In a quest to look for specifically immune biomarkers and the most studied immune biomarker currently, PD-L1. Specifically for TKIs and for cabozantinib we looked at PD-L1 status on the CABOSUN trial, the trial that led to the approval of cabozantinib versus sunitinib front line, and on the METEOR trial, the trial that led to the approval of cabozantinib second line or later, post TKI, that trial was against everolimus. When we looked at PD-L1 immunohistochemistry we looked at different cut-offs, 1% and 5%; we looked at PD-L1 immune density; we looked at PD-L1 on tumour cells, immune cells, several lines of PD-L1, what defines PD-L1 positivity. What we found is irrespective of the treatment PD-L1 positivity is associated with worse outcomes – worse progression free survival and worse overall survival. Nevertheless, compared to the control arm, sunitinib in CABOSUN, everolimus in METEOR, cabozantinib does better.

This will be important, in a way, to define what are the benchmarks for PD-L1 positivity in terms of clinical activity, PFS and OS, especially when we start combining agents with cabozantinib. There is one ongoing trial combining cabozantinib with nivolumab and potentially other trials are in the making. One led by Dr Pal is combining cabozantinib with atezolizumab, a smaller study. So this will be important for the future for combinations to try to benchmark in PD-L1 positive patients.

If PD-1 is associated with the worst result for cabozantinib is there some kind of synergy you’re hoping for with the nivolumab, atezolizumab arms or could it just end up making things a lot worse a lot quicker?

This is an interesting question. It’s possible that it’s simply that PD-L1 positive patients do worse; though they may do better with checkpoint blockers they do worse in terms of everything else. We know very well that in localised disease PD-L1 positive patients have a shorter recurrence free survival and overall survival. We know very well from looking at the COMPARZ study with Dr Motzer and Dr Signoretti that if you get front line sunitinib or pazopanib if you’re PD-L1 positive your overall survival and progression free survival is worse. So I’m not surprised that PD-L1 positive patients do fare worse on sunitinib, everolimus and cabozantinib and I’m also not surprised that cabozantinib maintains its efficacy over sunitinib and everolimus in the CABOSUN and METEOR trials.

With cabozantinib established as superior to sunitinib, that leads to the JAVELIN trial in which sunitinib was the control arm. Was that just due to the timing of trial design?

Absolutely, at that time it was due to trial design and there are still ongoing studies with sunitinib as a comparator arm which is a less than appropriate control at this point, perhaps not so much in the good risk but in the intermediate and poor risk. Nevertheless that study enrolled and finished accrual some time ago and clearly showed superiority with the combination over sunitinib – response rate, progression free survival, both in the PD-L1 positive patients and in the overall population. Overall survival, we’re eagerly awaiting that, hopefully we can report on that soon but there weren’t enough events to conduct really the analysis with much precision. The follow-up for survival was less than a year on the study.

Do you see perhaps a future in which a patient’s PD-1 positivity sorts them to either the JAVELIN trial or cabozantinib or is there any future for a headline comparison between those two as well?

It’s possible, although when you combine the drugs, especially PD-1, PD-L1, with a VEGF TKI I wonder if they do worse with a VEGF TKI but then they do better with a PD-L1, PD-1, if then at that time you do not need the biomarker. Fair enough, at least if you look at progression free survival it was the same in the whole population versus the one that is PD-L1 positive. So we may not actually need the biomarker at all.

The interesting thing about the JAVELIN study too is that a subgroup analysis looking at IMDC or the MSKCC risk group, there was benefit in all three risk groups while in the 214 trial, the nivo/ipi trial that led to the approval of the combination of nivo/ipi versus sunitinib in poor and intermediate risk, the benefit of sunitinib in terms of response rate and PFS, at least, outweigh nivolumab ipilimumab in the older general population.

That sounds like a very interesting and complicated time working in renal cell carcinoma then.

I think so, I think we’re going to have a lot of first line options, at least in the United States, approved or at least be able to use nivolumab/ipilimumab, axi/avelumab, axi/pembro potentially, cabo/nivo, lenvatinib/pembrolizumab. Also let’s not forget atezo/bevacizumab – that study showed that that regimen is actually pretty well tolerated and the study met its primary endpoint of progression free survival in PD-L1 positive.

So, to me, the unmet clinical need is the big question of what to do second line and can you still use these PD-1, PD-L1 inhibitors after progression on first PD-1, PD-L1 inhibitor. So in the past we have used TKI after TKI but these were different TKIs. We’ve never used almost, let’s say in RCC, everolimus after temsirolimus or temsirolimus after everolimus, how different they are. Can you put a patient, let’s say, who progressed on axi/avelumab, can you put them on single agent pembrolizumab or on pembro/axi or pembro/lenvatinib or atezo/bev? Do you need to put them or can you just use a plain TKI? Totally unmet need, very important questions and hopefully the next few years we’ll be able to answer that.