Identification and alleviation of pain is essential for welfare of all species. The most commonly used analgesic medications in horses include the alpha₂-adrenergic agonists, non-steroidal anti-inflammatory drugs (NSAIDs), and opioids. Despite well-documented analgesic properties, the alpha₂-adrenergic agonists profoundly decrease gastrointestinal motility [1–3] making these drugs a poor choice for prolonged administration. NSAIDs can lead to gastric and colonic ulceration and renal tubular necrosis, potentially life-threatening side effects. Severe pain may be refractory to single analgesic therapy; thus its management may require a multimodal approach, employing drugs with different mechanisms of action. Despite the potential for improved analgesia [4], use of drug combinations may also increase the potential for adverse effects, especially alterations in gastrointestinal motility or behavior [2] when drugs are combined at the dosages used for monotherapy.

Lidocaine, a local anesthetic, is commonly administered as a constant rate infusion (CRI) in horses following exploratory laparotomy for its potential analgesic and prokinetic properties [5–10]. Ketamine, an N-methyl-D-aspartate (NMDA) receptor antagonist, has anti-nociceptive properties when administered as a CRI at sub-anesthetic doses in both anesthetized and conscious horses [11, 12]. The opioid drug butorphanol, a kappa (OP2) agonist and competitive mu (OP3) antagonist, is commonly administered in the horse with varying success. No anti-nociceptive properties were noted in clinically normal horses in response to visceral distension or noxious thermal stimulus [13] but pain scores were significantly decreased in horses administered butorphanol as a constant rate infusion in the immediate post-operative period following exploratory laparotomy [14].

Overall gastrointestinal transit time from the ingestion of a meal to passage of feces is one of the most clinically applicable objective evaluations of gastrointestinal function as it incorporates all segments of the gastrointestinal tract. It is commonly measured by passage of either liquid or particulate markers administered via nasogastric tube, which are then either physically recovered or identified radiographically in feces [15, 16]. Clinical methods for assessing gastrointestinal motility include auscultation of gastrointestinal borborygmi and passage of feces. In this study we combined objective measures of total gastrointestinal transit time and fecal output and subjective assessment of borborygmi scores for a global assessment of gastrointestinal motility.

Thermal threshold (TT) testing has been successfully used in adult horses and foals to evaluate somatic nociception [10, 17, 18]. A wireless TT system allows testing to be performed with the horse unrestrained, [17] which allows the horse to display the full range of behavioral responses to a noxious stimulus and reduces the potential effects of restraint and presence of an investigator on the response. Thermal nociceptive threshold testing is one measure of nociception. The benefits of this mode of testing are that it is repeatable and causes no tissue damage thus reducing the animal welfare concerns with repeated nociceptive testing.

The purpose of the study reported here was to evaluate the effects of lidocaine, ketamine, and butorphanol, alone and in combination, on total gastrointestinal transit time, somatic nociception, and behavior in clinically normal horses. From an immediate clinical standpoint, this study provides clinicians already using these drugs in combination with knowledge of both the anti-nociceptive and systemic effects of the medications in healthy, non-painful, horses.

The infusion was discontinued for 3 horses in the LBK group (after a mean of 38 hours) and 1 in the BK group (after 16 hours) due to signs of colic; their data were excluded from analysis. Signs of pain were mild in all horses (flank watching, lying down), none developed nasogastric reflux, and all responded to medical therapy (flunixin meglumine, water and electrolytes via nasogastric tube). No other problems were observed.

The percentage of active drug, relative to time 0, detected in the combined solution (as prepared for all infusions) after 6, 12, 24, 48, and 72 hours of storage was as follows: butorphanol - 78, 92, 81, 83, 79%; ketamine - 72, 90, 75, 74, 69%; lidocaine - 88, 93, 108, 85, 87%. Each 1L bag was administered within 15 hours.

Mean fecal bead passage logistic regression curves are presented in Figure 1; time to passage of 1, 25, 50, and 75% of beads (T₁, T₂₅, T₅₀, T₇₅) are presented in Table 1. Transit time was significantly delayed in the LB and LBK groups.

