Baseline PET-Derived Metabolic Tumor Volume Metrics Predict Progression-Free and Overall Survival in DLBCL after First-Line Treatment: Results from the Phase 3 GOYA Study

Abstract

Abstract Introduction: Quantitative 18fluorodeoxyglucose positron emission tomography (PET)/computed tomography assessment using total metabolic tumor volume (TMTV) and tumor lesion glycolysis (TLG) measurements has been found promising as an objective method to predict survival in diffuse large B-cell lymphoma (DLBCL) patients (pts). However, the methodology for PET-derived metrics is still evolving, and their predictive value is yet to be proven in large-scale, prospective, multicenter studies. We investigated the prognostic value of baseline maximum standardized uptake value (SUVmax), TMTV and TLG for progression-free survival (PFS) in a large pt cohort treated with obinutuzumab (GA101; G) or rituximab (R) combined with cyclophosphamide, doxorubicin, vincristine, and prednisone (CHOP) in the Phase 3 GOYA study (NCT01287741; Vitolo et al. J Clin Oncol 2017). Methods: Pts aged ≥18 years, with previously untreated, CD20-positive DLBCL and an International Prognostic Index (IPI) score ≥2 and low-risk pts with IPI scores of 1 (not due to age alone) or 0 (with bulky disease) were randomized 1:1 to receive 8 x 21-day cycles of G (1000mg intravenous [IV] on Days [D] 1, 8, and 15 of Cycle [C] 1 and D1, C2-8) or R (375mg/m2 IV on D1, C1-8) plus 6 or 8 cycles of CHOP. All pts had a baseline and end of treatment (EOT) PET. PET images were segmented using an automated workflow program in MIM software, applying thresholds of 1.5 x liver background and a minimum volume of 1mL to the whole body PET images. The data were analyzed for the overall population and according to germinal center B-cell-like (GCB), unclassified, and activated B-cell-like (ABC) subtypes of DLBCL. TMTV, TLG, and SUVmax were split into 4 categories/levels according to the following quartiles: Q1, <25%; Q2, 25-50%; Q3, 50-75%; and Q4, 75-100%, which were obtained based on their distribution in the available population. The reported hazard ratios (HRs) refer to stratified log-rank tests comparing Q2, Q3, and Q4 to Q1, adjusted for stratification factors of the study: IPI score (low [0-2], intermediate [3], and high [4-5]) and number of planned CHOP cycles (6 or 8). Results: Of 1418 enrolled pts, 1346 had a baseline PET scan and 1334 had detectable lesions. There was no statistical difference in PFS between the treatment arms (G vs R), thus the entire cohort was analyzed as a whole. Results of the predictive value of baseline TMTV for PFS are presented in quartiles in Figure 1, and results of the predictive value of TLG for PFS are presented in quartiles in Figure 2, for the overall PET intent-to-treat population. After a median follow-up of 29 months TMTV and TLG were highly predictive of PFS when comparing Q4 vs Q1: HR=2.21, 95% CI 1.48-3.29, p<0.0001, and HR=1.91, 95% CI 1.28-2.85, p=0.0005, respectively. TMTV was also predictive of overall survival (OS): HR=2.63, 95% CI 1.55-4.46; p<0.0001. However, SUVmax-based prediction of PFS was not statistically significant (HR=0.84, 95% CI 0.57-1.23, p=0.3782). Three-year PFS for pts in TMTV Q1, 2, 3 and 4 was 86% (95% CI 81-89%), 84% (95% CI 78-88%), 78% (95% CI 72-83%) and 66% (95% CI 59-71%), respectively. TMTV also showed a trend for a better prediction of PFS (Figure 3) and OS in pts with the unclassified and ABC DLBCL subtypes when compared with those with the GCB subtype. Conclusions: This large prospective study confirms baseline TMTV and TLG as predictors of PFS and OS in DLBCL after first-line immunochemotherapy, while SUVmax may not be a predictor. Furthermore, TMTV and TLG appear to be better predictors of survival for pts with the unclassified and ABC subtypes of DLBCL than for those pts with the GCB subtype. Further analyses are underway comparing these results with the predictive value of percentage change from baseline to EOT PET, Deauville score-based analysis of EOT PET, and the various molecular DLBCL subtypes. Figure 1 Figure 1. Disclosures Kostakoglu: Roche: Consultancy, Other: GOYA is sponsored by F. Hoffmann-La Roche Ltd. Third-party editorial support, under the direction of Lale Kostakoglu and Denis Sahin, was provided by Helen Cathro of Gardiner-Caldwell Communications, and was funded by F. Hoffmann-La Roche Ltd. Sehn: Celgene: Consultancy, Honoraria; Seattle Genetics: Consultancy, Honoraria; Janssen: Consultancy, Honoraria; Abbvie: Consultancy, Honoraria; Amgen: Consultancy, Honoraria; Roche/Genentech: Consultancy, Honoraria. Chua: Lundbeck: Honoraria; Roche: Honoraria; Seattle Genetics: Honoraria; Gilead Sciences: Honoraria; Merck: Honoraria. Gonzalez-Barca: Gilead: Consultancy; Sandot: Consultancy; Janssen: Speakers Bureau; Roche: Speakers Bureau. Pinto: Millenium Takeda: Research Funding; Gilead: Honoraria; Roche: Honoraria; Bristol Myers Squibb: Honoraria; Merck Sharp Dome: Honoraria; Celgene: Honoraria; Helssin: Honoraria; Mundipharma EDO: Speakers Bureau. Fingerle-Rowson: F. Hoffmann-La Roche Ltd: Employment, Equity Ownership. Knapp: Roche: Employment. Mattiello: Roche: Employment. Nielsen: F. Hoffmann-La Roche Ltd: Employment, Equity Ownership. Sellam: Roche: Employment. Sahin: Roche: Employment, Equity Ownership. Vitolo: Janssen: Honoraria, Membership on an entity's Board of Directors or advisory committees; Gilead: Honoraria; Mundipharma: Honoraria; Takeda: Honoraria; Celgene: Honoraria, Membership on an entity's Board of Directors or advisory committees; Roche: Honoraria, Membership on an entity's Board of Directors or advisory committees, Research Funding. Trněný: Janssen: Consultancy, Honoraria; Abbvie: Consultancy, Honoraria; Gilead: Consultancy, Honoraria; BMS: Consultancy, Honoraria; Takeda: Consultancy, Honoraria; Roche: Consultancy, Honoraria, Research Funding; Celgene: Consultancy, Honoraria, Research Funding.