ORIGINAL RESEARCH

A phase II study of pembrolizumab plus carboplatin in BRCA-related

metastatic breast cancer (PEMBRACA)
L. Cortesi1*, M. Venturelli1, G. Cortesi1, F. Caggia1, A. Toss1, E. Barbieri1, U. De Giorgi2, V. Guarneri3,4, A. Musolino5,6,
E. De Matteis7, A. Zambelli8, G. Bisagni9 & M. Dominici1
1
Department of Oncology and Haematology, University Hospital Modena, Modena; 2Department of Medical Oncology, “Dino Amadori Scientific Institute of Romagna
for the Study of Cancer”, Meldola; 3Department of Surgery, Oncology and Gastroenterology, University of Padua, Padova; 4Division of Oncology, Istituto Oncologico
Veneto IRCCS, Padova; 5Department of Medicine and Surgery, University of Parma, Parma; 6Medical Oncology, Breast Unit and Cancer Genetics Service, University
Hospital of Parma, Parma; 7Oncology Unit, Hospital “Vito Fazzi”, Lecce; 8Papa Giovanni XXIII Cancer Center Hospital, Bergamo; 9Oncology Department, Azienda USL-
IRCCS di Reggio Emilia, Reggio Emilia, Italy

Available online 5 April 2023

Background: BRCA1/2-related metastatic breast cancers (mBC) are sensitive to DNA-damage agents and show high
tumor-infiltrated lymphocytes. We hypothesized that the association between pembrolizumab and carboplatin could
be active in BRCA-related mBC.
Patients and methods: In this phase II Simon’s design multicenter single-arm study, BRCA1/2-related mBC patients
received carboplatin at area under the curve 6 every 3 weeks for six courses associated with 200 mg
pembrolizumab every 3 weeks until disease progression or unacceptable toxicity. The primary aim at first stage was
overall response rate (ORR)
70%. Disease control rate (DCR), time to progression (TTP), duration of response
(DOR), and overall survival (OS) were the secondary aims.
Results: Among 22 patients enrolled at the first stage, 5 BRCA1 and 17 BRCA2, 16 (76%) were luminal tumors and 6
(24%) triple-negative BC (TNBC). In 21 patients, ORR and DCR were 43% and 76% (47% and 87% in luminal, 33%
and 50% in TNBC), respectively. TTP was 7.1 months, DOR was 6.3 months, and median OS was not reached. Grade

3 adverse events (AEs) or serious AEs occurred in 5/22 patients (22.7%). Since the primary aim was not met, the
study was terminated at the first stage.
Conclusions: Although the primary aim was not reached, data on efficacy and safety of pembrolizumab plus carboplatin
in first-line visceral disease BRCA-related luminal mBC were provided and they need to be further investigated.
Key words: BRCA1, BRCA2, pembrolizumab, carboplatin, phase II

INTRODUCTION of the two alleles, and in cancer cells the wild-type allele is
Despite efforts in targeted therapeutic approaches over the almost invariably lost, leading to a defect in homologous
past 20 years, metastatic breast cancer (mBC) remains a recombination DNA repair in the cancer. Platinum chemo-
lethal disease with a median overall survival (OS) of 39 therapy generates interstrand cross-links that can only be
months and the need for further development in person- adequately repaired by homologous recombination DNA
alized medicine.1 About 3%-6% of all BCs may present ge- repair, and consequently BRCA1-deficient and BRCA2-defi-
netic alterations in BRCA1/2 with higher prevalence for cient cells are highly sensitive to platinum chemotherapy
triple-negative BC (TNBC) (15.4%), with respect to hor- both in vitro and in vivo. Many studies showed an impres-
mone receptors-positive (HRþ)/HER2-negative BC (5.2%).2 sive overall response rate (ORR) to platinum-derived drugs
Both BRCA1 and BRCA2 are required for DNA double- in mBC with germline BRCA (gBRCA) mutations ranging
strand break repair by homologous recombination.3,4 between 54.5% and 68%.5,6 Another important character-
Inherited mutations in BRCA1 and BRCA2 inactivate one istic of BRCA-mutated BC is represented by an elevated rate
of programmed death-ligand 1 (PD-L1), as compensatory
up-regulation of the mechanism of inhibiting T-cell activa-
*Correspondence to: Dr Laura Cortesi, Genetic Oncology Unit, Department of
Oncology and Haematology University Hospital Modena Via del Pozzo 71,
tion at tumor sites, related to the genomic instability, and
41124, Modena, Italy. Tel: þ390594224334; Fax: þ390594224152 subsequent tumor surface neoantigen expression that leads
E-mail: hbc@unimore.it (L. Cortesi). to an increase in tumor-infiltrating immune cells. The PD-L1
2059-7029/© 2023 The Author(s). Published by Elsevier Ltd on behalf of expression in BRCA1 and BRCA2 mutations differs, being
European Society for Medical Oncology. This is an open access article under the higher in BRCA1 than in BRCA2 BC.7,8 Pembrolizumab was
CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

