Introduction

Currently all arthroscopic technique trends to establish itself as the gold standard for rotator cuff repair. In 2005, it was reported that AAOS surgeons performed cuff repairs with arthroscopic, mini-open, and open technique in 14.5, 46.2, and 36.6%, respectively [1]. In 2012, only 15.7% of surgeons’ respondent to ASES survey still used an open technique [2].

Only few studies compared open and arthroscopic rotator cuff repair and it remains unclear that arthroscopic is superior to open technique in term of clinical outcomes and tendon healing [3,4,5,6,7,8].

Therefore, the purpose of the study was to compare the postoperative clinical results and structural integrity of rotator cuff repairs performed either open or arthroscopically. We hypothesized that both techniques would achieve equivalent clinical outcomes and tendon repair integrity.

Methods

Study design and patients

The study was a prospective comparative cohort analysis of rotator cuff repairs between 2012 and 2013. Inclusion criteria were: patients operated on for a primary rotator cuff repair of supraspinatus and/or infraspinatus tendons tears, and reviewed with clinical examination and ultrasound evaluation with a minimum follow-up of 12 months.

Exclusion criteria were: prior surgical procedure on the shoulder, cuff tear extended to subscapularis tendon (full thickness tear), irreparable tears, and degenerative or inflammatory arthritis of the glenohumeral joint.

The institutional review board approved the research protocol (number 17-0313).

Eighty-seven patients were consecutively included and enrolled in an arthroscopic surgery (group A) or open surgery group (group O) according to the surgeon who performed the procedure (NB for arthroscopic and PM for open surgery).

Surgical technique

Patients received a preoperative interscalene block (single injection of Ropivacaine 7.5 mg/mL) and were placed in a beach-chair position under general anaesthesia in both group.

Arthroscopic technique: after intra-articular examination, a biceps tenotomy or tenodesis was done in every case because of tenosynovitis or instability. A subacromial decompression was performed systematically with a burr and the footprint on the greater tuberosity was abraded before tendon fixation. The cuff repair was performed with a single row of double-loaded threaded suture anchors placed lateral to the greater tuberosity with a tension band technique as described by Boileau et al. [5].

Open technique

Through a superior approach the anterior body of the deltoid was detached from the acromion to perform a subacromial decompression with an osteotome. The long head of the biceps was fixed to the transverse ligament in the groove, after the intra-articular portion has been resected. Then, the cuff was repaired onto the abraded greater tuberosity with trans osseous number 2 non absorbable sutures and Masson-Allen stitches. At the end of the procedure the deltoid was reattached to the acromion with trans osseous sutures.

Postoperative protocol

In both group a sling was placed for 6 weeks and passive range of motion as well as pendulum exercises were authorized from the day after the surgery. A physiotherapist controlled the rehabilitation program from the 2nd week to the 3rd month. Active motion started at 6 weeks. No heavy labour or activities were allowed before the 6 months.

Clinical assessment

All the patients were evaluated by an independent observer (XB or TTP) preoperatively, at 3 months, 6 months and 1 year after the surgery. At last follow-up, Constant score, American Shoulder and Elbow Surgeons (ASES) score and Simple Shoulder Value (SSV) were used [9, 10]. For active range of motion, elevation was measured with a goniometer (patient in a seated position) in the sagittal plane, external rotation was measured in the coronal plane with the arm by the patient’s side, and internal rotation was measured by level when the patient was asked to reach behind to the highest vertebral level.

Radiological assessment

Preoperatively, tendon retraction was staged on MRI or arthro-CT scan according to Patte’s classification and fatty degeneration according to Gouttalier [11, 12]. At follow-up, ultrasonographic evaluation was performed by two radiologists specialized in musculoskeletal disorders in a blinded double-check. Cuff repair integrity was classified into five types following Sugaya classification adapted for ultrasound [13]: type I, sufficient and homogeneous thickness; type II, sufficient, but heterogeneous thickness; type III, less than half the thickness without discontinuity; type IV, minor discontinuity; and type V, major discontinuity. In case of disagreement between the radiologists, the worst grade of healing was taken.

Statistical analysis

Descriptive statistical analysis consisted of means, maximum, minimum and standard deviations for continuous variables and percentages for discrete variables. Statistical comparisons between groups were performed with paired t tests or Wilcoxon test for continuous variables and the χ 2 or Fisher test for categorical variables. The significant difference was set at p < 0.05. All analyses were performed with the use of SAS software (SAS Institute, Cary, NC).

Results

Patients demographics

One hundred and eighteen patients were operated on for rotator cuff repair during the period of the study. Eighteen were excluded and 13 were lost before the minimum follow-up of 12 months, leaving 44 patients in group A and 43 in group O available for analysis.

