A few weeks ago, the Citizens Generic v. Branded Drugs Project sponsored and designed by the Mission of Ethics and Science in Healthcare (MESH) announced on social media, astonishing results of its study that tested 131 generic drugs before peer-review was completed. MESH has now made public the test reports for all 131 generic drugs, although the raw data such as the chromatograms have not yet been made public. A cursory review of the test reports raise some questions about the design and testing process based on which broad and generalized claims were made about the overall quality of our drug supply.
Analytical testing of these samples was conducted by a contract research organization (CRO) named Eureka Analytical Services Pvt. Ltd. These privately run laboratories survive on business from the pharmaceutical industry. Their entire business model is contingent upon securing testing contracts from the pharma industry for reasons ranging from capacity constraints, specialized testing methods to regulatory compliance. They perform testing for the pharma industry under a contractual agreement and the results are shared with the sponsor under a confidentiality agreement. These results become a part of the regulatory submissions to secure market authorization and for continued licensing to sell the sponsor’s product.
In addition to supplying data for regulatory filings, these CROs are sometimes involved in the public procurement process. Some programs like Jan Aushadhi require the pharmaceutical industry to procure testing certificates from CROs as a pre-requisite to shipping their product. Many times, government agencies, as a part of quality assurance process, will empanel different CROs for testing drugs supplied to their procurement programs.
Given this relationship between the CRO industry, the pharmaceutical industry and the government, nothing could be worse to the business of a CRO than publishing for public viewing, a report bearing its name and logo declaring drugs manufactured or distributed by its biggest customers to be ‘not of standard quality’.
Given these realities about the CRO business, selection of an academic laboratory or alternatively a government laboratory to eliminate a potential conflict of interest, especially when it was making the test reports publicly available would have been wise. An alternate method to tackle the conflict of interest issue would have been to “blind” the CRO to the identity of the manufacturers of the samples being sent for analysis – this was not done, as is now evident from the test reports published by MESH. The brand names and batch numbers are visible on the published test reports. The CRO knew the name of the manufacturer of the drugs it was testing, along with the identity of the pharmacy from where it was drawn. That this information was not blinded, is a serious methodological flaw in the design of the study and one that significantly undermines its credibility .
In addition to the above concerns, there is also the issue of the CRO industry fearing retaliation by the government. As I mentioned in my earlier piece, this government has a history of threatening people who raise quality problems with drugs products in India. Targets in the past have included American academics who conducted a similar study. Apart from attracting the wrath of government, a commercial CROs also has to factor in the contracts dangled by the government for labs to participate in the public procurement process.
Published reports raise questions about testing process
A second problem evident from a reading of the test reports is the quality of testing conducted by the laboratory. There are obvious issues which leap out from the test reports and raise doubts about the testing process adopted by the laboratory.
First, is the issue of impurities in the drugs tested. These impurities can be of two kinds. The first are “related impurities”, which are byproducts of the manufacturing process and these are permissible as long as they remain within the range laid down in the pharmacopeia. For example, the manufacture of Azithromycin generally results in the impurity B which has an acceptable threshold of 2%. Similarly, the manufacture of Clopidogrel results in the generation of an impurity C with an acceptable threshold of 1.5%. One industry expert I spoke to characterised these examples as “stubborn impurities” which are inevitable in the manufacturing process of even the best of pharmaceutical companies. Yet, as per the test reports published by MESH, not a single sample of either Azithromycin and Clopidogrel shows either of these impurities. Lack of evidence of known impurities in all the samples manufactured by different pharma companies is a manufacturing impossibility. I say this not only on the basis of my personal experience of decades in the industry but also on the basis of opinions of formulation developments experts who have spent a career in the industry.
