AACR 2014

New antibody-drug conjugate shows early promise against all forms of melanoma

Dr Jeffrey Infante - Tennessee Oncology, Nashville, USA

I represent a great study team both in the US and Australia and this is a first in human trial of DEDN6526A. So this is an early phase I escalation study, first in human dosing, of an antibody-drug conjugate to ETBR or anti-endothelin B receptor. It’s specifically for patients with unresectable or metastatic melanoma.

So we think about ETBR as a cell surface receptor, a G-protein coupled receptor that has three ligands. It does signal through the MAP kinase pathway, which you hear a lot about with melanoma with BRAF and MEK inhibitors. It does regulate proliferation in migration of melanocyte precursors during embryonic development from the neural crest. It is associated with malignant transformational melanocytes with the potentiation of metastatic spread. With that being said, I think it’s important to note that a naked antibody without any chemo or any toxin attached to it, if it were to link onto this it’s not expected to have anti-cancer therapy by itself. The receptor which this is going to be targeting is ETBR and it’s expressed in over 50% of patients with melanoma. That expression is much higher than in the normal skin.

So this is an antibody-drug conjugate and it’s shown here in the figure. The stars on the figure represent sites of potential conjugation. It’s a Seattle Genetics based linker and then the payload is MMAE or monomethyl auristatin E. When I think about these antibody-drug conjugates it’s similar to TDM1 or Kadcyla in that your whole goal is that the antibody is very specifically targeted to reach a cancer cell, it’ll carry chemo on the back of it, a very potent chemo normally, one that can’t be given as a straight IV infusion. When it’s carried on the antibody it’s in an inactive state, as it’s given as an infusion it will search through the blood stream looking for the place to attach to and then it eventually attaches; in this case it’s searching for ETBR receptor. One of the most important parts of this, it has to get internalised into the cell so that once it binds the whole complex will get internalised and then through the normal cell processes it will break down the linker and release the very potent chemo directly inside the cell. In this case it’s a microtubule disrupting agent like a vinca alkaloid or taxane therapy and then lead to apoptosis.

One of the real important things for drug development of all the antibody-drug conjugates is to have a companion diagnostic to go along with it. So you want to identify what’s the likelihood that patients will express that receptor which is targeting and see if you can have a tool to eventually select patients as you move further in development. In this case a prototype ETBR IHC assay is being developed. It can be done on archival specimens or fresh biopsies but it’s just basically paraffin embedded tissue that’s needed to do the IHC staining. These are examples of what a pathologist can look at under the microscope from the biopsy where if it has a very minimal staining on the left it’s zero and then it can be very intense staining that goes across. But this gives us an opportunity to stain the patient’s tumour biopsies and see if this marker is present and then eventually correlate it with outcome.

So this was a first in human trial, it’s only melanoma patients. I’ll say that many of the melanoma patients, over half, had greater than three prior therapies; 70% had two or more prior therapies. Most of them had had ipilimumab or other chemotherapy. When we looked at the breakdown, I’ll present that today, it did allow cutaneous melanoma as the primary but it did not specify that so there were eight patients with ocular melanoma and three with mucosal melanoma.

We escalated quickly from 0.3mg with three patients in each cohort up to 1.8mg and that was the first DLT, which is a dose limiting toxicity. That was infusion reactions and those don’t commonly occur per se with antibody-drug conjugates; there are some that have it. In this case we were not premedicating patients prior and we had a grade 3 infusion reaction. Eventually we amended the protocol to allow premedication with steroids and H1 and H2 blockers like we would for patients that were getting taxanes in the clinic and this helped mitigate that side effect. We continued to escalate to 2.4 and had liver function test elevations as a dose limiting toxicity. Then eventually we went to 2.8 and again had a second DLT of liver function abnormalities that included a bilirubin in addition to the transaminitis. We didn’t meet protocol defined MTD in this case because normally you’d need two out of six patients to have a dose limiting toxicity but based on the severity of that DLT the pharmacokinetic data which I’ll show you was getting to levels that we thought were needed and also comparable across the platform with the antibody-drug conjugates. The fact that three of the six patients at the highest dose level ended up needing to be dose reduced we felt that moving forward that 2.4mg/kg once every three weeks, so as an instant IV infusion once every three weeks at 2.4mg/kg, is the recommended dose to move forward and this is what the expansion cohort opened.

I will say I’m only presenting information today on the dose escalation portion. The expansion did complete enrolment at the end of January but the data was not mature.

