Sepsis-related laminitis


Sepsis, according to Wikipedia, is ‘a potentially deadly medical condition that is characterised by a whole-body inflammatory state (called a systemic inflammatory response syndrome or SIRS) and the presence of a known or suspected infection …’ Whereas, in man, the systemic inflammatory response results in failure of organs such as the liver and lung, the digital lamellae are a primary target in the horse suffering from sepsis. Sepsis-associated organ failure in man and laminitis in the horse arise as late sequelae to microbial diseases following other therapeutic interventions. In this sense the conditions are somewhat ‘man made’. Only domestication of the horse and modern medicine for both species has resulted in prolonging life of the diseased patient to the point that these sequelae take such an important place in medicine. Thus, working to discover effective therapies for disease entities for which the body has not evolved defences puts the researcher and clinician at a great disadvantage, probably playing a role in the failure of discovery of effective therapies for organ or lamellar failure due to sepsis.

Although not as widely reported as pasture-associated laminitis, sepsis-related laminitis still takes a significant toll on equine patients. These cases include the horse with pleuropneumonia, septic endometritis, and most commonly, gastrointestinal injury from insults ranging from caecal mucosal sloughing in horses with accidental exposure to excessive starch, to large or small intestinal injury from infectious types of enterocolitis to strangulating injuries such as torsion of the large colon.

Laminitis related to equine sepsis has enjoyed the most attention of any type of laminitis from researchers over the past few decades, primarily because the most commonly used models to study laminitis are models of equine sepsis. In the carbohydrate-overload models, including the traditional corn starch/wood flour and the more recent oligofructose models [1, 2], the excess starch that arrives in the large intestine is digested by intestinal bacteria resulting in a severe drop in intestinal pH, an ensuing die off of Gram-negative organisms, and a moderate to severe enterocolitis in which injury to the mucosal barrier results in the absorption of numerous substances including bacterial toxins [3, 4]. Endotoxaemia, a hallmark of sepsis, has been detected (in jugular blood) in both traditional carbohydrate- and oligofructose-overload models, as have the classic signs of SIRS associated with sepsis including increased heart rate and fever. Although infusions of endotoxin have resulted in detection of digital pain in horses [5], lamellar failure as occurs in clinical- and carbohydrate-overload models has not been reported with endotoxin administration. This has resulted over the years in investigators attempting to discover other ‘trigger factors’, which may be absorbed from the compromised gastrointestinal tract and lead to lamellar injury. Although other factors in addition to endotoxin including other bacterial toxins (i.e. flagellin) [6], and cellular proteins released from dying host cells can cause or exacerbate systemic inflammation in sepsis [7], the reality that, in the clinical equine patient, a diverse array of septic disease entities not involving the gastrointestinal tract (i.e. acute septic endometritis, pleuropneumonia) can lead to laminitis indicates that another gastrointestinal trigger factor outside the realm of bacterial components/toxins may play a role if present but is not necessary for induction of laminitis. The other experimental model that is likely to mirror sepsis-related laminitis, due to its rapid induction of a systemic inflammatory response, is the black walnut extract model [8-10]. Like the carbohydrate-overload models, the black walnut extract model was designed after it was discovered that clinical cases of laminitis were occurring due to horses being bedded on shavings from the heartwood of black walnut trees. Although the clinical cases exposed for several days to black walnut shavings can undergo severe lamellar injury and displacement of the distal phalanx, the model is associated with a mild, transient laminitis, probably due to the reality that a single exposure to the extract is given (similar to the difference of an endotoxin bolus vs. a long-term endotoxin infusion).

Similar to organ failure research in human sepsis, the main focus of laminitis research for several decades was based on the dogma that decreased blood flow was responsible for lamellar injury. Multiple papers emanating from studies of lamellar vasculature and blood flow suggest that lamellar haemodynamics are likely to be involved in lamellar pathophysiology in laminitis [11, 12], but may not play the primary role in the early stages of the disease. Again similar to human sepsis, the failure of most drugs aimed at improving blood flow to withstand scientific rigour or improve clinical outcomes provided incentive for researchers to address other pathophysiological mechanisms in an attempt to establish effective therapies [13, 14]. In the past decade, the 2 research focuses that have received the most attention of investigators studying sepsis-related laminitis have been: 1) inflammatory events; and 2) events at the interface of the lamellar basal epithelial cell (LBEC) layer and the underlying matrix molecules of the basement membrane (and closely attached lamellar dermis). Inflammatory events have been investigated due to the documented role of inflammation in organ injury in human sepsis [15], and events regarding adhesion of the epithelial cells to the underlying matrix have come to the forefront as this interface was documented histologically to be the point of structural lamellar failure in laminitis [16].

