% CI = 1.12–1.64, P = 2×10 −03 ) and severe IMC (OR meta =1.49 per SD, 95% CI = 1.18–1.88, P = 9×10 −04 ). PRS CD is not associated with IMC. Furthermore, PRS UC predicts severe IMC among patients treated with combination ICIs (OR meta =2.20 per SD, 95% CI = 1.07–4.53, P = 0.03). Overall, PRS UC can identify patients receiving ICI at risk of developing IMC and may be useful to monitor patients and improve patient outcomes.

12092 Background: In a recent meta-analysis, non small cell lung cancers (NSCLC) with ALK or ROS1 fusions had higher rates of VTE compared to KRAS or EGFR mutant NSCLC, and were associated with increased ATE risk, particularly around diagnosis. The underlying mechanisms of how these somatic fusions affect thrombosis remain unclear. We examined clinico-demographic and germline genetic factors associated with VTE and ATE in NSCLC patients with ALK/ROS1 fusions in this first comprehensive genome wide association analysis (GWAS). Methods: In this prospective cohort, germline DNA from whole blood was obtained from 150 patients with ALK fusions and 32 with ROS1 fusions at Princess Margaret Cancer Centre (recruited 2014- 2023). Clinico-demographic, treatment and outcome data were collected from electronic medical charts. Overall survival (OS) by VTE or ATE status was assessed via Cox regression, treating ATE and VTE as time-varying covariates. Genotyping utilized the Infinium Global Screening Array (v3.0 Illumina); quality-control removed 3 patients from final analysis. Using linear regression, outcomes were weighted; no VTE/ATE events (0), one VTE/ATE event (1), or multiple ATE/VTEs (>2). Global significance was set at p < 5 x10-8 and adjusted for age, sex, body mass index (BMI), and ethnic/population stratification (top 3 principal components). Results: There were 97 females (54%) and 82 males (46%), mean age was 57.4 years, 70% had stage 4 disease, 96% adenocarcinoma; 35% experienced VTE/ATE at any time; 13% had 2+ VTE/ATE events. For those with VTE, 32 scored 1, 22 scored 2 and 6 scored 3 on Khorana score. Higher BMI was a significant risk factor for VTE/ATE (adjusted odds ratio 1.59 per 5-unit increase, p=0.01). Shorter OS was observed in patients with VTE (Hazard ratio (HR)=3.15, p<0.001) and ATE (HR=2.51, p=0.036), compared to those without. Two novel GWAS gene peaks on the Manhattan plot (previously not reported to be associated with VTE/ATE) had globally significant associations with risk of VTE/ATE: at Chromosome 6q15 (intergenic region between RNGTT and LOC101928936; 5/10 top variants with top risk-allele p=3.594E-09) and at 15p22 ( TLN2 gene; 2/10 top variants p=1.095E-08). TLN2 is a cytoskeletal protein involved in the assembly of actin filaments and linkages with extracellular matrices. Additional identified variants potentially associated with VTE/ATE in ALK/ROS1 fusion patients were found in: CTBP2, NALCN, WDR7, PAX7, ZNF385D, SORBS2, and CSMD1. Conclusions: Two novel globally significant variants, associated with VTE/ATE, are unique to patients with ALK/ROS1 fusion NSCLC. If validated, these biomarkers may help identify ALK/ROS1 patients who are at highest risk of VTE/ATE who may benefit from prophylactic anticoagulation. Further investigation and clinical evaluation are warranted.

