Positron Emission Tomography guided therapy of aggressive lympomas (PETAL) - the predictive value of quantitative PET

Objectives The PETAL study (EudraCT 2006-001614-33) is a German multicentric randomized treatment optimization trial for patients with aggressive non-Hodgkin’s lymphoma, investigating treatment intensification in high-risk patients as stratified by FDG-PET.

Methods At first diagnosis of aggressive non-Hodgkin’s lymphoma (NHL) patients received a baseline FDG-PET and after 2 cycles of R-CHOP therapy an interim FDG-PET. Risk stratification was based on the decrease of the maximum SUV in lymphoma (ΔSUV), i.e., ΔSUV>66% responders, ≤66% non-responders. Responders continued with 4 cycles R-CHOP, non-responders were randomized between 6 further cycles of R-CHOP vs. 6 blocks of the B-ALL protocol. Patients from 57 oncology centers were scanned in 23 PET facilities (intent-to-treat population: 851 patients). The calibration of all PET scanners was centrally validated and patient preparation, scanning conditions, data acquisition, and processing were standardized. The predictive value of ΔSUV for time to treatment failure (TTF) and overall survival (OS) was assessed by survival ROC analysis at 3 years.

Results Survival was significantly higher for responders vs. non-responders (p<0.001), TTF/OS after 3 years were 73/87% vs. 43/62%. No significant difference (p>0.1) in TTF or OS was found between R-CHOP and B-ALL. Since the equivalent therapeutic efficacy of 6 and 8 cycles of R-CHOP is known from the literature, all arms can be analyzed together to assess the predictive value of FDG-PET. The area under the ROC curve for the prediction of TTF by ΔSUV was 0.66 and 0.63 for OS.

Conclusions The predictive value of ΔSUV for risk stratification was confirmed in a large prospective randomized controlled trial. Under standardized imaging conditions quantitative FDG-PET is suitable for use in the field, e.g. for future NHL trials.

Research Support Deutsche Krebshilfe 107592 and 110515