Zakharia Evaluates Available Combination Options for Non–Clear Cell RCC

Yousef Zakharia, MD

Medical Oncologist/Hematologist

Mayo Clinic Arizona

Phoenix, AZ

Peers & Perspectives in Oncology: What options do the National Comprehensive Cancer Network (NCCN) guidelines prefer for this patient?

Zakharia: We always try a clinical trial whenever you have a clinical trial available, but obviously that’s not always the case. Cabozantinib [Cabometyx], based on [findings from] the PAPMET study [NCT02761057], was compared with sunitinib [Sutent] and showed superiority, so it became the standard of care 4 or 5 years ago. Cabozantinib plus nivolumab [Opdivo] and lenvatinib [Lenvima] plus pembrolizumab [Keytruda] are the [more recent] preferred regimens.¹

Then we have very specific types, like hereditary leiomyomatosis with RCC [HLRCC]. Those specifically do well with erlotinib [Tarceva] and bevacizumab [Avastin]. This regimen came out of [findings from a phase 2 trial (NCT01130519)] about 5 years ago, and those patients specifically do well with this. Those patients are typically the ones who have a history of uterine fibroma or fibromas on their skin, so that could be a hint [because] in our case, they had [a family] history of fibroids. It raises the suspicion for HLRCC histology. There are other regimens [recommended], like everolimus/lenvatinib…but those are not the preferred regimens. There are other options if you have contraindication for immunotherapy, for example.

Could you describe the phase 2 KEYNOTE-B61 trial (NCT04704219)?

This clinical trial was presented at the [2023] American Society of Clinical Oncology Annual Meeting, combining pembrolizumab and lenvatinib in patients who had nccRCC, who were treatment naive and had locally advanced or metastatic disease with no prior therapies. They had measurable disease and good performance status. There were 158 patients, a sizable number of patients for nccRCC, yet it was a single-arm study. Those patients were given pembrolizumab at 400 mg every 6 weeks up to 2 years, along with lenvatinib at 20 mg [daily], which is the same dose that is [given in] clear cell RCC. The primary end point was overall response rate [ORR] per RECIST by blinded central review and multiple secondary end points included disease control rate, duration of response [DOR], and progression-free survival [PFS].

In the intention-to-treat population of 158 patients, the ORR was 50%; I found it quite surprising to see high responses in nccRCC [Figure²]. I was not expecting that when I first saw these data, and 8% of them had complete responses [CRs]. It’s a mixed bag of subtypes, including papillary [n = 93], chromophobe [n = 28], 20 unclassified, 6 with translocations, and [10 with] other subtype not specified. Papillary [histology] was in the majority of patients here, which makes sense.

I find it good to consider the more than 50% ORR. If you look at the depths of the response in the waterfall plot, the majority of patients had some shrinkage of their disease. [The depth of response was presented for] the papillary, the unclassified, and the chromophobe [histologies]. Despite the fact it’s a singlearm study with a smaller number of patients, I especially like to see this type of waterfall plot where the majority of patients are having some shrinkage of their disease, and some of these responses are durable.³ They are going beyond 12 months [median DOR, 19.5 months; 50.6% with 18-month or greater DOR].² We do not have granularity on which ones are more aggressive. Are the aggressive ones responding more than the favorable [disease cases]? We did not have those data per se.

Looking at the PFS, the median in the total population was approximately 17.9 months.² Interestingly, in the chromophobe [subgroup], it was about 26 months of median PFS. Median overall survival was not reached in [the total population or in] the papillary and the chromophobe [subgroups].

That was sufficient for the FDA to grant it as a label for firstline nccRCC, and it made it to the NCCN as a preferred regimen. Looking at the adverse event [AE] profile, we are experienced from clear cell RCC at utilizing this IO/TKI [immunotherapy plus tyrosine kinase inhibitor] regimen. [There is] nothing surprising: hypertension, diarrhea, hypothyroidism, hand-foot syndrome are the common AEs that you see with any TKI.³ [Although] they did document grade 4 AEs in 5 patients…no alarming AE profile was documented in this trial.

What data supported the other IO/TKI combination for nccRCC?

