PFS Curves Research Articles

Investigation of synergistic effects in trials of combination therapies with immune-checkpoint inhibitors in advanced gastric or gastroesophageal junction cancer.

405 Background: Combinations of immune checkpoint inhibitors (ICI) and chemotherapy (CT) have been approved for gastric cancer. However, there is a hypothesis that this combination may blunt antitumor immune responses because most chemotherapeutic agents also target lymphocytes. The primary objective was to investigate that the ICI and CT does not interfere each other’s therapeutic effects in advanced gastric cancer (GC) or gastroesophageal junction cancer (GEJC) patients. Methods: The reconstructed individual patient data was electronically extracted from the Kaplan-Meier curve of phase III randomized controlled trials (RCTs). The observed PFS curve of each constituent monotherapies was used to estimate simulated PFS curves expected under a model of independent drug action. If the observed curve demonstrated significantly better PFS than simulated curve, the combination of ICI and CT may have a synergistic effect, implying a superior outcome compared to simply adding the component monotherapy. Results: The study included 2,538 unresectable advanced, recurrent, or metastatic GC or GEJC patients from three RCTs comparing pembrolizumab (KEYNOTE-061, KEYNOTE-062 and KEYNOTE-859). In patients with programmed cell death ligand 1 (PD-L1) combined positive score (CPS) of 1 or greater, the 1-year and median PFS of the observed and simulated curves were 28.0% vs. 27.9%, and 6.89 months vs. 6.88 months, respectively. One sample log-rank test showed no significant differences between the observed and simulated curves (p = 0.107). In the subgroups with PD-L1 CPS ≥10 or <1, the 1-year PFS of the observed and simulated curves was 34.7% vs 32.8%, and 28.5% vs 26.8%. Conclusions: The observed PFS of ICT involving pembrolizumab was comparable to the simulated PFS estimated from the data for each monotherapy regardless of the magnitude of PD-L1 CPS. Although it was not clear whether potential synergies existed for ICT, these findings at least suggest that the benefits of ICI and CT are not interfering each other, thereby providing theoretical support for the efficacy of ICT in patients with advanced GC or GEJC.

Abstract PR009: Application of Liquid Biopsy-RECIST (LB-RECIST) criteria to track metastatic colorectal cancer (mCRC) molecular response in first-line (1L) setting: Findings from the PLATFORM-B study

Abstract Introduction: Circulating tumor DNA (ctDNA) analysis is a minimally invasive and reproducible method to explore tumor molecular biology. Recently, Gouda et al. proposed the LB-RECIST criteria to monitor molecular response, validating them on a heterogeneous metastatic population. However, data on mCRC are lacking. Here, we report the first attempt to apply LB-RECIST in the multicenter prospective PLATFORM-B study, which enrolled 130 mCRC patients (pts) receiving 1L standard therapy, with ctDNA analysis performed at different time points. Materials and Methods: Baseline (BL) and week 8 (W8) plasma samples were analyzed using a targeted NGS-based approach (Ion S5 system). Changes in ctDNA detectability were assessed to determine molecular response as per qualitative LB-RECIST criteria: Group1 (G1, detectable ctDNA at BL and W8); Group2 (G2, detectable BL and undetectable W8); Group3 (G3, undetectable BL and detectable W8); Group4 (G4, undetectable at BL and W8). Changes in aggregate variant allele frequency (VAF) - the sum of all VAFs of the studied genes in a sample - were used to assess quantitative LB-RECIST responses: ctDNA Complete Response (CCR, clearance after initial detectability), Partial Response (CPR, decrease >10%), Stable Disease (CSD, no change or decrease <10%), Progressive Disease (CPD, increase >10% or new mutations), and Non-measurable Disease (CND, undetectable at BL and W8). Molecular results were correlated with clinical outcomes. Statistically significant differences were assessed across different groups using 2-sided P values with an α ≤ .05 level of significance. Results: Overall, 79 pts had paired BL and W8 samples. Quantitative LB-RECIST identified 32 CCR, 18 CND, 20 CPR, 8 CPD, and 1 CSD cases. Qualitative LB-RECIST found 22 G1, 32 G2, 7 G3, and 18 G4 cases. Radiological RECIST identified 6 CR, 47 PR, 21 SD, and 5 PD. CCR showed the highest survival rates: with a median follow-up of 25 months (mo) (range: 1–49), median progression-free survival (mPFS) was not reached (NR) and median overall survival (mOS) was 41.8 mo. The CND group had similar rates (mPFS NR; mOS 41.1 mo). CPR and CPD had the poorest outcomes: for CPR, mPFS was 12.7 and mOS 16.4 mo; for CPD, mPFS and mOS were 11.9 and 25.5 mo, respectively. In the G1, G2, G3, and G4 groups, mPFS was 11.9, NR, 14.2, and NR, respectively, while mOS was 16.4, 41.8, 33.4, and 41.1 mo. Kaplan-Meier analysis of the quantitative groups showed statistically significant differences (P < .0001) for both PFS and OS curves. The same level of significance was found in the qualitative groups. Conclusion: LB- RECIST can be a rapid and reproducible method to assess molecular response in mCRC. However, concerns remain. The CND group (23% of cases) showed survival rates similar to the CCR group, as observed also between G4 and G2. These data suggest that absence of ctDNA at W8 is the best surrogate for OS. Moreover, CPD had higher OS rates than CPR, but similar PFS. Further investigations in larger populations are needed to refine these criteria and improve molecular response stratification Citation Format: Valentino Martelli, Joana Vidal, Maria Concepción Fernández-Rodríguez, Pilar García-Alfonso, David Páez, Joan Gibert, Jennifer Linares, Annarita Sibilio, Vicente Alonso, Maria Teresa Cano, Cristina Santos, Gema Duran, Elena Élez, José Luis Manzano, Rocio Garcia-Carbonero, Reyes Ferreiro, Ferran Losa, Estela Pineda, Javier Sastre, Auxiliadora Gomez-España, Rodrigo Toledo, Fernando Rivera, Beatriz Bellosillo, Josep Tabernero, Ramon Ramon Salazar, Enrique Aranda, Clara Montagut. Application of Liquid Biopsy-RECIST (LB-RECIST) criteria to track metastatic colorectal cancer (mCRC) molecular response in first-line (1L) setting: Findings from the PLATFORM-B study [abstract]. In: Proceedings of the AACR Special Conference: Liquid Biopsy: From Discovery to Clinical Implementation; 2024 Nov 13-16; San Diego, CA. Philadelphia (PA): AACR; Clin Cancer Res 2024;30(21_Suppl):Abstract nr PR009.

