Phase I clinical trial of the bispecific antibody MDX-H210 (anti-FcgammaRI x anti-HER-2/neu) in combination with Filgrastim (G-CSF) for treatment of advanced breast cancer

Abstract

A phase I study of the bispecific antibody MDX-H210 in combination with granulocyte colony-stimulating factor (G-CSF) was performed in stage IV breast carcinoma patients, overexpressing HER-2/neu. MDX-H210, constructed by crosslinking antigen binding fragments (F(ab') fragments) of monoclonal antibody (mAb) H22 to Fc gamma receptor I (FcgammaRI), and mAb 520C9 to HER-2/neu, respectively, mediates the lysis of tumour cells in vitro, and in human FcgammaRI transgenic mouse models. The proto-oncogene HER-2/neu is overexpressed in approximately 30% of breast cancer patients, and represents a promising target for antibody-based immunotherapy. Fc gamma receptor I (CD64) is an effective trigger molecule, which is expressed on monocytes/macrophages, immature dendritic cells, and G-CSF-primed polymorphonuclear cells (PMN). Patients received G-CSF (Filgrastim) for 8 consecutive days, and cohorts of three patients were treated on day 4 with escalating, single doses of MDX-H210. A total of 30 patients were included, and treatment was generally well tolerated, without reaching dose-limiting toxicity. Side effects consisted mainly of fever and short periods of chills, which were timely related to elevated plasma levels of interleukin 6 and tumour necrosis factor alpha. In the last two cohorts, MDX-H210 plasma levels exceeded 1 microg ml(-1), and on circulating myeloid cells >50% saturation of FcgammaRI was found until day 4. These effector cells were highly effective in antibody-dependent cell-mediated cytotoxicity. Immunohistochemical analyses of tumour biopsies in individual patients documented infiltration of monocytes and PMN after MDX-H210 infusion. Although the clinical course of the disease was not altered by the single dose of MDX-H210, a favourable toxicity profile--even at high doses--and remarkable biological effects were seen when combined with G-CSF. Therefore, the combination of G-CSF and MDX-H210 should be evaluated in further immunotherapeutical strategies.

Figure 1

Plasma levels of MDX-H210. Plasma levels of MDX-H210 in the different patient cohorts were determined 2, 4, 8, 24, 48, 72, and 96 h after BsAb infusion. The related pharmacokinetic parameters are described in Table 3. Data presented relate to all participating patients treated with ⩾10 mg m⁻².

Figure 2

Cell-bound MDX-H210. The total number of FcγRI expressed by circulating cells was determined after incubation with a saturating dose of MDX-H210. Patients were divided into three groups according to treatment with low (0.35–7 mg m⁻²), medium (10–30 mg m⁻²), and high (100–200 mg m⁻²) doses of MDX-H210. Cell-bound murine IgG without or with this preincubation was determined by incubation with FITC-labelled goat anti-mouse mAb. Significant amount of bound MDX-H210 was found on monocytes and PMN on day 1 in the low-dose group and days 1–4 in the medium- and high-dose group (P<0.05).

Figure 3

Effects of MDX-H210 on circulating white blood cells numbers. In all patients, ANC increased during Filgrastim application. A significant drop of ANC, monocytes, and lymphocytes was observed after infusion of MDX-H210 with a minimum 2–4 h after start of BsAb infusion. Data presented relate to all participating patients.

Figure 4

Receptor expression of PMN and monocytes during treatment with MDX-H210/Filgrastim. Receptor expression was determined by flow cytometry and expressed as RFI. Changes from days −3 to 0 are attributed to the effect of Filgrastim; further changes were noted after start of the 2 h during infusion of MDX-H210 on day 0. Data presented relate to all participating patients. ^**Significant changes (P<0.05) were seen for all tested antigens between days −3 and 0 and for all, except CD11b on monocytes between day 0 and the end of MDX-H210 infusion. No significant differences were seen between days −3 and 30.

Figure 5

ADCC of isolated PMN via MDX-H210. Isolated PMN from HD were compared with PMN from patients in their capacity to mediate cytotoxicity against SK-BR-3 breast cancer cells in the presence of 0.4 μg ml⁻¹ MDX-H210 at an E : T ratio of 80 : 1. Lysis of SK-BR-3 by PMN from patients on days 0, 1, and 4 is significantly higher compared to day −3 (before start of Filgrastim) as well as compared to healthy controls (P<0.005, indicated by ^*). Data presented (mean±s.d.) relate to a subset of 26 patients and 26 HDs.

Figure 6

Spontaneous cytotoxicity. For whole-blood ADCC, citrate anticoagulated, freshly drawn blood was added to ⁵¹Cr-labelled SK-BR-3 cells without the addition of antibody. Patients were divided into three groups according to treatment with low (0.35–7 mg m⁻²), medium (10–30 mg m⁻²), and high (100–200 mg m⁻²) doses of MDX-H210. Significant enhanced lysis was seen in patients of the high-dose group (day 1 vs day 0, P<0.05, indicated by ^*). The medium- and high-dose group showed enhanced lysis compared to the low-dose group on day 1 (#: P<0.05; ##: P<0.01). Patients treated with high doses of MDX-H210 (⩾100 mg m⁻²) showed a trend toward enhanced lysis of SK-BR-3 even 4 days after MDX-H210 infusion, although not statistically significant.

Figure 7

Phagocytosis of latex beads by PMN during treatment with MDX-H210. Phagocytosis of 1.0 μM fluorescent polystyrene beads, coated with human albumin and polyvalent human IgG, respectively, was measured by flow cytometry. Mean beads per neutrophil from a patient and an HD measured simultaneously were calculated as described in Material and Methods. The relation of beads per PMN of patient to HD is expressed. Phagocytosis of IgG-coated beads is significantly higher on days 0 and 1 (P<0.05, indicated by ^*) compared to baseline (day −3).

Figure 8

Cytokine plasma levels. Mean plasma levels of all patients are shown. Plasma levels of IL-6 and TNF-α increase after the infusion of MDX-H210 with a maximum 2 h after the start of infusion. Anti-inflammatory cytokine IL-10 is released with a similar time kinetic. G-CSF levels begin to rise after the start of Filgrastim therapy, with a further significant increase after MDX-H210. Plasma levels of sIL-2R and G-CSF were significantly different between days −3 and 0 (P<0.001). Significant differences (P<0.05) of plasma levels before (day 0) and after infusion of MDX-H2120 were seen for TNF-α (8 h and 1 day after start of MDX-H210), IL-6 (2 h and 1 day after start of MDX-H210), and sIL-2R (from 2 h up to 3 days after start of MDX-H210). No significant release of IL-1α, IL-1β, IL-2, IFN-γ, and IL-12 was seen.

Figure 9

Immunohistochemical staining of a skin biopsy. A skin biopsy taken 3 days after the administration of MDX-H210 (patient #21) was stained for macrophages and PMN. Macrophages (arrow) were stained with a CD68 antibody, and PMN (arrow) with chloroacetate esterase. A skin biopsy from the same patient taken before the start of treatment showed no relevant infiltration with macrophages and PMN.

Figure 10

Plasma levels of soluble HER-2/neu. Soluble HER-2/neu levels were measured during treatment with MDX-H210 and Filgrastim. A significant decrease of sHer-2/neu was seen between 4 h after the start of infusion and 7 days (P<0.05). Data are presented as mean±s.e.m. of all patients.