Keywords

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

1 Introduction

As the worst global human pandemic climaxed during the summer of 1918, a new disease entity began to be recognized in swine in the Midwestern United States. The clinical signs of this new disease readily differentiated it from classical swine fever (hog cholera), which was the infectious disease of most consequence for swine at that time. This new respiratory disease was tabbed “hog flu” because of the similarity of the disease to influenza in humans (Dorset et al. 1922).

An early description paints a memorable clinical picture: “The onset of hog flu, as already stated, is sudden, an entire herd coming down, as a rule, within a day or two…The first symptom noted is loss of appetite, the animals failing to come up for their feed. They are disinclined to move and lie around the straw stacks or in their houses. When temperatures are taken, the animals are found to have fever. A thumpy or jerky respiration soon develops which is best observed when the animals are lying down and at first may be so slight as to escape notice unless the animals are carefully watched; later, it becomes more pronounced and may be noted when animals are standing. The disease evidently has a very short incubation period and develops rapidly. The second or third day, the entire herd, as a rule, will be lying in their nests and often present a very sick appearance. Sometimes one may walk among the sick animals and even step over them without rousing them, and anyone viewing for the first time a herd suffering from hog flu at the height of the infection would probably think that most of the affected animals would succumb. When the sick animals are roused from their nests, they almost invariably cough. The cough is paroxysmal in character, the back being often arched, and the spells of coughing are sometimes of sufficient violence to induce vomiting; in this respect the disease resembles whooping-cough in the human. When the paroxysms of coughing have passed, the animals stand in a listless attitude with their heads down, their tails limp, and soon lie down as though tired. The sick animals usually rest on their bellies, and sometimes assume a partly sitting position with the body propped on the forelegs, as if to afford room for greater lung expansion. There is usually a conjunctivitis, characterized by a watery or gummy secretion from the eyes, and a nasal discharge may also be present……” (McBryde 1927).

2 Clinical Disease

2.1 Classical Epidemic Swine Influenza

Epidemic swine influenza as described above was the predominant presentation of the disease in the United States for nearly 70 years. Indeed, this classical virus and presentation of swine influenza in conventional swine populations continues today as an acute, high morbidity-low mortality infection that spreads rapidly through groups of pigs. For the first 1–2 days after infection, affected pigs develop high fevers (>105°F, 40.5 °C) with lethargy and anorexia. Close examination reveals clear nasal discharge and conjunctivitis. Tachypnea and expiratory dyspnea (thumping) are often pronounced, especially when pigs are forced to move. By days 3 and 4, pigs begin to acquire the hallmark clinical sign of this disease, a harsh deep barking cough that results from the extensive bronchitis and bronchiolitis. In many pigs, fevers will have begun to drop by this time. Pigs of all ages, from mature sows to nursery pigs, may be similarly affected, but clinical disease is often milder in nursing pigs. In some outbreaks, sows may be inappetent and lethargic and develop high fevers but have less prominent respiratory clinical signs. Pregnant animals may abort. In the absence of concurrent bacterial pneumonia, individual pigs recover quickly, usually within 6–7 days. Not all pigs in a group will be infected simultaneously, and the disease course for the entire group may require a 2+ weeks before clinical signs abate and pigs return to normal body condition and weight gain. Mortality is generally low although some virus strains exact a higher toll.

Historically, there has always been a distinct seasonality to swine influenza in the north central U. S., a part of the country that has widely divergent seasonal temperatures. In years past, outbreaks commonly occurred each fall, with little recognition of the disease at other times of year. Even under current confinement production systems, the disease still exhibits a consistent seasonality, with the greatest peak in the fall and a smaller peak in the spring, the seasons of transition for prevailing weather conditions (Janke et al. 2000). Wide swings in temperature over short periods of time make it difficult to modulate housing environments, and these climatic stresses may increase animal susceptibility to infection. Cool moist conditions contribute to environmental survival and aerosol spread of the virus. How and where the virus survived/survives between outbreaks has never been fully explained. A long-term carrier state has not been discovered, and most individual pigs appear to clear the virus within 2 weeks. The virus is likely maintained in herds by subclinical passage to naïve pigs or those with low or compromised immunity. Studies on vaccine efficacy have indicated that there is no absolute immunity threshold for swine influenza virus (SIV) infection. Pigs with sufficient immunity to prevent clinical illness can still be infected and shed virus, though it is much reduced in duration and titer (Richt et al. 2006; Van Reeth et al. 2001).

