Epidemiology of first-and second-line anti-tuberculosis drug resistance in new pulmonary tuberculosis cases in Addis Ababa metropolitan area, Ethiopia

Introduction

Tuberculosis (TB), caused by mycobacterium tuberculosis (M. tuberculosis) is the oldest disease known to affect humans with myriad clinical manifestations of diseases and fatalities in the world [1-4]. Mycobacterium. tuberculosis belongs to closely related mycobacterial species including M. bovis, M. africanum, M. microti, M. caprae, M. pinnipedii, M. canetti, and M. mungi, together comprising M. tuberculosis complex [5 - 7]. The origin and global spread of human TB had a contemporary notion that the genus Mycobacterium originated in animals and was transferred to humans during the neolithic transition 150 million years ago [1, 8, 9]. Recent evolutionary data suggested that M. tuberculosis emerged as a human pathogen in Africa and survived with modern human migrations approximately 70 000 years ago [4,10,11]. However, it was in 1882, when Robert Koch was able to isolate tubercle bacillus that determine a milestone in the fight against TB [12-14].

According to the WHO global tuberculosis report of 2022, an estimated 10.6 million persons became ill from TB worldwide in 2021, of which 5.3 million people were diagnosed with pulmonary TB. Similarly, an estimated 1.6 million deaths and an estimated 4 million people with TB missed undiagnosed in 2021[15]. This reversal of progress from the previous estimations is indicative of challenges in closing the gaps to end TB. The emergence of drug resistance M. tuberculosis strains against commonly used and less toxic standard anti-TB drugs continue to be a clear threat to the realization of TB elimination [16,17]. Globally, 3.6% of new TB cases and 18% of previously treated cases had MDR or RR-TB. Conversely, about 2 billion people are estimated to generate a latent TB infection, one of the most remarkable features of M. tuberculosis, and are thus at risk of developing active TB disease in their lifetime [4,15,18]. Ethiopia still remained among the 30 TB and TB/HIV high-burden countries, while transitioning out of the list of MDR/RR-TB for 2021 – 2025. According to the WHO global tuberculosis report in 2022, Ethiopia constitutes an annual TB incidence rate of 143 per 100,000 population. Similarly, an estimated proportion of 1.1% and 12% MDR/RR-TB in new and in previously treated TB cases were reported respectively [15].

Globally, a remarkable campaign to control tuberculosis was started by WHO in 1948, and in 1974 the WHO Expert Committee on Tuberculosis issued a policy guideline to control TB by discouraging both radiographic screening and tuberculin skin testing while promoting sputum microscopy and BCG vaccination [14]. Following years of increased incidences, the WHO declared TB a global emergency in 1993 and in 1995 devised a directly observed treatment (DOTs) as a control strategy with the aspiration to save lives and dream of TB elimination [13,19]. Efforts to control TB in Ethiopia began with the sequential entrenchment of TB centers in the early 1960s, the National Tuberculosis Control Program central office in 1976, and the National Tuberculosis and Leprosy Control Program in 1994. The country adopted the DOTs strategy in 1997 and endorsed the ‘Stop TB’ and ‘End TB’ aspirational strategies of the UN MDGs and SDGs [20, 21]. However, despite concerted efforts during the past decades, the rate of decline in tuberculosis incidence, and mortality is lower than what would be required to reach the local as well as the global target of pre-elimination by 2035 and elimination by 2050.

The microbiological detection of TB is critical as it allows people to be correctly diagnosed, is necessary for drug resistance detection, and ensures that the most effective treatment regimen can be selected as early as possible. Tuberculosis is mostly diagnosed using smear microscopy, rapid molecular methods, culture, and phenotypic and genotypic DST methods. However, the gold standard method for the bacteriological confirmation of TB and detecting resistance to anti-TB drugs is culture using commercially available liquid media [22, 23].

The standard treatment regimens for susceptible new TB patients are 2RHZE:4RH divided into intensive and continuation phases respectively. While, the updated standard shorter MDR-TB and longer MDR-TB regimen to treat MDR/RR-TB are recently adopted by the MoH as follows: 4–6 BDQ -LFX-CFZ-Z-E-HH-ETO:5 LFX-CFZ-Z-E; and 18 BDQ: LFX/MFX-LZD-CFZ-CS respectively [15, 18, 21,24].

