Fatty infiltration in the intact supraspinatus tendon; a normal physiological response with increasing age and female gender

Abstract

Background

Fatty infiltration and muscle atrophy of supraspinatus are used as markers of chronicity in rotator cuff tears and are known to both be independently related to poorer outcomes following surgical repair.¹ We hypothesized that supraspinatus muscle atrophy and fatty infiltration increases with age irrespective of whether the rotator cuff is intact and therefore cannot be used as accurate markers for chronicity.

Method

Retrospective review of 280 patients who underwent 3.0 T shoulder MRI’s with either a normal scan or rotator cuff tear. Two independent observers reviewed the images. Data collected included intact rotator cuff tendons looking specifically at supraspinatus muscle height/length: suprascapular fossa ratio, tangent sign and Goutallier grade for fatty infiltration.

Results

There were 90 scans with intact rotator cuff tendons. Mean age was 51 years (range 17–86); 52 males, 38 females. On multiple regression analysis, there was a positive correlation of age with fatty infiltration and muscle atrophy on all parameters in the normal intact cuff. Females were significantly more likely than males to have a higher grade of fatty infiltration.

Conclusion

Age and female gender are risk factors for rotator cuff atrophy and fatty infiltration in patients with normal rotator cuffs. Therefore, these parameters should be used with caution by surgeons when deciding on tear chronicity and the potential to repair the torn rotator cuff.

Introduction

Fatty infiltration (FI), originally described by Goutallier, is a degenerative process affecting the rotator cuff, the onset of which is thought to be a result of tearing of the tendon–muscle unit. ¹ The severity of FI has been shown to increase with time following injury to the tendon. ² FI and muscle atrophy are used as markers of chronicity in rotator cuff tears, and studies have shown that they independently correlate with poorer outcomes following surgical repair. ³

The onset of FI, however, is not fully understood. There is contradictory evidence of its relationship to injury, as Gueniche and Bierry also found the presence of FI in intact rotator cuff tendons. ⁴ The question is then raised; is FI a normal physiological response, and what factors influence its presence? Few studies have been performed to discover its relationship to other independent factors such as gender, age, BMI or smoking.

The aim of this study was to investigate fatty changes in the intact rotator cuff. We hypothesized that FI and muscle atrophy are normal physiological responses in ageing tendon and cannot therefore be used as an accurate marker for chronicity of a rotator cuff tear.

Material and methods

Patient selection

Local institution approval was obtained for this study.

A total of 280 3.0 T shoulder MRI’s were retrieved from the Hospital Database. All scans were performed using our standard protocol in neutral shoulder rotation based on the plane of the scapula – parallel and perpendicular to the scapular body, allowing true coronal and axial measurements at the glenoid. Slice thickness was set to 3 mm. Indication for the scan was determined by the treating clinician. Exclusion criteria included presence of partial or full-thickness rotator cuff tears, movement artifact, previous surgery on the shoulder and tumour. Scans were identified with normal supraspinatus tendons after review by a musculoskeletal radiologist and two orthopaedic surgeons.

Measurements

A number of measurements were taken on the T1 Fast-Spin Echo MRI sequence. FI was measured according to the Fuch modification of the Goutallier’s Classification. ⁵ Grade 0 no FI, Grade 1 some fatty streaks, Grade 2 less fat than muscle, Grade 3 equal fat and muscle and Grade 4 more fat than muscle.

The ratio of muscle belly in the supraspinous fossa was then measured. The maximum length and width ratio of the supraspinatus muscle belly to supraspinous fossa was measured on the same sagittal ‘Y- view’ where the body of the scapula, scapula spine and medial border of the coracoid process form (Figure 1).

Figure 1.

Example of increased fatty infiltration in lower length and height muscle:fossa ratio.

Finally, the Tangent sign was measured on the sagittal T1 oblique plane, with a line drawn between the upper surface of the scapular spine and the upper surface of the coracoid process. A positive Tangent sign was considered to be supraspinous muscle above this line (Figure 2). ⁵

Figure 2.

Negative tangent sign: MRI sagittal oblique T1 view. A line is drawn between the scapular spine and the coracoid process. A normal supraspinatus muscle belly lies above this line (negative sign) and an abnormal supraspinatus muscle belly (small from fatty atrophy) lies below this line (positive sign).

Statistical analysis

Shapiro–Wilk testing was used to test for data following a normal distribution. Spearman’s correlation coefficient was used to test the associations between the continuous variables and FI and the non-parametric Mann–Whitney U test was used to test the association by gender. Post-hoc tests were used to decrease rate of type 1 error. Intra-class correlation coefficient was 0.94 (95% C.I.: 0.90–0.96) suggesting excellent agreement between observers. p Values ≤ 0.05 were considered statistically significant.

