• Home/
  • Industry/
  • Listing/
  • Participants/
  • Public Release Docs/
  • 2021 10/
  • Pembrolizumab For Locally Advanced Or Metastatic Urothelial

Pembrolizumab for locally advanced or metastatic urothelial carcinoma, October 2021

Page last updated: 24 March 2022

Drug utilisation sub-committee (DUSC)

October 2021

Abstract

Purpose

Analysis of the predicted versus actual utilisation of pembrolizumab 24 months following its addition to the Pharmaceutical Benefits Scheme (PBS) for the treatment of urothelial carcinoma on 1 March 2019.

Data Source / methodology

PBS dispensing data for pembrolizumab was extracted from the PBS data maintained by the Department of Health, processed by Services Australia. This data was used to establish the number of prevalent and incident patients utilising pembrolizumab for urothelial carcinoma and time on therapy.

Key Findings

  • 483 prevalent patients were supplied pembrolizumab in its first year of listing, 680 prevalent patients in Year 2 and 564 prevalent patients in 2021 to the data cut-off date of 31 July 2021.
  • Since its listing, the utilisation of pembrolizumab has increased steadily with up to 50 first initiating (incident) patients per month and the number of prevalent patients was continuing to increase.
  • The patient numbers were substantially underestimated in the submission with a XXXXXX number in Year 1, XXXXXX in Year 2 and XXXXXX so far in Year 3 of listing.
  • The number of patients who are platinum therapy-ineligible is approximately 8% which is less than the disparity in the predicted versus actual utilisation.
  • The PBAC stated at the November 2017 meeting that the duration of pembrolizumab use may be longer than estimated. However the Kaplan-Meier analysis of treatment duration, both with breaks and without breaks, indicates that the median time on treatment for patients utilising pembrolizumab is 110 days or 101 days, respectively. This suggests that the duration of pembrolizumab treatment from the submission is overestimated.
  • The number of prescriptions were overestimated where actual numbers are XXXXXX XXXXXX in Years 1 and 3 despite greater patient numbers than predicted. In Year 2 there was a XXXXXX patient number than predicted however this resulted in only a XXXXXX prescription count. 

Full Report