Biocomposite Suture Anchors Remain Visible Two Years After Rotator Cuff Repair

Abstract

Background

Biocomposite suture anchors containing osteoconductive materials have gained popularity in rotator cuff repairs. However, little is known about the influence of the addition of osteoconductive materials on implant resorption, bone reaction, tendon healing, and clinical outcomes scores.

Questions/purposes

(1) What percentage of suture anchors were not completely resorbed 2 years after implantation? (2) What are the diameters of the bone bed in relation to the implant? (3) Is tendon integrity correlated with bone tunnel diameter? (4) Is there an association between tunnel widening, periimplant fluid film grade, biodegradation grade, and retear with clinical outcomes scores, such as the Western Ontario Rotator Cuff Index (WORC) and the Oxford Shoulder Score (OSS)?

Methods

Thirty-six patients were enrolled from August 2012 to January 2014. The following inclusion criteria were applied: (1) reparable full-thickness supraspinatus tendon tears, (2) double-row suture bridge techniques applied for supraspinatus repair, (3) use of biocomposites suture anchor implants composed of poly L-lactic acid (PLLA) and β-tricalcium phosphate (TCP) exclusively, and (4) a minimum of 2 years followup. Four patients met the exclusion criteria, and seven of 36 patients (19%) were lost to followup. Thereby, 25 patients (84 implants) were included in this retrospective study. To answer the study’s questions, the following methods were applied: (1) The resorption of the implants and periimplant fluid film were assessed on MRI using a four-stage scale system, (2) bone bed diameter was measured on MRI at three different points on the longitudinal central axis of each anchor, (3) tendon integrity was evaluated on MRI according to the Sugaya classification and correlated to bone tunnel diameter, and (4) assessed tunnel diameters, periimplant fluid film grade, biodegradation grade, and tendon condition were related to clinical outcomes scores at the time of followup (2.3 ± 0.3 years). The intraobserver reliability was 0.981 (p < 0.001) and interobserver reliability was 0.895 (p < 0.001).

Results

At 2.3 ± 0.3 years, most analyzed suture anchors (76 of 84 [90%]) were, with varying degrees of degradation, still visible. Bone tunnels showed minor widening (0.4 ± 1.4 mm) at the base, but osseous ingrowth was detected as narrowing at the middle (0.1 ± 1.1 mm) and at the apex (1.4 ± 1.7 mm) of the implants. Patients with retears (Sugaya Grades 4-5) had narrower tunnels (3.6 ± 1.8 mm) than patients without retears (Sugaya Grades 1-3; 4.4 ± 1.6 mm; mean difference, 0.782 [95% confidence interval {CI}: 0.009–1.6]; p = 0.050). WORC and Oxford scores were not associated with the tunnel widening amount, fluid film grade, biodegradation grade, or tendon retear.

Conclusions

In light of the results of the present study, surgeons should consider in their daily practice that the resorption process of these implants may be slower than assumed so far, but no association with severe implant-related complications has been found in the short term. Future studies should focus on the evaluation of the effects of osteoconductive materials on resorption, tendon healing, and clinical outcomes in the long term and on the integration process in different rotator cuff reconstruction techniques.

Level of Evidence

Level IV, therapeutic study.

Introduction

Selecting an anchor that produces a stable and safe construct in arthroscopic rotator cuff repair while not affecting the healing process of the rotator cuff is essential [10, 23]. Since they were introduced into daily practice, bioabsorbable suture anchors have been repeatedly associated with various complications, such as osteolysis, pain, and implant failure. However, bioabsorbable suture anchors have undergone a steady improvement during the development of three implant generations [12, 13, 36]. The first bioabsorbable anchors were made of polyglycolic acid (PGA) and showed rapid absorption after implantation, producing inflammatory responses in the surrounding bone bed [9, 15]. Additionally, numerous reports indicated that bioabsorbable suture anchors made of PGA may have led to tunnel widening and pronounced periimplant fluid [1, 11, 29]. To slow the degradation process and reduce the periimplant fluid film formation, second-generation anchors made of poly L-lactic acid (PLLA) were introduced [17, 25]. The slower biodegradation process and improved biocompatibility of these implants led to fewer osseous reactions [12]. However, poor reestablishment of original bone tissues after implant resorption have been reported [2, 19, 31]. To increase bone ingrowth after implant absorption, third-generation biocomposite anchors containing β-tricalcium phosphate or hydroxyapatite and the aforementioned materials were introduced [4, 7, 30].

