Macroscopic features
Of the 36 samples, 24 (67%) had tears present in the hip abductors (either the gluteus medius, minimus or both). Of these 24 samples with tears present, 11 (46%) were isolated gluteus minimus tears, 5 (21%) were isolated gluteus medius tears and 8 (33%) involved the tendon of both muscles. Overall, the most common macroscopic finding was an isolated gluteus minimus tear, followed by tears involving both the gluteus minimus and medius.
In the 13 samples where a gluteus medius tear was identified, 11 tears were found in the anterior/middle portion of the tendon which inserts at the lateral facet of the greater trochanter. Only 2 small partial tears were found in the posterior portion of the gluteus medius. This portion inserts separately through a cord-like tendon at the supero-posterior facet of the greater trochanter. In both cases, the tears were less than 5mm and no other abnormalities were identified in either the gluteus medius or minimus. In the gluteus minimus, all tears were found anteriorly, adjacent to the attachment of the gluteus medius tendon (Figure 3).
Both complete and partial tears were found in the gluteus medius and minimus tendons. Partial tears were only observed on the under surface of the tendons. All tears were found in the distal 1cm of tendon, while no tears were observed in either the gluteus medius or minimus tendons proximally. Interestingly, three samples with a complete gluteus minimus tear were also associated with partial under surface tears in the gluteus medius tendon immediately adjacent to the gluteus minimus tear.
The average width of a complete gluteus medius tear was 15.6mm (SD 7.2) and 9.8mm (SD 6.5) for partial tears. The average width of a complete gluteus minimus tear was 7.9mm (SD 4.3) and 6.5mm (SD 3.7) for partial tears. The differences in size between complete and partial tears was not statistically significant (p=0.24).
Other macroscopic features were compared between samples with and without tears (Table 1). Of samples with complete tears, 89% were observed to have gross fatty atrophy of the related muscle. Delamination of the tendon was observed predominantly in samples with complete or partial tears. Bony changes (spur formation) were only observed in one sample with a complete tear.
Table 1. Associated features of hip abductor tendon tears
|
|
Complete tear
|
Partial tear
|
Normal
|
p-value
|
Fatty atrophy of muscle
|
8 (89%)
|
2 (14%)
|
0
|
<0.001
|
Delaminated
|
5 (56%)
|
8 (57%)
|
1 (10%)
|
0.030
|
Bony changes
|
1 (11%)
|
0
|
0
|
NS
|
NS = not significant
Microscopic features
Lateral insertion of gluteus medius:
Comparisons were made between normal samples, samples with partial tears and samples with complete tears. The Modified Movin’s score was significantly different in normal (3.2), partial tear (8.6) and complete (8.9) tear samples (ANOVA p=0.0005) (Figure 4). Further post-hoc analysis separately demonstrated higher degeneration scores in the samples with complete (p=0.0049) and partial tears (p=0.0099) when compared to those with no tears (Table 2). The difference in degeneration between complete and partial tears was not statistically significant.
Table 2. Histology tendon and enthesis scores for lateral insertion of gluteus medius
|
Tendon (SD)
|
p-value
(post hoc)*
|
Enthesis (SD)
|
P-value
(post hoc)*
|
Complete Tear
|
8.9 (2.0)
|
0.0049
|
7.1 (1.5)
|
0.011
|
Partial Tear
|
8.6 (1.7)
|
0.0099
|
7.4 (1.9)
|
0.0075
|
No Tear
|
3.2 (1.8)
|
reference
|
3.0 (2.0)
|
reference
|
p-value (ANOVA)
|
0.0005
|
|
0.0011
|
|
*Post hoc for multiple comparisons
In the enthesis, the semi-quantitative degeneration scores (Appendix 2) in normal (3.0), partial (7.4) and complete (7.1) tear samples were significantly different (ANOVA p=0.0011). Again, post-hoc analysis demonstrated significantly higher scores in samples with both complete (p=0.011) and partial tears (0.0075) when compared to those with no tears (Table 2).
