None of the authors have a professional or financial affiliation that may be perceived to have biased the presentation of this research.
Effect of timing of surgical SSP tendon repair on muscle alterations
Article first published online: 28 JUL 2014
© 2014 Orthopaedic Research Society. Published by Wiley Periodicals, Inc.
Journal of Orthopaedic Research
Volume 32, Issue 11, pages 1430–1435, November 2014
How to Cite
Uhthoff, H. K., Coletta, E. and Trudel, G. (2014), Effect of timing of surgical SSP tendon repair on muscle alterations. J. Orthop. Res., 32: 1430–1435. doi: 10.1002/jor.22692
- Issue published online: 22 SEP 2014
- Article first published online: 28 JUL 2014
- Manuscript Accepted: 20 JUN 2014
- Manuscript Received: 13 FEB 2014
- Workplace Safety and Insurance Board of Ontario. Grant Number: 04031
- Canadian Institutes of Health Research. Grant Number: MOP-81395
- supraspinatus tendon;
- fat accumulation;
To investigate the impacts of delayed repairs of a supraspinatus tendon tear on the supraspinatus muscle, we used an animal model data from two previously published studies in which one supraspinatus (SSP) tendon was detached. In one cohort, the rabbits were killed in groups of 10 at 4, 8, and 12 weeks. In the other cohort, a repair was done at these time points, 12 rabbits each, and the animals killed were 12 weeks later. SSP fossa volume (Muscle belly plus extramuscular fat [e-fat] volume), percentage of intramuscular fat (i-fat), and muscle tissue volume (muscle belly volume minus i-fat), as well as CT determination of e-fat and i-fat of both cohorts, were compared. Fossa volume increased (p < 0.05). Muscle belly and muscle tissue volumes did not increase after repair (p > 0.05), but early repair prevented further volume losses, a fact not seen after 8 and 12 weeks delay of repair. No reversal of e-fat or of i-fat occurred, in fact i-fat almost doubled after 4 weeks delay of repair (p < 0.05). CT studies confirmed the fat results. We conclude that early repair prevented loss of muscle belly and muscle tissue volumes, but that it has no positive influence on fat accumulation. © 2014 Orthopaedic Research Society. Published by Wiley Periodicals, Inc. J Orthop Res 32:1430–1435, 2014.
Many clinical and experimental studies have shown that after rupture of the supraspinatus (SSP) tendon, this muscle undergoes retraction, atrophy and fatty accumulation, all worsening with the duration of SSP tendon disruption.[1-7] These muscle alterations influence the clinical outcome of SSP tendon repair surgery and therefore the decision for surgery. The optimal timing of surgical repair of a complete SSP tendon tear is unknown. Some surgeons may want an early repair to prevent atrophy and fatty accumulation of the SSP muscle or to ensure their reversal. However, prevention or reversibility of these muscles alterations with regards to the timing of repair is not known.
Clinical studies can assess the effects of rotator cuff repair by comparing the cuff's state at the time of surgery to the state at the moment of follow-up. Methods and outcomes employed for the assessment are non-invasive, and the patient serves as its own control.[9, 10] The exact evaluation of the clinical outcome, however, is hindered by the following facts: Timing of the cuff tear can sometimes not be determined with certainty, patients and cuff defect cannot be standardized, and exact timing of surgical repair cannot be standardized especially early after tendon tear because of constraints related to the delivery of medical services. Experimental investigations, on the other hand, can standardize these parameters and can use invasive and more precise methods to characterize the effect of SSP repair but preclude the use of the animal as its own control. Researchers have instead used the opposite intact shoulder or shoulders of non-operated animals for comparison. The drawback of the comparison with normal shoulders instead of shoulders with a torn cuff in need of repair limits the inference of the findings to clinical situations. An experimental design using two cohorts of animals circumvents this drawback; it allows using precise invasive methods to assess the effect of repair surgery on muscle changes that are relevant to clinical situations.
