February 10, 2021 PAO-02021-CL-05
All pharmaceutical manufacturing processes have risks associated. Some are derived from each specific process, while others have origins external to the process itself. The former are intrinsic and can generally be identified from the process description, including the raw materials used and reactions performed. The latter are extrinsic risks and are much more difficult to identify. They can include the potential for cross-contamination in multiproduct facilities or operator error, as well as the use of recycled solvents or specific cleaning agents or packaging materials. These risks are not obvious from reading the process description but exist nonetheless and must be controlled in GMP environments.
Intrinsic risks by their nature are more predictable and easier to identify and control. By considering the reagents and process conditions, it is possible to determine whether undesirable side products could potentially be formed. They can be addressed by finding alternative routes to the desired product.
Extrinsic risks are not inherent to the given manufacturing process and not easily identifiable by considering the manufacturing process. Consequently, they are less predictable and more difficult to detect and control. Nevertheless, they still must be understood mitigated by the manufacturer.
Nitrosamines were first discovered in valsartan-based blood pressure medicines. Larger amounts of N-nitrosodimethylamine (NDMA) were initially detected (up to 75 ppm) in a valsartan drug containing an active pharmaceutical ingredient (API) produced by a Chinese manufacturer. N-nitrosodiethylamine (NDEA) was also detected in valsartan, but generally in lower amounts. NDMA and NDEA are genotoxic and carcinogenic agents in animals and are classified as probably carcinogenic to humans by the International Agency for Research of Cancer.
All N-nitrosamines belong to the so-called "cohort of concern" as highly mutagenic carcinogens, according to the ICH M7 guidance for mutagenic impurities. For such compounds, acceptable limits must be derived from compound-specific carcinogenicity data.
Before June 2018, these contaminants were not expected in sartan (valsartan, losartan, and irbesartan) drug products. These drugs are used to treat a chronic condition, and some patients could have been taking medications contaminated with potential carcinogenic impurities for up to six years before the NDMA/NDEA was detected.
The finding of the N-nitrosamine contaminants in valsartan, losartan, and irbesartan led to direct actions by authorities worldwide in the form of requests to carry out analytical tests, to report the test results, and consequently to recall the respective drugs.
Notably, all of these active ingredients have a tetrazole ring as a structural element. Other sartans that have a bioisosteric carboxylic acid function at this position are not affected by nitrosamine contaminants. This difference led to identification of the root cause of the contamination: a change in the synthesis involving the use of sodium azide and zinc chloride instead of tributyl tin azide, accompanied by a switch in solvent from xylene to dimethylformamide (DMF) and the need to quench excess sodium azide via the addition of nitrite. This new synthetic route was patented and adopted by many other active ingredient manufacturers.
Use of the new process in combination with recycled DMF containing dimethylamine and/or diethylamine as impurities resulted in the formation of nitrosamines.
The special attention that nitrosamines had drawn in the sartans led to a reassessment of these contaminants in other medicinal products. NDMA was next found in the antidiabetic therapy pioglitazone, again due to the use of recycled DMF in a process involving a nitrite reagent. The quantities were very low, however, and no market action was taken.
However, when NDMA was detected in the heartburn medication ranitidine, it became clear that the contamination issues were not limited to the sartans. This discovery was important, because ranitidine is widely used as an OTC product and is also one of the top 50 most prescribed drugs in the United States. In this case, the nitrosamine forms as the result of oxidative degradation and may also be generated under in vivo conditions. Lower amounts of NDMA were subsequently found in metformin drugs in December 2019, with the cause in this case also thought to be degradation. Both ranitidine- and metformin-based drugs have been recalled in the United States and the European Union, even though clinical and epidemiological data have not yet indicated an increased carcinogenicity risk.
The risk posed by NDMA contamination was evaluated by both the EMA and the FDA. The limit for NMDA set by both agencies is 96 ng/day. The highest concentrations in valsartan drugs (60–75 ppm) exceeded this value by about 200 times. The EMA calculated that, in a worst-case scenario with 100,000 patients taking NDMA-contaminated valsartan at the highest dose every day for six years, one additional case of cancer would occur for every 4,500 patients. The FDA reached similar conclusions; over 4 years of taking the maximum daily dose, one additional case of cancer would occur for each 8,000 patients.
We should bear in mind that we are exposed to nitrosamines from other sources, such as tobacco, smoked sausage and meat products, and in beer. However, we accept these products voluntarily and can replace them with other products or do without them. Patients do not have much choice when it comes to pharmaceuticals, and that is why special safety standards and risk–benefit considerations apply. While the risk of cancer development due to nitrosamine contamination was found not to be high, it is still much greater than what is deemed acceptable (one in 100,000).
The difficulty in these cases is that the risk of contamination should have been detected and controlled. The process changes that were made for the production of valsartan should have raised red flags regarding the potential for nitrosamine formation, specifically for experts at the manufacturer. Similarly, the potential for degradation of ranitidine and metformin might have been considered if a dedicated risk evaluation was conducted.
