Agnes Shanley is senior editor of Pharmaceutical Technology.

Agnes Shanley

Keeping pharmaceutical facilities clean and germ free seems an obvious prerequisite for any company that manufactures medicines. However, regulatory citations suggest that pharmaceutical manufacturers and contract manufacturing and
development organizations (CDMOs) may take cleaning and cleaning validation for granted.

An April 2021 Form 483 issued by FDA to the Maryland-based CDMO, Emergent Biosolutions, which is under contract to Johnson & Johnson (J&J) and AstraZeneca to manufacture their COVID-19 vaccines, showed how deeply rooted this problem may be (1). J&J and AstraZeneca use very different manufacturing processes, but both involve the use of viral vectors.

FDA inspectors found fundamental deviations from good manufacturing practices (GMPs) at the facility. Among the observations noted were failure to maintain clean and sanitary conditions; lack of written procedures for cleaning and decontamination of the vessels used to transport and store materials at the site; and no formal requirements to clean and sanitize these vessels after each use.

In addition, no tests were conducted to determine how long viruses used in a client’s production process remained viable on equipment surfaces in one area within the facility.

If these omissions seem basic, it might be helpful to remember that in pre-pandemic 2019, the last full year when FDA performed routine site inspections, roughly 60% of FDA 483 regulatory observations involved deficiencies in cleaning and sanitization and their validation.

This fact was presented by Matt Cokely, senior global technical consultant manager for Ecolab Life Sciences in a best-practices webinar presented by the International Society for Pharmaceutical Engineering (ISPE) on April 9, 2021 (2). The most often cited problems, he said, were the lack of procedures for preventing contamination, deficiencies in cleaning, failure to keep cleaning records, and inadequate standard operating procedures (SOPs).

Cleaning validation consultant Rich Forsyth agrees. “Problems often boil down to a failure to follow cleaning procedures. Either people are doing something that is not in the SOP or there’s something in the SOP that they aren’t doing,” he says. For example, Forsyth notes, many procedures fail to specify cleaning frequency, or the amount of time that should be spent to clean a piece of equipment using documented procedures and materials. This forces operators to take the unprovable position that ‘it’s not clean until it’s clean’, he says.

Cleaning and cleaning validation practices need to be reproducible, Cokely noted, and employees must be trained, not only in the basics of microbiology and hygiene pertaining to cleaning, but in cleanroom procedures and gowning and in understanding potential risks and safety hazards. “They need to understand not just how, but why they are performing a task, and to be trained to actually perform it,” he said, rather than simply being given a pile of SOPs to read and sign off on. Given the trend to outsource cleaning and sanitization activities, Cokely noted, this training must be given to contract cleaning workers as well.

Elaborating further, Cokely noted that written instructions in SOPs should specify who is responsible for the duties, and set a schedule of frequency for cleaning, describe the materials and techniques to be used in detail, and the preparation and storage of cleaning materials. In addition, documentation must include the names and signatures of staff performing the work, as well as the product batch number, cleaner expiry date, and disinfectants used. Cokely also noted that disinfectants with different modes of action should be rotated. Care must also be taken, he said, to avoid interactions between different cleaning chemistries that might inhibit microbe-killing mechanisms and promote the buildup of cleaning solution residues.

The basic problem, Cokely noted on the webcast, is that cleaning and disinfection are undervalued in the pharmaceutical industry, despite their importance to patient and overall safety. “At some companies, cleaning procedures have been used for years, or the facility may have inherited practices that have never been reviewed or scrutinized,” he said.

Regulations for cleaning and its validation abound, supplemented by best practices circulated by ISPE, the Parenteral Drug Association (PDA), and now the American Society for Testing and Materials (ASTM), which has set standards for pharmaceutical cleaning operations and validation. Risk-based standards for cleaning validation, set in 2018 (3) began to establish a more scientific basis for practices at individual facilities.

These standards have had a positive impact, notes Forsyth, getting pharmaceutical companies to go beyond thinking simply about products, equipment, and detergents in their cleaning operations. “Risk-based standards forced people to formalize their risk assessments. Many of them began to consider peripheral risks that they hadn’t factored in before, such as the air-handling system’s impact or the material and equipment traffic in the facility,” he says.

However, in the end, some companies may view having a formal cleaning validation program and strategy as involving too much work. “Companies pay a lot of attention to their cleaning and cleaning validation programs when they get cited by regulators for those programs’ failings,” Forsythe says. But establishing a program or updating one that hasn’t been reviewed in several years, takes commitment, he notes, and it generally requires a year to do because everyone has their regular jobs to do, too. “Often, shepherding a cleaning validation project through to its end becomes more problematic the longer the projects continue. At many companies the success or failure of a cleaning and cleaning validation program may rest on one champion supporting it,” Forsyth says.

