Evaluation of experimentally induced injury to the superficial digital flexor tendon in horses by use of low-field magnetic resonance imaging and ultrasonography

William M. Karlin Department of Veterinary Clinical Medicine, College of Veterinary Medicine, University of Illinois, Urbana, IL 61802.

Search for other papers by William M. Karlin in
Current site
Google Scholar
PubMed
Close
 DVM, MS
,
Allison A. Stewart Department of Veterinary Clinical Medicine, College of Veterinary Medicine, University of Illinois, Urbana, IL 61802.

Search for other papers by Allison A. Stewart in
Current site
Google Scholar
PubMed
Close
 DVM, MS
,
Sushmitha S. Durgam Department of Veterinary Clinical Medicine, College of Veterinary Medicine, University of Illinois, Urbana, IL 61802.

Search for other papers by Sushmitha S. Durgam in
Current site
Google Scholar
PubMed
Close
 BVSc
,
James F. Naughton Department of Veterinary Clinical Medicine, College of Veterinary Medicine, University of Illinois, Urbana, IL 61802.

Search for other papers by James F. Naughton in
Current site
Google Scholar
PubMed
Close
 DVM, MS
,
Kristen J. O'Dell-Anderson Department of Veterinary Clinical Medicine, College of Veterinary Medicine, University of Illinois, Urbana, IL 61802.

Search for other papers by Kristen J. O'Dell-Anderson in
Current site
Google Scholar
PubMed
Close
 DVM, MS
, and
Matthew C. Stewart Department of Veterinary Clinical Medicine, College of Veterinary Medicine, University of Illinois, Urbana, IL 61802.

Search for other papers by Matthew C. Stewart in
Current site
Google Scholar
PubMed
Close
 BVSc, PhD

Abstract

Objective—To evaluate tendon injuries in horses over a 16-week period by use of ultrasonography and low-field magnetic resonance imaging (MRI).

Sample—Tendons of 8 young adult horses.

Procedures—The percentage of experimentally induced tendon injury was evaluated in cross section at the maximal area of injury by use of ultrasonography and MRI at 3, 4, 6, 8, and 16 weeks after collagenase injection. The MRI signal intensities and histologic characteristics of each tendon were determined at the same time points.

Results—At 4 weeks after collagenase injection, the area of maximal injury assessed on cross section was similar between ultrasonography and MRI. In lesions of > 4 weeks' duration, ultrasonography underestimated the area of maximal cross-sectional injury by approximately 18%, compared with results for MRI. Signal intensity of lesions on T1-weighted images was the most hyperintense of all the sequences, lesions on short tau inversion recovery images were slightly less hyperintense, and T2-weighted images were the most hypointense. Signal intensity of tendon lesions was significantly higher than the signal intensity for the unaltered deep digital flexor tendon. Histologically, there was a decrease in proteoglycan content, an increase in collagen content, and minimal change in fiber alignment during the 16 weeks of the study.

Conclusions and Clinical Relevance—Ultrasonography may underestimate the extent of tendon damage in tendons with long-term injury. Low-field MRI provided a more sensitive technique for evaluation of tendon injury and should be considered in horses with tendinitis of > 4 weeks' duration.

Abstract

Objective—To evaluate tendon injuries in horses over a 16-week period by use of ultrasonography and low-field magnetic resonance imaging (MRI).

Sample—Tendons of 8 young adult horses.

Procedures—The percentage of experimentally induced tendon injury was evaluated in cross section at the maximal area of injury by use of ultrasonography and MRI at 3, 4, 6, 8, and 16 weeks after collagenase injection. The MRI signal intensities and histologic characteristics of each tendon were determined at the same time points.

Results—At 4 weeks after collagenase injection, the area of maximal injury assessed on cross section was similar between ultrasonography and MRI. In lesions of > 4 weeks' duration, ultrasonography underestimated the area of maximal cross-sectional injury by approximately 18%, compared with results for MRI. Signal intensity of lesions on T1-weighted images was the most hyperintense of all the sequences, lesions on short tau inversion recovery images were slightly less hyperintense, and T2-weighted images were the most hypointense. Signal intensity of tendon lesions was significantly higher than the signal intensity for the unaltered deep digital flexor tendon. Histologically, there was a decrease in proteoglycan content, an increase in collagen content, and minimal change in fiber alignment during the 16 weeks of the study.

Conclusions and Clinical Relevance—Ultrasonography may underestimate the extent of tendon damage in tendons with long-term injury. Low-field MRI provided a more sensitive technique for evaluation of tendon injury and should be considered in horses with tendinitis of > 4 weeks' duration.

