Tuberculin reactions and pathological lesions in the pigs imported to Serbia indicated mycobacterial infection. More than 50% of the animals reacted positive for both avian and bovine tuberculin. Only three pigs reacted positive to bovine and negative to avian tuberculin. None of the FFPE lymph nodes originating from these three pigs were positive on real time PCR for IS6110, while two were positive for IS1245. Cross reactions between avian and bovine tuberculin are commonly observed in cattle , and our results indicate that such reactions are common in pigs as well.
In all but 12 pigs, pathological lesions were detected, indicating a very high infection rate in the herd. Only one of the PCR examined FFPE lymph nodes from the pigs without pathological lesions was positive for IS1245 by PCR, confirming that a thorough macroscopic and microscopic examination will detect most of the infected pigs. However, neither tuberculin reactions nor pathology can differentiate between infections with M. bovis and M. avium.
Culturing was unsuccessful, possibly due to low number of bacteria in the samples, uneven distribution of bacteria in the lesions or the lack of different growth media used for culture. As the samples were discarded, reculturing was not possible. Still, positive reactions for IS1245 on real-time PCR confirmed presence of M. avium in 55% of the FFPE lymph nodes. It was therefore assumed that the animals imported to Serbia were infected with M. avium.
Real-time PCR for IS6110 gave positive reactions in FFPE lymph nodes from ten pigs in Serbia. This might indicate an infection with the MTC, possibly before import. However, this is unlikely as Lithuania is officially free of bovine tuberculosis today and the last occurrence in domestic animals was reported in 2001 [23, 24]. In Serbia, bovine tuberculosis is still present in domestic animals , but considering the relatively short lapse of time between import to Serbia and sampling of the lymph nodes, and the strict quarantine measures applied, it seems unlikely that the animals were infected with MTC after import. The results are possibly due to false positive reactions or to cross reactions with DNA sequences yet to be described. However, without the possibility to culture mycobacteria from the IS6110 positive lymph nodes, it is not possible to conclude with 100% certainty whether any of these pigs were infected by bacteria within the MTC.
The sensitivity and specificity of the real-time PCRs for IS6110 and IS1245 cannot be determined as culturing in Serbia was not successful for these samples. Real-time PCR is considered an efficient, highly sensitive and specific diagnostic tool, using the appropriate controls and protocols. However, false negatives on real-time PCR might be due to impaired DNA quality, as formalin fixation and paraffin embedding are reported to compromise the integrity of DNA . To counteract the negative effect of formalin fixation and paraffin embedding on DNA quality, the target sequences were designed to be short (≤ 200 bp) and samples with Ct values up to 40 were considered positive. These measures lower the possibility of false negatives, but also increase the risk of false positive results. Seven of the ten samples positive for IS6110 were also positive for IS1245. This could indicate multiple infections or false positive reactions. The specificity of IS6110 PCR in clinical material have been described to be between 96.1% and 100% [26, 27], but could be lower for real time PCR in FFPE lymph nodes.
The animals were kept in quarantine in Serbia, and the first tuberculin testing was performed only 10 days after arrival. It was therefore assumed that any mycobacterial infection was contracted in Lithuania before export. Through culture and RFLP analysis of samples from the Lithuanian breeding facility, this suspicion was further confirmed, as M. avium subsp. hominissuis was detected in several samples from pigs and from their environment. The lymph nodes sampled from the Lithuanian breeding herd were randomly collected from healthy swine at slaughter, showing no visible external lesions. Still, all were positive for M. avium subsp. hominissuis, indicating a very high level of infection in the herd. M. avium subsp. hominissuis was also isolated from half of the sawdust and all peat samples, showing that the pigs lived in a contaminated environment.
Identical and closely related isolates from peat and porcine lymph nodes were detected by RFLP analysis, indicating peat as a possible source of infection for the pigs. Four porcine and four peat isolates identical on IS1311 RFLP, showed minor differences in the IS1245 RFLP profiles. Such differences have earlier been observed both on subculture and in multiple cultures from the same individual [28, 29]. Tenover et al.  demonstrated how one mutation could result in two or three band differences in pulsed field gel electrophoresis patterns and that such patterns could be considered subtypes of the outbreak pattern when isolates are epidemiologically linked. A similar interpretation may be valid for RFLP patterns, which would indicate a close genetic relationship between these eight isolates. The discrepancies in RFLP profiles between porcine isolates and isolates originating from sawdust indicate that mycobacteria in sawdust were not the source of infection in the pigs. Two isolates originating from the same animal clustered in two different clusters. This suggests an infection with two different variants, and could be related to a high infection pressure in the environment. Such infections have been described earlier, both in humans and in swine [31–34].
In the Lithuanian herd, iron enriched peat was fed to piglets as an iron supplement and for regulation of intestinal function, while sawdust was used for bedding. M. avium subsp. hominissuis has been isolated from both peat and sawdust used in swine farms earlier, and related RFLP types have been detected in both peat and swine and in sawdust and swine [35–37]. Feeding contaminated peat to small piglets could explain the high level of infection in this herd, because young animals are more susceptible to infectious agents than older animals. After the source of infection was identified, all peat and sawdust was removed, the facilities were cleaned and disinfected, and new heat treated peat was introduced in the herd.