Materials and Methods

Criteria for selection of cases—Medical records of the Cornell University College of Veterinary Medicine were reviewed to identify Thoroughbred racehorses examined between January 2000 and August 2004 in which SDF tendonitis had been diagnosed. Horses were included in the study if clinical signs and ultrasonographic findings confirmed a diagnosis of SDF tendonitis and the horse had been treated by means of intralesional injection of IGF-I within 13 weeks after the initial injury.

Medical records review—Medical records of horses included in the study were reviewed to obtain information on age, sex, affected limb, time between injury and treatment, concurrent injuries, and treatment. Results of ultrasonography performed at the time of initial examination and throughout the treatment course were reviewed by a board-certified radiologist (AEY). Information was recorded for lesion location (core vs margin), lesion echogenicity (scored on a scale from 0 to 4, where 0 = normal echogenicity and 4 = anechoic²⁴), lesion length (number of zones [1A to 3C] in which the lesion was visible), tendon CSA, lesion CSA, and maximal percentage of tendon CSA affected.

Treatment—For intralesional injection of IGF-I, sterile lyophilized powder^a was reconstituted with sterile saline (0.9% NaCl) solution to a concentration of 25 μg/mL and stored frozen until used. Hair was clipped from the palmar aspect of the metacarpus, and the area was prepared by means of standard aseptic technique. One or more 25-gauge or 27-gauge needles were used to administer 25 to 50 μg of IGF-I under ultrasound guidance, with needles inserted transversely every 2 cm along the length of the lesion. The exact amount of IGF-I administered was determined by the size of the lesion. Lesions extending up to and including 1 zone (approx 4 cm) were treated with 25 μg of IGF-I, whereas longer lesions were treated with 50 μg of IGF-I. A bandage was applied following injection of IGF-I, and injections were repeated every 2 days for a total of 4 or 5 injections, depending on initial lesion severity. Tenoscopic transection of the accessory ligament of the SDF tendon was performed under general anesthesia in some horses, as described.²⁵ In these horses, the first IGF-I treatment was performed at the time of surgery.

Recommendations for exercise during the rehabilitation period varied depending on lesion severity and followed published guidelines.³ Reassessment by means of clinical and ultrasonographic examination was recommended 3 and 6 months after treatment, prior to initiation of race training, after the first breeze, and after the first race. However, owner compliance with these recommendations varied.

Outcome assessment—Pretreatment and post-treatment race records were retrieved from an online database.^b The number of race starts before and after treatment, the value of each race, and the horse's earnings for each race were recorded. Attempts were made to contact the owner or trainer of each horse to determine whether the horse reinjured the tendon and, for horses that had been retired, the reason for retirement.

Statistical analysis—For horses that returned to racing, descriptive statistics were calculated for number of races after treatment, time to first race, posttreatment racing performance (number of races and earnings/race), and rates of reinjury, unrelated musculoskeletal injuries, and retirement for other (nonmusculoskeletal) reasons. Normality of continuous data was tested by examining skewness (normal < 2) and kurtosis (−2 < normal < 2) and by use of a Shapiro-Wilk test. Between-groups comparisons were performed with a χ² test for categorical variables and with a t test or ANOVA (or nonparametric equivalents as specified in the results) for continuous variables. A generalized linear model with log-link was used to evaluate whether demographic factors (age, sex, and pretreatment number of races), lesion descriptors (location [core vs margin], length, echogenicity, and CSA), and treatment factors (desmotomy or no desmotomy) were associated with categorical outcome variables (return to racing [yes or no], number of races after treatment [1, ≤ 4, or > 4], and reason for retirement [reinjury, no reinjury, or other musculoskeletal injury]). Variables for which the P value in univariate analyses was < 0.2 were subsequently incorporated into the multivariate model. All analyses were performed by use of statistics software.^c Values of P < 0.05 were considered significant.

