A RANDOMISED TRIAL IS NOT ENOUGH: PICO FEASIBILITY FOR FIRST-LINE DURVALUMAB IN HEPATOCELLULAR CARCINOMA
Author(s)
Hend Jedidi, DMD1, Aleksandra Matjaszek, MSc2, Aleksandra Caban, PharmD2, Kamilia Ben Abdallah, Eng1, Lylia Chachoua, PharmD, PhD3, Mondher Toumi, MSc, PhD, MD4.
1Clever-Access, Tunis, Tunisia, 2Clever-Access, Cracow, Poland, 3Clever-Access, Paris, France, 4Aix-Marseille University, Marseille, France.
1Clever-Access, Tunis, Tunisia, 2Clever-Access, Cracow, Poland, 3Clever-Access, Paris, France, 4Aix-Marseille University, Marseille, France.
OBJECTIVES: Does a large randomised pivotal trial guarantee that a Joint Clinical Assessment (JCA) can answer its PICOs? Using the 2024 JCA pilot exercise for first-line durvalumab in advanced or unresectable hepatocellular carcinoma, this study quantified the gap between the comparisons the scope requested and those feasible with the evidence available at marketing-authorisation submission.
METHODS: We analysed the thirteen consolidated PICOs of the MP01 exercise published by the HTACG, mapped the pivotal randomised HIMALAYA trial against each, and searched ClinicalTrials.gov and PubMed for comparator trials. Each PICO was classified, by a pre-specified algorithm, as addressable by direct comparison, potentially addressable by indirect treatment comparison (ITC), or not addressable, anchored to the evidence existing at submission.
RESULTS: The thirteen PICOs shared one intervention but spanned heterogeneous, overlapping populations (full label; Child-Pugh A; Child-Pugh B/C; locoregional-eligible; ECOG-restricted) and seven comparators, expanding to fifteen distinct analyses. Because HIMALAYA was confined to Child-Pugh A, locoregional-ineligible patients, only two PICOs were directly addressable (durvalumab plus tremelimumab; sorafenib). A further subset was potentially addressable by anchored ITC — REFLECT and IMbrave150 share sorafenib with HIMALAYA — but residual population differences and incomplete outcome and subgroup reporting limited interpretability. The Child-Pugh B/C and locoregional-eligible PICOs could not be addressed, those populations being excluded from the trial. Thus, despite a large randomised trial, high-quality evidence for one population and comparator did not translate into feasibility across the scoped set. The feasibility gap is therefore not confined to single-arm products: it arises whenever the scoped populations and comparators outrun the available trial base.
CONCLUSIONS: A randomised pivotal trial is necessary but not sufficient for a feasible JCA. When a broad, comparator-heavy scope outruns the trial population, most questions remain answerable only by indirect comparisons of limited interpretability — an argument for evidence-feasibility screening before the PICO scope is finalised.
METHODS: We analysed the thirteen consolidated PICOs of the MP01 exercise published by the HTACG, mapped the pivotal randomised HIMALAYA trial against each, and searched ClinicalTrials.gov and PubMed for comparator trials. Each PICO was classified, by a pre-specified algorithm, as addressable by direct comparison, potentially addressable by indirect treatment comparison (ITC), or not addressable, anchored to the evidence existing at submission.
RESULTS: The thirteen PICOs shared one intervention but spanned heterogeneous, overlapping populations (full label; Child-Pugh A; Child-Pugh B/C; locoregional-eligible; ECOG-restricted) and seven comparators, expanding to fifteen distinct analyses. Because HIMALAYA was confined to Child-Pugh A, locoregional-ineligible patients, only two PICOs were directly addressable (durvalumab plus tremelimumab; sorafenib). A further subset was potentially addressable by anchored ITC — REFLECT and IMbrave150 share sorafenib with HIMALAYA — but residual population differences and incomplete outcome and subgroup reporting limited interpretability. The Child-Pugh B/C and locoregional-eligible PICOs could not be addressed, those populations being excluded from the trial. Thus, despite a large randomised trial, high-quality evidence for one population and comparator did not translate into feasibility across the scoped set. The feasibility gap is therefore not confined to single-arm products: it arises whenever the scoped populations and comparators outrun the available trial base.
CONCLUSIONS: A randomised pivotal trial is necessary but not sufficient for a feasible JCA. When a broad, comparator-heavy scope outruns the trial population, most questions remain answerable only by indirect comparisons of limited interpretability — an argument for evidence-feasibility screening before the PICO scope is finalised.
Conference/Value in Health Info
2026-11, ISPOR Europe 2026, Vienna, Austria
Value in Health, Volume 29, Issue 12S
Code
HPR41
Topic
Health Policy & Regulatory
Disease
No Additional Disease & Conditions/Specialized Treatment Areas