High proportion of tuberculosis transmission among social contacts in rural China: a 12-year prospective population-based genomic epidemiological study

Abstract

ABSTRACTTuberculosis (TB) is more prevalent in rural than urban areas in China, and delineating TB transmission patterns in rural populations could improve TB control. We conducted a prospective population-based study of culture-positive pulmonary TB patients diagnosed between July 1, 2009 and December 31, 2020 in two rural counties in China. Genomic clusters were defined with a threshold distance of 12-single-nucleotide-polymorphisms, based on whole-genome sequencing. Risk factors for clustering were identified by logistic regression. Transmission links were sought through epidemiological investigation of genomic-clustered patients. Of 1517 and 751 culture-positive pulmonary TB patients in Wusheng and Wuchang counties, respectively, 1289 and 699 strains were sequenced. Overall, 624 (31.4%, 624/1988) patients were grouped into 225 genomic clusters. Epidemiological links were confirmed in 41.8% (196/469) of clustered isolates, including family (32.7%, 64/196) and social contacts (67.3%, 132/196). Social contacts were generally with relatives, within the community or in shared aggregated settings outside the community, but the proportion of clustered contacts in each category differed between the two sites. The time interval between diagnosis of student cases and contacts was significantly shorter than family and social contacts, probably due to enhanced student contact screening. Transmission of multidrug-resistant (MDR) strains was likely responsible for 81.4% (83/102) of MDR-TB cases, with minimal acquisition of additional resistance mutations. A large proportion of TB transmission in rural China occurred among social contacts, suggesting that active screening and aggressive contact tracing could benefit TB control, but contact screening should be tailored to local patterns of social interactions.

Keywords: Tuberculosis; rural China; social contacts; transmission; whole-genome sequencing.

Figure 1.

Sample enrolment in (A) Wusheng and (B) Wuchang. NTM: non-tuberculous mycobacteria.

Figure 2.

Phylogeny, clustering, and resistance profile of 1289 Mycobacterium tuberculosis strains isolated in (A) Wusheng and 699 Mycobacterium tuberculosis strains isolated in (B) Wuchang. The different colours on the branches indicate different lineages and sublineages. The outer grey circle indicates genomic-clustered strains differing by ≤ 12 single-nucleotide polymorphisms. The outer yellow-red circle indicates other drug resistance and multidrug resistance.

Figure 3.

Cumulative clustering rate in (A) Wusheng and (B) Wuchang, 2009-2020. The bar indicates the cumulative number of cases, the line indicates the cumulative clustering rate.

Figure 4.

Categories of confirmed epidemiological links and their time interval between diagnoses. The four colours of the circle indicate four categories of contacts with confirmed epidemiological links. The darker the colour, the closer the relationship. (A) The black border indicates family contacts (Family) and the grey border indicates social contacts (Relatives, Community and Aggregated settings outside the community). The number in the ring indicates the number and proportion of each category of contacts. The width of the ring indicates the proportion of each category. (B) The time interval between diagnoses in the four categories of contacts and students. Due to the small number of relatives, they were not compared with the other categories. **p < 0.01, NS refers to no significance (given by Wilcoxon rank sum test).

Figure 5.

Putative transmission network for 3 clusters based on the structure of the genomic phylogeny and the epidemiological links. (A) Transmission network of Cluster 60 of Wuchang. HLJ2014_435, HLJ2014_437, HLJ2010_148, HLJ2009_24_1, HLJ2013_345 and HLJ2015_422 all live in village A. HLJ2014_435 and HLJ2014_437 are couples and they often go to the village shop to chat and play card. HLJ2010_148, HLJ2009_24_1, HLJ2013_345 and HLJ2015_422 are neighbours of HLJ2014_435 and HLJ2014_437. HLJ2010_148, HLJ2009_24_1 and HLJ2013_345 also often go to the village shop to chat and play card. HLJ2017_705 and HLJ2017_666 live in village B and know each other. HLJ2019_894 and HLJ2014_399 live in village C and village D, and no epidemiological links was found with other patients. (B) Transmission network of Cluster 67 of Wuchang. SC2013_0552 and SC2013_0694 are friends. SC2013_0322 and SC2014_0573 are grandsons of SC2013_0552. SC2015_0169, SC2016_0302 and SC2018_0169 are classmates of SC2014_0573. (C) Transmission network of Cluster 14 of Wuchang. SC2013_0625, SC2009_0069, SC2011_0316, SC2013_0100 and SC2013_0659 live in different villages. They know each other and go to the teahouse in the township almost every day to drink tea, chat and play mahjong. The colour of the line indicates the categories of contacts in Figure 4. Epi-link: epidemiological link.