SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. AUTHOR CONTRIBUTIONS
  8. Acknowledgements
  9. REFERENCES

Objective

To evaluate the value (sensitivity and specificity) of 2 modified physical tests for the diagnosis of gluteal tendinopathy in patients with refractory greater trochanter pain syndrome (GTPS).

Methods

The 2 tests were prospectively evaluated by a single physician in all consecutive patients with persistent (≥4 months) GTPS and no hip joint arthropathy seen on radiography between 2002 and 2006. The 2 tests evaluated the occurrence of pain similar to spontaneous pain during a single-leg stance held for 30 seconds and resisted external derotation in a supine position (hip flexed 90°) then prone position (hip extended). A matched control population without hip pain was examined similarly. Tendinitis, tendon tear, and associated bursitis in the target group were documented by magnetic resonance imaging (MRI) in transverse, coronal, and sagittal planes, with MRI serving as the gold standard.

Results

Seventeen patients completed the study (mean ± SD age 68.1 ± 10.8 years, mean duration of symptoms 13 months). MRI revealed tendinopathy and/or bursitis of the gluteus medius and/or minimus tendons in all patients, with evidence of tearing in 15. Sensitivity and specificity were 100% and 97.3%, respectively, for the single-leg stance test and 88% and 97.3%, respectively, for the resisted external derotation test in the supine position.

Conclusion

The 30-second single-leg stance and resisted external derotation tests had very good sensitivity and specificity for the diagnosis of tendinous lesion and bursitis in patients with MRI-documented refractory GTPS.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. AUTHOR CONTRIBUTIONS
  8. Acknowledgements
  9. REFERENCES

Greater trochanter pain syndrome (GTPS) is considered to be due to tendinitis and/or bursitis involving the abductor apparatus of the hip (i.e., the gluteus medius [GMe] and gluteus minimus [GMi] tendons). The GMe tendon is composed of 2 parts: a strong main round tendon that attaches to the superoposterior facet of the greater trochanter (GT) and a thin lateral part that attaches to the lateral facet of the GT. The GMi tendon attaches to the anterior facet of the GT (1). A bursa is located under each of these tendons. The trochanteric bursa, the largest and most superficial bursa, is located beneath the gluteus maximus muscle and the iliotibial tract and covers the posterior and lateral facets of the GT and the GMe tendon (1). The locations of sub-GMe and sub-GMi bursae are indicated in Figure 1.

thumbnail image

Figure 1. A sketch of the anterolateral view of the greater trochanter (GT), which illustrates the gluteus medius (GMe) tendon insertion. a= main GMe tendon; b = lateral GMe tendon; c = gluteus minimus tendon insertion on the anterior facet of the GT; 1 = trochanteric bursa; 2 = subgluteus medius bursa; 3 = subgluteus minimus bursa. FH = femoral head.

Download figure to PowerPoint

GMe and GMi muscles act as hip abductors and internal rotators by their anterior fibers and as external rotators by their posterior fibers (2), the other external rotators being the pelvitrochanteric muscles. The term hip rotator cuff is used by analogy with the shoulder rotator cuff (3, 4). During the single-leg stance, the GMe acts as the main pelvic stabilizer, preventing the contralateral pelvis from tilting downward (Figure 2). Therefore, the GMe muscle contracts at each step.

thumbnail image

Figure 2. As weight is transferred to each leg during walking, contraction of the gluteus medius between the iliac wing (A) and attachment to the greater trochanter (B) keeps the pelvis horizontal by counterbalancing the weight of the “suspended” body (W). D = deviated weight from the median axis; R = resulting stress on hip joint.

Download figure to PowerPoint

Trochanteric tendinobursitis (TTB) is usually diagnosed on the basis of clinical signs, then imaging findings. Classic physical signs are of unequal diagnostic value, especially with a tendon tear. In the first surgical series of 7 women published in 1999, physical examination did not reveal the existence of tear (4). A tender zone over the lateral trochanter is almost always present but involves very poor specificity. Passive elongation of the involved tendon commonly results in pain. Thus, passive external rotation with the thigh flexed 90° is often the only painful maneuver during routine hip examination but only suggests TTB. The best tests are pain on resisted abduction and resisted internal rotation of the thigh. However, these tests lack sensitivity. Bird et al found sensitivities of only 72.7% and 54.5% for the resisted abduction and resisted internal rotation tests, respectively (5). Moreover, the precise technique used for physical examination is not always clearly described in the literature, although it can strongly influence the results.

