Charlotte Hopster-Iversen
Influence of mechanical manipulations on the local inflammatory reaction in the equine
jejunum and colon
Cuvillier Verlag Göttingen
Internationaler wissenschaftlicher Fachverlag
Bibliografische Information der Deutschen Nationalbibliothek Die Deutsche Nationalbibliothek verzeichnet diese Publikation in der
Deutschen Nationalbibliografie; detaillierte bibliografische Daten sind im Internet über http://dnb.d-nb.de abrufbar.
1. Aufl. - Göttingen : Cuvillier, 2013
Zugl.: Hannover (TiHo), Univ., Diss., 2013 978-3-95404-560-0
© CUVILLIER VERLAG, Göttingen 2013 Nonnenstieg 8, 37075 Göttingen Telefon: 0551-54724-0
Telefax: 0551-54724-21 www.cuvillier.de
Alle Rechte vorbehalten. Ohne ausdrückliche Genehmigung des Verlages ist es nicht gestattet, das Buch oder Teile daraus auf fotomechanischem Weg (Fotokopie, Mikrokopie) zu vervielfältigen.
1. Auflage, 2013
Gedruckt auf umweltfreundlichem, säurefreiem Papier aus nachhaltiger Forstwirtschaft.
978-3-95404-560-0
Equine Clinic
University of Veterinary Medicine Hannover
Influence of mechanical manipulations on the local inflammatory reaction in the equine jejunum and colon
THESIS
Submitted in partial fulfillment of the requirements for the degree
DOCTOR OF PHILOSOPHY (PhD)
awarded by the University of Veterinary Medicine Hannover
By
Charlotte Hopster-Iversen Understed (Denmark)
Hannover (2013)
Supervisors: Dr. Anna Rötting, PhD Prof. Dr. Karsten Feige
Equine Clinic, University of Veterinary Medicine Hannover, Germany
Supervision Group: Prof. Dr. Karsten Feige
Equine Clinic, University of Veterinary Medicine Hannover, Germany
Prof. Dr. Ralph Brehm
Funktionelle Histologie und Zellbiologie, Anatomisches Institut, Stiftung Tierärztliche Hochschule Hannover
Prof. Dr. David Freeman, PhD
Island Whirl Equine Colic Research Laboratory, Department of Large Animal Clinical Sciences, College of Veterinary Medicine, University of Florida
1st Evaluation: Prof. Dr. Karsten Feige Dr. Anna Rötting, PhD Prof. Dr. Ralph Brehm
Prof. Dr. David Freeman, PhD
2nd Evaluation: Prof. Dr. Anthony Blikslager, PhD,
Department of Veterinary Sciences, College of Veterinary Medicine,North Carolina State University
Date of the final exam: 31.10.2013
Parts of the thesis have been published in:
Equine Veterinary Journal 2011; 43 1-7
“Influence of mechanical manipulations on the local inflammatory reaction in the equine colon”
Hopster-Iversen, C., Hopster, K., Staszyk, C., Rohn, K., Freeman, D., and Rötting, A.
K.
American Journal of Veterinary Research
“Effects of experimental mechanical manipulations on the local inflammatory reaction in the equine jejunum”
Hopster-Iversen, C., Hopster, K., Staszyk, C., Rohn, K., Freeman, D., and Rötting, A.
K.
American Journal of Veterinary Research, Manuscript ID: AJVR-13-08-0238, under review
Parts of the thesis have been presented on conferences:
21. Arbeitstagung der DVG-Fachgruppe Pferdekrankheiten on Februar 12-13, 2010 in Hannover (in german)
European College of Veterinary Surgeons, 20th Annual Scientific Meeting on July 7–
9, 2011 in Ghent, Belgium (Poster)
10th International Equine Colic Research Symposium on July 26-28, 2011 in Indianapolis, USA (Poster)
5th ECEIM congress in February 2012 in Edinburgh
22. Arbeitstagung der DVG- Fachgruppe Pferdekrankheiten on March 16.-17, 2012 in Hannover, Germany (in German)
To Klaus
Table of Contents
1 SUMMARY ... 1
2 ZUSAMMENFASSUNG ... 3
3 INTRODUCTION ... 5
3.1 Colic in horses and its importance ... 5
3.1.1 Complications after colic surgery ... 5
3.1.1.1 Postoperative ileus (POI) ... 6
3.1.1.2 Endotoxaemia ... 6
3.1.1.3 Adhesion formation ... 8
3.1.1.4 Other complications ... 9
3.2 Postoperative ileus (POI) in the literature ... 10
3.2.1 Definition ... 10
3.2.2 Aetiology ... 10
3.2.2.1 Neurogenic phase ... 11
3.2.2.2 Inflammatory phase ... 12
3.2.3 Treatment of POI ... 14
3.2.3.1 Treatment of POI in human medicine ... 14
3.2.3.1.1 Factors enchaining recovery after surgery: ... 14
3.2.3.1.2 Factors contributing to delayed recovery from POI ... 16
3.2.3.2 Treatment of POI in equine medicine ... 17
3.3 Physiology of normal equine intestinal motility ... 21
3.4 Intestinal homeostasis and inflammation ... 22
3.5 Eosinophilic granulocytes ... 23
3.5.1 Origin, development, and localization of the eosinophilic granulocyte .. 23
3.5.2 Function of the eosinophilic granulocytes ... 24
3.5.2.1 Eosinophilic granulocytes and organ development ... 26
3.5.2.2 Function of the eosinophilic granulocytes in the gastrointestinal tract ... 26
3.5.2.3 Eosinophils and parasitism ... 27
3.5.3 Eosinophilic granulocytes and diseases ... 27
3.5.3.1 Gastrointestinal tract diseases involving eosinophilic granulocytes 27 3.5.3.2 Other diseases involving eosinophilic granulocytes in horses ... 30
3.5.4 Studies involving eosinophils in the equine gastrointestinal tract ... 31
3.6 Neutrophilic granulocytes ... 33
3.6.1 Origin, development, and localization of the neutrophilic granulocytes . 33 3.6.2 Function of the neutrophilic granulocyte ... 34
3.6.3 Neutrophilic granulocytes and diseases ... 35
3.6.3.1 Gastrointestinal tract diseases involving neutrophilic granulocytes 36 3.6.3.2 Other diseases involving neutrophilic granulocytes ... 37
3.6.4 Studies involving neutrophilic granulocytes in the equine gastrointestinal tract ... 39
3.7 Previous studies on the effects of mechanical manipulations on the intestine ... 40
3.7.1 Effect of laparotomy ... 40
3.7.2 Effect of small intestinal manipulation ... 40
3.7.3 Effect of large intestinal manipulation ... 43
3.7.4 Panenterits and systemic effects of intestinal manipulation ... 44
4 HYPOTHESIS AND AIM OF THE STUDY ... 45
5 Manuscript A ... 47
Abstract: ... 47
Introduction: ... 48
Material and methods ... 50
Horses ... 50
Study design: ... 50
Study 1: ... 51
Study 2: ... 52
Histological evaluation: ... 52
Statistical analysis: ... 55
Results ... 55
Study 1: ... 55
Study 2: ... 56
Discussion ... 62
Manufactures addresses: ... 65
References ... 66
6 Manuscript B ... 71
Abstract ... 72
Objective—To determine characteristics of the inflammatory reaction in jejunums mechanical manipulations. ... 72
ABBREVIATION . ... 72
Material and Methods ... 74
Study design— Histologic evaluation ... 76
Statistical analysis ... 80
Results ... 8
Discussion ... 85
Footnotes: ... 89
References ... 90
7 GENERAL DISCUSSION ... 95
7.1 Method ... 95
7.2 Material ... 98
7.3 Results ... 99
7.3.1 Eosinophils ... 99
7.3.2 Neutrophils . ... 101
7.4 Clinical consequences ... 104
7.5 Outlook . ... 105
8 REFERENCES ... 107
ACKNOWLEDGEMENTS ... 149 of horses in response to various
0
9
List of Abbreviations
5-HT 5-Hydroxytryptamin (Serotonin)
ACVS American College of Veterinary Surgeons ARDS Acute Respiratory Distress Syndrome BAL Bronchoalveolar Lavage
bwt Body Weight
CCR C Chemokine Receptor CF Cystic Fibrosis COX-2 Cyclo-Oxygenase 2 CRI Constant Rate Infusion CSF Colony Stimulating Factor DEE Diffuse Eosinophilic Enteritis
DIC Disseminated Intravasal Coagulopathy ECP Eosinophil Cation Protein
EDN Eosinophil Derived Neurotoxin
EGID Eosinophilic Gastro-Intestinal Disorders EPO Eosinophil Peroxidase
G-CSF Granulocyte-Colony Stimulating Factor GIT Gastro-Intestinal Tract
GM-CSF Granulocyte-Macrophage-Colony-Stimulating-Factor HOCl Hypochlorous Acid