Results

Forty horses met the criteria for inclusion in the study (Table 1). There were 14 females and 26 males, including 14 sexually intact males, 6 castrated males, and 6 males for which neutering status was not recorded. Mean ± SD age at the time of injury was 3.6 ± 1.2 years (range, 2 to 7 years). Time between injury and treatment was recorded for only 9 of the 40 horses and ranged from 8 to 90 days (median, 18 days; mean, 31 days; SD, 30 days). The left forelimb was affected in 19 horses, the right forelimb was affected in 16, and both forelimbs were affected in 5.

Mean ± SD lesion length at the time of initial examination was 3.9 ± 1.5 zones (range, 1 to 7 zones). On the basis of a previous publication,²⁶ length of each ultrasonographic zone was estimated to be 4 cm. Therefore, mean ± SD estimated lesion length was calculated to be 15.6 ± 6 cm. Tendon CSA at the zone of maximum injury ranged from 1.14 to 2.85 cm² (mean ± SD, 1.60 ± 0.65 cm²). Lesion CSA at the same site ranged from 0.02 to 1.94 cm² (mean ± SD, 0.51 ± 0.45 cm²). Percentage of tendon CSA that was affected ranged from 12% to 79% (mean ± SD, 26 ± 18%). Mean tendon CSA (P = 0.04), lesion CSA (P = 0.04), and percentage of tendon CSA that was affected (P = 0.01) were significantly higher in males than in females. Mean ± SD tendon CSA was 1.72 ± 0.55 cm² and 1.38 ± 0.11 cm², mean lesion CSA was 0.62 ± 0.51 cm² and 0.26 ± 0.17 cm², and mean percentage of tendon CSA that was affected was 36% ± 20% and 19% ± 13% in males and females, respectively.

Results of follow-up ultrasonography were available for 26 of the 40 horses (Figure 1). Time from initial to follow-up ultrasonography ranged from 6 to 18 days (mean ± SD, 9.6 ± 2.8 days). Twenty-three of the 26 horses had a significantly lower lesion echogenicity score at the time of follow-up examination (mean, 2.7 ± 0.8, median 3), compared with score at the time of the initial evaluation (mean, 2.9 ± 0.7; median, 3; P = 0.02, related-samples Wilcoxon signed rank test). No significant differences were found between initial and follow-up ultrasonographic findings for tendon CSA (P = 0.57), lesion CSA (P = 0.99), or percentage of CSA injured (P = 0.79).

Transverse (A, B, C, and D) and longitudinal (E and F) ultrasonographic appearance of the SDF tendon in a 3-year-old Thoroughbred racehorse with SDF tendonitis extending from zone 1B to zone 3C of the left forelimb. Images were obtained before (A, C, and E) and 7 days after (B, D, and F) 4 intralesional injections of IGF-I.

Citation: Journal of the American Veterinary Medical Association 239, 7; 10.2460/javma.239.7.992

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Transverse (A, B, C, and D) and longitudinal (E and F) ultrasonographic appearance of the SDF tendon in a 3-year-old Thoroughbred racehorse with SDF tendonitis extending from zone 1B to zone 3C of the left forelimb. Images were obtained before (A, C, and E) and 7 days after (B, D, and F) 4 intralesional injections of IGF-I.

Citation: Journal of the American Veterinary Medical Association 239, 7; 10.2460/javma.239.7.992

• Download Figure
• Download figure as PowerPoint slide

Transverse (A, B, C, and D) and longitudinal (E and F) ultrasonographic appearance of the SDF tendon in a 3-year-old Thoroughbred racehorse with SDF tendonitis extending from zone 1B to zone 3C of the left forelimb. Images were obtained before (A, C, and E) and 7 days after (B, D, and F) 4 intralesional injections of IGF-I.

Citation: Journal of the American Veterinary Medical Association 239, 7; 10.2460/javma.239.7.992

• Download Figure
• Download figure as PowerPoint slide

Twelve of the 40 (30%) horses had never raced before treatment, 10 (25%) had raced between 1 and 4 times, and 18 (45%) had raced ≥ 5 times. Posttreatment race data were available for 34 of the 40 horses, and 21 of the 34 (62%; 95% CI, 45% to 76%) horses raced at least once after treatment, including 10 of the 34 (29%; 95% CI, 17% to 46%) horses that raced between 1 and 4 times and 11 (32%; 95% CI, 19% to 49%) that raced ≥ 5 times. For the 21 horses that raced after treatment, mean ± SD time to first race after treatment was 388 ± 160 days (range, 112 to 862 days).

