Tuberculosis Cohort Avicenne Hospital

Globally, an estimated 10.0 million cases and 1.4 million deaths of tuberculosis diseases (TB) were reported in 2019. Despite a rate of latent tuberculosis estimated at 23% of the world population, there are very large disparities in incidence between continents and countries with an incidence of less than 10 cases per 100,000 inhabitants in Western Europe to more than 500/100,000 for countries like South Africa, the Philippines or Mozambique. Alt-hough the European Region accounted only for 2.5% of all cases, TB remains a common in-fection in immigration and poor areas. The overall objective of the WHO by 2035 (The End TB strategy) is to reduce the number of deaths from TB by 95% (compared to 2015) and to reduce the incidence rate of TB by 90% to less than 10/100,000 people. In France, there is a national notification rate of 7.6/100,000 but the North of Paris is the area of France with the highest incidence of tuberculosis (25.9/100,000 for 2016-2018). The decreasing notification rates observed in most countries are reassuring, but annual rates of decline are still insufficient to achieve the WHO target of TB elimination by 2050 in European low-incidence countries. TB control requires the early detection and treatment of patients, and investigation of potentially exposed contacts. These measures can prevent ongoing transmission of the infection and the development of drug resistance. Therefore, surveillance of TB treatment outcomes is also of great importance when evaluating TB programs. Since 2014, the target set by WHO is 90% of treatment success among new sputum smear-positive TB cases. In 2017, 30 of 31 EU/EEA countries notified TB 55,337 TB cases. Only two-thirds (67.6%) were treated successfully, and 7% died during TB treatment. Indeed, treatment outcome monitoring (TOM) is part of mandatory notification in France since 2007, with treatment outcome forms completed by clinician at 1 year. Unfortunately, in France, the available information on treatment outcome is limited by frequent missing data with only 64.8% of treatment outcomes files completed in 2018. Furthermore, thorough knowledge of a specific country's TB disease epidemiology is essential to map out a comprehensive national strategic plan for TB and could help identify determinants for unfavorable treatment outcome or loss to follow-up. The healthcare pathway of patients with tuberculosis presents glaring shortcomings with a large proportion of treatment outcomes unknown, including a large number of loss of follow-up resulting in secondary transmissions, recurrences of tuberculosis and emergence of resistance. Unknown treatment outcomes correspond to unspecified treatment outcomes 12 months from the start of treatment (lost to follow-up, transfer, absence of information). Health actors must mobilize to better understand the characteristics of patients with incomplete treatment outcomes and provide solutions to achieve better control of tuberculosis in our region. We also want to assess the relationship between plasma concentrations of major anti-TB drugs (rifampicin and isoniazid) and treatment outcomes.

We propose to constitute for the first time in France a cohort of patients with tuberculosis disease in order to meet these objectives.

The princeps study would recruit 75 patients per year for 4 years associated with 1 year of follow-up for a total duration of the study estimated at 5 years.By setting up a prospective cohort study of patients with tuberculosis disease in northern Paris, we aim to assess tuberculosis treatment outcomes and their determinants in the area with the metropolitan French highest TB incidence. At the same time, we want to evaluate the socio-demographic and clinical characteristics of our patients as well as the relationship between plasma concentrations of major anti-tuberculosis drugs (rifampicin and isoniazid) and treatment outcomes.

Methods Study population and data collection An observational, prospective, monocentric cohort study of adult patients treated for TB disease will be carried out in 1 centre in the north of Paris: the department of infectious diseases and respiratory medicine of Avicenne hospital (Bobigny).

Sociodemographic, physical, biological, and radiological data will be collected for patients meeting the inclusion criteria according to the study schedule: day 0 (D0), D15, month 1 (M1), M2, M6 and M12. The primary endpoint will be an unfavorable treatment outcome at 12 months. The categories of treatment outcomes were defined by adapting European and WHO recommendations to the French context. A patient who completed treatment within 12 months was considered to have favorable treatment outcome even if he did not attend the visit at M12. The secondary endpoints will include socioeconomic and clinical characteristics of TB cases, and rifampicin and isoniazid pharmacokinetic.

The inclusion criteria are TB out- or in-patient newly treated in one of the inclusion center for pulmonary or extra-pulmonary TB confirmed bacteriologically (direct examination, Gen-eXpert MTB/RIF and/or positive culture), suggestive histology (epithelioid gigantocellular granuloma with or without caseous necrosis), or strong clinical suspicion of tuberculosis not yet biologically confirmed but with a decision to introduce treatment. The non-inclusion criteria are: opposition to participate in the study, anti-tuberculosis treatment for more than 7 days at the inclusion visit and participation in another research protocol.

Data collected on inclusion will be: demographic, social and professional data (sex, age, country of birth, date of arrival in France, stay in a foreign country in the last 2 years, family situ-ation, profession, domiciliation, health insurance coverage, languages spoken, co-infections, clinical history, addictions (tobacco, alcohol, drugs), WHO performance status, clinical data (co-infections, WHO performance status, history of tuberculosis, tuberculosis location, presence of cough, fever, sweating nocturnal, diagnostic context), biological, microbiological and radiological data for monitoring the treatment and its tolerance. Pharmacokinetic data will be also collected on Day 15 from the start of treatment with peak dosing of rifampicin and isoniazid 3 hours after taking the treatment. The follow-up for each patient will last 12 months. Data will be entered into an electronic data capture system.

Statistical analysis Qualitative variables will be described as frequency (%) and quantitative variables as median and interquartile range (IQR). Treatment outcome at 1 year and associated socioeconomic and clinical factors will be studied by multivariate analysis. Univariate and multivariate analyses will be performed using logistic regression. All statistical tests will be based on two-tailed p values, with p<0.05 considered to indicate statistical significance. Associations will be expressed using crude and adjusted Odds ratios with 95% Confidence Intervals. In our final model, we will include all statistically significant after a backward selection using a threshold of 0.20 for the p-value. All analyses will be performed using RStudio statistical software (Version 1.4.869 © 2009-2020 RStudio, Inc).