HR+ HER2−: The First SERD in a Phase III Trial and Biomarker in CDK4/6 Inhibitors
ESR1 mutations for selection of endocrine counterpart for treatment with SERDs
With mounting evidence that an ESR1 mutation may be a resistance marker for aromatase inhibitor therapy [1 ], it is essential to understand whether this
biomarker can be of clinical relevance. A mutation in ESR1 leads to the estrogen receptor being constitutionally switched on, regardless of
whether the receptor complex is activated
by estrogen [1 ]. In this situation, treatment with aromatase inhibitors or a selective estrogen
receptor modulator (SERM) cannot downregulate the activity of
the estrogen receptor. However, SERDs are thought to be able to downregulate
both the wild-type and mutant forms of the estrogen receptor ([Fig. 1 ]).
ESR1 mutations are known to occur in only about 5% of cases in patients without prior
treatment, whereas they are detected in up to 30 – 40% of cases in patients whose
disease had
progressed to or have relapsed on an aromatase inhibitor [2 ], [3 ]. It must therefore be assumed that ESR1 mutations
accumulate during treatment with aromatase inhibitors and that this is one of
the resistance mechanisms that reduces the effectiveness of aromatase inhibitors.
Fig. 1 Possible mechanisms of endocrine resistance associated with an estrogen receptor
(ESR1) mutation.
One of the trials investigating the clinical utility of this approach is the PADA-1
trial [4 ]. The PADA-1 trial included patients treated with palbociclib and
an aromatase inhibitor as first-line therapy without evidence of endocrine resistance.
During treatment, circulating tumor DNA in blood (ctDNA) was tested for evidence of
an ESR1
mutation before treatment, 1 month after the start of treatment and every 2 months
thereafter. Only a few relevant mutations have been described for the ESR1 gene, so genotyping can
be restricted to a few genomic loci. In the case of the PADA-1 trial, mutations
were determined at the following locations: E380, P535, L536, Y537, D538 [4 ].
If ESR1 mutations were found and there was no clinical evidence of progression, patients
were randomized to one of two treatment arms. In one arm, the previous treatment was
continued, while in the experimental arm, endocrine treatment was substituted
for the SERD fulvestrant while continuing palbociclib.
The analysis of progression-free survival showed that patients who switched to fulvestrant
had a median PFS of 11.9 months (95% CI: 9.1 – 13.6), which was significantly longer
than patients
who continued treatment with an aromatase inhibitor. In this group, the median
PFS was 5.7 months (95% CI: 3.9 – 7.5). The hazard ratio was 0.61 (95% CI: 0.43 – 0.86)
[4 ]. Overall survival data have not yet been presented [4 ]. Patients who were not switched at the time of “molecular progression”
went on to receive fulvestrant and palbociclib once progression was detectable
on imaging; PFS in this group was 3.5 months (95% CI: 2.7 – 5.1), so the actual gain
in PFS in the intervention
arm was only 2.7 months. The question therefore arises as to whether the patients
in the intervention arm actually benefited or whether they were simply switched over
earlier. The concept of
“molecular progression” has been clinically validated for the first time in PADA-1
and proof-of-concept has been obtained. However, it is too early for clinical application,
and it remains
to be seen whether ongoing trials with more effective interventions will provide
a clearer picture.
Currently, several trial programs are pursuing a similar strategy with oral SERDs.
For example, one such ongoing trial is SERENA-6 [5 ].
EMERALD trial published
With a large number of SERDs currently in clinical development [6 ], [7 ], the first phase III trial, the EMERALD trial, has now
been published [8 ]. This study is of particular interest not only in terms of whether SERD treatment
can overcome endocrine resistance based on an ESR1
mutation, but also in terms of a comparison between the oral SERD elacestrant
and fulvestrant. The EMERALD trial included patients with advanced breast cancer who
had previously been treated
with at least one CDK4/6 inhibitor plus endocrine treatment. However, patients
who had received additional endocrine treatment or prior chemotherapy were also eligible.
After randomization,
patients were treated with either the SERD elacestrant or standard treatment
(fulvestrant, exemestane, letrozole, or anastrozole) [8 ]. Of particular interest
are the stratification factors of the ESR1 mutation in circulating ctDNA and prior treatment with fulvestrant.
