Bortezomib/docetaxel combination therapy in patients with anthracycline-pretreated advanced/metastatic breast cancer: a phase I/II dose-escalation study

The aim of this study was to determine the dose-limiting toxicities (DLTs) and maximum tolerated dose (MTD) of bortezomib plus docetaxel in patients with anthracycline-pretreated advanced/metastatic breast cancer. Forty-eight patients received up to eight 21-day cycles of docetaxel (60–100 mg m−2 on day 1) plus bortezomib (1.0–1.5 mg m−2 on days 1, 4, 8, and 11). Pharmacodynamic and pharmacokinetic analyses were performed in a subset of patients. Five patients experienced DLTs: grade 3 bone pain (n=1) and febrile neutropenia (n=4). The MTD was bortezomib 1.5 mg m−2 plus docetaxel 75 mg m−2. All 48 patients were assessable for safety and efficacy. The most common adverse events were diarrhoea, nausea, alopecia, asthenia, and vomiting. The most common grade 3/4 toxicities were neutropenia (44%), and febrile neutropenia and diarrhoea (each 19%). Overall patient response rate was 29%. Median time to progression was 5.4 months. In patients with confirmed response, median time to response was 1.3 months and median duration of response was 3.2 months. At the MTD, response rate was 38%. Pharmacokinetic characteristics of bortezomib/docetaxel were comparable with single-agent data. Addition of docetaxel appeared not to affect bortezomib inhibition of 20S proteasome activity. Mean alpha-1 acid glycoprotein concentrations increased from baseline at nearly all time points across different bortezomib dose levels. Bortezomib plus docetaxel is an active combination for anthracycline-pretreated advanced/metastatic breast cancer. The safety profile is manageable and consistent with the side effects of the individual agents.

