Keeping them out of the rough: Practical Insights into Hemorrhagic Bowel Syndrome

Hemorrhagic bowel syndrome (HBS) is newly-described disorder primarily affecting dairy cattle. Its cause is not known. In this study, the possibilities that Aspergillus fumigatus and Clostridium were involved in etiology of HBS were examined. Samples of feed, gastrointestinal (GI) contents, GI wall, lymph nodes and blood from HBS and control animals were collected in four states (IA, ID, OR, WA). Real-time SybrGreen quantitative polymerase-chain reaction (PCR) analysis indicated that all HBS cows were infected with A. fumigatus. Samples from control cows were negative. Multiplex PCR analysis of five clostridial toxin genes did not reveal a correlation with HBS. Specifically, clostridial toxin genes were detected in both HBS and control animals. A. fumigatus correlates closely with HBS and may play an important role in its etiology. Introduction. In 1991, Anderson reported a new disease in Idaho dairy herds (1). The syndrome, which he termed “Point Source Hemorrhage”, was observed in five high-producing Holstein cows from one dairy. Symptoms included point-source sub-mucosal hematomas, each affecting 10-20 cm of the jejunum. One of the five cows exhibited a ruptured hematoma with exsanguination into the lumen of the jejunum. The origin of the hematoma was traced to the jejunal submucosa which dissected mucosa from underlying connective tissue. Despite the hemorrhage, clotting time was normal. No bacteriological assays proved definitive and no ulcerative processes, parasitism or vaculitis were apparent. Feeding and management practices on the dairy were “exemplary”. Incidence. Since 1991, awareness of “Point Source Hemorrhage” (now more commonly known as “Hemorrhagic Bowel Syndrome (HBS)”), in dairy cattle has grown. Dennison et al (2) published a survey of HBS in 22 dairy cows in 2002. Dr. Bruce Anderson, in his practice with the University of Idaho in Caldwell, typically conducts necropsies on 10-20 cases per year and, in one dairy, has noted 35 cases since 1985. HBS incidence is increasing (3,4) and is responsible for 2% of the deaths of dairy animals in the US (5). Additional estimates of incidence may be unavailable because there is a marked seasonality to the disease (more cases occur in cooler winter months), many veterinarians, dairy producers and nutritionists are unfamiliar with the disease, symptoms mimic common ruminant digestive diseases and an unknown proportion of afflicted cattle are submitted for necropsy. Signalment and etiology. HBS is characterized by sudden drop in milk production, abdominal pain due to obstructed bowel and anemia (6). Death comes within 48 hours from the onset of the obstructing blood clot plug. Fatal factors are presumed to be the anemia combined with digesta stagnation in much of the severely dilated small intestine proximal to the plug (6). What is the cause of HBS? Dennison et al (2) conducted a retrospective analysis of all dairy cows examined at Colorado State University Veterinary Teaching Hospital which were submitted with dysentery, melena or colic. Cows with hemorrhagic enteritis were considered to have HBS if they displayed melena and clotted blood in feces or small intestine and displayed no evidence of intestinal or extraintestinal lesion. Of the 22 HBS cows, age ranged from 2-8 years and incidence occurred at a mean of 107.5 days post-parturition. Average milk production was 89.8 lbs. Two-thirds (14 of 22) of the cases arrived in the cooler months (September through February). Lethargy was noted in 21 of 22 cows. Abdominal distension was noted in 8 of 22. Nineteen of 22 were clinically dehydrated and 10 of 17 displayed ruminal hypomotility. Seven of 15 had bloody feces. Distended loops of the small intestine were noted in 7 of 14 cows and trans-abdominal ultrasonography revealed small intestinal ileus and distension in 12 of 12 cows. Seven of 8 cows treated medically died and nine of 13 cows treated with surgery died or were euthanized. Clostridium perfringens was isolated from fecal samples in 17 of 20 cows. Genotyping of the C. perfringens in 10 cows revealed Type A in five cows and Type A with the b2 toxin gene in the remaining five cows. The dairy from which the C. perfringens-positive cows had originated had vaccinated the cows with C. perfringens types C and D toxoid. Despite finding C. perfringens in most HBS cows, Dennison et al (2) commented that “it is unclear whether proliferation of C. perfringens is part of the primary disease process in cows with HBS or occurs as a secondary response.” Evidence against C. perfringens playing the primary etiologic role includes the observations that C. perfringens is ubiquitous (7,8). Furthermore, immunization against Clostridium spp. does not appear to protect animals from HBS. Many potential causes of HBS have been investigated and discarded. Agents which do not play an etiologic role