IAD Inflammatory Airway Disease IBD Inflammatory Bowel Disease ICAM Intra Cellular Adhesion Molecule ICC Intestinal Cells of Cajal
IFEE Idiopathic Focal Eosinophilic Enteritis IL Interleukin
iNOS inducible isoform of NO synthase I/R Ischaemia-Reperfusion
i.v. Intravenously LPS Lipopolysaccharide
LTB Leukotriene B MBP Major Basic Protein
MCP-1 Monocytic Chemotactic Protein-1
MEED Multisystemic Eosinophilic Epitheliotrophic Disease MHC Major Histocompatibility Complex
MMC Migrating Myoelectric Complex mRNA Messenger RNA
NO Nitric Oxide
NSAID Non-Steroidal Anti-Inflammatory Drug PAF Platelet Activating Factor
PCV Packed Cell Volume POI Postoperative Ileus
PONV Post Operative Nausea and Vomiting Prophylaxis SIRS Systemic Inflammatory Response Syndrome TBS Tracheobronchial Secretions
TNF-α Tumor-Necrosis Factor alpha TNF-β Tumor-Necrosis-Factor beta
List of Tables, Figures
Table 1 ... 25
Table A 1 ... 57
Table A 2 ... 59
Table A 3 ... 60
Figure B 1 ... 80
Figure B 2 ... 81
Figure B 3 ... 91
Table B 1 ... 85
Figure 1 ... 101
1 1 SUMMARY
Charlotte Hopster-Iversen:
Influence of mechanical manipulations on the local inflammatory reaction in the equine jejunum and colon
The aim of this research was to characterize the type and degree of intestinal inflammation induced by various mechanical stimuli in the equine ascending colon and jejunum as they could occur during colic surgery.
Median celiotomy was performed under general anaesthesia in 26 horses. In 12 horses, different types of mechanical manipulations were performed and intestinal biopsies collected before the start of manipulations, immediately after, and 30 minutes after the end of manipulations. In the other 14 horses, pelvic flexure biopsies were collected before (7 horses) and after (7 horses) experimental jejunal ischaemia- reperfusion-injury to evaluate the influence of jejunal manipulation on the non- manipulated colonic intestinal layers. Horses were euthanized at the end of both studies while under anaesthesia. Histological evaluation of the intestinal biopsies included analysis of the distribution and accumulation of all intestinal layers with eosinophils and neutrophils.
The mechanical manipulations of the colon caused a redistribution of mucosal neutrophils and eosinophils towards the luminal surface. Furthermore, there was an increased neutrophilic infiltration of the submucosa after serosal and mucosal irritation. All colonic manipulations resulted in serosal infiltration with neutrophils.
Laparotomy and small intestinal manipulation increased mucosal eosinophilic accumulation. The mechanical manipulations of the jejunum resulted in a neutrophilic infiltration of the serosa after all manipulation types and of the submucosa after placement of doyen forceps or mucosal irritation. Manual emptying of the jejunum resulted in neutrophilic infiltration of the circular muscular layer. An increased neutrophilic infiltration of the mucosa was seen after mucosal and after serosal irritation, placement of doyen forceps and after manual emptying of the bowel. An
2
increased eosinophilic infiltration of the submucosa was observed after mechanical irritation of the serosa and after manual emptying of the bowel.
The present studies showed a rapid local inflammatory reaction after mechanical manipulations. These changes could exacerbate existing inflammation or increase morbidity in horses after colic surgery. Colic surgery can lead to intestinal inflammation in non-manipulated intestine and this could contribute to a higher morbidity rate in horses after prolonged colic surgery.
Additionally, the eosinophilic response in the colonic mucosa observed after jejunal manipulation demonstrate a fast reaction of the intestine to distant manipulation. This lead to the recommendation of harvesting intestinal biopsies before any major manipulation to avoid the morphological changes observed to be the result of the manipulation performed and not the primary disorder.
In contrast to ischaemia-reperfusion injury, where the longitudinal muscle layer is the muscle layer mainly affected, the intestinal manipulations performed in the present study resulted in a neutrophilic infiltration of the circular muscle layer. To which degree the inflammatory changes observed in the present study has an effect on the motility of the intestine has to be evaluated in further study. Additionally, the expression of inflammatory mediators would also be an interesting topic of further research studies.
3 2 ZUSAMMENFASSUNG
Charlotte Hopster-Iversen:
Einfluss mechanischer Manipulationen auf die lokale Entzündungsreaktion im Jejunum und Kolon des Pferdes
Ziel der vorliegenden These war die Charakterisierung des Ausmaßes und Typs einer lokalen Entzündungsreaktion im linken dorsalen Kolon und Jejunum des Pferdes nach verschiedenen mechanischen Manipulationen.
Eine mediane Laparotomie wurde bei 26 Pferden in Allgemeinanästhesie durchgeführt. Bei 12 Pferden erfolgten verschiedene mechanische Manipulationen von Jejunum und Kolon wie sie auch während einer Kolikchirurgie vorkommen können. Es wurden Darmbiopsien vor, direkt nach und 30 Minuten nach Ende der Manipulationen entnommen. Die histologische Untersuchung der entnommenen Darmbiopsien beinhaltete die Analyse der Verteilung und Akkumulation eosinophiler und neutrophiler Granulozyten in allen Darmschichten.
Die mechanische Manipulation des Kolons resultierte in eine Umverteilung der eosinophilen und neutrophilen Granulozyten in der Mukosa in Richtung Lumen.
Außerdem ergab sich eine erhöhte Infiltration der Submukosa mit neutrophilen Granulozyten nach Irritation der Serosa und der Mukosa. Alle Manipulationen des Kolons resultierten in einer Infiltration der Serosa mit neutrophilen Granulozyten.
Eine Laparotomie und anschließende experimentelle Ischämie und Reperfusion des Dünndarms resultierte in einer erhöhten Infiltration der Mukosa des Kolons mit eosinophilen Granulozyten. Die mechanische Manipulationen des Jejunums verursachten eine Infiltration der Serosa mit neutrophilen Granulozyten nach allen Arten der Manipulation und der Submukosa nach Applikation der Doyenklemmen und nach Mukosairritation. Eine erhöhte Infiltration der Zirkulärmuskulatur mit neutrophilen Granulozyten konnte nach dem Ausstreichen beobachtet werden.
Zudem kam es zur einer erhöhten Infiltration der Mukosa mit neutrophilen Granulozyten nach allen Manipulationsarten mit der Ausnahme von einer alleinigen Enterotomie. Es zeigte sich eine erhöhte eosinophilen Infiltration der Submukosa
4
nach mechanischer Irritation der Serosa sowie nach dem Ausstreichen.
Zusammenfassend konnten die vorliegenden Studien zeigen, dass eine lokale mechanische Manipulation des Darms zur einen schnellen Entzündungsreaktion führt. Die beobachteten Veränderungen könnten zu einer Verschlimmerung einer Entzündung nach einer Kolikoperation beim Pferd führen und damit auch zu einer erhöhten Morbidität post-operationem. Eine Kolikoperation kann auch am nicht manipulierten Darmteilen zu einer Entzündungsreaktion führen.
Zusätzlich wurde in der hiesige Studie eine schnelle Reaktion der eosinophilen Granulozyten in der Kolonmukosa nach Manipulation des Jejunums beobachtet. Dies sollte bei der Entnahme von Darmbiopsien berücksichtigt werden.