Abstract Background: Adaptive Universal Platform for Real-world Observational Studies (AUPROS) is an emerging study design and platform for the generation of real-world evidence (RWE) particularly in malignancies with rare molecular aberrations. Here, we evaluate the efficiency of AUPROS as an innovative model for real-world observational studies in oncology. Methods: We conducted a mixed-methods study to evaluate three AUPROS studies at Princess Margaret Cancer Centre either locally or as a part of national and international collaborations: 1) METAL (Molecular Epidemiology of ThorAcic Lesions; n=11,378), 2) THANKS (Translational Head And NecK Study; n=3,506), and 3) CARMA (Canadian Cancers with Rare Molecular Alterations; NCT04151342; n=3,879). Specifically, we reviewed data collected, study elements, protocol language, coordination, institutional review boards (IRBs), and contracts. We also performed stakeholder-directed survey and discussions, analysis of funding, research output, and collaborations, as well as a Strengths-Weaknesses-Opportunities-Threats (SWOT) analysis. Results: AUPROS is an innovative study design that borrows design elements from master protocol trials, adaptive trials, and master observational trials as well as retrospective and ambispective designs, enabling comprehensive data collection. The universality of AUPROS allowed for multi-purpose analyses of various real-world data (RWD) including epidemiological, clinical, patient-reported outcomes, biospecimens, and imaging data. The adaptive nature created opportunities for multi-source funding (e.g., academic, pharmaceutical, philanthropic, among others), as well as national/international collaborations and involvement in consortia (e.g., ILCCO, TCGA, BEACON, and INHANCE). Additionally, AUPROS has allowed the development of an expanding national RWE platform that involves collaborations among medical and radiation oncologists and surgeons across different tumor sites. As assessed by research output analyses, AUPROS provided a platform for enhanced research productivity in different areas and utilizing various data sources (METAL=130; THANKS=31; CARMA=5 publications). Our survey and SWOT analysis identified several cost and operational benefits of AUPROS over conventional observational studies, as well as several challenges pertinent to ethics approvals, sustainability, complex coordination, and data quality. Conclusions: AUPROS is an innovative model for real-world observational studies with significant logistical and methodological benefits over conventional RWD study designs. Our findings may help broaden the use of AUPROS to address emerging needs in the scope of precision oncology and clinico-epidemiological research. Citation Format: Samir H. Barghout, Stavroula Raptis, Luna Jia Zhan, Faisal Al-Agha, M Catherine Brown, Devalben Patel, Geoffrey Liu. Adaptive universal platform for real-world observational studies (AUPROS): An emerging model for clinical, epidemiologic, and precision oncology research [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 926.

BackgroundThe 2018 ASCO pleural mesothelioma (PM) treatment guideline states that “a trial of expectant observation may be offered” in patients with asymptomatic inoperable epithelioid mesothelioma with low disease burden. The aim of our analysis was to evaluate clinical characteristics and outcomes in PM-patients managed with initial observation and deferred treatment initiation.MethodsWe retrospectively collected clinicodemograhic and outcome data of patients with inoperable PM. Patients were assigned to two treatment decision groups: decision to start immediate systemic treatment (Immediate Treatment Group) versus observation and deferring treatment (Deferred Treatment group).ResultsOf 222 patients with advanced PM, systemic treatment was started immediately in the majority of patients (189, 85%; immediate group); treatment was deferred in 33 (15%) patients (deferred group); systemic therapy was chemotherapy-based in 91% and 79% respectively. Patients in the deferred group were older (70 vs 67 years, p= 0.05), less likely to have stage IV disease (28% vs 51%, p=0.08) and more often had epithelioid histology (90% vs 70%, p=0.03). Nineteen patients (58%) in the deferred group eventually received treatment.With a median follow-up time of 10.9 months median overall survival (OS) in the entire cohort was 12.4 months and was significantly longer in the deferred group (20.6 months vs 11.5 months, p=0.02). No difference in median progression-free survival (PFS) in first-line treatment between groups was seen (5.4 and 5.3 months).ConclusionThis real-world analysis suggests that deferral of systemic therapy and close observation may not impact OS or physician-assessed PFS in selected PM-patients.Micro-Abstract• The 2018 ASCO pleural mesothelioma (PM) treatment guideline states, “a trial of expectant observation may be offered” in selected patients with inoperable PM.• We performed a retrospective analysis to evaluate outcomes in patients with inoperable PM managed with initial observation and deferred treatment initiation.• Our results suggest that deferral of systemic therapy may not impact overall survival in selected PM-patients.