Since their introduction, the third-generation biocomposite suture anchors have become widely used [27, 37]. Several studies have investigated their biomechanical properties [5, 32], but very few have been published to date on the effects of the addition of osteoconductive material on resorption, tendon healing, and clinical outcomes [24, 28]. In particular, with regard to the biodegradation of these implants, the manufacturers postulated a complete resorption of the suture anchors after 2 years, which has not been verified [6]. Furthermore, it is unclear whether the addition of osteoconductive materials avoids or reduces tunnel widening and the formation of a periimplant fluid film [11]. As a result, it would also be interesting to assess whether and how the resulting reduction in tunnel widening and the formation of periimplant fluid films affect the rotator cuff healing process and the clinical outcomes scores.

Therefore, we asked: (1) What percentage of suture anchors were not completely resorbed 2 years after implantation? (2) What are the diameters of the bone bed in relation to the implant? (3) Is tendon integrity correlated with bone tunnel diameter? (4) Is there an association between tunnel widening, periimplant fluid film grade, biodegradation grade, and retear with clinical outcomes scores, such as the Western Ontario Rotator Cuff Index (WORC) and the Oxford Shoulder Score (OSS)?

Patients and Methods

Patient Recruitment

This study was approved by our institutional ethical board (registration number 160/16). Since 2011, biocomposite implants with osteoconductive materials have been part of routine clinical use at our institute and have been used for all rotator cuff repairs. In all, 36 patients were retrospectively enrolled from August 2012 to January 2014 in the present study (Fig. 1). The following inclusion criteria were applied: (1) reparable full-thickness supraspinatus tendon tear (Bateman [8] grade 3 or less) verified by preoperative MRI at the time of surgery, (2) a double-row suture bridge technique applied for supraspinatus repair, (3) use of biocomposite suture anchor implants composed of PLLA and β-TCP exclusively, and (4) a minimum of 2 years of followup. The following exclusion criteria were set: (1) patients with shoulder instability, (2) patients with single-row rotator cuff repair, (3) patients with partial reconstruction of the rotator cuff, (4) patients with any history of previous shoulder surgery including rotator cuff repair. We excluded 11 patients from the study; four patients matched the exclusion criteria (three with partial reconstruction and one had previous surgery), and seven of 36 patients (19%) were lost to followup (two could not be reached, one died due to conditions not related to surgery, and four declined to participate in the study). Thereby, 25 patients with a mean followup of 2.2 ± 0.3 years were included in the present study. We analyzed 84 biocomposite suture anchors (44 Corkscrews® [Arthrex, Naples, FL, USA] and 40 SwiveLock® [Arthrex, Naples, FL, USA]) in the present study. In addition to the supraspinatus repair, 11 tenodesis/tenotomies of the biceps tendon, four lateral clavicular resections, 17 subacromial decompressions, and 12 subscapularis repairs were performed.

Fig. 1.

A flow diagram of the study is shown here.

