Abstract
The primary objective was to evaluate intracranial objective response rate (iORR) in patients receiving abemaciclib with brain or leptomeningeal metastases (LM) secondary to hormone receptor–positive (HR+) metastatic breast cancer (MBC). Secondary objectives evaluated extracranial response, abemaciclib pharmacokinetics, brain metastases tissue exposure, and safety.
This nonrandomized, phase II study (NCT02308020) enrolled patients in tumor subtype–specific cohorts A–D: A (HR+, HER2− MBC), B (HR+, HER2+ MBC), C (HR+ MBC LM), and D (brain metastases surgical resection). Abemaciclib 200 mg was administered twice daily as monotherapy or with endocrine therapy, or 150 mg twice daily with trastuzumab. Cohorts A and B used a Simon two-stage design.
In cohort A (n = 58), 3 patients were confirmed responders resulting in an iORR of 5.2% [95% confidence interval (CI), 0.0–10.9], and the intracranial clinical benefit rate (iCBR) was 24% (95% CI, 13.1–35.2). Median overall survival (OS) was 12.5 months (95% CI, 9.3–16.4). A volumetric decrease in target intracranial lesions was experienced by 38% of patients. In cohort B (n = 27), there were no confirmed intracranial responses. An iCBR of 11% (95% CI, 0.0–23.0) was observed. Median OS was 10.1 months (95% CI, 4.2–14.3). A volumetric decrease in target intracranial lesions was experienced by 22% of patients. In cohort C (n = 10), one confirmed complete parenchymal response was observed. In cohort D (n = 9), unbound brain metastases concentrations of total active abemaciclib analytes were 96- [cyclin-dependent kinase 4 (CDK4)] and 19-fold (CDK6) above in vitro IC50. Safety was consistent with prior studies.
This study did not meet its primary endpoint. Abemaciclib was associated with an iCBR of 24% in patients with heavily pretreated HR+, HER2− MBC. Abemaciclib achieved therapeutic concentrations in brain metastases tissue, far exceeding those necessary for CDK4 and CDK6 inhibition. Further studies are warranted, including assessing novel abemaciclib-based combinations.
This article is featured in Highlights of This Issue, p. 5269
Clinical data demonstrate abemaciclib, a selective cyclin-dependent kinase 4 and 6 inhibitor approved to treat patients with hormone receptor–positive (HR+), HER2− metastatic breast cancer (MBC), penetrates the blood–brain barrier resulting in comparable concentrations in cerebrospinal fluid (CSF) and plasma. Given that breast cancer is the second leading cause of metastases to the central nervous system (CNS), including brain metastases and leptomeningeal metastases (LM), this phase II study evaluated the activity and safety of abemaciclib as monotherapy and combined with endocrine therapy in patients with brain metastases or LM secondary to HR+ MBC. While this study did not meet the primary endpoint of a confirmed intracranial objective response rate, there was evidence of clinical benefit among patients with HR+, HER2− MBC. In addition, pharmacokinetic results demonstrated relevant concentrations of abemaciclib and its active metabolites in brain metastases and CSF. In patients with LM, abemaciclib treatment was associated with disease control, and overall survival longer than expected, compared with historical controls. Given treatment challenges among patients with MBC with CNS involvement, these results support further evaluations, including testing novel abemaciclib-based combinations.
Introduction
Brain metastases (BM) are the most common intracranial malignancy in adults (1). Breast cancer is the second leading cause of metastases to the central nervous system (CNS), including the brain and leptomeninges (2–4). Evidence of increasing intracranial recurrences in patients with metastatic breast cancer (MBC) is likely because of improved systemic treatment of extracranial disease leading to prolonged survival and higher intracranial disease detection rates (3). Prognosis after CNS recurrence varies by subtype and involvement, with 4-month estimated median survival in patients with leptomeningeal metastases (LM) and 3.4- to 25.0-month median survival in patients with predominantly parenchymal intracranial disease, depending on graded prognostic assessment scores (5, 6).
Current standard-of-care for parenchymal brain metastases include surgical resection, stereotactic radiosurgery (SRS), and/or whole-brain radiotherapy (WBRT; refs. 7, 8). While these local therapies palliate neurologic symptoms, they may cause neurocognitive deficits and leave extracranial disease untreated (7, 9). The role of systemic therapies to treat breast cancer brain metastases is increasingly important. There is some evidence suggesting a treatment benefit from HER2-targeted tyrosine kinase inhibitors in patients with HER2+ breast cancer (3, 10–14). Although some anticancer agents have shown promising extracranial results, their ability to cross the blood–brain barrier (BBB) and deliver therapeutic effects into the CNS presents challenges. Many drugs can penetrate a disrupted blood–tumor barrier, but more uniform penetration across the intact BBB could potentially delay drug resistance and prevent new brain metastases (15, 16).
Abemaciclib, a selective cyclin-dependent kinase (CDK) 4 and CDK6 inhibitor, has demonstrated efficacy in patients with hormone receptor–positive (HR+), HER2− MBC when combined with endocrine therapy (ET; refs. 17–19) and as monotherapy in patients pretreated with chemotherapy (20). In rodents, abemaciclib exhibited BBB penetration and prolonged survival in an orthotopic U87MG intracranial glioblastoma xenograft model (21). In a phase I study, abemaciclib achieved unbound cerebrospinal fluid (CSF) concentrations similar to time-matched plasma concentrations in patients with glioblastoma (22). Together, preclinical and clinical evidence indicates abemaciclib may have activity in brain metastases arising from HR+ MBC.
The objective of this analysis was to evaluate the activity and safety of abemaciclib in patients with brain metastases or LM secondary to HR+ MBC.
Patients and Methods
Study design
The primary objective of this open-label, nonrandomized, multi-cohort trial was to evaluate the intracranial objective response rate (iORR). Secondary outcomes included additional intra- and extracranial outcomes, as well as the safety and pharmacokinetics of abemaciclib and its metabolites.
Patients were enrolled into tumor-specific cohorts (Fig. 1). Cohorts A and B had brain metastases secondary to HR+ MBC, either HER2− (cohort A) or HER2+ (cohort B). Cohort C enrolled patients with LM secondary to HR+ MBC (HER2+/HER2−). Cohort D enrolled patients with planned surgical resection of brain metastases, secondary to HR+ MBC (HER2+/HER2−) or non–small cell lung cancer (NSCLC). This study included additional cohorts of patients with NSCLC or melanoma; results from these patients will be reported in a separate manuscript. The study protocol (Supplementary Materials and Methods) was approved by institutional review boards and ethics committees before initiation and conducted in accordance with the Declaration of Helsinki. All patients provided written informed consent.