QLTI-04. UPDATE ON HIGH-GRADE GLIOMA PATIENTS TREATED WITH TTFIELDS FROM NANFANG HOSPITAL IN CHINA

Abstract OBJECTIVE Tumour-treating fields (TTFields) have changed the first-line clinical practice of glioblastoma worldwide due to their promising therapeutic efficacy. The aim of this study was to investigate the clinical efficacy of TTFields therapy for HGG in Nanfang Hospital, China. In 2023 ASNO we reported PFS.Now we reporte update about OS from our single center study. METHODS From January 2019 to December 2022, a total of 25 patients with newly diagnosed HGG (ndHGG) treated with TTFields were retrospectively included in our study. Among these patients, 9 patients received concurrent radiochemotherapy in combination with TTFields and 16 patients received non-concurrent chemoradiotherapy in combination with TTFields. The primary endpoint of the study was mPFS and mOS (Kaplan-Meier method used to estimate PFS and OS survival curves). Secondary endpoint was the influences about KPS score in PFS and OS. RESULTS The 25 ndHGG patients treated with TTFields included nine males and sixteen females with a mean age of 47.1 years (12-68). Gross total resection (GTR) was performed in eighteen patients. Twenty-one patients were IDH wild type, eleven patients showed methylation of the MGMT promoter. Followed up until January 2024, the median PFS was 13.3 months [95% CI: 11.33-15.2] with TTFields combined and 4.0 months (95% CI, 3.8-4.4) without TTFields treatment in EF-14.Median OS was 26.53 months [95% CI:19.47-33.57] with TTFields and 16.0 months (95% CI, 3.8-4.4) control in EF-14. In the KPS stratification, OS was 27.5 months (95% CI: 22.95-32.1) vs 4.3 months (95% CI: 0-11.2), p=0.0001. The OS of patients with concurrent chemoradiotherapy combined with TTFields vs. without concurrent chemoradiotherapy combined with TTFields was 26.5 months (95% CI: 24.33-28.7) vs 21.2 (95% CI: 3.1-39.3) months, p=0.59. CONCLUSIONS TTFields therapy is an effective treatment for HGG. In this study the addition of TTFields resulted in a statistically significant improvement in PFS and OS.

A Combined Proteomic and Transcriptomic Signature Is Predictive of Response to Anti-PD-1 Treatment: A Retrospective Study in Metastatic Melanoma Patients.

Resistance biomarkers are needed to identify patients with advanced melanoma obtaining a response to ICI treatment and developing resistance later. We searched a combination of molecular signatures of response to ICIs in patients with metastatic melanoma. In a retrospective study on patients with metastatic melanoma treated with an anti-PD-1 agent carried out at Istituto Nazionale Tumori-IRCCS-Fondazione "G. Pascale", Naples, Italy. We integrated a whole proteome profiling of metastatic tissue with targeted transcriptomics. To assess the prognosis of patients according to groups of low and high risk, we used PFS and OS as outcomes. To identify the proteins and mRNAs gene signatures associated with the patient's response groups, the discriminant analysis for sparse data performed via partial least squares procedure was performed. Tissue samples from 22 patients were analyzed. A combined protein and gene signature associated with poorer response to ICI immunotherapy in terms of PFS and OS was identified. The PFS and OS Kaplan-Meier curves were significantly better for patients with high expression of the protein signature compared to patients with low expression of the protein signature and who were high-risk (Protein: HR = 0.023, 95% CI: 0.003-0.213; p < 0.0001. Gene: HR = 0.053, 95% CI: 0.011-0.260; p < 0.0001). The Kaplan-Meier curves showed that patients with low-risk gene signatures had better PFS (HR = 0 0.221, 95% CI: 0.071-0.68; p = 0.007) and OS (HR = 0.186, 95% CI: 0.05-0.695; p = 0.005). The proteomic and transcriptomic combined analysis was significantly associated with the outcomes of the anti-PD-1 treatment with a better predictive value compared to a single signature. All the patients with low expression of protein and gene signatures had progression within 6 months of treatment (median PFS = 3 months, 95% CI: 2-3), with a significant difference vs. the low-risk group (median PFS = not reached; p < 0.0001), and significantly poorer survival (OS = 9 months, 95% CI: 5-9) compared to patients with high expression of protein and gene signatures (median OS = not reached; p < 0.0001). We propose a combined proteomic and transcriptomic signature, including genes involved in pro-tumorigenic pathways, thereby identifying patients with reduced probability of response to immunotherapy with ICIs for metastatic melanoma.

Measurable residual disease (MRD) dynamics in multiple myeloma and the influence of clonal diversity analyzed by artificial intelligence

Minimal residual disease (MRD) assessment is a known surrogate marker for survival in multiple myeloma (MM). Here, we present a single institution’s experience assessing MRD by NGS of Ig genes and the long-term impact of depth of response as well as clonal diversity on the clinical outcome of a large population of MM patients; 482 MM patients at the University of California, San Francisco (UCSF) diagnosed from 2008 to 2020 were analyzed retrospectively. MRD assessment was performed by NGS. PFS curves were plotted by the Kaplan–Meier method. In the newly diagnosed group, 119 of 304, achieved MRD negativity at the level of 10−6 at least once. These patients had a prolonged PFS versus patients who were persistently MRD positive at different levels (p > 0.0001). In the relapsed disease group, 64 of 178 achieved MRD negativity at 10−6, and PFS was prolonged versus patients who remained MRD positive (p = 0.03). Three categories of MRD dynamics were defined by artificial intelligence: (A) patients with ≥3 consistently MRD negative samples, (B) patients with continuously declining but detectable clones, and (C) patients with either increasing or a stable number of clones. Groups A and B had a more prolonged PFS than group C (p < 10−7). Patients who were MRD positive and had not yet relapsed had a higher clonal diversity than those patients who were MRD positive and had relapsed. MRD dynamics can accurately predict disease evolution and drive clinical decision-making. Clonal Diversity could complement MRD assessment in the prediction of outcomes in MM.