Anchor Composition and Surgical Technique

All surgical procedures were performed by a single orthopaedic surgeon (TK) who subspecialized in sport orthopaedics and shoulder surgery. For the medial row, the surgeon used mostly biocomposite suture anchors (biocomposite Corkscrews), and for the lateral fixation, the surgeon used biocomposite knotless suture anchors (biocomposite SwiveLock). Both suture anchors were composed of 85% PLLA and 15% β-tricalcium phosphate. The biocomposite Corkscrew suture anchors were 14.7 mm long with a 5.5 mm outer diameter; the biocomposite SwiveLock suture anchors were 19.1 mm long with a 5.5 mm outer diameter. All implants studied were perforated by the manufacturer to allow for bony ingrowth. The biocomposite Corkscrew suture anchors were preloaded with number 2 FiberWire® (Arthrex, Naples, FL, USA) made of ultra-high molecular weight polyethylene (UHMWPE) and polyester. Implants were inserted according to the manufacturer recommendations. After preparing the bone bed with the punch supplied by the manufacturer, the surgeon screwed the implants into the bone until they were flush with the bone surface. The surgeon performed a double-row suture bridge technique in all patients. The medial row sutures were always tied before being incorporated into the lateral row.

Aftercare

After the procedure, the operated arm was held in an abduction pillow (Medi SAS 15, Bayreuth, Germany) for 6 weeks. The patients performed passive exercises, depending on the combination of procedures performed. For isolated supraspinatus tendon repairs, passive flexion and abduction to 90° was allowed for 6 weeks. External rotation with the arm at the side was limited to 0° for 3 weeks and to 20° for another 3 weeks. At 7 weeks, patients stopped using the abduction pillow and began active mobilization, without strain. At 13 weeks, patients started strengthening exercises.

Clinical and Radiological Assessment

Clinical and radiological assessment of all patients was obtained at 2 years followup by a clinician (TF) who specialized in sports orthopaedics. For clinical evaluation, we used the subjective shoulder value (SSV) [14], the WORC [20], and the OSS [26]. The preoperative and postoperative comparisons showed improvement at followup (Table 1).

Table 1.

Total scores of the preoperative and postoperative questionnaires

An MRI was performed at the time of followup (2.3 ± 0.3 years) using a 1.5 Tesla MRI-scanner (MAGNETOM TIM-Symphony, Siemens, Erlangen, Germany). The patients were positioned supine with the arm in neutral rotation by the side of the body. A dedicated standard shoulder coil was placed over the shoulder. The following MRI-scan-protocol was developed and applied in the present study for all MRI-scans: Localizer sequence in all three directions of space. Paracoronal T1-weighted sequence with TR of 555 ms, a TE of 11 ms, slice thickness of 3 mm, an interslice gap of 0.3 mm, an FoV of 180 x 180 mm and an image matrix 384 x 384. Paracoronal DESS 3D with water excitation, a TR of 23.2 ms, a TE of 8.1 ms, a slice thickness of 1.5 mm, an FoV of 160 x 160 mm and an image matrix of 256 x 256. Parasagittal PD TSE with fat saturation, a TR of 2560 ms, a TE of 43 ms, a slice thickness of 3.2 mm, an interslice gap of 0.3 mm, an FoV of 180 x 180 mm and an image matrix of 320 x 320 mm. Transversal T1-weighted SE sequence with a TR of 530 ms, a TE of 16 ms, a slice thickness of 3 mm, an interslice gap of 0.6 mm, an FoV of 160 x 160 mm and an image matrix of 512 x 512 mm. T2weighted MEDIC 2D sequence with a TR of 1090 ms, a TE of 21 ms, a slice thickness of 3 mm, an interslice gap of 0.3 mm, an FoV of 180 x 180 mm, and an image matrix of 448 x 448. The analysis of the MRI scans was performed by two blinded clinicians (TF, MS), the intraobserver reliability was 0.981 (p < 0.001) and interobserver reliability was 0.895 (p < 0.001). The radiological measurements were performed using the medical imaging viewer Osirix (Pixmeo SARL, Bernex, Switzerland). The radiological investigation of the biodegradation, tunnel widening, and periimplant fluid of the implants was performed using parasagittal T2-weighted and paracoronal PD-weighted sequences. A gradual degradation of the anchors occurred according to the Haneveld classification [17]. Tendon integrity was evaluated in paracoronal and parasagittal PD-weighted sequences.