2.2 Current Clinical Expression in Larger Swine Populations in Segregated Rearing Production Systems

Many pigs today are raised in segregated rearing systems . Sows are maintained, bred, and farrowed at a location separate from the farms on which younger pigs are fed to market weight. Once farrowed, pigs are only kept with the sows for about 3 weeks before they are weaned and moved to nurseries on another site (segregated early weaning). In two-site systems, pigs will be raised at that location until being sent to market. In three-site systems, pigs will be fed at the nursery site until about 10–12 weeks of age, after which they are moved to grow-finish units at a third location. At each location, the goal is to fill and empty each building completely at one time (all-in all-out). The objective is disease control, i.e., to minimize situations in which infections can be passed from older immune pigs on a site to younger naïve pigs as they are brought into the same buildings. In continuous flow systems, viruses and bacteria have greater opportunity to maintain contagious levels because of the periodic addition of susceptible hosts. This age-based segregation and movement of pigs must be considered when considering the clinical presentation of swine influenza as it currently manifests itself. In nearly all swine production units, one can find influenza virus in circulation or serologic evidence of previous exposure.

At the present time, because of the almost universal immunity against SIV at some level against some variant in all herds, and the number of subtypes, antigenic clusters within subtypes and reassortants that circulate within swine populations, expression varies widely. Outbreaks still occur but infections are more frequently endemic with clinical expression more muted and melded with that of other concurrent respiratory infections. Pigs of all ages from nursing pigs to sows may be infected, with blips of clinical disease appearing at different phases of production that vary with the situation. It becomes the task of veterinary practitioners and diagnosticians to determine whether increased clinical disease is due to the resurgence of a virus already circulating within the operation or to the introduction of a different virus against which only partial immunity may exist.

Expression of disease generally falls into one of the following presentations:

  1. 1.

    Acute, “fulminating” SIV—Resembling disease as it was first observed 90 years ago but now the least frequent presentation, this is the outbreak of a pathogenic variant in a population that has little or no specific antibody to attenuate the infection. There is very rapid spread, high fevers, anorexia, expiratory dyspnea with effort, mortality, and a brief course in the herd/group. Mortality occurs in pigs whose lungs fill rapidly with fluid; a typical ‘foam cast’ forms in the bronchial tree and is expelled through the nose. There is very rapid spread within and between sites and through pigs of all ages. Even nursing pigs are severely affected, but essentially all survive. Recovery is rapid and often complete, with very few secondarily infected or chronically affected pigs, and infection is followed by very high levels of specific immunity. This is rarely confounded with other agents even though they may be present.

  2. 2.

    Age-associated influenza in growing pigs—Infection can be relatively predictable in certain systems depending on how they are structured. Maternal passively acquired antibody is protective for weaned pigs against endemic virus, but that immunity wanes with maternal antibody decay. Usually occurring when sow herd immunity is uniform, and pigs have roughly equivalent levels of passive protection, infection that is clinically evident is delayed until the pigs are exposed to homologous or variant virus that can ‘break through’ the collective passive immunity. The classic pattern is robust pigs with no problems through the nursery phase until they are 10–12 weeks old and in finishers where they are exposed to other older growing pigs that are shedding virus. Approximately 2 weeks into the finishing phase, pigs develop cough and depression that does not explode, but rather works its way through the group over a protracted 2–3 week timeline as the individual passive antibody decay curves meet up with the various loads of virus that overcome it. The actual SIV-induced clinical disease expressed is usually not dramatic, of variable severity and commonly complicated by concurrent infections endemic in the group.

  3. 3.

