Ribociclib plus letrozole versus chemotherapy for postmenopausal women with hormone receptor-positive, HER2-negative, luminal B breast cancer (CORALLEEN):an open-label, multicentre, randomised, phase 2 trial
Aleix Prat, Cristina Saura, Tomás Pascual, Cristina Hernando, Montserrat Muñoz, Laia Paré, Blanca González Farré, Pedro L Fernández,
Patricia Galván, Núria Chic, Xavier González Farré, Mafalda Oliveira, Miguel Gil-Gil, Miriam Arumi, Neus Ferrer, Alvaro Montaño, Yann Izarzugaza, Antonio Llombart-Cussac, Raquel Bratos, Santiago González Santiago, Eduardo Martínez, Sergio Hoyos, Beatriz Rojas, Juan Antonio Virizuela, Vanesa Ortega, Rafael López, Pamela Céliz, Eva Ciruelos, Patricia Villagrasa, Joaquín Gavilá
Summary
Background In hormone receptor-positive, HER2-negative early stage breast cancer, cyclin-dependent kinases 4 and 6 (CDK4/6) inhibition in combination with endocrine therapy could represent an alternative to multiagent chemotherapy. We aimed to evaluate the biological and clinical activity of neoadjuvant ribociclib plus letrozole in the luminal B subtype of early stage breast cancer.
Methods CORALLEEN is a parallel-arm, multicentre, randomised, open-label, phase 2 trial completed across 21 hospitals in Spain. We recruited postmenopausal women (≥18 years) with stage I–IIIA hormone receptor-positive, Eastern Cooperative Oncology Group Performance Status 0–1, HER2-negative breast cancer and luminal B by PAM50 with histologically confirmed, operable primary tumour size of at least 2 cm in diameter as measured by MRI. Patients were randomly assigned (1:1) using a web-based system and permuted blocks of 25 to receive either six 28-days cycles of ribociclib (oral 600 mg once daily for 3 weeks on, 1 week off) plus daily letrozole (oral 2·5 mg/day) or four cycles of doxorubicin (intravenous 60 mg/m²) and cyclophosphamide (intravenous 600 mg/m²) every 21 days followed by weekly paclitaxel (intravenous 80 mg/m²) for 12 weeks. The total duration of the neoadjuvant therapy was 24 weeks. Randomisation was stratified by tumour size and nodal involvement. Samples were prospectively collected at baseline (day 0), day 15, and surgery. The primary endpoint was to evaluate the proportion of patients with PAM50 low-risk-of- relapse (ROR) disease at surgery in the modified intention-to-treat population including all randomly assigned patients who received study drug and had a baseline and at least one post-baseline measurement of ROR score. The PAM50 ROR risk class integrated gene expression data, tumour size, and nodal status to define prognosis. This trial was registered at ClinicalTrials.gov, NCT03248427.
Findings Between July 27, 2017 to Dec 7, 2018, 106 patients were enrolled. At baseline, of the 106 patients,
92 (87%) patients had high ROR disease (44 [85%] of 52 in the ribociclib and letrozole group and 48 [89%] of 54 in the chemotherapy group) and 14 (13%) patients had intermediate-ROR disease (eight [15%] and six [11%]). Median follow- up was 200·0 days (IQR 191·2–206·0). At surgery, 23 (46·9%; 95% CI 32·5–61·7) of 49 patients in the ribociclib plus letrozole group and 24 (46·1%; 32·9–61·5) of 52 patients in the chemotherapy group were low-ROR. The most common grade 3–4 adverse events in the ribociclib plus letrozole group were neutropenia (22 [43%] of 51 patients) and elevated alanine aminotransferase concentrations (ten [20%]). The most common grade 3–4 adverse events in the chemotherapy group were neutropenia (31 [60%] of 52 patients) and febrile neutropenia (seven [13%]). No deaths were observed during the study in either group.
Interpretation Our results suggest that some patients with high-risk, early stage, hormone receptor-positive, HER2- negative breast cancer could achieve molecular downstaging of their disease with CDK4/6 inhibitor and endocrine therapy.
Funding Novartis, Nanostring, Breast Cancer Research Foundation-AACR Career Development Award.
Lancet Oncol 2019
Published Online December 11, 2019 https://doi.org/10.1016/ S1470-2045(19)30786-7
See Online/Comment https://doi.org/10.1016/ S1470-2045(19)30806-X
SOLTI Breast Cancer Research Group, Barcelona, Spain (Prof A Prat MD, C Saura MD, T Pascual MD, M Muñoz MD,
L Paré PhD, N Chic MD, X González Farré MD,
M Oliveira MD, P Céliz MD,
E Ciruelos MD, P Villagrasa PhD, J Gavilá MD); Department of Medical Oncology (Prof A Prat, T Pascual, M Muñoz) and Department of Pathology
(B González Farré MD), Hospital Clinic of Barcelona, Barcelona, Spain; Translational Genomics and Targeted Therapeutics in Solid Tumors, August Pi i Sunyer Biomedical Research Institute, Barcelona, Spain (Prof A Prat, T Pascual,
P Galván MSc, N Chic); Department of Medical Oncology, Vall d’Hebron University Hospital and Vall d’Hebron Institute of Oncology, Barcelona, Spain (C Saura, M Oliveira,
M Arumi MD); Department of Medical Oncology, Hospital Clínico Universitario of Valencia, Valencia, Spain
(C Hernando MD); Department of Pathology, Hospital Germans Trials i Pujol,
Badalona, Spain
(P L Fernández MD); Department of Medical Oncology, Hospital General de
Introduction
Hormone receptor-positive, HER2-negative breast cancer accounts for approximately 70% of all invasive breast carcinomas and is responsible for most breast cancer- related deaths worldwide.1 In early stages, adjuvant
chemotherapy has consistently shown significant and long-term clinical benefits in disease-free survival and overall survival.2 However, substantial heterogeneity exists in hormone receptor-positive, HER2-negative breast cancer regarding patient prognosis and treatment benefit.3
Catalunya, Barcelona, Spain
(X González Farré); Department of Medical Oncology, Institut Català d’Oncologia Hospitalet, Hospitalet de Llobregat, Spain (M Gil-Gil MD); Department of Medical Oncology, Hospital
Universitari Son Espases, Palma
de Mallorca, Spain (N Ferrer MD); Department of Medical Oncology, Hospital Universitario Virgen del Rocío, Sevilla, Spain (A Montaño MD); Department of Medical Oncology, Hospital Universitario Fundación Jimenez Díaz, Madrid, Spain
(Y Izarzugaza MD); Department of Medical Oncology, Hospital Arnau de Vilanova, Valencia, Spain (A Llombart-Cussac MD); Department of Medical Oncology, Centro Oncológico Internacional MD Anderson, Madrid, Spain (R Bratos MD); Department of Medical Oncology, Hospital San Pedro de Alcántara, Cáceres, Spain
(S González Santiago MD);
Department of Medical Oncology, Consorcio Hospitalario Provincial of Castellón, Castellón de la Plana, Spain (E Martínez MD); Department of Medical Oncology, Hospital Rey Juan Carlos, Madrid, Spain
(S Hoyos MD); Department of Medical Oncology, Centro Integral Oncológico Clara Campal, Madrid, Spain
(B Rojas MD); Department of Medical Oncology, Hospital Virgen de la Macarena, Sevilla, Spain (J A Virizuela MD); Department of Medical Oncology, Fundación Privada Asil de Granollers, Barcelona, Spain (V Ortega MD); Department of Medical Oncology, Complejo Universitario de Santiago de
Compostela, Spain
