In dogs with idiopathic acute haemorrhagic diarrhoea syndrome (AHDS), a serious loss of intestinal mucosal barrier integrity occurs. However, the incidence of bacterial translocation in dogs with idiopathic AHDS is not known. Thus, the objectives of this prospective study were to identify the incidence of bacteraemia, to evaluate the frequency of septic events and the influence of bacteraemia on various clinical and laboratory parameters, duration of hospitalisation and survival of dogs with idiopathic AHDS. The study included 87 dogs with idiopathic AHDS. Twenty-one healthy dogs served as control group. To evaluate clinical significance of bacterial translocation, blood culture results were compared between patients and controls. Clinical and laboratory parameters were compared between patients with positive and negative blood cultures. There was no significant difference in either incidence of bacteraemia between patients with idiopathic AHDS (11 per cent) and controls (14 per cent) or in severity of clinical signs, laboratory parameters, duration of hospitalisation or mortality between blood culture-positive and culture-negative dogs with idiopathic AHDS. The results of this study suggest that the incidence of bacteraemia in dogs with idiopathic AHDS is low and not different from that of healthy control dogs. Bacteraemia does not influence the clinical course or survival and thus antibiotic treatment is not indicated to prevent sepsis.
- Bacterial diseases
- Intestinal disease
Statistics from Altmetric.com
Idiopathic haemorrhagic gastroenteritis (HGE), a syndrome of unknown aetiology, has recently been renamed to idiopathic ‘acute haemorrhagic diarrhoea syndrome’ (AHDS) (Unterer and others 2014). Idiopathic AHDS is characterised by acute-onset vomiting, anorexia and lethargy, progressing to haematemesis, and severe, malodorous, haemorrhagic diarrhoea (Burrows 1977, Spielman and Garvey 1993, Unterer and others 2014). A presumptive diagnosis of idiopathic AHDS is made by observation of characteristic clinical signs and exclusion of any disease known to cause haemorrhagic diarrhoea.
A number of risk factors for bacteraemia exist in patients with acute haemorrhagic diarrhoea. Necrosis of the mucosa leading to a breakdown of the gastrointestinal (GI) mucosa–blood barrier has been demonstrated in dogs with idiopathic AHDS (Unterer and others 2014). Additionally, the adherence of large numbers of bacteria to the necrotic mucosal surfaces has been demonstrated in some cases (Prescott and others 1978, Sasaki and others 1999, Unterer and others 2014). Hypoalbuminaemia due to intestinal loss, which is frequently observed in dogs with HGE (Will and others 2005), indicates a substantial destruction of the mucosal epithelial cell layer. It is believed that splanchnic and intestinal ischaemia leads to reduced intestinal barrier function (Wiest and Rath 2003). Severe dehydration and hypovolaemia, which are very likely associated with the hypoperfusion of the splanchnic organs, are characteristic findings in dogs with idiopathic AHDS (Spielman and Garvey 1993, Unterer and others 2011) and can also increase the risk of bacteraemia. Recent molecular studies revealed bacterial dysbiosis in faecal samples of dogs with idiopathic AHDS (Suchodolski and others 2012), and these microbial changes can be associated with an increased risk of bacterial translocation (Wiest and Rath 2003).
Transient bacteraemia can also occur in normal, healthy individuals (Harari and others 1993, Howe and others 1999). In these dogs, the liver usually effectively eliminates organisms from the blood as it passes through the sinuses as long as the bacterial load does not overwhelm the functional capacity of the hepatic reticuloendothelial system (Fox and others 1987). However, persistence or overload of pathogens in the systemic bloodstream with the release of toxic products can result in septicaemia. The GI tract is suspected to be the main source of sepsis with Gram-negative bacteria (Turk and others 1990, MacFie and others 1999, Lobetti and others 2002, Wiest and Rath 2003, Kenney and others 2010). Three primary mechanisms leading to enhanced bacterial translocation from the GI tract have been identified: (1) intestinal bacterial overgrowth, (2) deficiency in host immune defences and (3) damage to the GI mucosal barrier (Wiest and Rath 2003).
