The 2022 annual congress of the European Society of Medical Oncology (ESMO), held September 9-13, featured the results of three eagerly anticipated Phase 3 trials of adjuvant immunotherapy in renal cell carcinoma (RCC). All three trials failed to meet their primary endpoint, most disappointingly PROSPER RCC,*1 which took the novel approach of adding neoadjuvant therapy on one arm. Shortly after presentation of PROSPER RCC, Thomas Powles, MBBS, MD, (Barts Cancer Centre, London, UK) and Brian Rini, MD (Vanderbilt University, Nashville, TN) met up with the presenter, Mohamad E. Allaf, MD, Urologist-in-Chief of the Johns Hopkins Hospital, Baltimore and Professor of Urology and Oncology, to review the trial and discuss its implications for clinical practice (Uromigos podcast episode 191).
The Phase III RandOmized Study Comparing PERioperative nivolumab versus observation in patients with Renal Cell Carcinoma (PROSPER RCC trial, also known as ECOG-ACRIN EA8143 (NCT03055013), was a randomized open-label trial that compared neoadjuvant nivolumab prior to nephrectomy followed by adjuvant nivolumab versus surgery alone in patients with high-risk RCC.2,3 The premise of the trial, Dr Allaf explained, was to leverage neoadjuvant therapy and the promise of potentially eliciting a better and stronger immune response by priming the immune system, while also maximizing the immune effect by continuing immune engagement with the adjuvant doses. The study was conducted by the ECOG-ACRIN Cancer Research Group and funded by the National Cancer Institute and the Canadian Cancer Society. Expectations for PROSPER RCC were high following the positive results with pembrolizumab in the adjuvant setting in the KEYNOTE-564 trial.4
Between February 2017 and June 2021, 404 patients were randomized to perioperative nivolumab 480 mg IV administered every 4 weeks with 1 dose prior to surgery (partial or radical nephrectomy) followed by 9 adjuvant doses and 415 to surgery alone followed by surveillance without a placebo. Clinical stage at enrollment was 53% cT2, 47% cT3-4, 17% cN1, and 4% cM1; 83% of patients had clear cell RCC. Select oligometastatic disease was permitted if the patient could be rendered “no evidence of disease” within 12 weeks of surgery. Baseline tumor biopsy was required only in patients on the nivolumab arm.
The trial was stopped early by the Data and Safety Monitoring Committee (DSMC) due to futility. After a median follow-up of 16 months, the primary endpoint, recurrence free survival (RFS), was similar on the two study arms (HR 0.97; 95% CI: [0.74 – 1.28]; P1-sided = 0.43). Median RFS was not reached. Data for overall survival (OS), a secondary endpoint, were not mature at the time of analysis but showed no statistical difference between the two arms. Withdrawal rates were about 12% with the nivolumab and surgery-alone arms and 20% of patients treated with nivolumab experienced at least one Grade 3-4 drug-related adverse event.
During his presentation, Dr Allaf expressed pride in the involvement of patient advocates in the trial design. Initially, the patients were not enthusiastic about a placebo arm. Their other major concern was that they did not want to get nivolumab if they did not have RCC. Enrolling patients into PROSPER RCC was a tour de force, Dr Allaf said. “It took multiple years and there were a lot of logistical hurdles,” he admitted. “Initially we were a little slow. It was a unique trial design, urology-centered, and we needed urologists who would usually put patients on surgery within a week or two after seeing them to pause, think about a trial, and change their thought process and decision making,” he explained. Some patients were also concerned about delaying their treatment. “But with the help of urologists, teamed with oncologists and the Society of Urologic Oncology Clinical Trials Consortium (SUO-CTC), we were able to accrue fully,” he said.
Dr Rini wondered about the rationale for including patients’ non-clear cell RCC (nccRCC), observing that since half the patients did not have a biopsy before surgery, there was no choice but to include these patients. Dr Allaf explained that the number of nccRCC patients was capped, but that the threshold for stopping accrual of these patients was never reached. It will be an interesting subset analysis to look at that population alone, he suggested.
Dr Rini noted that on the nivolumab arm of the trial, as many as 47 patients did not receive the allocated intervention, including 20 who were ineligible and, surprisingly, 22 who refused before starting protocol therapies. “You would have thought that if they were randomized to nivolumab arm, they would be happy,” he remarked. “This was an early interim analysis and we need to go and look at these ineligible patients and dropouts, to see exactly what happened,” Dr Allaf replied. There were also 45 patients who did not receive surgery on the nivolumab arm compared with 28 on the observation arm, and 51 on the nivolumab arm who did not get the neoadjuvant dose and then, “surprisingly to me,” 90 who did not receive any adjuvant therapy, he recalled. “It’s not known whether these were patients who had a T1 low-grade tumor and were told by someone to get out of the trial and not take a toxic drug if they were cured, or people with high-grade disease who really needed it,” Dr Allaf said.
