Johnson, Nathalie A.

Background Immunotherapies targeting the PD-1/PD-L1 pathway have been shown to induce high response rates in patients with relapsed/refractory (r/r) classical Hodgkin lymphoma (cHL). However, most responses are partial and not durable. T cell immunoglobulin and mucin domain-containing protein-3 (TIM-3) is implicated as a resistance mechanism following anti-PD-1 therapy, and increased TIM-3 expression has been observed on T cells of patients with cHL post anti-PD-1 therapy. Therefore, we hypothesized that dual targeting of PD-1 and TIM-3 may reinvigorate immune responses and lead to more durable antitumor activity. Sabestomig is a monovalent, bispecific, humanized IgG1 monoclonal antibody that binds PD-1 and a unique epitope in the TIM-3 IgV domain without blocking phosphatidylserine binding, eliciting differential functionality. In preclinical mouse models, sabestomig was more effective at inhibiting growth of solid tumors than treatment targeting only PD-1, and sequential sabestomig after anti-PD-1 therapy increased antitumor responses. Here we report preliminary data from the ongoing dose escalation portion of a Phase I/II, open-label, multicenter study to assess the safety and preliminary efficacy of sabestomig in patients with r/r cHL (NCT05216835). Methods Eligible patients are aged ≥16 years with r/r cHL, an Eastern Cooperative Oncology Group performance status 0-1, at least one positron emission tomography-avid measurable lesion according to modified Lugano Criteria, and exposure to at least 2 prior lines of systemic therapy including a minimum of 3 cycles of an anti-PD-(L)1-based therapy. Sabestomig is intravenously infused over 1 hour following pretreatment with diphenhydramine and acetaminophen every 3 weeks, with planned doses ranging from 2 to 2000 mg across 8 cohorts (A1-A8); cohorts A1 to A4 (2, 7, 22.5 or 75 mg) follow an accelerated titration design with a single patient treated at each dose level, while cohorts A5 to A8 (225-2000 mg) follow a modified toxicity probability interval-2 algorithm. The primary endpoint of the dose escalation part of the study is safety, including dose-limiting toxicities (DLTs). Secondary endpoints include efficacy, pharmacokinetics, and immunogenicity. Data cutoffs were May 17, 2023 for safety and June 20, 2023 for efficacy. Results As of May 17, 2023, 14 patients were treated across the first 6 cohorts (A1-A6; 2-750 mg). They were predominantly male (64.3%) with Stage IV disease (85.7%) and a median age of 41.5 years (range, 25-77). Patients received a median of 6 prior lines of therapy (range, 3-13). Baseline disease characteristics are summarized in Table 1. Four patients received sabestomig 2-75 mg in cohorts A1-A4 (n=1 each), 5 patients received 225 mg in cohort A5, and 5 patients received 750 mg in cohort A6. Median duration of exposure to sabestomig was 8.9 weeks (range, 3.0-34.3) with a median of 3 cycles received (range 1-12). Sabestomig was well tolerated with no Grade ≥3 treatment-related adverse events (AEs) (Table 2). One Grade ≥3 AE occurred, which was fatal (sepsis secondary to gastric ulcer rupture, not related to sabestomig but deemed a DLT by the safety review committee). No immune-mediated AEs were reported, and no AEs led to treatment discontinuation. At the safety data cutoff, treatment was ongoing for 6 patients (42.9%). As of June 20, 2023, 11 patients had their first disease assessment and 4/11 had an objective response based on modified Lugano Criteria: 1 patient in cohort A3 (22.5 mg, a partial response [PR]) and all 3 patients who had their first disease assessment in cohort A6 (750 mg, 2 CRs and 1 PR). Two of these 4 responders (both in cohort A6) had experienced a best overall response of progressive disease during prior anti-PD-1 therapy; the other two initially had a response (1 CR [cohort A6] and 1 PR [cohort A3]) but subsequently progressed on anti-PD-1 therapy. Sustained PD-1 receptor occupancy (≥90%) was observed in peripheral blood at doses ≥225 mg. Conclusions Sabestomig was well tolerated with a manageable safety profile. Early efficacy data at 750 mg is encouraging with objective responses in 3/3 patients who had their first disease assessment, including those who were anti-PD-1 refractory or who relapsed on treatment. Updated data will be presented from ongoing and subsequent patients treated during dose escalation with doses of up to 2000 mg.