Today, multiproduct facilities, and particularly biopharmaceutical facilities, pose several cleaning validation challenges, Forsyth notes. “You worry more about bioburden in a biological plant than in a small-molecule facility, but generally if you have an effective cleaning validation program, you’re fine,” he says.

However, setting cleaning limits can be tricky because traditional methods fail to consider some of the most significant differences between biopharma and small-molecule facilities. “The most notable difference between the two types of plants is that biopharmaceutical facilities start processes using equipment with large surface area but generate small batches, both of which require the use of lower cleaning limits, so if you use the entire equipment train, your limits may, at first, appear to make no sense,” Forsyth says. Single-use systems and equipment have eliminated some of those problems, but they persist for hybrid plants still using stainless steel equipment.

Another question is how batches are finished and whether pre-cleaning equipment preparation processes are consistent from one batch to another. If they are not, varying amounts of product may be left on equipment surfaces after each run. “Some types of equipment will empty completely, but others won’t,” Forsyth says.

He recalls one of his first experiences working on cleaning validation in a pharmaceutical development lab. “We were concerned about pouring active ingredients down the drain, so we wound up scraping equipment and then vacuuming it before cleaning. By the time we started cleaning that equipment, it was almost visually clean,” he says, noting that this type of approach reduces the risk of cleaning failures and compliance problems. At some companies, varying amounts of material may be left on the equipment when cleaning officially begins.

CDMOs that focus on clinical trial drug manufacturing face unique challenges in setting cleaning limits. “For clinical operations, you are not making a lot of batches of the same thing, so you don’t make enough batches of any single product to validate the cleaning methods using standard approaches,” says Forsyth. “In this case, it is usually best to work with a more general model, establish a worst case, and then evaluate each new product coming in. However, this approach increases the effort and justification required to maintain the model, and a lot of CDMOs don’t do this very well,” says Forsyth.

Specific approaches are needed to prevent cross-contamination in multiproduct facilities, Forsyth says. “Best practices call for calculating the amount of residue that can safely be carried over for every product. For a multiproduct facility that makes 10 products, that means calculating the limit of one product versus those of the other nine, taking the lowest number and then using that, and then following the same procedure for the rest of the products,” he says.

ASTM standards have been focusing on use of statistics, which are important after a cleaning program has been validated and needs to be maintained, says Forsyth. Maintenance requires trending data. “The basic statistics needed for trending are fairly straightforward, and more advanced statistics can also be applied, but it has been difficult to motivate some companies to do that in the current environment,” he says.

More work is also being done to use analytical methods to reduce the subjectivity of visual inspection, but Forsyth hasn’t seen any consistently effective methods coming into general use. The best way to remove subjectivity from inspections has been to qualify those who must perform those inspections, a step that most regulatory agencies now require.

“I advise clients to set visual limits for residues and have staffers who might be assigned to perform visual inspections go through specialized training to recognize clearly what various levels of residues look like on equipment, compared with images of surfaces with residues that meet safe carryover limits. It’s important to have exercises in place that will allow people to visually recognize cleanliness levels to avoid patient safety risks,” Forsyth says.

Cleaning is not a popular activity, Cokely noted in his presentation, but it is essential. Taking cleaning and sanitization validation programs for granted only puts facilities and people at risk and increases the likelihood of costly noncompliance.

1. FDA, Form 483, Emergent Biosolutions, 

, April 20, 2021.
2. ISPE, “Regulatory Requirements and Expectations for Cleaning and Disinfection of Controlled Manufacturing Areas,” Webcast, April 9, 2021.
3. ASTM, E3106-18e1, Standard Guide for Science-Based and Risk-Based Cleaning Process Development and Validation (2018). 

Agnes Shanley is the former senior editor for 


Vol. 45, No. 6
June 2021
Pages: 44–46, 52

When referring to this article, please cite it as A. Shanley, “Prioritizing Cleaning Validation,” 

Articles

With COVID-19 vaccine facilities receiving recent citations, it is evident that failure to meet cleaning and sanitization requirements puts patients, facilities, and operators at risk. 

The Necessity of Prioritizing  Cleaning Validation

As recent COVID-19 vaccine facility citations make clear, failure to meet cleaning and sanitization requirements puts patients, facilities, and operators at risk. 

Prioritizing  Cleaning Validation

The past year has been a busy one for vendors and contract development and manufacturing organizations (CDMOs) that provide lyophilization equipment and services. On the bulk lyophilization side, demand for lyophilization for antibody-drug conjugates (ADCs), monoclonal antibody therapies (mAbs) for COVID-19, and some of the raw materials used in COVID-19 vaccines has been growing, says Scott Ross, global product specialist for bulk lyophilization with W.L. Gore and Associates.

