Bedaquiline is an anti-tuberculosis drug that is still being tested in clinical trials. It is not yet approved anywhere in the world, though it is steadily making progress towards approval in Europe and the United States.[^Janssen2012] Several organisations, including the Treatment Action Campaign, the Treatment Action Group, HIV i-Base, the Global Tuberculosis Community Advisory Board, Medecins Sans Frontieres and the Southern African HIV Clinicians Society have called for the drug to be made available to patients with drug-resistant tuberculosis (TB) before it is approved. This demand was made as far back as the World Lung Conference in Mexico in 2009. Yet little progress towards pre-approval access has been made in South Africa. The South African medicines regulatory authority, the Medicines Control Council (MCC), has responded sceptically. We make the case for pre-approval access in this article. Although we deal with bedaquiline specifically, the arguments made here can perhaps be applied to other drugs, including another new anti-TB drug, delamanid, which is also at an advanced stage in clinical trials. Pre-approval access has also been called compassionate care access and expanded access. We think pre-approval access is the most dispassionate and accurate way to describe what we are calling for. By pre-approval access we mean making bedaquiline available, with reasonable conditions, to patients with drug-resistant TB before the drug is approved. In South Africa the obvious mechanism for doing this would be through Section 21 authorisations approved by the MCC.[^Section21] Patients who immediately need life-saving treatment might also access the drug by taking part in a trial that has been proposed to examine the drug's safety. Pre-approval access has risks. While we have some understanding of bedaquiline's safety and efficacy, there is not yet enough data to say with the utmost confidence that it is safe and effective. Prescribing it to patients with drug-resistant TB is undoubtedly risky. The MCC is understandably cautious about pre-approval access. No phase III clinical trial of bedaquiline has been completed. Although results of a randomised controlled Phase II trial of bedaquiline have been published, the trial was small. Also, there is a clamour for the MCC to register thousands of dubious treatments for numerous ailments even though they have not been tested properly. It is understandable that the MCC is worried that giving pre-approval access to one drug will open the door for other drug-manufacturers to demand pre-approval access.[^Goldacre2012] These are legitimate concerns. Making medicines available to patients before they are properly tested, evaluated and approved is not something that the MCC can do at a whim. Nevertheless, we believe the arguments for making bedaquiline available are compelling and outweigh the concerns. ## The arguments for pre-approval access ### It is reasonable for patients with drug-resistant TB to choose to take experimental medicines Drug-resistant TB treatment outcomes vary a lot from location to location. Nevertheless, the risk of treatment failure and death is high in all settings. In a recent meta-analysis of over 9,000 patients with multi-drug resistant (MDR) TB from 23 countries, 46% died, relapsed, failed treatment or defaulted.[^Ahuja2012] Several studies of drug-resistant TB in South African settings have been published. In a study of a Durban cohort of 60 people with extensively drug resistant (XDR) TB, 25 (42%) died. Only 12 (20%) sputum-converted.[^Odonnell2009] A study of the Tugela Ferry TB register for the years 2007 to 2009 found that at one year, mortality with XDR TB was 82% and with MDR TB it was 69%. Over time MDR TB mortality dropped from 87% to 45%, perhaps reflecting improved treatment, but there was no significant decrease in XDR TB mortality.[^Gandhi2009] In Kwazulu-Natal, health workers are at particularly high risk. A study by Max O'Donnell and colleagues found that the incidence of MDR-TB hospitalization was about 65 per 100,000 health care workers versus 12 per 100,000 for non-health care workers (IRR: 5.46, 95%CI: 4.75-6.28). For XDR TB it was 7 per 100,000 health care workers versus 1 per 100,000 non-health care workers (IRR: 6.69, 95%CI: 4.38-10.20).