We investigated the prognostic value of interim 18F-FDG PET/CT (PET-2) in pediatric Hodgkin lymphoma (pHL), evaluating both visual and semiquantitative analysis.
Thirty pHL patients (age ≤16) underwent serial 18F-FDG PET/CT: at baseline (PET-0), after 2 cycles of chemotherapy (PET-2) and at the end of first-line chemotherapy (PET-T). PET response assessment was carried out visually according to the Deauville Score (DS), as well as semiquantitatively by using the semiquantitative parameters reduction from PET-0 to PET-2 (ΔΣSUVmax0–2, ΔΣSUVmean0–2). Final clinical response assessment (outcome) at the end of first-line chemotherapy was the criterion standard, considering patients as responders (R) or nonresponders (NR). Disease status was followed identifying patients with absence or relapsed/progression disease (mean follow-up: 24 months, range 3–78).
Sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and accuracy of visual and semiquantitative assessment were calculated; furthermore, Fisher exact test was performed to evaluate the association between both visual and semiquantitative assessment and outcome at the end of the first-line chemotherapy. The prognostic capability of PET-2 semiquantitative parameters was calculated by ROC analysis and expressed as area under curve (AUC). Finally, progression-free survival (PFS) was analyzed according to PET-2 results based on the 5-point scale and semiquantitative criteria, using the Kaplan–Meier method.
Based on the outcome at the end of first-line chemotherapy, 5 of 30 patients were NR, the remnant 25 of 30 were R. Sensitivity, specificity, PPV, NPV, and accuracy of visual analysis were 60%,72%,30%,90%,70%; conversely, sensitivity, specificity, PPV, NPV, and accuracy of semiquantitative assessment were 80%, 92%, 66.7%, 95.8%, 90%. The highest AUC resulted for ΔΣSUVmax0–2 (0.836; cut-off <12.5; sensitivity 80%; specificity 91%). The association between ΔΣSUVmax0–2 and outcome at the end of first-line chemotherapy resulted to have a strong statistical significance (P = 0.0026). Both methods demonstrated to influence PFS, even if the semiquantitative assessment allowed a more accurate identification of patients with a high risk of treatment failure (P = 0.005).
Our preliminary results showed that PET-2 visual assessment, by using Deauville criteria, can be improved by using the semiquantitative analysis. The SUV max reduction (ΔΣSUVmax0–2) evaluation might provide a support for the interpretation of intermediate scores, predicting with good confidence those patients who will have a poor outcome and require alternative therapies.