Jennifer Markarian is manufacturing editor of BioPharm International.

Jennifer Markarian

Pharmaceutical manufacturing equipment companies and technology providers are exploring how artificial intelligence (AI) can benefit process control systems. Two Italian companies—pharmaceutical machinery manufacturer Luperini and technology start-up Eigengran—partnered to launch AI-enabled equipment for oral solid dosage drugs that they have called Cognitive Pharma Machinery.

Luperini signed a licensing agreement to use Eigengran’s cognitive control technology in December 2022, following 18 months of testing and development. The initial work focused on high-shear wet-granulation (HSWG). In one-pot HSWG equipment, such as that made by Luperini, the entire granulation process, including the drying step, is performed within one unit. This set-up was well-suited for the initial project of developing a cognitive control system using process analytical technology (PAT) sensors to monitor the progression of the granulation process over time.

AI Algorithms Find Use in Pharma Equipment

“Our technology is a key element in the application of PAT guidelines for oral drug production and will be a key step towards achieving Pharma 4.0 practice in the pharmaceutical manufacturing process,” said Marco Luperini, chief technology officer of Luperini, in an announcement of the technology (1). Luperini’s new test center in Chiesina Uzzanese, Italy houses pilot machines for evaluation to simulate conditions that may be encountered in real-world scenarios.

Jennifer Markarian is manufacturing reporter for 

When referring to this article, please cite it as Markarian, J. AI Algorithms Find Use in Pharma Equipment. 

Pharmaceutical Technology Trends in Manufacturing

Articles

Machine learning provides real-time data for process control of high-shear wet granulation.

Pharmaceutical manufacturers should be familiar with the terms “Smart Factory,” or “Pharma 4.0,” which is the International Society for Pharmaceutical Engineering (ISPE)’s trademarked name for Industry 4.0 in pharmaceutical manufacturing. Pharma 4.0 moves beyond automation and strives to use digital technology, including the Industrial Internet of Things (IIoT) and software tools that mine data from large datasets in the cloud, to increase productivity. These tools can be implemented in pharmaceutical process development and in clinical and commercial manufacturing, and many manufacturers in the bio/pharma industry are using them to aid in digital transformation. Yet, misconceptions and limited knowledge remain a barrier for some because digitalization is a new way of thinking and working.

At its essence, Pharma 4.0 is a paradigm shift, says Gilad Langer, industry practice lead at Tulip Interfaces. He is currently leading an ISPE Pharma 4.0 Community of Practice subgroup that is preparing to publish a Pharma 4.0 best practice guide in 2023. Industry 4.0 equipment—such as a smart sensor—is autonomous (i.e., plug and produce) and collaborative (i.e., communicates with other equipment or with people using the IIoT), explains Langer. “Pharma 4.0 is a concept that, if adopted wholly and effectively, has the promise of an order-of-magnitude productivity increase,” he asserts.

“The connectivity of smart devices changes where, how, and by whom decisions are made in the manufacturing process,” adds Sandra K. Rodriguez, senior industry analyst at life-science consultancy Axendia. “The efficiencies gained by smart manufacturing techniques cannot be achieved using paper-based, paper-on-glass, or manual processes that rely on people. The use of manufacturing execution systems (MES) is a key component in smart manufacturing. People supervise the manufacturing process in the new standard, as opposed to driving it.”

Axendia researched the state of digital transformation in life sciences manufacturing (1) and found that 43% of respondents (representing
pharmaceutical, biotech, and medical device companies) are currently undergoing digital transformation, reports Rodriguez.

Business drivers for digital transformation initiatives include “speed to market, enabling data-driven decisions, enhanced operational excellence, improved regulatory compliance, and addressing sustainability goals,” adds Mark-John Bruwer, senior director of strategic product marketing, pharma, at AspenTech. Digital transformation is being accelerated by the need for more supply chain agility, the increase in remote work, and the tight labor market, he adds. “Pharma companies are increasingly adopting digital tools for continuous process verification (CPV), advanced planning and scheduling, electronic batch records, convergence of information technology (IT) and operational technology (OT) functions, process analytical technology, and predictive maintenance,” says Bruwer. Barriers to using digital tools persist, however, including limited knowledge or skill gaps, he adds.

Moving away from paper-based records and collecting digital data (i.e., digitization) so that it can be used is a crucial early step in digitalization. One of the roadblocks to implementing this shift is that people who are not “digital natives” (generally workers born before 1980) are still most comfortable with paper.

“If a non-digital native is asked a question in an audit, they want to point to data on a piece of paper. A digital native, however, is comfortable with the idea that a picture stored in the cloud and viewed on a device is tangible evidence,” suggests Langer. He says that a picture of a temperature reading or a video of an operator performing a task are examples of digital evidence that can be collected with Tulip’s apps. “We must move beyond a ‘document mindset’. Document-based data—whether on physical paper or a pdf file—can’t be mined the way digital data can,” says Langer.

Langer adds that in Pharma 4.0, all digitalization software will use the cloud, and eventually it will look different from the current MES and enterprise resource planning software. “In the new paradigm, programs will be app-based, intuitive, and democratized so that anyone can use them. All programs will reside in the cloud, and people will use devices in the edge to interact with data in the cloud,” he explains.

This shift is happening, says Bryon Hayes, director of smart manufacturing solutions at Grantek. “Companies are moving from ‘paper on glass’ toward a vertically integrated approach,” Hayes explains. “Many companies see the value of having a cloud-hosted data hub from which they can run analytics or populate dashboards.”

Rodriguez adds that, based on Axendia’s research and interactions with FDA officials, industry executives, and MES software vendors, “we are finally seeing a much-needed shift from paper batch records to implementing MES and electronic batch records (EBR) in support of release/review by exception.” Axendia’s research found that 32% of companies surveyed plan to digitally streamline batch execution and release in the next two years (1).

Automating EBR review is a key benefit of modernizing manufacturing operations, but some companies have held back because of misconceptions regarding FDA’s acceptance of review by exception, says Rodriguez. “In an Axendia ‘Straight from the Source’ webcast [2], Francisco (Cisco) Vicenty, program manager, case for quality at FDA, debunked many industry myths including that the agency requires EBR print outs for official records and the agency requires validation of every feature in MES.”