The company’s Lyoguard trays are being used in the production of mAbs; diagnostics tests for COVID-19 virus or antibodies; and some of the key raw materials used in messenger RNA (mRNA) vaccines, including synthetic oligonucleotides, adjuvants, and lipid nanoparticles, he says. “For ADCs, lyophilization is likely to become a de facto standard,” Ross says, noting that operator safety concerns and the need to reduce turnaround times at CDMOs are driving use of contained tray systems. In addition, real-time monitoring of temperature and other parameters is becoming routine, he says.

As it has lifted demand for bulk lyophilization, the COVID-19 pandemic has also highlighted deficiencies with vial-based lyophilization, a process that has seen few changes in the past 50 years. By offering a powdered form of vaccine that could be shipped at ambient temperatures and reconstituted at the point of administration, lyophilization could have simplified the distribution of new vaccines.

This is especially true for COVID-19 vaccines based on mRNA, which must be transported and stored as frozen liquids at extremely low temperatures, then mixed and manipulated before patients can be injected. Of the five COVID-19 vaccines that are now being used globally to prevent COVID-19, only Russia’s Sputnik V—which also comes in frozen liquid form—actually uses traditional vial freeze drying.

In the end, it’s a question of capacity. It would take three years for the industry to have enough capacity to lyophilize 300 million vials of product, says Alina Alexeenko, professor of chemical engineering, aeronautics, and astronautics at Purdue University and cochair of LyoHUB, a consortium of users and equipment and services vendors which was established four years ago to promote best practices and establish a roadmap for the future of this technology.

Fellow co-chair and Purdue professor Elizabeth Topp, who is also chief science officer at the Dublin, Ireland-based National Institute of Biopharmaceutical Research and Training (NIBRT), agrees.“Global lyophilization capacity wouldn’t allow us to process a billion vaccines a year, and stability testing is another hurdle. If you’re pushing a vaccine out the door in nine months for emergency use authorization, you clearly won’t have time to do a shelf stability study on a lyophilized formulation and guarantee 18 months stability,” she says.

However, new analytical methods and process development strategies promise to improve processing and simplify distribution, administration, and the cold chain. “We can expect to see more lyophilized vaccines in the future,” says Steve Nail, senior LyoHUB scientific advisor who was formerly a Purdue professor and is now principal scientist at Baxter Biopharmaceutical Solutions, a CDMO in Bloomington, Ind.However, he says, the equipment used to process the materials will look quite different from the traditional stainless steel lyophilizers, with heat-transfer fluid circulating through shelves.

“The technology is up for disruption,” says Alexeenko. She notes that innovative drying processes (e.g., spray- freeze drying [both continuous and non-continuous]; spin freezing; thin- film freezing; and microwave drying), show promise to simplify the processing, transportation, and distribution of biopharmaceuticals and vaccines.

Lyophilization poses challenges, particularly for some proteins. A number of stresses are associated with freezing and freeze drying, as well as with agitation and phase interfaces, says Nail.Another problem is the fact that lyophilization involves very slow cooling, at a rate of one degree Kelvin per second, says Bill Williams, a professor at the University of Texas and inventor of the thin-film freezing process, which he developed years ago at the Dow Chemical Co.

TFF Pharma licensed his technology and commercialized it in 2019. Williams and his team, with corporate and US government funding, are now focusing on research designed to optimize use of thin-film freezing to improve the processing and delivery of biologics, including vaccines, along with the cold chain. Some of the problems that can occur with lyophilization are due to ingredients within vaccine formulations. As Williams notes, 80% of all existing vaccines use aluminum hydroxide adjuvants. If these ingredients freeze and then melt, particles can then aggregate, affecting both vaccine immunogenicity and activity, he says.

However, Williams notes, slow freezing is another challenge with lyophilization that affects vaccine activity. “When the freezing rate is 100 to 1000 times faster than it is for traditional lyophilization, as it is in thin-film freezing, the integrity of protein materials is maintained,” he says.

Using thin-film freezing, the vaccine can either be reconstituted at the point of use, or administered via nasal or pulmonary delivery, he says. Currently, Williams’ lab is working with TFF Pharma and more than 25 different pharma companies on research that is formulating liquid mAbs, mRNA lipid nanoparticles, and other bioformulations into thin-film powders that can be delivered to the lungs. The company recently worked on a reformulation of remdesivir as an inhaled dry powder, and in July 2020, completed Phase I of a clinical study that studied formulation of voriconazole as an inhaled powder.

Protein destabilization is another issue with lyophilization. Although research has studied the mechanics of chemical destabilization, which occurs as the result of reactions such as deamidation, oxidation, and chain cleavage or “clipping,” the science behind physical destabilization processes such as aggregation is not yet fully understood, says Topp. “Aggregation is hard to control because it is part of the protein folding problem. We don’t really understand how proteins fold, and we don’t understand how they aggregate either,” she says.

