Advances in drug potency and tailored therapeutics are promoting pharmaceutical manufacturing to transition from a traditional batch paradigm to more flexible continuous processing. Here we report the development of a multistep continuous-flow CGMP (current good manufacturing practices) process that produced 24 kilograms of prexasertib monolactate monohydrate suitable for use in human clinical trials. Eight continuous unit operations were conducted to produce the target at roughly 3 kilograms per day using small continuous reactors, extractors, evaporators, crystallizers, and filters in laboratory fume hoods. Success was enabled by advances in chemistry, engineering, analytical science, process modeling, and equipment design. Substantial technical and business drivers were identified, which merited the continuous process. The continuous process afforded improved performance and safety relative to batch processes and also improved containment of a highly potent compound.

Continuous crystallization has gained interest in the pharmaceutical sector as part of the drive toward the transition from exclusive batch manufacturing to integrated continuous manufacturing in this industry. As a result, the design and operation of continuous crystallization processes for the preparation of pharmaceutical materials has been featured strongly in recent scientific literature. This review is an effort to gather together all of the published understanding on continuous crystallization with a pharmaceutical focus and to benchmark progress to date in realizing the potential benefits of transitioning this stalwart pharmaceutical unit operation from batch to continuous configurations.

No abstract

Development of a small volume continuous process that used a combination of batch and flow unit operations to manufacture the small molecule oncolytic candidate merestinib is described. Continuous processing was enabled following the identification and development of suitable chemical transformations and unit operations. Aspects of the nascent process control strategy were evaluated in the context of a 20 kg laboratory demonstration campaign, executed in walk-in fume hoods at a throughput of 5–10 kg of active pharmaceutical ingredient per day. The process comprised an automated Suzuki–Miyaura cross-coupling reaction, a nitro-group hydrogenolysis, a continuous amide bond formation, and a continuous deprotection. Three of the four steps were purified using mixed-suspension, mixed-product removal crystallizations. Process analytical technology enabled real-time or nearly real-time process diagnostics. Findings from the demonstration campaign informed a second process development cycle as well as decision making for what steps to implement using continuous processing in a proximate manufacturing campaign, which will be described in part 2 of this series.

The design, development, and implementation of a pilot-scale continuous Schotten−Baumann amide bond formation and reactive crystallization to afford LY2886721 is described. The material met all API quality attributes and was comparable to material produced by a defined batch process. The scalability of the reaction and crystallization processes was confirmed during the development process. The pilot-scale equipment set was contained in a walk-in fume hood and operated at a production rate of 3 kg/day in a 72 h continuous run. Significant technical and business drivers for running the process in continuous flow mode were proposed and examined during development. The continuous process provided for lab hood commercialization and provided for minimal material at risk in the process. The demonstration also confirmed the risk inherent to operation of a tubular reactor under supersaturated conditions, and fouling occurred in the plug flow reactor. Fouling also occurred in the crystallizer. Recognizing these deficiencies, the process operated within the footprint of a standard walk-in fume hood, providing a successful demonstration of the opportunities afforded by continuous processing for low volume pharmaceuticals.

Lilly Compound X (LCX) is an oncology drug that was tested in a phase I clinical study using starch blend capsules. The drug was given to a small patient population (4 patients) and showed large inter- and intra-patient variability. In order to evaluate the possible effect of stomach pH on exposure and ways to mitigate the variability issue, artificial stomach-duodenum (ASD) experiments were conducted to investigate the hypothesis that carefully selected dosing fluids would have an impact in minimizing exposure variability caused by the formulation, which could lead to more consistent evaluation of drug absorption in patients. The ASD data corroborates the observed variability, and was a good tool to investigate the effect of stomach pH and potential dosing solutions on duodenal concentrations. Administering capsules co-formulated with Captisol (10% drug load) along with Sprite was shown by the ASD to be an effective way to increase duodenal concentrations as well as to reduce the difference between duodenal concentrations for different gastric pH. The reduction in variability of duodenum AUC (in ASD) is expected to correlate well with a reduction of variability in patient exposure. The dosing regimen of Sprite/Captisol is therefore suggested for future clinical trials involving LCX. Furthermore, for design of early phase clinical trials, ASD technology can be used to assist in choosing the proper dosing solution to mitigate absorption and exposure variability issues.

A fully automated fill/empty reactor system for liquid–liquid biphasic Suzuki couplings is described. The system was capable of charging reactant and catalyst solutions to a heated vessel, heating reagent solutions by flow heat exchanger on the way into the reactor, allowing the reaction to occur, monitoring reaction completion, discharge of the product solution, and initiation of another cycle in a repeating fashion. A unique noncontact colorimetric method was used to monitor reaction completion. The reactor system exhibits many of the characteristics of a fully continuous reactor such as (1) high productivity from a small process footprint, (2) a large number of volume turnovers each day, (3) higher heat transfer area per unit volume compared to batch because the reactor is 50× smaller, and (4) rapid heat up and cool down of process streams enabled by heat exchangers. Downstream unit operations that are intended for eventual integrated end-to-end continuous production included a batch metal removal step and a continuous antisolvent crystallization to isolate the product in high yield and purity.

The molar quantity of adsorbed CO and H2 present on the surface of a mixed CuO x -CeO2 catalyst during CO preferential oxidation in H2 at 353 K was quantified using both the reactive titration method and the steady-state isotopic-transient kinetic analysis (SSITKA). For the reactive titration method, either CO or H2 was replaced by He during steady state reaction while monitoring the residual transient product formation of CO2 or H2O produced from the surface adsorbed CO or H2 via catalytic oxidations. For SSITKA, 12CO was replaced by 13CO during steady state reaction while monitoring the transient product formation of 12CO2. The amount of adsorbed reactive CO increases with increasing CO partial pressure or decreasing H2 partial pressure, while the amount of reactive H2 decreases with increasing CO partial pressure or decreasing H2 partial pressure, showing that the adsorbed CO and H2 compete for active redox sites and prohibit the other’s adsorption. Using reactive CO and H2 amounts, two models of coverage were defined with trends providing insight into the competitive redox mechanism between adsorbed CO and H2. CO oxidation is kinetically preferred over CuO x -CeO2, and the relative CO to H2 coverage is shown to be the determiner for CO2 selectivity. This new depiction of selectivity parameters provides a useful principle for the design of selective PROX catalysts.