Histopathologic patterns and etiologic diagnosis of porcine respiratory disease complex in Brazil

Porcine respiratory disease complex is a major health concern for the porcine industry, causing significant economic loss. In this study, a total of 156 samples from pigs referred to a diagnostic laboratory in Brazil for 15 months were analyzed by histopathology, bacterial isolation, PCR, and immunohistochemistry. Multiple infections were common, so 42.3% of the pigs had more than one pathogen detected in the lungs. Swine influenza virus was detected in 25.0% of the cases. Porcine circovirus type 2 was detected in 7.1% of the pigs, which was often associated with Pasteurella multocida . In addition, one case of porcine circovirus type 3 infection associated with granulomatous pneumonia was diagnosed. Bacteria were isolated in 125 cases, namely Pasteurella multocida (34.0%), Glaesserella ( Haemophilus ) parasuis (35.2%) , Streptococcus suis (13.5%), and Actinobacillus pleuropneumoniae (7.7%). Mycoplasma hyopneumoniae was identified in 7.0% of the cases, and 18.6% of pigs carried Salmonella sp. The most common patterns of pulmonary inflammation were broncopneumonia, bronchointerstitial pneumonia, and pleuritis, in that order. This study demonstrated that histopathology is an efficient tool along with other laboratorial diagnostic tests for establishing an etiologic diagnosis in cases of porcine respiratory disease complex.


INTRODUCTION
Respiratory diseases in pigs are one of the most important health problems in highly technified porcine production systems, being a cause of economic losses due to deaths, treatments, low weight gain, and condemnation of carcasses at the abattoir (Sobestiansky et al., 1987;Fraile et al., 2010;Opriessnig et al., 2011;Morés et al., 2015).Porcine Respiratory Disease Complex (PRDC) is a terminology that refers to swine mixed respiratory infections with multiple etiologic agents (Opriessnig et al., 2011).It is known that the environment, management practices, and the type of production system, might be relevant predisposing factors to PRDC supporting the notion of a multifactorial complex (Brockmeier et al., 2002;Fraile et al., 2010;Morés et al., 2015).
Several different pathogens can colonize the respiratory tract of pigs.Although innate immunity may control most primary infections, immunocompromised pigs may develop lesions due to primary pathogens or when lesions are exacerbated by the pathogenicity of secondary infectious agents (Brockmeier et al., 2002).Indeed, some viruses and bacteria can induce important lesions because of a primary infection, but polymicrobial respiratory diseases are more common than single infections in porcine herds (Opriessnig et al., 2011).Conclusive etiologic diagnosis of respiratory diseases is challenging and usually requires multiple diagnostic tests.Although viruses are often considered primary agents that predispose to secondary bacterial infections, pathogenesis of multiple infections in the respiratory tract of pigs can be more complex so identification of primary or secondary infections may not be achievable (Opriessnig et al., 2011).
Viruses including swine influenza virus (SIV), porcine circovirus type 2 (PCV2) and porcine reproductive and respiratory syndrome virus (PRRS) are important primary agents in PRDC (Brockmeier et al., 2002).It is known that PCV2 infections may cause respiratory lesions including interstitial pneumonia, bronchial and bronchiolar fibroplasia and necrosis (Kim et al., 2003).However, PCV2 can be involved in subclinical infections with lymphoid depletion and, consequently, immunosuppression, predisposing to pulmonary infections.
Although there are important primary pulmonary infections by bacteria, including, Mycoplasma hyopneumoniae, Actinobacillus pleuropneumoniae, and Bordetella bronchyseptica, numerous bacteria can cause secondary infections resulting in exacerbation of respiratory clinical signs and lesions (Brockmeier et al., 2002).Infections with A. pleuropneumoniae are usually associated with necrosis of pulmonary parenchyma and pleuritis.Necrotizing lesions in these cases are due to the exotoxins ApxI, II and III (Frey, 1995).pyogenes.Salmonella enterica serotype Choleraesuis may be associated to primary or secondary pulmonary infections, due to the systemic behavior of this bacterium (Brockmeier et al., 2002).