Study design and patient disposition. aPatients were not randomized to treatment. All patients received abemaciclib. bAdditional cohorts enrolling a total of 58 patients with brain metastases secondary to NSCLC or melanoma are not reported here. HER+/−, human epidermal growth factor receptor-2 positive/negative; HR+, hormone receptor positive; MBC, metastatic breast cancer; NSCLC, non–small cell lung cancer.
Study design and patient disposition. aPatients were not randomized to treatment. All patients received abemaciclib. bAdditional cohorts enrolling a total of 58 patients with brain metastases secondary to NSCLC or melanoma are not reported here. HER+/−, human epidermal growth factor receptor-2 positive/negative; HR+, hormone receptor positive; MBC, metastatic breast cancer; NSCLC, non–small cell lung cancer.
Patients
Patients in cohorts A and B were eligible with ≥1 new or not previously irradiated and measurable [per Response Assessment in Neuro-Oncology (RANO) brain metastases criteria; ref. 23] brain metastases ≥10 mm, or progressive previously irradiated brain metastases secondary to HR+ MBC who had not received prior treatment with any cyclin-dependent kinase (CDK) 4 and CDK6 inhibitor. For patients receiving ET at enrollment, continued ET was permitted if extracranial disease was stable for ≥3 months and CNS progressive disease (PD) occurred while on ET; initiation of a new ET was not allowed as there was concern this would make it difficult to assess attribution of response to study treatment. For patients in cohort B (HER2+ MBC), trastuzumab was allowed (per investigator discretion) if initiated before or simultaneously with abemaciclib because continuation of HER2-directed therapy beyond progression is an accepted treatment strategy (24). For patients in cohorts A and B, prior surgical resection or radiotherapy was permitted if local therapy was completed ≥14 days prior to initiating abemaciclib, and the patient had recovered from all acute effects.
Cohort C required LM presence (by positive CSF cytology, or clinical signs and symptoms associated with abnormal MRI features), secondary to HR+ (HER2− or HER2+) MBC, regardless of concomitant parenchymal metastases, with stable disease (SD) for ≥4 weeks following WBRT or SRS.
Patients in cohort D had brain metastases clinically indicated for surgical resection (per physician's judgement) secondary to HR+ (HER2− or HER2+) MBC or NSCLC. Patients with MBC in cohort D could continue or initiate ET with no additional restrictions. Prior treatment with either palbociclib or ribociclib, but not abemaciclib, was permitted.
Additional study-wide key inclusion and exclusion criteria are detailed in the study protocol (Supplementary Materials and Methods; redacted protocol, pages 27–31).
Treatment
Abemaciclib (200 mg) was administered twice daily orally as monotherapy, or with ET, for a 21-day cycle. For patients with HER+ MBC taking concurrent trastuzumab, twice daily 150 mg abemaciclib was given. Dose reductions were permitted for toxicity management (25). All patients remained on study until PD or other discontinuation criteria were met.
Patients in cohort D received abemaciclib 5–14 days before surgical resection, as described above, with the last dose taken 6–12 hours before surgical resection, allowing for drug concentration assessment in brain metastases tissue. Patients resumed abemaciclib 15–21 days postsurgery, allowing adequate wound healing, and remained on study until PD or other discontinuation criteria were met.
Efficacy and safety measures
Baseline brain metastases (cohorts A, B, and D) or LM (cohort C) were assessed locally by CT or MRI ≤28 days before treatment initiation. Subsequent tumor assessments occurred every 6 weeks for the first 24 weeks, and approximately every 12 weeks thereafter. Target (maximum, five) and nontarget lesions were assessed according to RANO-BM (23) composite criteria for intracranial tumors (RANO-LM composite criteria for LM; ref. 26) and RECIST v1.1 (27) for extracranial tumors. Safety evaluations at all patient visits included laboratory and adverse event (AE) assessments using NCI Common Terminology Criteria v4.0 (28).
Endpoints
The primary endpoint was iORR [complete response (CR) + partial response (PR)] according to RANO-BM criteria. An intracranial CR was defined as disappearance of all CNS target lesions sustained ≥4 weeks, appearance of no new lesions, no corticosteroid therapy, and patient condition classified as clinically stable or improved. An intracranial PR (iPR) was defined as a ≥30% decrease in the sum longest diameter sustained ≥4 weeks, appearance of no new lesions, treatment regimen including stable-to-decreased corticosteroid dose, and patient condition classified as clinically stable or improved. Secondary intracranial endpoints included best overall response, disease control rate (iDCR, CR + PR + SD), and clinical benefit rate (iCBR, CR + PR + SD ≥6 months). Extracranial endpoints included eORR, eDCR, and eCBR. Median progression-free survival (PFS), specifically bicompartmental PFS, iPFS, and ePFS were assessed. Bicompartmental PFS was measured from randomization date to the earliest date of CNS or non-CNS PD, per RECIST v1.1, or death from any cause. Overall survival (OS), measured in months, was defined as date of enrollment to the date of death of any cause. For patients not known to have died as of the data cutoff (November 8, 2018), OS was censored at the date of last contact prior to the inclusion cut-off date.
Pharmacokinetics
In all cohorts, blood samples were collected on the first day of each 21-day cycle, cycles 1–4. In cohort C, a CSF sample was drawn on cycle 3 day 1, before abemaciclib dose. For cohort D, blood, CSF, and brain metastases tissue samples were collected during surgery, and another blood sample taken 6 hours postsurgery. Validated LC/MS-MS methods (Q2 solutions) assessed concentrations of abemaciclib and its active metabolites [LSN2839576 (M2) and LSN3106726 (M20); refs. 29, 30], in plasma, CSF, and brain metastases tissue. Patients receiving ≥1 abemaciclib dose who had ≥1 concentration measurement were included in pharmacokinetic analyses. Total unbound active analytes were calculated by summating unbound abemaciclib, M2, and M20 concentrations at each timepoint.
Statistical analysis
Cohorts A and B used a Simon two-stage design as an interim analysis; the decision criteria to open the second stage was iORR, assuming a 0.05 one-sided type-I error, and 80% power. If stage I had ≥2 responders of 23 enrolled, accrual continued to the second stage until 33 additional patients were enrolled. Among the 56 enrolled patients, six responders were needed to warrant further investigation. This method tested the null hypothesis, a true abemaciclib iORR ≤ 5%, versus the alternative, a true iORR ≥ 15%.
All analyses included patients receiving ≥1 abemaciclib dose. The Kaplan–Meier method calculated point estimates and exact 95% confidence intervals (CI) for time-to-event outcomes. Descriptive statistics summarized safety data. All tests assumed a two-sided 0.05 alpha, and CIs assumed a two-sided 95% level. Analyses were conducted with SAS version 9.2 (SAS Institute Inc.).
