Impact of Trial Eligibility Criteria on Outcomes of 1,245 Patients with Follicular Lymphoma Treated in the Real-World Setting: A Danish Population-Based Study

Background

For most patients, follicular lymphoma (FL) is an incurable, chronic disease and time in remission gradually becomes shorter with each new treatment intervention. Survival after 1st line immunochemotherapy is excellent overall, but outcomes for the minority with progression of disease within 24 months of first line treatment (POD24) are dismal. The 10-year overall survival (OS) rate in clinical trials of 1st line immunochemotherapy is around 80%, making further survival improvements difficult to achieve. However, the unmet medical needs could be more substantial for the group of patients which are excluded from trial participation due to violation of inclusion/exclusion criteria.

Aim

To investigate the impact of commonly used in/exclusion criteria in clinical trials on the OS of real-world patients with FL.

Methods

This study is based on a nationwide population-based lymphoma registry (LYFO). Patients ≥18 years with grade 1-3 FL (including 3B as grade 3 is not subcategorized in LYFO) and diagnosed between 2000-2018 were included if they had received 1st line treatment with R-CVP, R-CHOP or R-Bendamustin within 90 days of date of diagnosis. The eligibility criteria from following trials were applied to the patient population: GALLIUM, RELEVANCE, BRIGHT, and StiL NHL1. First, patients were selected to match the age group and disease stages relevant for the trials. The impact of performance status (ECOG PS), hemoglobin, platelets, neutrophil count (estimated as total WBC-lymphocyte count), liver function (ALAT, bilirubin) and kidney function (estimated by eGFR) on OS were assessed. Shapley-values measuring an average decrease in 5-year restricted mean survival difference by omitting a criterion were used to assess the impact of the inclusion criteria on the OS difference between eligible and ineligible patients.

Results

A total of 3,702 adults were diagnosed with FL, of those 1,183 were included in the study. Median age was 63 and male:female ratio was 1.1. Firstline treatments were R-Bendamustin (20.3%), R-CVP (35.4%) and R-CHOP (44.3%), proportion treated with R-Bendamustin increasing to 50% after 2016. Rituximab maintenance was used in 47.3%. The 5-year OS for all patients was 82% with no differences between patients treated with R-Bendamustine (5-year OS, 79%, CI: 73-86%), R-CVP (81%, CI: 77-85%) and R-CHOP (84%, CI: 80-87%) (Table 1).

1,036 patients were relevant for the GALLIUM trial based on age (≥18 years) and Ann Arbor disease stage (2 with bulky disease, 3 or 4) alone. Of 870 patients with available data on in/exclusion criteria, 778 (89%) were trial eligible, 92 (11%) were ineligible. The outcome of eligible patients was superior to the outcome of ineligible patients with 5-year OS of 83% vs 56% (P<0.0001, Fig. 1) with results similar when stratifying on treatment. Adjustments for age differences between eligible and ineligible patients did not impact results (5-year OS difference of 24%, CI: 13-35%). Results were consistent for the other trials with 5-year OS for eligible and ineligible patients of 83 vs 62% for RELEVANCE (P<0.0001), 84% vs 66% for BRIGHT (P<0.0001) and 81% vs 53% for StiL NHL1 (P<0.0001). POD24 was observed more frequently in ineligible patients (except in RELEVANCE), even though deaths without disease progression were treated as a competing risk to reduce impact of comorbidity related deaths.

According to the Shapley-values, ECOG PS (ranging from 5.64 to 9.98 months) and creatinine (ranging from 1.27 to 11.0 months) were the most important contributors to the observed OS differences between eligible and ineligible patients.

The main reason for ineligibility for GALLIUM, RELEVANCE and BRIGHT was low creatinine clearance with 38%, 69% and 47% of ineligible patients violating this criterion respectively (24% in StiL NHL1). In StiL NHL1, 45% and 39% of ineligible patients did not fulfill the ECOG PS and ALAT criteria.

Conclusions

This study shows that commonly used in/exclusion criteria in clinical trials of FL lead to exclusion of a group of patients with 5-year OS outcomes significantly worse than trial eligible patients. A higher proportion of POD 24 events in the group of ineligible patients also suggests that disease control was inferior in those patients.

Collectively, these results show that a high unmet need remains for better 1st line therapies for the minority of FL patients ineligible for clinical trials with narrow in/exclusion criteria.

Brown:Roche: Consultancy; Novartis: Consultancy; Gilead: Consultancy. Jørgensen:Gilead: Consultancy; Roche: Consultancy; Novartis: Consultancy; BMS: Consultancy; Orion Pharma: Consultancy; Incyte: Consultancy. Larsen:Genentech: Research Funding; Roche: Consultancy; Novartis: Consultancy; Gilead Sciences: Consultancy; Bristol Myers Squibb: Consultancy. Jakobsen:Roche: Honoraria. El-Galaly:Abbvie: Other: Teaching in 2021; Roche: Ended employment in the past 24 months.