1245 Malignant Meningioma Survival, Recurrence and Functional Outcomes in an International Multicenter Cohort

INTRODUCTION: Meningiomas are the most common intracranial tumor. Grade 3 (malignant) meningiomas are the most aggressive subtype and represent 1-3% of all meningiomas cases. METHODS: Demographic, clinical and histopathological data from grade 3 intracranial meningioma cases were identified in the clinical databases from seven sites in North America and Europe from 1991-2022. Summary statistics and Kaplan-Meier OS and PFS curves were generated. RESULTS: We identified 108 patients, with a median age 65 years (IQR: 52, 72) and 53.7% were female. Median OS was 109 months (95% CIs: 88, 227) and 5-year OS rate was 65% (95% CIs: 56, 76). Median PFS was 38 months (95% CIs: 24, 56) and 5-year PFS rate was 37% (95% CIs: 28, 49). Older age (=65 years old) and male sex were independent predictors of worse OS and PFS in multivariate regression analysis. Postoperative radiotherapy was independently associated with improved OS in all patients and a sensitivity analysis including only patients =65 years old. Functional status, as measured by Karnofsky performance status, was stable between preoperative (median: 80 [IQR: 70, 90]), postoperative (median: 90 [IQR: 70, 98]) and 1-year postoperative (median: 70 [IQR: 50, 80]) timepoints. CONCLUSIONS: This study provides robust survival, recurrence, and functional outcome data for malignant meningiomas in North America and Europe over a 30-year period. We demonstrate that older age and male sex are associated with worse survival and recurrence outcomes. We further affirm that adjuvant radiotherapy confers a survival benefit in the treatment of malignant meningiomas, including in patients older than 65 years. Despite concerns around surgical and radiotherapy treatment complications, functional status remains stable in the first year following treatment of malignant meningiomas.

Texture analysis of 18F-FDG PET/CT and CECT: Prediction of refractoriness of Hodgkin lymphoma with mediastinal bulk involvement.

To recognize patients at high risk of refractory disease, the identification of novel prognostic parameters improving stratification of newly diagnosed Hodgkin Lymphoma (HL) is still needed. This study investigates the potential value of metabolic and texture features, extracted from baseline 18F-FDG Positron Emission Tomography/Computed Tomography (PET) and Contrast-Enhanced Computed Tomography scan (CECT), together with clinical data, in predicting first-line therapy refractoriness (R) of classical HL (cHL) with mediastinal bulk involvement. We reviewed 69 cHL patients who underwent staging PET and CECT. Lesion segmentation and texture parameter extraction were performed using the freeware software LIFEx 6.3. The prognostic significance of clinical and imaging features was evaluated in relation to the development of refractory disease. Receiver operating characteristic curve, Cox proportional hazard regression and Kaplan-Meier analyses were performed to examine the potential independent predictors and to evaluate their prognostic value. Among clinical characteristics, only stage according to the German Hodgkin Group (GHSG) classification system significantly differed between R and not-R. Among CECT variables, only parameters derived from second order matrices (gray-level co-occurrence matrix (GLCM) and gray-level run length matrix (GLRLM) demonstrated significant prognostic power. Among PET variables, SUVmean, several variables derived from first (histograms, shape), and second order analyses (GLCM, GLRLM, NGLDM) exhibited significant predictive power. Such variables obtained accuracies greater than 70% at receiver operating characteristic analysis and their PFS curves resulted statistically significant in predicting refractoriness. At multivariate analysis, only HISTO_EntropyPET extracted from PET (HISTO_EntropyPET ) and GHSG stage resulted as significant independent predictors. Their combination identified 4 patient groups with significantly different PFS curves, with worst prognosis in patients with higher HISTO_EntropyPET values, regardless of the stage. Imaging radiomics may provide a reference for prognostic evaluation of patients with mediastinal bulky cHL. The best prognostic value in the prediction of R versus not-R disease was reached by combining HISTO_EntropyPET with GHSG stage.

Impact of Folinic Acid Dosing on Efficacy and Toxicity of High-Dose Methotrexate in Central Nervous System Lymphoma

Treatment of primary central nervous system lymphoma (PCNSL) is centered around a backbone of high-dose methotrexate (HDMTX) with folinic acid (FA) rescue. Higher FA doses were associated with treatment failure in several retrospective studies in acute lymphatic leukemia (ALL), leading to the adoption of lower FA doses in ALL treatment protocols. In PCNSL, FA doses have varied widely, guidelines and consensus statements generally do not address FA dosing, and the effect of FA dosing on treatment outcomes has not been assessed. We retrospectively analyzed the medical records of patients treated at our institute for PCNSL between the years 2014 and 2022 to address this question. Patient characteristics are listed in table 1. Complete dosing data was available for 36 patients. The median number of HDMTX doses administered was 6.5 (range, 2-9;), with an average dose of 3.9 g/m 2 (SD +/- 0.74; range, 2.7-5.0). Per the local protocol, IV FA rescue began 24 hours after each MTX infusion at a fixed dose of 50 mg four times daily and continued until MTX clearance. The average cumulative dose of FA administered per treatment was 322 mg/m 2 (SD +/- 141; range, 140-700). 33 patients (92%) achieved either a partial or complete response on MRI following induction treatment. Three (8%) had progressive disease. 26 patients (72%) proceeded with consolidation treatment. Of these, 16 (44%) underwent autologous hematopoietic stem cell transplant. At a median follow up of 1.6 years (range, 0.3-8.1), 12 patients (33%) suffered progressive disease, disease relapse, or death. The 2-year PFS probability for the entire cohort was estimated at 64.5% (95% confidence interval [CI], 49.4-84.1%), and the 2-year OS as 70.7% (95% CI 56.0-89.2%). In univariate Cox regression analysis, average FA dose during MTX treatment was a significant predictor of both PFS and OS, with hazard ratios (HR) of 2.22 (95% CI 1.39-3.55; p &amp;lt; 0.001) and 2.07 (95% CI 1.31-3.29; p = 0.002), respectively, for each 100 mg/m 2 increase. The FA dose during the first treatment cycle, before assessment of response, was also significantly associated with worse PFS (HR 1.56; 95% CI 1.24-1.96; p &amp;lt; 0.001) and OS (HR 1.41; 95% CI 1.14-1.74; p = 0.001). A trend towards worse PFS and OS was seen with male gender, elevated serum LDH, and deep structure involvement. A Cox multivariate regression analysis using the variables age, gender, serum LDH, deep structure involvement, and the average MTX and FA doses, identified FA dose as an independent prognostic factor for PFS, with an estimated hazard ratio (HR) of 2.23 for each 100 mg/m 2 increase (95% CI, 1.25-3.97; p = 0.006). A smaller effect was seen when the FA dose during only the first treatment cycle was examined (HR 1.55; 95% CI 1.16-2.08; p = 0.003). Involvement of deep structures was also significantly associated with worse PFS, with an estimated HR of 8.59 (95% CI, 1.28-57.5; p = 0.027), and there was a borderline significant trend for worse PFS with male gender and elevated LDH. None of the examined variables were identified as statistically significant predictors of OS in the multivariate analysis. Logistic regression with 2-year PFS as the dependent variable was used to define a cutoff value of 350 mg/m 2 for FA dose. The group of patients who received FA doses above and below this threshold showed significantly different PFS curves (figure 1). The 2-year PFS probability was estimated at 56.6% (95% CI, 35.4-90.5%) for patients who received doses of FA above this threshold, and 70.5% (95% CI, 51.7%-96.2%) for patients who received doses below it. We identified no effect of FA dosing on hematologic, renal, hepatic, or gastrointestinal toxicity at the dose ranges examined. In summary, we found that high FA doses were associated with inferior PFS in PCNSL. Such high doses may be avoided through use of lower individual doses of FA, such as 15 mg/m 2 or 25 mg fixed doses as have been used in some protocols, rather than the 50 mg or higher doses used in our and other's practice. Our conclusions are moderated by the study's limitations, which include a small sample size limited to a single center and a retrospective design. However, they suggest that optimizing FA dosing in treatment of PCNSL might improve treatment efficacy without excess toxicity. Future larger and prospective studies will be required to better define the range of MTX and FA doses at which the balance between prevention of treatment toxicity and preservation of treatment efficacy is preserved.

