Estimating the Impact of Adjuvant Treatment With Nivolumab on Long-Term Survivorship Rates Compared With Surveillance in Muscle Invasive Urothelial Carcinoma: Mixture Cure Modeling Analyses of Disease-Free Survival From the Phase 3 CheckMate 274 Trial.

TPS9593 Background: NIVO (anti-PD1) alone or in combination with IPI (anti-CTLA-4) or RELA (anti-LAG3) are approved immune checkpoint blockade (ICB) agents for the treatment (tx) of patients (pts) with advanced, unresectable metastatic melanoma. Doublet combinations induce higher rates of durable disease control vs single agent, translating into nominal improvements in survival. While there is no established dose-response relationship for NIVO alone or with RELA, IPI at higher doses induces higher objective response rate (ORR) but increased grade ≥3 immune-related adverse events. Deeper mechanistic understanding points towards potential synergy given IPI’s role in expanding the TCR repertoire and modulating suppressive T cell populations while NIVO+RELA regulate the exhaustion signatures of activated T cells and allow for improved effector function. Recently, results from RELATIVITY-048 combining all three ICB agents (NIVO 480 mg Q4W + RELA 160 mg Q4W + IPI 1 mg/kg Q8W) demonstrated impressive efficacy with high response rates (59% ORR) and seemingly improved progression free and overall survival (PFS, OS) over previously reported doublet regimens. This study evaluated a markedly lower IPI dose than the approved regimen and did not include a dose escalation component to optimize the IPI dosing strategy. Our team seeks to optimize the dose and schedule of IPI to combine with NIVO+RELA in order to determine the recommended phase II dose (RP2D) for triplet ICB and maximize clinical benefit while maintaining a toxicity profile comparable to approved regimens. Methods: In this single center, investigator initiated, phase I/IIa study evaluating triplet ICB (NCT06683755), all pts will receive FDA approved regimen of NIVO 480mg + RELA 160mg IV Q4W along with escalating doses of IPI . Dose escalation (DE) with IPI will begin at 0.5mg/kg Q4W 4 induction doses and will incrementally escalate up to 2mg/kg Q4W. Maintenance tx will consist of NIVO+RELA Q4W. Bayesian optimal interval (BOIN) design will be used to identify the maximum tolerated dose (MTD) and RP2D (primary objective) in the DE portion, accruing an estimated 12-18 pts. The PhIIa portion will accrue an additional 12-18 pts at the RP2D to better characterize safety, and determine the ORR, (primary objective) by RECIST 1.1. Secondary objectives include PFS, OS, and tumor and immunological correlatives obtained on pre and post tx blood and tumor samples. Pts must be previously untreated, unresectable, or advanced melanoma. Non IPI containing prior adjuvant or neo-adjuvant tx will be permitted if the last dose has been >6 months. Pts with asymptomatic brain metastasis are allowed, provided no immunesuppressive doses of corticosteroids are required. Safely biopsiable lesions are required for pts enrolled in the PhII portion. This study is open for accrual at MD Anderson Cancer Center in Houston, Texas. Clinical trial information: NCT06683755 .

e16087 Background: Nivolumab (NIV) plus chemotherapy (CHT) is the standard first-line treatment for advanced gastric and gastroesophageal junction (GEJ) adenocarcinoma with PD-L1 combined positive score (CPS) ≥ 5, based on phase 3 trials. Real-world data are essential to validate these results. Methods: We conducted a multicenter retrospective study (2022-2024) on 90 patients (pts) with PD-L1 CPS ≥ 5 gastric/GEJ adenocarcinoma treated with first-line CHT plus NIV. Clinical pathological features are described in Table 1. Endpoints included progression-free survival (PFS), overall survival (OS), objective response rate (ORR), and disease control rate (DCR). We assessed treatment response using the RECIST (Response Evaluation Criteria in Solid Tumors) criteria: complete response (CR), partial response (PR), stable disease (SD), and progressive disease (PD). Results: At a median follow-up of 7 months (m), median OS was 14.6 m (95%CI: 11.9–21.24), and median PFS was 11.78 m (95%CI: 7.6–15.6). ORR was 46.7%, and DCR was 76.7%. Responders (CR/PR) had better outcomes than non-responders (SD/PD): median PFS of 15.6 m (95%CI: 11.5–25.5) vs. 4.3 m (95% CI: 3.2–9) (p < 0.0001) and median OS not reached vs. 10.2 m (95%CI: 6.8–13.9) (p < 0.0001). Early progressors (PD within 3–6 m) had worse outcomes, with median OS of 5.5 m (95%CI: 3.9–10.2) and median PFS of 2.3 m (95%CI: 1.9–3) vs. OS of 21.2 m (95%CI: 13.8–21.2) and PFS of 15.6 m (95% CI: 11.5–19) in other pts (p < 0.0001). Immune-related toxicity occurred in 21 pts (23.3%), leading to treatment discontinuation in 6 cases (7.1%). Conclusions: This real-world study confirmed the efficacy and safety of NIV plus CHT in advanced gastric/GEJ adenocarcinoma with PD-L1 CPS ≥ 5. Response to treatment emerged as a key prognostic factor for survival. Further follow-up is needed to validate these findings and identify predictive factors for response and progression. Clinicopathologic features. Clinicopathologic features N=90(%) Age>50≤50 78 (86.7)12 (13.3) GenderMF 58 (64.4)32 (35.6) ECOG PS0-1≥2 84 (93.3)6 (6.7) Primary tumor locationstomachcardia/GEJ 48 (53.3)42 (46.7) Primary tumor resectionYN 32 (35.5)58 (64.4) Histology Type IntestinalDiffuse signet ring cellsNot Assessed 23 (25.5)35 (38.8)19 (21.1)32 (35.5) MMR/MSI status pMMR/MSSdMMR/MSI-HNot Assessed 76 (84.4)10 (11.1)4 (4.5) PD-L1 CPS5≤CPS<10≥10 27 (30)57 (63.3) Metastasis Siteslivernodeslungperitoneumother 24 (26.7)38 (42.2)5 (5.5)24 (26.7)9 (10) Chemotherapy regimenFOLFOXXELOX5FU 81 (90)5 (5.5)4 (4.5)

