Lessons from Heparin

Sept. 17, 2008
Avoiding future drug quality disasters will require closer control over raw materials, use of more powerful analytics and IT, and a Quality by Design approach
A Heparin Timeline
Heparin is an anticoagulant that has been used for more than a half-century to prevent unwanted clotting in post-surgical patients and those on dialysis. In January of this year, Baxter International, which produces roughly half of the U.S. heparin supply, recalled nine lots of product after news broke of a spike in deaths in patients taking heparin. The cause: hypotensive/allergic reactions to the medication. A few weeks later, Baxter recalled all remaining lots.
In March, FDA attributed the source of the contamination as over-sulfated chondroitin sulfate, a common dietary supplement. The degree of contamination ranged from 2% to 50% in the samples FDA tested. As the over-sulfated version was not known to occur naturally and was significantly cheaper than raw heparin, deliberate manipulation of the product was suspected.
The contamination source proved to be a plant in Changzhou, China, owned by Wisconsin-based Baxter supplier Scientific Protein Laboratories (SPL). SPL subsequently passed the blame onto its vast network of suppliers. FDA admitted that it had failed to inspect the Changzhou plant, as it should have before the drug was approved for sale in 2004.
Though no official source of the contamination has been found, FDA has worked to clear up several issues. It has announced that it is opening an office in China (as well as India), will be dedicating more inspectors to the country, and has received funding from congress to do so. It has also coauthored published research—namely a New England Journal of Medicine article in April—that outlines the scientific rationale behind the OSCS deaths.
In June, the U.S. Pharmacopeia (USP) released revised monographs for heparin sodium and heparin calcium, accompanied by two new and two updated official USP Reference Standards, giving industry (and FDA) improved guidelines for assurance testing heparin sodium and calcium drug substances.
For now, the Agency says, the heparin supply is under control and sufficient testing methods are in place to keep it that way. “The heparin supply in the US is being tested and is free of this contaminant,” Janet Woodcock, director of FDA’s Center for Drug Evaluation and Research (CDER), has said.
“We now have a mechanistic link, and we are confident that this contaminant does trigger the adverse events seen, so we feel the adverse reactions will now cease.”

This year, the world was riveted by coverage of the recall of several lots of the blood thinner, heparin. Not only were patient deaths connected to tainted heparin, but the case became the subject of Congressional hearings, triggering severe criticism of the FDA and the drug industry. At a time when the U.S. is outsourcing more drug development and manufacturing offshore, the recall and its aftermath made more people doubt the safety of APIs and other raw materials made in China. The events of the heparin saga are well known (Box, right).

Rather than rehash the story, this article considers the practical lessons—in cGMP, vendor auditing and Quality by Design (QbD)—that the pharmaceutical industry might draw from this case. They’re lessons that must be learned fast, because, as pharmaceutical supply chains become more global, complex and opaque, heparin may be just the tip of the iceberg. A survey by Pharmaceutical Manufacturing and Marsh Consulting Group suggests that supplier auditing and supply chain safety are nowhere near where they need to be. And experts say the climate is such that another crisis is within reason. “It’s going to get worse before it gets better,” predicts Warren Perry, compliance advisor for Qumas Consulting (San Francisco). Research into the root causes of heparin contamination suggests intentional tampering, and in a way that differs somewhat from earlier cases of tainted drugs.

For one thing, several lots of contaminated product still passed basic analytical and compendial quality tests, suggesting that whoever contaminated the compound had a sophisticated understanding of the limits of both. The heparin case also occurred at a time when the boundaries separating counterfeit from substandard drugs are blurring. “The level of sophistication of counterfeit drugs has gone up,” says Patrick Lukulay, director of the U.S. Pharmacopeia’s (USP’s) Drug Quality and Information Program, designed to strengthen pharmaceutical QC and QA in developing countries. “Previously, in Africa and in Asia, counterfeit products were comprised of either nothing in the capsule or starch or some inferior raw material. Now counterfeiters are putting the right stuff in there, but inferior quality material that doesn’t meet the pharmacopeial standards that you’d expect for that product.”