Another misconception is that cloud storage could raise data integrity concerns not found in on-premise storage methods. To the contrary, says Bruwer, “cloud-based solutions reduce data integrity challenges by having the cloud data repository be the single source of the truth, with global access easily and securely provided to properly authenticated individuals. In addition, cloud service providers have the critical mass to ensure cutting-edge cyber-security technology and best practices are employed, which an IT department at a bio/pharma company or contract development and manufacturing organization may be more challenged to deliver, especially at smaller organizations with smaller budgets. Similarly, the software underpinning the data warehousing solution can more easily be kept current in a centralized cloud environment versus across multiple servers distributed geographically.”

Bruwer reports that survey respondents in an AspenTech study performed in collaboration with FT Longitude (3) listed cloud and software-as-a-service as key initiatives to improve operational agility in the next three to five years. “Pharma and biopharma companies are looking for long-term technology partnerships with key vendors rather than transactional software purchases,” concludes Bruwer.

Bruwer says that a data integrity challenge—in the cloud or in on-premise storage—can arise if data sources are integrated point-to-point, with many bespoke OT-IT connections accruing across sites. “These disparate implementations can become unmanageable over time, increasing the risk for data corruption,” he explains. “In response, the robust way to ensure long-term data integrity and regulatory compliance over a growing number of OT and IT nodes is to implement an intermediary data platform that
connects the disparate data sources, securely bridging firewalls, contextualizing and cleansing data in transit, and reliably delivering it to a centralized repository that provides a single source of the truth.”

Digitalization is important not only for GxP manufacturing processes, but also for process development and efficient technical transfer. “Documents and spreadsheets tend to be the norm for data collection, process knowledge capture, and the technical transfer process between the lab and production floor,” says Rodriguez. She adds, however, that there is a growing need to eliminate paper and manual processes to meet the increasing pressure to get to market faster. The ongoing shift to broader product mix portfolios, which include large-molecule products and smaller production volumes, adds to the need for digitalization.

Digital product lifecycle management (PLM) tools can accelerate commercialization and improve knowledge transfer (4). “With product and process data predominantly stored across siloed systems and paper on glass documents, data [are] difficult to maintain, reuse, and share with stakeholders across extended, global value chains,” says Rodriguez. “Managing data in a system like PLM helps biopharma companies transition from document-centric to data-centric materials and process knowledge management, establishing a seamless data flow from lab to plant. Structured data [are] easily and quickly evolved and enriched throughout the product lifecycle as the process moves across functional stakeholders, systems, and sites, both internal and external.”

Bruwer reports that nearly half of the 400 pharma professionals surveyed by AspenTech (3) listed data silos between departments as a challenge in digital transformation. “[These divides] can hinder effective information sharing across the product lifecycle, including during technology transfer and commercial manufacturing,” explains Bruwer.

In the new Pharma 4.0 paradigm, knowledge can be stored as data in a repository (i.e., a data lake) that can be mined by digital tools, adds Langer.

Software validation is moving into “Validation 4.0” with the use of computer software assurance (CSA) processes to achieve computer system validation. “Software that can be validated for intended use instead of following the traditional stepwise process will be prominent in life science manufacturing in 2023,” predicts Langer. In CSA, the burden falls on the software providers to show that critical data have integrity and are controlled, and manufacturers demonstrate control of their process using the software, he explains. “The new CSA methods use risk-based approaches to reduce the validation burden by orders of magnitude,” says Langer.

Pharma 4.0 technologies are also being applied to CPV, to enable predictive monitoring of manufacturing processes. “The cloud is a huge enabler for CPV,” says Hayes. “It’s elastic, so resource-intensive applications like CPV are perfect for the cloud. Gathering, storing, contextualizing, and analyzing large volumes of data requires heavy storage, memory, and processing power, and the cloud can scale up and down as needed.” Hayes says that the quality team should first identify their needs and then digital solutions can be found to meet these requirements. “Keep in mind that CPV is a quality-led function that is supported by technology, and not the other way around,” he cautions.

Perceptions are beginning to shift toward greater acceptance, according to AspenTech, as only 35% of respondents to their 2022 survey said that the digital transformation process was hindered by risk aversion (3), compared to a 2021 survey in which risk aversion was considered the top barrier among digital culture leaders (5).

Digital transformation is not a simple or quick change. The benefits, however, are clear, and the tools are available. 

Axendia Research Report: The Needle is Moving–Digital Transformation in Life Sciences Manufacturing

Axendia Straight from the Source Webcast: FDA Debunks Industry Myths on Electronic Batch Records and Review by Exception

Seizing New Opportunities: Pharma’s Roadmap for Smarter Manufacturing

Axendia Market Research Report: Accelerating Time to Market with Biopharma PLM

Culture Reimagined: How Pharmaceutical Firms Can Use Data and AI with

When referring to this article, please cite it as Markarian, J. Envisioning Digital Pharma Manufacturing. 

Digital transformation to the Pharma 4.0 paradigm moves forward in process development and clinical and commercial manufacturing. 

Envisioning Digital Pharma Manufacturing

PAT is crucial to process control and real-time release in continuous manufacturing of solid-dosage drugs.

Advances in process analytical technology (PAT) are enabling increased use of in-line PAT instruments for real-time measurement in continuous manufacturing (CM) of oral solid dosage (OSD) drugs. Broad use of PAT is crucial in development stages to build understanding of the product and process. In commercial manufacturing, manufacturer’s strategies for how and where to use PAT on continuous production lines varies (1), but the tools are an essential part of real-time quality control.

Near infrared (NIR) spectroscopy is one of the commonly used PAT instruments in OSD CM, typically for measuring the amount of API to allow real-time tracking of content uniformity. Raman spectroscopy is also finding use for real-time content uniformity and other measurements. These in-line spectroscopy tools use a probe inserted into the process stream or an interface that views the process material through a window to detect the signal to be measured. These in-line tools do not need sample collection or preparation; they are non-destructive and provide real-time data, which can be combined with residence-time distribution (RTD) models to allow real-time process monitoring and control.