Results
Patients and disease characteristics
Between April 2015 and October 2018, 104 patients were enrolled and received ≥1 abemaciclib dose (cohort A, n = 58; cohort B, n = 27; cohort C, n = 10; and cohort D, n = 9; Fig. 1). Among patients who received any prior systemic therapy for advanced disease, a median number of three lines was reported in both cohort A (range 1–10) and cohort B (range 1–12; Table 1). In cohort A, 16 patients received concomitant ET [12 (75.0%) of whom were taking an aromatase inhibitor] and in cohort B, 6 patients received concomitant trastuzumab. Patient, disease, and prior therapy characteristics, detailed in Table 1, indicate a heavily pretreated patient population.
Baseline patient and disease characteristics.
. | Cohort A . | Cohort B . | Cohort C . | . | |
---|---|---|---|---|---|
. | HR+, HER2− MBC . | HR+, HER2+ MBC . | HR+, HER2− MBC with LM . | HR+, HER2+ MBC with LM . | Cohort D, surgicala . |
n (%), unless otherwise stated . | n = 58 . | n = 27 . | n = 7 . | n = 3 . | n = 9 . |
Age, years, median (range) | 55 (30–79) | 51 (30–73) | 53 (39–60) | 41 (38–41) | 65 (30–82) |
Female sex | 57 (98.3) | 27 (100.0) | 7 (100.0) | 3 (100.0) | 8 (88.9) |
Disease durationb, years, median (range) | 7 (1–18) | 5 (1–16) | 9 (5–14) | 5 (3–6) | 3 (1–5) |
Target brain lesions, number, median (range) | 2 (1–5) | 2 (1–5) | 1 (1–3) | 1 (1–1) | 1 (1–2) |
Prior anticancer surgery or radiation for BMc | |||||
Surgical resection | 4 (6.9) | 2 (7.4) | 0 (0.0) | 0 (0.0) | 1 (11.1) |
WBRT | 27 (46.6) | 16 (59.3) | 2 (28.6) | 1 (33.3) | 0 (0.0) |
SRS | 20 (34.5) | 10 (37.0) | 3 (42.9) | 0 (0.0) | 1 (11.1) |
Prior lines of systemic anticancer therapy for metastatic disease | |||||
Any systemic therapy, n (%) | 51 (87.9) | 27 (100.0) | 5 (71.4) | 1 (33.3) | 5 (55.6) |
[median (range)]d | [3 (1–10)] | [3 (1–12)] | [4 (4–6)] | [6 (6–6)] | [1 (1–6)] |
ET, n (%) | 41 (70.7) | 12 (44.4) | 3 (42.9) | 1 (33.3) | 3 (33.3) |
[median (range)]d | [2 (1–4)] | [1 (1–3)] | [1 (1–2)] | [1 (1–1)] | [1 (1–1)] |
Chemotherapy, n (%) | 44 (75.9) | 25 (92.6) | 5 (71.4) | 1 (33.3) | 3 (33.3) |
[median (range)]d | [2 (1–6)] | [2 (1–6)] | [4 (2–5)] | [5 (5–5)] | [3 (1–4)] |
Target therapy, n (%) | 16 (27.6) | 27 (100.0) | 4 (57.1) | 1 (33.3) | 1 (11.1) |
[median (range)]d | [1 (1–3)] | [3 (1–11)] | [2 (1–2)] | [4 (4–4)] | [2 (2–2)] |
Prior trastuzumab | 3 (5.2) | 27 (100.0) | 0 (0) | 3 (100.0) | 2 (22.2) |
KPS | |||||
≥90 | 37 (63.8) | 17 (63.0) | 2 (28.6) | 2 (66.7) | 4 (44.4) |
80 | 16 (27.6) | 4 (14.8) | 4 (57.1) | 0 (0) | 1 (11.1) |
70 | 5 (8.6) | 5 (18.5) | 1 (14.3) | 1 (33.3) | 4 (44.4) |
60e | 0 (0) | 1 (3.7) | 0 (0) | 0 (0) | 0 (0) |
. | Cohort A . | Cohort B . | Cohort C . | . | |
---|---|---|---|---|---|
. | HR+, HER2− MBC . | HR+, HER2+ MBC . | HR+, HER2− MBC with LM . | HR+, HER2+ MBC with LM . | Cohort D, surgicala . |
n (%), unless otherwise stated . | n = 58 . | n = 27 . | n = 7 . | n = 3 . | n = 9 . |
Age, years, median (range) | 55 (30–79) | 51 (30–73) | 53 (39–60) | 41 (38–41) | 65 (30–82) |
Female sex | 57 (98.3) | 27 (100.0) | 7 (100.0) | 3 (100.0) | 8 (88.9) |
Disease durationb, years, median (range) | 7 (1–18) | 5 (1–16) | 9 (5–14) | 5 (3–6) | 3 (1–5) |
Target brain lesions, number, median (range) | 2 (1–5) | 2 (1–5) | 1 (1–3) | 1 (1–1) | 1 (1–2) |
Prior anticancer surgery or radiation for BMc | |||||
Surgical resection | 4 (6.9) | 2 (7.4) | 0 (0.0) | 0 (0.0) | 1 (11.1) |
WBRT | 27 (46.6) | 16 (59.3) | 2 (28.6) | 1 (33.3) | 0 (0.0) |
SRS | 20 (34.5) | 10 (37.0) | 3 (42.9) | 0 (0.0) | 1 (11.1) |
Prior lines of systemic anticancer therapy for metastatic disease | |||||
Any systemic therapy, n (%) | 51 (87.9) | 27 (100.0) | 5 (71.4) | 1 (33.3) | 5 (55.6) |
[median (range)]d | [3 (1–10)] | [3 (1–12)] | [4 (4–6)] | [6 (6–6)] | [1 (1–6)] |
ET, n (%) | 41 (70.7) | 12 (44.4) | 3 (42.9) | 1 (33.3) | 3 (33.3) |
[median (range)]d | [2 (1–4)] | [1 (1–3)] | [1 (1–2)] | [1 (1–1)] | [1 (1–1)] |
Chemotherapy, n (%) | 44 (75.9) | 25 (92.6) | 5 (71.4) | 1 (33.3) | 3 (33.3) |
[median (range)]d | [2 (1–6)] | [2 (1–6)] | [4 (2–5)] | [5 (5–5)] | [3 (1–4)] |
Target therapy, n (%) | 16 (27.6) | 27 (100.0) | 4 (57.1) | 1 (33.3) | 1 (11.1) |
[median (range)]d | [1 (1–3)] | [3 (1–11)] | [2 (1–2)] | [4 (4–4)] | [2 (2–2)] |
Prior trastuzumab | 3 (5.2) | 27 (100.0) | 0 (0) | 3 (100.0) | 2 (22.2) |
KPS | |||||
≥90 | 37 (63.8) | 17 (63.0) | 2 (28.6) | 2 (66.7) | 4 (44.4) |
80 | 16 (27.6) | 4 (14.8) | 4 (57.1) | 0 (0) | 1 (11.1) |
70 | 5 (8.6) | 5 (18.5) | 1 (14.3) | 1 (33.3) | 4 (44.4) |
60e | 0 (0) | 1 (3.7) | 0 (0) | 0 (0) | 0 (0) |
Abbreviations: BM, brain metastases; KPS, Karnofsky performance status; LM, leptomeningeal metastases.
aSurgical resection is clinically indicated per treating physician's judgement in these patients.
bDisease duration is time since initial diagnosis to study enrollment date.
cPrior therapies were measured from time of initial diagnosis to study enrollment date.
dMedian was calculated for patients who took the therapy.
eKPS of 60 at baseline is a protocol deviation.
