Support for third-party writing assistance for this manuscript was provided by Genentech, Inc.
We conducted a phase 1, multicenter, open-label, dose-escalation study (TDM3569g) to assess the safety, tolerability, and pharmacokinetics of single-agent trastuzumab emtansine (T-DM1) administered weekly and once every 3 weeks in patients with HER2-positive metastatic breast cancer previously treated with trastuzumab. The weekly dose results are described here.
Patients were administered escalating doses of T-DM1 weekly, starting at 1.2 mg/kg. Additional patients were enrolled at the maximum tolerated dose (MTD) to better characterize tolerability and pharmacokinetics.
Twenty-eight patients received weekly T-DM1, and the MTD was determined to be 2.4 mg/kg. In general, T-DM1 was well tolerated, requiring few dose modifications or discontinuations because of adverse events (AEs). Grade ≥3 AEs were reported in 19 patients (67.9%); treatment-related AEs occurred in 25 (89.3%) patients. Exposure to weekly T-DM1 was dose-proportional at ≥1.2 mg/kg, and accumulation of T-DM1 and total trastuzumab was observed. Objective partial tumor responses were reported in 13 (46.4%) patients; the median duration of response was 18.6 months, and the 6-month clinical benefit rate was 57.1%.
Trastuzumab emtansine (T-DM1) (Genentech, South San Francisco, CA) is an antibody-drug conjugate comprising the humanized anti-human epidermal growth factor 2 (HER2) immunoglobulin G1, trastuzumab, and DM1, a potent microtubule inhibitory maytansine derivative, linked through a stable thioether bond.1, 2 T-DM1 is unique in that its trastuzumab component has antitumor activity2 and is itself a widely used therapy for HER2-positive breast cancer.
T-DM1 has been studied in phase 1 and 2 clinical trials in patients with previously treated HER2-positive metastatic breast cancer. When administered every 3 weeks, T-DM1 at its maximum tolerated dose (MTD) of 3.6 mg/kg demonstrated antitumor activity and was well tolerated.3, 4 In a randomized phase 2 study in patients with HER2-positive metastatic breast cancer, this first-line T-DM1 regimen resulted in significantly longer progression-free survival and a favorable safety profile compared with the current standard of care, trastuzumab (Genentech, South San Francisco, CA) plus docetaxel (sanofi-aventis US LLC, Bridgewater, NJ).5
To assess the safety, tolerability, and pharmacokinetics of single-agent T-DM1, we performed a phase 1, multicenter, open-label dose-escalation study (TDM3569g). In addition to different doses, this study also evaluated 2 different T-DM1 administration schedules—weekly (qw) and once every 3 weeks (q3w)—as reports from studies of other DM1-based antibody-drug conjugates suggested that qw dosing could result in increased exposure while maintaining a tolerable safety profile.6, 7 Key results from those patients who received T-DM1 qw are described herein. The results from the patients who received T-DM1 q3w are reported elsewhere.3
The primary objectives were to assess the safety, tolerability, and pharmacokinetics of T-DM1. Secondary objectives were to assess the safety of repeated dosing, antitumor activity, and immunogenicity.
This study received approval from a local human investigations committee and the Department of Health and Human Services. Patients provided written informed consent according to federal and institutional guidelines. Eligible patients had HER2-positive (fluorescence in situ hybridization–positive or immunohistochemistry 3+), histologically documented, incurable, locally advanced, or metastatic breast cancer that had been previously treated with chemotherapy (in the metastatic setting) and trastuzumab (in any setting). Progression no later than 60 days after treatment with any prior trastuzumab-containing regimen was required, as was adequate organ function and limited prior exposure to anthracyclines (≤360 mg/m2 of doxorubicin or equivalent). A detailed description of the patient inclusion and exclusion criteria has been reported previously.3
Drug Administration and Dose Escalation
Evaluation of T-DM1 qw was initiated after the MTD with T-DM1 q3w was determined. The starting dose level for T-DM1 qw was 1.2 mg/kg, a third of the MTD of 3.6 mg/kg determined for T-DM1 q3w. Pharmacokinetic data suggested that at doses ≤1.2 mg/kg, clearance was still driven predominantly by target-mediated mechanisms. Therefore, the cumulative exposure of 3 weeks of the 1.2 mg/kg qw regimen was predicted to be less than that of the 3.6-mg/kg q3w regimen and was expected to be tolerable.
