THE VALUE OF ADDITIONAL FOLLOW-UP IN ONCOLOGY: A CASE STUDY OF HEALTH ECONOMIC ANALYSIS USING DANTE PHASE III TRIAL DATA COMBINED WITH ROUTINELY COLLECTED ELECTRONIC HEALTH RECORDS (HER)
Author(s)
Georgios Gkountouras, PhD1, David Meads, PhD1, Ruben Ernesto Mujica-Mota, PhD2, Sarah Danson, BMedSci, BMBS, MSc, PhD, FRCP3, Pippa Corrie, PhD, FRCP4.
1AUHE, University of Leeds, Leeds, United Kingdom, 2University of Exeter, Exeter, United Kingdom, 3University of Sheffield, Sheffield, United Kingdom, 4Cambridge University Hospital NHS Foundation Trust, Cambridge, United Kingdom.
1AUHE, University of Leeds, Leeds, United Kingdom, 2University of Exeter, Exeter, United Kingdom, 3University of Sheffield, Sheffield, United Kingdom, 4Cambridge University Hospital NHS Foundation Trust, Cambridge, United Kingdom.
OBJECTIVES: The UK DANTE clinical trial (ISRCTN15837212) investigated the optimal duration of anti-PD1-based immunotherapy as first line therapy for patients with metastatic melanoma. DANTE’s early termination due to recruitment challenges, urged us to explore the impact of using health records (EHR) of study participants in addition to trial data to extend follow-up in the cost-utility analysis of the study’s treatment strategies.
METHODS: In DANTE, 166 adults with stage III/IV melanoma receiving 1 year of first line anti-PD1 +/- anti-CTLA-4 therapy were randomised (1:1) to stop treatment, (Stop) or continue to at least 2 years (Control). We extended overall survival (OS) follow-up using EHR. We estimated OS using standard parametric-SP (Trial) and relative survival-RS (Trial+EHR) models. We adjusted RS estimates using UK lifetables.
Partitioned survival models (PSM) compared the annually discounted (at 3.5%) costs and benefits of Control vs Stop from a UK NHS perspective over a lifetime horizon, using trial data (PSM1) and extended follow-up data (PSM2). Models included trial-based health state utilities (EQ-5D) plus healthcare resource use with UK-specific unit costs.
Incremental cost-effectiveness ratios (ICERs) were calculated and compared against a quality adjusted life year (QALY) threshold value of £25,000.
RESULTS: EHR usage increased median follow-up time by 57% (Trial+EHR: 44.10 months; IQR: 34.7-60.1 vs Trial: 28.05 months; IQR: 17.9-41.4), with 10-year and 20-year survival prediction estimates increasing for both arms; Stop: 44.9%,13.6% (Trial+EHR) vs 19.1%,0.5% (Trial); Control: 53.4%,20.9%(Trial+EHR) vs 31.1%,2.4% (Trial). In PSM1, patients in Stop arm had 2.02 fewer QALYs (2.60 fewer LYs) and £61,957 less costs versus Control arm (ICER: £30,648/QALY); in PSM2, they had 1.10 fewer QALYs (1.20 fewer LYs) and £52,317 less costs versus Control arm (ICER: £47,650/QALY).
CONCLUSIONS: Securing EHR extended follow-up, impacted analytical approach, but data acquisition was prolonged. Without EHR, we would underestimate OS overall, particularly for the Stop arm, with potentially decision altering consequences.
METHODS: In DANTE, 166 adults with stage III/IV melanoma receiving 1 year of first line anti-PD1 +/- anti-CTLA-4 therapy were randomised (1:1) to stop treatment, (Stop) or continue to at least 2 years (Control). We extended overall survival (OS) follow-up using EHR. We estimated OS using standard parametric-SP (Trial) and relative survival-RS (Trial+EHR) models. We adjusted RS estimates using UK lifetables.
Partitioned survival models (PSM) compared the annually discounted (at 3.5%) costs and benefits of Control vs Stop from a UK NHS perspective over a lifetime horizon, using trial data (PSM1) and extended follow-up data (PSM2). Models included trial-based health state utilities (EQ-5D) plus healthcare resource use with UK-specific unit costs.
Incremental cost-effectiveness ratios (ICERs) were calculated and compared against a quality adjusted life year (QALY) threshold value of £25,000.
RESULTS: EHR usage increased median follow-up time by 57% (Trial+EHR: 44.10 months; IQR: 34.7-60.1 vs Trial: 28.05 months; IQR: 17.9-41.4), with 10-year and 20-year survival prediction estimates increasing for both arms; Stop: 44.9%,13.6% (Trial+EHR) vs 19.1%,0.5% (Trial); Control: 53.4%,20.9%(Trial+EHR) vs 31.1%,2.4% (Trial). In PSM1, patients in Stop arm had 2.02 fewer QALYs (2.60 fewer LYs) and £61,957 less costs versus Control arm (ICER: £30,648/QALY); in PSM2, they had 1.10 fewer QALYs (1.20 fewer LYs) and £52,317 less costs versus Control arm (ICER: £47,650/QALY).
CONCLUSIONS: Securing EHR extended follow-up, impacted analytical approach, but data acquisition was prolonged. Without EHR, we would underestimate OS overall, particularly for the Stop arm, with potentially decision altering consequences.
Conference/Value in Health Info
2026-11, ISPOR Europe 2026, Vienna, Austria
Value in Health, Volume 29, Issue 12S
Code
MSR198
Topic
Economic Evaluation, Methodological & Statistical Research, Real World Data & Information Systems
Disease
No Additional Disease & Conditions/Specialized Treatment Areas, Oncology