Intra- and extracranial activity
Cohort A (HR+, HER2−)
At data cutoff, 55 (94.8%) patients were off-treatment (Table 2). Discontinuation due to PD was most common, occurring in 74.1% of cohort A patients (53.4% iPD; 3.4% ePD; 17.2% iPD and ePD; Table 2). Two confirmed iPRs were observed in the first 23 patients, meeting the criteria for expansion to stage II; 35 additional patients were enrolled, totaling 58 who completed stage II. One additional iPR was observed during stage II, resulting in a confirmed iORR of 5.2% (95% CI, 0.0%–10.9%; Table 2). A best response of iSD in 35 patients resulted in an iDCR of 65.5% (95% CI, 53.3%–77.7%). In 11 patients, the iSD lasted ≥6 months, for an iCBR of 24.1% (95% CI, 13.1%–35.2%). In addition, 38% of patients experienced a decrease in the size of intracranial target lesions; 19% experienced a ≥20% volumetric decrease relative to baseline (Fig. 2A). The median iPFS was 4.9 months (95% CI, 2.9–5.6).
Reason for abemaciclib discontinuation, and overall responses and survival summary in patients with HR+, HER2− MBC or HR+, HER+ MBC.
. | Cohort A . | Cohort B . | Cohort C . | |
---|---|---|---|---|
. | HR+, HER2− MBC . | HR+, HER2+ MBC . | HR+, HER2− MBC LM . | HR+, HER2+ MBC LM . |
n (%), unless otherwise stated . | (n = 58) . | (n = 27) . | (n = 7) . | (n = 3) . |
Remained on treatment | 3 (5.2) | 0 (0) | 0 (0) | 0 (0) |
Discontinued abemaciclib | 55 (94.8) | 27 (100) | 7 (100) | 3 (100) |
Reason for discontinuation | ||||
AE | 5 (8.6) | 3 (11.1) | 1 (14.3) | 0 (0) |
Death | 3 (5.2) | 1 (3.7) | 0 (0) | 0 (0) |
PD | 43 (74.1) | 23 (85.2) | 6 (85.7) | 3 (100) |
Intracranial | 31 (53.4) | 17 (63.0) | 6 (85.7) | 1 (33.3) |
Extracranial | 2 (3.4) | 2 (7.4) | 0 (0) | 2 (66.7) |
Intra- and extracranial | 10 (17.2) | 4 (14.8) | 0 (0) | 0 (0) |
Withdrawal by patient | 4 (6.9) | 0 (0) | 0 (0) | 0 (0) |
Intracranial outcomesa | ||||
ORR (CR/PR) | 3 (5.2) | 0 (0) | 0 (0) | 0 (0) |
DCR (CR/PR/SD) | 38 (65.5) | 12 (44.4) | 2 (28.6) | 1 (33.3) |
CBR (CR/PR/SD ≥ 6 months) | 14 (24.1) | 3 (11.1) | 1 (14.3) | 0 (0) |
CR | 0 (0) | 0 (0) | 0 (0) | 0 (0) |
PR | 3 (5.2) | 0 (0) | 0 (0) | 0 (0) |
SD | 35 (60.3) | 12 (44.4) | 2 (28.6) | 1 (33.3) |
SD ≥6 months | 11 (19.0) | 3 (11.1) | 1 (14.3) | 0 (0) |
PD | 15 (25.9) | 13 (48.1) | 2 (28.6) | 0 (0) |
Extracranial outcomesb | ||||
ORRc (CR/PR) | 2 (3.4) | 0 (0) | 0 (0) | 1 (33.3) |
DCR (CR/PR/SD) | 30 (51.7) | 11 (40.7) | 4 (57.1) | 1 (33.3) |
CBR (CR/PR/SD ≥ 6 months) | 12 (20.7) | 1 (3.7) | 3 (42.9) | 1 (33.3) |
CR | 0 (0) | 0 (0) | 0 (0) | 0 (0) |
PR | 2 (3.4) | 0 (0) | 0 (0) | 1 (33.3) |
SD | 28 (48.3) | 11 (40.7) | 4 (57.1) | 0 (0) |
SD ≥ 6 months | 10 (17.2) | 1 (3.7) | 3 (42.9) | 0 (0) |
PD | 9 (15.5) | 3 (11.1) | 0 (0) | 2 (66.7) |
. | Cohort A . | Cohort B . | Cohort C . | |
---|---|---|---|---|
. | HR+, HER2− MBC . | HR+, HER2+ MBC . | HR+, HER2− MBC LM . | HR+, HER2+ MBC LM . |
n (%), unless otherwise stated . | (n = 58) . | (n = 27) . | (n = 7) . | (n = 3) . |
Remained on treatment | 3 (5.2) | 0 (0) | 0 (0) | 0 (0) |
Discontinued abemaciclib | 55 (94.8) | 27 (100) | 7 (100) | 3 (100) |
Reason for discontinuation | ||||
AE | 5 (8.6) | 3 (11.1) | 1 (14.3) | 0 (0) |
Death | 3 (5.2) | 1 (3.7) | 0 (0) | 0 (0) |
PD | 43 (74.1) | 23 (85.2) | 6 (85.7) | 3 (100) |
Intracranial | 31 (53.4) | 17 (63.0) | 6 (85.7) | 1 (33.3) |
Extracranial | 2 (3.4) | 2 (7.4) | 0 (0) | 2 (66.7) |
Intra- and extracranial | 10 (17.2) | 4 (14.8) | 0 (0) | 0 (0) |
Withdrawal by patient | 4 (6.9) | 0 (0) | 0 (0) | 0 (0) |
Intracranial outcomesa | ||||
ORR (CR/PR) | 3 (5.2) | 0 (0) | 0 (0) | 0 (0) |
DCR (CR/PR/SD) | 38 (65.5) | 12 (44.4) | 2 (28.6) | 1 (33.3) |
CBR (CR/PR/SD ≥ 6 months) | 14 (24.1) | 3 (11.1) | 1 (14.3) | 0 (0) |
CR | 0 (0) | 0 (0) | 0 (0) | 0 (0) |
PR | 3 (5.2) | 0 (0) | 0 (0) | 0 (0) |
SD | 35 (60.3) | 12 (44.4) | 2 (28.6) | 1 (33.3) |
SD ≥6 months | 11 (19.0) | 3 (11.1) | 1 (14.3) | 0 (0) |
PD | 15 (25.9) | 13 (48.1) | 2 (28.6) | 0 (0) |
Extracranial outcomesb | ||||
ORRc (CR/PR) | 2 (3.4) | 0 (0) | 0 (0) | 1 (33.3) |
DCR (CR/PR/SD) | 30 (51.7) | 11 (40.7) | 4 (57.1) | 1 (33.3) |
CBR (CR/PR/SD ≥ 6 months) | 12 (20.7) | 1 (3.7) | 3 (42.9) | 1 (33.3) |
CR | 0 (0) | 0 (0) | 0 (0) | 0 (0) |
PR | 2 (3.4) | 0 (0) | 0 (0) | 1 (33.3) |
SD | 28 (48.3) | 11 (40.7) | 4 (57.1) | 0 (0) |
SD ≥ 6 months | 10 (17.2) | 1 (3.7) | 3 (42.9) | 0 (0) |
PD | 9 (15.5) | 3 (11.1) | 0 (0) | 2 (66.7) |
Abbreviations: LM, leptomeningeal metastases; ORR, best confirmed overall response rate.
