Continuous digital hypothermia initiated after the onset of lameness prevents lamellar failure in the oligofructose laminitis model



Reasons for performing study

Prophylactic digital hypothermia reduces the severity of acute laminitis experimentally but there is no evidence for its efficacy as a treatment once lameness has already developed.


To investigate the therapeutic effects of digital hypothermia, applied after the onset of lameness, in an experimental acute laminitis model.

Study design

Randomised, controlled (within subject), blinded, experimental trial.


Eight Standardbred horses underwent laminitis induction using the oligofructose model. Once lameness was detected at the walk, one forelimb was continuously cooled (CRYO), with the other forelimb maintained at ambient temperature (NON-RX). Dorsal lamellar sections (proximal, middle and distal) harvested 36 h after the onset of lameness/initiation of cryotherapy were analysed by 2 blinded observers: laminitis pathology was scored (0 [normal] to 4 [severe]) and morphometric analyses performed.


Median (interquartile range) histological scores were greater (P<0.05) in NON-RX (proximal 2.8 [2.5–4]; middle 3.5 [2–4]; distal 2.5 [2–3.8]) compared with CRYO limbs (proximal 0.5 [0.5–1.4]; middle 1 [0.6–1]; distal 0.75 [0.5–1]). There was complete physical separation of lamellar dermis from epidermis (score of 4) in 4 of the NON-RX feet at one or more section level(s), which was not observed in any CRYO sections. Histomorphometry was thus limited to sections that remained intact; there was a trend of increased total (TELL) and secondary (SELL) epidermal lamellar length and decreased secondary epidermal lamellar width (SELW) in NON-RX limbs compared with CRYO at all 3 levels; differences were significant (P<0.05) for SELL and SELW in the distal sections.


Digital hypothermia reduced the severity of lamellar injury and prevented lamellar structural failure (complete dermoepidermal separation) when initiated at the detection of lameness in an acute laminitis model. This study provides the first evidence to support the use of therapeutic digital hypothermia as a treatment for acute laminitis.


Acute laminitis is a common complication of a variety of primary diseases in the horse; however, there is currently no evidence to support an effective treatment. Recommendations for the treatment of acute laminitis are focused on supportive therapy (including analgesia), while strategies for reducing the progression of the actual lesion (lamellar separation and ultimately mechanical failure of the suspensory apparatus of the distal phalanx) are limited to orthotic support and restriction of ambulation [1]. Lamellar separation occurs at the interface between epidermis and dermis (the basement membrane; BM) [2]. Although the mechanisms that initiate the lesion may differ according to the inciting cause (endocrinopathy, sepsis or painful conditions of the contralateral limb), there is evidence of some commonality during the acute phase of laminitis, with inflammatory and enzymatic processes identified in models of sepsis-induced [3-8] and, to a lesser extent, endocrinopathic (hyperinsulinaemic) [9-11] laminitis. Furthermore, once there is sufficient weakening of the lamellar attachment, traumatic damage to the tissue (due to the distractive forces of weightbearing) is inevitable, with secondary inflammation and enzymatic activation (either primary or as part of the remodelling process) potentially contributing to further damage. The aim in acute laminitis cases must therefore be to limit the lesion progression so that mechanical failure does not occur.

There is strong evidence from studies utilising the oligofructose (OF) model of laminitis that digital hypothermia initiated early in the developmental period has a preventative effect [12-14] and a recent clinical study showed that prophylactic digital cryotherapy reduced the incidence of laminitis in horses with colitis [15]. Hypothermia has potent effects on tissue, primarily causing analgesia, hypometabolism and vasoconstriction [16]. The profound anti-inflammatory effect of hypothermia reportedly occurs as a result of reduced production and activity of proinflammatory cytokines [17-19], increased production of anti-inflammatory cytokines [17, 20], reduced adhesion and extravasation of leucocytes from the microvasculature into affected tissues [21, 22] and reduced production of oxygen radicals by polymorphonuclear leucocytes [23]. These profound anti-inflammatory effects were demonstrated in the OF model of laminitis when hypothermia was initiated early in the developmental period [14].