Ultrasonography has been used to diagnose tendon injuries in horses for > 3 decades.1–4 During this time, high-resolution real-time ultrasonography has been the primary modality used to diagnose tendon injuries in horses and to determine the readiness for return to exercise and competition.2–4 Accepting that ultrasonography is a reliable and noninvasive method for evaluating equine tendons,5 image acquisition is extremely dependent on the operator and considerable variability in the determination of lesion boundaries is possible among operators.5 For this reason, MRI is the imaging modality of choice in human orthopedics because of image reproducibility and minimal variability among operators.6

Magnetic resonance imaging is the imaging modality of choice for musculoskeletal injuries in human sports medicine,7,8 but expense, availability, and size constraints of equipment have limited use of MRI in horses until recently. In horses, MRI is now considered the criterion-referenced standard for detection of injuries in tendons6,9 and ligaments.10–12 In horses, MRI provides a promising alternative imaging modality to ultrasonography for the evaluation of pathological changes in tendons. The fibrous nature and low water content of equine tendons result in a low signal intensity in the primary diagnostic MRI sequences.9 Acute tendon injuries, which are characterized by early cellular infiltration prior to fluid accumulation or hemorrhage, have increased signal intensities on T1- and T2-weighted images.9 Healing injured tendons have increased signal intensity on T1-weighted images but decreased to normal signal intensity on T2-weighted images.9

Several studies6,9,11 have been conducted to compare results for ultrasonography and MRI of tendon with histologic characteristics of the tissue. Results from these studies are contradictory. In 1 study,6 the investigators concluded that there was no clear benefit of MRI over real-time ultrasonography for the detection of tendon lesions in horses. In contrast, a more recent study12 revealed that MRI was the better imaging modality for the diagnosis of desmitis in horses in which chronic scar tissue was present; chronic scar tissue is a pathological condition that is difficult to detect via ultrasonography.12 In human medicine, MRI has been found to be more sensitive than ultrasonography.13 The objective of the study reported here was to compare ultrasonography and low-field MRI for evaluation of tendon injuries in horses over a 16-week period. We hypothesized that ultrasonography would reveal smaller areas of maximal cross-sectional injury, compared with results for MRI.

Materials and Methods

Animals—Eight clinically normal Quarter Horses (7 horses were 3 years old and 1 horse was 4 years old) with a mean ± SEM body weight of 447 ± 10.3 kg were used in the study. All horses were evaluated to ensure they were clinically normal without any signs of tendon injury or lameness. Lameness examinations and ultrasonographic evaluation of all flexor tendons were performed prior to inclusion in the study. All procedures were approved by the University of Illinois Institutional Animal Care and Use Committee.

Induction of tendinitis—On day 0, horses received phenylbutazonea (2.2 mg/kg, IV), penicillin G procaineb (22,000 U/kg, IM), and tetanus toxoidc before induction of tendinitis. The metacarpal regions were clipped of hair with a No. 40 clipper blade and aseptically prepared, and horses were sedated via administration of detomidine hydrochlorided (0.01 to 0.015 mg/kg, IV, as needed). Each SDF tendon in both forelimbs of all horses was injected by use of a 25-gauge needle with 2,000 U of filter-sterilized bacterial collagenasee at the midmetacarpal region via ultrasonographicf guidance, as described elsewhere.14–16 After injection of collagenase, all horses received an additional dose of penicillin G procaine and a dose of phenylbutazone. Horses were maintained in strict stall confinement for 8 weeks after collagenase injection, after which time they were hand walked (15 min/walk) twice daily until the conclusion of the study.

In addition, the SDF tendon in 1 randomly selected hind limb of each of 4 horses was also injected with collagenase at a point 12 cm distal to the tuber calcaneous at weeks 8, 10, 12, and 13, respectively (1 horse/time point). Similar to the injection procedure for the forelimbs, the metatarsal region was clipped of hair with a No. 40 clipper blade and aseptically prepared. Horses were sedated with detomidine, and the SDF tendon was injected with collagenase. After injection of collagenase, horses received a dose of penicillin G procaine and a dose of phenylbutazone. Following hind limb tendon injection, all horses were maintained in stall confinement for another 2 weeks, before resuming hand walking (15 min/walk) twice daily.

