Roundtable Discussion: Rini Reviews Later-Line Immunotherapy and TKI Combinations in Refractory RCC

Brian Rini, MD

During a Targeted Oncology Case-Based Roundtable event, Brian Rini, MD, Ingram professor of Medicine, chief of Clinical Trials, Vanderbilt-Ingram Cancer Center, Vanderbilt University Medical Center, discussed the case of 54-year-old man with refractory renal cell carcinoma.

RINI: You have a young, healthy patient who has somewhat indolent disease. It ultimately progresses and he is deemed to need systemic therapy. The question for the group is: When you’re choosing among these regimens, 3 to 4 regimens that have extended survival compared with sunitinib [Sutent] monotherapy, how do you choose? What factors are you using? Obviously, you can only choose 1, and they’re all good options, but what factors are you using to determine [which one], or what’s your practice? What’s your go-to frontline immunotherapy [IO]-based doublet?

AGARWAL: If it’s a low-risk patient, then I would use an [IO/tyrosine kinase inhibitor (TKI)] combination. But if [they are] intermediate or high risk, I go with IO/IO, [with] ipilimumab [Yervoy] and nivolumab [Opdivo].

RINI: I think this is a very common practice pattern. The IO/TKI combinations certainly have an effect in intermediate- and poor-risk patients, so what’s the reason that you choose ipilimumab/nivolumab in those patients? It’s a perfectly reasonable choice, and that’s a very common practice pattern; I’m just curious.

TIJANI: My personal experience is I also look at the bulk of disease and how many metastases they have. For patients who have a lot of metastases, my experience is that IO and TKI tend to work better than ipilimumab/nivolumab. So that’s 1 thing that I consider, the bulk of disease. If we want a quick response, I think putting a TKI plus IO will do better than nivolumab/ipilimumab. Most of the time, I go with axitinib and pembrolizumab.

RINI: Certainly the IO/TKI combinations have about a 20% to 30% greater response rate than ipilimumab/ nivolumab, so I think that’s pretty common as well; for patients with “bulkier” disease, or who have symptomatic disease, your chance of reducing their disease burden and hopefully making their symptoms better is certainly higher with those combinations. I think the definition of bulky disease is in the eye of the beholder, but I agree with you.

Do others use the IMDC [International Metastatic Renal Cell Carcinoma Database Consortium risk model] or specifically not use it? Do you calculate it, either online or in your head, and put it in your notes?

VARDHANA: I use it before making a decision about systemic therapy. The axitinib/pembrolizumab combination is reasonably well tolerated, other than hand/foot issues, and it’s convenient. Cost also sometimes plays a role.

KOVOOR: For low-risk patients, I do IO/TKI. For high risk, I use doublet IO. I usually just calculate it online.

RINI: I think that pattern developed because ipilimumab/ nivolumab was the first regimen approved, and the primary population was intermediate and poor risk. That practice pattern developed and when the IO/TKI combinations came along, they were effective in favorable-risk patients, and so were applied there.

KOKO: Another factor that I take into consideration is also the patient’s ability to be [adherent] with the oral therapy. For those patients who I feel are not going to be [adherent] with the oral therapy, I tend to go with the ipilimumab/nivolumab.

RINI: Fair enough. When you’re prescribing frontline treatment, and you choose it based on the factors, is one of the factors what you might give them in the second line? Do you think, “I don’t want to give cabozantinib/nivolumab in the front line because I like using cabozantinib second,” as an example? Does anybody use that consideration, the next therapy, in their choice of the first therapy?

KOKO: Yes, I think I do consider the second line when I’m recommending first-line therapy.

HEMPHILL: Typically, my go-to is still pembrolizumab/ axitinib. I think it’s probably from experience from when that first doublet came out; and just anecdotally, I feel like it had such a strong response. For the second line, I would potentially use dual IO, depending on how long the response rate was with the first line. I tend to calculate the [IMDC] score mostly on the prognostic standpoint than anything else.

RINI: Let’s talk about second-line therapy. Take a patient like this, who is a typical frontline IO/TKI patient. They’ve had some response, progression-free survival [PFS] for over a year, some of the normal toxicities; the pneumonitis is distinctly unusual with single-agent IO, but nonetheless [they] had an immune-mediated AE as well.

Now you have to choose a second-line therapy for this patient. We can talk about what specific drug you would choose, but I think maybe more important is what are the factors that influence decision-making? [For example], response, tolerance, burden, sites of metastases, etc.

