What do CDSCO officials do during vaccine-manufacturing site inspections?

This post is born out of sheer frustration. I don’t hope to achieve anything by writing this post. After all, everyone already knows what a hopelessly opaque organisation CDSCO is. This has been documented many times over, but CDSCO doesn’t care. What I do hope to document, however, is just the sheer, enthralling mystery of it all.

On 2 April 2022, Indians heard the news that the WHO had suspended all purchases of Covaxin from Bharat Biotech, after having found what were obviously major deviations in Good Manufacturing Practices. They also learnt that Bharat Biotech had helpfully offered to suspend all exports of Covaxin, because it was the right thing to do. However, the company seemingly didn’t think it was necessary to suspend production and sales of Covaxin in India.

This news bothered, but didn’t surprise industry insiders. It’s fairly well known that CDSCO imposes more lenient quality and manufacturing standards on Indian companies than ICH regulators do. And WHO tends to impose ICH standards on most medicinal products too. So, what seemed like a major violation to WHO perhaps didn’t mean anything to the Indian regulator.

But having followed the Bharat-Biotech-Covaxin saga for so long, and being the object of a defamation lawsuit by the company to boot, I was naturally curious. In April 2021, I had written in detail about Brazil’s ANVISA, a regulator who follows ICH standards, finding serious GMP violations in Bharat Biotech’s Covaxin manufacturing process. This led to Brazil cancelling its procurement order of Covaxin altogether. Even then, Bharat Biotech had brushed off the issue, saying that Brazil was politicising things. And CDSCO hadn’t said a word.

Now, here was the WHO making the same claim, that Bharat Biotech had violated GMP while making Covaxin. Was WHO also playing politics? I was inclined to believe it was not. And that Bharat Biotech had indeed screwed up. This got me wondering: what was it about the GMP violations that WHO considered to be serious enough to suspend vaccine purchases, but CDSCO considered to be a non-issue? I was seized with an itch to understand these violations in granular detail. But the WHO had published no inspection report on its website (the WHO does publish inspection reports sometimes, but it isn’t very clear when it chooses to do so, and when it doesn’t).

However, thanks to responses from WHO’s media team when I wrote this story, I knew that WHO had asked CDSCO officials to accompany them to the fateful inspection of Bharat Biotech’s Telangana Covaxin manufacturing site between 14th-22nd March 2022. WHO’s team had also told me that they had subsequently held meetings with CDSCO, so that both agencies could “coordinate our actions”. And while the WHO is, unfortunately, not covered by India’s Right to Information Act, the CDSCO is.

So, I filed an RTI query with CDSCO, asking them for the WHO inspection report, which I assumed they had a copy of. My expectation was that CDSCO would deny my request under section 8(1)(e), claiming it has a fiduciary responsibility towards Bharat Biotech (Yes, they have been known to use this section to deny information about the corporations they regulate!). But the reply I got was not one I was prepared for. After the usual forwarding of the RTI query from one department to another, CDSCO said they didn’t have the inspection report at all!

Now, this was decidedly odd. By then, Bharat Biotech had received an Emergency Use License from the WHO; infact, the WHO inspection was a part the agency’s post-EUL diligence. And according to this document on the EUL procedure, the WHO enters information-sharing agreements with the national regulators of record, when it issues EULs. The idea of this agreement is that the national regulator will share all the data on quality/efficacy/safety that it has with WHO, so that WHO can make a determination about whether the vaccine is fit to receive an EUL.

It’s obvious that some form of information-sharing agreement does exist between the WHO and CDSCO. Prior to awarding the EUL to Covaxin, the WHO had not inspected Bharat Biotech’s manufacturing site, which it typically does during the pre-qualification process. Instead, this inspection had been waived off due to the pandemic. As an alternative, the WHO had relied on the report from the CDSCO’s inspection of Bharat Biotech’s Covaxin manufacturing site. Given this, we already know that CDSCO was freely sharing information with the WHO. Which begs the question: why would such an information-sharing agreement not cut both ways? After all, one of the objectives of the pre-qualification/EUL process is to build capacity of national regulators. And what better way for WHO to build CDSCO’s capacity than to share its inspection report with CDSCO?

So, why was CDSCO claiming it didn’t have the inspection report? I tried to resolve this conundrum by asking the WHO media cell again if the WHO shared the inspection report with Bharat Biotech. However, the WHO media team didn’t respond this time.

Confused, I decided to file a second RTI query with CDSCO. After all, even if the WHO didn’t give CDSCO a copy of its inspection report, surely there must be some written communication related to the joint inspection, since both agencies were supposed to “coordinate” actions. At the very least, CDSCO officials inspecting the manufacturing facility would have made some notes of their own? Typed up a report of some sort, and submitted it to their seniors? A reasonable assumption, one might think.

