Materials and Methods

Horses—The study was conducted with 2 groups of horses. Group 1 was composed of 11 horses with a history of hind limb lameness and 5 nonlame horses. The nonlame horses were used for force-plate data acquisition only. Lame horses in group 1 underwent a complete lameness examination, and force-plate data were acquired; radiographic examinations of the tarsal joints were subsequently performed to determine the joints affected. Following intra-articular analgesia, lameness and force-plate examinations were repeated. On the basis of contrast arthrographic findings, joints were or were not injected with alcohol. Group 2 was composed of another 10 horses with a history of hind limb lameness. These horses underwent the procedure for alcohol-facilitated joint ankylosis after an initial lameness and radiographic examination (including intra-articular analgesia) and contrast arthrographic evaluation without initial force-plate examination and serial radiographic examinations. Treatment assessment for group 2 horses included information collected via telephone conversations with owners (n = 3) and reports of follow-up examinations performed by either the primary author (JLC) or veterinarians in practices where the primary author had performed the technique (7). The study was approved by an institutional animal care committee, and owners also completed an informed consent form prior to enrolling their horses in the study.

Lameness examination—The American Association of Equine Practitioners' lameness scale from 0 to 5 (in which 0 is nonlame [sound] and 5 is non—weight bearing) was used during assessment of all horses. The use of scores in increments of 0.5 within the American Association of Equine Practitioners' framework was allowed when the examining clinician thought that severity of the lameness was intermediate between 2 ordinal scores. This type of subdivision provides greater flexibility² and has been previously reported in the peer-reviewed veterinary medical literature.^9,10

Force-plate data acquisition and radiographic examination—Subsequent to the lameness examination, a quantitative gait analysis of each horse in group 1 was performed with an in-ground stationary force plate^b embedded in a 20 × 3-m runway. The data acquisition rate was set at 1,000 Hz. The force plate and runway were covered by mats to prevent horses from slipping. Each horse was led over the force plate at a consistent speed (1.8 to 2.2 m/s), which was confirmed via 2 photoelectric switches that were placed 5 m apart on either side of the force plate. Five valid repetitions were recorded to ensure precision. A valid foot strike was defined as one in which the hoof fully contacted the plate and landed within a 10-cm square boundary from the edges of the force plate within the speed range. For each horse, the mean of the 5 valid measurements per limb per time point were calculated and reported as Newtons per kilogram of body weight.

The lame horses were then restrained, and a complete series of tarsal radiographic views was obtained (4 standard views). The horses were subsequently placed in stocks, and the skin over the DIT and TMT joints of the affected limb (or bilaterally if necessary, as determined from results of the radiographic investigation) was aseptically prepared. A lip-twitch was applied, and intra-articular analgesia was achieved via injection of 3 mL of mepivacaine hydrochloride^c into each joint by use of a routine approach to arthrocentesis. In some instances, joints could not be accessed because of the severity of disease. After 20 minutes, the lameness and force-plate examinations were repeated.

Contrast arthrographic evaluation and joint injection with alcohol—Having localized the source of the lameness to the distal tarsal joints (by detection of a decrease in ASI calculated from force-plate data), the lame horses in both groups were sedated and a contrast arthrographic evaluation of the TMT or DIT (or both, based on radiographic evidence of disease) joint was performed following injection with 3 mL of iohexol 52% radiographic contrast material.^d Needles remained in situ during the interpretation of the contrast arthrographic findings. If there was no communication with the proximal intertarsal or the tibiotarsal joint, 3 mL of 70% ethyl alcohol (prepared in-house from 95% laboratory-grade ethyl alcohol)^e was immediately injected via the needles used for arthrographic evaluation. If there was suspicion or overt evidence of joint communication, needles were withdrawn without injection of alcohol. All horses received 2 g of phenylbutazone^f IV after joint injection. All owners continued using their horses as if no treatment had been given.

Postinjection assessments—Owners were instructed to observe their horses closely for 48 hours and to contact the primary investigator within 7 days to report on their progress. The lame group 1 horses were reexamined at 3, 6, 9, and 12 months after alcohol treatment, and at each time point, lameness, force-plate, and radiographic examinations were performed. Follow-up information for group 2 horses was obtained via reexamination by either the primary author or an experienced equine veterinarian at the practice where the horse had been treated (or failing that, through telephone conversation with owners at 1 year after treatment). No other treatments for bone spavin (eg, corrective shoeing or administration of NSAIDs or other analgesic agents) were permitted in either group.

