Abstract
Objective: To determine treatment regimens used in clinical practice and the associated clinical outcomes among third line (3L) follicular lymphoma (FL) patients in the United States (US).
Methods: This non-interventional, retrospective study used Optum electronic health records (EHR) database for FL patients in the US between 1 Jan 2007 and 31 Dec 2020. The start of this period was selected to align with the Morrison et al. 2019, with 5 years of additional data. The identification period was 1 Jan 2008 to 31 Dec 2019, to ensure at least 1 year of baseline before and 60 days of follow-up (unless death happens before) after the index date, defined as start date of 3L treatment. Adult patients (≥18 years) treated in integrated delivery networks with at least one of the 3L treatments of interest (rituximab, bendamustine and rituximab, phosphatidylinositol 3-kinase [PI3K] inhibitors [copanlisib, duvelisib, idelalisib], lenalidomide and rituximab [R2], tazemetostat, and stem cell transplant) were included. Patients with Diffuse Large B-cell Lymphoma (DLBCL) diagnosis or clinical trial enrollment on or before the index date or any other cancer diagnosis before the first FL diagnosis were excluded. All agents initiated within 90 days after the index diagnosis constituted the 1L treatment. A subsequent line of therapy (LOT) was defined as treatment initiated after ≥180 days following the runout date of all agents, or addition or substitution of a new agent in the prior LOT after 90 days. The primary endpoints were time to progression (DLBCL transformation, new LOT initiation, or supportive care), overall survival (OS) and progression-free survival (PFS), while time to next treatment (TTNT) and treatment patterns were the key secondary endpoints. The analyses were conducted for the overall cohort, patients with early progression within 24 months (POD24) after 1L treatment, patients with index date after and including year 2014, as well as for different 3L treatment regimens. The sub-group with 2014 as index date was selected based on idelalisib approval in 2014.
Results: The final cohort of patients (used one of the 3L treatments of interest and met inclusion/exclusion criteria) consisted of 687 patients: mean age 62.9 years (range 18 - 86), female (46.9%), Caucasians (87.3%), non-Hispanics (92.1%), and median Charlson Comorbidity Index (CCI) 3 (range 1 - 18). Rituximab-based regimens (73.7%) were the most common 3L treatments (mono 38.4%, combo 35.2%). Obinutuzumab was used as combination 3L therapy by 6 (0.87%) patients. Bendamustine, PI3K and lenalidomide monotherapies were administered to 3.1%, 2.2% and 1.9% patients, respectively (Figure 1). Rituximab-based regimens were also the most frequently used 1L, 2L, and 4L treatment options (50.8% moved to 4L and 33.6% had rituximab-based regimens).
The median time to progression, PFS, and TTNT for 3L in the overall cohort were 16.6 (95% CI 14.4, 18.1), 12.5 (95% CI 11.3, 14.4), and 18 (95% CI 15.8, 19.9) months, respectively. The 1-, 2-and 5-year OS were 83.1%, 74.8% and 61.4%, respectively. The outcomes of 3L among POD24 , non-POD24, as well as patients with index date after and including year 2014 were similar to that of the overall cohort. The median time to progression, PFS, and TTNT with rituximab treatment were 19.1 (95% CI 16.7, 21.7), 15.7 (95% CI 14.2, 17.5), and 18.8 (95% CI 17, 21.7) months respectively. The median OS with rituximab therapy was not reached while the 5-year OS was 67% (Table 1).
Moreover, we did not observe statistically significant differences in time to progression, OS, PFS, and TTNT for the 3L treatment between POD24 and non-POD24 patients using a Cox regression model with adjustment for baseline characteristics (age, gender, region, and CCI). The median time to progression, PFS, and TTNT among POD24 vs. non-POD24 were 15.7 vs. 17.9, 11.6 vs. 15.2, and 18 vs. 17.9 months, respectively.
Conclusion: Rituximab-based regimens were the most common 3L treatment options for FL patients. Bendamustine, PI3K, and lenalidomide monotherapies were used by a smaller proportion of patients. R2 was used by a small number of patients for 3L treatment, but it is becoming an important option for FL treatment since its approval in 2019. The majority of outcomes observed could be considered poor, newer agents undergoing clinical trials could provide additional treatment choices to physicians to balance treatment effectiveness with safety and patients' quality of life.
Dai: Novartis: Current Employment, Current equity holder in publicly-traded company. Heo: Genesis Research: Current Employment, Current equity holder in publicly-traded company. Rava: Genesis Research: Current Employment, Current equity holder in publicly-traded company. Jousseaume: Novartis: Current Employment, Current equity holder in publicly-traded company. Ramos: Novartis: Current Employment, Current equity holder in publicly-traded company. Bollu: Novartis: Current Employment, Current equity holder in publicly-traded company.