In the end, the wide-ranging recalls of a number of these drugs have required changes in therapies, and thus the potential contamination of pharmaceuticals with nitrosamines is considered one of the most important drug incidents in the recent years.
In fact, preventive risk assessments are now required for all medicinal products with chemically synthetic active substances by many regulatory authorities. The EMA,1 the FDA,2 and many others have adopted a three-stage process.
Marketing authorization holders are expected to work with pharmaceutical companies and drug manufacturers to assess the risk of nitrosamine contamination as the result of by-product formation during synthesis, degradation, or cross-contamination. If no risk is identified, the assessment must be documented accordingly and can be made available for review by the regulatory authority on request.
If a risk is identified, confirmatory testing must be performed using a validated analytical method. If a nitrosamine is detected, the regulatory agency must be notified immediately, and action must be taken to eliminate the risk. Any changes to processes that impact marketing authorizations must be reviewed by the relevant authorities. It is important to remember that analytical tests alone do not constitute a risk assessment, but instead may raise even more questions about what was tested for and why.
One difficulty with the guidances, however, is that the risk assessment requirements apply to API and drug product manufacturers, but not to excipient producers. However, risk assessments by drug product manufacturers must consider all of the ingredients in a formulation, including any excipients.
If there are no statements from raw material/excipient suppliers, other sources must be used for the assessment, such as audit reports, available quality information (i.e., residual solvents, quality management statements to avoid cross-contamination, history of contamination issues). In addition, scientific knowledge gained through experience or evaluation of literature (public and patent) on the manufacturing process should be applied.
MilliporeSigma has worked closely with industry associations to develop a common understanding of how to carry out risk assessments for active substances and excipients alike. Without a standardized approach, excipient producers and other raw material suppliers were faced with numerous and various questionnaires from their many different customers, all of whom expected quick answers.
For excipients, IPEC (International Pharmaceutical Excipients Council) Europe created a template in cooperation with the EFPIA (European Federation of Pharmaceutical Industries and Associations) to help their member companies prepare for nitrosamine risk evaluations.3 The aim of the IPEC Europe template is to provide information so that the marketing authorization holder (MAH) is able to perform effective risk assessments.
MilliporeSigma has approximately 100 active substances, 550 excipients, and 50 other raw materials for pharmaceutical production in our portfolio. Completion of 700 risk assessments within a six-month period as indicated by the latest regulatory guidelines requires tremendous effort.
We have established a working group with the aim of developing harmonized processes for completion of these risk assessments, which must be product-specific. We are also working to establish effective and consistent mechanisms for communicating with our customers at all 11 manufacturing sites in a uniform format.
The first step was to gain a common understanding of how the active ingredients and excipients should be taken into account. For active substances, it is quite clear that the assessment should show how the risk of formation or cross-contamination of nitrosamines is to be classified.
For excipients, there are no regulatory requirements or even indications that tests for nitrosamines would generally be necessary to assess the risk of nitrosamine formation. Within our working group at MilliporeSigma, however, we agreed that assessments also must be performed for excipients; they are components of the medicinal product and therefore must also be part of the risk evaluation. After all, the MAH is ultimately responsible for the quality of the drug, including the quality of the active ingredients, excipients, and raw materials used for the manufacture of the final drug product.
Next, the intrinsic and extrinsic risk factors for nitrosamine contamination of active ingredients, excipients, and finished medicinal products were considered in order to enable a rational, risk-based prioritization. Non-chemically synthesized substances were considered to have a low-to-negligible intrinsic risk of nitrosamine contamination. Similarly, nitrogen-free substances, because they are highly likely not synthesized with nitrosating agents, were considered to be of lower priority unless manufactured with nitrogen-containing organic solvents.
This prioritization was only used to establish an order for conducting assessments. From the outset, the plan was to conduct product-specific evaluations rather than mass assessments for similar type of compounds. We established our own assessment scheme and documented the risk factors using the IPEC template.
the use of nitrosating agents, such as nitrites and other substances that could give rise to nitrosating agents;
the presence of nitrosatable structures, which are primarily secondary and tertiary amines, but also quaternary amines or substances that can form secondary or tertiary amines as degradation or by-products;
the availability of analytical data on specific nitrates and nitrites; the presence of residues could lead to nitrosamine formation in compounds with nitrosatables structures, (a low nitrite/nitrate concentration can be assumed);
the quality of the process water used for production, if the process water was obtained using pharmacopeial methods;
recycled solvents, particularly nitrogen-containing solvents such as DMF solvents, especially if the recovery process was external or performed in non-dedicated facilities; and
the potential for cross-contamination in multipurpose equipment.