(R López MD); Department of Medical Oncology, Hospital 12 de Octubre, Madrid, Spain
(E Ciruelos); and Department of Medical Oncology, Instituto Valenciano de Oncología, Valencia, Spain (J Gavilá)
Correspondence to: Prof Aleix Prat, Hospital Clínic of Barcelona, Translational Genomics and Targeted Therapeutics in Solid Tumors, August Pi i Sunyer Biomedical
Research Institute, 08035, Barcelona, Spain [email protected]
Several prognostic tools allow for a better indivi- dualisation of systemic treatments in patients with early stage, hormone receptor-positive, HER2-negative breast cancer.4 For example, baseline gene expression-based assays such as OncotypeDX,5 EndoPredict,6 and PAM50 (now known as Prosigna)7 can help to identify patients at low risk of distant recurrence who do not need neo- adjuvant or adjuvant chemotherapy.8 Second generation genomic tests in early stage breast cancer, such as EndoPredict EPclin or PAM50 (Prosigna) risk of relapse (ROR), which include clinical parameters, such as tumour size and nodal status in the final risk assessment, might be more accurate in predicting long-term prognosis and, as a result, better discriminate patients who might not need chemotherapy from those who are likely to benefit.9
The neoadjuvant approach for early stage, hormone receptor-positive, HER2-negative breast cancer is an ideal setting to analyse both the clinical and biological heterogeneity that exists and better select patients in need of escalation or de-escalation of systemic therapy. Several on-treatment and post-treatment biological biomarkers, mostly tracking proliferation-related processes, have been shown to provide additional prognostic information beyond the classic endpoint of pathological response. Examples are Ki67 expression at day 15 after endocrine
therapy, and Ki67,10 oestrogen receptor levels, and PAM50 ROR after 4–6 months of neoadjuvant endocrine therapy.11–13 Similarly, adding data for Ki67 and oestrogen receptor concentrations to that for residual cancer burden following neoadjuvant chemotherapy improves the prediction of long-term outcomes.14,15
A new class of drugs, known as CDK4/6 inhibitors, has been introduced in 2015 as a treatment for advanced hormone receptor-positive, HER2-negative disease. CDK4/6 inhibitors in combination with endocrine therapy, provide clinically relevant results in terms of the proportion of patients with an objective response, progression-free survival, and overall survival compared with those for only endocrine therapy, and with good tolerability.16 Arguably, these agents delay the use of chemotherapy in the advanced stage setting. However, in the setting of early stage adjuvants, the strategy being tested in phase 3 studies is to combine CDK4/6 inhibitors with endocrine therapy in patients with high-risk disease who have completed locoregional therapy and neoadjuvant or adjuvant chemotherapy.17 Thus, the value of replacing neoadjuvant and adjuvant chemotherapy with CDK4/6 inhibitors is currently unknown.
Here, we aimed to test the hypothesis that the selective inhibition of CDK4 and CDK6 with ribociclib plus letrozole offers high biological and clinical activity in the
preoperative setting in patients with PAM50 luminal B early stage breast cancer.
Methods
Study design and participants
CORALLEEN is a parallel-arm, multicentre, randomised, open-label, phase 2 trial. 21 hospitals in Spain were used to recruit patients (appendix pp 1–2). Eligible patients were women (≥18 years) who were postmenopausal with histologically confirmed invasive hormone-sensitive breast cancer, defined locally as hormone receptor- positive, HER2-negative by American Society of Clinical Oncology and College of American Pathologists guide- lines, and centrally as luminal B by the standardised PAM50 (Prosigna) assay.7 Patients had operable stage I-IIIA breast cancer, with primary tumour size of at least
2 cm in diameter as measured by MRI. Other main inclusion criteria were baseline Eastern Cooperative Oncology Group (ECOG) performance status of 0–1, suitable haematological counts, and adequate hepatic and renal function. Patients with stage IV disease, or multifocal or bilateral breast cancer were ineligible. Patients were also excluded if they had other malignancies, inadequate bone marrow, impaired liver function, cardiac disease or history of cardiac dysfunction, including QT interval 450 ms or more, and uncontrolled hypertension.
The study was done in accordance with Good Clinical Practice guidelines and the World Medical Association Declaration of Helsinki. Patients provided written, informed consent. Approvals for the study protocol were obtained from an independent ethics committee (Vall d’Hebron University Hospital, Barcelona, Spain). The study protocol can be found in the appendix.
Randomisation and masking
Eligible patients were randomly assigned (1:1) using a secure web-based system to ribociclib plus letrozole or multiagent chemotherapy. Randomisation was stratified by tumour size (T1–T2 vs T3) and nodal involvement using permuted blocks of 25 and an internet-based tool used to generate a pseudorandom number. Investigators enrolled participants and used identifying information to register them in the interactive web-response system. Randomisation was concealed, but neither participants nor investigators were masked to group assignment. All outcomes assessors were masked to clinical data.
Procedures
Standardised PAM50 analysis was done at a central laboratory (Hospital Clinic of Barcelona, Barcelona, Spain). Patients with a subtype other than luminal B at screening were excluded from the trial. The ribociclib plus letrozole group consisted of six 28-day cycles of oral letrozole 2·5 mg once daily continuously plus oral ribociclib 600 mg once daily administrated on a
3 weeks on, 1 week off schedule. The chemotherapy
group consisted of four cycles of doxorubicin 60 mg/m² and cyclophosphamide 600 mg/m² administrated intravenously every 21 days followed by intravenous weekly paclitaxel 80 mg/m² for 12 weeks. The total duration of the neoadjuvant therapy was 24 weeks. Treatment was interrupted or delayed in case of adverse event occurrence and resumed if protocol-defined criteria for treatment resumption were met. Dose reductions were allowed for doxorubicin, cyclophos- phamide, paclitaxel, and ribociclib for toxicities pre- specified in the protocol under dose reduction criteria. Dose reductions were not permitted for letrozole because only the 2·5 mg pill is approved and therefore no other dose reductions were possible. In the presence of severe toxicity, the investigator decided whether to discontinue treatment and undertake immediate surgery. Patients had the right to withdraw from the study at any time for any reason. The investigator had the right to withdraw patients from the study in the event of intercurrent illness; adverse events; and treatment failure after a prescribed procedure, protocol violation, administrative reasons, or for other reasons.