In the present study, we aimed to assess the diagnostic value of 2 tests involving reproduction of spontaneous pain during abductor muscle contraction with 1) the single-leg stance held for 30 seconds and 2) resisted realignment of the externally rotated hip in well-defined postural conditions.

PATIENTS AND METHODS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. AUTHOR CONTRIBUTIONS
  8. Acknowledgements
  9. REFERENCES

Study design and population.

Once the 2 tests were finalized in clinical practice by one author (ML), we carried out a prospective study from 2003 to 2006 involving all consecutive outpatients with persistent (≥4 months) GTPS seen in the rheumatology unit of Leopold-Bellan Hospital. Magnetic resonance images were used as the gold standard to ascertain lesions of the hip abductor apparatus: tendinitis (with or without image of tendon tear) and bursitis. Indeed, our hypothesis was that the pain resulting from the physical tests is generated by tendinobursitis or an inflammatory focus close to the tendon, not by the tear itself. Physical signs were recorded before magnetic resonance imaging (MRI) in all consecutive patients meeting the inclusion criteria. Some patients received an injection of a nonfluoride corticosteroid into the bursitis or the inflammatory focus under sonographic control to identify the source of pain by induced remission.

Inclusion criteria were hip pain anterior, lateral, or posterior to the GT or involving the thigh, and physical signs of TTB: a tender zone on pressure to the superior or lateral aspect of the GT, pain in the external rotation of the thigh flexed at 90°, and a normal range of passive motion of the hip joint, apart from external rotation, that could be limited by pain. Other criteria were no evidence of hip joint disease such as osteoarthritis, avascular necrosis of the femoral head, or any other hip arthropathy seen on anteroposterior pelvis radiography, as well as pain and disability that were persistent (chronic or relapsing pain) over at least 4 months and refractory to standard management (analgesics, nonsteroidal antiinflammatory drugs, local injection of steroids, rest, physiotherapy). Patients with sciatica or crural neuralgia or with a history of hip surgery were excluded. Because this study focused on clinical tests designed to reproduce spontaneous pain, patients without pain but with a complete abductor tear syndrome (pseudoparalysis of the abductor muscles) were also excluded.

Physical examination for pain resulting from muscle contraction.

The 2 modified pain tests to be assessed are described below.

Single-leg stance held for 30 seconds.

Patients were asked to stand on the affected leg for 30 seconds and to say whether any pain occurred. Pain similar to spontaneous pain was recorded as immediate, early, or late if it occurred after 0–5 seconds, 6–15 seconds, or 16–30 seconds, respectively. No homolateral deviation of the trunk above the involved hip was allowed during this test, because such deviation can delay or prevent the onset of pain. To avoid unsteadiness during this posture, patients' hands were gently held by the examiner.

Resisted external derotation (resisted realignment of the externally rotated hip).

Patients were asked to lie supine on a table with the hip and knee flexed at 90° and the hip in external rotation (a position that is usually painful in patients with TTB). After slightly diminishing the external rotation just enough to relieve the pain (if any), patients were asked to return actively to neutral rotation, that is, to place the leg along the axis of the bed, against resistance (Figure 3). The test result was considered positive when the test reproduced spontaneous pain. If the result was negative, the test was repeated with the patient lying prone, hip extended and knee flexed at 90°.

thumbnail image

Figure 3. Resisted external derotation test. The hip is flexed 90°, and the patient is asked to return the leg to the axis of the table against resistance. The test result is positive when the usual pain is reproduced.

Download figure to PowerPoint

In addition, we assessed pain during active and resisted abduction of the thigh, with the patient lying on the contralateral side. Patients performed active abduction up to 25°, with the thigh extended, then pain was assessed during resisted abduction with the thigh successively extended, then flexed at 45° and at 90°.

Controls.