Ultrasonography—At week 16, ultrasonography of all experimentally injured tendons was performed by use of a 7.5- to 13-MHz linear transducerf set at 13 MHz. Transverse and longitudinal ultrasonographic imaging planes were acquired by use of a standoff to determine the site of maximal injury within each SDF tendon. The cross-sectional areas of the induced tendinitis lesions were obtained at the site of maximal injury and recorded as the percentage of the tendon area via measurement tools of the ultrasonography unit. Percentage was calculated as (cross-sectional area of injured area/cross-sectional area of tendon) × 100. All images were measured by use of a DICOM workstation.g

MRI—On the day after the ultrasonographic examination was performed at week 16, MRI was performed. The SDF tendons of both forelimbs (which had been injected with collagenase 16 weeks before MRI) were evaluated in all 8 horses. In addition, the experimentally injured SDF tendon in each of the 4 horses injected at weeks 8, 10, 12, and 13 (ie, injected with collagenase 8, 6, 4, and 3 weeks before MRI, respectively) were also evaluated.

All horses were medicated with xylazine hydrochlorideh (0.02 mg/kg, IV). Anesthesia was induced with IV administration of diazepami (0.08 mg/kg) and ketamine hydrochloridej (2.2 mg/kg) and was maintained with isoflurane.k Low-field MRI images were acquired with a 0.25-T MRI systeml and a custom-designed equine extremity coil. The MRI procedures used in the pulse sequences for the study were as follows: transverse spin-echo T1 weighted, TR = 1,240 milliseconds and TE = 26 milliseconds; transverse spin-echo T2 weighted, TR = 3,300 milliseconds and TE = 80 milliseconds; and transverse STIR, TR = 2,330 milliseconds and TE = 24 milliseconds (inversion time, 85 milliseconds). For all 3 sequences, flip angle was 90°, field of view was 180 × 180 mm, matrix size was 256 × 256, number of excitations was 2, slice width was 4 mm, gap width was 0.4 mm, and phase encoding was in the y-plane.

MRI analysis—Quantitative measurements, which included cross-sectional area (Figures 1 and 2) and signal intensity values (Figure 3), were obtained at the level of maximal cross-sectional tendon injury in all pulse sequences by use of the ROI measurement tool of the DICOM workstation. In each sequence, an ASI was calculated for the injured SDF tendon and the unaltered control DDF tendon at the level of maximal tendon injury by dividing the signal intensity value of the ROI (20 mm2) by the signal intensity value for a 20-mm2 box obtained from a hypointense area (background noise) outside the margins of the limb image. Measurements were repeated 5 times.

Figure 1—
Figure 1—

Ultrasonographic images obtained from a representative horse and used to determine the percentage of cross-sectional tendon injury calculated at 16 weeks after collagenase injection by use of an ultrasound machine (A) and a DICOM workstation (B). In panel A, the percentage was calculated by dividing the cross-sectional area of the injured tendon (small outlined area No. 2 [0.12 cm2]) by the total cross-sectional area of the SDF tendon (large outlined area No. 1 [1.69 cm2]) for the area of maximal injury. In panel B, the percentage was calculated by dividing the cross-sectional area of the injured tendon (yellow outline [27.09 mm2]) by the total cross-sectional area of the SDF tendon (red outline [148.85 mm2]) for the area of maximal injury. AR = Cross-sectional area.

Citation: American Journal of Veterinary Research 72, 6; 10.2460/ajvr.72.6.791

Figure 2—
Figure 2—

A T1-weighted image obtained from a representative horse and used to determine the percentage of cross-sectional area of injured tendon calculated by dividing the cross-sectional area of the injured tendon (yellow outline [179.37 mm2]) by the total cross-sectional area of the SDF tendon (red outline [246.14 mm2]).

Citation: American Journal of Veterinary Research 72, 6; 10.2460/ajvr.72.6.791

Figure 3—
Figure 3—

A T1-weighted image obtained from a representative horse and used to determine signal intensity whereby the ASI was calculated from the mean of 5 signal intensities derived from the ROI (depicted by the 20-mm2 yellow box) at the maximal zone of injury divided by the mean of 5 signal intensities derived from the background (surrounding air, depicted by the 20-mm2 red box). Mean ± SEM ASI was 798.08 ± 102.77 and 116.86 ± 25.05 for the injured tendon and background, respectively.

