Outcomes of footprint medialization and bone marrow stimulation in chronic retracted rotator cuff tears
Article information
Abstract
Background
This study compared the clinical and radiological outcomes of chronic retracted rotator cuff tears where complete footprint coverage cannot be achieved, using two treatments: footprint medialization with bone marrow stimulation and conventional incomplete repair.
Methods
This retrospective study included 87 patients who underwent arthroscopic rotator cuff repair with incomplete footprint coverage. The included patients were divided into group 1 (54 patients with footprint medialization and bone marrow stimulation) and group 2 (33 patients with conventional repair). Medialization and bone marrow stimulation were performed if the tendon did not cover a footprint of ≥1 cm2. Clinical outcomes, including a visual analog scale for pain score, range of motion, American Shoulder and Elbow Surgeons score, Constant-Murley score, and patient satisfaction, were evaluated preoperatively and at follow-up. Radiological healing was assessed using magnetic resonance imaging or computed tomography arthrography at 6 months and ultrasound at 2 years.
Results
Both groups showed significant improvements in clinical outcomes from preoperative levels. Group 1 demonstrated better outcomes in all measures at the final follow-up than group 2. Active range of motion improved significantly in both groups, with no significant postoperative differences. At 2 years postoperatively, group 1 had a significantly lower retear rate (14.8%) than group 2 (36.4%) (P=0.020).
Conclusions
In this study, the group that underwent footprint medialization and bone marrow stimulation for chronic retracted rotator cuff tears, in which complete footprint coverage was not possible, exhibited a lower retear rate and better clinical outcomes than the group that underwent conventional incomplete repair, with a minimum follow-up period of 2 years.
Level of evidence
III.
INTRODUCTION
The best surgical treatment method for a rotator cuff tear is to completely cover the anatomical footprint of the tendon to promote bone-tendon healing. However, in cases of chronic degenerative tears with associated rotator cuff atrophy and fatty degeneration, healing may be challenging, and high retear rates have been reported [1-3]. With advancements in surgical techniques, various methods have been developed to reduce tension on torn tendons and restore the anatomical footprint effectively. These include techniques such as adhesion release on both the articular and bursal sides, margin convergence, and the interval slide [4,5]. However, when complete coverage of the footprint cannot be achieved using these techniques, incomplete repair can be performed [6].
When it is challenging to position an atrophied tendon at its normal anatomical attachment site, or when excessive tension is placed on the tendon, a technique involving the removal of a portion of the joint cartilage and medialization of the footprint can be employed. This approach reduces tension on the torn tendon and recreates the footprint, increasing the tendon-to-bone contact area, which may enhance the potential for healing [7]. Reports have expressed concerns that footprint medialization can result in a shorter moment arm in the supraspinatus muscle, leading to limited mobility [8]. However, reports have suggested that this technique can be used to attach a torn tendon to the greater tuberosity when conventional repair methods are not feasible. Additionally, this approach has been shown to reduce the retear rate while achieving relatively good clinical outcomes [9-11].
One method known to promote tendon healing involves bone marrow stimulation of the greater tuberosity footprint. This technique uses mesenchymal stromal cells from the bone marrow to regulate growth factors and cytokines, thereby enhancing the rate of bone-tendon healing [12]. Several studies have reported that incorporating bone marrow stimulation during rotator cuff repair can reduce the retear rate and lead to favorable clinical outcomes [13,14]. Moreover, there are reports suggesting that, even when the rotator cuff is not fully repaired, applying bone marrow stimulation to the footprint can promote the formation of scar tissue composed of type II collagen in the uncovered area, thereby improving the overall healing rate of rotator cuff repair [15]. Footprint medialization and bone marrow stimulation are efficient and cost-effective techniques. Medialization repositions the torn tendon to the articular margin, reduces tension, and increases footprint coverage, while bone marrow stimulation promotes healing and enhances collagen integration over a larger area. Together, these methods may lower retear rates and improve clinical outcomes.
In this study, we aimed to compare the clinical and radiological outcomes of chronic retracted rotator cuff tears in which complete footprint coverage is not achievable between two treatment approaches: footprint medialization with bone marrow stimulation and conventional incomplete repair techniques. We hypothesized that combining footprint medialization and bone marrow stimulation would enhance the healing rate of the rotator cuff and result in superior clinical outcomes compared to conventional incomplete repair methods.
