In the present study, the occurrence of TB was investigated in slaughter pigs in two abattoirs from Addis Ababa and Bishoftu. Although the pig industry is growing in recent years and many piggeries are being established in many parts of the country public slaughtering is carried out only in Addis Ababa Abattoirs Enterprise slaughterhouse. The meat inspection activity is under close supervision by the federal ministry of agriculture where every carcass is inspected by recognized meat inspector. The slaughterhouse in Bishoftu is part of a private swine farm where small numbers of pigs are slaughtered. Carcass is not inspected by meat inspectors in this slaughterhouse. In addition to these two abattoirs are several backyard slaughters in many of small scale swine production systems which are favoured by increasing demand for pork by foreigners particularly by Chinese people dwelling in this regions.
This study is the first of its kind in Ethiopia to investigate the molecular epidemiology of TB in pigs. Post-mortem and bacteriological examinations as well as molecular typing were applied. The estimated prevalence of TB was approximately 6%, and is was lower than prevalence estimates reported from other African countries including Cameroon , Egypt  and Uganda . On the other hand, it is higher than those reported from developed countries including New Zealand , The Netherlands , Czech Republic  and USA . Such differences in the prevalence of TB in pigs between developing and developed countries could be associated with the eradication of TB in human and cattle in the developed world, where cattle served as sources of infection to pigs [30, 31].
A significant association was detected among the age classes and the prevalence of TB lesion in pigs. Tuberculous lesions occurred more frequently in older pigs than younger pigs, which may be due to chronic nature of the disease, which requires a longer time to produce detectable lesions, and due to a longer period of exposure in older pigs. This findings is consistent with reports from other countries and in other species of animals [29, 32, 33].
Pigs that were kept on free grazing and fed with swill, offal or left to roam for garbage were found to harbour mycobacterial infections twice more frequently than those fed on commercial mixed feed, although the difference was not found to be significant. Outbreaks of M. tuberculosis in pigs which have been associated with the feeding of uncooked garbage from hospitals or residences housing human cases have been reported .
A higher estimated prevalence was recorded in pigs from Addis Ababa and Special Oromia Zone as compared to those slaughtered at Bishoftu. This variation could be a reflection of the husbandry system in that pigs are reared in Addis Ababa and the nearby Special Oromia Zone, where animals are kept under poor husbandry practices such as feeding swill, poultry litter, abattoir offal and garbage; sheltering with other domestic animals and confinement in poor housing system and where they have close contact with other animals and humans. In Ethiopia, the traditional small scale system, which is the predominant pig production system particularly in Addis Ababa and its nearby Special Oromia Zone, is characterised by absence or minimal health care, supplementary feeding and proper housing. It has been reported that the pig production is aggregated in the central part of the country , which could also promote the transmission of mycobacterial infections among the different farms through the movement of pigs from one farm to another particularly during establishment of new farms.
Lesions suggestive of TB were observed in the lungs, liver, head and were associated with mesenteric lymph nodes. Lesions appeared more frequently in the head, the mesenteric lymph nodes and in the liver. The presence of tuberculous lesions in the liver, the lymph nodes of the head and mesenteries may indicate infection by ingestion of contaminated feed, offal or infection from scavenging on contaminated garbage [30, 32], while lesions in the lung and its associated lymph nodes suggest airborne transmission . The observation of a higher proportion (>65%) of lesions in the lymph nodes of the head and gastrointestinal organs suggests ingestion to be the most common route of transmission in pigs. Similar observations were previously reported from Uganda [29, 30]. However, this cannot exclude airborne transmission in pigs.
Only about 31% of the lesions yielded mycobacterial growth in culture. Lower rate of mycobacterial growth from tissues with gross TB lesions have been reported from Czech Republic , Egypt , Uganda  and many other countries . Failure to demonstrate tubercle bacilli may be due to the occurrence of healed processes that contain no longer viable tubercle bacilli, or to microorganisms other than tubercle bacilli causing the lesions, such as Rhodococcus equi or R. sputi or inadequacy of the methods used for isolating tubercle bacilli [30, 32]. It could also be due to subjective differences in identifying tuberculous lesions . Nevertheless, culturing tubercle bacilli from such lesions needs further investigation so as to enable a maximum yield of viable organisms from culture for a correct interpretation of the results and thereby to appropriately explain the magnitude of the disease.
Isolation of M. tuberculosis, which is a predominant agent of tuberculosis in humans , from pigs suggests transmission between human and pigs, which could occur as a result of close contact between the two species , feeding of undercooked garbage or by sputum or body secretions from infected individuals . The isolation of M. tuberculosis in the lung and its associated lymph nodes further supports the idea proposed by Parra and co-workers  that swine are not dead end hosts for mammalian tuberculosis. Previous studies suggested interspecies transmission of mycobacteria in Ethiopia [19, 20, 37–40]. Similar findings were also reported from other countries where the burden of TB in humans and animals is high and where transmission between these two is likely to occur on a regular basis [3, 41]. Nonetheless, most reports from developed countries showed that lesions of TB in swine are due to members of the MAC [17, 30, 42–44]. In view of the fact that such reports are from countries, where there is an on-going mammalian TB control program, this difference may be due to the control of TB achieved by these countries .
Early in the 20th century, when TB in cattle and humans was more prevalent, TB in swine was either due to M. bovis or M. tuberculosis. However, for example in the USA the avian type TB began to occur more frequently in swine by 1925. Today, isolation of mycobacteria other than M. avium from swine is uncommon in the USA. In rare cases, it occurs in pigs which are kept on the same premises with M. bovis infected cattle or which have been in close contact with M. tuberculosis infected humans . After the enforcement of strict regulations for a test-and-slaughter policy, the prevalence of bovine type of TB in swine gradually declined concurrently with the eradication of the disease in cattle in most western countries [17, 30, 31].
In this study, one of the AFB positive isolates had a PCR product size indicative for the Genus Mycobacterium but none for MTC or MAC upon genus typing using multiplex PCR. The isolate was hence assumed to be a member of the mycobacteria other than tuberculosis (MOTT). Isolation of MOTT from swine has previously been reported from other countries [3, 4]. In Ethiopia, MOTT have been isolated from cattle  and camels  with tuberculous-like lesions, which indicates their involvement in a broad range of animal species  and their role as a cause of tuberculosis.
The two isolates in one of the clusters with identified SIT1088 have been detected in Egypt , India, South Africa and Portugal, while the other single isolate identified as SIT1995 has been reported from India . In Ethiopia, several studies recently reported new strains of M. tuberculosis[20, 33, 40, 47].