Volume 8 - Issue 2 - 2023 https://doi.org/10.1016/j.esmoop.2023.101207 1

ESMO Open L. Cortesi et al.

administered in combination with poly (ADP-ribose) poly- (chemotherapy, hormonal therapy, immunotherapy, and
merase (PARP) inhibitors (PARPi) in BRCA-mutated patients radiotherapy, with the exception for symptomatic solitary
affected by metastatic TNBC (mTNBC).9 The TOPACIO trial lesion or brain). Live vaccines were prohibited within 30 days
provided, in gBRCA-mutated TNBC, an ORR and disease before the first dose of trial treatment and while participating
control rate (DCR) of 47% and 80%, respectively. in the trial, whereas administration of killed vaccines and RNA
Based on the improvement of progression-free survival vaccines was allowed. Finally, systemic glucocorticoids for
(PFS) derived from the platinum-derived drugs and also modulating symptoms from an event of clinical interest of
based on a median PFS of 8.3 months provided by pem- suspected immunologic etiology could be permitted.
brolizumab and niraparib in BRCA1/2-mutated patients, we Toxicity management and dose reduction followed a
designed the PEMBRACA phase II study with carboplatin summary of product characteristic recommendations and
plus pembrolizumab in metastatic HER2-BC with gBRCA1/2 local standard practice. Clinical and laboratory examinations
mutation. The primary aim was to evaluate the ORR; were carried out every 3 weeks after treatment initiation.
secondary objectives were to assess the time to The safety was assessed and graded by the National Cancer
progression (TTP), duration of response (DOR), DCR, and Institute-Common Terminology Criteria for Adverse Events
OS. The exploratory analyses regarded the CD8/tumor- version 5 (NCI-CTCAE v.5.0) every 3 weeks from treatment
infiltrated lymphocytes (TILs) and PD-L1 evaluation in the initiation until the end of treatment. Assessment of
metastatic biopsy and/or on primary surgical specimen, and response to treatment was based on investigator-reported
they will not be reported in this paper. measurements on target and non-target lesions and car-
ried out according to RECIST v1.111 and irRECIST v.1.1 (as
PATIENTS AND METHODS exploratory analysis in which changes in two dimensions are
used for target lesions)12 with computed tomography (CT)
Patients and study design scans or magnetic resonance imaging repeated every 9
PEMBRACA was a single-arm, open-label, multicenter, weeks along the combination therapy and every 12 weeks
phase II study built with a two-stage Simon’s design along the maintenance treatment.
(NCT03732391).10 The CONSORT diagram is represented in
Supplementary Figure S1, available at https://doi.org/10. Germline BRCA1/2 mutations, tumor immunostaining
1016/j.esmoop.2023.101207. All patients provided written The BRCA1/2 germline analysis was carried out at the
informed consent. Eligible patients were
18 years (males different centers that provided us the report. The immu-
and females) with HER2-mBC gBRCA1/2-mutated, previ- nostaining was carried out on primary breast cancer tissue,
ously treated with no more than one line of chemotherapy and when available on metastatic biopsy at the local labo-
for advanced BC. They needed to have received anthracy- ratory too. Immunohistochemical assessment was carried
cline and/or taxane in a (neo)adjuvant or metastatic setting; out to evaluate: HR measured by estrogen receptors (ER)
carboplatin could be offered in the (neo)adjuvant setting and progesterone receptors (PgR), HER2, and Ki67 expres-
without progression under treatment. In case of luminal BC, sion. HRþ were considered with ER and/or PgR
10%.
a first-line hormonal treatment for mBC can be adminis-
tered before. Patients had a BRCA1/2 deleterious germline
Statistical analysis
mutation (C4-C5 by ENIGMA classification) or with unknown
significance (C3 classification). Main exclusion criteria This phase II study was planned according to the two-step Si-
included benign gBRCA1/2 variants (C1-C2 classification), mon’s design: assuming as minimal interesting activity an ORR
having received more than one line of chemotherapy for of 70%, 12 objective responses among the first 20 enrolled
mBC, prior therapy with an anti-programmed cell death patients were necessary for the first step; to verify the hy-
protein 1 (PD-1), anti-PD-L1 or anti-PD-L2 agent, contrain- pothesis of ORR 70%, another 33 patients had to be enrolled in
dication to immunotherapy, <14 days from radiotherapy, the second step, with 34 objective responses among the total
chemotherapy or target small-molecule therapy, symp- 53 patients enrolled being necessary. The null hypothesis (p0)
tomatic or progressive brain metastases and/or carcino- has been set equal to 0.55 with a type I (alfa) error of 0.10
matous meningitis, previous history of pneumonitis/ (10%) and a type II (beta) error of 0.20 (power ¼ 80%). In case
interstitial lung disease that required steroids, hematopoi- the combination was unsatisfactory, the adopted design
etic function, or organ impairment. The local ethics com- allowed to stop the study at stage one with a probability of
mittee approved the study. 59%. By reaching an ORR of 70% at the end of the trial, the
probability error to declare the combination inefficacious was
19.8%, and the probability to stop the study at stage one was
Procedure and assessments 1.13%. We choose an alfa error of 0.10 rather than 0.05, since
Patients were treated with carboplatin intravenous (i.v.) at the specific population is very rare (BRCA mutation carriers).
area under the curve 6 every 3 weeks for six courses in Furthermore, the rate of non-responder patients could be
combination with 200 mg pembrolizumab i.v. every 3 weeks reduced by the increased number of CT scan due to the
until disease progression, unacceptable toxicity, or patient pseudoprogression phenomenon typical of immunotherapy
refusal. Subjects were prohibited from other concomitant treatments. We considered that the study by Isakoff,5 carried
therapy while starting and during the study treatment, out with platinum-derived drugs in first- or second-line

