A Phase I Study of Eribulin Mesylate (E7389), a Mechanistically Novel Inhibitor of Microtubule Dynamics, in Patients with Advanced Solid Malignancies

Eribulin is a non-taxane, structurally simplified, synthetic analogue of the marine natural product halichondrin B, a large polyether macrolide derived from a marine sponge, which exerts potent anticancer effects in cell-based and animal models (1). The structure of eribulin encompasses the biologically active portion of halichondrin B (2). Eribulin inhibits microtubule dynamics via a novel mechanism of action (3, 4), which is thought to involve binding to a unique binding site on tubulin (5), resulting in the suppression of microtubule polymerization, without effects on depolymerization, together with sequestration of tubulin into nonfunctional aggregates (3). By inhibiting mitotic spindle formation, eribulin causes irreversible mitotic block (which ultimately leads to cell cycle arrest in the G₂-M phase) and apoptosis (3, 5, 6). In addition, eribulin showed potent antiproliferative activity against a broad range of human cancer cell lines including breast, prostate, leukemia, melanoma, and colorectal cancer (2). Furthermore, eribulin treatment was associated with tumor regressions and eliminations in a variety of well-established human tumor xenograft models, including those derived from breast, ovarian, melanoma, and colon cancer (2).

Based on the novel mechanism of action of eribulin, which is distinct from other known classes of tubulin-targeted agents, and its encouraging preclinical activity, it was hypothesized that eribulin may have efficacy in patients with malignancies that are resistant to other tubulin-targeted agents together with a more favorable tolerability profile. Therefore, eribulin was selected for clinical evaluation. The objectives of this phase I study were to determine the maximum tolerated dose (MTD) of eribulin mesylate, given as a 1-hour i.v. infusion on days 1, 8, and 15 of a 28-day cycle, and to characterize its toxicity profile, preliminary anticancer activity, and pharmacokinetics.

Patients enrolled in the study were at least 18 y of age and had histologically or cytologically confirmed advanced solid malignancies that had either progressed following standard therapy or for which no standard therapy existed. Patients were permitted to have received prior chemotherapy and were required to have a Karnofsky Performance Status of at least 70%, a life expectancy of at least 3 mo, and were also required to have adequate renal and hepatic function. All patients provided written informed consent before any study procedure. Patients who had received chemotherapy or investigational drug therapy within 3 wk (6 wk for nitrosoureas) before study initiation were excluded from participation as were those who required therapeutic doses of anticoagulant therapy, women who were pregnant or breast-feeding, and patients who had not successfully completed local therapy for central nervous system metastases. Additional exclusion criteria included a positive HIV test, any medically uncontrolled cardiovascular illness or receipt of an organ allograft, or a history of hypersensitivity to the study drug.

The protocol was approved by the Institutional Review Board at each participating institution. The study was conducted in accordance with the Declaration of Helsinki, and all patients gave written informed consent before treatment.

Eribulin mesylate was supplied in a 1-mL vial containing 500 μg/mL solution in ethanol/water (5:95). Each vial was diluted with 4 mL of 0.9% sodium chloride to provide a 5-mL solution containing eribulin mesylate at a concentration of 100 μg/mL. Dosage calculations were based on the eribulin mesylate salt.

Patients received a 1-h i.v. infusion of study medication on days 1, 8, and 15 of a 28-d cycle. Concomitant use of other medications or treatments was allowed with the exception of antitumor directed therapies, radiotherapy in cycle 1, and any substrates, potent inhibitors, or inducers of CYP3A4. Prophylactic use of granulocyte colony-stimulating factor, granulocyte-monocyte colony-stimulating factor, or erythropoietin was permitted subsequent to eribulin treatment beyond the first course according to the American Society of Clinical Oncology guidelines. Use of therapeutic anticoagulant therapy was not permitted except for low doses of anticoagulants used to maintain patency of indwelling catheters.

