Discovery route

Reduction of the lactone 137 to the corresponding lactol 138 necessitated use of DIBAL quenching the reaction with methanol followed by workup with aqueous potassium sodium tartarate furnished the product as a clear viscous oil that solidified on standing to a white solid. The anomeric hydroxyl group was conveniently protected by the formation of terf-butyldimethylsilyl ether by treatment with rert-butyldimeth-ylsilyl chloride, imidazole, and A,A-dimethylaminopyridine. Further activation of the anomeric position in compound 139 with trimethylsilyl bronfide followed by treatment with l-t-butyldimethysiloxy-3-butyne and n-butyUithium yielded a transicis mixture (1 1) (140) that was used without further purification. Reaction of compound 140 with tetrabutylammonium fluoride led to deprotection of the hydroxy functionality. The resulting transicis mixture of the alkynols was subjected to extensive chromatography and repeated crystallization to obtain a tran -alcohol (141) as a white crystalline solid. Further elaboration to 131 was carried out by appropriate modifications of a literature procedure.  

The discovery route was fraught with several tandem difficulties, such as expensive and difflcult-to-source reagents, elaborate proteclion-deprotection sequences, many cryogenic and capricious reactions, poor diastereoselectivity, and unfavorable waste 

The interesting characteristic stereochemical features of the molecule coupled with the inefficiency of the existing synthetic protocols prompted us to undertake route selection. An ideal synthesis would encompass the facility and felicity of reaction conditions and ready accessibility of raw materials allied with high and reproducible yields in all steps of the synthetic sequences. 

) and the diol (145) (46%, 97% e.e.) after column chromatography. The diol was converted to the corresponding (/ )-144 following a one-step procedure developed by us. 

We next focused our efforts on introduction of the hydroxyureidyl moiety by a patent noninfringing process. We tried two different protecting groups of N and O in hydroxylamine in order to have a convenient appendage for selective manipulation at N and O, to generate substituted hydroxyureidyl substitutes for future studies. 0-4-Methoxybenzylhydroxylamine hydrochloride, prepared from A-hydroxyphthalimide in 

---

The first step on transfer of the sythesis is to evaluate the discovery route, looking particularly at overall yield and purity, as well as parameters such as cost of production (cost of starting materials, solvents, labor and overhead, and disposal of waste stream), ease of removal of impurities or catalyst from products, and the degree of hazard associated with solvents, reactants, intermediates, and products. The route used in discovery is... 

After preliminary evaluation of the discovery route (scheme 1), we concluded that the overall yield of RWJ-26240 should be improved and that the use of NaCNBH. should be eliminated, since it produces a waste stream containing HCN or NaCN. Replacement of the expensive silver reagent, silver acetate, would permit significant cost reduction. The copper-catalyzed palladium coupling step would lead to palladium as a contaminant in the final product.177 Since a drug substance containing palladium would not be acceptable, this step would also have to be revised. 

The drug discovery route to rizatriptan (1) began with the preparation of 1-(4 -nitrobenzyl)-l,2,4-triazole 5 using 4-nitrobenzyl bromide (4) and 1,2,4-triazole. (Scheme 4.1). Benzylation of the sodium salt of 1,2,4-triazole prepared with NaH was not regioselective and afforded a 1.5 1 mixture of l-(4 -nitrobenzyl)-1,2,4-triazole (5) and its regioisomer, 4-(4 -nitrobenzyl)-l,2,4-triazole. The desired isomer 5 was isolated in 52% yield after silica gel chromatography. Hydrogenation... 

The drug discovery route to compound 1 started out with the expensive and poorly available boronic acid 2, which was coupled with aryl bromide 3 (Scheme 8.1). Hydrogenation of the resulting pyrrole 4 provided the racemic pyrrolidine 5. At... 

Suffice it to say that almost all operations on scale are possible if money and time are not issues of course, money rarely flows freely, and time is one of the major issues for developing drugs. Operations are almost always changed in progressing from the discovery route to scale-up. The best approach for rapid, successful scale-up is to scale down operations to the lab, and then develop processes that can be scaled up by mimicking conditions that will subsequently be encountered on scale. 