“The use of RTD in continuous manufacturing of tablets was started at Rutgers more than 15 years ago, and it is now common to use this approach,” says Fernando Muzzio, distinguished professor at the School of Engineering at Rutgers, The State University of New Jersey and member of the Pharmaceutical Technology® Editorial Advisory Board. “PAT identifies a disturbance in the process, and you then use RTD to define a diversion interval to predict which portion of the product stream should be diverted. The RTD is calculated from the point where the sensor takes the measurement of the perturbation to the point of diversion,” he explains.

Muzzio’s research group at Rutgers has been working with eight industry participants—including pharmaceutical manufacturers, equipment suppliers, and a contract development and manufacturing organization (CDMO)—on an FDA-funded project to evaluate the RTD approach. “We have demonstrated that RTD can be implemented in different rigs for different products. It is a flexible tool,” Muzzio says.

The Rutgers research group has also evaluated next-generation, multi-point NIR spatially resolved spectroscopy (SRS) probes and found that they provide accurate and fast measurement, typically with a 3-ms acquisition time, that can detect even a single agglomerate in a blend, says Muzzio. “We have demonstrated that you can measure an agglomerate size distribution, which enables a more thorough characterization of blends. With the accuracy and speed of this technology, you can get close to 100% inspection of tablets coming through a tablet press at 200,000 tablets/hr,” he reports. The sensor technology is offered by multiple vendors and has been integrated into commercial tableting and inspection equipment, such as the inspection system from Pharma Technology (1).

Adoption of Raman within CM is increasing due to its applicability and affordability. “Recent developments have made Raman easier to use, more stable in calibration, and smaller in size [compared to earlier Raman technology]. These innovations have also made Raman scalable by reducing the cost to acquire and implement the technology,” explains Brian Marquardt, founder and CEO of MarqMetrix.

Marquardt says that Raman is currently used for the real-time monitoring of API production and for real-time release in final drug products. “Raman provides immediate chemical feedback for the identification of molecules and their corresponding concentrations that must be controlled for API, OSD, or other final product quality,” he explains. “By understanding the composition of the drug intermediate and/or final product, companies can reduce the number of failed batches and prevent significant waste, while simultaneously reducing costs and improving quality.”

Due to the mechanism of measurement, Raman spectroscopy is not impacted by common factors (such as particle size) that can cause variability in API lots. This advantage sets Raman spectroscopy apart from traditional PAT tools, according to Marquardt. Water does not interfere with Raman measurements, which is a significant benefit given the aqueous nature of many pharma applications, he adds. He further explains that some excipients may have fluorescence that can interfere with Raman signals and that moderate fluorescence can be managed mathematically with minimal effect on the Raman data, but in extreme cases, other analytical tools are required.

Improved sample interface technology enables reproducible measurements of solids, liquids, gases, powders, or slurries. In addition, new interfaces that can be applied in “stand-off mode” allow measurement through a process window for non-contact, in-process measurements. This mode can also measure through the wall of a package or container, which can simplify stability studies and avoid impacting sample sterility. Raman spectroscopy can be used, for example, to measure API concentration before tablet compression with either a contact probe in the powder stream or a contact-free interface that measures through a sapphire window. Raman spectroscopy using a high-volume, contactless probe can also be used to measure coating thickness on tablets inside a coater in real-time (2).

Physically integrating the PAT probe into the process for accurate measurement is important. Effort must also be put into connecting the instrument to the data collection system and to developing robust models that can use these data to calculate critical quality attributes (CQAs).

Thermo Fisher Scientific is working on PAT models for real-time release (RTR) as the CDMO prepares for its client’s expected 2024 commercial launch of a drug using CM. “We are planning for real-time release and using PAT data as part of our models,” explains Doug Hausner, senior manager of business development for continuous manufacturing at Thermo Fisher Scientific. “The model for content uniformity uses API measurements from the NIR in the tablet press coupled with the mass of the tablets.” The CDMO is integrating its existing stand-alone tablet testers so that data are connected and flow from the test equipment in real-time. In addition, microwave spectroscopy is used to measure moisture content. “This PAT instrument is located after tablet coating, which occurs in a separate suite and is a batch process. These data will also become part of the electronic batch record and be used for RTR,” explains Hausner.

Hausner says that the Thermo Fisher team is conducting a wide range of experiments up front to test how NIR responds to material variations so that anticipated variability can be built into the model to increase its robustness. “For long-term commercial use of CM, it would be useful to have a model-maintenance program to look ahead and identify upcoming changes (e.g., an API change) and to perform any lab work needed to adjust the model,” he suggests.

CDMO WuXi STA, a subsidiary of WuXi AppTec, started up its first CM line for oral solid dosage drug products at its Wuxi, China site, the company announced in January 2023 (3). The CM line uses continuous direct compression for tableting in both development and commercial production and incorporates PAT to monitor blend uniformity before the compression step.

Continuus Pharmaceuticals has been building up its capabilities in end-to-end CM and, more recently, has been developing its role as a CDMO. PAT tools are currently used both in development and pilot-scale manufacturing, says Qinglin Su, associate director of Manufacturing Science & Technology at Continuus. PAT tools being used include in-line NIR, mid-IR, Raman, UV, laser-based particle sizers, and pH and density meters.

“We are designing and building ‘plug-and-play’ unit operation modules to support flexible and efficient commercial continuous manufacturing, both drug substance and drug product. Unit operations are being built with ports in place for current and future PAT implementation,” says Su.

He explains that incorporating PAT into the equipment typically requires custom designs. Many of the PAT tools are interchangeable between development and pilot-scale manufacturing, but others must be designed specifically for the scale at which they will be used.

Implementing PAT for RTR requires a strong understanding of the PAT tools, adds Su. “To best use PAT for developing a RTR strategy, I believe we need greater understanding and more predefined acceptance criteria in the sampling rate/size/accuracy/precision of PAT tools for measurement of critical quality attributes (CQAs),” he says. “Better models for how in-process testing results and critical process parameters (CPPs) can be used to predict CQAs and release specifications are needed for RTR.Demonstrating comparability for in/online testing versus off-line testing is another challenge.”

