Phase I dose escalation pharmacokinetic assessment of intravenous humanized anti-MUC1 antibody AS1402 in patients with advanced breast cancer

MUC1 is a cell-surface glycoprotein that establishes a molecular barrier at the epithelial surface and engages in morphogenetic signal transduction. Alterations in MUC1 glycosylation accompany the development of cancer and influence cellular growth, differentiation, transformation, adhesion, invasion, and immune surveillance [1]. MUC1 is overexpressed in more than 90% of breast cancers and the majority of epithelial tumors and has prognostic value in a number of malignancies, including breast cancer [2]. MUC1 is expressed as a stable heterodimer composed of two subunits derived from a single polypeptide chain, after cleavage in the endoplasmic reticulum [3]. The MUC1 N-terminal subunit contains a variable number of 20-amino-acid tandem repeats that are modified by O-linked glycans. The MUC1 C-terminal subunit comprises a 58-amino-acid extracellular domain, a 28-amino-acid transmembrane domain, and a 72-amino-acid cytoplasmic tail. This C-terminal domain accumulates in the cytosol of transformed cells and is delivered to the nucleus and mitochondria [4]. The MUC1 cytoplasmic domain associates with β-catenin and with the p53 tumor suppressor and is subject to phosphorylation by the epidermal growth factor receptor, c-Src, and glycogen synthase kinase-3β, suggesting a role for MUC1 in the erbB receptor kinase and Wnt signaling pathways [5].

MUC1 is often highly overexpressed in breast cancer relative to normal breast epithelial cells. Recently, Wei et al. [6] demonstrated that the C-terminal fragment of MUC1 associates with the DNA-binding domain of the estrogen receptor. Such binding stabilized the estrogen receptor by reducing ubiquitination and proteasomal degradation of the estrogen receptor. MUC1 also increased recruitment of coactivators SRC1 and GRIP1 and was associated with increased ER-α-mediated transcription. Taken together, these data suggest a role for MUC1 oncoprotein in estrogen-mediated cell growth and survival of breast cancer cells.

Antibodies targeting MUC1 epitopes studied in human breast tumor biopsies bind to at least 90% of invasive breast neoplasms [7]. The overexpression of MUC1 correlates with adhesion and invasion of breast cancer cells in vitro [8]. Breast cancer patients who demonstrate MUC1 overexpression in greater than 75% of tumor cells and aberrant subcellular localization (cytoplasmic and membranous) have significantly poorer disease-free and overall survival [9]. AS1402 (formerly R1550) is a humanized IgG1κ monoclonal antibody (huHMFG1, [10]) which binds (K_d~1 nmol/L) the extracellular MUC1 peptide sequence, PDTR. These sequences are not exposed in normal cells because of full glycosylation, but aberrant glycosylation in cancer cells exposes the epitope to the antibody. AS1402 is a potent inducer of antibody-dependent cellular cytotoxicity (ADCC), specifically against MUC1-expressing tumor cells.

Snijdewint et al. [11] demonstrated ADCC elicited by AS1402 in vitro in the breast tumor cell line ZR-75-1, in three bone marrow-derived tumor cell lines from breast cancer patients (KM22, 1590, HG15), and in Chinese hamster ovary (CHO) cells transfected with the human MUC1 gene, by using peripheral blood mononuclear cells (PBMCs) from healthy donors. Cell lines that did not express human MUC1 were not susceptible to ADCC in the presence of AS1402. Very weak or no killing of the MUC1-positive target cell line with human PBMCs was noted in the absence of AS1402. These authors also demonstrated a strong reduction in specific AS1402-dependent cell killing of ZR-75-1 cells when the PBMCs were depleted of CD56⁺cells (that is, natural killer (NK) cells). Partial to complete depletion (> 50% to 90%) of either CD4⁺, CD8⁺, or CD19⁺cells (that is, T and B cells) from the PBMCs did not significantly reduce the ADCC. Moreover, the ADCC activity of AS1402 was shown to be dependent on the involvement of the Fcγ III receptor (CD16) on NK cells.

Immunotherapy for the treatment of cancer is an attractive alternative to cytotoxic chemotherapy. Since the approval in 1997 of rituximab (anti-CD20) for the therapy of non-Hodgkin lymphoma, several other antibodies have been licensed for different cancers. Many of these antibodies act by inhibiting signal transduction, and others, such as rituximab, trastuzumab (anti-HER2), and alemtuzumab (anti-CD52), act in addition through ADCC. NK cells of the body's immune system are directed to destroy antibody-targeted tumor cells. It has been reported that a natural humoral immune response to MUC1 protein in early breast cancer patients results in improved disease-free survival [12]. Interestingly, those patients with endogenous anti-MUC1 antibodies had a significantly higher probability of freedom from distant metastases, raising the possibility that the antibodies may be destroying circulating MUC1-positive tumor cells.