For the 34 horses for which race data were available, we did not detect significant associations between racing status (ie, did not race vs raced 1 to 4 times vs raced ≥ 5 times) before treatment and racing status after treatment (P = 0.14), between sex (male vs female) and racing status after treatment (P = 0.71), or between affected limb (left forelimb vs right forelimb vs both forelimbs) and racing status after treatment (P = 0.21). Mean age (P = 0.9) and mean lesion echogenicity score (P = 0.78) at the time of initial examination were not significantly different between horses that raced after treatment and those that did not.

Twenty-six horses underwent tenoscopic transection of the accessory ligament of the SDF tendon at the time of the initial IGF-I injection, and posttreatment race data were available for 23 of the 26 horses that underwent desmotomy (Table 1). Fourteen of the 23 (61%; 95% CI, 41% to 78%) horses raced at least once after treatment, including 8 of the 23 (35%; 95% CI, 19% to 55%) horses that raced between 1 and 4 times and 6 (26%; 95% CI, 12% to 47%) that raced ≥ 5 times. Post-treatment race data were available for 11 of the 14 horses that did not undergo desmotomy, and 7 of 11 (64%; 95% CI, 35% to 85%) raced at least once after treatment, including 2 of the 11 (18%; 95% CI, 4% to 49%) horses that raced between 1 and 4 times and 5 (45%; 95% CI, 21% to 72%) that raced ≥ 5 times. We did not detect a significant (P = 0.82) association between desmotomy (yes vs no) and return to racing (yes vs no). For the 27 horses for which race earnings data were available, the mean value of individual horse earnings (median for 3 races) was $6,446 (range, $0 to $23,220) before treatment and $3,379 (range, $0 to $24,997) after treatment.

Thirteen of 28 (46%) horses for which long-term follow-up data were available had a recurrence of tendonitis in the same limb or developed tendonitis in the opposite limb. Ten of these 13 horses had undergone desmotomy, and 3 had not. Eight of the 28 (29%) horses had a nontendinous musculoskeletal injury that ultimately ended their racing careers. Of these, 5 had undergone desmotomy, and 3 had not. One horse had a suspensory ligament injury; this horse had not undergone desmotomy. We did not detect a significant (P = 0.19) association between desmotomy (yes vs no) and recurrence of SDF tendonitis (yes vs no).

Sex (P = 0.10), affected limb (P = 0.14), pretreatment number of races (P = 0.22), and whether desmotomy was performed (P = 0.20) were identified as potential predictors of return to racing (yes vs no) during screening in univariate analyses. However, none were significant in the final multivariate model (sex, P = 0.10; affected limb, P = 0.13; desmotomy, P = 0.06; pretreatment number of races, P = 0.17).

Discussion

Results of the present study suggested that in Thoroughbred racehorses with SDF tendonitis, intralesional injection of IGF-I led to a decrease in ultrasonographic lesion severity. However, treated horses had only a moderate prognosis for return to racing, with 21 of 34 (62%; 95% CI, 45% to 76%) racing at least once after treatment and 11 (32%; 95% CI, 19% to 49%) racing ≥ 5 times after treatment. In addition, concurrent tenoscopic desmotomy of the accessory ligament of the SDF tendon did not appear to improve the prognosis, as no significant association was identified between desmotomy (yes vs no) and return to racing (yes vs no) and desmotomy was not identified as a significant predictor of a return to racing in the logistic regression analysis.