Using PAT with models to predict tablet dissolution is another area that is moving forward. “PAT data on tablet composition, along with weight and size data collected at the tablet press, can be used to predict dissolution for many formulations. We’ve developed dissolution models in the lab for both prescription and over-the-counter formulations,” says Muzzio. “These models, combined with RTD, enable prediction of dissolution for real-time release.”

Regulators are taking note of the work being accomplished in PAT for CM. For example, the International Council for Harmonisation Q13 guidance for continuous manufacturing, published Jan. 6, 2023 (4), points out that PAT is well-suited for CM (e.g., for process monitoring and control or material tracking and diversion), notes Su. “I believe this guidance will assure the pharmaceutical manufacturing industry of the full support of the regulatory agencies. Regulatory agencies will also start to expect to see PAT implementation for CM,” Su concludes.

Regulations, however, are not yet taking full advantage of the progress of PAT technology, says Muzzio. “Technology has advanced to the point of allowing 100% quality assurance—which could be used for either continuous or batch processes—but the regulatory framework still requires the old methods of sampling in the approval process. Eventually, PAT technology could drive a change both in quality assurance and in the way drugs are approved,” he concludes.

Markarian, J. Breaking Through Barriers to Continuous Manufacturing. Pharm. Tech.2022 46 (4) pp. 16-20.

Marquardt, B.Raman Detects Pharma Processing Fingerprints. Pharmaceutical Technology, Solid Dosage Drug Development and Manufacturing Supplement, April 2021.

STA Pharma. WuXi STA Launches First Continuous Manufacturing Line for Oral Solids. Press Release, Jan. 3, 2023.

ICH. Guideline Q13 on Continuous Manufacturing of Drug Substances and Drug Products, Step 5 (Jan. 6, 2023)

Jennifer Markarian is manufacturing report for 

When referring to this article, please cite it as Markarian, J. 

Implementing Process Analytical Technology

PAT is crucial to process control and real-time release in continuous manufacturing of solid-dosage drugs. 

Additive manufacturing, which is commonly called three-dimensional printing (3DP), offers the potential for manufacturing specialty dosage forms and personalized medicine in either a centralized, industrial environment or in distributed, point-of-care supply. The recent development of small-footprint, good manufacturing practice (GMP) 3D printers using various types of 3D-printing technologies further opens the possibility of portable and flexible point-of-care manufacturing, which could include in-hospital clinical trials.

Interest in 3DP technology for point-of-care supply is growing, says Thomas Kipping, head of Drug Carriers, Life Science business of Merck KGaA, Darmstadt, Germany. GMP equipment for distributed manufacturing should be simple to use, require low maintenance, and have easily cleaned, interchangeable product-contact parts, he suggests. Advances in automated quality control are especially
beneficial for GMP manufacturing. “Visual tools confirming the right geometry of the prints combined with enhanced analytical technology are already available, leading to 100% in-process control [and] resulting in higher safety for patients,” says Kipping.

New materials designed for 3DP could help advance 3DP technology adoption. “Standard polymer grades are often used as a basis for formulation development. These excipients are, in most cases, initially developed for other application fields and can only provide limited design space for developers,” Kipping points out. He says Merck KGaA is investigating the requirements for polymeric excipients and particle design for different types of 3DP technologies, including melt-drop deposition and selective laser sintering approaches.

As 3DP technologies are moving quickly from concept stage to industrial use, attention to quality controls is crucial. “Close collaboration between regulatory bodies, engineers, formulation developers, and respective quality functions is required to ensure a successful implementation,” concludes Kipping.

Developers have been refining pharmaceutical 3D printers for several years, and several types of GMP printers and systems have become available.

FabRx, created by University College London researchers to develop pharmaceutical 3D printing, launched the M3DIMAKER GMP printer in 2020 (1). The printer is designed to prepare personalized medicine in hospitals and pharmacies or to manufacture small batches for preclinical or clinical trials, says Alvaro Goyanes, CEO and co-founder of FabRx. He says that the printer is in use in several clinical trials around the world, including one in the Gustave Roussy hospital in France, a center for cancer research in Europe.

The M3DIMaker is a material extrusion type of 3D printer that melts a material and deposits it in layers on the printing bed to build up the desired structure. The machine has three different types of printheads that can be used, depending on the form of the input material. In fused deposition modeling (FDM) the API and other ingredients are combined in an extruder and formed into a filament in a separate manufacturing step, and then the filament is melted and deposited in the FDM printer. In direct powder extrusion (DPE), ingredients in powder form are fed directly into the 3D printer. Semisolid extrusion (SSE) extrudes a gel or paste from a syringe under pressure (2). SSE operates at a lower temperature than FDM or DPE, which makes it suitable for thermosensitive drugs.

GMP features of the M3DIMAKER include in-line quality control procedures and camera monitoring of the process. “These are tools that are controlled by our software to assure that each ‘printlet’ (3D-printed tablet) contains the dose that it should have. It is based on the use of balances, cameras, and near-infrared (NIR) spectroscopic methods,” explains Goyanes.

Triastek, a 3D-printing technology platform company based in China, invented a melt extrusion deposition technology, trademarked as MED, for pharmaceutical manufacturing. MED technology continuously converts powder into a softened, molten state, which is then deposited layer-by-layer to produce objects with well-designed geometric structures, explains Xiaoling Li, chief scientific officer at Triastek and professor at the University of the Pacific in California. He says it can be used to process thermosensitive APIs and print tablets at room temperature. The technology can be used to solve drug-delivery challenges. “Using MED technology, combinations of complex drug release profiles can be accessed by varied, layered structural designs and high-precision processing,” says Li. “Tablets can be precisely delivered to a specific, hard-to-reach gastrointestinal region to release APIs at the designed onset time, release rate, and specific amount.”

Two of Triastek’s own 3D-printed drug products, T19 (for rheumatoid arthritis) and T20 (for cardiovascular and clotting disorders), have received FDA approval under investigational new drug applications, and the company has started pharmacokinetic (PK) study trials with a clinical contract research organization. “The results of these pilot PK trials will help guide the formulation optimization to establish the to-be-marketed formulation. Once the formulation is finalized, Tristek will perform large-scale production for pivotal clinical studies to support NDAs [new drug applications],” explains Li.

