Comparison of the clinical and radiological outcomes of arthroscopic transosseous and transosseous-equivalent double-row rotator cuff repair techniques

Abstract

Objective

This study compared the clinical and radiological results of the arthroscopic transosseous (ATO) and transosseous-equivalent (TOE) double-row rotator cuff repair techniques.

Methods

Prospective data collected from patients treated with ATO (32 women and 7 men, mean age: 57.03±6.39 years) and TOE (36 women and 8 men; mean age: 57.86±7.81 years) techniques were retrospectively evaluated. The visual analog scale score, Constant score, and Oxford shoulder score were used to assess the clinical results. Anchor pullout on standard anteroposterior shoulder radiographs and rotator cuff re-tear on magnetic resonance images were examined at the final follow-up to evaluate the radiological results. Rotator cuff re-tears were graded as per the classification system described by Sugaya et al.

Results

The mean follow-up duration was 33.3±11.8 months. No difference was observed in the demographic data of the two groups. Significant improvement was observed in the postoperative shoulder scores of the groups; however, no difference was observed between the groups. Re-tear was detected in 10 patients of the TOE group and 9 patients of the ATO group. Age, tear size, and retraction level could cause re-tear.

Conclusion

In the treatment of rotator cuff tears, the ATO and TOE techniques may achieve considerable improvements in shoulder functions in the short term.

Level of Evidence

Level III, Therapeutic study

Introduction

Rotator cuff repair techniques have been developed since many years (1). The open transosseous repair technique was first defined by McLaughlin (2). With the development of arthroscopy and implants, mini open, arthroscopic single-row (SR) and double-row (DR) repair methods, arthroscopic transosseous-equivalent (TOE) suture bridge technique, and arthroscopic transosseous technique (ATO) with or without implants have been defined over the years (3–5). These changes were aimed at achieving better bone tendon healing and better functional results. Several biomechanical studies have indicated that the DR technique is superior to the SR technique because of enhanced footprint coverage, better construct stability with higher ultimate tensile load, and decreased gap formation (6–10). Moreover, compared with the DR technique, the TOE technique is a stronger method and achieves superior restoration of the footprint (11). Although some studies have revealed contradictory results, no clinically meaningful difference has been determined between the two techniques in level-1 randomized studies (12, 13).

The transosseous technique has become popular again owing to weak bone stock at the footprint, increasing pullout rates of anchors, remaining anchors at the footprint making revision more difficult in re-tear cases, high costs of using multiple anchors, and the development of devices that allow arthroscopic tunnel opening and suture passing (14–19). The ATO technique can be used with or without implants. In particular, the suture used in the technique without implant can cause bone laceration. An implant has been developed to prevent this complication and improve repair durability (1). Sharc-FT (NSC Lab, Carpi, Italy) allows the use of the ATO technique, and some devices help open a transosseous tunnel. Accordingly, medial and lateral row repair, similar to the TOE technique, as well as suture bridge can be performed. Although one study has reported on the results of the ATO technique with implant, to our knowledge, no comparative study exists.

This study aimed to compare the clinical and radiological results of the TOE and ATO techniques with implant in the short term.

Materials and Methods

Prospective data collected from patients treated with rotator cuff repairs using the ATO and TOE techniques between January 2013 and October 2017 were retrospectively evaluated. The operations were performed by two experienced surgeons (AF and MU). TOE and ATO implant repair techniques were used in 142 patients at our clinic for rotator cuff tear diagnosis. The patients were divided into TOE and ATO groups depending on the technique used.

The inclusion criteria were as follows: patient aged 18–65 years, according to Cofield (20) classification, those with medium, full thickness, isolated supraspinatus, or supraspinatus with infraspinatus extension tear, arthroscopic biceps tenotomy, or tenodesis and acromioplasty. Furthermore, patients with subscapularis tear; SLAP lesion, excluding type 1 and 2; SLAP repair; teres minor tear; irreparable tear; grade 3 fatty infiltration; frozen shoulder; arthrosis; osteonecrosis; osteomyelitis; and a history of shoulder surgery were excluded. A total of 39 ATO and 44 TOE patients were eventually evaluated on the basis of these criteria. Informed consent and institutional review board approval were obtained for the study.

