ARIEL3 is a trial with rucaparib, which is a PARP inhibitor, and it’s a randomised phase III trial looking at the effect of rucaparib maintenance therapy in women with recurrent ovarian cancer. This is a group of women with platinum sensitive recurrent ovarian cancer, that is they’ve responded to a course of platinum therapy, either with a partial or complete response and normalisation of CA125, and at the end of that treatment are then randomised either to rucaparib 600mg twice daily or placebo until progression.

Is there a standard of care for this indication already or would this be filling that niche?

There is in Europe a licence for maintenance olaparib but only in patients with a BRCA mutation. What we did in the ARIEL3 trial was take all women with recurrent ovarian cancer, high grade ovarian cancer, and we first looked at the tumours to see whether or not there was a BRCA mutation, either somatic or germline mutation.

The design of the trial was to look at the progression free survival in three nested cohorts. The primary endpoint was an investigator assessed PFS and the first cohort was in patients who had a BRCA mutation in their tumour, either a germline or somatic mutation. If that was positive we then added in a second cohort. These were BRCA wildtype patients who had been shown to have a high LOH score. LOH is loss of heterozygosity and it’s an indication of the HRD, homologous recombination deficient, phenotype and therefore believed to be greater susceptibility to a PARP inhibitor. So all tumours underwent LOH testing looking for the amount of genomic scarring and this was done by a Foundation Medicine assay as part of the analysis of the BRCA and also other HRR genes. So that was the second cohort and if that HRD cohort was positive we then went on and looked at an all-comer analysis, that’s all the patients, BRCA mutated, wildtype, including the LOH low and those in whom LOH could not be determined.

The results showed that in the patients with a BRCA mutation there was a highly significant improvement in the progression free survival with maintenance rucaparib, the hazard ratio was 0.23 which was significant at p to 10-4. The difference in the median progression free survival in that group was 11.2 months between the control arm, placebo, and those taking rucaparib. When we went on to look at the HRD group we also saw that rucaparib was effective and then we went on to look at the all-comer group, the intention to treat analysis, and still saw a significant benefit of rucaparib with a hazard ratio of 0.36. So the conclusion from this is that rucaparib is active in all patients with platinum sensitive recurrent ovarian cancer. The magnitude of the effect was greatest in the patients with a BRCA mutation.

With that tremendous effect one would expect there would be quite a lot of toxicity to come with it.

PARP inhibitors in general are very well tolerated and this was what we saw with rucaparib too. We know there is a spectrum of side effects that you see with PARP inhibitors, including rucaparib, such as fatigue, nausea, altered taste and anaemia, and the pattern of those side effects with rucaparib was similar to many of the other PARP inhibitors and, indeed, similar to what we saw with single agent studies with rucaparib.  Although these symptoms were reported they were usually of low grade, there were relatively few grade 3 or above side effects, and many of them disappeared with time. So, for example, fatigue and nausea are often worse in the first few months of treatment and then get better. As far as anaemia was concerned some patients needed a dose reduction or dose interruption for that as well as nausea. Overall about 50% of the patients had a dose reduction but only 13% of the patients discontinued treatment because of adverse events.

One of the special adverse events of note for rucaparib, as opposed to some of the other PARP inhibitors, was the presence of a raised ALT and AST. We know that alteration in transaminase is seen with rucaparib, it’s been seen in previous studies. Again, this tends to occur very early, in the first two cycles, and then spontaneously resolves. Having said that, there were a few patients in whom a dose reduction was needed so in fact about 11% of the dose reduction that we saw overall was due to alteration of the transaminase. I suspect that was also at an early part of the trial and we were getting more familiar with the drug and so as time went on we became more confident that the transaminase elevation would spontaneously settle.

With all of these results, PFS being the first endpoint, are there hopes that this will eventually translate to overall survival?

Overall survival is a difficult endpoint in ovarian cancer and it’s a difficult endpoint in patients on PARP inhibitors for many reasons. The key reason why it’s difficult to translate an improvement in progression free survival to an improvement in overall survival is because of crossover. Crossover is becoming more frequent now because of the availability of PARP inhibitors, licensed PARP inhibitors and, indeed, lots of other trials with PARP inhibitors. So although it is an important secondary endpoint for us and of course it is an endpoint that we would like to achieve, it is possible that that will not be different.

There are ways around this, for example the EMA have introduced a new coregulatory endpoint called the PFS2 which is the time to the second progression. So when a patient has a progression from a drug, here a PARP inhibitor, they will then restart probably chemotherapy, we hope they will have a response to that but at some point beyond that they will then progress. That is the PFS2: the second progression after the progression from the trial drug of interest. Why the EMA have raised that as an option to have as a coregulatory endpoint with PFS is because they know that patients will often crossover to a trial drug and, indeed, may have many lines of treatment beyond the trial drug in question. That too makes it harder to show differences in overall survival. So what you want to show is having improved the progression free survival with your drug of choice or your trial drug, you want to show that as you go through the next line of treatment that that benefit isn’t obliterated by the next drug. So, for example, if you had a drug that caused a lot of bone marrow toxicity and you then couldn’t treat the patient again with chemotherapy then you would question whether the benefit you see with the trial drug is worth having because you lose that benefit immediately. So PFS2 has been looked at now with olaparib and it’s been looked at with niraparib, you obviously have to wait until the data are sufficiently mature, and of course we will do that with rucaparib too.

By way of conclusion, you mention that there has been European approval for olaparib in a similar maintenance place. Is this another PARP to have or just another PARP to have?

The European licence for olaparib is specific for patients with either germline or somatic mutations. Here we’ve seen a group of patients without a germline mutation who respond very well to rucaparib. Indeed, that was the case with niraparib too. So I imagine that the plan would be for both niraparib and rucaparib, on the basis of the NOVA trial presented here last year and the ARIEL3 trial, to be submitted to the EMA to seek approval for a broad licence to cover all patients with platinum sensitive ovarian cancer. This has been the case already in the United States where niraparib is licensed in patients, all patients, with ovarian cancer. Actually very recently olaparib has been given a licence too for patients with or without a BRCA mutation in the United States as maintenance therapy but not in Europe yet.