Im Gegensatz zur Ischämie-Reperfusion-Schaden wo hauptsächlich die longitudinale Muskelschicht betroffen ist, war in der vorliegenden Studie die Zirkulärmuskelschicht eher betroffen. In wie weit die vorliegenden entzündlichen Veränderungen ein Einfluss auf die Motilität des Darm haben, bleibt das Ziel zukünftiger Studien.
Zusätzlich wäre die Untersuchung der Expression von Entzündungsmediatoren ebenfalls von Interesse.
5 3 INTRODUCTION
3.1 Colic in horses and its importance
In horses, colic is an important disease and one of the most frequent causes of death (TRAUB-DARGATZ et al. 2001). The annual incidence in the horse population has been reported as 4-10 colic events/100 horses per year (TINKER et al. 1997, KANEENE et al. 1997) with a case fatality rate of 11%, and 1.4 % of colic events resulting in surgery. Beside the fatality rate of colic, the event is also associated with a high economic burden with an estimated annual cost of $115,300,000 in the USA (TRAUB-DARGATZ et al. 2001). In most cases of colic (approx. 80-85%) the horses respond well to medical treatment or resolve spontaneously (WHITE 2009) and obstructing or strangulating diseases requiring surgery only represent 2 to 4 per cent of colic cases (WHITE 1990). Reported short-term survival rate (to discharge) in literature is between 49,4 % and 88% depending on the year of the study and population examined (HUNT et al. 1986, PASCOE et al. 1983, PHILLIPS and WALMSLEY 1993, FUGARO and COTÈ 2001, SEMEVOLOS et al. 2002, FREEMAN et al. 2000). Long-term survival rates are varying between 66 % and 84%
(SEMEVOLOS et al. 2002, MAIR and SMITH 2005c, PHILLIPS and WALMSLEY 1993, FREEMAN et al. 2000).
3.1.1 Complications after colic surgery
Complications after colic surgery are dependent on the involved intestinal segment (MAIR and SMITH 2005a). After small intestinal surgery and possibly anastomosis frequent complications are anastomotic obstructions, postoperative ileus (POI) and adhesions (FREEMAN et al. 2000, VAN DEN BOOM, R. and VAN DER VELDEN 2001). After large intestinal surgery, recurrence of the disorder, endotoxaemia, progression of ischaemia and peritonitis are some of the complications, which can occur after surgery (FREEMAN 2010).
6 3.1.1.1 Postoperative ileus (POI)
The prevalence of POI in horses after colic surgery has been reported as 10 to 19 % (FREEMAN et al. 2000, COHEN et al. 2004) and up to 27% after small intestinal surgery (HOLCOMBE et al. 2009). Risk factors associated with an increased likelihood of developing POI include small intestinal lesions, high PCV at admission, more than 8 L reflux at admission, high heart rate at admission, and increased duration of anaesthesia (COHEN et al. 2004, TORFS et al. 2009, TORFS 2012).
Factors which may reduce the odds of developing POI includes pelvic flexure enterotomy and intraoperative administration of lidocaine (COHEN et al. 2004).
The POI in horses is characterized by gastrointestinal reflux, small intestinal distension, haemoconcentration, tachycardia and abdominal pain (TORFS 2012).
Beside the longer recovery time phase after colic surgery, POI is associated with a higher complication rate, higher financial costs, and in some cases mortality rate (HOLCOMBE et al. 2009). The rate of POI after colic surgery is dependent on the population studied with the higher prevalence after small intestinal lesions with a reported rate of 10% to 33% (FREEMAN et al. 2000, HOLCOMBE et al. 2009, TORFS et al. 2009). A higher rate of POI with 30% is reported after resection of damaged intestine after small intestinal lesions compared to a rate of 15% in horses with small intestinal lesions without resection. The rate of POI after colic surgery without distinguishing between small and large intestinal lesions is reported to be 13.7% (MAIR and SMITH 2005b).
In a study by TORFS et al. (2009) the reported survival rate of horses with POI to discharge after colic surgery because of small intestinal lesions was 34%.
3.1.1.2 Endotoxaemia
In the equine intestinal tract there is a large resident population of gram-negative bacteria. This serves as a reservoir of endotoxin, normally confined to the lumen of the healthy intestine by the protective mucosal barrier (MORRIS 1991). If the intestinal wall is injured as in cases of acute gastrointestinal diseases with mural inflammation or ischaemia (such as volvulus coli), the endotoxin gains access to the
7
circulation. Through the release of lipopolysaccharide (LPS) from gram-negative bacteria and subsequent interaction with the immune system, many endogenous inflammatory mediators such as interleukins, TNF-α and prostaglandins are released into the circulation (MORRIS 1991, SMITH et al. 2005). The presence of endotoxin within the blood is called endotoxaemia (MOORE 2001). Clinical signs of endotoxaemia in the horse are dependent on the stage of the disease. In the early hyperdynamic phase of endotoxaemia, pulmonary hypertension and ileus are present with clinical signs of depression, restlessness, tachycardia, and colic. In the hypodynamic phase of endotoxaemia, the depression and anorexia are accompanied by fever, hyperaemic mucous membranes, prolonged capillary refill time, and hypotension caused by decreased systemic vascular resistance caused by the release of prostaglandins (BOTTOMS et al. 1981, LAVOIE et al. 1990). In severe cases, the endotoxaemia can result in collapse, SIRS, DIC and laminitis (MOORE 2001). A typical laboratory finding is a profound neutropenia with toxic neutrophil morphology (presence of Döhle bodies) and a left shift (BARTON and PERONI 2012).
The clinical management of endotoxaemia includes reduced transfer of endotoxins into the blood circulation by administering smectite orally to absorb bacterial toxins and bacteria and by removing the source of endotoxins (such as the ischaemic intestine). After endotoxins have entered into the circulation, the most effective treatment is the binding and neutralisation of the toxin before it can interact with the host´s receptors sites (MOORE 2001). This can be achieved through the administration of a hyperimmune serum containing anti-endotoxin antibodies or Polymyxin B (DURANDO et al. 1994). If the inflammatory cascade has been activated and clinical signs are present, additionally administration of NSAIDs is indicated to prevent further release of prostaglandins (KELMER 2009). Additionally, supportive fluid therapy to prevent cardiovascular shock as a result of the endotoxaemia is recommended.
8 3.1.1.3 Adhesion formation
A frequent long-term complication after colic surgery is adhesion formation with a reported incidence of clinically relevant adhesions requiring additional surgery or euthanasia as high as 22.1% in a population of horses after small intestinal surgery ( et al. 1990) and of 28.2% in a population of horses after colic surgery (GERHARDS 1990). The risk of adhesions is reported to be higher in juvenile horses after surgery (LUNDIN et al. 1989). However, a study by SANTSCHI et al. (2000) found only 8 % adhesions in foals which recovered after the first laparotomy. The adhesion formation was related to the initial lesion causing colic and the age of the foal at first surgery.
The highest risk of adhesion was found in suckling foals (15 days-6 months). Another study by VATISTAS et al. (1996) identified adhesions in 17% of the foals studied, which recovered from anaesthesia after laparotomy but were subsequently subjected to euthanasia because of recurrent colic.
The relationship between the primary site of the lesion and the site of adhesion formation in adult horses is not always related (GERHARDS 1990, GORVY et al.
2008). In the study of GORVY et al. (2008) a prevalence of adhesion of 32% was found at re-laparotomy with no association between the site of the original lesion or the performance of an intestinal resection at the first surgery and subsequent adhesion formation.
Different techniques are used to prevent adhesion formation including good tissue handling skills during surgery, frequent moistening of the exposed viscera, intra- operative lavage, avoiding exposed suture on the serosal surface of the bowel, minimizing contamination, applying aseptic surgical techniques, minimal tissue trauma, and omentectomy (SOUTHWOOD et al. 1997). Postoperatively, low-dose heparin therapy (PARKER et al. 1987) and appropriate use of antimicrobials and anti-inflammatory drugs have been suggested to prevent adhesion formation (SANTSCHI et al. 2000, SMITH et al. 2005a). The use of new surgical material and techniques such as carboxymethylcellulose, bioresorbable patches and indwelling abdominal drains for peritoneal lavage have been suggested to decrease adhesion formation after laparotomy (SMITH et al. 2005a).
