Read the Full Video Transcript

Rashid Sayyid: Hello, everyone. This is Rashid Sayyid. I'm a urologic oncology fellow at the University of Toronto and along with Zach Klaassen, associate professor of the Department of Urology. We'll be presenting the recently published CONTACT-03 trial which looked at the combination of atezolizumab plus cabozantinib versus cabo monotherapy for patients with RCC who progressed after prior immune checkpoint inhibitor treatment. This trial was recently published in The Lancet by Dr. Sumanta Pal. Over the last half a decade almost, we've seen numerous immune checkpoint inhibitor combination therapies emerge in the first-line treatment setting for patients with locally advanced or metastatic RCC. There's been numerous approved combination of this setting. We have the dual immune checkpoint inhibitors with nivolumab and ipilimumab based on the results of the CheckMate 214 trial. Now we have combinations of an immune checkpoint inhibitor and a VEGF TKI with pembro and axitinib as part of the KEYNOTE-426 trial.

We also have the avelumab plus axitinib combination based on the JAVELIN 101. There's nivo and cabo based on the CheckMate 9ER trial, and then pembrolizumab plus lenvatinib as part of the CLEAR trial. What I want to highlight here is that the prior second-line treatment options were mostly established in the pre-immune checkpoint inhibitor combination era. And as such, there's been a major effort over the last few years to further define second-line agents in this new era which may have different efficacy. So what is cabozantinib? Essentially, it's a VEGF TKI that also inhibits the TAM family of kinases. It is approved in the first, second, and third-line settings in many countries. Currently, it's the most commonly used agent in patients who progress following prior ICI therapy. This agent is approved in the second-line setting following results of the phase III METEOR trial.

I want to note that the majority of these patients, if not all, were treated in the sunitinib era. This trial randomized patients following progression on sunitinib to cabozantinib 60 milligrams full dose versus everolimus 10 milligrams in these patients. It showed that cabo, as opposed to everolimus, was associated with improved objective response rates of 17% versus 3% and the overall survival was improved by about five months from 16 and a half to 21 and a half of the hazard ratio about 0.66. Recently presented ASCO GU 2023, we have data from the CaboPoint trial, which is a phase II trial of cabozantinib in patients with advanced or metastatic clear cell RCC who had radiographic progression following first-line ICI therapy. If we go back to the METEOR trials, patients who progressed following the VEGF TKI therapy, here we have patients that have progressed following the ICI therapy, so more contemporary cohort of patients.

So patients in this trial were split into two cohorts, cohort A, progression following dual ICI therapy, essentially ipi and nivo, and then cohort B, which progressed following ICI plus VEGF-targeted therapy. As we can see here in the schematic, the overall response rate, the objective response rate is about 30%, so clearly, there's some signal in this setting. Well, what about the combination of cabozantinib and atezo in the post-ICI setting? Some may ask, well, these patients are already progressed on immune checkpoint inhibitors, is there any reason for us to rechallenge them with further treatment with similar mechanism of action? It appears that the answer is yes. There's evidence that immune checkpoint inhibitor rechallenge in patients with disease progression following prior ICI therapy maybe associated with durable responses. We see that across disease space, be it melanoma, lung cancer, urothelial cancer.

Atezo, which is a PD-L1 inhibitor, is not yet approved in the advanced RCC setting, but there's some precedent for the use of cabo and atezo and that comes from the COSMIC-021 trial where the combination of cabo 60 plus atezo 1,200 milligrams IV every three weeks, it was associated with an objective response rate of 58% in patients with advanced clear cell RCC. So based on this clinical need as well as some evidence that this combination may work, the CONTACT-03 trial aim to value the safety and efficacy of this combination versus cabo alone in those with metastatic RCC who had disease progression during or after prior ICI treatment. This was a multicenter, randomized, open-label phase III trial and included patients obviously greater than 18 with a good performance status and who had either histologically confirmed locally advanced or metastatic RCC.

It's important to note that a lot of these trials include only patient with clear cell RCC, but this trial include patients with non-clear cell histology as well, be it papillary, chromophobe, or unclassifiable. We also note that sarcomatoid features are permitted and there's quite large body of evidence to suggest that these patients benefit from early immune checkpoint inhibitor therapy. So all patients included had radiographic progression. This could have occurred in one of two ways, either during or after immune checkpoint inhibitor in the first or second-line locally advanced metastatic setting or patients who progressed during or within six months of adjuvant immune checkpoint inhibitor therapy. How did the authors define immune checkpoint inhibitor therapy, either anti PD-L1 or PD-1 antibodies with a minimum of two cycles administered. We also note that patients were not allowed prior cabozantinib, but other VEGF TKIs were permitted and no prior mTOR inhibitors, prior treatment with more than one immune checkpoint inhibitors or more than two lines of therapy in the locally advanced or metastatic setting. Essentially, what the authors are trying to do is make this a second-line setting trial.

