Mansfield, Aaron S.

125 Background: The advent of immunotherapy (IO) has led to significant improvement in outcomes for patients (pts) with advanced non-small cell lung cancer (adv NSCLC). The rate of crossovers and receipt of post-protocol IO in pivotal trials leading to FDA approvals of IO in NSCLC has not been systematically evaluated. Here, we evaluate crossover rates and post protocol therapies for pts with adv NSCLC across multiple first-line (1L) IO monotherapy and chemotherapy/IO (chemo/IO) combination trials. Methods: We utilized the publicly available data from pivotal clinical trials leading to approvals of IO or chemo/IO regimens in 1L treatment of adv NSCLC. We extracted data on outcomes, rate of crossover from control arm to IO, proportion of pts in control arm receiving IO and the start dates of these clinical trials. The primary outcomes were the rates of crossover and the proportion of patients in control arms who received post-control IO. Results: The study included 4 trials with IO monotherapy and 12 trials with chemo/IO combinations in 1L adv NSCLC. The primary endpoints for these trials were PFS (25%), OS (19%), and both PFS and OS (56%). The crossover rate from control arm to experimental arm (with IO) ranged from 0-74% in IO monotherapy trials and 0-49% in chemo/IO trials. Two IO monotherapy trials and five chemo/IO trials did not allow crossover; among them, 3 trials had a PFS/OS co-primary endpoint, while others had OS as primary endpoint. Ten of 16 trials provided explicit information on use of subsequent post-protocol therapies in their publications. Among the two IO monotherapy trials which did not allow crossover, post-protocol IO was administered in only 20-30% of patients. Among the six chemo/IO trials with information on post-protocol therapies, 30%-59% patients on the control arm subsequently received some form of IO on progression. Nine of 12 trials started accrual after 10/2015, when nivolumab was approved in the United States as second-line (2L) therapy for adv NSCLC regardless of tumor PD-L1 expression. Conclusions: Despite the highly significant OS benefit from 2L IO, which was the standard of care (SOC) in the United States, the rates of crossover and post-protocol IO administration was distressingly low in 1L IO monotherapy and chemo/IO trials for 1L adv NSCLC. This low rate of 2L treatment with IO may have been due to limited global availability prior to widespread regulatory approval during the conduct of these trials. There is an increased need for consistency in reporting of crossover treatment and post-protocol treatments to allow adequate assessment of the true 1L benefit with IO. Control arms in pivotal trials require scrutiny to ensure confirmation with SOC to provide access to optimal treatments for patients and prevent magnification of observed benefits in experimental arms. The difficulty lies in the global conduct of large randomized clinical trials with differing regulatory approvals.

6080 Background: Alterations in RET are targetable oncogenic drivers in TC. Pralsetinib is a highly potent, selective RET inhibitor, with demonstrated efficacy in pts with RET-altered TC. In a previous analysis of the Phase I/II ARROW trial (NCT03037385; data cutoff: 22 May 2020), overall response rates [ORR; measurable-disease population] with pralsetinib at 400 mg once daily (QD) were 60% (33/55) and 71% (15/21) in pts with RET-mutant medullary TC ( RET-mutant MTC) who had received prior multikinase inhibitors cabozantinib and/or vandetanib (C/V), and those who were treatment naïve, respectively, and 89% (8/9) in pts with previously treated RET fusion-positive TC ( RET-fp TC). Here we report an updated analysis of these cohorts in the intention-to-treat (ITT) population. Methods: Adult pts with RET-altered locally advanced/metastatic TC who had enrolled in ARROW and initiated pralsetinib at 400 mg QD, were included. Phase II primary endpoints: ORR (by blinded independent central review, per RECIST v1.1) and safety. Efficacy endpoints for this analysis were assessed in the ITT population. Safety was assessed in all pts with RET-altered TC who initiated pralsetinib at 400 mg QD. Enrolment cutoff: 23 August 2020 for the ITT population; data cutoff: 12 April 2021. Results: The ITT population comprised 145 pts with RET-mutant MTC (with or without prior systemic therapy, including C/V) and 22 pts with RET-fp TC, 21 of whom had received prior systemic therapy (including multikinase inhibitor[s] and/or radioactive iodine). In pts with RET-mutant MTC who had received prior C/V (n = 67), ORR was 51% (34/67; 95% CI 38–63; 2 complete responses [CR]; 32 partial responses [PR]), median duration of response (DoR) was 25.8 months (95% CI 18.0–not reached [NR]) and median progression-free survival (PFS) was 24.9 months (95% CI 19.7–31.2). In treatment-naïve pts with RET-mutant MTC, ORR was 72% (48/67; 95% CI 59–82; 4 CR; 44 PR), and median DoR and median PFS were not reached. In pts with previously treated RET-fp TC, ORR was 86% (18/21; 95% CI 64–97; 3 CR; 15 PR), median DoR was 17.5 months (95% CI 16.0–NR) and median PFS was 19.4 months (95% CI 13.0–NR). Median overall survival (OS) was not reached in any of the cohorts. The safety population comprised 172 pts with RET-altered TC treated at 400 mg QD. The most common treatment-related adverse events (TRAEs) were increased aspartate aminotransferase (n = 67; 39%), anemia (n = 60; 35%), hypertension (n = 57; 33%) and decreased white blood cell count (n = 52; 30%). Serious TRAEs were reported in 27 pts (16%); the most frequent was pneumonitis (n = 5; 3%). Nine pts (5%) discontinued pralsetinib due to a TRAE and one patient ( < 1%) died due to a TRAE ( pneumocystis jirovecii pneumonia) following 44 days ( < 3 cycles) on pralsetinib. Conclusions: In this updated analysis including more pts, pralsetinib continues to be efficacious with a manageable safety profile in pts with RET-altered TC. Clinical trial information: NCT03037385.