2 https://doi.org/10.1016/j.esmoop.2023.101207 Volume 8 - Issue 2 - 2023

L. Cortesi et al. ESMO Open

BRCA-related mTNBC, provided an ORR equal to 54.6% not enter into the study due to screening failure (Table 1).
whereas the KEYNOTE-012 study reported an ORR of 18.5% in Among the 22 eligible patients for the study, the male/fe-
patients with mTNBC treated only by pembrolizumab.13 Our male ratio was 1/21; median age, 50 years (35-69 years);
study hypothesis was generated to provide an ORR equal to Eastern Cooperative Oncology Group performance status 0,
70% as the sum of benefits from these two drugs. 22/22; BRCA1/2, 5/17; HRþ/HR-16/6; metastatic sites: liver
The ORR, TTP, DOR, and DCR were based on the in- 12 patients (29.2%), lung 7 patients (17.1%), lymph nodes 7
vestigator’s evaluation using the per-protocol population, by patients, (17.1%); local recurrence 1 patient (2.4%), bone 10
which patients were excluded if they did not receive at least patients (24.4%), others 5 patients (12.2%). Metastatic site
three courses of treatment (for reasons other than progressive was single in 9 patients (41%) and multiple in 13 patients
disease).The TTP was defined for each patient as the time from (59%). Single metastatic sites were: bone (two patients),
the first cycle until objective tumor progression (TTP does not lymph nodes (two patients), lung (two patients) and liver
include deaths) and was considered positive with at least (three patients). Lines of chemotherapies for mBC: 7 pa-
5 months. The other secondary objectives were the DOR, tients (32%) received a second-line treatment after taxane
measured from the first ORR to the date of progression, and it (3 patients) or anthracycline (2 patients) or capecitabine (2
was considered positive with 5 or more months, the DCR, patients), whereas 15 received experimental therapy as first
considered as the percentage of patients with ORR and stable line (68%). Totally, six patients received a previous hormonal
disease (SD) with an expected positivity rate of
80%, and OS, treatment for mBC, four with and two without cyclin-
defined as the time from the first cycle of treatment until death dependent kinase 4/6 (CDK4/6) inhibitors (27%).
from any cause (expected as at least 15 months). OS was
calculated by KaplaneMeier plots and summarized by median Activity and efficacy
and confidence intervals.
The preliminary analysis of efficacy was conducted among
The safety of the combination was evaluated as the
the first 21 patients (one patient was excluded for G3
secondary endpoint too, based on the intent-to treat pop-
hepatotoxicity after one cycle of treatment). The ORR was
ulation according to the toxicity grade reported throughout
43% (nine objective responses): eight partial responses
the whole treatment period. The toxicity descriptions and
(89%) and one complete response (11%). The study was
grading scales for adverse event (AE) reporting were using
terminated in the first stage according to the statistical plan.
the revised CTCAE v. 5.0).14
The DCR was equal to 76%: seven SD; five patients pro-
gressed (24%) (Figure 1A). The ORR in luminal patients
RESULTS (76%) was equal to 47% (7/15) with a DCR equal to 87%
(13/15), whereas in case of TNBC (24%) the ORR and DCR
Patient demographics were 33% and 50%, respectively (Figure 1B). The ORR and
From December 2019 to May 2022, 23 consecutive, unse- DCR in first-line therapy for mBC were 50% (7/14) and 100%
lected patients were screened for the study. One patient did (14/14), respectively, whereas in the second line the ORR