In vitro and in vivo studies^7,12 of the effects of IGF-I on equine tendon repair indicate that IGF-I increases cellularity and collagen deposition and enhances collagen type I expression. In the present study, 23 of 26 horses in which follow-up ultrasonography was performed had a lower lesion echogenicity score at the time of follow-up examination, compared with the score at the time of initial evaluation. A similar increase in echogenicity was identified in a study⁶ in which horses with SDF tendonitis were treated with intralesional injections of β-aminopropionitrile fumarate. Experimental data suggest that the improved ultrasonographic appearance in the present study may have been the result of increased collagen deposition and organization, although vascular and cellular proliferation may also have played a role.^7,12

The percentage of horses in the present study that returned to racing was only slightly higher than percentages reported in previous studies^6,9,19,27 in which horses received conservative treatment. Many factors can limit athletic longevity following treatment of SDF tendonitis, including injuries in other musculoskeletal sites, abnormalities beyond the musculoskeletal system, and the tendency to retire Thoroughbreds from competition for use in breeding programs. Despite these factors, 21 of 34 (62%) horses returned to racing following treatment with IGF-I. Importantly, a control group of horses that was not treated with IGF-I was not available for comparison with our data. However, a previous study¹⁹ involving Thoroughbred racehorses used for flat racing in New Zealand found that 34% raced ≥ 5 times after conservative management,¹ and a study⁹ from the United Kingdom reported reinjury rates as high as 67% among Thoroughbreds used for flat racing. In a study⁶ of SDF tendonitis in which control horses treated with saline solution were compared with horses that received intralesional injections of β-aminopropionitrile fumarate, 33% of the control horses raced ≥ 5 times after treatment. However, a substantial proportion of the control horses were Standardbreds, which generally return to racing in greater numbers than flat-racing Thoroughbreds.

In vivo and in vitro data support the use of intralesional injection of IGF-I for treatment of tendon lesions.^7,12,28 Repeated injection of core lesions was considered important in a previous equine study⁷ to overcome the short residence time in situ, and a similar injection regimen was used in the present study. However, the dynamics of IGF-I following injection into a largely avascular or even necrotic core lesion are unknown. Morphological and vascular perfusion studies indicate that the central core of the lesion during the acute stages of tendonitis is devitalized,²⁹ which provides a rationale for procedures such as tendon splitting, which are intended to revascularize these core lesions.²⁷ The impact of IGF-I on neovascularization is unknown, but our ultrasonographic findings suggest that either needle insertion or IGF-I itself may have improved the early phase of recovery. In addition, the short half-life of IGF-I may be partially overcome by ongoing IGF-I autoinduction³⁰ and binding to IGF-binding proteins.³¹ Equine tendon responds to injury by increasing production of IGF-binding proteins 2, 3, and 4.³¹ Although binding of IGF-I by these proteins sequesters it from IGF receptors, protection from IGF proteases and later ligand release may extend the action of IGF-I in vivo.^31,32 Overall, given the outcome data in the present study, mechanisms to extend IGF-I synthesis or residence time seem worthy of further investigation. Implantation of equine stem cells overexpressing IGF-I shows advantages over implantation of stem cells alone for short-term repair of SDF tendonitis.³³

In the present study, the addition of accessory ligament desmotomy to the treatment protocol was expected to bolster the mechanical capabilities of the SDF tendon and compensate for inelasticity at the site of the tendonitis lesion. In a study²² of Australian racehorses that underwent accessory ligament desmotomy without concurrent surgical or medical procedures, 52% returned to racing, and a study²¹ of Thoroughbreds in the United States treated with a combination of percutaneous tendon splitting and accessory ligament desmotomy found that 51% returned to sustained racing. Thus, in the present study it appears that the percentage of Thoroughbred horses that returned to racing after intralesional IGF-I injection, alone or in combination with accessory ligament desmotomy, was higher than percentages reported following conservative treatment and other pharmaceutical treatments,^6,9,27 but was not substantially different from the percentage reported following accessory ligament desmotomy alone. Sixty-four percent of horses treated with IGF-I alone in the present study returned to racing, but this represented only 7 of 11 horses.