Three patients initially were to be treated in each cohort. Patients were observed for dose-limiting toxicity (DLT) during a 21-day DLT observation period. The dose escalation schema is shown in Figure 1A. The MTD was defined as the highest dose level that resulted in a DLT in no more than 1 of 6 patients. After the MTD had been determined, additional patients were to be enrolled at that dose level to better characterize tolerability and pharmacokinetics.
Adverse events (AEs) were graded according to the National Cancer Institute Common Terminology Criteria for Adverse Events, Version 3.0 (CTCAE v3), except for cardiac troponin I, which is not included in CTCAE v3. A DLT was defined as a grade ≥3 nonhematologic, nonhepatic major organ AE; grade ≥4 thrombocytopenia, neutropenia, or anemia; or grade ≥3 increases in serum bilirubin or hepatic enzymes. The details of DLT assessment are reported elsewhere.3 For this evaluation of T-DM1 qw, a DLT was also defined as an AE that led the investigator to hold treatment at either cycle 1, day 8, or both days 15 and 22.
Safety and Response Assessments
Medical histories were obtained at baseline. Physical and laboratory assessments were performed at baseline and weekly thereafter. A detailed description of baseline physical and laboratory assessments has been reported.3
Left ventricular function was assessed via echocardiography or multigated acquisition scan at baseline, and between days 15 and 21 of cycles 1 and 2, every 2 cycles through cycle 8, every 4 cycles thereafter, and at study termination. Results were reviewed locally and at a central facility; treatment decisions were based on the local review.
Tumor assessments were performed at baseline and every 2 cycles through cycle 8, every 4 cycles thereafter, and at study termination. Responses were evaluated by either physical or image-based evaluation using the Response Evaluation Criteria in Solid Tumors.8 Any objective responses were confirmed by repeat assessments after ≥4 weeks. Duration of response (DOR) was defined as the time from the initial response to disease progression or death within 30 days of the last dose of T-DM1.
Pharmacokinetic and Immunogenicity Assessments
Serum concentrations of T-DM1 (conjugated trastuzumab) and total trastuzumab (conjugated and unconjugated trastuzumab) were measured using validated enzyme-linked immunosorbent assay methodology. DM1 was assayed in plasma samples using electrospray liquid chromatography-tandem mass spectrometry.
Pharmacokinetic parameters for T-DM1, total trastuzumab, and DM1 after the first dose of T-DM1 in cycle 1 were estimated, when possible, by standard noncompartmental modeling, using WinNonlin 5.1 software (PharSight, Sunnyvale, CA) in PharSight Knowledgebase Server (PharSight, Sunnyvale, CA). At subsequent cycles, peak and trough (preinfusion) concentrations were summarized using descriptive statistics. For the cohort receiving the 2.4-mg/kg dose, the area under the serum concentration-time curve (AUC) profile for T-DM1 was also estimated after the administration of all 3 doses in cycle 1 (the AUC preinfusion to 21 days postinfusion [AUC0–21d] or the AUC on the last sampling of cycle 1 [AUClast,cycle 1]).
Serum samples were assayed for the presence of anti–T-DM1 antibodies using a bridging antibody electrochemiluminescent assay. The pharmacokinetic and immunogenicity sample collection schedule is shown in Figure 1B.
Enrollment, Disposition, and Determination of MTD
In total, 28 patients were enrolled and received treatment with T-DM1 qw. The baseline characteristics for these patients are presented in Table 1. Twenty-five patients received ≥3 doses of T-DM1, and 3 received ≤3 doses before discontinuing the study. A median of 14.5 doses (range, 1-90 doses) was given over a median treatment duration of 18.9 weeks (range, 0-102 weeks).
