A Reply to Anand Grover on the Bedaquiline Issue

In a piece published on June 13, 2018 on the Wire, Anand Grover takes issue with several pieces that I have written regarding the approval of bedaquiline, a new drug meant to treat a certain type of multi drug resistant tuberculosis (MDR). Surprisingly, he doesn’t refer to my 6^th piece on the issue, published on Newslaundry, where I discuss the law on clinical trials.

While I’m glad that we are debating this issue, I disagree with Grover on multiple issues.

Cherry picking data on bedaquiline

To briefly recap the facts of this debate – Janssen & Janssen a pharmaceutical company discovered a new drug called bedaquiline. It is one of the first new drugs developed in the last 40 years that shown promise in treating MDR. The government is controlling access to the drug by administering it only through government hospitals – the data collected from these government hospitals is then given to Janssen (both parties signed a data sharing agreement). Like Grover, there are several other activists who are arguing for more patients to be given access to the drug on the grounds that it is the last hope for patients with XDR i.e. extensively drug-resistant tuberculosis. The problem however is that the drug has only cleared Phase II trials and is still undergoing Phase III trials which will conclude only in 2021.

This is an important point to note because a lot of drugs that clear Phase II, fail Phase III trials. An industry study revealed that only 55% of drugs clear Phase III trials because these trials are more rigorous since they have more patients and are meant to conclusively establish safety and efficacy of the drug. Since bedaquiline hasn’t completed Phase III trials, the medical community simply lacks the data to conclusively determine the safety and efficacy of the drug. One of the reasons this drug needs more safety data is because the cohort that was administered bedaquiline in the Phase 2 clinical trials reported a larger number of deaths than the cohort which did not receive the drug. While causation is yet to be proved between the drug and the deaths, there is enough reason to be apprehensive about the safety of the drug since the clinical trials have also reported that a side-effect of bedaquiline is increased Qt prolongation of the heart, which basically means that there were some alterations to the electrical activity of the heart. The Phase II study can be read here.

In countries like the US, which have granted bedaquiline conditional approval (a regulatory pathway that doesn’t exist in Indian law), Janssen is forced to carry a blackbox warning on the packaging of the drug stating the following: “An increased risk of death was seen in the SIRTURO treatment group (9/79, 11.4%) compared to the placebo treatment group (2/81, 2.5%) in one placebo-controlled trial”. The USFDA also requires Janssen to inform patients that the safety and efficacy of the drug is yet to be established. On the other hand, the consent form drawn up by the Indian Government for Indian patients being treated at government hospitals is entirely silent about these adverse side-effects. The consent form can be accessed here at page 4 and the patient information booklet is in the preceding pages – neither mention the past history of this drug.

Notwithstanding these serious issues raised by medical practitioners and regulators, activists like Grover describe bedaquiline in glowing terms. Grover justifies his enthusiasm on the grounds that the World Health Organisation (WHO) has recommended the drug. It should be stressed that the WHO guidance is ‘INTERIM’ in nature which means that even the WHO is yet to form a final opinion and will not do so until it has Phase III data. In the case of another new TB drug called delaminid, the WHO revised its interim guidance after the drug disappointed in Phase III clinical trials. So, I would take the interim guidance with a pinch of salt.

Grover also claims that the WHO’s latest assessment shows that 63% of the patients administered bedaquiline were cured. That is not an accurate description of the WHO’s report because bedaquiline was only one of the drugs in a multi-drug treatment i.e. bedaquiline is being prescribed along with other drugs. The same WHO assessment also mentions that 30% of the patients experienced Qt prolongations – meaning that the electrical activity of their heart was being affected as reported in earlier trials. The good news however, from the same document, is that there were no mortalities in the later trials and hence the WHO decided to downgrade the risk of undesirable effects from large to moderate, which although better, still represents a risk.

The question that flows from above discussion is how did Indian authorities approve this drug when Phase III trials were not complete especially when the drug has known safety issue? The issue of approval is different from the issue of whether the drug could be given to MDR patients in India if it was not first approved by the DCGI but since the latter issue has caused more concern let me tackle that issue first.

Could bedaquiline have been provided to patients in any other manner?

On reading Grover’s piece, one gets the sense that bedaquiline could not have been provided to patients unless the DCGI approved it immediately without waiting for results from Phase III clinical trials that are due in 2021. As I explained in the Newslaundry piece, it was perfectly possible for the Government of India to sponsor a clinical trial for all the patients. In the last 3 years the government has provided bedaquiline to approximately 1000 patients through six hospitals – that’s about the size of a large clinical trial.  If the government had sponsored the trial and registered the same as per the law it would have been forced to safeguard patient rights. This would have meant a right to compensation, under Rule 122DAB, in case the drug caused the death of patients or any other injury to them (as of now the bedaquiline consent form at govt. hospitals forces patients to waive any right to compensation). The entire treatment plan would have been vetted by an ethics committee. Any serious adverse events during the course of the trial would have to be mandatorily reported to the DCGI. The informed consent form would have to meet the requirement of Appendix V to ensure that patients had complete information. These are just some of the safeguards that would have been in place if the treatment was granted through a clinical trial process. Many of these safeguards have been inserted after the HPV scandal – people have lost their lives before these safeguards were inserted into the law.

In my interaction with one of my other critics, she claimed that only Janssen as the patentee and not the government could have sponsored a clinical trial. This is not true – as long as an institution has access to the drug it can seek approval to carry out a clinical trial. There is no requirement under the law for only the patentee to carry out the trial. In the case of bedaquiline, Janssen is anyway donating the drug to the Indian government for free. If the government had access to the drug, it should have nominated a hospital or research agency to sponsor a clinical trial. Instead, we have a scenario where Indian patients are being administered a drug by the government in an unethical manner while activists cheer on the government.