Force-plate data analysis—The following data from the force-plate examination were used for analysis. For unilaterally lame horses, the peak vertical forces in the hind limb were used to calculate an ASI (ASI-F_zHPeak) with the method created by Weishaupt et al.¹¹ The ASI is calculated as the difference between the peak vertical force of the left and right limbs divided by the sum of the peak vertical forces of the left and right limbs (results are expressed as a percentage). Thus, symmetric peak forces yield an ASI value of zero, whereas a higher peak force in either the left or right limb yields a positive or negative ASI value. In nonlame horses, the absolute force ASIs are generally < 5%.^11,12 For horses with bilateral hind limb disease, a force-based ASI was not considered appropriate because both lame limbs generally had the same loading patterns. Due to the nature of hind limb bilateral lameness, peak forces and impulses in the hind limbs of these horses were generally unchanged over the course of the study; therefore, no force-plate data from these horses were used to assess lameness.

Statistical analysis—Statistical analysis was performed with a commercial software package,^g with the level of significance set at a value of P < 0.05. Descriptive statistics were applied to the data, which were also examined for normality of variance. A repeated-measures ANOVA (controlling for horse) was used to examine the mean ASI-F_zHPeak over time, and if time was a significant factor, multiple t tests were subsequently used to determine the time point at which this occurred. A Kruskal-Wallis nonparametric ANOVA was used to determine whether there was a significant difference in ASI-F_zHPeak between the various modified lameness grades. A κ statistic was subsequently applied to a number of subjective ASI-F_zHPeak cutoff points for the distinction of lame versus nonlame states to determine the ASI-F_zHPeak value at which observers could reliably state that horses were not lame.

Results

Group 1 horses—None of the 5 horses used for force-plate validation were subjectively determined to be lame (all grade 0), a finding that was further substantiated by use of data acquired from the force plate. In these horses, hind limb ASI-F_zHPeak ranged from 0.13% to 3.85%, indicating that either complete symmetry was not a prerequisite for a diagnosis of a nonlame horse or that a subclinical lameness was present.

Of the 11 lame horses in group 1, 8 were castrated males, 2 were females, and 1 was a stallion (Table 1). The mean ± SD age was 8.23 ± 3.96 years (range, 1.5 to 13 years). Median lameness score at the start of the study was 3.0 (range, 3.0 to 4.0; interquartile range [25% to 75% percentiles], 3.0 to 3.0). On the basis of clinical and force-plate data, 7 horses were unilaterally lame (Figure 1) and 4 were bilaterally lame. However, radiographic examination indicated that 3 of the unilaterally lame horses had evidence of bilateral osteoarthritis. These horses were analyzed as unilaterally lame horses for the purposes of force-plate data examination despite having joints of both hind limbs treated intra-articularly with alcohol.

View Full Size

Mean ± SD peak hind limb vertical ground reaction forces for the affected lame hind limb (gray boxes) and unaffected hind limb (white boxes) in 7 horses with unilateral lameness referable to the DIT joint, TMT joint, or both joints before (0 months) and after injection of the affected joint with 3 mL of 70% ethyl alcohol. Three of these group 1 horses appeared unilaterally lame but had radiographic evidence of osteoarthritis in those joints bilaterally. Peak vertical ground reaction forces changed in both the affected and unaffected hind limbs in the unilaterally lame horses during the study, with a gradual increase in force in the affected hind limb and a gradual decrease in force in the unaffected hind limb. Peak vertical ground reaction forces in affected and unaffected hind limbs differed significantly (P < 0.05) at months 0, 3, and 9.

Citation: Journal of the American Veterinary Medical Association 240, 2; 10.2460/javma.240.2.199

• Download Figure
• Download figure as PowerPoint slide

Thirty joints (17 TMT and 13 DIT joints) in 11 horses had radiographic evidence of osteoarthritis. Contrast arthrographic evaluation was attempted in all joints. All 17 TMT joints were accessed, and none indicated a communication with either the DIT or the tibiotarsal joints. Of the 13 DIT joints undergoing contrast arthrographic evaluation, 4 could not be entered because of the degree of periarticular new bone. Of the remaining 9 DIT joints, 5 had no evidence of proximal joint communication and 4 had evidence of communication with the proximal intertarsal and tibiotarsal joints.

One horse was reported by the owner to have developed edema surrounding the injection site (TMT joint). No other complications were noted in this or any other horse in either of the 2 distinct study populations.

All of the horses in group 1 had a significant (P = 0.002) reduction in lameness after injection with alcohol. Peak vertical ground reaction forces changed in both the affected and unaffected hind limbs in the unilaterally lame horses during the study (Figure 1), with a gradual increase in force in the affected hind limb and a gradual decrease in force in the unaffected hind limb. Results of a repeated-measures ANOVA (controlling for the unilaterally lame horses) indicated that there was a significant (P < 0.001) reduction in mean ASI-F_zHPeak over time (Figure 2). The period determining this significance was between initial examination and 3 months after alcohol injection (P = 0.02). Owners were contacted via telephone follow-up within 7 days after injection, and most indicated that an improvement in lameness had been detected within 48 hours following treatment. At 12 months, 10 of 11 horses were free of lameness and 1 was mildly lame (lameness grade, 0.5; Figure 3).