What would you commonly choose for an IO/TKI failure, specifically axitinib/pembrolizumab failure, and what are some of the factors that you consider?

AGARWAL: In the second-line setting, I use cabozantinib as a single agent. There are a lot of data in the second-line setting, once you’ve failed with the IO/IO combination and the IO/TKI combination.

RINI: Cabozantinib is certainly the most commonly used second-line agent after a doublet.

AGARWAL: I’ve even had a patient [for whom] after I used IO/TKI, I used IO/IO. The patient had a response with ipilimumab/nivolumab.

SHULMAN: In this case, because he had toxicity to the IO, then I would be more inclined [to use] a TKI. If they have more rapidly progressive disease, lenvatinib [Lenvima] plus everolimus [Afinitor] is one combination that has an improved response rate in PFS over sunitinib.1 I’d consider that combination in this case. Adding another IO seems a little less attractive if they’ve had toxicity with it.

RINI: Yes, fair enough. There are a ton of options in the refractory setting. If you were using axitinib/pembrolizumab [in the] front line, what would you use second?

HAMMERS: In this patient, with this disease progression, who is symptomatic, and [with] the toxicity, a TKI is probably the next best choice. We have some data on the effectiveness of nivolumab/ipilimumab after nivolumab....If you want to go with a combination, maybe it’s better to use it up front than in the second- or third-line setting. But it is an option in carefully selected patients.

RINI: I think the frontline field of kidney cancer got settled recently; even though there are other options, the refractory field is wide open for the next frontier and that is where a lot of trials are happening now. But I think those are all reasonable considerations. I think a question for the field that hopefully will be answered in upcoming trials is: Can you use sequential IO, or do patients really just have 1 shot. We all have anecdotes and practice patterns, but there’s no real level 1 evidence yet, but there are some trials ongoing, so we will see.

SHULMAN: Do you think there’s a difference between either lenvatinib or cabozantinib vs axitinib as the partner TKI for frontline IO? Do you think axitinib is less active than the other 2 drugs?

RINI: It’s a good and complicated question. When I look at the axitinib/pembrolizumab data2 and the cabozantinib/nivolumab data,3 they look pretty similar to me. The response rates, PFS, and survival landmarks are pretty identical. When you look at the lenvatinib/pembrolizumab data, it does stand out.1 The response rate is about 10% higher, the complete response rate is 6% higher, the PFS is almost 2 years.

I don’t know whether that’s because lenvatinib is a different drug that pairs better with IO or whether that’s just the vagaries of data you get across trials that can vary. I don’t know, and I don’t have a lot of personal experience with the lenvatinib/pembrolizumab combination, so I think we’ll find out over time, as more data accumulate. But it’s a good question, and the clinical data would suggest lenvatinib is different, but I don’t know.

HAMMERS: This is a very complicated question, but the way I look at it is that axitinib is a fairly selective agent; cabozantinib and lenvatinib have been the classic salvage agent. The problem with cabozantinib is that you cannot give it at its highest dose level in combination with PD-1 inhibitors, because of liver toxicity. Lenvatinib stands out because it causes almost no liver toxicity in combination with PD-1 inhibitors. They could escalate the dose higher than what we normally use for lenvatinib. So, it’s like an escalated VEGF partner, if you will, in this combination. My feeling is that might drive some of the differences, but it’s certainly a very interesting data set, no doubt about it.

TAJUDDIN: Do you choose therapy based on metastatic sites such as delta metastases, and do you choose one over the other?

RINI: I don’t tend to choose based on site of metastases. Cabozantinib has this lore about working better in bone metastases; to me, those data are a bit all over the map. I choose more on my interpretation of the data, the toxicity, and what I know how to give and use, more than anything.

HAMMERS: I would say that all TKIs don’t tend to work so well in the bone. I go by tumor burden. The first question that I ask is: Does the patient need to have a high response rate? Is that my reason to go to a VEGF-based therapy or do I do IO/IO, for example? That’s the main question. And then, I pick my favorite combination.

RINI: Which second-line therapy are you likely to recommend: lenvatinib/everolimus combination, ipilimumab/ nivolumab combination, or something else? Single-agent TKI I still think is the standard of care, although not a wonderful standard. Lenvatinib/everolimus certainly has activity, although it comes at the cost of toxicity, so I think it depends on that balance for a particular patient. The immune checkpoint inhibitors [ICIs] probably have some role, although we lack level 1 evidence. There are 2 phase 3 trials that will look at TKI vs IO/TKI in this exact setting, so we should know within a couple of years.

