(This post relates to work done at Weill Cornell, where OneThree’s core technology was developed)
This project is a bit of a followup from our previous collaboration with the Giannakakou lab where we identified and validated novel candidates that could act on cancer cells resistant to one form of microtubule inhibition. We wanted to investigate how we could treat resistant patients by identifying a new combination therapy that could break the initial drug acquired resistance.
New York Presbyterian hospital had recently completed a clinical trial for gastric cancer patients treated with docetaxol (a drug that acts on microtubules), where they observed that approximately 50% of patients developed resistance to taxol treatment over the course of the trial. Throughout this trial they had collected genomic data from samples taken from the sensitive and resistant patient populations both before and after taxol treatment. However, simple differential expression analyses didn’t turn up many leads as to how it might be possible to break resistance in these patients. That’s where we came in.
We started off by ingesting all this genomic data in our platform. Then using a pathway based machine learning approach we identified a genomic signature specific to the resistant patient and identified an approved drug that was predicted to be reverse this signature. Unfortunately I can’t publicly share the name of this drug just yet, so lets just call it D1 for simplicity.
Interestingly COMP. #1 had previously been tested in gastric cancer patients and it had little to efficacy, however, when we tested COMP. #1 + Docetaxol in resistant samples, we observed a dramatic increase efficacy with cell survival dropping to ~30% in most samples.
This was exciting for us, because it was a clear example of how our platform could identify synergistic combinations for drug resistance, even when one of the drugs showed little to no efficacy on its own.
We’re working with the Shah and Giannakakou labs to start a clinical trial to specifically test this combination and to better understand the mechanism through which it breaks docetaxol related resistance (and I’ll update this post as we go along), but in the meantime we wanted to think about how we could generalize this approach to rank