Establishing a commercial batch size is a crucial decision in pharmaceutical operations. It is influenced by the type of manufacturing technology being used, regulatory filing commitments, supply chain demand, and operational planning factors. To understand batch size, the differences between “batch,” “continuous,” “semi-batch,” and “semi-continuous” manufacturing must first be defined (1). In batch manufacturing, all materials are charged before the start of processing and discharged at the end of processing. Examples include bin blending and lyophilization. Continuous manufacturing involves materials simultaneously charged and discharged from the process; examples are found in petroleum refining, food processing, and, more recently, in pharmaceutical manufacturing. Other manufacturing variations include semi-batch (i.e., fed-batch) manufacturing, as found in wet granulation, tablet coating, and fermentation, in which materials are added during processing and discharged at the end of processing. In semi-continuous manufacturing, materials are simultaneously charged and discharged, but for a discrete time period. Examples include roller compaction, tablet compression, and encapsulation. For semi-continuous manufacturing processes, the process output is independent of batch size as long as the material input is set up to produce consistent output as per the controlled process. Therefore, a fixed batch size is not required for semi-continuous manufacturing processes.

	Regulatory guidances provide clarity on batch requirements. FDA’s 

Guidance for Industry: Immediate Release Solid Dosage Forms Scale Up and Post Approval Changes 

(2) outlines the  maximum allowable batch size as 10 times the size of the pilot/bio batch. FDA’s cGMP guidance 

 (3) provides clarity on batch and lot definitions. A European Medicines Agency (EMA) draft guideline on manufacture of finished dosage forms (4) notes that the batch size for a product to be marketed should normally be compatible with qualified equipment. It should be sufficient enough to allow process capability to be established. For example, a commercial batch size for solid oral dosage forms should be at least 100,000 units unless justification is provided. The equipment capacity and maximum quantity allowed determines the maximum batch size.

	Operational efficiency increases with use of smaller batch sizes for such semi-continuous operations as compaction, tablet compression, and encapsulation. Taking the product out faster achieves higher asset and line capacity utilization. Having a minimum possible batch size can also increase supply chain velocity. Examples of unit operations that are not affected by having non-standard batch size include: particle size reduction, fluid-bed drying, roller compaction, extrusion, tablet compression, and encapsulation (5).

	FDA’s 2015 guidance on emerging technology (6) promotes adoption of continuous manufacturing technology, and the recent approval of tablet production on a Janssen continuous manufacturing line highlights regulators’ acceptance of continuous manufacturing, as long as product quality is maintained. Traditional tablet compression and encapsulation operations are semi-continuous manufacturing operations, which implies that the same duration for the tablet compression or encapsulation run should yield similar quality product. The following sections discuss approaches for determining and provide justification for an appropriate minimum batch size for semi-continuous dosing unit operations, such as tableting and encapsulation.

	Control strategies for tablet compression and encapsulation manufacturing include in-process quality control specifications for such quality attributes as tablet weight, hardness, thickness, friability, and disintegration time. These attributes should be measured frequently (i.e., at intervals of 15 or 30 min.) throughout the run. The quality attributes are guaranteed through the compression run by force control established during product set-up. Any tablet compressed beyond the upper and lower limits of the set main compression forces will be rejected. For both tablets and capsules, 100% or frequent weight sorting/check are also used to assure product quality over the entire run.

	Powders, however, exhibit complex behavior during processing. The internal dynamics of powder systems under dosing are specific to a product. The start and end fraction samples are considered “worst case” and are indicative of segregation issues that may occur during a routine manufacturing process. Solid-dose tablet compression/encapsulation processes do not typically include purging as a control strategy to reduce segregation risk. Therefore, along with the stratified process validation (PV) Stage 1 content uniformity data, the content uniformity results from PV Stage 2 batches provide a clear indication of homogeneity. Typically, stratified samples are collected from across the dosing run from 20 locations to verify consistency and reproducibility throughout the batch, and replicate samples are tested for uniformity per location. The unit dose is also tested to confirm product-specific physical and chemical quality attributes (see 

). In addition to content uniformity, active ingredient homogeneity is evaluated in powder blends (i.e., blend uniformity) in all products at each batch size and bin size at Stage 2. The drug-content uniformity is determined by blend uniformity, stratified dosage uniformity, and assay. In-vitro performance of the dosage form is assured by dissolution testing. Based on the assurance of multi-component uniformity and in-vitro performance, it can be concluded that a semi-continuous tablet compression/encapsulation process is capable of producing consistent unit doses with any dosing batch size.  

					Appearance, individual tablet weight, weight of 10 tablets, hardness, thickness, friability, disintegration 

					Stratified dosage uniformity (minimum of 20 samples), assay, dissolution profile, product-specific certificate of analysis and submission requirements. 

					Appearance, individual capsule weight, weight of 10 capsules 

	In a systematic approach to determine the minimum tablet compression batch size, the following factors can be considered: tablet weight, API percentage, manufacturing process, and set-up waste from similar products and/or processes. 

 The total run time for the dosing batch is based on the target tablet/capsule weight. For any assigned minimum batch size to be compressed, the products with minimum tablet weight will take more time to compress than products with maximum tablet weight. As per the Stage 2 stratified sampling, a total of 20 locations are sampled during the dosing run. Therefore, use of a minimum batch size can be evaluated with products of low and high unit weights. 

Robustness of unit dose weight uniformity.

 Tablet weight is typically monitored using force control mechanism throughout the tablet compression process, where rejection limits (S+ and S- rejection forces and M+ and M- adjustment forces) are defined. Any tablet outside the acceptable limits is rejected. Forces are specific to a tablet press model, although the concept is applied for all. Tablets or filled capsules can be subjected to weight check as well. Normally, higher tolerance limits for weight checking are maintained, which are tighter than the values in the master document.