Results

There were 90 scans with intact rotator cuff tendons. Mean age was 51.3 years (range 17–86); 52 males, 38 females. There were 31 patients with grade 0, 45 patients with grade 1, 13 patients with grade 2 and one patient with grade 3 FI. Primary pathology on the MRI scan is shown in Table 1.

Table 1.

Distribution of primary pathology on MRI scan.

Diagnosis	n
Subacromial bursitis	21
Normal	19
Glenohumeral joint OA	12
Adhesive capsulitis	11
AC joint pathology	10
Labral tear/instability	8
Calcific tendinitis	8
Long head biceps rupture	1
TOTAL	90

FI and muscle volume

Higher FI was associated with a lower length (R = −0.642, p < 0.001) and height (R = –0.539, p < 0.001) muscle:fossa ratio (Figure 3). Patients with a higher Goutallier grade were more likely to have a positive Tangent sign (R = 0.334, p = 0.002).

Figure 3.

Example of increased fatty infiltration in lower length and height muscle:fossa ratio.

Age

There was a positive correlation between FI and age (R = 0.423, p < 0.001). Muscle volume was also affected by age. Increasing age was associated with a lower length (R = –0.399, p < 0.001) and height (R = –0.402, p < 0.001) muscle:fossa ratio (Figure 4).

Figure 4.

Comparison of fatty infiltration in the supraspinatus muscle (Fuchs modification of Goutallier classification) with age. The mean age and error bars are displayed.

Female gender

There was a statistically significant difference between FI and female gender p = 0.004. Female gender was associated with a lower length (R = –0.321, p = 0.003) and height (R = –0.364, p = 0.001) muscle:fossa ratio.

Discussion

Previous studies have focused on supraspinatus fatty change as a poor prognostic marker for rotator cuff repair. ⁶ It has been established that FI is a sequelae of a chronic rotator cuff tear. However, we have established that it can be present in the intact rotator cuff and is likely to represent a normal physiological process in ageing tendon. As such, repair of a cuff which is already infiltrated with high levels of fat may have a higher retear rate despite an acute repair. The association of high pre-operative FI and functional outcomes is contentious, however, and satisfactory outcomes can still be achieved post-operatively. ⁷ Matson et al., showed that age, obesity as well as diabetes mellitus can also be predictive of FI in rotator cuff tendons on MRI. ⁸

Recent evidence suggests that age and disuse can influence the severity of FI. ⁹ With age, muscle involutes, resulting in an increase in protein degradation and muscle atrophy, that is replaced by fat. ¹⁰ Muscle atrophy also occurs in this instance and is not surprising that there was a significant correlation between FI and other measures of muscle volume, including muscle:fossa ratios and the tangent sign. Our study revealed that even in the intact rotator cuff, 16 patients had a Goutallier grade above 2, with a positive correlation of age with FI and also a decrease in height and length to fossa ration with increasing age.

Female gender was a more surprising finding. There was a positive correlation between FI and female gender. With females tending to have more FI than males, and also associated with a lower length and height muscle:fossa ratio. There is known to be differences in the fat distribution between gender, with a greater proportion of adipose tissue likely to be stored in the hips and thighs in women, and more around the trunk and abdomen in men. ¹¹ There is also shown to be a gender difference in the hypertrophy of skeletal muscle in response to exercise, as shown on whole body MRI scan. ¹² Little research however has been performed to support these findings in the shoulder, only that women with rotator cuff pathology suffer from higher levels of disability pre and post rotator repair. ¹³

There were several limitations in our study. First, this was a retrospective study, including patients from our database with possible shoulder pathology. Though the intact cuffs exclusion criteria is explicit, other pathology such as shoulder osteoarthritis or AC Joint arthritis may have been included. Patients and FI therefore may be contributed to by disuse through pain in patients with these other pathologies, and we did not objectively assess pain levels prior to the scan. We had more males than females in our cohort which may have introduced bias in our comparative analysis. Our population group above 60 years was small, and this is the age-group which most commonly suffers with rotator cuff pathology. Our study cohort was limited, and therefore we could not use a statistical model with the data to identify independent associations with FI grade. We believe our results, therefore, do not provide a definitive answer to our study aims, but pose important questions as to the clinical relevance of FI in the torn rotator cuff.

Conclusion

In conclusion, in patients with intact rotator cuffs, there is a positive correlation between age and FI. As age increases, FI also increases. When taking into account gender, females have significantly more FI within the intact supraspinatus muscle than males, and males have a higher supraspinatus muscle:fat ratio than females. Care should be taken when using FI as an indicator of chronicity in rotator cuff tears, as it is present with increasing age and female gender in the intact rotator cuff. Therefore, the treating surgeon should be cautious about using FI as a contraindicating parameter when deciding on whether to repair the torn rotator cuff or not.