At MilliporeSigma, therefore, we have conducted paper-based or theoretical assessments to determine if there is any risk of nitrosamine formation — or other toxic/carcinogenic/mutagenic compound generation — associated with our existing processes. In most cases, no risk was identified, because our process chemists are aware of the issues around impurity formation and specifically design processes to avoid such issues. However, in the rare event that a potential risk was identified, perhaps because full insight into the production of a raw material was lacking, confirmatory testing was performed. Fortunately, to date we have not detected the presence of any nitrosamines.
The Emprove® Program contains over 400 raw and starting materials, and the Emprove® Dossiers provide comprehensive, up-to-date documentation to help you navigate regulatory challenges, manage risks, and improve your manufacturing processes. Because there is no specific guidance for excipients, yet excipients are usually considered in the risk assessments for formulated drug products, MilliporeSigma has begun to include product-specific nitrosamine declarations within our Emprove® Dossiers for excipients, APIs, and other pharmaceutical raw materials. Using the IPEC template, we outline the relevant and detailed information required for assessment of drug product manufacturers containing those materials.
By providing this information, therefore, we make it possible for MAHs to conduct the necessary risk assessments on their finished products. In this way, we are making the necessary connections and interacting with customers in support of elimination of nitrosamine contamination.
The recent regulatory requirements issued by the FDA, EMA, and other regulatory authorities are likely to be supplemented by other regulations to address potential contamination by nitrosamines.
A draft revision of the Substances for Pharmaceutical chapter of the European Pharmacopoeia was published in the January edition of PharmEuropa.4 It provides that manufacturers of substances for pharmaceutical use (i.e. active substances and auxiliary substances including excipients) carry out risk assessments of their manufacturing processes and implement strategies for the detection and control of nitrosamine impurities. Since the pharmacopoeia is enacted by law, these requirements would be legally binding.
The concern is that some excipients are not mainly manufactured for pharmaceutical use and are only used in small quantities by the pharma sector. If the manufacturer of such a substance does not provide the risk assessment, it may no longer be usable as an excipient by pharmaceutical manufacturers. There may still be an adaptation of the text here. However, the intent of the pharmacopoeia, as far as one can see, is that an assessment of the risk of contamination of nitrosamines in the medicinal product should consider and begin at the level of the starting materials.
A draft of the European Pharmacopoeia published in April contained three proposed test methods (LC-MS/MS, GC-MS, GC MS/MS) for nitrosamines.5 The tests for five different sartans containing tetrazole rings have been developed and validated with a limit of 30 ppb for each of the listed N-nitrosamines. Importantly, the validation of these methods applies only to the specified sartans as drug substances and does not extend to any formulated drug products. A separate validation must also be performed for quantitative evaluation of medicinal products containing these APIs. In addition, for any other nitrosamines not listed in the draft text, application of any of the three methods would need to be checked and validated.
In the September Pharmacopeial Forum, USP published an informative chapter on nitrosamine contaminants6 with a numbering greater than 1000, and therefore has guideline character and will not be binding. Several analytical test methods for control compliance with limit values for nitrosamines are expected to be provided, as is a flow chart describing the proposed risk-based control strategy.
We can also expect ICH M7 on mutagenic impurities7 to be expanded to include nitrosamines as specific examples. It is important to remember that nitrosamines are only one type of mutagenic impurity, and that there are many others that could potentially be found in pharmaceutical products. As a result, the industry as a whole and individual manufacturers of pharmaceutical raw materials, intermediates, APIs, excipients, and final drug products must all be vigilant in assessing the risk of mutagen generation associated with their manufacturing, packaging, and storage and handling processes and practices.
Given the ongoing discussions and questions regarding some of the nitrosamine contamination incidents that have already occurred, it is obvious that more clarity is still needed on the root causes and other contributing factors in these cases. With metformin drugs, for example, nitrosamine contamination has been detected at elevated levels in extended-release dosage forms. Interaction with excipients of these dosage forms is likely to trigger the oxidative degradation to form NDMA, but further investigation will be required. We do not know whether other APIs will be found in the future to contain nitrosamine contaminants. There potentially could be other chemistries that we are not currently aware of that may present risk factors. As a result, this topic will front and center in the industry for a while yet.
Given the level of uncertainty, going forward we believe that customers will soon be asking for more extensive evaluations that include testing for not just nitrosamines, but other precursor compounds that could react to form nitrosamines, such as nitrites. As a result, MilliporeSigma is already exploring how we might generate this type of data.
Ulrich Reichert leads the Pharma & Food Materials organization of Regulatory Management within the Life Science sector of Merck KGaA, Darmstadt, Germany. He is pharmacist by training and after his doctoral degree in medicinal chemistry, he gained many years of analytical experience as laboratory manager for pharmaceutical starting materials. He graduated in Drug Regulatory Affairs with a Master’s degree (M.D.R.A.). He co-leads the global nitrosamine task force of his company and chairs the nitrosamine task force of IPEC Europe.