Baseline MRI and axillary lymph node and breast ultrasound were done, and ultrasound-guided fine needle aspiration or core biopsy was required in case of suspicious axillary lymph nodes. After 12 weeks of treatment, an ultrasound was done to assure no progressive disease. If disease progression was suspected, an MRI was mandatory to confirm progression, and the patient could either proceed to surgery or receive alternative neoadjuvant treatment.
Laboratory tests, 12 lead electrocardiogram, ECOG performance status, and vital signs were assessed at baseline, at 15 days after treatment initiation, and at every cycle in each group. Adverse events were assessed by the Common Terminology Criteria for Adverse Events, version 4.03.
Surgery was done within 7 days after the last dose of ribociclib or 2 weeks after the last dose of chemotherapy. Endocrine therapy in the ribociclib plus letrozole group was continued until the day of surgery. Standard adjuvant chemotherapy was administered according to the physician’s discretion.
At baseline, day 15, and surgery, collection of tissue samples were mandatory to study gene expression changes. A section of the formalin-fixed, paraffin- embedded breast tissue was first examined with haematoxylin and eosin staining to confirm the presence of invasive tumour cells (≥10%) and to determine the minimum tumour surface area (10 mm²).
For RNA purification (High Pure formalin-fixed paraffin-embedded RNA isolation kit, Roche Diagnostics Limited, West Sussex, UK), one to five 10 μm formalin- fixed paraffin-embedded slides were used for each tumour specimen (at baseline, day 15, and surgical). A minimum of about 125 ng of total RNA was used to measure the expression of the 50 PAM50 genes. Data
were analysed using the Prosigna assay (NanoString Technologies, Seattle, USA).
ROR prognostic score was based on gene expression data and tumour size and had a range from 0 to 100 (a higher score indicates high risk of distant recurrence at 10 years in the absence of chemotherapy).7 Specifically, the ROR score was calculated using weighted coefficients to the four subtypes (ie, luminal A, luminal B, HER2- enriched, and basal-like), a proliferation score, and tumour size. These weighted coefficients were obtained from a cohort of patients with breast cancer who did not receive adjuvant systemic therapy.7 The proliferation score was calculated using the average mean expression of a subset of the 50 classifier genes that are associated with the cell cycle.7 Tumour size was calculated by assigning a binary classification value to primary tumours measuring 2·0 cm in the greatest dimension versus those measuring more than 2·0 cm. In CORALLEEN, tumour size was measured at baseline by MRI.
Low-ROR disease was defined according to the CE certification marked PAM50 (Prosigna) assay, was at least 40 points if node-negative and at least 15 points if
one to three positive nodes. These cutoff points were validated to identify a population of postmenopausal patients with a risk of less than 10% of developing distant metastasis at 10 years if treated with local therapy and 5 years of endocrine therapy and without chemotherapy. Intermediate-ROR disease was defined as 41–60 points if node-negative and 16–40 points if one to three positive nodes. High-ROR disease was defined as 61–100 points if node negative and 41–100 points if one to three positive nodes. All patients with four or more positives nodes were considered high-ROR disease regardless of ROR score. To determine the axilla status at baseline, ultrasound was done in all patients, and if no nodal disease was suspected, the patient was considered node negative. If one node was suspected to have disease, we did a fine needle aspiration cytology assay. If the fine needle aspiration cytology result was negative for malignant cells, the patient was considered node- negative. If the fine needle aspiration cytology result was positive for malignant cells, the patient was considered to have one to three positive nodes. In patients with cN2 disease and positive fine needle aspiration cytology, the axilla was considered to have four or more positive lymph nodes.
Outcomes
The primary endpoint, after a protocol amendment (protocol version 3.0; June 15, 2018), was to evaluate the proportion of patients with low-ROR disease after neoadjuvant treatment according to the standard PAM50.
The nine key secondary endpoints that we report in this Article are (1) to determine the proportion of patients who had an objective response (defined as the sum of partial responses and complete responses according to Modified Response Evaluation Criteria in Solid Tumors, version 1.1) by MRI and by physical examination; (2) to evaluate the proportion of patients with a pathological complete response in the breast (pCRB; ypT0/Tis ypNx) and pCR in breast and axillary lymph nodes (pCRBL; defined as ypT0/Tis ypN0) in both groups; (3) to evaluate the proportion of patients with a residual cancer burden18 score (0–1) by central assessment; (4) to evaluate the proportion of patients with preoperative endocrine prognostic index score of 0; (5) to determine the proportion of subtype switching from luminal B to luminal A and the decrease in Ki67 from baseline to day 15 and surgery;
(6) to evaluate the change of low-ROR disease from baseline to day 15; (7) to report the frequency of breast conserving surgery; and (8) to evaluate the correlation between ROR score (as a continuous variable) and KI67; and (9) to evaluate safety and tolerability of treatment in both groups. Other secondary endpoints (residual cancer burden 0–1 as per local assessment, quality of life, and other biomarkers of response) will be reported elsewhere.
Statistical analysis
Assuming that 20·0–25·0% of the patients in each group would reach low-ROR disease, the study would require a sample size of 47 patients per group for estimating the expected proportion with a 95% CI of plus or minus 11·5–12·4%. Assuming a proportion of patients of 10% dropped out, a total of 104 patients was planned to be recruited. Each group was designed to be analysed independently and the trial was not powered for formal comparison between groups.
The main study analysis has been done using a modified intention-to-treat population, defined as all randomly assigned patients who received study medi- cation, had a baseline ROR score measurement, and had at least one corresponding post-baseline ROR score measurement.
The main analysis was the estimation of the proportion of patients with low-ROR at surgery described by means proportions and 95% CI using the Clopper-Pearson method.19 Mixed linear effects models were applied to evaluate the changes of ROR score between two time periods, adjusting by tumour size, and nodal involve- ment. The absolute changes in ROR score in the two groups was a post-hoc exploratory analysis.
The number and percentage of patients who had one or more treatment-emergent adverse events and the number of adverse event episodes have been tabulated by treatment group, system organ class, preferred term (according to Medical Dictionary for Regulatory Activities), seriousness, intensity, and relationship to the study drug. The safety population was defined as all patients who were randomly assigned and had at least one dose of the study medication.
In a post-hoc exploratory analysis, the association between ROR score and Ki67 was evaluated using Pearson’s χ2 test and the expression of oestrogen receptor at surgery was evaluated using percentage of positive tumour cells and the Allred score.
All statistical analyses were done with R (version 3.5.1). This study is registered with ClinicalTrials.gov,
NCT03248427.
