Birrer Shares Perspective on Folate Receptor Alpha-Targeting Agents in Ovarian Cancer

Michael Birrer, MD, PhD

Folate receptor alpha (FRα) is an active area of investigation in the ovarian cancer paradigm, as it is highly overexpressed in approximately 60% of patients with high-grade, serous ovarian cancer, explained Michael Birrer, MD, PhD.

Moreover, despite recent findings from the phase III FORWARD I trial, which showed that mirvetuximab soravtansine did not improve progression-free survival (PFS) versus chemotherapy in patients with FRα—positive, platinum-resistant ovarian cancer and in an overall patient population, its role continues to be explored, as the antibody-drug conjugate could have benefit in combination with other available agents.

In an interview with OncLive, Birrer highlighted the evolution of FRα-targeting agents, took a closer look at the phase III FORWARD I findings, and theorized where mirvetuximab soravtansine could optimally fit in the paradigm.

OncLive^®: Can you discuss some of the broader directions of ovarian cancer therapy in recent years and the importance of new targets for treatment?

Birrer: There are 2 major areas of discovery. The first is antiangiogenic agents, particularly bevacizumab, which is now approved for the treatment of ovarian cancer at every time during the natural history of the disease: upfront, recurrent, and platinum-resistant. The more recent step forward has been the discovery, recharacterization, and approval of PARP inhibitors, which are now approved for second-line maintenance, third-line and fourth-line treatment, and then, just recently, first-line maintenance. These have been big steps forward, particularly PARP inhibitors.

What are some common knowledge gaps in ovarian cancer when it comes to these newer approaches in research?

Given the rapid trajectory of research, how would you assess current efforts emphasizing folate receptor α (FRα) as a target for treatment?

How are FRα-targeting agents different from other agents currently used to treat ovarian cancer?

I think we still have a problem with the upfront treatment. We probably cure about 15%, but that number hasn’t changed in 35 years. This is why [the] SOLO-I [trial] is so exciting, in that the PFS curve is pretty flat at the end and had not even reached median PFS, so that is the kind of improvement we need to see more of, and, of course, ultimately stratifying patients according to who will really benefit from the treatment versus toxicity is really important. The other area [in which there] is still an unmet need is the effective treatment of platinum-resistant disease, which is on the other end of the spectrum. Once patients recur, they are incurable, and most tumors will eventually become platinum-resistant. [Patients with] platinum-resistant tumors have a median overall survival of about 11 months and a median PFS of about 6 months.There is no doubt that FRα is a validated target. It is highly overexpressed in about 60% of high-grade, serous ovarian cancers, probably 25% of endometrial, and some non—small cell lung cancer, also. But it has certainly been validated as a cell-surface target for ovarian cancer. Now, the use of it as a therapeutic target had some starts and stops in terms of using a folate mimetic, which didn’t work out too well. But with mirvetuximab soravtansine, a humanized antibody, the initial data, phase I, and expansion cohorts showed the target to be important and the agent to be quite effective.

The majority of folate gets into the cell using the reduced folate transporter and does not come through the FRα. We are not even fully sure what the FRα does and why it is overexpressed. Some data suggest that it might be a poor prognostic feature, but the data are pretty weak. So, we are not sure what role it is playing in the cancer cell.

Can you talk about the evolution of FRα-targeting agents, including the differences among them?

FRα is being used as a selective target with the cancer cell because it is not really expressed in any other normal tissue. It is a specific cell-surface target, and if you use an antibody—drug conjugate and the antibody binds to the receptor, you can deliver a payload, which is the way mirvetuximab soravtansine actually works. It binds to the receptor, it gets endocytosed, and then it releases its payload into the tumor cell. In that sense, the target itself is unique; it is just being used as a self-surface marker, and the antibody–drug conjugate is very unique. You are delivering the maytansinoid anti-tumor compound DM4 directly to the tumor cells. DM4 is so powerful that after the tumor cell dies, it releases active drugs, and this enables what is called a bystander effect, meaning that not only do you kill the cell that it binds it, but you kill the cells around it.

FRα was cloned in the late 1970s, and it really was not considered seriously as a target until more recently. The initial attempt to target this was using just a naked antibody, which would bind to the receptor, and the thought was there might be some biology which would make it clinically relevant. These naked antibodies also had the ability for antibody-dependent cytotoxicity, [in which] basically macrophages or other immune cells would bind to the Fc receptor and then kill the tumor cell. The take-home message was that the antibody trials were negative. The next version of this was vintafolide, a molecule that carried a chemotherapy payload and used a folate moiety to bind to the receptor. This molecule actually went to a randomized phase III trial. Unfortunately, it was negative, and it stalled some of the development of FRα as a target.