Precise experimental data on SSP muscle weight, volume and extramuscular fat (e-fat) and intramuscular fat accumulation (i-fat) have been gathered at various durations after tendon rupture. Similar data have also been assembled at various intervals after repair. However, never has a comparison between these states been attempted. As a prerequisite for such a comparison, the following conditions must be controlled: Identical age and sex of the animal and identical surgical procedures to produce the SSP tendon tear. Moreover, the time elapsed between tendon tear and killing of the first cohort must be identical to the time elapsed between tendon tear and repair of the second cohort. Finally, identical methods must be employed to determine muscle atrophy and fatty accumulation.
Our previous investigations met these prerequisites and allowed us to plan a comparison between two cohorts of animals: First cohort: One group killed after a short duration of rupture (4 weeks), one after a medium duration of rupture (8 weeks), one after longer duration of rupture (12 weeks); Second cohort: One group tendon reattachment after short duration of rupture and followed up for a 12-week period (4 + 12 weeks), one group reattached after a medium duration of rupture and followed up for a 12-week period (8 + 12 weeks) and one reattached after a longer duration of rupture and followed up for a 12-week period (12 + 12 weeks).
Comparing these two cohorts will permit a reliable assessment of SSP muscle alterations and the influence of various delays before repair on these alterations, results that may benefit evidence-based clinical practice. Our objective was to measure the impact of early versus delayed SSP tendon repair on muscular and adipose changes after SSP tendon tear. Our first hypothesis was that early repair preserves SSP muscle belly volume and muscle tissue volume. Our second hypothesis was that no reversal of e-fat or i-fat accumulation takes place after repair irrespective of its timing.
The findings of a preserved muscle belly volume and of muscle tissue volume in the group with a short delay of repair and the absence of any beneficial effect of repair on fat accumulation may influence the surgical decision-making.
MATERIALS AND METHODS
We compared data from two previously published investigations.[11, 12] Institutional approval from the animal care and use committee for these projects had been obtained. In the first cohort of 30 female New Zealand rabbits, one SSP tendon was divided and the animals killed in groups of 10 after 4, 8, and 12 weeks (4, 8, 12). Both shoulders of 15 age- and sex-matched non-operated rabbits served as controls. In the second cohort of 36 female New Zealand rabbits, the tendon of the supraspinatus was divided. In 12 rabbits, the tendon was repaired after 4 weeks (4 + 12), in the second group after 8 weeks of detachment (8 + 12), and in the last group after 12 weeks (12 + 12). These animals were killed 12 weeks after repair. Both shoulders of 18 age- and sex-matched non-operated rabbits served as controls (Fig. 1). Details of the surgical techniques were described in our previous papers.[11, 12]
SSP Tendon Detachment, First Cohort
After complete detachment of SSP tendon at the greater tuberosity, the tendon stump was enveloped with a polyvinylidene fluoride membrane to avoid spontaneous reattachment.
All animals received fentanyl and buprenorpnine for 3 days postoperatively and were allowed to roam freely in their cages with unlimited access to food and water.
SSP Tendon Repair, Second Cohort
In the second cohort, using the previous incision the polyvinylidene membrane was removed, a trough burred into the greater tuberosity and the tendon inserted into it.
Collection of Specimens
All animals were killed with a pentobarbital overdose at the specified times. All shoulders were harvested on bloc, taking care to leave the SSP muscle and e-fat intact. Specimens were frozen at −30°C until CT scanning.
All specimens were thawed for 12 h prior to CT scanning. To measure i-fat, we used a four slice helical scanner in the axial plane with a table speed of 3.75 mm/s, pitch of 0.75, 120 kV, 50 mA, 1.25 mm collimation and 0.625 mm slice overlap in a soft tissue algorithm (LightSpeed Plus, GE Healthcare, Milwaukee, WI). The SSP muscle was scanned at three levels for measuring fat content in Hounsfield units: Proximal quarter, mid-part and distal quarter. A 10 mm2 region of interest was placed on each of the three muscle segments by a radiologist blinded to the identity of the specimens to measure i-fat. I-fat was measured in Hounsfield units (HU); the lower the value of HU the greater the amount of fat. E-fat overlying either the proximal quarter, middle half or distal quarter of the muscle was assessed as either being absent or present (Fig. 2).