Since several mechanisms for bacterial translocation, which potentially could lead to sepsis, exist in dogs with haemorrhagic diarrhoea, antibiotics have generally been recommended. However, the frequency and clinical relevance of bacterial translocation and the role of bacteria as inciting agents in dogs with idiopathic AHDS as well as the necessity of treatment with antibiotics are not known. An inappropriate use of antibiotics can cause disruptions of the protective intestinal microbiota (Dethlefsen and others 2008), post-antibiotic salmonellosis, Clostridium difficile-associated diarrhoea (Rupnik and others 2009) and resistance to antibiotics (Lloyd 2007, Costelloe and others 2010; Gronvold and others 2010).
Thus, the objectives of this study were to identify the incidence of bacteraemia in dogs with idiopathic AHDS, to evaluate the frequency of septic events and the influence of bacteraemia on various clinical and laboratory parameters during the hospitalisation and survival of these dogs.
Materials and methods
This prospective clinical study was conducted according to the German animal welfare law. Owners were informed about the purposes of the study. Eighty-seven patients with idiopathic AHDS that were presented between April 2006 and September 2009 to the Clinic of Small Animal Medicine, LMU University of Munich, were enrolled in this study. The median age was 5.0 years (range 0.3–16 years), and the median body weight was 9.8 kg (range 1.5–54.8 kg). The breeds most commonly represented were mixed-breed dogs (n=30), Yorkshire terriers (n=5), maltese (n=5) and miniature pinschers (n=4). Thirty-eight patients were male (11 castrated), and 49 were female (26 spayed).
The inclusion criterion was presence of idiopathic AHDS with acute haemorrhagic diarrhoea for less than three days. Dogs pretreated with antibiotics and dogs with any disease known to cause haemorrhagic diarrhoea were excluded. To rule out these causes for haemorrhagic diarrhoea, a standardised history and physical examination, abdominal ultrasound examination (LOGIQ P6, GE Medical Systems, Milwaukee, Wisconsin, USA), complete blood cell (CBC) count (Cell-Dyn 3500 R, Abbott Diagnostics, Illinois, USA), manual differential blood cell count, serum biochemistry profile including serum bile acid concentration evaluation (Hitachi 912 Automatic Analyzer, Roche Diagnostics GmbH, Mannheim, Germany), prothrombin time and activated partial thromboplastin time (aPTT) determination (Coagulometer CL, BE Behnk Elektronik, Norderstedt, Germany) and faecal examination for nematode and protozoan parasites (29.5 per cent sodium nitrate flotation solution, Janssen-Cilag, Neuss, Germany, and a Giardia antigen ELISA, ProSpecT Giardia Microplate Assay, Remel, Lenexa, Kansas, USA) and for parvovirosis (electron microscopy) were performed. On the same day or overnight, a faecal culture from fresh faeces (less than one hour) was shipped to the reference laboratory to detect potential enteropathogenic bacteria (Salmonella species, Campylobacter species, Yersinia species, Escherichia coli).
Excluded from the study were dogs that had received any drug known to cause mucosal irritation (e.g. non-steroidal anti-inflammatory drugs, corticosteroids and doxycycline) within one week of presentation or that had an underlying disease that could potentially cause haemorrhagic diarrhoea. An underlying disease was suspected if one of the following criteria was met: bleeding disorders with a prothrombin time longer than 46 seconds, aPTT longer than 26 seconds or platelet count less than 100,000/μl; renal failure with blood urea nitrogen (BUN) higher than 17.3 mmol/l or creatinine more than 146.0 µmol/l in conjunction with a urine specific gravity less than 1.030; and liver disease with more than two parameters that reflected abnormal liver function (e.g. serum bile acids, albumin, bilirubin, BUN, cholesterol and glucose). Patients with parvovirosis, patients with a faecal culture positive for Salmonella, Yersinia or Campylobacter species and patients with endoparasite infestation (including Giardia species) also were excluded. To screen for focal intestinal (e.g. neoplasia, intussusception and foreign body) or other visceral diseases, abdominal ultrasound examination was performed in every dog. Special attention was paid to the pancreatic region to rule out pancreatic changes. In addition to the ultrasound examination, radiographs were performed at the discretion of the clinician managing the case to rule out foreign bodies. Pancreatitis (which also led to exclusion from the study) was suspected if typical ultrasonographic changes, substantial abdominal pain or both were present and if clinical improvement was not consistent with the expected clinical course of idiopathic AHDS.