View Full Size

Mean ± SD ASI values (calculated on the basis of force-plate data) for the 7 horses in Figure 1 with unilateral lameness referable to the DIT joint, TMT joint, or both joints before (0 months) and after injection of the affected joint with 3 mL of 70% ethyl alcohol. Results of a repeated-measures ANOVA (controlling for the unilaterally lame horses) indicated that there was a significant (P < 0.001) reduction in the mean ASI-F_zHPeak over time; the period determining this significance was between initial examination and 3 months after alcohol injection (P = 0.02).

Citation: Journal of the American Veterinary Medical Association 240, 2; 10.2460/javma.240.2.199

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Mean lameness grade among 11 horses (group 1) with unilateral or bilateral lameness referable to the DIT joint, TMT joint, or both joints before (0 months) and after injection of the affected joint with 3 mL of 70% ethyl alcohol. Lameness was assessed by use of the American Association of Equine Practitioners' lameness scale from 0 to 5 (where 0 is nonlame [sound] and 5 is non-weight bearing); use of scores in increments of 0.5 within the American Association of Equine Practitioners' framework was allowed when the examining clinician thought that the severity of lameness was intermediate between 2 ordinal scores. Error bars denote maximum values. Mean lameness grade at 0 months differed significantly (P < 0.05) from the value at each other time point. Mean lameness grades at 3 and 6 months did not differ from each other, but both were significantly different from the value at each other time point. Mean lameness grades at 6 and 9 months did not differ from each other, but both were significantly different from the value at each other time point.

Citation: Journal of the American Veterinary Medical Association 240, 2; 10.2460/javma.240.2.199

• Download Figure
• Download figure as PowerPoint slide

There was a highly significant (P < 0.001) association between modified lameness grade and mean ASI-F_zHPeak, indicating that subjective assessment was reflected in the objective determination of lameness as indicated by the force-plate data. When κ statistics were applied to a subjective determination of either a lame or nonlame state by use of arbitrary ASI-F_zHPeak, cutoff values of 3.5%, 4%, and 5%, results were 0.417, 0.541, and 0.608, respectively, indicating moderate agreement in the first 2 instances and substantial agreement in the latter.¹³

Radiographically, there was a substantial subjective reduction in the width of the joint space of alcohol-injected joints over time. Although treated joints in all group 1 horses were deemed fused, there was evidence of a joint space remnant in at least 1 standard radiographic view obtained from each horse at each time point.

Group 2 horses—Of the horses in group 2, 3 were stallions, 6 were geldings, and 1 was a mare (Table 2). Follow-up clinical examinations (n = 7) or telephone conversations with owners (3) of the horses were performed a mean interval after treatment of 11 months (range, 2 to 24 months). One horse was being overwintered on pasture and could not be returned for follow-up examination. Seven of the remaining 9 horses were not lame, and 2 horses remained mildly lame. One of the lame horses had a concurrent upward fixation of the patella in the injected limb, which was treated with medial patella ligament splitting; the other had severe pathological changes in a DIT joint (not previously injected or treated by the authors), which could not be injected because the periarticular bone prevented needle access.

Discussion

In the present study, joint injection of alcohol in horses with lameness attributable to arthritis of the distal tarsal joints resulted in a rapid reduction in lameness, such that horses could return to their intended use. Lameness resolved in 18 of the 21 treated horses, and 3 remained mildly lame. Up to the last update, (approx 18 months after the first horse entered the study), none of the horses had required joint reinjection for the purposes of pain resolution. Although 1 horse developed some edema surrounding the site of injection, no treatment was necessary. This was the only complication noted during the study.

Contrary to results of an experimental study by Shoemaker et al,⁸ the horses that underwent alcohol-facilitated ankylosis did not develop the degree of joint fusion expected. There are, however, 3 fundamental differences between the previous experimental study and the clinical trial reported here. In the former protocol, none of the horses underwent contrast arthrographic evaluation, they were all 3 to 4 years of age, and none had clinical disease. Explanations for the disparity in results may be as follows: there was a possible interaction between the alcohol and the contrast agent (either chemical or spatial) following joint injection; the younger horses in the experimental study⁸ had fundamentally different cartilage or subchondral bone plate characteristics, compared with those of the older horses, as described by Murray et al¹⁴; or there was an undetermined difference in the action of ethyl alcohol in osteoarthritic versus normal joints.

It has also been reported⁸ that radiographic evidence of joint fusion underestimates the true prevalence of ankylosis or fusion at 12 months, compared with results of postmortem examination. Initial protocols for the present study intended that horses would undergo a computed tomographic examination at the end of the 12-month study period in an attempt to correlate the extents of joint fusion determined via radiographic evaluation and computed tomographic imaging. Unfortunately, but not unexpectedly, most clients were unwilling to subject their horses to the risks of general anesthesia, especially because they were no longer lame at that time.

