Portal de Periódicos da CAPES

Linker selection in antibody-maytansinoid conjugates impacts bystander killing in mouse xenograft models

2007; American Association for Cancer Research; Volume: 6; Linguagem: Inglês

1538-8514

Hans K. Erickson, Michele Mayo, Wayne C. Widdison, Charlene A. Audette, Yelena Kovtun, Ravi Chari, Robert Lutz, Rajeeva Singh,

Radiopharmaceutical Chemistry and Applications

A86 Three antibody-maytansinoid conjugates (AMCs), huC242-DM4, huN901-DM1, and trastuzumab-DM1, are currently being evaluated in phase II clinical trials for the treatment of cancer. All three conjugates consist of 3-4 maytansinoid (DMx) molecules tethered to a humanized monoclonal antibody (huMab) via a stable linker. HuN901-DM1 and huC242-DM4 both contain disulfide links, while trastuzumab-DM1 contains a thioether link. The optimal linker for an AMC is established by preparing and testing conjugates with different linkers in preclinical animal models. The conjugate with the greatest activity in human xenograft tumor models in SCID mice and the widest therapeutic window - as defined as the largest difference between the minimal effective dose and the maximal tolerated dose - is then selected. Interestingly, it has been found with several AMCs, including huC242-DM4 and huN901-DM1, that the type of linker (disulfide or thioether) has little impact on the in vitro activity against antigen-positive cells, although in several xenograft models, the disulfide-linked conjugates to many (but not all) antigen targets display considerably more activity than thioether-linked conjugates. To further investigate these observations, we set out to quantify the maytansinoid metabolites delivered to tumor xenografts in mice following administration of huC242-[ 3 H]DM1, huC242-[ 3 H]DM4 or huC242-SMCC-[ 3 H]DM1. Tumors were isolated at 8 h, 2 d, 4 d, and 7 d following a single administration of 15 mg/kg of each conjugate. Portions of tumor homogenates were solubilized, mixed with scintillation cocktail and submitted to liquid scintillation counting (LSC) to determine the percent injected dose per gram (%ID/g). A separate portion of the homogenate was extracted for maytansinoid metabolites, which were identified by LCMS and quantified by HPLC and LSC. Efficacy studies were conducted with unlabeled conjugates under the same conditions as the metabolism studies. The disulfide-linked huC242-DMx conjugates displayed more in vivo activity than the thioether-linked huC242-DM1 conjugate, despite similar in vitro potency. By isolating and characterizing the conjugate metabolites present in tumors, we show that the different activities cannot be explained by unique metabolism in vivo _the metabolites found in the tumors were identical to those found following exposure of target cells to these conjugates in vitro . The different activities also could not be explained by greater net delivery, as the concentration of active metabolites at the huC242-SMCC-DM1- treated tumors was actually found to be slightly greater than the corresponding tumors treated with the disulfide-linked conjugates. We reported previously that disulfide-linked conjugates, but not thioether-linked conjugates, can be metabolized in vivo to form cell-permeable active metabolites. Our new findings support the hypothesis that the in vivo activity of disulfide-linked conjugates is enhanced by the impact of these metabolites on neighboring cancer cells (bystander killing) and not through altered or elevated production of active metabolites.