Introduction
Mycotoxins are naturally occurring toxic metabolites released by moulds and fungi. They grow on a variety of feed and crops, most commonly in wet and humid conditions. Over 500 different mycotoxins have been discovered to date (Alshannaq and Yu 2017). In animals, mycotoxins can contribute to respiratory, reproductive, immunological, gastrointestinal and other disorders resulting in signs ranging from reduced productivity to death (Raymond 2000). However, not all mycotoxins cause serious acute disease and the effects of many are not well understood. In contrast to intensively farmed animals, little is known about the impact of mycotoxins in horses. Being a monogastric non-ruminant species, it has been hypothesised that horses may be more sensitive than ruminants towards adverse effects of mycotoxins (Liesener et al. 2009).Increased liver enzymes are reported in response to mycotoxicosis in horses, as in other species (Raymond et al. 2003; Durham 2022).
Globally, the most widely detected mycotoxins in animal feed or forage are produced by fusarium species; the most commonly reported is deoxynivalenol (Smith et al 1997; Raymond et al. 2003). However, to date there is only one study reporting on mycotoxin found in commercial horse feed (Liesener et al. 2009). They concluded that “co-contamination with several mycotoxins is very common in commercial horse feed” (Liesener et al. 2009). However, in most samples the toxin concentrations were well below the levels which are usually considered as critical or even toxic (Liesener et al. 2009). There are only two studies to date that has investigated mycotoxin levels in forage (hay or grass) intended for horses (Raymond 2000, Durham 2022). In the North American study, they found deoxynivalenol, T2 toxin and zearalenone in forage, with deoxynivalenol present in the highest amounts that could impact horse health (Raymond 2000). Durham, 2022, found that fumonisin B1 may be associated with outbreaks of liver disease (Durham 2022). However, studies have also found mycotoxins in a high proportion of forage fed to the control groups (Dänicke et al. 2021; Durham 2022). Our understanding of what mycotoxins horses are exposed to in forage is limited and even less is published regarding which mycotoxins could be clinically significant in horses.
This retrospective study aimed to present the data collected from forage sampling undertaken on horses with increased liver enzymes between May 2019 to October 2021. The primary aim was to identify if mycotoxins are identified in forage of horses that presented with increased liver enzymes and which mycotoxins are commonly detected. Additionally, we aimed to investigate the forage mycotoxin concentrations of those detected. The information collected in this pilot study should provide a foundation for further, more in-depth, research into the mycotoxins commonly found in equine forage in the UK and their potential for causing disease.
Materials and methods:
Electronic patient records were manually searched to retrospectively collect data from client submission forms submitted with forage samples to Rossdales Laboratories prior to mycotoxin testing. Data collected:
Age, sex, breed
Geographical location (postcode) of pasture/forage sampling
Supplementary feeding, including if a mycotoxin binder has been used
Sample type: grass, hay or haylage
Clinical signs/ reason for testing
If increased liver enzyme concentrations had been detected
Forage samples were taken by clients and submitted to Rossdales Laboratories. All clients were advised to sample the centre of multiple different hay bales (five to six). For grass sampling, clients were advised to take small handfuls from across the whole pasture. To be included in the study, horses must have had clinical signs of liver disease and/or increased liver enzyme concentrations on a blood sample (confirmed by Rossdales Laboratories or the referring veterinary surgeon) within 120 days of forage sampling; and must have been greater or equal to two years of age at the time of sample collection.
Samples were sent to Alltech and tested for percentage dry matter and then tested for 54 mycotoxins (see appendix one for list of mycotoxins tested) using liquid chromatography and mass spectrometry techniques (Jackson et al. 2012). Samples were ground in a coffee grinder for 30 seconds to obtain consistent particle size. 400mg sub-samples were taken and equally distributed in glass reaction vials. The samples were centrifuged at 4000rpm for 30 minutes. 500uL of supernatant was collected and dried under a nitrogen stream for 30 minutes at room temperature. The samples were reconstitute in 500uL of loading buffer. The analysis was performed on Acquity UPLC/ESI-TQD MS/MS system utilising an ethylene-bridged hybrid C18 analytical column maintained at 40 degree centigrade. The analysis was carried out at a flow rate of 0.42ml/min over 16 minutes per samples injection with a gradient of water. 54 mycotoxins were anaylsed and the detection limits, lower quantification limits and standard deviations were set by Alltech for each mycotoxin.