Nonetheless, efficient monitoring response to ant-TB drugs is quite critical to ensure that patients are responding to treatment at an acceptable success rate, to guide policymakers, to effectively limit transmission of resistant strains in the community, to improve the quality of life of every TB patient and to recognize M. tuberculosis drug resistance pattern in a specific geographic area. However, monitoring responses to ant-TB drugs using molecular assay is not suitable, and culture and phenotypic DST is not routinely utilized in Ethiopia. Furthermore, while it is quite important to assign patients to appropriate treatment regimens, evidence of the agreement between phenotypic and WHO-recommended rapid molecular drug resistance detection is elusive. The aim of this study was to describe the epidemiology of first-and second-line anti-TB drug resistance situations in new presumptive pulmonary tuberculosis cases in Addis Ababa metropolitan area, Ethiopia.

Materials and Methods

Results

sociodemographic and clinical characteristics

M. tuberculosis was isolated from 216 new presumptive pulmonary TB cases with MTB-positive GeneXpert MTB/RIF Assay results diagnosed between October 2019 and July 2020. DST was performed till June 2021. We excluded 60 patients from the analysis based on the fidelity of liquid culture, mainly due to viability and contamination. Overall, 15.7% (34/216) and 12.0% (26/216) of specimens were found to be not viable and contaminated respectively (Fig 1).

• Download figure
• Open in new tab

Figure 1:

Flowchart showing clinical and experiement workflow.

A total of 156 Mycobacterium tuberculosis complex isolates were successfully recovered using a gold standard BD BACTEC™ MGIT™ 960 automated liquid culture system and were subjected to drug susceptibility testing. Among the study participants, males account for 53.8 % (84/156). The median age of the study participants was 30.0 (range 6 – 75) years with a mean and standard deviation of 33.56 and 12.65 years, respectively. Of all the study participants, 49.4 % (77/156) were in the age group below 30.0 years, 58.3 % (91/156) were married, and 76.9% (120/156) were urban residents (Table 1).

View this table:

• View inline
• View popup

Table 1. Characteristics of the study participants in Addis Ababa metropolitan area, Ethiopia, (n = 156)

Discussion

In this study, we identified a rate of 14.1% (22/156) resistance of M. tuberculosis strains to at least one first-line anti-TB drug, while 85.9% (134/156) were found to be pan-susceptible M. tuberculosis strains in new presumptive pulmonary TB cases. Similar to this finding, a comparable rate of resistance of M. tuberculosis strains to at least one first-line anti-TB drug among new cases was reported by Tessema B et al., (15.8%, 2012) [28], Adane K et al., (15.58%, 2015) [29], Lobie TA et al., (16.1%, 2020) [30], Worku G et al., (11.6%, 2022) [31] in Ethiopia; Doss M et al., (13.4%, 1999) [32] in Côte d’Ivoire; Antunes ML et al., (17.8%, 2000) [33] in Portugal; Glasauer S et al., (12.7%, 2019) [34] in Germany and Ullah I et al., (11.5%, 2016) [35] in Pakistan. Compared to our finding, a higher rate of resistance of M. tuberculosis strains to at least one first-line anti-TB drug among new cases was reported by Yimer SA et al., (30.1%, 2012) [36], Bedewi Z et al., (22.2%, 2017) [37], Seyoum B et al., (23.0%, 2015) [38], Dagne B et al., (26.8%, 2021) [39], Yigzaw WB et al., (34%, 2021) [40] in Ethiopia; Mende N et al., (23.5%, 2023) [41] in Zambia; Thapa G et al., (31.1%, 2016) [42] in Nepal; Li Y et al., (34.1%, 2016) [43] in China; Ndung’u PW et al., (30.0%, 2012) [44], Kerubo G et al., (30.0%, 2016) [45] and Yonge SA et al., (48.6%, 2017) [46] in Kenya; and Aung WW et al., (27.7%, 2015) [47] in Myanmar. However, a lower rate of resistance of M. tuberculosis strains in new presumptive pulmonary TB cases was reported in the previous studies by Biadglegne F et al., (6.7%, 2014) [48] in Ethiopia.