To assess specificity, 38 hips of 20 controls matched for age and sex with the TTB patients underwent the following tests: single-leg stance, resisted external derotation, resisted abduction, and search for a tender zone in the GT area. The control group comprised consecutive patients visiting the department of rheumatology for reasons other than hip, knee, ankle, or foot pain or radiculalgia of the ipsilateral lower limb.

Imaging.

Standard anteroposterior radiographic views of the pelvis and/or of the painful hip were systematically obtained. MRI was acquired using a 1.5T system, with a surface coil applied to the painful hip or a pelvic phased-array coil. T1-weighted and fat-saturated fast T2-weighted images were acquired in the transverse, sagittal, and coronal planes in 4-mm thick sections, with a 28 × 28–cm field of view and a 220 × 320–pixel matrix. Each part of the gluteal tendons (main GMe tendon, GMe lateral tendon, GMi tendon, and bursae; trochanteric bursa; sub-GMe and sub-GMi bursae) was systematically analyzed, following the criteria of Pfirrmann et al (1).

Tendinitis was defined as a thickening and/or an increased signal intensity of the tendon area seen on T2-weighted images, without tendon discontinuity or thinning. A full-thickness tear was considered complete discontinuity of tendon fibers with high signal intensity seen on T2-weighted images. A partial tear was recorded when tendon thinning was seen on T1- and T2-weighted images and was associated with high signal intensity seen on T2-weighted images. Partial and complete tears were pooled for analysis. Bursitis was defined as a collection of fluid (T2-weighted isosignal to water) or an area of high signal intensity (T2-weighted image) situated in a bursa location. Amyotrophy and/or fatty degeneration of GMe and GMi muscle were evaluated from T1-weighted images. Two musculoskeletal radiologists (VV-B and PM) interpreted the images by consensus.

Statistical analysis.

Quantitative data are presented as means, standard deviations, and ranges. Means were compared using Wilcoxon's rank sum test for 2 unmatched samples. A P value less than 0.05 was considered significant. Qualitative data are presented as numbers and percentages. Sensitivity and specificity data for the evaluated physical tests, based on the usual ratios, are expressed as percentages.

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. AUTHOR CONTRIBUTIONS
  8. Acknowledgements
  9. REFERENCES

Eighteen patients were included; 1 withdrew for personal reasons. Seventeen patients (16 women) completed the study. The main results, including MRI findings, are summarized in Table 1. Age, duration of symptoms, number of corticosteroid injections, and MRI findings did not differ significantly between the 8 patients who underwent surgery and the 9 patients whose disease was managed by conservative treatment. Mean duration of symptoms was 16.4 months and 9.8 months for the surgery and conservative-management groups, respectively. Mean ± SD body mass index was 22.1 ± 5.4 kg/m2 (range 16–34.6) (data not shown).

Table 1. Demographics, test results, and magnetic resonance imaging (MRI) findings in 17 patients with persistent tendinobursitis syndrome of the gluteus medius (GMe) and gluteus minimus (GMi) tendons*
 Surgery (n = 8)No surgery (n = 9)Total (n = 17)
  • *

    Values are the number (percentage) unless otherwise indicated. NR = not recorded.

  • Expressed per patient per year.

Age, mean ± SD (range) years71.1 ± 9.4 (55–81)65.4 ± 11.9 (46–81)68.1 ± 10.8 (46–81)
Sex, female/male8/08/116/1
Right side/left side3/56/39/8
Symptom duration, mean ± SD (range) months16.4 ± 10.4 (9–36)10 ± 10.5 (4–12)13 ± 10.5 (4–36)
No. of local corticosteroid injections, mean (range)2.1 (1–4)2.8 (2–5)2.6 (1–5)
Positive test results for pain   
 Single-leg stance8 (100)9 (100)17 (100)
  Pain within 1–5 seconds4 (50)1 (11)5 (29.5)
  Pain within 6–15 seconds2 (25)5 (55)7 (41)
  Pain within 16–30 seconds2 (25)3 (33)5 (29.5)
 Resisted external derotation7 (87.5)8 (89)15 (88)
 Resisted abduction6 (75)6 (66)12 (70.1)
MRI (surgery) findings, no.   
 Anterior GMe tendon tear8 (8)816
 GMi tendon tear2 (5)13
 Tendinitis6 (NR)915
 Bursitis8 (8)816

Clinical tests.