Addison's disease also led to exclusion. None of the included dogs had electrolyte changes, which are considered characteristic for hypoadrenocorticism. To further rule out Addison's disease, dogs were followed up for at least 12 months after discharge. Follow-up information was obtained by telephone communication with the owner or referring veterinarian. Two dogs had a recurrence of diarrhoea. In those two dogs, baseline cortisol concentration was measured to rule out Addison's disease, and the results were within the reference range.
Twenty-one dogs owned by students or employees of the LMU University of Munich, Germany, were included in the control group. Blood was obtained for unrelated purposes (e.g. a yearly health check examination). The median age was 4.0 years (range 1.0–11.0 years), and the median body weight was 23.8 kg (range 7.5–40.0 kg). Breeds included mixed-breed dogs (n=10), dachshunds (n=1), labrador retrievers (n=2), rottweilers (n=2), huskies (n=1), cairn terriers (n=1), Rhodesian ridgebacks (n=1), beagles (n=1), cocker spaniels (n=1) and Jack Russell terriers (n=1). Eight dogs were male (three castrated) and 13 were female (10 spayed).
All dogs were considered to be healthy based on their medical histories, physical examinations and laboratory analyses (CBC and biochemical profiles). Dogs pretreated with antibiotics over the last three weeks before entering the study were excluded from the control population.
For disinfection of the phlebotomy side, the skin was scrubbed twice for two to three minutes with a iodine washing solution (Jodosept,Vetoquinol, Ravensburg, Germany) followed by the application of isopropyl alcohol (Kodan, Schülke & Mayr GmbH, Norderstedt, Germany). Two separate blood cultures of 5–10 ml (5 ml in dogs <10 kg, 10 ml in dogs >10 kg) were taken from all dogs at a 30-minute interval from the jugular vein after sterile preparation of the skin. Blood samples were immediately inoculated into a blood culture system (Oxoid system blood culture medium, Oxoid, Basingstoke, UK) using a fresh needle. Care was taken that air did not enter the vacuum bottles. The blood was dispersed in the culture medium by gently inverting the bottle two to three times. Samples collected from 8:00 to 17:00 were immediately (<1 hour) submitted for aerobic and anaerobic bacterial cultures. Samples collected from emergency cases were stored at room temperature and submitted within 15 hours of collection.
The culture bottles were incubated at 37°C–39°C. As soon as a positive signal arose, 1 ml of the broth was used to produce subcultures on a set of routine primary media plates. Nutrient agar with and without six per cent sheep blood, Gassner-agar, Rambach-agar (all from Merck, Darmstadt, Germany) and Columbia colistin nalidixic acid blood-agar (Becton Dickinson, Heidelberg, Germany) were incubated at 37°C–39°C under aerobic conditions and were examined daily for at least two days. For anaerobic incubation, blood-agar for incubation with enhanced CO2 and chocolate agar plates were used (Anaerocult P and Anaerocult C, Merck, Darmstadt, Germany). For the biochemical differentiation of isolates, ID-32-Staph, API-20-NE, ID-32-E rapid (all from Bio Mérieux, Lyon, France) and BBL Enterotube (Becton Dickinson, Heidelberg, Germany) were used. In blood culture bottles without an indication of bacterial growth, culture methods were applied as described earlier (after five to seven days of incubation).
A blood culture was considered positive if clinically significant bacteria were isolated at least once (e.g. Enterobacteriaceae) or if a potential contaminant (e.g. coagulase-negative Staphylococcus species) was isolated twice from blood cultures of both blood samples collected within the 30-minute period. A single positive growth of an organism considered a normal skin resident was determined to be a contaminant (Dow and others 1989).
