This article is more than 5 years old. Information may no longer be current.
Guest Commentary: Targeted therapies appear effective in breast cancer subtypes
Click Here to Manage Email Alerts
In this guest commentary, Debu Tripathy, MD, professor and chair of the department of breast medical oncology at The University of Texas MD Anderson Cancer Center and a HemOnc Today Editorial Board member, discusses several breast cancer abstracts from the ASCO Annual Meeting that focused on targeting the PI3K/AKT/mTOR pathway.
The movement toward targeted therapy for all cancers, including breast cancers, is advancing rapidly. As we have dissected many of the growth and survival pathways, as well as other phenotypes of cancer such as invasion and metastasis, we are beginning to understand these pathways in more detail. Importantly, we are developing targeted drugs that specifically block different components of these pathways.
One important pathway that has been targeted for quite some time, the PI3K/AKT/mTOR pathway, is driven by several inputs, and is employed by many different growth factors, including EGFR and HER2, and even the insulin receptor and some metabolic sensors. The PI3K/AKT/mTOR pathway can also interact with other cellular pathways. As a result, when we examine one pathway, we may not be fully cognizant of the other inputs and outputs in the pathway. As we develop drugs targeting these pathways, we use laboratory models to validate their impact prior to moving them into the clinic, then use correlative studies from trials to identify predictive markers to better personalize treatment.
We took another step forward in this process at ASCO through the dissection of one component of this pathway known as the AKT protein. This is a serine/threonine kinase enzyme that conducts the signal from PI3 kinase downward, which ultimately results in growth of the cell and in survival pathways as well as other signals. Activating mutations in AKT are rare in breast cancer, but alterations upstream of this pathway (e.g., PI3 kinase and PTEN) are more common and can be targeted with agents that inhibit several of the pathway’s components.
In breast cancer, particularly in hormone receptor-positive breast cancer, 30% to 40% of cancers harbor a mutation in the catalytic subunit known as PIK3CA, and many drugs are being developed to target that pathway. In triple-negative breast cancer, lesions in this pathway are also common. However, targeting AKT is a more recent development. The results of two phase 2 trials targeting AKT and a phase 3 trial using a mutant-selective (but not mutant-specific) PI3 kinase inhibitor were presented at ASCO.
Ipatasertib improves OS in triple-negative disease
LOTUS is a randomized phase 2 trial testing the addition of the AKT inhibitor ipatasertib (GDC-0068; Genentech/Roche) to paclitaxel compared with paclitaxel alone for patients with triple-negative, metastatic breast cancer in the first-line setting. In triple-negative breast cancer, there are many abnormalities in the so-called PAM pathway that involve abnormalities in PI3 kinase, AKT, mTOR and some other regulatory proteins like PTEN; PTEN is a negative regulator of PI3K/AKT, and loss of the protein can amplify this pathway. The researchers analyzed the patients’ tumors in order to classify individuals according to whether they had alterations in the PAM pathway.
PFS data from this trial was published previously by Kim and colleagues in Lancet Oncology. An update, which included survival data, was provided by Rebecca Dent, MSc, MD, FRCP, senior consultant in the department of medical oncology at the National Cancer Center in Singapore.
In that analysis, there seemed to be a difference in survival, although the difference was not statistically significant for the overall group. The OS analysis showed that median survival was 18.1 months in the ipatasertib group and 16.1 in the placebo group. In the subgroup of patients with PTEN-low tumors, it was more pronounced. However, if PAM alterations were pooled together, there was a bigger difference; median OS was 19.7 in the ipatasertib group and not quite reached in the placebo group.
It is important to remember that these data are all preliminary. The safety data that was presented suggested that this class of drugs is similar to PI3 kinase inhibitors, although their similarity is arguable, especially in this trial, because ipatasertib was combined with paclitaxel, and nausea, vomiting and fatigue were observed at slightly higher frequencies in the combination arm compared with the group that received paclitaxel alone.
