Using molecular typing, twenty-five percent of relapse cases (65; 25%) were clustered in 29 clusters. A retrospective studies revealed that patients in clusters developed their second episode of TB within the same period. Further analysis of spoligopatterns identified Haarlem I and Beijing types of
M. tuberculosis strains in thirty eight percent of patients in clusters(38%)(table.
2). Therefore, the possibility of exogenous reinfection through transmission of particular
M. tuberculosis strains were highlighted. Although, due to non-availability of previous culture genotyping results, we could not confirm the exogenous reinfection in them. Recently, investigators have suggested that the relative contribution of exogenous reinfection increases in parallel with the incidence of disease [
6,
8,
18]. Most reported cases of exogenous reinfection observed among alcoholic residents of a homeless shelter or patients with advanced HIV infection [
6,
9,
19]. We found no particular risk factors between patients in cluster and non-cluster cases. Retrospective analysis of cluster cases identified 41% of intra- community transmission between Iranian and Afghan TB patients. In our previous study, the impact of intra- community transmission was much lower (13.7%) than present result. Furthermore, we found that the Haarlem I and Beijing type of
M. tuberculosis strains were the most frequent super families in intra-community transmission (figure.
1). The Haarlem I and Beijing strains have been reported in different geographical regions of the world and they thought to possess selective advantages in comparison to other
M. tuberculosis super families [
22,
23]. Previously, we demonstrated that 44% of MDR-TB patients in intra-community transmission were belonged to HaarlemI (73%) and Beijing (27%) superfamilies [
21]. Therefore, based on previous and present reports, it is clear that both Haarlem I and Beijing strains can cause epidemic and from epidemiological point it is necessary to conduct more extensive surveillance of MDR-TB strains because they might cause serious outbreaks [
12]. During the year 2000–2005, 32% of initial TB patients that referred to our unit were from Afghan born immigrants [
20]. Majority of these patients (58%) had either resistant to any drug or drug combination including MDR-TB [
20]. In present report also, the rate of resistance to any drug or drug combination in Afghan patients were more than two folds as compared to Iranian (table
1). These finding highlighted the need to reinforce the TB policy measures with regards to screening immigrants from neighboring countries, which is absent in the current system. Based on IS6110-RFLP, 74.8% of patients were grouped in non- cluster cases and it was assumed that reactivation of endogenous infection remains the more probable cause of active tuberculosis, in studied populations (table.
1). Today, it is known that sterilization of a pulmonary lesion is possible through effective treatment regimens. But, it is also accepted that subsequent episodes of TB are almost invariably caused by endogenous reactivation of resistant strains [
7,
24]. In other words, majority of those who returned for treatment after default might develop resistant in comparison to those who returned after exogenous reinfection [
7,
9,
25]. In this regards, we also found high number of MDR-TB strains among endogenous reactivation cases. The frequent superfamilies in MDR-TB cases of non-cluster cases were CASI (19; 31%) and EAI3 (12; 20%) (table.
2). Based on Sreevatsan
et al, the EAI and CAS are belong to genetic group I organisms which are evolutionary older and they are the most frequent superfamilies in Central Asia and Middle East [
26]. Our results also showed different susceptibility patterns for isolates in same clusters e.g., in the cluster numbered-27 (figure.
1); one strain was susceptible to all drugs tested and the other two strains were resistant to isoniazid. Mitchison [
27] has already described the emergence of such drug resistance strains, solely due to irregularity in administration of drugs. That means the strains became acquired drug resistance isolates, as classically defined [
24,
25]. In the studied population, no relation was found between the patterns of IS6110-RFLP and susceptibility results. In fact, the number and manner of IS6110 positioning along the genome of
M. tuberculosis does not have any relation with drug-susceptibility results [
2,
4]. Last but not the least, the major consideration of the usefulness of the IS6110-RFLP typing method is its specificity, which depends on the number of bands obtained. In the present study, 12% of collected strains had low copies of IS6110. Previously, we detected only 5.4% of strains with low IS6110 copy number [
20]. Thereby, the prevalence of
M. tuberculosis isolates with low or no IS6110 insert is not clear and further studies are required to show the real distribution of these strains within the country.