ASCO 2022: Assessing Intermediate Clinical Endpoints (ICE) as Potential Surrogates for Overall Survival (OS) in Men with Metastatic Hormone-Sensitive Prostate Cancer (mHSPC).

( Given the often long natural history of prostate cancer, randomized controlled trials in the hormone-sensitive setting can take many years to report overall survival (OS) results. It has therefore been of interest to determine if other clinical outcomes that can be measured earlier in the disease course could serve as surrogates for the gold standard outcome of overall survival. The international Intermediate Clinical Endpoints in Cancer of the Prostate (ICECaP) working group has previously published work establishing metastasis-free survival as a strong surrogate of overall survival in intermediate and high-risk localized prostate cancers1. This endpoint is being prospectively evaluated as part of the ongoing phase 3 PROTEUS clinical trial (NCT03767244) evaluating neoadjuvant apalutamide therapy and testosterone suppression prior to surgery in high-risk localized prostate cancer. How intermediate clinical endpoints (ICE) can be applied in more advanced prostate cancer as surrogates of overall survival remains to be fully established.

In this presentation of work by the STOPCAP M1 consortium funded by the Prostate Cancer Foundation ( and Prostate Cancer UK (, Dr. Halabi presented new work testing whether radiographic progression free survival (rPFS) by CT and bone scan and clinical progression free survival (cPFS) are intermediate clinical endpoints (ICE) for overall survival in metastatic hormone-sensitive prostate cancer (mHSPC). First, she defined the different endpoints. Overall survival, the gold standard, is defined as the date of randomization on trial to date of death from any cause. rPFS is defined as either progression by CT or bone scan or death. cPFS is defined as either progression by CT or bone scan, death, treatment switch, or progressive symptoms.

To analyze the relationship between ICE and OS, data analysis was conducted under two conditions. The first is whether the surrogate endpoints (rPFS and cPFS) are correlated at the trial and patient level, and the second is whether the treatment effect on surrogate and OS endpoints are correlated). To test condition 1, Dr. Halabi and colleagues chose to compare 5-year OS versus 3-year rPFS and cPFS. 5-year OS is at the upper limit of the median survival of mHSPC trials, and thus would allow for the inclusion of mature outcomes data. 3-year cPFS and rPFS were felt to be clinically important and within a time-frame that would expedite clinical trial readout.

A total of 51 randomized phase 2 or phase 3 trials that completed accrual of 14,695 patients after January 1992 were considered for this study so long as they did not have a primary endpoint of safety, quality of life or feasibility. Of these 51 trials, 9 trials (13 clinical comparisons, 8,612 patients) had data that was able to be collected for this study. The 9 trials included are shown below.

STOPCAP M1-0.jpg

The clinical characteristics of the patients included in this study are shown below.

STOPCAP M1-1.jpg

Overall across the trials, a total of 5,377 deaths were recorded, 74% of which were due to prostate cancer. The median OS was 4.1 years, the median rPFS was 2.3 years, and the median cPFS was 2.1 years.

For condition 1, the investigators determined that the tested ICEs and OS are correlated. Specifically, at the trial level, R2 for rPS and OS was 0.78 (0.43-0.99) and similar for cPFS and OS. Importantly, of the 9 studies, three studies with median follow-up of less than 4 years were excluded from this analysis. At the patient level using Kendall’s Tau, both rPFS and cPFS were correlated with OS (0.83 and 0.84, respectively).

For condition 2, the investigators found that the treatment effects on rPFS and cPFS were correlated with OS across trials.

STOPCAP M1-2.jpg

STOPCAP M1-3.jpg

Dr. Halabi concluded that in this study of over 8500 patients and 9 randomized clinical trials (including 5 comparisons from the STAMPEDE platform study), rPFS and cPFS appear to be valid surrogate endpoints for overall survival for men with mHSPC. Other strengths of this study include the inclusion of individual patient level data. However, there are several limitations including the fact that disease volume and timing of metastatic disease (synchronous or metachronous) were not collected for older trials, and the definitions of rPFS and cPFS may have varied across trial protocols. Importantly, Dr. Halabi stated that these surrogate endpoints are not yet ready for primetime as endpoints in new phase 3 studies. More validation is needed using contemporary data with more potent androgen pathway inhibitors, sensitivity analyses by class of therapy, and an expanded repertoire of statistical analyses.

Presented by: Susan Halabi, PhD, FASCO, Duke University Medical Center, Durham, NC, Duke University, Durham, NC

Written by: Alok K. Tewari, MD, PhD, medical oncologist at the Dana-Farber Cancer Institute, @aloktewar on Twitter during the 2022 American Society of Clinical Oncology (ASCO) Annual Meeting, Chicago, IL, Fri, June 3 – Mon, June 7, 2022. 


  1. Wanling Xie, Meredith Regan, Marc Guyse, et al. “Metastasis-Free Survival Is a Strong Surrogate of Overall Survival in Localized Prostate Cancer.” Journal of Clinical Oncology. 2017; 35(27):3097-3104.
email news signup