If you have an ACS member number, please enter it here so we can link this account to your membership. (optional)

ACS values your privacy. By submitting your information, you are gaining access to C&EN and subscribing to our weekly newsletter. We use the information you provide to make your reading experience better, and we will never sell your data to third party members.



Pushing Cancer Over The Edge

After many disappointing studies, PARP inhibitors are finally showing promise as killers of cells already damaged by mutations

by Lisa M. Jarvis
June 17, 2013 | A version of this story appeared in Volume 91, Issue 24

Credit: SPL/Science Source
A class of compounds called PARP inhibitors shows promise in killing ovarian cancer cells, like the ones shown in this SEM image.
This is a SEM image of an ovarian cancer cell.
Credit: SPL/Science Source
A class of compounds called PARP inhibitors shows promise in killing ovarian cancer cells, like the ones shown in this SEM image.

A class of experimental drugs staged a quiet comeback last month at the American Society of Clinical Oncology’s annual meeting, held in Chicago. A cluster of studies suggested that compounds that block PARP, an enzyme involved in DNA repair, could be useful in treating ovarian and breast cancers driven by a mutation in BRCA, a gene that recently made headlines when actress Angelina Jolie revealed she carried the mutation.

The data presented at the ASCO meeting were not earth-shattering. And the results are early; companies still need to show their compounds can prolong cancer patients’ lives, a feat that has for years eluded the drug class. But for a field that has endured a roller-coaster ride of hype and disappointment, the forward momentum is encouraging. At least four drugs are poised to enter Phase III clinical studies in both breast and ovarian cancer later this year, and oncologists say companies are finally figuring out how to use the compounds safely and thoughtfully.

The new enthusiasm for PARP inhibitors is in contrast to the lows researchers felt in 2011 after high-profile setbacks for drugs from Sanofi and AstraZeneca. But subsequent sleuthing determined that the Sanofi compound wasn’t actually inhibiting PARP. And researchers from AstraZeneca and other firms have become better at finding the people most likely to respond to PARP inhibitors.

“Everybody now has recalibrated, redevoted themselves, and perhaps everyone is on a trajectory that I’m more optimistic about,” says Mark E. Robson, a clinician at Memorial Sloan-Kettering Cancer Center who advises companies working on PARP inhibitors.

The concept of developing a drug to target the PARP family of enzymes has been around for decades. PARP, which stands for poly(ADP-ribose) polymerase, was first identified in the 1960s, but it wasn’t until the early 1980s that, using chemical probes to block its activity, scientists understood the enzyme’s role as a first responder in repairing single-strand breaks in DNA.

Researchers saw two potential uses for PARP inhibitors in oncology. The first wave of studies tried to exploit their ability to knock out cancer cells whose DNA repair capabilities had been hobbled by chemotherapy.

The chemotherapy “sensitization” route carried appeal. A drug that could be safely and effectively added to front-line chemotherapy across many types of cancer would be a major commercial success. As a consequence, “many dollars were spent trying to combine PARP inhibitors with all kinds of chemotherapies,” notes Mary Lynne Hedley, president of Tesaro, a biotech firm founded by former executives from MGI Pharma. Hedley was also involved in the PARP inhibitor program at MGI Pharma, a biotech firm that was acquired in 2007 by the Japanese firm Eisai.

But combining PARP inhibitors with chemotherapy proved too toxic. Putting the two drug classes together caused synergistic myelosuppression, or a drop in bone marrow activity. Companies couldn’t find a way to add PARP inhibitors to standard chemotherapy regimens at doses that were both potent and safe.

Credit: Shutterstock/C&EN
This graphic shows how PARP inhibitors work.
Credit: Shutterstock/C&EN

The second potential use exploited the concept of synthetic lethality—cell death that occurs when two or more critical genes are impaired. The idea was to cripple both methods that cancerous cells use to mend themselves: homologous recombination (HR), a way of fixing double-strand breaks in DNA, and single-strand DNA repair, during which PARP is an early responder.

In practice, this meant a PARP inhibitor could be used on its own in situations where HR was already weakened. In 2005, researchers showed that the BRCA mutation—a faulty inherited gene implicated in roughly 5% of breast cancers and 10–15% of ovarian cancers—decreases HR. In the lab at least, cells with the mutation were susceptible to PARP inhibitors, sparking a new wave of trials in BRCA-positive cancers.

