Bone Marrow Biopsy Impacts Response Assessment in a Minority of Patients with Follicular Lymphoma and Diffuse Large B-Cell Lymphoma Treated with Immunochemotherapy: Results from the Randomized Phase III GALLIUM and GOYA Trials

Introduction: Clinical trial response assessments for follicular lymphoma (FL) and diffuse large B-cell lymphoma (DLBCL) typically mandate bone marrow biopsies (BMBs) at baseline and for confirmation of complete response (CR) as assessed by computed tomography (CT) imaging according to International Working Group (IWG) criteria. BMBs are painful and expensive, and can deter patients (pts) from clinical trial participation. Rutherford et al. (Br J Haematol 2017) suggested that confirmatory BMBs do not impact response assessment in the majority of FL pts treated in clinical trials. We describe the impact of confirmatory BMBs on response in the GALLIUM study (NCT01332968), which randomized previously untreated FL pts to obinutuzumab (G)- or rituximab (R)-chemotherapy (CHOP, CVP, or bendamustine) followed by maintenance with the same antibody in responders, and in the GOYA study (NCT01287741), which randomized previously untreated DLBCL pts to G-CHOP or R-CHOP.

Methods: We conducted a retrospective analysis of GALLIUM and GOYA to evaluate the percentage of pts with CR as assessed by CT but with a positive BMB at end of induction (EOI), and the percentage of pts with complete metabolic response (CMR) on positron emission tomography (PET) by Lugano 2014 criteria but with a positive BMB at EOI. Response by PET was an exploratory endpoint and not mandatory.

Results: Of 1202 randomized FL pts in GALLIUM, 633 (52.7%) had a positive (613 pts, 51.0%) or indeterminate (20 pts, 1.7%) baseline BMB (data missing for 12 pts, 1.0%). Bone marrow involvement was not prognostic for progression-free survival (PFS; investigator (INV)-assessed PFS: HR 0.99 [95% CI 0.80, 1.23] for pts with positive vs non-positive BMB). At EOI, 209/633 (33%) pts had a CR on CT by IWG criteria, 179 of whom had a follow-up BMB to confirm response; 174/179 (97.2%) had a negative BMB that confirmed CR, 1/179 (0.6%) was positive, and 4/179 (2.2%) were indeterminate. Overall, only 5/179 (2.8%) pts with a positive/indeterminate baseline BMB and EOI CR by CT, or 5/1202 (0.4%) pts enrolled in the trial, had a repeat BMB result that was relevant for response assessment. Lugano 2014 criteria (determined by Independent Review Committee; IRC) were used for response assessment in 282/633 (44.5%) pts with positive/indeterminate baseline BMBs and EOI PET scans available; 251/282 (89.0%) pts had a CMR, and 213/251 (84.9%) underwent confirmatory BMBs at EOI (203 [95.3%] negative, 5 [2.3%] positive, and 5 [2.3%] indeterminate). Thus, BMB results only affected response by Lugano 2014 in 4.7% of pts having a repeat BMB (0.8% of all enrolled pts).

Of 1418 randomized DLBCL pts in GOYA, 167 (11.8%) had a positive (154 pts, 10.9%) or indeterminate (13 pts, 0.9%) baseline BMB; information was missing for 14 pts (1.0%). Bone marrow involvement was prognostic for PFS (INV-assessed PFS: HR 0.75 [95% CI: 0.56, 0.99]; 3-year PFS rate 55.3% vs 69.8% for pts with positive vs non-positive BMBs, respectively). At EOI, 69/167 (41.3%) pts had a CR on CT by IWG criteria, 49 of whom had a follow-up BMB to confirm response; 47/49 (95.9%) had a negative BMB and 2/49 (4.0%) were positive. Therefore, only 2/49 (4.0%) pts with a positive/indeterminate baseline BMB and EOI CR by CT, or 2/1418 (0.1%) pts enrolled in the GOYA trial, had a repeat BMB result that was relevant for response assessment. Lugano 2014 criteria (by IRC) were used for response assessment in 121/167 (72.5%) pts with positive/indeterminate baseline BMBs and EOI PET scans available; 96/121 (79.3%) pts had a CMR, and 70/96 (72.9%) underwent confirmatory BMB. Of these, 65/70 (92.9%) were negative. Thus, BMB results only affected response by Lugano 2014 in 7.1% of pts having a repeat BMB (0.4% of all enrolled pts).

Conclusions: In previously untreated FL and DLBCL pts in the GALLIUM and GOYA studies, post-induction BMB results impacted the CR rate by IWG criteria in only a minority of pts (<5% of pts with an initial positive/indeterminate BMB and EOI CR by CT who underwent repeat BMB and 0.1-0.4% of all enrolled pts). There was a trend toward similar findings when Lugano 2014 criteria were used, but conclusions are limited because only a subset of pts underwent PET imaging. In design of future FL trials, consideration should be given to eliminating the requirement for BMB to confirm CR. With increasing use of Lugano criteria, particularly in DLBCL, investigations should continue to evaluate the impact of BMBs when PET scans are used to assess response.