 Low API % w/w in the formulation is considered a high risk factor due to potential segregation after final blend and a high risk for content uniformity. Therefore, a product with low unit weight and low API content can be selected to further evaluate the minimum tableting batch size. This factor can also be used to determine variability within a batch for a particular product. Data of comparable campaigns can provide a reasonable indication of the anticipated batch-to-batch variability. The magnitude of the batch-to-batch variability can be dependent on several different factors; one factor in particular is the API content or product label claim. To gain an understanding of batch-to-batch variability, historical dosage uniformity and dissolution data from a large number of campaigns (e.g., more than 200 Stage 2 campaigns encompassing over 700 individual batches and approximately 100 distinct molecules) can be compiled. The batch-to-batch variability can be extracted from each campaign by separating the intra-batch variability from the overall total campaign variability. The data can be then segregated and analyzed to assess the relative influence of several factors (e.g., manufacturing process, strength, batch size). Product active content or strength has a high probability to be a factor that correlates to batch-to-batch variability. Typically, batch-to-batch variability for both content uniformity (CU) and dissolution significantly increases for low-strength products (<1 mg). Thus, one relevant factor for batch-to-batch variability is active content or strength (7).

Another factor that affects the minimum required dosing batch size is the manufacturing process. For example, the potential for segregation is minimal with hot-melt extrusion processes, in which the API is formed into agglomerates by melting of liquid binders. A direct mix manufacturing process has a higher potential for API segregation than a compacted process wherein the API is locked into granules. Therefore, a product with low API content and low unit weight from a direct-mix manufacturing process and high API content and high unit weight from a granulation manufacturing process may be considered as extreme cases for the assessment. 

Start-up waste from similar product/process.

 Start-up waste is generated due to initial adjustments to achieve the in-process quality attributes (i.e., tablet weight, hardness, thickness, friability, and disintegration) prior to starting the compression process. Process performance qualification samples are collected after start up, where all the in-process specifications are achieved at an established consistent tablet-press speed. In-process quality attributes are maintained by the force-control mechanism throughout the compression process. Because the tablets manufactured as start-up waste are disposed of, the start-up waste generated should be considered for the minimum batch quantity requirement.

	Based on these four factors, the product portfolio can be assessed and summarized as shown in 

	The assessment is based on an underlying assumption that the equipment qualification and calibration tolerances of key parameters were considered, and that a compliant quality management system is in place to detect any adverse trends or deviations. Additional factors can be added to the assessment. A minimum dosing batch size can be assigned based on the factors (tablet weight, API percentage, manufacturing process, start-up waste) as well as equipment capacity/capability (e.g., hopper feeder, powder bowl capacities) and regulatory market requirements, with the underlying assumption that the product has an established control strategy. Application of a minimum batch size based on the described strategy is of low risk for the product with a defined manufacturing process from Stage 1: process design, confirmed at Stage 2: process performance qualification, continually monitored as part of Stage 3: continued process verification, and with a control strategy established, including in-process controls. Jana Spes, vice-president of Technical Operations, Apotex, summarizes that variable batch sizes for semi-continuous processes such as tablet compression and encapsulation, without compromising product quality, allows for operational flexibility and efficiency. With regulatory acceptance of continuous manufacturing technologies and the associated minimal impact of batch sizes, it is time to reconsider the approach to batch-size determination for traditional semi-continuous manufacturing processes as well.

		S. Chatterjee, “FDA Perspective on Continuous Manufacturing,” IFPAC Annual Meeting (Baltimore, MD, January 2012).	

Guidance for Industry: Immediate Release Solid Dosage Forms Scale Up and Post Approval Changes

Draft guideline on manufacture of the finished dosage form

Draft Guidance for Industry: Advancement of Emerging Technology Applications to Modernize the Pharmaceutical Manufacturing Base


	Naheed Sayeed-Desta is manager, Process Validation, 

; Ajay Pazhayattil is associate director, Process Validation; and Srihari Chowdari is manager, Technical Operations, all at Apotex.

 Supplement
	September 2016
	Pages: s16–s19

:
	When referring to this article, please cite it as N. Sayeed-Desta, A. Pazhayattil, and S. Chowdari, "Determining Minimum Batch Size," 

Articles

Features

Flexible batch sizes for semi-continuous unit operations, such as tableting and encapsulation, can improve efficiency while maintaining quality.

Determining Minimum Batch Size

The pharmaceutical contract manufacturing market is expected to hit revenues of $80.5 billion in 2019, according to a report by business intelligence provider Visiongain (1). A notable trend is that pharmaceutical companies will continue to outsource more drug production operations. For several years, API manufacturing has formed the largest share of the pharmaceutical contract manufacturing market, driven by the increased use of generic drugs worldwide, the rise of biologics and biosimilars, and the growth of emerging markets such as India and China (1).

	To remain competitive in an increasingly demanding pharmaceutical market, contract service providers are constantly striving to increase the capacity and efficiency of their manufacturing activities. 

spoke to industry experts about process optimization in API manufacturing. Participants in this roundtable discussion include Joshua P. Van Kley, corporate sales and sales operations manager, Cambrex; David Goeddel, PhD, group leader, Process Development, MilliporeSigma; 
	Andreas Stolle, vice-president, Process Development Services API, and Peter Poechlauer, innovation manager, who are both based in Linz, Austria, which is currently Patheon’s largest API manufacturing site.

 What factors should be taken into account when developing an API manufacturing process? What guidelines have regulators provided?

 When developing an API manufacturing process, there are a number of important considerations from a practical and logistical standpoint. Firstly, can the chemistry be performed at the manufacturing site in terms of handling the necessary solvents and reagents, and does the plant have the capabilities to accommodate the temperature ranges of the process? It is also important to evaluate the specific hazards and safety implications of undertaking the process.