    Piglet influenza—A consistent/persistent clinical expression of SIV can be found in nursing to newly weaned pigs. Again associated with passive protection variance within groups and increasing frequency of SIV variants, clinically the infection is first appreciated as a cough in the 2–3-week-old pig about to be weaned. But the cough is in scattered pigs in a given room (maybe 10 % of litters have a pig that sporadically coughs) that are hard to find and almost too subtle to raise any alarm. When these pigs are moved to the nursery, the stress and activity exacerbate clinical expression. During the move and on first entry into the nursery, there is a common description “…newly weaned pigs came off the truck coughing…” even though the farrowing management would say “…no, we didn’t notice any cough; maybe a pig or two…”. Clinically in this situation, a subset of pigs suffers fever, anorexia, and after a couple days, cough due to SIV infection. Again, the severity of clinical disease is greatly dependent on the level and specificity of passive protection. Affected pigs are anorexic during the critical transition phase to solid food, i.e., they are sick for 36 to 48 h after the move, and become very hard to start on feed. When just a few pigs in a large group are affected, the infection probably goes unrecognized, but on many farms, a consistent 5–10 % or greater are affected and the problem is economically substantial. These situations are the purpose of sow immunization, not to protect the sow, but to try to extend the passive shield until pigs are well-started on feed and can handle the infection.

2.3 Clinical Disease in Experimentally Infected Pigs

The clinical disease induced by experimental challenge rarely reaches the severity of that observed in the field (Landolt et al. 2003; De Vleeschauwer et al. 2009). Clinical signs of illness (fever, depression, anorexia, tachypnea, serous nasal, and ocular discharge) develop to some degree after most challenges but vary under the multitude of protocols employed (Nayak et al. 1965; Winkler and Cheville 1986; Brown et al. 1993; Van Reeth et al. 1996; Thacker et al. 2001; Richt et al. 2003; Landolt et al. 2003; Jung et al. 2005; Vincent et al. 2006; De Vleeschauwer et al. 2009; Sreta et al. 2009). Under most experimental models, tachypnea is evident in pigs at least when aroused, but coughing is usually minimal, limited to an occasional soft cough that develops a day or two after the onset of other clinical signs. Fever is quite variable as is the onset of clinical signs. In general, with higher virus titers in inoculum (>106–107 TCID50, EID50 or PFU/pig) and intratracheal inoculation, the onset of illness occurs sooner [24–36 h postinoculation (PI)] and clinical signs are more apparent. With lesser challenge regimens (103–105 TCID50, EID50, PFU/pig) and intranasal inoculation, clinical disease may not be seen or may take 2–4 days to become evident. Duration of clinical illness is usually 2–4 days with most experimental infections.

3 Virus Infection, Replication, and Shedding

SIVs are spread between pigs through direct contact via nasal secretions and through inhalation of aerosolized virus in droplets generated by coughing and exhalation. The cellular targets of infection are the epithelial cells lining the nasal passages, trachea, bronchi, bronchioles, and alveoli. Numerous experimental infection studies have been conducted to define the progression of the infectious process.

A variety of inoculation methods have been employed in these experimental infection studies to administer virus: nebulization via nose cone or chamber, direct intranasal (IN) inoculation, and intratracheal (IT) inoculation. Nebulization is probably the most effective method of depositing large quantities of virus throughout the respiratory tract but this method is also the most labor intensive and is little used. Intranasal methods (the mechanical specifics of which are often not described) appear to be more variable in their efficacy of establishing infection, probably because with some techniques pigs swallow most of the challenge dose. Intratracheal methods provide the most consistent challenges, and if not deposited too far down the tract, the virus appears to be well-distributed throughout the lungs.

Infection and multiplication in host cells progresses very rapidly with influenza viruses. In an immunofluorescent study (Nayak et al. 1965), the first evidence of virus infection was a pale fluorescence in the nucleus of bronchial epithelial cells, as early as 2 h PI. By 4 h PI, virus antigen was abundant in both nucleus and cytoplasm of infected cells. In an ultrastructural study (Winkler and Cheville 1986), virus was observed budding from the surface of Type II pneumocytes as early as 5 h PI.