The starting dose of eribulin mesylate in this study was based on the findings of a previous phase I study in which grade 3 toxicity first appeared at 0.5 mg/m² (7). Dosing was scheduled to begin at 0.25 mg/m² to a maximum of 8 mg/m². Dose escalation rules followed the National Cancer Institute accelerated titration scheme design 4B (8). In this design, 100% dose increases were used during the initial accelerated stage of dose escalation, in which cohorts consisted of a maximum of one patient for each of the two study sites on condition that the second patient was enrolled no later than 1 wk after the first. On the observation of grade 2 toxicity, the cohorts were expanded by two patients. The accelerated stage ended and further escalation reverted to dose increases of 40% (design 1B) at the first instance of any dose-limiting toxicity (DLT) or the second instance of any grade 2 drug-related toxicity.

The nonaccelerated cohorts consisted of at least three patients, and intra-patient dose escalation was not permitted. If none of three patients experienced a DLT, three additional patients were treated at the next dose level. If a DLT was experienced by one of three patients, three additional patients (to a total of six) were treated at that dose level, and the dose was escalated if no additional DLT was observed in the expanded cohort. Dose escalation was terminated when at least two patients of a minimum of six fully evaluable patients experienced a DLT at any given dose level (8). If it was determined after discussion between the investigator and the sponsor that most of the patients recruited up to determination of MTD received three or more chemotherapy regimens before study inclusion, three additional patients previously treated with ≤2 regimens were to be treated at the MTD.

Safety, DLT, and MTD. DLT and toxicity assessments were conducted throughout the study, along with the assessment of overall adverse events. Vital signs, Karnofsky Performance Status score, clinical laboratory assessments, and complete physical examinations were done at screening; on days 1, 8, 15, and 21; and on study completion (at least 30 d after study initiation). Electrocardiograms were also done at baseline and study completion. Adverse event severity was graded using the National Cancer Institute Common Toxicity Criteria version 2 or the three-point scale of mild, moderate, or severe. Relationship to study treatment, as assessed by the investigators, was also evaluated.

A DLT was defined as (a) a nonhematologic toxicity of at least grade 3 in severity, excluding suboptimally treated nausea and vomiting, infertility, tumor flare, or tumor lysis syndrome; (b) a hematologic toxicity, which included grade 4 thrombocytopenia (defined as <25,000 platelets), grade 3 thrombocytopenia accompanied by clinically significant bleeding, grade 4 neutropenia not reversible to at least grade 3 within 5 d without the use of growth factors, febrile neutropenia, or neutropenia associated with bacteremia or sepsis; or (c) a treatment delay >14 d after which the patient had not recovered from study-related toxicities. The MTD was defined as the highest dose at which no more than one of six patients evaluable for toxicity experienced a DLT.

Pharmacokinetics. Plasma sampling for pharmacokinetic studies of eribulin was done on days 1 and 15 of cycle 1. Urine samples for pharmacokinetic analyses were obtained pre-dose on days 1, 8, and 15 and for time periods of 0 to 24, 24 to 48, and 48 to 72 h after drug administration on days 1 and 15. Blood samples were obtained by venipuncture (or by indwelling catheter in the arm contralateral to the infusion site) pre-dose and at 15, 30, 60, 65, 70, 90, 105 min, and 2, 4, 8, and 12 h after the start of the infusion on days 1 and 15. Blood sampling was done before treatment on day 8, 24 h after the start of the infusion (days 2 and 16), 48 h post-infusion (days 3 and 17), 72 h post-infusion (days 4 and 18), and 96 h post-infusion (days 5 and 19). On collection, blood samples were kept on wet ice and centrifuged within 1 h of collection at ∼2,000 × g and 4°C for 15 min to separate the plasma. Plasma samples were then stored in properly labeled polypropylene tubes at −70°C until analysis. At some study sites, plasma samples were stored at −20°C for less than a week and then shipped frozen on dry ice to the Eisai Research Institute for analysis.