Anthracyclines and taxanes are widely used in treating patients with metastatic breast cancer. In anthracycline-pretreated advanced/metastatic breast cancer, docetaxel administered once every 3 weeks is more active than the same schedule of paclitaxel (Jones et al, 2005). The efficacy of docetaxel as a single agent (Nabholtz et al, 1999;Sjostrom et al, 1999;Jones et al, 2005) or in combination with cytotoxic agents (O'Shaughnessy et al, 2002;Chan et al, 2005) has been established in phase III trials in this setting. The standard dose of single-agent docetaxel is 75 -100 mg m À2 every 21 days. In combination, the dose is reduced, typically to 75 mg m À2 (Aventis Pharmaceuticals Inc, 2005). Dose reduction potentially reduces efficacy; therefore, careful identification of dosing regimens that provide acceptable safety profiles without compromising efficacy is essential.
Combining agents with unique modes of action may improve outcomes and overcome chemoresistance without significantly increasing toxicity. Docetaxel acts by disrupting the microtubular network essential for cellular functions (Bissery et al, 1995;Eisenhauer and Vermorken, 1998). Bortezomib, which is approved for treating multiple myeloma and mantle cell lymphoma patients who have received at least one prior therapy (Kane et al, 2006), acts by inhibiting the 26S proteasome, the degradative enzyme complex involved in the catabolism of numerous intracellular regulatory proteins, including NF-kB (nuclear factor-kB)-inhibitor IkBa, p53, p21, and p27 (Adams, 2002;Cusack, 2003;Lenz, 2003;Boccadoro et al, 2005). Malignancies with high concentrations of activated NF-kB, such as breast cancer, are logical targets for agents that interrupt this pathway (Orlowski and Dees, 2003). Mutations in the tumour suppressor gene p53 occur in 20 -40% of sporadic breast cancers (Osin and Lakhani, 1999) and are associated with a poor prognosis (Pharoah et al, 1999;Overgaard et al, 2000) and poor response to treatment with certain chemotherapeutic and hormonal agents (Berns et al, 2000;Kandioler-Eckersberger et al, 2000). The cyclin-dependent kinase inhibitors p21 and p27 also play important roles in breast cancer (Osin and Lakhani, 1999), supporting the investigation of bortezomib in breast malignancies.
Bortezomib has demonstrated cytotoxic activity in breast, lung, pancreatic, prostate, and head and neck tumour models in vivo Teicher et al, 1999;Shah et al, 2001;Sunwoo et al, 2001;Nawrocki et al, 2002;Williams et al, 2003;Ikezoe et al, 2004). Preliminary in vitro and in vivo studies in a range of solid tumours demonstrated an additive antitumour effect of bortezomib with standard cytotoxic agents, including docetaxel (Gumerlock et al, 2002(Gumerlock et al, , 2004Mack et al, 2003;Nawrocki et al, 2004;Farneth et al, 2005). Bortezomib 1 mg kg À1 combined with docetaxel 5 mg kg À1 was active in pancreatic xenograft models (Nawrocki et al, 2004). Interestingly, sequential administration led to greater tumour growth inhibition than with either agent alone (Gumerlock et al, 2004). Preclinical studies in prostate and lung cancer models indicated that the sequence docetaxel-bortezomib was more effective than bortezomib-docetaxel (Gumerlock et al, 2002;Farneth et al, 2005). A possible explanation is that bortezomib promotes cell cycle arrest before the M phase, consequently interfering with docetaxel-induced apoptosis (Nawrocki et al, 2004). Additionally, in vitro studies have shown that docetaxel is extensively plasma protein-bound, especially to alpha-1 acid glycoprotein (AAG), albumin, and lipoproteins (Aventis Pharmaceuticals Inc, 2005). As AAG is an acute-phase protein that may be elevated by IL (interleukin)-6, which is inhibited by bortezomib (Hideshima et al, 2001), bortezomib may decrease AAG concentrations and improve docetaxel efficacy.
In clinical studies in patients with solid tumours, bortezomib/ docetaxel combination therapy demonstrated encouraging activity, and the side effects were predictable and manageable (Meluch et al, 2005;Fanucchi et al, 2006;Lara et al, 2006;Dreicer et al, 2007). There is no evidence of a drug interaction between bortezomib and docetaxel (Messersmith et al, 2006).
Based on the known single-agent antitumour activity of bortezomib and docetaxel, and their additive efficacy in preclinical models, complementary mechanisms of action (Nawrocki et al, 2004), different toxicity profiles, and feasibility in clinical studies in various solid tumours, we conducted the current study to explore the potential of bortezomib/docetaxel combination therapy in patients with advanced/metastatic breast cancer. Our primary aim was to identify the most appropriate regimen for further evaluation.

Study objectives
The primary objective was to establish the dose-limiting toxicity (DLT) and maximum tolerated dose (MTD) of bortezomib/ docetaxel combination therapy in patients with advanced and/or metastatic breast cancer that had previously been treated with anthracyclines. Secondary objectives were to assess tumour response rate (complete response (CR), partial response (PR), or stable disease (SD)), time to response, duration of response, and time to disease progression (TTP); evaluate disposition profiles of docetaxel and bortezomib administered in combination using pharmacokinetic analyses; evaluate pharmacodynamic properties of bortezomib in combination with docetaxel using 20S proteasome inhibition assay (Lightcap et al, 2000); and investigate the possible influence of bortezomib/docetaxel combination treatment on serum concentrations of AAG.

Eligibility
All study participants provided written, informed consent. Patients who had advanced/metastatic breast cancer, had received at least one anthracycline-containing regimen, were aged X18 years, and had a Karnofsky Performance Status X80% (ECOG 0 -1) were considered eligible. Prior treatment with endocrine or biologic agents was permitted. Patients were excluded if they had previously received docetaxel or paclitaxel as adjuvant treatment within the previous year or for metastatic disease at any time; radiotherapy to 435% of bone marrow (e.g., pelvic radiation), major surgery, or chemotherapy within 4 weeks of enrollment; nitrosoureas within 6 weeks or antibody therapy within 8 weeks of enrollment; or high-dose chemotherapy and peripheral blood stem cell transplantation at any time. Patients were excluded if they had grade X1 peripheral neuropathy; abnormal laboratory values within 2 weeks of enrollment; history of severe hypersensitivity reaction to docetaxel; not recovered from all toxic effects (except alopecia) of previous chemotherapy, radiotherapy, or antibody therapy. The following were prohibited during the study treatment: any investigational agent other than bortezomib, haematopoietic growth factors during the first cycle (except for haematologic DLT, or continued erythropoietin for pre-existing anaemia), immunotherapeutic agents, steroids (except dexamethasone), chemotherapeutic agents other than docetaxel, radiotherapy, or surgery for cancer.

