PLHIV at all stages have developed TB. The incidence of TB is 7.68 cases per 100 PLHIV year. Incident cases are high in patients aged 18 to 44 years (15.74 per 100 PLHIV-year), men (18.75 per 100 PLHIV-year) and patients with a CD4 less than 350/mm3 (19.01 per 100 PLHIV-year). This incidence is very high in non-adherent patients (28.57 per 100 PLHIV-years), pre-ART patients (38.99 per 100 PLHIV-years), WHO stage III and IV patients (61.80 per 100 PLHIV-years) and patients with weight loss (94.28 per 100 PLHIV-years). Among PLHIV on ARVs, half of the incident cases (n=41/81, or 50.7%) occurred during the first trimester of follow-up. WHO clinical stages III and IV, lack of ART, poor adherence and weight loss were the risk factors for TB in PLHIV.
This study has limitations. The first difficulty of this study is related to missing data due to its retrospective nature. However, even though the follow-up time was not long, almost a quarter of the HIV-infected patients (n=144 or 21.27%) had enrolled for treatment when they were already in stage III and IV of the disease. At the time of data collection, information on the weight of patients was well documented; however, data on the height of patients was only partially available. This did not allow us to take into account the body mass index in the study of risk factors. Therefore, information on dietary habits (tobacco and alcohol) as risk factors was not collected from PLHIV. PLHIV in communal settings (prisons, IDP sites and others) have a higher risk of developing TB than those living outside these settings [5,6]. We did not have sufficient information on the place of residence or occupation, or the level of education. However, according to the literature, precariousness, other infectious diseases, alcoholism and chronic illnesses are all factors that contribute to the lowering of the immune system and can influence the extra pulmonary spread of TB [5,6].
The incidence of TB, which was high in patients on ART in the first quarter of treatment, declined over time (Figure 2). We believe that the risk of contracting TB in PLHIV decreases as the time to start ART increases. Koch's bacillus (KB) is most likely already present but quiescent even before HIV infection in a hyper-endemic TB country like CAR.
The delay in ARV treatment probably facilitates the clinical expression of BK. The high annual incidence of TB and the predominance of the pulmonary form can be explained by the high prevalence of TB in CAR, which is predominantly pulmonary between 2008 and 2012 (data from the National TB Control Programme, 2012 Review, CAR). This high incidence of TB among PLHIV during the first quarter of ART is in line with the results of studies conducted in Côte d'Ivoire , Burkina Faso [8-10], South Africa [11,12], Uganda , Niger , and the study by Bonnet et al. conducted in five countries (Cambodia, Thailand, Kenya, Malawi and Cameroon) . This high incidence of TB thus reflects the need to introduce isoniazid preventive treatment among PLHIV in CAR as soon as ARVs are introduced. The deflation of incidence as one moves away from the date of ART initiation was also reported by Foucher and colleagues in Niger, who found a regressing incidence (5%) after 6 months of ARV treatment . Dembélé et al. in Burkina Faso reported the same for the decrease in incidence with 2.23 per 100 people after 180 days, 1.21 per 100 people after 365 days and 0.18 per 100 people after 365 days . The restoration of the immune system induced by antiretroviral treatment could explain the decline in TB incidence . In contrast, in Uganda, the incidence of TB in one cohort remained high beyond 365 days (2.18 per 100 PLHIV) after initiation of highly active antiretroviral therapy . The high prevalence of HIV, known in the literature as an aggravating factor in TB, would explain this situation in this country . WHO clinical stages III and IV, delay in ARV treatment and low body mass index (BMI ≤ 18.5) or wasting, considered as risk factors in our study, have also been reported in studies conducted in Niger, South Africa, Burkina Faso and Uganda [8-14]. Low CD4 T-cell count as a risk factor in some African studies was not identified as a risk factor in our results [9, 14, 15]. Male sex was considered a risk factor in the studies by Poda et al in Burkina Faso . In our study the risk was not significant (RR=1.33, p=0.06). Nationally, as in our study, the female sex among all PLWH was predominant (data from the National TB Control Programme, 2012 Review, CAR) and the most affected by TB among the 104 PLWH. In contrast to our data where female HIV-TB co-infected patients were predominant, male predominance of co-infections was reported in studies conducted in Africa [9,17-19] and Brazil [20,21]. The slightly higher incidence of TB in men than in women is consistent with the findings of Dembélé and colleagues . Incident TB cases in patients with CD4 less than 350/mm3 were higher than in patients with high CD4 (˃350/mm3); this is consistent with the literature [9,14]. The majority of PLWH had TB in WHO stages III and IV combined (85.58%), a result comparable to previous studies [9,14]. Patients with less advanced WHO stages (I and II) reported TB. This reflects the fact that due to immunosuppression, TB can occur at any stage of the disease . The predominance of microscopy-positive pulmonary TB (MPT+), followed by microscopy-negative pulmonary TB (MPT-) and extra pulmonary TB (EPT), respectively, is in line with national NTP data from 2008 to 2012, which place MPT+ in first position, followed by MPT- and EPT. Similar results have been found in Africa [10, 17, 23, 24], in contrast to studies in Niger which showed a predominance of EPT [14,24].
The incidence of TB in pre-ART PLHIV was almost 3 times higher than that of PLHIV on ART. According to WHO, in the context of monitoring TB/HIV activities, access to ARVs can reduce the incidence of TB in PLHIV by 50-80%. Similarly, isoniazid preventive treatment when given to PLHIV without active TB reduces the incidence of TB by 50% (WHO Collaborative Guide for TB/HIV Activities 2004). The incidence of TB in a study by Dembelé et al. in Burkina Faso where all PLWH were on highly active antiretroviral therapy was lower than ours . In our study the majority of pre-ART patients developed WHO stage III TB. The incidence of TB is high in this pre-ART group although they are less numerous than PLWHA on ART. The absence of ART in some PLWHA (59 or 8.71%) and the advanced WHO stage would explain this high incidence of TB at the Bangui CNRISTTAR.