Cost-effectiveness of nivolumab and ipilimumab versus pembrolizumab and axitinib in advanced renal cell carcinoma with intermediate or poor prognostic risk: a Brazilian private healthcare system perspective

ObjectiveNivolumab plus ipilimumab (NIVO + IPI) and pembrolizumab plus axitinib (PEM + AXI) have demonstrated significant clinical benefits as first-line (1 L) treatments for intermediate/poor-risk advanced renal cell carcinoma (aRCC) patients. This study aimed to assess the cost-effectiveness of NIVO + IPI versus PEM + AXI from a Brazilian private healthcare system perspective, utilizing a novel approach to estimate comparative efficacy between the treatments.MethodsA three-state partitioned survival model (progression-free, progressed, and death) was developed to estimate costs, life-years (LYs), quality-adjusted LYs (QALYs), and the incremental cost-utility ratio (ICUR) over a 40-year time horizon. In the absence of head-to-head comparisons between NIVO + IPI and PEM + AXI, clinical data for NIVO + IPI was obtained from CheckMate 214 (NCT02231749) and for PEM + AXI from KEYNOTE-426 (NCT02853331). A matching-adjusted indirect comparison was conducted to account for the imbalance of treatment effect modifiers between the trials. Patient characteristics, resource use, health state utilities, and costs were based on Brazilian-specific sources. Costs and health outcomes were both discounted by 5% annually in line with Brazilian guidelines. The robustness of the results was evaluated through extensive sensitivity analysis and scenario analyses.ResultsWhen comparing the matched versus unmatched OS, PFS, and TTD curves there was no noteworthy difference. NIVO + IPI was associated with cost savings (R$ 350,232), higher LYs (5.54 vs. 4.61), and QALYs (4.74 vs. 3.76) versus PEM + AXI, resulting in NIVO + IPI dominating PEM + AXI. Key model drivers were the treatment duration for PEM, NIVO, and AXI. NIVO + IPI remained dominant in all scenario analyses, which indicated that model results were robust to alternative modelling inputs or assumptions.ConclusionsThis analysis shows that NIVO + IPI is estimated to be a life-extending and potentially cost-saving 1 L treatment option when compared with PEM + AXI for intermediate/poor-risk a RCC patients in the Brazilian private healthcare system.

Measurable residual disease (MRD) and clonal diversity for multiple myeloma treatment monitoring.

8028 Background: MRD assessment is a known surrogate marker for survival in multiple myeloma (MM). However, some patients with non-detectable MRD, do relapse, and some MRD-positive patients with an extensive follow-up, do not relapse. Clonal diversity is defined as the number of unique Immunoglobulin (H, K, or L) sequences in each sample. Here, we present a single institution experience assessing MRD by NGS of Ig genes and the long-term impact of depth of response as well as clonal diversity on the clinical outcome of a large population of MM patients. Methods: 482 MM patients at University of California, San Francisco (UCSF) diagnosed from 2008 to 2020. Of them, 304 were newly diagnosed and 178 in ≥ 2nd line. MRD was assessed in patients achieving VGPR or better by IMWG criteria. MRD assessment was performed by NGS (Adaptive Biotechnologies, Seattle, WA). PFS curves were plotted by the Kaplan-Meier method, and the log-rank test was used to estimate statistical significance. Results: 1098 MRD samples were analyzed. MRD was available at 3 time points for 118 patients. Median follow-up was 26m. Overall, 181 of 482 patients (38%) achieved MRD- ( &lt; 10-6) on one or more samples. Clonal diversity was available in 87 of those patients. In the newly diagnosed group, 84 of 120 (39%), achieved MRD- at the level of 10-6 at least once. These patients had a prolonged PFS versus patients who were persistently MRD+ at different levels (p &gt; 0.0001). Notably, patients achieving a MRD &lt; 10-5 have longer OS than those with MRD &gt; 10-5 (p = 0.009). Of the 178 patients who received therapy for relapsed disease, 64 achieved MRD- at 10-6 (36%) and PFS was also prolonged versus patients who remained MRD+ (44 m v NR, p = 0.03). Then, we analyzed the effect of repeated MRD monitoring on PFS. Three categories were defined: (A) patients with ≥3 MRD- samples, (B) patients with continuously declining detectable clones, and (C) patients with a stable number of clones. Groups A and B had a more prolonged PFS than group C (NR vs NR vs 38m, p &lt; 0.0001). Finally, we analyzed clonal diversity at the moment of maximum response: patients MRD+ who have not relapsed have higher clonal diversity than MRD+ patients who relapsed. Moreover, patients who were MRD- who have not relapsed have higher clonal diversity than patients who were MRD- who did relapse. This was also observed independently for the 3 receptors analyzed (IgH p = 0.026; IgK p = 0.036 and IgL p = 0.036). Patients with more than 66000 IgH unique sequences at the moment of maximum response had a prolonged PFS (p = 00.5). Conclusions: MRD assessment in a real-world setting has the same predictive power as that seen in clinical trials even on OS. MRD dynamics can accurately predict disease evolution and drive clinical decision-making. Clonal Diversity could complement MRD assessment in the prediction of the outcome in MM.

Economic evaluation for the US of durvalumab plus gemcitabine and cisplatin (DGC) in advanced biliary tract cancer (BTC).

