AUTHORS
A. Vera De Jonge, MD, PhD, Marcel Nijland, MD, PhD2, Joost S.P. Vermaat, MD, PhD, MSc3, Marjolein van der Poel, MD, PhD4, Koen Heer, MD5, Clara Klerk, MD6, Rogier Mous, MD, PhD7, Erik D. van Werkhoven, PhD8, Brigitta AG Schuthof, MSc9, Marie Jose Kersten, MD10, Martine E.D. Chamuleau, MD PhD11
AFFILIATIONS
1 Amsterdam UMC Location Vrije Universiteit, Amsterdam, Netherlands, Cancer Center Amsterdam, Amsterdam, Netherlands
2 University Medical Center Groningen, Groningen, Netherlands
3 Leiden University Medical Center, Leiden, Netherlands
4 Maastricht University Medical Center, Maastricht, Netherlands
5 Flevoziekenhuis, Almere, Netherlands
6 Dijklanderziekenhuis, Hoorn, Netherlands
7 University Medical Center Utrecht, Utrecht, Netherlands
8 HOVON foundation, Rotterdam, Netherlands
9 HOVON foundation, Rotterdam, Netherlands
10 Amsterdam UMC Location Vrije Universiteit, Amsterdam, Netherlands
11 Amsterdam UMC Location Vrije Universiteit, Amsterdam, Netherlands, Cancer Center Amsterdam, Amsterdam, Netherlands
ABSTRACT
Introduction
Patients with double and triple hit (DH/TH) high-grade B-cell lymphoma (HGBL) face a poor prognosis with standard R-CHOP. We previously conducted two phase II trials evaluating alternative strategies -in this high-risk population.
The HOVON-130 trial demonstrated that adding lenalidomide, an oral immunomodulatory drug (patent expired) with indirect MYC suppressive activity, to R-CHOP (R2CHOP) was well-tolerated, with a 67% end-of-treatment (EOT) complete metabolic remission (CMR) rate and a 2-year disease-free survival of 75% (PMID 33256379). The subsequent HOVON-152 trial evaluated a higher-intensity induction regimen with DA-EPOCH-R.
Despite widespread use of DA-EPOCH-R, no randomized trials have demonstrated superiority over less intensive regimens. We therefore performed a non-randomized, risk-adjusted comparison between DA-EPOCH-R and R2CHOP.
Methods
In the single-arm, phase II HOVON-130 trial, n=85 patients with MYC-rearranged large B-cell lymphoma received 6 cycles R-CHOP + lenalidomide 15 mg day 1-15 (PMID 33256379). For the current comparison, 51 DH/TH patients were selected (single-hit cases were excluded).
From the HOVON-152 trial, 95 patients with DH/TH HGBL who received 6 cycles of DA-EPOCH-R were selected. Patients achieving PET-CT-based CMR after DA-EPOCH-R received 1 year of nivolumab consolidation.
Both trials had similar inclusion criteria: ≥ 18 years, Ann-Arbor stage II-IV, WHO performance status (PS) 0-3.
CMR rate, progression-free survival (PFS) and overall survival (OS) using unadjusted Cox regression and adverse events (AE as CTCAE ≥ gr 2) were compared. To account for baseline imbalances and treatment-selection bias, we applied three complementary statistical methods: 1:1 matching, multivariable Cox regression, and inverse probability of treatment weighting (IPTW) with a doubly-robust augmented IPTW estimator.
Adjustment variables included age, sex, stage, WHO PS, LDH, and BCL2/BCL6 DH/TH status.
Results
Patients treated with DA-EPOCH-R +/- nivolumab (n=95) and patients treated with R2CHOP (n=51) did not differ significantly in age (median 60.4 vs 62.2 years, resp.), sex (35% female in both cohorts), stage, WHO PS and LDH. Patients treated with DA-EPOCH-R +/- nivolumab less often had a high IPI score than patients treated with R2CHOP (9.5% vs 25.5%; p=0.04).
Incidence of MYC/BCL2 DH, MYC/BCL6 DH and TH was 68%, 12% and 16% in DA-EPOCH-R (4% missing 1 variable thus DH or TH) and 47%, 22% and 18% in R2CHOP treated patients (13% DH or TH) (p=0.03).
CMR rates at EOT were similar for patients treated DA-EPOCH-R +/- nivolumab (63/95, 66.3%) and R2CHOP (36/51, 70.6%; p=0.60).
Median PFS was 4.3 years (95% CI: 2.52–NR) in DA-EPOCH-R +/- nivolumab versus 3.6 years (95% CI: 2.11-NR) in R2CHOP (p=0.8) treated patients. Median OS was not reached in either group (p=0.3).
All four statistical analyses consistently showed no significant differences between DA-EPOCH-R +/- nivolumab versus R2CHOP for PFS (unadjusted HR 1.07 [95% CI 0.64-1.80]; multivariable HR 1.13 [0.65-1.96]; matched HR 1.29 [0.70-2.37]; IPTW HR 1.12 [0.65-1.93]) or OS (unadjusted HR 0.73 [0.40-1.33]; multivariable HR 0.82 [0.42-1.58]; matched HR 1.08 [0.53-2.19]; IPTW HR 0.77 [0.40-1.46]; all p=ns).
A subgroup analysis restricted to MYC/BCL2 DH patients yielded similar results.
A subgroup analysis restricted to patients who achieved CMR at end of DA-EPOCH-R (n=63) or R2CHOP (n=36), demonstrated no significant difference for PFS (unadjusted HR 0.64 [95% CI 0.26-1.59]; multivariable HR 0.69 [0.25-1.95]; matched HR 0.93 [0.31-2.81]; IPTW HR 0.95 [0.37-2.49]) and OS (unadjusted HR 0.33 [CI 0.09-1.14]; multivariable HR 0.43 [0.10-1.91]; matched HR 0.60 [CI 0.14-2.61]; IPTW HR 0.53 [0.15-1.92]; all p=ns).
Grade ≥2 toxicity occurred in 76/96 (78%) patients treated with DA-EPOCH-R (gr2 16%, gr3 22%, gr4 34%, gr5 7%) and 46/51 (90%) patients treated with R2CHOP (gr2 29%, gr3 41%, gr4 18%, gr5 2%) and Occurrence of gr4-5 toxicity was significantly different (41% vs 20% resp, p=0.01).
Conclusion
R2CHOP was equally effective as DA-EPOCH-R +/- nivolumab in patients with DH/TH HGBL across four statistical methods, with half the incidence of grade 4–5 toxicity. These data support R2CHOP as a valuable treatment option for patients with DH/TH HGBL, given its efficacy, favorable toxicity profile, and convenient outpatient administration requiring minimal hospital visits, particularly in settings where clinical trials evaluating novel therapies are unavailable.
DISCLOSURES
A. Vera De Jonge, MD, PhD: No relevant conflicts of interest to declare.