Triastek announced a partnership with Eli Lilly in July 2022 to use 3D-printing technology to create a structure that permits targeted and programmed release of drugs in the gastrointestinal tract (3). MED has the ability to create a multi-layered tablet with controlled release of each layer. The joint research project will study how excipient properties and process parameters impact drug chemical and physical stability, which influence the release profiles of APIs.

The company’s GMP, mass production-scale MED 3D printing system uses a 32 printing-nozzle array design that is capable of manufacturing 3D printed tablets with three different materials at a production rate of approximately 132 kg/72 hours, which is similar to the production rate of manufacturing a 150-kg batch of coated tablets. The array design enables mass production using multiple materials to fabricate complex structures. “The design enables scale-up to meet the manufacturing demands of a blockbuster product, or scale-down to manufacture a smaller quantity of a rare disease product,” says Li.

The small-footprint GMP MED 3D printer can also be used for clinical-trial manufacturing and for personalized medicines. “We see a future with our technology where patients can follow directions to print their own customized tablets,” says Li.

All GMP MED equipment integrates process analytical technology (PAT) to provide digital detection of product quality related attributes, says Li. NIR is used to measure mixture homogeneity, and a photographic monitoring system identifies the physical properties of the tablets; nonconforming tablets can be automatically eliminated.

Aprecia Pharmaceuticals has launched a small-footprint 3DP technology that uses binder jetting to form 3D-printed tablets directly in blister packaging cavities. Z-Form technology expands the formulation capabilities of Aprecia’s original ZipDose technology, which is used to make the FDA-approved rapidly-disintegrating Spritam (levetiracetam) using open-bed binder jetting 3D-printing equipment. Z-Form enables formulations with multiple powder and liquid blends, multiple APIs, micro doses of API, particles engineered to enhance solubility and bioavailability, particles coated to modify the drug release profiles, and special markings for anticounterfeiting, says Kyle Smith, president and chief operating officer of Aprecia. A patent-pending containment system enables use of highly potent compounds. The Z-Form Flex printing system can produce a range of finished dose formulation options, such as orally disintegrating, sublingual, buccal, and swallow-intact formulations.

Forming the tablet directly in the blister packaging offers several benefits, one of which is the ability to eliminate material waste and produce yields of 99%, reports Smith. Another benefit is flexibility—each individual blister cavity can have a custom formulation and tablet volume/size.

“Z-Form Flex allows researchers to produce small batches of tablets with different formulations or the same base formulation in different strengths,” explains Smith. “Rapid and efficient drug prototyping enables faster entry into first-in-human studies with a close-to-commercial dose form. This [technique] enables speedier transition to proof-of-concept trials or facilitates a quicker failure to conserve valuable time, money, and resources. The flexibility to modify formulations quickly and accurately through customized, layer-by-layer precision dosing at the individual blister cavity level better supports dose ranging and
adaptive clinical trials.”

Traceability features—barcodes on the blister cards and a QR code on each cavity—facilitate both clinical development and personalized drug delivery. “This digital fingerprint enables precise chain of custody throughout the supply chain,” explains Smith.

The small footprint of the Z-Form Flex—approximately 4 ft by 10 ft—lowers costs and environmental impact of construction and operation, says Smith. In addition, the GMP equipment is designed for easy cleanability and quick changeovers of less than 8 hours for better efficiency with small-batch production. The Z-Form Flex platform is designed for a wide range of production volumes, from one blister pack up to 2000 to 3000 finished blister package units/hr, with minimal to no scale-up.

Aprecia plans to further refine the platform by scaling it down to a laboratory version (Z-Form Lab) and up to high-volume GMP commercial production (Z-Form Pro). Both systems are currently in the design phase of development, and the intent is for Z-Form Pro to match the serial output of conventional process trains.

The platform’s PAT architecture uses non-destructive, in-line analytical assessment of product and process quality attributes. PAT data will be used in product and process development, and to establish in-process controls and product specifications that will enable real-time release.

“In terms of large-scale manufacturing, Z-Form Flex sets up the potential for continuous CGMP manufacturing and real-time release,” contends Smith. “Many of the discrete steps involved in conventional manufacturing, such as granulation, drying, and compression/encapsulation, are now contained within a small, compact footprint and occur within the blister cavity. Deployment of a revolutionary planar motion system enables all units of operation to happen simultaneously, so 3D-tablets are formed within three minutes. While drying times are product-dependent, the primary package is often completed within 60 minutes from start to finish, including final inspection and blister card lidding.”

“Aprecia is already performing feasibility studies with pharmaceutical companies using development-scale versions of the Z-Form Flex machine. We are now reaching an inflection point of adoption for our technology,” concludes Chris Gilmore, CEO of Aprecia. He says that Aprecia continues to search out companies with the “appetite and vision to be market disruptors”.

Gilmore believes the ongoing platform innovations at Aprecia are quickly advancing small-scale customization and personalization while also
establishing a solid foundation for addressing unmet need in large-scale, mass production.

1. FabRx, “FabRx’s Pharmaceutical 3D Printer for Personalised Medicines,” Press Release, April 6, 2020.

2. FabRx, “Semisolid Extrusion: A Disruptive Technology for the Medical and
Pharmaceutical Fields,” Press Release, April 1, 2021.

3. Triastek, “Triastek Announces Research Collaboration with Lilly to Explore the Application of 3D Printing Technology in Oral Delivery of Drugs,” Press Release, July 13, 2022. 


Volume 46, Number 12
December 2022
Pages: 24-27

When referring to this article, please cite it as J. Markarian, “Moving Forward with GMP 3D Printing for Drug Products," 

Unique dosage forms, personalized medicine, and flexible manufacturing are possible with 3DP. 

Moving Forward with GMP 3D Printing for Drug Products

Many areas of pharmaceutical and biopharmaceutical manufacturing—solid-dosage drugs, vaccines, cell and gene therapies, viral vectors, and others—are facing demands for faster time-to-market and the ability to have more flexibility, with faster changeovers from one product to the next, to match the shift to smaller batch sizes for more personalized medicine. Advances in equipment and data analytics technologies, along with digital tools and risk-based methods for qualification and validation of equipment and processes, are helping meet this need for speed.