include parasitism (1), BVD, coccidia, salmonella, coagulopathies, intestinal foreign bodies, physical obstructions and deformities including volvulus and intussuception (2). Furthermore, analyses of diets, ages of cow, levels of milk production and a full spectrum of blood chemistry and biochemical assays failed to reveal a consistent clinical correlate to HBS (2). An alternative etiology. Moldy feed was observed on several dairies which experienced cow losses due to HBS. In “human literature” Aspergillus fumigatus is described as pathogenic, causing intestinal bleeding and invasive apsergillosis in immunocompromised patients. We hypothesized that immunocompromised dairy cows might develop HBS when exposed to A. fumigatus –laden feed. Pathogenicity of Aspergillus fumigatus. Aspergillus is a large fungal genera containing > 100 species. Of these, A. fumigatus and flavus are the most pathogenic (9,10). Pathogenicity of Aspergillus is attributed to three virulence factors: 1) production of iron (Fe)-sequestering siderophores, 2) secretion of complementand phagocyticinhibitory lipids and 3) secretion of proteases (9,10). Specifically, A. fumigatus is able to meet its Fe requirement, and thereby maintain growth, by the actions of its proteolytic enzymes. These liberate the host’s Fe stores from transferrin and lactoferrin and allow Fe transfer to triacetylfusarimine and ferriciocin (10). In addition, invasiveness is facilitated by secretion of polar and neutral lipids, phenolic compounds and heterocyclic toxins (including aflatoxins and other toxins). Some of these inhibit phagocytosis while others suppress the immune response of the host by inhibiting complement factors C3a and C5a (10). Finally, pathogenic species of Aspergillus, like an invasive tumor, secrete proteases which facilitate hyphal penetration from a colonization site into the underlying parenchymal tissue (10). Aspergillus in the ruminant gut. Several studies have demonstrated potential for Aspergillus species to infect the ruminant gut at various sites and to cause enteric hemorrhage. Sheridan (11) reported aspergillosis in calf abomasum in 1981. In 1989, Jensen et al (8,12) reported that A. fumigatus infected the terminal gastric compartments, particularly the omasum. In 1991, Jensen et al (13) received a 4 year-old Jersey cow with suspected right-displaced abomasum (RDA) and acidosis. The RDA and acidosis were not confirmed by examination. The animal also did not respond to antibiotics or antiinflammatory drugs. The animal was euthanized and necropsy revealed hemorrhagic lesions in the reticulum, rumen, omasum, and small intestinal Peyer’s patches. Hemorrhagic necroses of the mesenteric lymph nodes, liver, kidney and lung were observed. Hyphal growth, thrombosed vessels and vasculitis were detected in all necrotic tissues. A. fumigatus and A. corymbifera were identified in all necrotic lesions. Jensen et al (13) surmised that the ruminant gut provided two portals for fungal invasion: intestinal Peyer’s patches and the pre-gastric digestive compartments and proposed that A. fumigatus was the primary invader. In more recent studies (14,15) Jensen et al evaluated predisposing factors for mycotic infections in ruminants. The most common mycoses included aspergillosis, candidosis and zygomycoses. Principle etiologic agents were A. fumigatus, Candida albicans and Absidia corymbifera, respectively. Mucor pusillus and Rhizopus spp. were also identified a common etiologic agents in zygomycoses (14). Portals for infection were identified in the respiratory and GI tracts. GI mycosis was identified with the omasum representing the main organ for infection. In one study (15), 32 of 694 cattle submitted for necropsy had gastrointestinal mycoses with an elevated incidence in cooler months (i.e., 75% incidence was in October through March, a time during which stored feed is fed to cattle). Hypomotility of the foregut was noted in 22 of 32 cattle, 23 of 29 cattle were post-partum, 22 of 29 had been given broad-spectrum antibiotics, 9 or 29 had been given antiinflammatory drugs, 26 of 29 displayed inappetance, 23 of 29 displayed diarrhea and 17 of 29 displayed fever. Aspergillus and zygomycetes were detected in gut wall vasculature with thromboses and vasculitis. Hematogenous spread of fungi to the liver, lung and kidney was detected. A. fumigatus was detected in 10 of 21 cows, A. corymbifera was detected in 8 of 23 cows and Candida was detected in 1 cow. Of interest, animals were never infected with more than one fungal species. Predisposing factors for mycotic infections included: 1) feeding of moldy feed, 2) immunocompromizing diseases, 3) acidosis, 4) antimicrobial therapy, 5) reflux of abomasal contents, 6) metabolic disturbances, 7) post-partum stress, 8) viral erosive diseases such as IBR, 9) antiinflammatory treatment, and 10) abortion. Asperillosis in immunocompromised humans. A. fumigatus is ubiquitous. Yet it