Recreating the Wheel

Heparin provides justification for the need for Quality by Design practices within the industry. When news of the heparin deaths first broke, FDA, USP and academic scientists were called upon to help determine what the heparin contaminant was and how it got there, and whether analytical and compendial tests were sufficient to protect the public (Box, below).

Sleuths on the Case
The analytical sleuthing to determine the nature and source of the heparin contaminant was impressive. Ram Sasisekharan, PhD, from MIT, at the behest of the FDA (in the person of Moheb Nasr), spearheaded a joint effort to discover what, if anything, was wrong with a number of lots of heparin, distributed by Baxter Healthcare. Work was already started by Dr. Nasr’s group at CDER, who had already used capillary electrophoresis (CE) and heparinases; they demonstrated that something was in the samples that did not qualify as heparin. He asked Sasisekharan to qualify exactly what the adulterant(s) was. One problem, of course, is that heparin is derived from pigs, and natural products are never exactly the same.
Heparin is mostly a polymer of disaccharides (two-ring sugars) with each unit containing uronic acid and glucosamide. When the various stereoisomers, sugars and sulfation patterns are combined, there are potentially 32 disaccharide units to be included in what is labeled “heparin.” The size of the task suggested two more research groups be
included: the Giuliana Ronzoni Institute for Chemical and Biochemical Research (Milan), which had been working with heparin for some time, and Momenta Pharmaceuticals (Cambridge, Mass.), equipped with state-of-the-art high throughput analysis equipment to perform numerous analyses in parallel. The work began with 10 samples from FDA: six determined to be adulterated and four “clean” controls.
These were blinded by the MIT researchers and sent for analysis. Preliminary work with 1H and 13C (one dimensional) NMR yielded some promising results: 1H NMR showed a peak at 2.16 ppm, signaling an acetyl group, determined to be not from heparin (2.06 ppm) or the common contaminant dermatan sulfate (2.08 ppm). 13C NMR gave unusual peaks at 25.6 ppm and 53.5 ppm. These uncovered the presence of an O-substituted N-Acetylgalactosamine. 13C also showed a signal between 103 and 105 ppm that implied a β-glycoside linkage between monosaccharides. When two-dimensional NMR (plotting spectrum/spectra from one technique versus another) was tried, even more information was gleaned.
Both sugars in the disaccharide unit contained two sulfate groups, a condition never before seen in “normal” heparin. To enrich/purify the contaminant, several methods were attempted: One group attempted enrichment by degrading the heparin. Another used an alcohol-based precipitation, while the third used chromatography, based on charge differences. These purified samples were subjected to a battery of NMR techniques: HSQC (Heterocyclic Singular Quantum Correlation), COSY (COrrelation SpectroscopY), TOCSY (TOtal Correlation SpectroscopY), ROSEY (Rotating-frame OverhauSer Effect spectroscopY), and HMBC (Heteronuclear Multiple Bond Correlation).
Basically, combinations of these techniques “map” the entire structure of the molecules, showing all links. The bottom line was the discovery of a “heparin-like” molecule, made of 2,3-O-sulfoglucuronic acid and 4,6-O-sulfo-N-acetylgalactosamine, with a β-1,3-linkage between the two sugars in the disaccharide and a β-1,4-linkage between adjacent disaccharide units. The molecule was named oversulfated chondroitin sulfate (OSCS). The molecule is not a natural product and was determined to have been synthesized and introduced into the heparin. The question still remained whether this was what caused the reactions in patients.
Since similar compounds (e.g., the drug Arteparon, an arthritis medicine manufactured by a German company and which was removed from market after causing adverse reactions in patients) were found, it appeared that one of these might be the culprit. Another team of researchers from FDA, Harvard, MIT and VPI were assembled to determine the toxicity of the newly found compound. The material was tested, ironically, on pigs and was found to cause the symptoms reported by the patients/doctors using the questionable heparin. Apparently, the chemical does not affect rodents and was not seen in routine screenings. The findings were published in April, two days after Chinese regulators held a press conference stating that OSCS was not the cause of the adverse reactions. – Emil Ciurczak

The detective work required to determine the contaminant was challenging, particularly since there was no developmental data available on the drug, says Moheb Nasr, PhD, director of FDA’s Offce of New Drug Quality Assessment (ONDQA), who led a team that had to, essentially, recreate the QbD data for heparin. “We had to start from scratch, conducting very detailed quality risk assessment of all aspects of the manufacturing process and the product characterization in order to identify what could be the potential cause or causes of this problem,” he says.

