Materials and Methods

Subjects—A total of 332 Thoroughbred racehorses consecutively admitted to Rood and Riddle Equine Hospital and treated by SCLD for SDFT over a 15-year period (1989 to 2003) were followed-up through the end of their racing career or for a minimum of 4 years in the most recently operated horses. Superficial digital flexor tendonitis diagnosis was confirmed by 2 clinicians (the referring veterinarian and the admitting surgeon) and documented ultrasonographically. If a recent or diagnostic ultrasonographic examination was not available for the horse at the time of evaluation, a third examiner confirmed the diagnosis and tendon damage via ultrasonography. The grading system used by Genovese et al⁵ of lesion grades I to IV was used to determine surgical candidates in this group of horses.¹² All horses selected for surgery had grade II, III, and IV, moderate to severe, ultrasonographic lesions.⁵ Grade I lesions were treated medically rather than surgically (data not shown). Breed, age, sex, limb affected, and postoperative complications were determined from the medical records. All horses were evaluated for treatment when the trainers, owners, or veterinarians deemed the degree of tendonitis to be limiting to the horse's performance. Most horses had acute clinical signs, although some were acute exacerbations of previous episodes. The duration of clinical signs and information on number of episodes of tendonitis were not accurately discernible and varied widely. Ultrasonography was used to document evidence of acute inflammation in all horses.

Surgical procedures—All surgeries were done at Rood and Riddle Equine Hospital via the same technique; 312 of the 332 horses were done by 1 surgeon (LRB). The remaining 20 surgeries were performed by 5 other surgeons at our hospital. Superior check ligament desmotomy via a medial approach through the flexor carpi radialis tendon sheath was performed in 1 or both forelimbs.¹⁶ Horses with < 5 starts at the time of injury were prophylactically operated on the contralateral limb during the same procedure. This was routine protocol for surgery for horses that had not completed ≥ 5 starts during the time of this study.

Postoperative care and rehabilitation—In the perioperative period, horses received potassium penicillin (20,000 U/kg [9,091 U/lb], IV, q 6 h for 24 hours), gentamicin (6.6 mg/kg [3.0 mg/lb], IV, q 24 h for 2 doses), and phenylbutazone (6 mg/kg [2.73 mg/lb], IV, q 24 h before surgery, then 2 mg/kg [0.9 mg/lb], PO, q 12 h for 10 days after surgery). Postoperative care consisted of full limb bandaging until suture removal at 10 to 12 days after surgery. Recommended rehabilitation consisted of stall rest during the first 2 weeks, followed by 2 weeks of stall rest and hand walking, and then paddock exercise for 4 weeks. At 60 until 120 days after surgery, 10 minutes of jogging under tack 3 times weekly in addition to free-choice paddock exercise was recommended. Follow-up ultrasonography was recommended at 120 days, with images evaluated by the surgeon or referring veterinarian. If the tendon contained no lesions above grade I and fiber alignment was considered good at this time, horses were allowed to return to normal training schedules. If this level of healing was not achieved, horses were reexamined monthly until the tendon had improved to this extent.

Performance analysis—The lifetime performance records of all 332 horses were obtained for a minimum of 4 years after surgery.^a All horses were categorized simply as raced or unraced after surgery. No horses were excluded from analysis for any reason. All horses were grouped according to whether they raced prior to injury (starters) or did not race prior to injury (nonstarters) for comparison between these categories and followed-up through the end of their racing careers or for a minimum of 4 years after treatment, whichever was longer, to determine the number of races competed after injury and treatment. The days from surgery to first postoperative race were calculated. For descriptive statistics (sex, age, and surgery), values of P were determined with the χ² test to compare differences in the proportions of return to racing between starters and nonstarters within these categories. The Wilcoxon rank sum test was used to evaluate return to racing statistics and postoperative success for starters versus nonstarters in the data categories. Values of P < 0.05 were considered significant. Statistical analyses were performed with a commercial software package.^b