Bioanalytic methods. Plasma and urine concentrations of eribulin were determined using validated liquid chromatography/mass spectrometry/mass spectrometry methods (9). Analysis of eribulin (in free-base equivalents) in human plasma and urine was done using liquid chromatography/mass spectrometry/mass spectrometry on a Micromass Quattro Ultima (Micromass Limited) and an API4000 (Applied Biosystems) triple quadrupole mass spectrometer, respectively, under positive ion mode. The separation of eribulin and the internal standard ER-076349, an analogue of eribulin (4), used liquid-liquid extraction with 90:5:5 ethyl acetate/methanol/ethanol (v/v/v) followed by reverse-phase chromatography. A Polaris C18 3-μm (2.0-mm i.d. × 30-mm length) column was used for plasma analysis, and a Betasil Phenyl 3-μm (2.1-mm i.d. × 50-mm length) column was used for urine analysis. The gradient mobile phase used for plasma assay consisted of the aqueous phase (a mixture of 13% acetonitrile in water with 0.1% formic acid) and the organic phase (a mixture of 70% acetonitrile and 30% tetrahydrofuran with 0.1% formic acid) and was pumped at 0.3 mL/min. An isocratic mobile phase was used for the urine assay, which consisted of 90% methanol and 10% water containing 0.1% formic acid, and was pumped at 0.3 mL/min. Eribulin was monitored at precursor ion m/z 730.5 and product ion m/z 712.5 in plasma, and precursor ion m/z 730.5 and product ion m/z 712.5 in urine.

The assay was validated according to the U.S. Food and Drug Administration published guidelines for bioanalytic method validation (10). The quantifiable range of the assay was from 0.2 ng/mL (plasma) or 0.4 ng/mL (urine) to 88.4 ng/mL of the free base of eribulin. The intra- and inter-day assay accuracy was <10% of the nominal values for plasma or 20% for the urine samples. The intra- and inter-day assay precision was <16% for plasma samples or <20% for urine samples.

Tumor assessment. Tumor assessments were done at screening and every 8 wk thereafter or on symptomatic evidence of disease progression. Tumor response and progression were evaluated using the international criteria proposed by the Response Evaluation Criteria in Solid Tumors Committee (11). The duration of stable disease was measured from the start of the treatment until the criteria for either disease progression or partial response were met. Duration of overall response was measured from when the time measurement criteria were met for complete or partial response (whichever was first recorded) until the first date that recurrent or progressive disease was objectively documented.

Safety and pharmacokinetic analyses were conducted for all patients who received at least one dose of study medication. Summary statistics are presented for adverse events and clinical laboratory values reported during the study period (SAS 8.2). Plasma eribulin concentration versus time data were subjected to noncompartmental analysis using WinNonlin version 4.1 (Pharsight). The amount and percent of the eribulin dose recovered in urine was also calculated.

Demographics and treatment. Thirty-two individual patients with advanced solid malignancies were treated with eribulin during this study. Primary tumor sites included colorectal (eight patients), ovary (five patients), uterus (two patients), breast (two patients), cervix (two patients), lung (two patients), and liver (two patients). The demographic and baseline disease characteristics are summarized in Table 1; patients had received a median of 2 (range, 2-13) prior chemotherapy regimens.

Patients received eribulin mesylate in five-dose cohorts at doses of 0.25, 0.5, 0.7, 1.0, and 1.4 mg/m² (Table 2). The median number of cycles received was 2 (range, 1-10).

Toxicities and MTD. A DLT of grade 3 fatigue was observed in one patient at 0.5 mg/m², resulting in further patient treatment at this dose level (Table 2). At the highest dose level (1.4 mg/m²), two patients developed grade 4 neutropenia (duration 7 days) and one of these patients also experienced grade 3 fatigue. Although not meeting the protocol definition of a DLT, three other patients at the 1.4 mg/m² dose level experienced grade 3 neutropenia leading to the omission of the planned cycle 1 day 15 dose. These events showed an inability to administer 1.4 mg/m² eribulin mesylate on the planned schedule; therefore, the MTD in this study was regarded as 1.0 mg/m². Of the nine patients who were treated at this dose level, only one of the six evaluable patients experienced a DLT (grade 3 fatigue).