Table 1. Characteristics of patients at baseline

Patients Age Sex ECOG PS BRCA1/2 Phenotype Metastatic site Previous therapies for mBC
1 45 F 0 1 TNBC Lymph nodes None
2 50 F 0 2 Luminal Liver, bone HT þ CDK4/6i
3 54 F 0 2 Luminal Lung Taxane
4 47 F 0 2 Luminal Liver None
5 50 F 0 2 Luminal Lung None
6 57 F 0 2 Luminal Liver, lung, bone, lymph nodes None
7 46 F 0 2 Luminal Liver, bone HT þ CDK4/6i
8 63 F 0 2 Luminal Lung, liver, lymph nodes, bone Taxane
9 46 F 0 2 Luminal Lung, liver, pericardial effusion, bone None
10 40 F 0 1 TNBC Breast, lymph nodes None
11 53 F 0 2 Luminal Liver, pleural effusion, bone None
12 59 F 0 2 Luminal Lymph nodes None
13 51 F 0 2 Luminal Bone Capecitabine
14 60 F 0 2 Luminal Liver, bone Anthracycline
15 69 F 0 2 TNBC Pleura, peritoneum Capecitabine
16 46 F 0 2 Luminal Liver, bone HT þ CDK4/6i
17 35 F 0 1 TNBC Liver Anthracycline
18 50 F 0 1 Luminal Liver HT þ CDK4/6i
19 69 M 0 2 Luminal Bone HT
20 44 F 0 2 Luminal Liver, lung HT
21 29 F 0 1 TNBC Lymph nodes, lung, pleura Taxane
22 47 F 0 2 TNBC Lymph nodes None
23 (s.f.) 34 F 0 2 Luminal Bone HT
CDK4/6i, cyclin-dependent kinase4/6 inhibitors; ECOG PS, Eastern Cooperative Oncology Group Performance Scale; F, female; HT, hormonal therapy; M, male; mBC, metastatic
breast cancer; s.f., screening failure; TNBC, triple-negative breast cancer.

Volume 8 - Issue 2 - 2023 https://doi.org/10.1016/j.esmoop.2023.101207 3

ESMO Open L. Cortesi et al.

A 80% B 80%
60% 60%
40% 40%

% Change from baseline

% Change from baseline

20% 20%
0% 0%
-20% -20%
-40% -40%
-60% -60%
-80% -80%
-100% -100%
Patients Patients

Figure 1. Overall response rate and disease control rate in all patients and according to phenotype. Waterfall plot of maximum changes in tumor size (diameter)
from baseline in all individual patients (A) and according to tumor phenotype (B) during the treatment. Pink bars represent TNBC, violet bars represent luminal BC. The
dash line represents -30% of change from baseline.

and DCR were equal to 29% (2/7) (Figure 2A). Finally, the duration of 6-10 months; and 3 additional patients had a
ORR and DCR in visceral disease were equal to 43% (6/14) response duration of <6 months. Two of seven patients
and 79% (11/14), respectively, whereas in the non-visceral with SD continued without progression for >6 months
metastasis the ORR and DCR were 43% (3/7) and 72% (5/ (Figure 3B). The median duration of treatment with carbo-
7), respectively (Figure 2B). platin and pembrolizumab was 4 months, whereas the
After a median follow-up of 19 months, two patients are median maintenance therapy with pembrolizumab alone
still on treatment. The median TTP was 7.1 months (1.7- was 3.1 months. The median OS was not reached: 5 patients
18.9 months): 19 events occurred and one patient (6%) was died (23%) and 17 patients (77%) are still alive; 45% of
progression-free for >12 months (Figure 3A). The DOR was patients (10 patients) were alive >18 months.
measured on 16 patients, excluding 5 patients who pro-
gressed at the first evaluation. The median DOR was equal Safety
to 6.3 months (1.2-17 months). Of the 10 responders, 1 All 22 patients were assessable for safety. The overall inci-
patient had a response duration longer than 1 year; 6 pa- dence of AEs of any grade was 100% (22/22 patients), and
tients (two with ongoing treatment) had a response the incidence of grade
3 AEs or serious AEs (SAEs) was

A 80% B 80%
60% 60%
40% 40%
% Change from baseline%

% Change from baseline

20% 20%
0% 0%
-20% -20%
-40% -40%
-60% -60%
-80% -80%
-100% -100%
Patients Patients

Figure 2. Overall response rate and disease control rate according to the line of treatment and to the metastatic site. Waterfall plot of maximum changes in tumor
size (diameter) from baseline according to the line of treatment (A) and to the metastatic sites (B). The blue bars represent the first line of treatment for the
experimental therapy, the green bars correspond to second line of treatment after taxane, the pink bars after anthracyclines, and the violet bars after capecitabine.
The yellow bars represent visceral disease and the gray bars, the non-visceral metastasis.