Volume of SSP Muscle
The combined volume SSP muscle belly with overlying fat (e-fat) was determined by water displacement. We termed the volume of muscle belly with its overlying e-fat: The SSP fossa content. The volumes of muscle belly and e-fat were then measured separately. The muscle belly volume contained both muscle tissue and i-fat. Determining muscle tissue content is important because the force generated depends only on content of muscle tissue. To arrive at muscle tissue content, the percentage of i-fat was measured on three histologic sections of the SSP muscle (see Histology), proximal quarter, middle half and distal quarter, and averaged. The muscle tissue volume was calculated by subtracting the averaged percentage of i-fat from muscle belly volume, therefore providing a three-dimensional estimate of muscle tissue volume.
The muscles were fixed for a minimum of 1 week in 4% paraformaldehyde. Three cross-sectional blocks of 2 mm in width were cut at the proximal quarter, the mid-part, and the distal quarter of the muscle belly. The blocks were placed in an osmium tetroxide solution on a rotating plate for 1 week; the solution was then changed and blocks left in the solution for another week. The tissues were then washed in a 50 ml centrifuge tube with water and placed thereafter in 60% alcohol. The blocks were embedded in paraffin in a vacuum oven, cut in 6 µm thin sections and counterstained with hematoxylin–eosin. The transverse sections were photographed using a Pixelink Megapixel FireWire camera (Viana Corporation, Ottawa, Canada), mounted onto an Olympus SZ61 dissection microscope (Olympus Corporation, Tokyo, Japan). The magnification used was X6.7. SSP entire muscle cross-sectional areas were measured using ImageJ (1.34s, National Institutes of Health, Bethesda, MD). The areas of fat, stained black, were measured the same way using the entire cross-sectional area (Fig. 2). The percentage of the entire cross-sectional area occupied by fat was established.
Data and Statistical Analysis
Animal weights at arrival varied from 2.5 to 3.3 kg. These variations affect the volumes of muscle and fat. In this paper, in order to compare identical conditions, we corrected muscle and fat data to a baseline weight of 3.1 kg. Data were analyzed using SPSS statistical software (version 20.0). Outcome measures were compared between the detached SSP and repaired SSP cohorts. We used one-way ANOVAs to assess the effect of timing of surgical repair on groups 4 versus 4 + 12; 8 versus 8 + 12, and 12 versus 12 + 12 for outcome measures: SSP fossa content, SSP muscle belly volume, SSP muscle tissue volume, e-fat volume, i-fat cross-sectional area, i-fat signal on high-resolution CT, and presence of e-fat on CT. A p value of <0.05 was considered to be statistically significant.
None of the rabbits were lost; the final sample size was 99 rabbits or 132 shoulders. All repairs were found to be intact at sacrifice.
Volume of SSP Muscle and Fat
The SSP fossa volume was larger at 12 weeks post-repair than at surgery of 4-week-old tendon tears (9.9 ± 0.2 vs. 11.2 ± 0.3 ml; p ≤ 0.05, Fig. 3). No such increase happened after reattaching 8 or 12 weeks old tears.
Muscle belly volume did not improve at 12 weeks post-repair of 4, 8, or 12 weeks old SSP tendon tears (all p > 0.05). However, repair after a short duration of rupture (4 weeks) prevented a muscle belly volume decrease, such a prevention of decrease not seen in repairs after medium or long durations of rupture (7.9 ± 0.3 ml compared to 6.5 ± 0.4 ml and 7.1 ± 0.5 ml; p > 0.05).
Muscle tissue volume similarly did not improve at 12 weeks post-repair of 4, 8, or 12 weeks old SSP tendon tears (all p > 0.05; Fig. 3). And, the earlier repair at 4 weeks prevented a muscle tissue volume decrease, such a prevention of decrease not seen in repairs after medium or long durations of rupture (7.1 ± 0.3 ml compared to 5.8 ± 0.4 ml and 6.5 ± 0.5 ml; p > 0.05). The muscle tissue volume remained below control levels at 8 + 12 and 12 + 12 (Fig. 3).