The most common eribulin-related adverse events included fatigue [experienced by 53% of patients (40% grade 1/2, 13% grade 3, no grade 4)], nausea [experienced by 41% of patients (all grade 1/2)], and anorexia [experienced by 38% of patients (35% grade 1/2, 3% grade 3, no grade 4); Table 3]. Grade 3 or 4 toxicities that were considered related to study treatment included neutropenia (six patients) and fatigue (four patients). Four patients reported a total of six serious adverse events that were considered by the investigators to be eribulin related. In the 1.0 mg/m² dose group, one patient with ongoing pleural effusion experienced hypoxia and anoxic encephalopathy starting at days 14 and 16, respectively, after the last study treatment. The patient then developed grade 3 and grade 4 multiorgan failure and died 5 days later. Another patient in this cohort reported grade 3 vomiting. In the 1.4 mg/m² dose group, one patient developed febrile neutropenia and anemia and another patient experienced a catheter-related infection. Thrombocytopenia was not observed in any patient in this study. Three patients died during the study; none of the deaths were considered treatment related.

Clinical manifestations of neuropathy, including numbness, tingling, paresthesia, hypoesthesia, hyperesthesia, or sensory loss, which were considered to be possibly or probably related to study treatment, were observed in eight (25%) patients. These symptoms were assessed as mild (grade 1) for six patients and grade 2 for the remaining two patients. Of these eight patients, two had a prior history of neuropathy. However, eight other patients entered the study with a history of neuropathy and did not report any neuropathy or related symptoms during the study. The reasons for treatment discontinuation, other than progressive disease [23 (70%) patients], were adverse events unrelated to the study drug (three patients), withdrawal of consent by the patient [one patient who withdrew before receiving the study drug, and three patients who had worsening adverse events at the time of the discontinuation (neuropathy, fatigue, and vomiting)], and physician decision to discontinue therapy (three patients).

Clinical outcome. An unconfirmed partial response, which lasted 79 days (2.5 months), was observed in a woman with cervical cancer metastatic to the lungs. This patient progressed before her response was confirmed at the next tumor assessment. She had previously received treatment with cisplatin, carboplatin, and paclitaxel, to which her best response was progressive disease. Stable disease, as a best response, ranging from 39 to 234 days, was observed in 10 patients with appendiceal (one patient), breast (one patient), cervical (one patient), uterine (one patient), liver (two patients), and ovarian (four patients) cancer. Two of these patients experienced stable disease for >180 days. One patient in the 0.7 mg/m² cohort with endometrial carcinoma previously treated with doxorubicin/paclitaxel and cisplatin experienced stable disease for 219 days, and a patient in the 1.4 mg/m² cohort with ovarian adenocarcinoma previously treated with carboplatin/docetaxel was stable for 234 days. Time to progression ranged from 1 to 234 days.

Pharmacokinetics. Following a 1-hour i.v. infusion, eribulin mesylate pharmacokinetics were linear and dose-proportional over the dosing range of 0.25 to 1.4 mg/m² (Fig. 1A and B). Eribulin exhibited consistent pharmacokinetic parameter estimates between the first and third i.v. doses administered on days 1 and 15 at each dose level (Table 4). The plasma concentration-time profile exhibited a rapid distribution phase with a mean distribution half-life of ∼0.43 hours followed by a slower elimination phase with a half-life of 38.7 hours (Fig. 1C). Overall urinary excretion of eribulin was minimal with 5% to 6% of the administered dose eliminated in urine over a 72-hour period after a single dose. This suggests that urinary excretion may be a minor route of elimination for eribulin.

This phase I study was one of three dose-finding studies that have been conducted with eribulin, each assessing different dose regimens in patients with advanced solid malignancies. In the current study, the MTD of eribulin mesylate was 1.0 mg/m² when administered as a weekly 1-hour i.v. infusion. Nine patients were treated at this dose level. Neutropenia was dose-limiting in two patients at the 1.4 mg/m² dose level and led to termination of dose escalation. Other frequently reported treatment-related adverse events were fatigue, nausea, and anorexia, occurring at a severity of grade 1/2 in the majority of cases. Eribulin exhibited a low incidence of severe (grade 3/4) neurotoxicity or cumulative toxicities in the 12 patients treated for three or more cycles.