rarely causes serious disease in healthy individuals. Immunoincompetence is the primary predisposing factor in Aspergillus infection (invasive aspergillosis; 16-19) in humans. Patients with AIDS, cancer and those receiving organ transplant are particularly susceptible to invasive aspergillosis (17-19). For example, invasive aspergillosis occurs in 2.6 – 10.3% of all bone marrow transplant patients and has a mortality rate of 56 to 88.1% (16). Immunosuppression in dairy cows. Mallard et al (20) have reported that immunosuppression is common in dairy cows and accounted for the high incidence of disease. Changes in both immune function and nonspecific host defense mechanisms have been reported in dairy cows at onset of lactation (21-25). Stressors in lactation include 1) a high energy diet (and potential acid reflux), 2) ketosis, 3) milk fever, 4) lameness, 5) regular handling, 6) post-partum stress, 7) potential poor feeding practices, (26), 8) social isolation when sick animals are placed in a “hospital pen” (27) and 9) artificial insemination (28). A stressed, immunocompromised dairy animal is susceptible to mycotic infection. The provision of A. fumigatus-infected feed to this animal may be a “trigger” which elicits HBS. Feed-borne Aspergillus fumigatus infects the GI tract, tissues and blood of ruminants. Results of A. fumigatus genomic analysis from eight HBS cows, one abomasal hemorrhage (AH) cow and one AH gazelle are shown in Table 1. A. fumigatus was detected in 3 of 3 (3/3) feeds submitted for analysis. It was also detected in the gut contents (7 of 7 cases), gut wall (5 of 5 cases), and mesenteric lymph node (3 of 5 cases). Invasive aspergillosis was indicated by detection of A. fumigatus DNA in blood (6 of 6 cases) and in liver (1 of 1 case). One case of HBS was associated with a late-term abortion. Cotyledons from the case were sampled and also found to harbor A. fumigatus. Of interest, A. fumigatus DNA was also detected in two cases of abomasal hemorrhage (AH), one in a dairy cow and another in a Dama gazelle (Table 1). Several negative control cows (i.e., non-HBS; n=17) have also been processed and, of these, 14 have tested negative for A. fumigatus (data not shown). The remaining cows (n=3) contained very low levels of A. fumigatus DNA; near the detection limit of our assay (i.e., < 0.02 X 10 A. fumigatus genomic units/ml of blood). These levels were 1/20 to 1/50,000 of the levels detected in HBS cows. Two of the negative control cows (which tested negative for A. fumigatus) were from cows which had died from unknown causes at two different dairies. Both had exhibited rumen stasis and sudden death but did not have HBS. Local feeds were tested for the presence of A. fumigatus. Feed samples have included mill run, ground corn, grass and corn silages and dried grass hay. Many were infected with A. fumigatus. Infection is not always visible (A. fumigatus on moist feed (e.g., corn) is dark blue-green). Hence, exposure to A. fumigatus may be common. An additional factor, in tandem with A. fumigatus, (possibly immunosuppression) predisposes dairy cattle to HBS. Clostridial toxins. Genotype analysis of five clostridial toxins indicated no correlation between HBS and toxin genes (Table 2). In HBS and AH cases, genes encoding toxins a and e were detected in 3 of 9 analyses. Toxin genes ? and enterotoxin were not detected in any samples. The a and e toxin genes were detected in blood (1 of 3 analyses), jejunal or abomasal clot (2 of 4 analyses) and GI wall (1 of 5 analyses). In negative control cows, the ß toxin gene was detected in blood of 3 of 17 animals (Table 2). Of interest, two of the three negative control cows which harbored the Clostridium ß-toxin gene had aborted. Study A. In vivo assessment. Efficacy of OmniGen-AF was tested on local dairies. To date, OmniGen-AF has been introduced at five dairies in the Pacific-Northwest. Dairies ranged in size from 100 to 950 cows. Each had experienced HBS, mycotic abortions, or both. While definitive conclusions from these studies cannot be drawn (because studies on farms are not “controlled” and because dairy numbers remains low) incidence of HBS and mycotic abortions were eliminated following the introduction of OmniGen-AF on these dairies. Study B: Infection of steers with A. fumigatus. Moldy feed from one HBS-afflicted dairy (WA #1; Table 1) was cultured in large-scale and used to infect rolled corn. Infected corn was used as a culture medium for fungal growth by incubating at 27 C for 1-month. A. fumigatus was detected in the molded corn product after 1 month of culture. The infected corn was fed to beef steers for 21 days (with and without OmniGen-AF) and then removed from the diet. A. fumigatus DNA concentrations were monitored in the blood. Infection of steers occurred rapidly. Within 2 days, A. fumigatus levels reached nearly 40 million genomic units/ml in infected steers. Infection with A. fumigatus caused