4081 Background: BTCs are a group of relatively rare malignancies comprised of intrahepatic cholangiocarcinoma (CCA), extrahepatic CCA, and gallbladder cancer. The first-line standard of care of gemcitabine and cisplatin (GC) has remained unchanged for the past decade, highlighting the need for novel therapies. The TOPAZ-1 trial demonstrated improved progression-free (PFS) and overall survival (OS) with durvalumab added to GC. This trial-based economic evaluation estimated the cost-effectiveness/utility of DGC versus GC from a US payer perspective. Methods: A three state partitioned survival model (Progression-free, Progressed, Death) was developed to compare costs and overall survival outcomes associated with both treatments. PFS and OS curves were digitized, and parametric functions fitted. A 5-year time horizon with a 3% discount rate/year was considered. Costs of treatment (average sales price), administration, and monitoring parameters were sourced from Centers for Medicare &amp; Medicaid Services databases; costs of adverse event management (grade 3/4 with rate ≥ 5%) were sourced from prior BTC economic evaluations. Life years (LY), quality adjusted life years (QALY), incremental cost-effectiveness and utility ratios (ICER/ICUR) were estimated in a base case (BCA) and probabilistic sensitivity analyses (PSA). A cost-effectiveness acceptability curve (CEAC) was plotted to determine the probability of either treatment being cost-effective over the other at different willingness to pay (WTP) thresholds. Results: Exponential regression was used to extrapolate DGC and GC survival curves. As shown in the table, the BCA (PSA) shows an incremental cost of DGC over GC of 131,350 (131,212), incremental LY of 0.38 (0.38), and incremental QALY of 0.26 (0.28). The BCA and (PSA) ICERs reveal an additional 345,658 (345,295) per LY gained (g) and an additional 505,192 (468,614) per QALYg. The CEAC curve shows that DGC treatment has a 50% probability of being cost-effective at a WTP threshold value of 525,000 and 100% probability at a threshold of 1,375,000 or above. Conclusions: This economic evaluation demonstrates that, in the setting of advanced BTC, DGC is associated with a slight improvement in LY and QALY, yet at a marked incremental cost, requiring a very high WTP threshold. [Table: see text]

Is HER2 expression related to CDKi efficacy in HER2-negative luminal metastatic breast cancer?

e13073 Background: Cyclin kinase inhibitors (CDKi) are the standard of care in the first line of luminal (Lum) metastatic breast cancer (MBC). HER2 overexpression is a classic risk factor in BC, however, the use of targeted therapy has changed the prognosis of this disease, even with anti-HER2 efficacy data in cases of low HER2 expression. We wonder if the level of HER2 expression can be a predictor of efficacy in treatment with CDKi. Methods: Descriptive, observational, retrospective and analytical study of patients with MBC treated with CDKi, in the period 2018-2022. Descriptive and analytical statistical analysis (IBM SPSS Statistics 22) using logistic regression and Kaplan-Meier with long-rank test to analyze the relationship between CKi efficacy and HER2 expression (negative = 0+; HER2 low = 1+ and 2 + with SISH negative). Results: We analyzed 152 patients (p) with Lum MBC under treatment with CDKi (median 58 years), in first and successive lines. HER2 expression: 31% 46p (0+), 54% 81p (1+) and 14% 21p (2+ and SISH negative). 41% were treated with palbociclib, 34% with ribociclib, and 26% with abemaciclib. Radiological response (RR) to CDKi was achieved by 34.1% (CRR 2.4%, PRR 31.7), and the rest stable disease (SD) or progressive disease (PD): SD 42% and PD 22%. In the overall OR, there was no evidence of a risk association between the response to CDKi and the level of HER2 expression. We obtained similar results in the analysis stratified by type of CDKi. In the multivariate study, for the variables CDKi type and HER2 expression, the logistic regression model was not statistically significant either. Regarding the Kaplan-Meier analysis, we compared the median PFS and OS using the Long-rank test, based on the expression of HER2. In the complete series we did not find statistically significant differences between the curves. In the palbociclib subgroup the curves split in favor of HER2 0+ vs. HER2 low: 44 months 95%CI (7.6-80) vs 16 months 95%CI (7.9-24); p = 0.01. Regarding OS, we found no significant differences. With ribociclib, the curves separate in favor of HER2 low versus HER2 0+: 33 months 95%CI (7.2-18) vs. 13 months 95%CI (23-42); p = 0.03. In OS we did not find significant differences either. In the subgroup with abemaciclib, we observed a trend towards greater survival in HER2 0+ patients, but with no significant difference between the PFS or OS curves. Conclusions: Based on our study, the expression of HER2 could be considered a predictor of efficacy depending on the CDKi used. We assume greater survival with palbociclib in HER2 negative patients (0+) and with ribociclib in HER2 low patients. We consider it essential to develop studies with a larger sample size and clinical trials that allow us to demonstrate the hypothesis of our work with scientific evidence.

Comparative analysis of outcomes following craniotomy and expanded endoscopic endonasal approach resection of tuberculum sellae meningiomas: a single-institution study.

Traditionally, supratentorial craniotomy has been used to sever tuberculum sellae meningiomas (TSMs), but there has been a remarkably increasing tendency of extended endoscopic endonasal approach (EEEA) used to treat TSMs in the recent decade. Several documents have described the advantages and disadvantages of both approaches, but there is no consensus on whether one is superior to the other. This study aimed to compare surgical outcomes between craniotomy and EEEA for TSMS treated at our institution. From January 2015 to December 2021, a total of 84 cases of TSMs were included in this study. Cases were separated into two groups: the craniotomy group and the EEEA group. Their anamneses and surgical records were reviewed. Demographic data, presenting symptoms, tumor volume, extent of resection, visual outcomes, and follow-up data were tabulated. The Kaplan-Meier curves were constructed for the PFS for both cohorts. Complete data were available for 84 surgeries; 39 cases were treated via craniotomy, and 45 were treated via EEEA. Patient demographic data, pre-operative symptoms, and tumor characteristics were similar between the two cohorts. The extent of resection was similar between the two groups (GTR: 91.11% EEEA vs. 87.18% craniotomy; STR 8.89 vs. 12.82%, p = 0.91). There was no difference in visual outcomes between both groups (92.1 vs. 84.84%, p = 0.46). An increased rate of cranial nerve injury was noted in the craniotomy group (0 vs. 10.25%, p = 0.04). Post-operative CSF leak rate occurred in one patient in the EEEA group. The PFS curves (p = 0.52) and recurrence/progression rates (13.33 vs. 20.51%, p = 0.39) were similar between the two groups. Both EEEA and craniotomy can successfully sever TSMs. The recurrence/progression rate and PFS appear to be similar between the two groups. Although there are no differences in EOR and visual outcomes between the two groups, there was a clear trend in the EEEA group to obtain a better outcome. CSF leakage was common in the EEEA cohort, whereas the rate of cranial nerve injury was found to be higher in the craniotomy cohort. We believe that our data support the conclusion that EEEA surgery is the preferred approach for the removal of TSMs.