Role of the funding source
The study was conceived and designed by SOLTI investigators. Novartis funded the study and provided ribociclib for the study. NanoString provided CE marked Prosigna tests for the study. The funders had no role in study design, data collection, data analysis, data interpretation, or writing of the report. The corresponding author had full access to all of the data in the study and final responsibility for the decision to submit for publication.
Results
From July 27, 2017 to Dec 7, 2018, 198 patients were assessed for eligibility and 106 were finally enrolled (figure 1, table 1).
Ribociclib and letrozole group (n=52) Chemotherapy group (n=54)
Median age, years 63·0 (56·5–70·3) 64·0 (58·3–71·8)
Tumour size
The mean turnaround time of the PAM50 assay from the receipt of the tumour block at the central laboratory to verification of the results was 4·7 days (IQR 3·0–6·0). The distribution of intrinsic subtypes in the 188 screened patients with tissue samples for PAM50 was 131 (70%) luminal B, 48 (26%) luminal A, five (3%) basal-like, and four (2%) HER2-enriched. Only 106 of 188 were randomly assigned and all were subtype luminal B. Patients’ characteristics were well balanced between the two groups. Among the 106 patients were recruited; the baseline median age was 63·5 years (IQR 58·0–71·0), median tumour size by breast MRI was 31 mm (IQR 25–43), and most patients had T2 tumours and no
Primary endpoint
ROR groups
Ribociclib and letrozole group (n=49) Chemotherapy group (n=52)
n (%) 95% CI n (%) 95% CI
group. One sample in the chemotherapy group was excluded due to low quality. At surgery, 24 (46·1% [95% CI 32·9–61·5]) of 52 patients in the chemotherapy group and 23 (46·9% [95% CI 32·5–61·7]) of 49 patients in the ribociclib plus letrozole group were low-ROR (table 2;
Low 23 (46·9%) 32·5–61·7 24 (46·1%) 32·9–61·5
Intermediate 15 (30·6%) 18·2–45·4 16 (30·8%) 19·1–45·9 High 11 (22·5%) 11·8–36·7 11 (21·2%) 11·2–35·2
Missing 0 NA 1 (1·9%) NA
Secondary endpoints
ROR score (continuous variable)
Median (IQR) 18·0 (12·0–35·0) ·· 25·0 (12·0–45·0) ·· Mean 25·0 (19·3) ·· 29·4 (21·0) ··
Ki67 (central)
appendix p 3).
Consistent with the primary endpoint, intrinsic subtype conversion to luminal A occurred in 43 (82·7% [95% CI 69·7–91·8]) of 52 patients in the chemotherapy group and 43 (87·8% [75·3–95·4]) of 49 patients ribociclib plus letrozole group (appendix p 3). Ki67 at surgery was also lower at surgery than at baseline in both groups (table 2). The correlation coefficients between ROR score at surgery (as a continuous variable) and Ki67 immuno-
Median proportion of cells with Ki67 in a sample (IQR)
Mean proportion of cells with Ki67 in a sample (SD)
RCB (central)
histochemistry expression was 0·35 in the chemotherapy group and 0·71 in the ribociclib plus letrozole group. All residual tumours were oestrogen receptor-positive by central immunohistochemistry with oestrogen receptor concentrations between 50–100% and Allred scores of 5–8 (data not shown).
0–1 3 (6·1%) 1·3–16·8 6 (11·8%) 4·5–27·8
2–3 46 (93·9%) 83·2–98·7 45 (88·2%)s 76·2–95·5
We evaluated the individual absolute changes in ROR score between baseline and surgery in both groups
pCRBL
0 (0%) 0·0–7·25 3 (5·8%) 1·4–16·6
(figure 2). Scores at surgery were 29·4 (95% CI 23·5–35·3)
in the chemotherapy group and 25·0 (19·5–30·5) in the
pCRB 1 (2·0%) 0·05–10·8 3 (5·8%) 1·4–16·6
PEPI score
0 11 (22·4%) 11·7–36·6 9 (17·3%) 8·6–31·4
1–3 25 (51·0%) 36·3–65·6 24 (46·1%) 33·6–62·6
>4 13 (26·6%) 15·0–41·2 17 (32·7%) 21·2–48·7
Missing 0 ·· 2 (3·9%) ·· Overall response by MRI
Complete response 7 (14·3%) 6·6–30·1 10 (19·2%) 10·4–34·9 Partial response 21 (42·9%) 32·4–63·3 31 (59·6%) 49·5–77·9 Stable disease 16 (32·6%) 22·4–52·3 7 (13·5%) 6·1–27·8 Progression disease 0 NA 0 NA
NA 5 (10·2%) NA 3 (7·7%) NA
Overall response by physical examination
Complete response 16 (32·7%) 19·9–47·5 16 (30·8%) 18·7–45·1 Partial response 15 (30·6%) 18·2–45·4 12 (23·1%) 12·5–36·8 Stable disease 8 (16·3%) 7·3–29·7 5 (9·6%) 3·2–21·0 Progression disease 1 (2·0%) 0·1–10·8 3 (5·7%) 1·2–15·9 NA 9 (18·4%) 8·8–32·0 16 (30·8%) 18·7–45·1
Data are n (%); 95% CI; median (IQR); or mean (SD). NA=not applicable. pCRB=pathological complete response in the breast. pCRBL=pathological complete response in breast and axillary lymph nodes. PEPI=preoperative endocrine prognostic index. RCB=residual cancer burden. ROR=risk of relapse.
Table 2: Primary and secondary endpoints determined at surgery
axillary node involvement (table 1). Overall of the 106 patients, 92 (87%) patients had high-ROR disease and 14 (13%) patients had intermediate-ROR disease. Median follow-up was 200·0 days (IQR 191·2–206·0).
51 patients in the chemotherapy group and 49 patients in the ribociclib and letrozole group had samples at surgery for genomic analysis (figure 1). Tissue samples were not available for three patients in the ribociclib and letrozole group and two patients in the chemotherpay
ribociclib plus letrozole group. In the chemotherapy group, the mean ROR score at surgery in patients with low-ROR disease was 14·0 (95% CI 9·3–18·6), with intermediate-ROR disease was 35·0 (28·5–41·5), and with high-ROR disease was 54·9 (42·1–67·7). In the ribociclib plus letrozole group, the mean ROR score in patients with low-ROR was 14·0 (95% CI 10·3–17·6), with intermediate-ROR disease was 28·9 (21·4–36·3), and with high-ROR disease was 42·8 (24·7–60·9). Two patients did not show a decrease in ROR score following ribociclib plus letrozole (figure 2B) and their tumours at surgery were luminal B and HER2 enriched. In addition, two patients with a decrease in ROR score of more than
60 points in the ribociclib plus letrozole group were identified as high-ROR at surgery due to four or more positive residual nodes.