The most likely justification for these differences in the rate of drug resistance reported by the previous studies in the same as well as different geographic regions might be due to the difference in the study period, study design, study subject, sample size, health system efficiency to TB control activities, patient’s immunological and economic status and virulence factors of the Mycobacterial strain circulating in the community.

The rate of MDR-TB identified in this study was 5.8%, (9/156) among new presumptive TB cases and this was congruent to the rate of MDR-TB reported in the previous studies by Tessema B et al., (3.7%, 2012) [28], Yigzaw WB et al., (5.0%, 2021) [40] in Ethiopia; Yonge SA et al., (4.8%, 2017) [46] in Kenya and Dosso M et al., (5.3%, 1999) [32] in Côte d’Ivoire. However, compared to our study findings, a higher rate of MDR-TB was reported in the previous studies by Welekidan LN et al., (11.6%, 2020) [49], and Dagne B et al., (61.9%, 2021) [39] in Ethiopia; Ullah I et al., (9.3%, 2016) [35] in Pakistan; Monde N et al., (9.8%, 2023) [41] in Zambia, Thapa G et al., (15.6%, 2016) [42] in Nepal, Li Y et al., (13.2%, 2016) [43], and Zhou L et al., (18.8%, 2022) [50] in China and Aung WW et al., (18.3%,2015) [] in Myanmar. On the other hand, a lower rate of MDR-TB was reported by Yimer SA et al., (1.0%, 2012) [36], Seyoum B et al., (1.1%, 2015) [38], Bedewi Z et al., (1.5%, 2017) [37], Adane K et al., (1.29%, 2015) [29], Mekonnen F et al., (2.3%, 2015) [51], and Lobie TA et al., (2.3%, 2020) [30] in Ethiopia; Glasauer S et al., (3.3%, 2019) [34] in Germany. Notwithstanding this observation, the WHO estimate of 1.1% of MDR-TB in new TB cases revealed in the 2022 global tuberculosis report is much lower than the findings demonstrated in this study [15].

The most plausible justification for the higher rate of MDR-TB among new presumptive pulmonary TB cases in our study than in the previous studies may contemplate the emergence and transmission of multi-drug resistant TB in Ethiopia remain a challenge and shows that the battle against TB is far from over within the anticipated local as well as global initiatives. Furthermore, in addition to the biological factors leading to the high prevalence of MDR-TB, such as the fitness of DR-MTB strains, mutation ability, and virulence variation, the rate of poor treatment adherence and inefficient health service system have a significant influence on the development of MDR-TB. Nevertheless, the DOTS strategy has been proven to be one of the most effective strategies to manage TB and limit the emergence and transmission of MDR-TB in the community. However, despite the tremendous efforts and encouraging progresses in widening DOTs coverage in Ethiopia in public as well as private TB treatment initiation centers, evidence showed active TB case surveillance, as well as rapid detection of drug resistance, remained low to end TB by 2035 and eliminate TB by 2050 partly because culture and universal DST are not widely accessible and routinely done in Ethiopia.

In this study, 3.8% (6/156) of isolates were identified as resistant to all first-line anti-TB drugs tested, including STM, INH, RIF, and EMB. This proportion of resistance was higher than the reported proportion in the previous study done by Yigzaw WB et al., (2.38%, 2021) [40] in Ethiopia. In addition, 7.1% (11/156) of isolates were mono-resistant M. tuberculosis strains, and the highest mono-drug resistance isolates to first-line anti-TB drugs were from INH (5.1%, 8/156) and STM (1.9%, 3/156). A congruent monoresistance rate was reported in the previous study by Zürcher K et al., (7.0%, 2019) [52] in seven high TB-burden countries. This rate of INH and STM monoresistance respectively was lower than the reported rate in the previous studies by Seyoum B et al., (9.5%, 7.0%, 2015) [38] and Bedewi Z et al., (6.3% and 7.5%, 2017) [37] in Ethiopia. However, compared to our study findings, variably a higher rate of INH and a lower rate of STM monoresistance were reported in the previous study by Monde N et al., (9.8% and 0.8%, 2023) [41], respectively, in Zambia.