4 https://doi.org/10.1016/j.esmoop.2023.101207 Volume 8 - Issue 2 - 2023

L. Cortesi et al. ESMO Open

A 18.9 B 17
11.4 9.7
11.2 8.9
10.3 8.6

Duration of response
9.9 7.7
Time to progression

9.8 7.4
8.2 6.8 ONGOING
7.5 6.7
7.3 ONGOING
7.1 5.9 X
5.8 5.4
5.1 5.1 X
4.7 4.7 X
3.2 4.2
2.5 Median TTP 7.1 months 3.2 Median DOR 6.3 months
2.3 3.1
2.2 1.2
2.1
1.7
0 5 10 15 20
0 5 10 15 20
Months Months

Figure 3. Time to progression and duration of response. The TTP is shown on the left bar graph (A). The median TTP was 7.1 months. The duration of response is
shown on the right bar graph (B). The green bar represents the CR, the blue bars represent the PR, and the orange bars, the SD. The pink crosses represent death. Two
patients, one with CR and one with PR, are still on treatment after 6.8 and 6.7 months, respectively. The median DOR was 6.3 months.
CR, complete response; DOR, duration of response; PR, partial response; SD, stable disease; TTP, time to progression.

22.7% (5/22 patients). The most common toxicities were DISCUSSION

the increase in alanine transaminase (ALT) and/or aspartate Chemo-immunotherapy has been shown to significantly
transaminase (AST) (27.3%), followed by nausea (22.7%), prolong survival in first-line metastatic or locally advanced
fatigue (13.6%), and vomiting (13.6%) (Table 2). One patient TNBC.15-17 In the IMpassion-130 trial, adding atezolizumab
with grade 3 vomiting along chemotherapy required dose to nab-paclitaxel in the first line significantly improved OS of
reduction. One patient with grade 3 hepatotoxicity inter- mTNBC patients with PD-L1
1, compared with chemo-
rupted treatment after the first cycle of treatment, whereas therapy alone (median OS 25.4 months versus 17.9
one patient showed a grade 3 autoimmune nephritis SAE months).15,16 In the KEYNOTE-355 study, adding pem-
after three courses of chemo-immunotherapy, leading to a brolizumab to nab-paclitaxel, paclitaxel, or gemcitabinee
permanent discontinuation. Finally, one patient showed a carboplatin in advanced TNBC patients significantly
grade 2 interstitial lung disease along pembrolizumab increased OS over chemotherapy alone in patients with or
maintenance, treated by prednisone for 1 month at a de- without PD-L1 overexpression (median OS 23 months
escalating dosage and resumed (still ongoing). versus 16 months),17,18 regardless of BRCA1/2 mutation
status.8 Data on platinum-derived drugs in BRCA-related
mTNBC have shown an impressive ORR, also compared to
Table 2. Summary of AEs of pembrolizumab also in association with taxane.5,6 These results suggest a synergistic action of
chemotherapy immunotherapy in association with chemotherapy in
Any grade N Grade ‡3 (%)
mTNBC, but no trial for BRCA1/2-mutated patients,
regardless of the phenotype, was designed.
N ¼ 22 (%)
To our knowledge, this is the first prospective multicenter
Any AE 22 (100) 5 (22.7)
phase II study evaluating efficacy and toxicity of pem-
AST/ALT increase 6 (27.3) 1 (4.5)a
Nausea 5 (22.7) 0 brolizumab in combination with carboplatin, in the first- and
Fatigue 3 (13.6) 0 second-line setting in metastatic BRCA patients, in both
Vomiting 3 (13.6) 1 (4.5)a TNBC and luminal. Our study was interrupted at the time of
Abdominal pain 2 (9.1) 0
Headache 2 (9.1) 0 the first-stage analysis because the primary endpoint was not
Systemic allergic reaction 1 (4.5) 0 achieved (43% versus expected 70%). The ORR of 70% may
Cutaneous rash 1 (4.5) 0 have been too ambitious. We hypothesized more consecu-
Arthralgia 1 (4.5) 0
Autoimmune nephritis 1 (4.5) 1 (4.5) (SAE)b
tively enrolled TNBC, but surprisingly they were very few. In
Diarrhea 1 (4.5) 1 (4.5) our series, 76% of patients were luminal, more than the
Interstitial lung disease 1 (4.5) 1 (4.5)c prevalence usually reported in the literature (50% in the
AE, adverse event; AST, aspartate transaminase; ALT, alanine transaminase; SAE, OlympiAD and 56% in the EMBRACA trials, respectively).19,20
serious adverse event.
a
AE leading to dose reduction. However, in the luminal cancers the ORR was greater than
b
AE leading to treatment interruption. that in the TNBC subgroup (47% versus 33%). In the
c
AE leading to treatment delay.
KEYNOTE-028 study, single-agent pembrolizumab exhibited

