Purpose: We aimed to compare the standardized central review of (18)F-fluorodeoxyglucose (FDG) positron emission tomography (PET)/CT scans performed after induction therapy for follicular lymphoma (FL) in the PRIMA study (Salles et al., Lancet 377:42-51, 2011; Trotman et al., J Clin Oncol 29:3194-3200, 2011) to scan review at local centres.
Methods: PET/CT scans were independently evaluated by two nuclear medicine physicians using the 2007 International Harmonization Project (IHP) criteria (Cheson et al., J Clin Oncol 25:579-586, 2007; Juweid et al., J Clin Oncol 25:571-578, 2007; Shankar et al., J Nucl Med 47:1059-1066, 2006) and Deauville 5-point scale (5PS) criteria (Meignan et al., Leuk Lymphoma 50:1257-1260, 2009; Meignan et al., Leuk Lymphoma 51:2171-2180, 2010; Barrington et al., Eur J Nucl Med Mol Imaging 37:1824-1833, 2010). PET/CT status was compared with prospectively recorded patient outcomes.
Results: Central evaluation was performed on 119 scans. At diagnosis, 58 of 59 were recorded as positive, with a mean maximum standardized uptake value (SUVmax) of 11.7 (range 4.6-35.6). There was no significant association between baseline SUVmax and progression-free survival (PFS). Sixty post-induction scans were interpreted using both the IHP criteria and 5PS. Post-induction PET-positive status failed to predict progression when applying the IHP criteria [p = 0.14; hazard ratio (HR) 1.9; 95 % confidence interval (CI) 0.8-4.6] or 5PS with a cut-off ≥3 (p = 0.12; HR 2.0; 95% CI 0.8-4.7). However, when applying the 5PS with a cut-off ≥4, there was a significantly inferior 42-month PFS in PET-positive patients of 25.0% (95% CI 3.7-55.8%) versus 61.4% (95% CI 45.4-74.1%) in PET-negative patients (p = 0.01; HR 3.1; 95% CI 1.2-7.8). The positive predictive value (PPV) of post-induction PET with this liver cut-off was 75%. The 42-month PFS for patients remaining PET-positive by local assessment was 31.1% (95% CI 10.2-55.0%) vs 64.6% (95% CI 47.0-77.6%) for PET-negative patients (p = 0.002; HR 3.3; 95% CI 1.5-7.4), with a PPV of 66.7%.
Conclusion: We confirm that FDG PET/CT status when applying the 5PS with a cut-off ≥4 is strongly predictive of outcome after first-line immunochemotherapy for FL. Further efforts to refine the criteria for assessing minimal residual FDG uptake in FL should provide a reproducible platform for response assessment in future prospective studies of a PET-adapted approach.