METHODS
This study was approved by the National Health Insurance Medical Center Institutional Review Board (No. NHIMC 2021-03-023), and informed consent was obtained from each patient. This study was carried out in accordance with the ethical standards in the 1964 Declaration of Helsinki and relevant regulations of the U.S. Health Insurance Portability and Accountability Act.
Patient Selection
We analyzed the medical records of patients who underwent arthroscopic rotator cuff repair between March 2013 and December 2019 in whom complete coverage of the footprint of the supraspinatus tendon was not achieved. Incomplete repair was defined as a space of ≥1 cm² remaining in the footprint despite adequate soft tissue release. We excluded patients (1) who underwent only partial repair for an irreparable rotator cuff tear; (2) who underwent superior capsular reconstruction or patch augmentation; (3) with a follow-up period of <2 years; (4) in whom healing of the repaired tendon was not confirmed using magnetic resonance imaging (MRI) or computed tomography (CT) arthrography after 6 months, as well as those who did not undergo ultrasound examination follow-up at 2 years postoperatively; (5) with grade 3 or higher joint arthritis according to the Hamada classification; (6) who underwent concurrent subscapularis tendon repair; and (7) who had a history of surgical treatment of the ipsilateral shoulder, respectively. A total of 87 patients were finally analyzed after applying the inclusion and exclusion criteria. Footprint medialization and bone marrow stimulation were performed from August 2015; therefore, 54 patients who underwent surgery between February 2016 and December 2019 were classified into group 1 (footprint medialization and bone marrow stimulation group), while the remaining 33 patients were classified into group 2 (conventional repair group).
Surgical Techniques for Bone Marrow Stimulation and Medialization
All patients underwent arthroscopic rotator cuff repair under general anesthesia in the lateral decubitus position, which was performed by a single surgeon (SHK). The procedure began with the use of standard posterior and anterior portals for careful examination of the glenohumeral joint. For partial tears of the subscapularis tendon, debridement was performed using a radiofrequency device and a shaver. A tenotomy or tenodesis was performed for lesions involving the bicep tendon. Using the posterolateral portal for visualization, rotator cuff tears on the bursal surface were assessed, and the medial and lateral aspects of the torn tendon were mobilized through subacromial decompression using a shaver and a radiofrequency device. The coracohumeral ligament was released in all patients, and margin convergence was performed for U-shaped and L-shaped tears. Medialization was performed if the torn tendon did not cover the greater tuberosity by ≥1 cm2. Approximately 5–10 mm of articular cartilage was removed with a ring curette, with subsequent removal of the remaining soft tissue from the entire greater tuberosity performed using a radiofrequency device. The remaining cortical bone at the rotator cuff attachment site was abraded with a burr, and bone marrow stimulation was performed using a microfracture pick with a spacing of approximately 5 mm. Finally, rotator cuff repair was performed using the suture bridge technique with 2–3 double-loaded suture anchors (Healix Advanced BR and Healix Advanced Knotless; Depuy Mitek). The size of the uncovered attachment site after the repair was measured and recorded using a measurement probe.
Postoperative Rehabilitation
Postoperative rehabilitation was carried out identically in both groups. All patients underwent shoulder brace immobilization for 6 weeks after surgery and were not allowed to perform pendulum exercises. After 6 weeks, the brace was removed, and active-assisted range-of-motion exercises were started. After confirming a recovery of motion at 3 months postoperatively, isotonic strengthening exercises using elastic bands were initiated. Patients were gradually encouraged to start exercises and activities at the same as before surgery beginning 6 months after surgery.
Clinical and Radiological Assessments
The medical records of the 87 included patients were reviewed to evaluate demographic information, including sex, age at the time of surgery, body mass index, duration of symptoms before surgery, dominant side, and total follow-up period. Clinical outcomes were assessed by measuring pain (visual analog scale [VAS] for pain score, 0–10 points) and active range of motion of the shoulder joint before surgery and at each follow-up visit. Functional outcomes were evaluated using the American Shoulder and Elbow Surgeons (ASES) score, Constant-Murley (Constant) score, and patient satisfaction score (0–10 points) at the final follow-up visit. The active range of motion was measured by assessing forward flexion with the scapula stabilized, external rotation with the elbow fixed to the side, and internal rotation based on the highest point on the spine that the patient’s thumb could reach. For statistical analysis, these reference points were assigned consecutive numbers: the first thoracic vertebra to the 12th thoracic vertebra was numbered 1–12, the first lumbar vertebra to the fifth lumbar vertebra was numbered 13–17, and the coccyx was assigned the number 18. To ensure objectivity, all preoperative and postoperative parameters were recorded by a physician’s assistant who was not involved in the study.
