Herzog Highlights Approval of Olaparib/Bevacizumab as Frontline Maintenance in HRD+ Ovarian Cancer

Thomas Herzog, MD, discusses the FDA approval of olaparib plus bevacizumab for the frontline maintenance treatment of patients with homologous recombination deficiency–positive ovarian cancer.

Thomas Herzog, MD

The FDA approval of the combination of olaparib (Lynparza) and bevacizumab (Avastin) for the frontline maintenance treatment of patients with advanced ovarian cancer is a welcome addition to the treatment paradigm, according to Thomas Herzog, MD.

The approval is specifically for patients who are in complete or partial response to frontline platinum-based chemotherapy and whose cancer is associated with homologous recombination deficiency (HRD) positivity defined by either a deleterious or suspected deleterious BRCA mutation, and/or genomic instability.

“In terms of the clinical implications of the approval, it provides us with more choices, which is always helpful. If you had started the patient on bevacizumab [and chemotherapy] and you see they're responding beautifully, would you really want to drop the bevacizumab and do switch maintenance to a PARP if you found out the patient had HRD positivity, for example? Or would you want to continue?” questioned Herzog. “By continuing, you then have then the ability to add a PARP inhibitor into maintenance. I think it's wonderful that we now have the ability to be able to add the PARP inhibitor into the maintenance portion of the bevacizumab.”

The approval of the combination was based on data from a biomarker subgroup analysis of patients with HRD-positive tumors from the pivotal phase 3 PAOLA-1 trial.1,2 In this subgroup of patients with HRD-positive tumors, including tumors that had BRCA mutations, the combination reduced the risk of disease progression or death by 67% compared with bevacizumab alone (HR, 0.33; 95% CI, 0.25-0.45). The median progression-free survival (PFS) with the combination versus bevacizumab alone was 37.2 months versus 17.7 months, respectively. In those with HRD-positive tumors without BRCA mutations, the median PFS was 28.1 months with olaparib plus bevacizumab and 16.6 months with bevacizumab alone (HR, 0.43; 95% CI, 0.28-0.66).

Moreover, in the intent-to-treat population, the combination was found to reduce the risk of disease progression or death by 41% versus bevacizumab alone (HR, 0.59; 95% CI, 0.49-0.72; P <.001). After a median follow-up of 22.9 months, the median PFS was 22.1 months and 16.6 months with the combination and bevacizumab alone, respectively.

In an interview with OncLive, Herzog, professor of Obstetrics and Gynecology at the UC College of Medicine and deputy director of the University of Cincinnati Cancer Institute, discussed the rationale for the combination, provided further insight into the data that led to the approval, and shed light on further subgroup analyses that came out of the trial.

OncLive: In light of the FDA approval, could you discuss the rationale of the doublet therapy and previous findings with this combination?

Herzog: It certainly is an exciting time in ovarian cancer. By building off the results of the SOLO-1 trial, where we saw the significant advantage to adding a PARP inhibitor for those with a germline or somatic mutation, we now see more potential advantages from the data that were reported from the PAOLA-1 study at the 2019 ESMO Congress that are now published in the New England Journal of Medicine. Not only do patients with germline and somatic mutations benefit from the addition of PARP inhibition to their treatment, but [so do] all comers. Those data are very important to understand.

If you look at the BRCA group in the PAOLA-1 trial, you see a hazard ratio of 0.31, which compares very favorably with what we saw in SOLO-1; the [hazard ratio] was essentially the same. That’s very reassuring. However, what’s interesting is, and many people forget about this, you're not just comparing [the regimen] with a control arm in in PAOLA-1; that control arm is an FDA-approved active drug in the case of bevacizumab.

In this study, patients were randomized 2:1 to receive either olaparib plus bevacizumab or bevacizumab plus placebo. In that control arm, you had bevacizumab, which we know very well is active; it’s very interesting in terms of looking at that. Then, we have this whole concept of HRD; these are patients who don't have a germline or somatic mutation but have other genes and other pathways where they have the phenotype as if they had a BRCA mutation. We expect them to behave similarly, maybe not quite as great in terms of magnitude of effect, but very similarly in terms of the patterns of a response to a PARP inhibitor. We want to capitalize upon the leverage of these DNA pathways that we can drug.

If you look at the results from PAOLA-1, specifically the HRD population that included those who had a BRCA mutation, we didn't see really much movement; [the hazard ratio was] 0.31 for those with BRCA-mutated [disease] and it only went up to 0.33 if we looked at all patients with HRD-positive tumors. What's interesting is, if you pull out those patients with BRCA mutations, so now you don't have those in your group, and you just look at those with HRD alone, you still see a significant effect. In fact, the hazard ratio was 0.43 and the medians were quite large, with almost 12 months gained. It was 28.1 on the median for PFS versus 16.6 and this was based on investigator [assessment]. So, you're seeing a 57% reduction in PFS, even when you pull out the patients with BRCA mutations; this is very impressive.