The other clinical trial [was for] the combination of cabozantinib plus nivolumab. This is a slightly different design [but] still [a phase 2] single-arm trial [NCT03635892] in nccRCC. They allowed [patients who were] treatment naive or had 1 prior line of therapy; that’s the difference between the 2 trials. Here, the primary end point was ORR with multiple secondary end points…. This trial had 2 cohorts. Cohort 1 was papillary, unclassified, or translocation. That was 40 patients. Cohort 2 [for chromophobe histology] was closed early at 7 patients because of no signal [0% ORR].⁴

In looking at the type of responses we are seeing, this is the unfortunate part of this clinical trial. Because it’s a rare disease, there is slow accrual, and then you have a mixed bag. You put everything in the same bucket: papillary, chromophobe…. They are all different [disease types], but given the rarity of these tumors, we still look at the data as an aggregate. To add on top of this, this trial allowed patients [treated in the] first line and second line. What I would take from this is that it’s reassuring to see the responses across the groups [ORR: 54%], despite the fact some of these numbers are small, like 2 patients [with translocation], 6 patients [unclassified], 32 patients [with papillary histology]. It’s reassuring to see a significant number of patients are benefiting from this combination.

From the waterfall plot, the majority of them are having some shrinkage of their tumor, and with longer follow-up, many patients are beyond even 2 years. So [findings from] a smaller trial give us a hint that this is a regimen that is acceptable to use given the lack of large clinical trials.

If you look at the median PFS in the intention-to-treat population, it was 13 months, and in the first line, 11 months, and in the second line, 13 months.… It does not make sense that in the second line, they seem to do better, but you are looking at a smaller number of patients. Nevertheless, I think the signal is there, so it’s an acceptable regimen to use in your practice, especially extrapolating from another IO/TKI regimen showing interesting results. It’s reassuring to see that these data are not contradicting each other.

We are experienced in how to handle these AEs, nothing alarming here. It’s the same thing [as with lenvatinib/pembrolizumab]: fatigue, hand-foot syndrome, diarrhea, dry mouth, hypertension. We have seen it all in clinic.

I don’t like to do cross-trial [comparisons]. My takehome [message] here is that it’s reassuring to see that these 2 regimens are not total opposites of each other. You’re seeing a 50% ORR with both regimens and quite good numbers of CR rates. This is almost mimicking the [data in] clear cell RCC, although I was not expecting to see the higher responses in nccRCC based on the biology of these different tumors.

How do you approach the dosing of the TKIs?

[Cabozantinib was given at] 40 mg; 60 mg in combination was not studied. Both [cabozantinib and lenvatinib] can be toxic at higher doses. In CheckMate 9ER [NCT03141177], cabozantinib was studied at 40 mg in combination with nivolumab. When it was studied in CABOSUN [NCT01835158] and METEOR [NCT01865747] on its own, it was studied at 60 mg. Almost universally, nobody tolerates 60 mg. With the combination, they went down to 40 mg, and in my practice, a lot of patients go down to 20 mg. They do not do well even on 40 mg. It’s the same thing with lenvatinib. We start at 20 mg with pembrolizumab, but we de-escalate to 14 mg or even 10 mg, and I have a few patients on 8 mg, because it can be rough. Usually, having a low threshold to hold the dose and lower the dose is key in order to keep the patient for a longer period on these treatments.

We have reported real-world data with axitinib [Inlyta] and pembrolizumab 2 years ago, and what we have shown is the patients who have dose interruption and lower doses tend to do better in the long run.⁵ In your practice, don’t be afraid to lower the dose or give them a good period of interruption, especially in those patients who are responding well to the treatment, especially with cabozantinib, [which has] a longer half-life.

In what circumstances would lenvatinib plus everolimus be used?

[There was] another study that was testing lenvatinib/ everolimus. This was run prior to the IO/TKI combinations, and this was lenvatinib/everolimus as a first-line treatment for nccRCC. The number of patients for each disease group: 20 [papillary], 9 [chromophobe], and 2 for unclassified, so it was a smaller study here. The results were not as impressive as the IO/TKI combination, especially with papillary [histology], but 70% of them had stable disease.6 In the presence of the other IO/TKI regimen, this would be more of an alternative regimen, especially for those patients who are not fit for or have a contraindication to immunotherapy, like with autoimmune disease. This could be a fine alternative as a first-line treatment, but if they are fit to have an IO/TKI, I would go with that first.