Common use of Good Manufacturing Practice Clause in Contracts

Good Manufacturing Practice. Manufacturing of FMT-like products and defined LBPs intended for oral or rectal administration (and thus, likely based on anaerobic organisms) has a few shared challenges. These include minimizing exposure to oxygen, in particular in steps of the process where the organisms are metabolically active and preserving the viability of bacterial cells during processing and storage. A variety of factors influence the viability of bacteria during the manufacturing process and subsequent storage, including oxygen exposure, growth media, shearing, composition of the buffer solutions used to suspend the bacteria before freezing or freeze-drying, cooling rate, and freeze-thaw cycles, among others. The problem of maintaining cell viability during freezing and particularly during freeze-drying for long term storage deserves special attention, as it is one of the most technically challenging steps of manufacturing an LBP. During freezing and freeze-drying, the bacterial cell wall is exposed to mechanical forces due to formation of ice crystals inside and outside the cell, which can disrupt the membrane and kill the cell. During freeze-drying, furthermore, the process of removing water by sublimation generates osmotic pressures that can damage the cell membrane. Optimization of freeze- drying cooling cycles and development of buffer solutions containing cryoprotectants or lyoprotectants is therefore an important step to ensure the long-term preservation of LBPs. While preservation conditions for a number of aerobes and some facultative anaerobes such as E. coli, and Lactobacillus and Lactococcus species has been described in the literature, there is very little published on the topic of preservation of anaerobic gut commensals [32]. Further complicating the matter of long-term preservation of LBPs, the efficiency of cooling regimes and cryoprotectant and lyoprotectant substances can be highly bacterial species-specific. There are certain manufacturing considerations that are unique to FMT-like products. Feces are a heterogenous substance composed of bacteria, viruses, fungi, food, and host secretions, which does not naturally yield itself to precise characterization. Consequently, manufacturing considerations emphasize rigorous donor screening and processing of stool donations, and relatively de-emphasize in-depth characterization of the composition, which would vary with every donation. FMT is performed using suspensions made of donor stool from carefully selected and screened healthy individuals. Donors undergo extensive health questionnaires and their blood and stool samples are analyzed for a list of known pathogenic viruses, bacteria, and parasites before being accepted. Recently, some amendments have been introduced to this process as a result of FDA’s issuance of a series of safety alerts on the potential risks of life-threatening infections with the use of FMT [33, 34] and on the risk of transmission of Sars-CoV-2 with FMT, leading to a halting of FMT studies in the US during 2020. Processing of stool donations varies depending on whether the final formulation is intended for oral or rectal administration. Stool samples may undergo a series of steps to filter the non-microbial components of stool, or the non-spore forming bacterial components of stool, depending on the product. FMT drug product may be released after meeting a specification of potency consisting of an estimate of the total aggregate of viable organisms present in the product, and the same assays may be used to demonstrate the FMT product stability for the planned duration of the clinical studies in which it is being used. Defined LBP manufacturing considerations, by virtue of the composition being known and standardized, can put increased emphasis on the characterization of the component strains and less emphasis on an in-depth understanding of the donor from whom the strains were originally isolated. FDA expects a description of the drug substance including the biological name of each of the strains and strain designations, the original source of each of the strains, their passage history, and a description of the phenotype and genotype of the product strains [1]. Furthermore, sponsors are expected to characterize their LBPs using assays that assure the identity, purity, and potency of the drug substance and final drug product, and to apply these same assays over time as part of a stability program to ensure the product remains within specification for the duration of the proposed clinical studies. Identity tests are expected to detect each of the bacterial strains that compose the LPB, and to discriminate among LBP component strains. High quality genome sequences for each strain can provide an authoritative identification of each organism and enable comprehensive identification of potentially undesirable safety traits such as antibiotic resistance genes or virulence factors. A further assessment of the risk of transmission of such genes to relevant microbial flora (for example, based on proximity to mobile elements) is of particular interest. Sponsors are also expected to determine the antibiotic resistance phenotypes of the LBP strains, with a particular focus towards identifying clinically relevant antibiotics that can be used as rescue therapies in the event of an infection suspected to be caused by LBP strains. Purity tests are expected to show the absence of contaminating bacteria or yeast above acceptable limits. Potency tests commonly used for LBPs assess the product viability, for example in terms of viable CFUs per dose. Defined LBPs are manufactured starting from clonal cell banks via fermentation, which may require optimization of growth media and physiological parameters like mixing, temperature, pH, retention time, and redox potential. After fermentation, bacteria are harvested by downstream steps such as filtration, which may require selection of appropriate filtration membranes and optimization of process variables such as transmembrane pressure and flow rates to minimize shear-induced damage to the bacterial cell. A further challenge inherent to multi-strain defined LBPs manufactured as monocultures is that the number of banking campaigns, production runs, and characterization assays required scales linearly with the number of strains in the product. Taken together, these considerations impose a significant burden on drug developers but also create an opportunity to innovate: a non-trivial amount of the advances made by LBP developers will originate in their process development and GMP manufacturing activities. Preclinical and clinical models to discover LBPs and study their pharmacology While not strictly required by FDA, use of in vivo and in vitro models to test the efficacy and characterize the mechanism of action of LBP candidates prior to use in humans can be a sensible business decision. A challenge in use of animal models to study efficacy of microbiome drugs is that it is not always clear what microbiome endpoints are the most relevant surrogates of therapeutic efficacy. For example, pinpointing a specific microbiome endpoint most predictive of efficacy in treating immune or metabolic disease is not straightforward. An advantage of designing LBPs for use in AMR is the relative clarity of the microbiome endpoints used to quantify efficacy and their relation to the therapeutic goal: the microbiome endpoint of an animal model used for efficacy testing may be reduction or elimination of MDRO carriage in the gut (e.g., carbapenem-resistant Enterobacteriaceae [CRE], extended spectrum beta-lactamase [ESBL], or VRE), and the therapeutic goal may be to prevent infection outcomes with that same MDRO. Rodents, for example, have been colonized (at least temporarily) with pathogenic MDRO strains that infect humans, without resorting to surrogate mouse pathogens [14], and used to rationally select defined bacterial consortia that reduce intestinal colonization. Whether the surrogate endpoints of decolonization models truly predict clinical outcomes of LBPs will have to be demonstrated in future clinical studies. A challenge in measuring efficacy of LBPs in AMR applications is the difficulty in anticipating which patients will be exposed to the pathogen, become colonized, and develop disease, which complicates execution of clinical studies powered on the basis of disease outcome endpoints. CHIM, where carefully selected human volunteers are deliberately infected with well-characterized infectious agents in a controlled setting can be an effective way of measuring the efficacy of a drug agent in these circumstances. CHIM have the advantage of decreasing the number of patients needed to detect efficacy in phase 2 and 3 trials, and have been used for testing vaccines in early in clinical development, dating back to 1900 [35]. CHIM offer the opportunity to study the physiological, immunological and metabolic changes that occur upon infection, including potentially assessing the role of the gut microbiome in transmission of antibiotic resistance and virulence genes. Conclusion Identification of commensal bacteria that can restore gut colonization resistance after antibiotics in high-risk patients is an important new strategy to prevent infection and transmission of MDROs. Use of LBPs as anti-infectives could circumvent a key limitation of antibiotics, namely the need for stewardship driven by selective pressure on resistant strains, while providing a potentially safe and convenient way of restoring the microbiota after antibiotic use in high risk patient populations. Notes