Extended Intensified Post-ASCT Consolidation with Daratumumab, Bortezomib, Lenalidomide and Dexamethasone (Dara-VRd) for Ultra-High Risk (UHiR) Newly Diagnosed Myeloma (NDMM) and Primary Plasma Cell Leukemia (pPCL): The UK Optimum/Muknine Trial

Background: Outcome for patients with ultra-high risk (UHiR) newly diagnosed multiple myeloma (NDMM) or primary plasma cell leukemia (pPCL) remains unsatisfactory. Recent results from MASTER (NCT03224507) and other trials unfortunately confirmed the sustained risk of post-ASCT relapse for UHiR MM patients in context of contemporary combination induction therapy, including relapse from undetectable minimal residual disease (MRD). The OPTIMUM/MUKnine (NCT03188172) trial was designed to reduce (pre- and) post-ASCT relapse risk in UHiR patients by continued Dara-VR(d) consolidation for 18 cycles, followed by Dara-R maintenance. We report here outcomes for OPTIMUM patients until end of consolidation therapy and contextualise results with outcomes of UHiR MM patients treated in the Myeloma XI (ISRCTN49407852) trial. Methods: In OPTIMUM, patients identified to have UHiR NDMM (≥2 high risk lesions: t(4;14), t(14;16), t(14;20), gain(1q), del(1p), del(17p) or SKY92 risk signature) or with pPCL (circulating plasmablasts >20%) through central molecular screening of an all-comer population received up to 6 cycles of Dara-CVRd induction and V-ASCT. All patients then received 18 cycles of post-ASCT consolidation, of which 6 cycles Dara-VRd (Cons1) and 12 cycles Dara-VR (Cons2), before moving to monthly Dara-R maintenance until progression. OPTIMUM recruited 107 UHiR patients, including 9 pPCL, from 472 screened patients with suspected NDMM from 39 UK NHS hospitals between Sep 2017 and Jul 2019. We previously reported the OPTIMUM primary endpoint analysis, which demonstrated improved PFS at 18 months post start of induction for OPTIMUM in a pre-specified comparison against UHiR patients treated in the Myeloma XI (MyXI) trial. We report here a planned analysis of OPTIMUM patient PFS and safety data after all patients completed consolidation. Results are contextualised with outcome for UHiR patients treated in MyXI with carfilzomib/cyclophosphamide-Rd (KCRd) or CRd induction, ASCT and lenalidomide maintenance or observation. Further analyses including overall survival are underway and will be presented at the meeting. Results: With median follow-up of 40.0 months (95% CI: 37.3-42.7), median PFS was not reached for OPTIMUM patients. The PFS estimate at the end of consolidation, 30 months from start of induction therapy, was 77.0% (95% CI: 68.8-85.1). For context, PFS at 30 months for UHiR patients in Myeloma XI was 39.8% (95% CI: 30.7-48.9) for the overall group, and specifically 49.7% (95% CI: 33.2-66.2) for MyXI patients randomised to KCRd and 34.1% (95% CI: 23.4-44.8) for those randomised to CRd treatment arms (Figure 1). We observed a progressive separation of PFS curves over time between OPTIMUM and MyXI, suggesting a sustained beneficial effect of ongoing post-ASCT consolidation for UHiR patients (Figure 1). Of the 85 OPTIMUM patients who received at least one cycle of consolidation post-ASCT, 80 patients completed Cons1 and 74 patients Cons2. Progressive disease was the commonest reason for trial discontinuation. Most frequent CTCAE grade 3/4 adverse events (AEs) during Cons2 included neutropenia (43.8%), thrombocytopenia (27.5%) and infection (15%). Grade 4 events were rare (≤5%) for all categories; there were no treatment-related deaths. Peripheral neuropathy rate grade ≥2 was 13.8% (Grade ≥3: 2.5%) for Cons2; no cumulative increase from Cons1 was noted. Of the initial 107 patients who started therapy in OPTIMUM, 47 had an MRD test result available at end of Cons2 at time of abstract submission. Of these, 44 (93.6%; 41.1% of overall trial population) had undetectable MRD at 10-5 sensitivity by flow cytometry. In the majority of patients undetectable MRD was sustained since ASCT. Conclusions: Prevention of relapse remains a key challenge in the treatment of UHiR MM/PCL patients, including for patients achieving undetectable MRD. OPTIMUM outcomes for extended Dara-VR consolidation in UHiR/PCL patients demonstrate 30 months PFS of 77% and compare favourably with MyXI, but also results from the MASTER trial, which recently reported 24-month PFS of 58% for UHiR patients defined by same criteria (JCO 2021). Safety data demonstrate manageable side effects with extended Dara-VR consolidation. Our results support extended, risk stratified post-ASCT therapy for UHiR MM/PCL patients. Figure Legend: PFS Kaplan-Meier plot for UHiR/PCL patients treated in OPTIMUM and Myeloma XI. Figure 1View largeDownload PPTFigure 1View largeDownload PPT Close modal

Cost-Effectiveness Analyses of First Line Ibrutinib Versus Acalabrutinib Versus Zanubrutinib Followed By Second Line Venetoclax Plus Rituximab in Previously Untreated Chronic Lymphocytic Leukemia (CLL) Patients