Only low numbers of randomly scattered cells are observed in the nasal turbinates and trachea during the first 24–72 h PI (Nayak et al. 1965). The most extensive infection occurs in epithelial cells lining the bronchi and bronchioles, with peak infection occurring at 48–72 h PI. Although some virus reaches the alveolar level early, especially with nebulization or high dosages with other methods, more extensive spread of virus to alveolar epithelial cells tends to occurs later in the course of infection, at 72–96 h PI (Jung et al. 2005; Van Reeth and Pensaert 1994).

Multiple studies (Brown et al. 1993; Van Reeth et al. 1996; Landolt et al. 2003; Vincent et al. 2006, 2009a, b; De Vleeschauwer et al. 2009; Ma et al. 2010) report nasal virus shedding by 1–3 days PI, regardless of the route of inoculation, and the duration of shedding, for 4–5 days, occasionally to 7 days PI. In one study, intranasal inoculation (107 EID50) resulted in virus shedding by 24 h PI, whereas intratracheal inoculation resulted in nasal shedding being delayed until 72 h PI and the peak titer of virus shed was much lower (De Vleeschauwer et al. 2009). In another study using intranasal inoculation, lower virus titers (103–104 TCID50) resulted in 24 h delay in onset and peak of nasal shedding compared to pigs given 105–106 TCID50 although peak titers of shed virus were similar (107 TCID50/ml) (Landolt et al. 2003). The amount of virus shed in nasal secretions tended to be fairly consistent through days 2–4 PI with peak titers described in the range of 103.5–107.5 TCID50, EID50, or PFU/ml of fluid used to flush nasal passages or to wash virus from nasal swabs.

Determination of virus titers in lung homogenate or bronchoalveolar lavage fluids is often used in experimental studies to monitor the dynamics of virus production in the lung. Peak virus load in the lung, as measured in studies that cover the first few days of infection, occurs at about day 3 PI with titers varying from 104.5 to 108.3 TCID50, EID50 or PFU/ml. Titers hold at relatively similar levels through 5 days PI. (Van Reeth et al. 1996; De Vleeschauwer et al. 2009; Vincent et al. 2009a, b; Ma et al. 2010).

The narrow time frame of virus replication and shedding described above is consistent in most challenge trials for most influenza viruses isolated from swine regardless of subtype. Experimentally, the dose of virus that reaches the lung initially may affect the course and severity of experimental infection. If low doses are given, either intranasally or intratracheally, the virus may initially spread more slowly which may result in a delayed onset and ultimately milder course of clinical disease. With most viruses, the course of infection is short and essentially complete within 5–7 days. The comparative effects of virus titer in inoculum, route of inoculation and/or age on the dynamics of infection have been described in several studies (Landolt et al. 2003; Richt et al. 2003; De Vleeschauwer et al. 2009).

In swine, influenza virus infection is generally considered to be limited to the respiratory tract, but a few studies have reported virus in extra-respiratory sites. A few infected cells were detected by immunohistochemistry in mediastinal lymph nodes, but none were detected in tonsil (Nayak et al. 1965). Influenza virus was isolated from the serum of all five inoculated pigs, for only one day each, at 1–3 days PI (Brown et al. 1993). In a more recent study, virus was detected by RT-PCR in spleen, ileum, and colon but not by virus isolation. Virus was detected in brainstem by both PCR assay and virus isolation, but no specific infected cells were detected by IHC (De Vleeschauwer et al. 2009). In this paper, researchers referred to unpublished data from in vitro studies indicating this virus could infect porcine trigeminal ganglion via the axons.