When evaluating the degeneration scores in the superficial, middle and deep layers of the tendon, no significant difference between the layers were found in samples without tears (p=0.19). However, in samples with complete and partial tears, the deep and middle layers scored significantly higher when compared to the superficial layer (p=0.0002, table 3). A similar pattern was also seen in the enthesis, where samples with tears were found to have significantly higher degeneration scores in the middle and deep layers (table 4). In the middle and deep layers, significant differences in degeneration scores were found between samples with tears and no tears, where this was not evident in the superficial layer of the enthesis.
Table 3. Histology scores for lateral gluteus medius tendon layers
|
Superficial
|
Middle
|
Deep
|
p-value*
|
Tears
|
6.0 (2.4)
|
8.7 (2.2)
|
11.5 (3.0)
|
0.0002
|
No tears
|
2.0 (1.7)
|
3.4 (1.6)
|
4.2 (2.5)
|
0.19
|
p-value⁺
|
0.0018
|
0.00050
|
0.00031
|
|
⁺ p-value for tears vs no tears for each tendon layer
*p-value for difference in layers for samples with and without tears
Two way ANOVA - column factor (tendon layer) p=0.0002, row factor (presence of tears) p<0.0001, interaction p=0.14
Table 4. Enthesis scores for lateral gluteus medius layers
|
Superficial
|
Middle
|
Deep
|
p-value*
|
Tears
|
4.0 (2.7)
|
5.9 (2.3)
|
8.3 (3.0)
|
0.0030
|
No tears
|
2.4 (1.8)
|
2.8 (2.2)
|
3.7 (2.6)
|
0.54
|
p-value⁺
|
0.28
|
0.0029
|
0.0004
|
|
⁺ p-value for tears vs no tears for each tendon layer
*p-value for difference in layers for samples with and without tears
ANOVA - column factor (tendon layer) p=0.014, row factor (presence of tears) p<0.0001, interaction p=0.0815
Two-way ANOVA was unable to detect a significant interaction between layer examined and the presence of tears in either the tendon (interaction p=0.14) or the enthesis (interaction p=0.082).
Supero-posterior insertion of gluteus medius:
In both samples where an isolated partial tear was found in the supero-posterior insertion, minimal degenerative changes were observed; in one sample, minimal regional variation in cellularity was found in the mid-substance of the tendon. The other sample displayed mild nuclei rounding and vascular infiltration, again in the mid-substance of the tendon. The enthesis of both samples exhibited no degenerative changes.
In samples of the supero-posterior insertion where a concurrent lateral insertion tear was present, the average degeneration score was 4.9 (SD 2.6) in the tendon and 3.8 (SD 2.2) in the enthesis. In contrast, the average degeneration scores in the same samples at the lateral insertion was 9.0 (SD 1.7) in the tendon and 7.6 (SD 2.0) in the enthesis. These differences were statistically significant (p=0.012 and p=0.013 respectively). No degenerative changes were found in the supero-posterior gluteus medius when tears were not present in the lateral portion of gluteus medius.
Gluteus minimus:
In samples where a complete or partial tear were identified, the average modified Movin’s score was 13.4 and 11.1 respectively compared to 3.0 in normal samples. These differences were statistically significant (p=0.0095) (Table 5). Again, enthesis involvement was seen in all samples with tears. The average enthesis degeneration score was 10.5 in complete tears, 9.6 in partial tears and 3.3 in normal samples (p=0.021).
Table 5. Histology and enthesis scores for gluteus minimus insertion
|
Tendon (SD)
|
p-value
(post hoc)
|
Enthesis (SD)
|
p-value
(post hoc)
|
Complete Tear
|
13.4 (1.7)
|
0.012
|
10.5 (1.7)
|
0.025
|
Partial Tear
|
11.1 (2.4)
|
0.2
|
9.6 (1.1)
|
0.16
|
No Tear
|
3.0 (1.8)
|
reference
|
3.3 (1.5)
|
reference
|
p-value (ANOVA)
|
0.0095
|
|
0.021
|
|