Despite the efficacy of M. hyopneumoniae vaccines in reducing lung lesions [26] and bacterial load from the respiratory tract [27], they are not capable to fully eliminate the bacteria from the animal giving rise to typical MLL in a variable percentage of pigs at abattoir in different countries [10, 14, 16, 17]. To corroborate the usefulness of a mycoplasma-like scoring system to be adapted to abattoirs with a high number of slaughtered pigs per hour we investigated the prevalence of mycoplasma lung lesions in Spain and Portugal in a study with an elevated number of pigs (199,678) involving a high number of farms (221). Additionally, a valuable sample of 788 lungs was considered to corroborate microscopically the presence of lesions characteristics of the infection by M. hyopneumoniae.
In our study, 50.26% from all lungs examined at abattoir did not exhibit any gross lesion and roughly 31% showed cranioventral consolidations compatible with mycoplasma lesions, with similar values for Spain (31.14%) and Portugal (29.95%). These prevalence rates of MLL were higher than those observed in Belgium (23.85%) [14] but lower than those previously found in Spain (44.61 to 55.69%) [9, 16], Italy (46.38%) [10], France (69.30%) [15] or Germany (72.60%) [17]. The main difference between studies was the number of examined lungs, with nearly 200,000 lungs evaluated in this study in comparison with numbers ranging from 600 [16] to 10,404 [9]. However, other factors such as the epidemiological scenario in each country play also a role in these differences. Consequently, taking into account the high number of farms and the different pig production areas, we consider that our data provide a precise percentage of prevalence of MLL in Spain and Portugal. Moreover, if we consider the results obtained in the histopathological study and quantify the percentage of selected lungs with MLL that really presented microscopic lesions compatible with M. hyopneumoniae infection (78.17%), the real prevalence of gross lesions caused by the bacteria would be reduced from 30.97% to 24.21%. Other lesions, such as suppurative bronchopneumonia, as shown in Fig. 2, have a similar gross appearance [28, 29] and, hence, could be mixed up with the lesions caused by M. hyopneumoniae. In this sense, those studies in which a microscopic confirmation has not been performed, could bias the results overestimating the percentages of prevalence. In the study carried out by Luhers et al. [17], 78.30% out of 400 lungs with MLL collected at the abattoir were positive to M. hyopneumoniae by PCR, a percentage very similar to the one found with compatible microscopic lesions in our study (78.17%). These results highlight the interest of performing additional studies on lungs with MLL to confirm the diagnosis of EP or infection by M. hyopneumoniae.
Pleurisy, associated to dorsocaudal infarcts or alone, was the second most prevalent gross lesion found in our study affecting 18.77% of the examined lungs, a percentage very similar to that found in France (15.00%) [15] but lower than those previously reported in Italy (25.10%) [10] and in Spain (26.80%) [9]. In our case, most of the pleurisy (66.65%) was associated with dorsocaudal infarcts, a hallmark of A. pleuropneumoniae infection, therefore, the measures to control this pathogen (management, therapeutic or vaccination) should be revised and/or implemented in the farms included in the study.
According to our findings, most of the lesions of bronchointerstitial and interstitial pneumonia (characteristics of mycoplasma and viral infections, respectively) had a chronic course, by contrast, the majority of the lesions of suppurative and fibrinous bronchopneumonia (characteristics of bacterial infection) had an acute course and a higher degree of severity. These results suggest that viruses and mycoplasmas could have acted in an earlier stage of the life of piglets, probably during nursery, and later on, during the fattening period, bacteria would have taken action. Ruggeri et al. [30] reported that the most prevalent microscopic lesions in fattening pigs were pleurisy, followed by pleuropneumonia, catarrhal bronchopneumonia and bronchointerstitial pneumonia, but animals included in that study died because of respiratory diseases, that is, samples were not collected from healthy animals at abattoir as in the present study. However, to a certain extent, it agrees with our observations since most of the deaths were consequence of an acute process of fibrinous and suppurative bronchopneumonia. The recent study carried out in Brazil by Galdeano et al. [13], also found the characteristic lesions of M. hyopneumoniae infection as the most prevalent ones, which were detected in 63.75% of the lungs examined microscopically, but different to our study, the following observed lesion was chronic bronchopneumonia (57.14%) and other lung lesions common in our study such as suppurative bronchopneumonia, interstitial pneumonia or fibrinous bronchopneumonia, were observed in a lesser extent (15.63%, 3.61% and 0.15%, respectively). Nevertheless, it must be taken into account that in Brazil porcine reproductive and respiratory syndrome virus (PRRSV) has never been detected [31], then the effect of one of the main primary agents involved in the PRDC [5], alone or in combination with other pathogens, is not taken place in those farms. Therefore, the clinical and lesional picture at abattoir in Brazil would be totally different when compared with those countries where PRRSV is endemic such as Spain or Portugal.
Besides to allow us having a more accurate idea of the real prevalence of lesions caused by M. hyopneumoniae, the histopathological study provided information about the action of other pathogens involved in the PRDC that sometimes go clinically unnoticed. By contrast, some lesions are unspecific and could be caused by different pathogens, but in association with other techniques, such as serology, bacteriology or PCR, histopathology could provide a more accurate information about which agent is causing the lesion, since a positive result in those diagnostic techniques against some pathogens (i.e. M. hyopneumoniae or porcine circovirus type 2 (PCV2)) does not necessarily always mean that they are causing any lesion.
Lesions of bronchointerstitial pneumonia compatible with M. hyopneumoniae infection were found in the six more frequent combinations of lesion patterns in our study, representing 66.13% of the lungs examined, what support the fact that this agent plays a relevant role in the PRDC as primary agent as well as enhancing the action of other pathogens involved in as PRRSV [32], PCV2 [33] or swine influenza virus [34].
The scoring system from 0 to 5 points used in this study for MLL, adapted from a previous one [35], has been shown as a simple and repeatable method that can be easily applied in cases of abattoirs with fast slaughterline (sometimes more than 500 pigs per hour). By contrast, it is not as precise as other methods which express the proportion of affected lung area in percentages [19, 20, 21, 22]. Therefore, in the case of a disease with an evident economic importance as EP, to complete a study of prevalence and severity of lung lesions with their relationship with productive parameters as ADWG, it is necessary to translate the score 0 to 5 to percentage of affected area, as was done in our study.
The average loss of ADWG calculated in our study was 22.33 g/day (22.44 g/day for Spain and 22.10 g/day for Portugal). In a study carried out in Spain by Bringas et al. [24] involving almost 50,000 pigs, they used a similar scoring system that the herein proposed (0 to 5 points) and divided the batches according the percentage of maximum lung lesions (lungs with score 3 or higher) in four quartiles. Comparing with our results, we would have an 8.90% of lungs with a score 3 or higher and would be included in their second quartile (5-12.50% of lungs with maximum lesions), and the difference with batches included in the first quartile (0–5% of lungs with maximum lesions) would be of 23 g in the ADWG, almost the same loss found in our study but calculated according the surface of lung affected. In the light of the above mentioned, both methods have been revealed as a valuable approach to calculate the loss in growth rate associated to MLL using a 0 to 5 scoring system.