4 Pathology

4.1 Macroscopic Lesions

The most common macroscopic manifestation of influenza virus infection is a cranioventral bronchopneumonia that can affect a variable amount of the lung. The lesions are similar in both experimental inoculation and uncomplicated natural infections (Janke 1998) (Fig. 1a, b). Such expression would be expected since the virus enters the lung via the airways rather than through viremia. In milder infections, dark red multilobular to coalescing, often somewhat linear, foci of consolidation are evident in the hilar area and more dorsal portions of the cranial and middle lung lobes. More extensive infections involve larger, usually more ventral, portions of the cranial and middle lobes and cranioventral portions of the caudal lobe; as much as 40 % of the total lung volume may be affected. In field cases, the lesions often involve concurrent bacterial bronchopneumonia which results in more extensive lesions. In an occasional pig, a few hemorrhagic emphysematous bullae distending interlobular spaces may be evident. Tracheobronchial lymph nodes are variably swollen and congested. The trachea and nasal turbinates may be congested but are usually unremarkable. Although virus infects the epithelial lining of these upper airways, grossly visible necrosis does not develop.

Fig. 1
figure 1

a Swine influenza in a grow-finish pig (field case). Lobular and sublobular consolidation affecting a large portion of cranioventral lung. b Lungs from a 6-week-old pig experimentally inoculated with H3N2 SIV and euthanized 5 days postinoculation. Multifocal to coalescing consolidation in cranioventral portions of lung

Full size image

Less frequently encountered in field situations, and not reproduced by experimental challenge, severe acute influenza infections may result in a diffusely congested and edematous lung with abundant foam in the trachea and larger airways (Janke 1998). In such an acutely affected lung, cranioventral lobular consolidation may be obscured by the diffuse inflammation.

In experimental studies (Winkler and Cheville 1986; Van Reeth et al. 1996; Thacker et al. 2001; Richt et al. 2003; Landolt et al. 2003; Vincent et al. 2006, 2009a, b; De Vleeschauwer et al. 2009; Sreta et al. 2009; Ma et al. 2010) the extent of lung involvement also is quite variable and usually is expressed as percent of total lung affected, calculated either by addition of the portions of each lobe (Halbur et al. 1995) or as the average for all lobes, with described values ranging from <1 to 58 %. As with clinical signs and virus distribution and shedding, lesion severity is influenced by the route of inoculation and the virus titer in the inoculum. An inoculum containing 106 or higher infectious doses of virus introduced intratracheally will often result in 10–30 % lung involvement. Inocula with 103–104 TCID50 or EID50 of virus, especially if administered intranasally, may result in <10 % lung involvement. In some experimental infections, the dorsocaudal aspect of the caudal lobe also is affected, most likely an artifact of the method of inoculation as this presentation is unusual in field cases.

4.2 Microscopic Lesions

The two most detailed microscopic descriptions of the effects of SIV infection on swine respiratory tract are a histopathologic and immunofluorescent study (Nayak et al. 1965) and an electron microscopic (ultrastructural) study (Winkler and Cheville 1986), both of which are experimental challenge studies with classic H1N1 virus. Less comprehensive but similar descriptions of microscopic lesions induced by SIV infection, some with concurrent immunohistochemical (IHC) studies describing virus distribution in tissues, have been reported by numerous researchers in studies characterizing isolates of interest (Brown et al. 1993; Van Reeth et al. 1996; Landolt et al. 2003; Jung et al. 2005; De Vleeschauwer et al. 2009; Sreta et al. 2009). Additional descriptions of the effect of SIV infection in pigs can also be found in many other studies, often in comparison to human and avian viruses inoculated into swine or in vaccine trials. The descriptions below are composites drawn from these studies as well as the author’s experience with both field cases (Janke 1998) and experimental trials (Thacker et al. 2001; Richt et al. 2003, 2006; Solorzano et al. 2005; Vincent et al. 2006, 2007, 2008, 2009a, b; Kitikoon et al. 2009; Ma et al. 2010). The hallmark microscopic lesion of influenza infection, consistently present, is necrotizing bronchitis and bronchiolitis (Fig. 2a). Interstitial pneumonia, though usually evident to some degree, is quite variable in severity in both field cases and in experimental trials, often with pig-to-pig variation.

