We conducted a retrospective cohort study from January 2016 to December 2021 at the Department of Pulmonology at Sylvanus Olympio Teaching Hospital (CHUSO) in Togo. Togo is a West African country split into 5 administrative regions ( Figure 1 ). The majority of the population (62%) lives in rural areas, while 44% live in the peri-urban coastal region, which includes the capital town Lomé. The referral center for drug-resistant TB patients’ treatment and management is located in Grand-Lomé within the pulmonology department of CHUSO. According to the national TB program, all patients diagnosed with TBR should be hospitalized at least for the intensive phase of the treatment regimen and culture conversion.

All patients diagnosed with DR-TB and hospitalized during our study period were included, and their medical records were used to collect data. Patients diagnosed with at least a Rifampicin Resistance after Xpert/MTB Rif, treated with the 9-month MDR-TB/RR-TB regimen that encompassed an injectable drug, were selected for the study. The regimen was 4 months of intensive phase with amikacin, moxifloxacin, prothionamide, clofazimine, pyrazinamide, isoniazid, ethambutol (Am-Mfx-Pto-Cfz-Z-H-E) following by a 5-months continuation phase with moxifloxacin, clofazimine, ethambutol and pyrazinamide (Mfx-Cfz-E-Z). Drug dosage administration protocol followed the national TB program guidelines for TBR management [PNLT Togo].

Patients treated for XDR-TB, with oral drug regimen only, and those with missing Xpert results or outcome data were excluded.

The sampling was exhaustive, including all the MDR TB patients who fulfilled our inclusion criteria and registered during the study period. Their medical records served as support to collect demographic data such as age, gender, marital status, clinical data (symptoms, history of prior TB treatment, and BMI), baseline Blood Cell Count (CBC) results, and chest X-ray radiological characteristics were also collected. Bacteriological results with monthly sputum smear microscopy and culture, patients’ sputum conversion assessed at month 4 and month 9 of treatment, and if available, drug susceptibility test (DST) for first-line drug DST and/or second-line drug Line Probe Assay (LPA) for drug resistance patterns. Patients’ treatment outcomes were defined as:

A structured questionnaire was developed and tested for the study. Data were entered and checked on the Koboocolect system, then extracted and de-identified on an Excel sheet before analysis. Data were analyzed using SPSS 23.0.

For analysis, patients were categorized into two groups based on their history of TB treatment. Group 1 (G1) or primary/transmitted resistance for patients naïve to any first-line anti-tuberculosis drugs (new cases). Group 2 (G2) or secondary/acquired resistance for patients with a history of first-line TB treatment with the previous outcome of cured, defaulted, failed, and relapse (retreatment cases). We have also classified outcomes into two categories: good outcomes, including cured and treatment completed, and poor outcomes for treatment failure, loss to follow-up, and death.

Fisher’s exact test was used to compare proportions between the two groups. Descriptive analyses were done to determine proportions; the univariate association test was performed to assess relationships. Multivariate analysis was performed to identify the independent associated factors. Adjusted Odds Ratios (aORs) were calculated when the p-value was less than 0.05. We included variables that had a relationship with secondary Drug-Resistant TB with a p-value less than 0.05. Differences were considered statistically significant if at a 95% confidence interval (CI) the p-value < 0.05.

The use of retrospective data for this study was approved by the scientific committee and the head of the department of pulmonology of the CHUSO. Data were extracted from the TBR patient’s records or the registry and were coded before analysis to respect anonymity and confidentiality. No patients were physically seen, and no additional data were collected from them.

During four years, 96 Drug-resistant TB patients were initiated on MDR-TB treatment of whom 7 were excluded due to incomplete information, and data from 89 patients were included in the study’s final analysis.