Abstract Seven different anti-PD-1 and PD-L1 monoclonal antibodies are now widely used in the US to treat a variety of cancer types, but there have been no clinical trials comparing them directly. Furthermore, because many of these antibodies do not cross react between mouse and human proteins, no preclinical models exist in which to consider these types of questions. Thus, we produced humanized PD-1 and PD-L1 mice in which the extracellular domains of both mouse PD-1 and PD-L1 were replaced with the corresponding human sequences. Using this new model, we sought to compare the strength of the immune response generated by FDA-approved antibodies. To do this, we performed an in vivo T cell priming assay in which anti-PD-1/L1 therapies were given at the time of T cell priming against surrogate tumor antigen (OVA), followed by subsequent B16-OVA tumor challenge. We found that anti-PD-1/L1-treated mice exhibited significantly better tumor rejection than controls, although both the control and antibody-treated mice generated comparable numbers of OVA-specific T cells at the time of priming. To determine what could mediate this strong antitumor immune response, we identified the increased production of CX3CR1+PD-1+CD8+ cytotoxic T cells in the anti-PD-1/L1-treated mice, the same subset of effector T cells known to increase in clinical responders to PD-1/L1 therapy. Thus, our model permits the direct comparison of FDA-approved anti-PD-1/L1 monoclonal antibodies and further correlates successful tumor rejection with the level of CX3CR1+PD-1+CD8 + T cells, making this model a critical tool for future studies to optimize and better utilize anti-PD-1/L1 therapeutics.

9008 Background: Amivantamab, a fully human bispecific antibody targeting epidermal growth factor receptor (EGFR) and MET, is approved for the treatment of non-small cell lung cancer (NSCLC) with EGFR exon 20 insertion after prior platinum-based chemotherapy. Given its bispecific nature, amivantamab is being explored in patients (pts) with primary MET exon 14 skipping mutation (METex14) in the MET-2 cohort of the CHRYSALIS study. Methods: CHRYSALIS (NCT02609776) is an ongoing phase 1 dose escalation/dose expansion study of amivantamab in pts with advanced NSCLC. Pts with primary METex14 whose disease progressed on or who declined current standard of care therapy were treated with amivantamab 1050 mg (pts <80 kg) or 1400 mg (pts ≥80 kg) weekly in cycle 1 and biweekly thereafter. Response was assessed by investigators using RECIST v1.1. Results: As of 2 Dec 2021, 43 pts with METex14 had received amivantamab. Median age was 70 y (range, 43-88), 58% were women, median prior lines of therapy was 2 (range, 0-10) [eg, crizotinib (n=13), capmatinib (n=11), tepotinib (n=5), anti-MET antibody (n=1)], and 23% had history of brain metastases at baseline. In 36 pts with ≥1 postbaseline disease assessment, median duration of follow-up was 5.8 months (range, 0.3-15.8); 6 pts had no prior treatment, 11 had no prior MET inhibitor, and 19 had a prior MET inhibitor. Overall response rate was 33% (50% [3/6] in treatment-naïve pts, 46% [5/11] in pts with no prior MET inhibitor, and 21% [4/19] in pts with prior MET inhibitor therapy). Clinical benefit rate was >54% regardless of prior treatment (Table). Median duration of response (DOR) was not reached (range, 2.1-12.2 months); 67% (8/13) had DOR ≥6 months. Ten of the 12 responders remain on treatment (6.0-14.4 months) with ongoing responses; 2 discontinued after 2 and 12 months, respectively. Safety profile was consistent with previously reported experience of amivantamab (Sabari 2021 JTO 16(3):S108-109). Treatment-related adverse events leading to dose reduction or discontinuation occurred in 3 pts, each. Conclusions: Amivantamab demonstrates anti-tumor activity in primary METex14 NSCLC including after prior MET inhibitor treatment. Enrollment is ongoing and updated data will be shown. Clinical trial information: NCT02609776. [Table: see text]