This study further demonstrated that the rate of resistance to any of the first-line anti-TB drugs to be 9.0% (14/156), 11.5% (18/156), 5.8% (9/156), and 4.5% (7/156) for STM, INH, RIF, and ETM respectively. A comparable rate of any resistance to first-line anti-TB drugs was reported in the previous studies by Seyoum B et al in 2015 [38], and Adane K et al., in 2015 [29] in Ethiopia, and an incomparably higher rate of drug resistance (any) related to INH and RIF was reported in the previous study by Thapa d et al., (23.0% and 17.8%, 2016) [42], respectively, in Nepal. Interestingly, monoresistance to RIF and EMB was not demonstrated in all isolates in this study. The absence of monoresistance against RIF and EMB was reported in the previous studies by Agonafir M et al., (2010) [53], Tessema B et al., (2012) [28], and Worku G et al., (2022) [31] in Ethiopia. A higher rate of sensitivity to RIF alone may be a good indicator of the success of the DOTs program in the study area in Ethiopia. However, the higher rate of INH monoresistance observed in new presumptive TB cases in this study is critical as INH is the main monotherapy drug used to treat LTBIs and the main chemoprophylaxis drug used in immunosuppressed TB patients in Ethiopia.

The findings of this study demonstrated that all MDR-TB isolates were susceptible to all recently recommended second-line anti-TB drugs tested and none of these isolates were found to be Pre-XDR or XDR-TB. Compared to this finding, similar results were reported in previous studies conducted by Tessema B et al., (2012) [28], and Worku G et al., (2022) [31] in Ethiopia. However, a significant proportion (81.8%) of isolates exhibited resistance to Ethionamide in this study. Although higher rate resistance against second-line anti-TB drugs was reported in the previous studies by Agonafir M et al., (73.9%, 2010) [53], and Dagne B et al., (5.6%, 2021) [40] in Ethiopia, and Monde N et al., (35.5%, 2023) [41] in Zambia, the absence of resistance to appropriate second-line anti-TB drugs in this study may reflect that the existing treatment of MDR-TB cases in the study area was encouragingly effective. This observation is critical because of the positive implication that initiation of second-line drugs without initial DST might not be beneficial to the patient and requires reemphasizing the need to perform drug susceptibility testing prior to granting anti-TB treatment. Nevertheless, previous studies reported that diagnosed, untreated, and inappropriately managed MDR-TB contributes to a continuously sustained high prevalence of drug-resistant M. tuberculosis strains circulating in the community [54].

However, compared to this study finding, a high rate of resistance against Ethionamide was reported in the previous studies by Ongaya VA et al., (56.3%, 2012) [55] in Kenya, Rueda J et al., (95.0%, 2017) [56] in Cambodia and Islam MM et al., (90.8%, 2019) [57] in China. Whereas, a lower rate of resistance to Ethionamide was reported by Tan Y et al., (39.1%, 2017) [58] and Wu X et al., (2.05%, 2019) [59] in China. The higher rate of resistance to Ethionamide observed in this study may be explained by the fact that Ethionamide is a derivative of isonicotinic acid having a similar structural analogue with isoniazid share common pathways that lead to cross-resistance and resistance to INH seemingly will mediate co-resistance to both isoniazid and ethionamide [60,61]. Notwithstanding this fact, all Ethionamide-resistant isolates were identified as either monoresistant or any-resistant to INH in our study. Meanwhile, there appear recent guidelines recommending against its inclusion within the standard second-line anti-TB treatment regimen, Ethionamide is a second-line pro-drug recommended to be conditionally prescribed for the treatment of MDR-TB. Nevertheless, a higher rate of resistance against Ethionamide observed in this study would mean that the use of ethionamide in the second-line treatment regimens may not have therapeutic benefit in this study area.

