Background
TB is a major global health problem with 8.6 million new cases and almost 1.3 million deaths attributed to the disease every year [
1]. WHO recommended standardized first-line anti-TB drug regimen is a single 6-month or 8-month regimen composed of isoniazid (H), rifampicin (R), ethambutol (E), and streptomycin (S) with pyrazinamide (Z) [
2]. This regimen is recommended for all new TB cases and previously treated cases in more than 90 countries. This strategy is designed and evaluated as the cost-effectiveness regimen used in resource-limited settings for decades for the consideration of convenient treatment management based on same number, dose, and types of medication. Though high cure rate was achieved of SCC for cases with drug-sensitive TB [
3], the emergence of drug resistant TB, especially MDR-TB, and acquisition of additional drug resistance during treatment, brought much less efficacy of SCC both in trials [
4‐
7] and under program conditions [
8]. Few cohort studies, though the sample size was small, worked on the treatment of mono- or poly-resistant TB with SCC and presented poor results [
9]. In most low- and middle-income countries, DST is not routinely performed for new cases nor for most previously treated cases [
10,
11]; therefore, cases carrying drug-resistant strains of
Mycobacterium tuberculosis (MTB) might be at greater risk for SCC failure and disseminating drug resistant strains [
12]. China is one of the 27 countries with high MDR-TB burden, early detection and treatment could help prevent its transmission. Currently, China’s NTP [
13] provides new and previously treated cases with SCC under directly observed treatment (DOT) and monitors their response to SCC by sputum smear (SS) rather than sputum culture and DST due to limited resources. Sputum culture and DST are recommended to be performed for cases with initial treatment and retreatment failure, where possible, to develop appropriate chemotherapy regimens. Hence, little is known whether drug resistant TB occurs during treatment and few data available links acquired drug resistant TB to treatment failure with SCC. To address these issues, we conducted a prospective observational study tracing the emergence of acquired drug resistant TB at indicated time points among new and previously treated cases receiving first-line standard regimens to determine the characteristics and its effect on treatment failure.
Discussion
Up to date, this study is the first to evaluate traits of acquired drug resistant TB emergence and its impact on treatment failure of SCC under China’s routine NTP condition. Majority of acquired drug resistant TB cases developed early within 2 months with SCC and late acquired drug resistant TB emergence after 2 months is more likely associated with the drug resistance pattern of MDR-TB. Furthermore, our results indicated a good response to SCC in cases with pan-susceptible strains throughout therapy which reconfirmed the SCC effectiveness for cases with susceptible strains [
8,
16]. However, our findings presented poor response of SCC in cases with acquired drug resistant TB, especially with late drug resistant TB. It is shown that late acquired drug resistant TB emergence after 2 months is 25.7 times higher to result in treatment failure than that within 2 months.
In our study, the data from 8 provinces in China indicating a overall drug resistant TB prevalence of 17.6 % in new cases and 29.1 % in previously treated cases and a MDR-TB prevalence of 2.6 % and 14.0 % in new and previously treated cases respectively are lower than that from China’s national drug resistance survey in 2007 [
17]. The reduction of drug resistant TB and MDR-TB prevalence in our study could be explained by two possible reasons. Firstly, given the prevalence of TB and MDR-TB not balanced which was highest in western China, 2 provinces from 4 main areas in the study may not be representative of the overall situation in China. Secondly, the drug resistant TB cases in the study were from TB Control and Prevention Centers system not TB specialized hospitals system where more previously treated cases tend to access for further second line anti-TB drugs due to MDR-TB pretreatment [
18].
It is well known that previous TB treatment is a strong determinant of drug resistance [
17,
19‐
22]. Current data suggests that previously treated cases are more likely to develop acquired drug resistance, especially MDR-TB, during therapy than that of new cases. However, it is impossible for us to determine the identified reason of acquired drug resistant TB during SCC without DNA fingerprinting analysis of resistant strains compared to that of original strains pretreatment. This represents a limitation of the study. Two reasons might contribute to the acquired drug resistant TB emergence under qualified DOT and quality-assured drug apply under China’s NTP. First, the selection of resistant mutants in mixed bacterial population infected pretreatment due to killing of susceptible strains by anti-TB drugs of SCC. Second, infection of new drug resistant strains may be another explanation for the acquired drug resistant TB emergence during SCC [
23].
