Simplifying Syntheses Is Always A Key Goal
A. MAUREEN ROUHI, C&EN WASHINGTON
Process R&D departments are kept busy by medicinal chemists, whose job is to make compounds by whatever means. "Syntheses coming out of medicinal chemistry are usually highly diluted and generate a lot of waste. They use exotic starting materials, chlorinated solvents, and hazardous chemistries," says Hans-Rudolf Marti, head of development at Siegfried, a Swiss supplier of pharmaceutical intermediates and active ingredients. "Our job is to squeeze a process down to the bare necessities."
"We try to make the synthesis as lean as possible," says Lukas Utiger, head of Lonza Exclusive Synthesis. A catalytic process for the intermediate called vince lactam provides an example.
Lonza makes vince lactam through a Diels-Alder cycloaddition between cyclopentadiene and methanesulfonyl cyanide (MSC). Initially, MSC was prepared separately and used in stoichiometric amounts, says John F. McGarrity, director of outsourcing R&D. But that step had low productivity.
Better productivity was achieved through a two-phase cyclic process. MSC is generated in catalytic quantities--0.1 equivalent--and reacted continually in situ with cyclopentadiene. Spontaneous hydrolysis of the intermediate formed in the methylene chloride/water system liberates the lactam and methanesulfinic acid, which is returned to a new round of MSC formation.
By not separately preparing one reagent, Lonza reduced the number of unit operations from 17 to 12 and the quantity of waste by 35%, McGarrity says. And because MSC is relatively unstable, not having to isolate it was advantageous. Lonza produces 50 metric tons of vince lactam annually. "The economic advantages realized in this step have played an important role in Lonza winning a major position in the supply chain for a reverse transcriptase inhibitor," McGarrity says.
The literature shows that vince lactam is a precursor to the side chain of the nucleoside reverse transcriptase inhibitor abacavir (Ziagen) from GlaxoSmithKline.
Similarly, at Hovione, process chemists continually look for opportunities to telescope reactions--that is, eliminate unit operations. "We will try not isolating solids, not drying them, seeing if we can use wet solids in the next step, because drying an intermediate is a bottleneck," Bill Heggie, Hovione's vice president for process chemistry, tells C&EN. He provides an example involving minocycline, a key generic product for Hovione. "As with all generic products, there is considerable pressure on the price," Heggie says.
Hovione's original route was encumbered by two bottlenecks in which solid intermediates were isolated and dried. In both cases, process chemists showed that the wet filtered solids could be moved to their respective next steps. Eliminating the drying step not only slashed production time, but it also solved other problems caused by the solid intermediates. In one step, the intermediate is heat sensitive, making it vulnerable to degradation during drying. In the other step, the intermediate corrodes normal drying equipment.
Finally, toward the end of the synthesis, a sequence of precipitation and solid isolation was avoided by extracting product into, and crystallizing directly from, the crystallization solvent. "The net results were a three-day savings in cycle time and elimination of expensive drying equipment," Heggie says. Because the intermediates no longer were isolated, much effort was expended to establish that the changes do not erode quality and yield and that the new method is robust.
Meanwhile, Diosynth has combined the advantages of classic solid- and solution-phase synthesis of peptides in a new high-speed method. According to Ralf van Dijk, Diosynth's product manager for peptides and chemicals, classic solution-phase peptide synthesis has an edge in being scalable, but each protocol is specific to a particular peptide. It also requires lots of solvent because intermediates must be isolated between amino acid additions. Classic solid-phase synthesis, on the other hand, involves a generic protocol and therefore is amenable to automation. But it is costly--because of solid supports and side-chain protections--and harder to scale up.
Combining the strengths of both methods, Diosynth has invented a new method in which the peptide is anchored in an organic phase through interactions of its hydrophobic C-terminal and side-chain-protecting groups with the solvent, usually ethyl acetate. After a new amino acid is added, the organic phase is washed with water to remove unreacted amino acid and by-products. The next amino acid is then added to the growing peptide. Further improvements will be realized with automation, van Dijk says.
The new method is highly scalable, he points out: The chemistry for 1 g is the same as for 1 kg. That's important, he emphasizes, because the impurity profile of the product does not change, whether it's gram quantities for preclinical studies or kilogram quantities for clinical trials. Van Dijk can't say by how much this method cuts the cost of goods. "But at the very least," he adds, "manufacturing efficiency and flexibility absolutely will go up."