TPS8604 Background: Chemoimmunotherapy followed by durvalumab maintenance yields a median overall survival of 12.9 months in patients with extensive stage Small Cell Lung Cancer (ES SCLC), which is an improvement over chemotherapy alone. However, 90% of these patients will have progressive disease. While topotecan and lurbinectedin have established modest activity in the second line, it is unknown whether continuing immunotherapy in this setting confers additional benefit. In preclinical studies lurbinectedin, a DNA minor groove binder, used with immune checkpoint inhibitors has synergistic effects. Methods: This phase 2 trial is enrolling patients with ES SCLC who have progressed on platinum based chemoimmunotherapy, to three treatment groups. Group 1 includes patients with platinum sensitive SCLC who will receive durvalumab (1500 mg given as an intravenous [IV] infusion once every 3 weeks) and topotecan (1.25 mg/m2/day IV for 5 consecutive days every 3 weeks). In Groups 2A and 2B, patients with platinum sensitive and platinum resistant disease respectively, receive durvalumab and lurbinectedin (3.2 mg/m2 IV on Day 1 of every 21-day cycle). Patients with platinum sensitive disease are assigned to Groups 1 or 2A based on the preferences of the treating physician and the patient. Patients with treated/stable CNS metastases are eligible. The primary endpoint is the proportion of patients who are alive at 6 months (6OS) for Group 1 and the proportion of patients who are alive and progression-free at 6 months (6PFS) in Groups 2A and 2B. Secondary endpoints include safety, adverse event profile, response rate, PFS, and OS. The sample size is based on a 2-stage Simon Optimal Design. For Treatment Group 1, with 22 eligible patients there is 80% power to detect a true 6-month OS rate (6OS) of 75%, with 10% alpha under the null hypothesis that the true 6OS is at most 50%. For Treatment Group 2A, with 20 eligible patients this design has 80% power to detect a true 6-month PFS rate (6PFS) of 65%, with 10% alpha under the null hypothesis that the true 6PFS is at most 40%. For Treatment Group 2B, with 22 eligible patients this design has 80% power to detect a true 6-month PFS rate (6PFS) of 40%, with 10% alpha under the null hypothesis that the true 6PFS is at most 19%. To account for possible drop-outs, accrual targets will be 24, 22, and 24 patients to Groups 1, 2A, and 2B respectively. For the safety analyses, 6 patients will be enrolled at the starting dose level for each treatment group (1, 2) and then briefly closed to accrual to assess adverse events. If we observe 2+ DLTs in these 6 treated patients during Cycle 1 within a treatment group (1 vs. 2), we will declare the combination treatment too toxic and lower the starting dose of chemotherapy for the next 6 patients. The study was open for all 3 groups as of January 2022 and has accrued 2 patients. Clinical trial information: NCT04607954.