12129 Background: Despite the transformative impact of ICIs on cancer treatment, their efficacy remains limited by adverse events (AEs), underscoring the need for reliable biomarkers. Here, we aimed to investigate the effect of a previously identified IL7 SNP in predicting irAEs across two clinical trials and an East Asian pan-cancer cohort. Methods: In this pooled analysis, we included 1,205 pts from the CheckMate-025 trial (CM025, NCT01668784) with renal cell carcinoma (RCC) who received either nivolumab (NIVO) or everolimus (EVE), from the BinTA-0037 (BTA-037, NCT03631706) in non-small cell lung cancer (NSCLC) treated with pembrolizumab (PEMBRO), and from the Asan ICI-treated pan-cancer cohort. The rs7816685 SNP dosages were inferred from blood and/or tumor whole exome sequencing (WES) using STITCH for CM025, and Minimac4 for Asan. For BTA-037, a surrogate SNP (rs16906062, R 2 =1.0) was extracted from tumor WES. The association between the SNP carrier status and the time to incident AEs was investigated via multivariable cause-specific Cox regression models. RNA-sequencing (RNA-seq) was performed on blood samples from the Asan cohort collected pre- and post-initiation of ICI. Blood immune cell fractions were estimated from RNA-seq data using ImmucellAI. Results: The frequency of the risk allele was 15% in CM025, 17% in BTA-037, and 24% in Asan. IL7 SNP carriers demonstrated a significantly higher risk of AEs when treated with ICI therapies in all 3 cohorts, but not with EVE (non-ICI control) (SNP´treatment P interaction =0.0012 in CM025) (Table). The SNP showed a consistent effect across different tumor types and irAE profiles, with no apparent impact on survival outcomes. RNA-seq data revealed the expression of a novel IL7 cryptic exon in carriers, and a significant increase in peripheral cytotoxic T-cell post-ICI (q=0.002). Both features were significantly correlated (R=0.29, P=1x10-11), suggesting a potential mechanistic link. Conclusions: The IL7 SNP (rs7816685) is associated with a higher risk of immune toxicity in pts treated with ICI. Overall, our findings support the use of this germline biomarker for irAE risk stratification, and pave the way for future functional studies. Adjusted hazard ratios (HRs) from multivariable Cox models adjusting for baseline covariates in each cohort. Cohort Cancer type Treatment N SNP adjusted HR for irAEs CM025 RCC NIVO 189 rs7816685 3.01 [1.59-5.68], P =0.0007 CM025 RCC EVE 193 rs7816685 0.65 [0.33-1.28], P =0.22 BTA-037 NSCLC PEMBRO 152 rs16906062 2.3 [1.16-4.6], P =0.017 Asan Pan-cancer Any ICI 671 rs7816685 1.12 [1.02-1.2], P =0.015

Immune checkpoint inhibitor (ICI) therapy, although effective in treating patients with a variety of advanced malignancies, can result in potentially severe or even fatal immune-related adverse events (irAEs). This study aimed to generate real-world evidence on irAE characteristics, clinical management and clinical outcomes among patients with advanced (unresectable or metastatic) malignant melanoma (a/mMM) or advanced renal cell carcinoma (aRCC) treated with nivolumab (NIVO) and/or ipilimumab (IPI) in the UK. This was a multi-centre, retrospective cohort study of adult patients with a/mMM or aRCC, who received NIVO and/or IPI at one of five specialist treatment centres in the UK between 1 January 2016 and 31 March 2020. The incidence and grading of irAEs were described, as well as time to irAE onset, the management of irAEs and overall survival (OS). In total, 199 patients were included in the study: 162 with a/mMM and 37 with aRCC. The majority of patients in both a/mMM (75.3%) and aRCC (62.2%) cohorts reported any irAE, while 45.9% and 30.4% in the a/mMM and aRCC cohorts reported grade 3 or 4 irAEs, respectively. Colitis/diarrhoea, skin reactions and hepatitis were most frequently reported, and the predominant treatment prescribed for any irAE was corticosteroids only. Analysis indicated a positive association between the development of an irAE and longer OS. Findings from this study support current literature, provide further insights into the characteristics and clinical management of irAEs and support an association between the development of an irAE and improved OS in these patient groups.