9 3.1.1.4 Other complications
After colic surgery, other common non-fatal complications include jugular vein thrombosis and ventral midline incisional infection (PROUDMAN et al. 2002, FREEMAN 2003). In some case, laminitis also occurs, especially in cases of severe endotoxaemia (FREEMAN 2010). Horses with DIC are predisposed to the development of jugular vein thrombosis (DIAS and NETO 2013).
10 3.2 Postoperative ileus (POI) in the literature
3.2.1 Definition
In human medicine, POI is defined as “a transient cessation of coordinated bowel motility after surgical intervention, which prevents effective transit of intestinal contents and/or tolerance of oral intake” (DELANEY et al. 2006). Clinical symptoms usually include abdominal pain, nausea and vomiting, distension and bloating, delayed passage of flatus and stool, and an inability to progress to an oral diet (BOECKXSTAENS and DE JONGE 2009). The duration of POI correlates with the degree of surgical trauma and is most severe after colonic surgery (DELANEY et al.
2006). Most cases rapidly resolve after few days but in complicated cases resolution may take days to weeks. This is associated with a high economical importance in the health system in human medicine (IYER et al. 2009).
In equine medicine POI has been defined as a nasogastric reflux volume > 20 L during a 24-hour period after surgery or a reflux volume > 8 L at any single sampling time after surgery (ROUSSEL et al. 2001). Clinical signs in horses with POI include abdominal pain, absence of gastrointestinal sounds, absence of defecation, and accumulation of stomach fluid (ROUSSEL et al. 2001). The treatment of POI in horses is cost-intensive with a high morbidity and mortality rate, ranking first among fatal postoperative complications in horses after colic surgery with a reported survival rate to discharge of only 34% (TORFS et al. 2009).
3.2.2 Aetiology
Previous studies in animal models have suggested two major mechanisms involved in the development of POI: neurogenic and inflammatory mechanisms both related to the surgical procedure itself (BOECKXSTAENS and DE JONGE 2009). These two mechanisms probably combine in causing POI with different time frames, considerable overlap, and possible interactions (BAUER et al. 2002). A third component discussed in the development of POI, is the effect of different drugs (e.g.
anaesthetics) on the motility of the intestine after surgery (KEHLET 2008). However,
11
the effects of volatile anaesthetics is short lasting and therefore only of marginal importance. Perioperative use of sedatives as alpha-2-agonists (e.g detomidine) can affect the motility of the gastro-intestinal-tract and gastric emptying (SUTTON et al.
2002). The combined use of ketamine and xylazine for induction of anaesthesia in horses resulted in a complete loss of intestinal motility for 3-6 hours (SINGH et al.
1996). The use of post-operative opioids as pain treatment has a greater impact on the gastro-intestinal motility, with a delayed gastric emptying and prolonged intestinal transit after administration of morphine (BOSCAN et al. 2006) or butorphanol (SELLON et al. 2004).
3.2.2.1 Neurogenic phase
The normal function of the gastrointestinal tract is dependent on the coordination of activating parasympathetic stimuli and inhibition though sympathetic activation.
Through hyperactivation of the autonomic nerve system in the early postoperative phase with an increased sympatheticus activity the motility of the intestine is decreased (MIEDEMA and JOHNSON 2003). The degree of inhibition of motility is dependent on the severity of the stimuli, with only a short-term inhibition of motility after skin incision and laparotomy through activation of a low-threshold adrenergic inhibitory pathway (BOECKXSTAENS and DE JONGE 2009). This probably involves a spinal loop with afferent splanchnic nerves synapsing in the spinal cord and efferents travelling back to the entire intestinal tract, abolishing the motility of the entire gastrointestinal tract (LIVINGSTON and PASSARO 1990). In contrast, more intense stimuli such as handling and manipulation of the intestine lead to an activation of high-threshold supraspinal pathways with a much longer inhibition of intestinal motility (BARQUIST et al. 1996, BOECKXSTAENS et al. 1999). This includes activation of additional pathways actively mediated by the brainstem. After transmission of afferent signals to the brainstem, the signals trigger an increased autonomic output to the neurons of the intermediolateral column of the thoracic cord to the location of sympathetic preganglionic neurons. Activation of these nerves subsequently inhibits the motility of the entire gastro-intestinal tract through the effect of secreted noradrenaline on adrenergic receptors and subsequently inhibition of the
12
migrating motor complexes (MMC) (SAGRADA et al. 1987, BOECKXSTAENS and DE JONGE 2009). Additionally, intense stimulation of the splanchnic afferents triggers an inhibitory non-adrenergic, vagally mediated pathway (THE et al. 2005).
The role of neurogenic activation for the development of POI is important, but is not the only contributing factor as it cannot explain the long-lasting dysfunction of the intestinal motility in some cases (PANTELIS and KALFF 2007).
The “first” or neurogenic phase usually ceased at the end of surgery (BOECKXSTAENS and DE JONGE 2009).
3.2.2.2 Inflammatory phase
In 1978 FIORAMONTI and RUCKEBUSCH observed two phases of inhibition of intestinal activity in dogs and sheep after abdominal surgery. The first phase consisted of complete inhibition of electrical spiking activity during and after surgery, which transiently ceased after the end of surgery. The second phase of inhibition was observed 3-4h after surgery with the duration being dependent on the nature of surgery, with a long-lasting reduction (48-72h) of spiking activity after resection of the small intestine in dogs and sheep (BUENO et al. 1978a). It could be demonstrated, that an inhibitory neural pathway mediated the first phase; however the exact origin of the second phase was unclear.
KALFF et al. (1998) suggested that the timeframe of the second phase of POI correlates with an activation and infiltration of inflammatory cells in the intestinal wall.
Intestinal manipulation resulted in the activation of macrophages and subsequent release of cytokine and chemokine with an influx of leukocytes approx. 3-4h after surgery. Other authors demonstrated an interaction between the immune system, the autonomic nervous system and the muscular function of the gastrointestinal tract (BOECKXSTAENS et al. 2009, DE WINTER and DE MAN 2010, WEHNER et al.
2012).
Discussed mediators involved in the pathogenesis of inflammatory-mediated ileus include nitric oxide (NO) and prostaglandins (DE WINTER and DE MAN 2010). The inducible form of NO synthase (iNOS) has been suggested to mediate LPS-induced motility disturbances in mice models of sepsis-induced ileus (DE WINTER et al.
13
2002, 2005) and in surgically induced POI (KALFF et al. 2000). There was evidence, that the effects of iNOS on motility disturbances in septic induced ileus were partly mediated by NO-mediated oxidative stress mechanisms, by the use of anti-oxidant molecules (DE WINTER et al. 2005). In a study by SCHWARZ et al. (2001) surgical manipulation of the intestine of rats resulted in an expression of COX-2 mRNA and proteins within resident muscularis macrophages, together with an increased prostaglandin level in the peritoneal fluid and in the circulation. The expression of COX-2 mRNA was associated with a decrease of jejunal circular muscle contractility in-vitro and increased gastrointestinal transit time. Both of these could be alleviated pharmacologically by the use of a selective COX-2 inhibitor (DFU (phenyl-2(5H)- furan one)). In humans, KALFF et al. (2003) also demonstrated an expression of COX-2 mRNA in the muscularis externa of jejunal specimens after prolonged abdominal surgery (approx. 3h). In addition to the local effect on gastrointestinal motility, prostaglandins are suggested to modulate afferent nerve signalling from the intestine to the spinal cord and higher brain centres, Thereby modulating sensitivity disturbances and pain signalling pathways (WANG et al. 2005).
Manipulation of small intestine in rodents resulted decreased contractility compared to control specimens, which was accompanied by an accumulation of neutrophilic granulocytes, monocytes and mast cells in the muscularis externa and an activation of resident macrophages (KALFF et al. 1998).