Pain occurred on passive external rotation (hip flexed 90°) in all 17 patients and on pressing the region of the GT in all but 1 patient. Nine patients exhibited the Trendelenburg gait. The single-leg stance reproduced immediate, early, and late spontaneous pain (see Patients and Methods section) in 5, 7, and 5 patients, respectively. Resisted external derotation produced immediate pain in all but 2 patients lying in the supine position. In 1 of these 2 patients, pain was produced in the prone position.

The control population of 20 subjects included 19 women (mean ± SD age 69.9 ± 11.2 years). Thirty-eight hips free of any spontaneous pain met the criteria for control hips. Both the single-leg stance and resisted external derotation tests gave negative results in 37 hips and positive results in 1 hip (results for the first test were positive for 1 subject and those for the latter test were positive for another subject), resulting in a specificity of 97.3% for both tests. The resisted abduction test gave negative results in 36 hips and positive results in 2; tenderness in the GT region was absent in 25 control hips and present in 13, resulting in specificities of 94.7% and 66%, respectively.

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. AUTHOR CONTRIBUTIONS
  8. Acknowledgements
  9. REFERENCES

The 2 modified clinical tests of TTB assessed in this study have very good sensitivity and specificity: 100% and 97.3%, respectively, for the single-leg stance test and 88% and 97.3%, respectively, for the resisted external derotation test in the supine position. For the latter test, the sensitivity increased to 94% with positive results in the prone position in the case of negative results in the supine position.

These high scores may have several explanations. First, the patients had refractory GTPS of long duration (mean duration 13 months, with no significant difference between groups with or without surgery). This long duration implies advanced lesions. Indeed, MRI findings demonstrated high signal intensity on T2-weighted images interpreted as tendinitis and/or bursitis in all cases, with tendon tear in 15 of 17 patients. This definite inflammatory reaction may cause pain when the tendon is actively stretched. Second, tendon stretching was optimized by the posture conditions we proposed in the tests. The single-leg stance was more sensitive when the time available for the pain to occur was long enough: only 29.5% of patients recognized their pain within 5 seconds, 41% within 5–15 seconds, and 29% within 15–30 seconds. The resisted external derotation test was enhanced by the initial position: knee and hip flexed 90°, and hip in external rotation. When producing pain, the effect is immediate, probably because the initial position for the test already maximizes the stretching of the GMe and GMi tendons. Because of this initial position, we named our maneuver resisted realignment of the externally rotated hip or, preferably, resisted external derotation, in contrast to classic maneuvers starting without any hip external rotation and usually called resisted internal rotation.

Demographic and disease characteristics of our patient population were similar to most of those found in previously published series of GTPS (4–9): marked female predominance, mean age approximately 60–70 years, and long duration of symptoms (Table 2). Bilateral involvement is rare.

Table 2. Case series of patients with persistent symptoms of trochanteric tendinopathy*
Authors (ref.), yearNo. of casesSex, F/MAge, mean (range)Symptom duration, mean (range)Findings by MRI/ ultrasonography
  • *

    MRI = magnetic resonance imaging; US = ultrasonography; GT = greater trochanter.

Kagan (4), 199975/269 (52–81)41 months (21–60)MRI: tendon enlarged or torn, bursitis
Chung et al (6), 199966/070 (62–79)A few weeks to a few monthsMRI: avulsion: 4; tear: 2
Kingzett-Taylor et al (8), 19993528/764 (40–86)A few weeks to yearsMRI: gluteus medius complete tear: 8; partial tear: 14; tendinitis: 13
Bird et al (5), 20012424/058 (36–75)12 months (2–60)MRI: gluteus medius; tear: 14; tendinitis: 9; bursitis: 2
Connell et al (7), 20035337/1657 (37–79)6.7 months (1.5–36)US: gluteus medius tear: 25; tendinitis: 28; bursitis: 8
Cvitanic et al (9), 20041513/270 (51–81)Not statedMRI: high T2-weighted signal, above or lateral to GT

