PET-Based Quantification of Baseline Metabolic Tumor Burden Improves Risk Stratification in High-Risk Hodgkin Lymphoma: A Children's Oncology Group Study

Abstract

<h3>Purpose/Objective(s)</h3> COG AHOD0831 was a multi-center, prospective study that used a response-adapted approach for patients < 22-years-old with high-risk Hodgkin lymphoma. After 2 cycles of chemotherapy, response evaluation was performed by ¹⁸F-fluorodeoxyglucose (FDG) positron-emission tomography (PET) scan. Rapid early responders (RERs) received less intensive therapy than slow early responders (SERs). In this study, both 1^st & 2^nd event-free survival (EFS) rates were below the pre-specified targets. We explored if measures of baseline tumor burden improve risk stratification. <h3>Materials/Methods</h3> Patients from AHOD0831 were identified who had baseline PET scans that were amenable to quantitative analyses. For each patient, metabolic tumor volume (MTV), total lesion glycolysis (TLG), maximum standardized uptake value (SUVmax), and peak SUV (SUVpeak) were obtained for mediastinal (m) and total body (t) disease. MTV was defined based on thresholds of SUV > 2.5 or > 40% SUVmax. TLG was defined as MTV * tumor SUVmean. EFS was assessed by Kaplan-Meier analyses. 2^nd EFS was defined as the time to a second event, reflecting rates of successful salvage after a 1^st relapse. Receiver operating characteristic analyses estimated the ability of PET parameters to predict EFS; optimized cutoff values were identified using a Youden index. <h3>Results</h3> Of 166 patients enrolled on AHOD0831, 94 (57%) had PET scans evaluable for quantitative analysis. For this subset: median age 15.5 years; 62% male; 67% white; 45% stage III, 55% stage IV; 86% bulk; 60% nodular sclerosis histology; 54% RER, 46% SER. These characteristics did not differ significantly from the complete AHOD0831 cohort. At a median follow-up of 49 months, 4-year EFS was 76% for the complete cohort (76% RERs, 75% SERs). Patients with high tMTVs and tTLGs, based on each threshold, were significantly more likely to be SERs (all <i>P</i> < 0.02). tTLG_2.5 was significantly associated with EFS, with an optimal cutoff value of 1841 to distinguish low- and high-risk groups (<i>P</i> = 0.04). tTLG_2.5 was significantly associated with EFS in RERs (<i>P</i> = 0.01) but not in SERs (<i>P</i> = 0.8). Patients were stratified by RER vs. SER and high vs. low tTLG_2.5. The best EFS was associated with RER & low tTLG_2.5, the worst EFS with RER & high tTLG_2.5, and intermediate EFS with SER regardless of tTLG_2.5 (4-year EFS: 92%, n = 25; 60%, n = 26; 79%, n = 14; 74%, n = 29, respectively). EFS was superior in patients with tTLG_2.5 ≤1841 & RER when compared to the rest of the cohort (94% vs. 71%, <i>P</i> = 0.02). For this favorable subgroup, 2^nd EFS was 100%. <h3>Conclusion</h3> RERs with a low baseline metabolic tumor burden experienced excellent EFS with less intensive therapy. Conversely, RERs with a high baseline tumor burden experienced poor EFS that was even worse than that of SERs. Thus, patients with a high metabolic tumor burden upfront may benefit from intensified therapy, even if they achieve a RER. PET-based measures of initial disease burden may contribute to risk-based treatment algorithms and improve outcomes in HL.