BACKGROUND:Checkpoint inhibitor pneumonitis (CIP) is a highly morbid complication of immune checkpoint immunotherapy, characterized by acute lung injury leading, in severe cases, to hypoxic respiratory failure and death. CIP incidence in lung cancer is high (10%-15%). Yet, the pathophysiology of CIP is poorly understood. OBJECTIVE/METHODS:To investigate the mechanisms underlying alveolar inflammation in patients with CIP, human bronchoalveolar lavage fluid (BALF) samples from control patients and patients with CIP were analyzed using flow cytometry, single-cell RNA sequencing (scRNA-seq), and ELISA. Findings were validated using multiple external cohorts. In vitro experiments and in vivo rodent models were employed to investigate the mechanisms driving alveolar inflammation in CIP. RESULTS:Analysis of scRNA-seq and flow cytometry data demonstrated increased macrophages in patients with CIP compared to controls. Several distinct proinflammatory alveolar macrophage subsets were increased in CIP. CIP macrophages expressed increased CCL18 at the transcript (scRNA-seq), cellular (flow cytometry) and secreted protein (BALF ELISA) levels. BALF CCL18 levels were associated with clinical CIP severity. CCL18 overexpression in mice promoted lung inflammation that phenocopied human CIP, including upregulation of proinflammatory macrophage subsets. CONCLUSION:These findings suggest that BALF macrophages and CCL18 protein levels are increased in patients with CIP and associate with greater CIP severity. Additionally, CCL18 promotes lung inflammation in mice that mimics human CIP, suggesting a causal role for CCL18 in CIP.

Immune checkpoint inhibitors (ICI), a type of cancer immunotherapy, can cause side effects including inflammatory arthritis (ICI-IA). Previous studies of ICI-IA do not include a thorough characterization of associated immune responses to provide potential targets for treatment. We aimed to identify cytokines uniquely increased in ICI-IA and determine correlations with IA severity and persistence. We evaluated patients diagnosed with ICI-IA by a rheumatologist (n = 80); control serum was obtained from ICI-treated cancer patients without any diagnosed irAEs (n = 17) or diagnosed with an unrelated irAE (n = 19). Serum was assayed to quantify 9 cytokine levels (IFN-γ, IL-4, IL-6, IL-10, IL-12p70, IL-1α, TNF-α, IL-17a, VEGF-A) using MSD U-PLEX assay. Mann-Whitney U tests were performed to evaluate differences in cytokine levels between control and ICI-IA groups. The Kruskal-Wallis test and multivariable ordinal logistic regression were used to determine difference in cytokine levels between patients of differing disease activity. VEGF-A and TNFα were significantly elevated in patients with ICI-IA compared to ICI-controls; results persisted when restricting analyses to patients not treated with immunosuppressants at the time of sampling. ICI-IA patients were stratified by IA severity using CDAI score; there was significantly higher VEGF-A in those with higher disease activity. Ordinal logistic regression showed higher levels of IL-6 and VEGF-A were associated with higher disease activity. Elevated levels of VEGF-A and TNFα are associated with ICI-IA. There was also higher IL-6 and VEGF-A among those with higher disease activity when controlling for confounding. These cytokines could be used as biomarkers of ICI-IA severity and present therapeutic targets.