Fig. 2
figure 2

a Subacute necrotizing bronchiolitis in the lung of a 6-week-old pig inoculated intratracheally with H3N2 SIV and euthanized at 3 days postinoculation. Extensive necrosis and sloughing of epithelial cells from a segmental bronchiole is evident. HE x40. b Immunohistochemical staining of a similar bronchiole from the same pig identifying virus-infected epithelial cells sloughing into the lumen. IHC x40. c Immunohistochemical staining of alveoli from the lung of a 6-week-old pig inoculated with H1N1 SIV by nebulization and euthanized at 24 h postinoculation. Virus has penetrated deep into the lung and numerous pneumocytes lining alveolar walls are infected. IHC x40

Full size image

The earliest response to infection is neutrophil infiltration. By 4–8 h PI, neutrophils are emigrating through airway epithelial layers and accumulating in the lumens of alveolar capillaries. Endothelial cells lining the capillaries are swollen and pavementing of vessel walls by marginated neutrophils is evident. Alveolar walls are widened by vascular congestion and lymphatic dilation. Although ballooning degeneration and cytoplasmic vacuolization of some epithelial cells lining smaller bronchioles may be recognized as early as 8–16 h, these changes are subtle and scattered, and most airways are still intact.

By 24 h PI, extensive infection of epithelial cells lining scattered airways of variable size can be detected by IHC. In a few studies, airway epithelial necrosis was described at this time, but in most studies, disruption of the epithelial layer in a significant number of airways has not yet occurred. Small numbers of neutrophils may be clustered in some airway lumens, accompanied by light infiltration of lymphocytes around some bronchioles. Alveolar walls may be more prominently thickened by congestion, edema, and leukocyte infiltration, predominantly peribronchiolar in distribution.

By 48 h, there is extensive necrosis and sloughing of epithelial cells into airway lumens accompanied by more obvious neutrophil accumulation. Loose infiltration of lymphocytes around affected airways is more prominent but still light. The epithelial cells remaining attached are swollen or attenuated and the layer is irregular in outline. Thickening of alveolar walls, if prominent, is more diffuse. Pneumocytes lining alveoli may be swollen with some sloughing into the lumen. Numerous epithelial cells in affected airways (Fig. 2b), both attached and sloughed, contain virus antigen by IHC, but only a few individual to small clusters of infected cells will be observed in alveoli. Some of these cells are obviously swollen pneumocytes still attached to alveolar walls or sloughed into the lumen (Fig. 2c), but other cells within the alveolar wall or loose in the alveolar lumen appear to be macrophages. By close examination, necrotic constituent cells may be identified in alveolar walls, but the septa remain intact. In occasional severe infections, clumps of necrotic cell debris are evident in clusters of alveoli. Consistent with the macroscopic appearance, lobules within the same section of lung may differ in the degree of involvement. Severely affected lobules are frequently atelectatic.

In both field cases and experimental challenge studies, the sizes of airways that are affected may vary. Most likely due to the dynamics of airflow that affect droplet suspension and deposition or to receptor distribution, the largest airways without cartilage (primary bronchioles) are most consistently and most severely affected (Thacker et al. 2001). In some animals, both experimentally and in field cases, the larger lobar or segmental bronchi may be spared. Conversely, in some mild infections, only these larger airways may be infected. The smallest terminal or respiratory bronchioles may be spared or may be necrotic, and the degree of alveolar involvement varies. Except in severe cases, the lesions are multifocal and unaffected lobules sit adjacent to severely affected lobules.

By 72 h PI, some airways are in active necrosis and filled with debris, but many airways are lined by an intact hyperplastic epithelial layer. Peribronchiolar lymphocytic cuffs are well-developed. Alveolar walls may still be thickened as described above with a light loose mixed population of sloughed pneumocytes, macrophages and leukocytes residing in alveolar lumens. Leukocyte populations have shifted to predominantly mononuclear cells. By this time, few infected airway epithelial cells will be identified by IHC but more numerous scattered infected cells, often limited to certain lobules, will be detected in alveoli. In some infections, alveoli may be little affected.