Solid-state hydrogen deuterium exchange (HDX), an analytical method that is being used in research at Purdue and NIBRT, allows proteins to be studied in the solid state at high resolution, and is improving understanding of protein stability, says Topp.“A lot of what seems to drive the performance of these materials has to do with the hydrogen bonds that help make up the structure of the proteins, and the hydrogen bonds between the protein and the matrix that surrounds it. HDX interrogates the hydrogen bond network, and, so far, results suggest that the better the network, the more stable the material,” she says.

Evidence suggests that aggregation may be driven by partial protein unfolding. In the future, Topp says, this research might yield a metric that would reveal how well folded each individual protein is, which could help determine whether a particular biopharmaceutical formulation will perform well or not, whether it is made bylyophilization or another process.

In addition, Topp notes, research suggests that measurements made with HDX correlate closely with the stability of the protein in storage. “If we can characterize these materials and the environment that the protein sees with sufficient resolution, then maybe we can start to replace time consuming 18-month to two-year stability studies with advanced analytical characterization,” she says.

This approach would likely start with comparability protocols and then move to other parts of regulatory filings. “Stability studies impede all kinds of process changes and improvements. We hope that we can put tools in place to change that,” Topp says.

Not only HDX but nuclear magnetic resonance (NMR) spectroscopy and neutron back scattering can be advanced to study stability. “In order for progress to be made, companies need to adopt newer technologies and include them in regulatory filings, to put the technology in front of regulators and the industry,” she says.

In a lighter vein, Topp envisions a “solid-state protein characterization smackdown,” in which the three methods would be used to analyze a blinded sample of a lyophilized protein such as an mAb, and results then compared. The method or combination of methods that best correlated with storage stability would then be deemed the winner(s).

On a smaller scale, similar work evaluating solid-state HDX has already been done with Roche Genentech. In this case, however, HDX measurements were taken for blinded samples of four different mAb formulations, and the results were then compared to results of 2.5 years of stability data that had already been gathered for the same formulations.In two weeks, the HDX came up with the same rank order of results as stability testing, Topp says. “HDX measurement metrics showed nearly perfect correlation with aggregation during mAb storage,” she recalls.

Purdue and NIBRT have also been working on projects supported by the National institute for Innovation in Manufacturing Biopharmaceuticals (NIIMBL) that use HDX. This research has evaluated the impact of process changes as well as the use of alternative drying methods on stability.

LyoHUB was never intended to be a research consortium. Its first goal was technology road mapping (i.e., discerning ways to revitalize the technology), and another primary activity has been disseminating best practices to improve use and design of lyophilization equipment. “The fact is that there is a lot of poor practice going on in the industry, due to a lack of understanding of the fundamentals of heat and mass transfer,” says Nail.

LyoHUB’s latest best practices paper focuses on the formulation of lyophilized
products, but validation and equipment qualification are areas that will be updated soon, says Alexeenko. LyoHUB is actively facilitating research and the adoption of new technologies and advanced analytics and equipment among its 25 member companies. To advance this goal, the organization opened up a new demonstration laboratory in 2019 to allow its members to work with new process equipment and analytical systems.

Process analytical technology (PAT) remains an area of focus. One technology being developed is smart vials, wireless sensors that would go into each individual vial being lyophilized and measure such data as temperature, pressure, and extract lyophilization rate, and communicate the information back to the freeze drier, says Alexeenko.

“This is important because, in lyophilization, batches are qualified, and a whole batch can be trashed if something goes wrong in just one or two vials. This sensor allows individual vials to be monitored, providing more control,” says Topp. The vials can also be used to measure and control ice nucleation and to understand its dynamics. “This is the one area in lyophilization where things happen very quickly, and we need to better understand and control that process,” Alexeenko says. Another method that is being studied in projects at LyoHUB’s new demonstration facility is in-situ residual gas analyzer mass spectroscopy, a technology that Purdue has been using since 2018. The method enables the study of problems that can occur in freeze drying when components in a formulation decompose and a pH shift or other issues result.It can detect parts-per-millon-level changes in the composition of the gas during lyophilization, and then allow them to be correlated to process changes.

The first applications of the technology used ammonia, but Nail is working on a project that uses organic co-solvents, typically required when processing poorly soluble drugs. This work will study the time course release of ethanol vapor from product, which has been shown to have a critical influence on product reject rates.

Other projects are focusing on improving sensors and process modeling. “You can’t do one well without the other,” says Alexeenko. In process modeling, LyoHUB members are working on a virtual thermocouple, which uses high speed computing and high-fidelity methods to model heat transfer, not only in porous materials but also in and between glass containers, and between glass and plastics containers (e.g., SiO2-type vials). The project combines noninvasive measurement and process modeling so that the temperature within the vial can be reconstructed with a much higher level of detail, says Alexeenko.A new area of interest is microwave heating using volumetric electromagnetic rather than conductive heat transfer. So far, the method has been found to reduce impact on large biomolecules.