Table 1. Patient Baseline Demographics and Disease Characteristics
Time since metastatic diagnosis, mo, median (range)
Prior agents (adjuvant and/or metastatic setting), n, median (range)
Duration of prior trastuzumab therapy, wk, median (range)
Time since last trastuzumab therapy, wk, median (range)
Received prior lapatinib therapy, n (%)
Received prior anthracycline therapy, n (%)
Distinct sites of metastasis, n (%)
ER-positive and/or PR-positive, n (%)
Patients with measurable disease at baseline, n (%)
Sum of longest tumor diameter, cm, median (range)
Site of disease, n (%)
T-DM1 was administered at 1.2 mg/kg, 1.6 mg/kg, and 2.0 mg/kg for the first 3 cohorts. None of the 3 patients enrolled in each of these cohorts experienced a DLT. Of the 6 patients who were enrolled in the subsequent 2.4-mg/kg cohort, 1 experienced a DLT (grade 2 thrombocytopenia) that led the investigator to hold dosing on day 8 of cycle 1. In addition, 2 of 3 patients who received T-DM1 2.9 mg/kg experienced AEs preventing retreatment on day 8 of cycle 1: grade 3 thrombocytopenia and grade 3 elevated aspartate aminotransferase (AST). Therefore, the MTD for the T-DM1 qw dosing schedule was determined to be 2.4 mg/kg, for a cumulative dose of 7.2 mg/kg in a 21-day cycle; 16 patients were enrolled at that dose level.
Twenty-two patients (79%) discontinued the study prior to completing a year of study treatment, owing to progressive disease (14 patients), AEs (3 patients), patient decision (1 patient), physician decision (2 patients), and other reasons (2 patients: clinical progression and “lack of clinical benefit and requiring palliative radiation treatment”). At the end of the study, 6 patients (21%) were still receiving treatment (2 in the 1.2-mg/kg cohort, 1 in the 2.0-mg/kg cohort, and 3 in the 2.4-mg/kg cohort). Of these 6 patients, 3 continued treatment in an extension study (TDM4259g), and the other 3 had disease progression prior to enrolling in the extension study.
In general, weekly T-DM1 was well tolerated, requiring few dose modifications and/or discontinuations because of AEs. Regardless of relationship to study drug, at least 1 grade ≥3 AE was reported in 19 patients (68%); the most common of which were anemia (14%), thrombocytopenia (11%), pneumonia (11%), and increased AST (11%). At least 1 all-grade AE that was related to the study drug occurred in 25 patients (89%). Study drug-related AEs that occurred in more than 10% of patients are shown in Table 2.
Table 2. Adverse Events Related to Study Drug Occurring in >10% of the Study Population (>3 patients)a
MedDRA System Organ Class Preferred Term for Adverse Event
Two patients developed grade 4 AEs while receiving the 2.4-mg/kg dose. One of these patients developed grade 4 thrombocytopenia on day 171; it was considered to be related to the study drug but resolved without major sequelae. The other patient was diagnosed with pulmonary embolism on day 40, which was not considered to be related to the study drug.
DLTs were observed in 3 patients as described previously. Of the 12 serious AEs (SAEs) reported in 11 patients, 1 (grade 1 pneumonitis; 2.0-mg/kg dose level) was considered related to the study drug. Other reported SAEs were confusion, cellulitis, dyspnea, hepatic encephalopathy, influenza, osteomyelitis, pain, pneumonia, and pulmonary embolism. There were 4 treatment discontinuations due to AEs: 3 patients in the 2.4-mg/kg cohort and 1 in the 2.9-mg/kg cohort. One of these events (grade 3 pain) was an SAE, though it was not deemed to be related to the study drug. The other 3 events were considered to be related to the study drug (grade 3 fatigue, grade 2 dyspnea, and grade 3 thrombocytopenia [a DLT]).
Eleven patients experienced grade ≥3 laboratory abnormalities. Those in categories not represented by the AEs listed previously were increases in serum alkaline phosphatase (ALP), calcium, glucose, potassium, and uric acid, and decreases in absolute neutrophil count, and serum potassium and sodium.
Prespecified AEs of special interest included hepatic dysfunction, gastrointestinal toxicity, peripheral neuropathy, eye disorders, pulmonary disorders, and hemorrhage. Hepatic events (ie, those reported as increased AST, alanine aminotransferase [ALT], or ALP abnormal liver function test; or hepatic encephalopathy) were reported in 15 patients (54%). There were no grade 4 hepatic dysfunction events, but there were 3 grade 3 events: hepatic encephalopathy, increased AST, and increased ALT. Eye disorders were experienced by 13 patients, 2 of whom had grade 3 events (cataract, ocular surface disease, and punctuate keratitis). In addition, 5 pulmonary AEs were reported, of which 3 were grade 3 pneumonia, 1 was grade 2 pneumonitis, and 1 was grade 1 pneumonitis. All 3 grade 3 events resolved with medication. The 2 patients with pneumonitis continued to receive treatment and completed the study (7 and 14 months after the date of AE onset, respectively).