There was no significant treatment effect on thermal threshold (Table 3), any vital parameter, behavior or gastrointestinal borborygmi scores, or hay or water consumption (data not shown). There were significant time and/or period effects on skin temperature, rectal temperature, borborygmi scores, heart rate, and body weight. Skin and rectal temperatures were higher in the early experimental periods (the order of experiments, regardless of drug administered) (Figure 2). Body weight was significantly lower in periods 1 and 2 (492.0±9.5 and 494.1±9.4 kg, respectively) relative to periods 4-6 (508.9±9.5, 509.4±9.4, 507.7±9.5 kg). Borborygmi scores were lower in periods 1-3 (7.3±0.6, 7.3±0.6, 6.3±0.6), relative to 5-8 (10.6±0.6, 10.6±0.6, 10.8±0.6, 11.2±0.6). Heart rate was higher in period 1 (41.9±1.2 bpm), relative to periods 5 (37.0±1.2) and 7 (36.9±1.2) and tended to increase then decrease over the 96-hour study period. Skin and rectal temperatures also had a significant time effect, such that temperatures were higher in the afternoon/evening (Figure 3).

Time (hr)
	-0.5	-0.25	0.5	6	12	24	36	48	60	72	84	96
P	45.7 ±3.4	50.0 ±4.5	52.5 ±3.2	46.4 ±3.9	49.2 ±4.8	51.7 ±4.0	50.8 ±3.8	47.3 ±3.5	51.8 ±3.0	49.7 ±2.3	51.3 ±2.0	52.0 ±4.5
L	47.2 ±1.9	46.9 ±3.0	52.5 ±2.8	49.7 ±3.6	52.2 ±4.1	49.7 ±4.6	50.9 ±2.7	49.2 ±1.4	49.7 ±4.1	51.1 ±3.7	53.2 ±2.5	49.0 ±3.3
B	50.5 ±3.5	49.0 ±3.9	47.9 ±4.2	49.1 ±2.6	52.1 ±2.7	50.8 ±2.7	51.2 ±3.5	50.4 ±3.2	49.1 ±2.5	52.3 ±3.5	53.3 ±2.5	51.1 ±5.2
K	47.0 ±4.1	47.4 ±5.1	48.4 ±5.3	47.8 ±2.4	49.3 ±5.5	50.3 ±2.6	49.7 ±3.4	49.2 ±1.8	52.2 ±3.7	52.1 ±2.8	50.1 ±3.3	49.4 ±4.5
LB	46.4 ±3.8	48.1 ±2.4	53.7 ±2.9	50.1 ±3.8	50.2 ±3.7	51.3 ±2.2	49.8 ±1.0	50.7 ±3.6	49.8 ±3.5	52.0 ±3.2	49.4 ±6.2	52.6 ±2.9
LK	48.7 ±3.0	47.3 ±3.5	53.6 ±2.7	48.1 ±4.3	49.5 ±3.3	49.8 ±3.9	47.6 ±5.6	49.9 ±5.4	49.1 ±4.5	49.7 ±2.4	52.1 ±2.0	49.4 ±3.7
BK	46.2 ±2.6	51.2 ±4.4	50.6 ±3.8	47.5 ±2.8	50.5 ±3.6	48.4 ±3.2	49.7 ±4.8	50.1 ±1.8	50.6 ±3.5	50.7 ±2.8	51.0 ±3.6	51.6 ±3.9
LBK	46.1 ±2.3	45.6 ±5.5	49.7 ±4.4	48.8 ±3.4	51.4 ±3.9	49.1 ±5.0	51.3 ±3.8	49.8 ±1.9	49.2 ±5.0	50.3 ±4.9	48.1 ±5.2	51.6 ±4.9

See Table 1 for key to treatments. No significant period, time, or treatment effects were detected.

We report a significant delay in total gastrointestinal transit time and a decrease in cumulative fecal weight as a result of prolonged continuous rate infusion with lidocaine/butorphanol and lidocaine/ketamine/butorphanol combinations. Each drug combination containing butorphanol caused a decrease in cumulative fecal weight during the latter portion of the infusion. The negative effects on cumulative fecal weight do not appear associated with decreased intake, as neither hay nor water consumption was significantly affected.

An infusion time of 96 hours was chosen to ensure that all beads passed during drug infusion. While this duration of infusion may be longer than that commonly used for post-operative patients, horses experiencing severe pain (i.e. laminitis, pleuropneumonia, orthopedic trauma) may require multimodal analgesic therapy of prolonged duration. In fact, we observed significant effects on fecal passage only in the last 24 hours of the 96-hour period with butorphanol, butorphanol/ketamine, and lidocaine/butorphanol/ketamine, whereas there were no substantial effects of these drug combinations (except LBK at T₂₅) on bead passage. This suggests that the effects of drug combinations on gastrointestinal motility worsen with longer duration of administration.

A significant treatment effect was not evident for any other measured parameter. Significant period and time effects were observed with rectal temperature, skin temperature and heart rate; a significant period effect was noted with body weight. The changes in skin and rectal temperature appear related to ambient temperature and/or relative humidity, as temperatures were higher in the afternoon/evening hours and during the summer months. These effects were small and likely clinically insignificant, and did not apparently affect treatment or time-treatment interactions because of the orthogonal study design.