E-fat volume did not decrease at 12 weeks post-repair of 4, 8, or 12 weeks old SSP tendon tears (all p > 0.05; Fig. 4).
Computed Tomography i-fat
Similarly, i-fat determined by high-resolution CT at the distal quarter of SSP muscles increased 12 weeks post-repair of 4 and 8 week old SSP tendon tears (4 + 12 vs. 4: 44.5 ± 4.7 vs. 60 ± 1.6 HU (p ≤ 0.05) and 8 + 12 vs. 8: 29.2 ± 7.2 vs. 58.5 ± 3.1 HU (p ≤ 0.05; Fig. 5).
No reversal occurred in repairs after medium or long durations of rupture. I-fat almost doubled 12 weeks post-repair of 4 weeks old SSP tendon tears (13.7 ± 2.1 vs. 7.0 ± 1.1%; p ≤ 0.05; Fig. 5).
Computed Tomography e-fat
Presence of e-fat as determined by CT at the distal quarter of SSP muscles was more prevalent at 12 weeks post-repair than 4 weeks after a tendon tear (67% vs. 20%; Fig. 4).
Early repair of SSP tendon tears preserved the muscle belly and muscle tissue volumes, confirming our first hypothesis. Importantly, however, repair did not lead to reversal of the muscle atrophy at 12 weeks follow-up. The degree of muscle atrophy that persisted after successful repair of medium and long duration tendon rupture (8 + 12 and 12 + 12) becomes evident when comparing the muscle belly volume of the experimental groups to that of the control animals (Fig. 3).
SSP tendon repair, early or late, also failed to reverse the e-fat and i-fat accumulation created by SSP tendon tear. This confirmed our second hypothesis. Not only did repair but also reverse fat accumulation, 4 and 8 weeks after tear did not even prevent further fat accumulations measured at 12 weeks follow-up by high-resolution CT.
These findings remind surgeons that SSP muscle atrophy and fat accumulation are ongoing while a SSP tendon is not in continuity. How do they influence decision-making? Should an early surgery be carried out to prevent these changes? Our comparisons of experimental data suggest that early surgery may in fact prevent progression of muscle tissue atrophy. There was no actual reversal of atrophy, but the repaired tendon limited the accrual of muscle atrophy. The limited atrophy preserved muscle tissue volume which in turn can predict better strength generation and better muscle function.
The situation was different regarding muscle fatty changes. Successful surgical repair of the torn SSP tendon did neither reverse nor prevent fatty accumulation both outside and inside SSP muscles as measured 12 weeks post-repair. Fat accumulation in SSP muscles decreases the contractile force as described by Gerber et al. Its effect on function is poorly documented. However, our data suggest that surgical decision-making cannot be based on the hope of reversing fat accumulation.
Our determination of the fossa content in this study is based on the volume of the entire SSP muscle belly with its overlying fat (e-fat), a tri-dimensional measurement. In contrast, Thomazeau et al. and Tae et al. based their measurements of the SSP fossa on one MRI oblique-sagittal cross-section made at the midpoint between the bottom of the scapular notch and a point just above the spino-glenoid notch, a two dimensional measurement. This demonstrates the enhanced precision of the animal model. Thomazeau et al. defined the limits of the SSP fossa as an area bordered by the scapula, the scapular spine, the clavicle, and the deep surface of the trapezius muscle. They measured the cross-sectional area of the SSP muscle belly and that of the entire surface of the fossa and calculated the occupation ratio, the lower the ratio the greater the muscle atrophy at this level. These authors did not describe the nature of the tissue occupying the remaining area of the fossa. It fell to Tae et al. using automated measurements based on the principles described by Thomazeau et al. to identify the tissue occupying the remainder part of the SSP fossa as fat (called e-fat in our study). They also showed that decrease of the muscle belly area was accompanied by fat accumulation. However, none of the above-mentioned authors considered the fat content inside the muscle (i-fat) which can be measured clinically using signal density. Consequently, fullness of the supraspinatus fossa on clinical examination may not reflect the status of the muscle belly.