Troubleshooting would have been much easier had the right information been on hand. Instead, the team had to redevelop information from the application, with additional information from reviewing the drug master file (DMF), to restructure the type of information needed. Essentially, they had to remap the entire development, manufacturing and product characterization. A systematic, scientific approach was prudent rather than “shooting in the dark” with different analytical methods, Nasr says. Risk assessment was a critical part of the process, Nasr explains. It confirmed that the contamination stemmed from manufacturing quality problems and allowed the team to rule out any impact from the clinical side.

The team used fishbone diagramming to look at every possible impact, whether in raw material, chemical reagents, equipment, manufacturing process or testing methods. “We started doing testing on the biological side and the analytical side, trying to decide what could be an external contaminant that acts like heparin in some ways and has some structural similarities but also has differences, and we focused on that,” he says. Results of the research were published last April in Nature Biotechnology and the New England Journal of Medicine.

The main finding: the source of the contamination was over-sulfated chondroitin sulfate (OSCS), a common dietary supplement, and the adulteration was likely intentional.

USP Responds

As a result of these efforts, FDA contacted the USP and asked it to revise its monographs on heparin; these were published in June. Instead of the ambiguous tests previously required, manufacturers now must subject heparin to testing via electrophoresis and nuclear magnetic resonance (NMR). Rather than broadening efforts to other biologicals, USP has taken a focused approach, restricting monographs to heparin products, striving to balance overly prescriptive measures and potentially expensive analytics with product quality.

“My favorite analogy is with the International Olympic Committee, which tries to identify foreign substances that athletes might choose to take,” says Darrel Abernethy, CSO for USP. “It has become abundantly clear that even when you’re dealing with closely chemically related compounds, such as androgenic substances, that the next one you synthesize can be incredibly hard to detect and hard to figure out.”

There will be no silver bullets, he says, or “easy analytical solutions to allow you to look for things when you don’t know what they are yet.” He adds, “We could, in theory, design tests to guarantee the purity and identity of every material known to man, but the costs would be prohibitive, and everyone would have to have a 900-megaherz NMR and the latest, greatest mass spec.”

The goal, he says, is methods that will allow one to unequivocally identify the material of interest with orthogonal tests—physical tests such as hardness, density or melting point, chemical tests such as high-pressure liquid chromatography, nuclear magnetic resonance or mass spectrometry, and a biological test for active materials. “Any adulterant would likely not pass all three or four orthogonal tests,” he continues. USP’s monograph establishes an initial method that does physical and chemical as well as biological characterization, as an interim measure to protect the public from the current contamination, he says, adding that advisory groups are evaluating a broader range of methodology to meet requirements and specify material purity in physical, chemical and biological terms.

Mission Accomplished?

Measures by FDA, USP, and heparin suppliers and manufacturers to establish greater control over the product have been “pretty impressive,” according to Jian Liu, PhD, a leading heparin researcher and associate professor at the University of North Carolina School of Pharmacy. Nevertheless, Liu says, the current way of making heparin will always be vulnerable to environmental (e.g., porcine disease) and human (e.g., purposeful adulteration) factors. Three main factors are involved in control of the supply, Liu notes: the global pig supply, the heparin harvesting and purification process, and the analytical control. Like USP’s Abernethy, Liu draws parallels with antidoping procedures for Olympic athletes.

“You have to know what you’re looking for,” he says. But just as athletes find ways to circumvent drug tests, there may always be heparin contaminants that will fall outside of materials inspection parameters. FDA “jumped the gun” in suggesting that it had a handle on the heparin cases and the root of contamination, says Jawed Fareed, PhD, professor of pathology and pharmacology at Loyola University Medical Center (Maywood, Ill.) and whose group has been researching heparin for 30 years. Fareed suspected deliberate contamination the minute the news broke, but also believes that multiple contaminants could be involved beyond OSCS and that FDA and other researchers still know little about the physiologic mechanisms that led to the deaths.