The second type of impurity are those that result from an unstable chemical entity or impurities in the API. For example, Nitrosamines (NDMA) which are carcinogens are often a by-product of manufacturing in the pharmaceutical sector. Low levels are found in food and water and are not harmful, but high-level exposure poses health risks. Ranitidine, which is sold in India today was withdrawn from most markets because of a high presence of this impurity in this drug. In a number of test reports published by MESH where the assay tested at less than 100% for the active ingredient, the test results claim that no impurities were detected. For example, a test for ranitidine hydrochloride tested positive for 97.2 % of the active ingredient (i.e. 145.8 mg out of 150 mg) but the same report also states that no other impurities were detected. What then is the remaining 2.8% content of the tablet if not impurities? The test reports are silent. Standard practice when issuing a Certificate of Analysis is not only to identify the impurity, but characterize it as well.
Lack of information regarding the impurities is not the only problem. The reports also do not provide enough information to categorially affirm results of dissolution testing which are meant to evaluate the ability of the drug to dissolve in the human body within a certain timespan. For example, Omeprazole, Atorvastatin, Telmisartan and Montelukast have very low solubility. For such drugs, dissolution needs multiple stages of testing – for example, stage 1 will usually fail, after which stage 2 testing is conducted with a larger sample size. None of the test reports published by MESH show whether testing proceeded to Stage 2 and if so, what the results were. It is highly unlikely for these drugs to clear dissolution testing in Stage 1.
A third set of problems with the test reports is with regard to the seven samples of Febuxostat. The test reports for all seven samples contain contradictory assertions. In the main body of the test report, the lab states that it was tested as per “in-house specification”, while at the bottom of the report is a line claiming the samples were tested per the Indian Pharmacopeia, 2022 edition. These are contradictory claims and both cannot be true. The bigger problem is the claim of “in-house specification” – this goes against the very concept of standards in pharmaceutical quality control. Standards are always set by an external body like the Indian Pharmacopeia or the United States Pharmacopeia (USP). For example, Febutaz-40 manufactured by Sun Pharma follows USP standards, as per its packaging data available on 1mg.com which means that it has to be tested per standards laid out by USP – this requires purchasing reference standards from USP and following the test protocol laid down in the USP. This was not done – the test report declared it was tested as per some undisclosed “in-house specification”, followed by another line at the bottom stating that it was tested as per the IP specification. There is no mention of the USP standard.
All these issues raise serious questions about the credibility of the testing process followed at Eureka, the CRO.
Why Phamacopieal testing is no substitute for Bioequivalence
In addition to those identified above, there is also a problem with some of the statements made by MESH on its website regarding bioequivalence wherein generics are first tested on human subjects in order to establish the dissolution profile and bioavailability of drugs. In an earlier piece, I had identified the lack of bioequivalence testing in India as a serious limitation to the claim that all generics are as good as innovator products on the basis of mere laboratory testing. Dismissing this claim, the MESH team states that bioequivalence testing is not necessary for “common, standard medicines” which dissolve quickly and that the government in any case verifies whether a generic drug is bioequivalent while approving it. While it is true that pharmacopeial testing does not include testing for bioequivalence, it should be noted that laboratory based dissolution testing is a poor surrogate for testing the drug on actual human subjects. This is the reason why bioequivalence testing was mandated in the first place for drugs which do not easily dissolve.
The bigger problem in India is that there is no guarantee that government actually verifies whether a drug is bioequivalent. India made bioequivalence testing compulsory only in 2017; and we do know that drugs approved prior to 2017 were not tested for bioequivalence. Unlike the American regulator, the Indian regulator does not publish an Orange Book type database informing citizens whether the formulations being sold in the market have been declared bioequivalent to innovator brands. We do not even know whether the new requirement in 2017 is being enforced by all state drug regulators before approving generic formulations. We simply do not have any information on this issue.
Given this lack of information, it would be prudent for the team at MESH to refrain from making claims that generics in India can be trusted solely on the basis of pharmacopeial testing.
While the intent of the MESH initiative was noble, it has erred in making broad and categorical statements about the quality of our drug supply based on such a small and unrepresentative sample. It is important for those in the media to be careful in how they report these results for its casual readers who may lack scientific training to critically evaluate these test results. This episode is also a reminder on the importance of following the established norms of science communication. Social media advocacy should follow, not precede, peer review.