Surgical techniques

All the surgeries were performed under general anesthesia and in the beach chair position. First, diagnostic arthroscopy was performed; thereafter, entry was made to the subacromial space. Bursectomy was performed, and a rotator cuff tear was revealed. The arthroscope was moved to the lateral portal, and the tear type, size, and retraction degree were identified (Figure 1a, b.). When necessary, the rotator cuff was released from the supraglenoid area and rotator interval. The footprint and tear ends were refreshed using a shaver blade and punch. When required, biceps tenodesis with anchor or biceps tenotomy was performed for patients with biceps structural lesion, instability, or subluxation.

Figure 1. a, b.

(a) Preoperative magnetic resonance imaging and (b) arthroscopic view of a patient with supraspinatus tear in the TOE group

In terms of repair with the TOE technique, two standard pieces of metallic 5-mm anchors (each loaded with two pieces of reinforced suture) were placed (Twinfix, Smith & Nephew) at the medial row. With the use of various suture passer devices, the sutures were passed through the rotator cuff and tied. The suture ends at the medial row were loaded to the 5.5-mm knotless anchor (Footprint, Smith & Nephew) and placed at the lateral of tuberculum majus to create a lateral row and suture bridge. A single anchor was used for the lateral order tears of 1 and 2 cm, and two knotless anchors were used for tears sized 3 cm. One from each suture of the knots at the medial row with four ends was loaded to the single knotless anchor of the lateral row, and the other empty sutures were cut. In cases where double knotless anchors were used in the lateral row, all the suture ends knotted at the medial row were loaded and placed to the knotless anchor (Figure 2).

Figure 2.

Intraoperative arthroscopic view of a patient in the TOE group after repair

In terms of the ATO technique, an implant entrance was opened at a distance of 18–20 mm distal to the lateral tubercular corner, with an alignment of the middle of the tear side. Then, a transosseous tunnel was opened using a tool loaded with a carrier suture at its end. The exit location was close to the cartilage border. Two reinforced sutures were loaded to the transosseous implant (Sharc-FT, NCS lab, Carpi, Italy). Four ends of these sutures were passed through the tunnel via the carrier suture (Figure 3). These sutures were drawn from the anterior portal, and the implant was placed on the cortex. Using suture passer devices, the sutures coming from the tunnel were passed through the rotator cuff and tied. One end of each suture tied at the medial row was passed through the hole behind the implant and tied with the other end. Thus, the lateral row repair and suture bridge were completed (Figure 4). Minimal anteroinferior acromioplasty was performed, and coracoacromial ligament was protected in both the techniques.

Figure 3.

Intraoperative arthroscopic view of a patient with a supraspinatus tear in the ATO group

Figure 4.

Intraoperative view of a repaired rotator cuff using the ATO technique

Rehabilitation

A standard program was applied for both the groups. A shoulder sling with abduction pillow was used by the patients for four weeks postoperatively. Elbow, wrist, and hand exercises were started on the first day. Passive range of movement (ROM) exercises were started in the 3^rd week after the surgery. Active-assisted exercises were started after the 4^th week, and active exercises were started after the 6^th week. Strengthening exercises were started after the 8^th week. The patients were allowed to return to sports and perform high-level activities after the 6^th month.

Outcome assessment

Clinical and radiological assessments were performed by two surgeons experienced in the field of shoulder surgery (AO, OT); all decisions were based on a joint consensus. Based on the patient’s medical records as well as radiological (magnetic resonance imaging: MRI) and surgery records, the tear type, size, retraction level, fatty degeneration degree on MRI, intra-joint additional pathology (SLAP1 and 2) existence, additional procedures (biceps tenotomy or tenodesis, acromioplasty), and complications were evaluated and recorded. The tear type was determined according to the Ellman and Garstman classification (21). The tear size was determined as the value measured by the probe during the surgery. Preoperative coronal magnetic resonance (MR) cross-sections and the surgery records were evaluated for retraction level, and evaluation was made according to Patte classification (22). The Thomazeau method was used for fatty degeneration in T1 sagittal MR cross-sections (23).