Nail notes increased industry interest in applying lyophilization technology to human stem cells. In 2020, Alexeenko says, Purdue researchers developed lyophilization processes for use with the mRNA and CRISPR Cas9 reagents that are used in some COVID-19 diagnostic testing kits.

When referring to this article, please cite it as A. Shanley, “Beyond Lyophilization,” 

Breakthroughs in analytics and alternatives to traditional freeze drying promise to reshape biological development and the cold chain.

Beyond Lyophilization

Since the COVID-19 pandemic struck, there has been little news of serialization breakthroughs, as manufacturers and their supply partners have focused on getting new therapeutics and vaccines to patients. However, progress has been ongoing behind the scenes, and the past two years have seen a significant increase in use of two-dimensional (2D) barcodes (1) and even adoption of blockchain, says Peter Sturtevant, senior director of community engagement at GS1 in the United States. Globally, GS1 works with pharmaceutical manufacturers, wholesalers, distributors, retail pharmacies, hospital dispensaries, trade associations, and regulatory agencies to help advance track-and-trace implementations and drive supply-chain efficiency.

“[In the US,] we have been supporting the [pharmaceutical] industry for the final milestone, the Drug Supply Chain Security Act (DSCSA) deadline of November 27, 2023,” says Sturtevant, and advances have continued through 2020 and 2021. “At this point, manufacturers are meeting requirements for 2D barcodes on packaging, and including package identifiers, National Drug Code (NDC) code embedded into the GS1 Global Trade Item Number (GTIN) for identifying products, serial numbers, lot numbers, and expiration dates. For medicines that come directly from manufacturers, you can already see the serial numbers on some products,” he reports.

COVID-19 vaccines, approved under emergency use authorization, were not required to meet serialization requirements, but they have been properly labelled to meet FDA requirements, he says, complete with 2D barcodes with GTIN, lot, and expiry date. In the future, serialization may be required.

He notes that several manufacturers, including Fresenius Kabi, opted to go beyond basic traceability requirements. The company decided to serialize its Diprovan anesthetic, a workhorse generic product, using RFID tags containing the four identifiers (2).

Although counterfeiting remains a threat, the growing use of barcoding and serialization promises to have an impact. Beyond serialization, the requirement that companies only do business with “authorized trading partners” will have a major impact, he says.

In April 2021, GS1 US issued a new implementation guideline (3), emphasizing best practices for maintaining chain of custody.This practice is particularly important for brandowner agents such as contract packagers, contract manufacturing organizations (CMOs), third-party logistics providers (3PLs), and reverse logistics providers. “If an agent is handling your product on your behalf, they need to leverage GS1 Standards including GTINs for products [and] global location numbers (GLNs) for physical locations, and share data electronically using electronic product code information services (EPCIS) to capture events from manufacturing to serialization [and] capping to shipping.EPCIS can capture those key events and share them with downstream trading partners,” Sturtevant explains.

Although the level of information technology (IT) infrastructure among contract service providers varies widely, adopting the technology may eventually become a requirement for doing business.Although GS1 Standards are not legally required of these entities, manufacturers may restrict business in the future to partners that can meet these requirements, just to avoid liability. “Some companies are still using fax machines. That’s the reality of the world,” he says, noting that cloud-based solutions have made technology more readily accessible.

Blockchain is also being evaluated and adopted where appropriate for use in product verification for resale and returns. “It’s not a panacea but another tool in the toolkit,” says Sturtevant.

Currently GS1 US’s Rx Secure Supply Chain workgroup is made up of 50 companies, including 12 manufacturers, seven dispensers, vendors, and all “Big Three” pharmaceutical whosesalers. The group is now developing new guidelines for applying GS1 Standards for serialization, which Sturtevant expects to be published later in 2021.

Generally, he sees the industry as prepared for the DSCSA deadline. “Nobody is taking it lightly. The industry is committed to it.If you look at our 2020 barcode assessment, four years ago 7% of pharma companies met requirements for labeling with 2D barcodes. By 2020 we had over 90%, both for packages and homogeneous cases, with proper barcodes.”

Real-time transportation visibility opens up

At the same time, on the vendor side, there has been an explosion of new technologies designed to track pharmaceutical cargo in transit. Dubbing the new market “real-time transportation visibility platforms,” as opposed to track-and-trace technology, the tech market analysts at Gartner predict that, by 2023, half of all global product-centric enterprises will have invested in these technologies (4).

In February 2020, Tive established the Open Visibility Network (OVN) with project44 and FourKites to enable open data access for mutual customers. The OVNhelps connect shippers, logistics service providers (LSPs), brokers, and customers by allowing open, collaborative data sharing. Instead of each partner working separately with clients, the open platform will allow the partners’ customers to easily access data acrossplatforms, so that Tive’s shipment data (which focuses mainly on hyper-accurate load location and condition), and project44’s or FourKites’ data (which focuseson other data facets) might be openly accessible.

