Risk Assessment in Diffuse Large B-Cell Lymphoma by Combining Baseline Metabolic Tumor Volume and Peking Criteria When Evaluating Series 18F-Fluorodeoxyglucose Positron Emission Tomography Scans

This study aimed to determine the predictive and prognostic value of baseline metabolic tumor volume (MTV) and the Peking criteria from serial positron emission tomography (PET) scans in diffuse large B-cell lymphoma, including 300 newly diagnosed patients who were prospectively treated with 2-4 cycles of standard first-line treatment (clinicaltrials.gov identifier: NCT02928861). PET/computed tomography (CT) examinations were performed at baseline, after two (PET-2) or four cycles (PET-4). PET during the interim was evaluated using Deauville 5-point scales (5-PS), Delta SUVmax criteria, and the Peking criteria which interpreted based on the maximum standard uptake of the liver (SUVmax-liver). Peking criteria had better accuracy, positive predictive value (PPV), and specificity than other two methods. The MTV and Peking criteria both significantly predicted progression-free survival (PFS) and overall survival (OS). An MTV > 191 cm(2) and Peking criteria of PET-2 and PET-4 > 1.6-fold SUVmax-liver was used as the cutoff for a positive result. PET-4 achieved higher accuracy, PPV, and specificity for 2-year PFS (83.3%, 86.7%, and 98.4%, respectively) and OS (92.6%, 73.3%, and 97.2%, respectively) than PET-2. Various prognostic models containing different risk factors were established via Cox regression analysis. The MTV and PET-2/PET-4 results were used to categorized patients into low-risk, intermediate-risk, and high-risk prognostic groups (with 0, 1, and 2 risk factors, respectively) (P < 0.0001). High burden MTV and positive PET-2 and PET-4 (>1.6-fold SUVmax-liver) could identify high-risk patients with 2-year PFS and OS of 0.0% and 26.3% (95% confidence interval [CI]: N/A to 54.3%). When PET-2 and PET-4 were evaluated by 5-PS, the 2-year PFS and OS from high risk patients of three-parameters model achieved 31.4% (95%CI: 6.9%-55.9%) and 42.7% (95%CI: 14.6%-70.7%). In conclusion, combining baseline MTV and any regular response on PET/CT evaluated using the Peking criteria can improve prognostic value. Serial PET/CT from baseline MTV to PET-4 may have relatively greater predictive power for poor prognosis in diffuse large B-cell lymphoma.