In the clinical setting, it is not always practical to initiate cryotherapy early in the developmental period (prior to clinical signs of laminitis) and laminitis cases frequently only receive veterinary attention after the development of lameness of adequate severity to be detected at the walk. There is currently no evidence demonstrating a therapeutic effect for hypothermia that is initiated once laminitis has progressed to the extent that lameness is evident. We hypothesised that digital hypothermia would reduce the progression and ultimate severity of acute laminitis when applied shortly after the onset of lameness. Using the OF laminitis model, we assessed the efficacy of continuous digital hypothermia on decreasing the progression of lamellar injury in acute laminitis, when the treatment was initiated after the onset of lameness (detectable at a walk).

Materials and methods

Animals, laminitis induction and monitoring procedures

Eight Standardbred geldings (aged 3–11 years; 430–548 kg bwt), with no lameness and no gross or radiographic abnormalities of the feet, were housed and fed in stables for 4 weeks prior to the experiment. Laminitis induction was by alimentary overload with OF as previously described [24]. Immediately after nasogastric administration of the bolus dose (10 g/kg bwt of OF, up to a maximum dose of 4.2 kg), each horse was confined to stocks for the remainder of the experiment.

Forelimb hoof temperature was monitored using hoof wall thermistors attached to data-logging devices as previously described [14]. Human pedometer devices (Yamax digiwalker sw700)a were taped onto the antebrachium of both forelimbs to provide data on frequency of weight shifting as previously described [14]. The total pedometer readings were recorded every hour and the pedometers were then reset.

The horses were monitored carefully for the onset of lameness using a combination of subjective gait evaluation and pedometer counts. Beginning 12 h after the bolus OF dose, the horses were examined every 4 h by 2 investigators for lameness at the walk. The horses were walked toward and away from the observer and circled to the right and left on a concrete surface: the presence of lameness in one or more forelimb(s) and the limb on which the horse was judged to be most lame was recorded (unless symmetrical). Mutual recognition and agreement by both investigators of lameness in one or both forelimbs detectable at the walk (Obel grade 2 [OG2] laminitis) [25] was the criterion for initiation of digital hypothermia and perineural analgesia.

At 36 h after the detection of lameness and initiation of digital hypothermia, each horse was subjected to euthanasia with pentobarbital sodium (20 mg/kg bwt i.v.). The forelimb dorsal lamellae were dissected from the hoof and third phalanx and sections were fixed and processed for light microscopy using formalin fixation as previously described [2].


Upon first recognition of OG2 lameness (immediately prior to initiation of cryotherapy), analgesia was achieved via a combination of continuous peripheral nerve block (CPNB) of palmar nerves in both forelimbs, phenylbutazone 8 mg/kg bwt i.v. (Phenylarthrite)b and, if required, supplementary intermittent perineural anaesthesia of the plantar and plantar metatarsal nerves using 2.5 ml each of 0.25% bupivacaine (Bupivicaine Injection BP)c and 2% mepivicaine (Mepivicaine)d per site. The hindimb perineural analgesia was initiated only if there was a subjective increase in weight shifting noted in the hindlimbs. Continuous peripheral nerve block of the palmar nerves was initiated in both the continuously cooled (CRYO) and maintained at ambient temperature (NON-RX) forelimbs according to a previously described technique [26]. Briefly, horses were sedated with detomidine (Detomo Vet)d, 10 µg/kg bwt i.v. and a perineural palmar block performed distal to the communicating branch of each limb using 5 ml 2% mepivacaine hydrochloride (Mepivicaine)d. The sites where the medial and lateral catheters were to be placed were identified using anatomical landmarks as previously described by Zarucco et al. [27]. The area was then surgically prepared and 3 ml 2% mepivacaine infiltrated into the skin and subcutaneous tissue. An 18 gauge 3.25 inch (8.25 cm) Tuohy needle (Perican)e was placed subcutaneously and then angled to penetrate the palmar fascia to lie adjacent to either the medial or lateral palmar nerve. A preheparinised closed tip epidural catheter (Perifix catheter)e was then threaded through the needle and advanced 3 cm past the needle tip. The Tuohy needle was removed and the catheter secured by tunnelling subcutaneously then fixing to the skin using cyanoacrylate. An initial volume of 2 ml of local anaesthetic solution (bupivacaine 0.25%, 0.04% NaHCO3 and epinephrine [adrenaline] 1:200,000 in 0.9% saline) was then injected into each catheter. A continuous perineural block was achieved by injecting 2 ml of local anaesthetic solution into the catheters every hour.