Still, I think these data are exciting. These are phase 2 data, of course, so we need to await the results of the larger randomized trial. A phase 3 randomized trial that uses a similar design is looking specifically at patients with advanced, triple-negative breast cancer who have alternations in the PAM pathway; it will also include patients with hormone receptor-positive, HER-2-negative breast cancer. That is known as the IPATunity trial.
PAKT trial underscores importance of tumor testing
Another Phase II trial of an AKT inhibitor, which had a similar study design, examined AZD5363 (capivasertib, AstraZeneca). The data were presented by Peter Schmid, MD, PhD, lead of the Center for Experimental Cancer Medicine at Barts Cancer Institute, Queen Mary University of London.
The placebo-controlled PAKT trial enrolled 140 patients with previously untreated, metastatic, triple-negative breast cancer. Patients were randomly assigned to treatment with paclitaxel plus AZD5363 (n = 70) or paclitaxel plus placebo (n = 70). The primary endpoint was PFS. The study was powered to detect a one-third reduction in the number of progression-related events.
The researchers presented safety data at ASCO, which aligned with results observed for other drugs in this class. Diarrhea incidence increased; about 13% of patients treated with paclitaxel plus AZD5363 experienced diarrhea compared with 1% of patients receiving paclitaxel plus placebo. There was also some fatigue, rash and neuropathy. The toxicities observed here are slightly less than those seen with the PI3 kinase inhibitors, although that’s difficult to say for sure in smaller studies.
In terms of efficacy, PFS reached 5.9 months for patients in the treatment group and 4.2 months for patients in the placebo group. These results are not statistically significant; they narrowly missed the one-sided P value (.06).
This trial also analyzed tumors based on alterations in the PAM pathway; alterations were found in PIK3CA, AKT and PTEN. Altogether, about 25% of patients had mutations in some part of this pathway. When you look at PFS results based on the presence of alterations, there was a greater difference that was intriguing. Also, in the 25% of patients who had alterations in the PAM pathway, PFS was 9.3 months in the active treatment group and 3.7 months in the placebo group. This resulted in an HR of .3 and a P value of .01. In the group with no alterations in the PAM pathway, there was no real difference in PFS.
I think, with this class of drugs, we’re learning that one does have to test the tumor to see if there are any mutations or aberrations in the PAM pathway. It’s also important to recognize that we can’t detect all of this with next-generation sequencing alone. We can identify mutations, deletions, amplifications and other genomic-level changes, but we can’t pick up on things like loss of protein that might be due to epigenetic phenomena. With PTEN loss in particular, you may miss some of these changes, so immunohistochemistry would be more effective. This is something that will have to be sorted out as these drugs go through development. Predictive gene expression profile signatures may emerge as a better way to personalize many targeted therapies.
Finally, I will mention the OS results that were presented for this group of patients. Not only was there a difference overall, but it was even more dramatic in the group of 25% of patients who had an activation in the PAM pathway. The HR for mortality also decreased by nearly two-thirds, which was exciting. It wasn’t quite statistically significant, but these data are still immature and the numbers are small. Also, the survival benefit was limited to those with alterations in the PAM pathway. I think we’re seeing some activity with the AKT inhibitors, similar to what we’re seeing with the PI3 kinase inhibitors.
SANDPIPER demonstrates positive results, examines benefit vs. side effect profile
José Baselga, MD, PhD, FASCO, physician in chief of Memorial Sloan Kettering Cancer Center, presented results of the long-awaited registrational phase 3 trial of the mutant-selective PI3 kinase inhibitor taselisib (GDC0032/RG7604, Genentech/Roche). The SANDPIPER trial enrolled 480 patients whose tumors had known PIK3CA mutations. All patients received fulvestrant (Faslodex, AstraZeneca) and were randomized 2:1 to receive taselisib 4 mg daily or placebo. Another cohort of 120 patients without PIK3CA-mutated tumors was randomized in the same fashion.