[^Odonnell2010] The state has a duty to do everything it can for its employees who have likely become ill because of their work. The current standard of care is a great burden for patients. The World Health Organisation (WHO) guidelines recommend a 20 month treatment regimen. Five recommendations are given for composing the regimen. All the recommendations are graded *very low quality evidence* which is the WHO's lowest level of evidence. This grading means, "Any estimate of effect is very uncertain."[^WHO2011] The side-effects of the standard drugs used to compose MDR TB regimens are awful. In one South African MDR TB cohort, more than half the patients taking aminoglycosides became hearing impaired.[^Harris2012] In another, 28% had severe adverse events.[^Odonnell2009] In a Turkish cohort, side effects were severe enough to cause drug changes in more than half the patients.[^Torun2005] This is how an MDR TB patient, Colisile Lushaba, who kept a blog of her progress, described taking her regimen: > I take 10 pills in the morning, together with the injection. In the evenings I only take antiretrovirals. Although I am slightly better than I was at the start of my treatment, I am not yet feeling very well. I am still coughing, though not as much as I used to. I still vomit a lot, especially after taking the pills. I have to say that this treatment is very difficult and, right now, I am still feeling very weak and unsure of where I will be in the next few months.[^Lushaba2011] In a nutshell, patients on MDR TB regimens typically take a side-effect ridden regimen with a poor evidence base for more than a year and a half and still only have a slightly better than half chance of a successful treatment outcome. Mortality and morbidity are extremely high. Given this situation, it is reasonable for patients with drug-resistant TB to consider taking experimental medicines that have some good quality safety and efficacy evidence. Below, we explain that bedaquiline is such a medicine. The MCC should be more open to considering pre-approval access for people facing high morbidity or mortality on current best approved treatments. This is not a frequent situation. It no longer applies to diseases like HIV or diabetes for example. But it does apply to drug-resistant TB. ### The evidence shows bedaquiline is promising The evidence showing that bedaquiline's benefits outweigh its risks needs to be developed, but so far it is promising. Bedaquiline is a diarylquinoline and is the first drug in this class for the treatment of drug-resistant TB, so there is unlikely to be any existing resistance to it. Bedaquiline inhibits mycobacteria ATP synthase. It is patented by Janssen Pharmaceuticals, a subsidiary of Johnson & Johnson.[^Patent2005] In a phase II randomised controlled trial bedaquiline was given to 23 patients and placebo to 24 patients, for eight weeks. All patients received standard MDR TB regimens as well. Bedaquiline significantly reduced the time to culture conversion over 24 weeks (HR 2.3; 95%CI:1.1-4.7; p=0.03). Nearly half the patients on bedaquiline became sputum-negative versus less than 10% of patients on placebo. The number of colony forming units was much lower in the bedaquiline arm too. Nausea was the only adverse event reported to occur significantly more often in the bedaquiline group (26% vs. 4%, p=0.04).[^Diacon2009] A study with two year follow up data of this trial has been published and confirms the initial promising outcomes.[^Diacon2012] An open-label safety trial of bedaquiline with about 200 patients has also been conducted. Janssen has presented 24 week data from this trial. The 2012 Pipeline Report explains: > The data indicated that adding bedaquiline to an individualized MDR-TB regimen was safe and well tolerated and resulted in an overall 81% culture conversion rate at week 24, with median times to culture conversion of 8 weeks for patients with MDR-TB, 12 weeks for patients with pre-XDR-TB, and 24 weeks for patients with XDR-TB.[^Pipeline2012] As far as interactions with antiretrovirals go, which is obviously an important concern in South Africa where many drug-resistant patients are co-infected with HIV, a small Phase I study found that bedaquiline is well-tolerated with efavirenz and that the effect of efavirenz on bedaquiline concentrations is unlikely to be clinically significant. Bedaquiline was associated with QT prolongation.