BackgroundPD‐1 or BRAF + MEK inhibition is considered the current gold standard in adjuvant melanoma therapy. Little is known if, after the recurrence of the disease and surgery, a second course of adjuvant therapy might be beneficial.MethodsA multicenter, retrospective study investigating a second course of adjuvant therapy after recurrence and surgery in stage III–IV melanoma patients. Patients received nivolumab (NIV), pembrolizumab (PEM) or dabrafenib plus trametinib (D + T) between 01/2017 and 10/2021. The primary endpoint was 12‐month recurrence‐free survival (RFS2). Further analyses included descriptive and correlative statistics.ResultsSixty‐six patients from 22 centers in Germany, Austria and Switzerland were included. Thirty‐two patients received D + T as second‐course adjuvant therapy, 9 patients received PEM and 25 patients received NIV. Recurrence‐free survival for the second‐course adjuvant treatment (RFS2) was assessed after 12 and 24 months and showed a superiority of adjuvant BRAF + MEK over PD‐1 therapy (12‐months RFS2: 90.6% vs. 70.6%, HR 4.226 [95% CI 1.154–15.48]; p = 0.030; 24‐months RFS2 71.9% vs. 52.9%, HR 3.154 [95% CI 1.374–7.242]; p = 0.007). There was no significant decrease in OS with either BRAF + MEK or PD‐1 treatment (12‐months OS: 100% both, 24‐months OS: 100% vs. 93.8%).Furthermore, therapy sequences were investigated. For better comparability, only BRAF V600 mutated patients were assessed: RFS2 was significantly better for patients with a class switch from PD‐1 to BRAF + MEK compared to BRAF + MEK to PD‐1 (HR 4.401 (1.04–18.63), p = 0.044). No new safety signals were detected.ConclusionsIn the investigated cohort, a second course of adjuvant melanoma treatment is feasible and provides similar RFS compared to an initial course of adjuvant therapy using BRAF + MEK inhibitors; however, RFS2 is reduced for PD‐1 antibodies. In addition, both treatments were convincing with a 24‐month OS of almost 100%. Switching from adjuvant PD‐1 to BRAF + MEK treatment provided better overall RFS compared to switching from adjuvant BRAF + MEK to PD‐1 treatment.

Bempegaldesleukin plus nivolumab in first-line advanced/metastatic urothelial carcinoma: Results from a phase II single-arm study (PIVOT-10)

Abstract Background: In CheckMate 816, lower %RVT in primary tumor (PT) and lymph node (LN) after neoadjuvant (neoadj) NIVO + chemo correlated with improved EFS in patients (pts) with resectable NSCLC. To further evaluate %RVT as a surrogate for EFS, we report an exploratory analysis of efficacy with adjuvant (adj) NIVO after neoadj treatment (tx) by LN involvement, nodal (N) status, and %RVT in PT and LN in CheckMate 77T. Methods: Pts with resectable stage IIA-IIIB NSCLC were randomized to neoadj NIVO + chemo Q3W (up to 4 cycles [cyc]) followed by adj NIVO Q4W (up to 13 cyc) or neoadj placebo (PBO) + chemo Q3W (up to 4 cyc) followed by adj PBO Q4W (up to 13 cyc). Primary endpoint: EFS per BICR. This analysis, which included pts with pathologically evaluable samples who had definitive surgery and ≥ 1 adj tx dose, assessed EFS by LN involvement, N status, %RVT in PT and LN, and associations between %RVT and EFS per time-dependent ROC curve analysis. Results: BL characteristics were similar between tx arms (NIVO, 123; PBO, 134; median f/u, 33.3 mo). NIVO improved EFS v PBO regardless of LN involvement or N status (Table). A higher proportion of pts treated with NIVO had 0% RVT in PT and/or LN v PBO (52% v 20%). In pts with LN involvement, 2-y EFS rates with NIVO were higher in pts with 0% RVT in both PT and LN (90%) or 0% RVT in PT or LN (85%) v > 0% RVT in both PT and LN (76%). Area under the ROC curve for %RVT-PT in pts with PT-only disease was 0.83. 2-y EFS rates with NIVO were 94%, 77%, and 50% in pts with 0-5%, > 5-80%, and > 80% RVT-PT, respectively; similar results were seen in all pts with pathologically evaluable samples whether they received adj tx or not. Conclusions: In this exploratory analysis, NIVO improved EFS v PBO, particularly in pts with LN involvement and regardless of N status. %RVT also associated with EFS in a continuous manner, supporting %RVT as a surrogate for EFS and highlighting its prognostic value in pts who receive perioperative NIVO. Citation Format: Julie Stein Deutsch, Ashley Cimino-Mathews, Elizabeth Thompson, Edward Gabrielson, Peter Illei, Jaroslaw Jedrych, Ezra Baraban, Alex S. Baras, Mariano Provencio Pulla, Tina Cascone, Jonathan D. Spicer, Mark M. Awad, Fumihiro Tanaka, Jie He, Shun Lu, Cinthya Coronado Erdmann, Vipul Devas, Sumeena Bhatia, Janis M. Taube. Associations between percent residual viable tumor (%RVT) and efficacy with perioperative nivolumab (NIVO) for resectable NSCLC in CheckMate 77T [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 2 (Late-Breaking, Clinical Trial, and Invited Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_2):Abstract nr CT097.

