Page 213 - 18F-FDG PET as biomarker in aggressive lymphoma; technical and clinical validation

Optimal timing and response criteria of Interim-PET in DLBCL
Introduction
Diffuse large B-cell lymphoma (DLBCL) is characterized by an aggressive clinical course. Standard first-line therapy consists of rituximab plus cyclophosphamide, doxorubicin, vincristine, and prednisone (R-CHOP). Up to one-third of patients relapse or fail to achieve complete remission. These patients have a poor prognosis and low response rates to salvage treatment [1,2]. Early identification of patients with poor prognosis is an important step toward testing alternative treatment options. Interim 18F-fluorodeoxyglucose positron emission tomography/computed tomography (18F-FDG PET/CT) can be used to differentiate good and poor responders during treatment to modify therapy and improve outcome for poor responders and de-escalate treatment for good responders [3].
Current tools for predicting outcome in DLBCL, such as the International Prognostic Index (IPI) [4], which captures pretreatment clinicopathologic features, have limited precision. Many studies have investigated the potential of metabolic imaging with 18F-FDG PET in the context of treatment evaluation using end-of-treatment PET/CT scans [5,6] or of prediction of therapy success using on-treatment (interim 18F-FDG PET [I-PET]) evaluation. End-of-treatment PET is the current clinical standard, but the impact of I-PET is less clear. A recent systematic review and meta-analysis concluded that I-PET has predictive value in DLBCL patients, but small sample sizes, use of different response criteria, different timings, and other methodologic variations among studies hamper the ability to draw firm conclusions [3].
Analyzing the individual patient data (IPD) from various studies made it possible to re-analyze clinical data and 18F-FDG PET scans, which reduced variability and thus allows for a statistically more robust analysis of prediction or prognosis, subgroup analyses, and identification of potential effect modifiers. To this end, we established the PETRA database (www.petralymphoma.org), for collecting individual patient data and PET/CT scans from high quality international clinical studies. The aim of this IPD meta-analysis was to determine the optimal timing and PET response criteria for I-PET in DLBCL.
211
8