aComposite RANO-BM or RANO-LM was used for response criteria.
bRECIST 1.1 was used for response criteria.
cThe best overall response without confirmation is not displayed but included 6 patients with PR (10.3%).
Change in intracranial (A) and extracranial (B) tumor size and best overall response, and duration of exposure to abemaciclib (C) in cohort A (HR+, HER2− MBC). Patients who did not have a postbaseline scan (for intracranial), or who did not have an extracranial lesion or a postbaseline scan (for extracranial) are not shown on waterfall plots. *, tumor shrinkage, but PR was not confirmed at subsequent tumor assessments. o, patient received prior WBRT. +, patient received prior SRS. #, patient received concomitant ET. CR, complete response; NE, not estimable; PD, progressive disease; PR, partial response; SD, stable disease.
Change in intracranial (A) and extracranial (B) tumor size and best overall response, and duration of exposure to abemaciclib (C) in cohort A (HR+, HER2− MBC). Patients who did not have a postbaseline scan (for intracranial), or who did not have an extracranial lesion or a postbaseline scan (for extracranial) are not shown on waterfall plots. *, tumor shrinkage, but PR was not confirmed at subsequent tumor assessments. o, patient received prior WBRT. +, patient received prior SRS. #, patient received concomitant ET. CR, complete response; NE, not estimable; PD, progressive disease; PR, partial response; SD, stable disease.
With respect to extracranial disease, two PRs were observed, for a confirmed eORR of 3.4%, (95% CI, 0.0–8.1; Table 2). The eDCR and eCBR were 51.7% and 20.7%, respectively, and 59% of patients had a decrease in size of their target extracranial lesions (Fig. 2B). The median ePFS was 6.6 months (95% CI, 4.3–12.4). Median bicompartmental PFS was 4.4 months (95% CI, 2.6–5.5). Median OS was 12.5 months (95% CI, 9.3–16.4).
Cohort B (HR+, HER2+)
At data cutoff, all patients (n = 27) had discontinued abemaciclib, 85.2% because of PD (63.0% intracranial; 7.4% extracranial; 14.8% intra- and extracranial; Table 2). There were no confirmed intracranial responses, so cohort B was closed at the end of stage I. A best response of iSD occurred in 12 patients, resulting in a 44.4% iDCR (95% CI, 25.7%–63.2%). The iSD lasted ≥6 months in 3 patients, for an iCBR of 11.1% (95% CI, 0.0%–23.0%); 1 of these 3 patients was taking concurrent trastuzumab. A decrease in size of target intracranial lesions was observed for 21.7% of patients (Supplementary Fig. S1A). The median iPFS was 2.7 months (95% CI, 1.4–4.0). Median OS was 10.1 months (95% CI, 4.2–14.3).
Consistent with intracranial observations, no patients had an eCR or ePR (Table 2; Supplementary Fig. S1B). The median ePFS was 7.3 months (95% CI, 3.3–13.3) and median bicompartmental PFS was 2.1 months (95% CI, 1.4–3.3).
Cohort C (LM)
All patients in cohort C had clinical features with abnormal MRI findings and 2 (20%) had CSF cytology data. At data cutoff, all patients (n = 7 with HR+, HER2− MBC and n = 3 with HR+, HER2+ MBC) had discontinued abemaciclib. Discontinuation due to PD was most common, occurring in 90% of all patients. Discontinuation due to ePD occurred in 66.7% of patients with HR+, HER2+ MBC, and iPD occurred in 85.7% of patients with HR+, HER2− MBC (Table 2). One patient among 7 with HR+, HER2− MBC had a confirmed parenchymal CR per RANO-BM, and a best overall LM response of SD per RANO-LM. No intracranial responses (partial or complete) were observed in 3 patients with HER2+ disease. A best response of SD occurred in 1 patient with HR+, HER2+ MBC (Table 2).
Among patients with HR+, HER2− MBC, median bicompartmental PFS was 5.9 months (95% CI, 0.7–8.6) and median OS was 8.4 months (95% CI, 3.3–23.5).
Cohort D (surgical)
For 9 surgical patients, reasons for discontinuation included iPD (n = 5), iPD and ePD (n = 1), death (n = 2, causes of death listed as primary study disease and sepsis not related to study treatment), and noncompliance (n = 1). One (11.1%) patient received a prior CDK4 and CDK6 inhibitor. The median bicompartmental PFS was 4.3 months (95% CI, 0.5–8.4). Median OS was 31.7 months (95% CI, 0.5–not estimable).
Exploratory outcomes
To assess the impact of any prior or concomitant treatment, exploratory subgroup analyses were performed among 52 cohort A patients who were iORR evaluable, defined as patients who received ≥1 abemaciclib dose, with ≥1 measurable brain lesion at enrollment, and for whom ≥1 postbaseline intracranial disease overall response assessment was available. Patients without prior SRS had a 28.1% iCBR, compared with 20.0% among those with prior SRS. Patients with no prior WBRT had a 30.8% iCBR, versus 19.2% in those with prior WBRT. Patients with no prior SRS or WBRT compared with patients with both had 28.2% and 15.4% iCBRs, respectively (Supplementary Table S1).
Patients taking concomitant ET, compared with those who did not, had 35.7% and 21.1% iCBRs, respectively (Supplementary Table S1). Patients taking concomitant corticosteroids, compared with those who did not had similar iCBRs (25.8% and 23.8%, respectively).