For clinical assessment, the visual analog scale (VAS), constant, and Oxford shoulder scores were evaluated based on the preoperative and last follow-up records of the patients.

For radiological assessment, anchor pullout was examined on standard anteroposterior shoulder radiography (Figure 5), and re-tear was assessed using MR images (Figure 6) taken during the last checkup. Rotator cuff re-tear as a repair failure was graded as per the classification system described by Sugaya et al. (24). Sugaya type III is defined as insufficient thickness without discontinuity and suggests a partial tear (Figure 7). Type IV indicates a minor discontinuity and suggests a small tear. Type V is defined as a major discontinuity and represents an obvious tear. In this study, Sugaya type III MRI findings suggesting a partial tear were assigned to the failed repair group. This is because numerous studies have indicated that postoperative Sugaya type III lesions could develop into obvious tears (25, 26).

Figure 5.

Postoperative anteroposterior shoulder radiograph of a patient in the ATO group

Figure 6.

Postoperative coronal magnetic resonance imaging view of a patient in the TOE group

Figure 7.

Sugaya type III re-tear pattern of magnetic resonance imaging view of a patient in the ATO group

Statistical analysis

All statistical analyses were performed using the Statistical Package for Social Sciences version 20.0 (IBM Corp.; Armonk, NY, USA). The normality of the data was determined using the Shapiro Wilk test. The descripts were presented as the mean±SD or as frequency tables for quantitative and qualitative parameters, respectively. Differences between the ATO and TOE groups were reviewed either using the Student’s t-test or Mann–Whitney U test. Pearson’s Chi-square test and Fisher’s exact tests were used to compare the qualitative data of the groups. Paired sample t-test was used to evaluate the change in the quantitative data over time. A logistic regression model was constructed using the parameters of age, sex, tear type, size, retraction level, fatty degeneration degree, re-tear existence, biceps tenotomy, and tenodesis existence using the Enter method. Power analysis was determined as 95% in all analyses, and p≤0.05 was considered to indicate statistical significance.

Results

The mean follow-up duration was 33.3±11.8 months. No difference was observed between the two groups in terms of the distribution of demographic data, tear type and size, fatty degeneration degree, intra-joint additional pathology (SLAP 1 and 2) existence, and additional procedures (biceps tenotomy or tenodesis, acromioplasty) (Table 1).

Table 1.

Demographic data

Variable		ATO repair (n=39)	TOE repair (n=44)	p
Age, y		57.03±6.39	57.86±7.81	0.597
Sex	Female	32 (82.1%)	36 (81.8%)	0.794
Follow-up, months		33.72±12.14	32.95±11.76	0.772
Site of rotator cuff tear		n (%)	n (%)	0.812
	Left	28 (71.80)	30 (68.19)
Smoking status		n (%)	n (%)	0.853
	Yes	9 (23.08)	12 (27.27)
	No	30 (76.92)	32 (72.73)
Diabetes mellitus		n (%)	n (%)	0.839
	Yes	11 (29.73)	14 (31.82)
	No	26 (70.27)	30 (68.18)
Tear Type (Ellman and Garstman classification)		n (%)	n (%)	0.941
	U Shape	10 (25.64)	11 (25.00)
	Crescent	19 (48.72)	24 (54.55)
	L Shape	8 (20.51)	7 (15.91)
	Reverse L	2 (5.13)	2 (4.55)
Tear size Anteroposterior dimension cuff tear, mm		22.87±5.77	21.77±6.11	0.403
Amount of retraction (Patte classification)		n (%)	n (%)	0.299
	Stage 1	20 (51.28)	22 (50.00)
	Stage 2	14 (35.90)	11 (25.00)
	Stage 3	5 (12.82)	11 (25.00)
Fatty infiltration of supraspinatus, (Thomazeau method)		n (%)	n (%)	0.673
	Grade 1	27 (69.23)	27 (61.36)
	Grade 2	12 (30.77)	17 (38.64)
Biceps procedure		n (%)	n (%)	0.891
	Tenotomy	15 (50.00)	13 (44.83)
	Tenodesis	15 (50.00)	16 (55.17)

Considerable improvements were noted in the VAS, Constant, and Oxford shoulder scores after the surgery. However, no difference was observed in the preoperative and postoperative scores between the groups (Table 2).

