Outcomes and Prognostic Factors of Transformed Follicular Lymphoma: An Analysis from the LEO Consortium

Abstract

Introduction: Histologic transformation (HT) from follicular lymphoma (FL) to an aggressive lymphoma is infrequent (rates in the rituximab era, ~ 2% per yr over the first 5 yrs after diagnosis) with unfavorable outcomes. Sarkozy (2019) showed approximately half of lymphoma-related deaths in the first decade following diagnosis follow HT. Reported survival outcomes following HT are heterogeneous, with median survival ranging from 22 to 50 mos. and 5-yr survival ranging from 24-60%. Establishing prognosis at the time of HT would be useful, but previous research efforts have been limited by small sample sizes. Data on factors associated with outcomes following HT have included time to progression, prior therapy, anthracycline exposure, and early HT, but contemporary analyses examining these factors in large, expert-practice datasets is lacking. Herein we utilized the scope of the Lymphoma Epidemiology of Outcomes (LEO) Consortium to examine prognostic factors for and outcomes following HT. Methods: This was a multicenter observational cohort study from the LEO Consortium. The cohort consists of prospectively enrolled patients in the University of Iowa/Mayo Clinic SPORE MER and LEO Cohorts, as well as patients with HT retrospectively identified at LEO centers. Data were collected from medical records by standard protocol with all events verified by lymphoma clinicians. Eligible patients were diagnosed between 2002-2019, had a biopsy proven grade 1-3A FL at diagnosis with a subsequent biopsy showing HT to an aggressive lymphoma (FL 3B, diffuse large B cell lymphoma (DLBCL), or high-grade B cell lymphoma (HGBCL)); patients with a component of FL3B, DLBCL, or HGBCL at diagnosis were excluded. The primary outcome was overall survival (OS), defined as the time from HT until death from any cause. Associations between clinical factors at time of HT and/or diagnosis with OS were evaluated using Kaplan Meier curves and Cox models; cause of death was evaluated using a competing risk approach. Results: 306 patients were identified having a HT event between 2002 and 2021. Histology at diagnosis was FL grade 1-2 in 86% and FL 3A in 14%. Median (IQR) age at time of HT was 63 (56-71) yrs; 39% of patients were female. Median time from diagnosis to HT was 28 (11-60) mos. 65% of patients (n=199) had prior immunochemotherapy (IC), of whom 59% (n=117) received an anthracycline. 20% (n=61) patients had HT prior to any systemic treatment and 53% (n=163) had HT after only 1 line. 69% (n=138) of HT following IC occurred within 24 mos of diagnosis. Transformed lymphoma histologies were predominantly DLBCL (83%), followed by HGBCL (11%) and FL3B (6%). At median follow-up from HT of 86 mos (62-99), 158 patients died, with 114 of 136 known causes of death (84%) lymphoma related. Median OS following HT was 61 mos (95% CI 50-97), and 5-yr OS estimate was 52% (95% CI: 46-58%). Early HT (within 24 mos of diagnosis) was associated with inferior outcomes (5-yr OS 42% vs 59%, HR=1.86 (1.36-2.55)) (Figure 1). Age >70 years was associated with worse OS compared to age<60 (5-yr OS 37% vs 61%, HR=2.19 (1.48-3.23)). No significant association with OS was observed for age 60-69 yrs vs younger patients. Anthracycline exposure prior to HT was associated with worse OS (5-yr OS 41% vs 58%, HR=1.51 (1.10-2.08)). Exposure to IC (5-yr OS 42% vs 67%, HR=1.68 (1.20-2.36)) or any systemic therapy (5-yr OS 46% vs 71%, HR=1.76 (1.15-2.70)) was also associated with inferior OS. Gender, histologic grade at diagnosis, or subtype at HT did not demonstrate association with OS (Table 1). Among patients who died, cause of death was predominantly due to lymphoma-related causes (progression, n=95 (65%); therapy-related, n=14 (10%)) compared to non-lymphoma (n=16, 11%) or missing/unknown causes (n=22, 15%). Cause of death patterns were consistent across all clinical groups evaluated. Conclusions: HT conveys a troubling prognosis for lymphoma patients. These data confirm previous reports of early HT, prior IC, and anthracycline exposure as negative prognostic indicators and extends the observation to any systemic therapy. HT at age >70 adds to inferior OS, but a more refined prognostic model will require more detailed clinical variables from time of diagnosis and HT. These data highlight HT as continued area of unmet need with high lymphoma-related mortality. Further research is needed with additional prognostic factors at the time of HT and the influence of patterns of care on outcomes. Figure 1View largeDownload PPTFigure 1View largeDownload PPT Close modal