	Availability of key raw materials must be evaluated to ensure that they are readily available from existing suppliers, or whether new suppliers can be established, to avoid a situation where you are limited by supply of a key raw material or unable to import it.

	It is also important to look at the process from an environmental point of view, to ensure that all waste can be handled and disposed of properly, and also to ensure that the process is scalable from laboratory through to the commercially projected scale. Hazards, by-products, and waste products that are not as consequential at smaller scale can become major issues at large scale, thus, it is important to factor these considerations in from the beginning.

	There are also many other considerations that come into play, such as clinical phase, cycle time, the control of the product’s particle size, polymorphism, and handling issues such as the filterability of steps within the process. All these factors can have an impact on the quality of the product as well as cost of goods, therefore, it is important to bear them in mind when providing a quality product as well as meeting the customer pricing demands.

 Several factors should be taken into account when developing the manufacturing process for an API. Careful focus should be placed on ensuring that the overall purity, purity profile, and individual impurity levels are at acceptable levels to ensure the safety of the patient. Guidance has been provided in this area by the International Council for Harmonization (ICH) on threshold limits for impurity identification and qualification in API drug substances--ICH Q3A. In addition to organic impurities, process chemists should also pay close attention to residual solvent levels and elemental impurities when developing a process for API manufacturing. This factor is particularly important when metal catalysts are used in the API synthesis, and remediation techniques (scavengers, charcoal, or crystallization) are often required to reduce these impurities to the acceptable levels outlined in ICH Q3D.

	In addition to impurities, another aspect that needs to be taken into consideration when developing an API process is the potential reactive hazards. Performing a thorough safety evaluation and modifying the chemistry as appropriate will enable the API to be made safely, which will help prevent operator injuries, plant or equipment damage, and potential supply-chain interruptions. The raw material supply chain is another important factor. Not only does the vendor need to be qualified, but they must also be able to ensure the long-term timely delivery of needed raw material quantities in the required quality.

	Finally, throughput is another important factor that should be taken into account when developing a process. Improvement in yield lowers API cost and reduces waste, which will consequently benefit the patient, company, and the environment.

	API purity, impurity levels, raw material supply chain, yield, and process safety are all important factors that should be taken into account when developing a process for API manufacturing. Focusing on those key areas will help secure both patient and employee safety while completing efficient chemical syntheses that reduce cost and minimize the impact on the environment during API production.

The development of a 
	pharmaceutical manufacturing process has to meet different requirements depending on the development phase of the product:

		In early clinical development (CT I), the primary goal is to deliver the required amounts of material quickly and in reproducible quality.

		Later on (in CT II), when the route is frozen, the production process must be reliable, well understood, and again deliver the product in the required quality. The appearance of potentially genotoxic byproducts in the final product must be excluded in a safe and scientifically sound way.

		Finally, the process for the final clinical trials (CT III) and launch of the product must be scalable to deliver the required product volumes with predictable quality, and it has to be environmentally benign and economical on resources.

		After product launch, the process must have room for continuous improvement without major changes.

	In addition to the process figures, the following factors must be taken into account:

		Availability and safe supply of starting materials in constant quality

		Investments in equipment to operate the envisaged process at the desired scale

		Start-up time and time demand for capacity changes

		Easy process transfer between different manufacturing sites to meet local demand and support supply-chain optimization.

	The current regulatory environment supports advancing regulatory science and innovation, which may include abandoning some traditional manufacturing 
	practices in favor of cleaner, more flexible, and more efficient continuous manufacturing. Regulatory authorities in the three ICH regions and beyond are encouraging the industry to adopt new technology as supported by ICH Q8(R2), Q9, Q10, and Q11; and the introduction of quality-by-design (QbD) concepts, emphasizing science and risk-based approaches to assure product quality. The regulatory expectations for assurance of reliable and predictive processing, which is technically sound, risk-based, and relevant to product quality in a commercial setting, are the same for batch and continuous processing.

 How do you optimize process chemistry in API manufacturing? What are the key considerations?

 Our key considerations in optimizing process chemistry are driven by optimizing the service to our clients. This approach comprises considerations such as:

		Are the involved chemicals and reagents reliably available?

		Is their price acceptable in view of the target product price?

		Does the generated waste pose a problem?

		Can we develop a sufficiently detailed process 

		Do the chosen conditions allow quick and safe scale-up?

		Is the process easily transferable between sites or does it require specialized equipment?

Batch-wise versus continuous processing concept:

		Will the process yield/throughput profit from continuous manufacturing techniques?

		Will process control profit from continuous manufacturing?

		Can we develop a sufficiently quick and easy process control system?

		Are the chosen analytical methods reliable and easily transferable?

		Does the process control system support continuous process verification?

	Our optimization strategy comprises both classical determination of proven acceptable ranges (PAR) values and, in tight collaboration with clients, strategies of multivariate analysis and other elements of process analytical technologies. In addition to technical aspects of optimization, there are aspects related to client requirements, such as use of innovative but proven technologies to provide maximum value.

 Initially, the process is carried out in its current state using the conditions provided by our clients. This approach allows us to observe the chemistry and get a feel for how it performs. From there, the next stages of development investigate ways to reduce solvent volumes, increase yields, reduce cycle times, lower raw material costs, and lower waste costs. These steps are crucial to improving product quality and the economics of the process, which allows us to pass efficiencies and qualityon to our clients. Most of this work is undertaken in the chemical development laboratory prior to going into production. Once in production, the chemist and engineer assigned to the program will further work on optimization of the process based on observations made during production. In addition, our continuous improvement/six sigma group will also contribute to the optimization process once the program is in validation or commercial launch. The group will help in managing the lifecycle of the program along with looking at ways to continually improve the efficiency of production by data mining.

