Prognostic value of end-of-induction PET response after first-line immunochemotherapy for follicular lymphoma (GALLIUM): secondary analysis of a randomised, phase 3 trial
Abstract
Introduction: The prognostic value of 18F-FDG PET-CT (PET) response assessment after first-line (1 L) immunochemotherapy for advanced-stage symptomatic follicular lymphoma (FL) has been reported in several smaller trials. We evaluated the prognostic value of PET complete remission (PET-CR) status for the large FL patient (pt) cohort enrolled in the prospective Phase III GALLIUM study (NCT01332968; Marcus 2016).
Methods: 1202 pts with previously untreated FL (ITT population) were randomised 1:1 to receive induction therapy comprising chemotherapy plus 1000 mg obinutuzumab (G; D1, 8, 15 C1 then D1 subsequent cycles) or 375 mg/m2 rituximab (R; D1 each cycle), for 8 x 21-day cycles (CHOP, CVP) or 6 x 28-day cycles (bendamustine). PET scans, introduced after an early protocol amendment (July 2011), were taken at baseline and end of induction (EOI) visits and assessed by the investigator (INV) and an independent review committee (IRC) comprising two radiologists, with a third adjudicator; final response was determined by a clinician.
Response was assessed by CT and PET plus bone marrow biopsy, applying the revised International Working Group (IWG) criteria (Cheson 2007, Juweid 2007). EOI PET-CR status was compared with pt characteristics, CT-based response, PFS and OS.
Results: Of 609 pts with a baseline PET scan, 595 had detectable lesions, and 535 also had an evaluable PET at EOI. Baseline disease and demographic characteristics were similar in the PET-evaluable and non-PET populations.
Pts with non-available (n = 52) and non-evaluable (n = 8) scans were considered as non-responders; these pts and those who progressed prior to EOI were excluded from landmark PFS analyses. At EOI 390/595 (65.5%) pts had achieved a PET-CR according to IRC, comprising 212/297 (71.4%) G-chemo pts and 178/298 (59.7%) R-chemo pts.
After a median follow-up of 34.5 months, EOI PET-CR status was highly prognostic of both PFS (PET-CR vs PET-non CR: HR 0.39; 95% CI 0.25-0.60; p < 0.0001) and OS (HR 0.41; 95% CI 0.19-0.86; p = 0.018; see Figure). 2.5-year PFS from EOI was 87.6% (95% CI 83.5-90.8) for PET-CR pts compared with 70.9% (95% CI 61.3-78.6) for PET-non CR pts; corresponding OS was 96.6% (95% CI 94.1-98.1) vs 90.9% (95% CI 84.7-94.6). IRC PET status was prognostic in both G- and R-treated populations.
Concordance between INV and IRC evaluation was 68.6%. Conclusions: This large prospective analysis confirms that PET status after 1 L immunochemotherapy, applying IWG 2007 criteria, is an early prognostic factor for PFS and OS in FL.
Further analyses, including PET assessment by the INV, according to treatment arm, and IRC review applying a GREATER-THAN OR EQUAL TO 4 point cut-off on the recommended 5-point scale for response assessment (Barrington 2014) will be presented. Pooled analyses of these and other data with longer follow-up may determine PET response as a reliable surrogate for PFS and OS, providing a platform for study of response-adapted therapy.