Another study by KALFF et al. (2003) investigated the initiation of an inflammatory response within the human intestinal muscularis externa intraoperatively. Numerous macrophages within the intestinal muscularis externa with an expression of increased lymphocyte immunreactivity were shown after prolonged abdominal surgery (approx.
3h). This was correlated with a time-dependent expression of intestinal cytokines and enzymes (IL-6, IL-1, TNF-α, iNOS and COX-2 mRNAs). In cases of re-laparotomy 24 or 48 h after the first laparotomy, a leukocytic infiltration (neutrophils and monocytes) primarily recruited to the circular muscle layer was observed in the jejunal specimens. This was not identified in the early specimens (30 min after incision) from the first surgery. The muscle strips after re-laparotomy also demonstrated a marked
14
decrease of in-vitro spontaneous activity and response of circular muscle to stimulation with bethanechol.
Type and severity of mechanical manipulations can induce various degrees of intestinal inflammation, as has been documented in rodents (KALFF et al. 1998, 1999 a,b, 2000, SCHWARZ et al. 2001, 2004, WEHNER et al. 2007), mice (THE et al. 2005), pigs (HIKI et al. 2006), and human beings (de JONGE and THE 2004, KALFF et al. 2003, THE et al. 2008).
3.2.3 Treatment of POI
3.2.3.1 Treatment of POI in human medicine
The concept for treatment of POI is based on a multimodal postoperative rehabilitation together with techniques to reduce the occurrence and/or duration of POI (KEHLET 2008).
3.2.3.1.1 Factors enchaining recovery after surgery:
Preventive techniques
The use of thoracic epidural local anaesthetics the first 2-3 days after laparoscopic sigmoid resection or colonic resection through an epidural catheter has been shown to reduce the duration of POI, and the use of analgesics (NEUDECKER et al. 1999, KEHLET and KENNEDY 2006). The use of epidural analgesia was more important after open surgery than after laparoscopy. Whenever possible the laparoscopic approach to major abdominal surgery is recommended because of the important beneficial effects as reduced pain and minimizing the inflammatory response. This also should reduce the incidence and duration of POI (KEHLET and KENNEDY 2006).
Prokinetic drugs and laxatives
TRAUT et al. (2008) performed an analysis of randomized controlled parallel-group trials comparing the effects and efficacy of different systemic acting pro-kinetic drugs for the treatment of adynamic postoperative ileus in adult patients. The positive
15
effects of alvimopan, a novel peripheral μ-receptor antagonist, were supported in 6 trials. However, the drug is still in an investigational stage. The use of erythromycin, an antibiotic and a motilin agonist, was not beneficial. Metoclopramide, a dopamine antagonist and cholinergic agonist, had inconsistent effects in most trials. Side effects include drowsiness, dystonic reactions, and agitation. Cisapride showed modest prokinetic effects, however, it has been withdrawn from the market due to adverse cardiac events in many countries. The evidence was insufficient to recommend the use of cholecystokinin-like drugs, dopamine-antagonists, propranolol or vasopressin. The use of intravenous lidocaine might provide a potential prokinetic effect, but more evidence on clinically relevant outcomes was needed (TRAUT et al.
2008). SUN et al. (2012) suggested lidocaine to be a useful adjunct for postoperative pain management by decreasing postoperative pain intensity, reducing opioid consumption, improving gastrointestinal function, and shortening length of hospital stay after performing a meta-analysis on randomized controlled clinical trial studies.
The use of laxatives has not been widely studied in randomised, prospective trials and is not part of the common fast-track treatments (LUBAWSKI and SACLARIDES 2008).
Laparoscopic surgery
The use of laparoscopy has the advantage of limiting tissue trauma. It has been suggested that laparoscopic surgery causes a lesser degree of mast cell activation and inflammation, and thereby limit the degree and duration of POI (THE et al. 2008, AUGESTAD and DELANEY 2010). Potential advantages of laparoscopy versus conventional open surgery include smaller incisions, earlier recovery of the gastrointestinal tract, shorter hospital stay, and less pain (DELANEY et al. 2003).
Early postoperative feeding
Early oral feeding is recommended after major abdominal surgery to attenuate catabolism and potentially decrease infectious complications (ANDERSEN et al.
2006).
16
Postoperative nausea and vomiting (PONV) prophylaxis
The use of serotonin 5-HT3 receptor antagonists, droperidol and glucocorticoids for postoperative nausea and vomiting prophylaxis has been supported in several studies. Untreated, PONV occurs in 20-30% of the general surgical population and are related with adverse effects such as patient discomfort to the point of increased post-operative morbidity (GAN et al. 2007).
Chewing gum
The use of chewing gum in postoperative patients has been reported to reduce the duration of POI. The mechanism of action is probably stimulation of oral and gastrointestinal major reflexes (CHAN and LAW 2007).
Opioid-sparing analgesia
The use of perioperative opioids for pain management has been reported to result in prolonged POI (KEHLET 2008). This finding has led to the development of multimodal non-opioid analgetic strategies. NSAIDs have been suggested to reduce opioid requirements and the occurrence of PONV (MARRET et al. 2005). The use of lidocaine as an anaesthetic-sparing drug during anaesthesia and as an additional analgetic during the post-operative phase has been analysed in humans. KABA et al.
(2007) demonstrated, using a randomized, controlled, double-blind design, that infusion of intravenous lidocaine during surgery and for 24h post surgery could produce effective analgesia after laparoscopic colectomy, improved postoperative fatigue, bowel function, and allowed a more rapid rehabilitation and quicker hospital discharge.
3.2.3.1.2 Factors contributing to delayed recovery from POI Nasogastric tubes
In the classic perioperative care after major abdominal surgery in humans, a prophylactic nasogastric tube has been inserted to reduce gastric retention, nausea, vomiting and POI. However, systemic reviews and trials have demonstrated the ineffectiveness of a nasogastric tube to achieve these goals, and in contrast an
17
increased pulmonary morbidity was associated with the use of the nasogastric tube (NELSON et al. 2005).
3.2.3.2 Treatment of POI in equine medicine
In equine medicine, supportive therapy is the most important treatment for POI because of the therapeutic limitations of available prokinetic agents. As in human medicine, there is discussion about nasogastric intubation and possible negative effects on the gastrointestinal motility. Therefore it has been suggested to use the nasogastric tube only when indicated by signs of pain, an elevated heart rate, or an elevated respiratory rate (HARDY and RAKESTRAW 2002).
Use of prokinetic drugs
Prokinetic agents are pharmaceutical substances that promote gastrointestinal transit time through the release of an excitatory neurotransmitter, stimulating the release of a gastrointestinal hormone or supressing the inhibitory effect of a biologically active substance (LONGO and VERNAVA 1993).
Parasympathomimetics
Bethanechol is an acetylcholinreceptor agonist and induces contraction of smooth muscle cells (SAUNDERS et al. 1997, MARTI et al. 2005). Reported effects in healthy horses include improved gastric emptying time and increased myoelectric activity of ileum, caecum and right ventral colon (LESTER et al. 1998). However, in an equine model of postoperative ileus there was only limited effectiveness in motility enhancement (GERRING and HUNT 1986).
The inhibition of the acetylcholinesterase by neostigmine can be expected to enhance gastrointestinal motility through stimulation of the parasympathetic nervous system. On the contrary, neostigmine in horses has been demonstrated to delay gastric emptying (ADAMS and MACHARG 1985) and is therefore suggested to be contraindicated in horses with POI. However, a more recent study by NIETO et al.
(2013) in healthy horses evaluated the effects of a constant-rate-infusion (CRI) of neostigmine. The authors observed an increased faecal production and urinary
18
frequency during CRI of neostigmine without decreasing gastric emptying time. An in- vitro evaluation of jejunum and pelvic flexure strips demonstrated a stimulation of the contractile activity of the smooth muscle by neostigmine.