Abstract Background: The composition of the tumor microenvironment (TME) can regulate T cell infiltration and potentially immunotherapy efficacy. Versican (VCAN), an extracellular matrix proteoglycan, regulates the presence of tumor infiltrating lymphocytes (TILs) and their activation, while its proteolytic byproduct, versikine (Vkine), can stimulate an immune response. Here we aim to validate our prior findings that VCAN status correlates with CD8+ T cell infiltration in intestinal cancers and evaluate the predictive role of VCAN/Vkine for patients undergoing immunotherapy for microsatellite instability high (MSI-H) cancers. Methods: A total of 28 consented subjects were identified with intestinal cancers that were MSI-H, high tumor mutation burden, or with alterations in mismatch repair genes as part of an IRB-approved protocol. Baseline characteristics, immunotherapy response, and survival data were collected. Cancer samples were obtained and stained for VCAN, Vkine, and CD8+ TILs. An intensity binning system was used to score VCAN and Vkine, ranging from 0-3. TILs were averaged across high power fields (HPF). Cancers with high VCAN (2-3) and any Vkine were categorized as VCAN proteolytic weak (VPW), cancers with low VCAN (0-1) and high Vkine (2-3) as VCAN proteolytic predominant (VPP), and cancers with both low VCAN (0-1) and low Vkine (0-1) as VCAN/Vkine low (VVL). Results: Clinical characteristics of the 28 subjects studied: median age 62.5 years (32-90); 17 females; 19 colon cancers; six rectal cancers; one rectosigmoid cancer and two small intestinal cancers. Further, nineteen were stage 3, nine were stage 4 and 71% were MSI-H. Six were VVL, one was VPP and 21 were VPW. The VVL and VPP cancers had a median of 27 CD8+ TILs/HPF, while the VPW cohort had a median of 15.7 (p=0015). CD8+ TILs in the MSI-H VPW subset trended higher than the microsatellite stable VPW cancers (21.7 versus 8.5; p=0.35). Thirteen subjects with MSI-H cancers received immunotherapy (10 pembrolizumab (PEMBRO), three ipilimumab (IPI) and nivolumab (NIVO)). Of these cancers, six were VPW and completed PEMBRO. None of the VPW cancers achieved a complete response to PEMBRO alone with most developing progressive disease. An additional VPW subject is still receiving therapy with a partial response to PEMBRO to date. Two patients with VPW cancers received IPI and NIVO and had complete responses. The four subjects with VVL or VPP cancers have all achieved a complete response and are alive without cancer recurrence. Three were treated with PEMBRO and one with IPI and NIVO. Conclusion: The VCAN status in the TME of intestinal cancers shows promise as a biomarker for T cell infiltration and response to immune checkpoint blockade. The ability of VCAN status to identify patients who are more likely to benefit from dual immune checkpoint blockade deserves further investigation. Citation Format: Elizabeth L. Field, Sean Kraus, Madeleine Buratti, Heidi Koehnke, Cheri A. Pasch, Fotis Asimakopoulos, Dustin A. Deming. Versican abundance/proteolysis as a biomarker for immunotherapy efficacy in intestinal cancers. [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 3282.

Approximately 50% of patients with advanced Merkel cell carcinoma (MCC) have primary or acquired resistance to PD-(L)1 blockade, which may be overcome using combination immune checkpoint inhibition (ICI) with anti-cytotoxic T lymphocyte antigen-4 antibody. We present results from the recurrent/metastatic MCC cohort in CheckMate 358, a nonrandomized, multicohort, phase I/II study of nivolumab (NIVO) with or without ipilimumab (IPI) in virus-associated cancers (ClinicalTrials.gov identifier: NCT02488759). ICI-naïve patients with recurrent/metastatic MCC and 0-2 previous systemic therapies were administered NIVO monotherapy at 240 mg once every 2 weeks or combination therapy with NIVO 3 mg/kg once every 2 weeks + IPI 1 mg/kg once every 6 weeks. The primary end point was objective response. Secondary end points included duration of response (DOR), progression-free survival (PFS), and overall survival (OS). Sixty-eight patients received NIVO (n = 25) or NIVO + IPI (n = 43). The objective response rate (95% CI) and median DOR (95% CI), respectively, were 60% (38.7 to 78.9) and 60.6 months (16.7 to not applicable [NA]) with NIVO and 58% (42.1 to 73) and 25.9 months (10.4 to NA) with NIVO + IPI. The median PFS (95% CI) and OS (95% CI), respectively, were 21.3 (9.2 to 62.5) and 80.7 (23.3 to NA) months with NIVO and 8.4 (3.7 to 24.3) and 29.8 (8.5 to 48.3) months with NIVO + IPI. The incidence of grade 3/4 treatment-related adverse events was 28% with NIVO and 47% with the combination. This nonrandomized study showed frequent and durable responses with both NIVO and NIVO + IPI in patients with ICI-naïve advanced MCC. However, it did not show improvement in efficacy with the combination, thus contradicting previous study reports that had suggested clinical benefit with combination ICI. A randomized trial of NIVO + IPI versus NIVO monotherapy is warranted.

208P: Pathological response after neoadjuvant chemotherapy (chemo) and nivolumab (NIVO) in resectable non-small cell lung cancer (NSCLC): Initial results from a French multicenter real-world (rw) retrospective study

138P: Outcomes by treatment (tx) cycles (cyc) with perioperative nivolumab (NIVO) in patients (pts) with resectable NSCLC from CheckMate 77 T study

127TiP: RELATIVITY-1093: A randomized, phase III study of first-line nivolumab (NIVO) + relatlimab (RELA) high dose (HD) with platinum-doublet chemotherapy (PDCT) vs pembrolizumab (PEMBRO) + PDCT in patients with metastatic NSCLC having tumor PD-L1 expression ≥ 1% and non-squamous (NSQ) histology

