Lasofoxifene Emerges as Understanding of ESR1 Mutations Expands in ER+/HER2- Breast Cancer

Tremendous progress has been made to understand the strategies that can be used to overcome ESR1 mutations in estrogen receptor (ER)–positive breast cancer, said Debu Tripathy, MD, who added that potent selective ER modulators (SERMs) may be the key to combat these mutations.

“Selective [ER] modulators can probably, to some extent, be effective [against] ESR1-mutant cells,” said Tripathy. “The effect to which they can reverse resistance probably depends on many aspects of the agent. These need to be formally studied not just in the lab, but in clinical trials. We are starting to get some hints of that [work], but it’s just the beginning.”

ESR1 mutations generally arise following treatment with an aromatase inhibitor (AI), explained Tripathy, so acquiring genomic information after a patient progresses is critical to inform treatment selection.

The novel SERM lasofoxifene is currently being investigated in combination with the CDK4/6 inhibitor abemaciclib (Verzenio) in the phase 2 ELAINEII trial (NCT04432454) in patients with advanced or metastatic ER-positive, HER2-negative breast cancer whose tumors harbor an ESR1 mutation. Moreover, in May 2019, the FDA granted a fast track designation to lasofoxifene for the treatment of female patients with ER-positive, HER2-negative metastatic, ESR1-mutant breast cancer.

In an interview with OncLive^®, Tripathy, professor of medicine and chair of the Department of Breast Medical Oncology at The University of Texas MD Anderson Cancer Center, discussed the role of CDK4/6 inhibitors in ER-positive, HER2-negative breast cancer, the emergence of SERMs, and the growing understanding of ESR1 mutations in this space.

OncLive^®: Resistance to frontline therapy with a CDK4/6 inhibitor plus endocrine therapy is a huge area of ongoing research. Is there any reason to believe that patients could have non-overlapping sensitivities to CDK4/6 inhibitors?

Tripathy: When patients progress on a CDK4/6 inhibitor, I will generally discontinue the use of CDK4/6 inhibitors. No clear data [show] that patients who become resistant [to a CDK4/6 inhibitor] may be sensitive to other [CDK4/6] inhibitors; however, some anecdotal data from individual hospital series suggest there could be some non-overlapping sensitivity.

However, we have many other options, which is why I will generally move on to another form of therapy. We need to formally study the impact of 1 CDK4/6 inhibitor after progression on another because we may find that some of these agents perform better in that setting.

Do you order tumor profiling after each progression? When you order the test, do you favor liquid or tissue biopsy?

I tend to get genomics at the very beginning when patients are newly diagnosed, although I may not use that information in the first-line setting. However, it is good to have [the genomic information], especially for truncal mutations. [Truncal mutations] tend to be prevalent in most of the cells, such as PIK3CA mutations. ESR1 mutations tend to be subclonal and may not be present [up front] but may arise later after the patient has been on an [AI].

If I haven’t gotten [genomic information] at the beginning, I will certainly get it at the time of progression on first-line therapy.

How do ESR1 mutations differ from other somatic mutations?

ESR1 mutations tend to arise in the presence of pretreatment, particularly with estrogen-deprivation therapy, which pharmacologically, would be AIs. The Cancer Genome Atlas published that ESR1 mutation rates were very low, at perhaps 1% or so at diagnosis and based on the patient’s primary tumor. However, [ESR1mutations] really [arise] after exposure to AIs.

That makes sense because AIs work by depriving estrogen from the environment. To be activated, ERs need to bind to estrogen. So, [AIs] select for any mutations that may occur where a mutation [is present] in the ER that allows it to be activated without the presence of estrogen; that is what most activating ESR1mutations do. Specifically, upon exposure to AIs, is when we start to see these mutations arise.

Lasofoxifene is an agent that is being evaluated in the setting of an acquired ESR1 mutation. How did the findings from the PADA-1 trial help to set the stage for this agent?

The PADA-1 trial showed that ESR1 mutations can be acquired and they can be subclonal, meaning that you can see a very low fraction of all the tumor DNA that has that mutation. However, you can’t underestimate that power of a small fraction of activated ESR1-mutant cells. We do see that pattern of resistance. [Additionally, the trial showed] that patients can have multiple clones.

How do you combat ESR1 mutations? One way is to totally take them out of the equation and to target them for downregulation or proteasomal degradation, which is what selective ER downregulators like fulvestrant [Faslodex] do. However, they don’t do it completely; these drugs don’t completely circumvent resistance. Another way is to use [SERMs] that may not necessarily inactivate the ER, but instead, modulate the ER in the way it behaves, the conformation of it, and how it binds to co-activators, co-receptors, and co-repressors. [Doing this], ultimately, mediates the expression of genes that guide what estrogen does, which is generally a growth signal. 

Have any in vitro data provided insight into the differences between lasofoxifene and other SERMs?

Lasofoxifene does have the ability to downregulate the transactivating nature of the ER, including ESR1mutations. In the lab, we can study how well an ER is able to activate the genes it targets. ERs are known to target numerous genes of interest that lead to cell growth. We can measure to what extent a drug will interact with ESR1 so that it negatively regulates it. In other words, the genes that are normally transcribed by estrogen are not transcribed. More importantly, growth is arrested. That has been demonstrated in vitro and in vivo using ESR1-mutant models.