Adrenergic antagonists and beta-receptor antagonists
Adrenergic antagonists block the inhibitory effects of the sympathetic nervous system on the gastrointestinal tract (SUPRENANT 1994). Especially in the early phase of POI, the stimulation of the sympathetic system leads to a decrease of motility (MIEDEMA and JOHNSON 2003). Yohimbin acts as an α2-adrenergic antagonist and has been shown to prevent the delay in gastric emptying after endotoxin administration in horses (MEISLER et al. 2000). However, the clinical use of yohimbin for treatment of POI has not been evaluated in horses.
Propranolol is a β-adrenergic antagonist and has been suggested to enhance intestinal motility. In an experimental model of POI, propranolol did not alter the delayed transit of spheres through the gastrointestinal tract of horses (GERRING and HUNT 1986).
Benzamides
Metoclopramide hydrochloride is a first-generation benzamide, which acts as dopamine 1 and 2 receptor antagonist and 5-HT4-receptor agonist and 5-HT3
antagonist (ALBIBI and McCALLUM 1983). In horses, metoclopramide has been reported to restore coordination of gastric and small intestinal activity in an experimental model of POI (GERRING and HUNT 1986). Side effects observed in horses include transient reversible excitement and in humans drowsiness and anxiety (PINDER et al. 1976). In a study by DART et al. (1996) the CRI of metoclopramide in horses following small intestinal surgery significantly reduced the total volume of reflux, the duration of POI, and hospitalisation when compared to horses not receiving metoclopramide or only intermittent infusion of metoclopramide.
Finally, the authors concluded that metoclopramide given as a continuous intravenous infusion decreased the incidence and severity of ileus following small intestinal resection and anastomosis in horses.
19
Cisapride has been suggested to function as an indirect cholinergic stimulant and thereby increase the release of acetylcholine. It is a 5-HT4 receptor agonist and 5-HT3
receptor antagonist (WISEMANN and FAULDS 1994). In equine medicine, cisapride has been suggested to reduce the incidence of POI (GERRING and KING 1989, DE GEEST et al. 1991, VELDEN and KLEIN 1993). In human medicine, cisapride has been reported to cause cardiac side effects as fatal arrhythmias (CARLSON et al.
1997). Reported side effects in horses include mild colic, increased frequency of faecal passage, increased bowel sounds, and a dose-related increase in heart rate (WONG et al. 2011).
Motilin Agonists
Erythromycin is an antibiotic, which acts as motilin receptor agonist in the intestine with a profound influence on gastroduodenal motor activity (CATNACH and FAIRCLOUGH 1992). An in-vitro study on jejunal strips from the horse demonstrated an increased smooth muscle contractile amplitude after administration of erythromycin (NIETO et al. 2000). An increase of gastric emptying after administration of 0.1 to 1.0 mg/kg bwt i.v. was observed in a study by RINGGER et al. (1996) in healthy horses. LESTER et al. (1998) demonstrated a more rapid caecal emptying in horses receiving erythromycin compared to a placebo infusion of saline.
A survey of use of prokinetic drugs by diplomates of the ACVS in horses with gastrointestinal injury by VAN HOOGMOED et al. (2004) identified erythromycin as the second frequent pro-kinetic drug used after lidocaine. Side effects observed in horses treated with erythromycin include colic and colitis (GUSTAFSSON et al.
1997). In Germany, erythromycin is listed as an antibiotic and its use is regulated by the guidelines for use of antibiotics. Additionally, erythromycin is not allowed to be used in horses, intended for use as food animals. In human medicine, the long-term use of erythromycin however is limited by its anti-bacterial action and desensitization to the therapeutic effects of the drug by change of the motilin receptor (GALLIGAN and VANNER 2005).
20 Sodium channel blockers
The local anaesthetic agent lidocaine infused into the peritoneal cavity or administered i.v. in humans has been shown to reduce the duration of POI (RIMBÄCK et al. 1990). Lidocaine may act by reducing the level of circulating catecholamines and through inhibition of the sympathoadrenal response.
Additionally, lidocaine can suppress the activity of the primary afferent neurons involved in reflex inhibition of gut motility (ROBERTSON et al. 2005). Lidocaine also decreases inflammation in the bowel wall through inhibition of prostaglandin synthesis, inhibition of granulocyte migration and the release of lysosomal enzymes by these cells and inhibition of free radical production (HARDY and RAKESTRAW 2002, SEDDIGHI 2010).
In horses, an in-vitro study on intestinal specimen demonstrated an increased contractility in proximal duodenum smooth muscle strips after administration of lidocaine (NIETO et al. 2000). Another study compared the effects of lidocaine infusion on ischaemia-reperfusion injured jejunal muscle strips in-vitro compared to control samples and lidocaine resulted in an increase of contractility of the circular muscle in both groups, but more pronounced in the injured jejunal samples (GUSCHLBAUER et al. 2010). Several studies evaluated the effects of lidocaine infusion in postoperative colic patients on the occurrence of POI (BRIANCEAU et al.
2002, COHEN et al. 2004, MALONE et al. 2006, HOLCOMBE et al. 2009, TORFS et al. 2009). There was a positive influence of lidocaine on outcome and rate of POI in horses receiving lidocaine in some of the studies. On the other side, studies in healthy horses failed to demonstrate a clinical effect of lidocaine with no effect on myoelectrical activity (MILLIGAN et al. 2007, RUSIECKI et al. 2008). Side effects of lidocaine administration include muscle fasciculations, trembling, and ataxia (RAKESTRAW 1998).
Symptomatic therapy
Supportive therapy aimed at restoring fluid and electrolyte balance should be provided. In addition, analgesia and limiting of bacteraemia and endotoxaemia is essential in horses with POI (SMITH et al. 2005a).
21
3.3 Physiology of normal equine intestinal motility
In the intestinal tract the cells of Cajal (ICC) are non-neural cells of mesenchymal origin responsible for generating electrical activity (SANDERS et al. 2006). The cyclic changes in membrane electrical potential initiated by the ICC are called “slow waves”
or “pacemaker” potential. The slow waves spread passively into longitudinal and circular muscle cells through the gap junctions between ICC and the smooth muscle cells. The slow waves are initiated orally and propagate aborally (RAKESTRAW 1998). The depolarization of the smooth muscle cells by the slow waves is sub threshold, as they do not depolarize the cells sufficiently to generate an action potential (OLSSON and HOLMGREN 2001). For generation of an action potential additional depolarizing or excitatory stimulation from the enteric or autonomic nervous system is necessary. These additional depolarizing stimulations are called
“spike potentials” and are usually superimposed on the slow waves (NAVARRE and ROUSSEL 1996).
The activity of the intestinal tract is not constant but changes between periods of quiescence and periods of spiking activity. This is called migrating myoelectric complex (MMC) (WONG et al. 2011). There exist 4 phases of the MMC (LESTER et al. 1992). Phase 1 is a period without spike potentials and no contractions of the intestine. Phase 2 is characterized by intermittent spike potentials and in the horse it is associated with propagation of ingesta (GERRING and HUNT 1986). Phase 3 contains regular spiking activity and is associated with propulsion of ingesta. Phase 4 is associated with a rapidly diminishing contractile activity. Each phase migrates along the stomach and small intestine. MMCs are not evident in the caecum and large intestine, where short spike bursts occur during mixing and long spike bursts occur during propulsion of ingesta (MERRITT et al. 1995, KOENIG and COLE 2006).
22 3.4 Intestinal homeostasis and inflammation
The intestinal tract is populated by a huge population of microbes (DALY et al. 2001).
The ability of the immune system to coexist and coevolve with these microorganisms during life is essential for both sides. Failure to achieve or maintain an equilibrium between host and microorganisms has negative consequences for the intestinal health (CHERVONSKI 2012). To maintain a barrier between the dense population of microorganism and blood circulation, the mucosal barrier is essential (FARHADI et al. 2003). The essential part of the mucosal barrier are the enterocytes. Additionally, the mucosal barrier is lined with goblet cells, Paneth cells, M cells, and enteroendocrine cells. The regulation of the junctional integrity and paracellular permeability is of special importance for the immune system, as pathogenic viruses, bacteria, and parasites exploit opportunities for breaching the epithelial barrier by entering through tight junctions (GARRETT et al. 2010). The release of inflammatory mediators is an important part of the host mucosal defence. Inflammatory mediators can accelerate epithelial replacement as a host defence mechanism. Beside the mucosal barrier function, the cellular level of the intestine with intestinal epithelial cells, myofibroblasts, stromal cells, and T and B cells also participate in a network of interactions regulating the inflammatory response of the intestinal tract (GARRETT et al. 2010).