Volume 8 - Issue 2 - 2023 https://doi.org/10.1016/j.esmoop.2023.101207 5

ESMO Open L. Cortesi et al.

modest activity (ORR ¼ 12%) in a subset of patients with PD- who are still under combination treatment, the DOR of >6
L1-positive, HRþ, HER2 mBC.21 Data in favor of better ORR months is slightly inferior to that in the TOPACIO trial (71%
in luminal metastatic BRCA-related tumors were found in the versus 90%, respectively). Our results are probably limited
MEDIOLA phase I/II trial, evaluating olaparib þ durvalumab by the low number of cases in study. Finally, although the
in patients with gBRCA1/2 mBC. The ORR in HRþ was equal OS data were not mature at the time of this analysis, 50% of
to 69% (all partial responses) compared to 59% (nine partial alive patients overcame 18 months, confirming the ex-
and one complete responses) in TNBC.7 Furthermore, the pected hypothesis of 15 months.
DCR in our HRþ patients was 87%, which was higher than Although 59% of patients had two or more metastatic
that in TNBC patients (50%), as in the case of the MEDIOLA sites at diagnosis and, most importantly, 29% of patients
data (92% versus71%). Data provided by our study and had liver metastases that are associated with a poor prog-
MEDIOLA trial seem support the hypothesis that gBRCAm nosis,26 the best ORR and DCR were obtained in visceral
could drive the response to pembrolizumab in HRþ BC, disease rather than in non-visceral. Our results are likely
combined with chemotherapy or PARP inhibitors (PARPi). opposite to those obtained with pembrolizumab plus
Studies on PD-L1 expression and TILs count are ongoing and chemotherapy in non-small-cell lung cancer, where liver
could provide more information on immunoresponse in metastases are associated with shorter OS and PFS
gBRCA-HRþ tumors. compared to other metastatic sites.27,28 Once again, gBRCA
In our study, an improvement in ORR (50%) and DCR mutation could guide unexpected responses with respect to
(100%) was observed when the combination was offered in BRCAwt conditions, providing a better prognosis for this
the first compared to the second line of treatment for mBC. poor setting of patients.
As recently reported by the JBCRG 22 TR study, patients Adverse effects were modest, with grade 3 and 4 events
with BRCA mutations who received neoadjuvant treatment reported in 22.7% of patients, mostly attributable to
with eribulin plus capecitabine lost TILs compared to those chemotherapy than to the immunotherapy. We mainly
treated with eribulin plus carboplatin.22 Based on this observed hepatotoxicity (27.3%) that usually is rare for the
evaluation, capecitabine could have reduced the subse- pembrolizumab-alone treatment (1%-2%), but it can in-
quent immunocompetence to pembrolizumab in both pa- crease until 15% in case of combination regimens.29 How-
tients who progressed along treatment with chemo- ever, only one grade 3 hepatotoxicity led to permanent
immunotherapy combination. On the other side, patients treatment discontinuation, whereas in five other 1/2
with BRCA mutations seem to be more sensitive to drugs grades, the delay week and AST/ALT monitoring, as Care
affecting the homologous recombination repair system, like Step Pathways suggest,30 were enough for recovering
carboplatin or PARPi, when offered in the first line, as therapy. The only grade 3 vomiting was due to carboplatin
shown in the OlympiAD trial, where olaparib provided a therapy, recovered with dose reduction at 75%. The grade 2
significant improvement in OS.23 interstitial pneumonitis accounted for 1.5% of all patients
Our study firstly evaluates the maintenance therapy after treated with pembrolizumab according to literature data
a chemotherapy induction with immunotherapy, in the and required administration of corticosteroid for 1 month to
mBC BRCA-related tumors. In the same population, the restart immunotherapy without consequences. The SAE due
BROCADE-3 study with veliparib and carboplatinum/taxane to nephritis was related to pembrolizumab treatment, as
and maintenance with PARPi had shown a significantly already reported in literature data,31 causing permanent
better PFS for the maintenance arm compared to chemo- discontinuation.
therapy alone (19.3 months versus 13.5 months).24 Also, in In summary, this trial, although did not meet the primary
the neoadjuvant setting for TNBC patients, the combination endpoint, provides data on some efficacy of immuno-
of chemotherapy and immunotherapy followed by mainte- therapy plus carboplatin in BRCA-mutated HRþ mBC. The
nance with pembrolizumab provided a statistically signifi- benefit appears particularly evident in first-line chemo-
cant benefit in event-free survival compared to therapy for mBC and in case of visceral disease. The limit of
chemotherapy alone (84.5% versus 76.8%), particularly in this study was represented by the low number of enrolled
patients who did not obtain the pathological complete patients and by the lack of comparison with PARPi.
response to the preoperative treatment.25 The results of
these trials force us to consider whether a strategy of
platinum-based induction combination chemotherapy fol- ACKNOWLEDGEMENTS
lowed by maintenance treatment might ultimately lead to We would like to thank and express our gratitude to Tamara
superior outcomes for this group of patients, regardless of Sassi for her editorial assistance and Associazione Angela
HR expression. Serra per la Ricerca sul Cancro for supporting wigs to the
Comparing our study to the TOPACIO trial,9 where tu- patients.
moral BRCA-mutated mTNBC patients were treated with
pembrolizumab plus niraparib, the TTP (equivalent to PFS,
since no death for other causes were registered), was FUNDING
similar (TTP 7.1 months versus PFS 8.3 months). After This work was supported by MSD Italy S.r.L (grant number
exclusion of seven patients with SD and two other patients MISP#53981).