TPS8598 Background: Peritoneal mesothelioma is a rare and poorly studied disease with few treatment options. For patients who are not surgical candidates, treatment recommendations for systemic therapy have been extrapolated from clinical trials for pleural mesothelioma that commonly exclude patients with peritoneal mesothelioma. Recently, the combination of the PD-1 inhibitor nivolumab and the CTLA-4 inhibitor ipilimumab received FDA-approval for the frontline treatment of non-resectable pleural mesothelioma. Additionally, a prospective, non-randomized phase 2 trial demonstrated activity with combined PD-L1 (atezolizumab) and VEGF (bevacizumab) blockade in peritoneal mesothelioma. In parallel, encouraging activity with combined chemo-immunotherapy has been reported in pleural mesothelioma. Given the benefits observed with immunotherapy, and the potential to improve upon those with chemotherapy and VEGF inhibition, we seek to determine whether the addition of the PD-L1 inhibitor atezolizumab improves outcomes with chemotherapy and bevacizumab in patients with newly diagnosed peritoneal mesothelioma. Methods: A092001 is a prospective, randomized phase 2 clinical trial. All patients with newly diagnosed peritoneal mesothelioma will be randomized 1:1 using a dynamic allocation Pocock-Simon procedure to receive carboplatin, pemetrexed, and bevacizumab, with or without atezolizumab, every 21 days for four cycles. Patients who are eligible to proceed with surgery after four cycles of therapy will then do so. Patients who are not eligible to proceed with surgery may receive maintenance bevacizumab and atezolizumab, or second-line atezolizumab with bevacizumab until progression of disease or toxicity. The primary objective is to determine whether frontline treatment with carboplatin, pemetrexed, bevacizumab and atezolizumab results in a superior best response rate (RR) to carboplatin, pemetrexed and bevacizumab as determined by RECIST. With 31 eligible patients per arm (62 eligible total), this randomized design has 80% power to detect an improvement in the RR from 20% to 45%, with a 1-sided significance level of 0.10 where an interim futility analysis will be conducted after 32 patients are enrolled. As stratification factors we have included eligibility for cytoreductive surgery at diagnosis, and histologic subtype. Secondary endpoints include assessment of progression-free survival, overall survival, and adverse events. As integrated biomarkers, we will determine if soluble mesothelin-related peptides and megakaryocyte potentiating factor correlate with responses. This trial was recently approved by the National Cancer Institute Central IRB and is activating at sites across the country. Support: U10CA180821, U10CA180882. Clinical trial information: NCT05001880.

8549 Background: First-line treatment of early stage non-small cell lung cancer (NSCLC) is surgical resection, but with a 5-year survival of only 54%, rates of future relapse are high. Identifying patients at high risk of relapse can help guide adjuvant treatment decisions. Deep convolutional neural network (DCNN) AI models trained on tumor histology have shown incredible flexibility as potential biomarkers, both in lung cancer and more generally across many malignancies. While DCNN models can obtain extremely accurate results when used for routine purposes such as diagnosis, subtyping, and grading of malignancies, most models trained for prognostication or treatment response prediction do not reach performance sufficient for clinical application. Uncertainty quantification (UQ) – a family of techniques that give DCNN models the ability to report confidence alongside predictions – is an underexplored avenue in cancer AI that may help further improve performance and clinical application of models designed to provide clinicians with estimations of risk. Methods: To explore the potential use of UQ in clinically-oriented DCNN models, we trained models on a single-institution, retrospective digital tumor histology slide cohort from patients with Stage I-III NSCLC who underwent surgical resection to predict risk of future relapse. Estimation of uncertainty was performed using dropout as a Bayesian approximation, and uncertainty thresholds were calculated from training sets using a novel method to identify and remove low-confidence predictions. For comparison, a separate multivariate logistic regression was trained in cross-validation using known clinical risk factors. Results: We trained DCNN models on slides from 198 patients (40 relapsed, 158 without relapse). In this cohort, 130 patients had stage I disease (65.6%), 42 had stage II (21.2%), and 26 had stage III (13.2%). The average age was 69 years, 85% were current or previous smokers, and 38 received guideline-concordant adjuvant chemotherapy. Without UQ, a DCNN model predicted risk of future relapse with an average area under receiver operator curve (AUROC) of 0.74 across three-fold cross-validation. Using UQ estimation, 63% of slides were reported with high confidence. Relapse prediction was significantly improved in the high-confidence cohort, with an average AUROC of 0.83 in cross-validation. With a specificity of 70%, this corresponds to an average sensitivity of 86.8% across the three cross-folds (79.1%, 85.5%, and 95.7%). In comparison, a clinical-only multivariate regression model predicted relapse with a cross-validated AUROC of 0.67. Conclusions: This method of uncertainty quantification appears to be a powerful tool to predict lung cancer recurrence risk from digital histopathology while simultaneously providing clinicians with a measurement of algorithm trustworthiness.

