Brentuximab Vedotin with RCHOP As Frontline Therapy in Patients with High-Intermediate/High-Risk Diffuse Large B Cell Lymphoma (DLBCL): Results from an Ongoing Phase 2 Study

Introduction

RCHOP is the standard frontline treatment for patients with DLBCL.  However, for patients with high-intermediate or high-risk disease (IPI 3-5), outcomes are suboptimal, with an estimated 3-year PFS of about 55%.  In addition to IPI, an additional factor associated with an increased risk include non-GCB cell of origin.  Brentuximab vedotin, an ADC targeting CD30, has demonstrated single-agent activity in relapsed/refractory DLBCL patients.  Of note, higher ORR and CR, as well as longer duration of PFS, were observed in those patients with CD30 detected on tumor cells by routine IHC.

Methods

This phase 2 study was designed to evaluate combination therapy using brentuximab vedotin (BV) and standard RCHOP chemotherapy as frontline treatment in patients with high-intermediate/high risk (standard IPI score 3–5 or age-adjusted IPI [aaIPI] score 2–3) DLBCL, regardless of CD30 expression (ClinicalTrials.gov NCT01925612).  Patients were randomized to receive either 1.2 mg/kg BV+RCHOP or 1.8 mg/kg BV+RCHOP for up to 6 cycles of treatment.  BV was given on Day 1 of every 21-day cycle.  Response was assessed per investigator using Cheson 2007.  AEs, physical examination findings, and laboratory testing were utilized to characterize safety.  Tumor cell CD30 expression was measured by IHC and gene expression analysis; soluble CD30 was measured in plasma.  Tumor microenvironment was also evaluated for frequency of immune-infiltrating cells.

The primary endpoints for this study were the tolerability of each regimen and the CR rate at the end of treatment (EOT).  Key secondary endpoints included progression-free survival (PFS).  Additional exploratory endpoints included CD30 expression on tumor specimens (CD30– defined as <1% per IHC) and correlation of CD30 expression and response.

Results

Twenty-nine patients were treated with 1.2 mg/kg BV+RCHOP and 22 patients were treated with 1.8 mg/kg BV+RCHOP.  Over half (63%) were high-intermediate risk (IPI 3, aaIPI 2) and 37% were high risk (IPI 4-5, aaIPI 3).  Most had Stage IV disease (71%) and 27% had an ECOG score of 2.  AEs in ≥50% of patients included peripheral sensory neuropathy (63%), fatigue (61%), nausea (55%), and diarrhea (53%).  Grade 3 or higher events occurred in 76% of patients; the most frequent were neutropenia (33%) and febrile neutropenia (31%).

The PET-negative CR rate was 69% overall; CD30+ patients had a CR rate of 76% (19/25) compared with 63% (12/19) in CD30– patients by IHC.  Although PFS data are still immature (median follow up of 8 months), the estimated PFS rate at 12 months was 82% (95% CI:  58, 93) in CD30+ patients and 56% (95% CI: 32, 75) in CD30– patients.  60% (3/5) of patients with CD30– ABC-DLBCL progressed versus 27% (3/11) of CD30+ ABC-DLBCL patients.  Four patients with EBV+ DLBCL were treated and all achieved CR; 3 EBV+ patients remain in remission after a median follow-up of 12 months.

Within the microenvironment, a higher percentage of CD3+ T-cells was observed in patients with CR (23%; range, 4-80%) than in those with PR (14%; range, 10-15%) or PD (11%; range, 5-15%).  Of the 32 patients who have not yet progressed, 50% had ≥20% CD3+ cells compared to 18% of patients with PD (2/11).  Other factors, including soluble CD30 levels and infiltrating CD68+ cells, were not associated with clinical outcomes.

Conclusions

Interim results demonstrate that adding BV to RCHOP results in a high rate of CR in this population of IPI 3-5 DLBCL.  Patients with CD30-expression in the tumor appear to have a higher CR rate and fewer early progression events than patients with CD30– DLBCL.  Furthermore, subsets of patients who have a particularly poor prognosis (CD30+ ABC subtype and EBV+ DLBCL) appeared to have a favorable outcome with BV+RCHOP.  Evaluation of the tumor microenvironment showed that higher frequencies of infiltrating CD3+ cells were observed in the CR group, suggesting possible immunologic correlates of response.  However, CD30 expression appears to have greater prognostic significance in this study despite the relatively small sample size.  These results are intriguing and merit further testing in a randomized trial.