Digital hypothermia

Immediately upon first recognition of OG2 lameness, each horse had one of the forelimbs continuously cooled (CRYO) and the other (NON-RX) forelimb maintained at ambient temperature for the duration of the experiment. The forelimbs of each horse were randomly assigned to CRYO or NON-RX groups prior to the start of the experiment. Cooling was achieved by placing the CRYO limb in a rubber boot (Bigfoot Ice boot)f containing a mixture of 50% cubed ice and 50% water (constantly maintained over the 36 h period) to a level just below the carpus, as previously described [12].

Histological evaluation

Formalin-fixed sections were stained with H&E and PAS, randomised and coded for histological analysis performed by 2 blinded observers. The severity of laminitis pathology was scored for each section using a system based on the 0–3 scale previously described by Pollitt [2], with the addition of a fourth category (score of 4), defined as complete physical separation of lamellar epidermis from dermis, with no association between epidermal and dermal tissues on the section (see Fig 1b). Morphometric analyses were also performed on 10 individual primary (PEL) and secondary (SEL) epidermal lamellae per section. At each level, the total epidermal lamellar length (TELL), secondary epidermal lamellar length (SELL) and secondary epidermal lamellar width (SELW) were recorded according to the technique of de Laat et al. [10] using computer software (Image Pro)g. The presence of complete separation of epidermis from dermis (score of 4) precluded those sections from histomorphometric analysis.

Figure 1.

Gross (left) and H&E-stained light microscopic appearance (right) of continuously cooled (CRYO) (a) and maintained at ambient temperature (NON-RX) (b) lamellar tissue from the middle lamellar level of the same horse. Complete loss of association between epidermis and dermis (indicating total failure of the lamellar attachment) is evident in the NON-RX limb (b) whereas mild histological changes were present in the opposite (CRYO) limb (a) from this horse. Complete lamellar separation (grade 4 histological score) was present in 4 NON-RX feet at one or more section levels, but was not observed in any of the CRYO feet sections.

Data analysis

The median histological scores and histomorphometric measurements were compared between the treated (CRYO) and untreated (NON-RX) limbs using Wilcoxon signed ranks tests. Pedometer counts were compared over time using Friedman analysis with Dunn's post tests and between CRYO and NON-RX limbs at each time point using Wilcoxon signed ranks tests. Interevaluator agreement for the histological grades was assessed using a weighted kappa test. Statistics were performed using computer software (Graphpad Prism)h. Significance was set at P<0.05.


Clinical and hoof temperature data

All horses developed Obel grade 2 lameness between 17 and 21 h after the bolus OF dose (median 18.5 h) and none had more severe lameness. The forelimb lameness was judged to be symmetrical in 3 horses, while 3 appeared more lame in the forelimb assigned to be the CRYO limb and 2 appeared more lame in the forelimb predesignated as NON-RX. All horses received a dose of phenylbutazone at the onset of lameness and again between 12 and 24 h after onset of lameness. In addition to the CPNB of the forelimbs (initiated at the onset of lameness in all horses), 3 of the horses received perineural analgesia in both hindlimbs (at 24 h after lameness onset, repeated every 6 h).

