This retrospective study evaluates the clinical outcome after open repair of the gluteus medius tendon using the single-row suture anchor technique. In addition, clinical outcomes were compared between subjects with a gluteus medius tear without prior surgery and after hip arthroplasty, as well as considering preoperative muscle status and tear pattern. All subjects in this study reported persistence of symptoms for more than 6 months. Until then, conservative treatments did not provide any relief of the symptoms.
Inclusion criteria were present for Trendelenburg limp, in which the contralateral side of the pelvis tilts downward during stance and/or tenderness over the gluteal insertion. The diagnosis was confirmed by partial- or full-thickness tears. In addition, a local anesthetic was injected into the trochanteric area to prove that the cause of the pain was the tendon lesion.
Exclusion from the study existed only in the event of patient refusal to participate or in the absence of contact information.
Using medical history collected from the hospital information system with preoperative detailed physical examination findings and demographic data (eg. age, sex, laterality, date of surgery, and presence of total hip arthroplasty) were evaluated. The preoperative quantification of the tendon lesion was performed using the MRIs of the pelvis from the clinic’s internal Picture Archieving and Communications System. Tear pattern was classified into a full thickness lesion or a partial-thickness tear. The muscle condition was assessed in comparison to the opposite side, taking into account fatty degeneration (classified according to the Goutallier-Fuchs [11]) as well as tendon retraction (yes/no). When images were no longer available, this information was taken from the preoperative and intraoperative medical reports.
In the study subjects, six cases had a full-thickness tear pattern, three of which had retraction of the tendon. In 37 cases, a partial-thickness tear pattern was present. There was muscle atrophy in nine cases, of which three had 1° and one had 2° fatty infiltration. In none of the cases was a 3° or 4° fatty infiltration found. Furthermore, in five cases there was a low-grade calcification in the region of the tendon insertion.
Using patient self-administered questionnaires, the following variables were retrospectively reviewed for preoperative and current status:
-
Visual analog scale for pain (0 is considered no pain - 10 is considered worst pain; VAS)
-
Subjectively assessed gait safety with evaluation based on VAS (0 = absolutely safe gait; 10 independent gait not possible; VGS)
-
Subjective Hip Value (0–100%; SHV) [18]
-
Modified Harris Hip Score (mHHS) [12]
-
Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) in German version [23]. After adjustment, 100 points are achievable, representing the worst possible outcome.
Current status:
-
Detection of previous hip arthroplasty surgeries.
-
Functional status of the affected gluteal musculature evaluating the presence of a Trendelenburg sign (y/n), possibility of a safe one-legged stance (y/n), and abduction ability of the hip against gravity (not possible, 1–10 times, > 10 times).
This study was approved by the institutional review board (No. 3660–2017).
Surgical technique
All subjects were operated on by a senior surgeon. The tendon was repaired using a single-row suture anchor technique with open surgical approach [19]. The patient was positioned in the lateral position. The leg was positioned on a TRIMANO FORTIS® Support Arm (Arthrex Inc., Naples, FL) and placed in neutral position of the hip joint. A 4–5 cm incision in the midline from the trochanter was made, the IT band was presented and dissected along the fibers. In the presence of trochanteric bursitis, it was excised while protecting the tendon. The tendon was inspected and the rupture identified, which could be optimally performed through the open approach. In case of adhesions, they were debrided. To visualize articular tendon defects, the tendon was split along the fibers in the midline. This was followed by mobilization of all tendon parts and debrading by removal of adhesions until the tendon could be placed in the anatomical position at the footprint on the greater trochanter, in neutral position of the hip joint, under a low residual tension. For repair, the limb was positioned in 20°- 25° abduction using the TRIMANO FORTIS® Support Arm. The entry points for the anchors were prepared with a 4.5 mm punch and tap for Corkscrew anchors (Arthrex Inc., Naples, FL). The single-row suture anchor technique on the gluteus medius tendon was performed with three 4.5-mm Bio-Composite Corkscrew anchors (Arthrex Inc., Naples, FL) considering the posterior and anterior border of the lateral facet of the footprint (Fig. 1). The repair was tested in neutral position of the hip joint, following in 90° flexion and 20° extension to ensure stability. Afterwards, the wound was irrigated with normal saline. Finally, the IT band was closed by single sutures, followed by subcutaneous sutures and skin clamps.
Post treatment
All patients received the same postoperative treatment regimen. Partial weight-bearing with plantar to floor contact without additional weight-bearing was recommended for 6 weeks. Adduction above neutral and active abduction were not allowed for 8 weeks. Subsequently, pain-adapted weight-bearing was allowed.
Statistical analysis
Data were evaluated retrospectively. The statistical analysis was performed using IBM SPSS Statistics for Windows (Version 24; IBM Corp., Armonk, NY). The chosen significance level for all variables examined was p < 0.05. For analysis of the independent ordinal variables (Visual analogue scale, Paris Hip Score and WMOMAC Score) between the two groups (with and without hip arthroplasty) the Mann- Whitney- U- Test was used. Examining the categorial variables (such as physiotherapy, necessity of further surgery, positive trendelenburg sign, etc.) contingency tables were used. For investigation of changes from preoperatively to postoperatively of the depended ordinal variables the Wilcoxon test was used.