An in-depth assessment of lamellar inflammation was first performed in the black walnut extract model of laminitis, where increased mRNA concentrations of inflammatory mediators were reported in the lamellae at both developmental time points and at the onset of clinical signs of lameness [17, 18]. Further investigation into the black walnut extract model delineated an extremely rapid lamellar inflammatory response, including the activation of lamellar endothelial cells [9], lamellar leucocyte infiltration [19, 20], and marked increases in lamellar cytokine gene expression [9, 21], with these changes occurring as early as 1.5 hours after receiving the bolus of black walnut extract via nasogastric tube [9]. As investigators have begun to focus more on the carbohydrate-overload models of laminitis, this virtual issue includes 2 recent investigations on lamellar inflammatory events in the traditional starch and oligofructose models of laminitis. These papers depict a similar fulminant inflammatory event occurring in the lamellae in the carbohydrate-overload model [22] as reported in the black walnut extract model. However, the onset of inflammation appears to be delayed when compared to the black walnut extract model, with the majority of inflammatory events including endothelial activation, leucocyte infiltration [23], and increased mRNA concentrations of pro-inflammatory cytokines and COX-2 expression not increasing until the onset of clinical signs of laminitis [22]. Interestingly, another paper in this virtual issue reports that the same degree of inflammation appears to take place in the hindlimb lamellae as occurs in the forelimb lamellae [24], indicating that the same cellular signalling events probably occur in all digits, and that the increased incidence of structural failure in the forelimbs is only due to increased physical forces sustained by the forelimb lamellae.

Although histological evidence of the importance of failure of the adhesion of the LBEC from the underlying basement membrane has been in the laminitis literature for approximately 2 decades [25], this virtual issue contains a recent article in which the investigators used a serial biopsy technique to detail the loss/derangement of matrix components of the basement membrane at earlier time points (prior to clinical signs of lameness) than previously reported [26], indicating that these changes are likely to play a decisive role in the dysadhesion process. The primary focus for several years on events occurring at the epidermal/matrix interface that might result in loss of basement membrane components was the role the matrix-degrading enzymes, especially matrix metalloproteinases (MMPs), were playing in the onset of structural failure of the lamellae. Early focus on 2 central MMPs in basement membrane degradation, MMP-2 and MMP-9, occurred due to increased concentrations reported in the lamellar explants in organ culture when induced by a general MMP activator [27]; this caused a great deal of excitement about the potential of using these enzymes as therapeutic targets. However, as well delineated by another paper in this virtual issue [28], these MMPs have since been found to be unlikely to play a causative role in the lamellar breakdown due to either only being present in inactive form, or only being present in active form late in the disease process well after the onset of dysadhesion of the LBEC from the underlying basement membrane [29, 30]. Interest then appeared to be shifting to other types of proteases possibly involved in matrix breakdown due to recent reports of increased lamellar expression of ADAMTS-4 in the experimental and natural cases of laminitis [31]; ADAMTS-4 is a protease that breaks down the large polysulphated proteoglycans aggrecan and versican (both of which have been confirmed to be present in the lamellar tissue) [32]. However, recent reports on this topic do not support a direct role for ADAMTS-4 in basement membrane breakdown and other derangements of the extracellular matrix. This reflects the atypical distribution of aggrecan and versican in the equine digital lamellae; both proteoglycans are primarily localised within the lamellar epithelial cells (not in the matrix). Furthermore, increased ADAMTS-4 in LBECs of laminitic lamellae does not appear to result in elevated degradation of aggrecan. Rather, it results in elevated degradation of versican only, which together with suppressed versican gene expression in the laminitic tissue results in the almost complete clearance of versican from the basal epithelium [33]. Versican expression is implicated in development of the epithelial cell phenotype [34] raising the possibility that elevated ADAMTS-4 expression and versican depletion contributes to lamellar failure by affecting critical LBEC functions [33].

