Failure of polymerized lactic acid tacks in shoulder surgery

Failure of polymerized lactic acid tacks in shoulder surgery John P. Wilkerson, MD, John E. Zvijac, MD, John W. Uribe, MD, Matthias R. Schürhoff, MD, and Jeremy B. Green, BA, Coral Gables, FL The purpose of this study was to evaluate 4 cases in which bioabsorbable polymerized lactic acid tacks failed after arthroscopic shoulder surgery. Four male elite athletes with recurrent shoulder pain were seen a mean of 7.5 months (range, 3-10 months) after initial arthroscopy. Three of the cases involved superior labrum anterior-to-posterior (SLAP) lesion stabilization, and the fourth case was a rotator cuff (RTC) repair. In the three labral repairs, the implant had broken and the unabsorbed fragments were visible with magnetic resonance imaging. The device used in the RTC repair showed no signs of absorption. All 4 patients underwent arthroscopic removal of the polymer tack fragments to alleviate their symptoms, 2 of whom had foreign-body reactions that required synovectomy. On the basis of clinical examination and magnetic resonance imaging, 2 of the SLAP lesions and the RTC tear had healed. The third patient with a SLAP lesion required arthroscopic debridement of a portion of the labrum. The intact RTC implant had backed out of its insertion point. In all 3 labral repairs, the polymerized lactic acid implant experienced a mechanical failure near the head-shaft junction. We theorize that the labral implants failed because of the variable rate of degradation along the shaft of the devices from the intraarticular to intraosseous regions. (J Shoulder Elbow Surg 2003;12:117-21.) The use of biodegradable implants in orthopaedics is a relatively recent application of a technology that has been used since the height of the Roman Empire. While serving as physician to Emperor Marcus Aurelius and the Roman gladiators, Grecian Claudius Galenus (ad 129-210), better known as Galen, first used a resorbable catgut suture made from animal intestines in 175 AD. Arthroscopic repair of pathologic conditions of the shoulder with biodegradable From the UHZ Sports Medicine Institute. Reprint requests: John E. Zvijac, MD, UHZ Sports Medicine Institute, 1150 Campo Sano Ave, Suite 200, Coral Gables, FL 33146, USA (E-mail: jzvijac@uhzsmi.com). Copyright 2003 by Journal of Shoulder and Elbow Surgery Board of Trustees. 1058-2746/2003/$35.00 0 doi:10.1067/mse.2003.16 suture anchors and tacks has become increasingly more popular. There are multiple reports of the success of arthroscopic shoulder surgery with biodegradable tacks, with follow-up ranging from 1 to 5 years. 1,6,9 Although there have been reports of biodegradable implant failure, to our knowledge, there are no published reports of polymerized lactic acid (PLA) implant failure in shoulder surgery. In a review of patients treated with bioabsorbable fixation devices, we have identified 4 patients who, 3 to 10 months after surgery, demonstrated clinical failure of the PLA tack. All 4 implants failed to absorb, and 3 failed near the head-shaft junction. We present these 4 cases of implant failure in the arthroscopic repair of 3 superior labrum anterior-to-posterior (SLAP) lesions and one rotator cuff (RTC) tear in elite athletes. CASE REPORTS Case 1 A 16-year-old, male, right hand dominant water polo player and swimmer presented with a 3-month history of pain with throwing and other overhead activities such as swimming. Magnetic resonance imaging (MRI) revealed a type II SLAP lesion in the right shoulder. The patient s shoulder was refractory to nonsurgical treatment. Diagnostic arthroscopy of the right shoulder confirmed the MRI diagnosis of a type II SLAP lesion. The bony glenoid was burred to healthy bleeding bone and served as a bed for the fixation of the displaced labrum with the use of a bioabsorbable poly-l,d-lactic acid (PLDLA) tack (TissueTak II; Arthrex, Naples, Fla). The patient was pain-free 2 months postoperatively. Three months after surgery, he complained of right shoulder pain while doing light swimming exercises. The shoulder was once again refractory to conservative measures including nonsteroidal anti-inflammatory drugs. Physical examination revealed pain with RTC exercises, especially abduction and external rotation activities. Plain radiographic findings were negative. MRI examination revealed healing at the site of the labral tear and a T-shaped foreign body in the subscapularis fossa (Figure 1). Diagnostic arthroscopy of the right shoulder confirmed the MRI results, as the head and small portion of the neck of the bioabsorbable tack were visible when viewed from the anterior portal down the subscapularis fossa. During abduction and external rotation of the shoulder, the fragment contacted the subscapularis. The single-fragment foreign body was removed with a grasper. During arthroscopy, the labral repair appeared to be solid and stable. Within 3 weeks of surgery, the patient had competed in 2 swim meets and had no complaints. 117