In healthy horses, identified resident immune cells of the jejunal mucosa include plasma cells, eosinophilic granulocytes, macrophages, and B cells (PACKER et al.
2005). In the acute inflammatory reaction, cells of the innate immune system are the main initiators. After injury, resident macrophages recognize damage-associated signals, invading microbes, and toxins. As a reaction to these signals, macrophages attract large numbers of neutrophils to the site of injury by release and initiation of chemoattractans such as IL-8, TNF, and LTB4 (SMITH et al. 2005b). Additionally, eosinophils and mast cells are potent immunomodulatory cells, activated by similar cells. After activation of mast cells, these release toxic metabolites, initiate inflammation, and attract other immune cells to the site of injury (LUSTER et al.
2005). Frequently, eosinophils are found together with activated mast cells.
23 3.5 Eosinophilic granulocytes
3.5.1 Origin, development, and localization of the eosinophilic granulocyte The eosinophilic granulocyte is a minor component of the peripheral blood leukocyte population, usually compromising only 0% to 3% of the total leukocyte count in horses (CARRICK and BEGG 2008). Most of the eosinophils are located in tissues, primarily the skin, gastrointestinal tract, and lungs (GIEMBYCZ and LINDSAY 1999).
Eosinophils have a bilobed nucleus and granular cytoplasm. The eosinophils are derived from bone-marrow CD34+ haemopoetic progeniator cells. The lineage selection and maturation is dependent on IL-5, IL-3 and GM-CSF, with IL-5 being most specific for eosinophil selection (ROTHENBERG and HOGAN 2006). Bone marrow production of eosinophils usually takes 2 to 6 days in horses (CARRICK and BEGG 2008), and the release from the bone marrow occurs in response to IL-5 and eotaxin (ROTHENBERG 1999). After release from the bone marrow, eosinophils remain in the circulation for a few days before they traffic into specific tissues, predominantly the gastrointestinal tract (CARRICK and BEGG 2008). The regulation of eosinophil development, migration, and effectors function is mediated by IL-3, IL-5, and GM-CSF. IL-1, IL-4, IL-13, and TNF-α regulate eosinophil trafficking by promoting adhesive interactions with the endothelium. The eosinophil trafficking is promoted by chemoattraction through IL-5, chemokines and cysteinyl leukotriene (NAGATA and SAITO 2003, ROTHENBERG and HOGAN 2006). Most specific for eosinophils are IL-5 and eotaxin (CONROY and WILLIAMS 2001). The regulation of eosinophil accumulation is regulated by the mediator eotaxin 1-3, which act on the CCR3 receptor (BENARAFA et al. 2000, 2002). Eotaxin-1 promotes eosinophil accumulation in the gastrointestinal tract (MATTHEWS et al. 1998). Eotaxin-1 is critically involved in regulation of the baseline homing of eosinophils into the gastrointestinal tract in mice (MISCHRA et al. 1999). Additionally, eotaxin-1 has been shown to have an essential role in regulating eosinophil-associated gastrointestinal diseases (HOGAN et al. 2000). Other preferred tissue localizations of eosinophils include thymus, haematopoietic organs, and mammary gland (during pubertal
24
development) (JACOBSEN et al. 2007). Homing of eosinophils into the intestinal tract occurs prenatal and seems to be independent of the presence of intestinal parasites (MISHRA et al. 1999, ROTHENBERG 2004).
3.5.2 Function of the eosinophilic granulocytes
Eosinophilic granulocytes can modulate immune responses through a multitude of mechanisms (WOERLY et al. 1999). Eosinophils can act as antigen-presenting cells with expression of co-stimulatory molecules, secretion of cytokines capable of stimulating lymphocytes (IL-2, IL-4, Il-6, IL-10, IL-12), (KITA 1996) and expression of MHC class-II molecules (LUCEY et al. 1989). Furthermore, eosinophils can have pro- inflammatory effects by releasing a multitude of cytokines (IL-2, IL-4, IL-5, IL-10, IL- 12, IL-16, and TGF-α, β) and lipid mediators (Table 1) as such PAF and leukotriene C4 (ROTHENBERG and HOGAN 2006).
The granular cytoplasm of eosinophils is toxic to a variety of tissues. The granules contain a crystalloid core composed of major basic protein (MBP) 1 and 2, and a matrix composed of eosinophil cationic protein (ECP), eosinophil-derived neurotoxin (EDN), and eosinophil peroxidase (EPO). MBP, EPO and ECP have cytotoxic effects on the epithelium (ROTHENBERG 2004). ECP can insert voltage-insensitive toxic pores into the membranes of target cells, which may facilitate the entry of other toxic molecules (YOUNG et al. 1986). MBP trigger degranulation of mast cells and basophils, and directly increases smooth muscle reactivity (JACOBY et al. 1993).
25 Table 1
Product Example Biological function
Granules
Eosinophil peroxidase (EPO)
Production of reactive oxygen species, cytotoxic, release of histamine from mast cells
Eosinophil cationic protein (ECP)
Modification of collagenmatrix Eosinophil-derived
neurotoxin (EDN) Neurotoxic activity
Major basic protein (MBP)
Toxic for parasites, cytotoxic, histamine release from mast cells
Cytokines IL-3, IL-5, GM-CSF
Increase production of eosinophils from the bone marrow
Activation of eosinophils
Chemokine IL-8 Chemotaxis of neutrophils
Lipid mediators
Leukotriene C4
Contraction of smooth muscle cells
Leukotriene D4
Increase vascular permeability Leukotriene E4
Increased mucus production
Platelet-activating-factor (PAF)
Chemotaxis of leukocytes, increased production of lipid mediators
Activation of eosinophils, neutrophils, platelet
Table 1 Examples of mediators produced by eosinophils and their biological effects.
26
Eosinophils can be triggered by a variety of cytokines, immunoglobulins and complement factors. This can lead to the generation of a wide range of inflammatory cytokines (JACOBSEN et al.2007).
Circulating eosinophilic granulocytes are capable of activating mast cells through the release of mediators such as eosinophil cationic protein and eosinophil peroxidase (ZHEUTLIN et al. 1984) and synthesis of nerve growth factor, a mast cell survival and activating factor (KOBAYSHI et al. 2002). The activation of the mast cells leads to release of histamine and further inflammatory reactions. The mast cell activation suggests that eosinophils contribute to the perpetuation of allergic inflammation (FOSTER et al. 1996). Cytokines expressed by eosinophils such as IL-2, IL-4, and IL-12 can induce the proliferation, and/or maturation of T cells. Additionally, eosinophils can also present antigens to T cells (LUCEY et al. 1989). Attraction of other inflammatory cells such as neutrophils can be initiated by the release of IL-8 or PAF by eosinophils (ROTHENBERG 2004).
3.5.2.1 Eosinophilic granulocytes and organ development
The expression of eotaxin-1 and the subsequent homing of eosinophils into the gastrointestinal tract during gestational development may suggest that eosinophils have a role in tissue or organ development (MISHRA et al.1999).
3.5.2.2 Function of the eosinophilic granulocytes in the gastrointestinal tract
Eosinophilic granulocytes are part of the resident cell population of the intestinal submucosa and mucosa of horses. Lowest numbers of eosinophils are found in the stomach, and numbers increase from there to the caecum. From the caecum, the numbers of eosinophils slowly decrease from the right ventral colon, left ventral colon, pelvic flexure, left dorsal colon, and right dorsal colon to the small colon (RÖTTING et al. 2008b). Most of the eosinophilic mucosal granulocytes are located in the lamina propria close to the lamina muscularis mucosae.
27 3.5.2.3 Eosinophils and parasitism
Eosinophils in the gastrointestinal tract are essential for defending the host against parasites. This function is based on the ability of the eosinophils to mediate an antibody-dependant cellular toxicity against helminths (BUTTERWORTH 1977).