To answer our first research question about the resorption process of biocomposite suture anchors, the visible structure of the anchors was analyzed on MRI and evaluated using a scale system of four stages according to Haneveld et al. [17] (Table 2). A suture anchor was defined as “completely resorbed” if no residual structure on MRI (stage 4 according to Haneveld et al. [17]) was seen. Also, the presence of periimplant fluid was assessed in the MRI and graded according to Haneveld et al. [17] (Fig. 2). In the present study, the tip of the implant was referred to as “apex” and the beginning of the implant, which ends with the cortex, called “base” (Fig. 3).

Table 2.

Classification of biodegradation and amount of periimplant fluid of biocomposite suture anchors according to Haneveld et al. [17]

Fig. 2.

A T2 weighted frontal MRI of the shoulder showing periimplant fluid is presented here. (A) Grade 1, no periimplant fluid, (B) slight periimplant fluid, (C) continuous line of fluid < 1 mm thick, (D) continuous line of fluid > 1 mm thick.

Fig. 3.

A T2 weighted axial MRI of the shoulder displaying the measurement technique of tunnel width and length is shown here. The longitudinal central axis of each anchor was marked; three measurements were performed at 5 mm apart perpendicular to this central axis, respectively.

To investigate our second research question concerning the relation between implants and diameters of the bone beds, the longitudinal central axis of each anchor was first marked. Three measurements of the width of the bone beds were placed perpendicular to this central axis. The first width measurement was performed at the middle of the anchor, 10 mm from the anchor head. The other two width measurements of the bone bed were carried out 5 mm above and below this point, respectively (Fig. 3). Those three measurements were then used to estimate the relationship between the implants and the diameter of bone beds.

To answer our third research question about the condition of the rotator cuff repair, the morphology of the cuff repair was evaluated on an MRI according to the Sugaya classification [35]. A retear of the cuff repair was defined as a discontinuity at the footprint corresponding to Grades 4 and 5 of the Sugaya classification [35].

To answer our fourth research question concerning the association between the tunnel widening, the degree of periimplant fluid film, the degree of biodegradation, and the rotator cuff retear with the clinical results scores, the measurements obtained were correlated with clinical outcome scores at the time of followup (2.3 ± 0.3 years) [17]. Cystic periimplant formations, with a continuous diameter greater than twice the implant diameter, were defined as gross osteolysis [19]. There were no relevant cystic formations in preoperative MRI of the analyzed patients.

Statistical Analysis

We analyzed the collected data using intergroup differences and the chi-square test for nominal and Mann-Whitney U test for ordinal variables. The intra- and interobserver reliabilities were analyzed using intraclass correlation coefficient (ICC). Differences were considered significant for p values less than 0.05. Statistical analysis was performed using SPSS (version 21, IBM Corp, New York, USA).

The sample size was calculated assuming a confidence interval (CI) of 95%, with a width of 0.1 and an effect size of 0.3, resulting in a sample size of at least 82 implants with a power of 0.88 [16, 21]. Since the authors of the present study considered that the main goal of the research was to analyze the behavior of each suture anchor, it would seem more appropriate to consider each implant separately and thus determine the sample size based on the number of anchors, not just the patient number.

Results

At a minimum of 2.3 ± 0.3 years, most analyzed suture anchors (76 of 84 [90%]) were still visible (Fig. 4) and in most implants (72 of 84 patients, [86%]) only a moderate formation of periimplant fluid film, corresponding to Haneveld [17] grade three or below, was observed (Fig. 5). We did not observe any malpositioned suture anchors or positional changes of the suture anchors in the present study. In particular, there were no suture anchors with a portion of the anchor outside the cortex. Bone tunnels showed minor widening (0.4 ± 1.4 mm) at the base, but osseous ingrowth was detected as narrowing at the middle (0.1 ± 1.1 mm) and at the apex (1.4 ± 1.7 mm) of the implants. The mean width of the bone tunnels was 4.1 ± 1.7 mm at the apex, 5.4 ± 1.1 mm in the middle, and 5.9 ± 1.4 mm at the base of the implant. Mean bone bed length was 19.0 ± 2.4 mm. One minor osteolysis adjacent to a Corkscrew suture anchor was observed (Fig. 6) The patient was asymptomatic, had an average clinical result, and was satisfied with the outcome of surgery. No other osteolysis was observed in the present study. The radiological analysis and comparison between lateral and medial row (Table 3) or between knotted and knotless implants (Table 4) showed no differences.