Thus, although it is still possible that matrix injury due to proteases plays a role in epithelial dysadhesion, investigators are beginning to change their focus to the possibility that the primary failure is at the level of regulation of epithelial cell physiology particularly the expression of adhesion proteins responsible for adhering the epithelial cells to each other and, probably most importantly, to the underlying basement membrane, i.e. the proteins that constitute the desmosomes, adherens junctions and hemidesmosomes. The hemidesmosomes, which are the primary protein complex for adhesion of the basal epithelial cell to the underlying basement membrane, are now known to be dynamically regulated structures that can undergo rapid disassembly when different signalling mechanisms are activated in the cell. Lamellar basal epithelial cells have been reported to undergo a loss of hemidesmosomes in a small number of animals administered oligofructose [35], thus indicating that regulation of hemidesmosomes may play a critical role. Recent interest in this area led to an international workshop on laminitis in which veterinary and human medicine researchers focused on the possible roles of the regulation of the hemidesmosome (and other epithelial adheshion molecules) in lamellar failure in laminitis (Havemeyer Equine Laminitis Workshop II, Key Largo, Florida, USA, 2012). Thus the possibility that a key part of lamellar failure is injury to the LBEC leading to dysregulation of its adhesion apparatus to the underlying basement membrane and dermis demands further investigation.

The first Havemeyer Equine Laminitis Workshop in 2007 was attended by a well-known physician scientist Elliott Crouser, from the human sepsis field, who had stated in an article on human sepsis, ‘Despite billions of dollars invested, no specific drug or therapy has been developed to effectively prevent the onset of SIRS or MODS [multiple organ dysfunction syndrome]’ [36]. At the closing of that meeting, he surprised the veterinary researchers by stating that he thought that we would come up with an effective therapy for sepsis-related laminitis before the human researchers did the same for organ failure in human sepsis. Although we have had the distinct disadvantage of a severe lack of funding compared with our human counterparts (and therefore a severe deficit of the number of the laboratories around the world to perform the research), equine laminitis researchers and clinicians have the noticeable advantage over our human physicians of having a peripheral tissue to treat (the digital lamellae vs. the visceral tissues in the human). Additionally, Crouser was surprised by the level of collaboration between the laboratories in the world working on this disease, and stated that this would also speed our progress. As presented by van Eps et al. in this issue, Dr. Crouser was correct in that we have a local therapy, termed digital ‘cryotherapy’ or hypothermia, which has withstood scientific rigour as an effective treatment in preventing sepsis-related laminitis [37]. That study, which provided clinical and histological evidence of the protective effect of hypothermia on the digital lamellae when instituted prior to the onset of the disease process, was followed by a study demonstrating that cryotherapy effectively blocks most inflammatory processes documented to be occurring in the early stages of sepsis-related laminitis [38]. The same investigators have recently presented preliminary evidence that cryotherapy is protective even when not instituted until the onset of clinical signs of lameness in the oligofructose model. Thus, after decades of laminitis therapies being introduced but not withstanding clinical or scientific testing, we now have one therapy that is well supported in the scientific literature to be an effective therapeutic option. As this therapy is a time- and work-intensive therapy that is not practical in many situations, the challenge is ongoing to find effective therapeutic agents that can be administered either locally (i.e. via regional limb perfusion) or systemically that reach the same or better levels of efficacy.

Thus, the last decade has been an exciting time to be in sepsis-related laminitis research, mainly due to the veterinary researchers throughout the world not only working together, but also being able to take advantage of the rapid advances in scientific methodology and the vast amount of data emanating from the human sepsis literature. There is certainly a great deal of work left to do in order to further our understanding of the sepsis and laminitis disease processes in our search for new therapeutic targets, but, as long as funding is available to support this increasingly expensive methodology, the veterinary clinician and researcher will continue to benefit from the rapidly advancing world of scientific discovery