In the chemotherapy group, the proportion of patients at surgery with a pCRBL was 5·8% (95% CI 1·4–16·6), with residual cancer burden of 0–1 was 11·8% (4·5–27·8), and with preoperative endocrine prognostic index of 0 was 17·3% (8·6–31·4; table 2). In the ribociclib plus letrozole group, the proportion of patients at surgery with pCRBL was 0%, with residual cancer burden of 0–1 was 6·1% (95% CI 1·3–16·8), and preoperative endocrine prognostic index 0 was 22·4% (11·7–36·6; table 2). The proportion of patients achieving an objective response (as assessed by breast MRI) was 41 (78·8%; 95% CI 65·3–88·9) of 52 patients treated with chemotherapy and
28 (57·1%; 42·2–71·2) of 49 patients treated with ribociclib plus endocrine therapy. The proportion of patients who had breast conserving surgery was 39 (72·2%; 95% CI 58·4–83·5) of 54 patients in the chemo- therapy group and 44 (85·7%; 73·3–92·9) of 52 patients in the ribociclib plus letrozole group. No cases of progressive disease were observed.
An ROR score at day 15 was successfully identified in
97 (92%) of 106 patients (data were missing for the remaining eight patients because tissue samples were not available). 24 (37%) of 52 patients had switched to luminal A subtype by day 15 in the chemotherapy group and 47 (96%) of 49 patients had switched in the ribociclib plus letrozole group (appendix p 4). We also evaluated individual changes in ROR score in each group (appendix p 4). The results showed a decrease in ROR score at day 15 in the most patients in both treatment groups.
94 (89%) of 106 patients (46 of 54 in the chemotherapy group; 48 of 52 in the ribociclib) had ROR scores across the three timepoints (ie, baseline, day 15, and surgery; appendix p 5). The mean number of days the last dose of paclitaxel and surgery was 19·0 (SD 10·4) and of the last dose of ribociclib and surgery was 13·1 (14·6). In this subgroup of patients, we explored the changes in ROR score during and after treatment. In the chemotherapy group, the mean change in ROR score in the first 15 days was –17·2 (SD 17·4) compared with –27·0 (29·4) between day 15 and surgery. In patients treated with ribociclib plus letrozole, the mean change in ROR in the first 15 days was –49·7 (SD 13·8), and from day 15 to surgery was 3·3 (20·6). Specifically, after comparing the absolute changes in ROR score from day 15 to surgery in 48 patients treated with ribociclib plus letrozole, 21 (44%) had a decrease, 10 (21%) had an increase of 1–10 ROR points, and 14 (29%) had an increase of more than 10 ROR points (range 12–61). In 46 patients treated with chemotherapy, 40 (87%) had a decrease, two (4%) had an increase of 1–10 ROR points, and four (9%) had an increase of more than 10 ROR points (range 17–60).
The most frequent grade 1–2 adverse events were alopecia (52 [100%] of 52 patients), neutropenia (36 [69%]), and asthenia (28 [54%]) in the chemotherapy group and neutropenia (29 [57%] of 51 patients), increased alanine
aminotransferase (13 [26%]), and asthenia (13 [26%]) in the ribociclib plus letrozole group (table 3). Nearly all of the most frequent adverse events were deemed possibly related to study treatment (table 3). Grade 3–4 toxicities were observed in 29 (57%) of 51 patients in the ribociclib plus letrozole group and 36 (69%) of 52 patients in the chemotherapy group. The most common grade 3–4 adverse events in the ribociclib plus letrozole group were neutropenia (22 [43%] of 51 patients) and increased alanine aminotransferase concentrations (ten [20%]). In the ribociclib plus letrozole group, grade 1–3 QTc prolongation was observed in two (4%) of 51 patients and eight (16%) of 51 patients discontinued study treatment because of grade 3–4 increased alanine and aspartate aminotransferases. The mean number of days of treatment completion (the number of days that patients received treatment that were planned according to the protocol) in eight (16%) of 51 patients who stopped ribociclib early was 55·0% (SD 26·2). Four (8%) of
Figure 2: Absolute changes in ROR score between baseline and surgery
(A) Chemotherapy group; (B) ribociclib and letrozole group. The ROR risk group was determined at surgery. ROR=risk of relapse.
52 patients required a dose reduction of ribociclib (appendix p 6). The most common grade 3–4 adverse events in the chemotherapy group were neutropenia (31 [60%] of 52 patients) and febrile neutropenia (seven [13%]). In the chemotherapy group, ten (19%) of 52 patients discontinued study treatment because of grade 3–4 peripheral neuropathy and colitis, both during paclitaxel treatment. The mean number of days of treatment completion in eight (19%) of 52 patients who stopped paclitaxel early was 80·2% (SD 16·8). 16 (31%) of 52 patients required a dose reduction. 13 serious adverse events have been reported in ten patients, two (4%) of 51 patients in the ribociclib plus letrozole group and eight (15%) of 52 patients in the chemotherapy group. The most reported serious adverse events in the ribociclib plus letrozole group were lung abscess and increased aminotransferase concentrations (table 3). The most reported serious adverse event in the chemotherapy group was febrile neutropenia, in four patients. No deaths were observed during the study in either group.
Discussion
The results from the CORALLEEN study suggest that a high proportion of patients achieve a molecular
Ribociclib and letrozole group (n=51) Chemotherapy group (n=52)
Grade 1–2 Grade 3 Grade 4 Grade 1–2 Grade 3 Grade 4 determine the success of combining novel biological
agents as done in the I-SPY 2 framework20 before large prospective trials with long follow-ups are undertaken.