6 https://doi.org/10.1016/j.esmoop.2023.101207 Volume 8 - Issue 2 - 2023

L. Cortesi et al. ESMO Open

DISCLOSURE 13. Nanda R, Chow LQ, Dees EC, et al. Pembrolizumab in patients with
advanced triple-negative breast cancer: phase Ib KEYNOTE-012 study.
LC: Honoraria: Astra Zeneca, Pfizer, Advisory Board: Pfizer, J Clin Onco. 2016;34(21):2460-2467.
Novartis, MSD, Gilead; AT: Advisory Board: Pfizer, Eli-Lilly; 14. National Cancer Institute-Evaluation Cancer. Therapy Program 2021.
Novartis; MSD, AstraZeneca, Travel Grant: Gilead; UDG: Available at https://ctep.cancer.gov/protocoldevelopment/electronic_
Advisory Board: MSD, Bristol Myers Squibb, Janssen, applications/ctc.htm.
15. Schmid P, Adams S, Rugo HS, et al. Atezolizumab and nab-paclitaxel in
Astellas, Sanofi, Bayer, Pfizer, Ipsen, Novartis, and Pharma- advanced triple-negative breast cancer. N Engl J Med. 2018;379(22):
mar. Institutional research grants: AstraZeneca, Sanofi, and 2108-2121.
Roche; VG: Advisory Board: Eli-Lilly, Novartis, Roche, MSD, 16. Emens LA, Adams S, Barrios CH, et al. First-line atezolizumab plus nab-
Gilead, Eisai; Speaker’s Bureau: Eli-Lilly, Novartis, GSK, paclitaxel for unresectable, locally advanced, or metastatic triple-
Amgen; AM: Honoraria: Eli-Lilly, Pfizer, Macrogenics, Sea- negative breast cancer: IMpassion130 final overall survival analysis.
Ann Oncol. 2021;32(8):983-993.
gen, and Daiichi Sankyo; institutional research funding: 17. Cortes J, Rugo HS, Cescon DW, et al. Pembrolizumab plus chemo-
Roche, Eli-Lilly; AZ: Honoraria: Novartis, AstraZeneca, Lilly, therapy in advanced triple-negative breast cancer. N Engl J Med.
Pfizer, Daiichi Sankyo, MSD, Roche, Seagen, Exact Sciences, 2022;387(3):217-226.
Gilead, Istituto Gentili. All other authors have declared no 18. Cortes J, Cescon DW, Rugo HS, et al. Pembrolizumab plus chemo-
conflicts of interest. therapy versus placebo plus chemotherapy for previously untreated
locally recurrent inoperable or metastatic triple-negative breast cancer
(KEYNOTE-355): a randomised, placebo-controlled, double-blind, phase
ROLE OF THE FUNDING SOURCE 3 clinical trial. Lancet. 2020;396(10265):1817-1828.
19. Robson M, Im SA, Senkus E, et al. Olaparib for metastatic breast cancer
The sponsor provided financial support for study conduc-
in patients with a germline BRCA mutation. N Engl J Med. 2017;377(6):
tion and reviewed the final report of study findings and 523-533.
conclusions. 20. Litton JK, Rugo HS, Ettl J, et al. Talazoparib in patients with advanced
breast cancer and a germline BRCA mutation. N Engl J Med.
2018;379(8):753-763.
REFERENCES 21. Rugo HS, Delord JP, Im SA, et al. Safety and antitumor activity of
1. Grinda T, Antoine A, Jacot W, et al. Evolution of overall survival and receipt pembrolizumab in patients with estrogen receptor-positive/human
of new therapies by subtype among 20 446 metastatic breast cancer pa- epidermal growth factor receptor 2-negative advanced breast cancer.
tients in the 2008-2017 ESME cohort. ESMO Open. 2021;6(3):100114. Clin Cancer Res. 2018;24(12):2804-2811.
2. Armstrong N, Ryder S, Forbes C, Ross J, Quek RG. A systematic review 22. Ueno T, Kitano S, Masuda N, et al. Immune microenvironment, ho-
of the international prevalence of BRCA mutation in breast cancer. Clin mologous recombination deficiency, and therapeutic response to
Epidemiol. 2019;11:543-561. neoadjuvant chemotherapy in triple-negative breast cancer: Japan
3. Cortesi L, Rugo HS, Jackisch C. An overview of PARP inhibitors for the Breast Cancer Research Group (JBCRG)22 TR. BMC Med. 2022;20(1):
treatment of breast cancer. Target Oncol. 2021;16(3):255-282. 136.
