For more than 100 years, endocrine therapy has been used for the treatment of breast cancer; since that time, we have learned much about the role of estrogen in the biology of normal breast cancer development.
Debu Tripathy, MD
For more than 100 years, endocrine therapy has been used for the treatment of breast cancer, initially with the use of oophorectomy when breast cancer was commonly seen in younger patients and at a more advanced stage. Since that time, we have learned much about the role of estrogen in the biology of normal breast development and in breast cancer.
The development of pharmacological therapy with tamoxifen in the late 1960s and early 1970s was a significant inflection point in the overall management of breast cancer. There was a clear demonstration of benefit as adjuvant therapy, cutting the risk of recurrence by half and breast cancer deaths by one-third. However, it is becoming clear that patients with early-stage breast cancer treated with endocrine adjuvant therapy face an ongoing risk of metastasis that does not seem to attenuate for up to 20 years.
In the metastatic setting, both endocrine and cytotoxic therapies can bring about responses and temporary remissions, but their impact on mortality has never been formally assessed because controlled trials are not appropriate given the palliative effects of these treatments. Other trials comparing combinations of treatments have shown incremental benefits in progression-free survival (PFS), but impacts on overall survival (OS) have been modest at best.
For endocrine-sensitive metastatic breast cancer, defined as the presence of estrogen receptors (ERs) or progesterone receptors (PRs) on 1% or more of the tumor cells, additional options have become available. The results of population studies suggest that patients with metastatic disease are living longer and better lives, but the exact impact of various therapies remains murky given changes in imaging technology that increasingly detect asymptomatic or minimally clinically evident metastatic recurrence along with improvements in supportive care. Furthermore, individual trials and tumor registries do not follow patients over the entire sojourn of their disease through multiple lines of therapy, limiting our understanding of the evolving natural history of advanced breast cancer.In the last 3 decades, aromatase inhibitors (AIs) and the downregulator fulvestrant (Faslodex) were added as options for advanced disease. Comparative trials of AIs, initially tested against tamoxifen, have shown small benefits in disease-free survival and OS but fewer complications such as thrombotic events; thus, AIs have been adopted as standard therapy for postmenopausal patients. Additionally, numerous consensus panels, supported by findings from several randomized trials, have advocated for endocrine therapy versus chemotherapy as initial treatment for most cases of ER-/PR-positive metastatic breast cancer.
Exceptions to this approach would include patients with symptomatic and rapid progression, including those with “visceral crisis.” These recommendations have clearly affected treatment patterns worldwide, with a higher proportion of patients now receiving endocrine therapy as first-line and even subsequent therapy.
Fulvestrant is commonly used in the second line and, with the appreciation that the dose of 500 mg every 4 weeks is better than the initially approved dose of 250 mg, the pivotal FALCON trial, which compared this agent with the AI anastrozole, showed a slight PFS benefit with fulvestrant. The population studied had no prior exposure to endocrine therapy for breast cancer. Patients in this study were therefore enriched for those with de novo disease presenting initially with advanced disease as opposed to recurrent from prior early-stage breast cancer. At the current time, both agents are appropriate as first-line therapy, and fulvestrant is recommended for patients recurring on or within 1 to 2 years of adjuvant AI therapy.Discoveries of mechanisms of resistance to endocrine therapy from the laboratory and correlative tumor analyses have shown that activation of growth factor—signaling pathways is a key mediator. Therapies that block these pathways, most notably through the inhibition of mammalian target of rapamycin (mTOR), a nodal point of this network, with the mTOR inhibitor everolimus (Afinitor), have been shown to be effective in nearly doubling PFS when added to either tamoxifen, exemestane, or fulvestrant.
However, no survival advantage has been demonstrated. Everolimus is commonly used in the second-line setting and is accompanied by toxicities such as stomatitis, significantly reduced with the use of a preventive steroid mouthwash containing 1% dexamethasone; hyperglycemia, fatigue, and nausea and, less commonly, pneumonitis, skin rash, and immune suppression. Newer strategies are targeting other components of signal transduction, including AKT (protein kinase B), PI3K (phosphoinositide 3-kinase), FGFR (fibroblast growth factor receptor), and other pathways.Efficacy Studies
The cell cycle comprises a set of tightly controlled signaling networks that coordinate the expression of genes needed to govern entry into different phases of cell division (eg, DNA replication, mitosis) when appropriate, such as stimulation by nutritional factors or growth factor activation, or to stop cell division, including DNA damage and nutrient inavailability. It is therefore expected that cancers can possess abnormalities in the cell cycle.
Efforts to target the cell cycle for cancer therapy, in particular by inhibiting cyclin-dependent kinases (CDKs), since kinases are easier to “drug”, were stalled for many years because the nonspecific CDK inhibitors tested were too toxic. However, as CDK4/6 inhibitors that focus on entry into the S-phase (Figure) were developed, it became apparent that one of the types of cancers most amenable to treatment in preclinical models was hormone receptor—positive/HER2-negative breast cancer.
The first of these inhibitors to enter phase III trials was palbociclib (Ibrance) in postmenopausal patients with ER- or PR-positive, HER2-negative advanced breast cancer without prior therapy for advanced disease. The drug demonstrated a clear improvement in PFS: The median PFS increased from 14.5 months with letrozole alone to 24.8 months with the addition of palbociclib, for a hazard ratio (HR) of 0.58.