Considering the economic losses due to PRDC, an accurate diagnosis is critical for the swine industry.Once the frequency and importance of different respiratory pathogens may vary among countries or regions, studies describing correlating lesions and specific agents are important to support veterinary pathologists, clinicians, and farmers (Fraile et al., 2010;Hansen et al., 2010;Morés et al., 2015).Although there are some studies on the etiology of PRDC in various countries (Hansen et al., 2010;Jimenez et al., 2014;Haimi-Hakala et al., 2017;Cheong et al., 2017) only a few of them include histopathology (Hansen et al., 2010).In Brazil, reports describing etiology of PRDC in Brazilian porcine herds are scarce (Schimidt et al., 2016;Galdeano et al., 2019) and none of them describe the association of the etiologic and pathologic diagnosis.Therefore, the aim of this study was to evaluate cases of PRDC from Brazilian swine herds and correlate the morphologic lesion with the etiologic agent.For isolation of G. parasuis, samples were plated on blood agar with Staphylococcus aureus nurse streak, to prevent satellite growth.Isolates were morphologically analyzed by Gram staining (Gram-negative cocobacillus), and biochemical analysis included catalase, CAMP, and urease.Isolation of S. suis was performed by plating samples on blood agar to identify alpha or beta hemolytic strains.Gram-stained smears smears of isolates demonstrated Gram-positive cocci arranged in chains, and biochemical analysis included catalase, amylase and NaCl.

Swine
For isolation of P. multocida samples were cultivated on blood agar to identify colonies with gray mucoid pattern or with gray non-mucoid pattern.Gram-stained smears demonstrated small Gram-negative coccobacillus, and biochemical analysis included catalase, oxidases, hyaluronidase and acriflavine to differentiate serotypes type A and D.
For isolation of A. pleuropneumoniae samples were plated on blood agar with 5% sheep blood streaked with Staphylococcus aureus, as NAD donor, generating hemolysis.Selected colonies (morphologically compatible with A. pleuropneumoniae) were then plated on tryptose agar.Gram-stained smears demonstrated Gramnegative coccobacilli, and biochemical analysis included NAD requirement, hemolysis, catalase, CAMP test, urea, acid production from arabinose, lactose, maltose, mannitol, melibiose, sucrose, and trehalose.
For isolation of T. pyogenes, samples were cultivated on blood agar to identify flat small colonies with mild hemolysis.Gram-stained smears demonstrated pleomorphic Gram-positive coccobacillus, and biochemical analysis included catalase and NaCl.
Occasionally samples were processed for isolation of Salmonella spp.by incubating samples in selective culture growth medium followed by culture on green shine agar, to identify pink small and irregular colonies.Gram staining demonstrated Gram-negative bacilli, and biochemical analysis included indol, citrate, urease, and lysine as well as motility.
Lung lesions were morphologically categorized based on the pattern of inflammatory changes in the pulmonary parenchyma as follow: (i) septal inflammatory infiltrate characterizing interstitial pneumonia; (ii) intraluminal alveolar and/or bronchiolar inflammatory infiltrate, characterizing bronchopneumonia, (iii) intraluminal and septal inflammatory infiltrate, categorizing bronchointerstitial pneumonia; and (iv) pleural inflammation, characterizing pleuritis or pleuropneumonia.Pulmonary parenchyma necrosis, bronchial and/or bronchiolar epithelial damage, including intraepithelial microabscesses, inflammation and necrosis, and bronchial associated lymphoid tissue (BALT) hyperplasia were assessed separately from the predominant morphologic pattern of pulmonary inflammation.
Frequencies of agents were compared among different categories (age groups) using the Fisher's exact test (Sampaio, 2010).

RESULTS
Data on age, sex, histopathologic findings, diagnostic techniques and results of each pig included in this study are summarized in Supplementary Table S1.