Results

Descriptive statistics—Of the 332 Thoroughbred racehorses in this study, at the time of surgery, there were 96 (29%) 2-year-olds, 139 (42%) 3-year-olds, 58 (17%) 4-year-olds, and 39 (12%) horses ≥ 5 years old (Figure 2). Of the 332 horses, 73 (22%) were sexually intact females and 259 (78%) were sexually intact males or castrated males. Fifty-eight of the 332 (17%) horses underwent unilateral SCLD surgery on the right forelimb and 84 (25%) on the left forelimb; 190 (57%) had bilateral surgery. Prior to injury, 214 of the 332 (64%) horses were starters in a race and 118 were (36%) nonstarters.

Return to racing—Of the 332 horses, 228 (69%) raced following injury and surgery and 104 (31%) did not. Comparatively, 78 of 118 (66%) nonstarters raced following injury and surgery and 150 of 214 (70%) starters returned to racing after surgery. There was no difference (P = 0.46) between the starters and nonstarters in their ability to resume racing (Table 1).

View Full Size

Age (A) and sex (B) distribution of 332 Thoroughbred racehorses with SDFT consecutively treated by means of SCLD from 1989 to 2003 grouped according to age and sex and whether they raced prior to injury (starters) or did not race prior to injury (nonstarters).

Citation: Journal of the American Veterinary Medical Association 244, 12; 10.2460/javma.244.12.1441

• Download Figure
• Download figure as PowerPoint slide

Association between sex and age and return to racing—The horses were analyzed for sex, age, and racing status prior to injury and the effect on return to racing. For simplicity, all sexually intact male and castrated male horses were considered together. For the variable sex, 181 of 259 (70%) males and 47 of 73 (64%) females returned to racing; there was no difference (P = 0.39) in percentage of return to racing for males or females. Fifty-two of 76 (68%) males that were nonstarters prior to injury went on to race. Of 183 males that raced prior to injury, 129 (70%) raced following injury and surgery (P = 0.77). Twenty-six of 42 (62%) females that were nonstarters prior to injury raced after injury, and 21 of 31 (68%) female starters returned to racing following injury and surgery (P = 0.63). None of the combined variable pairs of sex and starting status were different from each other.

Seventy-two of 96 (75%) 2-year-olds, 96 of 139 (69%) 3-year-olds, 43 of 58 (74%) 4-year-olds, and 17 of 39 (44%) horses ≥ 5 years old raced following injury and surgery. The horses ≥ 5 years old had a lower rate of return to racing (P = 0.003). There were no differences among the percentage of 2-, 3-, and 4-year olds that returned to racing. For nonstarters, 53 of 76 (70%) 2-year-olds, 33 of 37 (89%) 3-year-olds, 2 of 4 four year olds, and a single 5-year-old raced following injury and surgery. Of the starters, 19 of 20 (95%) 2-year-olds, 74 of 102 (73%) 3-year-olds, 40 of 54 (74%) 4-year-olds, and 16 of 38 (42%) horses ≥ 5 years old returned to race following injury and surgery. Two-year-olds that had previously raced had a better rate of return to racing than did 2-year-old nonstarters (P = 0.039). There was no significant difference between starters and nonstarters in any of the other age groups.

Association between limb and return to racing—Of the 332 horses, 114 (34%) sustained SDFT injury on the right forelimb and 159 (48%) on the left forelimb; 59 (18%) had bilateral SDFT injuries (Table 2). Of 332 horses, 58 (17%) underwent surgery on the right forelimb and 84 (25%) on the left forelimb; 190 (57%) had bilateral surgery performed (P = 0.86). Forty-one of 58 (71%) horses that had surgery performed on the right forelimb returned to racing following injury and surgery. Of the horses that had right forelimb surgery, 5 of 7 nonstarters and 35 of 51 (69%) starters raced (P = 1.0). Fifty-nine of 84 (70%) horses that had surgery performed on the left forelimb returned to racing. Of these horses, 4 of 7 nonstarters and 55 of 77 (71%) starters returned to race (P = 0.67). For bilateral surgery, 128 of 190 (67%) the horses that had bilateral surgery raced. Of the horses that had bilateral surgery, 69 of 104 (66%) nonstarters and 59 of 86 (69%) starters raced. Bilateral surgery had no effect on the return to racing, compared with surgery on 1 limb (P = 0.79).