A comparative cost analysis showed that the classical resolution route (Scheme 8.2) was 12 times cheaper than the discovery route (Scheme 8.1). The classical resolution route was successfully scaled up and used to launch the product and provide the first year s market supply. However, using a final-stage resolution meant that by definition half of the synthetic materials were thrown away. When an E factor analysis [8] was performed on the pregabalin synthesis it was found that 86 kg of waste was being produced for every kilogram of the desired product, and this inspired a search for more efficient chemistries. 

With the key pyrrole 27 in hand, the team was in a position to complete the synthesis. Hydrolysis and decarboxylation of the tert-butyl ester was initially attempted using the conditions employed in the discovery route for the conversion of 14 to 15. While these reaction conditions effected the hydrolysis and decarboxylation of 27 to 28 in good yield, the formation of impurities resulting from dimerization of the pyrrole was also observed. After screening various acids, the team eventually found that side product formation could be completely suppressed using 1 M H2S04 in 3 1 Me0H/H20 at 65 °C to afford the pyrrole 28 in quantitative yield. 

All that remained at this point was the formylation of pyrrole 28 followed by condensation with 5-fluorooxindole (10). While the initial discovery route had accomplished this in two separate steps, the process team developed a one-pot procedure. Thus pyrrole 28 was added to a solution of the Vilsmeier reagent 29 in acetonitrile at room temperature (Scheme 7.5). Following the completion of the formylation reaction,... 

Through their refinements to the synthesis of 1, the Astellas process group ultimately developed a multikilogram-scale process for the production of 1, which both decreased the cost and increased the safety of the synthesis relative to earlier discovery and process routes.34 The resulting process additionally provided conivaptan HCl (1) in 56% overall yield from cyanobenzazepinone 19, representing a four-fold increase in yield relative to the first-generation process synthesis and sixfold increase in yield relative to the initial discovery route. 

In the original discovery route by Sugimoto,19,46 aldol condensation of 5,6-dimethoxy-l-indanone (25) and l-benzyl-4-piperidine-carboxaldehyde (26) in the presence of freshly prepared LDA and hexamethylphosphoramide (HMPA) in THF provided exocyclic enone 27 in 62% yield (Scheme 2). Hydrogenation with 10% Pd/C followed by salt formation with hydrochloric acid afforded donepezil hydrochloride (3) in 86% yield. 

Mathad et al.47 recently disclosed preparation of 3 following a similar process to the discovery route with only modifications at the condensation stage (Scheme 3). They achieved aldol condensation of 25 and 26 using sodium hydroxide as a base under phase transfer conditions to provide the intermediate 27 in excellent yield (88%). Overall yield of this route was 37%, an improvement over the original route. 

In addition to the discovery route,21 52 several alternate syntheses of rivastigmine have been reported. The discovery route involved synthesis of key intermediate phenol 39 by following Stedman s procedure,53 as shown in Scheme 7. This was then treated with N-ethyl-jV-methyl carbamoyl chloride (40) to afford racemic rivastigmine ( 2), which was resolved using di-p-toluoyl-D-tartrate (DTTA) to give (. -rivastigmine (2). 

Development usually starts with the discovery synthesis, a process designed to produce only small quantities, and often involving chemicals and procedures not amenable to a manufacturing process. Early development converts the discovery route to a synthetic route that does not have chemical, safety, environmental, or operational issues that would prevent it from being commercially viable. This must be done before the drug substance solid form and impurity profile become set by formulation development and toxicological studies. 

As evident above, the discovery route to varenicline is a linear, nine-step synthesis that takes advantage of symmetry. Notably, the first step of the route is the only one that involves carbon-carbon bond formation. Benzazepine 6, a key synthetic intermediate, was known in the literature, and its preparation is the primary subject of this chapter. The latter steps have proven robust and safe on scale, providing a six-step regulatory synthesis from 6 to varenicline delivered in the desired salt form. 

The discovery route utilized the pyridinium chlorochromate (PCC) oxidation of 2-cyclohexylethanol in CH2CI2 in presence of molecular sieves. It is a simple process, as the aldehyde is simply isolated by filtration of the reaction mixture through silica gel. However, this process was proven to be difficult to scale up due to difficulties of filtration of the chromium salts. Furthermore, the environmental issues created by the large amount of toxic chromium salts make this process unsuitable for large-scale synthesis. Two other processes (Scheme 6.7) were therefore developed and tested to prepare the required 2-cyclohexyl acetaldehyde at the pilot-plant scale ... 

---