Males were predominant in the study population with a sex ratio of 1.3. Informal working activities constituted 49.4% of the subjects. The mean age was 36.02 years (±14.21 years), ranging from 5 months to 75 years. Of these patients, 41.6% were from Grand Lomé and 27% were from the maritime region, which are both considered urban and peri-urban areas, respectively. Among the cohort participants, 31.5% patients were HIV-co-infected, and 2 of them were ART-naïve. Among the study participants, 11.2% were smokers among them 2 were current smokers. Close contact with a patient diagnosed with tuberculosis (TB) was found in 22.5%. Based on clinical records, 51.7% had received first-line TB treatment; among them 20.2% had first-line treatment failure and 5.6% had second-line treatment failure. Furthermore, 20.2% of patients relapsed after the first treatment, and 3.4% of patients relapsed after the second treatment. The mean duration between the diagnosis of drug resistance (RR) and treatment initiation was 41.7 ± 7.1 days, with a median of 22 days. The overall mean BMI of patients was 17.64 ± 3.74 kg/m² ( Table 1 ).

In univariate analysis, the following factors: male gender (OR: 2.61, 95% CI: 1.1 -6.2); delay before second line treatment initiation (OR: 2.86, 95% CI: 1.21 - 6.78); dyspnea (OR: 3.67, 95% CI: 1.20 - 11.24), HIV positive (OR: 0.31; 95% CI 0.11 - 0.79) were associated with the secondary resistance group. Poor treatment outcome was less likely to be associated with secondary resistance group (OR: 0.24; 95% CI: 0.07 - 0.83).

Considering the chest X-ray results, the following factors were found to be associated with secondary resistance group: cavities (OR: 3.75, 95% CI: 1.39 - 9.18); Nodules (OR: 3.03, 95% CI: 1.23 - 7.49), infiltrates (OR: 3.92, 95% CI: 1.52 - 10.07) ( Table 2 ).

A multivariate logistic regression was performed to assess independent factors associated with Secondary Drug resistant TB. Four independent factors were identified namely, dyspnea, lung nodules and Infiltrates at imaging and long second-line treatment delay. However, lung cavity represented 3-times higher in secondary drug-resistant TB but was not statistically significant.

The multivariate analysis test revealed that the following factors: Delay before Second-line TTT (days) (aOR: 3.04, 95% CI: 1.08 - 8.58); dyspnea (aOR: 3.88; 95% CI: 1.01 - 14.89) were found to be significantly associated with secondary resistance group as well as 2 types abnormalities on chest X-ray Nodules (aOR: 3.62, 95% CI: 1.23 - 10.67) and Infiltrates (aOR: 4.66; 95% CI: 1.51 - 14.38) ( Table 2 ).

Among the 89 MDR/RR-TB cases, 59 had DST results, and 29 benefited from LPA. Based on the DST results, 54 RR-TB patients were confirmed MDR-TB. Except for the resistance to Streptomycin (S) and pyrazinamide (Z), MDR-TB and resistance to any first-line TB drug were common in the secondary resistance group. Four (4) cases of polyresistance were found in the primary resistance group. MDR-TB was found among 91.5% of whom having DST, which is 60.6% of our study sample. Based on LPA results, there were 5 pre-XDR-TB and 2 XDR-TB, with one in each resistance group ( Table 3 ).

At baseline, 94.4% had positive sputum microscopy, while 5.6% had negative esults. Smear microscopy and culture conversion rates at month 4 were 76.6% and 92.2% of the 77 patients tested respectively. At month 9, smear microscopy and culture conversion rate were 98.6% and 98.6% of 73 patients tested, respectively.

The overall death rate during the follow-up was 15.7% of patients, among them 78.5% occurred before the month 4 visit, including one XDR-TB. Patients declared cured and treatment completed were 80.9%, and 1.12%, respectively. Lost to follow-up and treatment failures were recorded in 1.1% case each. The treatment failure case was one of the two patients who were diagnosed with XDR-TB. Globally, the case-fatality rate of resistant TB was 16.8% in our study.

