Sussing Out Adjuvant ICI's Utility in RCC

Does adjuvant immune checkpoint inhibition (ICI) truly benefit patients with localized renal cell carcinoma (RCC) in terms of curbing recurrence? The picture is not entirely clear: Several recent phase III trials, all with negative results, have cast some doubt on this approach.

Following partial or complete nephrectomy, patients with stage II/III tumors “face a substantial risk of relapse,” said Robert Motzer, MD, of Memorial Sloan Kettering Cancer Center in New York, NY. Given the efficacy of combinations such as ipilimumab (Yervoy) and nivolumab (Opdivo)—both Bristol Myers Squibb drugs—in patients with metastatic RCC, “there's been a high level of enthusiasm for studying these agents as adjuvant therapy.”

“The story with ICI and advanced kidney cancer has been pretty straightforward so far,” agreed Thomas Powles, MBBS, of Barts Cancer Centre in London, UK. “We've had terrific combination trials that improved outcomes, with complete responses seen.”

Powles added that positive data from the phase III KEYNOTE-564 study led to the FDA greenlighting adjuvant pembrolizumab (Keytruda; Merck) in November 2021 for patients with early-stage RCC at an increased risk of recurrence. A 30-month follow-up analysis confirmed the PD-1 inhibitor's disease-free survival (DFS) advantage in this population, with overall survival “trending in the right direction too” (Lancet Oncol 2022;23:1133–44).

At the ESMO 2022 Congress in September, however, findings from three other trials proved disappointing on the adjuvant front. In one, CheckMate-914, Motzer reported that among 816 patients randomly assigned to receive ipilimumab–nivolumab or placebo, there was no difference in DFS between study arms.

Similarly, in the IMmotion010 trial, which assessed adjuvant atezolizumab (Tecentriq; Genentech) versus placebo in 778 patients, “we unfortunately found no evidence that atezolizumab reduced recurrence risk,” said Axel Bex, MD, PhD, of the Royal Free Hospital, also in London.

Meanwhile, the PROSPER study, albeit with a different design—patients either received neoadjuvant as well as adjuvant nivolumab or underwent standard observation after surgery—ended up being halted early for inefficacy. “Our idea was to prime the immune system preoperatively with continued postoperative engagement,” explained Mohamad Allaf, MD, of Johns Hopkins University in Baltimore, MD. Among 805 evaluable patients, though, “we saw no separation of the [DFS] curves at a median follow-up of 16 months.”

With adjuvant ICI in operable RCC, “the story's become clouded, and the dashboard is flashing red,” Powles remarked. “We need to figure out why.”

To James Larkin, PhD, of London's The Royal Marsden, CheckMate-914 is similar enough in design to KEYNOTE-564 that the former's negative findings “came as a bit of a surprise.” One possible explanation may be that both CTLA4 and PD-1 inhibitors were deployed, he said.

In the ipilimumab–nivolumab arm, “only 57% of patients received all planned doses,” Larkin noted, “and there was a high discontinuation rate [43%] due to toxicity.” Nearly one quarter needed daily high-dose prednisolone to manage side effects, compared with 8% of KEYNOTE-564's patients, and he wondered if “this level of immunosuppression could have affected efficacy.”

Motzer agreed. Based on past studies of VEGF-targeting agents as adjuvant therapy for RCC, “we've found in this setting, patients don't tolerate toxicity to the same degree as those with metastatic disease,” he observed, “and physicians tend to stop rather than continue treatment.”

“Of course we can justify more toxicity in principle, but only if there's a clear upside,” Larkin added. “The bottom line here is we have to be better at selecting patients who do need CTLA4 in addition to PD-1 blockade.”

IMmotion010 being also comparable to KEYNOTE-564, the fact that “there wasn't a whiff of a signal” is perplexing, Powles said. “We know atezolizumab does have activity” in advanced RCC, “but it just doesn't seem to have performed in the adjuvant setting.” Most of IMmotion010's patients had clear-cell RCC, so “it's plausible,” he mused, that “PD-L1 inhibitors are simply less effective in this subtype.”

Regarding PROSPER, Powles pointed out that its population “included many more patients with low-risk disease.” As well, “those who didn't have surgery for whatever reason—perhaps they opted out after achieving a response with neoadjuvant nivolumab—were nonetheless counted as progressors, adding noise to the data.” Overall, “due to design and methodology issues, it's impossible to judge nivolumab's activity” or lack thereof in this trial.

If nothing else, these studies have yielded large data sets “that I hope we can start to mine,” Motzer said. “We need to do a deep dive for underlying biology clues to direct us in moving forward” with perioperative ICI for RCC.

“In these adjuvant trials, about half the patients won't relapse” and could face unnecessary harm from ICI, Larkin added. “We still don't have any clinically useful molecular markers to help us better identify those at risk who'd actually benefit from treatment.”

For now, physicians should “make patients aware that although adjuvant pembrolizumab does prolong DFS, this hasn't been replicated with other checkpoint inhibitors, creating some uncertainty,” Powles concluded. “To paraphrase Tolstoy, positive trials look much the same, characterized by robust design and effective drugs. But negative trials occur for a plethora of reasons; it's not always possible to pin failure down to a single factor.” –Alissa Poh