Can you discuss the trajectory of mirvetuximab soravtansine in clinical trials and what the findings from these studies have revealed?

The reason this trial was negative is likely 2-fold. First, it relied on an imaging approach that was very subjective; the second problem is that the folate-binding molecule does not bind as solidly to the receptor as I think one needs, so the on/off properties may also have undermined the drug.

Mirvetuximab soravtansine represents the next generation of FRα-targeting molecules. It is a humanized antibody linked to a payload, and it binds very tightly to the receptor. I was a primary investigator in the phase I trial, and the only observation of note in the dose escalation was that the toxicity was pretty narrow and somewhat unique. We had a subset of patients who developed pseudocysts in the cornea, and these are dense enough that they cause blurred vision. We were able to show quite quickly that it was completely protein kinase (PK)—dependent. These patients had high serum concentrations of the drug and the toxicity was completely reversible. There is no FRα expressed in the cornea, but this is a drug effect from DM4 and based upon careful review of the PKs, we developed an ideal adjusted body weight dosing, which significantly eliminated a lot of the high toxicity. Then we added lubricating eye drops and some steroid eye drops, and at this point, the toxicity is below 30% for the eyes and all of it is grade 1 or 2. So, we think we’ve managed it quite well.

Once the phase I escalation phase was complete, we advanced to an expansion cohort, which is when we initially observed the data indicating that this might be effective. The response rates were impressive, pushing 50%. We went from the phase I expansion cohort to phase II with really good data and we analyzed them really carefully. Then, we decided that the best response rates [were] seen in patients who had platinum-resistant disease with 1 to 3 prior lines of treatment, nothing more than that. When we treated patients with up to 4 or 5 prior lines, we noticed that the response rates dropped off. In addition, for maximum effectiveness, 2- to 3-plus intensity staining for the FRα was optimal.

That was the basis for the randomized phase III FORWARD I trial, in which the experimental arm was mirvetuximab soravtansine and the control arm was physician’s choice of single-agent pegylated liposomal doxorubicin, topotecan, or weekly paclitaxel treated until progression (and 2- to 3-plus staining on their tumor for the FRα). That trial opened and closed quite rapidly, so we accrued very well.

Although mirvetuximab soravtansine did not hit the targeted PFS outcome, [a closer look at the data reveals that] the high expressors demonstrated a PFS prolongation and an OS difference that was statistically significant. The design of the trial did not allow for the usual P value of .05 but .025, which it did not meet. The high expressor data will likely be discussed with the FDA and a confirmatory trial [will take place].

Another step, potentially [in the development of mirvetuximab soravtansine], is to combine it with one of the accepted single agents that we use for platinum-resistant disease, which is the basis of the FORWARD II trial. Then, there is a movement to try to take mirvetuximab soravtansine and move it earlier in the disease, so that would be the basis for combining with carboplatin. The take-home message is that the mirvetuximab soravtansine/carboplatin combination is very well tolerated and in many ways is better tolerated than paclitaxel because there is no hair loss or neuropathy. The early data [suggest that this combination] is exceedingly active in platinum-sensitive patients.

If mirvetuximab soravtansine is approved, how would it affect the treatment spectrum?

The mirvetuximab soravtansine/bevacizumab combination is also interesting in that the response rate is as good as with mirvetuximab soravtansine alone, but [with the added] duration of response of bevacizumab, which is also encouraging. The one combination that has been disappointing is mirvetuximab soravtansine/pembrolizumab. Initial data suggested a very a strong response rate, but as the data matured, the response rate was observed to be more modest.

In the short term, we will have another active agent with which you can treat platinum-resistant patients, which is certainly going to change the standard of care. It is not going to cure anybody, but it is always helpful for platinum-resistant disease to have another active agent. This drug has the potential to completely change the paradigm for ovarian cancer. We are already in trial design to explore [the possibility] of substituting mirvetuximab soravtansine with paclitaxel at multiple places in the natural history of the disease, especially earlier in the disease, which would be the first major change in the upfront treatment of ovarian cancer in about 20 years. This will allow women to go through that treatment without any hair loss, without developing neuropathy. To me, that’s a big leap forward, and you could actually save paclitaxel until later in the course of the disease.

One does need to recognize that this agent is not going to be for all patients, because you have to have FRα expressed, but this brings up another interesting issue regarding resistance processes. It looks like the main resistance to this drug is the down-regulation of the receptor; if that is the case, finding agents that modulate it and that up-regulate the receptor could sensitize patients to this drug even after they have become resistant.