In April 2021, Tive, Persequor, and Itrazo joined forces in a similar collaboration designed to help improve traceability in the cold chain, with an emphasis on biopharmaceuticals. “With the COVID-19 vaccine and strains on supply chains, we needed to get to a new level of visibility and traceability,” says Tive Director of Marketing Jim Waters.Companies will share data so that any of their customers can access data from any of the three platforms or their transportation, warehouse, or other management systems, bringing data on shipment load and location from Tive, from Itrazo’s data processing platform, and Persecor’s production and manufacturing data platform, so that customers needn’t switch platforms and vendors. The April announcement focused on the Asia Pacific and Australian regions, but the collaboration’s benefits are already available to users in North America and Europe, says Waters.

Tive offers non-Lithium industrial Internet of Things (IoT) trackers and software that apply machine learning and artificial intelligence to provide real-time data on shipments in transit, sending alerts when a shipment experiences an excursion, whether temperature excursion, route deviation, container breach (light), or shock. Using a cloud-based approach, Tive’s trackers upload data to its data cloud, streamed to customers using application programming interfaces (APIs), to allow users to take corrective action in real time.

Tive sees collaboration, real-time hardware, and connection to users as the crucial needs for bio/pharma traceability technology today. Without collaboration, he says, companies cannot hope to move at the velocity that this new world is going to require, (i.e., getting vaccines to end users on time and in full). He also sees the need for the industry to move beyond passive data trackers, which can at best provide a post-mortem after a problem has happened with a load. There will also be a need to better serve users through services that can help companies receive alerts so that they can respond more quickly to problems. Tive has launched a 24/7 support team designed to provide an extension of customers’ teams and help it manage responses.

1. GS-1 US, “Pharmaceutical Industry Makes Steady Proress toward Meeting the 2023 DSCSA Interoperability Requirements, GSIUS,.org. 28 Feb. 2021.
2. GS-1 US, “Case Study: Fresenius Kabi,” gs1us.org, 2020.
3. GS-1 US, “Implementation Guidelines,” gs1us.org, April 2021.
4. C. West and B. De Muynck, “Magic Quadrant for Real-Time Transportation Visibility Platforms,”

Agnes Shanley is senior editor for Pharmaceutical Technology Group.

When referring to this article, please cite it as A. Shanley, “Track-and-Trace Progress Benefits Supply-Chain Security Efforts,” 

Despite the COVID-19 pandemic, the industry has made great strides in implementing serialization technologies, while technology developers are enabling more open access to real-time transportation data.

Track-and-Trace Progress Benefits Supply-Chain Security Efforts

Pressure is mounting on FDA to step up the pace of on-site inspections that were postponed due to COVID-19 safety issues and restrictions. An open question is whether the agency will use new remote auditing technologies to improve its ability to inspect facilities during the COVID-19 pandemic. 

Pharmaceutical Technology's Regulatory Sourcebook 


eBook: Regulatory Sourcebook, March 2021
March 2021
Pages: 4-5

When referring to this article, please cite it as A. Shanley, “Can FDA Reinvent Inspections,” 

Pharmaceutical Technology Regulatory Sourcebook eBook 

A plant inspection backlog caused by COVID-19 restrictions threatens innovation and public health.  Now pressured to step up on-site inspections, will FDA embrace new methods?

Can FDA Reinvent Inspections

Despite the limitations set by the COVID-19 pandemic, bio/pharmaceutical manufacturers achieved unprecedented speed in developing new products in 2020. This was evident with the first two COVID-19 vaccines, which reached the public under Emergency Use Authorization in December 2020. Where developing a vaccine has traditionally required 10 years, Pfizer-BioNTech and Moderna Therapeutics cut that time to less than a year (1,2) with their messenger RNA (mRNA) vaccines.

Developers attribute this speed to their willingness to assume more risk from the start, and to leverage years of past research and clinical experience with the platforms themselves (e.g., mRNA or, in the case of Johnson & Johnson’s (J&J’s) COVID-19 vaccine candidate, with human adenoviral vector [Ad26]). With J&J’s vaccine, cell-line optimization was also crucial, not only for speeding the vaccine’s development but for optimizing its manufacturability and stability, as Paul Stoffels, executive vice president and chief scientific officer told attendees at the 2020 Galien Forum (3). Stability was optimized to levels of 2-8 °C, simplifying global distribution to developing nations.

However, with COVID-19 vaccines and therapies, the greatest increases in developmental speed and agility were achieved on the clinical side. New adaptive clinical study models were used; some early testing in humans was done concurrently with animal testing; some safety tests were run simultaneously with efficacy tests, and elements of Phase II and Phase III trials were combined.