Any adverse event* 50 (98%) 29 (57%) 4 (8%) 52 (100%) 36 (69%) 27 (52%) Luminal B early stage breast cancer needs attention in
Neutropenia† 29 (57%) 22 (43%) 0 36 (69%) 31 (60%) 26 (50%) the clinic for several reasons. First, about 30% of all
Alanine aminotransferase 13 (26%) 7 (14%) 3 (6%) 4 (8%) 0 0 breast tumours are luminal B subtypes.21 Second, the risk
of presenting a distant recurrence at 10 years from
increase
Asthenia 13 (26%) 0 0 28 (54%) 2 (4%) 0
Alopecia 12 (24%) NA NA 52 (100%) NA NA diagnosis is more than 10–20%.22 Third, no targeted
therapies have been introduced in the past decade in luminal B early stage breast cancer beyond endocrine
Rash 12 (24%) 1 (2%) 0 10 (19%) 0 0
Aspartate aminotransferase increase 11 (22%) 5 (10%) 0 5 (10%) 0 0
Nausea 10 (20%) 0 0 16 (31%) 0 0
Pruritus 8 (16%) 1 (2%) 0 4 (8%) 0 0
Diarrhoea 7 (14%) 0 0 14 (27%) 1 (2%) 0
Fatigue 7 (14%) 0 0 10 (19%) 0 0
Hot flush 7 (14%) 0 0 0 0 0
Musculoskeletal pain 7 (14%) 0 0 4 (8%) 0 0
Constipation 6 (12%) 0 0 13 (25%) 0 0
Mucosal inflammation 5 (10%) 0 0 18 (35%) 1 (2%) 0
Anaemia 4 (8%) 0 0 18 (35%) 1 (2%) 0
Decreased appetite 4 (8%) 0 0 11 (21%) 0 0
Dysgeusia 2 (4%) 0 0 14 (27%) 0 0
Insomnia 2 (4%) 0 0 6 (12%) 0 0
Vomiting 2 (4%) 0 0 9 (17%) 0 0
Electrocardiogram QT prolonged 2 (4%) 1 (2%) 0 0 0 0
Neuropathy peripheral 1 (2%) 0 0 21 (40%) 2 (4%) 0
Onycholysis 1 (2%) 0 0 11 (21%) 0 0
γ-glutamyltransferase increased 1 (2%) 0 1 (2%) 0 1 (2%) 0
Lung abscess 0 1 (2%) 0 0 0 0
Lipase increased 0 0 1 (2%) 1 (2%) 0 0
Neurotoxicity 0 0 0 10 (19%) 2 (4%) 0
Febrile neutropenia 0 0 0 0 1 (2%) 6 (12%)
Thrombocytopenia‡ 0 0 0 3 (6%) 2 (4%) 0
Cellulitis 0 0 0 1 (2%) 1 (2%) 0
Colitis 0 0 0 0 2 (4%) 0
Deep vein thrombosis 0 0 0 0 1 (2%) 0
downstaging in clinically high-risk luminal B breast cancer treated with neoadjuvant ribociclib plus letrozole. Our results provide evidence for further studies in high- risk disease that aim to evaluate the long-term survival outcomes of a chemotherapy-free strategy after selecting patients based on variables such as the patient’s baseline risk of relapse and the pathological and biological response of the tumour to treatment. In addition, the concept of molecular downstaging could be explored to
therapy and chemotherapy.23 Fourth, large clinical and biological heterogeneity exists within this group of tumours.23 Fifth, this heterogeneity needs to be analysed and strategies to de-escalate or escalate locoregional and systemic therapies with or without the incorporation of novel agents need to be explored.23 For these reasons, we decided to focus on patients with luminal B early stage breast cancer in the CORALLEEN trial.
Contrary to the early stage disease setting, advanced or metastatic hormone receptor-positive, HER2-negative breast cancer has seen the successful introduction of targeted therapies in combination with endocrine therapies. Among them, CDK4/6 inhibitors have provided outstanding and clinically meaningful results in the endocrine naive and endocrine pretreated settings.16,24 For example, ribociclib in combination with goserelin and endocrine therapy has been shown to improve overall response and progression-free survival24 and, more recently, overall survival, compared with endocrine therapy alone.16 This class of drugs appears to be able to delay the use of chemotherapy in advanced setting.16
The evaluation of CDK4/6 inhibitors in combination with endocrine therapy in early stage, hormone receptor- positive, HER2-negative breast cancer is currently ongoing. The PENELOPE-B (NCT01864746) phase 3 trial is evaluating the value of 1 year of adjuvant palbociclib in patients who have not achieved a pCR following standard neoadjuvant chemotherapy. The PALLAS (NCT02513394) and MonarcHER (NCT03155997) phase 3 trials are assessing the value of 2 years of adjuvant palbociclib and abemaciclib, respectively. Finally, the NATALEE phase 3 trial (NCT03701334) is testing the efficacy of 3 years of adjuvant ribociclib in clinically high-risk disease. Thus, the strategy of these trials is to escalate therapy in luminal B breast cancer beyond chemotherapy and endocrine therapy.
The neoadjuvant setting offers a unique opportunity to tailor locoregional and systemic therapy. For example, patients who achieve a pCR after standard chemotherapy- based therapy have a better survival outcome compared with those without a pCR.25 The association between pCR and prognosis is seen across hormone receptor-positive and HER2-negative, HER2-positive, and triple negative subtypes.25 In addition, patients with HER2-positive or triple negative breast cancer can be offered additional treatment options in the adjuvant setting if they have residual disease following chemotherapy-based therapy.
For example, adjuvant T-DM1 in HER2-positive disease26 and capecitabine in HER2-negative disease27 can improve survival outcomes in this particular clinical scenario.
Although the proportions of patients with a pCR in HER2-positive or triple-negative disease are high (about 40–80%) with standard therapies, the proportion of patients with a pCR following neoadjuvant chemotherapy or endocrine therapy in hormone receptor-positive, HER2-negative breast cancer are low (eg, <10–20% with chemotherapy28 and <5% with endocrine therapy13). Consequently, additional predictors of long-term outcome are needed for patients with residual disease. Indeed, the biological study of residual tumours after neoadjuvant therapy in hormone receptor-positive, HER2-negative disease is of value. For example, the levels of Ki67 and oestrogen receptor by immunohistochemistry following 4–6 months of neoadjuvant endocrine monotherapy provide independent prognostic information beyond tumour size and nodal status.12 These findings led to the creation of a composite prognostic score in residual disease, known as preoperative endocrine prognostic index, which integrates all of these four variables into a single score. Preoperative endocrine prognostic index can identify patients with an outstanding outcome at 7–10 years if treated with endocrine therapy alone, in the absence of chemotherapy.29 In the context of neoadjuvant chemotherapy, these four variables evaluated in residual tumours also predict long-term survival outcomes in hormone receptor-positive, HER2-negative disease.14 Integration of pathological and biological response data after neoadjuvant treatment is prognostic in hormone receptor-positive, HER2-negative disease.