Yet just as with combinations of PARP inhibitors and chemotherapy, the approach failed to show sufficient activity in early studies. Researchers working in the field blame poor clinical trial design, in part caused by an insufficient understanding of how to use the compounds.

“At the beginning, these two development paths sometimes got conflated in people’s minds when they were thinking about trial design,” Robson says. “I don’t think people were always clear about which potential avenue of utility they were trying to explore.”

Nicola J. Curtin is a professor of experimental cancer therapeutics at Newcastle University, in England, who has worked in the field for decades. She explains that PARP inhibitors work differently when administered with chemotherapy than they do on their own.

Tests in mice show that PARP inhibitors, if given alongside another DNA-damaging agent, need to be given only in very low doses and for short periods of time; in fact, high doses cause the animals to die. On the other hand, as single agents, PARP inhibitors must be administered at very high dosages and for a prolonged period to keep cancer at bay. This is acceptable, though, because when taken alone the drugs are easily tolerated.

“The initial mistakes companies made were expecting the dose and duration sufficient for chemosensitization to be adequate for single-agent activity,” Curtin says. Conversely, a safe dose for a single agent is likely to be toxic in a combination.

The other wrinkle in the development of PARP inhibitors was the disappointing trajectory for BSI-201, a compound developed by BiPar Sciences. The compound entered the limelight in April 2009 when Sanofi announced it would pay up to $500 million to acquire BiPar.

The deal looked like a bargain a few months later, when early results from a Phase II study of the drug, later called iniparib, were presented in a plenary session at ASCO’s 2009 annual meeting. During that prime-time slot, the study’s investigator, Texas Oncology breast cancer expert Joyce A. O’Shaughnessy, reported that adding the compound to a chemotherapy regimen of gemcitabine and carboplatin prolonged survival in women with triple-negative breast cancer, a particularly tough-to-treat patient group. Furthermore, the addition of the PARP inhibitor caused no safety issues.

Final data from the study, released at the European Society for Medical Oncology meeting in October 2010, showed that addition of the drug extended survival by nearly five months.

Then things started to fall apart. In early 2011, Sanofi reported that iniparib had failed to prolong survival in a Phase III study of patients with triple-negative breast cancer. The disappointment deflated much of the hype around PARP inhibitors.

Around that same time, Astra­Zeneca said it would no longer pursue development of olaparib, its own PARP inhibitor, for BRCA-positive breast cancer. And at the end of 2011, AstraZeneca halted development of olaparib for ovarian cancer after a Phase II study found the drug delayed progression of the disease but failed to improve overall survival.

“It took us all—that is to say the clinical community—by surprise,” says Jonathan Ledermann, the University College London (UCL) Cancer Institute oncologist who served as lead investigator for the Astra­Zeneca study. “Here we had a drug that was clearly active in ovarian cancer, and on the basis of some immature survival data, the company had a very negative view.”

AstraZeneca’s decision to bench olaparib was also related to a formulation problem. To achieve the needed dose, patients had to swallow 16 large capsules a day, a tough sell for a drug that could be used as a maintenance therapy. The task of understanding the right dose of a pill form that would mean fewer pills each day took longer than anticipated.

“I wouldn’t suggest everything we’ve done is optimal,” says Susan Galbraith, head of AstraZeneca’s oncology unit, who concedes there was frustration in the cancer community around the firm’s difficulty in reformulating olaparib and in understanding how to properly design studies. The company was faced with balancing the “strong desire to rapidly move forward” with taking time to assess what the data mean, she adds.

In the midst of the Sanofi and Astra­Zeneca disappointments, Pfizer out-licensed its PARP inhibitor, rucaparib, to Clovis Oncology. Merck & Co. followed suit in 2012, out-licensing niraparib to Tesaro.

With so much uncertainty, momentum was lost. Clinicians were reluctant to start new studies of PARP inhibitors, and patients became less interested in participating in trials of drugs in the class, says Vincent L. Giranda, a project director for AbbVie’s oncology development program. AbbVie, formerly part of Abbott Laboratories, has been involved in PARP inhibitors since its acquisition of BASF’s pharmaceutical business in 2001.

But in 2012, the drug class got two boosts. The first came when two separate groups of academic scientists showed that Sanofi’s iniparib isn’t actually a PARP inhibitor, a revelation that helped both oncologists and investors feel more comfortable about the prospects for other PARP inhibitors in development.