[^Dooley2012a] The 2012 Pipeline Report further explains: > Janssen now plans to start a phase III trial of 600 subjects with sputum smear-positive pulmonary MDR- or pre-XDR-TB (confirmed by rapid diagnostic test). Participants in the first arm will receive 9 months of bedaquiline and a background regimen. Those in the control arm will receive placebo and the background regimen. Participants in a third rollover arm, which will capture the failures from the first two arms, will receive an individualized salvage regimen. The primary endpoint will be relapse-free cure at 15 months for those in the first two arms. The final analysis will look at relapse-free cure at 21 months. > Janssen is also taking into consideration TB/HIV-coinfection and pediatric drug-resistant TB in its development plans.[^Pipeline2012] ### The evidence base for bedaquiline is better than most other MDR TB drugs The South African MDR TB treatment guidelines recommend a regimen that includes kanamycin, ethionamide, pyrazinamide, levofloxacin and terizidone.[^SA2012] Other drugs like linezolid are also often used for MDR TB. The efficacy of pyrazinamide has been well established, but there is likely much resistance to it amongst MDR TB patients because it is also part of the standard first-line regimen. We searched for controlled clinical trials of the remaining drugs to treat TB, i.e. kanamycin, ethionamide, levofloxacin, terizidone and linezolid. **Kanamycin** We can find one clinical trial of kanamycin in people with TB, though it is more accurate to describe it as a prospective case-controlled study. It was published in 1958 and compared kanamycin to streptomycin in 162 patients. It does not appear to have been randomised and the two groups of patients appear not to have been matched at baseline. The data showing the effectiveness of kanamycin is no better, and perhaps worse, than the data on bedaquiline.[^Donomae1958] The available data shows that the side effect profile of kanamycin is much worse than bedaquiline. Kanamycin causes hearing problems in 3 to 10% of patients and it also causes kidney problems.[^Global2008] [^Dooley2012b] **Ethionamide** A tiny clinical trial of 27 people compared ethionamide against thiacetazone in 1963. Cycloserine was given to patients in both arms. Nine out of 14 versus three out of nine patients on the ethionamide and thiacetazone arms respectively had what the authors call "bacteriologically quiescent disease" after one year. The authors state that the ethionamide arm performed statistically significantly better than the thiacetazone one.[^Angel1963] A Japanese controlled clinical trial that compared ethionamide against prothionamide was published in 1968. However, since all patients also received isoniazid and streptomycin and nearly all patients sputum-converted in all arms of the trial, it is impossible to calculate the effectiveness, if any, of ethionamide in this trial.[^Coop1968] There are other clinical trials of drug-resistant TB patients that use ethionamide as part of a treatment regimen, but we can find no other clinical trial evidence in people with TB in which ethionamide is tested against a control. Ethionamide is associated with serious side effects. Liver toxicity in particular is common and may continue even after patients stop taking the drug.[^Global2008] It is also associated with peripheral neuropathy. **Levofloxacin** Besides a seven day early bactericidal activity trial, we can find no clinical trials of levofloxacin that have considered the drug for the treatment of MDR TB. It also has several side effects, including phototoxicity, glucose disturbances, QT prolongation and others.[^Global2008] **Terizidone** We can find one reference to possible clinical trials of terizidone in a 1972 paper written in Croation, but we are unable to get the paper.[^Zrilic1972] The evidence base for terizidone is poor.[^Dooley2012b] **Linezolid** Linezolid has been used in South Africa for drug-resistant TB for some time. Yet the first randomized controlled trial of linezolid in patients with drug-resistant TB was only published in October 2012 and it was a smaller trial (n=41) than the Phase II bedaquiline trial described above. The results are promising and significantly more patients taking linezolid sputum-converted compared to the controls, but the drug was also associated with more serious adverse events than bedaquiline.