All patients underwent preoperative MRI to assess the tear size, retracted levels, fatty infiltration, and muscle atrophy of rotator cuff muscles. The anteroposterior (AP) tear size was estimated as the longest distance between the anterior and posterior margins of the tear on oblique sagittal images. The mediolateral (ML) tear size was estimated as the longest distance between the most medial margin of the retracted cuff and the greater tuberosity on the coronal images. The degree of atrophy and fatty infiltration in both the supraspinatus and infraspinatus muscles was assessed on sagittal images [16,17]. Six months after surgery, the healing of the repaired tendon was evaluated using MRI or CT arthrography, and the state of the repaired tendon was observed using ultrasonography 2 years after surgery. In this study, a rotator cuff retear was defined as an occurrence when complete discontinuity was observed. All preoperative and postoperative radiological evaluations were conducted by an independent orthopedic surgeon to ensure unbiased assessment.
Statistical Analysis
For the analysis of demographic, clinical, and radiological data between the two groups, a two-sample t-test was performed for continuous variables, and a chi-square test was used for categorical variables. All statistical analyses were performed using IBM SPSS ver. 23 (IBM Corp), and the statistical significance level was set at P<0.05.
RESULTS
Demographic Data
Of the 1,058 patients who underwent arthroscopic rotator cuff repair for symptomatic rotator cuff tears during the study period, 168 patients were confirmed to have achieved complete footprint coverage. Of these, 116 patients had incomplete repairs with an uncovered footprint area of ≥1 cm² after repair. Fifteen patients who underwent subscapularis repair were excluded. Six patients did not undergo CT arthrography or MRI during the follow-up period, and eight patients were lost to follow-up. Finally, 87 patients who met the inclusion criteria were included in the study.
Among these 87 patients, 54 in group 1 underwent footprint medialization and bone marrow stimulation, with a mean age of 65.1±9.5 years (range, 49–81). Of these, 23 were male and 31 were female. In group 2, 33 patients underwent conventional incomplete repair, with a mean age of 63.2±8.8 years (range, 51–80), including 15 men and 18 women. The dominant arm was operated on in 39 patients (72.2%) in group 1 and 25 patients (75.8%) in group 2. The time from symptom onset to surgery was 12.2±13.7 months in group 1 and 11.9±13.2 months in group 2, with final follow-up periods of 27.1±5.3 months and 30.1±8.8 months, respectively. On the preoperative MRI, the tear size in group 1 was AP 24.7±4.8 mm and ML 30.2±4.1 mm, while, in group 2, it was AP 26.4±5.9 mm and ML 31.8±3.1 mm. The degree of fatty degeneration on preoperative MRI was 2.7±0.7 in group 1 and 2.7±0.8 in group 2 for the supraspinatus muscle and 1.7±0.6 in group 1 and 1.6±0.7 in group 2 for the infraspinatus muscle, respectively. No significant differences in demographic or radiological data were observed between the two groups (Table 1).
Patient demographics
Functional and Radiological Outcomes
Both groups showed statistically significant improvements in clinical outcomes, including VAS for pain, ASES, Constant scores, and satisfaction scores, compared to preoperative levels. In group 1, the VAS for pain score significantly improved from 5.3±1.1 points preoperatively to 0.7±0.9 points at the final follow-up, while in group 2, it improved from 5.6±1.0 points to 1.7±1.4 points (P<0.001). Additionally, group 1 had significantly lower pain scores at the final follow-up (P<0.001). ASES scores in group 1 improved significantly from 53.0±11.0 points preoperatively to 89.9±6.3 points at the final follow-up, compared to an improvement from 51.9±8.8 points to 83.4±11.6 points in group 2 (P<0.001). Group 1 demonstrated significantly better results at the final follow-up (P=0.001). Regarding Constant scores, those in group 1 improved from 49.1±11.2 points preoperatively to 87.2±7.8 points at the final follow-up, while those in group 2 improved from 46.3±11.1 points to 79.8±8.4 points (P<0.001). Group 1 showed significantly better results at the final follow-up (P<0.001). Satisfaction scores (range, 0–10 points) increased, with those in group 1 improving from 3.6±1.0 points preoperatively to 8.5±1.7 points at the final follow-up and those in group 2 improving from 3.3±1.2 points to 7.8±1.2 points (P<0.001). Group 1 had significantly higher satisfaction scores at the final follow-up (P=0.032) (Table 2).