	Typically, we will see programs that have chromatography steps within the process, high volume issues, filtration issues, and/or long cycle times. Our development efforts are centered on removal of any chromatography processes if present for scaling purposes, volume reductions, faster filtrations, and cycle time reduction, either for efficiency or the possibility of telescoping steps to reduce unnecessary isolation steps if the process lends itself.

 We strive to perform phase-appropriate process optimization for API manufacturing. Process optimization means very different things for Phase I clinical programs compared with programs that are entering validation. For an API that will be entering Phase I, the key objective is usually to rapidly develop a process that can safely yield the required API with the necessary quality attributes. This way, clinical evaluation of the API can begin quickly, which is important for both drug developers and patients who seek successful treatment. As the program advances toward validation and commercial launch, greater emphasis is placed on improving yield and gaining greater process understanding to support process validation and eventual launch.

	No two programs are the same, but there are some consistent factors that generally apply to most programs. We consider several factors when deciding whether the incoming synthetic route can be used or if a new synthesis should be developed. Raw material supply chain, process safety, projected future API manufacturing costs, likely commercial scale, and timing all play an important role in the decision-making process. After a route is selected, proof-of-concept studies are performed to determine whether or not the proposed route can generate the API. We then optimize the process to reliably and safely generate API in adequate quality. This objective is achieved by building process understanding through many techniques, including impurity origin and control, identifying critical parameters, and setting appropriate limits on operating ranges. We develop this chemistry with an eye on the intended commercial manufacturing scale, because APIs that will be manufactured on a smaller scale will have more processing options available than those that will be made on a larger scale. During the final phase of optimization, experimentation is performed to determine if the API can consistently be manufactured with the required quality attributes. Statistical design of experiments is a particularly useful technique for these studies, because interdependent variables can readily be identified. The successful completion of these phase-appropriate process optimization efforts enables us to deliver high-quality clinical batches and commercial supply in a timely manner, which is important for our customers and their patients.

 How can manufacturers ensure that APIs of the intended quality are consistently produced?

 Manufacturers take steps that span from early research and development through commercial manufacture to ensure that APIs of consistent quality are produced. The general pathway for this process is outlined by FDA and involves three phases: process design, process qualification, and continued process verification.

	During the process design phase in development, great effort is made to understand what parameters are critical. Building upon that knowledge, the process is optimized as necessary to enable the desired quality attributes to be consistently achieved. Manufacturers then perform a failure modes effects analysis (FMEA) on the process to identify processing risks that could impact quality attributes. From that exercise, additional experiments can be designed to address risks identified in the FMEA to ensure that the critical quality attributes are reliably met. During this stage of development, manufacturers perform stress testing, stability studies, design of experiments, and range-finding studies to help ensure that the intended quality is 
	consistently produced in subsequent manufacturing.

	The process qualification phase involves an assessment of whether or not the process is reproducible. There are two major components to process qualification. The first part involves the qualification of the plant and equipment to ensure everything works as intended. The second part involves the qualification of the process itself through an activity known as process performance qualification (PPQ). PPQ involves drafting a protocol, execution of the protocol for the specified number of batches under current good manufacturing practices (cGMPs), and issuance of a report. Following successful completion of the process qualification, the process can be used for commercial supply of the API.

	The process to ensure product quality does not end with product launch. Manufacturers use systems that enable them to track process data and identify any sort of trend that may require intervention. Furthermore, an adequate facility and equipment maintenance program ensures that the plant and equipment are functioning at the desired level. By performing all of the aforementioned activities, manufacturers can ensure that APIs of the intended quality are consistently produced.

 The key to consistent quality product is a sound process understanding combined with effective process control. Process understanding suffers if the features of the processing equipment mix with features of the actual chemical reaction, blurring them and interfering with precise process control. Consistent production of APIs of intended quality starts with a kinetic and thermodynamic analysis of the synthesis reaction. The rate, energy balance, and kinetics of by-product formation and factors such as equilibria of phase distribution determine the requirements of the process. They in turn determine the features of the processing equipment and ultimately the equipment selection. They also determine the control strategy to effectively safeguard consistent product quality. A sound process understanding allows the 
	conscious choice of proven acceptable ranges for reaction parameters and intermediate product quality. It avoids overly narrow parameter ranges or unnecessary tight intermediate product specifications and thus allows for continuous improvement without putting API quality at risk. In many cases, continuous processing simplifies the precise control of process conditions even for processes that are very exothermic or require quick mixing to establish the correct stoichiometry and avoid byproduct formation.

	Continuous processing equipment can be tailored to meet the respective requirements of a chemical reaction or work-up section with moderate effort. Its combination with state-of-the art methods of continuous analytics allows precise and reliable control of product quality.

	There appears to be a paradigm change: instead of slowing down the chemistry to a degree to allow large-scale batch processing equipment to cope with heat evolution etc., the developer determines ideal conditions for the respective chemical transformation and defines (or if necessary constructs) suitable processing equipment and control instruments. This approach leads to more modern ways of processing and, as the PAT guideline (2) stipulates, ‘development and implementation of innovative pharmaceutical development, manufacturing, and quality assurance.’

 An important step is performing critical process parameter studies. A critical process parameter study is initiated to identify those critical parameters in the process that affect the final product quality and reproducibility. Typical critical parameters to be investigated at each step requiring validation may include: temperature, charge ratios, pressure, vacuum, time (duration), flow rate, cooling rate, and agitation speed, among others.

	Critical process parameters are usually identified and studied after initial laboratory work, or after initial manufacturing campaigns, when the chemists can observe the behavior of the chemistry at scale. Our customers typically dictate when they want this work to be performed to tighten up the operating parameters.