Table 2.

Comparison of functional outcomes between groups and in-groups

Variable		ATO repair (n=39)	TOE repair (n=44)	p
VAS score	Preoperative	8.10±1.35	8.20±1.23	0.720
	Postoperative	3.15±1.23	3.18±1.13	0.914
	p	<0.001	<0.001
Constant score	Preoperative	31.59±6.07	33.48±7.03	0.197
	Postoperative	88.56±9.85	87.23±14.62	0.631
	p	<0.001	<0.001
Oxford shoulder score	Preoperative	16.51±4.24	16.14±4.65	0.702
	Postoperative	42.87±5.00	40.48±8.89	0.130
	p	<0.001	<0.001

Re-tear was identified in the MR examination of patients during their final follow-up; in the TOE group: 10 (Sugaya type 3 n=2, type 4 n=4, type 5 n=4) and ATO group: 9 (Sugaya type 3 n=6, type 4 n=2, type 5 n=1). In the final follow-up, the Constant and Oxford shoulder scores of the patients with re-tear were 64.4 and 25.9 in the TOE group and 76.33 and 37.2 in the ATO group, respectively. Considerable improvements were observed in the functional scores of patients with a re-tear after the surgery; however, improvement in the ATO group was more significant than that in the TOE group (p<0.05).

After testing for multicollinearity, the factors that could cause re-tear were investigated for both the groups using multivariate logistic regression analysis. Age, tear size, and retraction level were significantly associated with cuff repair failure in both the groups (p<0.05).

With respect to the complications, a superficial infection was found in one patient of the TOE group, which was treated with antibiotic therapy. Cortical penetration of implant occurred in one patient of the ATO group; however, it did not cause any failure. No anchor pullout or implant failure was observed in any patient.

Discussion

The most important finding of this study was the similar clinical success rate at a mean follow-up duration of 33.3±11.8 months in patients with rotator cuff rupture treated using ATO and TOE techniques. No difference was observed in the re-tear rates of the groups. However, the Oxford shoulder scores of those in the ATO group with re-tear improved more than those in the TOE group.

Traditional open transosseous repair gives excellent results for the fixation of the tendon to the bone and has represented the gold standard for rotator cuff repair with excellent long-term results (21). With the development of devices that allow arthroscopic transosseous repair, the ATO technique is now being preferred because of its biological and biomechanical features (27–29). Moreover, the TOE technique could support biological healing at a repaired rotator cuff tendon insertion by increasing the contact pressure between the repaired tendon and footprint, as in the traditional open transosseous repair (4, 11). A recent comparative study has revealed that TOE repair resulted in a significantly higher mean failure load than ATO repair in a cadaveric model (30). Additionally, the TOE technique has superior repair reduction and tendon bone contact capacity (31, 32). Thus, ATO and TOE techniques represent high capability of restoration of the footprint anatomy, and a comparison of these techniques did not yield any clinical or radiological difference in our study.

Various re-tear rates have been reported in numerous studies based on the fixation method. Plachel et al. stated that an overall re-tear rate of 33% was obtained in patients treated with ATO repair, with an average follow-up duration of 15 years (33). Baudi et al. reported that no re-tear occurred in their ATO patients with implant series at an average follow-up of 18.6 months (1). Kuroda et al. reported a re-tear rate of 7% in 384 shoulders treated with the ATO technique at the one-year follow-up (29). Sugaya et al. determined a re-tear rate of 25.6% after surgery using the arthroscopic SR technique and that of 9.7% after surgery with the DR technique (24). Cho et al. determined a re-tear rate of 22.5% after arthroscopic SR repair and identified patient age, fatty degeneration, and tear size as related factors (34). Randelli et al. stated that no significant difference was observed between the ATO and anchor repair techniques in their randomized prospective study (35). Numerous studies have suggested that the DR technique can increase tendon healing. The systematic review by Nho et al. stated that the current data do not support the DR technique (36). A prospective randomized study revealed little difference between both the techniques in terms of tendon healing (5). In our study, the re-tear rates were 22.72% for the TOE group and 23.1% for the ATO group. We did not determine any difference between the groups. The factors that contributed to failure were identified as patient age, tear size, and retraction level.