And so, in my second RTI query, I asked CDSCO for all written communication from the WHO to CDSCO, and all reports written by CDSCO. A month later, I got their response. Guess what? The vaunted apex drug regulator of India again claimed it had nothing!

Which brings me to the enthralling mystery I spoke about in paragraph 1. What *exactly* were CDSCO officials doing throughout the nine-day inspection they accompanied the WHO on? My imagination fails to come up with an answer. Were they just sauntering around, cleaning their teeth with toothpicks? Were they serving chai and sandwiches to WHO officials, as @grumpeoldman jokingly(?) suggested on Twitter? Is it that the WHO did give CDSCO a copy of their inspection report, and the CDSCO officials absent-mindedly wrapped the sandwiches in this report and gave it back? Did CDSCO officials doodle in their notebooks all the time? If so, why didn’t they do me the kindness of sending me their doodles in the RTI reply, as reward for my labours?

I really cannot think of any explanations. Somehow, I find it hard to believe that CDSCO officials would be foolish enough to lie in an RTI response. I am not saying it’s impossible, but how brazen that would be! And yet, the alternative explanation is equally scary to contemplate. Is this how CDSCO officials carry out inspections of manufacturing facilities?

Such an inspection is not just critical, but complex. Because vaccine manufacturing is complex. Such an inspection would require the full alertness of the CDSCO officials. And certainly, it requires them to take *some* notes. But, increasingly, it seems like CDSCO officials did nothing more than have a nine-day picnic at the manufacturing site.

PS: I asked the Drug Controller General of India, VG Somani, how it was possible, as the RTI reply claimed, that CDSCO officials had no record of the inspection. Feeling cheeky, I also asked if they did nothing more than have chai and samosa (/sandwich). I didn’t get a response. I also asked CDSCO’s Central Public Information Officer, B Sarala Devi, whose signature was on one of the RTI responses, for clarifications. I didn’t get a response from her either.


India and the nitrosamine saga

In my recent article published in Mint (behind a metered paywall reset every month), I wrote about how India was dealing with the issue of nitrosamines – carcinogenic compounds that can contaminate drugs. Several wealthy countries, like the US, European Union countries, and Australia, are taking nitrosamine contamination very seriously. India, on the other hand, has been dragging its feet. Neither the central Indian drug regulator, CDSCO, nor state drug regulators, have undertaken nitrosamine testing; nor have they mandated that pharmaceutical firms keep nitrosamines within safe limits.

Soon after I wrote this article, I received feedback from several sources in a Clubhouse discussion and through other forums. I am summarising some of the feedback and questions raised here:

  1. When we talk about regulating nitrosamine impurities, we also need to talk about far more basic problems with the Indian drug supply

Nitrosamine impurities are a bit of a ‘cutting-edge’ problem, whereas the India regulator and the Indian Pharmacopoeia are yet to fix basic drug-quality problems like the enforcement of Good Manufacturing Practices or the use of up-to-date technologies to test for other, arguably more toxic, impurities.

Think of the diethylene glycol (DEG) poisoning that killed atleast 20 children in Jammu and Himachal Pradesh in 2020. DEG is a deadly chemical, which destroys the kidneys, and the children had consumed a cough syrup contaminated with it. DEG is an extremely well known pharmaceutical adulterant, there are suitable methods available to test for it, and Indian law has safeguards to prevent such poisoning. In other words, DEG is not like nitrosamines – it’s a far older story.

And yet, more than eighty years after the first recorded case of drug contamination with DEG in the USA, in which over 100 people died, India is still seeing children die from DEG-contaminated medicine. We already know that many very basic GMP failures occurred in the Jammu-Himachal-DEG case (look out for my article on this topic, coming soon). When we aren’t ensuring such basic quality checks yet (i.e pharma companies testing their raw materials for well-known toxins like DEG), can we really expect pharma companies to respond to the nitrosamine problem?

What’s more, nitrosamines are not the only area in which the CDSCO and the Indian Pharmacopoeia Commission (IPC) are lagging behind. The Indian Pharmacopoeia is frequently slow in keeping up with global scientific and regulatory changes, a pharma expert who runs an API/formulations firm told me. They cited the example of so-called elemental impurities. Elemental impurities, as the name suggests, are a range of elements that shouldn’t be present in medicines, because they are toxic and not of much therapeutic use. The most toxic among them are heavy metals like lead and arsenic, which can creep into medicines through mined excipients like talc and titanium dioxide.

Until very recently, most global pharmacopoeias used a now obsolete, 100+ year old test for these impurities. This test was neither particularly sensitive or specific – it couldn’t differentiate heavy metals from each other. To conduct it, a chemist must visually compare the colour of a solution containing the elemental impurity with the colour of a lead sulphide solution. If the colour of the test solution is lighter than the lead sulphide solution, it means total amount of heavy metals is within limits. One can imagine how crude this test would be; the expert told me that its results vary widely, depending on the lab lighting, glassware quality or even the eyesight of the person performing the test!