Citation: American Journal of Veterinary Research 72, 6; 10.2460/ajvr.72.6.791

Histologic examination—After the MRI was completed at week 16, all 8 horses were euthanized by IV administration of a lethal dose of sodium pentobarbital. Tendon specimens were collected from all 4 limbs of each of 4 horses for histologic evaluation of the SDF tendons. These were the horses that had been given injections in the SDF tendon in a hind limb at 8, 6, 4, and 3 weeks before MRI, respectively. The tendon specimens were obtained from each of the SDF tendons of the forelimbs (16 weeks after injection) and the injected hind limb at the site of maximal injury (ie, location of ultrasonography and MRI). Specimens from the unaltered control SDF tendons were obtained from the contralateral hind limb at the same level as for the collagenase-induced lesions. Tendon samples were fixed in 4% paraformaldhyde and embedded in paraffin by use of standard techniques.17

Collagen content, proteoglycan content, and collagen fiber alignment—Sections were cut to a thickness of 6 μm with a mictrotome and stained with picrosirius red and toluidine blue stains,m which are specific for collagen and proteoglycan, respectively. For each specimen, 10 sections that were separated by at least 150 μm were prepared. In each section, proximal, middle, and distal sites were evaluated, which resulted in 30 images assessed for collagen content, proteoglycan content, and collagen alignment in each tendon. Collagen and proteoglycan contents were evaluated at 10× magnification by use of commercially available software.n Collagen fiber alignment was determined at 2.5× magnification by use of images stained with picrosirius red stain via a defined analysis algorithm to assess areas for collagen, as described elsewhere.17,18 Similarly, images stained with toluidine blue were analyzed via a defined algorithm to assess areas for proteoglycan, as described elsewhere.19 Three-point angles were drawn on images stained with picrosirius red to assess changes in collagen fiber alignment and obtain numeric values in degrees, as described elsewhere.19

Statistical analysis—A commercial software packageo was used to perform a 2-way repeated-measures ANOVA for evaluation of differences between the total cross-sectional areas determined via ultrasonographic and MRI (T1, T2, and STIR) analyses. In addition, MRI signal intensities of the injured SDF tendons were compared with signal intensities of the unaltered DDF tendons over time within each sequence (T1, T2, and STIR). Holm-Sidak post hoc tests were used when indicated. Values of P ≤ 0.05 were considered significant. Descriptive data were used to report the signal intensities of each sequence (T1, T2, and STIR) as well as the histologic changes in collagen content, proteoglycan content, and collagen alignment. All data were reported as mean ± SEM.

Results

Percentage of cross-sectional tendon damage—The percentage of damaged tendon was not determined on the 4 SDF tendons in the hind limb at 6 and 8 weeks after injection because of data loss. At 4 weeks after collagenase injection, there was no significant (P = 0.809) difference in the cross-sectional percentage of damaged tendon at the site of maximal injury between ultrasonography and MRI (Table 1). At 16 weeks after collagenase injection, the percentage of damaged tendon at the site of maximal injury was significantly (P < 0.001) less in ultrasonographic images (approx 18% less) than in MRI images.

Table 1—

Mean ± SEM percentage of tendon cross-sectional damage in the SDF tendon of horses as assessed via measurements obtained by use of ultrasonography and MRI sequences at ≤ 4 and 16 weeks after injection of collagen.

Time (wk)UltrasonographyT1T2STIRP value
≤451.2 ± 8.954.2 ± 9.652.4 ± 7.360.8 ± 0.70.809
1641.8 ± 6.059.5 ± 13.2*59.5 ± 13.5*55.6 ± 14.1*< 0.001

Percentage was calculated as (cross-sectional area of injured area/cross-sectional area of tendon) × 100.

Within a row, value differs significantly (P < 0.05) from the value for ultrasonography.

Collagen content—Minimal changes in collagen content were evident over time (Figure 4). Specifically, there was a 10% increase in collagen content between 3 and 8 weeks after collagen injection. At 16 weeks after collagen injection, collagen content in the injured SDF tendon was only 60% of the collagen content in the unaltered control tendon.

Figure 4—
Figure 4—

Photomicrograph of a representative section of unaltered control tendon tissue (A) and the mean ± SEM percentage area for collagen staining (green) calculated at various time points after induced tendon injury (B). Samples were collected from 1, 1, 1, 1, and 4 tendons at weeks 3, 4, 6, 8, and 16 after injury, respectively, and from the unaltered normal (control) tendon at these same times. In panel A, picrosirius red stain; bar = 200 μm.

Citation: American Journal of Veterinary Research 72, 6; 10.2460/ajvr.72.6.791

Proteoglycan content—A gradual decrease in proteoglycan content was detected over time (Figure 5). A 45% decrease in proteoglycan content was seen between 3 and 8 weeks after collagen injection. At 16 weeks after collagen injection, there was a 2.3-fold increase in proteoglycan content in injured SDF tendon, compared with the proteoglycan content in the unaltered control tendon.

