Interim Safety Analysis of ALLG-Pacific (NHL35): An Open Label Phase II Study of Pembrolizumab and Chemo-Immunotherapy As First-Line Therapy for Primary Mediastinal B-Cell Lymphoma

Background: Primary mediastinal B-cell lymphoma (PMBL) comprises 10% of diffuse large B-cell lymphoma (DLBCL), primarily affecting young adults. Combination chemoimmunotherapy is effective, but there is no consensus on the optimal therapeutic regimen. The randomized IELSG37 study demonstrated no improvement to progression free survival (PFS) when consolidation radiotherapy was administered following chemoimmunotherapy. However, PFS was inferior for patients (pts) treated with R-CHOP-21 (rituximab, cyclophosphamide, doxorubicin, vincristine, prednisolone every 21 days) compared with other regimens, supporting intensification of therapy beyond R-CHOP-21 in this entity. Programmed-death- (PD-1) ligands PD-L1/PD-L2 are frequently upregulated in PMBL, and the PD-1 inhibitor pembrolizumab (pembro) blocks interaction of PD-1 and PD-L1/2, and subsequent signalling. Pembro has encouraging activity in relapsed/refractory PMBL with an overall response rate of 43% (complete response 23%) and modest toxicity in a phase II study, and is predicted to be synergistic with rituximab. Therefore, there is compelling rationale for combining R-CHOP and pembro as first-line therapy for PMBL.

Methods: ALLG-PACIFIC (ACTRN12621001529831p) is a 35 patient, phase II, single arm, open-label study of R-CHOP in combination with pembro for pts with newly diagnosed PMBL. Pts receive 2 cycles of rituximab (375mg/m²) plus pembro (200mg) every 21 days (‘window phase’) followed by 6 cycles of R-CHOP plus pembro every 21 days (‘induction phase’) followed by 5 cycles of pembro 400mg every 42 days (‘consolidation phase’). Adults (≥18 years) with treatment naïve histologically confirmed PMBL and adequate organ function, no contraindications to R-CHOP and with no active autoimmune disease are eligible.

The primary endpoint is 18-month event free survival. A key secondary endpoint is safety of treatment including rates of early discontinuation due to toxicity. Here, we present the results of a planned interim safety analysis.

Results: As of 10^th July 2024, 12 pts have completed ‘window’ and ‘induction’ treatment phases (ie. 8 cycles). Three pts have withdrawn consent; 1 due to rash during the first window cycle, one due to patient preference to receive treatment off study, and one due to alopecia and nausea following completion of window treatment, who has subsequently consented to remain on the study for efficacy and survival follow up. All 15 pts are included in this safety analysis. All pts had ≥1 adverse event (AE), with 7 pts (47%) experiencing AEs leading to dose delay. Eight pts (53%) experienced ≥1 possible immune related AE (irAE); rash (G2 n=3, G3 n=3) and alanine aminotransferase (ALT) elevation (G2 n=1, G3 n=3), all with alternate possible causative concomitant medications. Twenty five ≥G3 AEs occurred in 12 pts; rash (n=3), ALT elevation (G3 n=3, G4 n=1), sinus tachycardia (G3 n=2), lipase elevation (G3 n=2), infections (G3 n=3), neutropenia (G3 n=2, G4 n=1) febrile neutropenia (G3, n=2), fever (G3, n=1), and one each G3 of anemia, headache, syncope, chylothorax. There was 1 G3 disease related pericardial effusion. There were no G5 AEs or AEs leading to treatment discontinuation per protocol.

Conclusions: This interim analysis of ALLG-PACIFIC suggests that R-CHOP-21 in combination with pembro as first-line therapy for PMBL has a manageable safety profile consistent with that expected of the agents involved. There were no unexpected toxicities, however rates of rash and ALT elevation during ‘window’ phase have led to a recommendation to avoid concomitant administration of supportive medications during ‘window’ phase that may precipitate these AEs. (eg. Allopurinol, co-trimoxazole). Study enrolment, safety and efficacy analyses are ongoing.