Postoperative infections—Six of the 332 (2%) operated horses (6/522 [1%] procedures) developed a postoperative infection at the surgical site; 5 cases resolved with antimicrobial treatment. One horse was euthanized owing to extension of infection into the carpal canal. Two horses subsequently raced, and 3 did not race.

Convalescence—For the 228 of 332 (69%) horses that raced following injury and surgery, the median time to first start was 302 days (range, 48 to 1,120 days; mean ± SD, 341 ± 153.4 days). The median time to first race for the 78 nonstarters that returned to race following injury and surgery was 324.5 days (range, 68 to 976 days; mean, 350 ± 155.7 days) and for the 150 horses that raced before surgery was 298.5.days (range, 48 to 1,120 days; mean, 336 ± 152.5 days). Convalescent time was not significantly (P = 0.33) different between starters and nonstarters.

Career longevity after injury and surgery—Horses that raced following injury and surgery had a median of 8 starts (range, 1 to 109 starts; mean ± SD, 14.2 ± 15.8 starts). Nonstarters had a median of 7 starts (range 1 to 109 starts; mean, 13.5 ± 16.6 starts) following injury and surgery. Starters had a median of 8 starts (range, 1 to 69 starts; mean, 14.5 ± 15.3 starts) following injury and surgery. Of 228 horses that raced, 159 (70% [48% of all horses]) raced ≥ 5 times after surgery; 69 of 228 (30%) horses returned to race but competed in < 5 races.

Discussion

This study was undertaken to objectively document the likelihood of return to racing, career longevity, and convalescent time of horses with moderate to severe (grade II to IV) SDFT treated by means of SCLD. Grade I lesions were treated medically during this time period due to the better prognosis previously documented for these less severe lesions.⁵ Results for this large group of flat-racing Thoroughbreds offer a starting place for comparison of the outcome of surgical SCLD treatment of tendonitis with other treatments for SDFT via objective quantitative documentation of results. In this study, the number of races as determined by the horse's race record was used as an objective measure of performance. In 2 more preliminary studies^7,10 on SCLD as a treatment for SDFT in horses, the results were encouraging. In one study⁷ inability to complete 2 races was used as the failure point; in the other study,¹⁰ recurrence rate was used as a failure point. However, analysis of recurrence proved to be very difficult because results of ultrasonographic examinations were not available on all horses, the investigators were unable to access all horses, and telephone interviews proved inaccurate when compared against race records. Therefore, in the present study, 332 horses were followed up through the end of their career or through a minimum of 4 years on the basis of objective third-party racing data. It could be argued that some horses raced longer than the analysis period and that the results are thus understated, but the most recently operated horses in the study were a minimum of 6 years old at the time of the record acquisition (at least 2 years old at injury, plus 4 complete years of follow-up), minimizing the number of uncounted starts beyond the analysis period. Removal from analysis was not allowed, although some horses would have had their careers terminated by other diseases or injuries. It is impossible to document accurately the exact reasons that each of the horses did not resume performance, so horses were simply categorized as having raced or not raced, counting horses that did not race for any other reason (disease, recurrent or unrelated injury, or economic reasons) as did not race. This gives an accurate representation of the likelihood of an injured horse returning to productive racing.