Data for the LBK group represent only 5 horses, as the infusion was discontinued in 3 due to colic. Although decreased fecal output and delayed transit were still evident in this group, the negative effects on gastrointestinal function may have been underestimated because the 3 omitted horses likely had the most severely altered gastrointestinal function, resulting in colic. In addition, these horses still displayed signs of abdominal pain despite receiving all three putative analgesic agents together, suggesting either limited visceral analgesia with this drug combination or overriding abdominal pain. Alternatively, it is possible that affected horses experienced some degree of analgesia thus continued to eat hay despite developing ileus, until a more severe problem developed. In prior studies, we were unable to demonstrate visceral anti-nociceptive effects with lidocaine [10] or butorphanol [13] administered as a CRI to healthy horses. It is not known whether anti-nociceptive properties would exist with this drug combination in ill or painful horses.

A significant decrease in cumulative fecal weight for each combination including butorphanol was observed in the current work. We also observed a decrease in fecal weight during the final 24 hours of the 96-hour infusion for horses treated with butorphanol alone, after all of the beads had been passed. This supports prior work in healthy horses [20] and horses post-celiotomy [14], which also showed decreased fecal output. But, in healthy horses, a short duration butorphanol CRI did not significantly alter duodenal motility [13]. The observed decrease in fecal weight and prolonged passage of beads for combinations that included butorphanol may be due to numerous factors including the potential for accumulation of any drug or its metabolites during the infusion, competitive metabolism, or a potential synergistic inhibitory effect on motility of drugs given in combination. Further evaluation of the effects of butorphanol and its metabolites and lidocaine and metabolites on the equine gastrointestinal tract in combination are warranted.

The reported effects of lidocaine administration on gastrointestinal motility in the horse have been variable. Lidocaine administration following exploratory laparotomy has been associated with reduced small intestinal diameter and peritoneal fluid accumulation [8], decreased duration of gastric reflux and hospital stay [7], and reduced incidence of post-operative ileus and improved survival [21]. However, in normal horses lidocaine causes decreased jejunal motility [6] and delayed gastrointestinal transit time after prolonged administration [22]. In this study, we observed no effects of a prolonged lidocaine CRI on gastrointestinal transit, except when combined with butorphanol.

In human and small animal intensive care units, sub-anesthetic doses of ketamine are commonly used for multi-modal pain management and it is believed that ketamine has few adverse systemic effects [23]. Sub-anesthetic doses of ketamine, when administered as constant rate infusions, significantly decreased overall gastrointestinal transit time in comparison with saline control in a prior study [19]. In the study reported here, a lower dose of ketamine was used with no observed adverse effects on behavior scores or total gastrointestinal transit. However, along with the lack of adverse effects, there were no demonstrated anti-nociceptive effects associated with infusions of any drug or combination at dosages used in this study. The lack of antinociceptive effects of ketamine may be due to the lack of central sensitization in these non-painful animals. Further studies in animals with existing pain are warranted and may assess the antinociceptive effects of ketamine infusion more accurately.

The results of this study, although performed in clinically normal horses, have potential direct clinical application. First, the negative effects of drug combinations on fecal weight began within 36 hours, but apparently worsened with prolonged infusion duration. Second, 3 of the 8 clinically normal horses receiving a combined infusion of lidocaine, butorphanol, and ketamine developed colic. Lidocaine and butorphanol in combination also resulted in a severe delay in gastrointestinal transit and reduced fecal weight, but not colic. Finally, despite the observed delay in gastrointestinal transit, there was no apparent effect on gastrointestinal borborygmi scores. This highlights the poor sensitivity of gastrointestinal borborygmi as a sole clinical marker of gastrointestinal motility. These results are consistent with those of a previous study that showed poor agreement of auscultation scores with electrointestinography [24]. Thus, clinicians using lidocaine, butorphanol, and ketamine in combination should carefully monitor fecal output and be aware of the potential for colic in these patients. The reader should also note that the doses studied likely represent the maximal dosage used, and the effects upon gastrointestinal motility may be limited if lower dosages are used in combination. The lack of observed thermal threshold effects likely represent a lack of drug effect on thermal nociception at doses used in this study. Thermal nociceptive threshold testing is only one measure of nociception and may poorly recapitulate severe clinical pain. Further work is required to determine the effects of these drugs in combination on nociception and gastrointestinal motility in critically ill and painful horses.