In the overall analysis, elacestrant was shown to improve progression-free survival
compared with standard treatment, with a hazard ratio of 0.697 (95% CI: 0.552 – 0.880).
However, median
PFS only improved from 1.9 months to 2.8 months in this heavily pre-exposed population.
Also, more than 40% of the patients treated with elacestrant had primary progression.
In the group of
patients with an ESR1 mutation in ctDNA, median PFS improved from 1.9 months to 3.8 months (HR = 0.546;
95% CI: 0.387 – 0.768). Comparing patients treated with elacestrant to patients
on fulvestrant therapy, the HR was 0.684 (95% CI: 0.521 – 0.897) in the overall
group and 0.504 (95% CI: 0.341 – 0.741) in the group with ESR1 mutation [8 ].
It was therefore shown that in this heavily pre-exposed population, the oral SERD
elacestrant was more effective than fulvestrant or treatment with aromatase inhibitors.
However, in a
population like this, primary progression rates are very high and median progression-free
times are very short, so it is not possible to make a robust assessment of whether
endocrine
resistance can be overcome with this treatment. However, the results are certainly
promising, especially for future combination therapies, for which elacestrant may
be a better candidate
than intramuscular fulvestrant. Studies carried out under previous treatment
regimens will show whether oral SERDs such as elacestrant, bring substantial further
benefit to the treatment of
patients with HR+, HER2− breast cancer.
Interim analyses in the MONARCH-3 trial
The Summary of Product Characteristics, published in January 2022, includes a new
interim analysis of the MONARCH-3 trial [9 ], which compared abemaciclib plus
an aromatase inhibitor and aromatase inhibitor alone as first-line therapy. The
most recent interim analysis of 255 deaths showed a median OS of 54.5 months with
an aromatase inhibitor
alone, and 67.1 months with combination therapy ([Fig. 2 ]). This corresponded to a hazard ratio of 0.754 (95% CI: 0.584 – 0.974; p = 0.0301).
The p-value
did not achieve the threshold for significance [9 ] required for the interim analysis, so the next analysis is awaited.
Fig. 2 Overall survival in the MONARCH-3 trial after 255 deaths (data from [9 ]).
CDK4/6 inhibitor therapy and BRCA mutations
Based on retrospective analyses, it has been suggested that a germline mutation in
BRCA1/2 (gBRCA1/2 ) may reduce the effectiveness of CDK4/6 inhibitors. When comparing
gBRCA1/2 mutation carriers and patients without mutations, these retrospective analyses showed
a hazard ratio of 1.50 (95% CI 1.06 – 2.14) [10 ].
The same question was now studied in a high-quality cohort of patients with advanced
breast cancer using retrospective-prospective data. A total of 223 gBRCA1/2 mutations were
detected in 4460 patients (101 in BRCA1 and 122 in BRCA2 ). This resulted in a mutation rate of 4.8%, which is similar to the figures from
a large real-world analysis of
gBRCA1/2 mutation rates [11 ].
The aforementioned cohort included a total of 1005 patients who had been treated with
a CDK4/6 inhibitor. 45 patients with a gBRCA2 mutation had worse PFS with a hazard ratio of 2.12
(95% CI: 1.48 – 3.03). When restricted to patients treated as part of first-line
therapy (n = 439), patients with a gBRCA2 mutation (n = 24) also had worse PFS (HR = 2.32; 95% CI:
1.38 – 3.91).
Overall, however, the patients treated in this cohort appeared to have a worse prognosis.
The median PFS for first-line therapy in wild-type patients was 14.7 months. Although
this analysis
provides good evidence that a gBRCA1/2 mutation is associated with somewhat poorer efficacy with a CDK4/6 inhibitor, the
results should be confirmed before they are applied more
broadly. However, in the presence of a gBRCA1/2 mutation, this study provides good arguments for case-by-case decision-making for
possibly starting first-line therapy with PARP
inhibitors.