PAM50 (Prosigna) low-ROR was used as the primary endpoint in the CORALLEEN trial for several reasons. First, the same endpoint was needed for both groups. Second, no clear validated endpoint exists for both types of therapy. Third, preoperative endocrine prognostic index has been validated retrospectively in patients treated with endocrine monotherapy;12 however, its value in the context of CDK4/6 inhibition is unclear. Fourth, residual cancer burden has been validated retrospectively in patients treated with chemotherapy,18 but not in patients treated with endocrines. Fifth, the residual cancer burden 0–1 rates in the NEOPAL neoadjuvant trial following CDK4/6 inhibition and endocrine therapy were very low (ie, <10%).30 For these reasons, we decided to choose a standardised and validated prognostic tool that integrates biology and tumour burden into a single score and is associated with long-term prognosis independently of treatment received.7 Of note, ROR score was originally trained in a population of patients that did not receive any adjuvant systemic therapy.7 In addition, a retrospective study evaluating ROR score in residual tumours following endocrine neoadjuvant therapy showed a strong association with long-term prognosis and independently of preoperative endocrine prognostic index.11
Whether CDK4/6 inhibitors in combination with endocrine therapy can replace multiagent chemotherapy is unknown. In the metastatic setting, the young PEARL (KCSG-BR15-10) phase 2 trial31 randomly assigned
184 premenopausal patients to palbociclib plus endo- crine therapy versus capecitabine and showed that the chemotherapy-free strategy was superior in terms of progression-free survival (20·1 months vs 14·4 months). In early stage disease, the NeoPAL phase 2 trial randomly assigned 106 patients with luminal A or B disease to
19 weeks of palbociclib and letrozole or fluorouracil, epirubicin, and cyclophosphamide for three cycles followed by docetaxel for three cycles.30 The primary objective was the proportion of patients who had a residual cancer burden of 0–1. Residual cancer burden of 0–1 was observed in four patients (8%) in the palbociclib group and eight patients (16%) in the chemotherapy group. Preoperative endocrine prognostic index of 0 was observed in nine (18%) of 51 patients in palbociclib group and four (8%) of 50 patients in the chemotherapy group. According to both endpoints in the NeoPAL trial, less than 20% of patients might not need chemotherapy. The results suggest that the clinical and biological activity of ribociclib in combination with letrozole is very high; although, the trial was not powered to compare both treatment groups. In addition, this high activity of the combination is reported very early at day 15. Overall, the results of the NeoPAL and CORALLEEN trials suggest that a treatment strategy free from chemotherapy is an alternative worth exploring in future neoadjuvant or adjuvant trials in patients with high-risk luminal B disease. However, whether preoperative endocrine prognostic index, ROR, or both, is the best biomarker to identify this patient population is unknown.
Our study has several limitations worth noting. First, the study is exploratory and was not designed to formally compare the proportion of patients with low-ROR disease in both groups. However, the proportion of low-ROR disease in the no-chemotherapy group was high at 24 weeks (about 47%), and this subpopulation of patients could be the focus of future trials. Second, the trial ended after surgery, no long-term follow-up was done and the use of chemotherapy was left at the discretion of the physician. Therefore, this underpowered trial will not be informative regarding the long-term survival outcome of patients with low-ROR residual disease treated without chemotherapy. Third, although PAM50 ROR measured at diagnosis has level 1b (when the biomarker shows consistent results in at least two retrospective studies using tumour samples from prospective clinical trials) evidence for predicting 10-year distant metastasis-free survival in the absence of chemotherapy and after 5 years of endocrine therapy, whether this measurement has utility as a prognostic biomarker when measured after neoadjuvant CDK4/6 inhibition and endocrine therapy is still unclear. However, other commonly used biomarkers such as preoperative endocrine prognostic index score or
residual cancer burden do not have level 1b evidence either and are not standardised.
Following our results, the question remains as to whether neoadjuvant or adjuvant chemotherapy could be replaced in high-risk, luminal B, early stage breast cancer using a CDK4/6 inhibitor plus endocrine therapy. Our opinion is that the findings are not complete to support this strategy. However, the results of the CORALLEEN trial do suggest that the neoadjuvant approach is feasible and might allow a better identifi- cation of the patients who might have a good outcome without chemotherapy. The definitive clinical study should be powered to show that the short-term and long-term survival outcomes in patients with low- ROR disease at surgery after 6 months of CDK4/6 inhibition and endocrine therapy are outstanding in the absence of chemotherapy. Another strategy would be a randomised adjuvant trial with a non-inferiority design of chemotherapy versus not in high-risk luminal B disease treated with 2–3 years of CDK4/6 inhibition. However, this study design would require thousands of patients and a risk is that some patients might still benefit from both treatment strategies, especially about 20% of patients in CORALLEEN who presented with high ROR scores in residual disease after CDK4/6 inhibition.
Contributors
AP and JG had the idea for and designed the study. AP, JG, TP, LP, PG, NC, PV, and PC contributed to data collection and assembly. AP, JG, TP, PV, and LP interpreted and analysed data. All authors wrote and reviewed the report and approved the final version for submission.
Declaration of interests
AP reports grants and personal fees from Novartis during the study; grants and personal fees from Pfizer and Amgen; personal fees from Lilly, Nanostring technologies, Oncolytics Biotech, Daiichi Sankyo, Puma, and Bristol-Myers Squibb; and grants from Roche, outside of the submitted work. CS reports non-financial support from ArticulateScience during the study; personal fees from Puma Biotechnology, Pfizer, Roche, AstraZeneca, Celgene, Daiichi Sankyo, Eisai, Genomyc Health, Novartis, Pierre Fabre, Synthon Biopharmaceuticals; and grants from Roche-Genentech, Macrogenics, Pfizer, Piqur Therapeutics, Puma biotechnology, Synthon biopharmaceuticals, Novartis, outside of the submitted work.
MM reports grants from Novartis, Breast Cancer Research Foundation- AACR, Breast Cancer Now Career Catalyst, during the study and advisors from Novartis, Roche, Pfizer and Lilly, outside of the submitted work. XGF reports personal fees from SOLTI, during the study; personal fees and non-financial support from Roche; personal fees from Eisai, outside of the submitted work. MO reports personal fees from Roche, GlaxoSmithKline, Puma Biotechnology, Grünenthal Group, Novartis, Pierre Fabre, GP Pharm; and grants from Philips Healthcare, outside of the submitted work. MG-G reports personal fees from Genentech, Novartis, Pfizer, Daiichi, Eisai during the study and outside of the submitted work. MA reports personal fees from Roche, Eisai, MSD, outside of this work. YI reports grants from Novartis, Breast Cancer Research Foundation-AACR Career Development Award, Breast Cancer Now Catalyst Grant during the study and personal fees and non-financial support from Roche, Novartis, Pfizer, AstraZeneca, and Eisai outside of the submitted work. AL-C reports grants, personal fees and non- financial support from Novartis, Roche, AstraZeneca, Eli Lilly, and Pfizer; grants and non-financial support from EISAI, grants and personal fees from Genomic Health, GlaxoSmithKline Tesaro, personal fees from MSD, personal fees and non-financial support from Bristol- Myers Squibb outside of the submitted work. EM reports personal fees
from Roche. JAV reports personal fees from Roche. RL reports grants and personal fees from Roche, Merck, personal fees from AstraZeneca, Bayer, Pharmamar, Leo and personal fees and non-financial support from Bristol-Myers Squibb and Novartis, outside of the submitted work. EC reports personal fees from Roche, Lilly, Novartis, and Pfizer during the study. JG reports grants and personal fees from Novartis, Pfizer, Roche outside of this work. All other CGS 20267 authors declare no competing interests.
Data sharing statement
Data sharing statement is in the appendix (p 7).
Acknowledgments
Novartis funded the study and provided ribociclib for the study. Breast Cancer Research Foundation-AACR funded the translational research.
References
1 Carioli G, Malvezzi M, Rodriguez T, Bertuccio P, Negri E,
La Vecchia C. Trends and predictions to 2020 in breast cancer mortality in Europe. Breast 2017; 36: 89–95.