The objectives of this multicenter phase I study were to determine the safety and pharmacokinetics as well as the maximal tolerated dose (MTD) of AS1402 in patients with metastatic breast cancer.

These are the first data from a clinical study of a MUC1-targeting naked antibody. In total, 26 evaluable patients were recruited into the study; doses of 1 mg/kg, 3 mg/kg, 9 mg/kg, and 16 mg/kg were tested. The 16-mg/kg dose was considered to be the maximum viable dose. Repeated i.v. administration of AS1402 was well tolerated, with an MTD exceeding 16 mg/kg. The majority of drug-related AEs were NCI CTC grade 1 or 2. Infusion-associated reactions were generally CTC grade 1. No incidents of patients with positive titers for anti-human antibodies were found during the study.

Systemic exposure of AS1402 appears to be linear with respect to dose within the 1- to 16-mg/kg dose range assessed. This linear relation between exposure and dose was suggestive that the pharmacokinetic disposition of this humanized anti-MUC1 antibody is not dependent on saturable distribution or clearance processes, which might be expected if significant binding to circulating antigen occurred in vivo. Previous in vitro studies have shown that the affinity of AS1402 for its antigen in solution is 100-fold lower than that for cell-bound antigen, which provides a good rationale for the lack of binding to circulating antigen in vivo [14]. The mean terminal half-life of AS1402 was measured to be approximately 5 days, which suggests that weekly dosing of AS1402 at doses greater than 3 mg/kg would be required to achieve serum antibody concentrations at or above the 10 μg/ml observed for maximal ADCC activity in vitro. A 5-day half-life for AS1402, a humanized antibody (approximately 90% human), is comparable with the published values of 4.7 days for cetuximab (chimeric anti-EGFR, approximately 66% human) and 7.5 days for panitumumab (fully human anti-EGFR) [15]. Although the initial proposed dosing interval was 21 days, a weekly administration would be an acceptable addition to standard treatment regimens.

The best overall response was stable disease (achieved by five patients), and the median time to disease progression ranged from 39 to 44 days. Patients enrolled into this study had been heavily pretreated and had progressive disease on entry. In light of this, the fact that stable disease was observed in five patients warrants further investigation in less heavily pretreated or chemotherapy-naïve patients.

Antibodies have, in general, been most successfully applied as elements of combination regimens. An antibody against MUC1 has the attraction in this context that its target is overexpressed in around 90% of human breast cancers. Moreover, some data suggest a rationale for combination with existing breast cancer therapies, especially anti-estrogens. The MUC1 C-terminal subunit interacts with ERα, and this interaction is stimulated by 17β-estradiol (E2). Direct binding of MUC1 to the ERα DNA-binding domain stabilizes ERα by blocking its ubiquitination and degradation. Furthermore, MUC1 stimulated ERα-mediated transcription and contributed to the E2-mediated growth and survival of breast cancer cells [6].

Reports from a phase III trial of a breast cancer vaccine, Theratope, indicated that aromatase inhibitors may increase ADCC, adding to the rationale for combining anti-MUC1 antibodies with this approach to estrogen reduction. In a study reported by Braun et al. [16], tumor cells incubated with the aromatase inhibitor, formestane, became sensitized to killing by monocyte-mediated, antibody-dependent cellular cytotoxicity by an anti-MUC1 antibody. These observations led the authors to conclude that a hormone-based treatment may collaborate with antigen-specific tumor immunity to produce improved tumor control in patients with breast cancer. Results from the phase III study showed that patients receiving Theratope plus concomitant hormone therapy had a prolonged survival over those patients receiving a control vaccine plus hormone therapy [17]. Survival in these patients was also positively associated with immunoglobulin G titers to the underglycosylated mucin-associated glycoprotein, an antigen similar to that recognized by AS1402. A phase III trial combining trastuzumab with the aromatase inhibitor anastrazole found that patients treated with the combination therapy had a significantly higher response rate and progression-free survival than did patients receiving anastrazole alone [18].

Data obtained from the AS1402 phase I clinical trial, together with the role of the MUC1 oncoprotein in stabilization and activation of the estrogen receptor and the potential for aromatase inhibitors to increase ADCC, provide a clear basis for a phase II study combining an anti-MUC1 antibody with endocrine therapy. Letrozole has been shown, in three separate trials, to be superior to anastrazole in reducing circulating estrogen levels [19]. A phase II randomized, open-label, multicenter study of weekly infusions of 9 mg/kg AS1402 in combination with letrozole as first-line treatment in postmenopausal women with locally advanced or metastatic breast cancer is ongoing.