The hoof wall surface temperature and pedometer data (normalised to be relative to the time of lameness onset in each horse) are shown in Figure 2. Limb movement data from the pedometers documented a trend of steady increase in the count frequency for both NON-RX and CRYO limbs after the bolus dose, peaking at the onset of OG2 lameness (neither NON-RX nor CRYO limbs were in ice during that time period prior to OG2 lameness; Fig 2a). After initiation of the CPNB and cooling at 0 h, the CRYO limb count frequency was significantly decreased at 5–14, 17 and 22 h compared with the onset of lameness (0 h). There was also a trend of decreased count frequency after 0 h in the NON-RX limbs, but this was not significant. From 2 h after lameness onset until the end of the experiment, the CRYO limb count frequency was significantly less than that of the NON-RX limbs with the exception of 4, 32 and 33 h time points. The horses were noted visually to consistently stand still on the CRYO limb while shifting the NON-RX forelimb.

Figure 2.

The median (± IQR) pedometer count frequency (a) showed a trend of increase in both continuously cooled (CRYO) and maintained at ambient temperature (NON-RX) limbs to peak at the onset of lameness (0 h), after which there was a trend of decrease for the remainder of the experiment, with the CRYO count frequency significantly less than NON-RX at all time points except 4, 32 and 33 h. The (median ± IQR) hoof wall surface temperature (b) was similar in both limbs until the onset of cooling at 0 h, after which the temperature was 7.1 (6.8–7.7)°C for the CRYO feet and 30.2 (29.7–31.0)°C for the NON-RX feet.

After the initiation of cryotherapy (from 0 h until the end of the experiment), the median [interquartile range] hoof wall surface temperature was) 7.1 [6.8–7.7]°C for the CRYO feet and 30.2 [29.7–31.0]°C for the NON-RX feet (Fig 2b.).

Histological scores

Median histological scores (Fig 3) of the lamellar sections at each section level were significantly greater in the NON-RX limbs (proximal 2.8 [2.5–4]; middle 3.5 [2–4]; distal 2.5 [2–3.8]) compared with the CRYO limbs (proximal 0.5 [0.5–1.4]; middle 1 [0.6–1]; distal 0.75 [0.5–1]). There was complete physical separation of dermal lamellae from epidermal lamellae (score of 4) in 4 NON-RX feet at one or more section levels, which was not observed in any of the CRYO feet sections (Fig 1). The calculated kappa statistic for the blinded evaluators was 0.73, which can be interpreted on the scale developed by Landis and Koch [28] as ‘substantial’ agreement.

Figure 3.

Median histological scores for proximal, middle and distal dorsal lamellar sections from continuously cooled (CRYO) (circles) and maintained at ambient temperature (NON-RX) (triangles) feet. The scores were significantly decreased in the CRYO sections at each level compared with NON-RX sections (*P<0.05).


Histomorphometry measurements were not possible in 9 of the 24 NON-RX sections (3 proximal, 4 middle and 2 distal) due to the complete lamellar dermoepidermal separation (histological score of 4) present in these sections. The results for the remaining sections are shown in Table 1. There was a trend of increased PEL and SEL length, and decreased SELW in the NON-RX limbs compared with CRYO limbs at all levels; however, the differences were only significant for SELL and SELW in the distal sections.

Table 1. Median [IQR] lamellar histomorphometry results for CRYO vs. NON-RX limbs, for sections at each of the 3 levels (proximal, middle and distal). Analysis was not possible on sections with complete separation of epidermis and dermis (histological score of 4). There was a trend of increased TPELL, increased SELL and decreased SELW at all levels in the NON-RX limbs compared with the CRYO limbs
MeasurementSection levelNON-RXCRYOP value
  1. *Denotes significant difference between NON-RX and CRYO at that level. CRYO = continuously cooled; NON-RX = maintained at ambient temperature; TPELL = total primary epidermal lamellar length; SELL = secondary epidermal lamellar length; SELW = secondary epidermal lamellar width.
TPELL (μm)Proximal3500 (3335–3936)3209 (3021–3693)0.0625
Middle3318 (3146–3667)3112 (2934–3248)0.125
Distal3188 (2885–3855)2985 (2758–3265)0.0938
SELL (μm)Proximal302 (272–401)120 (118–140)0.0625
Middle263 (192–316)208 (160–320)0.625
Distal322* (228–392)149 (130–181)0.0313
SELW (μm)Proximal9.5 (5.8–11.8)26.6 (22.8–38)0.0625
Middle14 (0.1–17.9)19.9 (18.2–23.9)0.25
Distal14.9* (13.2–17.3)27.8 (20.7–30.3)0.0313


