| Purpose: This study aimed to explore the prognostic value of interim and final 18F-FDG PET/CT in diffuse large B cell lymphoma(DLBCL) using the 5-PS and ?SUVmax.Methods: A total of 69 patients with diagnosed DLBCL were evaluated in this confirmatory study. All patients underwent 18F-FDG PET/CT at least twice.Scans were interpreted by two observers using the 5-PS and ?SUVmax. Explanation were evaluated for overall survival(OS) and progression-free survival(PFS) prediction. ROC were determined to assess the optimal cutting-off value of ?SUVmax. Kaplan-Meier method and the Log-rank test were used to evaluate and contrast survival rate. Clinical relevant factors were assessed by COX regression analysis of proportional hazards. The Wald?2 test presented relative risk(RR).Results: All the 69 DLBCL patients made baseline PET/CT examination, 51 patients underwent PET/CT after 2-4 cycles and 44 patients after 8 cycles. 1. Interim 18F-FDG PET/CT was predictive to the outcome of using 5- PS and ? SUVmax. The optimal cutting-off for predicting the progression or the death was 78.9% for ?SUVmax.For 5-PS, the 5-year PFS was 51.9% in the PET-positive patients(score>3) vs.82.9% in the PET-negative patients(score?3)(P=0.030); the 5-year OS was 61.4% in the PET-positive patients(score>3) vs.87.2% in the PET-negative patients(score?3)(P=0.030). The 5-year PFS of patients with ?SUVmax>78.9% and ?78.9% was found to be 87.5% and 40.0%(P<0.001). The 5-year OS of patients with ?SUVmax >78.9% and ?78.9% was found to be 89.3% and 57.1%(P=0.003). 2. The positive predictive value(PPV), negative predictive value(NPV) and accuracy of the prediction PFS using 5-PS were 45.5%, 85.0% and 76.4%. And PPV, NPV and accuracy of the OS were 36.4%, 90.0% and 78.4%. The PPV, NPV and accuracy of the prediction PFS by ?SUVmax were 58.3%, 89.7% and 82.4%. The PPV, NPV and accuracy of the OS were 41.7%, 92.3% and 80.4%. 3. A total of 42 DLBCL patients after 2-4 cycles underwent PET/CT for estimating survival. The 5-year PFS in the PET-positive and the negative group of patients were 53.3% and 90.2%(P=0.009).The 5-year estimate for OS was 92.4% in the PET-negative patients compared to 64.8% in the PET-positive patients(P=0.009).40 patients with DLBCL after 4-6 cycles evaluated the response. The 5-year PFS was 81.4% in the PET-negative patients compared to 31.3% in the PET-positive patients(P=0.004). The 5-year OS in the PET-positive and the negative group of patients were 45.0% and 83.5%(P=0.004). 4. COX regression analysis of proportional hazards showed ?SUVmax could be used to predict OS(RR = 16.835, P = 0.020).?SUVmax and age were the main factors influencing PFS(RR = 16.632, P = 0.015; RR = 6.104, P = 0.009). 5. 44 patients with DLBCL were used to assess the prognostic value of the final PET/CT after 8 cycles of chemotherapy. The 5-year PFS were 40.0% in the PET-positive patients(score>3) and 89.5% in the PET-negative patients(score?3)(P<0.050); the 5-year OS were 40.0 % in the PET-positive patients(score>3) and 92.0 % in the PET-negative patients(score?3)(P < 0.050). In other words, ?SUVmax analysis(> 56.5% vs ?56.5%) identified patients with significantly different 5-year PFS(89.7% vs 25.0%; P<0.001) and OS(92.2% vs 25.0%; P<0001).Conclusion: 1. 5- PS and ?SUVmax are valid for assessing the prognostic value of interim and final PET/CT in DLBCL. 2. The prediction, after 4-6 cycles, leads to a better performance than the 2-4 treatment.3. For OS,?SUVmax may have predictive relevance. Both age and ?SUVmax are important factors affecting PFS. |
PDF Full Text Request