Considering all 156 pigs included in this study, lung etiologic evaluation demonstrated that 42.3% (66/156 pigs) had more than one pathogen detected, 53.8% (84/156 pigs) had only one pathogen detect and 3.8% (6/156 pigs) had no pathogens detect in their lungs.From agents considered as primary infections, SIV was the most frequent, and it was detected in 25% of the cases (39/156 pigs), followed by A. pleuropneumoniae that was detected in 7.7% of the cases (12/156 pigs).M. hyopneumoniae was detected in 7.0% (11/156) of pigs included in this study, affecting only nursery and finishing pigs, with 6 and 5 cases in these categories, respectively.
Among those pathogens considered secondary agents, G. parasuis and P. multocida were the most frequently detected, corresponding to 35.2% (55/156) and 34.0%(53/156) of the cases, respectively.Interestingly, no other pathogen was detected in 52.7% (29/55 pigs) and 43.4% (23/53 pigs) of the pigs with G. parasuis and P. multocida infections, respectively.In contrast, considering all 156 pigs, Salmonella spp. was isolated from 18.6% (29/156 pigs) of the pigs.Among these cases, 89.7% (26/29 pigs) were associated to other pathogens.Figure 1 demonstrates distribution of single and coinfections among SIV and the two most common secondary bacteria.PCV2 was detected in lymph nodes of 7.1% (11/156 pigs) of pigs with any pattern of pulmonary lesions.Among these cases, 90.9% (10/11 pigs) had PCV2 associated to other pathogens.There was one finishing pig that was PCR positive to PCV3 with granulomatous interstitial pneumonia containing multiple multinucleated giant cells as well as granulomatous nephritis and lymphadenitis.
Starch granules, characterized as foreign body, were observed in association to histiocytic inflammation, with epithelioid macrophages and multinucleated giant cells in six pigs (3.8% -6/156; Figure 11).In these association to different pathogens was diagnosed including G. parasuis, Streptococcus suis, Salmonella spp., P. multocida, and SIV.

DISCUSSION
In this study, the association of histopathology evaluation with molecular, immunohistochemical and/or bacteriologic analysis was a crucial tool for an accurate diagnosis of porcine respiratory disease complex.A relevant infectious agent was identified in more than 95% of the cases of pigs with clinical respiratory disease in this study, which indicates that appropriate laboratorial diagnostic approach, based on histopathology as a screening method, is efficient for establishing an etiologic diagnosis.According to Caswell and Williams (2016), increasing number of auxiliary tests available does not decrease the importance of morphological analysis even if the principal agent involved in death is had been already identified, mainly because underlying causes are also important and need to be investigated as well.Importantly, coinfections or superinfections are very common in cases respiratory diseases of pigs (Saade et al., 2020).Therefore, this study represents a contribution since it is important to recognize morphological patterns associated with natural and often multiple infections, which contrasts with morphologic descriptions of experimentally induced lesions by a single pathogen under controlled conditions and specific time-points post infection (Müller et al., 2003;Takemae et al., 2018).In addition, under experimental conditions animals are challenged with one specific strain of the pathogens, which may not represent the variability among field strains (Oliveira Filho et al., 2018).
SIV infection is an important cause of respiratory disease in pigs.In this study, 29% (39/156) of the pigs were carrying SIV.However, among those pigs 59% (23/39 pigs) had no bronchiolar lesions, which possibly represent animals with early SIV infection, since necrosis and neutrophilic intraepithelial infiltrate in bronchioli is observed beginning only at 24 to 48 hours post infection with SIV (Janke, 2014).In these cases, molecular analysis is an important tool to detect the virus.Interestingly, ten animals with bronchiolar epithelial hyperplasia and necrotizing bronchiolitis were SIV-negative.Importantly, SIV-induced lesions may remain after viral clearance (Janke, 2014) so histopathology is very useful tool for a proper differential diagnosis in those cases.