The results indicate that continuous digital hypothermia, initiated at the detection of lameness at a walk (Obel grade 2 laminitis) in an acute laminitis model, dramatically reduced the severity of lamellar injury. Although digital hypothermia resulted in a consistent protection from severe lamellar damage in all horses, the most remarkable effects were noted in the 50% of horses in which complete separation of lamellar epidermis from dermis was present in the NON-RX limbs and the lesions in corresponding CRYO limbs were limited to only the mildest histological changes observable with light microscopy (see Fig 1). Despite the fact that histomorphometric analysis was limited to a small number of sections (4 middle, 5 proximal and 6 distal sections) due to the inability to analyse sections which had undergone complete lamellar separation, significant differences in SELL and SELW were detected in distal NON-RX sections compared with CRYO sections. These changes were similar to those previously reported in studies using OF and hyperinsulinaemia laminitis models [10, 13, 29]. It was necessary to add the histological score of 4 to the grading system of Pollitt [2] (originally 0–3) in order to document a complete lack of any remaining lamellar connection in the severely affected sections. Because the authors took care not to place distractive forces on the lamellar tissue during the harvesting process, the severity of separation noted in this experiment likely reflects the severity of the lamellar structural failure (not artefact from harvesting). Total separation was more common in the middle sections than the distal and proximal sections, where it is possible that the terminal and coronary papillae (which consistently maintained their connections) respectively stabilised the tissue sufficiently to allow it to remain intact.

Continuous peripheral nerve block of the palmar nerves appeared to be an effective research tool to improve the animals' comfort during a laminitis induction protocol that extended after the onset of lameness in this study. However, as CPNB was not performed in the hindlimbs, careful monitoring was important and hindlimb perineural analgesia was performed in the few animals which demonstrated signs of hindlimb discomfort. Systemic phenylbutazone administration was used to provide further analgesia and did not appear to interfere with the pathophysiology of the disease, as evidenced by the severity of lesions present in the NON-RX limbs. In contrast to the continuous (pump) infusion previously described [27], CPNB was performed by hourly manual bolus for logistical reasons in this study (specifically the lack of availability of 4 suitable infusion pumps per horse). The pedometer data demonstrated increased limb movement frequency in the NON-RX limbs compared with the CRYO limbs after the initiation of CPNB and hypothermia. The difference may be attributable to mild residual pain despite CPNB which may be expected to have been more profound in the NON-RX limbs due to the relative severity of the developing laminitis lesions in those limbs (compared with the CRYO limbs); the additional well-documented analgesic effects of hypothermia may also have played a role in decreased movement of the CRYO limbs [15, 30, 31], as noted during laminitis induction in a previous study [14]. Lameness was not evaluated without perineural anaesthesia at the end of the experiment as it was a humane consideration to ensure the continuous perineural blocks were maintained through to the time of euthanasia. However, the minimal degree of discomfort noted in the forelimbs of all animals in the face of severe lamellar injury of all NON-RX limbs (and total lamellar failure in 50% of animals) indicates that CPNB is an important tool for providing analgesia in laminitis research.