Fareed notes that the Agency attributed just 3 of the 83 reported deaths to heparin containing OSCS, and for those has not provided verifiable supportive data about the deaths. The NEJM article is speculative and over-interpreted, he says. USP deserves credit for its efforts to update its heparin monographs, Fareed says. “They’ve held several important meetings and have discussed the contaminant quite extensively, and expressed remorse over the limitation of the methods USP has been practicing over the past 40 or 50 years,” he says. Fareed and colleagues at Loyola continue to search for answers. “We have learned that it is easy to contaminate heparin with biological substances,” he says. The popularity of low-molecular-weight heparins makes them prime targets for contamination, as are other animal sources, particularly bovine, as porcine stocks decline, Fareed says. “What I’m trying to say is the saga will continue, and people will find other things to contaminate heparin which are not so easy to detect.” USP’s monographs spell out the use of NMR and capillary electrophoresis as tests for heparin specifically.

In the case of heparin, Fareed suggests that polymerase chain reaction (PCR) testing should become a standard practice. It wouldn’t hurt to find an alternative to heparin as an anticoagulant in. for instance, hemodialysis and open-heart surgery, he says. Synthetic heparin is one possibility, since it would be animal-free and could be produced start to finish within one facility.

Liu of North Carolina and research colleagues have produced milligram-level synthetic heparin, and are seeking partners for developing the compound for pilot- and commercial-scale manufacturing.

Where To From Here?

FDA’s Nasr understands that additional guidance might be needed, to spell out more clearly what the Agency expects in regards to heparin safety. For now, Nasr says, requirements are outlined in ICH Q10, which addresses the responsibility of manufacturers and their quality sys-tem to ensure quality throughout the supply chain. There are important lessons that both industry and regulators can learn from heparin, Nasr says, to minimize the chance of future crises:

  • Manufacturers have the primary responsibility for product quality. They’ll need to assure the quality of all raw materials they purchase, and make sure the quality system they have is robust enough to assure quality throughout the supply chain.
  • Quality by Designtype information is critical for appropriate regulatory oversight. Regulatory submissions will need to include detailed development and manufacturing information, and documentation of appropriate quality control strategy and risk assessment of the product and process.
  • Modern analytical technologies must be used— not just to satisfy regulatory requirements, but as good business practice.
  • Regulators must find a way to encourage and, in some cases, require the use of modern methodologies to assure product quality. That will mean periodic updating of compendia for all pharmaceutical products.

Others have their own advice. Among the primary lessons learned from the heparin affair is that supply chains have gotten increasingly complex and hard to manage. Heparin exposed the multi-layered nature of most supply chains and the sheer uncertainty that many manufacturers have in regards to their raw materials, says Jaime Velez, head of the supply chain management practice at Tunnell Consulting, Inc. (New York, N.Y.).

“Unfortunately, it’s more common than one would like to believe,” he says. “Many products are made with raw materials that are sometimes three, four or more levels deep—the end manufacturer doesn’t have any idea what’s in the materials.” Velez advises that manufacturers must create more robust raw materials oversight processes. This doesn’t just mean adding staff or creating an external supply group, he cautions.

Velez has seen many a drug manufacturer beef up supply oversight without a clear strategy and then become consumed by non-value-added work. What’s lacking is a basic methodology for tracking materials from their origins, one that includes tracking systems, increased visibility of supply operations and a modus operandi of working through multiple levels of the supply chain to get to the roots of materials. It’s a “tedious exercise,” Velez admits, but one that has to be done. Of the top 25 pharma companies, Velez estimates that two-thirds have external supply organizations, though perhaps only a half-dozen have effective, efficient practices. The rest exhibit a “fantastic level of duplication”, he says, as QA, QC, technical and supply chain representatives, planners and schedulers wrestle with who should do what.

A growing number of drug manufacturers are extending control of suppliers by sending their QA auditors to third-party manufacturers and suppliers, says Eize de Boer, PhD, global manager of Life Science Auditing at the consulting firm SGS (Princeton, NJ). This holds true not just for the oversight of active ingredients, but excipients as well. Many major multinationals have already taken these steps, he says.