KM curvces for recurrence based on stage and RVT

BACKGROUND:Immune checkpoint inhibitors (ICI) have revolutionized cancer treatment but can trigger immune-related encephalitis. We report one of the largest case series of patients with immune-related encephalitis and review of the literature. METHODS:Retrospective series of patients with immune-related encephalitis and literature review. RESULTS:Fourteen patients with cancer treated with ICI (50% combination therapy) developed immune-related encephalitis. Diagnostic testing revealed cerebral spinal fluid (CSF) lymphocytic pleocytosis (85%) and elevated protein (69%), abnormal brain magnetic resonance imaging(MRI) (33%) or brain FDG-PET (25%), electroencephalogram (EEG) abnormalities (30%), and autoantibodies (31%). Encephalitis treatment included: corticosteroids (86%), intravenous immunoglobulin (IVIg) (36%), plasmapheresis (7%), and rituximab (29%). There were no deaths and 12 patients had significant recovery, although long-term complications were observed. All patients discontinued ICI. Longitudinal follow-up demonstrated anti-cancer response to ICI at 3 months (85%) and 6 months post-ICI initiation (77%). A literature review identified 132 patients with immune-related encephalitis. Most were treated with PD-1 inhibitors (18% combination). Common abnormalities included elevated CSF protein (84%) or pleocytosis (77%), abnormal brain MRI (65%), or autoantibodies (47%). Nearly all were treated with corticosteroids, many required additional therapy with IVIg (26%) or rituximab (12%). Most patients had clinical improvement (81%) but a minority (10%) had a clinical relapse after completing corticosteroid taper. ICIs were resumed in 7 patients (5%), with relapse in 3. CONCLUSIONS AND RELEVANCE:Immune-related encephalitis is treatable and improves with corticosteroids in most cases but may require additional immunosuppression. Re-emergence of encephalitis is rare and does not typically result in adverse outcomes, and this should be considered in neurological immune-related adverse event management guidelines.

Fatigue is a common and often debilitating adverse event among patients receiving immune checkpoint inhibitor (ICI) therapy, yet its underlying immunologic mechanisms remain poorly defined. We prospectively collected clinical data and blood samples from patients with solid tumors receiving ICI therapy. Patients were prospectively surveyed at month 2, 4, or 6 on treatment to assess for the presence and severity of fatigue as compared with treatment baseline. We analyzed peripheral lymphocyte populations by cytometry by time of flight and peripheral levels of 39 cytokines with a Luminex multiplex assay to identify dynamic immune changes associated with ICI-related fatigue. Of 53 patients enrolled, 31 (58.5%) reported worsening fatigue during ICI therapy. Patients reporting fatigue exhibited broad early-treatment elevations in circulating cytokines, with the most prominent increases observed in the Th1-associated cytokine cluster, including IFN-γ, IL-2, and IL-12. In parallel, several clusters of cytotoxic effector CD8+ T cells expanded significantly from baseline in the fatigued group. Fatigue was not associated with objective tumor response or with the development of other clinically meaningful immune-related adverse events. In a pan-tumor cohort treated with ICIs, increases in clusters of cytotoxic effector CD8+ T cells in parallel with related Th1-associated cytokines were associated with ICI-related fatigue, implicating fatigue as a potential marker of immune activation in this population. SIGNIFICANCE:This study illuminates dynamic changes in peripheral cytokines and immune cell clusters that are associated with ICI-related fatigue. Namely, this study implicates the Th1 pathway as a novel contributor to ICI-related fatigue and identifies ICI-related fatigue as a clinical surrogate for immune activation in patients receiving ICI therapy. Recognizing that fatigue may be a biomarker of heightened immune activity influences monitoring strategies and informs supportive care interventions during immunotherapy.

Immune checkpoint blockade (ICB) is standard of care in advanced diffuse pleural mesothelioma (DPM), but its role in the perioperative management of DPM is unclear. In tandem, circulating tumor DNA (ctDNA) ultra-sensitive residual disease detection has shown promise in providing a molecular readout of ICB efficacy across resectable cancers. This phase 2 trial investigated neoadjuvant nivolumab and nivolumab/ipilimumab in resectable DPM along with tumor-informed liquid biopsy residual disease assessments. Patients with resectable epithelioid/biphasic DPM enrolled sequentially to nivolumab 240 mg every 2 weeks (q2w) for three cycles (Arm A, n = 16) or nivolumab 3 mg kg-1 q2w for three cycles plus ipilimumab 1 mg kg-1 on cycle 1 (Arm B, n = 14), followed by surgery, optional chemotherapy and/or radiotherapy, and nivolumab 480 mg q4w for 1 year. Co-primary endpoints included safety and feasibility; key exploratory endpoints included progression-free survival (PFS), overall survival (OS) and ctDNA analyses. The trial met its primary endpoints, and, in Arms A and B, 81.3% and 85.7% of patients proceeded to surgery, respectively. Treatment was safe, with a single dose-limiting toxicity in each arm. In Arm A, median PFS and OS were 9.6 months (95% confidence interval (CI): 2.5-27.7) and 19.3 months (95% CI: 14.9-34.7), respectively. In Arm B, median PFS and OS were 19.8 months (7.1-not reached) and 28.6 months (20.4-not reached), respectively. Persistent ctDNA was detected during neoadjuvant therapy in patients who did not undergo complete surgical resection due to disease progression (Fisher's exact test, P = 0.00013). Patients with detectable ctDNA on cycle 3 and pre-surgery had shorter PFS (log-rank test, P = 0.027 and P = 0.0059, respectively); this association was more pronounced when quantitative ctDNA changes were considered (log-rank test, P = 1.8 × 10-6). Our findings support the feasibility of neoadjuvant ICB and the clinical utility of ctDNA analyses to capture residual disease in resectable DPM. ClinicalTrials.gov identifier: NCT03918252 .

