A benefit-risk trial of 1-month rifapentine and isoniazid to prevent tuberculosis and reduce morbidity in people with non-communicable multimorbidity
Lead Research Organisation:
University College London
Department Name: Institute for Global Health
Abstract
TB causes the most deaths than any other infectious disease worldwide. People who live in poor countries are most
affected; this is at a large personal and financial cost to the individuals, their families and communities. Disease can be
prevented by giving TB drugs to people who are infected with the bacteria that causes TB, this is called TB preventive
treatment.
The World Health Organization (WHO) currently recommends TB preventive treatment for people who are most at risk of
developing TB such as people who live in the same house as a person with TB, and also for people living with HIV. For
these individuals, it is acknowledged that TB preventive treatment will achieve more benefit than harm. People with
diabetes also have an increased chance of developing TB disease and dying from TB. However, WHO does not currently
recommend TB preventive treatment since it is not clear if the benefit would outweigh the risks of treatment. There is not
enough evidence to inform this decision especially since the only option on offer is the standard TB prevention treatment
with 6 or 9 months of isoniazid which is taken daily and can cause harm especially to the liver. Moreover, benefit may be
limited since providing good WHO standard care for diabetes could already reduce the risk of TB in the future. The aim of
WHO standard care for diabetes is to control blood sugar levels and by so doing, research suggests the risk of developing
TB would also be reduced. In addition, standard care for diabetes includes a medicine called metformin which has been
recently shown to reduce the risk of developing TB in people prescribed this medicine compared to those not prescribed
the medicine.
The new short-duration treatment with one month of the TB drugs rifapentine and isoniazid (1HP) was shown to prevent TB
disease in people infected with HIV. Moreover, 1HP has other benefits since more people complete the short treatment,
and it causes fewer side effects. We think this new treatment may alter the balance of benefit against the harms. This
would result in a change in WHO policy in favour of providing TB preventive treatment to people living with diabetes. The
treatment has not not been studied in people with diabetes or people without HIV. We plan to find the best strategy for TB
prevention for people with diabetes who are HIV-uninfected and likely infected with TB (as shown by a test of exposure to
TB). A positive test would indicate an increased risk to develop TB in the future. We will carry out a study that will place
people in a random manner, like tossing a coin, to a group that receives the standard care for diabetes with 1HP or a group
that receives diabetes care alone. That type of study is called a randomised controlled study. In this study, we will assess
who gets TB disease and who develops side effects from the treatments in each group. If 1HP reduces the chance of
developing TB and does so without substantial side-effects, we think WHO policy will change quickly to recommend
prevention of TB in people with diabetes. However, if giving 1HP in addition to standard care for diabetes is not better than
standard care alone, the current recommendation would not change thus avoid unnecessary harm from TB drugs. Our trial
will recruit 4,130 male and female individuals aged 15 years and older with diabetes, and who reside in Philippines and
South Africa. The project will run for 5 years.
affected; this is at a large personal and financial cost to the individuals, their families and communities. Disease can be
prevented by giving TB drugs to people who are infected with the bacteria that causes TB, this is called TB preventive
treatment.
The World Health Organization (WHO) currently recommends TB preventive treatment for people who are most at risk of
developing TB such as people who live in the same house as a person with TB, and also for people living with HIV. For
these individuals, it is acknowledged that TB preventive treatment will achieve more benefit than harm. People with
diabetes also have an increased chance of developing TB disease and dying from TB. However, WHO does not currently
recommend TB preventive treatment since it is not clear if the benefit would outweigh the risks of treatment. There is not
enough evidence to inform this decision especially since the only option on offer is the standard TB prevention treatment
with 6 or 9 months of isoniazid which is taken daily and can cause harm especially to the liver. Moreover, benefit may be
limited since providing good WHO standard care for diabetes could already reduce the risk of TB in the future. The aim of
WHO standard care for diabetes is to control blood sugar levels and by so doing, research suggests the risk of developing
TB would also be reduced. In addition, standard care for diabetes includes a medicine called metformin which has been
recently shown to reduce the risk of developing TB in people prescribed this medicine compared to those not prescribed
the medicine.