Even on the preclinical side, work on Moderna’s COVID-19 vaccine was accomplished within a few months, CEO Stéphane Bancel explained at the 2020 Galien Foundation Forum (3). Scientists had sketched the experimental design in January 2020; the first test product was sent to the US National Institutes of Health in February; and Phase I testing started in early March. By the autumn of 2020, chemistry, manufacturing, and control (CMC) documentation had been submitted to FDA while safety testing was still going on.

Clearly, achievements in time reduction for COVID-19 therapies and vaccines have proven what is possible and set the bar high. One question is whether achievements made during a time of crisis can be sustained after the pandemic and move into mainstream bio/pharmaceutical development. New approaches, and the use of innovative technology, promises to enable additional gains in speed and efficiency.

A number of trends have already been in place to speed new drug development, including increased use of biomarkers and better patient targeting during clinical stage planning. In addition, more bio/pharma companies are using data analytics. One platform that is seeing increased adoption is artificial intelligence/machine learning (AIML), which can be applied to real-world data, including such diverse data sources as electronic health records, lab test results, and insurance claims, to ensure that trial design and patient recruitment reflect actual conditions.

Efforts by FDA and other regulatory agencies to speed review of important new drugs have also helped shrink development timeframes. Where the first decade of the 2000s saw slower development times and higher attrition rates for new drug candidates, notable increases in efficiency were made between 2010 and 2020, according to a recent study (4).

In a speech on June 1, 2020 (5), former FDA Commissioner Stephen Hahn pledged to make permanent some innovations that the industry and agency made to increase agility during the pandemic. He emphasized FDA’s interest in advancing practices such as decentralized clinical trials and increased use of real-world evidence in clinical trials. Technology is becoming increasingly important in optimizing study design, site selection, and patient recruitment, and also in predicting R&D and clinical outcomes, to help prioritize, early on, the most promising candidates, says Natalia Kotchie, vice president of R&DS Applied Data Science Center at the contract research organization, IQVIA.

Moderna used predictive models in the earliest phases of its COVID-19 vaccine development planning (3). As a CRO, IQVIA has been developing AIML-based tools and systems for the past five years to optimize clinical trials, for example prioritizing the highest potential sites. During the pandemic, IQVIA-developed COVID-19 models informed trial operational planning choices depending on areas where COVID-19 outbreaks were most likely to occur, says Kotchie.

For bio/pharma companies, one fundamental change that is speeding drug development has been increased collaboration and cross-learning through partnerships, says Lance Minor, national co-leader of life sciences practice for the management consultant, BDO USA LLP. The industry has engaged in collaborative research for decades, but during the current pandemic, the level of collaboration intensified, bringing together competitors such as Roche and Regeneron, Gilead and Roche, Gilead and Pfizer, Sanofi and GlaxoSmithKline (GSK), as well as companies working in incubators set up by the World Health Organization, the Global Alliance for Vaccines and Immunizations, the Coalition for Epidemic Preparedness Innovation, and the Gates Foundation (6).

Close work involving bio/pharma companies, CROs, and healthcare organizations has added another dimension to collaboration, Kotchie says. Another important factor speeding development of COVID-19 treatments and vaccines has been drug developers’ increased tolerance for risk, not only on the capital and financial side, but in the clinic, Minor says. Pfizer-BioNTech and Moderna spent billions of dollars on development for COVID-19 vaccines, but they also took risks by changing the traditional sequential approach to clinical testing, Minor says. In 2020 assuming these risks, and managing them, paid off in shortened development times. In 2021, novel approaches to enrolling patients and use of remote patient monitoring should also have an impact, he says.

Although efforts in 2020 focused on therapies and vaccines for COVID-19, the pace of development projects outside that area will pick up again in 2021, says Minor. Success in a very challenging year has led to optimism for 2021. In BDO’s 2021 survey, 100 chief financial officers (CFOs) at medium-to-large life sciences companies, the company found that 69% plan to increase spending on R&D, most of them singling out three or four promising products as commercialization targets. Pharma companies are still planning to increase investment in the development of new COVID-19 treatments, but a larger number also plan to invest in other therapeutic areas, with 57% mentioning cell therapies (up from 41% in 2020) and 56% targeting immunotherapies (up from 32% in 2020) (7).

Significantly, many respondents to the 2021 survey see digitalization as important to achieving their goals, with 63% responding that digital transformation would be their leading growth strategy in 2021.

One innovation that has helped shorten drug development timeframes has been use of biomarkers, indicators of a patient’s medical state that can be readily measured that have become a mainstay for clinical trials. Biomarkers were used in 55% of oncology clinical trials alone in 2018, up from 15% 20 years ago (8). FDA has recognized biomarkers’ importance since the agency first established its Critical Path initiative in 2004. More recently, however, FDA has begun to recognize other innovative drug development tools that are expected to help shorten drug development times, a mandate of FDA’s 21st Century Cures Act. On Nov. 30, 2020, FDA established a new pilot program, the Innovative Science and Technology Approaches for New Drugs (ISTAND) (9).