Clinicians had always thought the early clinical data for iniparib were strange. The compound didn’t cause unsafe levels of myelosuppression as expected when administered with carboplatin and gemcitabine.

At the time, BiPar and Sanofi had kept their program close to the vest, clinicians say, and didn’t release research quantities of the compound, as drug companies often do. As a result, academic scientists were unable to play around with it in the lab to confirm its activity or compare it with other agents.

Prakash Jagtap, who headed medicinal chemistry efforts at Inotek Pharmaceuticals, the first company to put a PARP inhibitor into the clinic, was not surprised to learn that iniparib isn’t a PARP inhibitor. He and his Inotek colleagues had dug through patent applications and synthesized iniparib themselves. Their subsequent experiments showed the compound was clearly not blocking its purported target.

Despite clinical failures, companies have continued to pursue compounds that block the PARP enzyme. SOURCE: Chemical Abstracts Service
This bar graph shows the PARP drug patents for the last decade.
Despite clinical failures, companies have continued to pursue compounds that block the PARP enzyme. SOURCE: Chemical Abstracts Service

Others also had suspected iniparib data were leading the field astray and decided to push forward. As Clovis was doing its due diligence on Pfizer’s PARP inhibitor before licensing it in 2011, company scientists concluded that the lack of myelosuppression in the iniparib study “raised a red flag that iniparib wasn’t a PARP inhibitor,” says Andrew R. Allen, Clovis’ chief medical officer. “We developed a point of view that said, ‘Forget iniparib; it’s irrelevant,’ ” he recalls.

The second boost came later in 2012, when results were revealed from a deeper analysis of a Phase II study that had earlier caused AstraZeneca to halt development of olaparib to treat ovarian cancer. The preliminary results showed some patients were in fact benefiting from the drug.

Final details of the retrospective analysis, which looked at response to the drug based on a woman’s genetic status, were unveiled earlier this month at the ASCO meeting. Researchers were able to go back and compare drug response among three groups of women: those who inherited the BRCA mutation, those who spontaneously developed a BRCA mutation, and those who lack the errant protein.

Several trends emerged that have corporate researchers excited about PARP inhibitors’ potential to treat tumors linked to BRCA mutations. First, more patients in the AstraZeneca trial tested positive for BRCA mutations than expected: 136 out of the 218 patients tested carried either an inherited or a spontaneous mutation. People with one of the BRCA mutations fared the best on olaparib, which extended progression-free survival for 6.9 months compared with a placebo.

As in the earlier Analysis, olaparib kept the disease from progressing but did not prove the compound kept patients alive longer. Ledermann, the study’s author, points out that a trend toward an improvement in survival did emerge: Patients with the BRCA mutation and who took the drug instead of the placebo lived on average 3.0 months longer.

“As a clinician, I’m not worried about the fact that we’re not seeing stunningly different overall survival data,” Ledermann says. One reason for his confidence is that the data are recent—40% of the women who participated are still alive. Furthermore, the survival results are confounded by the fact that 23% of the patients in the placebo group switched over to a PARP inhibitor, either olaparib or a competing compound, late in the study.

That retrospective analysis, along with success in finally achieving a better formulation of the drug, prompted AstraZeneca to revive the ovarian cancer program for olaparib. Last month the company said it plans to put the compound into Phase III studies later this year.


Early data for other PARP inhibitors were revealed at the ASCO meeting, adding to optimism that the drugs will eventually find their way into oncologists’ tool kits. The most robust data were for ovarian cancer, though breast cancer results were also encouraging. Clovis’ rucaparib elicited response in a small Phase I/II study of people with solid tumors. Of the subset of people with ovarian cancer, 89% benefited, even though the study didn’t reach the maximum dose.

Tesaro presented results from a Phase I study of niraparib, the PARP inhibitor it bought from Merck. In a small group of women whose ovarian cancer responded to platinum-based cancer therapies and who had already tried, on average, six other therapies, the disease was kept stable for more than a year, regardless of the women’s BRCA mutation status.

And BioMarin Pharmaceutical offered an early glimpse of data from its Phase I/II study of BMN-673 as a treatment for breast and ovarian cancer in people who inherited a BRCA mutation. They found that 44% of the patients with ovarian cancer responded to the drug, whereas 39% of the 18 breast cancer patients saw a response.