Evaluation of clinical and laboratory parameters
At presentation, clinical signs were assessed and quantified using a simple scoring system, the ‘AHD activity index’ (Table 1). This index includes the parameters of activity, appetite, vomiting, stool consistency, stool frequency and dehydration. Each parameter was scored (0=normal, 1=mild, 2=moderate, 3=severe), and the sum of the scores yielded a total cumulative score (Unterer and others 2011). Systemic inflammatory response syndrome (SIRS) was suspected when two or more of the following criteria were met: hypothermia or fever, tachycardia or bradycardia, tachypnoea, leucocytosis, leucopenia or increased band neutrophil counts in animals with a normal total white blood cell count (Hauptman and others 1997, de Laforcade and others 2003) (Table 2).
Some patients were treated with antibiotics (45 dogs). Additional therapy was equal for all dogs: fluid therapy (crystalloids; fluid amount depended on dehydration, maintenance demands and ongoing losses), antiemetics, gastric antacids and disseminated intravascular coagulation prophylaxis.
Dogs fulfilling the following criteria were discharged from the hospital: normal activity, no dehydration, no vomiting and no watery diarrhoea (only dogs with grades 1 and 2 diarrhoea according to the AHD activity index were allowed to leave the hospital).
CBC and serum biochemistry analysis for evaluation of the influence of bacteraemia included the following parameters: total white blood cell count, banded neutrophil count, platelet count, alanine aminotransferase, albumin, bilirubin and glucose.
Statistical analyses were conducted with GraphPad Prism V.5.0 (San Diego, USA). Normal distribution of data was determined with the D’Agostino-Pearson omnibus normality test. The values of two groups (patients v controls, dogs with positive blood culture v dogs with negative blood culture) were compared using an unpaired t test (or, if the data were not normally distributed, the Mann-Whitney U test) and were performed twice with regard to comparing dogs with positive and negative blood cultures. All dogs with positive cultures were included in the positive culture group in the first analysis. In a second analysis, only the dogs that had a positive culture with a clinically significant organism were included. With multiple comparisons, an adjustment of the P value is indicated. Due to the 18 comparisons for blood culture-positive versus blood culture-negative dogs, a Bonferroni correction to P<0.0028 (P<0.05 divided by 18 comparisons) was applied. A similar correction was applied to the comparison of age, sex and weight between the patients and controls leading to a P<0.017 (P<0.05 divided by the three comparisons). The number of dogs with positive blood cultures in the control versus the affected group was compared with Fisher's exact test, and a P<0.05 was considered significant.
Comparison of the signalment of patients and controls
Age and sex did not differ significantly between the patient and the control group. However, the mean body weight of the control group was significantly higher than that of the patient group (P=0.015).
Blood culture results
Fourteen of the 87 dogs (16 per cent) with idiopathic AHDS had positive blood cultures. In each patient, only one of the two blood samples was culture positive. Organisms found on blood cultures included Clostridium species (n=4), Corynebacterium accolens (n=1), other Corynebacterium species (n=1), Gram-positive coryneform bacteria (n=2) and Staphylococcus pseudintermedius (n=2). In four dogs, the detected bacterial organisms were classified as contaminants (Staphylococcus warneri (n=1), Staphylococcus xylosus (n=1), Staphylococcus epidermidis (n=1), Arthrobacter species (n=1)). Therefore, the incidence of bacteraemia was 11 per cent (10/87).
In three of the 21 dogs of the control group, bacteria were found in one of the two cultures (14 per cent). Organisms detected on blood cultures included Clostridium species (n=1), Corynebacterium species (n=1) and haemolysing Escherichia coli (n=1). None of these organisms was assessed as contaminants.
There was no significant difference when the prevalence of bacteraemia between patients and control animals was compared (P=0.714).