2 Early Breast Cancer Trialists’ Collaborative Group. Comparisons between different polychemotherapy regimens for early breast cancer: meta-analyses of long-term outcome among 120 000 women in 123 randomised trials. Lancet 2012; 379: 432–44.
3 Cejalvo JM, Pascual T, Fernández-Martínez A, et al. Clinical implications of the non-luminal intrinsic subtypes in hormone receptor-positive breast cancer. Cancer Treat Rev 2018; 67: 63–70.
4 Prat A, Ellis MJ, Perou CM. Practical implications of gene- expression-based assays for breast oncologists. Nat Rev Clinl Oncol 2012; 9: 48–57.
5 Sparano JA, Gray RJ, Makower DF, et al. Adjuvant chemotherapy guided by a 21-gene expression assay in breast cancer. N Engl J Med 2018; 379: 111–21.
6 Dubsky P, Brase JC, Jakesz R, et al. The EndoPredict score provides prognostic information on late distant metastases in ER+/HER2− breast cancer patients. Br J Cancer 2013; 109: 2959–64.
7 Wallden B, Storhoff J, Nielsen T, et al. Development and verification of the PAM50-based Prosigna breast cancer gene signature assay. BMC Med Genomics 2015; 8: 54.
8 Krop I, Ismaila N, Andre F, et al. Use of biomarkers to guide decisions on adjuvant systemic therapy for women with early-stage invasive breast cancer: American Society of Clinical Oncology Clinical practice guideline focused update. J Clin Oncol 2017;
35: 2838–47.
9 Sestak I, Buus R, Cuzick J, et al. Comparison of the performance of 6 prognostic signatures for estrogen receptor-positive breast cancer: a secondary analysis of a randomized clinical trial. JAMA Oncol 2018; 4: 545–53.
10 Dowsett M, Smith IE, Ebbs SR, et al. Prognostic value of Ki67 expression after short-term presurgical endocrine therapy for primary breast cancer. J Natl Cancer Inst 2007; 99: 167–70.
11 Gil MJG, Perez FJ, Monso TS, et al. Prognostic value of PAM50 in residual breast cancer following neoadjuvant endocrine therapy (NET): a retrospective analysis with long follow-up.
Proc Am Soc Clin Oncol 2019; 37 (suppl): 575.
12 Ellis MJ, Tao Y, Luo J, et al. Outcome prediction for estrogen receptor–positive breast cancer based on postneoadjuvant endocrine therapy tumor characteristics. J Natl Cancer Inst 2008; 100: 1380–88.
13 Ellis MJ, Suman VJ, Hoog J, et al. Randomized phase II neoadjuvant comparison between letrozole, anastrozole, and exemestane for postmenopausal women with estrogen receptor– rich stage 2 to 3 breast cancer: clinical and biomarker outcomes and predictive value of the baseline PAM50-based intrinsic subtype— ACOSOG Z1031. J Clin Oncol 2011; 29: 2342–49.
14 Sheri A, Smith IE, Johnston SR, et al. Residual proliferative cancer burden to predict long-term outcome following neoadjuvant chemotherapy. Ann Oncol 2014; 26: 75–80.
15 von Minckwitz G, Schmitt WD, Loibl S, et al. Ki67 measured after neoadjuvant chemotherapy for primary breast cancer.
Clin Cancer Res 2013; 19: 4521–31.
16 Im S-A, Lu Y-S, Bardia A, et al. Overall survival with ribociclib plus endocrine therapy in breast cancer. N Engl J Med 2019; 381: 307–16.
17 Pernas S, Tolaney SM, Winer EP, Goel S. CDK4/6 inhibition in breast cancer: current practice and future directions.
Ther Adv Med Oncol 10: 1758835918786451.
18 Symmans WF, Wei C, Gould R, et al. Long-term prognostic risk after neoadjuvant chemotherapy associated with residual cancer burden and breast cancer subtype. J Clin Oncol 2017; 35: 1049–60.
19 Clopper CJ, Pearson ES. The use of confidence or fiducial limits illustrated in the case of the binomial. Biometrika 1934; 26: 404–13.
20 Barker A, Sigman C, Kelloff G, Hylton N, Berry D, Esserman L.
I-SPY 2: An adaptive breast cancer trial design in the setting of neoadjuvant chemotherapy. Clin Pharmacol Ther 2009; 86: 97–100.
21 Prat A, Perou CM. Deconstructing the molecular portraits of breast cancer. Mol Oncol 2011; 5: 5–23.
22 Parker JS, Mullins M, Cheang MCU, et al. Supervised risk predictor of breast cancer based on intrinsic subtypes. J Clin Oncol 2009;
27: 1160–67.
23 Ades F, Zardavas D, Bozovic-Spasojevic I, et al. Luminal B breast cancer: molecular characterization, clinical management, and future perspectives. J Clin Oncol 2014; 32: 2794–803.
24 Tripathy D, Im S-A, Colleoni M, et al. Ribociclib plus endocrine therapy for premenopausal women with hormone-receptor-positive, advanced breast cancer (MONALEESA-7): a randomised phase 3 trial. Lancet Oncol 2018; 19: 904–15.
25 Cortazar P, Zhang L, Untch M, et al. Pathological complete response and long-term clinical benefit in breast cancer: the CTNeoBC pooled analysis. Lancet 2014; 384: 164–72.
26 von Minckwitz G, Huang C-S, Mano MS, et al. Trastuzumab emtansine for residual invasive HER2-positive breast cancer. N Engl J Med 2018; 380: 617–28.
27 Masuda N, Lee S-J, Ohtani S, et al. Adjuvant capecitabine for breast cancer after preoperative chemotherapy. N Engl J Med 2017;
376: 2147–59.
28 Prat A, Fan C, Fernández A, et al. Response and survival of breast cancer intrinsic subtypes following multi-agent neoadjuvant chemotherapy. BMC Med 2015; 13: 303.
29 Ellis MJ, Suman VJ, Hoog J, et al. Ki67 proliferation index as a tool for chemotherapy decisions during and after neoadjuvant aromatase inhibitor treatment of breast cancer: results from the American College of Surgeons Oncology Group Z1031 Trial (Alliance). J Clin Oncol 2017; 35: 1061–69.
30 Cottu P, D’Hondt V, Dureau S, et al. Letrozole and palbociclib versus chemotherapy as neoadjuvant therapy of high-risk luminal breast cancer. Ann Oncol 2018; 29: 2334–40.
31 Park YH, Kim T-Y, Kim GM, et al. Palbociclib plus exemestane with gonadotropin-releasing hormone agonist versus capecitabine in premenopausal women with hormone receptor-positive,
HER2-negative metastatic breast cancer (KCSG-BR15-10):
a multicentre, open-label, randomised, phase 2 trial. Lancet Oncol
2019; 20: 1750–59.