Clinical scores for both groups
Active forward elevation improved significantly in both groups: group 1 showed an improvement from 131.9°±13.6° preoperatively to 160.9°±11.6° at the final follow-up, while group 2 showed an improvement from 132.2°±12.8° to 160.1°±6.9° (P<0.001). External rotation improved significantly in both groups, with results in group 1 improving from 52.6°±10.1° preoperatively to 60.1°±6.9° at the final follow-up and those in group 2 improving from 52.4°±11.5° to 59.1°±9.9° (P<0.001). Internal rotation also improved significantly in both groups, with that in group 1 improving from 12.3°±2.8° to 9.6°±2.1° and that in group 2 improving from 12.0°±2.7° to 10.2°±2.7° (P<0.001). No statistically significant difference in the active range of motion was observed between the two groups after surgery (Table 3).
Active range of motion in both groups
Six months postoperatively, the healing status of the rotator cuff was confirmed using MRI or CT arthrography in all patients. In group 1, which underwent footprint medialization and bone marrow stimulation, 8 of 54 patients (14.8%) did not heal, whereas, in group 2, 11 of 33 patients (33.3%) did not heal. A statistically significant difference was observed in the retear rate between the two groups (P=0.043). Two years after surgery, ultrasound was used to confirm the presence of a retear; in group 1, eight patients (14.8%) maintained an unchanged retear rate from 6 months, while, in group 2, one more patient had no healing, resulting in 12 patients (36.4%) with no healing. A statistically significant difference was observed in the 2-year retear rate between the two groups, with group 1 showing a lower rate (P=0.020) (Table 4).
Retear rate in both groups
DISCUSSION
In this study, both footprint medialization with bone marrow stimulation and conventional incomplete repair groups showed clinical improvement after rotator cuff repair; however, the group that underwent footprint medialization and bone marrow stimulation demonstrated superior clinical outcomes, as indicated by the ASES, Constant, and satisfaction scores. Additionally, the radiological healing rate was higher in the footprint medialization and bone marrow stimulation group at both 6 months and 2 years postoperatively, supporting our initial hypothesis.
Bone marrow stimulation in rotator cuff repair facilitates tendon healing using mesenchymal stromal cells and various growth factors and cytokines contained in the bone marrow of the humeral head [12,18,19]. Jo et al. [13] reported a 2-year follow-up study on clinical outcomes and retear rates in a group of patients who underwent bone marrow stimulation during rotator cuff repair compared to those who did not, with a total of 124 patients included. Ultimately, they reported no significant differences in the clinical outcomes between the two groups; however, the retear rate was significantly lower in the bone marrow stimulation group than that in the control group (22.2% vs. 45.2%). They concluded that bone marrow stimulation could improve the structural stability of rotator cuff repair. Recent meta-analyses by Ajrawat et al. [19] and Li and Zhang [20] reported no significant difference in clinical outcomes between patients who received bone marrow stimulation and those who underwent conventional rotator cuff repair. Kim et al. [15] reported no significant differences in the clinical outcomes and overall retear rates between groups that underwent intraoperative bone marrow stimulation and those that did not during a 2-year follow-up study. In their study, single-row repair was performed on incomplete tears that could not be fully covered by attaching the tendon to the humeral head; they speculated that bone marrow stimulation in areas where the rotator cuff was not attached would not contribute to the healing process. They suggested that if the bone marrow stimulation area is sufficiently covered by the suture bridge repair technique, healing could be affected. To promote the healing of rotator cuff tears, the area of contact between the tendon and bone and the pressure exerted on this area are important [21-23]. When bone marrow stimulation is performed, if the sutured rotator cuff sufficiently covers the bone marrow stimulation site, it can assist in healing using mesenchymal stromal cells and growth factors. In this study, we included only cases in which the rotator cuff was sufficiently released to allow for incomplete repair, even if a defect remained in the footprint. Cases where the supraspinatus tendon could not be repaired and only partial repair was possible were excluded. Therefore, suture bridge repair was performed in all patients to ensure the maximum contact area and sufficient pressure on the repaired tendon. Although this study included only cases with incomplete repair, where the rotator cuff attachment site was not fully covered, all patients in group 1 underwent approximately 5–10 mm of medialization to place a medial anchor. This was followed by suturing using the suture bridge technique, which allows the bone marrow stimulation site to be covered by the rotator cuff, thereby significantly reducing the retear rate.
