Safety analysis of entrectinib for NTRK fusion tumours

Bookmark and Share
Published: 26 Oct 2018
Views: 2014
Dr George D. Demetri - Harvard Cancer Center, Boston, USA

Dr Demetri speaks with ecancer at ESMO 2018 in Munich about the pooled efficacy and safety results from the STARTRK-1, STARTRK-2 and ALKA-372-001 of entrectinib in NTRK fusion tumours.

He outlines his professional experience in treating sarcomas, in which TRK fusions are rare compared to other tumour sites, and the high overall response from sarcoma patients in these trials.

Dr Demitri notes the activity of entrectinib in the central nervous system to control metastases, and considers the future development of drugs to combat treatment resistance.

The data I presented here was about drugging with a new inhibitor, known as entrectinib, the TRK fusion proteins that are increasingly recognised as validated drivers of multiple kinds of cancers in both adults and paediatrics. So entrectinib is one of two very front running inhibitors being developed to drug TRK fusion proteins that are activated by chromosomal translocations. Extraordinarily interesting, extraordinarily rare in most cancers. In my world of sarcomas they happen to be enriched in certain undifferentiated sarcomas as well as gastrointestinal stromal tumours and related spindle cell malignancies that are not driven by other kinase activations such as mutated KIT or mutated PGDF receptor-alpha. But we’re finding these in lung cancers, in certain types of breast cancer, salivary gland tumours, thyroid, pancreas, cholangiocarcinoma s, all over the place and what’s fascinating is that when you find a fusion we can drug it and the patients get really remarkable levels of benefit in general. The presentation I made here was on entrectinib, it followed a presentation with the other kinase inhibitor known as larotrectinib and both of these are pursuing very rapid clinical development towards commercialisation in both the US and Europe.

Could we get into some of that results data then?

Absolutely. The thing that caught everybody’s attention were the high levels of objective response rates you see in a tumour agnostic manner which means that very similar responses can be seen regardless of whether it’s a sarcoma, a lung cancer, a thyroid cancer, breast. If the fusion is there they seem to have the same type of ability to respond.

Recognise that the data so far is based on 54 patients because these TRK fusion patients are so hard to find but just with the larotrectinib data having both expanded and enhanced its dataset with a year’s worth of extra accrual we’re seeing the same thing happen with entrectinib. With more patient recognition of that patients are asking their physicians for this sort of molecular testing; physicians are looking for these patients more so we’re finding more.

With entrectinib in particular it’s an interesting drug because it was designed to actually enter the CNS and lead to anti-nervous system coverage which is particularly important for diseases like lung cancer which has a high propensity to go with metastatic spread into the central nervous system.

What’s fascinating in the analyses I was able to present today is the high levels of objective response, more than 55%, 57.4% to be exact, and that’s likely to go up because patients over time can develop these responses. A number of patients who were in the safety database were not eligible to be evaluated because they had not long enough follow up on the drug yet. So these numbers are very mobile. We saw that as well in the larotrectinib data because they presented longer term data and, in fact, response rates went up because people had time to mature their responses from sometimes stable disease to a partial response over time.

With our data the duration of response in general has not been reached, the median duration has not been reached, which means that half the patients have not had failure of the drug yet. That’s always a good thing, I think a lot of physicians look at the data and say, ‘How come you didn’t report duration of response?’ Well if everybody is still on the drug it’s hard to respond in a median way. So that’s exciting.

Then the other important thing is that the patients responded regardless of whether or not they had metastatic disease to the brain which is extraordinarily unusual in cancer trials. First of all, cancer trials that even allow patients who have CNS disease to enter at study baseline are often not allowed into such clinical trials so it’s exciting that this trial was able to accrue patients with metastatic disease in the brain and it’s even more exciting for those patients that they had essentially an equivalent chance of having anti-tumour efficacy in their brain as they did in the systemic spread of the disease.

It sounds like a wonder drug which makes me wonder where things might start to go wrong, the first being toxicities. How is it being tolerated?