Several researchers have reported that amplification of resistance to additional anti-TB drugs while receiving WHO recommended SCC [
24]. Few data was reported on the traits of acquired drug resistant TB emergence receiving SCC with susceptible strains pretreatment. However, it is important to identify drug-resistant cases in time with standard treatment and prevent its dissemination. We found the time point of acquired drug resistant TB emergence was associated with drug resistance patterns. Cases with MDR-TB development were 8.3 times more likely to be late emergence compared to the non-MDR-TB pattern. Moreover, the cases with late emergence of acquired drug resistance are of high risk to contribute to SCC failure. In line with other studies, anti-TB drug resistance especially MDR-TB has a negative impact on treatment outcome of SCC [
8,
12,
16,
25‐
29]. Treatment success rate of MDR-TB cases was 58 % in Peru and 60 % in Hongkong with SCC [
8]. One study in rural counties of eastern China indicated that the cure rate of MDR-TB and other drug resistant TB were 58.3 % and 91.0 % of SCC [
30]. Different from these studies, we excluded the drug resistant TB cases pretreatment to target on the characteristics of drug resistant TB emergence during SCC and further explore its impact on treatment outcome. Treatment success rate of cases with acquired MDR-TB was 52.9 % while 82.2 % with non-MDR-TB, a little lower than that in Peru, Hongkong and eastern China. We also analyzed treatment success rate of 3 subgroups of non-MDR-TB indicating that cases with any H resistance and E/S resistance have higher treatment success rate around 90 % while much lower treatment success rate of 61.5 % with any R resistance with SCC. Some discordant impacts of drug susceptibility patterns on treatment success are reported in the literature [
25,
31,
32]. This could be explained by the different target population in our study. The greatest finding of this study is that the time point of acquired drug resistant TB emergence significantly impacted treatment outcomes with SCC. Cases with acquired drug resistant TB at 3–5 months were 25.7 times higher (OR, 25.7; 95 %CI, 4.3–153.4;
P < 0.001) than that of cases with acquired drug resistant TB within 2 months of SCC to experience treatment failure.
Besides drug resistant TB emergence time point, many other factors including low BMI, smoking, and some behaviors such as alcohol consumption and drug abuse are also associated with poor treatment outcomes [
33‐
36]. Diabetes and baseline disease severity of TB also have been shown to be independent risk factors for poor treatment outcomes in previous studies [
18,
26,
37‐
39].
As reported, the use of standard first-line anti- TB treatment on cases with drug resistant TB pretreatment have greater likelihood to get relapse, treatment failure and acquired drug resistance [
12,
40‐
42]. Therefore, pretreatment DST were carried out for individual patients and those with any drug resistance were excluded from our study and transferred to a DOTS-Plus program with the consideration of providing a more tailored regimen for optimal treatment outcomes. Without doubt, DST performed before and during SCC could provide information to recognize drug resistant TB, particularly MDR-TB. Our studies suggested DST should be taken before treatment starting and subsequent DST should be checked regarding patients who remain bacteriological positive at the month 2 or 3 and it is better to provide accordingly effective regimen rather than keeping using SCC in the setting with drug resistant TB emergence later after 2 months as these cases are more likely to fail with SCC. The pressing need to prevent MDR-TB warrants this recommendation and this approach will help decrease its transmission.
Our study has several limitations. First, we did not provide longer follow-up information to evaluate recurrence rates and correlated risk factors for long-term prognosis among successfully treated acquired drug-resistant TB cases. Second, this study was limited by the relatively small number of acquired drug resistant TB cases during SCC. A larger sample could better present the association of treatment failure and its impact factors. Hence, larger scale cohort studies are still needed to further verify the findings of our study. Third, we were not able to measure certain factors possibly related to treatment failure, for example, comorbidities including diabetes, chronic obstructive pulmonary disease, chronic hepatitis, and bacterial load. Fourth, our study is lack of DNA fingerprinting examination to differentiate the origin of acquired drug resistant TB.
Despite these limitations, to date, this might be the first study to evaluate the characteristics of acquired drug resistant TB and its effect on treatment failure with SCC. This study demonstrated later emergence of acquired drug resistant TB during SCC is the prognostic risk factor for treatment failure. Our findings may help relevant policy makers to take more consideration of treatment management on TB cases with potential failed outcomes. Early detection of treatment failure will decrease transmission and decrease likelihood of additional drug resistance acquisition, providing more probability to choose effective regimen. Administration of effective regimen may optimize cure rates and drug resistance acquisition may be avoided. Rapid, feasible and economical culture and molecular biology methods such as GeneXpert are imperative to be applied for identifying drug resistant TB in time pretreatment or during SCC. Effective and comprehensive TB control strategies with adapted DOT is needed to prevent drug resistance especially MDR-TB development. More strict infection control and health education measures should be taken to minimize the transmission of TB and drug resistant TB bacilli in public and ensure patients adherence to treatment preventing drug resistance development. In addition, we expect more robust predictors developed which could evaluate factors that could affect underlying pathological process of the disease being treated and measure the effects of interventions on clinical outcomes in multiple aspects [
43].
Competing interests
The authors declared that they have no competing interests.
Authors’ contributions
LL and GFZ conceived and designed the study. JTG, YM, JD, GFZ, SYT, YYF, LPM, LYZ, FYL, DYH, YLZ and LL performed the study. JTG, LL and QL analyzed the data. JTG prepared the final version of the manuscript. All authors read and approved the final manuscript.