There are several mechanisms by which hypothermia might protect lamellar tissue from damage and reduce the progression of laminitis in the OF model, which is essentially a model of sepsis [24, 32, 33]. Inflammation and protease activity are implicated in the developmental as well as acute phases of sepsis-associated laminitis [3-8, 34]. When applied before the onset of clinical lameness in the OF model, continuous digital hypothermia dramatically inhibited transcription of inflammatory mediators including cytokines, chemokines and cyclo-oxygenase in lamellar tissue [14]. In addition, hypothermia had an inhibitory effect on lamellar matrix metalloproteinase expression in the OF model [12], a finding consistent with experimental studies of hypothermia in brain trauma [35] and cardiac arrest [36]. Therapeutic hypothermia is well established in the management of human patients undergoing cardiac bypass, in the treatment of stroke and in the prevention and treatment of neonatal encephalopathy, where it is purported to protect brain tissue through attenuation of cellular energy failure, oxidative injury, apoptosis and inflammation [37-39]. Mild, whole body hypothermia (5–6°C below normal) has been shown to increase survival in rodent models of sepsis [40, 41] and the experimental evidence is prompting discussion of clinical application of therapeutic hypothermia in septic human patients [42, 43]. Preservation of cellular energy metabolism and mitochondrial function [44], reduced inflammation [16, 19, 45, 46] and inhibition of apoptotic pathways [47] are effects of hypothermia that protect against sepsis-related end organ damage in the liver, lung and heart. The contribution of cellular energy failure (nonischaemic) to the different forms of laminitis is unclear, but hypothermia of the equine digit may exert protective effects on lamellar tissue by modulating mitochondrial function and reducing metabolic demands. Many of these events will be assessed in the lamellar samples from this project in the future.

There are several limitations to the current study, particularly when extrapolating to clinical scenarios. Firstly, hypothermia was initiated immediately after the onset of lameness, whereas in clinical cases there are likely to be relative delays in both the recognition of lameness and also the initiation of therapy. It is reasonable to assume that the therapeutic effects of hypothermia are limited to the reduction of lesion progression (rather than reversal of pre-existing structural/morphological derangement); therefore, the severity of pre-existing lesions at the time cooling is initiated is likely to dictate the efficacy of the treatment in clinical cases. Secondly, the study was continued for only 36 h after the onset of lameness for logistical and ethical reasons. It is therefore unclear whether the laminitis lesions in the CRYO feet may have progressed after discontinuation of the cooling if the experiment was continued beyond this. This concern was partially addressed in a previous study where the protective effect of prophylactic digital hypothermia (72 h) was still apparent 7 days after OF administration [13].

In conclusion, this study demonstrates the efficacy of continuous digital hypothermia as a treatment for acute laminitis in an experimental model. The findings support the use of therapeutic hypothermia in clinical cases of acute laminitis; however, there remains a need for studies that document clinical efficacy and safety in order to develop recommendations for the clinical environment.

Authors' declaration of interests

No competing interests have been declared.

Ethical animal research

The project was approved by the University of Queensland Animal Ethics Committee (AEC) that monitors compliance with the Animal Welfare Act (2001) and The Code of Practice for the care and use of animals for scientific purposes (current edition). All animals were monitored continuously by the investigators.

Source of funding

This project was funded by a grant from the Grayson Jockey Club Research Foundation.


Supported by a Grayson Jockey Club Research Foundation Grant. Presented in part at the American College of Veterinary Internal Medicine Conference, New Orleans, 2012.


A.W. Van Eps and J.K. Belknap contributed equally to study design, data analysis and interpretation and preparation of the manuscript. All authors contributed to the study execution and approved the final manuscript.

Manufacturers' addresses

  1. aYamasa Tokei Keiki Co., Ltd., Tokyo, Japan.

  2. bAusrichter Pty Ltd, Anandale, Australia.

  3. cPfizer Australia Pty Ltd, West Ryde, Australia.

  4. dNature Vet Pty Ltd, Glenorie, Australia.

  5. eB. Braun, Melsungen, AG, Germany.

  6. fBigfoot Ice Boots, Esk, Queensland, Australia.

  7. gMedia Cybernetics, Silver Spring, Maryland, USA.

  8. hGraphPad Software, San Diego, California, USA.