The first of the other eribulin phase I studies was conducted in 38 patients and included a rapid-titration design with real-time pharmacokinetics to guide dose escalation (7). Eribulin mesylate was administered as a weekly 1- to 2-minute i.v. bolus for 3 of 4 weeks, beginning at 0.125 mg/m² and continuing on a standard 3 × 3 dose escalation schedule until grade 2 or higher toxicities were observed. With this dosing schedule, the MTD was determined to be 1.4 mg/m²/wk, following the emergence of two DLTs at 2.0 mg/m²/wk: one grade 3 febrile neutropenia and one grade 4 neutropenia. Another phase I study of eribulin, reported elsewhere in this issue, was conducted in 21 patients, using a 1-hour infusion on day 1 of a 21-day cycle (12). Doses ranged from 0.25 to 4 mg/m² and dose escalation involved an accelerated algorithm. All three patients enrolled in the 4 mg/m² cohort developed DLTs of febrile neutropenia (one of these patients also developed grade 2 mucositis). Furthermore, when the dose was reduced from 4.0 to 2.8 mg/m², two of three patients experienced febrile neutropenia, whereas only one of seven patients treated at the 2.0 mg/m² dose experienced febrile neutropenia. Thus, the MTD with this dosing schedule was found to be 2.0 mg/m².

The tolerability profile for eribulin observed in the study reported here is also consistent with preclinical toxicity observations, which showed reversible myelosuppression in dogs given eribulin at 0.03 mg/kg/d on days 1, 5, and 9.⁶

The aqueous solubility of eribulin allows for a ready-to-use formulation that can be administered in a short period of time and does not require the use of excipients associated with toxicity or hypersensitivity. In addition, eribulin inhibits microtubule dynamics via a novel mechanism (3) by suppressing microtubule polymerization without affecting microtubule depolymerization and sequestering tubulin into nonfunctional aggregates. Eribulin also possesses a unique chemical structure compared with approved tubulin-targeted agents.

Although efficacy was not a primary objective of this dose-finding study, we observed an unconfirmed partial response lasting 2.5 months in a patient with cervical cancer, and 10 other patients with a variety of solid malignancies achieved stable disease as their best response. Two patients with previously treated tumors, one with endometrial carcinoma and the other with ovarian adenocarcinoma, exhibited stable disease for more than 6 months.

An additional objective of this study was to evaluate the pharmacokinetic profile of eribulin in adults with advanced solid tumors. The disposition of eribulin mesylate follows a linear kinetics over the dose range studied (0.25-1.4 mg/m²), as indicated by the proportional increases in C_max and area under the concentration-time curve. The plasma concentration-time profile exhibited a rapid distribution phase followed by a slow elimination phase. The linear kinetics and slow elimination reported here are consistent with observations reported in the other phase I studies (7, 12), the first of which reported triphasic elimination of eribulin and a prolonged terminal t_1/2 of 36 to 48 hours (7), with the second reporting a prolonged elimination phase (plasma half life of 34-73 hours), and a small fraction of eribulin (4-12%) excreted unchanged in urine (12). Eribulin inhibits cell growth with subnanomolar to low-nanomolar inhibitory concentrations in a wide range of established human cancer cell lines in vitro (2); therefore, this prolonged half-life may provide longer exposure to the antitumor activity of eribulin.

Overall, the manageable tolerability profile combined with the encouraging activity in this dose-finding study supports further clinical development of eribulin for the treatment of cancer. Eribulin is currently being investigated in several phase II and phase III trials involving multiple tumor types, including breast, non–small-cell lung cancer, prostate, and sarcoma.

S. Mani, commercial research grant, consultant, Eisai. C. Takimoto, commercial research support, Eisai. E. Rowinsky, commercial research support, consultant, Eisai. S. Goel, consultant, Eisai. The other authors have no conflicts of interest.

We thank Drs. Chris Takimoto and Sridhar Mani for their significant contribution to the design and interpretation of this study as well as the development of the paper; Tracey Lonergan, Ph.D., of Complete Medical Communications, who provided medical writing support funded by Eisai Corporation of North America; and the participating patients, their families, and the study investigators for their invaluable contribution to this research.