“Industry 4.0” technologies, which collect large volumes of data from sensors installed on equipment and use these data in data analytics software to optimize equipment and processes, continue to penetrate pharma manufacturing. Equipment and facility control systems can be designed for connection using standard protocols, providing better communication between control systems.

“The module type package (MTP) communication standard provides a framework of standardized connectivity between OEM [original equipment manufacturer] equipment and centralized control systems to streamline connectivity and interoperability,” says Michalle Adkins, director of life sciences consulting at Emerson. “For example, it can be used to enable much easier integration between a programmable logic controller used to control OEM equipment, such as a deionized water skid, and a distributed control system.” Easier integration enables flexible manufacturing. “Operations teams can use MTP to unlock modularity, using different equipment on demand, with MTP automating and simplifying the integration process,” says Adkins. “And users of these OEM packages can tie multiple systems together without the need to re-engineer solutions, making new equipment less costly to instal, validate, and maintain.”

“Plug-and-produce initiatives are driving towards increased data availability and accessibility, where data is interfaced and contextualized in a smarter way to embrace data integrity and analytics,” says Petter Mörée, managing director for EMEA and industry principal for Life Sciences and Pharmaceuticals at Seeq. Data analytics enable predictive or condition-based maintenance, which improves equipment reliability by reducing unplanned downtime. Data analytics also drive improved time-to-market, quality, and product and equipment availability, continues Mörée. Digital asset models (sometimes called digital twins) can be used to monitor and predict equipment functions, so that overall equipment efficiency in the facility can be improved. For example, a model of process chromatography can be used to predict when a chromatography column needs to be repacked.

More data and their analysis lead to increased knowledge and improved risk assessment. Use of data analytics, along with first principles, prior experience, and risk assessment, enhances the ability of process experts to improve operations and to perform qualification and continued process verification, adds Mörée.

When travel was restricted during the pandemic, many equipment vendors and users pivoted to using video and web conferencing tools for activities such as factory acceptance testing (FAT). Companies are continuing their use of remote FAT even when travel is possible because it speeds qualification and eliminates duplication, for various types of equipment, including the automation systems.

“Instead of testing systems at the supplier’s site and then testing again when equipment is installed at the plant, teams are eliminating duplicate work by using cloud-based FAT tools to remotely connect to equipment at the supplier, run tests using the actual system configuration, and generate documentation,” says Adkins. “Once testing is complete and equipment is on site, the team can use those same cloud tools to move the configuration from the development system to the production system.”

In addition, simulation technologies, such as digital twins and augmented reality, are being successfully used to speed timelines for equipment installation and counteract the effect of long lead times and supply-chain delays, contends Adkins. “Fit-for-purpose simulations allow engineers to test before the equipment is installed, and operators to train as though they were working with real equipment. Using these tools, teams can ensure their processes work properly and their people are ready even as they wait for equipment to arrive, taking training and testing off the critical path,” she explains. “Emerson customers are often surprised by the amount of time they save by using cloud technologies and simulation to support some of their qualification steps remotely. The ability to perform stages in parallel not only saves time, but it also contributes to lower costs by addressing problems earlier in the project, bringing new manufacturing lines online more quickly, while eliminating waste.”

Digital tools also save time by eliminating the time-consuming aspects of paperwork associated with input/output (I/O) designs and verification of connection testing. Modern control systems provide flexibility with automated “I/O on demand,” emphasizes Adkins. “These systems typically have interchangeable terminal blocks installed in the field and are connected to the control system via a single Ethernet cable. I/O is automatically mapped to the control system and field wiring of any signal type can be terminated on any I/O module. This simplifies wiring documentation and eliminates hundreds of hours labelling wire terminations and verifying cabinet drawings. Teams can move right to loop testing without spending significant amounts of time on documentation and testing because complex wiring and mapping are eliminated.”

Class-based systems can streamline validation of equipment. In automation systems, class-based control creates unit classes containing all the operations associated with a single unit, says Adkins. “Teams test the class-based unit once, and then only need to test any minor differences among units, instead of testing each unit individually,” she explains.

OEM equipment with integrated control systems can also reduce validation requirements. “Commercial off-the-shelf (COTS) control systems that are integrated using libraries of firmware-based objects are generally considered to be a configured product, rather than a solution based on custom code,” says John Hatzis, life sciences industry consultant at Rockwell Automation. “More OEM equipment is being built with COTS control systems as a standard offering. This reduces the validation scope of the producer since the process control comes validation ready from the OEM. Producers can get a bioreactor installed, commissioned, and validated for use in production in as little as six months. This OEM equipment is designed to fit into several processes with minimal reconfiguration.”

The industry has been increasingly moving toward risk assessment and risk management in validation, directed by International Council for Harmonisation (ICH) guidelines and, specifically for equipment, ASTM E2500 (1). Recent International Society for Pharmaceutical Engineering (ISPE) good practice guides, such as the second edition of the 

 (2) and the second edition of the Commissioning and Qualification baseline guide (3), for example, include risk management concepts.

“Recent regulatory changes focus on an increasing practice of risk assessment and risk management while reducing reliance on strictly worded guidance for any given topic,” suggests Moria Feighery-Ross, senior validation engineer at Pharmatech Associates. The new revision to the European Union’s Annex 1 (4), for example, “…demonstrates a reduction in prescriptive practices and an increased emphasis on risk assessment and sound scientific justification for how subjects are addressed throughout an organization’s quality management system,” she says.

Protocol templates that use a one-size-fits-all approach and don’t consider risk should be avoided, says Feighery-Ross. “For example, not every HVAC [heating, ventilation, and air conditioning] system is the same. Single-use aseptic processes are frequently performed in isolators and closed cartridge systems that reduce the reliance on the environmental quality. Therefore, HVAC in a facility with closed systems, [such as a personalized cell and gene therapy facility], should not be qualified using a template designed for an oral solid dose facility that uses open processing,” she explains.