The significant association between subjective lameness score and mean ASI was welcomed by the authors, as it reinforced our ability to assign a modified lameness grade to the horses that was reflected in the objective assessment. This has repercussions separate from the present study, with potential application in clinical practice. We further determined that a mean ASI cutoff value between 4% and 5% was useful in the determination of a lame versus nonlame state in horses, a value previously supported by information in the veterinary medical literature.^11,12

Although the results of the present study suggested that a single intra-articular injection of alcohol results in rapid and sustained reduction in lameness, the treated horses in group 1 were followed up for only 12 months. However, a group 1 horse had been monitored for 18 months and continued to be free of lameness attributable to the distal tarsal joints, and treated horses in group 2 were followed up for as long as 24 months with no return of lameness.

Undoubtedly, results of the present study could create a bias toward intra-articular injection of ethyl alcohol as a first-line treatment when osteoarthritis of the DIT or TMT joint is confirmed. Although this treatment may seem encouraging, the authors suggest caution. In discussions with an investigator involved in 2 studies^6,a of joint fusion via intra-articular injection of sodium monoiodoacetate, it was indicated that several of the treated horses that had negative contrast arthrographic findings (ie, no evidence of communication of the DIT or TMT joint with the tibiotarsal joint) developed osteoarthritis of the tibiotarsal joint 5 to 10 years after treatment.^h Possible conclusions were that a small interjoint communication was overlooked during review of the contrast arthrograms, the treatment could have stimulated an inflammatory response that triggered osteoarthritis in the tibiotarsal joint, or restricting movement of the DIT and TMT joints resulted in abnormal loading of the tibiotarsal joint such that osteoarthritis developed. The latter argument is not robust because there is no evidence in the veterinary medical literature that arthritis in proximal joints develops after horses undergo fusion (surgical or chemical) of the distal tarsal joints; however, long-term follow-up reports are lacking. Clearly, if the degree of pathological change is refractory to medications that are routinely injected intra-articularly, then injection of ethyl alcohol may be the only noninvasive option remaining, but care should be exercised in case selection.

Similarly, if only one of the distal tarsal joints (either the DIT or TMT joint) had clinical evidence of disease, the debate would be whether to treat the affected joint only or to simultaneously treat the affected joint and the ipselateral nonaffected joint with ethyl alcohol (assuming the contrast arthrogram did not indicate communication with the tibiotarsal joint). It is our opinion (although not practiced with the horses included in the study of this report) that when a horse with clinical evidence of disease in only one of the distal tarsal joints participates in sports that predispose to the development of osteoarthritis of the tarsal joints (eg, reining and cutting), then both the DIT and TMT joints of the affected limb should be treated at the same time.

Of the 13 affected DIT joints in the present study, only 5 were injected (4 could not be entered because of interference from new bone deposits, and 4 communicated with the proximal intertarsal joint). Despite the inability to medicate those affected joints, lameness ultimately resolved in all 8 horses. When both joints (DIT and TMT) had radiographic evidence of osteoarthritis initially, then diagnostic intra-articular analgesia was performed on both joints simultaneously (prior to completing the second force-plate data acquisition) in an effort to expedite the examination process. Thus, the pain contribution from each joint could not be assessed independently. The reason for the resolution of lameness in horses in which affected joints were not injected remains speculative. Possible explanations are that the pain component was primarily from the TMT joint; therefore, when ankylosis was achieved in these joints, the horses appeared nonlame. Alternatively, there may have been a communication between DIT and TMT joints that was inapparent via contrast arthrographic evaluation such that alcohol entered the DIT joint and effected ankylosis. One mechanism of action, namely that therapeutic concentrations of ethyl alcohol diffused into the DIT joint following TMT joint injection as has been detected following methylprednisolone injection of TMT joints,¹⁵ is unlikely. The drug causes nonspecific protein denaturation and cell protoplasm precipitation with dehydration, all of which occur at the drug-tissue interface. In addition to this, if clinical concentrations of ethyl alcohol were to diffuse out of the injected joint, then evidence of proximal intertarsal or tibiotarsal joint changes might have been observed (subsequent to injection) in the horses of the present study and those in the study by Shoemaker et al,⁸ which was not the case.

The results of this investigation of alcohol-facilitated ankylosis indicated that in contrast to current treatments, horses receiving an intra-articular injection of alcohol undergo a rapid and sustained reduction in lameness such that they can be trained (ie, brought into work) or compete within days following treatment. In the present study, owners were satisfied with the outcome of the horses' treatment, and within the initial 18-month period after treatment, we have had to administer only 1 injection.