Liu et al. achieved significant healing in the shoulder and pain scores of 27 patients treated with the ATO technique, with a minimum follow-up duration of 2 years (37). Plachel et al. reported good functional results with a high satisfaction rate at an average follow-up of 15 years (33). Randelli et al. could not determine any difference between the ATO and anchor repair techniques in terms of pain, function, and integrity (35). In a prospective study of the ATO technique with implant, Baudi et al. achieved significant improvements in the shoulder functions of 34 patients within an average follow-up of 18.6 months (1). Flanagin et al. stated that the ATO rotator cuff repair technique led to a statistically significant midterm improvement in the ROM and a satisfactory midterm subjective outcome score in 108 patients (38). Fox et al. achieved good short-term result in 50 patients treated with ATO using the button technique (39). Kuroda et al. used the ATO technique without implant on 380 patients and identified significant healing in the shoulder functions at an average follow-up of 3.3 years and found a re-tear in 24 patients (29). We achieved significant improvement in the shoulder functions of both the groups; however, no significant differences were observed between the groups. When the shoulder functions of the patients with re-tear were evaluated, the patients in the ATO group had better functional results than those in the TOE group. This situation can be explained by the fact that six patients with re-tear in the ATO group were Sugoya Type III. Sugoya III patients are actually regarded as being healthy in many studies (34). However, due to the potential of full layer progress of this pattern in the future, it must be accepted as a re-tear. Further, failures usually occur as tears at the musculotendinous junction and medial row anchor pullout in the TOE group (40). This failure type may be worsened by the functional results because of total failure of repair (Sugaya IV and V).

The failure is usually at the tendon level in repairs with anchor; however, it occurs at the tunnel in the transosseous technique (41, 42). In the latest techniques, tunnel placement extending to the footprint’s medial that covers a large area has reduced tunnel breakage complications (37). Caldwell et al. reported that cortical augmentation could increase the suture pullout strength by approximately two-fold (43). The use of bone augmentation to avoid suture pullout is well supported in several biomechanical studies (44, 45). The advantages of the implant that we used in the ATO group are reported as cortical fixation, the extension capacity of suture passers up to the medial, and traction-compression effect in the system because of the sutures passing via the hole behind the implant and the reduction of movement at tendon-footprint interface (1). No intraoperative bone laceration was observed in any patient. Only one patient’s implant had cortical penetration, which was solved by turning the implant 90° and placing it on the cortex.

This study has certain limitations. First, this retrospective study enrolled a relatively small number of patients and may be underpowered to detect differences in certain factors. Second, although no significant difference was observed in the clinical outcomes of patients with and without bone laceration, a type II error may exist due to the relatively small number of patients. Third, we did not record the operation time; however, based on our experience, both the techniques can be performed within a similar duration. Fourth, we did not compare the invoiced cost of both the techniques. However, the estimated cost of an ATO implant equals the price of three anchors in our country. Generally, both the techniques cost the same, although four anchors are used in the TOE technique.

In conclusion, the ATO and TOE techniques help achieve significant improvements in the shoulder function in the short term. However, studies that constitute longer follow-up periods and include a large series are warranted.

MAIN POINTS.

• This is the one of the small number of studies which compared ATO and TOE techniques in the litearature.

• Significant functional and radiological results were achieved with both techniques in the short term. Also better functional results were determined in patients with rerupture in ATO group.

• Although no difference was observed in the re-tear rates of the groups, better functional results were determined in patients with re-tear in ATO group.