The results of the non-invasive CT evaluations of e-fat and i-fat used in this study were comparable to those of the invasive histologic methods. This permitted the conclusion that fat determinations done in patients are reliable surrogates of real fat content both for e-fat and i-fat.
The responses of the SSP to transection and to repair were different in respect to muscle tissue and to fat accumulation leading to our assumption that the processes of muscle tissue changes and those of fat accumulation are differently modulated. For example, an earlier repair prevented further muscle losses whereas fat accumulation increased.
In a previous study, the muscle status after delayed repair was compared to that of non-operated control shoulders. In clinical practice, the outcomes of delayed rotator cuff repair are compared to the muscle's status at the time of surgery. In the current study, the sizeable difference between a normal control versus the muscles' status after varying durations of detachment was well reflected by all graphs. Data collected after experimental tendon detachment instead of normal rabbit muscles allowed a more realistic appreciation of the effects of delayed tendon repair to derive evidence-based decisions.
The pathomechanism of i-fat accumulation after rotator cuff tear has been explained by two different phenomena, both based on the precondition of muscle retraction. One stipulates that the retraction causes a tethering of the suprascapular nerve [15, 16] leading in turn a muscle atrophy and fat accumulation. The other measured the pennation angle that increases during retraction, widening the spaces between the muscle fibers, and thus creating loci of fat accumulation.[6, 17] The pennation angle was not measured in the current study. We could show in a recent study that neither tethering of the suprascapular nerve nor an increase of the pennation angle are necessary prerequisites for fat accumulation. In that experiment, we detached the supraspinatus tendon and repaired it immediately, thus preventing any muscle retraction; intramuscular fat still accumulated appearing already after 1 week and increasing over the six ensuing weeks.
The choice of different rabbits for comparison before and after repair may have impacted on the results of our investigation. We tried to minimize this impact by establishing stringent and identical prerequisites and volume corrections for both cohorts as well as identical sex and age of all rabbits. We also used identical surgical and laboratory techniques as well as the same methods of evaluation.
Our study being limited to a 12-week period of delay after SSP tendon tear and to a duration of 12 weeks after repair cannot predict muscle changes in instances of repair of chronic tears nor follow ups of repairs exceeding 3 months. However, our results using precise and invasive measures, support two recent longitudinal clinical studies where reversal of neither fat accumulation nor of muscle atrophy could be observed on imaging with an average follow-up of 1 year.[8, 17] Shoulder surgeons may consider this evidence in the decision and timing of rotator cuff repair. In clinical practice, rotator cuff repair is rarely considered an urgent intervention, and the delay between onset of tear and surgery often amounts to months. Yet, in younger and active persons, repair should be done as soon as possible, and the outcome of our investigation supports this advice. After all, the postoperative SSP muscle strength depends on the preservation of muscle power.
We thank Julie Courchesne and Philippe Poitras for the surgical procedures, Ying Nie for tissue processing, Mélanie Aubé for the histologic measurements, Greg Latourelle for CT scanning and image postprocessing, and Tim Ramsay, PhD, for consultation on statistical analyses. Special thanks to Stephen E. Ryan, radiologist, Department of Radiology, University of Ottawa, for his valuable advice.
- 71989. L‘évaluation par le scanner dela trophicité des muscles des coiffes des rotateurs ayant une rupture tendineuse. Rev Chir Orthop 75:126–127., ,
- 122012. Muscle tissue atrophy, extramuscular and intramuscular fat accumulation, and fat gradient after delayed repair of the supraspinatus tendon: a comparative study in rabbits. J Orthop Res 30:781–786., , , et al.
- 152010. Development of fatty atrophy after neurologic and rotator cuff injuries in an animal model of rotator cuff pathology. J Bone Joint Surg Am 92-A:2270–2278., , , et al.
- 172004. Effect of tendon release and delayed repair on the structure of the muscles of the rotator cuff: an experimental study in sheep. J Bone Joint Surg Am 86:1973–1982., , , et al.