o Case Study (2): Identification of a new payload release pathway of DS-0821a using the untargeted LC-HRMS method

Introduction: In vitro study of payload release from an ADC plays a critical role in supporting lead optimization, bioanalytical development, and ADME and DDI studies of ADCs. Targeted LC-HRMS methods commonly used for metabolite profiling of small-molecule drugs based on their predicted molecular weights, fragmentation patterns, or mass defects are generally unsuitable for profiling payload-containing components. In addition, the low abundance of these components, combined with high levels of peptides from ADC hydrolysis and formulation excipients, makes their detection particularly challenging. Thus, we developed an untargeted LC-HRMS method for the detection and identification of payload-containing components formed via ADC catabolism or linker deconjugation.

Experimental: DS-8201a (ENHERTU) and its antibody component (trastuzumab) at 10 mg/mL were incubated separately with human liver S9 at pH 5.0 and 37 °C for 0, 24, and 48 hours. LC-HRMS datasets from these incubation samples were acquired. Detection of payload-containing components of DS-8201a was accomplished using an in-house, untargeted background subtraction method. Structures of payload-related species were determined by interpreting their MS/MS fragmentation data.

Results: As shown in Fig. 3A, the LC-MS profile of the DS8201a incubation in the human liver S9 did not showed software, three payload-containing components (M579, DXd, and M1154) were clearly revealed (Fig. 3E). DXd is the payload of DS-8201a. M579 and M1154 are new payload-containing components from DS-8201a, and their formation has not been reported in the literature.

Fig. 3. Detection of payload-containing components in incubation of DS-8201a with human

liver S9 by processing three LC-MS datasets using background subtraction filter

Fig. 4. Proposed payload release mechanisms of DS-8201a

Conclusion: The case study discovered a new DXd formation pathway from DS-8201a (Fig. 4), demonstrating that the BSF data processing method is capable of both sensitive and selective detection of payload-containing components in vitro, regardless of their molecular weights, mass defects, charge states, structures, or fragmentation patterns.