Darolutamide/ADT Significantly Improves OS in Nonmetastatic CRPC, Irrespective of Crossover

The addition of darolutamide (Nubeqa) to androgen deprivation therapy (ADT) significantly improved overall survival (OS) over ADT alone in patients with nonmetastatic castration-resistant prostate cancer (CRPC), and crossover from the control arm to the investigative arm only had a small impact on this benefit, according to Neal Shore, MD, FACS.

Data from the final analysis of the phase 3 ARAMIS trial (NCT02200614) showed that darolutamide resulted in a 31% reduction in the risk of death in this population (hazard ratio [HR], 0.69; 95% CI, 0.53-0.88; P = .003). Of the 554 patients who were randomized to placebo, 170 (30.7%) crossed over to receive darolutamide after the trial was unblinded.

The iterative parameter estimation (IPE) and rank-preserving structural failure time (RPSFT), were performed as pre-planned sensitivity analyses to adjust for the treatment effect of crossover. A censoring at crossover analysis was also conducted. Results demonstrated that the impact of the crossover on OS was small. The HRs were 0.66 (95% CI 0.51-0.84; P < .001), 0.68 (95% CI 0.51-0.90; P = .007), and 0.59 (95% CI, 0.45-0.76; P <.0001), respectively.

“The consistency of the findings between the final OS analysis and the sensitivity analyses indicate that the crossover from placebo to darolutamide has minimal impact on the OS benefit [derived with] darolutamide vs placebo,” Shore said. “To me, these findings further indicate that darolutamide is an effective and well-tolerated androgen receptor pathway inhibitor [for use in the] early treatment of our patients with nmCRPC.”

In an interview with OncLive^®, Shore, the medical director of the Carolina Urologic Research Center, discussed the impact of crossover from placebo to darolutamide on OS benefit in patients with nmCRPC enrolled to the ARAMIS trial, and next steps for the androgen receptor inhibitor.

OncLive^®: Could you highlight some key points of the ARAMIS trial, and what led to the crossover analysis from placebo to darolutamide?

Shore: ARAMIS was a global phase 3 study of [patients with] nmCRPC. At the time that the study was designed, no approved treatments [were available] for patients with nmCRPC [who had] rising prostate-specific antigen [PSA], a castrate level of testosterone, and negative imaging on a conventional CT scan and technetium bone scan.

A little over 1500 patients were [enrolled], and [they underwent] a 2:1 randomization. In the [investigative] arm, patients received continuous androgen deprivation therapy [ADT] and darolutamide, given at 2 pills twice daily. In the control arm, [patients received] continued ADT and placebo [in the form of] 2 pills twice daily.

The [goal of the] primary analysis was to look at metastasis-free survival, which was a combination of radiographic progression and/or death. Most of the events were triggered by progression. It was a very successful analysis, and a paper [on those data] was published in the New England Journal of Medicine [NEJM].

We subsequently performed an OS analysis, where in the hierarchical ordering, OS was the first secondary end point. [As such,] if that was not met, we could not have proceeded. [However,] OS was met, and that led to a second [paper published in the] NEJM. Those data were presented in 2020 at the ASCO Annual Meeting, as well as at the ESMO Congress. In the presentation delivered at the 2021 Genitourinary Cancers Symposium, we examined the effect of the crossover of the placebo arm to darolutamide, and [whether] that [impacted] OS benefit within the trial.

Could you expand on what was analyzed in the crossover analysis?

Once we unblinded the study at the primary analysis, 170 of the 554 patients who were in the placebo arm were crossed over to receive darolutamide. We looked at concepts of 4 sensitivity analyses, 2 of which were planned in the study protocol. [One was] the RPSFT, and the other one was the IPE. Essentially, these were done to then recreate the Kaplan-Meier curves to see what would happen for [those in] the control arm that switched over, and what type of effect that would have on OS compared with the [investigative] arm.

Two other sensitivity analyses used censored data. In the OS-adjusted censoring analysis, patients in the placebo arm were censored at the time of crossover to darolutamide. Then, we did another sensitivity analysis using that censored data, [which is] called the inverse probability of censoring weighting method.

These [analyses] are a bit on the wonky academic side of statistics, but they are important—especially if you are a biostatistician. The observations in this censored dataset are inversely weighted according to a censoring probability predictor for baseline characteristics.

What were the key takeaways from this research?

At the end of the day, and very importantly, in the final analysis, darolutamide was associated with a significant improvement in OS compared with placebo, with a hazard ratio of 0.69. The sensitivity analyses showed that the crossover likely had only a small dilutive effect on OS.

In terms of overall safety, it remained very favorable. No unexpected findings [were observed] in either the darolutamide arm, or the patients who crossed over from the control arm to receive darolutamide.

What is the clinical significance of these findings?

This [research] is important because it continues to demonstrate the findings that we saw from the primary analysis…it’s basically saying that these are all consistent findings between the final OS and the sensitivity analyses of the crossover. The crossover had minimal impact on the OS benefit that we saw with darolutamide vs placebo. We were very happy to see that. [We] did this in-depth analysis because in other nmCRPC trials, there are differing degrees of crossover.

What are the next steps for darolutamide?

Throughout most of the world, darolutamide has approval [for use] in the nmCRPC population. I certainly believe this to be a very powerful androgen receptor pathway inhibitor, much like other approved agents like enzalutamide [Xtandi] or apalutamide [Erleada].

[Additionally,] I look forward to [seeing additional data with] darolutamide that will expand its approval for [use in] patients with mCRPC, and for those with mCSPC. It is very important for my colleagues to continue to get real-world experience using all these androgen receptor pathway inhibitors [in practice]. They all have an excellent effect on delaying radiographic progression, and they have all showed demonstrable benefit [in terms pf] OS in the nmCRPC indication. There are broader indications for enzalutamide and apalutamide; I look forward to darolutamide expanding its indications.

Reference

1. Shore N, Fizazi K, Tammela T, et al. Analysis of the effect of crossover from placebo (PBO) to darolutamide (DARO) on overall survival (OS) benefit in the ARAMIS Trial. J Clin Oncol. 2021;39(6):240-240. doi:10.1200/JCO.2021.39.6_suppl.240