Analysis: Descriptive statistics were carried out for categorical data and summary statistics for quantitative data. If normally distributed (as determined by Shapiro Wilk Normality Test) means and confidence intervals were presented for quantitative data. If the data was not normally distributed medians and interquartile ranges were presented. The frequency of each mycotoxin detected was recorded to establish the most commonly detected mycotoxins and the median levels detected of those identified. Any detected values were reported as μg/kg. Adverse performance risks associated with multiple mycotoxins in feed were evaluated by calculating a risk equivalent quantity (REQ) (Yiannikouris 2013). REQ represents the sum of the mycotoxin risk based on the mycotoxin concentration and respective risk factor (Yiannikouris 2013). A species-specific risk equivalence factor (REF) is assigned to each mycotoxin relative to the most toxic mycotoxin (aflatoxin B1). The total toxicity of multiple mycotoxins can then be hypothesised as a single risk equivalent quantity (REQ), which is calculated by summing the products of individual REFs and their respective concentrations (Yiannikouris 2013).
Results: A total of 78 forage samples were submitted to Rossdales Laboratories for testing by Alltech between May 2019 and October 2021. Of those tested, 52 samples fulfilled the selection criteria (see appendix 2). 27 samples of grass (52%), one sample of haylage (2%) and 24 samples of hay (46%) were submitted from 46 cases (six horses were submitted with two or more forage samples). Ages of horses ranged from 2-32 years old, with age unspecified in 8 horses (median age 12 years old, with an interquartile range of 6.75-19 years). The predominant breed was cobs (n=10), with mixed representation from other breeds (pony = 8, warmblood = 7, miniature = 3, thoroughbred = 3, Irish sport horse = 2, arabian = 1, hackney = 1, suffolk = 1, unspecified = 10). All horses had increased liver enzyme concentrations confirmed on blood serum analysis by either Rossdales Laboratories or by the referring veterinary surgeon with a median 28 days (IQR 21-60 days) between liver enzyme analysis and mycotoxin forage analysis. 18/52 samples had some or all data available for liver enzymes values (see table one).
Geographical distribution was predominant focused in the southeast of England with all but one sample (300 miles) within 120 miles of Newmarket, UK. Mycotoxins were detected in 50/52 samples. Two or more groups were detected in 42/52 samples, with the highest number of six mycotoxins groups detected (n=1). Toxins were detected from all groups except aflatoxins. The median number of mycotoxin groups detected in each sample was three (see figure one). The most commonly detected groups were emerging mycotoxins (n=39), fusaric acid (n=25), followed by type b trichothecenes (n=24) (see table two and figure two).
Based on current research and published data for other species, Alltech quantify individual mycotoxins risk to the animal as lower, medium or higher risk. All individual mycotoxins groups identified were detected at median concentration levels of ‘lower’ or below except Ochratoxins/citrinin (AB,B) which were ‘higher’ with a median concentration of 66 μg/kg [IQR 22-66 μg/kg] (see table three). Type B trichothecenes was most commonly found at significant concentrations, with 8/24 samples type B trichothecenes identified at medium or high-risk concentrations (see figure three). 14/52 (27%) samples had one or more mycotoxin group that was detected at the concentration above the ‘higher’ risk threshold, 22/52 (42%) samples had one or more mycotoxins groups that were detected at concentrations at ‘medium’ or ‘higher’ risk.
Limitations: The major limitation to the study was case selection bias and a lack of a control group. This was impossible to mitigate due to the method of data collection and retrospective nature of the study. Incomplete data sets were also a problem and was the most common reason for samples not meeting inclusion criteria. Data demonstrating degree of increase in liver enzyme concentrations was only available in 18/52 samples. The growth of mycotoxins is affected by multiple factors such as environmental conditions such as temperature, moisture conditions, geography and agricultural practices. As these factors vary both seasonally and annually, levels of mycotoxins will also vary. Due to the short study duration and low sample numbers, it was not possible to investigate this further. There was no data available for mycotoxins in hard or concentrate feed, which may have also been a source of mycotoxins for horses fed concentrates in addition to forage. Due to the retrospective nature of the study, it was impossible to control the method and timing of the forage sampling. All clients were given the same advice for sampling, but the timing of mycotoxin testing after increased liver enzyme detection could not be controlled. Due to the lack of evidence regarding mycotoxins and their effects in horses, reference ranges were extrapolated from food animals. Alltech evaluate the impact (lower/moderate/higher) of mycotoxins concentrations detected where an impact on performance and health has been observed at chronic levels of exposure in farm animals, rather than toxicological limits. No such data is available in horses. Alltech set the reference limits based on a variety of sources including research and government regulations, with support from commercial observations. There is very little data to demonstrate effects of mycotoxins on horses. The spectrum mycotoxins tested are those the most commonly affect food animals which were selected to be based due to a lack of data in horses that indicate which mycotoxins are prevalent or clinically significant.