Characteristics of cancer drug-indication pairs recommended for funding on the Australian Pharmaceutical Benefits Scheme (PBS), 2005-2025.

Abstract

e23039 Background: The cost of cancer drugs is substantial and increasing, intensifying scrutiny of regulatory and health technology assessment (HTA) processes to ensure that reimbursement decisions align with meaningful clinical benefit and value. Australia has a universal healthcare system in which medicines are assessed for federal funding on the Pharmaceutical Benefits Scheme following evaluation by the Pharmaceutical Benefits Advisory Committee (PBAC). No prior study has described HTA outcomes for cancer medicines in Australia. Methods: We conducted a retrospective cohort study of solid cancer drug–indication pairs submitted to the PBAC between July 2005 and July 2025. Submission characteristics, supporting clinical evidence, efficacy outcomes (overall survival (OS) and progression-free or disease-free survival (PFS)), and economic analyses were extracted. Hazard ratios for OS and PFS were meta-analysed. Multivariable logistic regression was used to assess factors associated with a funding recommendation. Results: Of 410 solid cancer drug–indication pairs submitted to the PBAC, 184 (44.9%) were recommended for funding. Recommended drugs were commonly supported by randomised (96.2%), phase III (87.5%), and unblinded (57.4%) trials, with overall survival (OS) as the most frequent (co-)primary endpoint (48.9%). Across recommended drugs, pooled hazard ratios indicated a 24% relative reduction in the hazard of death and a 36% reduction in the hazard of progression, corresponding to gains in mOS of 3.2 months (IQR 0.0–6.9) and mPFS of 4.1 months (IQR 1.8–9.6). Recommended drugs were more likely to be resubmissions (46.7% v 35.6%, p = 0.03), involve curative-intent (20.1% v 11.9%, p = 0.03) or first-line therapy (57.6% v 44.7%, p = 0.03), and present a comparator deemed appropriate by the PBAC (94.3% vs 85.9%, p = 0.01). mOS gains were modestly greater among recommended drugs (3.2 vs 2.3 months, p = 0.04). Recommended drugs were more frequently supported by cost-minimisation analyses (33.2% v 11.9%, p < 0.01) and lower ICER ranges ( p < 0.01). In multivariable logistic regression, no single clinical or economic variable independently predicted funding recommendations. Conclusions: Cancer drugs recommended for federal funding in Australia typically demonstrate statistically significant but modest survival benefits and are supported by late-phase randomised evidence. PBAC recommendations appear to reflect a deliberative HTA process incorporating comparative credibility, economic framing, and contextual considerations beyond efficacy or cost-effectiveness thresholds.