Pharmacokinetics
In cohort D, 8 patients had brain metastases tissue, CSF, and plasma pharmacokinetics evaluated simultaneously. Patients in cohorts A, B, and C contributed further pharmacokinetic data, with simultaneous CSF and plasma from 14 patients, and plasma alone from 68 patients (Supplementary Table S2). Time-matched abemaciclib concentrations in brain metastases tissue (Fig. 3) demonstrated abemaciclib and its active metabolites penetrated the BBB, with an average ratio between unbound brain metastases tissue and unbound plasma concentrations (Kp,uu) of 5.6 (range, 0.6–14.0). In all patients, unbound brain metastases tissue concentrations of total active analytes exceeded the historic in vitro IC50 for both targets, with concentrations achieving an average of 96- and 19-fold above CDK4 and CDK6 IC50 values, respectively (30). Similarly, average concentrations in CSF exceeded CDK4 and CDK6 IC50 values by 21- and 4.3-fold, respectively. Plasma and CSF concentrations were approximately equivalent (Fig. 3), and comparable with previous observations in patients with glioblastoma (22). Plasma concentrations were consistent with efficacious concentrations in preclinical xenograft and clinical studies in patients with HR+ HER2− MBC (17, 19–21).
Steady-state brain metastases tissue, plasma, and CSF concentrations for total unbound active analytes of abemaciclib in the surgical cohort, cohort D. Total unbound active analytes were calculated by summating unbound abemaciclib, M2, and M20 concentrations at each timepoint: |$Total\ {\rm{ }}Unbound\ {\rm{ }}Active\ {\rm{ }}Analytes{\rm{ }} = {\rm{ }}abemaciclib{\rm{ }}*{\rm{ }}f{u_{x,\, abemaciclib}} + {\rm{ }}M2{\rm{ }}*{\rm{ }}f{u_{x,M2}} + {\rm{ }}M20{\rm{ }}*{\rm{ }}f{u_{x,M20,}}$| where, x is either plasma (p), brain metastases tissue (b), or CSF (csf). fup,abemaciclib = 0.0557, fup,M2 = 0.0814, and fup,M20 = 0.0206; fub,abemaciclib = 0.012, fub,M2 = 0.009, and fub,M20 = 0.021; fucsf was assumed to be 1 for each analyte. IC50 (30). CDK4, cyclin-dependent kinase 4; CDK6, cyclin-dependent kinase 6; CSF, cerebrospinal fluid; IC50, half maximal inhibitory concentration; M2, abemaciclib active metabolite LSN2839576; M20, abemaciclib active metabolite LSN3106726.
Steady-state brain metastases tissue, plasma, and CSF concentrations for total unbound active analytes of abemaciclib in the surgical cohort, cohort D. Total unbound active analytes were calculated by summating unbound abemaciclib, M2, and M20 concentrations at each timepoint: |$Total\ {\rm{ }}Unbound\ {\rm{ }}Active\ {\rm{ }}Analytes{\rm{ }} = {\rm{ }}abemaciclib{\rm{ }}*{\rm{ }}f{u_{x,\, abemaciclib}} + {\rm{ }}M2{\rm{ }}*{\rm{ }}f{u_{x,M2}} + {\rm{ }}M20{\rm{ }}*{\rm{ }}f{u_{x,M20,}}$| where, x is either plasma (p), brain metastases tissue (b), or CSF (csf). fup,abemaciclib = 0.0557, fup,M2 = 0.0814, and fup,M20 = 0.0206; fub,abemaciclib = 0.012, fub,M2 = 0.009, and fub,M20 = 0.021; fucsf was assumed to be 1 for each analyte. IC50 (30). CDK4, cyclin-dependent kinase 4; CDK6, cyclin-dependent kinase 6; CSF, cerebrospinal fluid; IC50, half maximal inhibitory concentration; M2, abemaciclib active metabolite LSN2839576; M20, abemaciclib active metabolite LSN3106726.
Safety and tolerability
The median treatment durations were four cycles (range, 1–49), two cycles (range, 1–11), four cycles (range, 1–14), and four cycles (range, 1–11) for cohorts A, B, C HR+ HER2−, and C HR+ HER2+, respectively (Fig. 2C; Supplementary Figs. S1C and S2C). Observed AEs were consistent with the previously reported abemaciclib safety profile (Table 3; refs. 17–20). No additional safety signals were uncovered in this population of patients with brain metastases or LM. In cohort A, the most common any grade AEs were diarrhea (77.6%), fatigue (48.3%), and nausea (44.8%). Grade ≥3 AEs occurring in ≥10% of patients included diarrhea (17.2%) and neutropenia (15.5%).
Treatment emergent AEs, ≥grade 3, occurring in ≥5% of all patients, by decreasing frequency in patients with HR+, HER2− MBC.
. | Cohort A . | Cohort B . | Cohort C . | Cohort D . | |
---|---|---|---|---|---|
. | HR+, HER2−MBC . | HR+, HER2+ MBC . | HR+, HER2− MBC LM . | HR+, HER2+ MBC LM . | Surgical resection . |
n (%) . | (n = 58) . | (n = 27) . | (n = 7) . | (n = 3) . | (n = 9) . |
Any | 37 (63.8) | 19 (70.4) | 4 (57.1) | 1 (33.3) | 8 (88.9) |
Diarrhea | 10 (17.2) | 4 (14.8) | 0 (0) | 1 (33.3) | 1 (11.1) |
Neutropenia | 9 (15.5) | 2 (7.4) | 1 (14.3) | 0 (0) | 3 (33.3) |
Hypokalemia | 5 (8.6) | 2 (7.4) | 1 (14.3) | 0 (0) | 2 (22.2) |
Lymphopenia | 4 (6.9) | 3 (11.1) | 0 (0) | 0 (0) | 1 (11.1) |
Thrombocytopenia | 4 (6.9) | 2 (7.4) | 0 (0) | 0 (0) | 0 (0) |
Vomiting | 3 (5.2) | 0 (0) | 2 (28.6) | 0 (0) | 0 (0) |
Alanine aminotransferase increased | 2 (3.4) | 2 (7.4) | 0 (0) | 0 (0) | 1 (11.1) |
Fatigue | 1 (1.7) | 3 (11.1) | 1 (14.3) | 1 (33.3) | 0 (0) |
. | Cohort A . | Cohort B . | Cohort C . | Cohort D . | |
---|---|---|---|---|---|
. | HR+, HER2−MBC . | HR+, HER2+ MBC . | HR+, HER2− MBC LM . | HR+, HER2+ MBC LM . | Surgical resection . |
n (%) . | (n = 58) . | (n = 27) . | (n = 7) . | (n = 3) . | (n = 9) . |
Any | 37 (63.8) | 19 (70.4) | 4 (57.1) | 1 (33.3) | 8 (88.9) |
Diarrhea | 10 (17.2) | 4 (14.8) | 0 (0) | 1 (33.3) | 1 (11.1) |
Neutropenia | 9 (15.5) | 2 (7.4) | 1 (14.3) | 0 (0) | 3 (33.3) |
Hypokalemia | 5 (8.6) | 2 (7.4) | 1 (14.3) | 0 (0) | 2 (22.2) |
Lymphopenia | 4 (6.9) | 3 (11.1) | 0 (0) | 0 (0) | 1 (11.1) |
Thrombocytopenia | 4 (6.9) | 2 (7.4) | 0 (0) | 0 (0) | 0 (0) |
Vomiting | 3 (5.2) | 0 (0) | 2 (28.6) | 0 (0) | 0 (0) |
Alanine aminotransferase increased | 2 (3.4) | 2 (7.4) | 0 (0) | 0 (0) | 1 (11.1) |
Fatigue | 1 (1.7) | 3 (11.1) | 1 (14.3) | 1 (33.3) | 0 (0) |
Abbreviation: LM, leptomeningeal metastases.
