Ipilimumab with and without Nivolumab in Patients with Classic Hodgkin Lymphoma with Progression after PD-1 Blockade

Background: CTLA-4 and its ligand, CD86, are expressed within the classic Hodgkin Lymphoma (cHL) tumor microenvironment (TME) at even higher frequency than PD-1 and its ligands, suggesting that the CTLA-4 pathway plays a key role in immune evasion. Monotherapy with the CTLA-4 monoclonal antibody (mAb), ipilimumab (ipi), yielded responses in patients (pts) with cHL relapsing after allogeneic stem cell transplantation; however, trials combining ipi with a PD-1 mAb (+/- brentuximab vedotin) in earlier lines of therapy failed to demonstrate a clear benefit with the addition of ipi. These results suggest that ipi has therapeutic potential in cHL, but that combination treatment in unselected pts may not be a successful strategy. To further investigate the role of ipi in cHL, we conducted a phase II clinical trial testing ipi with or without nivolumab (nivo) among pts with R/R cHL who progressed after PD-1 blockade.

Methods: Adult pts with R/R cHL who had received ≥ 2 prior therapies including a PD-1 mAb were eligible. The trial was initially designed with 2 cohorts - Cohort 1 for pts with a best response of stable disease (SD) or partial response (PR) after at least 18 weeks of PD-1 mAb monotherapy and Cohort 2 for pts with progressive disease (PD) on PD-1 based treatment. Cohort 1 was closed for slow accrual (no pts enrolled). In cohort 2, pts received ipi 3 mg/kg every 3 weeks for 4 doses and then underwent restaging with positron emission tomography. Ipi responders (complete response [CR] or PR) continued ipi maintenance (3 mg/kg every 12 weeks for up to 8 doses). Ipi non-responders (SD or PD) who were clinically stable received 4 cycles of ipi (1 mg/kg) and nivo (3 mg/kg) dosed every 3 weeks followed by ipi maintenance, as above. The primary endpoint was objective response rate (ORR) to ipi monotherapy, assessed using 2014 Lugano criteria.

Results: 13 pts were enrolled to Cohort 2 at 3 centers. The median age was 39 (23-76) and pts had received a median of 4 (3-17) prior therapies, including a PD-1 mAb-based regimen for all pts (median 1, range 1-3). 6 pts received PD-1 monotherapy only, 2 received PD-1-based combinations, and 5 received both PD-1 monotherapy plus 1 or more PD-1 combinations. The best ORR for PD-1 monotherapy was 27% (3/11) and for PD-1 combinations 50% (5/10). A PD-1 based regimen directly preceded study treatment for 9 pts. Pts received a median of 2 cycles of ipi monotherapy (range 1-4). 3 pts achieved a PR (ORR 23%) with no CRs observed. 3 of 10 ipi non-responders received combination treatment with nivo and ipi and none experienced an objective response. The most common reasons for treatment discontinuation were PD (n=8), toxicity (n=4), and pt decision (n=1). With a median follow-up of 11.5 months, the median progression-free survival for ipi monotherapy was 3.3 months. The duration of response for individual responders was 27.8, 3.2, and 1.5 months. The 1-year overall survival was 59%. The most common treatment-related adverse events (TRAE) with ipi monotherapy were fever (n=6) (including 3 cases in the absence of an identifiable infection), ALT elevation (n=5), cough (n=4), diarrhea (n=4), anorexia (n=3), dyspnea (n=3), fatigue (n=3), headache (n=3), nausea (n=3), pneumonitis (n=3), and maculopapular rash (n=3). Grade 3+ TRAEs occurred in 6 pts and included ALT/AST elevation in 1 pt; abdominal pain, cytokine release syndrome, dyspnea, lung infection, and thrombocytopenia in 1 pt; urinary tract infection in 1 pt; colitis in 1 pt; diarrhea and hyponatremia in 1 pt; and anorexia in 1 pt. TRAEs for the 3 pts who received ipi + nivo included 1 case each of hypothyroidism, bilateral hand rash, infusion-related reaction, and peripheral sensory neuropathy (all grade 1-2). Pre-treatment biopsy samples from all pts were analyzed using multiplex immunofluorescence (to define TME features among PD-1 relapsed cHL and to describe features associated with response) and will be presented at the meeting.

Conclusions: Despite strong preclinical data supporting a key role for the CTLA-4 axis in cHL, we observed limited clinical activity with ipi monotherapy in this high-risk cohort of pts with multiply relapsed cHL. The toxicity profile in this heavily pre-treated population was notable for frequent immune-related AEs and higher rates of treatment discontinuation due to toxicity than seen with ipi in other clinical settings. Correlatives studies are ongoing which may shed further light on predictors of response.