540 Background: Ipilimumab (IPI) and nivolumab (NIVO) are standard first-line systemic therapy for patients with metastatic renal cell carcinoma (RCC). The regimen is administered in combination once every 3 weeks for 4 doses, followed by NIVO maintenance. The dose and frequency of IPI appears to correlate with treatment safety and tolerability across cancer types. Further, studies in advanced melanoma have demonstrated that the efficacy of IPI + NIVO is largely driven by the first two doses in many patients (Postow MA et al., J Clin. Oncol. 2022). We assessed outcomes with IPI + NIVO by number of doses given in patients with metastatic RCC. Methods: We conducted a retrospective study of patients with metastatic RCC at Memorial Sloan Kettering Cancer Center treated with first-line IPI + NIVO. Baseline characteristics and treatment outcomes were obtained from electronic health record review. We calculated overall survival (OS) by the Kaplan-Meier method starting at 12 weeks after initiation of combination therapy, including all patients who were still alive and being followed at that time point and excluding those who had disease progression prior to completing 4 doses. We compared survival rates at 12 and 18 months and median OS for patients who received 4 doses versus those who received fewer than 4 doses. Results: Patients with metastatic RCC treated with first-line IPI + NIVO were included (N=222); 77% were male, 85% had clear cell RCC, 48% had sarcomatoid and/or rhabdoid features, and 87% had IMDC intermediate or poor risk disease. Regarding IPI + NIVO, 145 patients (65%) received all 4 doses, 30 (14%) received 3 doses, 21 (9%) received 2 doses, and 26 (12%) received 1 dose. The most common reasons for not completing all 4 doses (77, 35%) were toxicity (57%) and disease progression (21%). All 145 patients who received 4 doses and 44 who received fewer than 4 doses for reasons other than early progression or death were included in the analysis. OS in the 4 dose and <4 dose group at 18 months was 83% (95% CI: 76%, 89%) and 79% (95% CI: 63%, 88%), respectively. Conclusions: In this observational analysis, we found comparable OS rates in those patients who received all 4 doses of IPI + NIVO compared to those who received fewer than 4 doses for reasons other than disease progression, primarily toxicity. Four Doses of IPI + NIVO (N=145) Fewer Than Four Doses of IPI + NIVO, Excluding Early Disease Progression (N=44) 12-month OS (95% CI) 89% (82%, 93%) 86% (72%, 94%) 18-month OS (95% CI) 83% (76%, 89%) 79% (63%, 88%) Median OS, months (95% CI) 67.1 (40.1, NR) 82.5 (27.1, 109.3) Log-rank p value = 0.595

658 Background: In the phase 3, randomized, double-blind CheckMate 274 trial, adjuvant NIVO demonstrated statistically significant and clinically meaningful disease-free survival (DFS) benefit vs PBO in pts with high-risk MIUC after radical surgery (RS) ± prior neoadjuvant cisplatin-based chemotherapy (NAC). With extended 3-y median follow-up, continued improvements in DFS were seen with NIVO vs PBO in the primary efficacy populations (intent-to-treat [ITT], tumor programmed death ligand 1 [PD-L1] expression ≥ 1%) and in pts with MIBC. Early trends in interim OS favored NIVO vs PBO in ITT and tumor PD-L1 ≥ 1% pts. Here we report additional efficacy outcomes for pts with MIBC. Methods: Pts were randomized 1:1 to NIVO 240 mg every 2 wk or PBO for ≤ 1 y of adjuvant treatment, stratified by tumor PD-L1 expression, nodal status, and prior NAC. Primary endpoints were DFS in ITT and tumor PD-L1 expression ≥ 1% pts. OS in ITT and PD-L1 ≥ 1% pts was a secondary endpoint. Analysis of MIBC pts was exploratory. MIBC OS data are from preplanned interim analyses of ITT and PD-L1 ≥ 1% pts. OS follow-up is ongoing as the prespecified statistical boundaries for significance in ITT and PD-L1 ≥ 1% pts were not crossed at the time of these analyses. Results: Of 709 randomized pts (ITT), 560 (79%) had MIBC (NIVO, n = 279; PBO, n = 281); 284 (51%) of MIBC pts had prior NAC. With median follow-up of 36.1 mo (ITT), DFS improvement with NIVO vs PBO was consistent between all pts with MIBC (hazard ratio [HR] 0.63) and those with (HR 0.58) and without prior NAC (HR 0.69; Table). For OS, HRs favored NIVO vs PBO in all pts with MIBC (HR 0.70) and the tumor PD-L1 ≥ 1% subgroup (HR 0.48), as well as in pts with MIBC with (HR 0.74) and without prior NAC (HR 0.67). Safety was consistent with previous data in ITT pts; no new safety signals were identified. Conclusions: With 3-y median follow-up, consistent benefit in DFS was observed with NIVO vs PBO in all MIBC pts and across prior NAC subgroups. The HR for OS favored NIVO in all MIBC pts, in those with PD-L1 ≥ 1%, and regardless of prior NAC status. These results continue to support adjuvant NIVO as a standard of care for high-risk MIUC and MIBC, potentially providing an opportunity for a curative outcome. Clinical trial information: NCT02632409 . NIVOn NIVOMedian(95% CI), mo PBOn PBOMedian(95% CI), mo HR (95% CI) DFS All MIBC 279 25.6 (19.2–41.8) 281 8.5 (7.3–13.7) 0.63 (0.51–0.78) With prior NAC 142 19.6 (15.6–48.2) 142 8.3 (5.6–11.2) 0.58 (0.43–0.79) No prior NAC 137 25.9 (19.2–51.5) 139 13.7 (7.8–22.1) 0.69 (0.50–0.94) OS All MIBC 279 NR (45.0–NE) 281 39.9 (29.8–52.1) 0.70 (0.55–0.90) PD-L1 ≥ 1% 113 NR (NE–NE) 117 37.6 (26.9–NE) 0.48 (0.29–0.77) With prior NAC 142 55.2 (41.8–NE) 142 40.2 (28.8–53.7) 0.74 (0.53–1.03) No prior NAC 137 NR (40.7–NE) 139 37.7 (28.7–65.2) 0.67 (0.47–0.95) NE, not estimable; NR, not reached.