Postmortem examination of SCLD surgical sites has shown that the transected check ligament is reconstituted during the healing process, but the transected ligament is lengthened after surgical transection⁷ (Figure 1). This increased the length of the check ligament–superficial digital flexor tendon unit (radius to second phalanx), by a finite amount and presumably shifts some of the load to the suspensory ligament, increasing the risk to that structure. Experimental studies^8,9 have demonstrated this to occur acutely in cadaver limbs. Acute transient alteration of carpal anatomy with transection of the check ligament has been noted clinically, but no clinical evidence of carpal injuries in horses following SCLD has been documented^{1,8,9,11–14} and the carpal anatomy returns to normal within days after the surgery. As a medical treatment, β-aminopropionitrile fumarate lowers ipsilateral reinjury rates, compared with lesions treated with only controlled exercise and intralesional and systemic polysulfated glycosaminoglycans.⁴ However, injuries to the contralateral limb offset the benefits of treatment with β-aminopropionitrile fumarate.⁴ Data in that group of horses demonstrated contralateral injury as a significant reason for nonperformance. An increase in the occurrence of suspensory ligament desmitis in horses racing after SCLD, as has been described,^1,12 can also limit performance. We have also seen suspensory ligament desmitis occur occasionally in horses that have had previous SCLD performed and have returned to racing. In our clinical experience, it is infrequent; however, these predispositions to other injuries demonstrate the need to assess horse performance following injury and treatment rather than look solely at the reoccurrence of tendonitis. If the induced injuries substantially reduce the horse's ability to race, the treatment simply trades one cause of retirement for another. Therefore, to assure that any effect of the injury and the treatment was accounted for, no horses were excused from the analysis and the documentation of racing was objectively determined by a nonbiased third party.^a

Some studies exclude horses from analysis when they are retired from training or change careers prior to racing. This markedly alters the interpretation of the results. In 1 study,¹² 37% of the horses that were treated medically were removed from analysis because of retirement or were lost to follow-up. In the same study,¹² 12% of the horses in the SCLD-treated group were similarly excused. If these horses had been retained, the interpretation may have been different.¹² One would expect similar groups of horses to have similar interfering conditions. In another study,⁶ 20% of horses were removed from analysis for the same reasons. Similarly, in a third study,¹⁸ in which the quality of return to racing was evaluated, 209 of 401 (52%) horses were removed from analysis because of loss to follow-up or retirement. Retirement and loss to follow-up can be an indication of poor response to treatment; elimination of these horses alters the return to racing data. External interference, such as other disease, other lameness, or retirement for economic reasons occur, but separation of these conditions from any relationship to the initial tendonitis can be difficult and is subjective, so rather than make a subjective judgment, we retained all horses in the study; if horses did not race for any reason, they were counted as not having raced. The racing documentation system is so sophisticated that if a horse races, it is recorded objectively by a third party.^a If a horse is not recorded, it can be certain that horse did not race. Use of race data removes all author or treating or examining veterinarian bias and is more accurate than phone conversations with the horse's connections,¹⁰ in the authors' experience.

The check ligament–tendon unit assumes load as weight is applied to the forelimb and transmitted to the palmar support structures of the fetlock joint. In vitro studies^8,9 that have looked at strain in frozen and then thawed limbs show acute increased strain of the tendon and the suspensory ligament when the check ligament is cut and the fetlock joint increases its excursion. Investigators in those studies^8,9 have assumed that the increased strain is detrimental to the flexor tendon.

The premise of the SCLD surgery is to cause lengthening of the proximal portion of the tendon unit by disabling the check ligament. Transection acutely reduces the amount of resistance to loading of the fetlock joint that the tendon unit with the intact check ligament provides, leaving the remaining portion of the SDFT that is attached to the muscle proximal to the superior check ligament and the suspensory ligament to support the fetlock joint. This compromised support system allows increased excursion of the fetlock joint, increasing the palmar strain on the suspensory ligament and remaining superficial digital flexor unit.^8,9

In live horses, the viable muscle fibers and tendon proximal to the check ligament can elongate with the increased load. This is the basis of the use of SCLD and inferior check ligament desmotomy for treatment of contraction of the flexor tendons. Transection of the superior check ligament results in an acute increase in strain experimentally in cadaver limbs,^8,9 but the frozen and thawed limbs used in the cadaver limb experiments would have no viable muscle to elongate and no time was allowed for this to occur. In the clinical case of contracted superficial digital flexor tendon, the response to check ligament desmotomy is not instantaneous lengthening. The proximal portion of the superficial digital flexor lengthens gradually during the postoperative period. We believe this same process occurs in this application. The gradual lengthening presumably reduces the increased strain in the superficial digital flexor back to or possibly below normal as lengthening of the superficial digital flexor unit occurs.