Figure 5—
Figure 5—

Photomicrograph of a representative tissue section from a tendon lesion collected 16 weeks after collagenase injection and stained to reveal proteoglycan (which appears as green staining; A) and the percentage area for proteoglycan staining calculated at various time points after induced tendon injury (B). Samples were collected from 1, 1, 1, 1, and 4 tendons at weeks 3, 4, 6, 8, and 16 after injury, respectively, and from the unaltered control tendon at these same times. In panel A, toluidine blue stain; bar = 200 μm.

Citation: American Journal of Veterinary Research 72, 6; 10.2460/ajvr.72.6.791

Collagen fiber alignment—Minimal changes in collagen fiber alignment were seen in the injured tendon during the 16 weeks after collagen injection (Figure 6). However, fiber alignment in the injured tendon differed markedly at all time points after injection from the alignment seen in the unaltered control tendon. Angles for the collagen fiber alignment varied from 19° to 34°, compared with small variations (3.5°) in angles derived for the unaltered control tendon.

Figure 6—
Figure 6—

Photomicrographs of representative tissue sections collected from an unaltered control tendon (A) and a tendon 3 weeks after injury (B) and the angle of collagen fiber alignment calculated at various time points after induced tendon injury (C). Notice the various angles of collagen fiber alignment in the histologic sections. Samples were collected from 1,1,1,1, and 4 tendons at weeks 3, 4, 6, 8, and 16 after injury, respectively, and from the unaltered control tendon at these same times. In panels A and B, picrosirius red stain; bar = 3 mm.

Citation: American Journal of Veterinary Research 72, 6; 10.2460/ajvr.72.6.791

ASI for lesions detected via MRI—At all time points, signal intensity of lesions was greatest for T1-weighted images, intermediate for STIR images, and lowest for T2-weighted images (Figure 7). On the T1-weighted images, signal intensity of the injured SDF tendon was significantly (P = 0.001) higher than the signal intensity of the unaltered DDF tendon at all time points after injury. There was no significant (P = 0.264) decrease in the T1-lesion signal intensity of injured tendons over time. However, the T1-weighted images had the largest differences in signal intensity (17.0 to 12.9) between the injured and unaltered control tendon. On the STIR images, the signal intensity of the injured SDF tendon was significantly (P = 0.037) higher than the signal intensity of the unaltered control DDF tendon at all time points following injury. There was no significant (P = 0.663) decrease in the STIR signal intensity of injured tendons over time. On the T2-weighted images, signal intensity of the injured SDF tendon was significantly (P = 0.001) higher than the signal intensity of the unaltered control DDF tendon at 16 weeks after injection. There was increased signal intensity for the injured SDF tendon, compared with signal intensity for the unaltered control DDF tendon, at 3 to 8 weeks after injection; however, these values did not differ significantly (P = 0.087). In addition, there was no significant (P = 0.498) decrease in the signal intensity in lesions for T2-weighted images of injured tendons over time. However, the T2-weighted images had the smallest difference in signal intensity (2.0 to 3.7) between the injured and unaltered control tendons. Ultrasonographic and low-field MRI images of tendon injury were evident at 4 (Figure 8) and 16 (Figure 9) weeks after collagen injection.

Figure 7—
Figure 7—

Mean ± SEM signal intensity of lesions for T1-weighted (black circles), T2-weighted (black inverted triangles), and STIR (black squares) images (A) and T1-weighted (B), T2-weighted (C), and STIR (D) images of injured SDF tendon (black symbols) and unaltered control DDF tendon (white symbols) obtained from horses at various time points after induced tendon injury. Notice that the scale on the y-axis differs among panels. *Within a time point, signal intensity differs significantly (P < 0.05) between the injured SDF tendon and unaltered control DDF tendon.

Citation: American Journal of Veterinary Research 72, 6; 10.2460/ajvr.72.6.791

Figure 8—
Figure 8—

Ultrasonographic image (top) and T1-weighted (bottom left), T2-weighted (bottom middle), and STIR (bottom right) images of cross-sectional damage of a representative SDF tendon at 4 weeks after injection of collagenase. The percentage of cross-sectional injury for this tendon was 64.1%, 63.8%, 58.4%, and 61.3% for ultrasonographic, T1-weighted, T2-weighted, and STIR images, respectively. Signal intensity for the injured tendon was 24.9, 9.3, and 16.6 for T1-weighted, T2-weighted, and STIR images, respectively.

Citation: American Journal of Veterinary Research 72, 6; 10.2460/ajvr.72.6.791

Figure 9—
Figure 9—

Ultrasonographic image (top) and T1-weighted (bottom left), T2-weighted (bottom middle), and STIR (bottom right) images of cross-sectional damage of a representative SDF tendon at 16 weeks after injection of collagenase. The percentage of cross-sectional injury for this tendon was 56.0%, 67.1 %, 67.1 %, and 67.0% for ultrasonographic, T1-weighted, T2-weighted, and STIR images, respectively. Signal intensity for the injured tendon was 23.6, 5.8, and 19.9 for T1-weighted, T2-weighted, and STIR images, respectively.