INTRODUCTION:Treatment-free survival (TFS) characterizes periods of disease control and durable clinical benefit after treatment discontinuation in patients treated with immunotherapy. In CheckMate 227 Part 1, nivolumab plus ipilimumab reported long-term durable overall survival (OS) benefit versus chemotherapy in patients with metastatic NSCLC. Here, we report updated long-term TFS results. METHODS:This analysis included all patients randomized (tumor programmed death ligand 1 [PD-L1] expression ≥1% and <1%). TFS was estimated as the restricted-mean survival time (between Kaplan-Meier curves for time to treatment discontinuation and time to subsequent systemic therapy or death) over 6 years after randomization. TFS was further divided into periods with or without ongoing toxicity (grade 3 or greater treatment-related adverse events) and estimated over 2 and 6 years after randomization. RESULTS:At 6 years after randomization (minimum follow-up: 73.5 months [∼6.1 years]), the estimated OS rate was 20% with nivolumab plus ipilimumab versus 11% with chemotherapy; 13% versus 2% of patients were treatment free. The 6-year mean TFS was 12.2 versus 5.0 months (difference 7.2 [95% confidence interval: 5.4-9.2]), with 17% versus 7% of the 6-year period spent in TFS. The 6-year mean TFS without grade 3 or greater treatment-related adverse events was 11.6 versus 4.8 months (difference, 6.9 [95% confidence interval: 5.1-8.9]). The proportion of mean TFS time increased from 15% of a 2-year to 17% of a 6-year period with nivolumab plus ipilimumab but decreased from 14% to 7% with chemotherapy. Similar results were observed by tumor PD-L1 expression. CONCLUSIONS:Nivolumab plus ipilimumab improved TFS versus chemotherapy, regardless of tumor PD-L1 expression, supporting its use as an efficacious first-line treatment for metastatic NSCLC.

Non-small cell lung cancer (NSCLC) with anaplastic lymphoma kinase rearrangement (ALK+) has a high affinity to form brain metastases (BMs). The cumulative incidence of BMs in ALK + lung cancer is over 50

BACKGROUND:Patients with a high neutrophil/lymphocyte ratio (NLR) have poor prognosis in non-small cell lung cancer (NSCLC). Limited data are available on the contribution of other immune cells. This analysis assessed the prognostic importance of NLR and other peripheral blood cells in patients with advanced NSCLC receiving the PD-1 inhibitor cemiplimab in 2 large phase III studies. METHODS:The impact of baseline immune cells on survival was assessed in patients with complete blood cell counts. Cox proportional hazard regression and Kaplan-Meier methods assessed the relationships between baseline blood cell counts and survival. Data were randomly split into training (70 %) and validation (30 %) cohorts to allow for independent evaluation of the Cox model. RESULTS:Multivariable analyses revealed that a higher NLR (HR: 1.09; 95 % CI: 1.06-1.12, P < .001) and monocytes (HR: 1.49; 95 % CI: 1.15-1.93, P < .001) were strongly associated with an increased risk of death. Higher levels of eosinophils (HR: 0.93; 95 % CI: 0.88-0.99, P < .001) were associated with a reduced risk of death. A calibration curve of observed and predicted probabilities in the unseen test set using independent data revealed that the Cox model was well-calibrated up to a 1-year mortality probability of approximately 30 %. Harrell's concordance index was 0.61, indicating a modest predictive performance. CONCLUSIONS:Our data confirmed the detrimental impact of a high NLR on survival and revealed the importance of monocyte levels in anti-tumor responses, providing useful information to physicians treating advanced NSCLC that may help tailor immunotherapy regimens and provide more accurate prognostic assessments.