As the check ligament heals (Figure 1), the check ligament partially reassumes its role as part of the superficial digital flexor unit. However, a longer check ligament results in decreased efficiency of load absorption, compared with the shorter intact check ligament–tendon unit. This decreased efficiency would transfer some of the load to the suspensory ligament as the fetlock joint is loaded.^8,9 If this load shift partially protects the injured tendon, it requires more of the suspensory ligament as training resumes. Clinical data of increased incidence of suspensory ligament desmitis as the horses resume racing would support this hypothesis.^11,12

In horses we have observed, and which have been reported^1,12 with the described suspensory ligament desmitis on return to racing after SCLD, we regarded this as evidence that the superficial digital flexor was being unloaded and the suspensory ligament was assuming added load as horses were performing. Considering that the weight and exercise of a horse after surgery would be similar to before surgery, if the superficial digital flexor tendon is unloaded, that load has to be transferred somewhere and that would most likely be to the suspensory ligament. This is the treatment strategy of making the superficial digital flexor tendon unit less efficient and expecting the suspensory ligament to take some of the load. If this were a disabling effect of the surgery in all horses, it would eliminate the usefulness of the surgery. The deep digital flexor probably participates in this load shift to a lesser degree, but it has not been studied experimentally. The previous study¹² showed that the SCLD horses were more likely to race than nonsurgically treated horses and had a higher incidence of suspensory ligament desmitis and a lower incidence of recurrent tendonitis while racing. This would support the hypothesis that the SCLD was sparing the tendon and shifting the weight-bearing load to the suspensory ligament.

The reduction of load in the superficial digital flexor and the transfer of load to the suspensory ligament emphasize the need for assessment of performance (not reoccurrence) in the documentation of results. If only the reoccurrence of tendonitis is assessed, the outcome would be considered successful in a horse with suspensory desmitis. If performance is assessed, the outcome would be considered failure for these horses, when performance is limited by the suspensory ligament desmitis and therefore indirectly affected by the tendonitis and its treatment. A similar rationale would exist for assessment of any treatment which predisposes to iatrogenic ipsilateral or contralateral injury.⁴

A recent study¹⁸ documented that the completion of 5 races, as suggested in a previous publication,¹⁰ is an appropriate indicator for successful treatment and return to productive performance. We avoided the categorization as successful or failed and presented the actual number of races, but we also assessed the horses for the ability to complete 5 races. Ideally, the study would have been conducted in a blinded, randomized fashion with a control group of non-treated horses. However, the nature of clinical referral practice makes this difficult. A comparable study of flat-racing Thoroughbreds treated by other means is not available in the literature, so although the data are presented, definitive benefit cannot be assured without controls.

In the present study, 150 of 214 (70%) starters and 78 of 118 (66%) nonstarters raced following injury and surgery. Unlike some injuries, starting a race prior to injury was not a significant factor in the ability to return to racing following injury and treatment.

In this group of horses, 181 of 259 males and 47 of 73 females returned to racing following injury and surgery. No significant (P = 0.39) difference was noted between sexes. The higher breeding value potential of females likely reduced the number of females treated surgically; however, no significant difference was detected in the response to treatment. Analysis for age and starting status showed that horses ≥ 5 years old had a significantly (P = 0.003) lower (44%) likelihood of returned to racing. Age had no effect on the ability of horses ≤ 4 years old to return to racing. No significant difference was noted between 2-, 3-, and 4-year olds. Tendons of horses ≥ 5 years of age may have diminished healing capacity or may sustain injuries caused by a differing pathological process. Older racehorses in general may have a worse prognosis for any injury. This tendency may not be unique to tendonitis and may be multifactorial, but the percentage of success with SCLD in older horses was significantly lower in the present study.