Dr Powles wondered how much that influenced the results, because they were “a bit disappointing,” he said. Many people were expecting positive results, similar to those of the melanoma trial, SWOG S1801, presented earlier at the ESMO congress, in which neoadjuvant therapy with pembrolizumab in patients with high-risk resectable disease was found to improve event-free survival compared with adjuvant pembrolizumab.5
Dr Powles wondered why the neoadjuvant approach was not effective in this disease. “Had we done it differently with different methods, and I’m not sure what they would be, could we get a positive study that way? Do you think we’ve answered that question definitively? Is this the end of neoadjuvant studies?” he asked. Dr Allaf hopes that “we are at the beginning of more investigations in the neoadjuvant space.” He pointed out that when PROSPER RCC was designed, much less was known about neoadjuvant therapy, particularly about the toxicity of PD-1 inhibitors.
He cited the ASSURE trial ((ECOG-ACRIN E2805), which started in 2006 and in which over 55% of patients treated with a VEGF receptor tyrosine kinase inhibitor (sunitinib or sorafenib) had grade ≥3 adverse effects.6 “Now we know how to take care of those,” he noted. He also suggested that the dosing of nivolumab in PROSPER RCC might have been too low. Initially he had favored a trial with three doses of nivolumab on the neoadjuvant arm, but that design had been rejected. “It may be difficult to achieve CR with one neoadjuvant cycle,” Dr Powles agreed. There are also issues of the study’s being open label and not having a placebo arm, he added. Dr Allaf defended the trial as a real world experience. “Unlike a well-curated adjuvant trial where everybody gets surgery and then you pick the ones to get it, in a neoadjuvant trial, how many aren’t going to get the surgery? How many are going to refuse therapy or for whatever reason? I think we’re going to learn a lot from this,” he predicted.
Dr Powles was concerned that the trial definition of an event was a progression event, which is standard, but also not having surgery. “That is obviously challenging, because many of those patients who didn’t have surgery did not have it because their cancer was getting worse; it was the reverse,” he noted. “How do you address that bias and do you think that this is a robust surrogate for OS?” Dr Allaf replied that they don’t know what happened to the patients who did not have surgery. “I doubt that the patients who responded dropped off because they were doing well, they may have enrolled in the trial and had surgery locally and it’s not recorded. We need to look into that further,” he commented. He added that they had re-examined the data “and our statisticians are very comfortable saying that it met futility, at least as designed,” he said. “The question is, where do we go from here?”
Dr Rini observed that four studies with adjuvant immune checkpoint inhibitor therapy in RCC have now reported, three negative (CheckMate-914,7 IMmotion010,8 and PROSPER RCC1) and one positive (KEYNOTE-5645). Dr Powles would now like to see a biomarker study examine whether it would be possible to identify patients who would benefit from therapy. “I feel that we are reaching a plateau with immune therapy,” he stated. “I’m not sure that pursuing it any further in this disease is necessarily the right thing to do.” He noted that the results of the LITESPARK022 trial (NCT05239728) with the hypoxia-inducible factor (HIF)-2α inhibitor belzutifan added to pembrolizumab,9 and RAMPART (NCT03288532), comparing durvalumab (a PD-L1 inhibitor) plus tremelimumab (a CTLA-4 inhibitor) with durvalumab monotherapy,10 are pending.
“I suspect the PROSPER RCC study had some flaws and that under those circumstances, there is scope for a second, really clean PD-1 inhibitor study. But I don’t think that would move the field forward,” he said. “I would like to see a randomized phase 2 of three cycles of neoadjuvant therapy rather than one, but that wouldn’t be top priority,” he added. “I think we should move the field forward with new-generation drugs and new-generation biomarkers. Now that we have hit this plateau, I don’t think we can go much further with VEGF-targeted therapy or with single agent immune checkpoint inhibition.
Linda Brookes, MSc is a freelance medical writer/editor based in New York and London.
*PROSPER RCC” to avoid confusion with the PROSPER trial (A multinational, phase 3, randomized, double-blind, placebo-controlled, efficacy and safety study of enzalutamide in patients with nonmetastatic castration-resistant prostate cancer; NCT02003924).