The new short-duration treatment with one month of the TB drugs rifapentine and isoniazid (1HP) was shown to prevent TB
disease in people infected with HIV. Moreover, 1HP has other benefits since more people complete the short treatment,
and it causes fewer side effects. We think this new treatment may alter the balance of benefit against the harms. This
would result in a change in WHO policy in favour of providing TB preventive treatment to people living with diabetes. The
treatment has not not been studied in people with diabetes or people without HIV. We plan to find the best strategy for TB
prevention for people with diabetes who are HIV-uninfected and likely infected with TB (as shown by a test of exposure to
TB). A positive test would indicate an increased risk to develop TB in the future. We will carry out a study that will place
people in a random manner, like tossing a coin, to a group that receives the standard care for diabetes with 1HP or a group
that receives diabetes care alone. That type of study is called a randomised controlled study. In this study, we will assess
who gets TB disease and who develops side effects from the treatments in each group. If 1HP reduces the chance of
developing TB and does so without substantial side-effects, we think WHO policy will change quickly to recommend
prevention of TB in people with diabetes. However, if giving 1HP in addition to standard care for diabetes is not better than
standard care alone, the current recommendation would not change thus avoid unnecessary harm from TB drugs. Our trial
will recruit 4,130 male and female individuals aged 15 years and older with diabetes, and who reside in Philippines and
South Africa. The project will run for 5 years.
Technical Summary
Diabetes affects millions in poor countries and increases the risk of TB ~3-fold, worsening the TB epidemic. The WHO
recommends against TB preventive treatment (TPT) unless individuals are HIV-infected or contacts of persons with TB,
given insufficient evidence to inform the benefit-risk balance for TB prevention using isoniazid for 6/9 months. Moreover,
benefit may be limited since effective mitigation of the TB risk might be achieved with optimal diabetes care which includes
metformin, a host-directed therapy.
One-month daily rifapentine with isoniazid (1HP) appears non-inferior to isoniazid in HIV. Moreover, the regimen is safe
and promotes adherence. These advantages may alter the clinical benefit-risk for prevention but this is unknown. 1HP may
likely be cost-effective than existing TPT; its ultra-short duration would promote fewer visits, limited clinical monitoring and
reduce care burden on the health system and indiviuals. If shown effective, this would make 1HP a favourable option,
particularly during the COVID-19 pandemic.
BALANCE is a 1:1 N=4,130, 5-year, multi-centre, pragmatic, randomised controlled trial of a novel WHO short-course TPT
with 1HP compared to no TPT to prevent TB and reduce morbidity in: adults with diabetes that receive WHO standard care,
are HIV-uninfected, test positive for TB infection, and reside in Philippines and South Africa. The sample size provides 90%
power to assess superiority in effectiveness, with 5% 2-sided alpha, assuming: 1.5% 2-year TB risk in controls; 70% 1HP
efficacy; 10% lost to follow-up. This is a phase 3 open-label trial that enables real-world evaluation; use of a microbiological
definition of TB would reduce bias in the primary outcome. Secondary outcomes will assess effectiveness to reduce
clinically diagnosed TB, deaths, the durability of protection, safety, benefit-risk, impact on health-systems and costeffectiveness.
Applicants' engagement in WHO guideline development will maximize impact.
recommends against TB preventive treatment (TPT) unless individuals are HIV-infected or contacts of persons with TB,
given insufficient evidence to inform the benefit-risk balance for TB prevention using isoniazid for 6/9 months. Moreover,
benefit may be limited since effective mitigation of the TB risk might be achieved with optimal diabetes care which includes
metformin, a host-directed therapy.
One-month daily rifapentine with isoniazid (1HP) appears non-inferior to isoniazid in HIV. Moreover, the regimen is safe
and promotes adherence. These advantages may alter the clinical benefit-risk for prevention but this is unknown. 1HP may
likely be cost-effective than existing TPT; its ultra-short duration would promote fewer visits, limited clinical monitoring and
reduce care burden on the health system and indiviuals. If shown effective, this would make 1HP a favourable option,
particularly during the COVID-19 pandemic.
BALANCE is a 1:1 N=4,130, 5-year, multi-centre, pragmatic, randomised controlled trial of a novel WHO short-course TPT
with 1HP compared to no TPT to prevent TB and reduce morbidity in: adults with diabetes that receive WHO standard care,
are HIV-uninfected, test positive for TB infection, and reside in Philippines and South Africa. The sample size provides 90%
power to assess superiority in effectiveness, with 5% 2-sided alpha, assuming: 1.5% 2-year TB risk in controls; 70% 1HP
efficacy; 10% lost to follow-up. This is a phase 3 open-label trial that enables real-world evaluation; use of a microbiological
definition of TB would reduce bias in the primary outcome. Secondary outcomes will assess effectiveness to reduce
clinically diagnosed TB, deaths, the durability of protection, safety, benefit-risk, impact on health-systems and costeffectiveness.
Applicants' engagement in WHO guideline development will maximize impact.