Although the primary objective of this study was not to validate rapid molecular testing against the reference standard method, we compared the rate of RR-TB detected using GeneXpert MTB/RIF Assay with BD BACTEC™ MGIT™ 960 SIRE liquid culture system for automated drug susceptibility testing. The sensitivity of 100% of GeneXpert MTB/RIF Assay in detecting rifampicin resistance in our study was in line with previous studies reported by Meawed TE et al., (100%; 2016) [62] in Egypt, Jahan H et al., (98.5%; 2016) [63] in Bangladesh, Pandey P et al., (98.6%; 2017) [64] in Nepal and Mekkaoui L et al., (100%; 2021) [65] in Belgium; but higher than the sensitivity reported in previous studies by Tang T et al., (84.0%; 2017) [66] in China and Feliciano CS et al., (94%; 2019) [67] in Brazil. Whereas, the specificity of 94.6% of GeneXpert MTB/RIF Assay in detecting rifampicin resistance in our study was lower than previous study reported by Meawed TE et al., (100%; 2016) [62] in Egypt, Jahan H et al., (100%; 2016) [63] in Bangladesh, Pandey P et al., (100%; 2017) [64] in Nepal, Mekkaoui L et al., (99.2%; 2021) [65] in Belgium and Feliciano CS et al., (98%; 2019) [67] in Brazil; but higher than the specificity reported by Tang T et al., (87.8%; 2017) [66] in China.

In this study, 52.9% (9/17) of cases showed concordance between GeneXpert MTB/RIF Assay and BD BACTEC™ MGIT™ 960 SIRE liquid culture system for automated phenotypic DST in detecting rifampicin resistance. However, a significant proportion, 47.1% (8/17) of discordance rate was observed. This result was lower than the concordance rate reported by Feliciano CS et al., (90.5%; 2019) [67] in Brazil, but comparable with what was reported in the previous studies by Williamson DA et al., (31%; 2012) [68] in New Zealand and Ngabonziza JCS et al., (47.0%; 2020) [69] in Rwanda. In other words, the rate of detection of rifampicin resistance among GeneXpert MTB/RIF Assay and BD BACTEC™ MGIT™ 960 SIRE liquid culture and DST System was determined at 10.9% (17/156) and 5.8% (9/156) respectively.

However, this study’s findings showed comparable sensitivity, but inconsistent specificity with previous similar studies might be due to the nature and limitations of GeneXpert MTB/RIF Assay as a positive test does not necessarily indicate the presence of viable tubercle bacilli, the presence of mutation conferring rifampicin resistance within as well as outside of the RRDR of the rpoB gene, the presence of silent mutations in the rpoB gene, and low bacterial DNA in the clinical specimens contribute to the increased false rifampicin resistance observed in GeneXpert MTB/RIF Assay in this study. Needless to say, the classical GeneXpert MTB/RIF Assay System is playing a critical role in the screening and detection of M. tuberculosis in clinical specimens. However, its efficiency in identifying resistance against rifampicin introduces an unannounced high rate of false positive results leading a significant number of patients to be placed on a second-line anti-TB treatment regimen.

This study has some limitations that need to be noted. Firstly, the study was conducted in new presumptive pulmonary TB cases and this finding could not be generalized to other types of TB disease classification. Secondly, we were not able to include PZA susceptibility testing as we encountered shortages of validated and commercially available reagents. Thirdly, we were unable to follow patients in order to look into treatment outcomes due to logistic reasons.

Conclusion

The rate of MDR-TB among new pulmonary TB cases remained high at fivefold the national and nearly twofold the global estimated rate. The rate of monoresistance against anti-TB drugs was also high. The absence of resistance against recommended second-line anti-TB drugs was quite encouraging. However, the high rate of resistance against Ethionamide would mean that its inclusion in the regimen may not have therapeutic benefit in this geographic area. Furthermore, the use of GeneXpert MTB/RIF Assay alone for the detection of M. tuberculosis might introduce 15.7% false positivity and 47.1% false rifampicin resistance leading patients erroneously assigned in the MDR-TB category and placing on an unnecessary second-line anti-TB-treatment regimen. Enhanced validation and progressive harmonization of rapid molecular diagnostics against reference methods are recommended to improve patient outcomes.

Supporting information

S1File. Sociodemographic, culture, and DST raw data xls.