The study we conducted was a retrospective study to look at if patients were on drugs that affected the angiotensin II pathway would that affect their response rates to anti-PD1 or anti-PD-L1 therapy. We did that because there is a lot of preclinical data suggesting that the angiotensin II pathway has overlap with both the VEGF pathway and the TGF-β pathway. Those two pathways have been connected to tumour progression and resistance to immune attack. We know that angiotensin II binds to one of two receptors, either AT1 or AT2, and preclinical data suggests that when angiotensin II binds to AT1 it leads to upregulation of both TGF-β and VEGF which are both bad in that they are linked to tumour progression, again preventing the immune attack on the tumour.
So we figured that drugs that block AT1 may potentially be helpful in patients getting other immunotherapies like anti-PD1 and anti-PDL1 inhibitors. The most common drugs that block AT1 are angiotensin receptor blockers and ACE inhibitors. There is a slight difference between ACE inhibitors and angiotensin receptor blockers in that both block AT1 but only angiotensin receptor blockers block AT1 by itself. ACE inhibitors also indirectly block AT2 and why that is important is because preclinical data suggests that when angiotensin II binds to AT2 it actually leads to a decrease in VEGF and TGF-β expression – a kind of counter-regulatory pathway. So ideally if you want to be targeting the angiotensin II pathway leading to reduction of VEGF and TGF-β as much as possible, then you’re going to want to be targeting specifically AT1 and not AT2. Then, in theory, angiotensin receptor blockers will be potentially a better drug to use in combination with anti-PD1 and anti-PD-L1 therapies than ACE inhibitors.
What we looked at, we looked at 600 patients with advanced cancer who were receiving anti-PD1 and anti-PD-L1 therapies. We broke them up by patients who were receiving ACE inhibitors, patients who were receiving angiotensin receptor blockers and patients who were receiving none of the above. We saw that response rates for patients who were receiving angiotensin receptor blockers were statistically improved compared to patients who were receiving no such agents. ACE inhibitors had a modest improvement but it was not statistically significant.
We also looked at survival and we saw, again, with patients receiving angiotensin receptor blockers that they had statistically improved survival. ACE inhibitors also had a modest improvement in survival which was not statistically significant.
So this seemed to be in line with our hypothesis that ACE inhibitors potentially may have a very modest effect because they target the AT1 but it may be negated by the fact that they also target AT2 whereas ARBs seem to have more of an effect by singly targeting AT1. The actual response benefit and survival benefit we saw was about a doubling of response rate and about a doubling of median survival. But, again, this is all retrospective data and retrospective data is best used for hypothesis generation. So all of this has to be evaluated in prospective studies and it’s very possible that use of these types of agents may help some tumour types and not others. So it probably has to be evaluated prospectively in individual tumour types. We did see a hint of that when we looked at individual tumour types, that potentially bladder cancer may be a tumour type that may benefit from this and other tumour types we didn’t get as strong a signal. Tumour types that traditionally don’t respond to anti-PD1 and PD-L1 inhibitors didn’t seem to be much improved by adding this or it seemed to be not statistically significantly improved with prostate and ovarian cancer. But, again, the numbers may be low for individual tumour types but it’s more of a hypothesis generation that has to be evaluated further prospectively in tumour types. At the moment bladder cancer looks to be the most promising but a lot of further study is needed.
How did you carry out the study and what are the next steps?
In addition to the standard chi-squared test to look at statistical significance among response rate and survival, we conducted a multiple regression analysis as well as Cox regression analysis to look to see different versions of multivariate analysis to see how much of these response rates and survival benefits were potentially due to the drug versus potentially other confounding factors. Other confounding factors that we looked at were age and weight and gender which have all been linked in the literature potentially as to having an effect on likelihood of responding or surviving after receiving PD-1 and PD-L1 inhibitors.
We also looked at other factors. We looked at the specific tumour types and we looked at other drugs that patients may have been receiving. A lot of these patients weren’t receiving only PD-1 and PD-L1 inhibitors but they were on trials where they were receiving anti-PD1 or PD-L1 inhibitors in combination with some other anti-cancer agent such as a CTLA4 inhibitor or a tyrosine kinase inhibitor. So we tried to take into account all these variables and still on our multiple regression and Cox regression analysis we saw the same effect with angiotensin receptor blockers seeming to statistically significantly improve response rates and overall survival; not so much so with ACE inhibitors.
So I would say that the next step is probably to further evaluate this data prospectively. The most simple concept is probably to evaluate some sort of version of angiotensin receptor blockers in combination with anti-PD1 and PD-L1 inhibitors for tumour types where anti-PD1 or PD-L1 inhibitors are already improved like bladder cancer. You could evaluate the combination and compare to historical controls or it would be potentially even better academically to evaluate the combination randomised to just anti-PD1 and PD-L1 inhibitors alone. That is one traditional way to go.
There are also several people I know who, just based on the literature, are developing more targeted angiotensin receptor blocker molecules. The issue with angiotensin receptor blockers is that they are prescribed generally most commonly for blood pressure so a common side effect of them would be hypotension. So it would be difficult to prescribe to people who have normal blood pressure or slightly low blood pressure at baseline. So in order to make effective use of these drugs for a larger public it would probably have to be a new design of angiotensin receptor blockers that are specifically targeted to the tumour and have less off-tumour toxicity. There are drugs that are in development that are meant to do this and these would potentially have even more of an effect with less toxicity.
Just to state again that while this data is very promising, it is just hypothesis generating and a lot of further study is needed before this can be acted on clinically for patient care. That’s all, just to summarise where we’re at.