The National Comprehnsive Cancer Network [NCCN] guidelines for subsequent therapy for clear cell RCC [show] the preferred regimens include nivolumab as a single-agent TKI in the front line, but also cabozantinib and lenvatinib/ everolimus, which I think most [of us] talked about using.4 But there are other recommended regimens, and the other drugs we have are useful in certain circumstances. Most drugs have some data in the refractory setting.

RINI: I usually start patients at 60 mg; there will be patients that I’ll start at 40 mg if I’m worried I’m going to [overwhelm them] at 60 mg. I’m not sure what the group’s practice is in terms of the starting dose. Unfortunately, this patient does not respond to cabozantinib, but the patient’s well enough to continue active therapy and there are other options.

Tivozanib’s regulatory history is about a decade long. It’s been around for a while, and finally achieved regulatory approval in kidney cancer.5 [Based on] the data, it is a very useful drug. I think its tolerability profile makes it stand out.

If you’re like me, as patients are generally in reasonable shape to get first-line therapy, you want to be aggressive. When you get to the second line, there are still options and patients still want treatment. As I’m transitioning into the third line, many times I’m asking myself…whether I am actually helping this patient. Are they well enough to get therapy, am I giving them more toxicity than benefit, etc? I start to have that conversation somewhere in the third to fourth line.

RINI: The goal of therapy is certainly to control disease. We’re probably not curing anybody who’s third line or later, so we’re trying to control disease. As you are seeing these patients, [such as] this very typical patient [who received] IO/TKI then had progression on single-agent TKI, what are you using, and why are you choosing that?

VARDHANA: I think one of the big considerations…for third-line therapies is toxicity. Clinical trials are something I always look for. I also use next-generation sequencing [NGS] to see whether there are any other targets. Most of the time we don’t find anything, but I think clinical trials are something we look for in this setting.

RINI: Yes, clinical trials are always the right answer, of course, but what [else] would you consider in a patient who progressed on pembrolizumab/axitinib then cabozantinib?

VARDHANA: Maybe lenvatinib. I would use that as a single agent.

RINI: So another TKI single agent, lenvatinib or otherwise… is very reasonable. There are not a whole lot of third-line data. [There are the] tivozanib data, but there’s not a ton of level 1 evidence out there.

AGARWAL: I mostly refer the patient for clinical trials, but I do an NGS panel, looking for NTRK fusion and all that.

TIJANI: I’ve used that lenvatinib and everolimus.

RINI: Certainly, lenvatinib/everolimus is active in this setting. I don’t use a ton of lenvatinib/everolimus, mostly because of the balance of toxicity, but there are certainly others who use a lot of it; it’s a very reasonable regimen. [Others] have mentioned NGS testing. I don’t do a lot of it up front or in the second line because I don’t really find

it helpful, but usually as I’m getting to the third line is…where I start to do it.

HAMMERS: [I don’t use NGS] up front. When I get desperate, like you…I look at NGS. In my practice, I do like to use lenvatinib/everolimus. I typically use IO/IO, sometimes certain TKIs, and then essentially go down my list. Once I go beyond cabozantinib, I tend to look at lenvatinib/everolimus for patients who are in a little bit better shape, who I think can tolerate it. Tivozanib certainly becomes an option then. So I try to tailor it to who’s sitting in front of me.

RINI: [It’s] mostly based on what you think they can tolerate and their fitness level.

KOKO: Is bevacizumab [Avastin] with an mTOR inhibitor an option?

RINI: There are not a whole lot of data for bevacizumab and everolimus, or temsirolimus [Torisel]. There are some data on bevacizumab/erlotinib, which I don’t tend to use much of, and I think [they’re] mostly in non–clear cell RCC. I don’t tend to use those regimens in this refractory setting. The story of bevacizumab development in kidney cancer is a bit of a sad one because it’s a great drug that’s well tolerated. It just didn’t get well developed, frankly, and it struck out. But I don’t know of many data for bevacizumab in this setting.

RINI: So the patient took pembrolizumab/axitinib [first line], then cabozantinib [second line]. I think once you get to the third- and fourth-line setting, all of these options are reasonable. Sunitinib and pazopanib [Votrient] tend to be used as frontline agents or not at all. It’s the way they were developed, although they certainly would have activity here. Lenvatinib/everolimus, we have talked about. Nivolumab, probably less so, given this patient has already failed IO and nivolumab by itself probably wouldn’t do much. Does anybody want to say what the “other” vote was?