The regulatory approval granted to bedaquiline

In his piece Grover disputes my interpretation of the law, in particular my assertion that Phase III trials cannot be waived under Rule 122A of the Drugs & Cosmetics Rules, 1945 and that the DCGI could not have approved bedaquiline until Phase III trials are complete. He contradicts my conclusion by saying that “Nowhere do the rules require that data from Phase III clinical trials be available to the DCGI. They only say that relevant “data is [to be] available from other countries.” In plain English, Grover makes the point that as long as “data” is available it is of no consequence whether that data is from a clinical trial or from other sources – to burnish his point he cites data from the WHO to claim that the data was in fact available.

His interpretation of the law is incorrect, as I will explain later, but let’s presume for a moment that he is correct in his interpretation that only data is required and not clinical trials. The problem with the “data” that he quotes is that the DCGI approved Bedaquiline on 14 January 2015 while the WHO data that downgraded the risk associated with bedaquiline from “large” to “moderate” is from 2017. In any event, it is obvious from the minutes of the Technical Committee that the government considered only the US and EU approval and not WHO data while deciding to waive clinical trials. Even on this count, the Technical Committee was incorrect in its assessment because Bedaquiline got only conditional approval after Phase II trials data in both jurisdictions – Phase III trials were not complete and hence bedaquiline could not get final approval.  The Committee also appears to be ignorant of the difference between conditional approval and final approval. Don’t forget that the USFDA made it clear to Janssen that it had to inform patients that the safety and efficacy of the drug was not established since Phase III trial data was pending. How can the Technical Committee depend on such an approval to waive clinical trials for a drug with known safety issues?

Rule 122A read with Schedule Y mandate Phase III clinical trials

Grover also argues that Rule 122A never mentions the phrase “clinical trial” and that it only requires “data” which can come in other forms and not just through clinical trials. Grover has clearly missed the part in Rule 122A that requires all new drug applications to be accompanied by “information and data as required by Appendix I or Appendix IA of Schedule Y, as the case may be.” Schedule Y lists all data that is to be submitted to the DCGI – in particular Clause 1(iv) of Schedule Y, identifies the following as relevant data: “…human clinical pharmacology data as prescribed in Items 5, 6 and 7 of Appendix I”. A reference to Items 5, 6 & 7 of Appendix I indicates that those items reference Phase I, Phase II & Phase III clinical trials as described in Schedule Y which can be accessed here. It is thus clear that data for Phase I, Phase II and Phase III clinical trials is required to be submitted to the DCGI.

I therefore do not understand how Grover is claiming that Rule 122A requires only “data” and not “Phase III clinical trials”.

Can Phase III trials be waived?

The issue now is whether Phase III trials can be waived. Under Indian law, Phase I trials are safety trials meant to establish the toxicity of the drug on healthy volunteers. Phase II is a therapeutic exploratory trial on a small number of patients to understand the efficacy of the drug. If Phase II results are promising the drug moves to Phase III trials which are therapeutic confirmatory drugs that establish the efficacy of the drug on a large number of patients and helps determine the risk/benefit of the drug.

Phase III trials in India are usually split into two parts: clinical trials conducted outside the country and local clinical trials conducted within the country. Schedule Y, states in pertinent part that “For new drugs approved outside India, Phase III studies need to be carried out primarily to generate evidence of efficacy and safety of the drug in Indian patients when used as recommended…”. These are the local clinical trials referred to in Rule 122A.

Thus, even if a drug has been approved in foreign jurisdictions based on clinical trials in those countries, the law requires the drug to be tested on an Indian population to validate the drug in the Indian context. This requirement for local clinical trials however can be waived under the proviso to Rule 122A provided the drug is already approved in foreign countries. The proviso states in relevant part:

“Provided that the requirement of submitting the results of local clinical trials may not be necessary if the drug is of such a nature that the licensing authority may, in public interest decide to grant such permission on the basis of data available from other countries”.

It should be noted that the above Proviso does not state “Phase III trials” may be waived – it only states that “local clinical trials” may be waived, which clearly means that the rulemaker was expecting some kind of Phase III trials. Thus, if the drug had approvals from the US or EU based on Phase III data the local clinical trials in India could be waived. But as discussed earlier, the US and EU approvals were conditional since their regulators were still awaiting Phase III data.

Apart from the proviso to Rule 122A, Grover in his latest piece makes a new argument by citing the following provision from Schedule Y to defend the DCGI’s decision:

“For drugs indicated in life threatening/serious diseases or diseases of special relevance to the Indian health scenario, the toxicological and clinical data requirements may be abbreviated, deferred or omitted, as deemed appropriate by the Licensing Authority.”

He claims that the bedaquiline fits the above criteria and hence he supports the DCGI’s decision to approve the drug sans trials. Open ended provisions like the one above do not give the government a blanket licence to act as it pleases. As Grover will surely admit, all administrative decision making has to be reasonable not capricious or arbitrary. In the case of bedaquiline, when the USFDA and WHO have both flagged the safety concerns, it is rather incredible for Grover to cite the above provision to defend waiving all clinical trials of a drug that demonstrated cardiotoxicity in Phase II trials. What then is the point of clinical trials and drug regulation if the regulator can waive trials of a drug with known safety issues? What of the human rights of the patient?

Now that a former UN Special Rapporteur on the Right to Health has offered the above provision as a get out of jail free card expect more pharmaceutical companies to invoke the provision to seek complete waivers.