End of induction [18F]FDG PET is prognostic for progression-free survival and overall survival in follicular lymphoma patients enrolled in the FOLL12 trial

Purpose: To evaluate the reliability of the Deauville score (DS) in therapy response assessment and to define the prognostic value of the metabolic response of end of induction (EOI) [18F]FDG PET (PET) in follicular lymphoma patients. Methods: Adult patients with untreated grade 1–3a FL/ stage II‐IV enrolled in the multicentre, prospective, phase III FOLL12 trial (NCT02063685) were randomized to receive standard immunochemotherapy followed by rituximab maintenance (standard arm) versus standard immunochemotherapy followed by response-adapted post‐induction management (experimental arm). Baseline and EOI PET were mandatory for the study. All PET scans were centralized on the WIDEN® platform and classified according to DS in a blind independent central review. DS1–3 was considered negative (CMR), whereas DS4‐5 was considered positive (not CMR). The primary endpoint was PFS. The main secondary endpoint was overall survival (OS). Results: Overall, 807 follicular lymphoma patients—52% women, 89% stage III-IV disease, 40% with a high-risk FLIPI-2 score (3–5)—were enrolled in the study; 729 (90.4%) baseline and EOI PET were available for the analysis. EOI PET was positive (DS4–5) in 88/729 (12.1%) cases. Overall inter-reviewer agreement on PET pos/neg result was 0.92, while agreement on positive and negative cases was 0.77 and 0.94, respectively. The median follow-up was 69 months; 247 events were registered in the 5-yr follow-up, with a 5-yr PFS of 67% (95%CI: 63%–70%). The 5-yr PFS rate for PET neg (DS1–3) and PET pos (DS4–5) patients was 71% (95%CI: 67%–75%) and 36% (95%CI: 25%–46%), respectively, with HR 3.49 (95%CI: 2.57–4.72). Five-year PFS was worse as DS increased, with 74% (70%–78%), 58% (48%–67%; HR 1.71; p = 0.001)] and 36% (25%–46%; HR 3.88; p < 0.001) in DS1–2, DS3 and DS4–5, respectively. EOI PET maintained its prognostic value in both the standard and experimental arms. In the whole population, 5-yr OS was 94% (95%CI: 92%–96%), with 96% (95%CI: 94–97) and 82% (95%CI: 72%–89%) in EOI PET negative (DS1–3) and positive (DS4–5), respectively (HR 4.48; p < 0.001). When DS was associated with FLIPI-2, patients with DS3 or DS1–2 with high FLIPI-2 (3–5) experienced worse OS than patients with DS1–2 and low FLIPI-2 (1–2) (p = 0.003). Conclusion: This study shows that DS is a reliable prognostic tool to evaluate EOI PET in follicular lymphoma patients, with prognostic value maintained both in the standard and experimental arms, making metabolic imaging a robust tool to assess response in FL. Moreover, although preliminary, this study provides further information on DS3 patients, who are considered as CMR but show a less favourable PFS than DS1–2 patients.