Age, like sex and starting status prior to injury, had no significant effect on horses < 5 years old and also had no significant effect on likelihood to return to racing when comparing age groups year to year. A significant (P = 0.039) difference was noted between nonstarters (53/76 [70%]) and starters (19/20 [95%]) within the 2-year-olds but not in any other age group. Two-year-old starters had a higher likelihood of returning to racing following injury and surgery than did 2-year-old nonstarters. A unique reason for this difference was not readily apparent.

Convalescent time and care for injured racehorses are expensive. Recovery time for nonsurgical treatment of SDFT has been described as 7 to 9 months of stall rest with gradually increasing exercise prior to return to training.¹ Two weeks of stall rest and 2 weeks of stall rest with hand walking followed by free-choice paddock exercise for 90 days was recommended for these SCLD patients, for a postoperative convalescence time of 120 days. Not all owners adhered to this recommendation (1 horse raced 48 days after surgery); the median time to first race was 302 days (range, 48 to 1,120 days; mean ± SD, 341 ± 153 days). The median time to the first race for nonstarters (324.5 days; mean, 350 ± 155 days) was slightly longer than that for starters (298.5 days; range, 48 to 1,120 days; mean, 336 ± 152 days) as might be expected. However, no significant (P = 0.33) difference was noted. This is a much quicker to return to training and racing than recommended in the literature.^1,19 However, no detailed data on convalescent time or time to first race with any other treatment were available for comparison. Shortening the recovery time is a viable reason for considering a treatment method, especially in racehorses because the horse racing prize money available declines as horses age.

Of the total horses in the present study, most had unilateral SDFT of the left forelimb. This may be reflective of the counterclockwise direction of horse racing in the United States. Of the 332 horses, 142 (43%) had unilateral surgery and 190 (57%) underwent bilateral surgery. Bilateral surgery was routine protocol for horses that had not completed > 5 starts during the time of this study. The results of the present study do not support this protocol, and it is likely not warranted as a preventative measure because no significant (P = 0.86) difference was noted in the percentage of horses that underwent unilateral or bilateral surgery in their return to racing, regardless of racing status prior to injury in this study. The authors no longer prophylactically operate contralateral normal tendons.

In conclusion, 228 of the 332 (69%) horses in this study returned to racing, a percentage similar to that found in previous studies^7,10,14 of SCLD by the same author (LRB). In this study, of 228 horses (48% of all horses treated) that raced after surgery, 159 (70%) raced ≥ 5 times after surgery. All horses in the present study were racing Thoroughbreds, and other breeds and other uses may respond differently. Because tendon injuries vary in type and severity in various disciplines, these data may not be applicable to nonracing uses. Of note is that only 17 of 39 (44%) horses ≥ 5 years old raced following injury and treatment, so older horses or chronic injuries may not respond as well to this approach. This may limit use of SCLD in most horses in other disciplines, given that they are generally much older than this racing population.

We did not evaluate the success of treatment of tendonitis, but rather the effect of injury and treatment on the horse's ability to perform; however, in our clinical practice, this is the information the owner wants to know when assessing prognosis prior to treatment. The practicality of the surgery is low if it cannot produce a competitive athlete after surgery.

No control population of sham-operated patients with identical aftercare and case selection was compared with the horses treated with SCLD in the present study, so the treatment efficacy of SCLD in the treatment of SDFT cannot be determined from this study. Apparent treatment efficacy could have been attributable to other factors or to chance. However, the reporting of the results without exclusion gives a basis for objectively comparing horses treated by SCLD with other treatments, if they are reported similarly, when control groups are not available. If reported in this manner, data for these horses will give an objective basis for comparison of newer biological treatments currently commonly used in a similar population.

The use of objective third party data is likely superior than subjective data or data obtained by telephone interview in assessing response to treatment. Superior check ligament desmotomy is compatible with concurrent use of newer biological repair techniques, physical therapies, and rehabilitation advances. These combined treatment methods may further improve results.