459 Background: The TiNivo-2 study assessed TIVO 0.89 mg + NIVO vs TIVO 1.34 mg in patients with RCC that progressed after ICI therapy. This study did not meet its primary endpoint of demonstrating a benefit of adding NIVO to TIVO vs TIVO after prior ICI exposure; however, clinically meaningful outcomes were observed with TIVO as a second-line (2L) and third-line (3L) treatment following ICI. In the intent-to-treat population, the median progression-free survival was 5.7 months (95% CI, 4.0-7.4) with TIVO + NIVO and 7.4 months (5.6-9.2) with TIVO (hazard ratio, 1.10; 95% CI, 0.84-1.43; P =.49), with fewer treatment-emergent adverse events in the TIVO + NIVO vs TIVO arm. Quality of life (QOL) data are reported here (NCT04987203). Methods: The Functional Assessment of Cancer Therapy Kidney Cancer Symptom Index–Disease-Related Symptoms (FKSI-DRS) and European Organisation for Research and Treatment of Cancer (EORTC) QLQ-C30 questionnaires were administered at baseline (BL), day 1 of each cycle, and end of treatment. The PRO-evaluable set was defined as all randomized patients with a BL and ≥1 post-BL assessment. Descriptive statistics were provided for the FKSI-DRS and EORTC QLQ-C30 data. The number and percentage of patients categorized as having improved (I), stable (S), or deteriorated (D) QOL were summarized. Results: As of April 1, 2024, median follow-up was 12.0 months. Median duration (range) of treatment was 6·3 months (2.68-10.12) with TIVO + NIVO and 7·4 months (2.83-11.04) with TIVO. Completion rates for FKSI-DRS and EORTC QLQ-C30 were >90% at BL and >50% at week 24 (≈6 mo.) in each arm. Compliance rates for both instruments at BL and week 24 were >90%. PRO results are presented in the table. Conclusions: There were no differences in the PRO outcomes between the combination therapy and monotherapy arms or between the 2 different TIVO doses. PRO data suggested that TIVO maintained the FKSI-DRS and EORTC QLQ-C30 mean scores from BL to week 24. In the TIVO arm, the proportion of patients who had improvement in FKSI-DRS and EORTC QLQ-C30 scores was numerically better in patients receiving 2L vs 3L treatment, while the portion of patients with a deterioration was smaller in the 2L than in the 3L. Clinical trial information: NCT04987203 . PRO variables TIVO + NIVO TIVO ITT 2L 3L ITT 2L 3L FKSI-DRSBL mean (SD) 28.8 (5.6) 29.5 (4.9) 27.6 (6.5) 29.3 (5.3) 29.1(5.5) 29.5 (5.0) Week 24 mean (SD) 29.9 (4.9) 30.1 (4.7) 29.5 (5.4) 29.4 (5.3) 29.3 (4.8) 29.6 (6.3) I/S/D, % 28.2/47.2/24.6 28.0/52.7/19.4 28.6/36.7/34.7 22.9/53.5/23.6 27.5/53.8/18.7 15.1/52.8/32.1 EORTC-QLQ-C30BL mean (SD) 63.4 (23.4) 65.0 (21.8) 60.5 (26.1) 66.2 (21.4) 67.1 (21.9) 64.9 (20.7) Week 24 mean (SD) 68.7 (17.4) 68.7 (16.1) 68.6 (20.6) 64.8 (21.1) 64.6 (20.7) 65.1 (22.4) I/S/D, % 27.7/48.9/23.4 27.8/52.2/20.0 27.7/42.6/29.8 21.3/53.7/25.0 23.0/56.3/20.7 18.4/49.0/32.7

LBA143 Background: The CheckMate 8HW study met its dual primary endpoint with NIVO + IPI demonstrating superior progression-free survival (PFS) by blinded independent central review (BICR) vs chemotherapy (chemo) in patients (pts) with centrally confirmed MSI-H/dMMR mCRC in the first-line (1L) setting (HR 0.21; 95% CI 0.14–0.32; P < 0.0001). We report first results from the other dual primary endpoint of PFS for NIVO + IPI vs NIVO across all lines of therapy in pts with centrally confirmed MSI-H/dMMR mCRC. Methods: Immunotherapy-naive pts with unresectable or mCRC and MSI-H/dMMR status by local testing who had received 0 or 1 prior line of therapy were randomized 2:2:1 to (i) NIVO (240 mg) Q2W (6 doses, then NIVO 480 mg Q4W), (ii) NIVO (240 mg) + IPI (1 mg/kg) Q3W (4 doses, then NIVO 480 mg Q4W), or (iii) chemo ± targeted therapies. Pts who had received ≥ 2 prior lines of therapy were randomized 1:1 to the NIVO + IPI or NIVO arms. Treatments continued until disease progression or unacceptable toxicity (all arms), or a maximum of 2 years (NIVO ± IPI arms). Results: Across all lines of therapy,707 pts were randomized to NIVO + IPI (n = 354) or NIVO (n = 353); 55% and 52% received study treatment in the 1L setting, respectively. Of all randomized pts, 296 in the NIVO + IPI arm and 286 in the NIVO arm had centrally confirmed MSI-H/dMMR status. With 47.0 months (mo) of median follow-up (range, 16.7–60.5), NIVO + IPI demonstrated clinically meaningful and statistically significant improvement in PFS by BICR vs NIVO (HR 0.62; 95% CI 0.48–0.81; P = 0.0003) and higher 12-, 24-, and 36-mo PFS rates vs NIVO (Table). Objective response rate (ORR) by BICR was significantly higher with NIVO + IPI vs NIVO (71% vs 58%; P = 0.0011; Table); best overall response of progressive disease was reported in 10% and 19% of pts, respectively. No new safety concerns were identified (Table). Conclusions: In the first randomized study to compare dual- vs single-agent immunotherapy in MSI-H/dMMR mCRC, NIVO + IPI demonstrated superior PFS vs NIVO across all lines of therapy, with a manageable safety profile. These results establish NIVO + IPI as the potential new standard-of-care treatment for MSI-H/dMMR mCRC. Clinical trial information: NCT04008030 . Efficacy by BICR (all lines; centrally confirmed MSI-H/dMMR by IHC and/or PCR test) NIVO + IPI(n = 296) NIVO(n = 286) Median PFS (95% CI), mo NR (53.8–NE) 39.3 (22.1–NE) HR (95% CI); P value 0.62 (0.48–0.81); 0.0003 PFS rate (12/24/36-mo), % 76/71/68 63/56/51 ORR, n (%); 95% CI, % 209 (71); 65–76 165 (58); 52–64 P value 0.0011 Safety (all lines; all treated), n (%) NIVO + IPI (n = 352) NIVO (n = 351) Any-grade/grade 3–4 TRAEs 285 (81)/78 (22) 249 (71)/50 (14) Any-grade/grade 3–4 TRAEs leading to discontinuation 48 (14)/33 (9) 21 (6)/14 (4) Treatment-related deaths 2 (< 1) 1 (< 1) IHC, immunohistochemistry; NE, not estimable; NR, not reached; PCR, polymerase chain reaction; TRAE, treatment-related adverse event.