Fig. 4.

The biodegradation degrees, according to Hanevald [17], of all suture anchors (n = 84) are presented here. Grade 1: clearly visible structure; Grade 2: visible structure; Grade 3: partial visible; Grad 4: not visible structure.

Fig. 5.

The periimplant fluid film formation degrees, according to Hanevald [17], of all suture anchors (n = 84) are shown here. Grade 1: no periimplant fluid; Grade 2: slight periimplant fluid; Grade 3: continuous line of fluid < 1 mm thick; Grade 4: continuous line of fluid > 1 mm thick.

Fig. 6.

A T2 weighted coronal MRI scan of the shoulder showing one minor osteolysis, involving a Corkscrew suture anchor is shown here.

Table 3.

Comparison between suture anchors of the medial row vs lateral row, concerning biodegradation and periimplant fluid and diameter of the bone beds

Table 4.

Comparison between and knotted vs knotless suture anchors, concerning biodegradation and periimplant fluid and diameter of the bone beds

Patients with Sugaya [35] Grades 4 to 5 retears had narrower tunnels (3.6 ± 1.8 mm) than patients without retears (Sugaya [35] Grades 1-3; 4.4 ± 1.6 mm; mean difference: 0.782 [CI: 0.009–1.6]; p = 0.050). However, there was no association between retear and diameters at the base and middle (Table 5). Additionally, there were no associations between tendon integrity and periimplant fluid or biodegradation of the suture anchors. Overall, retear of the supraspinatus tendon was detected in eight of 25 patients [32%] (Fig. 7).

Table 5.

Correlation between tendon integrity and periimplant fluid, biodegradation and tunnel diameter of the biocomposite anchors

Fig. 7.

Tendon integrity degrees, according to Sugaya classification [35], of all patients (n = 25) are presented here. Grade 1: sufficient thickness, homogeneous tendon; Grade 2: sufficient thickness, partial high-intensity from within the tendon; Grade 3: insufficient thickness without discontinuity; Grade 4: minor discontinuity on more than one slice; Grade 5: major discontinuity.

WORC and Oxford scores were not associated with tunnel widening amount, fluid film grade, biodegradation grade, or tendon retear (Table 6). No differences were found between patients with isolated supraspinatus repair versus those with additional procedures, such as acromioplasty, lateral clavicle resection, subscapularis repair, or biceps tenodesis.

Table 6.

Correlation between sum scores and tunnel diameter, amount of periimplant fluid, biodegradation and tendon integrity

Discussion

Biocomposite implants, composed of PLLA and β-TCP, are seeing wider use for rotator cuff repairs due to their perceived biocompatibility, biomechanical properties, and clinical safety [5, 6, 24]. However, only few studies have been published to date on how osteoconductive materials effect the resorption process, tendon healing, and clinical outcome [4, 30]. Therefore, we aimed to investigate the resorption process of the biocomposite suture anchors, the bone tunnel characteristics around the implant, and their association with clinical outcomes scores. The results of the present study showed that the resorption process is slower, as indicated by the manufacturers. Furthermore, minor widening at the base but narrowing at the middle and at the apex of the bone tunnels was observed. Clinical outcome scores were not associated with tunnel widening amount, fluid film grade, biodegradation grade, or tendon retear.

