This trial was in the ESMART platform trial, is assessing the combination of a PARP inhibitor, olaparib, and an ATR inhibitor, ceralasertib. The aim was of this first part of this study to find a dose which is well tolerated in children and to assess preliminary activity. Given that we have molecular profiles at relapse from the patients, we are in a position to trace back and to identify molecular biomarkers for response.

The background, why we wanted to assess this combination, is that in paediatric cancers responses to single agent PARP inhibitors are not really expected. This is in contrast to adult cancers where a lot of patients in specific tumour types, like breast cancer, ovarian cancer, prostate cancer, pancreatic cancer, who have alterations in genes called BRCA1 and 2 and a few other ones which are really driving this disease and where the effect of the PARP inhibitor is to kill these cells. These patients really benefit from single agent PARP inhibitor treatment. In paediatrics the situation is very different – we don’t have these mutations but we still have evidence from work in the laboratory that PARP inhibitors might sensitise these tumours and that they have a role to play, but that these tumours have a primary resistance to these PARP inhibitors because they over-activate another pathway which is driven by a key protein or kinase which is called ATR.

So assessing this combination of ATR and PARP inhibitors we were thinking could overcome this primary resistance and really be of benefit to these children with relapsed/refractory tumours.

The study design was in this ESMART platform study. All trial arms are assessed separately and they have a two-stage design. The first stage is a phase I dose escalation which uses a Bayesian design, in this case it was a BOIN design, which is meant to identify a safe dose to move forward in the phase II part which is the assessment of the preliminary activity of the combination. In this specific combination of the ATR and the PARP inhibitor it's a Simon two-stage design which moves on to the second stage if a specific number of responses has been observed in the first stage. 

We are looking at two cohorts in this second phase and these cohorts are tumours with deficiency in homologous recombination in the broad way and sense, looking at our molecular profiling data. The second cohort is tumours which have high replication or transcription stress based on proven presence of, for example, aberrant transcription factors like AWS, Fli-1 or PAC-3, FOXA1.

At AACR this year we have reported the phase I part of the study which was the dose finding part. We have recruited 18 patients and we have identified a safe dose which is comparable to the dose used in adult patients. Already within this limited number of patients and the dose finding we have seen preliminary responses. So we had a partial response which was confirmed in a patient with a pineoblastoma and in a patient with a neuroblastoma. We also had prolonged stable diseases in three more patients for several cycles, one of the patients being on study for more than a year and ongoing.

What are the next steps for this study?

These results are extremely important and also very exciting because they show us in this first in child study that we can combine safely a PARP with an ATR inhibitor and that we, indeed, can see responses. We have proven a safe dose for the older patient cohort with 12 years and onwards. We don’t yet have enough patients of younger age to declare the safe dose for the younger patients so work in this direction is ongoing. We are also awaiting the results of the pharmacokinetic analysis of both drugs to see if the dosing schedule is the right one where we are and if we can further optimise it.

Most importantly, we have the opportunity to go now back to our tissue blocks of the patients and the molecular information which is already available from sequencing from the tumour samples at relapse. By going back to the laboratory and to the bioinformatics colleagues we are able to see which alterations in the tumour can really be linked with the responses or the rapid progressions in the clinic to identify future markers which can help us to choose the right patients for this combination and to avoid treating patients who very unlikely will benefit from this combination.

I would like to thank for the opportunity to speak to you and share the data. I do think that for paediatric oncology the approach to have comprehensive molecular sequencing available and doing an early phase clinical trial and then go back to the bench and to the thinking process and refine selection criteria for a trial is key going forward for us to drive forward paediatric oncology drug development. Another opportunity out of this is that if we have hints which could be the right biomarkers we also can go back in the laboratory and do more research with this gained knowledge. So it’s an iterative process between preclinical work and clinical work; in this way one day we hopefully are beating cancer in childhood.