Quality Contracts: The Devil’s in the Details

Contracts are also essential, and establish the specific requirements made by the manufacturer to the supplier. In the technical contract, both parties can agree upon a checklist, in which the company contracting the services lays down all the requirements for the supplier. “A checklist with an extreme level of detail is necessary, and should be an attachment to the technical contract,” de Boer says. “For example, if there is a specific concern for cross contamination, the auditor can put special emphasis on air handling units, hygienic control, environmental monitoring, etc.” A site should be visited by two auditors, so that they may confer on findings, de Boer explains.

Off-site, the auditors will study the site master file (SMF) and drug master file (DMF), and inspection documents, previous audit reports and annual product quality review. Finally, auditors and their suppliers must assess deviations and complaints. This includes a closing session where the auditors disclose deficiencies as transparently as possible, and seek management comment. This information will be contained within a final audit report.

Based upon the seriousness of the deficiencies, the plant will propose a plan for corrective action and, if approved, follow through with it. Process knowledge is also critical, Tunnell’s Velez says. Manufacturers must characterize all processes, monitor process variables (in real-time if possible), and anticipate how product and process will behave in this multivariate environment.

What are the implications of a raw material that approaches the upper side of release specifications, for example? Process analytical technologies (PAT) may be useful in this regard, and have often been too narrowly defined in industry and excluded for raw material characterization, he says.

IT as a Solution

IT implementation is always easier said than done, but it can play an important role in safeguarding the supply chain. This year, Cephalon (Frazer, Penn.), a maker of small-molecule oncology and central nervous system therapies, began implementing both SAP for enterprise management and Sparta Systems’ Trackwise for supply chain management.

So far so good, says Roger Bakale, senior director of worldwide chemical process R&D. The company is already reaping rewards from the systems—Trackwise documentation was used to clear up a situation in which a shipper had inadvertently routed material through Canada—but it will be some time before Cephalon will be able to fully integrate and share data with suppliers, Bakale says. All manufacturers need two IT essentials, says Qumas’s Perry: a content management system, and a CAPA system for tracking deviations and preventing further occurrences. Depending on the training required, that’s a six- or seven-figure investment, he acknowledges, which many manufacturers can’t afford.

Manufacturers need to make the effort to collaborate with suppliers, and educate them if needed, says Perry. This is especially true for suppliers overseas. “Sometimes they still need computers and an IT department that understands how to keep networks up and running. Employees need basic training on operating systems, even logins and passwords—things that we take for granted,” he says. “I don’t blame manufacturers for feeling overwhelmed” by these challenges, he says.

Suppliers may not always be amenable to IT integration and increased transparency. Perry tells of a Chinese client that wanted to install a supply chain management (SCM) system, but changed its mind after realizing the system was “closed” and couldn’t be doctored. “They felt it was better to be able to trace materials some of the time rather than all of the time,” he says.

Despite the challenges, Perry says it’s important for companies to lay the foundation for IT sharing now so that, down the road, they can overlay partners’ systems. Cephalon’s Bakale says it’s not so important where a supplier is from, as what its track record is and how it manages its own quality control. Cephalon sources APIs and raw materials from across the globe, including India and China, but maintains rigorous oversight in terms of performing supplier audits, making on-site visits, and maintaining open lines of communication (including a penchant for videoconferencing). The company shies away from suppliers who subcontract, and, for materials that are considered high-risk or are used in late-stage manufacturing, prefers to source from US and European companies with proven DMFs for APIs.

Another Heparin?

Will any of these measures work? Are we doomed to experience another material contamination crisis equal to or greater than the heparin one? Probably yes. The pressure to reap profits is still much greater than any concerns manufacturers might have about crises and their consequent damage control, Perry says.

“Until manufacturers have their feet held to the fire by regulatory agencies, things are not going to change.” And, cynically, Perry says regulators won’t do this until another crisis occurs. “Not enough people have died yet” to prompt real change, he says. “I would not be surprised to see another heparin-like event occur,” says Tunnell’s Velez.

About the Author

Agnes Shanley | Paul Thomas and Michele Vaccarello Wagner with Emil Ciurczak