Manufacturing expansion puts demands on validation

One example of an area that has benefited from risk-based validation is viral vector manufacturing. Manufacturers have been expanding viral vector production capacity to meet the industry’s needs that have been driven by rapid growth in cell and gene therapies and in use of viral vectors for large-scale vaccine production (5). Subsequently, the need for commissioning and validation of new or renovated manufacturing facilities has also increased, says Wai Wong, vice-president of validation at Pharmatech Associates.

Viral vector facilities typically use lab-scale equipment because batches are smaller in size, scaling out to multiple lines, he explains. Utility requirements are often less complex compared to traditional pharma or biopharma facilities. “Requirements for process gasses and purified water or water-for-injection are minimal, so many companies can rely upon bottled water rather than implementing skid-based utility systems,” says Wong. With these differences, validation should also be different. “Validating this new type of facility requires a nimble and flexible approach,” says Wong. At the same time, there have been diminishing resources of personnel, equipment, and raw materials. “Moving to a risk-based approach to validation based on critical quality attributes for the facility, utilities, and equipment is essential to meet demand,” he concludes.

“The trend toward risk-based validation reduces the CQV [commissioning, qualification, and validation] burden as requirements are based on specific attributes for the product rather than a broad brushstroke of requirements that apply to all types of facilities, systems, and utilities,” says Wong. “Risk-based validation allows us to focus our qualification activities on areas that provide the most impact to product quality rather than on unnecessary tests that are unlikely to affect critical attributes.”

Flexibility is crucial for contract development and manufacturing organizations (CDMOs) because they must accommodate a wide variety of products and processes. “A significant aspect of flexibility within an operation such as Catalent is segregation and multi-unit configuration, which allows multiple products to be processed simultaneously, in the same facilities, and often in the same production unit,” says John Machulski, vice-president of engineering, biologics, and gene therapy at Catalent. “An example would be segregating formulation from filling, or the ability to handle multiple product configurations [e.g., vials, syringes, or cartridges in various sizes] in one filling area.”

Catalent was one of the winners of ISPE’s 2022 Facility of the Year Social Impact Category for its Project Mercury, which rapidly transformed a shell space at its Bloomington, Ind. manufacturing site into flexible vaccine manufacturing capacity during the pandemic. Speed of validation was crucial, because every day made a difference in the ramp-up process to meet demand for COVID-19 vaccines and therapies, explains Machulski. Risk-based approaches, supported by documented risk analyses, were a key aspect of making the validation process efficient. A second key was leveraging prior work. The project referenced activities performed earlier in the project lifecycle (e.g., FAT and site acceptance testing) in later qualification phases, to avoid the need for repetition. A third key was using a family approach to equipment validation, which leveraged the work done on one piece of equipment in the validation of similar or identical equipment, again supported by risk analyses, explains Machulski.

Risk-based approaches can also be used in changeover between products in an existing facility. “Family validation approaches allow quickly qualified changeovers, reducing down-time, which, in turn, increases efficiency and the contribution made by that equipment,” concludes Machulski.

ASTM E2500-20: Standard Guide for Specification, Design, and Verification of Pharmaceutical and Biopharmaceutical Manufacturing Systems and Equipment

Good Practice Guide: Good Engineering Practice

Baseline Guide Vol. 5: Commissioning and Qualification,

4. EU, Annex 1: Manufacture of Sterile Medicinal Products (August 2022).

Jennifer Markarian is manufacturing editor for 


Volume 46, Number 11
November 2022
Pages:

When referring to this article, please cite it as J. Markarian, “Streamlining Equipment Quality and Flexibility,” 

Data analytics, modular equipment, digital tools, and risk-based validation improve speed, flexibility, and quality. 

Streamlining Equipment Quality and Flexibility

Extractables and leachables (E&L) must be analyzed for any product-contact material to determine the risk of leachable compounds contaminating a drug product. Container-closure systems, which contact the drug product during long-term storage, are often higher risk than process equipment, which has a short product-contact time. Processing equipment is not exempt, however. As polymer single-use systems (SUS) are increasingly being used in 
biopharmaceutical manufacturing, there is a growing need to analyze E&L of this equipment. Industry organizations and consortiums have worked
to standardize analysis protocols that can be used throughout the supply chain. Standard methods facilitate comparison between suppliers.

BioPhorum, for example, first published a standardized extractables protocol in 2014, which was updated in 2020 (1). Because the organization was previously known as BioPhorum Operations Group (BPOG), the protocol is often called the BPOG protocol. SUS suppliers can use the protocol to measure the extractables in representative components. The purpose of the extractables testing information package, says BioPhorum, is “to allow the SUS end-user to rigorously estimate the types and amounts of leachables that could be generated by the SUS component during its intended use” (1).

Another set of standard methods was recently published in the 

General Chapter <665> Plastic Components and Systems Used to Manufacture Pharmaceutical Drug Products and Biopharmaceutical Drug Substances and Products was published in 2022 (2), but the date for it to come into effect was postponed to 2026. 

 accompanying procedural chapter <1665> became official on May 1, 2022 (3).

 documents is the use of risk assessment for end-users to determine the risk level of a component in a particular process. The analysis method depends on this risk level.

The International Council for Harmonisation (ICH) is working on a document related to E&L testing. A final concept paper for ICH Q3E was published in 2020 (4). Industry experts say that questions about how this upcoming ICH guidance will align with the USP and BioPhorum documents are yet to be answered.

The Extractables and Leachables Safety Information Exchange (ELSIE) Consortium, made up of 33 biopharma/medical device companies, is also active in this area and aims to provide expert input to efforts such as 

new chapter and the in-progress ICH Q3E, through working groups and formal commenting processes as well as white papers, workshops, and other public meetings. “ELSIE provides a collaborative environment where member companies can discuss and benchmark the approaches that companies are taking,” explains Lee Nagao, senior director of science, regulation, and policy at the ELSIE Secretariat. Harmonized guidelines or compendial chapters are crucial, because without them, “regulatory expectations can vary depending on the health authority and the experience of the specific [biologics license application] reviewer,” says Nagao.

