In Chongqing, 23 of all 200 MDR/RR-TB patients died with a high death rate. The risk factors analysis for death showed that the elder age and cavitary disease are the major risk factors. Same with another research which studied the treatment outcomes of MDR-TB patients in Zhejiang, China, patients who were age over 60 years, relapse, receiving treatment after failure, cavitary disease or adverse events are more likely to get poor treatment outcomes [20]. In our research, we classified all patients into four groups based on the treatment results: favored outcome group that the patients were cured or completed treatment; pending group that the patients cannot obtain treatment outcome, such as: in treatment, adverse reaction, treatment refusal or lose to follow; died group that the patients were death from TB and died with other reasons group. In all the 49 cases in treatment, only 3 cases were new cases, the others were treatment failure or relapse cases that have to take a prolonged treatment course. Therefore, the patients aged 35–54, urban residents or previously treated with second-line drugs should be enhanced management and supervised to complete standard treatment regimens.
The control of TB depends on the identification and treatment of infectious patients and their contacts [21]. In China, a high burden of TB setting, epidemiological data are often challenging to obtain, WGS offers the prospect of identifying plausible transmission events between patients without prior recourse to epidemiological data [4, 13, 22, 23]. Among WSG analysis pipelines for detection of epidemiologically linked tuberculosis cases, the cg-MLST approach is standardisation and easy to available comparison to the SNP-based pipelines [24]. In Chongqing, 80 (42.8%) of all 178 MDR/RR-TB patients were caused by RT using a WGS-based cg-MLST typing approach. Patients who affected with lineage 2 mycobacterial strains had high risk of RT. A research about RT of MDR strains showed that 103 (32%) of the 324 MDR strains were clustered, with another 132 were treatment-naïve patients, totally 235 (73%) patients with MDR tuberculosis probably had transmission of MDR strains in Shanghai, China [12]. In our research, 21 were treatment-naïve patients (new cases) in all 98 gnomically unique isolates. According to this rule, 101 (54.0%) patients with MDR/RR-TB probably had transmission of MDR/RR strains, in Chongqing, China.
Patients who had delayed diagnosis or were older than 45 years had high risk of RT of MDR-TB in Shanghai, China [12]. Another research in Shanghai showed that local incidence of TB in urban centres is local transmission between both migrants and residents [25]. In Indian, transmission of particular pre-XDR/XDR lineage 2 strains is the main driver of the pre-XDR/XDR-TB epidemic [26]. In our research, the diagnosis delay defined by the interval between the diagnosis of TB and the diagnosis of MDR/RR-TB had little effect on RT of MDR/RR-TB. The local transmission between migrants and residents also affected little on RT of MDR/RR-TB. The patients who lives in urban and are occupied in rolling stone are more likely to happen RT in MDR/RR-TB cases. Therefore, the transmission contributes most of the MDR/RR-TB epidemic, and especially in dominant lineage 2 strains is main driver of the MDR/RR-TB epidemic in Chongqing, China.
Except identification of transmission clusters, WGS can predict drug resistance based on known mechanisms of resistance and has been proved a useful tool for predicting drug resistance of Mtb, especially for first-line tuberculosis drugs [27–29]. A number of web-tools for detection of genotype SNPs and indels in Mtb have been developed to predict resistance. These include PhyResSE, KvarQ, MTBseq, and TB-Profiler [30]. In our research, TB-profiler was used to predict the genotypic resistance. In all 180 RIF genotypic resistant strains, 171 (95.0%) were harbored in rpoB435 (21.7%, 39/180), rpoB445 (16.1%, 29/180) and rpoB450 (57.2%, 103/180) mutations. The most mutation site was rpoB_p.Ser450Leu that was same with most research [31]. In all the 176 INH genotypic resistant strains, 72.7% (128/176) was katG, 2.3% (4/176) was ahpC, 2.8% (5/176) was fabG1 (also called inhA Promotor), 7.4% (13/176) was ahpC plus katG, 14.8% (26/176) was fabG1 plus katG. Totally, 152 (86.4%) strains were observed mutations in katG315, fabG1_c.-15C > T and fabG1_c.-8T > C. Furthermore, some paper showed that WGS was more accurately predicts susceptibility of Mtb to first-line drugs than phenotypic testing [27]. For the new and repurposed tuberculosis drugs, no resistant mutations were observed in phenotypic BDQ, DLM, LZD and CFZ resistant strains. The mutations associated with resistance to the new and repurposed Mtb drug should be updated in their resistance gene catalogue database [32–35].
The genomic clusters might represent transmission of an MDR/RR strain, or initial transmission of a non-MDR/RR strain that later developed multidrug resistance [12]. We all know that the drug resistance of tuberculosis is mainly conferred by mutations in genes coding for drug targets or converting enzymes [36]. Previous mutations may accumulate along the chain of transmission and acquire new mutations that further increase resistance. In some research, it has highlighted that both resistance transmission and amplification are contributing to disease burden globally [37]. In China, it has proved that additional drug resistance amplified for MTB during turnaround time for drug-susceptibility testing [38].