Dynamics of Mycobacterium tuberculosis Lineages in Oman, 2009 to 2018

Study aim. Effective Tuberculosis (TB) control measures in Oman have reduced the annual incidence of tuberculosis cases by 92% between 1981 and 2016. However, the current incidence remains above the program control target of <1 TB case per 100,000 population. This has been partly attributed to a high influx of migrants from countries with high TB burdens. The present study aimed to elucidate Mycobacterium tuberculosis infection dynamics among nationals and foreigners over a period of 10 years. Methods. The study examined TB cases reported between 2009 and 2018 and examined the spatial heterogeneity of TB cases and the distribution of M. tuberculosis genotypes defined by spoligotypes and MIRU-VNTR among Omanis and foreigners. Results. A total of 484 spoligoprofiles were detected among the examined isolates (n = 1295). These include 943 (72.8%) clustered and 352 (27.2%) unique isolates. Diverse M. tuberculosis lineages exist in all provinces in Oman, with most lineages shared between Omanis and foreigners. The most frequent spoligotypes were found to belong to EAI (318, 30.9%), CAS (310, 30.1%), T (154, 14.9%), and Beijing (88, 8.5%) lineages. However, the frequencies of these lineages differed between Omanis and foreigners. Of the clustered strains, 192 MTB isolates were further analysed via MIRU-VNTR. Each isolate exhibited a unique MIRU-VNTR profile, indicative of absence of ongoing transmission. Conclusions. TB incidence exhibits spatial heterogeneity across Oman, with high levels of diversity of M. tuberculosis lineages among Omanis and foreigners and sub-lineages shared between the two groups. However, MIRU-VNTR analysis ruled out ongoing transmission.


Introduction
TB remains a major global health problem, among the top 10 leading causes of death worldwide.
Reduced access to health facilities, as a result of the COVID-19 pandemic in 2020, has resulted in the slowing down of declines in TB incidence and increased TB mortality, with The population data used in this study were obtained from the National Centre for Statistics and Information (NCSI) based on the 2020 population census [9]. The data presented are aggregated for all provinces. A total of 2,557 unique M. tuberculosis isolates were initially considered for this study; however, 18 were excluded, one BCG and 17 without DST. The remaining 2,539 isolates were obtained from TB diagnosed cases, Omanis (n = 1,341), and foreigners (n = 1,198), between 2009 and 2018 ( Figure 1).   Ethical approval for the study was granted by the Medical Research and Ethics Committee (MREC) of the College of Medicine and Health Sciences, Sultan Qaboos University, Oman, under the reference number SQU-EC/075/18.

Spoligotyping and MIRU-VNTR Typing
Successfully sub-cultured isolates (n = 1295) were subjected to heat-killing for DNA extraction and subsequent Spacer Oligonucleotide Typing (spoligotyping) analysis [10]. These samples included MTB isolates from different years, different provinces, and both Omanis and foreigners. Spoligotyping (the classical 43-spacer format) was performed as previously described [11]. DNA samples of the M. tuberculosis H37Rv and M. bovis BCG strains were included as positive controls. Molecular biology-grade water was used as a negative control. The spoligotypes were then recorded in 43-digit binary format and compared with those recorded in the SITVIT2 database (http://www.pasteur-guadeloupe. fr:8081/SITVIT2/, accessed on 10 March 2022) to identify the spoligotype International Type (SIT) and family [12].
Out of the 943 clustered MTB isolates, 192 belonging to major clusters were randomly selected for Mycobacterial Interspersed Repetitive Unit-Variable Number Tandem Repeat (MIRU-VNTR) analysis. These samples comprised 45 clusters out of the total 132 clusters detected by spoligotyping, with 53 (24.9%) belonging to the two major clusters (Beijing/SIT 1 and CAS1-Delhi/SIT26). MIRU-VNTR typing based in the standard 24 loci was performed as described by Supply et al. (2006) on 192 M. tuberculosis samples to determine genetic relationships among isolates that belonged to 41 spoligo-lineages (clusters). First, each locus was amplified individually, as previously described. Then, the PCR products were detected in a 1.5% agarose run with a 100 bp DNA ladder, a molecular weight standard. The number of tandem repeats was calculated based on the length of the repeat and flank sequences for each locus [13].

Data Analysis
The aggregated TB cases data at each province were used to calculate the TB incidence by dividing the total notification for each province for a particular year by the population of that given year, then multiplying by 100,000 [14].
Data management and analysis were performed using SPSS version-23 (SPSS Statistics 23, IBM: Corporation 1 New Orchard Road, Armonk, NY, USA). Factors associated with clustering, e.g., nationality, gender, province, and DST profiles were assessed by chi-square and Fisher's exact tests. A p-value < 0.05 was considered as evidence of a significant difference.
Genetic clustering was defined as two or more isolates with identical spoligotypes. Recent transmission was estimated by calculating the clustering rate (CR) as: CR = (nc − c)/n; where nc = number of clustered isolates, c = number of clusters, and n = total number of isolates [15].
Binary logistic regression analysis was performed using R-packages software commands (version 1.4.1717) to find the association between the province of isolation and MTB sub-lineages. A statistical level of α < 0.05 was considered as evidence of significant association between variables. The strength of the association between MTB spoligotype and province was determined using the odds ratio and 95% confidence interval of the odds ratio.
An association was seen between some spoligo-sub-lineages and nationality (Omani vs. foreigner), except for the T and LAM clades (p > 0.5). Beijing, H, and Orphan were significantly higher among foreigners (p < 0.05), while CAS and EAI were significantly higher among Omanis (p < 0.05) ( Table 2).

