Article
Author(s):
Brian I. Rini, MD, FASCO, highlights key clinical trial updates from the genitourinary cancers landscape presented at the 2023 ASCO Annual Meeting, details the implications of the CONTACT-03 data, expands on the long-term data from the phase 3 KEYNOTE-426 trial in clear cell RCC, discusses findings from the phase 3 TALAPRO-2 trial in patients with metastatic castration-resistant prostate cancer, and more.
Although the phase 3 CONTACT-03 trial (NCT04338269) evaluating atezolizumab (Tecentriq) plus cabozantinib (Cabometyx) compared with cabozantinib in patients with advanced renal cell carcinoma (RCC) who received a prior immune checkpoint inhibitor proved to be a negative study, the findings provided crucial information regarding immunotherapy rechallenge in this patient population, according to Brian I. Rini, MD, FASCO.1
“This probably, at least for now, puts an end to the practice of giving sequential immune therapy [in advanced RCC],” Rini said. “There is another big trial pending [the phase 3 TiNivo-2 trial (NCT04987203) of tivozanib [Fotivda] plus nivolumab [Opdivo)] vs tivozanib monotherapy], and there will most likely be some other datasets, so the story [of sequential immunotherapy] isn't completely told, but [results from CONTACT-03] are a cautionary tale.”
In an interview with OncLive® News Network: On Location during the meeting, Rini highlighted key clinical trial updates from the genitourinary (GU) cancers landscape presented at the 2023 ASCO Annual Meeting, detailed the implications of the CONTACT-03 data, expanded on the long-term data from the phase 3 KEYNOTE-426 trial (NCT02853331) in clear cell RCC, discussed findings from the phase 3 TALAPRO-2 trial (NCT03395197) in patients with metastatic castration-resistant prostate cancer (mCRPC), and more.
Rini is the Ingram Professor of Medicine in the Division of Hematology Oncology and chief of Clinical Trials at Vanderbilt-Ingram Cancer Center in Nashville, Tennessee.
Rini: In kidney cancer, the most impactful data [were from] CONTACT-03. Although it was a negative [trial], these were important negative data. As you may know, that was a randomized trial in [patients with] immune-oncology [IO]–refractory kidney cancer, which is increasingly common these days, [evaluating] atezolizumab plus cabozantinib vs cabozantinib alone. TKI monotherapy would be the standard of care [in the IO-refractory setting], and the question was, does continuing immune therapy matter? There was no difference in progression-free survival [PFS], overall survival [OS], or overall response rate [ORR], and there were deaths [due to toxicity] in the combination arm.
Pembrolizumab and axitinib was the first IO/TKI [combination] to report out, so it has the longest follow-up, now with a minimum of 5 years and a median of 67 months. What stands out is that there are some clinical effects that persist, [including] an ORR advantage, a complete response [CR] advantage, and a PFS advantage. The OS signal holds up, although the hazard ratio is going up over time. What we're seeing is that a lot of the patients [who had received] sunitinib are getting salvaged by [multiple] lines of therapy, including immunotherapy. This is still a subset of patients, and you would still rather be on the pembrolizumab/axitinib OS curve, so to speak, since OS is still greater at any of the landmarks.
As these data are maturing, we're seeing a tail to the curves, as about 20% of patients are progression free at 5 years. I use these data talking to patients. We can't cure kidney cancer, but we can have long-term disease control. We need to do better, of course, but it's nice to finally start seeing some of these long-term data.
It's very similar [to KEYNOTE-426], right? It's a year behind with a 4-year minimum follow-up, but a very similar story. There are strong ORRs, a PFS advantage, and an OS signal that's still there but not quite as strong as it was at the first analysis. These IO/TKI regimens have the greatest effect early, both in tumor shrinkage in the immune component, and then over time, you see a wash out of that effect. The data sets are quite similar.
The PARP inhibitor story in prostate cancer has been very active in the last 1 to 2 years. As we sit back and look at the data, very consistent themes are emerging in that patients with BRCA-mutated [disease] benefit wildly [in terms of] PFS and OS, and the TALAPRO-2 data show that. When getting into [patients with] non-BRCA homologous recombination repair [HRR] mutations, there's benefit—not as much as those with BRCA mutations, but better than non-HRR–mutated [patients].
Importantly, what they showed is that [HRR mutations] are not all the same. There are HRR mutants that are more sensitive, and then there are some that probably aren't sensitive at all. As these data mature and as the field matures, it won't just be HRR [mutated vs unmutated]. That's not granular enough. We'll dig down into individual mutations.
PARP inhibitors are important in prostate cancer, and they are really active in patients with BRCA mutations. There are trials moving them into earlier [settings], which makes all the sense in the world, especially for patients with BRCAmutations. The data [from TALAPRO-2] are pretty consistent with other datasets.
Prostate cancer has a number of novel mechanisms: lutetiums, PARP inhibitors, chemotherapy, etc. Therefore, it's very natural as oncologists to try to combine these. We're in the early days of combining novel drugs and figuring out when [to give them], patient selection, and biomarkers. Now we have some have tools to play with. Notably, circulating tumor DNA [ctDNA] is an emerging technology. When we're sitting here in 5 or 10 years, ctDNA is going to be routine, and we're going to be doing it in a lot of different settings to inform treatment.
We're not quite there [yet], at least in the GU diseases I treat, but the technology, the detection methods, and the quantification have advanced. It is going to be routine [in the future], much like genomic testing is routine now, whereas not quite so 10 years ago.
Belzutifan is a HIF2α inhibitor, and it's approved for [patients with] von Hippel-Lindau disease [who require treatment for associated RCC and other select cancers]. In sporadic kidney cancer, belzutifan is being tested in a number of settings—almost every setting you could [imagine]. The combinations with VEGF [inhibition] have produced some interesting results. They are both in the same pathway, but there is some mechanistic rationale as to why inhibiting both would be better. Although we only have single-arm data—we don't yet have randomized data—the response rates for those combinations are robust, including data presented [from KEYMAKER-U03B], where there was a [50%] ORR and that's ostensibly higher than a TKI alone, which might be as high as 40%, and this was in an IO-refractory setting.
However, we've been burned by shiny, new ORRs in single-arm trials before, so we need the randomized data. There is a large, [phase 3], randomized trial [NCT04586231] of that combination vs cabozantinib in the IO-refractory setting. That trial is far along in accrual, so we'll get the randomized data soon, but [this combination] has been very promising.
The biggest data here was the large, randomized [phase 3 THOR trial (NCT03390504)] of erdafitinib [Balversa] compared with chemotherapy in [patients with FGFR-altered,] refractory [urothelial cancer]. Erdafitinib received accelerated approval on based on ORR [observed in the phase 2 BLC2001 trial (NCT02365597)], and then [THOR] was the confirmatory trial. It showed a quite strong OS signal with a hazard ratio of [0.64], so this is clearly an active drug.7
[Erdafitinib] is now working its way into the treatment pattern of bladder cancer in clinical trials in the combinations and other settings. Of course, there is toxicity and a learning curve. It is oral, but it has unique toxicities; they're not what we're used to in terms of chemotherapy. The comfort level around managing toxicity is important, but the results were very consistent with the phase 2 data, which you sometimes don't see. In phase 3, things sometimes drop down a notch, but with these numbers for PFS and OS were almost identical [to the phase 2 data].