INTRODUCTION: The utility of PET imaging in aggressive lymphomas has led to its increased use in FL, although guidelines do not recommend routine PET in FL. The NLCS is an observational study that includes 2700+ FL patients who enrolled from 2004-2007. We report on imaging patterns after first-line rituximab (R)-based induction therapy.
METHODS: 1438 patients with FL completed R-induction therapy and were checked for disease assessments performed between 2 cycles prior to and 12 weeks post therapy. Baseline patient demographic and disease features, center type, geographic region, therapy regimen and duration, overall group difference, and pairwise comparisons between imaging groups were compared using Pearson Χ2 tests. A generalized logistic model with backwards stepwise selection (P>0.10) was used to identify factors related to the choice of imaging.
RESULTS: PET and CT were each performed on average 26 days after end of therapy. PET and/or CT were more common in patients <70 y, stage III/IV, > 5 nodal sites, R-Chemo, therapy >85 days, at academic sites, and was less common in the West (P<0.01). On univariate analysis, sex, race, FLIPI, LDH, Hgb, extranodal disease, marrow involvement, ECOG PS and B symptoms were not associated with imaging choice. On multivariate analysis, non-clinical factors such as duration, region, center type and therapy dominated imaging choice (P<0.05). Half of the patients who did not obtain a PET or CT were treated for <85 days. Compared with the West, the Midwest was less likely to use PET+CT vs PET (OR 0.49) or CT (OR 0.39). Academic sites imaged via PET vs CT less than the community (OR 0.50). PET+CT vs CT was more common with R-CHOP (OR 2.63) and R-CVP (OR 2.06), and PET vs CT was more likely with R-CHOP (OR 2.63) than with R-Mono.
CONCLUSIONS: Imaging choice after first-line therapy for patients with FL varies. Non-clinical factors such as duration, region, center type and therapy, rather than patient and disease features, drive imaging choice. Further follow-up is ongoing to predict whether prognostic value of response determinations differ by imaging choice.