Eosinophils respond to helminth infection with an increased accumulation in the mucosa and intra-epithelially (COLLOBERT-LAUGIER et al. 2002), and eosinophils aggregate and degranulate in the presence of helminths (BEHM and OVINGTON 2000).
The adjacent location of eosinophils and lymphocytes in the gastrointestinal tract suggests that a regulation of lymphocytes by the eosinophils through the release of cytokines seems possible (ROTHENBERG 2004).
3.5.3 Eosinophilic granulocytes and diseases
Through the multitude of mediators expressed by eosinophils, the eosinophil function can be beneficial (killing parasites and viruses) or detrimental (causing chronic inflammation) (JACOBSEN et al. 2007). Prominent blood and tissue eosinophilia is manifested in a number of inflammatory states (ROTHENBERG and HOGAN 2006).
3.5.3.1 Gastrointestinal tract diseases involving eosinophilic granulocytes
In human medicine as well as in veterinary medicine inflammatory bowel disease (IBD) is a known disease complex, which includes several diseases (HANAUER 2006). IBD are characterized through a histological visible inflammation of the gut (CRAVEN et al. 2004). The diseases manifest with persistent or recurrent gastro- intestinal symptoms.
In numerous gastrointestinal disorders eosinophil accumulation is present (GUARJARDO et al. 2002). Examples include IgE-mediated food allergy, eosinophilic gastroenteritis, allergic colitis, eosinophilic esophagitis, inflammatory bowel disease, and gastro oesophageal reflux disease (ROTHENBERG 2004). In IBD, eosinophils only represent a small percentage of the leukocytic infiltration, but are proposed to be
28
negative prognostic indicator, depending on their localization (NISHITANI et al.
1998).
Primary eosinophilic gastrointestinal disorders (EGID) are a particular form of IBD in humans defined as disorders primarily affecting the gastrointestinal tract with eosinophil-rich inflammation in the absence of a known cause for eosinophilia. The disease complex includes eosinophilic gastritis, eosinophilic gastroenteritis and eosinophilic colitis (ROTHENBERG 2004). Similar for all those manifestations is an increase in eosinophil granulocyte concentration in the region of the lamina propria.
Symptoms in patients with EGID include failure to thrive, abdominal pain, irritability, gastric dysmotility, vomiting, diarrhoea, and dysphagia (GUAJARDO et al. 2002).
The treatment for EGID includes trial of specific food antigen and allergen avoidance, glucocorticoids in acute exacerbations, and neutralisation of gastric acidity. Beside the traditional treatments, immunmodulatory agents (azathioprine, 6-mercaptopurine) have been suggested to down regulate or inhibit eosinophil growth factors, reducing eosinophilic infiltration, and improving symptoms and are therefore recommended in severe, refractory, or steroid-dependent eosinophilic colitis (ALFADDA et al. 2011).
Inflammatory bowel disease is also known in dogs as a cause of persistent gastro- intestinal disease (CRAVEN et al. 2004, FUCHS 2007). IBD of the small intestine includes lymphocytic-plasmacytic enteritis (LPE), eosinophilic enteritis, and eosinophilic gastro-enteritis (EGE). In the large intestine, four main conditions are recognized as IBD: lymphocytic-plasmacytic colitis, eosinophilic colitis, histiocytic ulcerative colitis, and regional granulomatous colitis (GERMAN et al. 2003).
Frequently, the diseases involve both small and large intestine, or even include the stomach (CERQUETELLA et al. 2010).
In dogs, the small and large intestine can be affected, whereas in cats the small intestine is predominantly affected.
In equine medicine, eosinophilic intestinal conditions without known causes for the eosinophilia have been grouped within the IBD complex (PASS and BOLTON 1982, KEMPER et al. 2000). Eosinophilic conditions grouped within IBD include
29
multisystemic eosinophilic epitheliotrophic disease (MEED), diffuse eosinophilic enteritis (DEE), and idiopathic focale eosinophilic enteritis (IFEE), depending on the localization and organ involvement (MAKINEN et al. 2008). DEE and IFEE have also been grouped under the heading idiopathic eosinophilic enterocolitis (SCHUMACHER et al. 2000). Clinical signs of IBD in horses include weight loss, often despite ravenous appetite, oedema, diarrhoea, signs of abdominal pain, and lethargy (SCHUMACHER et al. 2000). The cause of the disease remains unclear;
however, an immune-mediated reaction to a dietary allergy or parasitic infection has been discussed.
Focal granulomatous enteritis with eosinophil contribution has been described after Pythium infection in the USA. The core of the lesion consists of degenerating or necrotic eosinophilic debris with a marked granulomatous reaction around the cores.
The usual localization of the lesions is subcutaneously, but case reports describe the formation of jejunal masses with partial or complete obstruction of the lumen (MORTON et al. 1991).
Idiopathic focal eosinophilic enteritis is characterized by single or multiple, focal, circumferential or plaque-like wall thickening and/or constrictions, mainly in the small intestine (MAKINEN et al. 2008). Possible aetiologies discussed include hypersensitivity, changes in management, dietary factors, and death of migrating parasites (SOUTHWOOD et al. 2000, ARCHER et al. 2006). The lesions are characterized by an accumulation of leucocytes in the tunica submucosa and the tunica muscularis, mainly eosinophils and macrophages (MAKINEN et al. 2008).
Clinical signs in the described cases were mainly colic.
Diffuse eosinophilic enteritis (DEE) is also considered a primary eosinophilic intestinal condition and is characterized by a diffuse mucosal and submucosal inflammatory infiltration in the small and/or large intestine. MAKINEN et al. (2008) described 5 cases of DEE (4 horses small intestine, 1 horse colon), histological characterized by a moderate to severe inflammation of the submucosa dominated by eosinophils. The mucosal inflammation was characterized by mainly lymphocytic infiltration, followed in number by macrophages, eosinophils, and plasma cells. The
30
tunica muscularis exhibited a mild multifocal interstitial lymphocyte-dominated inflammatory infiltration.
3.5.3.2 Other diseases involving eosinophilic granulocytes in horses
In contrast to IFEE and DEE, multisystemic eosinophilic epitheliotropic disease is characterized by eosinophilic infiltration not only affecting the gastro-intestinal tract, but also the skin, liver, pancreas, and respiratory system (WILKIE et al. 1985). MEED is an uncommon disease mainly affecting young horses. Clinical signs include chronic weight loss, diarrhoea, exfoliative dermatitis, and respiratory symptoms. The diagnosis is made by histological examination of affected tissue. There only exist few reports on cases in equine medicine, mostly with a poor prognosis for survival (WILKIE et al. 1985, SANFORD 1989, La PERLE et al. 1998, BOSSELER et al.
2013). McCUE et al. (2003) described a case of MEED in a 4-year old paint mare, which was successfully treated with high dose dexamethasone, trimethoprim- sulfadimethoxazole against secondary infections, and hydroxyzine hydrochloride as an anti-histamine. The cause of MEED in horses is unknown. Suggested aetiologies include allergic, toxic, viral, and parasitic causes. In humans, premalignant or malignant expansion of type 2 helper cells is commonly found in cases of hypereosinophilic syndrome (ROUFOSSE et al. 2000). In horses, there also exists a report of concurrent intestinal T-cell lymphosarcoma with MEED (La PERLE et al.
1998).
In humans, eosinophilic granulocytes are the predominant airway inflammatory cells in chronic asthma. In horses, pulmonary eosinophilia has been identified in bronchoalveolar lavage (BAL) from young racehorses with inflammatory airway disease (IAD) (MOORE et al. 1995a, HARE and VIEL 1998). In addition, eosinophilia in tracheobronchial secretion (TBS) or BAL has also been related to infection with Dictyocaulus arnfieldi, the lungworm of horses (MACKAY and URQUHART 1979).
Another manifestation of eosinophilic infiltration is in skin disorders. Insect bite hypersensitivity (sweet itch) in horses is a seasonally recurrent, pruritic skin disorder