MIRU-VNTR Analysis
A total of 192 M. tuberculosis isolates of 41 clusters were subjected to MIRU-VNTR analysis and 24-digit MIRU profiling to test whether these clusters consisted of identical genotypes, indicative of ongoing transmission, or if they were related but genetically distinct strains (Figure 1).
Multi-locus genotypes (MLGs) of all isolates in the above clusters were highly diverse; each isolate carried a novel genotype. All of the above 41 spoligotype clusters were broken down into unique MIRU-VNTR genotypes (Supplementary Figure S2), implying absence of recent transmission among the examined TB cases.

Discussion
This study has revealed heterogeneous TB incidence in different provinces in Oman between 2009 and 2018, with a highly diverse reservoir of M. tuberculosis lineages comprising 45 clades in 1295 isolates. Most of the isolates (72.8%) were clustered in shared spoligotypes. The vast majority of the clusters (58.3%) comprised isolates of both Omanis and foreigners, indicative of close genetic relatedness between the M. tuberculosis lineages infecting the two groups. However, analysis of the major clusters did not detect any evidence of recent transmission.
TB burden in Oman exhibits special heterogeneity, being disproportionately higher in Muscat (8.2 per 100,000 population) and Dhofar (8.4 per 100,000 population), respectively (Figure 2), and has been consistent across the study years (Supplementary Table S2). These findings are in line with a pervious analysis that revealed heterogeneity in TB incidence between 1981 and 2016, with the highest incidences in Muscat (105 per 100,000), North Batinah (an average of 63 per 100,000), and Dhofar (an average of 54 per 100,000) [5]. Muscat and Dhofar provinces have large foreign communities, which are associated with overcrowding of habitats and possibly heightened disease incidence. The large foreign communities, the majority of whom are from high-TB burden countries, can sustain latent TB reservoirs [17]. Dhofar and Muscat regions are also considered to have a high burden of multidrug-resistant TB (MDR-TB), defined as resistance to at least rifampicin (RIF) and isoniazid (INH) [18]. MDR strains have been linked to direct household transmission and clusters of recent transmission [19]. Additional characteristics associated with dissemination of the disease and high TB risk include demographic factors (e.g., living habitat), susceptibility to infection (immune status), and environmental factors-all major drivers of the heterogeneity of TB incidence [20]. A better insight into characteristics that drive high TB incidence in Oman would allow interventions targeting areas of high-risk groups.
The present study reaffirms the decreased TB incidence over the study period (2009 to 2018) that has been reported before demonstrating the effectiveness of interventions to reduce TB transmission [5,6]. Nonetheless, a pronounced reduction in TB incidence was noted among Omanis (p < 0.05), while TB rates remained consistent among foreigners (p = 0.126) ( Figure 2B).
The major spoligotype sub-lineages, EAI (318, 30.9%), CAS (310, 30.1%), T (154, 14.9%), and Beijing (88, 8.5%), detected in the present study are common in the Indian subcontinent and East Africa. The above lineages are common in the Indian subcontinent (India, Pakistan, Bangladesh) and South East Asia (Indonesia, Philippines) [21][22][23][24]. Similarly, the T, LAM, and H lineages, which together account for 26.8% of the examined isolates in Oman, are common in Africa [25]. Again, this is in accordance with the fact that a large proportion of foreigners in Oman are from Africa. These lineages are common in East African countries, such as Tanzania, which has a historical link with Oman. The distribution of M. tuberculosis lineages in Oman mirrors that in Saudi Arabia, which has a large foreign community with similar demographic characteristics [26]. Similar to Oman, predominant M. tuberculosis sub-lineages in Saudi Arabia, Delhi/CAS, and Beijing, along with EAI sub-lineages, prevail in the Indian subcontinent and Southeast Asia [26]. The structure of M. tuberculosis sublineages in Oman and Saudi Arabia can also be attributed to the central geographical position of the Arabian Peninsula at the cross-roads between Asia and Africa and its ancient trade links with Southeast Asia and Africa [27][28][29].
The major spoligotype clades remained stable in space and time, with the exception of EAI (p = 0.019) and Beijing (p = 0.022), which fluctuated significantly over the study period (Table 1). The high diversity of M. tuberculosis in Oman and the stability of sub-lineages parallel those seen in areas of high TB burden [25].This suggests the presence of a large pathogen reservoir. The introduction of novel M. tuberculosis sub-lineages into the region via foreigners from high TB burden areas can enhance the diversity and effective population size (Ne), as there is a direct relationship between the expected level of diversity and Ne [30]. A large proportion of foreigners from high-TB burden countries may harbor the infection as latent TB. During the study period, 2009-2018, the population of foreigners has increased from 1156.358 to 2030.194 (NCSI. 2022) Thus, TB imported via foreigners represents not only a risk for augmenting local transmission but also the ability to disseminate novel strains that can escape the effects of current drug regimens.