The primary endpoint of PFS significantly improved with taselisib from a median of 5.4 months to 7.4 months (HR, 0.68, P = .0037). In the PIK3CA non-mutant cohort, median PFS was non-significantly improved from 4.0 to 5.6 months (HR, 0.67, P = .1062).
Toxicity was notable, with nearly 50% of patients experiencing grade 3 or higher toxicities compared to 16% in the placebo group; grade 3 or higher diarrhea was seen in 11.5% of patients receiving taselisib vs. 0.5% receiving placebo. Nausea and hyperglycemia were seen in 60% and 40.4% of patients in the taselisib arm, respectively, compared with 19.7% and 9.4%, respectively, in the placebo arm. Also, 16.8% of patients had to discontinue taselisib due to adverse events.
Upcoming trials will address side effects, add to survival data
At the current time, benefits from PI3K and AKT targeting in delaying PFS seem rather small, especially in view of the toxicities. What does all this mean?
There was a very interesting discussion about this question during a panel at ASCO. We know that the PI3 kinase pathway is essential for survival. It is used by normal cells to transmit growth and survival signals among many other phenotypes. But is it a pathway that we can successfully target with a drug? Is it a cancer-specific pathway?
We know this pathway is co-opted by cancer cells. When you look at whether a pathway is important to cancer, the first thing you do is see if tumors harbor aberrations in the genes/proteins involved, because – if there are alterations (including, most notably, activating mutations) – that means there is a selective advantage to activating that pathway. We also know, as I mentioned earlier, that there are many cancer-specific aberrations that lead to hyperactivation of the PAM pathway.
However, if we target the pathway, we are also going to induce collateral damage, because it is also used by normal cells. We are already seeing this happen in clinical practice; we’ve observed a lot of gastrointestinal toxicities and some skin toxicities, including rash, as well as hyperglycemia, because the insulin receptor uses this pathway as well. Psychological side effects, including depression, have also been observed, especially with the less selective inhibitors like the PI3 kinase inhibitors. There have even been reports of suicide attempts with the non-specific PI3 kinase inhibitors, which makes for a large spectrum of side effects with these agents.
We are also awaiting data on another PI3 kinase inhibitor, alpelisib (BYL719; Novartis), which is alpha-selective. This agent is being examined in the SOLAR trial, a large, randomized study for which results are expected in 2 or 3 months.
The discussion at ASCO focused on whether we will ever be able to selectively target this pathway without toxicities. We don’t know the answer to that question yet, but we do know that we have the ability to make more specific drugs, such as the alpha-specific PI3 kinase inhibitors for breast cancer, where the mutations are primarily in the alpha isoform (encoded by PIK3CA). That is a big step forward.
However, these drugs are not specific for the mutant PI3 kinase complex, so we may need to await the development of such molecules to fully realize the potential of PAM pathway targeting.
The AKT inhibitors are a little further behind. They also, based on the ipatasertib and capivasertib data that were presented at ASCO, might be a little more selective. The intriguing difference in survival that’s been observed in the early stages will, of course, need to be confirmed in larger, phase 3 trials, but I think this approach is going to result in another treatment advance. When this trial was conceptualized, PAM pathway aberrations were being targeted primarily in hormone receptor-positive breast cancer, but the ipatasertib and capivasertib trials are both being conducted in the setting of triple-negative breast cancer, and we have to remember that the PAM pathway is activated in these patients as well. It is also active in patients with HER-2-positive cancers. The results of additional trials in this area are going to be very informative and should lead to an addition to the armamentarium of treatments across different subtypes of breast cancer.
References:
Baselga J, et al. Abstract LBA1006. Presented at: ASCO Annual Meeting; June 1-5, 2018; Chicago.
Dent R, et al. Abstract 1008. Presented at: ASCO Annual Meeting; June 1-5, 2018; Chicago.
Schmid P, et al. Abstract 1007. Presented at: ASCO Annual Meeting; June 1-5, 2018; Chicago.
Disclosures: Tripathy reports serving as a consultant for Novartis and Pfizer and receiving research funding through his institution from Novartis.