ACCEPTANCE OF SURROGATE ENDPOINTS IN ONCOLOGY WITHIN EARLY BENEFIT ASSESSMENT IN GERMANY: AND YES, THERE ARE EXCEPTIONS
Author(s)
Sophie Striebinger, MSc1, Charalabos-Markos Dintsios, MA, MPH, MSc, RPh, PhD2.
1Institute for Health Economics and Clinical Epidemiology, University of Cologne, Medical Faculty, Cologne, Germany, 2Head of Strategy and Operations Market Access, Bayer Vital GmbH, Leverkusen, Germany.
1Institute for Health Economics and Clinical Epidemiology, University of Cologne, Medical Faculty, Cologne, Germany, 2Head of Strategy and Operations Market Access, Bayer Vital GmbH, Leverkusen, Germany.
OBJECTIVES: The acceptance of surrogate endpoints (SE) in oncology has become an important issue within the German early benefit assessment. However, almost all SE are not accepted being considered as insufficiently validated. There exists only a very limited number of exceptional cases, in which SE were accepted and served for determining an added benefit. We aimed at analyzing specific characteristics and underlying determinants of these cases.
METHODS: Cases with accepted SE in oncology from 2011 until April 2026 were identified and subsequently clustered according to their type. Descriptive and agreement statistics (dichotomous and ordinal Fleiss’ and Cohen’s kappa) were applied to check for SE acceptance and agreement between SE-specific (i) manufacturers’ claims, (ii) IQWiG assessments and (iii) FJC appraisals.
RESULTS: Thirty-four cases (8 orphan drugs) with accepted SE over a broad spectrum of oncological indications and mainly in the curative context could be identified and clustered in the following 3 categories: 7 event-free survival (EFS), 20 disease-free survival (DFS) and recurrence-related endpoints, and 7 exceptional cases. FJC granted a SE-specific added benefit in all cases which varied from non-quantifiable to major (IQWiG recommended in 6 cases no added benefit). A fair dichotomous agreement between manufacturers and IQWiG (kappa 0.294; n.s.) was achieved whereas manufacturers and FJC agreed 100%. The slight ordinal agreement between manufacturers and FJC (kappa 0.178; n.s.) became fair and statistically significant when weighted (linear kappa 0.234; 0.047 - 0.421 & quadratic kappa 0.261; 0.08 - 0.514). The ordinal agreement between manufacturers and IQWiG remained fair irrespective of weighting whereas only a slight agreement was achieved between IQWiG and FJC. Fleiss’ agreement between three raters (kappa 0.157; n.s.) was slight.
CONCLUSIONS: By identifying recurring patterns among the cases, meaning predominantly a curative therapeutic context, acceptance of SE could be characterized. Only a slight or fair agreement was reached between the raters.
METHODS: Cases with accepted SE in oncology from 2011 until April 2026 were identified and subsequently clustered according to their type. Descriptive and agreement statistics (dichotomous and ordinal Fleiss’ and Cohen’s kappa) were applied to check for SE acceptance and agreement between SE-specific (i) manufacturers’ claims, (ii) IQWiG assessments and (iii) FJC appraisals.
RESULTS: Thirty-four cases (8 orphan drugs) with accepted SE over a broad spectrum of oncological indications and mainly in the curative context could be identified and clustered in the following 3 categories: 7 event-free survival (EFS), 20 disease-free survival (DFS) and recurrence-related endpoints, and 7 exceptional cases. FJC granted a SE-specific added benefit in all cases which varied from non-quantifiable to major (IQWiG recommended in 6 cases no added benefit). A fair dichotomous agreement between manufacturers and IQWiG (kappa 0.294; n.s.) was achieved whereas manufacturers and FJC agreed 100%. The slight ordinal agreement between manufacturers and FJC (kappa 0.178; n.s.) became fair and statistically significant when weighted (linear kappa 0.234; 0.047 - 0.421 & quadratic kappa 0.261; 0.08 - 0.514). The ordinal agreement between manufacturers and IQWiG remained fair irrespective of weighting whereas only a slight agreement was achieved between IQWiG and FJC. Fleiss’ agreement between three raters (kappa 0.157; n.s.) was slight.
CONCLUSIONS: By identifying recurring patterns among the cases, meaning predominantly a curative therapeutic context, acceptance of SE could be characterized. Only a slight or fair agreement was reached between the raters.
Conference/Value in Health Info
2026-11, ISPOR Europe 2026, Vienna, Austria
Value in Health, Volume 29, Issue 12S
Code
CO158
Topic
Clinical Outcomes, Health Policy & Regulatory, Health Technology Assessment
Topic Subcategory
Clinical Outcomes Assessment, Comparative Effectiveness or Efficacy, Relating Intermediate to Long-term Outcomes
Disease
No Additional Disease & Conditions/Specialized Treatment Areas, Oncology