398 Background: At 4-y follow-up, 1L NIVO + chemo continued to demonstrate clinically meaningful overall survival (OS) and progression-free survival (PFS) benefit vs chemo with acceptable safety in patients (pts) with advanced non-HER2+ GC/GEJC/EAC from CheckMate 649. We report efficacy and safety results of NIVO + chemo vs chemo at 5-y follow-up. Methods: Adults with previously untreated, unresectable, advanced or metastatic, non-HER2+ GC/GEJC/EAC were enrolled, regardless of programmed death ligand 1 (PD-L1) expression. Pts were randomized to NIVO (360 mg Q3W or 240 mg Q2W) + chemo (XELOX Q3W or FOLFOX Q2W), NIVO + ipilimumab, or chemo. Primary endpoints for NIVO + chemo vs chemo were OS and PFS by blinded independent central review (BICR) in pts with PD-L1 combined positive score (CPS) ≥ 5. Results: Pts were randomized to NIVO + chemo (n = 789) or chemo (n = 792). NIVO + chemo continued to show OS and PFS benefit vs chemo in pts with PD-L1 CPS ≥ 5, pts with PD-L1 CPS ≥ 1, and all randomized pts at 60-month (mo) minimum follow-up (Table). OS rates at 60-mo were higher with NIVO + chemo vs chemo in pts with PD-L1 CPS ≥ 5, pts with PD-L1 CPS ≥ 1, and all randomized pts (Table), and OS benefit with NIVO + chemo continued to be observed in most prespecified subgroups. Objective response rates (ORRs) were higher and responses were more durable with NIVO + chemo vs chemo in pts with PD-L1 CPS ≥ 5, pts with PD-L1 CPS ≥ 1, and all randomized pts (Table). No new safety signals were identified. Conclusions: These results represent the first report of 5-y follow-up for anti–PD-1 + chemo combination therapy in GC/GEJC/EAC to our knowledge. NIVO + chemo continued to provide sustained long-term survival vs chemo with an acceptable safety profile after 5 y of follow-up. These data continue to support the use of NIVO + chemo as a standard 1L treatment for advanced GC/GEJC/EAC. Clinical trial information: NCT02872116 . Efficacy PD-L1 CPS ≥ 5 PD-L1 CPS ≥ 1 All randomized NIVO + chemo(n = 473) Chemo (n = 482) NIVO + chemo (n = 641) Chemo (n = 656) NIVO + chemo (n = 789) Chemo (n = 792) mOS (95% CI), mo 14.4 (13.1–16.2) 11.1 (10.1–12.1) 13.8 (12.4–14.8) 11.4 (10.7–12.3) 13.7 (12.4–14.5) 11.6 (10.9–12.5) HR (95% CI) 0.71 (0.61–0.81) 0.76 (0.67–0.85) 0.79 (0.71–0.88) 60-mo OS rate (95% CI), % 16 (12–19) 6 (4–9) 13 (11–16) 5 (4–7) 12 (10–14) 6 (4–8) mPFS a (95% CI), mo 8.3 (7.0–9.4) 6.1 (5.6–6.9) 7.5 (7.0–8.5) 6.9 (6.2–7.1) 7.8 (7.1–8.6) 6.9 (6.7–7.2) HR (95% CI) 0.71 (0.61–0.82) 0.77 (0.68–0.87) 0.79 (0.71–0.89) ORR a,b (95% CI), % 60 (55–65) 45 (40–50) 60 (55–64) 46 (42–51) 58 (54–62) 46 (42–50) mDOR a,c (95% CI), mo 9.6 (8.3–12.4) 7.0 (5.7–8.0) 8.6 (7.9–10.5) 6.9 (5.8–7.6) 8.5 (7.7–9.9) 6.9 (5.9–7.6) a Per BICR. b In pts with measurable target lesions at baseline. c In all measurable responders. DOR, duration of response; m, median.

