So What Happened to the PARP Inhibitors in Breast Cancer Anyway? ASCO 2011

Many of you may recall that we described our studies with the small molecules BSI201 (iniparib) and AZD2281 (olaparib) (Nagourney, et al. ASCO 2011). Based upon the exciting Phase II data reported by Dr. Joyce O’Shaughnessy, first at the ASCO meeting, then in the NEJM, describing the remarkable efficacy of BSI201 (iniparib) combined with carboplatin and gemcitabine in triple negative breast cancer (TNBC), we initiated a study of both iniparib and olaparib in human breast cancer specimens. Our results were reported at the American Society of Clinical Oncology meeting.

Despite the enthusiasm that surrounded Dr. O’Shaughnessy’s initial observations, the confirmatory clinical trial using iniparib combined with carboplatin and gemcitabine, then compared with carboplatin and gemcitabine did not achieve statistical significance. That is, the trial was negative and the combo of inabirib with carboplatin plus gemcitabine was not proven superior.

So, what happened? Quite a few things.

It turned out that BSI201, a member of the benzamine chemical family, at physiological concentrations achievable in humans is not a PARP inhibitor. This, in retrospect, should have been obvious because a full-dose PARP inhibitor, plus a potent combination of carboplatin plus gemcitabine would not likely be tolerable if PARP inhibition were achieved.

Second, the patients receiving the drug are probably not a homogeneous population. That is, some TNBC patients may be similar to the BRCA patients, while others may not have the DNA repair deficiencies associated with PARP inhibitor response.

Finally, it was our group that originally reported the carboplatin plus gemcitabine combination in breast cancer, as a split-dose doublet in 2008 (Nagourney, Clin Breast Cancer Research, 2008). We observed, in that original clinical trial, that even a lower starting dose of gemcitabine (i.e. 800mg/ml2 vs. the O’Shaughnessy 1000 mg/m2) resulted in significant toxicity and in our concluding comments in that paper, we suggested 600mg/ml2. At 1000 mg/m2, Dr. O’Shaughnessy’s trial nearly doubled our recommended dose in this patient population.

While our abstract did not receive the fanfare of the clinical trial, it was, in fact, remarkably prescient. We, like other investigators, entered into our original studies of these molecules believing iniparib to be a PARP inhibitor. To our surprise, and, in retrospect, to our credit, a direct comparison of olaparib (AZD2281) to inapaprib (BSI201) revealed no correlation. We described this in our abstract, “Of interest, BSI201 & AZD2281 activity did not correlate in parallel analyses (R = 0.07, P > 0.5).”  Thus, our human tumor primary culture analysis scooped the ASCO investigators. Unfortunately, it appears they weren’t listening.

So, what have we learned? First, we’ve learned that iniparib is not a true PARP inhibitor.

Second, we learned that the combination of platins plus gemcitabine in breast cancer is synergistic, highly active and can be toxic (particularly at the doses chosen for this trial).

Finally, we learned that TNBC, indeed all breast cancers, even more to the point, all cancers in general, are heterogeneous. That is precisely why the use of human tumor primary culture analyses are so instructive and should be incorporated into clinical trials for these and other targeted agents.