	From a quality perspective, Cambrex follows ICH Q7 guidelines. In addition, we have a strong analytical method validation program in place for all analytical methods, including cleanout methods for each isolated intermediate as well as finished goods. The validated cleanout methods not only ensure quality for the current product being manufactured, but also ensure the quality and integrity of the plant for the next product to be produced, as we operate a multipurpose facility with non-dedicated production streams.

Pharma Contract Manufacturing Services Market Will Reach $80.5 Billion in 2019

,” accessed July 19, 2016. 
	2. FDA, Guidance for Industry, PAT--A Framework for Innovative Pharmaceutical Development, Manufacturing, and Quality Assurance (Rockville, MD, September 2004).

 Supplement
	September 2016
	Pages: s6–s10

	When referring to this article, please cite it as A. Siew, " Optimizing API Manufacturing," 

Sound process understanding and having effective controls in place are crucial in ensuring that consistent product quality is obtained during API manufacturing.

Optimizing API Manufacturing

After years of taking excipients for granted, the pharmaceutical industry is beginning to realize just how complex these “fillers,” and the products into which they are formulated, can be. The physicist Kim Christensen has noted the following as signs of the complexity of any system (1): 

		Many degrees of freedom (i.e., many variables that are free to vary in a dynamic system)

		Emergent behavior that has not been encoded in the rules.

	In formulation and product development, the industry must currently deal with the following:

		The variability (degrees of freedom) found in both excipients and finished products

		Special cause variation in finished products, independent of excipient compliance.

	Compendial specifications are not enough

	Pharmaceutical manufacturers today contend with such complexity in the face of increasing analytical capability and regulatory pressure for better quality. The FDA has stated that pharmacopeial compliance is a necessary, but insufficient, prerequisite for pharmaceutical quality (2):

	“When a USP or National Formulary (NF) monograph material is used, the associated specifications may not always provide adequate assurance with regard to the assay, quality, or purity of the material or its performance in the drug product. In these cases, monograph specifications should be supplemented with appropriate controls (e.g., particle size distribution, crystal forms, amorphous content, foreign particulates) to ensure batch-to-batch reproducibility of these components.”

	Simply citing pharmacopoeial compliance for excipients under the Specifications section of the Chemical Manufacturing and Controls (CMC) portion of a new drug application may result in requests for justification of reliance on compendial specifications.

	One of the major degrees of freedom in excipients is polydispersity, in terms of their molecular weight, composition, and particle size distribution. Additional degrees of freedom arise from the raw materials (including additives) and processes used to manufacture the excipients.

	Many common excipients are continuously manufactured in high volumes, so the variability behind the results on the certificate of analysis is an additional degree of freedom. Any attributes not specified by the monograph also represent degrees of freedom.

	All of these degrees of freedom illustrate why compliance alone cannot determine whether a given excipient will be fit for purpose in a particular application. The products into which the excipients are formulated also have multiple degrees of freedom.

	Paradoxically, attempts to reduce finished product variability, by fixing the processes and formulas, have been counterproductive, since raw material variability can feed forward in a rigid system and affect finished-product quality. Any design or control strategy that does not have compensatory mechanisms to deal with raw material variability is at higher risk from the impact of excipient variability.

	Complex systems, such as pharmaceutical products, may exist in multiple states. The transition from one state to another is known as a criticality.

	Properties or performance on one side of the transition may not allow manufacturers to predict performance on the other side. Such critical transitions are another degree of freedom, and are also known as latent conditions because their existence is unknown to the designer.

	The most common sources of criticalities in pharmaceutical systems are percolation effects, or conflicting technological requirements.  Percolation effects abound in powder mixing and tabletting physics. Overgranulation is an example of conflicting technological objectives. If the minimum effective granulation level is too close to the level that causes overgranulation, a criticality will result.

	If minor excipient variability, possibly a known attribute within its norms of variability, interacts with a finished product criticality, a disproportionate, non-linear response may result, such as an out-of-specification excursion. Not all sources of critical variability will be identified at the time of filing, and excipients can cause special cause variation in finished product quality. Such variation is unanticipated, outside the historical experience base, and inherently unpredictable, even probabilistically.

	Traditionally, excipients have been categorized as critical or non-critical.  Critics have said that this approach to classifying process parameters and quality attributes is overly simplistic, and does not adequately reflect science and risk-based thinking (3).

	A better approach is to divide the excipients into design-critical or not design-critical. Design-critical implies a direct cause and effect on finished product performance, where the level of the excipient is titrated above a minimum, or to an optimal level, for specific performance in the finished product. Examples would be disintegrants, suspending agents, or the rate-controlling polymer in a controlled release matrix. Variability in such excipients can directly impact finished product performance.

	By definition, “non-critical” excipients and their variability have no observed effect on the critical quality attributes of the finished product during development. However, just because there is no evidence of a problem does not mean there is no problem. Because there is no reason to expect an impact, a null finding is inconclusive.

	History is full of examples in which “non-critical” excipients have become critical during a product lifecycle, often due to their interaction with finished product criticalities. Because impact is based on variability plus drift plus criticality, the impact of “non-critical” excipients is indirect.

	Indirect impact from “non-critical” excipients highlights the weakness of conventional change control. A pharmaceutical product is subject to the cumulative effect of multiple changes throughout its lifecycle.

	Univariate change control observes no impact from each individual change, but may not detect drift due to the cumulative effect of the multiple changes. Ultimately, one change too many triggers a criticality correlating with an excipient variability. It is important to note that the excipient variability has not caused the problem, but is now merely coincident with the criticality. Continuous multivariate monitoring of the finished product is required to detect product drift.