Medialization of the supraspinatus footprint can enable bone–tendon attachment in patients in whom reattachment to the original site is not feasible because of chronic degenerative rotator cuff tears. This technique is particularly useful in reducing excessive tension in sutured tendons. Similar to margin convergence, it facilitates attachment of the tendon to the bone and is particularly beneficial in challenging cases [9,24,25]. Biomechanically, the medialization of the footprint can result in a decrease in the moment arm or a change in the joint reaction force. In cadaveric studies, Yamamoto et al. [26] reported that a medialization of ≥10 mm can cause significant limitations in the range of motion. Recent biomechanical studies have reported similar results, stating that a medialization of 5 or 10 mm does not result in a significant change in the moment arm [27]. These studies suggest that a 5–10-mm medialization can effectively reduce tension on the repaired rotator cuff without significant complications and increase the contact area at the bone–tendon interface. Both Kim et al. [9] and Zampeli et al. [11] reported good clinical outcomes and lower retear rates following footprint medialization procedures in cases of rotator cuff tears that could not be repaired using conventional techniques. Lee et al. [25] compared the outcomes of patients with chronic degenerative rotator cuff tears who underwent rotator cuff repair with and without medialization and reported no difference in clinical outcomes; however, the retear rate was significantly lower in the group that underwent medialization procedures (8.3% vs. 30.6%). In this study, medialization of approximately 5–10 mm was performed even when the rotator cuff attachment site could not be covered by approximately 10 mm despite adequate soft tissue release. Through medialization, a sufficiently wide tendon-to-bone contact area was created to promote healing in the contact area through bone marrow stimulation. We believe that this approach achieved sufficient contact with reduced tension compared to that in the control group and that bone marrow stimulation facilitated tendon healing.
In several comparative studies on bone marrow stimulation techniques, no significant differences were observed in the final clinical outcomes. However, many studies have reported a reduced retear rate using this technique. Consistent with these findings, our study observed a decrease in the retear rate. Additionally, the clinically evaluated results, including VAS pain, ASES, and Constant scores, were significantly better in the group that underwent bone marrow stimulation and medialization than the control group. In our study, medialization followed by bone marrow stimulation to enhance tendon healing was performed in all patients in group 1. This approach resulted in a significant reduction in the retear rate compared to that in the control group. This difference in the clinical outcomes may be attributed to the combined effects of medialization and bone marrow stimulation. Although no difference was observed in active range of motion between the two groups after ≥2 years of follow-up, group 2 showed poorer pain-related measures.
This study had several limitations. First, it was a retrospective study and lacked randomization between the two surgical methods, which may have introduced a selection bias. Procedures involving bone marrow stimulation and medialization were performed later, potentially affecting comparability. Second, the study included a relatively small number of patients. This is due to the specific inclusion criteria: only patients with chronic degenerative rotator cuff tears who could not achieve complete coverage of the supraspinatus tendon footprint by ≥1 cm2 despite sufficient soft tissue release were included. This limitation made it challenging to include a larger cohort of patients undergoing complete rotator cuff repair, which may have affected the statistical power of the results. Finally, the follow-up period for both groups was relatively short, which may have affected the long-term assessment of clinical outcomes and retear rates.
CONCLUSIONS
The group that underwent footprint medialization and bone marrow stimulation for chronic retracted rotator cuff tears, in which complete footprint coverage was not possible, exhibited a lower retear rate and better clinical outcomes than the group that underwent conventional incomplete repair, with a minimum follow-up period of 2 years.
Notes
Author contributions
Conceptualization: SHK. Data curation: HBK, SHK. Investigation: SHK. Methodology: SHK. Project administration: SHK. Supervision: SHK. Writing – original draft: SHK. Writing – review & editing: HBK. All authors read and agreed to the published version of the manuscript.
Conflict of interest
None.
Funding
None.
Data availability
Contact the corresponding author for data availability.
Acknowledgments
None.