rapid clearance of A. flavus from the blood (Figure 6). Furthermore, addition of the antifungal product enhanced the clearance of A. fumigatus from the blood of A. fumigatus-infected steers (Figure 7). Figure 6. Blood fungal DNA concentrations in beefs steers fed A. fumigatus-laden feed for 21 days. Blood was analyzed for A. fumigatus and A. flavus DNA prior to introduction of feed (Day 0) and after 21 days of feeding. Note that A. fumigatus appearance was coincident with clearance of A. flavus. Observations are means of two steers/treatment. Figure 7. Clearance of A. fumigatus from the blood of infected steers. Animals had been fed A. fumigatus-infected feed after which feed was withdrawn and clearance of fungal load was monitored over 15 days. The anti-fungal product (OmniGen-AF) enhanced clearance from the blood (P<0.05). The experiment included three steers/treatment. This study demonstrates that HBS is associated with high levels of A. fumigatus in the gut, gut wall and blood. In human studies, A. fumigatus is recognized for its invasive properties and pathogenicity in immunocompromised patients. We propose it has similar potential in ruminants. Analysis of clostridial toxin genes did not indicate a correlation between these and HBS. Limitations of this analysis include the small data set (i.e., eight HBS cows, two AH animals, 17 controls) and lack of GI tissues from negative control cows. A larger data set and completion of controlled studies with dairy animals are needed to definitively ascribe HBS to A. fumigatus. A further limitation is that >100,000 fungal species are known. There are >100 Aspergillus species. Only a very small proportion of their ITS domains have been published. Hence, our detection of A. fumigatus could include unrecognized (non-sequenced) fungal species. Another issue with HBS is that many cows consume infected feed yet only a small percentage of these cows develop HBS. We do not know what the other predisposing factor(s) may be. However, we propose that stressinduced immunoincompetance may play a role in the etiology of HBS. Finally, we developed and tested a product which inhibits fungal growth in vitro and in vivo. Application of this product on dairies appears to hold potential for prevention of HBS and, possibly, other mycotic infections. Limiting Exposure from Improperly Ensiled Forages. Following best management practices for proper storage and rotation of feed ingredient and ensiled forage inventories can greatly reduce exposure to Aspergillus fumigatus growth. Attention to detail when ensiling forages is paramount to the prevention of mold growth. Maturity is important when harvesting forages. Goals for moistures should be 68 to 72% for corn silages and 64 to 68% for legume haylages going into bunker or drive over silos. Silage pits should be continuously and rapidly filled, and immediately packed with a heavy wheeled tractor for best results. Avoid interruptions in the filling process, if possible, to prevent layers of spoilage from forming. Properly fermented corn silages should reach a pH of 4.0 or less and legume forages will come in below pH 4.5. Innoculants have been beneficial in more rapidly lowering silage pH’s. Significant reductions in dry matter losses have been reported with the use of some silage inoculants, however results have historically been variable. Molds will rapidly grow as the lactic acid from fermentation flashes off from surface layers exposed to the weather. Bunker silos should be immediately covered with quality plastic and weighted down with adequate tires to ensure that the cover will remains firmly in place. Bunker silos should be designed to allow filling from the back, sloping away from the intended active feeding area to prevent rain runoff from draining into exposed feed. The spoiled layer at the top of the silo should be discarded, whenever possible, providing it doesn’t pose a safety risk to employees. In a 2003 OSU survey of Pacific Northwest feedstuffs, extremely high concentrations of Aspergillus fumigatus spores (>1.25 million per gram) were isolated from the spoiled layer from the top of corn silage bunkers (See Table 3). High levels of A.fumigatus spores were also isolated from separated manure solids used to anchor silo covers (>3 million spores per gram), making this a questionable practice, especially in cases where bunkers are sloped towards the active feeding face. Compromising hygiene at the feed bunks and water troughs often leads to explosive growth of molds including Aspergillus. Regular cleaning of feed bunks and water troughs should be part of every dairy’s HAACP program. Other recommendations include avoidance of feeding spoiled or molded grains and other feedstuffs to pregnant or lactating animals. Feeding excessively high starch rations can lead to ruminal acidosis and severely damage the gut mucosa, predisposing animals to colonization by invasive molds and other pathogens.