Predictive Value of Interim 18f-FDG PET-CT Scans on Diffuse Large B-Cell Lymphoma Treated with R-CHOP: A Prospective Study

¹⁸F-PET-CT is clinically recommended for monitoring therapeutic response in DLBCL patients. But the role of interim PET-CT remains controversial, and most of the previous researches were retrospective. We designed this study to prospectively evaluate whether interim PET-CT was a valid prognostic tool for patients with DLBCL treated with R-CHOP regimen and if yes, try to determine the more appropriate time and interpretation method for interim PET-CT.

This study was a sub-study of the parental study "A prospective, multicenter randomized phase III clinical trial of intensified chemotherapy in improving the treatment efficacy of patient with diffuse large B-cell lymphoma" (NCT01793844). The sub-study included patients that have already been enrolled in the parent study at Sun Yat-Sen University Cancer Center prospectively. Patients were evaluated with PET-CT scans before treatment and after every 2 cycles of R-CHOP and after the completion of first-line treatment. Regular follow-up starts from the enrollment.

Between Jan. 2008 and Aug. 2014, 221 patients in Sun Yat-Sen University Cancer Center were enrolled in this sub-study, among whom 203 patients were included in the analysis and the other 18 were excluded for lacking the necessary raw data of PET-CT scan. PET evaluation would be applying the visual criteria of International Harmonization Project(IHP) and the Deauville 5-point scale(5-PS). The results showed PET positive in 103 patients and negative in 100 patients after 2 cycles of R-CHOP chemotherapy with IHP criteria. Among the 103 patients with positive PET-2, 53 patients turned negative after 4 cycles of chemotherapy and still 50 patients remain positive. At the evaluation of end-of-first-line-treatment, 165 patients achieved CR, while 30 achieved PR and 8 PD . According to 5-PS criteria, 146 patients were PET negative with 57 were positive after 2 cycles of chemotherapy. And 173 patients were negative in PET-4 evaluation, while 30 patients remained positive. With a median follow-up of 25.46 months (range 3.60~77.33 months) and according to IHP criteria, patients with negative PET-2 had superior 3-year PFS (84.7% vs. 63.8%, p < 0.001) and OS (89.8% vs. 80.4%, p = 0.045) than those with positive results. Patients with negative PET-4 also had a better clinical outcome compared with the positive group with 3-year PFS (84.1% vs. 43.8%, p < 0.001) and OS (90.7% vs. 67.8%, p < 0.001). A further analysis showed that patients who achieved PET negative just after 2 cycles of chemotherapy (Early responder, ER) had a similar prognosis comparing with those who achieved PET negative after 4 cycles (Later responder, LR). There were no significant differences in the persistent CR rates (87.00% vs. 86.79%, p = 0.971), 3-year PFS (84.7% vs. 82.2%, p = 0.867) and 3-year OS (89.8% vs. 92.9%, p = 0.638) between the ER group and the LR group. Patients who remained PET positive after 4 cycles of chemotherapy (Interim non-responder, I-NR) had the worst prognosis. Their persistent CR rate (42.00% vs. 87.00%, p < 0.001; 42.00% vs. 86.79%, p < 0.001), 3-year PFS (43.8% vs. 84.7%, p < 0.001；43.8% vs. 82.2%, p < 0.001) and 3-year OS (67.8% vs. 89.8%, p < 0.001；67.8% vs. 92.9%, p = 0.002) were significantly lower, comparing with the ER and LR group. And according to 5-PS criteria, the results were similar. Patients with negative PET-2 had superior 3-year PFS (82.9% vs. 51.6%, p < 0.001) and OS (89.7% vs. 72.9%, p = 0.001) than those with positive results. Patients with negative PET-4 also had a better clinical outcome compared with the positive group with 3-year PFS (81.8% vs. 30.0%, p < 0.001) and OS (90.1% vs. 54.2%, p < 0.001). In the multivariate analysis, PET-4 with IHP criteria, PET-4 with 5-PS criteria and IPI score were independent predictive factors for PFS of patients with DLBCL. A further analysis of positive predictive value (NPV) and negative predictive value (PPV) showed PET-4 was superior than PET-2, especially interpretated with 5-PS. The PPV and NPV of PET-4 with 5-PS criteria were 70.00% and 83.82% respectively.

These data indicates that Interim ¹⁸F-FDG PET-CT scan could predict the prognosis of DLBCL patients treated with R-CHOP regimen. And it was recommended that interim ¹⁸F-FDG PET-CT scan be done after 4 cycles of chemotherapy, and that 5-PS criteria be applied in the interpretation of interim PET-CT scan rather than IHP criteria.