This is why the world is moving to better technologies and methods for elemental impurities. In a process that began almost a decade ago, and culminated in March 2019, the International Council for Harmonisation of Technical Requirements for Pharmaceuticals for Human Use (ICH) adopted new guidelines for measuring them. These guidelines set far more stringent upper limits for 24 elemental impurities, including many that cannot be detected in the color test. The colour test cannot detect the low levels prescribed by ICH anyway; if a pharma company wants to adhere to these limits, it must adopt more modern methods such as Inductively Coupled Plasma – Mass Spectroscopy (ICP-MS). Moreover, the ICH-Q3D, as the ICH’s elemental impurity guidelines are called, recommend a risk-based approach to testing. Instead of testing every product for every element, whether nickel, cadmium or arsenic, a manufacturer is advised to follow a more sophisticated approach: they must first assess the risk that the impurity is present, and then device a testing strategy for it.

Soon after, and in some instances, before, the ICH-Q3D was adopted, ICH members, like Europe, Brazil, the US and Japan have adopted these guidelines and updated their pharmacopoeias. The US subsequently knocked out the 110-year old color test for heavy metals. But India, as usual, is lagging behind. It continues to use the outdated color test, which also makes life hard for manufacturers who supply to US or Europe. Even though they may have conducted risk assessments for their medicines, and tested the medicines using ICP-MS, they still have to perform the obsolete color test to satisfy Indian regulators. “Liters of lead solutions used needlessly. Not good for the environment also,” the expert told me.

This is not the only example. There are many more monographs in the Indian Pharmacopoeia that have failed to keep up with global changes. Nitrosamines are just the latest in this line. More than one expert told me that given the costs and complexity of controlling for nitrosamines, the IPC and the drug regulator may drag their feet on bringing in new regulation, because many small Indian firms may not be able to keep up.

Still, there is some movement on the elemental-impurity front. The IPC website has published a general chapter on elemental impurities, along the lines of the ICH-Q3D, for comments from stakeholders. It’s hard to say if this general chapter will become official soon though, because complying with it would require small and medium manufacturers to conduct risk assessments, and perhaps invest in new infrastructure, like ICP-MS systems. All that is not necessarily easy.

2. How much will it cost Indian manufacturers to keep nitrosamine contamination in check?

When it comes to quality in the Indian market, the issue of cost is never far behind. If India adopts the FDA’s and EMA’s approach to nitrosamine contamination, all pharma companies will need to assess the risk that *any* of their products can potentially contain nitrosamines. This will, no doubt, cost small and medium firms a pretty penny. And India is a price sensitive market, with government programs like Jan Aushadhi trying to drive prices down further, without necessarily keeping an eye on quality.

Here is a look at why controlling for nitrosamines is neither easy, nor cheap for manufacturers. Consider a pharma firm, say ABC, which manufactures some 100 formulations and 50 APIs. For every formulation, this company will need to contact its API supplier and its excipient supplier, to understand the risk that these raw materials contain either nitrosamines, or the building blocks of nitrosamines (nitrites, amines etc). These building blocks could be present in the API or the excipient, or they could be impurities in the API or excipient.

Recall that this is how NDMA ended up in Zhejiang Huawei’s valsartan – a chemical (sodium nitrite) used to quench leftover sodium azide reacted with an impurity in the solvent dimethylformamide. In other words, NDMA was formed in a side reaction of the manufacturing process. To predict this kind side reaction, a manufacturer must have in-depth knowledge of their synthetic route, and every possible impurity that can arise from it. And I am told that many manufacturers, especially small firms who don’t invest in research, don’t have such knowledge.

Once manufacturer ABC completes its risk assessment and finds that one of its drugs has a high chance of being contaminated by, say, NDMA, they must test a few batches of the drug to show that the total NDMA levels are a small percentage of the limit allowed by regulators. If this testing reveals that NDMA is approaching the upper limit in multiple batches, ABC must start testing every batch of the final product, routinely.

To do this, it must invest in expensive apparatus – maybe a liquid chromatography-high resolution mass spectroscopy system (LC-HRMS), that costs upto Rs 2 crore. Then it must find a skilled professional to operate it, and that skilled professional must develop an appropriate method to test for NDMA in the drug. This is because a pharma company cannot directly copy publicly available testing methods: each drug and formulation requires a custom-test.

ABC must repeat this process for all of its 100 formulations and 50 APIs. As I reported in Mint, even large pharma companies like Sun Pharma and Zydus Cadila, are struggling with this challenge. How will small and medium firms deal with it?

Which brings us to the issue of price. Any pharma company that does all this, and invests in new skills and infrastructure to control nitrosamines, may want to raise prices. Will Indian consumers accept this increased cost? Will government schemes like Jan Aushadhi pay for it?

I think the answer to this question will determine whether Indians will ever get nitrosamine-free drugs.