PFS can reliably predict OS in immunotherapy trials: NHMRC study


By Michael Woodhead

9 Sep 2020

The surrogate outcome of progression free survival (PFS) can reliably predict overall survival (OS) in trials of immunotherapy for solid cancers, an Australian study has shown.

Researchers at the NHMRC Clinical Trials Centre at the University of Sydney have developed a prediction model that uses 6-month PFS from the immunotherapy arm of randomised clinical trials (RCTs) to estimate 12-month OS for various different tumour types with a good degree of performance.

But objective response rate  (ORR)  proved to be a poor predictor of 12-month OS, according to their systematic review and meta-analysis using data from 60 published immunotherapy RCTs.

Writing in JAMA Network Open, Dr Peey-Sei Kok and colleagues said the exponential increase in trials of immune checkpoint inhibitors (ICI) – more than 2000 RCTs since 2006 – had raised concerns about the validity of surrogate outcomes such as PFS as a basis of approval and adoption of novel therapies into clinical practice.

And to obtain robust OS outcomes requires phase 3 trials with large sample sizes that take years to produce results, they noted.

To investigate whether phase 2 trial surrogate outcomes could be used to predict OS, they extracted data from RCTs of anti–PD-1, anti–PD-L1, or anti–CTLA-4 ICIs in 17,000 patients with unresectable locally advanced or metastatic solid tumours, excluding trials in neoadjuvant, adjuvant, or maintenance settings.

The estimation model they used for 12-month OS using 6-month PFS was: (1.06 × PFS6) + 0.16 + (0.04 × melanoma) − (0.03 × NSCLC) + (0 × other tumours).

When applied on a validation data set, there was good calibration between the 6-month PFS model estimated vs observed 12-month OS (r = 0.89; Brier score, 0.008). The reliability was consistent across all ICI subgroups, and was particularly good for ICI-only arms (r = 0.94).

However calibration was poor for the ORR model (r = 0.47; Brier score, 0.03), although stronger for ICI-only arms (r = 0.80).

The study investigators said they believed their model could have widespread application as it was based on data from a variety of ICI agents, from first- to fourth-line treatment, in 12 different tumour types and covered both positive and negative trials.

“This study has several important implications. The ability to accurately estimate for OS from observed 6-month PFS using our model may allow smaller future ICI studies with shorter follow-up and earlier results saving resources,” they wrote.

It may also help better select and prioritise ICI agents, either as monotherapy or in combination with other treatments, for testing in phase 3 RCTs and reduce failure rates, they suggested.

The model may also assist regulators, policy makers, and funding bodies to better assess treatment efficacy based on 6-month PFS results where phase 3 RCTs are not feasible due to limitations such as small sample size in rare cancers.

“Importantly, our work may serve as a platform for future estimation models that incorporate multiple surrogate end points, including molecular surrogates and pharmacodynamic markers for a more sophisticated estimation of OS with ICI therapy,” they concluded.

Already a member?

Login to keep reading.

Email me a login link