Dose reduction because of AEs occurred in 13 (22.4%) patients; most commonly caused by diarrhea (n = 7) and neutropenia (n = 3). Of the 13 patients who experienced a dose reduction, 4 patients experienced a second dose reduction to 100 mg twice daily. Discontinuations due to AEs that were considered drug related occurred in 5 patients (8.6%); neutropenia was the only drug-related AE leading to discontinuation in ≥1 patient (n = 2). Of the 5 patients who discontinued, one also had a dose reduction prior to discontinuation, and all five experienced a dose omission prior to discontinuation. All fatal events while on therapy or within 30 days of discontinuation were due to study disease (n = 4, 6.9%), except one case of pneumonitis in a patient with a prior history of dyspnea. A similar safety profile was observed in cohorts B–D.
Discussion
This uncontrolled, multi-cohort study evaluated intracranial abemaciclib activity in patients with brain metastases secondary to HR+ MBC. While this study did not meet its primary endpoint, iORR, abemaciclib demonstrated antitumor activity in heavily pretreated HR+, HER2− MBC patients with an iCBR of 24%, and ≥20% volumetric reductions in brain metastases target lesions in 19% of patients with HR+, HER2− MBC, relative to baseline.
There is clear evidence abemaciclib and its metabolites crossed the BBB. The ratio between unbound brain metastases tissue and unbound plasma concentrations averaged approximately 6-fold, well in excess of reported ratio in rats (0.1), although actual exposure levels in both species were at least 5-fold higher than in vitro CDK4 IC50 (21). These clinical results are consistent with preclinical findings (21) that abemaciclib crossed the BBB and reached unbound levels expected to produce CDK4 and CDK6 inhibition and cell-cycle arrest. Steady-state plasma and CSF exposures achieved were consistent with those associated with reductions in RB protein and DNA topoisomerase II alpha G1 arrest in xenograft models (21, 22).
The cohort A patient population enrolled in this study was heavily pretreated, with multiple lines of prior therapy for advanced disease [systemic therapy median, 3 (range, 1–10)], including chemotherapy [median, 2 (range, 1–6)], SRS (n = 20, 34.5%), or WBRT (n = 27, 46.6%). Median bicompartmental PFS in cohort A (HR+, HER2−) was 4.4 months and median OS was 12.5 months. These compare favorably with patients in MONARCH 1 (20), where median PFS was 6.0 months (95% CI, 4.2–7.5) and median OS was 17.7 months (95% CI, 16.0–not reached); of note, MONARCH 1 patients were also heavily pretreated [median of 3 (range, 1–8) prior lines of systemic therapy in the metastatic setting; ref. 20]. Patients in MONARCH 1, a phase II single-arm open-label study, were women with HR+, HER2− MBC without a history of CNS metastases who progressed on or after prior ET and had one or two chemotherapy regimens in the prior metastatic setting (20). While selection criteria in MONARCH 1 were generally similar to this study, fully understanding and interpreting these comparisons are difficult given the more strict selection criteria in this study. The observed 3.4% eORR in cohort A differs from MONARCH 1, which demonstrated a 19.7% ORR (20). In this study, most patients discontinued because of iPD, which may explain the lower eORR. Another possibility, especially given the similar iORR and eORR, is acquired resistance mechanisms or competing comorbidities.
In exploratory subgroup analyses, data suggested patients with no prior radiotherapy and those taking concurrent ET had a numerically greater effect from abemaciclib. One potential explanation for this association with ET is that ET may induce G1-phase arrest (31), similar to abemaciclib (32), and the combination of both leads to enhanced G1-phase arrest and greater efficacy. Although this post hoc analysis is based on small sample sizes and is exploratory in nature, it is hypothesis generating for further evaluation of abemaciclib, in patients with MBC and secondary brain metastases who are less pretreated, and in combination with other molecules that can cross the BBB. For instance, patients with breast cancer whose disease harbors HER2− overexpression are at a higher risk for brain metastases (33). Development of anti-HER2–targeted therapies that can cross the BBB has led to a substantial survival gain in these patients (34). There was no significant activity seen in patients with HR+, HER2+ MBC examined here. However, no patients received concomitant HER2-targeted tyrosine kinase inhibitors with known CNS penetration (10–14). Abemaciclib plus fulvestrant and trastuzumab has demonstrated clinical activity and PFS prolongation in heavily pretreated HR+, HER2+ MBC patients, compared with trastuzumab plus chemotherapy (35). Future studies exploring combinations of abemaciclib with ET and brain-penetrating HER2-directed tyrosine kinase inhibitors are indicated in patients with HR+, HER2+ brain metastases, including assessing tolerability of the combination, given potential overlapping toxicities, such as gastrointestinal AEs.
Patients with HR+, HER2− with LM experienced a median PFS of 5.9 months and median OS of 8.4 months. This compares favorably with historical median OS estimates of 3.5–4.4 months for patients with LM from breast cancer (5). Two of the patients in the LM cohort, both of whom had HR+, HER2− disease, were considered long-term survivors, 1 patient surviving for more than a year, and the other patient for nearly 2 years.
Abemaciclib safety was consistent with prior studies (17–20) with no new safety signals in patients with brain metastases or LM. The most common AE was diarrhea, which can be well managed with antidiarrheal medication at the first sign of loose stool, and subsequent, as-needed, dose adjustments (36).
Some limitations should be considered when interpreting this study. Results from this relatively small, uncontrolled study of heavily and heterogeneously pretreated patients with inherent or acquired therapy-resistant disease are difficult to interpret due to the small number of patients, and without an appropriate control arm. Particularly, it is difficult to determine the clinical relevance of outcomes such as iSD and iCBR. This study assessed time-to-worsening of patient-reported quality of life with the MD Anderson Symptom Inventory and neurocognitive function with Trail Making Tests A and B. There were no consistent differences between CR/PR and SD patients. The multi-cohort design included a mix of HR+ MBC subtypes, making it difficult to make comparisons with more homogenous populations.