Mutation analysis of ctDNA and the efficacy of ribociclib therapy
The excellent outcomes seen with CDK4/6 inhibitors in treating metastatic disease
[12 ], [13 ], [14 ], [15 ], [16 ], [17 ], [18 ], [19 ], [20 ], [21 ], [22 ], [23 ], [24 ], [25 ], [26 ], [27 ] highlight the importance
of this treatment for patients with advanced HER2−/HR+ breast cancer. It has
become the gold standard in first-line therapy [28 ] and has largely replaced
endocrine monotherapy and chemotherapy as first-line therapy [29 ]. It is likely that drug development will continue to be guided by this standard
for many
years to come. This makes it all the more important to increase our understanding
of resistance mechanisms and molecular patterns of efficacy. This will help identify
new drug targets and
establish surveillance mechanisms. One trial with these objectives is BioItaLEE,
the preliminary results of which have been published [30 ].
The BioItaLEE trial included patients treated with ribociclib and letrozole in an
endocrine-sensitive setting during first-line therapy. Extensive biomarker sampling
was performed before
treatment, after 15 days, and on the first day of the second cycle. Using the
blood samples, exons from a total of 39 breast cancer-related genes were sequenced.
A total of 263 patients were
included in the trial. In one of the genes analyzed, mutations were found in
113 patients. Patients without a genetic mutation had a significantly better prognosis,
with a hazard ratio of
0.41 (95% CI: 0.27 – 0.61) [30 ]. In 49 of the patients who had a mutation prior to starting treatment, the treatment
was able to eliminate the ctDNA carrying
the mutation in the blood. By grouping the patients according to their mutation
status before treatment and after 15 days, it was possible to identify different prognostic
groups:
no mutation before treatment → no mutation after 15 days (n = 115)
no mutation before treatment → mutations after 15 days (n = 19)
mutations before treatment → no mutation after 15 days (n = 44)
mutations before treatment → mutations after 15 days (n = 60)
Patients with persistent mutations had the worst prognosis with a median PFS of 12.3
months. Patients with no mutations at either point in time had the best prognosis,
with a median
follow-up of 26.9 months and a median PFS in the overall population of 23.4 months
in the group that had not yet attained median PFS. Median progression-free survival
is shown in [Fig. 3 ]
[30 ].
Fig. 3 Prognosis on treatment with ribociclib and letrozole according to detection status
of mutations before starting treatment and after 15 days (data from [30 ]).
Future studies are needed to determine whether other treatments may be a better option
for patients with a poor prognosis. It must be kept in mind that CDK4/6 inhibitors
are a very
effective treatment with an acceptable side effect profile and that a worse prognosis
based on biomarkers does not automatically imply that a better outcome can be achieved
with an
alternative treatment.
Choosing Treatment After Biomarker Testing
The use of biomarkers to choose a targeted treatment instead of chemotherapy
As our knowledge of biomarkers increases and potential treatments emerge based on
them, the question arises as to whether the available treatments and knowledge of
so-called “actionable
genetic variants” (mutations/amplifications/translocations that indicate the
efficacy of a targeted treatment) are sufficient to decide which patients need chemotherapy
and which patients
would be better served by a targeted treatment.
One trial that investigated this is SAFIR02 [31 ]. Patients with advanced HER2-negative breast cancer who were stable and without
progression after 6 – 8
cycles of chemotherapy were enrolled in this trial. These patients were tested
for the following so-called “actionable genetic alterations” for the following treatments:
alpelisib, olaparib,
capivasertib, vistusertib, AZD8931, vandetanib, bicalutamide, AZD4547 and selumetinib.
However, it must be kept in mind that there is a probable ESCAT category I or II association
for only a
proportion of these drugs [32 ]. According to these ESCAT I and II categories, the following targeted treatments
were administered in the presence of the
corresponding mutations (number of patients treated in parentheses): Olaparib
for a gBRCA1/2 mutation (n = 60), alpelisib for a PIK3CA mutation (n = 31), capivasertib for an
AKT1 mutation (n = 21), and sapitinib for an EGFR mutation (n = 3). In this group, patients treated with targeted treatment had a longer
median PFS of 9.1 months (95% CI:
7.1 – 9.8) compared with the group in which chemotherapy was continued (median
PFS: 2.8 months; 95% CI: 2.1 – 4.8). It should be noted, however, that the patients
who accounted for a large
proportion of the overall effect were those treated with olaparib (HR = 0.29;
95% CI: 0.17 – 0.49) [31 ].