If we look at the safety profile of anti-ETBR ADC we see that fatigue is the most common adverse event, usually grade 1/2 but there were really no grade 3 fatigues. Chills, alopecia, a little bit of diarrhoea and sensory neuropathy were all in the 30-40% range. Some decreased appetite, headache, nausea and vomiting were also seen. If we look at grade 3 adverse events there really aren’t that many. So grade 3 and 4 together neutropenia was the most frequent, it often did not cause a dose reduction so it was manageable, and then the liver function abnormalities, the infusion reactions, which most of these were prior to the steroid premedication, and then neuropathy and that’s based on the payload which is a microtubule disrupting agent. So patients that are on study for periods of time that can develop.

This was the pharmacokinetics and basically the summary is that it’s mainly determined by the total antibody, so this is concentration versus time curves. The blue curve is the total antibody, the green curve is the antibody-drug conjugate with the conjugated MMAE and then the red curve is the unconjugated MMAE. If we look at the half-life it’s about 3½ days for the ADC and 3 days for the unconjugated MMAE. The concentrations of the MMAE are about 100-fold lower than the antibody-drug conjugate in the blood. There was no accumulation. So it behaved similar to other antibody-drug conjugates in the pharmacokinetic fashion.

This is the waterfall plot looking at outcome data for the 24 patients that made it, it says 28, it should be 24, patients that made it to restaging scans. Again, each bar represents an individual patient and you guys have seen a lot. This one is colour coded by dose because this is a first in human trial with many patients at different dose levels. So you can see there’s a few patients that do not have clinical benefit, there are many that have stable disease and many that had minor radiographic reductions actually. Then we have four patients that were partial responses. If we look at other factors that could influence it… I’m sorry, before I do that if it’s an M above the bar graph that’s for mucosal melanoma and if it’s an O that’s for ocular melanoma, if there’s nothing there it’s cutaneous.

If we look at the patients that had prior ipilimumab they’re shaded on this table, you can see that most of them did. There does not appear to be a difference in radiographic changes. This on the next one is prior DTIC or temozolomide as chemotherapy and again there does not appear to be a relationship and many of the patients that had clinical benefit had prior chemotherapy. Then there were two patients that were BRAF mutant on the trial, the majority of the patients were BRAF wildtype, partially because of the other therapies that were out there for those patients. But the two patients that had BRAF mutations both had BRAF and/or MEK inhibitors and there were a few patients with BRAF wildtype melanoma that had MEK inhibitors. Again it does not seem to play a differentiation factor in who is going to respond or not.

Then this was the development of the companion diagnostic that I was explaining, that you’re trying to develop a tool where you could select patients that highly express ETBR that they would be more likely to get the antibody-drug conjugate delivered appropriately to their tumour and then maybe have better outcomes. If you look on here, the orange shaded areas are those that by this IFC assay had high expression of ETBR, the green ones were considered low. So there is this trend that if you have high expressing on your tumour specimens by the IHC that you may be one that’s more likely to respond to the therapy.

This is a spider plot looking not just at responses but at duration on study and it’s of the 19 patients that are at, and again the end is wrong on this, it’s 19 patients at the 1.8mg/kg dose or higher and so that’s towards the end where we felt like we were getting levels that were predicted to be meaningful. If you look on these, each patient is an individual line and where there are circles these are response restaging times where they’re getting scans. If you’re going up that means the tumour is growing and if you’re flat, stable, and if it’s going down it’s shrinking. So if you look at this you see again there’s some patients where the drug does not seem to help, however there are many patients where they can have stable disease or there’s minor radiographic changes and some of them can stay on for a good duration. Then there are patients that develop full or we consider partial response with a 30% more decrease in their tumour and many of those are staying on for more than six months. So at least a third of the patients were able to stay on this first in human trial for more than six months.

So, in summary, this is an antibody-drug conjugate to ETBR. This is a first in human study and first in class for that. We did have a good safety profile, I’d say, it’s not water. We have fatigue and that’s probably the most common thing and neutropenia is common too but it’s manageable and often not a cause for dose reduction. Then if you look at infusion reactions that was a little hurdle we were able to get through with the premedications moving forward. If we looked at the anti-tumour activity there are four partial responses and again most of these patients were heavily pretreated. It did not seem to depend on whether they had a BRAF mutation or prior ipilimumab. I’ll say there are only two patients that had prior PD1 therapy so the population is evolving in melanoma because there are so many different options. 9 of the 28 patients remained on study for more than six months and that was across all dose levels, then the development of the ETBR assay, if we could further refine or identify the right patients to go on study, would potentially give more benefit. But the recommended phase II dose is 2.4mg/kg intravenously once every three weeks and the expansion cohort enrolled 24 patients which completed but the data is too immature to present.

With that I’d like to say thank you and I’m happy to take questions.