Objective: The predictive value of interim PET/computed tomography (I-PET/CT) in diffuse large B-cell lymphoma (DLBCL) is controversial. Our aim was to evaluate the predictive value of I-PET/CT for an event-free survival.
Patients and methods: We analyzed patients with DLBCL included in a prospective clinical trial who were treated with six cycles of dose-dense R-CHOP followed by pegfilgrastim and who had undergone an I-PET/CT (after two cycles) and a final PET [F-PET/CT (60 days after the sixth cycle)]. Event was defined as nonresponse, relapse, or death.
Results: A total of 69 patients were included. Their median age was 60 years; 54% were male, 25% had bulky disease, and 67% had an International Prognostic Index of 0–2. The median follow-up duration was 28.8 months. I-PET/CT was positive in 34 (49%) patients and F-PET/CT was positive in 12 (17.4%). The 3-year event-free survival was 86% for patients who were I-PET/CT negative as against 64% for those who were I-PET/CT positive (P=0.036). The negative and positive predictive values, sensitivity, and specificity of I-PET/CT for an event were 83, 32, 65, and 56%, respectively. In a multivariate analysis including baseline characteristics, I-PET/CT, and F-PET/CT, F-PET/CT was the only significant predictor (P<0.0005).
Conclusion: In patients with DLBCL treated with dose-dense R-CHOP plus pegfilgrastim, a negative I-PET/CT is highly predictive of a favorable outcome and a positive I-PET/CT is of limited clinical value. These results do not support treatment intensification after a short course of chemotherapy based solely on a positive I-PET/CT.