Interim PET after 4 cycles predicts outcome in histomolecularly confirmed primary mediastinal B-cell lymphoma

Abstract

The GAINED study was a randomized phase 3 trial comparing obinutuzumab (G) with rituximab (R) plus ACVBP (doxorubicin, cyclophosphamide, and prednisone, combined with either vindesine or bleomycin) or CHOP14 (cyclophosphamide, doxorubicin, vincristine, and prednisone, administered on a 14-day schedule) induction, followed by positron emission tomography (PET)-guided consolidation. This post hoc analysis aimed to detail the outcomes of patients with primary mediastinal B-cell lymphoma (PMBL), verified through expert pathological review and the use of gene expression profiling (GEP) and next-generation sequencing. Of 620 centrally reviewed patients, 138 (22.3%) confirmed PMBL cases were analyzed. Baseline characteristics included a median age of 33.5 years, 63.8% female, 55.1% stage III to IV, 90.6% elevated lactate dehydrogenase, 87.6% Eastern Cooperative Oncology Group performance status score of 0 to 1, 62.3% extranodal involvement, 52.6% age-adjusted International Prognostic Index (aaIPI) of 2% to 3%, and 53.6% bulk (>10 cm). Induction regimens were R/G-CHOP14 (56.9%) and R/G-ACVBP (43.1%). Postinduction treatments, based on interim PET results, included: standard consolidation chemotherapy (59.8%) if change in maximum standardized uptake value (ΔSUVmax) of >66% after cycle 2 and >70% after cycle 4 (PET2-/4-), intensive treatment and autologous transplantation (26.8%) if PET2+/4-, and salvage therapy (13.4%) if PET4+ (ΔSUVmax of ≤70%). Among patients with GEP data (n = 107), 38 (35.5%) were PDL1high/PDL2high. Key somatic mutations data (n = 87) included SOCS1 (70.1%), B2M (56.3%), STAT6 (49.4%), TNFAIP3 (47.1%), GNA13 (39.1%), CIITA (37.9%), CD58 (36.8%), and TP53 (29.9%). After a median follow-up of 39.5 months, 2-year progression-free survival (PFS) and overall survival (OS) rates were 86.2% and 93.2%, respectively. In a multivariate model including bulk, aaIPI, and ΔSUVmax PET2/PET4, only bulk and ΔSUVmax PET4 of ≤70% were associated with shorter PFS (hazard ratio, 4.39 [95% confidence interval (CI), 1.28-15.11] and 4.95 [95% CI, 1.71-14.3], respectively), whereas none were associated with OS. The ΔSUVmax-based interim PET4 response emerged as the strongest predictor of patient outcomes in this selected clinical trial population. This trial was registered at www.ClinicalTrials.gov as #NCT01659099.

Graphical abstract

Figure 1.

Population selection from the GAINED cohort.

Figure 2.

Oncoprint showing the distribution of genomic alterations in patients with PMBL from the GAINED cohort with available targeted NGS (n = 87), clustered based on the PDL1/PDL2 gene expression status (GEP). cfDNA, baseline cfDNA level; Mut, Mutation; NA, Not available; sPDL1, baseline sPDL1 level; Synon, Synonymous mutation; Trunc, Truncating mutation.

Figure 3.

Treatments received according to iPET responses. ∗One patient died before treatment start. ∗∗In patients with missing PET2 and/or PET4 data, consolidation treatments data were also missing. ∗∗∗One patient experienced treatment discontinuation after 3 cycles of R-ACVBP (toxicity). Switch for R-CHOP (n = 2 cycles) then etoposide-ifosfamide (2 cycles).

Figure 4.

Progression-Free Survival and Overall Survival Curves (Kaplan-Meier Analysis). (A) PFS and (B) OS in the PMBL set.

Figure 5.

Progression-Free Survival and Overall Survival According to Interim PET2 and PET4 Response (Kaplan-Meier Analysis). (A) PFS and (B) OS according to interim PET2 and PET4 response in the PMBL set.