OBJECTIVES

This study aimed to analyze oncology-related health technology assessments (HTAs) in Germany in terms of endpoint evaluations and their association with G-BA’s assessment.

METHODS

For all oncology-related HTAs (10/2011-03/2019), information on submission characteristics was collected (INGRESS HTA databank). Multivariable logistic regression was performed, with HTA assessment by G‑BA (benefit/no benefit) as an outcome and significant improvements with respect to evaluated endpoints [overall survival (OS), progression-free survival (PFS), response rate (RR), quality of life (QOL), serious adverse events (SAE) and therapy discontinuation (TD)] as main independent variables. Control variables included characteristics of presented evidence [presence of randomized controlled trials (RCT), indirect treatment comparison (ITC), and patient preferences, mutation-specific assessment and real-world evidence] as well as general characteristics of the submitted technology (orphan status, treatment line and submission year).

RESULTS

153 HTAs were included, most of them completed within recent years (79.1% from 2015 onwards), and the majority dealt with treatments for blood and bone marrow cancers (24.2%), lung cancer (16.3%) and skin cancer (14.3%). 73.9% of HTAs were associated with a positive recommendation. Conditional on the above-mentioned control variables, showing evidence on improvements in OS and QOL significantly increased the probability of a positive decision (OR=6.1, p<0.01; OR=9.0, p<0.01; respectively). Instead, improvements in PFS (OR=0.8; p=0.66), RR (OR=2.5; p=0.17), SAEs (OR=1.0; p=0.97) and TD (OR=1.0; p=0.96) were not significantly associated with the probability of a positive assessment. These results were robust to different model specifications; a probit regression confirmed the significant effects of OS and QOL and insignificant effects of PFS, RR, SAE and TD.

CONCLUSIONS

For oncology-related HTAs, demonstrating improvements in OS and QOL were found to increase the probability of a positive G-BA assessment. No further contributing factors as per evaluated endpoints could be found, with the caveat that HTA decisions could be driven by other undetected factors.