Purpose: Co-mutations of the Kirsten rat sarcoma virus (KRAS) and serine/threonine kinase 11 (STK11) genes in advanced non-small cell lung cancer (NSCLC) are associated with immune checkpoint blockade (ICB) resistance. Although neoadjuvant chemoimmunotherapy is now a standard-of-care treatment for resectable NSCLC, the clinical and immunologic impacts of KRAS and STK11 co-mutations in this setting are unknown.Experimental Design: We evaluated and compared recurrence-free survival of resectable KRAS-mutated NSCLC tumors, with or without co-occurring STK11 mutations, treated with neoadjuvant ICB. Single-cell transcriptomics was performed on tumor-infiltrating T cells from seven KRASmut/STK11wt tumors and six KRAS and STK11 co-mutated (KRASmut/STK11mut) tumors.Results: Relative to KRASmut/STK11wt tumors, KRASmut/STK11mut exhibited significantly higher recurrence risk. Single-cell transcriptomics showed enhanced oxidative phosphorylation with evidence of decreased prostaglandin E2 signaling and increased IL-2 signaling in CD8+ tumor-infiltrating lymphocytes (TIL) from KRASmut/STK11mut tumors, a finding that was mirrored in KRASwt tumors that relapsed. TILs from KRASmut/STK11mut tumors expressed high levels of molecules associated with tumor residence, including CD39 and ZNF683 (HOBIT).Conclusions: These divergent T-cell transcriptional fates suggest that T-cell maintenance and residence may be detrimental to antitumor immunity in the context of neoadjuvant ICB for resectable NSCLC, regardless of KRAS mutation status. Our work provides a basis for future investigations into the mechanisms underpinning prostaglandin E2 signaling and IL-2 signaling as they relate to T-cell immunity to cancer and to divergent clinical outcomes in KRASmut/STK11mut NSCLC treated with neoadjuvant ICB.

PURPOSE:Neoadjuvant immune checkpoint blockade with nivolumab plus ipilimumab improves overall survival (OS) in non-small cell lung cancer (NSCLC); however, randomized data for resectable lung cancer are limited. We report results from the exploratory concurrently randomized nivolumab plus ipilimumab and chemotherapy arms of the international phase III CheckMate 816 trial. METHODS:Adults with stage IB-IIIA (American Joint Committee on Cancer seventh edition) resectable NSCLC received three cycles of nivolumab once every 2 weeks plus one cycle of ipilimumab or three cycles of chemotherapy (on day 1 or days 1 and 8 of each 3-week cycle) followed by surgery. Analyses included event-free survival (EFS), OS, pathologic response, surgical outcomes, biomarker analyses, and safety. RESULTS:A total of 221 patients were concurrently randomly assigned to nivolumab plus ipilimumab (n = 113) or chemotherapy (n = 108). At a median follow-up of 49.2 months, the median EFS was 54.8 months (95% CI, 24.4 to not reached [NR]) with nivolumab plus ipilimumab versus 20.9 months (95% CI, 14.2 to NR) with chemotherapy (HR, 0.77 [95% CI, 0.51 to 1.15]); 3-year EFS rates were 56% versus 44%. Higher rates of EFS events were initially seen, with later benefit favoring nivolumab plus ipilimumab; 3-year OS rates were 73% versus 61% (HR, 0.73 [95% CI, 0.47 to 1.14]); pathologic complete response rates were 20.4% versus 4.6%, respectively. In the respective arms, 83 (74%) and 82 patients (76%) underwent definitive surgery. Grade 3-4 treatment-related adverse events occurred in 14% and 36% of patients, respectively. CONCLUSION:Neoadjuvant nivolumab plus ipilimumab showed potential long-term clinical benefit versus chemotherapy, despite early crossing of EFS curves in the preoperative phase and a lower rate of high-grade toxicity.