Study design
This prospective, phase I/II, open-label dose-escalation study was conducted at six centers in Europe (Belgium, Spain, UK) in accordance with the International Conference on Harmonisation for Good Clinical Practice. The protocol and informed consent were approved by the Institutional Review Board at each clinical site before study initiation.
Each 3-week treatment cycle consisted of docetaxel (Taxotere s ; Sanofi-Aventis, Paris, France) infusion on day 1 and bortezomib (Velcade s ; Millennium Pharmaceuticals Inc., Cambridge, MA, USA and Johnson & Johnson Pharmaceuticals, Rariman, NY, USA, Research and Development, L.L.C) injections on days 1, 4, 8, and 11. Bortezomib was administered 1 h after docetaxel. Treatment comprised eight cycles. Docetaxel was administered at doses of 60, 75, or 100 mg m À2 with standard oral dexamethasone premedication (six 8-mg doses administered the night before, morning of, immediately before, evening after, and morning and evening of day 2 after each docetaxel infusion). Bortezomib doses were 1.0, 1.3, or 1.5 mg m À2 , based on tolerability and efficacy in heavily pretreated patients with solid tumours. All patients received appropriate supportive therapy. Patients achieving a tumour response at the end of the treatment could receive further treatment at the investigator's discretion. Patients with progressive disease discontinued treatment.
Patients were evaluated for toxicities before each scheduled study drug dose. Adverse events (AEs) were graded according to the National Cancer Institute Common Toxicity Criteria (version 2.0). The following AEs during the first cycle were considered DLTs: platelet count p25 Â 10 9 l À1 ; febrile neutropenia (absolute neutrophil count [ANC]o1 Â 10 9 l À1 with temperature X38.51C); ANCo0.5 Â 10 9 l À1 for days 1 -7, or o0.2 Â 10 9 l À1 without fever for X7 days starting on or after day 8; any other grade 4 haematologic toxicity; grade 2 peripheral neuropathy; any grade 3 or 4 non-haematologic toxicity (except inadequately treated nausea, vomiting, and diarrhoea). In patients experiencing grade 4 haematologic or grade 3/4 non-haematologic toxicity, the start of the next cycle was delayed for up to 2 weeks or bortezomib therapy was interrupted for up to 2 weeks until toxicity returned to baseline or better. Treatment was re-initiated at a reduced dose (25% dose reduction of the drug considered by the investigator to have caused the toxicity). A maximum of two dose reductions for each drug was recommended. If toxicity did not resolve, the patient was excluded from further study.
Doses of both agents were reduced if patients experienced grade 2 neuropathic pain or peripheral neuropathy. In patients with either toxicity at grade 3 intensity or both toxicities at grade 2, bortezomib treatment was interrupted until resolution to grade p1, when dose was reduced, and at resolution, treatment was continued using weekly bortezomib (days 1 and 8 only). In patients experiencing grade 4 neuropathic pain or peripheral neuropathy, or both toxicities with grade 3 intensity, bortezomib and docetaxel were discontinued.
Doses were escalated in a stepwise fashion if fewer two out of three patients in a dose cohort experienced a DLT. The MTD was defined as the dose below that causing DLTs in at least two out of three patients. Up to 10 additional patients were to be enrolled at the MTD.

