Diagnostic accuracy of FDG PET/CT for clinical evaluation at the end of treatment of HL and NHL: a comparison of the Deauville Criteria (DC) and the International Harmonization Project Criteria (IHPC)

To evaluate the accuracy and prognostic value of FDG PET/CT for response assessment after treatment in patients with Hodgkin lymphoma (HL) and non-Hodgkin lymphoma (NHL) when using the Deauville Criteria (DC) and the International Harmonization Project Criteria (IHPC). This retrospective study included 101 patients (35 HL, 66 NHL) who underwent early restaging FDG PET/CT after treatment. Scans were evaluated using the IHPC and DC. Two thresholds of positivity for the DC were used: a score of at least 3 (DC3, i.e. scores 3 - 5) and a score of at least 4 (DC4, i.e. a score of 4 or 5). Accuracy was assessed using conventional diagnostic procedures, multidisciplinary team case notes, further PET/CT scans and/or follow-up. Progression-free survival and overall survival were computed using the Kaplan-Meier method. The Cox proportional hazards model was used to identify predictors of outcome. Sensitivity, specificity, positive predictive value, negative predictive value and accuracy of FDG PET/CT for early restaging were, respectively, 92 %, 87 %, 74 %, 92 % and 86 % using DC4, 97 %, 76 %, 64 %, 98 % and 84 % using DC3, and 97 %, 67 %, 57 %, 98 % and 76 % using the IHPC. FDG PET/CT positivity was associated with a worse cumulative survival rate over a 2-year period when using DC4 in comparison with the IHPC (20 % vs. 49 %; p < 0.05) and DC3 (47 %; p < 0.05). Cox regression analysis showed different risks of progression in patients positive on FDG PET/CT using the IHPC, DC3 and DC4 (hazard ratios 1.57, 0.7 and 3.2, respectively). FDG PET/CT using DC4 showed higher diagnostic accuracy for HL and NHL than FDG PET/CT using either the IHPC or DC3, indicating its value in predicting clinical outcome after treatment.