Background. Targeted therapy with Bruton's tyrosine kinase inhibitors (BTKI) has significantly improved progression-free (PFS) and overall survival (OS) outcomes when compared with chemoimmunotherapy in the front line setting of CLL. Following the approval of the first generation of BTKI (ibrutinib, IBR) in 2016 as first line (1L) treatment of CLL, second generation BTKIs with more favorable safety profiles were approved/recommended in 2020 (acalabrutinib, ACA) and 2021 (zanubrutinib, ZAN). We performed a network meta-analysis (NMA) of these three BTKIs in terms of PFS in support of cost-effectiveness analyses from US payer perspective. Methods. We conducted a NMA of 5 trials (Woyach 2018, Burger 2015, Sharman 2020, Tam 2021, Goede 2014) on the PFS hazard ratios (HR) of the 3 BTKIs and any other CLL interventions that connected in the network. We specified a 5-state Markov model to capture the disease and clinical pathway for CLL patients over a 10 years' time horizon: progression free (PF) on 1L treatment [PF1]; progression before initiating second line (2L) treatment [P1]; PF on 2L treatment [PF2]; progression after 2L treatment [P2]; and death. Patients started PF on 1L treatment with a BTKI and moved to other states based on extracted transition probabilities from lognormal parametric distributions fitted to Kaplan-Meier PFS curves of phase III trials of IBR, ACA and ZAN. Venetoclax plus rituximab (VR) was considered the 2L treatment with PF2, P2 and death in 2L extracted from Seymour 2018. The transition probabilities of PF on 1L treatment and progression after 1L were adjusted based on the NMA results. OS and time to next treatment (TTNT) after first progression were considered the same for the three BTKIs. Costs of regimens (wholesale acquisition cost), adverse event (AE) management, and disease progression were incorporated into the appropriate health states. Utilities of PF1 (0.799), P1 (0.66), PF2 (0.55) and P2 (0.42) were obtained from literature. Annual discount rate of 3.0% was applied. Life years (LY), quality adjusted LY (QALY), and incremental cost-effectiveness (ICER) and cost-utility ratios (ICUR) were estimated. Probabilistic sensitivity analyses (PSA; 10,000 iterations) were performed to obtain cost-effectiveness acceptability curves (CEAC). Results. Compared with IBR, ACA (HR=0.54; 95%CI=0.41-0.72) but not ZAN (HR=1.03; 95%CI=0.80-1.34) was associated with a statistically significant lower HR of progression or death. At 10 years, 77% of ACA patients were in PF1 and 9% in PF2, compared to 49% and 28% of ZAN patients, and 50% and 27% of IBR patients. As shown in the table, in base case analyses, total costs of IBR treatment were higher by $104,331 over ACA and by $350,340 over ZAN; with ACA being higher by $246,009 over ZAN. Survival was estimated to be 6.83 LYs for IBR, 7.02 LYs for ACA (+0.19) and 6.82 LYs for ZAN (-0.01). QALYs were estimated to be 5.00 for IBR, 5.46 for ACA (+0.46) and 5.09 for ZAN (+0.09). This yielded negative (cost saving) ICER of $-549,110/LYg and ICUR of $-226,806/QALYg for ACA over IBR. The statistically non-significant survival benefit, the 0.01 LY lost but 0.09 QALY gained, and the lower cost of ZAN over IBR yielded an ICER of $35,034,000/LYg and a negative (cost saving) ICUR of $3,892,666/QALYg. Compared to ZAN, treatment with ACA was associated with incremental costs of $246,009, 0.20 LY and 0.37 QALY differentials, yielding ICER of $1,230,045/LYg yet ICUR of $664,889/QALYg. All estimates were confirmed with negligible differences in PSAs. In a scenario analysis for the PF1 state only, QALY gains were 5.25 for ACA, 4.40 for ZAN and 4.34 for IBR. Considering cost differentials of $-69,138 for ACA and $-352,297 for ZAN relative to IBR, the corresponding ICURs were $-75,975/QALYg and $-5,871,617/QALYg, resp. Conclusion. Clinically over the course of 1L treatment with a BTKI followed by 2L treatment with VR, ACA in 1L is associated with greater reductions in the risk of progression or death compared to IBR and ZAN over 10 years. This benefit comes at lower cost than IBR, which resulted in cost-saving ICUR and ICER estimates; but at higher cost than ZAN, which resulted in an ICUR of ~$665,000 per QALY gained. Proportionately, QALY results came out more favorably than LY results. This rather unusual result is attributable to the better safety profiles and lower AE costs of ACA and ZAN, coupled with lower pricing. Figure 1View largeDownload PPTFigure 1View largeDownload PPT Close modal

Prognostic Analysis of the IDH1 G105G (rs11554137) SNP in IDH-Wildtype Glioblastoma.

The G105G SNP (rs11554137) in the IDH1 gene is observed in about 10–15% of patients with a diffuse glioma. Data regarding its impact on gliomas are poor and partially conflicting, possibly due to the evolving classification of CNS tumors. The aim of this study was to investigate the G105G SNP prognostic significance in a homogenous cohort of IDH-wildtype glioblastomas, in agreement with the 2021 WHO classification. The study analyzed 211 patients by collecting several clinico-pathological and molecular characteristics, including the age, lesion localization, number of involved lobes, type of surgical treatment, disease outcome and MGMT promoter methylation status. PFS and DSS curves were plotted according to the Kaplan–Meier method and statistical analyses were performed using parametric and non-parametric tests. A total of 32 patients out of 211 (15.2%) were found to be G105G SNP carriers. No significant impact of the IDH1 G105G SNP on patients’ outcomes was observed in terms of PFS and DSS, while MGMT promoter methylation and gross total resection resulted as key prognostic factors in our cohort as expected. No prognostic impact of the IDH1 G105G SNP was detected in this strict cohort of IDH-wildtype glioblastomas. Analysis of larger cohorts is warranted to address the sample size limitations.

Glassy cell carcinoma of the cervix (GCCC): a single institution review of treatment patterns and outcomes (184)

Glassy cell carcinoma of the cervix (GCCC): a single institution review of treatment patterns and outcomes (184)

P1462: POPULATION-BASED REAL WORLD RESULTS OF CD19-DIRECTED CAR T-CELL THERAPY FOR PATIENTS WITH RELAPSED OR REFRACTORY LARGE B-CELL LYMPHOMA: A DUTCH CAR T-CELL TUMORBOARD EXPERIENCE

Background: CD19-directed CAR T-cell therapy has become standard of care for patients with relapsed/refractory (R/R) large B-cell lymphoma (LBCL) and durable responses are seen in up to 40% of treated patients. Optimal selection of patients who will benefit from CAR T-cell therapy is important. In the real world reasons for ineligibility of patients who meet the registration label are often not reported. Additionally, data on health-related quality of life (HR-QoL) after infusion are scarce. Real world (in)eligibility and HR-QoL can play an important role in guiding treatment choices with novel therapeutic options becoming available. In the Netherlands, population-based referral and eligibility assessment for CAR T-cell therapy are centralized in a national tumorboard. A multicenter expert team discusses cases and data are collected in a national registry from time of referral. The aims of the tumorboard and registry are to capture population-based data on patients potentially eligible for CAR T-cell therapy and to optimize (cost-) effective, high quality care. Aims: To evaluate real world results of axicabtagene ciloleucel (axi-cel) for R/R LBCL on a population level, including reasons for ineligibility, efficacy, toxicity and (in a subset of patients) HR-QoL. Methods: For this prospective multicenter cohort study, all adult patients referred to the tumorboard for treatment of R/R LBCL with axi-cel were included from May 1st 2020 – October 19th 2021. Descriptive statistics were used for baseline characteristics, Kaplan Meier methods for survival analysis, regression models for univariate/multivariate analysis and linear mixed models for HR-QoL over time. Results: In total, 186 patients were referred to the tumorboard, 57 were not approved for axi-cel during the online tumorboard meeting, mostly (n=23) due to rapid progression (RP). The 129 approved patients were screened at a CAR T center, 32 were rejected (n=21 RP). Ninety-seven patients underwent apheresis of whom 7 did not receive axi-cel, mostly due to RP (n=6). Ninety patients were infused with CAR T-cells. Data of 74 of these patients are currently available for toxicity/efficacy analysis and updated results, including HR-QoL, will be presented during the EHA meeting. Median age was 59-years (min-max 27-79), 81% received bridging therapy. Best overall response rate was 82.4 %, including 66.2% complete remissions (CR). Median time to response was 1 month and all patients achieved their best response within 3 months. Four patients with PR at 1 month achieved a CR at month 3. The PFS and OS curves are shown in Figure 1. All grades CRS occurred in 89%, including ≥grade 3 in 2.7%, median time to onset was 3 days (min-max 0-14) and median duration 6 days (min-max 1-17). All grades ICANS occurred in 58%, including ≥ grade 3 in 27%, with a median time to onset of 7 days (min-max 1-28) and a median duration of 7 days (min-max 1-28). Non-relapse mortality was 1.4% due to one grade 5 ICANS event. Image:Summary/Conclusion: The merit of a national tumorboard is to enable prospective, in-dept evaluation of rejected as well as CART-infused patients. Twenty-eight percent of the referred patients in the Dutch population that fulfilled the indication criteria for R/R LBCL treatment with axi-cel could not receive the therapy due to rapidly progressive disease. A relatively low percentage of patients, 7%, underwent leukapheresis but eventually did not receive axi-cel. Efficacy and toxicity in our cohort are comparable with ZUMA-1 study results and other large real world cohorts.