Study assessments
Karnofsky Performance Status assessments, physical examinations, and laboratory sampling were undertaken during screening, during the first cycle, at the end-of-therapy visit (10 days after the last dose of study drug), and at the end-of-study visit (3 weeks after the end-of-therapy visit).
Lesions were assessed by CT/MRI every 6 weeks using the Response Evaluation Criteria in Solid Tumours (Therasse et al, 2000). Response was confirmed 6 weeks later in patients with CR/PR. Efficacy was also assessed based on the expression of tumour marker CA15.3, Karnofsky Performance Status, and C-reactive protein, IL-6, and AAG concentrations. Blood samples were collected at screening, on day 1 of cycles 3, 5, and 7 (and all odd-numbered cycles thereafter in responding patients receiving further treatment), and at the end-of-study visit.
In patients participating in pharmacokinetic analysis, blood samples were taken on day 1, cycle 1, to determine the plasma concentration time profile of docetaxel and bortezomib. Samples were collected immediately before docetaxel infusion and bortezomib injection and at prespecified intervals on day 1. Samples were also taken for bortezomib pharmacokinetic analysis on day 11 of cycle 1. Whole-blood samples were taken for pharmacodynamic analysis using 20S proteasome inhibition assay (Lightcap et al, 2000) immediately before and 1 h after bortezomib dosing on days 1 and 11 of cycles 1 and 2, and on days 2 and 12 of cycle 1, in a subset of patients participating in pharmacokinetic analysis. Samples for determination of AAG concentrations were collected at screening and before docetaxel administration on day 1 of cycles 1, 2, 4, and 6.

Statistical analysis
Statistical analyses were primarily descriptive; the aim of the study was to establish the MTD of bortezomib/docetaxel combination therapy. Planned enrollment was up to 70 patients.
Safety and efficacy were evaluated in all patients who received any amount of either study drug. Patients who underwent dose reduction were analysed in the dose group in which they were initially treated. The MTD-evaluable population included all patients in the dose-escalation phase with sufficient safety assessments during cycle 1 to determine whether a DLT occurred. Patients were excluded from the MTD-evaluable population if they had discontinued during cycle 1 for reasons other than DLTs or had received alternate antineoplastic therapies during that period.

Safety
All 48 patients experienced at least 1 AE, most commonly diarrhoea, nausea, alopecia, asthenia, and vomiting (Table 3). Grade 3/4 AEs occurred in 37 patients (77%) and were considered drug-related in 34 (71%) patients. The most common grade 3/4 AEs were neutropenia, febrile neutropenia, and diarrhoea (Table 3). Eighteen (50%) out of the 36 patients who withdrew prematurely had unacceptable AEs listed as the primary reason for treatment discontinuation: 4 at the bortezomib 1.0 mg m À2 dose level, 6 at the 1.3 mg m À2 dose level, and 8 at the 1.5 mg m À2 dose level. Events leading to discontinuation included peripheral neuropathy (nine patients), neuralgia (four patients), paraesthesia and diarrhoea (two patients each), and neutropenia (one patient). Twelve patients (33%) withdrew because of progressive disease. Dose reductions were most common in the second and third cycles of treatment (19 and 15% patients, respectively). Two patients died on study, one due to cardiac tamponade 3 days after the last dose of study drug and the other from disease progression 22 days after the last dose of study drug. MTD ¼ maximum tolerated dose; DLT ¼ dose-limiting toxicity. a The MTD-evaluable population included all patients in the dose-escalation phase with sufficient safety assessments during cycle 1 to determine whether a DLT occurred. b Three patients did not complete cycle 1 of treatment, and one patient was excluded due to pre-existing peripheral neuropathy. c One patient had study drug held in cycle 1 and was not used in the determination of MTD. d Ten additional patients enrolled following determination of dose level as MTD.