3052 Background: Genomic profiling through liquid biopsies (LB) has enabled precision oncology decision making, however a key challenge lies in critically interpreting LB data to optimize patient care. Methods: We report results from the first planned interim analysis of an observational biomarker trial, designed to evaluate the clinical utility of serial LB in patients with advanced/metastatic solid tumors (NCT05585684). Primary endpoints were to determine feasibility, prevalence of actionable alterations in LB and the fraction of patients with enacted genotype-matched therapies. Secondary endpoints included progression-free (PFS) and overall survival (OS), time to subsequent therapy and concordance between LB and tumor next generation sequencing (NGS). Exploratory endpoints included correlation of ctDNA dynamics with survival. Serial LBs were obtained at baseline, 1-3 weeks on therapy and at progression using a CAP/CLIA validated NGS panel (Labcorp, MD). Patient-matched white blood cell (WBC) NGS was utilized to identify clonal hematopoiesis (CH)-derived variants. Actionability of genomic alterations was assessed by an ensemble multi-resource programmatic approach; results were reviewed at the Johns Hopkins Molecular Tumor Board (JH MTB). Results: Between March 2023 and July 2024, 51 patients with NSCLC, SCLC and esophageal cancer were enrolled, with 45 evaluable baseline and 12 progression LBs reviewed at JH MTB. Median turnaround time from baseline and progression LB to MTB recommendation was 14 and 13 days respectively. Patient-matched analyses of baseline WBC samples revealed 30.1% (n = 22) CH-derived alterations. The frequency of actionable variants was 28.8% (n = 21) at baseline, 23.3% (n = 7) on therapy and 21.7% (n = 5) at progression. Of the 45 patients reviewed at baseline, 33 received a recommendation for genotype-matched therapies; 48.5% (n = 16) based on tumor molecular profiling, 15.2% (n = 5) based on LB alone and 36.3% (n = 12) based on LB and tissue NGS. Thirteen patients were treated according to MTB recommendations. Patients who were treated with genotype-matched MTB recommended therapies had longer OS and PFS compared to those who received alternate therapies (not reached-NR vs. 14.8 months, log-rank p = 0.028 and NR vs 6.2 months, log-rank p = 0.21 respectively). Among the 12 patients reviewed at progression, 5 received an MTB recommendation for genotype-tailored therapies based on LB alone (n = 3) or in combination with tissue NGS (n = 2). Early on-therapy ctDNA clearance was associated with longer PFS and OS (log rank p = 0.02 and p = 0.06). Conclusions: Our findings highlight the value of a multidisciplinary MTB when supported by comprehensive liquid biopsy molecular information to inform therapy selection and improve patient outcomes. Clinical trial information: NCT05585684 .