Comparison of clinical and laboratory parameters between patients with positive and negative blood culture results
The results were similar for both statistical analyses (only considering the relevant pathogens as positive and considering all dogs positive when bacteria were isolated on blood cultures). Further, only the results of the evaluation considering the dogs with clinically presumably relevant bacteria positive are displayed. There were no significant differences between the dogs with bacteria-positive and bacteria-negative blood cultures regarding severity of clinical signs or laboratory parameters in this study (Table 3). Even when omitting the α-inflation and setting the significance level at P<0.05, only two parameters (AHD score and rectal temperature) were different between the dogs with positive and negative blood cultures. An F-test was performed for each parameter in both groups. The variances were not significantly different. Additionally, the differences concerning these two parameters between dogs with bacteria-positive (median: AHDS index 10; temperature 38.8°C) and bacteria-negative blood cultures (median: AHDS index 12; temperature 38.3) were considered to be clinically irrelevant.
No dog died of idiopathic AHDS; therefore, every dog was discharged as soon as the prospectively defined criteria were met. No significant differences were found when comparing the duration of hospitalisation (P=0.434) between the blood culture-positive and culture-negative patients.
Based on the criteria defined by de Laforcade and colleagues (2003) and Hauptman and others (1997) (Table 2), 48 out of the 87 patients (55 per cent) were classified as having SIRS. Three of 10 dogs with bacteria-positive blood cultures (30 per cent) and 45/77 dogs with bacteria-negative blood cultures (58 per cent) fulfilled the criteria for SIRS. The number of dogs fulfilling the criteria was significantly higher in the group of culture-negative dogs (P=0.035).
Blood cultures are warranted whenever clinical signs indicate the possibility of bacteraemia or sepsis. Blood cultures are essential for selecting the appropriate antimicrobial treatment (Dow and others 1989). The unnecessary administration of costly and potentially toxic antibiotics to patients may result in adverse drug reactions, increased length of hospitalisation and antimicrobial resistance in microorganisms. Therefore, information regarding the frequency of bacteraemia and the necessity of blood cultures in patients at risk for sepsis is important. Dogs with idiopathic AHDS meet several criteria for being at risk. First, they show clinical signs (haemorrhagic diarrhoea) and histological lesions (necrotising enterocolitis) (Prescott and others 1978, Unterer and others 2014) reflecting the severe damage of the intestinal mucosal barrier, which can lead to increased bacterial translocation. Second, when applying the modified criteria established for dogs (Hauptman and others 1997, de Laforcade and others 2003) at presentation, more than 50 per cent of patients with idiopathic AHDS would be classified as being septic or as suffering from SIRS.
Despite including a population with several factors predisposing them for bacterial translocation in this study, the prevalence of bacteria-positive blood cultures was low, and no difference was observed when comparing the study group with the healthy control dogs. Either bacteraemia is an uncommon event in dogs with idiopathic AHDS or blood cultures are an insensitive diagnostic tool to detect bacteraemia. False-negative blood cultures are frequently observed in chronic bacteraemia (e.g. infective endocarditis), even in dogs with clinical signs (Hoen and others 1995). In diseases associated with acute bacterial translocation, the frequency of false-negative blood culture results is not known. However, the bacterial load would be expected to be high in the event of massive translocation and, thus, it appears unlikely that systemic bacteraemia was frequently missed in dogs with idiopathic AHDS in the present study. Additionally, the sensitivity of blood cultures used in this study does not appear to be low as bacteria cultured in 14 per cent of the healthy dogs. Molecular methods for the detection of pathogens in blood have been developed with the aim to improve sensitivity and to detect bloodstream infection earlier. However, at the time the study was started, blood cultures were the gold standard and molecular-based methods had not passed the threshold for clinical practice (Avni and others 2010, Skvarc and others 2013).
Bacteria are normally eliminated from the blood by host defence mechanisms. On occasion, they circumvent these barriers in clinically healthy individuals and cause transient bacteraemia. The first report documenting the prevalence of bacterial translocation in a large series of apparently healthy dogs showed that bacterial translocation to mesenteric lymph nodes occurred in 52 per cent of dogs undergoing elective ovariohysterectomy (Dahlinger and others 1997). Additional studies confirmed portal and systemic bacteraemia in clinically normal dogs (Harari and others 1993, Howe and others 1999, Winkler and others 2003). These findings are in agreement with the results of the present study, indicating a rate of bacterial translocation in 14 per cent of healthy dogs.