So the TRK inhibitors are extraordinarily well tolerated, it’s remarkable. The side effects are a bit of fatigue in some people, that was very well tolerated, a bit of anaemia, a bit of weight gain. Interestingly, these drugs can actually stimulate appetite in some people and that can be an interesting side effect that we’ve not been used to dealing with in cancer medicine before. Other than that the patients have actually enjoyed being on these oral medications, it’s really quite remarkable.

I guess we have to wait to find out more about it because it sounds very promising. When can we expect more data from this trial or any others?

We’re seeing an incredibly rapid evolution of data in this field, partly because the patients are staying on the drug so it’s easy to continue to build the database when the patients are all remaining on the trial. We do expect resistance, let’s be honest. This is a kinase and, just like with other kinases like the KIT driven GIST diseases that develop resistance mutations, even before this went into humans we predicted what the resistance mutations would be in laboratory studies. In fact, on any waterfall plot on one side of the waterfall you see all the patients whose tumours are shrinking, there’s always three or four patients on the left unfortunately who had the disease continue to grow. With all of the data one should always ask what’s different about those patients? Well, several of those patients had resistance mutations at baseline. That’s the kind of data that’s going to be much more commonly accessible and that’s the data that’s going to help doctors really fulfil the wishes of precision medicine to get the right drug to the right patient at the right time and to understand who might not benefit from a drug like this and then find the right drug for those patients. Already we’re seeing in the laboratory many interesting types of drugs be developed to overcome resistance in the TRK kinase. So as we’re going to see at every meeting right now we’ll see more follow up on these patients and likely much more experience with new patients who are going to find their way into these and other expanding trials so that we can learn more about them.

The other interesting thing is that with any kind of a disease agnostic drug, if this drug works regardless of cholangiocarcinoma, sarcoma, lung cancer, whatever, physicians are still going to look at it in the eyes of physicians who are specialised in lung cancer or sarcoma or colorectal cancer. Or if they’re not specialised oncologists they’ll say, ‘Well, I’m interested for my patient who has a cholangiocarcinoma,’ so they’re going to focus in on that part of the curve, as they should. So what’s going to be interesting is as we develop more data how we really get that data out and educate physicians so that they can make wise decisions with their patients about how to use these drugs, what to expect with them, what the risk factors are and what the long-term prognosis could actually be. It’s fun to talk about long-term prognosis in patients with metastatic cancer. I still have some of my patients who first went on Gleevec for metastatic GIST back in the year 2000, it’s 2018 right now, and they still visit me in Boston once a year still on Gleevec. That’s only about 10-12% of them but still the fact that it’s possible means that some people can do extraordinarily well if their disease is genomically stable. With the TRK inhibitors we think that the paediatric diseases are likely to have a more stable genome so we’re hoping for the good of those kids that they will be able to remain on these drugs long-term without long-term side effects.

Is there anything else that you would like to say about entrectinib at this point?

No, I think for the world the most exciting thing is that we have now seen two very effective kinase inhibitors, entrectinib and larotrectinib, both of which are very well tolerated and both of which should hopefully find a way of finding their way to the right patients. The challenge with these is how do we find these patients in a cost-effective way that’s sensitive to society’s needs. We can’t spend $2 million to find one patient, that’s just not sustainable. But I do believe that technology will find an answer to this; we’re bringing down the cost of next generation profiling of tumours to the degree that it’s much more affordable than it was even five years ago; it’s going to get even cheaper next year. And then we’ll use other kinds of technologies to try to hone in on exactly who these patients are. For example, in GIST I can tell you where these fusions are not going to be found. If a GIST patient has a KIT mutation you don’t have to look further, there will be no TRK fusion. If a GIST patient has a PDGF receptor-alpha mutation there’s no fusion. So what’s easy to say to people is if you’ve got a GIST that has no KIT mutation, no PDGF receptor-alpha mutation, no mutation in BRAF and no deficiency in the succinate dehydrogenase, the SDH system, that’s known as a quad negative GIST and those are the ones where the TRK fusions are going to hide. So that actually gives doctors a lot of guidance about how to screen in that particular rare disease. We have to do that level of detail for lung cancer, for cholangiocarcinoma and all the others.