	Left unchecked, product drift may adversely affect process capability, leading to out-of-trend or out-of- specification results. If a criticality is triggered, there may be a non-linear, disproportionate impact invalidating models or control logic. The potential impact could become more significant with the move to continuous processing and real-time release. It becomes more difficult to demonstrate control. 

	A well-thought-out control strategy is the only way to deal with so many excipient-related degrees of freedom. The application of materials science and risk-management principles will drive the need for more and better data on excipient properties and performance. This will require closer collaboration between suppliers of excipients and their customers.

	Ranging studies may be useful in identifying the sensitivity to excipient variability vs level of incorporation. Products with an excipient close to the minimum effective level will be more sensitive to variability in that excipient.

	Design of experiments should focus on design-critical excipients, but the control strategy must cover all excipients and related application-specific failure modes. Good designs build in compensatory flexibility. Usually, this is in terms of process parameters, but future designs could focus on formulation.

	Access to supplier data going beyond the certificate of analysis is needed in order to understand supplier process capability and to enable continuous multivariate monitoring. Beause product drift and criticalities cannot be verified experimentally, it is up to users and suppliers to work together on due diligence. Such collaboration provides the best, lowest risk way for incorporating excipients into the product and process design.

	1. K. Christensen and N. Moloney, Complexity and Criticality, Imperial College Press, London, 2005.
	2. FDA Guidance for Industry, 

Nasal Spray, Inhalation Solution, Suspension and Spray Dried Products

,” (FDA, July 2002).
	3. D. O’Keefe et al., “

A Spectrum of Importance: Challenging the Concept of Critical/Non-Critical in Qualification and Validation Activities

,” Institute of Validation Technology, ivtnetwork.com, Feb. 29, 2016. Retrieved July 26, 2016.

	Brian Carlin is director of open innovation at FMC Corp.

 Supplement
	September 2016
	Pages: s12–s14

	When referring to this article, please cite it as B. Carlin, " Dealing With Complexity in Excipients and Formulations," 

One of the major challenges in working with excipients today is understanding, and adjusting to, complexity in materials and formulations.

Dealing With Complexity in Excipients and Formulations

Needle-free injection technologies can improve patient experience and compliance with treatment, eliminate needlestick injury, and enable self-administration. Several needle-free technologies such as liquid jet and powder injectors have been developed, while future alternatives such as microneedle patches remain in development (1, 2). 
	
	This study investigates the use of a novel solid-dose formulation delivered as a whole entity into the skin, by a needle-free delivery method (Glide SDI, Glide Pharmaceutical Technologies). With the needle-free delivery method, the potential dose size available is greater than with microneedle patch delivery, and the entire dose is received each time. Solid-dose formulations also have the potential for increased stability of the API compared to liquid injections, reducing costs and inconvenience to the user.

	One example of an extended release solid-dose injection is Zoladex (goserelin acetate), for the treatment of female infertility and some breast and prostate cancers. Zoladex, however, is administered by a healthcare professional via a needle. Due to the needle, minimal stress is applied to the solid dose during injection, but this device does not offer the benefits of needle-free technology. In a needle-free system, the solid dose requires significant mechanical strength to penetrate the skin, while maintaining the required drug release (immediate- or sustained-release) profile.

	In this study, solid doses containing octreotide were injected into pigs and the subsequent pharmacokinetic release profiles measured. Octreotide is currently approved for the treatment of several serious disorders, including acromegaly and certain neuroendocrine tumours (3).
	 

A solid-dose injector (SDI) system (Glide Pharmaceutical Technologies) was used to inject the solid doses. This system consists of a reusable spring-powered actuator and a disposable cassette containing the solid dose. Each cut dose consists of a rod of less than 1 mm diameter with a pointed V-shaped tip (

	The cassette containing the dose connects to the actuator via a bayonet fitting. The actuator is then used to press the cassette against the injection site at 90 degrees, loading the actuator spring (

). Once a defined force is reached, the spring energy is released pushing the solid dose into the subcutaneous layer.

 The solid doses were prepared by extruding a paste containing a mixture of stabilizing and bulking agents, all of which are generally regarded as safe (GRAS), and the API (100 µg/dose octreotide). The resulting extrudate was then dried and cut. The solid doses were visually inspected under magnification to check for defects prior to releasing for use.

	The solid-dose formulations were loaded into disposable cassettes and sealed in foil pouches containing desiccant to prevent uptake of moisture. The active solid doses designated for use in the pharmacokinetic study were then e-beam sterilized (10-25 kGy).

The mechanical strength of the formulations was assessed by compressing 10 2-mm long rods on a materials tester (AGS-X, Shimadzu) at 0.5 mm/min until fracture.

	The disintegration time was measured by removing a single sterilized solid dose from its packaging and cassette and placing it in a well with 200 µL of distilled water at room temperature.

	Photographs were taken at regular intervals (approximately one a minute) until the solid dose had fully dissolved. This was repeated for a total of three different samples. 

	To test for content and purity, 10 samples were weighed and fully dissolved in 10 mL purified water (18 MΩ.cm) before being analyzed using high-performance liquid chromatography (HPLC).

	For content uniformity, 10 samples doses were each dissolved in 1 mL 18 MΩ.cm before being analyzed individually. The acceptance value (AV) was then calculated in accordance with 

 Reported models used in literature to determine skin and tissue penetration forces, include cadaver (2, 4), porcine, ovine (2), and murine (5, 6). Porcine skin is considered most comparable to human skin due to its anatomical and physiological similarity (7, 8). Some studies have compared the fundamental biomechanical properties of pig and human skin (9), and higher needle penetration forces have been measured into pig at superficial depths. Porcine skin is a standard model for skin penetration at various scales, from knife wounds (10) through to microneedle array design (11).