An epidemiologically relevant finding was the high similarity of M. tuberculosis linages distributed among Omanis and foreigners (Supplementary Figure S1). Out of 45 spoligotyping patterns (clades) detected, 37 (82.2%) were shared between the two groups, while only 3 (0.5%) clades (EAI2, LAM5, and Manu3) and 5 (1.1%) (EAI2-nonthaburi, LAM8, Manu1, Ural-1, and X1) were unique to Omanis and foreigners, respectively (Supplementary Figure S1). These findings are in line with the findings from Saudi Arabia, where no differences were seen among phylogenetic lineages infecting nationals and foreigners [26]. However, the pattern seen in Oman and Saudi Arabia contrasts with that reported in areas with large immigrant communities in Europe and the US, where the Euro-American lineage strongly predominates among the isolates from native patients, while lineages associated with immigrants are rarely found among these patients [31,32]. Furthermore, mixed-cluster lineages (i.e., including isolates from both native and foreigners) constituted a large proportion of the total number of patient strain clusters in Oman (58.3%) and Saudi Arabia (59.5%) [26], while the above pattern of clustering has rarely been seen in Western countries [33][34][35]. The high rate of mixed-cluster phylogenetic lineages among nationals and foreigners in Oman and Saudi Arabia suggests a prevailing culture of close interaction that augments the probability of transmission between the two groups [36]. A large proportion of foreigners in Oman and Saudi Arabia occupy household jobs and are in daily contact with local families [26], whereas socio-economic and cultural barriers reduce contact between foreigners and native populations in Europe and US settings [37].
In the present study, the vast majority of examined M. tuberculosis isolates 943/1296 (72.8%) were grouped into 132 clusters of variable sizes, whereas 352 (27.2%) were distinctive. Of the 132 clusters, 85 were minor spoligo-patterns (2 to 4 isolates), while 47 were regarded as major spoligo-patterns (5 or more isolates). The biggest clusters included CAS1_Delhi, SIT 26 (n = 128), Beijing SIT = 1 (n = 82), EAI5 SIT = 8 (n = 40), and CAS1_Delhi SIT = 25 (n = 40). Analysis of correlations between clustering and demographic variables showed that clustering was not influenced by demographic variables, including patient gender, age, province, and DST profile (p > 0.05). The lack of association between clustering and M. tuberculosis drug response profile has previously been reported in other countries [36]; however, a comparatively higher clustering was seen among Omanis (74.4%) compared to foreigners (68.8%) (p = 0.021) ( Table 3). These findings are consistent with those among TB patients in Saudi Arabia indicating that nationals had a higher rate of clustering (50.3%) compared to non-Saudi patients (45.8%) . This contrasts with the pattern of clustering in Western countries, where limited shared clustering is generally found between native and immigrants TB populations [37].
We further analyzed M. tuberculosis in large clusters in Oman, using standard 24-locus MIRU-VNTR typing, known to provide superior discriminatory power, to test the hypothesis that the persistence of TB in Oman is driven by transmission between foreigners and Omanis [6,38]. All of the 192 isolates, in 41 clusters, were found to have a unique combination of the examined MIRU-VNTR loci, hence there was no evidence of transmission between Omanis and foreigners or among any of the two groups. Thus, in the case of Oman, MIRU_VNTR offers an advantage over spoligotyping for genotyping M. tuberculosis to assess areas of clustering and active transmission for targeted control. Nonetheless, whole genome sequencing, which has a much higher discriminatory power than MIRU-VNTR, can be considered for clinical strains exhibiting identical MIRU-VNTR genotyping profiles for better interpretation of transmission patterns.
In summary, the present study has revealed heterogeneous spatial TB risk in Oman and a high extent of genetic diversity in M. tuberculosis, consistent across the ten years of the study period. Nonetheless, there is high similarity between M. tuberculosis lineages infecting Omanis and foreigners and a high proportion of shared clustered lineages containing isolates of the two groups, indicative of close social interaction favorable for TB transmission. Nonetheless, no evidence of ongoing transmission was evident from MIRU-VNTR alleles. Further analysis of recent incidences in high-TB risk areas, in Muscat and Dhofar, using highly discriminatory markers, MIRU-VNTR or whole genome, is critical for targeted intervention. Data Availability Statement: The authors confirm that the data supporting the findings of this study are available within the article if required.

Conflicts of Interest:
The authors declare that they have no competing interest.