There was a low incidence of gastrointestinal toxicity, and very few patients experienced infusion reactions. No deaths were reported during the study or within 30 days after patients received their last dose of T-DM1. One death due to progressive disease occurred after study discontinuation, 39 days after the last day of treatment. No significant declines in left ventricular ejection fraction or segmental wall motion abnormalities were reported.
Pharmacokinetics and Immunogenicity Analyses
Serum T-DM1, serum total trastuzumab, and plasma DM1 concentrations are presented in Figure 2 with nominal time profiles after all 3 doses in cycle 1 for the 2.4-mg/kg qw cohort. Table 3 shows the mean serum T-DM1 pharmacokinetic parameters for each cohort after the first dose in cycle 1. Serum T-DM1 reached maximal concentration (Cmax) at the end of infusion; thereafter, concentrations decreased, with a mean terminal phase half-life ranging between 2.3 and 3.4 days. Modest accumulation was observed for T-DM1 following the 3 doses in cycle 1. Systemic exposure (Cmax and AUC) increased proportionally with dose (Table 3). Mean clearance values were similar across the dose levels, and the volume of distribution of T-DM1 approximated the physiologic serum volume. Taken together, the pharmacokinetics of T-DM1 were linear after the administration of the first dose at doses ranging from 1.2 to 2.9 mg/kg.
Table 3. Serum T-DM1 Pharmacokinetic Parameters After the First Dose of T-DM1 in Cycle 1 for Different Dose Cohorts
No. of Patients
AUCinf, day × μg/mL
AUCinf, area under the serum concentration-time curve from time zero extrapolated to infinity; Cmax, maximum serum concentration; t½, terminal half-life; T-DM1, trastuzumab emtansine; Vss, volume of distribution, steady state.
All data are presented as mean (standard deviation).
Trastuzumab was detected in the pretreatment serum samples of 13 patients (46%). When compared within the same dose group, the Cmax and AUC extrapolated to infinity values of total trastuzumab were higher than those of T-DM1 (Table 4). In addition, the elimination half-life values were longer and the clearance values were lower for total trastuzumab than for T-DM1 (see Table 4). The mean volume of distribution approximated the physiologic serum volume and was generally independent of the T-DM1 dose administered. Total trastuzumab, like T-DM1, accumulated with qw dosing.
Table 4. Pharmacokinetic Parameters for T-DM1, Total Trastuzumab, and DM1 After the Cycle 1, Day 1 Dose at the MTD (2.4 mg/kg)
AUCinf, area under the serum concentration-time curve from time zero extrapolated to infinity; Cmax, maximum serum concentration; MTD, maximum tolerated dose; NE, not evaluable; PK, pharmacokinetic; SD, standard deviation; t½, terminal half-life; T-DM1, trastuzumab emtansine; Vss, volume of distribution, steady state.
All data are presented as mean (standard deviation).
Plasma DM1 concentrations were consistently low, and, by day 7, DM1 concentrations were below the limit of detection in the majority of patients. Consistent with this observation, no accumulation of DM1 was observed with the T-DM1 qw dosing regimen and systemic exposure was consistently <10 ng/mL. By molar equivalents, systemic exposure of DM1 was more than 50-fold less than that of T-DM1.
Twenty-five patients (89%) had at least 1 postdose sample that was evaluable for anti–T-DM1 antibody screening; all samples tested negative.
Confirmed partial tumor responses were reported in 13 of the 28 patients enrolled (46%) (Figure 3). Of these, 6 occurred in the 2.4-mg/kg cohort (n = 16), 2 in the 2.0-mg/kg cohort (n = 3), 2 in the 1.6-mg/kg cohort (n = 3), and 3 in the 1.2-mg/kg cohort (n = 3). The DOR ranged from 1.4+ to 19.1+ months (where + represents a censored value). Four of the patients with objective responses had a DOR >14 months. The DOR for the other patients ranged from 1.4+ to 7.7 months. Stable disease >6 months was reported in 3 patients, and the 6-month clinical benefit rate was 57% (16 of 28 patients).
