乌鸦传媒

Sunitinib (Sutent) received FDA approval for first-line metastatic renal cell carcinoma (RCC) over a decade ago, but now three new combinations using immune checkpoint inhibitors have toppled the former standard and sparked a debate over which combination clinicians should gravitate toward, and for whom.

"There are different ways of looking at this problem," said Thomas Powles, MD, of the Barts Cancer Institute in London. "It's a good problem to have -- that's a good place to start."

In interim analyses of two phase III trials presented earlier this month at the Genitourinary Cancers Symposium (GuCS), axitinib (Inlyta) plus PD-1/PD-L1 checkpoint inhibition improved response rates and delayed disease progression over single-agent sunitinib, and across all risk groups.

"The majority of patients with advanced clear cell renal cell carcinoma will be eligible to receive this combination," said discussant Lori Wood, MD, of Dalhousie University in Nova Scotia, Canada.

These are "exciting times," she said, noting that importantly, axitinib should be used and not another VEGF tyrosine kinase inhibitor (TKI).

In JAVELIN Renal 101, patients were randomized to axitinib plus the PD-L1 inhibitor avelumab (Bavencio) or sunitinib. Overall response rates (ORRs) were 51.4% and 25.7%, respectively, and median progression-free survival (PFS) was 13.8 months in the axitinib-avelumab group compared to 8.4 months with sunitinib (HR 0.61, 95% CI 0.47-0.79, P<0.001).

This PFS benefit was seen across all International Metastatic Renal Cell Carcinoma Database Consortium (IMDC) risk groups:

• Favorable: HR 0.54, 95% CI 0.32-0.91

• Intermediate: HR 0.74, 95% CI 0.57-0.95

• Poor: HR 0.57, 95% CI 0.38-0.88

Overall survival (OS) data were still immature (HR 0.78, 95% CI 0.55-1.08, P=0.14).

But in the other trial, KEYNOTE-426, the combination of axitinib plus the PD-1 inhibitor pembrolizumab (Keytruda) demonstrated benefit in ORR and PFS over sunitinib, and also improved OS. At a median follow-up of 12.8 months, the estimated 1-year OS rate was 89.9% in the investigational arm versus 78.3% in the control group (HR 0.53, 95% CI 0.38-0.74, P<0.0001).

And the OS benefit with the axitinib-pembrolizumab combination was seen across all IMDC risk groups:

• Favorable: HR 0.64, 95% CI 0.24-1.68

• Intermediate: HR 0.53, 95% CI 0.35-0.82

• Poor: HR 0.43, 95% CI 0.23-0.81

"We saw benefit across all risk groups," said Powles, who presented the KEYNOTE data at GuCS. "Pembrolizumab plus axitinib should be a new standard of care in this setting."

"To me, the immediate practice-changing part is that axitinib plus pembrolizumab should be considered a standard of care for patients with favorable-risk disease," Che-Kai Tsao, MD, of Icahn School of Medicine at Mount Sinai in New York City, told 乌鸦传媒.

"It's clearly going to alter the good-risk management of patients very quickly," said Robert Dreicer, MD, of the University of Virginia in Charlottesville.

Across the whole study population, ORR was 59.3% with axitinib-pembrolizumab versus 35.7% with single-agent sunitinib, and median PFS was 15.1 versus 11.1 months, respectively (HR 0.69, 95% CI 0.57-0.84, P<0.001).

Toni Choueiri, MD, of Dana-Farber Cancer Institute in Boston, who presented the JAVELIN data at GuCS, pointed to "nuances" between the two studies, noting that there were more favorable-risk patients in the pembrolizumab study (31.2% vs 21.4%) and that while discontinuation of both combinations was low in the two trials, it was higher in KEYNOTE-426 (10.7% vs 7.6%).

He said that the OS data from axitinib-avelumab are "very early" and that a second interim analysis was planned for 2019. "We will have an OS analysis," he said, "and then we can talk."

But Powles argued that the two studies had very similar designs and follow-up times. "While you can argue that one OS signal is not there yet because it's immature, the reality is the other study has -- at a similar time point -- managed to come in with a 0.53," he said.

"Five words," Choueiri said, channeling Yogi Berra. "Ain't over 'til it's over."

Coinciding with their presentations, results from and were published simultaneously in the New England Journal of Medicine.

"Both combinations are expected to become new standards of care and to be incorporated into future guidelines," wrote Bernard Escudier, MD, in an that accompanied the papers.

He noted that single-agent axitinib comparator arms would have been beneficial in order to assess the added benefit of checkpoint inhibitors, and that axitinib monotherapy demonstrated a 48% ORR and median PFS of 14.5 months in a phase II trial of metastatic RCC.

Of "great interest" will be whether longer follow-up reveals an OS benefit in the avelumab study, he said.

"Right now the bar is higher," said Tsao. "Therapies may not necessarily get approved based on PFS."