At the virtual 2020 AACR Annual Meeting, the University of Chicago presented a poster on in vivodata looking at lasofoxifene alone and in combination with palbociclib [Ibrance] vs fulvestrant. What did we learn from these data and how could they potentially translate to the clinic?

Those data are compelling. They show that we can get growth inhibition in ESR1 mutations and that it can be potentially aided with CDK4/6 inhibitors. Although I don’t have the exact data at hand, I have seen examples of this and other data that show [tumor growth] can be reversed in cell-line models.

What we see in those cell-line models—even in animal models, which are more accurate but may not be the whole story—may not turn out to be the case in the tumor microenvironment and all of the other factors involved. Therefore, although this is very supportive and raises enthusiasm for getting these types of drugs into clinical trials, we need to wait to see what happens when patients are taking these drugs.

The ELAINE trial set the stage for ELAINEII, which is evaluating lasofoxifene plus abemaciclib. What is the anticipated impact of the ELAINEII trial?

The ELAINE trial is comparing fulvestrant with lasofoxifene in patients with ESR1 mutations. It’s a direct test of the hypothesis that [lasofoxifene] may be more favorable than [fulvestrant], which we now consider a reasonable drug to use when we see ESR1 mutations or in the second line in general. 

ELAINEII is designed to move [lasofoxifene] further, combining it with a CDK4/6 inhibitor. It will allow patients who have seen prior CDK4/6 inhibitors [to enroll]. There is some anecdotal evidence that responses have been seen with abemaciclib after progression on other CDK4/6 inhibitors.

The approval of abemaciclib was based on a trial that showed responses in refractory hormone receptor–positive disease as a single agent, but those patients had not been previously exposed to CDK4/6 inhibitors. We believe that the combination of these 2 different strategies could bring about more responses and longer time to progression.

The PARSIFAL trial didn’t look at lasofoxifene, but rather an AI plus palbociclib versus fulvestrant. How could those findings potentially affect the landscape?

One of the issues is pairing CDK4/6 inhibitors with different endocrine therapies; we haven’t formally compared that. We do know from the FALCON trial that, when using endocrine therapies alone, fulvestrant is slightly better than AI therapy, particularly in patients who haven’t been treated before or have non-visceral disease. Then, the MONALEESA-3 study was a trial that took advantage of the FALCON findings and compared treatment with fulvestrant alone versus fulvestrant plus ribociclib [Kisqali] regardless of first- or second-line therapy. That study showed a significant improvement in outcomes with a hazard ratio of around 0.5, which has been seen before.

The PARSIFAL trial, on the other hand, was a direct comparison with palbociclib. Patients received an AI or fulvestrant. The trial did not show a difference in progression-free survival. We are left without knowing what population fulvestrant might be best for, but certainly fulvestrant is a reasonable option in the first- or second-line setting. We have less data regarding what to do in the second line for patients who received fulvestrant up front, but the PARSIFAL and MONALEESA-3 studies showed that [ribociclib and palbociclib] are both reasonable agents to use and could be building blocks for [combinations] with other targeted agents.

How do you anticipate genomic classification will continue to affect the research landscape?

The sub-classification of the breast cancer subtypes is very important in general. In breast cancer, not all patients with luminal A or luminal B [disease] are the same. There may be certain aspects about the different pathways involved in growth and other cellular functions, such as invasion and apoptosis that may vary depending on different gene profiling. Gene profiling is basically a way to categorize cancers, but if we can categorize them in functionally relevant ways, then we are really making advances.

This takes studying large numbers of patients, ideally in the context of a clinical trial where they are being treated similarly or we are comparing 1 treatment with another. Then, we can dissect the different molecular profiles that may predict a benefit. More importantly, we can use that information about the molecular profiles to identify mechanisms of resistance that could inform newer strategies.

What other ongoing research is needed in the ESR1-mutant breast cancer space?

Not all mutations are the same. There may be certain ESR1 mutations that are generated in certain situations and under certain treatments that we need to understand better. What is really fascinating is trying to understand the functional consequences of ESR1 mutations, not only in terms of proliferation, but in terms of many other phenotypes that we are interested in targeting, such as metabolic activity, DNA repair, or other hallmarks of cancer. Whether it is invasion, the ability to resist apoptosis, or immunogenicity is critical. Studies in the basic aspects of what ERS1 mutations do [are important]. [Also, evaluation of] the nitty gritty molecular structure and how the ER interacts with co-activators, co-repressors, and other aspects of the transcriptional machinery that make the ER work [is needed] to understand what ESR1 does. 

Of course, we are aware of many other mutational isoforms, such as RAS mutations and growth factor mutations, like exon 20. We need to understand the inner workings of what these mutations cause. When you think about it, cancer is selection of the fittest. Cancer cells do have a higher mutational rate. Many of the mutations that arise out of accident lead to cell death. However, it is those that lead to cell advantage that perpetuate. By understanding the most common driver mutations that arise over time, we can start to develop therapies against each one. Although, there will always be rare mutations that are going to be more difficult to understand and target.