[^Lee2012] This is not a comprehensive literature review or comparison, but except for pyrazinamide the controlled clinical trial evidence supporting the use of bedaquiline for the treatment of MDR TB is better than any other medicine recommended for this indication in the South African treatment guidelines. For XDR and pre-XDR TB patients, the additional drugs that are often used have even less compelling evidence supporting their use than the ones discussed here. Surely, it is reasonable for people to request and receive bedaquiline to try and treat their drug-resistant TB infection. ## Precedents In Europe and North America there are several pre-approval access precedents, particularly with antiretrovirals. For example, more than 35,000 people received didanosine before it was approved by the FDA in 1991.[^Harrington2012] This was controversial but likely prolonged many lives. Other antiretrovirals were also available before approval. There are of course risks with pre-approval access. There were cases of didanosine-associated pancreatis during the drug's pre-approval phase. Thousands of people took adefovir as an antiretroviral but it ended up not being approved for HIV treatment. In South Africa, lopinavir/ritonavir (branded as Kaletra) was made available to patients on Section 21 authorisation before the drug was registered. Pre-approving access to bedaquiline would therefore not be an unprecedented step by the MCC. ## Responsibility for serious adverse events or treatment failure A concern with pre-approving drugs is who should bear responsibility for the risk of patients experiencing severe adverse events. It is our view that until a drug is registered, if it is used outside of a clinical trial setting as part of pre-approval access, then the pharmaceutical company that manufacturers, tests or holds the patent on the drug should not be responsible for the risk. This means that it is very important that doctors tell patients of the risk of taking an experimental drug. Patients have to be aware that the experimental drug they are taking has not yet been fully tested, that it might not work, that it might cause severe adverse events and that it might be worse than not taking it at all although this is unlikely. The one exception is if the pharmaceutical company held back important safety and efficacy data that might have influenced a patient's decision to take the experimental drug or a doctor's decisions to prescribe it. In that case, the pharmaceutical company must be held responsible. Because of the risk of taking an experimental drug and the need for patients to be properly informed of the risks, as well as the public interest to limit resistance to bedaquiline, the drug should only be made available to institutions which are likely to have the capacity to monitor patient adherence and inform them of the risk. In our view, health units run by or partnered with academic institutions and proven medical delivery organisations, for example Medecins Sans Frontieres, meet these criteria. ## The benefits of pre-approval access outweighs the risks There are benefits and risks of pre-approval access to bedaquiline. On the current evidence, the benefits of pre-approval access outweigh the risks. Pre-approval access is likely to give patients hope, increase their chances of being cured, reduce the time that they are infectious and possibly reduce their risk of death. The MCC should acknowledge the risks but give drug-resistant TB patients who attend well-run health facilities and who have been properly informed, the opportunity to add bedaquiline to their treatment regimen. **The original document was published on 24 November 2012. The following amendment was published on 21 December 2012.