Escudier pointed out that in contrast to these two trials, CheckMate-214 -- which tested ipilimumab (Yervoy) plus nivolumab (Opdivo) -- did not show a significant PFS benefit, but showed an OS benefit in intermediate- and poor-risk patients and "more importantly, the complete response rate among patients in that trial was high."

Ipilimumab-nivolumab received FDA approval in first-line metastatic RCC last year for intermediate- and poor-risk patients, and a recent analysis reported that in this setting.

In 30-month follow-up data presented at the meeting by Nizar Tannir, MD, of MD Anderson Cancer Center in Houston, the complete response (CR) rate with ipilimumab-nivolumab was 10.5% in the intention-to-treat population -- 11.2% in the intermediate- and poor-risk group and 8.0% in the favorable-risk group.

By comparison, CR rates were 5.8% with the pembrolizumab combination in the KEYNOTE trial and 3.4% with the avelumab combination in JAVELIN. In the sunitinib arms of the three trials, CR rates ranged from 1.8% to 1.9%.

Tannir said he would urge the FDA to approve ipilimumab-nivolumab in favorable-risk patients if longer follow-up revealed an OS advantage over sunitinib in this group, and said he had "no problem" recommending the combination for favorable-risk patients based on the "impressive" CR rate in all risk groups, and the fact that the vast majority of these responses proved durable.

At a median follow-up of 32.4 months, 88% of those with CRs had ongoing responses to ipilimumab-nivolumab.

"We need to provide the best chance of cure to our patients," said Tannir. "I advocate to recommend nivo-ipi for all patients if they have no contraindication to receive nivo-ipi, and they're willing to go through the ride with us."

Since the results of CheckMate-214, ipilimumab-nivolumab "is considered the gold standard" for the first-line treatment of intermediate- or poor-risk RCC, said Tsao.

In these groups, the OS rates favored ipilimumab-nivolumab over sunitinib at all time points (80% vs 72% at 12 months, 66% vs 53% at 24 months, and 60% vs 47% at 30 months). Median OS was not reached in the ipilimumab/nivolumab arm compared with 26.6 months in the sunitinib arm (HR 0.66, 95% CI 0.54-0.80, P<0.0001).

Most experts agreed that ipilimumab-nivolumab should be reserved for these poorer-risk patients.

"I do think we still have to look at favorable risk differently," said Wood, noting that across these three treatment combinations, the hazard ratios for PFS and OS were in favor of all combinations and in each risk group with one exception -- the favorable-risk group in the ipilimumab/nivolumab-treated patients. "Based on the data, I would say that I would not recommend it for that patient population."

In the favorable-risk group, median OS was not reached for either ipilimumab-nivolumab or sunitinib (HR 1.22, 95% CI 0.73-2.04, P=0.4426), but the OS rates were numerically worse at all time points with the immunotherapy combination compared with sunitinib (94% vs 96% at 12 months, 85% vs 88% at 24 months, and 80% vs 85% at 30 months).

Powles said that comparing single-digit CR rates when making treatment decisions "is not a good game to play."

"Survival is really important -- I think it's important for patients," he said. "I see the OS signals in the two big studies that currently stand driving decision-making."

Decisions in IMDC Intermediate and Poor Risk

With CheckMate-214 and KEYNOTE-426 both showing improved OS for intermediate- and poor-risk groups, which combination should be selected?

That's the "million dollar question," Tsao said. "How do you pick between the two ... which both show in randomized phase III trials to have an overall survival benefit in those intermediate/poor-risk disease groups?"

"It's going to take a little bit of time to sort out what patient may do better with certain therapies with higher-risk disease," Dreicer told 乌鸦传媒. "We're not gonna know the answer immediately."

Tsao noted that CR rate, which he too called "impressive" with ipilimumab-nivolumab, would certainly be a consideration.

"What a lot of us got very excited about with immunotherapy was the CR rate," said Wood. "I think all of us are happy with the 10%, but we sure would have liked 20%."

Wood said that when it comes to choosing, multiple factors need to be taken into account: efficacy, tumor or patient characteristics, toxicity, cost and drug access, and logistics. "There's no clear winner in my mind at this current time."

She pointed out that as checkpoint inhibitors require intravenous infusions, the nine scheduled visits with sunitinib, an oral therapy, increased to 13 with ipilimumab-nivolumab, 18 with pembrolizumab-axitinib, and 26 with avelumab-axitinib. And there were many unscheduled visits for the immunotherapy combinations, while complications with sunitinib could usually be managed over the phone.

Biomarkers currently appear to be of little help in this setting.

"We don't have a specific biomarker to guide us," said Tsao, adding that more research is needed.

PD-L1 expression was looked at in each of the studies, but in nearly all cases, both PD-L1-positive and PD-L1-negative saw PFS and OS benefit with the checkpoint inhibitor combinations compared with sunitinib.

"I don't think this is going to help us at all," said Wood.

"Some recent data suggest that angiogenesis, the T-cell effector response, the interferon-γ response, and myeloid inflammatory gene-expression signatures may help to predict response to inhibitors of VEGF, PD-1, and PD-L1," Escudier wrote in his editorial.