The second CDK4/6 inhibitor tested and approved was ribociclib (Kisqali), which was evaluated in a trial of similar eligibility and design as the palbociclib study and produced roughly equivalent benefits. A third CDK4/6 inhibitor, abemaciclib (Verzenio), which has a slightly higher affinity for CDKs 4 and 6, as well as a small amount of activity against CDKs 1, 2, 5, and 9, also showed similar outcomes in a trial that, again, had an essentially equivalent design as the other 2 studies. This drug is dosed continuously without a week off, which is not necessary given its lower hematological toxicity profile. The indications for the approved CDK4/6 inhibitors are listed in Table 1. Summaries of the key trials that have been conducted with these drugs are provided in the Table 2.1-11
For premenopausal patients, the benefit of adding medical therapy for ovarian blockade gonadotropin-releasing hormone (GnRH) therapy to tamoxifen for advanced hormone-sensitive breast cancer was shown nearly 20 years ago. The rapid adoption of AIs and fulvestrant, which are only effective in the post-menopausal state, further consolidated the use of either ablative (ie, surgical removal of or radiation to the ovaries) or medical ovarian suppression in premenopausal advanced breast cancer.
One of the CDK inhibitor trials focused on pre/perimenopausal patients, where all patients received the GnRH analogue goserelin and palbociclib was compared with placebo in combination with either a nonsteroidal AI or tamoxifen. This trial yielded equivalent treatment effects as were seen in the other trials, including with either AI or tamoxifen as the endocrine partner. A subset of premenopausal patients treated with GnRH in first-line therapy trials with palbociclib and abemaciclib also showed outcomes comparable to the overall study group.
In second-line therapy, each of the clinically available CDK4/6 inhibitors have shown a benefit, similar in terms of HRs, although the absolute differences in PFS were not as large due to the overall shorter PFS as expected in the second line. However, the MONALEESA-3 trial was the only study that tested fulvestrant with or without the CDK inhibitor ribociclib in both the first and second lines; the results showed a longer PFS in all arms and a similar ribociclib HR benefit of 0.56 as the other trials. The better outcome in all arms might be due to the small superiority of fulvestrant over anastrozole seen earlier in the FALCON trial for endocrine therapy-naïve patients.
Abemaciclib has also been tested in later lines of therapy in patients refractory to multiple therapies and shown single-agent activity (response rate, 20%; clinical benefit rate, 42%), so it is approved for this indication.
On the basis of these trials, the ASCO and other guidelines organizations include the use of CDK 4/6 inhibitors as options for first- and second-line therapies, but they also recommend AI or fulvestrant therapy alone since differences in survival have not yet been proven.
Toxicities, Monitoring, and Dose Adjustments
As expected with inhibition of the cell cycle, rapidly dividing hematopoietic progenitor cells will be affected. A notable impact seen thus far with CDK4/6 inhibitors has been on neutrophil counts, which motivated the 1 week off at the end of every 28-day cycle for palbociclib and ribociclib. Therefore, blood counts are recommended every 2 weeks for the first 2 cycles and every 4 weeks thereafter for palbociclib and ribociclib, with a delay and recheck 1 week later for neutrophil count under 1000. If the drug is delayed 2 or more times or for more than 2 weeks, dose reduction should be implemented as it would be for other grade 3 toxicities that are more than transient.
Abemaciclib has less in the way of hematological adverse effects, although it has more gastrointestinal (GI) effects. GI symptoms, particularly diarrhea, need to be monitored closely and treated promptly with standard antidiarrheals.
Ribociclib and, to a lesser extent, the other CDK inhibitors, can cause prolongation of the corrected QT interval (cQT). Ribociclib use requires EKG testing for the first 2 cycles and interruptions/cessation for cQT prolongation of >480/500 ms. Since tamoxifen alone was shown to prolong the cQT interval, it is not recommended or approved in combination with CDK inhibitors. Ribociclib also requires monitoring and possible dose adjustments for liver function test abnormalities.
All of the agents can show potential drug interactions with strong inducers/inhibitors of CYP3A4, and concomitant use of such drugs should be avoided.
Predictive Factors and Mechanisms of Resistance
Little is known about prognostic and predictive factors specifically for CDK inhibitors. In all the trials, subset analyses were unable to reveal any striking differences in the PFS HR in any of the clinical and pathological subsets. There was a suggestion from the first-line abemaciclib trial that subsets with a disease-free interval from early stage to metastasis of >36 months have quite a long PFS in either arm with significant differences, but this finding is still tentative as an exploratory observation.
Although it was speculated that cyclin D amplification and retinoblastoma (Rb) gene functional loss would predict differential benefit based on the known biology, this was not seen in early smaller studies. However, the results of larger studies seem to show some trends, and early data suggest that elevated cyclin E mRNA may predict relative resistance as it bypasses CDK 4/6 via CDK2.
Mutations in ESR1 and PIK3CA do not appear to affect treatment benefit (ie, PFS HRs), but studies of serially measured circulating tumor DNA suggest complex enrichment and depletion of subclones involving many genes over the course of treatment. Data are starting to become available from tissue- and blood-based tests from large trials, and these may better elucidate drivers of resistance that could inform future combinatorial trials for targeted therapies after progression.Currently, several early-phase trials are testing biological combinations that are rationally derived based on correlative tissue studies from trials or emerging data from the laboratory. For example, a neoadjuvant study with abemaciclib induction of immune infiltrates with CD8-positive cells suggested that combination with immune checkpoint inhibition may be effective and led to promising results in an early-phase trial.
Combinations with mTOR, PI3K, FGFR, and other tyrosine receptor kinase and signal transduction inhibitors are also in progress. There are also trials assessing CDK inhibitor benefit with HER2-targeted agents in patients with HER2-positive and ER-/PR-positive disease. Given the elevated and persistent recurrence risk for high-risk early-stage ER-/PR-positive/HER2-negative breast cancer, large phase III adjuvant trials combining standard endocrine therapy with and without each of the approved CDK inhibitors for 2 to 3 years are also ongoing.