The value of physical signs of TTB syndrome has rarely been studied in the literature. To our knowledge, this is the first study of the single-leg stance test maintained up to 30 seconds for assessing the occurrence and delay of pain. Such a delay was proposed for assessing delayed positive test results for the Trendelenburg sign (10). However, the Trendelenburg test response is not pain but, rather, pelvis tilting. We did not observe this tilting in our patients, probably because there was no complete tear of the GMe: we found only 1 instance in which the main GMe tendon was torn (Table 1). The resisted internal rotation maneuver was tested by Bird et al (5) in a series of 24 patients with GTPS, 11 with MRI evidence of a GMe tear. Sensitivity was 54.5% and specificity was 69.2% for these patients. However, details of implementation differed from those used in our study: the hip was flexed at only 45° and was not placed in external rotation at the onset of the maneuver; moreover, the test was performed only in the supine position. In this previous study, the Trendelenburg sign had the highest sensitivity (72.7%), and specificity was 76.9% in patients with complete or partial tears; resisted abduction sensitivity was 72.7% and specificity was 46.2%. However, a comparison with our specificity values is not possible, because the controls in the previous study were not healthy subjects: data for 24 patients, those with and without MRI evidence of a tendon tear, were compared. Moreover, the proportion of patients with a complete tear of the GMe tendon was not stated, whereas in our study the tear involved only the lateral part of the GMe tendon in 7 of 8 surgical patients. In addition to the present series, we observed 2 patients with a complete tear of the GMe (posterior and anterior sections) that was surgically confirmed. The patients were without pain and were impaired by the weakness of the abductor apparatus, which resulted in important Trendelenburg gait and early tiring on walking.

The value of MRI as a gold standard is based on the comparison of MR images versus surgical findings both in previous studies (4, 6, 7, 9) and in our study (Table 1). The role of bursitis or inflammatory focus as the source of pain is confirmed by the results of corticosteroid injections: in patients receiving injection, the pain disappeared or clearly subsided transitorily (data not shown).

Our study has several limitations. We did not assess reproducibility of the tests, but we intend to complete this step in a separate study. We do not know whether these clinical tests would show similar results in common TTB syndrome of only a few days' or weeks' duration. Moreover, another study of specificity should be performed, closer to actual practice, with control patients exhibiting hip joint disease and negative results on radiography as very early osteoarthritis, labrale lesion, or avascular necrosis of the femoral head. In fact, difficulties would arise in patients with both tendon tear and hip osteoarthritis, the association not being rare and often noticed during surgery for hip arthroplasty (11, 12).

In conclusion, our study presents the first step of the validation of 2 modified physical tests of TTB, recorded prospectively before MRI or sonography. Positive test results allow for immediately choosing the most appropriate imaging process: MRI of the painful side using a coil to obtain detailed small views. The availability of new clinical tests is not unnecessary: the shoulder rotator cuff examination is rich in the range of tests for diagnosis. Likewise, when clinicians are confronted with a possible hip rotator cuff problem, a range of tests would allow for choosing the most relevant. A specific strategy of clinical and imaging diagnosis steps could allow for recognizing and treating hip rotator cuff tendinobursitis and tears much earlier than after several months or years.

AUTHOR CONTRIBUTIONS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. AUTHOR CONTRIBUTIONS
  8. Acknowledgements
  9. REFERENCES

Dr. Lequesne had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study design. Lequesne, Vuillemin-Bodaghi, Bard.

Acquisition of data. Lequesne, Mathieu, Vuillemin-Bodaghi, Djian.

Analysis and interpretation of data. Lequesne, Mathieu, Vuillemin-Bodaghi, Bard.

Manuscript preparation. Lequesne.

Statistical analysis. Lequesne.

Acknowledgements

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. AUTHOR CONTRIBUTIONS
  8. Acknowledgements
  9. REFERENCES

We thank Mrs. Squires Lind for her assistance in the English translation and Drs. L. Bellaiche and E. Maheu for their comments and suggestions for this article.

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. AUTHOR CONTRIBUTIONS
  8. Acknowledgements
  9. REFERENCES