Citation: American Journal of Veterinary Research 72, 6; 10.2460/ajvr.72.6.791

Discussion

Analysis of results of the study reported here revealed that the area of maximal injury assessed on cross section in acute tendinopathies was similar between ultrasonographic and low-field MRI images for the 4 weeks after collagenase injection. However, in lesions with a longer duration, ultrasonography revealed significantly smaller areas of maximal cross-sectional injury (approx 18% smaller), compared with the cross-sectional area of the injury for MRI images of the same site. These findings suggested that ultrasonography is similar to MRI for identifying the cross-sectional area of the damaged tendon during the acute phase of injury but detects smaller areas at later times. Thus, our hypothesis that ultrasonography would reveal smaller areas of maximal cross-sectional injury, compared with areas obtained by use of MRI, was only supported at 16 weeks following induced injury. It appeared that use of MRI enabled us to detect a larger area of tendon damage in injuries of > 4 weeks' duration. Investigators in another study9 also found that MRI more accurately represented the severity of tendon injury evident during histologic examination following euthanasia. One of the limitations of the study reported here was that the palmar border of the tendon injury was less well-defined for collagenase-induced tendon injury than for naturally occurring tendon injury.

Although the signal intensity of the lesion for all 3 MRI sequences decreased over time, this decrease was not significant. Signal intensity of the lesion on T1-weighted images was the highest of the 3 sequences and had the greatest difference from the signal intensity of the unaltered control DDF tendon. Signal intensities of the lesion for the STIR images were slightly less than signal intensities for the T1-weighted images. Signal intensities of the lesion for T2-weighted images were lowest and had the smallest difference from the unaltered control DDF tendon. In another study,9 investigators found that equine tendon had a persistent increase in T1-weighted signal intensity for > 6 months following the onset of injury, whereas the T2-weighted signal intensity decreased over time. The present study revealed similar findings, although the horses were only monitored for 4 months after injury. The horses with tendon injury of 6 months' duration still had evidence of slight neovascularization and irregular collagen fiber arrangement, as determined during histologic assessment.9 The persistence of increased T1-weighted signal intensity has been described in healing tendons of humans for months to years after an injury.20,21

Clinically, T2-weighted images may be more useful for assessment of tendon healing because the signal intensities of T2-weighted images decrease over time, whereas the signal intensities of T1-weighted images remain high for years.9,21 The T2-weighted images are also less susceptible to artifact.22 Tissue T2-weighted images have a prolonged TR and are more sensitive to changes in water content and cellular infiltration.23 On the basis of this evidence, T2-weighted images may be more useful for evaluation of lesions.6 In contrast, evaluation of tendons over time by use of STIR sequences has not been described. In the present study, the STIR signal intensity of the tendon lesion closely mirrored the T1 signal intensity of the tendon lesion over time, except at 6 weeks after collagenase injection. The increase in STIR signal intensity at 6 weeks was mediated by an increase in STIR lesion signal for 1 horse. Given that STIR sequences are designed to eliminate fat-derived signals and that adipose tissue is not a major component of normal or healing tendon, it could be expected that STIR data correspond closely with T1-weighted profiles. Similar to results of other studies,9,21,22 results of the study reported here support the contention that multiple MRI sequences are necessary to provide the most information about injured tendons. In the present study, proton density sequences were not used because of the amount of time required for low-field magnet acquisition of the T1, T2, and STIR images of all 4 tendons.

Healing tendons contain increased amounts of GAG-substituted proteoglycans deposited in the repair matrix.9 The negative charges associated with the GAG sulfate residues bind water molecules and contribute to the increased signal intensity of the lesion seen with MRI over time. This GAG-mediated increase in tissue hydration would provide an increased lesion signal on T2-weighted images and could be misinterpreted as resulting from active inflammation, persistent edema, or neovascularization.9 Four horses in the present study had histologic evaluations of tissues obtained from the site of maximal injury at 3, 4, 6, 8, and 16 weeks following injury. Histologically, there was a decrease in proteoglycan content, minimal changes in collagen content, and minimal change in fiber alignment during the 16 weeks of the study. This decrease in proteoglycan content corresponded with the slow decrease in T2 signal intensities of the lesions seen on MRI images over time. Signal intensities of lesions detected on T1-weighted images could have been a result of changes in collagen fiber alignment. Although the histologic changes were quantitative, these changes were described as a pattern based on low numbers (1 tendon at 3, 4, 6, and 8 weeks after injury and 8 tendons at 16 weeks after injury). Thus, our study was limited by the small number of horses and histologic samples.

