PFS in Follicular Lymphoma Predicted by Gene Expression Model

Gilles A. Salles, MD, PhD

Gilles A. Salles, MD, PhD

Progression-free survival (PFS) was successfully predicted by a 23-gene expression-based model for patients with follicular lymphoma enrolled in the phase III randomized PRIMA trial. Investigators believe the model can allow clinicians to better adjust existing treatment options according to patient risk.

Investigators identified 395 genes that were associated with the risk for progression, 23 of which were included in a predictive model based on both B-cell biology and tumor microenvironment. In a multivariate Cox model for PFS adjusted on rituximab (Rituxan) maintenance treatment and Follicular Lymphoma International Prognostic Index 1 (FLIPI-1) score, the model independently identified patients at high risk for progression compared with those at low risk (adjusted HR [aHR], 3.68; 95% CI, 2.19-6.17;P<.0001). The 5-year PFS was 26% (95% CI, 16-43) in the high-risk group and 73% (95% CI, 64-83) in the low-risk group.

The predictor performances were confirmed in each of 3 individual validation cohorts. The aHR comparing high-risk to low-risk groups was 2.57 (95% CI, 1.65-4.01) in cohort 1, 2.12 (95% CI, 1.32-3.39) in cohort 2, and 2.11 (95% CI, 1.01-4.41) in cohort 3.

The median PFS was 3.1 years (95% CI, 2.4-4.8) for the high-risk group and 10.8 years (95% CI, 10.1-not reached) for the low-risk group (P<.0001) in the combined validation cohort. The risk for lymphoma progression at 2 years was 38% (95% CI, 29-46) in the high-risk group and 19% (95% CI, 15-24) in the low-risk group. In a multivariate analysis, the score predicted PFS independently of anti-CD20 maintenance treatment and of the FLIPI score (HR, 2.30; 95% CI, 1.72-3.07).

&ldquo;This predictor can be used in routine practice and captures multiple aspects of the biology of the tumor and the heterogeneous composition of the tumor microenvironment,&rdquo; Gilles Salles, MD, Service d&rsquo;Hématologie Clinique, Centre Hospitalier Lyon Sud, Pierre-Bénite Cedex, France, and colleagues wrote.

&ldquo;Together with clinical parameters such as the FLIPI index, this score might allow clinicians to better adjust existing therapeutic options according to the patient risk category. For patients at low risk of progression, short treatments with a low toxicity profile should be considered. For patients with high-risk FLIPI and 23-gene scores, having a 50% risk estimate of lymphoma progression at 2 years, new treatment options should be developed,&rdquo; added Salles et al.

Salles and colleagues collected fresh-frozen tumor biopsies from 160 treatment-na&iuml;ve patients participating in the PRIMA trial evaluating rituximab maintenance following rituximab plus chemotherapy induction in patients with high-tumor-burden follicular lymphoma.

Patients were first treated with 1 of 3 regimens: R-CHOP (rituximab, cyclophosphamide, doxorubicin, vincristine, and prednisone), R-CVP (rituximab, cyclophosphamide, vincristine, and prednisone), or R-FCM (rituximab, fludarabine, cyclophosphamide, and mitoxantrone). After this induction phase, patients with a complete or partial response were randomly assigned to observation or rituximab maintenance for 2 years.

Investigators collected RNA from 150 patients. Formalin-fixed, paraffin-embedded biopsies were also available for 53 patients for the technical validation of gene-expression quantification on fixed tissue.

As mentioned above, the model generated from the training cohort was further evaluated in 3 independent validation cohorts of patients for whom formalin-fixed, paraffin-embedded biopsies obtained at diagnosis were available, including a distinct validation set of patients in PRIMA who were not part of the training cohort (cohort 1; n = 178), 201 patients from the University of Iowa/Mayo Clinic Lymphoma SPORE trial (cohort 2), and 109 patients from the Hospital Clinic University of Barcelona (cohort 3). Cohorts 2 and 3 consisted of newly diagnosed patients who were prospectively recruited in observational studies.

When investigators tested the 23-gene expression model in the independent cohorts, the digital gene-expression data met quality criteria for 460 (94%) of 488 formalin-fixed, paraffin-embedded samples, with performances ranging from 93% to 97% across the 3 cohorts.

&ldquo;Despite recent progress in the stratification and management of patients with follicular lymphoma, a substantial proportion of patients are still underserved by existing standard treatment and have rapid progression of their disease,&rdquo; wrote Salles et al. &ldquo;Our gene-expression predictor could be valuable in the clinical setting to identify patients at high risk or low risk of progression so as to adjust the therapeutic strategy and enrollment for innovative treatments.&rdquo;

Reference:

Huet S, Tesson B, Jais JP, et al. A gene-expression profiling score for prediction of outcome in patients with follicular lymphoma: a retrospective training and validation analysis in three international cohorts [published online February 20, 2018].Lancet Oncol.doi: 10.1016/S1470-2045(18)30102-5.