The most frequently isolated bacterial pathogens in this study were G. parasuis and P. multocida.G. parasuis is a common pathogen in the technified swine industry in all pig ages (Oliveira and Pijoan, 2014) and bronchopneumonia with fibrinous pleuritis is the expected histopathological pattern in those cases (Caswell and Williams, 2016;Vahle et al., 1995).The large number of pigs with respiratory disease associated to P. multocida infection in this study is in good agreement with a previous study in Brazil (Paladino et al., 2017).In these cases of pulmonary damage by P. multocida bronchopneumonia was the most common histopathological pattern, which has also been previously described (Pors et al., 2011(Pors et al., , 2013;;Caswell and Williams, 2016).Although considered a secondary pathogen (Brockmeier et al., 2002;Opriessnig et al., 2011), P. multocida were isolated in eight cases of necrotizing pulmonary lesions in this study.One possibility is that those pigs were infected with a recently described highly pathogenic P. multocida (Oliveira Filho et al., 2018), which may cause necrotizing lesions as a single agent (Paladino et al., 2017), resulting in similar histopathological patterns of parenchyma necrosis as observed in this study.An important differential in these cases is A. pleuropneumoniae infection.
Mycoplasma hyopneumoniae is a pathogen associated with the condition known as enzootic pneumonia, although very often this organism is associated with other pathogens in the lung (Hillen et al., 2014).Microscopically, BALT hyperplasia is an important morphologic marker of M. hyopneumoniae infection (Hillen et al., 2014;Know et al., 2002).However, only two out of 11 pigs positive for M. hyopneumoniae in this study had BALT hyperplasia, which is in good agreement with previous reports indicating that the diagnosis and interpretation of M. hyopneumoniae infection may be quite challenging (Buddle and O'Hara, 2005).In some cases in this study, the infectious agent detected by different diagnostic tests was associated with lesions that did not quite are expected for those agents, as occurred in some cases of pigs infected with A. pleuropneumoniae and M. hyopneumoniae.Importantly, the expected patterns of lesions are largely based on studies based on experimental infections, which contrasts with naturally occurring disease that often is a result of coinfections as demonstrated in this study.
PCV2 is a common and very important cause of multisystemic disease in pigs, which is often associated with pulmonary lesions (Oppressing et al., 2020).Importantly, PCV3 was detected in one case in this study, associated with granulomatous pneumonia, a presentation similar to PCV2-induced systemic disease.PCV3 is a recently described virus, associated with multiple lesions, particularly the condition known as porcine dermatitis and nephropathy syndrome (Arruda et al., 2019;Jiang et al., 2019).PCV3 has been previously diagnosed in Brazil (Tochetto et al., 2017).
T. pyogenes is not a specific agent of porcine respiratory disease complex, being considered an opportunistic bacterium, part of respiratory, skin and reproductive microbiota.However, T. pyogenes is an emerging pathogen with potential to cause economic losses for the swine industry (Jarosz et al., 2014).
Animals with pulmonary interstitial reaction and septicemic salmonellosis were frequently observed in this study, a situation which histopathology may also be useful to confirm systemic and pulmonary lesions of salmonellosis.
Finally, cases of lesions attributable to starch aspiration were also identified in this study.Similarly to previous reports, here we found a bronchopneumonia with epithelioid or multinucleated cells in alveolar lumen, containing negative cytoplasmic particles morphologically compatible with starch (Corner and Jericho, 1972).Starch-associated pneumonia was considered in this study in spite of not being associated with any specific infectious agent, but because it is a relevant differential diagnosis for infectious agents.As a matter of fact, it was considered an incidental finding in pigs affected by different pathogens in this study.
In conclusion, association of histopathology and traditional microbiologic methods, along with other ancillary diagnostic methods, particularly immunohistochemistry and PCR, are powerful tools for the differential and etiologic diagnosis of porcine respiratory disease complex, which is highly desirable since an accurate diagnosis in these cases is critical for implementing appropriate control and mitigation protocols.

Figure 1 .
Figure 1.Venn diagram demonstrating cases of single or co-infections of swine influenza virus (SIV) and the two most common secondary bacterial agents, Glaesserella parasuis and Pasteurella multocida.