The present study has several limitations. First, when calculating the sample size in this study, the number of anchors was considered. This may seem unusual, due to the fact that each patient had an average of three suture anchors, and thus each analyzed anchor was not completely independent. This may have violated certain statistical assumptions about this study. But, given that the purpose of this study was to analyze the behavior of each suture anchor, it seemed more appropriate for the authors to analyze each implant separately and to determine the sample size based on the number of anchors. Second, the present study has a short-term followup, as the MRI was performed at a minimum 2 years after the rotator cuff repair. This was due to the fact that the manufacturers assumed a complete resorption of suture anchors between 18 and 24 months and bone ingrowth 24 months after implantation [3, 4]. However, this assumption does not match the daily experiences of the authors in revision surgery. Therefore, to confirm the presumption of the manufacturers, the authors decided to conduct this examination 2 years after the implantation of suture anchors. As well, in relation to this limitation, accounting for the fact that degradation is an ongoing process, the selection of a minimum of 2 years may have increased the variance in the degradation results. Considering the low SD, however, this variance must to have been minimal. Third, we did not analyze a control group of nonosteoconductive implants in the present study. Studies of other suture anchors that do not contain osteoconductive materials have already been adequately performed [17, 19]. Therefore, an additional comparison with nonbiocomposite implants would not add any important new information. Fourth, the radiological analysis of the osseous integration was studied on MRI. To minimize radiation exposure, no CT scans were performed in the present study. Previous studies have demonstrated that it is possible to indirectly estimate the osseous integration of biocomposite implants using MRI [30, 33, 34]. The narrowing of the tunnels on MRI, from the point of view of the authors, may only have arisen because of true osseous ingrowth in the implants, as this already demonstrates an osseous integration of the anchors. The implementation of a CT-scan evaluations would have exposed the patients to unnecessary radiation and would not have been ethical. Finally, the radiological results of the present study concerning osseous reactions were not confirmed histologically. MRI is certainly not able to reproduce all histological reactions between implant and bone bed. However, the main purpose of the present study was to investigate the clinical feasibility of these implants, and the most reproducible method, in this regard, appeared to be an MRI. A histological examination would certainly have provided further interesting information, but would have been difficult to reproduce in clinical practice.

The most important finding of the present study was that nearly all anchors were not completely resorbed 2 years after implantation. Manufacturers postulated a complete resorption between 18 and 24 months and bone ingrowth by 24 months after implantation. These results have not been adequately investigated so far. Barber et al. [4] studied the degradation process and tissue restoration of suture anchors made of PLLA/PGA and β-tricalcium phosphate in 20 patients at an average of 37 months after the rotator cuff repair. In this study, suture anchors were completely degraded, and no remnant was detected in all patients. In a review including 13 studies and 668 patients, Barber at al. [6] reported that the biocomposite interference screws or suture anchors composed of PLGA/ β-tricalcium phosphate were almost fully absorbed after a followup of 28 months. Bearing in mind the different implant compositions of the analyzed anchors in Barber et al. [6] and based on the results of our study, the resorption process appears to be slower than previously reported. This finding reflects the everyday experiences of the authors in shoulder revision surgery. From the authors´ point of view, this finding raises the question of whether or not the implants will be completely resorbed and replaced by autologous bone, as this was the main indication for the addition of osteoconductive materials. Therefore, to definitively clarify if the addition of osteoconductive materials affects the resorption of the suture anchors, further studies with longer followup are required.

Evidence for osteoconductivity was found at the middle and the apex of the implants. In a study by Lee et al. [22], the radiological results of two different tibial fixations performed using bioabsorbable screws with added osteoconductive materials (hydroxyapatite) and pure PLLA screws were compared in 394 patients after ACL reconstruction. In this study, screws with hydroxyapatite showed reduced tibial tunnel widening, improved osteointegration, and lower foreign body reactions. In the study by Barber et al. [4], osteoconductivity was seen in 20 of 28 [71%] of the anchor sites and was nearly complete in 14 of 28 [50%]. In the present study, the narrowing in the middle and apex of the bone tunnels may indicate true osseous ingrowth, as all implants were intentionally perforated by the manufacturer to allow bone channel formation to the core of the implants. We suggest that the different width measurements observed at different levels of the bone tunnel may derive from the fact that implant bone integration does not occur homogeneously, but starts at the apex and progresses towards the base of the implants. Therefore, considering the minor tunnel widening at the base and the tunnel narrowing at the apex and at the middle we observed, the results of the present study confirm that the addition of osteoconductive material in suture anchors allows the process of implant integration beginning at the apex. This may be advantageous in revisions surgeries in which new implant placements are requested. To clarify this issue, cadaveric studies with a histological analysis of biocomposite implants containing osteoconductive materials and related surrounding bone beds are needed.