In the group of patients treated with targeted therapy based on proven mutations that
did not fall into ESCAT categories I or II, no improvement in PFS was observed compared
with continuing
chemotherapy (HR = 1.15; 95% CI: 0.76 – 1.75). In this respect, SAFIR02 represents
an important proof-of-concept for molecular tumor boards, but also further clarifies
that this is limited
exclusively to ESCAT tier I or II alterations.
In principle, the SAFIR02 trial confirms the results of the OlympiAD trial, in which
patients treated with olaparib monotherapy had better PFS than patients treated with
chemotherapy [33 ]. However, the SAFIR02 trial shows that it may also be appropriate to switch from
chemotherapy to which patients had responded or even stable disease to
treatment with olaparib in the case of a gBRCA1/2 mutation.
Role of PD-L1 expression on progression-free survival and overall survival in the
KEYNOTE-355 trial
The KEYNOTE-355 trial evaluated the addition of pembrolizumab to standard chemotherapies
in the first advanced line of therapy in patients with metastatic triple-negative
breast cancer
(TNBC [34 ]. Due to the study design, the primary analysis focused on patients with a PD-L1
expression of ≥ 10 as determined by the Combined Positive Score
(CPS). However, patients with lower PD-L1 expression were also included. In this
context, the question is whether patients with lower levels of PD-L1 expression might
also benefit from
treatment with pembrolizumab, taking account of its side effects. In this context,
extensive analyses of KEYNOTE-355 have now been published [35 ].
In the overall population of KEYNOTE-355 patients regardless of PD-L1 expression,
median progression-free survival improved from 5.6 months to 7.5 months (HR = 0.82;
95% CI: 0.70 – 0.98),
while overall survival improved from 15.5 months to 27.2 months (HR = 0.89; 95%
CI: 0.76 – 1.05) [34 ]. In the pre-specified population of patients with a CPS
of ≥ 10, analyses showed statistically significant superiority of pembrolizumab
combination therapy with respect to both outcome parameters [34 ].
[Fig. 4 ] shows the hazard ratios for PFS and OS as a function of PD-L1 expression (CPS < 1;
CPS 1 – 9; CPS 10 – 19; and CPS ≥ 20). There is evidently a
consistent improvement in the hazard ratio in favor of pembrolizumab therapy
in progression-free survival from 1.09 at a CPS of 0 to a HR of 0.62 at a CPS of ≥ 20
[34 ]. These significant effects could not be shown with respect to overall survival.
This showed hazard ratios of approximately 1 in patients up to a CPS of 0 – 9 and
an HR of
approximately 0.7 in both groups with a CPS ≥ 10 [34 ].
Fig. 4 Hazard ratios for comparison of pembrolizumab + chemotherapy vs. chemotherapy as
a function of CPS score in the KEYNOTE-355 trial [34 ].
Thus, the established CPS cut-off of 10, at which it can be assumed that a benefit
in terms of overall survival and progression-free survival has been achieved, therefore
seems to be
appropriate.
Triple-negative Patients – Further Development of Antibody-Drug Conjugate
A new ADC with Trop2 as the target structure
The impressive results of therapy with the anti-Trop2 antibody-drug conjugate (ADC)
sacituzumab govitecan in patients with heavily pretreated advanced TNBC in the ASCENT
trial have focused
attention on this target. In the ASCENT trial, in patients with advanced triple-negative
breast cancer, median progression-free survival was significantly improved with sacituzumab
govitecan
compared with chemotherapy of the physicianʼs choice (capecitabine or eribulin
or vinorelbine or gemcitabine) (HR = 0.41; 95% CI: 0.32 – 0.52) and median overall
survival approximately
doubled-from 6.7 months to 12.1 months (HR = 0.48; 95% CI: 0.38 – 0.59) [36 ].
Trop2 is an antigen that is overexpressed in some cancers such as breast cancer, some
thyroid cancers, pancreatic cancer, colon cancer, urothelial cancer, and other tumors
[37 ], [38 ], [39 ]. It is thought to be involved in various signal transduction pathways ([Fig. 5 ]).
Fig. 5 Signaling pathways described in the context of Trop-2 (A – C ) (Source: Liao S, Wang B, Zeng R et al. Recent advances in trophoblast cell-surface
antigen 2 targeted
therapy for solid tumors. Preprints 2020; 2020120062. https://www.preprints.org/manuscript/202012.0062/v1 . Creative
Commons License CC BY 4.0).