This study used a quantitative approach to show differences between patients with primary and secondary drug resistance. We compared their epidemiological, clinical features, and outcomes under the 9-month second-line treatment regimen with an injectable drug. Despite a small study population, our data suggested that secondary resistance was associated with dyspnea, more radiological lung lesions, and longer duration before MRD treatment initiation.

The clinical characteristics of our population have a similar pattern to the general TB population; thus, MDR/RR-TB seems to affect male and younger people. Studies across African regions found that most patients with RR-TB are aged under 40 years [11] - [13] , suggesting that drug-resistant TB may affect patients’ capacity to work as well as a country’s economy. Tanimura et al., and Menzies et al. have found a high financial cost for TB patients living with the disease in low- and middle-income countries. Loss of work capacity and income, due to temporary disabilities and TB complication-related treatment costs, are some of the financial issues encountered by them [14] [15] . Patients living in the urban area accounted for more than half of our population. Dickson et al. in South Africa and Mbuh et al. in Cameroon found the same findings in their studies [16] [17] . In Africa, people assume that the quality of health care is better in town, thus they often move there when sick [17] . However, younger age, male gender, and inappropriate behaviors may impede this health quality access opportunity [11] [17] . Moreover, Longo et al. have found an association between urban/periurban setting and a risk of developing MDR/RR-TB in the Republic of Central Africa [12] .

More than half of our population had a history of first-line TB treatment intake. Similarly, Fara et al. in the Central African Republic found a resistance rate of up to 46.6% among those previously treated. In their study, a TB relapse was associated with a 3 times higher risk of developing TBR [18] . Even with TB incidence varying from one country to another, results show that poor previous treatment outcomes and lack of adherence remain the main triggers of TB drug resistance [9] [12] [18] [19] .

In our study, no association was found between having a former close contact with a TB patient and the resistance group. This contrasts with findings of studies held in Ethiopia and Namibia, where close contact with a TB case was strongly associated with TB drug resistance [18] [19] . This fact is highlighted in Chinese studies that have unveiled a community transmission of DR strains and exogenous reinfection among former TB patients [20] [21] . Hence, the importance of TB contact tracking. The low number of close-contact cases in our study could be due to patients being unaware of TB cases in their surroundings or difficulty recalling.

Univariate analysis has found an association between HIV-positive status, being male gender, having dyspnea, lung scars at X-Ray (cavity, nodule, and infiltrate), delay in DR treatment initiation, and treatment outcome with secondary resistance group. However, after multivariate analysis, factors remaining associated were dyspnea, having nodules or infiltrates at X-Ray, and delay in DR treatment initiation.

Wang et al. found that pulmonary lesions among MDR-TB patients tend to be extended and bilateral [22] . The finding in the chest Tomography scan was a tree-in-bud sign, patchy or lobular areas of consolidation, cavitation, and bronchiectasis [22] [23] . In our study, the secondary resistant group patients were more likely to present nodules and infiltrates on chest X-rays. Oladimeji and al in Nigeria found that MRD TB patients with previous treatment are 6.9 times more likely to have abnormalities in chest X-ray. Lesions found in their studies were mostly cavitation, fibrosis, and infiltrates [24] . Even if the type of lesions were different, multiple radiological patterns were observed among resistant TB patients previously treated [22] - [24] . Each pattern may have different therapeutic implications. Nodules and infiltrates express more disease current activity, while cavities are associated with slow sputum conversion, increased risk of transmission, and developing additional drug resistance [22] . The more extensive the cavities, the worse the outcome of patient during second-line treatment. [1] [20] [22] . Therefore, initial radiological evaluation may help to screen patients who need additional treatment supplies and follow-up investigation. Our data also shows that dyspnea was associated with the secondary resistant group, which can be in line with the extension of lung damage on chest X-ray.

Our study also revealed the importance of drug susceptibility testing to avoid misdiagnosis of resistance patterns that can lead to ineffective treatments and worse outcomes [1] [8] .

