Review of cancer related surrogate outcomes used for Pharmaceutical Benefits Advisory Committee (PBAC) decision making

Page last updated: 8 September 2023


Surrogate outcome measures have an increasingly important role in cancer medicine research providing evidence to support registration and subsidy of new medicines. Where validated, surrogate outcome measures may allow researchers to extrapolate short-term trial results into long-term clinically relevant patient outcomes. In May 2022, the PBAC recommended a research project into the use of surrogate outcome measures in PBAC submissions for cancer medicines. The Department contracted this work to Monash University in September 2022.


The final Report, ‘A review of cancer related surrogate outcomes used for PBAC decision making’, is available below.



All PBAC Public Summary Documents (PSDs) for cancer medicines (excluding treatments for pain, nausea, or adverse events) considered between January 2012 and May 2022 were reviewed. Submissions (including resubmissions) were included for detailed data extraction if a surrogate outcome was used in lieu of overall survival (OS) or if the PSD stated that a surrogate outcome was relied upon for the clinical claim or PBAC decision. A narrative review was conducted to describe the use of surrogate outcomes in PBAC decisions for cancer medicines.

Final OS data, limited to Phase II/III clinical trials, was sought for trials where the PBAC had considered immature or interim OS data. These data were compared to the OS data considered by the PBAC.

A literature review was conducted to identify recent high-quality meta-analyses that assessed the validity of surrogate outcomes in cancer therapies.

Key Findings

The narrative review of PBAC PSDs for cancer medicines found that 50% (247 out of 498) of submissions (including resubmissions) for cancer medicines between January 2012 and May 2022, were based primarily on a surrogate outcome. These included 91 medicines for 22 broad cancer types.

Of the submissions that relied on a surrogate, 44% received a positive recommendation for PBS listing, 12% were deferred and 44% were not recommended. Of those not recommended by the PBAC, in 62% of cases this was due to immature OS data.

Of the submissions that relied on a surrogate, 65% presented OS data based on interim results, of which 41% have now published final results. Final OS results were generally consistent with the interim results and no submission with worse final OS results had received a positive recommendation for PBS listing.

In submissions that presented modelled economic evaluations, surrogate outcomes were used in 85% (140 out of 165) of cases. The cost-effectiveness was likely heavily dependent on the assumed relationship between the surrogate outcome and OS or other clinically meaningful outcome.

The literature review of surrogate measure validation studies presented mixed results with most studies showing low to moderate correlation between OS and the most common surrogates, progression-free survival (PFS) and objective/overall response rate. Issues identified with the validation studies included:

  • use of different criteria for assessing correlation between the surrogate and OS
  • inconsistent thresholds to establish that a surrogate is sufficiently validated
  • thresholds that may not be robust enough for health technology assessment and economic modelling purposes
  • unclear disease stage and type.

The Report found that it was difficult to characterise surrogate outcomes based on information in the PSDs and found limited use and mention of the PBAC’s guidance on surrogate outcomes (i.e., Appendix 5 of the PBAC Guidelines).

ESC and PBAC consideration

The Report was considered by the PBAC Economic Sub-committee (ESC) in June 2023, and the PBAC in July 2023. The PBAC considered that the Report was informative for health technology assessment and recommended that the Report and PBAC consideration should be published on the PBS website and provided to relevant stakeholders.

The PBAC Minutes for this item are available below.