Some ADCs are currently in clinical development [40 ]. Data from a study investigating a different ADC, datopotamab deruxtecan, have now
been published [41 ]. Similar to sacituzumab govitecan, the payload is a topoisomerase I inhibitor. In
the TROPION-PanTumor01 trial, 44 patients with advanced triple-negative
breast cancer were treated, among other cancer types, 30 (68%) of whom had undergone
two or more prior treatments for advanced TNBC [41 ]. A response was seen
in 15 patients (34%) and 17 had stable disease. Interestingly, 14 of 27 patients
(52%) who had already been pretreated with another topoisomerase I inhibitor-based
ADC also responded to
datopotamab deruxtecan. Nausea/vomiting and stomatitis were the most common side
effects, whereas hematological toxicity and diarrhea were uncommon, occurring in only
15 – 20% [41 ].
In the ASCENT trial of sacituzumab govitecan, the response rate was 31%, which is
very similar to the response rate seen in the TROPION-PanTumor01 trial [36 ], [41 ]. It is hoped that ADCs of this kind will be developed for earlier disease stages
as soon as possible. Sacituzumab govitecan, for example,
is already being tested in the post-neoadjuvant setting in the SASCIA trial,
which is currently recruiting patients [42 ].
Brain Metastases in Focus in Patients with HER2+ aBC
Brain metastases in the HER2CLIMB trial
Data from the HER2CLIMB trial, which investigated therapy with tucatinib, trastuzumab,
and chemotherapy in the treatment setting after trastuzumab/pertuzumab and T-DM1,
showed an
improvement in PFS and OS compared with trastuzumab and chemotherapy even in
the primary analysis. This study was interesting in that it also enrolled patients
who had newly diagnosed or
progressive (“active”) cerebral metastasis without prior local treatment when
it was not immediately necessary. The presence of brain metastases (yes vs. no) was
also a preplanned
stratification factor. In addition, patients with brain metastases at baseline
were divided into those with active and stable brain metastases. All patients underwent
MRI of the brain at
baseline and were assigned to the following groups: [treated and stable], i.e.
patients who had received prior local treatment and had not progressed at the time
of enrollment; treatment may
have been given during the screening period; and [treated and progressive], i.e.
patients who had been treated for brain metastases in the past and had progressed
at the time of enrollment.
Patients who had not received local pretreatment were also included in this group.
In addition, the inclusion and exclusion criteria listed in [Table 1 ]
were applied. A total of 117 patients with stable brain metastases and 174 patients
with active brain metastases were enrolled in the HER2CLIMB trial.
Table 1 Original texts for the inclusion and exclusion criteria related to brain metastases
in the DESTINY-B03 and HER2CLIMB trials (according to [46 ] and [47 ]).
Original text
Inclusion criteria
HER2CLIMB trial
CNS Inclusion – Based on screening contrast brain magnetic resonance imaging (MRI), patients must
have one of the following:
1. No evidence of brain metastases
2. Untreated brain metastases not needing immediate local therapy. For patients with
untreated CNS lesions > 2.0 cm on screening contrast brain MRI, discussion with and
approval from the medical monitor is required prior to enrollment
3. Previously treated brain metastases
a. Brain metastases previously treated with local therapy may either be stable since
treatment or may have progressed since prior local CNS therapy, provided that there
is no
clinical indication for immediate re-treatment with local therapy in the
opinion of the investigator
b. Patients treated with CNS local therapy for newly identified lesions found on contrast
brain MRI performed during screening for this study may be eligible to enroll if all
of
the following criteria are met:
i. Time since whole brain radiation therapy (WBRT) is ≥ 21 days prior to first dose
of study treatment, time since stereotactic radiosurgery (SRS) is ≥ 7 days prior to
first dose
of study treatment, or time since surgical resection is ≥ 28 days
ii. Other sites of disease assessable by RECIST 1.1 are present
4. Relevant records of any CNS treatment must be available to allow for classification
of target and non-target lesions
Original text
Exclusion criteria