When we look at the patients with homologous recombination—proficient disease, in other words, they don't score positive on the HRD test, we then see a difference in results. Here, the hazard ratio is above 0.90 and if we take the unknown group, it's 1.0. The medians are essentially the same; they're less than 1-month difference and we don't really see that effect. Remember, these are subgroups analyses.

If you look at the entire population, you see the hazard ratio was 0.59, which was really impressive in [that this] was a large trial of over 800 patients. When you put that into context, and you look at the FDA approval, on one level, it does make sense to take patients who have HRD because that's actually where you see the greatest benefit. On the other hand, you can argue that these small subgroups are just that, and if they filed based on the intent-to-treat analysis you would expect that the entire population would be approved for the combination of bevacizumab along with olaparib. There are 2 ways to look at it, but the label is what the label is.

Could you speak to the analysis from the PAOLA-1 trial presented at the 2020 SGO Annual Meeting, which looked at efficacy with the combination by timing of surgery and residual tumor status?

For this research, investigators looked into the PAOLO-1 data and did some subgroup analysis. Christoph Grim, MD, of the Medical University of Vienna, presented these data and examined the role of the timing of surgery and the influence of residual disease on patient outcomes. In general, you would expect a selection bias in [terms of] who receives primary surgery and who receives neoadjuvant treatment. Generally, those patients with a greater volume of disease are going to receive neoadjuvant therapy. Then, of course, we’ve known for a long time that residual disease has an influence on outcome, as well.

What was interesting is that in this study, investigators broke it down by the subgroups of whether they received the bevacizumab plus the olaparib versus just the bevacizumab alone and they found that the patients who had primary surgery had a hazard ratio of 0.52 versus 0.66, which was interesting [but] not surprising. If they had no residual disease [and complete surgical debulking in the up-front setting], the hazard ratio was 0.47 versus 0.61 [in those who underwent interval surgery and had no residual disease]. In [those with] residual disease, we didn't see much of a difference, at 0.74 [in those who underwent up-front surgery] versus 0.70 [in those who underwent interval surgery. The numbers are a little bit smaller there.

None of the findings were really unexpected. However, [this analysis] did confirm that the treatment effect was present across all cohorts. [The effect] was the greatest in those who had a primary cytoreduction and where there was no residual disease; that was the group who had the best outcome.

What does the safety profile of this combination look like?

In the PAOLA-1 study, one of the concerns is the toxicity of putting 2 active agents together. Certainly, when looking at the treatment-emergent adverse events (AEs), a significant number was observed in both arms. Grade 3 events were only minimally elevated in those [who received] the combination; it was 57% in the olaparib plus bevacizumab arm versus 51% in the placebo plus bevacizumab arm. Serious adverse events were similar [between the arms], as were the deaths.

AEs that led to dose interruptions were higher in the combination arm, occurring in 54% of patients versus 24% of those in the bevacizumab arm; reductions were also higher with the combination compared with bevacizumab alone, at 41% versus 7%. Discontinuations that were due to AEs occurred in 20% of those on the combination versus 6% of those on bevacizumab alone. You can see that there is a difference.

The most common toxicities were the usual PARP inhibition—related AEs, such as fatigue, nausea, hypertension, and anemia. In terms of grade 3 [events], the [most common were] hypertension and anemia. [Those events were] highest in the combination arm. However, interestingly, the grade 3 hypertension was actually lower [in the combination arm]. You would think this was a bevacizumab effect, but grade 3 hypertension was 19% in the combination arm and yet was 30% in the group that had just bevacizumab alone. Is there a protective effect? Nobody really knows but it was very interesting to see those data in terms of the grade 3 hypertension.

What are the clinical implications of this approval?

In this country, we see that 45% to 50% of patients receive bevacizumab; that’s on average. Some people use bevacizumab in 90% of their patients, while others use it in less than 10% of their patients. Nonetheless, if we look at all the patients, a little under half currently receive bevacizumab. When a patient comes into the office, the first thing you have to do is decide when you're going to perform the surgery or whether you're going to start them with neoadjuvant chemotherapy and then do an interval cytoreduction.

Then, right with that, especially if you're doing neoadjuvant chemotherapy, you need to make that decision as to whether you're going to add bevacizumab. Then, when you get to the end of chemotherapy, you're making that decision for maintenance. I think it's wonderful that we now have the ability to be able to add the PARP inhibitor in the maintenance portion of the bevacizumab.


  1. Highlights of olaparib (Lynparza) prescribing information. FDA. Updated May 8, 2020. https://bit.ly/2Wfjsgz. Accessed May 8, 2020.
  2. LYNPARZA® (olaparib) Approved by FDA as First-Line Maintenance Treatment with Bevacizumab for HRD-Positive Advanced Ovarian Cancer. Published May 8, 2020. https://bit.ly/3cft3tu. Accessed May 8, 2020.