 Using a crossover design with a 12-day washout between doses, female Large White pigs (18 to 25 kg) received a subcutaneous dose (nominal) of either 100 μg octreotide (Sandostatin) as a liquid control or 100 μg octreotide Glide solid dose. Blood samples were taken at pre-dose and at the regular intervals post-dose.

	The octreotide (100 μg/mL) comparator was injected using a 21-gauge needle at a 45-degree angle after ‘tenting’ skin. Both sets of treatment were subcutaneous at the inguinal site.

	Blood samples were collected into tubes containing K2EDTA and Aprotonin (stabilizing agent) and frozen prior to analysis.

	Porcine plasma samples were extracted and analyzed using the validated liquid chromatography-tandem mass spectrometry (LC-MS/MS) method.

 One hundred injections were carried out with a placebo version of the formulation into the neck region of an anaesthetised female Large White pig (22.0 kg), which is understood to provide a tougher barrier to penetration than the inguinal. Where a small injection mark and no other debris was visible, the actuation was considered successful.

	Solid-dose preparation test data. The solid-dose preparation test data were:
	Purity = 99.3%;
	Nominal dose = 101.5%
	Content uniformity (AV) = 6.6
	Disintegration time = 840 s (standard deviation = 92 s)
	Pharmacokinetic study compression strength = 135 MPa (standard deviation 13 MPa)
	Penetration study placebo compression strength = 160 MPa (standard deviation 9 MPa)

 values indicate no significant difference between the solid dose and octreotide reference treatments at α = 0.05 level. All dosed animals showed exposure to octreotide with no analyte measured in pre-dose samples. 

	The plasma concentration vs. time profiles were as expected for subcutaneous administration of octreotide, with the exception of female 4 following administration of octreotide. This individual demonstrated a higher C

 of 12.9 ng/mL compared to an average of 8.16 ng/mL for the rest of the group. A shift in T

 from 20 (median) to 5 min was also observed for this animal. These observations are consistent with part of the dose directly entering the systemic circulation rather than the intended subcutaneous route. Female 4 octreotide treatment was included in all tables and figures.

) pharmacokinetic profiles for the solid-dose formulation and octreotide reference treatments were similar, with the solid-dose treatment exhibiting a delay in T

 (53 minutes compared to 20 minutes for octreotide). The solid-dose formulation was not optimized for a rapid release profile as one was not required for this treatment. To achieve a faster release profile, a modified formulation would be required. 

values, the needle-free delivery method treatment demonstrated a relative bioavailability of 93.7% and 94.3% when compared to the octreotide reference, including and excluding individual 4, respectively. Overall, the exposure based on area under the curve (AUC) values, half-life, and clearance parameters were comparable between time points.

 All 100 of the penetration tests were successful with no debris visible after injection, which is consistent with the dose being fully delivered under the skin.

 values indicated no significant difference (α < 0.05) between the solid-dose formulation delivered by the needle-free delivery method and the octreotide reference treatment. This verifies that the needle-free delivery method delivers the dose to the target site and that the pharmacokinetic parameters relevant for bioequivalence in the clinic are comparable between formulations.

 (53 min compared to 20 mins for octreotide), but this should not alter the efficacy or safety of the treatment for non-emergency treatments such as octreotide. If a more rapid release rate is required, the formulation characteristics would require optimization for this.

	The data gathered demonstrate the potential for an effective solid-dose formulation treatment to be delivered subcutaneously without the aid of a needle. The next phase of development is a safety and pharmacokinetic study using the needle-free delivery method and octreotide solid-dose formulation in human volunteers.

	The in-vivo tests were performed by Quotient Bioresearch Ltd. (Rushden, Northants) in accordance with its Ethical Review Process (ERP) as detailed within the Animals (Scientific Procedures) Act 1986. 

	The study was funded and sponsored by Glide Pharmaceutical Technologies Ltd. D. Grant, D. Smith, and I. Collis are employees of Glide, and J. Sharman was an employee at the time of this study. A. Lyness completed his doctorate with Glide. S. Constable is an employee of Quotient Bioresearch (Rushden) Ltd who were contracted by Glide to conduct the 

	The authors would like to acknowledge Quotient Bioresearch Ltd for support of the in-vivo trials and LGC Ltd which conducted the pharmacokinetic sample analysis.

	1. Y.-C. Kim, J.-H. Park, and M.R. Prausnitz, 

 64, 1547-1568 (2012).
	2. P.N. Brett, et al., 

 211, 335-347 (1997).
	3. A. Rinke, et al., 

27, 4656-4663 (2009).
	4. T.B. Frick, et al., 

 34, 1335-1340 (2001).
	5. M. L. Crichton, et al., 

 32, 4670-4681 (2011).
	6. M. L. Crichton, et al., 

 34, 2087-2097 (2013).
	7. N. J. Vardaxis, et al., 

190, 601-611 (1997).
	8. T. P. Sullivan, et al., 

 9, 66-76 (2001).
	9. J. Jachowicz, R. McMullen, and D. Prettypaul, 

 13, 299-309 (2007).
	10. S. Jones, L. Nokes, and S. Leadbeatter, 

 59-63 (1994).
	11. K. Cheung, T. Han, and D.B. Das, 

	David Grant is technical lead, new technologies, david.grant@glide-technologies.com; David Smith is CMO manager; and Imogen Collins is chief scientific officer, all at Glide Pharmaceutical Technologies Ltd., Abingdon, UK; Johannah Sharman was pre-clinical manager at Glide Technologies during the study and is currently regulatory affairs manager at Kalvista Pharmaceuticals Ltd. Alex Lyness is an enterprise fellow at Loughborough University, Healthcare Engineering Group, Centre for Biological Engineering, Wolfson School of Mechanical and Manufacturing Engineering, Loughborough University, Leicestershire, UK. Scott Constable is principal project co-ordinator, Quotient Metabolism at Quotient Bioresearch (Rushden) Ltd., Northamptonshire, UK.