AGARWAL: That would be for a clinical trial.

RINI: Clinical trial is always the right answer.

RINI: What do you think of tivozanib? It’s hard to wrap your head around yet another TKI because it’s the sixth or seventh one approved, but is it something that you would try in a refractory setting? Is it just another TKI and you don’t really want to spend the time or effort to learn it?

PATEL: I would definitely consider using tivozanib in the third-line setting. I don’t really have much reservation in…using it.

SHULMAN: I think I’m impressed with the tail on the [Kaplan-Meier] curve, and also the low toxicity. It has a fairly low incidence of hand-foot syndrome, which I sometimes find the most difficult thing to manage when they develop it. Once patients get hand-foot syndrome, it’s hard to stay on top of it. You sometimes have to get them off therapy for a prolonged period of time and the dosing becomes difficult. Sometimes it will even occur again at the lower dose, or it just sneaks up and gets bad all of a sudden.

RINI: I have been involved with the development, but I’ve used a lot of tivozanib. Conceptually, as patients get more and more refractory, I’m more worried about hurting the patient. So tolerance is No. 1 to me. I tend to choose what I think are better-tolerated drugs as I move into later lines of therapy.

SHULMAN: Would you consider this drug for the second line over cabozantinib?

RINI: It’s a good question. It would be off label, because patients have to have failed 2 prior systemic treatments. Would it be active in the second line? Absolutely, just like any other TKI would. And so my off-label answer is yes. I haven’t been using it that way; I’ve generally been using it in the third line. There is a trial called COSMIC-313 [NCT03937219], which is ipilimumab/nivolumab/cabozantinib vs ipilimumab/nivolumab. If that’s positive, and cabozantinib moves to the front line, you could imagine this going in the second line behind it. Every few years the landscape gets mixed up, so we’ll see how it plays out as some of these pending phase 3 [trials] read out.

TAJUDDIN: Is tivozanib studied with IO?

RINI: Good question. There’s a phase 1/2 study called TiNivo [NCT03136627]. It was a French study that was tivozanib plus nivolumab that, not surprisingly, given what we know about IO/TKIs, showed activity.7 There’s a phase 3 trial [TiNivo-2 (NCT04987203)], which is in a refractory setting, and it’s second-line tivozanib vs the tivozanib/ nivolumab combination. So [it’s] mostly asking the question, “Is second-line IO active after prior IO?” There are data on this combination, but it’s well behind some of the other combinations.

I would say to give it a try; it’s really well tolerated. My nurses love it because they get very few calls. So I’m biased, but I tend to use a lot of this agent.

_REFERENCES:

_{1. Motzer R, Alekseev B, Rha SY, et al; CLEAR Trial Investigators. Lenvatinib plus pembrolizumab or everolimus for advanced renal cell carcinoma. N Engl J Med. 2021;384(14):1289-1300. doi:10.1056/NEJMoa2035716}

_{2. Rini BI, Plimack ER, Stus V, et al; KEYNOTE-426 Investigators. Pembrolizumab plus axitinib versus sunitinib for advanced renal-cell carcinoma. N Engl J Med. 2019;380(12):1116-1127. doi:10.1056/NEJMoa1816714}

_{3. Choueiri TK, Powles T, Burotto M, et al; CheckMate 9ER Investigators. Nivolumab plus cabozantinib versus sunitinib for advanced renal-cell carcinoma. N Engl J Med. 2021;384(9):829-841. doi:10.1056/NEJMoa2026982}

_{4. NCCN. Clinical Practice Guidelines in Oncology. Kidney cancer, version 2.2022. Accessed November 4, 2021. https://bit.ly/3nXn5Ev}

_{5. FDA approves tivozanib for relapsed or refractory advanced renal cell carcinoma. FDA. March 10, 2021. Accessed November 4, 2021. https://bit. ly/3CM4NMK}

_{6. Rini BI, Pal SK, Escudier BJ, et al. Tivozanib versus sorafenib in patients with advanced renal cell carcinoma (TIVO-3): a phase 3, multicentre, randomised, controlled, open-label study. Lancet Oncol. 2020;21(1):95-104. doi:10.1016/ S1470-2045(19)30735-1}

_{7. Albiges L, Barthélémy P, Gross-Goupil M, Negrier S, Needle MN, Escudier B. TiNivo: safety and efficacy of tivozanib-nivolumab combination therapy in patients with metastatic renal cell carcinoma. Ann Oncol. 2021;32(1):97-102. doi:10.1016/j.annonc.2020.09.021}