Pharmacokinetics
Bortezomib plasma profiles exhibited an initial rapid distribution phase followed by a slower elimination phase, demonstrating apparent multi-exponential decay characteristics, similar to those reported previously. Bortezomib was eliminated from plasma with similar terminal half-lives across dose levels, with mean values of 15 -18 h following the first dose in patients with measurable plasma concentrations at 24 h post-dose (Table 5). Pharmacokinetic characteristics of bortezomib/docetaxel were comparable with single-agent data (Millennium Pharmaceuticals Inc.; Rosing et al, 2000).

Pharmacodynamics
In total, 33 patients provided whole-blood samples for measurement of 20S proteasome inhibition at baseline. Inhibition of proteasome activity was greatest at 1 h post-dose (46.14 -88.80%). Immediately prior to the dose on day 11, percent inhibition was 10.93 -51.18%. Percent inhibition was greater in cycle 2 than cycle 1 for all dose groups. Coadministration of docetaxel with bortezomib did not appear to affect 20S proteasome inhibition by bortezomib. The degree of inhibition observed is similar to that observed in studies of bortezomib monotherapy (Aghajanian et al, 2002;Orlowski et al, 2002;Blaney et al, 2004;Papandreou et al, 2004;Hamilton et al, 2005).
Limited blood samples were collected for IL-6 and CRP analysis. Therefore, no analyses of these data were conducted. Mean AAG concentrations increased from baseline for nearly all time points across all three bortezomib dose levels ( Figure 1).