Abstract Introduction : Follicular lymphoma (FL) remains an incurable malignancy as patients eventually experience progressive disease. A subset of patients is at risk of early lymphoma-related mortality due to histological transformation (TFL) to aggressive lymphoma (2-3% of patients per year) or early progression after immuno-chemotherapy, each of which leads to shortened survival. Mutations associated with transformation and/or early progression have been reported in the literature (e.g. CDKN2A, TP53, B2M) and the m7-FLIPI clinico-genetic risk model defines a high-risk patient group with poor failure-free survival after first-line treatment. However, the current knowledge of genetic alterations underlying transformation and/or early progression is inadequate to explain the majority of transformed cases and accurately predict early progressive disease. Materials and Methods: We performed targeted capture-based sequencing of 94 genes in a large cohort of fresh-frozen and formalin-fixed and paraffin-embedded patient specimens (402 samples in total). This cohort comprised 277 samples from 159 patients that experienced transformation (including 128 samples at T1 (the time of FL) and 149 samples at T2 (the time of transformation) with 118 paired biopsies) and 125 samples from 125 patients (pre-treatment samples only) presenting with either early progression within 2.5 years after starting immunochemotherapy (n=41) or late or never progression for at least 5 years after starting immunochemotherapy or observation (n=84). Mutations were called using MutationSeq and putative single nucleotide polymorphisms were filtered out using either matching germline samples (n=84) or dbSNP (v147). Bayesian proportion tests were used to compare the prevalence of gene mutations between groups. Results : We first compared T1 (n=128) and T2 (n=149) samples from 159 patients experiencing transformation. Eleven genes were more commonly altered in transformed lymphoma. These included genes that had previously been associated with transformation, such as TP53, B2M, MYC and EBF1, as well as genes that have not yet been implicated as contributing to this process, for example EZH2, CCND3, PIM1 and ITPKB. Moreover, mutations in GNA13, S1PR2 and P2RY8, that have been implicated in dissemination of germinal centre B cells, were enriched in T2 samples. In addition, cell-of-origin classification was available for 108 of the TFL cases with DLBCL histology, 18 and 90 of which were of the ABC and GCB subtype, respectively. Although the number of ABC-TFL cases was small, we observed higher percentages of BCL10 (16% versus 1%, Fisher P=0.004), CD79B (22% versus 4%, Fisher P=0.005) and MYD88 mutations (28% versus 9%, P=0.006) in ABC-TFL than in GCB-TFL, suggesting that B-cell receptor signaling and NF-κB signaling are important contributors to the ABC phenotype in TFL. Next, we assessed the association of gene mutations with patient outcomes contrasting early progressers (n=41) and late or non-progressers (n=84). Eleven genes were mutated more commonly in early progressers than in late progressers, including KMT2C, TP53, BTG1, MKI67, XBP1 and SOCS1. Overall, 32 out of 41 early progressers (78%) had mutations in at least one of the 11 early progression-associated genes, but none of the individual early progression-associated genes were mutated at a frequency > 22%. Thus, early progression appears to be related to relatively infrequent genetic alterations. None of the early progression-associated gene mutations form part of the m7-FLIPI outcome predictor. Furthermore, in our cohort that was enriched for clinical extremes, the m7-FLIPI was similarly associated with early progression when compared to the FLIPI but it was not superior, having higher specificity (89% versus 76%), but lower sensitivity (36% versus 63%). Conclusions: We found novel associations of gene mutations with transformation and showed that TFL is molecularly heterogeneous, with the ABC subtype being characterized by differential gene mutation when compared to the GCB subtype. With regards to progressive disease after immuno-chemotherapy, our approach identified early progression as a distinct clinico-genetic disease category that is imperfectly captured by traditional prognostic tools. Disclosures Connors: Seattle Genetics: Research Funding; Bristol Myers Squib: Research Funding; F Hoffmann-La Roche: Research Funding; NanoString Technologies: Research Funding; Millennium Takeda: Research Funding. Scott:NanoString Technologies: Patents & Royalties: named inventor on a patent for molecular subtyping of DLBCL that has been licensed to NanoString Technologies.