** # Post-publication amendment to Case for Bedaquiline Following the publication of the Case for Bedaquiline we became aware of a further phase II clinical trial of bedaquiline that we did not include. Information on this trial was made available in November in a report to the US Food and Drug Administration. [^FDA] In this trial 79 people were randomised to take bedaquiline and 81 to take placebo. All participants had MDR TB and were on a standard background regimen. Participants in the bedaquiline arm had significantly faster time to sputum conversion at 24 weeks (p<0.0001). This benefit was still present at 72 week follow-up. At week 24, 79% vs. 58% of people had sputum-converted on the bedaquiline and placebo arms respectively (p=0.008). However, a significant safety issue was reported. Ten people died on the bedaquiline arm and two died in the placebo arm (p=0.02, Fisher's exact test). This result is possibly a statistical anomaly but it is concerning and emphasises the need for a phase III study with adverse events as an endpoint. All but one of the deaths on the bedaquiline arm occurred after the drug was stopped and the time range over which the deaths occurred was wide, but bedaquiline has a long half-life and a causal effect cannot be ruled out. The only cause of death that was reported more than once was TB, which supports the possibility of the mortality result being a fluke. **End of amendment.** [^Janssen2012]: In July the drug was submitted for approval to the US Food and Drug Administration through its *Accelerated Approval* process. See *Janssen Research & Development Submits New Drug Application to FDA for Investigational Multi-Drug Resistant Tuberculosis Treatment Bedaquiline (TMC207).* [http://www.jnj.com/connect/news/all/janssen-research-and-development-submits-new-drug-application-to-fda-for-investigational-multi-drug-resistant-tuberculosis-treatment-bedaquiline-tmc207](http://www.jnj.com/connect/news/all/janssen-research-and-development-submits-new-drug-application-to-fda-for-investigational-multi-drug-resistant-tuberculosis-treatment-bedaquiline-tmc207) [^Section21]: Section 21 of the South African Medicines and Related Substances Act 101 of 1965 empowers the MCC to authorise the sale of unregistered medicine for specific purposes. This is commonly called a Section 21 authorisation. [^Goldacre2012]: This concern is articulated by Ben Goldacre with respect to the understandable and justified accelerated approval of AIDS drugs in the United States in the late 1980s and early 1990s in the United States. Unfortunately, this resulted in drug companies, sometimes aided by patient groups, using relaxed approval mechanisms to push through non-AIDS drugs that should have been better tested. Goldacre B. 2012. Bad Pharma. At page 137. [^Ahuja2012]: Ahuja SD et al. (2012) Multidrug Resistant Pulmonary Tuberculosis Treatment Regimens and Patient Outcomes: An Individual Patient Data Meta-analysis of 9,153 Patients. PLoS Med 9(8): e1001300. doi:10.1371/journal.pmed.1001300 [^Odonnell2009]: O'Donnell MR et al. Improved early results for patients with extensively drug-resistant tuberculosis and HIV in South Africa. Int J Tuberc Lung Dis. 2009 Jul;13(7):855-61. [http://www.ncbi.nlm.nih.gov/pubmed/19555535](http://www.ncbi.nlm.nih.gov/pubmed/19555535) [^Gandhi2009]: Gandhi N et al. High Early Mortality among HIV-infected Patients with Extensively Drug-resistant or Multidrug-resistant TB in Rural South Africa. 16th CROI, 2009. Poster abstract 784. [http://www.retroconference.org/2009/Abstracts/36299.htm](http://www.retroconference.org/2009/Abstracts/36299.htm) [^WHO2011]: World Health Organisation. Guidelines for the Programmatic Management of Drug-Resistant Tuberculosis: 2011 update. [http://whqlibdoc.who.int/publications/2011/9789241501583_eng.pdf](http://whqlibdoc.who.int/publications/2011/9789241501583_eng.pdf) [^Odonnell2010]: O'Donnell MR et al. High incidence of hospital admissions with multidrug-resistant and extensively drug-resistant tuberculosis among South African health care workers. Ann Intern Med. 2010 Oct 19;153(8):516-22. [http://www.ncbi.nlm.nih.gov/pubmed/20956708](http://www.ncbi.nlm.nih.gov/pubmed/20956708) [^Lushaba2011]: Lushaba, C. 2011. TB and me. Real stories of people living with multidrug resistant tuberculosis. [http://blogs.msf.org/tb/author/colisile/](http://blogs.msf.org/tb/author/colisile/) [^Harris2012]: Harris T et al. Aminoglycoside-induced hearing loss in HIV-positive and HIV-negative multidrug-resistant tuberculosis patients. South African Medical Journal, Vol 102, No 6 (2012). [http://www.samj.org.za/index.php/samj/article/view/4964/4128](http://www.samj.org.za/index.php/samj/article/view/4964/4128) [^Torun2005]: Törün T et al. Side Effects Associated with the Treatment of Multidrug-resistant Tuberculosis. The International Journal of Tuberculosis and Lung Disease: The Official Journal of the International Union Against Tuberculosis and Lung Disease 9, no. 12 (December 2005): 1373–1377. [^Patent2005]: Patent EP1527050. Filed 2002. Published 2005. Quinoline derivatives and their use as mycobacterial inhibitors. [http://www.google.com/patents/EP1527050A1?cl=en](http://www.google.com/patents/EP1527050A1?cl=en) [^Diacon2009]: Diacon AH et al. The Diarylquinoline TMC207 for Multidrug-resistant Tuberculosis. The New England Journal of Medicine 360, no. 23 (June 4, 2009): 2397–2405. [http://www.ncbi.nlm.nih.gov/pubmed/19494215](http://www.ncbi.nlm.nih.gov/pubmed/19494215) [^Diacon2012]: Diacon AH et al. Randomized Pilot Trial of Eight Weeks of Bedaquiline (TMC207) Treatment for Multidrug-resistant Tuberculosis: Long-term Outcome, Tolerability, and Effect on Emergence of Drug Resistance. Antimicrobial Agents and Chemotherapy 56, no. 6 (June 2012): 3271–3276. [http://www.ncbi.nlm.nih.gov/pubmed/22391540](http://www.ncbi.nlm.nih.gov/pubmed/22391540) [^Dooley2012a]: Dooley KE et al. Safety, Tolerability, and Pharmacokinetic Interactions of the Antituberculous Agent TMC207 (bedaquiline) with Efavirenz in Healthy Volunteers: AIDS Clinical Trials Group Study A5267. Journal of Acquired Immune Deficiency Syndromes (1999) 59, no. 5 (April 15, 2012): 455–462. [http://www.ncbi.nlm.nih.gov/pubmed/22126739](http://www.ncbi.nlm.nih.gov/pubmed/22126739) [^Dooley2012b]: Dooley KE et al. Old Drugs, New Purpose: Retooling Existing Drugs for Optimized Treatment of Resistant Tuberculosis. Clinical Infectious Diseases: An Official Publication of the Infectious Diseases Society of America 55, no. 4 (August 2012): 572–581. doi:10.1093/cid/cis487. [^Angel1963]: Angel, J.H et al. A Controlled Comparison of Cycloserine Plus Ethionamide with Cycloserine Plus Thiacetazone in Patients with Active Pulmonary Tuberculosis Despite Prolonged Previous Chemotherapy. Tubercle 44, no. 2 (June 1963): 215–224. doi:10.1016/S0041-3879(63)80115-4. [^Pipeline2012]: Clayden P et al. 2012 Pipeline Report. [http://www.pipelinereport.org/browse/tb-treatments/bedaquiline-tmc207](http://www.pipelinereport.org/browse/tb-treatments/bedaquiline-tmc207) [^SA2012]: Republic of South Africa Department of Health. Guidelines for Management of Drug-Resistant Tuberculosis in South Africa. 2010. [http://www.tbonline.info/media/uploads/documents/mdr-tb_sa_2010.pdf](http://www.tbonline.info/media/uploads/documents/mdr-tb_sa_2010.pdf) [^Lee2012]: Lee M et al. Linezolid for Treatment of Chronic Extensively Drug-Resistant Tuberculosis. New England Journal of Medicine 367, no. 16 (2012): 1508–1518. [http://www.nejm.org/doi/full/10.1056/NEJMoa1201964](http://www.nejm.org/doi/full/10.1056/NEJMoa1201964) [^Harrington2012]: Harrington M. TAG at 20 - Part II: On a Darkling Plain—The Years of Despair. tagline Fall 2012. [http://www.treatmentactiongroup.org/tagline/2012/fall/tag-20-darkling-plain](http://www.treatmentactiongroup.org/tagline/2012/fall/tag-20-darkling-plain) [^Donomae1958]: Donomae I et al. Clinical studies of kanamycin treatment of pulmonary tuberculosis. Annals of the New York Academy of Sciences Volume 76, The Basic and Clinical Research of the New Antibiotic, Kanamycin pages 166–187, September 1958. [http://onlinelibrary.wiley.com/doi/10.1111/j.1749-6632.1958.tb54705.x/abstract](http://onlinelibrary.wiley.com/doi/10.1111/j.1749-6632.1958.tb54705.x/abstract) [^Global2008]: Global Alliance for TB Drug Development. 2008. Handbook of Anti-Tuberculosis Agents. [^Coop1968]: The Co-Operative Study Unit On Chemotherapy Of Tuberculosis Of The National Sanatoria In Japan. Comparison of the clinical usefulness of ethionamide and prothionamide in initial treatment of tuberculosis: tenth series of controlled trials. Tubercle. Volume 49, Issue 3, September 1968, Pages 281–290. [^Zrilic1972]: Zrilić V et al. Results of clinical trials of a new antitubercular agent, terizidone (Terivalidine). Plucne Bolesti Tuberk. 1972 Apr-Jun;24(2):89-98. [http://www.ncbi.nlm.nih.gov/pubmed/4565191](http://www.ncbi.nlm.nih.gov/pubmed/4565191) [^FDA]: Anti-Infective Drugs Advisory Committee Meeting Briefing Document. TMC207 (bedaquiline) Treatment of Patients with MDR-TB. NDA 204-384. 28 November 2012. http://www.fda.gov/downloads/AdvisoryCommittees/CommitteesMeetingMaterials/Drugs/Anti-InfectiveDrugsAdvisoryCommittee/UCM329260.pdf