LBA8000 Background: NIVO + chemo is an established standard of care neoadjuvant treatment (tx) for eligible patients (pts) with resectable NSCLC and has shown statistically significant and clinically meaningful improvements in EFS and pCR in the phase 3 CheckMate 816 study. Here, we report the planned final analysis of OS from CheckMate 816 at 5-y follow-up (f/u). Methods: Adults with stage IB (≥ 4 cm)–IIIA (per AJCC v7) resectable NSCLC, ECOG PS ≤ 1, and no known EGFR / ALK alterations were randomized 1:1 to receive neoadjuvant NIVO + chemo Q3W or chemo alone Q3W for 3 cycles, followed by surgery. Primary endpoints were EFS and pCR (both by blinded independent review). OS was a key prespecified, statistically powered secondary endpoint that was tested hierarchically. Exploratory analyses included OS by ctDNA clearance and pCR status. Results: At a median f/u of 68 mo (range, 60–85; database lock, 23 Jan 2025), neoadjuvant NIVO + chemo demonstrated a statistically significant OS benefit vs chemo alone (median [95% CI], not reached [NR] vs 73.7 mo [47.3–NR]; HR [95% CI], 0.72 [0.523–0.998]; P = 0.0479); 5-y OS rates were 65% vs 55%. OS favored NIVO + chemo in the subgroups defined by tumor PD-L1 expression, baseline disease stage, and histology (Table). In an exploratory analysis in pts with ctDNA+ at baseline (NIVO + chemo, n = 43; chemo, n = 43), pts with presurgical ctDNA clearance (56% vs 35%) had continued OS improvement vs those without across both tx arms (HR [95% CI]: NIVO + chemo, 0.38 [0.15–1.00]; chemo, 0.39 [0.14–1.11]). Furthermore, pts who had pCR with NIVO + chemo had sustained OS improvement vs those without (HR [95% CI], 0.11 [0.04–0.36]; 5-y OS rates, 95% vs 56%). Neoadjuvant NIVO + chemo continued to improve EFS vs chemo (median [95% CI], 59.6 [31.6–NR] vs 21.1 mo [16.5–36.8]; HR [95% CI], 0.68 [0.51–0.91]); 5-y EFS rates were 49% vs 34%. No new safety signals were observed at this long-term f/u. Conclusions: CheckMate 816 is the only neoadjuvant-only immunotherapy phase 3 trial to demonstrate a statistically and clinically significant OS benefit at 5 y for a resectable solid tumor. Pts with pCR with neoadjuvant NIVO + chemo had a ~90% reduction in their risk of death by 5 y compared with those without pCR. The findings show long-term survival benefit from a short course of neoadjuvant NIVO + chemo and affirm a paradigm shift in the tx of resectable NSCLC without actionable genomic alterations. Clinical trial information: NCT02998528 . All pts PD-L1 < 1% PD-L1 ≥ 1% Stage IB/II Stage IIIA Squamous Non-squamous NIVO + chemo (N = 179) vs chemo (N = 179) NIVO + chemo (n = 78) vs chemo (n = 77) NIVO + chemo (n = 89) vs chemo (n = 89) NIVO + chemo (n = 65) vs chemo (n = 61) NIVO + chemo (n = 113) vs chemo (n = 116) NIVO + chemo (n = 87) vs chemo (n = 95) NIVO + chemo (n = 92) vs chemo (n = 84) Median OS, mo NR vs 73.7 NR vs 61.8 NR vs 73.7 NR vs 76.8 NR vs 73.7 NR vs 73.7 NR vs NR HR (95% CI) 0.72 (0.523–0.998) 0.89 (0.57–1.41) 0.51 (0.31–0.84) 0.77 (0.44–1.35) 0.70 (0.47–1.05) 0.71 (0.46–1.11) 0.72 (0.45–1.16)

