Robert L Coleman, MD: PARP is an interesting enzyme that’s part of the process of DNA repair called “base excision repair.” In 2005, it was discovered that for cell lines that lacked BRCA, BRCA homozygous knockouts, when you gave those cells a PARP inhibitor, they had about a 1000-fold sensitivity for their cytotoxicity when they were administered. This was an experiment that was done in a dish. They took these cells and looked to see if they had the machinery, which we call “homologous recombination.” If that was intact, those cells did fine until very high doses of a PARP inhibitor were given. But when they lacked it, there was about a 1000-fold increase in sensitivity. That was strong proof that this damaged HRD (homologous recombination deficiency), and the damage of that process, made these cells extremely vulnerable to PARP.
Then, we started to think about how this happens. As it turns out, when you inhibit a base excision repair process, over time, the cells, under the cellular stress of division, are more prone to develop double-strand breaks. There are a few ways in which the cell can repair a double-strand DNA break. The one that has the highest fidelity and is the most often used is the one that’s governed by the BRCA genes. BRCA1 and BRCA2 act, in concert, with other genes to initiate this repair of a double-strand DNA break. There are some other repair processes that are also included, but this is the one that has the highest fidelity and the most opportunity for cell survival. In a normal situation, if BRCA is intact, this double-strand DNA break, which is induced by the PARP blockade, can be repaired. But if the cell is damaged and doesn’t have that machinery intact, it will die. So, neither one of those events is lethal, meaning a lack of BRCA or use of a PARP inhibitor. But if you combine them, you get what’s called “synthetic lethality.”
We started to look at a broader audience. We knew about the germline mutation, but we knew that there were patients in whom the tumors had the mutation but the germline was revealed as negative. They were wild type for BRCA in the germline setting, but the tumors actually were BRCA mutated. We found that 7% or so of patients were also sensitive to PARP inhibitors. This was shown with the use of one of the drugs that was recently approved, called rucaparib. A large phase II trial, called ARIEL2, showed that patients who had both germline and somatic mutations had a response. And, it was about the same degree of response that we would see in a patient who had a germline mutation.
Then, we also started to look at this in patients who had multiple lines of prior therapy. The drug olaparib, which is now approved for fourth-line therapy and beyond treatment of germline BRCA mutations, showed a response in patients who’ve had multiple lines of prior therapy. They were still responding. And, it wasn’t that they were just responding, but they were responding at a rate that was substantially higher than what was expected with chemotherapy. I think that was influential in the FDA‘s decision to grant accelerated approval to olaparib in patients that carried a germline mutation and had 4 or more lines of therapy. Actually, the response rate in that trial, called the Kaufman report, showed about a 34% response. This is impressive. But even more impressive was the fact that the 95% confidence interval, around that 34%, goes down to right around 25%. And if you look at what the expectation is for chemotherapy in the same lines of patients, it’s more than 4 standard deviations away.
So, while the number doesn’t seem very high, and that lower limit of confidence doesn’t seem very high, it is so much better than what is expected from chemotherapy in a general population. It was persuasive, and I think it was rightfully done. So, those 2 drugs, rucaparib and olaparib, are approved in patients who have specific characteristics. In other words, olaparib, for germline, and rucaparib, for germline and somatic mutations for BRCA.
In the background, we were also doing trials in patients that were wild-type. Now, in some of these trials, we didn’t know whether they were necessarily somatic mutations or not. But we started to see patients who were wild-type for BRCA in the germline setting having responses. So, it kind of launched 2 different investigative angles. One was to look for the somatic mutations which, as I mentioned, happen in about 7% of patients. But we started to look at the other genes that are responsible for homologous recombination and that central component that drives the repair of double-strand DNA breaks. We found that there may be as many as 30 genes that are responsible for that. What’s interesting is that not all of those genes have the same sensitivity to a PARP inhibitor. In fact, some of these genes have absolutely no sensitivity to a PARP inhibitor, even though they’re responsible for that DNA repair process.
So, that was an interesting observation, which has now started to come forward. It told us that there are patients in the general population that are not identifiable by germline or somatic testing, who respond to PARP inhibitors. That was a big win. It told us that maybe we could expand this to a broader audience. One of those audiences was patients who had responded to a platinum-based drug. As we’ve known for so many years, patients that have BRCA mutations actually show exquisite sensitivity to platinum-based therapy.
These trials, which banked on the particular physical characteristics of response to platinum-based therapy, said, “What do we do when we finish that treatment and then randomize those patients to a PARP inhibitor? They’ve responded to platinum-based therapy or the placebo.” That bore the concept of switch maintenance therapy.