As documented in Table 1, there was no statistical difference between group A and O regarding gender, trauma incidence, work compensation context, tobacco use, diabetes, dominance side involved, active range of motion, preoperative Constant score, ASES score, and SSV. However, patients of group O were older and with smaller tears in group A (Table 2).

Table 1 Preoperative analysis of group O and group A
Full size table
Table 2 Preoperative analysis of tear patterns in group O and group A
Full size table

Clinical assessment

At mean follow-up of 13 months (12–17), Constant score, ASES score and SSV significantly improved postoperatively in both groups. No significant difference was detected between the two groups regarding total Constant score, ASES score, SSV (Table 3). However, active internal rotation was significantly better in group O, as well as the mobility according the Constant score.

Table 3 Postoperative outcomes
Full size table

Tendon integrity

The rate of re-tear (stage IV and V) was 7 and 9%, respectively, for group A and O (p = 0.8) (Table 4).

Table 4 Ultrasound evaluation in both group and results according Sugaya’s classification
Full size table

Complications

One patient in each group developed capsulitis. Neither patient required further surgical intervention. No complication related to the deltoid detachment was observed in group O. No patient needed revision surgery (even in case of tendon re-tear) at the last follow-up.

Discussion

This study compared open and arthroscopic cuff repair in a prospective case series. Because arthroscopic technique became the gold standard over the past decade, the debate is to know if open surgery is still warranted or not. Therefore, we demonstrated that there was not any significant difference between both techniques, regarding functional outcomes and cuff integrity at the minimum follow-up of 12 months.

This study has some limitations. Firstly, patients were not randomly included in arthroscopic or open group. Thereby, two different surgeons performed two different technique of cuff repair. However, each of them had a wide experience in his own technique and could attest of its reproducibility. This allowed to compare properly the outcomes of both technique. Secondly, the number of patients lost of follow-up was high (13 patients). The main reason was the lack of clinical and ultrasound examination at 12 months for many of them. Because the number of patient lost was equivalent in each group (group O n = 7; group A n = 6), this bias should not impact our comparative analysis and as far as the 6th postoperative month, none of them needed a revision surgery or complained of unsatisfactory results. In addition, both group were not perfectly similar in terms of age and tear size, which could have specifically an impact on tendon healing and clinical outcomes. Finally, the follow-up for clinical and ultrasound evaluation was only 1 year. However, it was previously reported that re-tears mainly occur during the first postoperative year [14, 15].

Previously several studies compared arthroscopic and mini-open surgery, and did not find any statistical differences in outcomes measurement [16,17,18,19,20,21,22]. Only few studies compared clearly open surgery to all arthroscopic procedure. Buess et al. [6] compared clinical results after open and arthroscopic rotator cuff repair. They concluded that arthroscopic cuff repair yielded equal or better results than open. However, the open group mixed conventional technique and mini-open (12 and 18 patients, respectively). In the same way, Bishop et al. [4] constructed in their prospective study a heterogeneous cohort of 24 open and 8 mini-open surgery compared to an arthroscopic cohort. They suggested that open technique should be chosen for large tears because of a lower rate of re-tears. Walton and Murrel [8] published a large series of 200 true open and 200 arthroscopic cuff repair. They excluded lesion that exceeded 16 cm2 and did not report how they managed long head of the biceps’ pathologies. A decompression was performed systematically in open group compared to 76% in arthroscopic group, and two different configurations of single row arthroscopic repair were used. However, they concluded that arthroscopic repair was superior to open regarding operative time, cuff integrity, and speed of postoperative recovery.

In our study, the clinical evaluation at follow-up did not highlight any difference between the two groups, except a greater active internal rotation after open surgery. However, this parameter did not influence overall clinical scores. This finding was contrary to many studies which were in favour of better range of motion after arthroscopic technique [6, 8]. However, it was observed that passive range of motion converging at 2 years of follow-up between both technique [8].

One relevant information of this study is that open surgery did not increase the rate of postoperative complications. We did not identify clinical deltoid impairment despite the detachment of its proximal origin [23]. This finding confirmed the study of Cho et al. [24] who explored with MRI postoperative thickness of the deltoid after open and arthroscopic surgery and did not find out any difference between both techniques.

However, in a large series of more than 1800 cuff repairs, Vopat et al. [25] reported that open technique was more at risk of infection than arthroscopic surgery, as well as to be a male and to have a high body mass index. Owens et al. [26] confirmed a higher incidence of complication and hospital readmission after open cuff repair in a veteran population.

Conclusion

This study did not demonstrate any difference of arthroscopic over open surgery in case of rotator cuff repair regarding clinical outcome and cuff integrity at 1-year follow-up. Moreover, the rate of postoperative complication remained low in both group and this cannot be an argument to support one technique.