392 Background: NIVO + chemo demonstrated clinically meaningful survival benefit and an acceptable safety profile vs chemo in previously untreated Chinese pts with advanced GC/GEJC/EAC from CheckMate 649, consistent with the overall study population. 1L NIVO + chemo is currently approved for pts with advanced non-HER2+ GC/GEJC/EAC in China and other countries. We report 5-y results of NIVO + chemo vs chemo in Chinese pts from CheckMate 649. Methods: Adults with previously untreated, unresectable advanced or metastatic, non-HER2+ GC/GEJC/EAC were enrolled regardless of programmed death ligand 1 (PD-L1) expression. Randomized pts received NIVO + chemo, NIVO + ipilimumab, or chemo. Dual primary endpoints for NIVO + chemo vs chemo were overall survival (OS) and progression-free survival (PFS) by blinded independent central review (BICR) in pts with PD-L1 combined positive score (CPS) ≥ 5. Results: 208 Chinese pts were randomized to NIVO + chemo or chemo. At 61-month (mo) minimum follow-up, NIVO + chemo continued to demonstrate OS and PFS benefit vs chemo in pts with PD-L1 CPS ≥ 5 and all randomized pts (Table). 5-y OS rate was 24% with NIVO + chemo vs 8% with chemo in pts with PD-L1 CPS ≥ 5 and 20% vs 7% in all randomized pts. Objective response rate (ORR) was higher and responses were more durable with NIVO + chemo vs chemo in pts with PD-L1 CPS ≥ 5 and all randomized pts (Table). No new safety signals were identified with longer follow-up. Conclusions: NIVO + chemo continued to demonstrate clinically meaningful long-term survival benefit, more durable responses, and acceptable safety vs chemo in Chinese pts after 5 y of follow-up, consistent with earlier reports and with the overall study population of pts with advanced non-HER2+ GC/GEJC/EAC. These results further support NIVO + chemo as a standard 1L treatment option for Chinese pts. Clinical trial information: NCT02872116 . Efficacy PD-L1 CPS ≥ 5 All randomized NIVO + chemo(n = 75) Chemo (n = 81) NIVO + chemo (n = 99) Chemo (n = 109) mOS (95% CI), mo 15.5(11.9–21.1) 9.6(8.0–12.1) 14.3(11.5–16.5) 10.3(8.1–12.1) HR (95% CI) 0.57 (0.40–0.82) 0.63 (0.46–0.85) mPFS a (95% CI), mo 8.5(6.0–14.0) 4.3(4.1–6.5) 8.3(6.2–12.4) 5.6(4.2–6.8) HR (95% CI) 0.51 (0.34–0.76) 0.57 (0.41–0.80) ORR a,b (95% CI), % 68 (56–79) 48 (36–60) 66 (55–76) 45 (35–56) mDOR a,c (95% CI), mo 12.5 (7.2–23.4) 6.9 (3.9–8.5) 12.5 (7.2–17.7) 5.6 (4.4–8.3) a Per BICR. b In pts with measurable target lesions at baseline. c In responders. DOR, duration of response; m, median.

520 Background: In the phase 3 CheckMate 9DW study (NCT04039607), 1L NIVO + IPI demonstrated significant overall survival (OS) benefit vs LEN/SOR, higher objective response rate (ORR) with durable responses, and manageable safety in uHCC. We present efficacy by best overall response (BOR) subgroups and baseline characteristics, and additional safety analyses from the preplanned interim analysis. Methods: Patients (pts) with previously untreated HCC not eligible for curative surgical or locoregional therapies, Child-Pugh score 5 or 6, and ECOG performance status 0 or 1 were randomized 1:1 to receive NIVO 1 mg/kg + IPI 3 mg/kg Q3W (up to 4 cycles), then NIVO 480 mg Q4W or LEN 8 mg or 12 mg QD or SOR 400 mg BID until disease progression or unacceptable toxicity. NIVO was given for a maximum of 2 years. The primary endpoint was OS; secondary endpoints included ORR and duration of response (DOR) per blinded independent central review (BICR) using RECIST v1.1. Results: A total of 668 pts were randomized to NIVO + IPI (n = 335) or LEN/SOR (n = 333). At a median follow-up of 35.2 (range 26.8–48.9) months (mo), median OS (95% CI) was 23.7 (18.8–29.4) mo with NIVO + IPI vs 20.6 (17.5–22.5) mo with LEN/SOR (HR 0.79 [95% CI 0.65–0.96]; P = 0.0180). ORR (95% CI) per BICR was significantly higher with NIVO + IPI vs LEN/SOR (36% [31–42] vs 13% [10–17]; P < 0.0001); median DOR (95% CI) was 30.4 (21.2–not estimable [NE]) mo vs 12.9 (10.2–31.2) mo. Survival benefit of NIVO + IPI vs LEN/SOR was observed across BOR subgroups at the 24-week landmark timepoint (Table). In subgroup analyses, ORR (95% CI) per BICR was higher with NIVO + IPI vs LEN/SOR across HCC etiologies (uninfected: 35% [26–44] vs 8% [4–15]; HBV infected: 25% [17–34] vs 17% [10–25]; HCV infected: 50% [39–61] vs 16% [9–25]) and in pts with Barcelona Clinic Liver Cancer stage ≤B (33% [23–43] vs 13% [6–21]) or stage C (37% [31–44] vs 14% [10–19]). Safety data are shown in the Table. Additional exploratory analyses will be presented. Conclusions: These additional analyses from CheckMate 9DW demonstrate the efficacy and manageable safety of 1L NIVO + IPI in uHCC and further support its use as a potential standard-of-care treatment option in this setting. Clinical trial information: NCT04039607 . OS by BOR at week 24 landmark NIVO + IPI LEN/SOR BOR CR + PR (n = 101) SD a (n = 105) PD (n = 47) CR + PR (n = 28) SD a (n = 212) PD (n = 31) Median OS (95% CI), mo NR (44.4–NE) 30.0(23.5–37.8) 16.0(12.0–18.7) 28.3(20.6–NE) 22.5(20.5–24.8) 13.5(8.7–25.3) All treated pts NIVO + IPI (n = 332) LEN/SOR (n = 325) Any-grade/grade 3–4 TRAEs, n (%) 278 (84)/137 (41) 297 (91)/138 (42) Hepatobiliary 44 (13)/35 (11) 15 (5)/10 (3) Cardiovascular 10 (3)/3 (< 1) 138 (42)/39 (12) Hemorrhagic 2 (< 1)/1 (< 1) 20 (6)/5 (2) a Includes non-CR/non-PD. CR, complete response; NR, not reached; PD, progressive disease; PR, partial response; SD, stable disease; TRAE, treatment-related adverse event.