Although the long-term biomechanical implications of accessory ligament desmotomy remain unknown, 26 of the 40 horses in the present study underwent this procedure. The decision to perform or not perform desmotomy was largely based on economics, in that desmotomy was recommended for all 40 horses. The low number of horses limited our ability to draw conclusions regarding the role of accessory ligament desmotomy in the treatment of racehorses with SDF tendonitis. Recent in vitro data suggest that accessory ligament desmotomy actually increases strain on the SDF tendon, rather than elongating the tendon-ligament unit.²³ It has also been suggested that increased strain on the suspensory ligament explains the increased incidence of suspensory ligament desmitis following accessory ligament desmotomy.¹⁹ We did not identify a similar incidence of suspensory ligament desmitis in our population, and only 1 horse in the present study was known to develop an injury of the suspensory ligament following IGF-I treatment. This horse had not undergone concurrent accessory ligament desmotomy.

Accessory ligament desmotomy did not appear to change racing outcome, with our data failing to identify a significant association between desmotomy (yes vs no) and return to racing (yes vs no). In addition, although the stated goal of accessory ligament desmotomy is to reduce the recurrence rate in horses with SDF tendonitis, we did not detect a significant association between desmotomy (yes vs no) and recurrence (yes vs no). Additionally, bilateral desmotomy was instituted early in this study, as 2 horses developed tendonitis in the opposite forelimb after returning to racing. Similar recommendations have been made by others.^19,21,22

In the present study, mean lesion echogenicity score at the time of initial examination was not significantly different between horses that raced after treatment and those that did not. However, most horses had moderate to severe lesions, with a mean estimated lesion length of 15.6 cm and mean percentage affected tendon CSA of 26%, which may explain the lack of association. There is a paucity of data on lesion CSA and lesion length in previous studies,^9,22 and in 1 study,⁶ severity scores were summed over multiple levels of assessment, so single-level CSA information was lost and lesion length was confounded.

On a broader scale, the relationship between lesion severity and likelihood of successful return to athletic competition has not been fully elucidated in the literature. It is widely accepted that few racehorses treated conservatively successfully return to racing without tendon reinjury (with success defined as > 4 races completed).^27,34 In addition, moderate to severe lesions are thought to be associated with an even worse prognosis.³⁵ The overrepresentation in our population of moderate to severe lesions therefore means that the results of this study may not be a fair reflection of the potential of IGF-I treatment. Male horses were also overrepresented in our study, which has been a common finding in other studies,^19,21 and this may have important implications for outcome. Male horses may be more likely to be treated after severe tendon injury because retirement from racing for breeding purposes is less common. Conversely, female horses may not be treated because retirement for breeding is considered earlier. In addition, male horses in our study had more severe lesions at the time of initial examination. A larger case series and randomized controlled clinical trial would be required to draw stronger conclusions on the impacts of lesion severity and sex on outcome of IGF-I treatment.

Appropriate measures of success following treatment of a condition such as tendonitis in horses remain controversial. A variety of measures of success have been used previously, with some more lenient than others. In the present study, we evaluated 2 measures of success: whether the horse raced after treatment and how many races a horse completed after treatment, with ≥ 5 races being considered highly successful. However, it is unknown what the normal career path is for Thoroughbred racehorses, particularly those that are ≥ 4 years old. Thus, our criteria may have been too stringent for most horses in the study, particularly given their age.

Return-to-race data after tendonitis have varied with previous treatments. Historically, counterirritation (pin firing or external blistering) combined with 6 months' rest allowed 23 of 54 (43%) Thoroughbreds to return to racing.²⁴ Ten of these 23 (43%) raced without reinjury to the tendon.³⁵ Intralesional treatment with β-aminopropionitrile fumarate and controlled exercise allowed a small proportion of Thoroughbred racehorses to return to flat racing, with reinjury or new tendon and ligament injuries developing in > 62% of horses.⁹ Outcome was better in other horse disciplines. Other pharmaceuticals have been injected into tendonitis lesions in racehorses. Intralesional administration of hyaluronan did nothing to reduce reinjury rate in Thoroughbred flat racehorses and National Hunt racehorses in the United Kingdom.⁹ Together, these clinical studies may represent benchmarks against which the current IGF-I-treated group can be measured, and in that context, a return to racing of 62% seems beneficial.