Early Serum and Hematological Responses to Pembrolizumab Therapy as Predictors of Survival in Metastatic Urothelial Cancer.

In order to search for predictive biomarkers of efficacy of pembrolizumab therapy for metastatic urothelial cancer (UC), we investigated the relationship between treatment outcomes and early neutrophil-lymphocyte ratio (NLR), lactate dehydrogenase (LDH), and C-reactive protein (CRP) responses. Medical records of 101 patients with metastatic UC who started pembrolizumab as a second-line or later treatment were reviewed. NLR, LDH, and CRP were recorded after 3 weeks of therapy. In addition, we investigated whether these factors had an association with prolonged progression-free (PFS) or overall (OS) survival. The objective response rate, median PFS, and median OS were 25.7%, 6.3 months, and 15.2 months, respectively. PFS and OS were significantly shorter in patients with NLR>3, LDH>upper limit of normal (ULN), and CRP>0.5 mg/dl after 3 weeks of pembrolizumab treatment (p<0.05). A predictive model comprising these factors (favorable risk group: 0 risk factors; intermediate-risk group: 1-2 risk factors; poor-risk group: 3 risk factors) revealed distinct PFS and OS curves (p<0.001). In the favorable risk group, 12-month OS was 79.6%; in the poor-risk group, it was 12.8%. Harrell's C-indices for NLR >3, LDH >ULN, CRP >0.5 mg/dl, and all three combined for predicting OS were 0.656, 0.625, 0.633 and 0.678, respectively. Early responses were also non-significantly associated with ORR (p=0.37). Pembrolizumab treatment outcomes are associated with early NLR, LDH, and CRP responses in metastatic urothelial cancer.

DREAMseq (Doublet, Randomized Evaluation in Advanced Melanoma Sequencing): A phase III trial—ECOG-ACRIN EA6134.

356154 Background: Combinations of immune checkpoint inhibitors (CPI) blocking PD-1 and CTLA-4 or BRAF/MEK inhibitors have both shown significant antitumor efficacy and overall survival (OS) benefit in patients (pts) with BRAFV600-mutant metastatic melanoma (MM), leading to broad regulatory approval. Little prospective data exists to guide the choice of one over the other as initial therapy or the preferred treatment sequence in this population. The DREAMseq Trial was designed to compare the efficacy and toxicity of the sequence of nivolumab/ipilimumab (N/I) followed by dabrafenib/trametinib (D/T) to the converse sequence.Methods: Eligible pts with treatment-naive BRAFV600-mutant MM were stratified by ECOG Performance Status (PS) 0 or 1 and LDH level and randomized 1:1 to receive Step 1 with either N/I (Arm A) or D/T (Arm B) and at disease progression (PD) were enrolled in Step 2 receiving the alternate therapy, D/T (Arm C) or N/I (Arm D), respectively. Pts received N (1mg/kg)/I (3 mg/kg) q3 wks x 4 doses followed by N 240 IV q2 wks for up to 72 wks (Arms A and D) or D 150 mg po BID and T 2 mg po qD until PD (Arms B and C). In 2019, investigators were given the option to use alternate induction dosing of N (3mg/kg)/I (1 mg/kg) q3 wks x 4 doses for Arms A and D. Cycles were every 6 wks and imaging was obtained at baseline and q12 wks on each arm. Primary endpoint was 2-year OS. At the 4th Interim Analysis with 59% of pts being 2 yrs from enrollment, the DSMC and NCI CTEP recommended halting accrual and releasing the data.Results: Beginning 7/2015, 265 out of a proposed 300 pts were enrolled (133 Arm A and 132 Arm B). Median age was 61 (25-85) and 63% were male. Demographics for Arm A and B were balanced with 67% PS 0 and 60% with normal LDH. As of 7/16/21, at a median follow-up of 27.7 mos, 27 pts had switched to Arm C and 46 to Arm D. Overall Grade 3+ toxicity was 60% in Arm A and 52% in Arm B. Grade 5 treatment-related AEs included 2 on Arm A and 1 on Arm C. ORR to date is: Arm A 46% (52/113), Arm B 43% (49/114), Arm C 48% (11/23) and Arm D 30% (8/27). 37/42 assessed pts in Arm A and 19/37 in Arm B remain in response. Median DOR: Arm A- Not reached; Arm B-12.7 mos (95% CI: 8.2, -) (p &lt;0.001). There were 100 deaths (Arm A to C- 38/Arm B to D- 62). 2-yr OS rate for those starting with Arm A was 72% (95% CI: 62-81%) and for Arm B 52% (95% CI: 42-62%) (log-rank p= 0.0095). PFS showed a trend in favor of Arm A (log-rank p=0.054). Both the PFS and OS curves show a biphasic pattern with Arm B being above Arm A until 6 and 10 mos, respectively. For the 115 pts with documented progression on Step 1 (Arm A-44/Arm B-71), 60 (52%) had registered for Step 2. The principal reason for not enrolling on Step 2 was death from PD within 6 mos (Arm A:15/23; Arm B: 25/32). Conclusions: For pts with advanced BRAFV600-mutant MM, the treatment sequence beginning with the CPI combination of N/I resulted in superior OS, which became evident at 10 mos, with longer Step 1 DOR and more ongoing responses than the treatment sequence beginning with D/T. Clinical trial information: NCT02224781.