By 96 h and beyond, airways are in repair, lined by hyperplastic or nearly normal epithelium and surrounded by moderate-sized lymphocytic cuffs. Alveolar inflammation is also resolving. By this time, very little virus can usually be detected by IHC, in occasional isolated airways or in scattered individual cells in alveoli. In field cases, in some severely damaged bronchioles, repair is accompanied by fibrosis and endobronchial polyp formation (bronchiolitis obliterans). Such lesions are rarely observed in experimental infections. Over the following days, the epithelial hyperplasia resolves and peribronchiolar lymphocytic cuffing and partially atelectatic alveoli with variable leukocyte populations are all that remain. Though somewhat dependent on the extent of damage, lungs return to normal by 2 weeks PI.

In trachea, infected epithelial cells, as identified by IHC, are usually few in number and widely scattered. Damage to the tracheal epithelial lining, as characterized by attenuation or squamous metaplasia, if present at all, tends to be focal to multifocal. Very few viruses induce extensive epithelial injury, and even then, not consistently in all pigs. Subepithelial lymphocyte infiltration may be intense in the latter situations, but in most tracheas with minimal or focal epithelial attenuation, inflammation exhibits the same range of variation in severity as that observed in pigs not challenged with SIV. Although pigs shed quantities of virus over multiple days in nasal secretions, only mild attenuation of the epithelium lining the inside of nasal turbinate scrolls is observed and that inconsistently. Only low numbers of infected cells are usually identifiable in histopathologic sections by IHC. Infection of tonsil and tracheobronchial lymph node has been reported in some studies described above. In the author’s experience, low numbers of infected cells can be identified by IHC in the lymph nodes but rarely in tonsil.

This rapid sequence of events will occur in both the individually affected lobules in most naturally infected pigs and in pigs experimentally infected with high doses of virus. Under experimental challenge conditions in which less virus may be given, examination of multiple sections of lung may reveal asynchronous infection with different lobules at different stages of infection.

5 The Question of Virulence

Although some outbreaks of epidemic swine influenza suggest that certain viruses may be more pathogenic or virulent than others, defining or discovering the basis for virulence has proven to be difficult. Even viruses from severe field outbreaks have tended to be rather tame in captivity (Ma et al. 2010). Most comparative studies conducted in tandem under the same protocols have revealed only minor differences between viruses. More virulent viruses have been deemed so, not so much on increased severity of clinical disease, but on higher or more prolonged fever, higher or prolonged virus shedding or titers in lung, and more extensive macroscopic lesions. Microscopically, lesions tend to be similar but with more lobules and more airways within lobules affected.

Comparative evaluation of results of studies from different researchers must be done cautiously because of the multitude of factors (virus titer, route, and method of inoculation, age of pigs, etc.) that can affect results. Thus, the most significant observations in this regard are likely to come from researchers who have used similar protocols in multiple animal trials with many different viruses. Clinical differences are usually subtle but analysis of parameters that can be quantified can yield clues to virulence differences, e.g. fever, cytokine levels, virus shedding, virus titers in lung tissue or BALF, gross and microscopic lesion scoring. (See Landolt et al. 2003; Richt et al. 2003 for examples of scoring). Although differences may be minimal or below statistical significance because of the number of animals that researchers can reasonably afford to use in such trials, the various parameters often correlate well. In the field situation, such differences would likely be amplified. Some studies suggest the newer triple reassortant viruses may be more virulent than preceding classic viruses (Vincent et al. 2006; Ma et al. 2010). Several recent triple reassortant viruses recovered from situations in which both pigs and people were infected also appear to be slightly more virulent in swine from which they originated (Vincent et al. 2009b).

Most of the research into virulence factors in influenza viruses has been initiated with two unquestionably virulent viruses: H5N1 highly pathogenic avian influenza virus and reconstructed 1918 human pandemic virus (Tumpey et al. 2005; de Wit and Fouchier 2008; de Wit et al. 2008; Basler and Aguilar 2008; Lycett et al. 2009; Janke et al. 2010). The reverse genetics techniques now available allow researchers to replace specific genes or gene sequences in influenza viruses to compare the relative contribution of each gene to virulence or infectivity. Pigs are one of the models (mice, ferrets, chickens, primates) used for such studies, and information gained from such studies may eventually benefit understanding and control of SIV infection (Memoli et al. 2009; Solorzano et al. 2005; Richt et al. 2006).