Standardized protocols help in setting baselines and comparing differences. For end-users of SUS, standard protocols facilitate “like-for-like assessments between each supplier’s single-use technology (SUT)
products and components,” says Dan Rosen, vice-president and general manager of bioproduction at Thermo Fisher Scientific.

Extractables testing of the materials or components is used to identify which leachables might come from the material under standard conditions and in a simulation study. A challenge, however, is that different leachables could be formed during sterilization, storage, or use.

“Key challenges in characterizing E&Ls include identifying unknown molecules that arise during primary sterilization of the SUT products,” says Rosen.

The response of materials to commonly used gamma irradiation is generally well established, but due to limitations in capacity for gamma-radiation sterilization, X-ray sterilization is being considered as an alternative (5). Current theories suggest that the E&L profile should be similar after X-ray or gamma sterilization, but it is important to perform extractables testing, says the Bio-Process Systems Alliance (BPSA).

BPSA published a guide for E&L in cell and gene therapies (CGT) in July 2020 (6) to support companies developing manufacturing lines for scale up or scale out. “Since CGT was a new modality, the drug developers were almost exclusively using single-use components for economic, safety, and technical performance. This [situation] led to a need for BPSA to help coordinate a guide to assist developers on appropriate choice of materials and components for their process development,” says Brendan Lucey, chair of BPSA’s CGT Committee. CGT processes differ from other biopharma manufacturing using SUS because in classical biopharma, process-related impurities, including process-equipment related leachables, are removed in the downstream purification process. In CGT processes, however, there may be washing steps, but not the same kind of downstream purification. Because of this process set-up and the variety of contact surfaces and contact times, E&L testing to assess and then control leachables is crucial, concluded BPSA (6). The type of polymers used in SUS for CGT manufacturing also differs from classical biopharmaceutical manufacturing. In particular, the report noted that polyvinyl chloride (PVC) polymer may be used in transfer bags and tubing sets in CGT collection, processing, and storage. Because flexible PVC typically contains plasticizers and other additives that are not present in other polymers used in SUS, leachables from these additives should be considered (6).

The rapid pace of change in guidelines has required a quick response by contract labs to create experiment designs that are compliant, says Dujuan Lu, manager and global lead of extractable and leachable testing at SGS Health Science. SGS labs are following the 

 <1665> procedures, because these have been official since May 2022, but the postponement of 

 <665>’s official date is helpful for the industry, Lu says. “The transition period gives pharma/biopharma manufacturers time to requalify as needed. We are seeing lots of activity now because clients want to be ready for 2026,” she explains.

 and BioPhorum protocols, says Lu, who explains that the two differ slightly in recommended solvents for acid and base conditions. Lu points out that 

 considers the different solvents for acid conditions to be equivalent; for the base conditions, 

 recommends choosing the alkaline solvent depending on the specific process pH value.

Design of extraction studies for single-use components is complicated by multi-component assemblies and complex geometries. “The recommended ratio of surface area to volume is 6 cm2/mL. Calculating the surface area is straightforward for a tube or a bag, but can be challenging for a connector or valve, for example,” explains Lu. In designing an extraction study for single-use filters, both the housing and the membrane must be evaluated. “The procedure depends on the shape. For a capsule filter, for example, we can fill the filter with solvent and cap the ends. For a smaller filter, we might need to remove the membrane and submerge it in solvent,” explains Lu.

Because many different polymeric materials are used to construct single-use components, the extractable profile is complex and multiple analytical techniques are needed to identify and quantify the compounds, says Lu. She explains that volatile organic compounds can be measured with headspace gas chromatography (HS-GC) coupled with high-resolution mass spectrometry (HRMS) or flame ionization detection (FID). Semi-volatile organic compounds can be measured with direct sample injection in GC-HRMS/FID. Non-volatile organic compounds can be measured with ultra high-performance liquid chromatography (UHPLC) coupled with HRMS and a photodiode-array detector. “Unknown identifications are crucial for the extractables testing, so high-resolution mass spectrometers are recommended for both GC and LC testing,” says Lu. She adds that elemental impurities can be measured with inductively coupled plasma-mass spectrometry.

Calculating the analytical evaluation threshold (AET) for processing equipment is another challenge. “The AET is set up based on the toxicology concerns; above [the AET], analytes need to be identified and quantified for health risk assessment,” explains Lu. “The instrument detection limits need to be at, or lower than, the AET to make sure all compounds above the AET can be detected. The process equipment AET is typically calculated based on the final product AET by applying the dilution factor during the process. The difference between the batch size and the extraction stoichiometry during the extractable study will be considered during the calculation.”

To perform risk assessment, it is crucial to have a good understanding of the manufacturing process. “[Understanding] any clearance, concentration, or dilution steps [is] critical, particularly to help narrow down the number of materials requiring in-depth characterization,” says Nagao.

Questions remain, says Nagao, about what point in the process a risk assessment should begin, how to define component risk, and what the minimum data would be to qualify low-risk components. “For example, gaskets, seals, or connectors used upstream of a [filtration] step will be inherently low risk and unlikely to contribute significantly to the leachables profile. Is it necessary, then, to fully characterize the extractables profile? This [type of question] is where a harmonized guidance that is implemented across the globe, perhaps based on risk-management principles, would be helpful,” Nagao suggests.

BioPhorum Best Practices Guide for Extractables Testing of Polymeric Single-Use Components Used in Biopharmaceutical Manufacturing 

2. USP, General Chapter <665>, “Plastic Components and Systems Used to Manufacture Pharmaceutical Drug Products and Biopharmaceutical Drug Substances and Products,” 

3. USP, General Chapter <1665>, “Characterization and Qualification of Plastic Components and Systems Used to Manufacture Pharmaceutical Drug Products and Biopharmaceutical Drug Substances and Products,” 

Final Concept Paper ICH Q3E: Guideline for Extractables and Leachables

X-Ray Sterilization of Single-Use Bioprocess Equipment, Part 1: Industry Need, Requirements, and Risk Evaluation

Extractables/Leachables Considerations for Cell & Gene Therapy Drug Production Development

Single-use systems can benefit from standardized risk assessment and analysis protocols, which can facilitate the way in which processing equipment components are compared.