Evaluation of the m7-FLIPI in Patients with Follicular Lymphoma: FOXO1 Mutational Status May be a Predictive Marker of Early Progression and Long-Term Outcome

BACKGROUND:

In pts with follicular lymphoma (FL) who received chemoimmunotherapy (CIT) as first-line treatment, progression of disease within 24 months of diagnosis (POD24) was associated with increased risk of death (Casulo, 2015). A clinicogenetic risk model that includes the mutation status of 7 genes (EZH2, ARID1A, MEF2B, EP300, FOXO1, CREBBP, and CARD11), FLIPI and ECOG, known as the m7-FLIPI, was shown to improve risk stratification in frontline pts with advanced stage FL treated with CIT (Pastore, 2015), as well as predict POD24 after first-line CIT (Jurinovic, 2016). With the hypothesis that there are distinguishing features in the mutational landscape between pts with FL who experience POD24 vs those who don’t, we evaluated the prognostic utility of the m7-FLIPI and its components in a retrospective series.

METHODS:

MSK-IMPACT Heme (Ptashkin, 2019) is a 468 targeted gene panel that includes the 7 genes required to calculate the m7-FLIPI. We identified all pts with FL, grade 1-3A, who had IMPACT testing, excluding pts with FL grade IIIB, mixed histology, or de novo transformation at diagnosis (n = 295). Three pts were excluded for inadequate clinical data (n = 292). The m7-FLIPI was calculated as previously described. Progression-free survival (PFS) was calculated from date of diagnosis till date of progression after first-line treatment, death, or censored at date of last follow-up. Overall survival (OS) was calculated from date of diagnosis till death or censored at date of last follow-up. Cox proportional hazard models were used to compare differences in survival and the Fisher’s exact test for categorical variables in SAS 9.4.

RESULTS:

A cohort of 292 pts was analyzed. Genomic alterations were identified in the 7 genes. Median age of diagnosis was 57 years (range 26-92) with a slight male predominance (53%). The median follow-up was 5.2 years (95% CI, 4.6-5.7). Initial treatment included monoclonal antibody + chemotherapy (57%), single-agent rituximab (17%), chemotherapy (2%), radiotherapy alone (8%), and other treatments (5%). Thirty-four pts (12%) were never treated.

Out of 292 pts, 86 (29%) were high-risk by FLIPI and 42 (14%) were high-risk m7-FLIPI. Forty-four out of 86 pts with high-risk FLIPI (51%) were reclassified as low-risk by the m7-FLIPI. High-risk FLIPI was associated with shorter PFS (HR 1.54, 95% CI 1.12 – 2.11, p = 0.0085) and OS (HR 3.73, 95% CI 1.74 – 7.98, p = 0.0007). High-risk m7-FLIPI was associated with shorter PFS (HR 1.55, 95% CI 1.05 – 2.28, p = 0.027) but not OS (HR 2.09, 95% CI 0.92 – 4.76, p = 0.08). C-statistics showed that m7-FLIPI did not outperform FLIPI alone as a prognostic tool for PFS (FLIPI c-statistic = 0.55, 95% CI 0.51 – 0.59; m7-FLIPI c-statistic = 0.53, 95% CI 0.50 – 0.56) or OS (FLIPI c-statistic = 0.68, 95% CI 0.58 – 0.78; m7-FLIPI c-statistic = 0.55, 95% CI 0.47 – 0.63).

Mutations in ARID1A as a single factor were associated with shorter PFS (HR 1.52, 95% CI 1.01-2.29, p=0.047), but this association did not remain significant after adjusting for FLIPI and ECOG (HR 1.49, 95% CI 0.97-2.28, p=0.07). Mutations in FOXO1 were associated with shorter PFS (HR 1.48, 95% CI 1.00-2.19, p=0.049) and OS (HR 2.65, 95% CI 1.21-5.79, p=0.015) as a single factor. This association remained significant after adjusting for FLIPI and ECOG: PFS (HR 1.60, 95% CI 1.08-2.37, p=0.020) and OS (HR 2.65, 95% CI 1.21-5.83, p=0.015; Table 1).

High-risk m7-FLIPI and FLIPI pts were not more significantly likely to experience POD24, OR 1.44, 95% CI 0.73-2.83, p=0.37 for m7-FLIPI and 1.48, 95% CI 0.87–2.51, p=0.17 for FLIPI. Pts with mutated FOXO1 were more likely to experience POD24, OR 2.47, 95% CI 1.29-4.71, p=0.008.

DISCUSSION:

The m7-FLIPI neither improved risk stratification compared to FLIPI nor did it predict POD24 in this series. Mutations in FOXO1 were associated with shorter PFS and OS independent of FLIPI and ECOG. Pts with mutated FOXO1 were enriched among those with POD24, suggesting that FOXO1 mutation may be a predictive marker for early failure of first-line treatment and death, including pts not fulfilling the POD24 criteria. Similar negative impact of FOXO1 was reported in a comprehensive genomic analysis of FL (Mozas, 2023). Both series are limited by patient numbers. Additional efforts to confirm this observation may be useful for understanding the mechanisms behind the prognostic effect of POD24 on survival and improving long-term outcome in this high-risk FL group.