PURPOSE This investigation sought to evaluate the prognostic value of pretreatment of circulating tumor DNA (ctDNA) in metastatic biliary tract cancers (BTCs) treated with platinum-based first-line chemotherapy treatment. MATERIALS AND METHODS We performed a retrospective analysis of 67 patients who underwent ctDNA testing before platinum-based chemotherapy for first-line treatment for metastatic BTC. For analysis, we considered the detected gene with highest variant allele frequency as the dominant clone allele frequency (DCAF). Results of ctDNA analysis were correlated with patients' demographics, progression-free survival (PFS), and overall survival (OS). RESULTS The median age of patients was 67 (27-90) years. Fifty-four (80.6%) of 67 patients evaluated had intrahepatic cholangiocarcinoma; seven had extrahepatic cholangiocarcinoma, and six gallbladder cancers. Forty-six (68.6%) of the patients were treated with cisplatin plus gemcitabine, and 16.4% of patients received gemcitabine and other platinum (carboplatin or oxaliplatin) combinations, whereas 15% of patients were treated on a clinical trial with gemcitabine and cisplatin plus additional agents (CX4945, PEGPH20, or nab-paclitaxel). TP53, KRAS, FGFR2, ARID1A, STK11, and IDH1 were the genes with highest frequency as DCAF. The median DCAF was 3% (0%-97%). DCAF > 3% was associated with worse OS (median OS: 10.8 v 18.8 months, P = .032). Stratifying DCAF in quartiles, DCAF > 10% was significantly related to worse PFS (median PFS: 3 months, P = .014) and worse OS (median OS: 7.0 months, P = .001). Each 1% increase in ctDNA was associated with a hazard ratio of 13.1 in OS when adjusting for subtypes, metastatic sites, size of largest tumor, age, sex, and CA19-9. CONCLUSION DCAF at diagnosis of advanced BTC can stratify patients who have worse outcomes when treated with upfront platinum-based chemotherapy. Each increase in %ctDNA decreases survival probabilities.

Background This phase I multicenter study was designed to evaluate the safety, tolerability, efficacy, and translational effects on the tumor microenvironment of itacitinib (Janus-associated kinase 1 (JAK1) inhibitor) in combination with epacadostat (indoleamine 2,3-dioxygenase 1 (IDO1) inhibitor) or parsaclisib (phosphatidylinositol 3-kinase δ (PI3Kδ) inhibitor). Methods Patients with advanced or metastatic solid tumors were enrolled and received itacitinib (100–400 mg once a day) plus epacadostat (50–300 mg two times per day; group A), or itacitinib (100–400 mg once a day) plus parsaclisib or parsaclisib monotherapy (0.3–10 mg once a day; group B). Results A total of 142 patients were enrolled in the study. The maximum tolerated dose was not reached for either the combination of itacitinib plus epacadostat (n=47) or itacitinib plus parsaclisib (n=90). One dose-limiting toxicity of serious, grade 3 aseptic meningitis was reported in a patient receiving itacitinib 300 mg once a day plus parsaclisib 10 mg once a day, which resolved when the study drugs were withdrawn. The most common treatment-related adverse events among patients treated with itacitinib plus epacadostat included fatigue, nausea, pyrexia, and vomiting, and for patients treated with itacitinib plus parsaclisib were fatigue, pyrexia, and diarrhea. In the itacitinib plus epacadostat group, no patient had an objective response. Among patients receiving itacitinib 100 mg once a day plus parsaclisib 0.3 mg once a day, three achieved partial response for an objective response rate (95% CI) of 7.1% (1.50 to 19.48). Treatment with itacitinib plus epacadostat demonstrated some increase in tumor CD8+ T cell infiltration and minor changes in six plasma proteins, whereas treatment with itacitinib plus high-dose parsaclisib resulted in downregulation of 20 plasma proteins mostly involved in immune cell function, with no observed change in intratumoral CD8+ T cell infiltration. Conclusion Adverse events with JAK1 inhibition combined with either IDO1 or PI3Kδ inhibition were manageable, but the combinations demonstrated limited clinical activity or enhancement of immune activation in the tumor microenvironment. Trial registration number NCT02559492.