FL is the most common low-grade B-cell NHL. Although indolent, it has heterogeneous biological behavior and variable clinical outcomes. Histologic transformation (HT) to high-grade B-cell NHL is associated with dismal survival. The therapeutic management of FL and its prognosis are highly dependent on its staging and tumor burden. Here we aimed to describe clinical characteristics, assess outcomes, determine predictors of survival and HT, and compare responses between different therapies applied in a large cohort of FL patients.

This retrospective study involved 223 patients with FL G1-3A, diagnosed at USP, Brazil, from 2006 to 2022. FL patients were categorized into early-stage (I/II) (ES), advanced-stage (AS) (III/IV) with low tumor burden (LTV), and AS with high tumor burden (HTV) according to the GELF criteria. Endpoints were OS, PFS, early-relapse (< 24 months) and HT rates. Survival curves were constructed using the KM method and the Log-Rank test was used to assess the relationship between variables and outcomes. Univariate and multivariate analysis were performed. The results were presented in HR and 95% CI and p ≤ 0.05 was considered significant.

The median age at diagnosis was 60 years and 54.8% were female. 15% of cases had ES, 18.4% had AS-LTV, and 66.8% had AS-HTV. BM involvement, leukemic presentation, splenomegaly, and serous effusions occurred in 48.9%, 11.7%, 12.6%, and 21.1%, respectively. Bulky ≥ 7 cm, B-symptoms, ECOG ≥ 2, and involvement of ≥ 4 nodal areas were observed in 44.4%, 51.6%, 15.2%, and 56%, respectively. 49% of patients were categorized as high-risk according to the FLIPI score. With a median follow-up of 79.5 months (95% CI 67.0-92.0) the OS medians were 18 years (95% CI 14.9-21.1), 11 years (95% CI 9.2-12.9), and 10.5 years (95% CI 9.3-11.8) for ES, AS-LTV, and AS-HTV, respectively, p = 0.138. Similarly, the PFS medians were 6.7 years (95% CI 3.6-9.8), 8.4 years (95% CI 2.0-14.8), and 3.5 years (95 CI 1.8-5.3) for ES, AS-LTV, and AS-HTV, respectively, p = 0.171. Early-relapses occurred in 36.3% of cases, being 18.2%, 31.7%, and 41.0% for ES, AS-LTV, and AS-HTV, respectively, p = 0.032. The overall mortality rate was 29.1% for the whole cohort. HT was documented in 16.7% of cases, being 3.0%, 14.6% and 20.1% for ES, AS-LTV, and AS-HTV, respectively, p = 0.053. Among patients with AS-HTV, the R-CHOP regimen did not promote increased OS compared to the R-CVP regimen (p = 0.415), however it was associated with a substantial increase in PFS (p = 0.005). Age ≥ 60 years (HR 1.06, p < 0.001), ≥ 2 comorbidities (HR 3.85, p = 0.014), B-symptoms (HR 2.35, p = 0.015), HT (HR 2.77, p = 0.002) and thrombocytopenia (HR 2.96, p = 0.028) were predictors of poor OS. Similarly, age ≥ 60 years (HR 1.08, p < 0.001), involvelment of ≥ 4 nodal areas (HR 1.27, p < 0.001), and high LDH (HR 1.65, p = 0.002) predicted decreased PFS. Serous effusions (HR 2.09, p = 0.05), albumin < 3.5 g/dL (HR 2.82, p = 0.05), B-symptoms (HR 2.00, p = 0.006), involvement of ≥ 4 nodal areas (HR 1.21, p = 0.014), and BM infiltration (HR 2.11, p = 0.05) were the main predictors for HT.

Although it is characteristically an indolent disease, we demonstrated that a significant portion of FL patients have shortened survival. We confirmed this prognostic heterogeneity, particularly considering the clinical staging and tumor burden. Therefore, FL patients with AS-HTV had higher HT rates and early-relapses, both classic adverse prognostic markers, as well as a tendency to higher mortality. Additionally, we identified clinical and laboratory predictors for HT, which may in a near future direct adapted-risk therapeutic strategies.