In patients with reruptures after rotator cuff repairs, greater narrowing of the bone tunnel at the apex was observed. A possible increase of reruptures after rotator cuff repairs due to osseous reactions around biodegradable anchors has been suggested [18, 19]. A study by Kim et al. [18] reported a retear percentage of 31%; however, they observed no correlation between cyst formation and occurrence of a retear. In addition, Haneveld et al. [17] found a retear percentage of 37% and no correlation with tunnel widening, periimplant fluid film grade, and biodegradation grade. In addition, in the study by Haneveld et al. [17], the authors found no difference in rerupture of the rotator cuff between biocomposite suture anchors (37%) and nonabsorbable implants made of PEEK (35%). In the present study, the incidence of retorn rotator cuff tendons reflects the results of the studies mentioned previously. However, in the present study, we found an association between bone tunnel width at the apex and tendon integrity. We postulate that this was due to the fact that the traction of the healed tendon loading the implants led to micromovements of the implants and caused a slower bony integration process beginning at the apex compared with retorn rotator cuff tendons. These findings reveal that the traction of the reconstructed rotator cuff has a relevant impact on the integration of suture anchors in the surrounding bone, demonstrating the importance of distributing this tensile force on multiple anchors for the long-term stability of rotator cuff reconstruction. Future studies should compare the integration behavior of biocomposite suture anchors in different rotator cuff reconstruction techniques with several implanted suture anchors.

All outcomes scores improved after surgery, there was no association with the grade of periimplant fluid, biodegradation, bone tunnel diameter, or tendon integrity. The association between implant behavior through time and clinical outcomes scores was debated in previous studies [4, 6]. In a study that included 20 patients, Barber et al. [4] reported an improvement of all outcomes scores 37 months after rotator cuff repairs with biocomposite suture anchors made of PLLA/PGA and β-tricalcium phosphate. In this regard, the results of the present analysis are consistent with the study mentioned above. However, the present study has extended the analysis of clinical outcomes to the investigation of the relation between each behavior aspect of bioabsorbable suture anchors (such as the degree of resorption, the widening, and narrowing of the bone beds, the formation of periimplant fluid film) and clinical outcome scores. Considering the long-term development of bioresorbable anchors since their introduction, and bearing in mind the initially reported complications [9, 13, 31], such as the formation of osteolysis, implant failure, or chondrolysis, the results of the present study confirm that after a short-term followup, biocomposite implants with the addition of osteoconductive materials appear to be safe, with little periimplant film formation or bone bed widening. Studies are needed to evaluate the long-term effects of osteoconductive material on resorption, tendon healing, and clinical outcomes.

In light of our findings, surgeons should consider in their everyday practice that: the resorption process of these implants may be slower than assumed by the manufacturers; the addition of osteoconductive materials allows bone integration, which may be beneficial in revision operations where implantation of a new implant is required; the tensile force of the reconstructed rotator cuff may affect the integration of suture anchors in the surrounding bone; biocomposite suture anchors in short-term follow-up are not associated with severe implant-related complications. Future studies should focus on suture anchors resorption process in the long term, on the histological analysis of implanted anchors and their surrounding bone beds, on the integration process in different rotator cuff reconstruction techniques with different numbers of implanted suture anchors, and on the effects of osteoconductive material on bone reactions, tendon healing, and clinical results after prolonged followup.