8526 Background: The treatment paradigm for metastatic non-small cell lung cancer (mNSCLC) without actionable genomic alterations does not differentiate between histologic subtypes for the use of checkpoint inhibitors. However, there is growing recognition of differences between squamous (SQ) and non-squamous (NSQ) lung cancer that may impact response to treatment. For example, in the RELATIVITY-104 study, addition of the LAG-3 inhibitor relatlimab to anti-PD-1 + platinum-doublet chemotherapy (PDCT) showed improved clinical benefit among patients with PD-L1 ≥1%, which was further enriched with NSQ histology. There is an unmet need to understand differences in tumor biology between NSQ and SQ histologies in patients with mNSCLC to inform on mechanisms underlying differences in clinical activity of anti-PD-1 + PDCT, alone or in combination with a LAG-3 inhibitor. Methods: Data were obtained from molecular profiling of baseline tumor samples of treatment-naive patients enrolled in the phase 3 CheckMate 227 (NCT02477826) study. PD-L1 (N=1739) and LAG-3 expression (N=540) were evaluated using immunohistochemistry. Somatic mutations and copy number alterations were assessed using the FoundationOne panel (N=1368). Gene expression, analyzed through RNA sequencing (N=465), was used to characterize differences in tumor immunobiology, including differential gene expression, pathway enrichment, and calculation of cell type-specific scores representing different immune and stromal cell types. Results: Transcriptional and mutational analyses revealed clear differences between NSQ and SQ tumors. NSQ tumors showed enrichment of immune pathways (e.g., antigen presentation and T cells), while SQ tumors exhibited enrichment of pathways consistent with rapid cell growth and numerous oncogenic alterations (e.g., p53 and PIK3CA). Differences in the relationship between tumor PD-L1 expression and the tumor microenvironment by histology were observed; PD-L1 expression was positively correlated with immune infiltration scores in NSQ but not SQ tumors, suggesting that drivers of PD-L1 expression may differ by histology. Consistent with previously published reports for mNSCLC, PD-L1 expression enriched for 1L anti-PD-1 + PDCT benefit in NSQ but not SQ tumors (CheckMate 227 Part 2). Differences in LAG-3 ligand expression by histology and PD-L1 expression were noted. Both canonical LAG-3 ligands, MHC-II and FGL-1, were expressed at higher levels in NSQ tumors. Within NSQ tumors, relative expression of each ligand varied by PD-L1 expression, with high MHC-II expression specifically in NSQ, PD-L1 ≥1% tumors. Conclusions: These data provide a supporting mechanistic rationale for the use of tumor histology in addition to PD-L1 expression to identify patients who would benefit from the addition of a LAG-3 inhibitor to PD-1 inhibitor + PDCT. Clinical trial information: NCT02477826 .

Probabilistic spatial modelling techniques developed on large-scale tumor-immune Atlases (~35M individually mapped cells; 50,000 high power fields) were used to characterize predictive features of treatment-responsive lung cancer. We identified CD8+FoxP3+ cell density as a robust pre-treatment biomarker for outcomes across disease stages and therapy types. In parallel, single-cell RNAseq studies of CD8+FoxP3+ T-cells revealed an activated, early effector phenotype, substantiating an anti-tumor role, and contrasting with CD4+FoxP3+ T-regulatory cells. A spatial biomarker was developed using an empirical probabilistic model to define the immediate cell neighbors or niche surrounding CD8+FoxP3+ cells and proximity to the tumor-stromal boundary. The resultant 'Diversity of Niches Unlocking Treatment Sensitivity (DONUTS)' are more prevalent than the CD8+FoxP3+ cells themselves, mitigating sampling error in small biopsies. Further, the DONUTS only require four markers, are additive to PD-L1, and associate with tertiary lymphoid structure counts. Taken together, the DONUTS represent a next-generation predictive biomarker poised for clinical implementation.

KRAS G12C-mutated non-small cell lung cancer (NSCLC) is linked to higher risk of brain metastases (BM). However, clinicopathologic and genomic characteristics associated with BM in this population are poorly understood. This is a single-center, retrospective analysis of patients with KRAS G12C-mutated NSCLC. Clinicopathologic and genomic characteristics including PD-L1 status; tumor mutation burden (TMB); and specific co-mutations in TP53, STK11, and KEAP1 were assessed. Treatment patterns were characterized for the BM sub-cohort. Chi-squared, ANOVA, and Kaplan-Meier analyses were used to evaluate association of covariates with BM occurrence and overall survival. Among 460 patients with KRAS G12C-mutated NSCLC treated from 2013-2023, 107 (23.3%) developed BM, with 54.2% of cases identified synchronously (within 60 days of initial diagnosis). Median time to metachronous BM diagnosis from primary diagnosis was 16.9 months. Solitary BM were present in 37.4% of all patients, while 28.0% had ≥4 lesions. For patients receiving comprehensive genomic testing (N=197), BM were associated with increased rates of co-mutations in TP53, STK11, and KEAP1 (p=0.049), but no differences in TMB or PD-L1. Patients with synchronous BM had median OS of 8.0 months from BM diagnosis while those with metachronous BM had median OS of 11.1 months, though this difference was not statistically significant. Upfront interventions following BM diagnosis were surgical resection (33.6%), stereotactic radiation (24.3%), systemic therapy (16.8%), and whole brain radiation (12.1%). Whole brain radiation differed for patients with one vs. multiple lesions (p=0.029). Patients who received no cancer-directed treatment due to poor performance had more frequent co-mutations. Discussion: BM in KRAS G12C-mutated NSCLC was associated with higher prevalence of co-mutations in TP53, STK11, and KEAP1. Given the heterogeneity of features influencing treatment response and observed variability in treatment strategies, larger studies examining patients with active CNS lesions are needed to inform management of BM in this population.