Patients were randomized in a one-to-one fashion to cabo 60 milligrams, again the full dose, plus atezo 1,200 milligrams IV every three weeks or cabo alone. No crossover was permitted, so gives it more of a cleaner trial design. Randomization was stratified by the IMDC risk group. What was the prior ICI line of therapy? Was it adjuvant? Was it first-line? Was it second-line or the RCC histology as well? Was it clear cell without sarcomatoid? Was it non-clear cell without sarcomatoid or sarcomatoid? The importance of this randomization to be stratified by these variables to make sure that they're well-balanced across the treatment arms and then you don't have one group that has more sarcomatoid versus clear cell, so makes it again, a bit more organized. How are these patients followed? It was mainly cross-sectional imaging, so CT or MRI. That was done at baseline and then every nine weeks for the first year and a half and then every 12 weeks thereafter.

We're at the study endpoints, so co-primary in the setting. We have progression-free survival, which was defined from random assignment to disease progression. This was assessed by a blind independent central review and overall survival as well. What about secondary endpoints? Investigator assessed PFS as opposed to the blind independent central review. They looked at objective response rate defined as either complete or partial response as well as the duration of response to the respondent. Again, they use the RECIST version 1.1 criteria. They also looked at the time to response and then the safety outcomes and adverse events. In this schematic, very nicely summarized the trial design that we talked about with the key eligibility criteria, the randomization, the endpoints, and how the randomization was stratified.

Looking at the statistical analysis of a target sample size of 500 patients. The primary analysis of progression-free survival was planned following 325 events had occurred in the intent-to-treat population. Just by way of background, intent-to-treat, meaning if you're assigned to a treatment irrespective of what happened, you receive it or don't receive it or receive the other drug, you're still included in that first group that you're assigned to. Then this design would allow for a 90% power to detect an improvement in PFS rate by about 33%, so a hazard ratio 0.67, which is quite classic with these trials. This was powered at two-sided P-value alpha 0.02. That's the false positive rate. And you may ask, well, why didn't they use a 5%? Well, it's important that you don't overspend your false positive rate allowance. So they split that between different analyses.

We'll see that at first interim, overall survival analysis was also performed this study after 176 death events had occurred. The allocated alpha was quite small at 0.0019. Then they performed survival analysis using the Kaplan-Meier curves with comparisons made by the log-rank test. They generated hazard ratios using Cox proportional hazards modeling and this was stratified by the IMDC risk grouping to adjust for any differences in outcomes that may be related to the underlying IMDC risk group. At this point, I'll turn it over to Zach to go over the results and discussion of this trial.

Zach Klaassen: Thanks so much, Rashid. This trial had 692 patients that were assessed for eligibility. Ultimately, 522 are randomly assigned including 263 to the atezo, cabo arm and 259 to the cabo arm alone. If we look at the bottom, we can see that 161 patients in the atezo, cabo arm were still alive or ongoing treatment at the time of cutoff compared to 157 in the cabozantinib arm. This looks like the baseline patient demographics in this trial. We can see that atezo plus cabo had a median age of 62 compared to the cabo arm at 63 years of age. We could see that the majority of patients, over three quarters were male in each of these arms with the majority of patients over 80% being white with the second most common demographic being Asian patients at 13% in the atezo, cabo arm and 9% in the cabo arm.

With regards to most recent immune checkpoint inhibitor therapy, just over 50% were locally advanced or metastatic first-line with just under 50% being locally advanced or metastatic second-line and just a smattering of patients receiving only adjuvant therapy. The most common histology was dominant clear cell without sarcomatoid at just under 80%. The second most common was dominant non-clear cell without sarcomatoid, 11% for atezo, cabo and 12% for cabo alone. The most common IMDC score was one to two at 65% in the atezo, cabo arm and 59% in the cabozantinib alone arm. Moving down to previous VEGF TKI use, most commonly was one round of therapy with VEGF TKI at 63% for atezo plus cabo and 61% for cabo alone. With regards to regimens of the previous first-line therapy, we see four different regimens that were quite commonly used, but the most common being ipi plus nivo, 31% in the atezo plus ccabozantinib arm, and 27% in the cabozantinib arm.