The drug susceptibility test was performed on 59 positive cultures of our patients. Resistance to any first-line drugs and resistance patterns were most frequent in the secondary resistance group. Anyway, finding any resistance patterns in the primary resistance group may be due to the circulation of resistant strains within the community. This may reflect the efficacy of the TB programs on DOTS implementation [1] [2] . Resistance to both rifampicine (R) and isoniazid (H) were found among 60% of those having DST and classified as the secondary resistance group. In addition, one patient had rifampicin resistance associated with other drugs different from isoniazid. Our results are close to those from an Ethiopian study, where people with a history of TB treatment had 8 times the risk of developing resistance to any first-line drugs and were at least 7 times more likely to have MDR-TB [25] [26] . R and H are the most powerful anti-TB drugs, mutations are the main noticeable way of resistance occurrence. Some mutation regions associated with resistance to R and H are responsible for greater transmission and cross-resistance to ethionamide, respectively [27] .

Patients with MDR TB, compared to those with monoresistance to INH or RIF, are more likely to develop resistance to another drug during the treatment. This new resistance often occurs within the first 6 months of treatment and could be a source of ultra-resistance [28] [29] . Follow-up DSTs can exhibit MDR TB, pre-XDR, and XDR during treatment. In this study, participants were admitted to the hospital during the first 4 months for DOT and amikacin injections. The long waiting time to obtain the DST result could delay the use of the appropriate drugs in case of resistance. These situations could increase the risk of new resistance development and transmission of resistant strains in the community.

It appears necessary to perform DST at treatment initiation and follow-up to track additional drug resistance, as some of our study participants were carrying pre-XDR and XDR strains.

Overall, the treatment success rate was 82,03%, which is higher than the MDR/RR-TB global treatment success rate in 2023 and lower than 86% of Mbuh et al. in Cameroon [2] [17] . Both Togo and Cameroon have implemented TBMR management, which encompasses hospitalization and directly observed treatment during the intensive phase of treatment, along with comorbidities screening [7] [17] . The higher success rate in Cameroon can be explained by the inclusion of a narrow population from one region rather than the whole country. In South Africa (SA), Ndjeka et al. had a success rate of 60% which could be explained by a higher number of HIV-infected patients in their study [30] . Globally, our results are better than those in West and Central Africa, with a success rate ranging from 69 to 81% [13] . Reasons for poor outcomes of R-TB treatment are linked to HIV infection disease stage, occurrence of side effects, lack of adherence, unemployment, and high duration between Drug-resistant confirmation and treatment initiation [1] [13] [30] [31] .

In our study, we had a death rate of 16.8%, which is slightly below the global trend [2] . Surprisingly, a poor outcome was associated with primary resistance in univariate analysis. Even if the association dropped after multivariate analysis, our results are contrary to what was observed in other studies [31] [32] . Our results are close to those of Ockenga et al., who found that the primary resistance group tends to be under 40 years old, have bilateral and cavitary diseases on lung imaging, and are associated with poor outcomes [33] . Thus, the poor outcome of the primary resistance group observed in our study could be explained by the younger age, HIV positivity rate, and the influence of female gender. Our results show that being HIV positive seems to be protective for the secondary group participant at univariate analysis. Despite the loss of that protection, we assume that being already in the TB management system for a longer time was an opportunity for better care. Knowing that the Togo TB program is associated with HIV one for coinfection management in health care facilities [7] [9] . The secondary resistance group had had much time to benefit from both care compared to their counterpart from the primary resistance group. The latter included more females, and some have also shown that socio-cultural barriers, stigma, lack of financial autonomy, or inequity often impede women’s access to care compared to men [32] . Studies found that bilateral lung damage and cavities are associated with poor outcomes and death [22] [31] [33] . In our study, cavities were not associated with the secondary group that had better outcomes. The radiological abnormalities and clinical signs associated with a negative impact could have been counterbalanced by the low proportion of HIV, the higher BIM, male gender, and age of the secondary group participants.