8518 Background: In previously treated KRAS G12C MUT NSCLC, the allosteric KRAS G12C inhibitor sotorasib had superior outcomes compared to docetaxel (ORR 28% vs 13%). S1900E was the first to prospectively test sotorasib in KRAS G12C MUT NSCLC according to co-mutations (CO-MUT) in tumor suppressor genes such as TP53, STK11, and KEAP1 . We report on the results of TP53 and STK11 CO-MUT cohorts of S1900E and hypothesized that CO-MUT would not impact the efficacy of sotorasib. Methods: Pts with KRAS G12C MUT identified by FoundationOne CDx tissue assay in the LUNGMAP screening master protocol were assigned to S1900E. Pts with stage IV/recurrent non-sq NSCLC who had progressed after ≥1 line of systemic therapy, and were ECOG PS 0-1 were eligible. There were 3 biomarker cohorts: 1 ( TP53 CO-MUT & wild type [WT] STK11 , KEAP1 / NFE2L2 / CUL3) ; 2 ( STK11 CO-MUT & WT TP53 , KEAP1/NFE2L2 / CUL3) ; 3 (all others). The primary objective was to evaluate the confirmed objective response rate (ORR) per RECIST 1.1 in each cohort. Accrual goals for Cohorts 1 and 2 were 40 and 25 evaluable pts, respectively, based on a 1-stage binomial design with 90% power to rule out a 14% ORR (historical second-line docetaxel ORR) at the 1-sided 5% level. Results: S1900E completed accrual with 118 total pts and 103 evaluable from Apr 2021-Dec 2024; 59 (57%) were female and 86 (84%) were non-Hispanic white. In the TP53 CO-MUT (N=48; 40 evaluable) and STK11 CO-MUT (N=28; 25 evaluable) cohorts, respectively, 48% and 68% received only one prior line of therapy, 70% and 76% received both platinum chemotherapy and PD-(L)1 immunotherapy, 68% and 24% were female, known PD-L1 expression (≥1% / ≥50%) was 95%/45% and 43%/0%, and almost all had smoked. In the TP53 CO-MUT cohort, confirmed ORR was 35% (CI 23-47). In the STK11 CO-MUT cohort, confirmed ORR was 16% (CI 4-28). Disease control rate (DCR), duration of response (DOR), investigator progression-free survival (PFS), and overall survival (OS) (Table 1) had numerically higher values in the TP53 CO-MUT cohort. Adverse event rates ≥ Grade 3 were similar to prior reports of single agent sotorasib. Conclusions: TP53 CO-MUT cohort met its primary endpoint, while the STK11 CO-MUT cohort did not, suggesting that STK11 CO-MUT have detrimental effect on sotorasib in KRAS G12C NSCLC. S1900E Cohort 3, which may include KEAP1/NFE2L2 and other CO-MUT, will be reported later, as will resistance patterns identified through ctDNA analysis. Clinical trial information: NCT04625647 . TP53 CO-MUT (N=40) STK11 CO-MUT (N=25) ORR (90% CI) 35% (23-47) 16% (4-28) DCR (90% CI) 78% (67-88) 60% (44-76) Follow-Up Median mo 19.5 16.8 DOR [Median mo (95% CI)] 7.1 (2.7-11.5) 6.2 (1.6-NA) PFS [Median mo (95% CI)] 5.7 (3.0-8.4) 4.1 (2.6-7.1) OS [Median mo (95% CI)] 18.2 (12.2-33.7) 8.0 (5.1-14.2)