DISCUSSION
In this phase I/II study in patients with pretreated advanced/ metastatic breast cancer, the MTD was bortezomib 1.5 mg m À2 on days 1, 4, 8, and 11, plus docetaxel 75 mg m À2 on day 1 of a 21-day cycle. The DLTs were bone pain and febrile neutropenia. The bortezomib dose was slightly higher than the approved 1.3 mg m À2 on the same schedule, but similar to that identified in phase I studies evaluating single-agent bortezomib administered on the same schedule (Aghajanian et al, 2002;Dy et al, 2005). Slightly higher doses were tolerated when bortezomib was administered weekly (Papandreou et al, 2004) or on days 1 and 4 every 14 days (Hamilton et al, 2005). In other combination dose-escalation studies in advanced non-small-cell lung cancer and other refractory solid tumours (Lara et al, 2006), MTDs were bortezomib 1.0 mg m À2 plus docetaxel 75 mg m À2 on the same schedule, and bortezomib 0.8 mg m À2 twice weekly for 2 weeks plus docetaxel 25 mg m À2 on days 1 and 8 of a 21-day cycle (Messersmith et al, 2006). The docetaxel dose in our study is consistent with that used in combination with chemotherapeutic agents in anthracyclinepretreated breast cancer (O'Shaughnessy et al, 2002;Chan et al, 2005).  3 (100) 10 (91) 15 (88) 10 (91) 4 (100) 14 (93) 12 (75) 41 (85) Time to response (days) N The safety profile of the combination was consistent with that observed in phase II studies in other solid tumours evaluating bortezomib alone (Shah et al, 2004;Maki et al, 2005) or in combination with docetaxel (Meluch et al, 2005;Fanucchi et al, 2006;Dreicer et al, 2007). The most common grade 3/4 toxicities were neutropenia, febrile neutropenia, diarrhoea, and peripheral neuropathy. Dose modifications were implemented to try and prevent evolution of neurotoxicity and worsening of existing symptoms. Nevertheless, in these heavily pretreated patients, 15 patients discontinued treatment due to neurological complaints, a known side effect of bortezomib and, to a lesser extent, docetaxel. The high incidence of neutropenia was expected given the typically high incidence of neutropenia and febrile neutropenia with singleagent docetaxel (Nabholtz et al, 1999;O'Shaughnessy et al, 2002) and the partially overlapping side-effect profiles of the two agents. Supportive therapies and strategies for side-effect management can prevent worsening of these symptoms, thereby avoiding treatment delays or discontinuations (Colson et al, 2004). Notably, bortezomib-associated neutropenia has been shown to be transient and reversible (Lonial et al, 2005); consistent with this, only one patient discontinued the study drug due to neutropenia.
Bortezomib/docetaxel combination treatment demonstrated antitumour activity. The response rate was 29% and median TTP was 5.4 months. These data are similar to those reported for docetaxel (75 mg m À2 ) in combination with either gemcitabine or capecitabine, which have been compared in a phase III trial (Chan et al, 2005). Recent studies have shown response rates of 62.5% with docetaxel plus cisplatin (Lin et al, 2007) and 50% with docetaxel plus epirubicin (Hainsworth et al, 2006). The 38% response rate at the MTD compares favourably with docetaxel and paclitaxel administered as single agents in anthracycline-pretreated metastatic breast cancer: in a recent study comparing docetaxel and paclitaxel, response rates were 32 vs 25%, and median TTP was 5.7 vs 3.6 months, respectively (Jones et al, 2005). The activity demonstrated in the present study does not appear to arise from the mechanism proposed, by which bortezomib reduces AAG concentrations through inhibition of IL-6 and consequently increases docetaxel efficacy, as mean AAG concentrations increases from baseline at nearly all time points across bortezomib dose levels. Other cellular and molecular effects of bortezomib may be involved; recently published results demonstrating its differential effects in breast cancer cells may be relevant in designing mechanism-based combination treatments (Codony-Servat et al, 2006).
Bortezomib has been studied in combination with trastuzumab in HER2-positive breast cancer cell lines (Cardoso et al, 2006). Although its value in HER2-postive cells is well established, the response rate with single-agent trastuzumab in metastatic breast cancer is o40% (Vogel et al, 2002). A preclinical study in four cell lines demonstrated that bortezomib acts synergistically to increase the effect of trastuzumab in HER2-positive cells (Cardoso et al, 2006). The potential clinical application of this combination is currently under investigation in a phase I trial. Additionally, a phase II study is underway to investigate bortezomib plus pegylated liposomal doxorubicin in patients with anthracyclinepretreated metastatic breast cancer.
The baseline characteristics of patients in the present study were typical of patients with anthracycline-pretreated advanced/metastatic breast cancer. The p53 and estrogen receptor status were not recorded but may be of interest in future trials, as in vitro studies suggest proteasome inhibition is at least partially dependent on the p53 status in breast cancer (MacLaren et al, 2001), and  (47) 538 (42) Docetaxel dose (mg m À2 ) C max (ng ml À1 ) AUC 0 -t (h ng ml À1 ) AUC 0 -N N (h ng ml À1 ) T1 2 ,z (h) T1 2 ,z 24-h (h) CL (l h À1 ) V ss (l) V z (l) 75 (n ¼ 7) 1 1620 (38) 1704 (50) (85) 1514 (55) C 0 ¼ back-extrapolated time 0 plasma drug concentration (bortezomib alone); C max ¼ observed maximum plasma drug concentration (docetaxel alone); AUC 0Àt ¼ area under plasma concentration time curve from time 0 to last time point with measurable drug concentration; AUC 0ÀN ¼ area under plasma concentration time curve extrapolated to infinity; T1 2 ,z ¼ apparent plasma half-life; T1 2 ,z (24-h) ¼ apparent plasma half-life in patients with measurable plasma concentrations at 24 h post-dose; CL ¼ total body clearance; CV ¼ coefficient of variance; V ss ¼ volume of distribution at steady state; V z ¼ apparent volume of distribution in terminal elimination phase. Mean change from baseline in alpha-1 acid glycoprotein concentrations by bortezomib dose level during the study. The number of patients available for each analysis are shown above or below each column; notably, the significant decrease from baseline seen at cycle 6, day 1, in the bortezomib 1.5 mg m À2 group arises from evaluation of data from only one patient.
estrogen-receptor-negative status plus a dysregulated IkBa/NF-kB system is associated with greater bortezomib antitumour activity (Tapia et al, 2005).
In conclusion, the study demonstrated that bortezomib and docetaxel combination therapy is feasible, tolerable, and active in anthracycline-pretreated metastatic breast cancer.