HER2CLIMB trial
CNS Exclusion – Based on screening brain MRI, patients must not have any of the following:
1. Any untreated brain lesions > 2.0 cm in size, unless discussed with medical monitor
and approval for enrollment is given
2. Ongoing use of systemic corticosteroids for control of symptoms of brain metastases
at a total daily dose of > 2 mg of dexamethasone (or equivalent). However, patients
on a
chronic stable dose of ≤ 2 mg total daily of dexamethasone (or equivalent)
may be eligible with discussion and approval by the medical monitor
3. Any brain lesion thought to require immediate local therapy, including (but not
limited to) a lesion in an anatomic site where increase in size or possible treatment-related
edema may pose risk to patient (e.g. brain stem lesions). Patients who undergo
local treatment for such lesions identified by screening contrast brain MRI may still
be eligible for
the study based on criteria described under CNS inclusion criteria 19b
4. Known or suspected leptomeningeal disease (LMD) as documented by the investigator
5. Have poorly controlled (> 1/week) generalized or complex partial seizures, or manifest
neurologic progression due to brain metastases notwithstanding CNS-directed
therapy
Original text
Exclusion criteria
DESTINY-B03 trial
Spinal cord compression or clinically active central nervous system (CNS) metastases,
defined as untreated or symptomatic, or requiring therapy with corticosteroids or
anticonvulsants to control associated symptoms.
Subjects with clinically inactive brain metastases may be included in the study.
Subjects with treated brain metastases that are no longer symptomatic and who require
no treatment with corticosteroids or anticonvulsants may be included in the study
if they
have recovered from the acute toxic effect of radiotherapy. A minimum
of 2 wk must have elapsed between the end of whole brain radiotherapy and study enrollment.
Using this categorization, patients with active brain metastases treated with trastuzumab
and chemotherapy had a median PFS of 4.1 months (95% CI: 2.9 – 5.6) and patients with
stable brain
metastases had a median PFS of 5.6 months (95% CI: 3.0 – 9.5). In both populations,
PFS was improved by tucatinib with a hazard ratio of 0.36 (95% CI: 0.22 – 0.57) for
patients with active
brain metastases and a hazard ratio of 0.31 (95% CI: 0.14 – 0.67) for patients
with stable brain metastases [43 ].
Detailed data have now been published for patients who had brain metastases at the
start of treatment in the DESTINY-B03 trial (n = 82) [44 ]. The exclusion
criteria relating to brain metastases are listed in [Table 1 ]. Accordingly, patients with untreated or symptomatic brain metastases were not eligible
for
inclusion in the study. Nevertheless, patients in the comparator arm (T-DM1)
with these criteria for brain metastases had a median PFS of only 3 months (95% CI:
2.8 – 5.8). The median PFS
for patients with brain metastases was improved by T-DXd to 20.9 months (95%
CI: 8.7 – 36.6) (HR = 0.25; 95% CI: 0.13 – 0.45) [44 ]. Of 82 patients with brain
metastases, 72 had a target lesion in the brain, 36 in the T-DXd arm and 36 in
the T-DM1 arm. In the T-DXd arm, 10 patients (27.8%) achieved complete remission compared
with one patient
(2.8%) in the T-DM1 arm [44 ].
Without comparing HER2CLIMB and DESTINY-B03 in terms of the efficacy of their investigational
substances, patients treated with T-DM1 in the DESTINY-B03 trial do not appear to
have been a
more stable population in terms of progression than patients treated in the comparator
arm of the HER2CLIMB trial. In quantitative terms, the median PFS in the DESTINY-B03
trial for this
subgroup in the comparator arm was even shorter (3 months) than in the HER2CLIMB
trial (4.1 months). It should be noted that the population of patients in the DESTINY-B03
trial was much
smaller than in the HER2CLIMB trial, so the data on brain metastases will certainly
require improvement. It is also unclear whether therapy with T-DM1 may be less effective
than therapy with
trastuzumab and chemotherapy. In the KAMILLA trial, patients with brain metastases
predominantly after trastuzumab pretreatment had a response rate of 21.4% and a median
PFS of 5.5 months
(95% CI: 5.3 – 5.6 months) [45 ]. In the DESTINY-B03 trial, the response rate with T-DM1 was 33.4% and the median
PFS was 3.0 months [44 ].