4. Cortesi L, Piombino C, Toss A. Germline mutations in other homolo- 23. Robson ME, Tung N, Conte P, et al. OlympiAD final overall survival and
gous recombination repair-related genes than BRCA1/2: predictive or tolerability results: olaparib versus chemotherapy treatment of
prognostic factors? J Pers Med. 2021;11(4):245. physician’s choice in patients with a germline BRCA mutation and
5. Isakoff SJ, Mayer EL, He L, et al. TBCRC009: A multicenter phase II HER2-negative metastatic breast cancer. Ann Oncol. 2019;30(4):558-
clinical trial of platinum monotherapy with biomarker assessment in 566.
metastatic triple-negative breast cancer. J Clin Oncol. 2015;33(17): 24. Diéras V, Han HS, Kaufman B, et al. Veliparib with carboplatin and
1902-1909. paclitaxel in BRCA-mutated advanced breast cancer (BROCADE3): a
6. Tutt A, Tovey H, Cheang MCU, et al. Carboplatin in BRCA1/2-mutated randomised, double-blind, placebo-controlled, phase 3 trial. Lancet
and triple-negative breast cancer BRCAness subgroups: the TNT Trial. Oncol. 2020;21(10):1269-1282.
Nat Med. 2018;24(5):628-637. 25. Schmid P, Cortes J, Dent R, et al. Event-free survival with pembrolizumab in
7. Domchek SM, Postel-Vinay S, Im SA, et al. Olaparib and durvalumab in early triple-negative breast cancer. N Engl J Med. 2022;386(6):556-567.
patients with germline BRCA-mutated metastatic breast cancer 26. Tian C, Liu S, Wang Y, Song X. Prognosis and genomic landscape of liver
(MEDIOLA): an open-label, multicentre, phase 1/2, basket study. Lan- metastasis in patients with breast cancer. Front Oncol. 2021;11:
cet Oncol. 2020;21(9):1155-1164. 588136.
8. Emens LA, Molinero L, Loi S, et al. Atezolizumab and nab-paclitaxel in 27. Takeyasu Y, Yoshida T, Shibaki R, et al. Differential efficacy of pem-
advanced triple-negative breast cancer: biomarker evaluation of the brolizumab according to metastatic sites in patients with PD-L1
IMpassion130 study. J Natl Cancer Inst. 2021;113(8):1005-1016. strongly positive (TPS
50%) NSCLC. Clin Lung Cancer. 2021;22(2):
9. Vinayak S, Tolaney SM, Schwartzberg L, et al. Open-label clinical trial of 127-133.e3.
niraparib combined with pembrolizumab for treatment of advanced or 28. Kitadai R, Okuma Y, Hakozaki T, Hosomi Y. The efficacy of immune
metastatic triple-negative breast cancer. JAMA Oncol. 2019;5(8):1132- checkpoint inhibitors in advanced non-small-cell lung cancer with liver
1140. metastases. J Cancer Res Clin Oncol. 2020;146(3):777-785.
10. Simon R. Optimal two-stage designs for phase II clinical trials. Control 29. Dadu R, Zobniw C, Diab A. Managing adverse events with immune
Clin Trials. 1989;10(1):1-10. checkpoint agents. Cancer J. 2016;22(2):121-129.
11. Eisenhauer EA, Therasse P, Bogaerts J, et al. New response evaluation 30. Fazer C. Checkpoint inhibitor immunotherapy for head and neck can-
criteria in solid tumours: revised RECIST guideline (version 1.1). Eur J cer: incorporating care step pathways for effective side-effect man-
Cancer. 2009;45:228-247. agement. J Adv Pract Oncol. 2019;10(suppl 1):37-46.
12. Nishino M, Gargano M, Suda M, Ramaiya NH, Hodi FS. Optimizing 31. Ribas A, Puzanov I, Dummer R, et al. Pembrolizumab versus
immune-related tumor response assessment: does reducing the investigator-choice chemotherapy for ipilimumab-refractory melanoma
number of lesions impact response assessment in melanoma patients (KEYNOTE-002): a randomised, controlled, phase 2 trial. Lancet Oncol.
treated with ipilimumab? J Immunother Cancer. 2014;2:17. 2015;16(8):908-918.

Volume 8 - Issue 2 - 2023 https://doi.org/10.1016/j.esmoop.2023.101207 7