On the basis of results of the present study, both low-field MRI and ultrasonography are valuable diagnostic modalities for use in assessing acute tendon injury. In this study, horses were managed with stall rest and hand walking after collagenase-induced injury, and extrapolation of these results to horses with naturally occurring tendon injury and increasing exercise should be made with care. Allowing for this, ultrasonography may underestimate the extent of tendon damage in horses with injury of > 4 weeks' duration. It is possible that the flexor tendon reinjury rate is high because practitioners increase the amount of exercise in a horse prematurely on the basis of ultrasonographic findings for pathological changes in tendons. Perhaps a more critical evaluation of ultrasonographic findings could prevent tendon reinjury. For example, it has been reported24 that the longitudinal fiber pattern seen on ultrasonographic evaluation is the most accurate predictor for successful resumption of racing activity in Thoroughbreds. Other investigators have suggested25 the use of off-angled ultrasonographic images in which the ultrasound beam is not oriented perpendicular (ie, 90°) to the tendon during cross-sectional evaluation. The off-angle image takes into consideration the fact that the echogenicity of normal tendon becomes less echogenic when the angle of the beam is not oriented perpendicular to the fiber pattern. The off-angled technique provides an additional method for detecting small changes in echogenicity indicative of subtle injuries.24 Accepting these possibilities, low-field MRI provides a more sensitive determination of tendon abnormalities and a more detailed understanding of the cellular and tissue pathological changes in tendons at > 4 weeks after injury. It is unknown whether the T1-weighted signal intensity of injured tendon would ever return to values similar to that for unaltered normal tendon. In fact, it has been suggested in 2 studies9,12 that the T1 signal intensity for the lesion will remain high. The present study was limited in duration, and ideally, these tendon lesions would have been monitored for several years. Not surprisingly, the STIR images were similar to the T1-weighted images in which fat suppression in tendon would have minimal effect. Similar to results of another study,6 the T2-weighted images in the study reported here had the best subjective correlation with the ultrasonographically determined progression of healing. In this respect, MRI should be considered for evaluation of chronic tendon injuries in horses for which a successful return to competition is critical.

ABBREVIATIONS

ASI

Average signal intensity

DDF

Deep digital flexor

DICOM

Digital imaging and communications in medicine

MRI

Magnetic resonance imaging

ROI

Region of interest

SDF

Superficial digital flexor

STIR

Short tau inversion recovery

TE

Echo time

TR

Repetition time

a.

Sparhawk Laboratories Inc, Lenexa, Kan.

b.

Phoenix Pharmaceutical, St Joseph, Mo.

c.

Fort Dodge, Fort Dodge, Iowa.

d.

Pfizer Animal Health, New York, NY.

e.

Worthington Biochemical Corp, Lakewood, NJ.

f.

Model SSD-4000, Aloka, Tokyo, Japan.

g.

Kodak Carestream Pacs, version 10.2, Carestream Health, Rochester, NY.

h.

Akon, Decatur, Ill.

i.

Diazepam, Hospira Inc, Lake Forest, Ill.

j.

VetaKet, IUX Health, St Joseph, Mo.

k.

Hospira Inc, Lake Forest, Ill.

l.

0.25-T low-field strength VetMR Grande unit, Esaote Biomedica, Genova, Italy.

m.

Sigma Aldrich, St Louis, Mo.

n.

AxioVision, version 4.7, Carl Zeiss Int, Thornwood, NY.

o.

SigmaStat, version 3.0, Systat Software Inc, San Jose, Calif.

References

  • 1.

    Genovese RLRantanen NWHauser ML, et al. Diagnostic ultrasonography of equine limbs. Vet Clin North Am Equine Pract 1986; 4:145226.

  • 2.

    Rantanen NW. The use of diagnostic ultrasound in limb disorders of the horse: a preliminary report. J Equine Vet Sci 1982; 2:6264.

  • 3.

    Rantanen NWHauser MLGenovese RL. Superficial digital flexor tendinitis: diagnosis using real-time ultrasound imaging. J Equine Vet Sci 1985; 5:115119.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 4.

    Palmer SEGenovese RLLongo KL, et al. Practical management of superficial digital flexor tendinitis in the performance horse. Vet Clin North Am Equine Pract 1994; 10:425481.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 5.