The effort could break through some of the industry’s reluctance to use technologies that have not first been applied in approved new drug applications. Through ISTAND, technology developers can apply to FDA for qualification of a novel technology for a specific use. The pilot would cover new, nontraditional data collection or analysis methods; technologies such as tissue chips that serve as micromodels of specific organ systems for drug safety assessments; and use of artificial intelligence and wearable sensors, which could be used during clinical trials.

To strengthen its foundation for data analytics, FDA has been working on initiatives in food import screening and medical devices. For pharma, key targets are drug safety testing and development of real-world data approaches.

In December 2020, FDA signed contracts with Palantir and SAS to build the database foundation required (10,11). Bio/pharma companies, including Novartis, Roche, Merck, Pfizer, AstraZeneca, and GSK, have also been actively working in this area, with more firms using AI for clinical trial optimization (12).

Among the challenges for pharma companies that have already embraced new technology is finding, cleaning, and reorganizing data to be research-ready so that AIML platforms can yield useful results, says Kotchie. Another new capability that will be required, she says, is the ability to translate business problems to how users make decisions, and ensure that the user experience of new systems is able to maximize the application of the models developed. “This is an area that the new discipline of ‘decision intelligence’ is helping us to address,” Kotchie says.

But even without the advantage of new technology, clinical trial flexibility has had to be balanced by strict risk evaluation and management. Business continuity programs have been crucial, says Teresa Lamentia, senior vice-president of operational transformation and performance at IQVIA, who also leads the company’s COVID-19 task force for R&D.

IQVIA’s clients range from small innovators to large pharma companies, and although some firms had to halt projects in 2020, most have restarted efforts, she says. “We’re not back to previous levels, but recruitment is definitely increasing and this trend should continue in 2021,” she says.

Lamentia and her team take a three-pronged approach, examining every trial protocol from the standpoints of patient safety, data integrity and robustness, and day-to-day realities affecting site visits (e.g., to ensure that subjects receive trial drugs and have necessary tests performed when COVID restrictions cause a site to be shut down).

Detailed reports were then developed to ensure that risks would not affect pharma clients or patients. As she notes, the company had to adopt remote approaches, for dosing or for patient testing, and wound up taking a hybrid approach combining on-site and remote elements.

This method demanded that both the CRO and pharma company pay close attention to supply chain issues. For IQVIA, an especially challenging problem was the sourcing of personal protective equipment and ancillary products for technical and business staff. Outside the United States, particularly in the United Kingdom and European Union, conformance with patient data protection regulations was another challenge.

In general, clinical trials are being run to optimize flexibility. “Taking an ‘umbrella’ approach to protocols allowed us to be able to perform more tests concurrently,” Lamentia notes.

It is not yet clear which new methods adopted during the pandemic’s first year will become permanent best practices, she says, but she expects to see more remote study approaches and more flexible clinical trial designs dominating drug development in the future. Based on regulatory and corporate plans, data monitoring and analytics promise to play an increasingly important role as well.

1. J. Hopkins, “How Pfizer Delivered a COVID Vaccine in Record Time,” 

, Dec. 11, 2020.
2. S. Mukherjee, moderator, “Can a Vaccine for COVID-19 Be Developed in Pandemic Time”, 

 June 9, 2020.
3. A. Shanley, “The Last Mile: COVID-19 Vaccine Developers Prepare to File for Emergency Use Applications,” 

18 (162) (2020).
5. S. Hahn, “Remarks by Commissioner Stephen Hahn MD on the COVID-19 Pandemic—Finding Solutions and Applying Lessons Learned,”

 June 1, 2020.
6. C. Leaf, “The Whole World is Coming Together: How the Race for a COVID Vaccine is Revolutionizing Big Pharma,”

, Sept. 21, 2020.
7. BDO Life Sciences, “2021 BDO Life Sciences CFO Survey,” Report, January 2021.
8. A. Vadas, T.J. Bilodeaux, C. Oza, “Special Report: The Evolution of Biomarker Use in Clinical Trials of Cancer Therapies,” 

, 2019.
9. FDA, “Innovative Science and Technology Approaches for New Drugs (I-Stand)”

, Nov. 30, 2020.
10. C. O’Donnell, “Palantir Wins $44 million FDA Contract,” 

 Dec. 7, 2020. 
11. SAS, “FDA Inks $49.9-Million Dollar Deal with SAS”, 

, Jan. 11, 2020.
12. Novartis, “AI at Novartis,” 


Vol. 45, No. 2
February 2021
Pages: 16-18

When referring to this article, please cite it as A. Shanley ,“Drug Developers Put Pedal to the Metal,” 

New approaches to clinical trials, spurred by the COVID-19 pandemic, have dramatically shortened development timeframes. Can these gains be sustained once conditions return to normal?