With regards to previous second-line therapy, the majority was nivolumab alone at 87% in the combination arm and 93% in the cabo alone arm. This is the Kaplan-Meier estimate of PFS by blinded central review in the intention-to-treat population. We can see just looking qualitatively at this Kaplan-Meier curve, these are essentially overlapping curves. We can see that the atezo, cabo group had a median PFS of 10.6 months and the cabo alone arm was 10.8 months with a hazard ratio of 1.03, so not statistically significant. This is the forest plot of PFS by blinded central review in key subgroups. We can see looking at the dots and the hazard ratio of one, there was very little benefit among any of these subgroups to the atezo plus cabo arm versus cabo alone.

This is the Kaplan-Meier estimate of overall survival in the intention-to-treat population. Again, overlapping curves. The median OS in the combination arm was 25.7 months and was actually not reached in the cabozantinib arm with a hazard ratio of 0.94 and a 95% confidence interval of 0.70 to 1.27. Again, looking at forest plot of the OS and key subgroups, no benefit between these subgroups, although we do see perhaps a slight signal IMDC score one to two, hazard ratio of 0.67, 95% confidence interval, 0.45 to 0.99. But in the context of the subgroups and the intention to treat population analysis, no significant benefit amongst these patients.

This is the secondary efficacy outcomes and we'll look at the RECIST 1.1 per central review. We see the exact same confirmed objective response for each group, 41%, very few to none with regards to complete response. Partial response, 41% for the combination arm and 40% for cabo arm. Stable disease, slightly higher in the atezo plus cabo group at 51% versus 48% similar progressive disease, and ongoing objective response at cutoff basically the same between these two groups. We actually see a slightly longer median duration of response for cabozantinib alone at 14.8 months compared to 12.7 months for atezolizumab plus cabozantinib.

With regards to adverse events, there're several key points to highlight here. If we look at any-cause adverse events, essentially all patients had at least one. Grade 3, 4 was similar between the two groups, 68% for atezo plus cabo compared to 62% for cabo alone. But when we look at serious adverse events, this is where we see some differences which are important to point. Serious adverse events, 48% for the combination arm compared to 33% for the cabozantinib alone arm. Serious adverse events related to treatment, 24% for atezo plus cabo compared to 12% for cabo alone and adverse events leading to withdrawal from the trial, 16% for the combination arm and only 4% for cabo alone. So we do see that there's some additional toxicity leading to serious adverse events with the addition of atezolizumab. Most common adverse events was diarrhea, 65% in the combination arm and 71% in the cabo alone arm.

CONTACT-03 is the first randomized phase III trial in patients with any type of cancer to examine the safety and efficacy of rechallenge PD-1 inhibitor following progression or previous PD-1 or PD-L1 therapy. These results showed an absence of benefit and increased toxicity with continuation of PD-L1 inhibitor use in patients receiving TKI therapy for checkpoint inhibitor resistant renal cell carcinoma. Adding atezolizumab led to an absolute increase of 15 percentage points in the rate of serious adverse events and an absolute increase of eight percentage points in treatment-related grade 3, 4 adverse events versus cabozantinib alone. This suggests that routine use of checkpoint inhibitor rechallenge outside of clinical trials should be discouraged. However, we have the ongoing phase III TiNivo study comparing tivozanib plus or minus nivolumab in patients with metastatic RCC with previous checkpoint inhibitor exposure, which will provide additional data and hopefully find a benefit in this disease space.

However, CONTACT-03 is practice changing. Currently, we know that anti-PD-1 or PD-L1 should not be used after progression on prior PD-1 or PD-L1, and we must continue looking for novel targets. So in conclusion, the data from CONTACT-03 do not support the addition of atezolizumab to targeted therapy beyond progression on previous checkpoint inhibition in patients with RCC. This is despite encouraging phase II data that supported this approach. So it shows the importance of performing phase III trials. This data highlights the importance of randomized prospective assessment of rechallenge with checkpoint inhibitors in patients with RCC and highlights the importance of further research and clarifying the role of immunotherapy in other tumor types. We thank you very much for your attention. We hope you enjoyed this UroToday Journal Club discussion of the CONTACT-03 trial recently published in The Lancet.