    Pickersgill CHMarr CMReid SW. Repeatability of diagnostic ultrasonography in the assessment of the equine superficial digital flexor tendon. Equine Vet J 2001; 33:3337.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 6.

    Crass JRGenovese RLRender JA, et al. Magnetic resonance, ultrasound and histopathologic correlation of acute and healing equine tendon injuries. Vet Radiol Ultrasound 1992; 33:206213.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 7.

    Crass JRCraig EVFeinberg SB. Ultrasonography of rotator cuff tears: a review of 500 diagnostic studies. J Clin Ultrasound 1988; 16:313327.

  • 8.

    Fleckenstein JLWeatherall PRParkey RW, et al. Sports-related muscle injuries: evaluation with MR imaging. Radiology 1988; 172:793798.

    • Search Google Scholar
    • Export Citation
  • 9.

    Kasashima YKuwano AKatayama Y, et al. Magnetic resonance imaging application to live horse for diagnosis of tendinitis. J Vet Med Sci 2002; 64:577582.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 10.

    Brokken MTSchneider RKSampson SN, et al. Magnetic resonance imaging features of proximal metacarpal and metatarsal injuries in the horse. Vet Radiol Ultrasound 2007; 48:507517.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 11.

    Bischofberger ASKonar MOhlerth S, et al. Magnetic resonance imaging, ultrasonography, and histology of the suspensory ligament origin: a comparative study of normal anatomy of Warm-blood horses. Equine Vet J 2006; 38:508516.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 12.

    Sampson SNSchneider RKTucker RL, et al. Magnetic resonance imaging features of oblique and straight distal sesamoidean desmitis in 27 horses. Vet Radiol Ultrasound 2007; 48:303311.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 13.

    Berg WAGutierrez LNessAiver MS, et al. Diagnostic accuracy of mammography, clinical examination, US, and MR imaging in preoperative assessment of breast cancer. Radiology 2004; 233:830849.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 14.

    Williams IFMcCullagh GDGoodship AE, et al. Studies on the pathogenesis of equine tendinitis following collagenase injury. Res Vet Sci 1984; 36:326328.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 15.

    Redding WR. Booth LC, Pool RR. The effects of polysulphated glycosaminoglycan on the healing of collagenase-induced tendinitis. Vet Comp Orthop Traumatol 1999; 12:4855.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 16.

    Keg PRvan den Belt AJMMerkens HW, et al. The effect of regional nerve block on the lameness caused by collagenase induced tendonitis in the midmetacarpal region of the horse: a study using gait analysis and ultrasonography to determine tendon healing. Zentralbl Veterinarmed A 1992; 39:349364.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 17.

    Rich LBraz PW. Collagen and picro sirius red staining: polarized light assessment of fibrillar hue and spatial distribution. J Morphol Sci 2005; 22:97104.

    • Search Google Scholar
    • Export Citation
  • 18.

    Whittaker PKloner RABoughner DR, et al. Quantitative assessment of myocardial collagen with picro sirius red staining and circularly polarized light. Basic Ref Cardiol 1994; 89:397410.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 19.

    Fung DTWang VMLaudier DM, et al. Subrupture tendon fatigue damage. J Orthop Res 2009; 27:264273.

  • 20.

    Shalabi A. Magnetic resonance imaging in chronic Achilles tendinopathy. Acta Radiol Suppl (Stockholm) 2004; (432): 145.

  • 21.

    Shalabi AKristoffersen-Wiberg MAspelin P, et al. MR evaluation of chronic Achilles tendinosis. A longitudinal study of 15 patients preoperatively and two years postoperatively. Acta Radiol 2001; 5;42:269276.

    • Search Google Scholar
    • Export Citation
  • 22.

    Hayes CWParellada JA. The magic angle effect in musculoskeletal MR imaging. Top Magn Reson Imaging 1996; 8:5156.

  • 23.

    Benjamin MToumi HRalphs JR, et al. Where tendons and ligaments meet bone: attachment sites (‘entheses’) in relation to exercise and/or mechanical load. J Anat 2006; 208:471490.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 24.

    Genovese RL. Equine limb ultrasonography: notes and tidbits, in Proceedings. Fall Conf Vet Univ Ill 2008;153.

  • 25.

    Werpy NMCharles BRantanen N. Should I throw away my ultrasound machine now that MRI is here? A review of ultrasound for the diagnosis of musculoskeletal injury in the equine patient, in Proceedings. 54th Annu Meet Am Assoc Equine Pract 2008;439446.

    • Search Google Scholar
    • Export Citation
All Time Past Year Past 30 Days
Abstract Views 60 0 0
Full Text Views 830 566 210
PDF Downloads 346 217 30
Advertisement