At baseline, 15 of the 16 patients in the 2.4-mg/kg dose cohort had measurable disease. Of these, 6 patients (40%) had an objective partial response, with a median DOR of 5.6 months.
In this phase 1 dose-escalation study, qw administration of T-DM1 at an MTD of 2.4 mg/kg was well tolerated and clinically active in patients with HER2-positive metastatic breast cancer who had previously progressed while receiving treatment with a trastuzumab-containing regimen.
As with the T-DM1 q3w regimen, exposure to T-DM1 qw was dose-proportional at doses of ≥1.2 mg/kg. As expected from the half-life of T-DM1 and total trastuzumab, modest accumulation of both analytes was observed with T-DM1 qw in cycle 1. However, the lack of increased T-DM1 concentration during subsequent cycles suggested that steady state was attained in cycle 1. T-DM1 has a shorter elimination half-life compared with total trastuzumab3, 9; this is thought to be a reflection of the differences in clearance mechanisms. T-DM1 is eliminated via both deconjugation of DM1 from intact antibody, as well as typical catabolic processes associated with antibodies, whereas trastuzumab is subject only to the latter.9
A comparison of T-DM1 q3w and qw exposures at MTD during cycle 1 (AUC0-21d or AUClast,cycle 1) showed that the cumulative exposure in cycle 1 with the T-DM1 qw MTD of 2.4 mg/kg was approximately twofold of that observed with the T-DM1 q3w MTD of 3.6 mg/kg, which is consistent with the relative cumulative doses administered in cycle 1 (3.6 mg/kg in the q3w schedule vs 7.2 mg/kg in the qw schedule). Interestingly, this did not appear to be associated with a significant increase in toxicity. The incidence of SAEs was low, and 4 (14%) patients discontinued treatment as a result of AEs. The AEs experienced by patients receiving the T-DM1 qw regimen were similar in nature, grade, and incidence to those experienced by patients receiving the T-DM1 q3w regimen.3 Grade ≥3 AEs were experienced by 19 of 28 (68%) patients who received the T-DM1 qw regimen and by 12 of 24 (50%) patients who received the T-DM1 q3w regimen. The incidence of AEs typical of maytansinoids (eg, sensory neuropathy, diarrhea, vomiting) was low, which is consistent with the minimal systemic exposure to DM1. This can be attributed to the unique, stable thioether linker used to conjugate DM1 to trastuzumab.1
In general, the tolerability of T-DM1 qw also appeared to be similar to that of T-DM1 q3w. For patients receiving T-DM1 qw, the median dose intensity (the total dose received divided by the total dose expected to be administered) was 82.0% compared with 99.6% for patients who received the T-DM1 q3w regimen, and the median duration of treatment was 18.9 vs 16.7 weeks, respectively.
Thrombocytopenia was a DLT for both regimens. While the mechanism of action is currently unknown, population pharmacokinetic modeling and in vitro studies suggest that altered platelet production by megakaryocytes, rather than an effect on circulating platelets, likely accounts for the observed thrombocytopenia.10, 11 In addition to thrombocytopenia, mild-to-moderate increases in hepatic aminotransferases were reported on both schedules, which is consistent with findings from studies of T-DM1 in cynomolgus monkeys (Genentech, data on file).
The qw and q3w schedules of T-DM1 both demonstrated substantial activity in patients with previously treated HER2-positive metastatic breast cancer, with clinical benefit rates of 57% and 50%, respectively. While further conclusions about the relative efficacy of these schedules cannot be drawn because of the small sample sizes involved, the T-DM1 qw dosing schedule is being evaluated further in ongoing clinical studies, including a phase 1b dose-escalation trial of a combination therapy (T-DM1, paclitaxel, and pertuzumab) in patients with HER2-positive, locally advanced, or metastatic breast cancer.12
This study was sponsored by Genentech, Inc.
CONFLICT OF INTEREST
Dr. Krop has received funding for clinical trials from Genentech. Dr. Girish is an employee of Genentech and has limited stock in the company. Dr. Yu is an employee of Genentech and holds stock in Roche. Dr. Lu is an employee of Genentech and owns stock in Roche. Dr. Holden is an employee of Genentech and holds stock in Roche, the owner of Genentech, the maker of T-DM1. Dr. Beeram, Dr. Burris, and Dr. Modi have no conflicts to declare.