In vitro cytotoxicity assays are powerful tools for initial proof of concept and candidate ranking/selection. IC50 values (i.e., concentrations required for 50% growth inhibition) are routinely used as an indicator of potency. However, because of the intrinsic simplicity of the system, in vitro assays do not always predict the magnitude of in vivo responses; in many cases, they lead to underestimation of the in vivo efficacy for ADCs. In studies developing prostate stem cell antigen (PSCA) ADC for prostate cancer therapy, Ross and colleagues compared in vitro cytotoxicity of maytansinoid-conjugated Abs with their in vivo tumor growth inhibition (141). In vitro cytotoxicity assays in cell lines expressing different levels of PSCA showed that the IC50 values increased as the expression level decreased. In some cases, ADCs showed minimal activity in low-expressing lines. However, significant activity was observed in a xenograft model derived from the same cell lines. Both the duration of drug exposure to the tumor and the dose level of the ADC may be responsible for this phenomenon. Similar findings were reported in the colorectal cancer target EphB2, wherein Mao and coworkers related the in vitro binding, internalization, and cytotoxicity to in vivo tumor suppression (142). In the case of EphB2, a 100-fold difference in potency in vitro led to a <5-fold difference in vivo. In these cases, in vitro efficacy is more directly related to expression level, reflecting that the number of binding sites per cell is likely a major determinant in the corresponding cell-killing response. In contrast, in vivo efficacy of ADCs is complicated by multiple other factors including differences in PK, drug exposure, and metabolic pathways. As such, the receptor copy number or expression level is not always the limiting factor for in vivo efficacy.
Evaluation of CD79 as an NHL target provides another interesting example of an IVIVC derived from an ADC. CD79 is a covalent heterodimer consisting of two subunits designated CD79a and CD79b. Both anti-CD79a and anti-CD79b ADCs have been explored using an in vitro cell-killing assay and an in vivo xenograft model (143). The relative efficacies of the ADCs did not correlate with the affinities of the Abs. An anti-CD79a Ab having higher affinity translated to inferior in vivo efficacy relative to CD79b. Anti-CD79b was internalized and specifically targeted to the lysosome-like MHC class II-enriched compartment (i.e., MIIC) where active metabolites were released (143), suggesting that the superiority of anti-CD79b may be related to intracellular drug release mechanisms. The determinant factor for in vivo efficacy in this example is effective ADC metabolism and release of drug inside the cells instead of binding affinity.
Was this article helpful?