Although this study did not meet its primary endpoint, iORR, abemaciclib was associated with an iCBR of 24% in heavily pretreated HR+, HER2− MBC patients with secondary brain metastases. To our knowledge, this study is the first to demonstrate pharmacologically relevant concentrations of abemaciclib and its active metabolites in brain metastases and CSF in patients with brain metastases secondary to MBC. Furthermore, in an exploratory cohort of patients with LM, abemaciclib treatment was associated with disease control and OS longer than expected, compared with historical controls. Given treatment challenges among patients with MBC with CNS involvement (3–7, 16), these results indicate further evaluations, including testing novel abemaciclib-based combinations, are warranted.
Disclosure of Potential Conflicts of Interest
S.M. Tolaney reports grants, personal fees, and other from Eli Lilly and Company (travel expense reimbursement, advisory board) during the conduct of the study; and reports personal fees from AstraZeneca (travel expense reimbursement, advisory board), Merck (travel expense reimbursement, advisory board), Nektar (travel expense reimbursement, advisory board), Novartis (travel expense reimbursement, advisory board), Pfizer (travel expense reimbursement, advisory board), Genentech/Roche (travel expense reimbursement, advisory board), Immunomedics (travel expense reimbursement, advisory board), Bristol-Myers Squibb (advisory board), Eisai (travel expense reimbursement, advisory board), and NanoString (travel expense reimbursement, advisory board), grants from Exelixis and Cyclacel, personal fees and other from Puma (travel expense reimbursement, advisory board) and Sanofi (advisory board), and personal fees from Celldex (travel expense reimbursement, advisory board), Paxman (advisory board/consultant), and Seattle Genetics (steering committee) outside the submitted work. S. Sahebjam reports grants from Merck (covering cost of investigator-initiated clinical trial), Bristol Myers Squibb (covering cost of investigator-initiated clinical trial), and Brooklyn ImmunoTherapeutics (covering cost of investigator-initiated clinical trial), personal fees from Merck (advisory board), other from Eli Lilly [covering the cost of trip to present data (melanoma and lung cancer cohorts of study) at annual ESMO meeting], and Boehringer-Ingelheim (protocol advisory group) outside the submitted work, and has stock ownership for AbbVie. E. Le Rhun reports personal fees from Tocagen (advisory board) and Daiichi Sankyo (advisory board) and personal fees and non-financial support from AbbVie (advisory board, travel) outside the submitted work. T. Bachelot reports grants, personal fees, and non-financial support from Roche, AstraZeneca, Pfizer, and Novartis, reports grants and personal fees from Lilly, and personal fees from Seattle Genetics outside the submitted work. P. Kabos reports grants from Eli Lilly (clinical research), Pfizer (clinical research), Sanofi (clinical research), AstraZeneca (clinical research), Radius Health (clinical and translational research), and Genentech (clinical research) during the conduct of the study. A. Awada reports personal fees from Lilly (advisory board) during the conduct of the study; and reports personal fees from Pfizer (advisory board + travel grant), Genomic Health (advisory board), Bayer (advisory board), Leo Pharma (advisory board), Merck (advisory board), Daiichi Sankyo (advisory board), and Seattle Genetics (advisory board), grants from Roche and BMS, and personal fees from ESAI (advisory board) outside the submitted work. P. Conte reports grants from Merck KGA, grants and personal fees from Roche and Novartis, personal fees and other from Lilly (educational activities) and AstraZeneca (educational activities), and personal fees and non-financial support from BMS (drug supply) during the conduct of the study. V. Diéras reports personal fees from Eli Lilly (advisory boards and lectures) during the conduct of the study; and reports personal fees from Pfizer (advisory boards and lectures), Novartis (advisory boards and lectures), AbbVie (advisory boards), AstraZeneca (advisory boards and lectures), MSD (advisory boards and lectures), Daiichi Sankyo (advisory boards and lectures), Seattle Genetics (advisory boards), and Roche (advisory boards and lectures) outside the submitted work. N.U. Lin reports grants from Genentech, Merck, and Pfizer, grants and personal fees from Seattle Genetics, and personal fees from Puma, Denali Therapeutics, California Institute for Regenerative Medicine, and Daiichi Sankyo outside the submitted work. M. Bear reports shareholdings from Eli Lilly and Company (employment) during the conduct of the study and outside the submitted work, and reports employment with Eli Lilly and Company. S.C. Chapman reports personal fees from Eli Lilly and Company outside the submitted work, and shareholdings in Eli Lilly and Company. Z. Yang reports personal fees from Eli Lilly and Company (employment) during the conduct of the study. C.K. Anders reports grants from Lilly (clinical trial funding) during the conduct of the study; and reports grants from PUMA (clinical trial funding), Merck (clinical trial funding), Seattle Genetics (clinical trial funding and consulting), Genentech (DSMC), Nektar (clinical trial funding), and Tesaro (clinical trial and preclinical funding), and other from Eisai (educational program), IPSEN (advisory board), AstraZeneca (advisory board), Novartis (consulting), UpToDate (royalties), and Jones and Bartlett (royalties) outside the submitted work. No potential conflicts of interest were disclosed by the other authors.
Authors' Contributions
S.M. Tolaney: Conceptualization, resources, data curation, formal analysis, supervision, methodology, writing-original draft, writing-review and editing. S. Sahebjam: Data curation, methodology, writing-review and editing. E. Le Rhun: Data curation, methodology, writing-review and editing. T. Bachelot: Data curation, writing-review and editing. P. Kabos: Data curation, methodology, writing-review and editing. A. Awada: Data curation, methodology, writing-review and editing. D. Yardley: Data curation, formal analysis, methodology, writing-review and editing. A. Chan: Data curation, methodology, writing-review and editing. P. Conte: Data curation, methodology, writing-review and editing. V. Diéras: Data curation, writing-review and editing. N.U. Lin: Conceptualization, data curation, methodology, writing-review and editing. M. Bear: Data curation, methodology, writing-review and editing. S.C. Chapman: Data curation, methodology, writing-original draft, writing-review and editing. Z. Yang: Data curation, formal analysis, writing-original draft, writing-review and editing. Y. Chen: Formal analysis, writing-original draft, writing-review and editing. C.K. Anders: Data curation, methodology, writing-review and editing.
Acknowledgments
This work was funded by Eli Lilly and Company. Eli Lilly and Company contracted with Syneos Health for writing and editing support from Andrea Metti, PhD, MPH; Angela Lorio, ELS; and Cynthia Abbott.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.