Process Development toward the Continuous Manufacturing of Baloxavir Marboxil

Continuous processing offers significant advantages to the pharmaceutical industry, enhancing outcomes from research and development to final dosage forms. Often, transitioning a single unit operation from batch to continuous mode yields clear benefits. For instance, flow chemistry improves mass and heat transfer, directly enhancing safety and scalability by enabling challenging or previously unfeasible chemical reactions. Additionally, small reaction volumes facilitate rapid substrate screening and diversification during the discovery phase. However, in integrated continuous manufacturing (ICM)—where all unit operations are continuous and integrated with real-time monitoring and feedback control—the benefits can be more context-dependent. Traditionally, continuous manufacturing was deemed most suitable for high-volume products to leverage economies of scale and reduce production costs. Nonetheless, other benefits have emerged, such as reduced environmental impact, improved quality outcomes, streamlined supply chains, shorter regulatory reviews, and more.

The Centers for Disease Control and Prevention (CDC) collaborates with various partners at state and local levels to monitor influenza viruses annually. Xofluza (Baloxavir Marboxil) is an effective antiviral medication from Roche/Genentech, Inc., acquired under global licensing from Shionogi in May 2018. It inhibits the cap-dependent endonuclease activity of the influenza polymerase enzyme and is administered orally as a single dose at the onset of symptoms. During winter months, the CDC observes significant spikes in viral activity, leading to increased medical visits, hospitalizations, and deaths. For example, symptomatic illnesses from in-season disease burden estimates range from 9.3 million to 41 million (2010–11 through 2022–23), creating a highly variable market demand for the medication. Currently, ensuring a sufficient on-demand supply involves stockpiling the active pharmaceutical ingredient (API) and reserving drug product (DP) manufacturing slots. This approach can lead to mismanaged plant time, resulting in unmet patient needs or economic losses from expired surplus. Unfortunately, Baloxavir Marboxil requires a complex four-step batch synthesis from its registered starting materials (RSMs) and necessitates a 12-month lead time. Adding to the complexity, API, DP, and finished goods are decoupled in the current manufacturing process, outsourced to various contract manufacturing organizations (CMOs) across multiple locations. Consequently, Roche explored ICM as an ideal platform to address these challenges, aiming to unify manufacturing branches into a single location, reduce lead times from years to days, and enable a highly agile and flexible supply chain to meet fluctuating market demands. In early 2019, Roche and CONTINUUS Pharmaceuticals initiated a collaboration to develop an end-to-end ICM process for Baloxavir Marboxil.

Baloxavir Marboxil (S-033188) is a prodrug metabolized into its active form, Baloxavir acid (S-033447). Its structure comprises two tricyclic core fragments: a 4-pyridone (S199AL) and a fluorinated thiepane (S199AR), which are registered starting materials in the manufacturing process and follow a four-step synthetic route. The key transformation in this route is the stereoselective coupling of these components through a crystallization-induced diastereomer transformation (CIDT). In this process, fragment condensation is mediated by an amide coupling reagent and a Brønsted acid under equilibrium, preferentially crystallizing the desired diastereomeric salt. Diastereoselectivity is achieved through fundamental differences between solubility equilibria and crystallization behaviors of the salts. Shionogi encountered challenges running the CIDT directly from S199AL due to partial deprotection of the benzyl-protected alcohol, resulting in low yield, selectivity, and recovery. To overcome these issues, they first performed a protecting group interconversion to n-hexyl to improve the CIDT outcome. Subsequently, in the final two synthesis steps, the n-hexyl group is removed to form S-033447 before being etherified to the prodrug S-033188.