 Supplement
	September 2016
	Pages: s36-s38

	When referring to this article, please cite it as D. Grant et al., "A Novel Solid-Dose Injection," 

A study demonstrates the feasibility of a novel solid-dose formulation injection technology to effectively deliver a therapeutic treatment without the aid of a needle.

A Novel Solid-Dose Injection

The past few years have seen, not only the introduction of major supply-chain regulations in the United States and Europe, but changes in the way that pharmaceutical excipients manufacturing sites are evaluated to assess regulatory compliance. Manufacturers can now choose to undergo third-party audits to show that their excipient manufacturing operations comply with current good manufacturing practice (cGMP) regulations. Irwin Silverstein, consultant to the International Pharmaceutical Excipients Council (IPEC), shared his views with 
	

	Are mandates needed to drive standardization?

 Has the pharmaceutical industry fully embraced the idea of standards for ensuring that excipient manufacturers follow cGMPs? 

 Excipient makers still seem to be developing plans to meet the standard. I’m not sure how many pharmaceutical users expect their suppliers to comply. 

	In addition, there is still some confusion about the differences between requirements for standards under the American National Standards Institute’s (ANSI) 363 vs EXCiPACT. In my personal opinion, there won’t be a lot of motivation for manufacturers to get certified, or to attempt to demonstrate compliance with cGMP standards, if certification is not made a regulatory mandate.

When IPEC ran the International Pharmaceutical Excipient Auditing group IPEA (now part of NSF International), I would have expected all companies certified by IPEA as complying with cGMPs to embrace the standards. 

	I’m not sure of the exact number, but I would think that, by now, the seven sites that IPEA had certified in the US would be in substantial compliance with the ANSI standard.

	More facilities are being certified to EXCiPACT. Expectations for both certifications are essentially the same. Even companies in the US are now choosing to be certified by EXCiPACT, and these are not US-based subsidiaries of European firms, but US companies, which must pay to fly auditors over from Europe for the certification. Despite FDA’s support for standards, certification with ANSI hasn’t yet gained wide industry acceptance. 

 Can you comment on the industry’s readiness for European regulations under the Falsified Medicines Directive (FMD)? How are manufacturers and customers responding?  Where do you see gaps?

 From what I’ve heard, customers are still somewhat confused about the role that excipient manufacturers should play. The regulations call for risk assessments, which require collaboration between manufacturers and customers, and, so far, I’m hearing that there hasn’t been that much contact between the two groups. 

	If you manufacture pharmaceuticals for sale in Europe, whether you make the product in Europe or elsewhere, you have to comply with terms of the FMD for excipients.  Compliance requires that users perform a risk assessment of each excipient and excipient supplier, as well as separate risk assessments of each excipient’s role in the formulations in which it is used. Based on these two risk assessments, supplier quality indicators, and evidence that the supplier meets cGMPs, customers are supposed to get an idea of the extent to which the supplier should implement cGMPs. The supplier’s role is nothing more than supplying the information that customers need for the risk assessments. This type of information would include product data and specifics about grade and quality (e.g., whether ingredients are derived from animals, or whether the material was produced in a dedicated or multi-use facility, using dedicated or multi-use equipment). 

	Their only obligation is to ensure that this information is sent to each pharmaceutical manufacturer customer, for each excipient they sell to them. They can meet this requirement by reviewing and updating their product information packages, and including all the information that will be needed for customers to be able to perform the required risk assessments.

	When pharmaceutical manufacturers have determined their suppliers’ level of cGMP compliance, they are supposed to notify suppliers of any gaps through revised quality agreements. If many of these risk assessments had been completed, there would have been more agreements stipulating the degree of cGMP compliance expected. 

	The apparent lack of information sharing involving users, manufacturers, and suppliers does not support the idea that excipient suppliers and customers have been aggressively implementing FMD requirements. Yet regulators, notably the United Kingdom’s Medicines and Health products Regulatory Agency (MHRA), have said that they will enforce this rule rigorously and would cite violating companies, as FDA would with a 483.  Even the threat of regulatory action doesn’t seem to be enough to motivate some suppliers.

	The FMD’s intent is to require site visits, where customers audit supplier sites. Given the resources available, regulators have had to allow provisions for qualified third-party audits. 

	As time goes on, there is a strong likelihood that suppliers will rush to certification, since European regulators will allow use of certification as a component of the risk assessment. Excipient manufacturers have too many customers to be able to collaborate with more than a very significant few. 

  Regulators have voiced support and acceptance of third-party audits, in lieu of on-site inspections and audits by customers.  Why aren’t more companies adopting this approach?

 Everyone seems to have reached the conclusion that third-party audit and certification is essential to meet expectations for on-site inspections. However, third-party audits will only become the norm when pharmaceutical manufacturers that rely on certification are confident that the audits will be accepted by regulators.

IPEC consultant Irwin Silverstein discusses third-party audits and regulatory compliance issues facing suppliers and their customers.

Pharmaceutical Technology-09-01-2016

Determining Minimum Batch Size

Optimizing API Manufacturing

Dealing With Complexity in Excipients and Formulations

A Novel Solid-Dose Injection

Data Integrity Emerges as an Excipients Issue

Contract Service Providers Tackle Biopharma Challenges

Topical Formulation: Moving From Art to Science

Advances in the Regulated Pharmaceutical Use of Dimethyl Sulfoxide USP, Ph.Eur.

Pharmaceutical Technology, 2016 APIs, Excipients, and Manufacturing Supplement (PDF)