For pharmaceutical companies, there are few things worse than a product recall. Even if a drug has not negatively impacted patient safety, managing a recall can be complicated and costly, and can damage customer confidence. Alongside failed specifications and mislabelling, one of the most common reasons for a recall is particulate contamination, which is defined by the US Pharmacopeia (USP) as the process by which “mobile undissolved particles become unintentionally present in a parenteral solution”. No matter the cause, this can have serious financial and reputational consequences, and can lead to a host of problems.

The sources of particulate matter are varied. Some originate from foreign or extraneous contaminants, such as product containers and packaging products made from materials like glass, plastic rubber, aluminium and paper. Others are inherent to the active pharmaceutical ingredient and are often caused by protein aggregation.

While all injectable drug products contain some level of particulate matter, contamination by either foreign matter or protein aggregates can create issues for patient safety, says Erwin Freund, the executive director for drug product engineering at Amgen, the US-based multinational biopharmaceutical company. “It is an important quality attribute,” says Freund. “These particles can be non-sterile, they can be immunogenic and they can cause local reactions at the site of injection.”

Robust regulation

While recalls do happen, according to Freund, the danger posed by particle contamination has been significantly reduced in recent years thanks to the attention of regulators and subsequent efforts by manufacturers to improve their inspection processes. The rise of biotherapeutic medicines in the 1980s, for example, increased the risk of contamination through aggregation, but saw a quick response from FDA as well as European counterparts.

“FDA has increased its efforts to make manufacturers introduce a particle prevention strategy that would limit, if not totally eliminate, the presence of protein particles,” says Freund. “They expect the manufacturer to carefully analyse adverse events in clinical trials. There is an allowance for the presence of a certain amount and type of protein particle in the final drug product if you have demonstrated that this is the typical behaviour of the drug product, without impact to safety and efficacy.”

More recently, there was a sharp uptick in regulatory observations following an increase in the presence of glass particles in vials. According to a report by West Pharmaceutical Services, between 2006 and 2011, more than 20 product recalls were caused by glass, resulting in 100 million units of drugs withdrawn from the market. But again, the industry bounced back fast, according to Freund.

“This event suddenly intensified the concern that the agencies had,” he says. “The industry rapidly learns from those specific issues and puts surveillance methods in place to apply best practices. Adverse events in the past ten years due to particles in parenterals have been very low.”

FDA regulations governing particles have become progressively clearer in recent years too. For a while, the industry struggled to understand the USP’s requirement that parenteral drug products should be “essentially free” of visible particulates. While this acknowledged that injectable drug products could never be totally free of particles, many felt the requirement was too vague.

However, in 2017 “FDA improved the clarity of the conditions under which particles should be examined, which gets closer to the European guidelines,” says Freund. “Now, we have a scientific definition of what it means for a solution to be practically free of particles.”

You can easily injure an animal by infusing large numbers of particles, but the data you get is not really relevant to understanding what happens in a human.

Improving inspection

More can be done in others areas, however. According to Freund, one major problem with current inspection processes is how invasive they are. To examine a syringe, vial, cartridge or IV bag, for example, pharmaceutical companies must first open them and analyse the contents.

“It is a challenging procedure,” Freund says. “There is work to be done to introduce non-invasive analysis, where you can examine without opening the container. These [methods] are in an early stage of technology development.”

Predicting whether a protein particle may cause patient safety issues is also made difficult by unreliable animal research. As things stand, pharmaceutical companies use information that is based on injecting animals with billions of foreign particles – far larger than what gets found in injections used on humans.

“You can easily injure an animal by infusing large numbers of particles, but the data you get is not really relevant to understanding what happens in a human,” says Freund. “Amgen and other companies are working with academics to find out to what extent those risks are a reality. I expect, in the next few years, there will be more articles published in peer-review journals that give more clarity to the industry on how to better predict the outcome.”

When working out the risk of a particle, Freund recommends taking into account what he calls the route of administration. “The level of tolerance for particles depends on these routes,” he says. “Among the most sensitive routes is injection in the eye. The next most sensitive is an IV injection, because you are distributing the particles straight into the system and bloodflow.” Calculating risk also depends on the specific patient in question. Are they immunocompromised? What is their level of defence? What is the volume of the injection being used? “You need to take into consideration all of those factors,” says Freund. “There is not one rule or set of assumptions to be applied to all cases. It really depends on a range of parameters. You also need to document the risk assessments, so that when FDA asks for justification, you have reports or monographs in it.”

For biopharmaceutical companies, Freund recommends creating processes and technical reports that describe what the normal appearance of a drug product is. “If you have no protein aggregation, it is pretty straightforward,” he says. “But if you have some aggregation, the key is to characterise the protein product over time during stability studies and, where possible, enrich the qualitative observation with quantitative data. If you do see particles, ask: ‘Do I have any information I can provide on the shape, size and number of those protein particles, so that I know what is normal?’”

Now, we have a scientific definition of what it means for a solution to be practically free of particles.

Manual and automated inspection

The industry is also struggling with manual and automated inspection processes. While recent USP publications have given greater clarity on how to execute a manual inspection, humans will always be fallible. Two people may produce inconsistent results, for example, and even the same person’s work may vary as time goes by. “This is not an approach that gives you completely consistent data,” says Freund.

While automated inspection is considered faster and more consistent, it typically yields a much higher false rejection rate – sometimes up to 30% – thanks to the use of high-resolution cameras.

“If I have 20% false rejections based, for example, on tiny air bubbles, I may have to manually inspect 20,000 of them,” says Freund. “The validation of automated inspection is very tedious and takes a lot of time. Thankfully, technology is progressing and, in the future, I see more use of artificial intelligence and machine vision.” With the adoption of automated inspection machines expected to continue rising, Freund recommends that companies hire machine-vision experts who truly understand the software and optics involved. These experts can then work closely with vendors to dictate how particular machines should be equipped.

“You need to have your engineer with experience in optics and physics dictate to the vendor the specification or the components that they are going to install,” Freund says. “If you just rely on the vendor, you are going to be dissatisfied.”

While new technologies are likely to emerge in the coming years, the ultimate objective for the industry will remain the same: reducing particles and removing potential risk to patients. “Although the number of negative cases is very low, we must keep raising the bar to ensure parenteral injections are as particle-free as possible,”says Freund.

Drug recall: the statistics

  • The number of UK product recalls increased 48% to 575 in 2015–16, from 388 in 2014–15, according to law firm RPC.
  • This leap was heavily influenced by new EU legislation on the labelling of food allergens. The legislation, which was introduced in late 2014, requires all food labels to display information on 14 different allergens.
  • In the US, recalls are almost always voluntary, meaning a company discovers a problem and recalls a product on its own. At other times, a company recalls a product after FDA raises concerns.
  • Only in rare cases will FDA request or order a recall. But in every case, its role is to oversee a company’s strategy, classify the recalled products according to the level of hazard involved and assess the adequacy of the recall. Recall information is posted in the Enforcement Reports once the products are classified.
  • A study performed by the US Food Marketing Institute and Grocery Manufacturers Association indicates the average direct cost to a company for every recall is $10 million – and incidental costs linked to outsourcing and the supply chain can be much higher.

Source: City A.M., FDA.