TRIAL-INELIGIBLE PATIENTS RECEIVING IMMUNE CHECKPOINT INHIBITORS FOR ADVANCED NSCLC: A SYSTEMATIC REVIEW OF REAL-WORLD EVIDENCE
Author(s)
Jimin Son1, Dongwon Yoon, ScD2, Seungjin Bae, ScD3.
1Masters student, Ewha Womans University, Seoul, Korea, Republic of, 2National Institutes of Health, Baltimore, MD, USA, 3Ewha Womans University, College of Pharmacy, Seoul, Korea, Republic of.
1Masters student, Ewha Womans University, Seoul, Korea, Republic of, 2National Institutes of Health, Baltimore, MD, USA, 3Ewha Womans University, College of Pharmacy, Seoul, Korea, Republic of.
OBJECTIVES: Pivotal immune checkpoint inhibitor (ICI) trials in advanced non-small cell lung cancer (NSCLC) used restrictive eligibility criteria, yet trial-ineligible patients are treated routinely. We mapped the reporting of prespecified trial-ineligible criteria and summarized overall survival (OS) differences among trial-ineligible patients in real-world ICI studies.
METHODS: MEDLINE, Embase, and Web of Science were searched for real-world studies published from 2014 to 2026 that evaluated ICI use in advanced/metastatic NSCLC and reported trial eligibility or outcomes in patients with trial-ineligible characteristics. Domains were prespecified from pivotal trials: Eastern Cooperative Oncology Group (ECOG) performance status ≥2, brain metastases, autoimmune disease, corticosteroid use, and organ dysfunction. The primary outcome was the reporting frequency of each trial-ineligible criterion; the secondary outcome was OS differences reported for trial-ineligible patients. Due to heterogeneity, findings were synthesized narratively.
RESULTS: Of 19,335 records identified and 10,450 screened, seven studies met inclusion criteria: first-line (n=3), later-line (n=1), and mixed/all-line (n=3). Domain reporting was uneven: ECOG ≥2 was reported in all studies (7/7), brain metastases in 4/7. Autoimmune disease (2-13%), corticosteroid use (8-15%), and organ dysfunction (26%) were mainly reported as prevalence. In studies that explicitly classified trial eligibility, trial-ineligible patients comprised 46-47% of ICI recipients. Domain-specific OS was reported most consistently for ECOG ≥2, which was associated with shorter OS across lines, including later-line nivolumab (3.0 vs 10.2 months) and first-line pembrolizumab (6.1 vs 19.6 months; adjusted HR 1.89). One study linked autoimmune disease to OS (not reached vs 22.7 months, p=0.79), and none reported domain-specific OS for corticosteroid use or organ dysfunction.
CONCLUSIONS: Across real-world ICI studies in NSCLC, trial-ineligible criteria were reported unevenly: poor performance status was consistently reported and linked to shorter survival, whereas other recognized trial-ineligible criteria remain sparsely linked to outcomes. Further real-world evidence is needed for patients treated outside pivotal-trial criteria to inform treatment and coverage decisions.
METHODS: MEDLINE, Embase, and Web of Science were searched for real-world studies published from 2014 to 2026 that evaluated ICI use in advanced/metastatic NSCLC and reported trial eligibility or outcomes in patients with trial-ineligible characteristics. Domains were prespecified from pivotal trials: Eastern Cooperative Oncology Group (ECOG) performance status ≥2, brain metastases, autoimmune disease, corticosteroid use, and organ dysfunction. The primary outcome was the reporting frequency of each trial-ineligible criterion; the secondary outcome was OS differences reported for trial-ineligible patients. Due to heterogeneity, findings were synthesized narratively.
RESULTS: Of 19,335 records identified and 10,450 screened, seven studies met inclusion criteria: first-line (n=3), later-line (n=1), and mixed/all-line (n=3). Domain reporting was uneven: ECOG ≥2 was reported in all studies (7/7), brain metastases in 4/7. Autoimmune disease (2-13%), corticosteroid use (8-15%), and organ dysfunction (26%) were mainly reported as prevalence. In studies that explicitly classified trial eligibility, trial-ineligible patients comprised 46-47% of ICI recipients. Domain-specific OS was reported most consistently for ECOG ≥2, which was associated with shorter OS across lines, including later-line nivolumab (3.0 vs 10.2 months) and first-line pembrolizumab (6.1 vs 19.6 months; adjusted HR 1.89). One study linked autoimmune disease to OS (not reached vs 22.7 months, p=0.79), and none reported domain-specific OS for corticosteroid use or organ dysfunction.
CONCLUSIONS: Across real-world ICI studies in NSCLC, trial-ineligible criteria were reported unevenly: poor performance status was consistently reported and linked to shorter survival, whereas other recognized trial-ineligible criteria remain sparsely linked to outcomes. Further real-world evidence is needed for patients treated outside pivotal-trial criteria to inform treatment and coverage decisions.
Conference/Value in Health Info
2026-11, ISPOR Europe 2026, Vienna, Austria
Value in Health, Volume 29, Issue 12S
Code
RWD56
Topic
Clinical Outcomes, Health Policy & Regulatory, Real World Data & Information Systems
Disease
Oncology, Respiratory-Related Disorders (Allergy, Asthma, Smoking, Other Respiratory), Systemic Disorders/Conditions (Anesthesia, Auto-Immune Disorders (n.e.c.), Hematological Disorders (non-oncologic), Pain)