It can be assumed that dogs with a severe destruction of the intestinal barrier, as observed in idiopathic AHDS, can still effectively clear bacteria entering the portal blood as long as the liver and immune system are functioning as a cellular back-up system. This is in contrast with patients with parvoviral enteritis, in which the combination of intestinal damage and neutropenia usually results in sepsis with Gram-negative bacteria from the intestinal microbiota (Goddard and Leisewitz 2010). In cases of parvovirosis, neutropenia severely impairs the elimination of bacteria from the bloodstream (Goddard and Leisewitz 2010). The importance of clearing the portal blood from pathogens by the liver was shown in a case series of dogs with congenital portosystemic shunts. Recurrent fever reflecting transient sepsis was the only or the predominant clinical sign in these patients with portosystemic vascular anomalies (Wess and others 2003).
All dogs included in the present study survived and recovered uneventfully. There was no difference in the outcome between dogs with positive and negative blood cultures for bacteria. These facts and the low rate of positive blood cultures indicate that sepsis is uncommon in dogs with idiopathic AHDS, although clinical signs suggestive of sepsis (e.g. hypothermia, tachycardia and tachypnoea) are observed in half of the dogs with and without bacteraemia. The list of criteria proposed by Hauptman and others (1997) is a sensitive tool in diagnosing sepsis in dogs (Hauptman and others 1997). However, the high sensitivity is associated with a low specificity and thus a high rate of false-positive categorisations. Dogs with idiopathic AHDS are frequently dehydrated due to a rapid and intense loss of fluid into the gut lumen (Burrows 1977). Hypothermia, tachycardia and tachypnoea are common physical examination findings in patients in hypovolaemic shock (Boag and Hughes 2005). Thus, it becomes obvious that these criteria cannot be used in hypovolaemic dogs with idiopathic AHDS to detect SIRS. The low rate of bacteraemia in dogs with idiopathic AHDS and the fact that antibiotics do not change the outcome or time to recovery in dogs with idiopathic AHDS (Unterer and others 2011) suggest that dogs with idiopathic AHDS should not routinely be treated with antibiotics. There was no difference in severity of clinical signs, duration of hospitalisation or outcome between patients with positive and negative blood cultures, providing evidence that blood cultures are not helpful in assessing the clinical course and the need for antibiotic treatment.
An extensive diagnostic work up was performed in every dog included in this study to rule out any disease known to cause bloody diarrhoea. However, it is possible that different aetiologies were responsible for this syndrome in different patients.
The rather small number of dogs in the control group is a possible limitation of this study. Nevertheless, previous studies showed the same or an even higher prevalence of bacteria growth in the blood cultures of healthy dogs (Harari and others 1993, Dahlinger and others 1997). Consequently, it appears unlikely that the results would have changed with a larger control group.
It is difficult to interpret a positive blood culture with bacteria that are considered contaminants, as those bacteria might be present on normal skin and are not normally involved in the pathogenesis of sepsis. However, it cannot be completely excluded that in some dogs, those bacteria might indeed contribute to the clinical signs. To address this dilemma, two statistical analyses were conducted that compared either all dogs with positive bacterial cultures or only the dogs with likely pathogens in their blood cultures with the dogs with negative blood cultures. The results of the two analyses were identical; therefore, it appears unlikely that changes in the definition of contaminants and pathogens would change the results and their interpretation.
The results of this study suggest that the incidence of bacteraemia in dogs with idiopathic AHDS is low and is not different from healthy control dogs. Bacteraemia neither influenced the clinical course nor the survival. Thus, blood cultures are not useful in dogs with idiopathic AHDS because they do not influence therapeutic decisions. Specific sepsis criteria not depending on the hydration status of the patient might be better suited to identify the dogs with idiopathic AHDS that could benefit from antibiotic treatment. Antimicrobial administration cannot be recommended as a routine treatment in dogs with AHDS.
Provenance: not commissioned; externally peer reviewed
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.