Pheast Therapeutics is using a platform-based approach to scaling manufacturing for its anti-CD24 checkpoint inhibitor, PHST001. By compressing the chemistry, manufacturing, and control (CMC) steps into 11 months, down from the more typical 18 to 24 months, the company is performing several development activities in parallel while preparing for GMP operations.
“It’s a calculated risk, and, ultimately, it’s working,” Kiren Khanduja, director, CMC at Pheast, tells GEN.
The greatest risk comes from making decisions based on limited information. Therefore, Pheast builds in the flexibility to adjust processes as information becomes available.
That includes conducting a small pilot process alongside its GMP operations. The pilot-scale work is predictive of the larger-scale GMP processes and identifies areas that need improvement, Khanduja says. Process engineers, therefore, are adjusting the GMP process based on real-time updates from the pilot process.
Some of those process adaptations involve information that’s proprietary to Lonza, the contract development and manufacturing organization (CDMO) with which Pheast is working. Others are involved in proactively managing the timeline to ensure informed decisions can be made quickly and effectively. There’s a short turnaround time, Khanduja stresses.
Process development is challenged by two particular factors. The target, CD24, “is heavily glycosylated and lacks a secondary structure, and the glycosylation pattern can really vary between cell types that express the receptor.” And, she adds, the monoclonal antibody IgG4 subtype has known challenges.
Extra assessments build-in flexibility
To account for them, “Early in process development, we assessed our molecule against our CDMO’s platform process,” Khanduja says. “At the beginning of the program, we performed a robust developability assessment to understand any specific liabilities that needed to be addressed.”
She and her colleagues also built in additional assessment phases to increase flexibility. For example, antibody manufacturing commonly includes a low pH hold step for viral inactivation. “Rather than take that for granted, we built in an assessment for that detergent-based inactivation step,” she says. They assessed the polishing step too, suspecting this molecule may need “something more selective” than usual.
This compressed parallel approach may be applied to future molecules, Khanduja speculates. “There are definitely some learnings.”
To apply this compressed approach to other programs, Khanduja recommends identifying questions early on, such as “how the CDMO developed its platform, when it was updated, and how it was de-risked.” She also advises conducting a robust developability assessment up front that determines how the molecule of interest behaves under various stress conditions.
Pheast plans to file an investigational new drug application later this year for this program and anticipates the first human dosing of PHST001 in early 2025.