On the basis of the encouraging data in the various trials, firms are pushing forward into the final tests needed for regulatory approval. AstraZeneca, Bio­Marin, Tesaro, and Clovis all plan to begin Phase III studies of their compounds in the second half of this year. “This is an active class of drugs. It’s unambiguous now,” Clovis’ Allen says.

As companies head into the next wave of trials, the race to market—once easily led by AstraZeneca—is suddenly tight, and differences between the molecules are subtle. Research published last year by National Cancer Institute scientists showed that PARP inhibitors work by trapping the enzyme at the site of DNA, creating a complex that effectively kills cancer cells. AstraZeneca’s Galbraith points out that the compounds don’t all trap with the same efficiency, a possible clue to their efficacy as drugs.

And although BioMarin’s compound is widely acknowledged as the most potent of the crop, clinicians say potency isn’t necessarily the be-all and end-all in giving DNA repair its fatal blow. Some compounds “have better pharmacokinetics [PK] than others, and it may be that for some indications you want short exposure and for others you want long exposure,” Newcastle’s Curtin says. “I think it’s going to come down to PK. I hope it’s going to come down to something rational rather than who gets there first.”

One challenge for the field going forward will be moving beyond BRCA-positive cancers, which are a small subset of the overall breast and ovarian cancer universe. Returning to the roots of PARP inhibitor development, scientists are again pondering ways to exploit the concept of synthetic lethality—the method of killing a cancer cell by knocking out both single-strand DNA repair and HR repair pathways.

For example, the Cancer Genome Atlas, a National Institutes of Health consortium devoted to large-scale genome sequencing, found in 2011 that half of the nearly 500 ovarian tumors tested had a deficiency in a DNA repair pathway and could benefit from treatment with a PARP inhibitor.

“We know that germ-line BRCA isn’t the whole story,” Clovis’ Allen says, using a term to describe an inherited BRCA mutation. The AstraZeneca data on olaparib confirm that a PARP inhibitor is effective in both inherited and spontaneous BRCA mutations while also suggesting that the drug has some activity in people without any BRCA mutation. “This tells us that PARP inhibitors can work beyond just BRCA mutants,” Allen says. “The question is: Can we identify those patients?”

Last year Clovis teamed with the cancer diagnostics firm Foundation Medicine to try to answer that question by analyzing an upcoming clinical study of its PARP inhibitor, rucaparib. The study will be similar to AstraZeneca’s olaparib study: Roughly 500 ovarian cancer patients will be enrolled and will be randomized to receive the drug or a placebo. The goal will be to demonstrate progression-free survival.

Clovis will collect tissue from each participant and, using a test developed by Foundation, sort patients into three categories: BRCA mutant, HR deficient, and biomarker-negative people—those who test negative for aberrations in either pathway. “Then we will see whether the drug has differential activity in these three groups,” Allen says. “Our hypothesis is that it works best in BRCA mutants, very well in HR deficiencies, and will have a minimal effect on the biomarker-negative group.”

Although Clovis is excited about its clinical strategy, not everyone in the field thinks enough is yet known about treating patients with HR deficiency. “I think there’s BRCA and everything else,” says Henry J. Fuchs, chief medical officer at BioMarin. “The state of our knowledge about BRCA and vulnerability of a cancer-bearing cell to a PARP inhibitor is really well described. Everything else is a distant second.”

A big concern shared by many is whether scientists can develop an assay to identify patients that the Food & Drug Administration will approve alongside the PARP inhibitor. Although AstraZeneca is assessing ways to select patients with HR deficiencies, “the challenge around trying to find a patient-selection tool when any part of a complex pathway could be deficient shouldn’t be underestimated,” says Mark J. O’Connor, senior principal scientist at AstraZeneca.

While companies cautiously move forward with patients who aren’t BRCA positive, researchers hope data from new studies involving BRCA-positive tumors will put to rest lingering doubts about PARP inhibitors. “I do think that the drugs will be approved in the next two to three years,” UCL’s Ledermann says. “I sincerely hope so since it’s an opportunity to treat a number of patients with a pretty nontoxic drug on a chronic basis.”


This article has been sent to the following recipient:

Chemistry matters. Join us to get the news you need.