118 Wilkerson et al J Shoulder Elbow Surg March/April 2003 Figure 2 Case 2: The polymerized lactic acid tack shattered into multiple fragments that had failed to biodegrade. Figure 1 Case 1: Magnetic resonance image 4 months postoperatively shows the T-shaped polymerized lactic acid head-shaft fragment. Case 2 A 23-year-old, male, right hand dominant professional football player presented with a 2-week history of right shoulder pain. His symptoms were consistent with posterior right shoulder instability. Plain radiographic findings were negative. An MRI examination revealed a type II SLAP lesion. The posterior labrum was completely separated from the glenoid. At this time, one bioabsorbable poly-l lactic acid (PLLA) tack (Bankart Tack; Bionx Implants, Blue Bell, Pa) was placed posteriorly, and a suture anchor and another tack were placed anteriorly. Grade 3 and 4 chondromalacia of the glenoid was visible. Three months postoperatively, the patient had a full range of motion and full strength. He returned to football 5 months after surgery. Eight months postoperatively, he presented with new and increasing right shoulder pain. MRI examination revealed a recurrent tear of the posterior labrum and postsurgical lucencies in the glenoid, as well as degenerative changes of the glenohumeral joint with subchondral sclerosis. Intraoperatively, both Bankart Tacks were found to be shattered and not absorbed. The head and part of the stem of each had broken off and fragmented. Sections of the stems protruded from their original position in the glenoid. Other fragments were either freely floating or trapped in synovitis. Both loose implant fragments and the still-fixated stems were removed by suction irrigation and grasper (Figure 2). The patient returned to his pre-injury level of activity 6 weeks after implant removal. Case 3 A 29-year-old, male, right hand dominant professional football player presented with a 1-week history of right shoulder pain. Physical examination revealed posterior instability and signs of a SLAP lesion of the shoulder. Radiographic findings were negative. MRI examination demonstrated a type II SLAP lesion, with the posterior labrum completely detached from the glenoid. He underwent diagnostic arthroscopy. The posterior labrum was completely detached up to and including the biceps anchor. A torn inferior flap of labral tissue was also found and debrided. Two PLLA Bankart Tacks were placed posteriorly, stabilizing the labrum. The patient returned to full athletic activity 3 months postoperatively. Nine months after surgery, he presented with shoulder pain. MRI examination at that time revealed that the head and part of the stem of the Bankart Tack had broken off and had settled in the subscapularis fossa. Although the patient was counseled to undergo arthroscopic removal of the foreign body, he chose to continue playing through the pain and elected not to have surgery. Eleven months after initial arthroscopic surgery and two months after his presentation with recurrent shoulder pain, the patient could no longer tolerate the inflammation and pain and underwent season-ending surgery to remove the tack fragments. During arthroscopy, the labral repair appeared to be solid and stable. Six weeks postoperatively, the patient returned to his pre-injury level of activity. Case 4 An 18-year-old right hand dominant college basketball player presented with a 2-week history of right shoulder pain and loss of range of motion. Physical examination revealed pain with forward flexion and extension. Radiographic findings were negative. MRI examination revealed a complete tear in the supraspinatus tendon at its insertion onto the greater tuberosity. The injury was initially treated conservatively with an RTC-strengthening program and nonsteroidal anti-inflammatory drugs. The patient elected to undergo arthroscopy, at which time the diagnosis of a torn RTC was confirmed. A bony bed was prepared and the torn RTC tendon was debrided and fixed to the greater tuberosity with 2 PLLA CuffTack bioabsorbable implants (Mitek, Westwood, Mass). The repair was noted to be solid and stable. Six months postoperatively, he was cleared to return to full activity. Ten months after surgery, he presented with recurrent right shoulder pain, especially during overhead activities. Plain radiographic findings were negative. MRI

J Shoulder Elbow Surg Wilkerson et al 119 Volume 12, Number 2 Figure 3 Case 4: During arthroscopy 11 months after initial rotator cuff repair, the polymerized lactic acid tack was visualized. There is no indication of implant biodegradation. examination revealed no significant RTC pathology. One of the tacks remained in its original position in the posterosuperior margin of the joint, while a portion of the other was prominent intraarticular. Diagnostic arthroscopy showed that the RTC tendons were intact. However, neither of the 2 tacks had signs of absorption (Figure 3). Both implants were removed, and the extensive synovitis in the region was debrided. Three weeks later, the patient returned to full activity. DISCUSSION Initially, nonabsorbable orthopaedic devices were used during open shoulder procedures. In recent years, arthroscopy with biodegradable implants has been used in shoulder surgery with positive outcomes. 1 Pitfalls associated with nonabsorbable implants have included difficult postoperative imaging, infection, impingement, hardware loosening with the possibility of metallic loose bodies leading to articular cartilage damage, and increased difficulty of a revision if the repair fails. 10,15 These problems fueled the development of absorbable fixation devices hoping to avoid the risks of metallic implants around joint surfaces. Biodegradable implants are attractive as orthopaedic fixation devices because they provide solid initial fixation and progressively lose strength as the injury heals and requires less rigid stabilization. Furthermore, after they begin to degrade, these implants become viscoelastic, meaning they isolate vibrations and absorb shock as their physical properties change with the rate of applied load. 5,10 Postoperatively, biodegradable implants do not interfere with imaging and allow for easier revision surgery, as they can be drilled through and do not need to be removed. In theory, because the polymers degrade to natural body metabolites, they are not subject to adverse foreign-body reaction and should require no secondary surgery for removal. The biodegradative pathway of polyglycolic acid (PGA) and PLA and their copolymers is well known. First, water molecules in the local tissue fluid nonspecifically cleave the ester linkages of the polymer chain, creating small, low-molecular-weight polymers and monomers that are more easily metabolized. As a result of this hydrolysis, the implant first loses strength and then mass integrity. In the next stage a mild inflammatory response consisting of giant cell and phagocyte infiltration facilitates the absorption of the implant. 17 The short chains are converted to lactic acid, which is a natural byproduct of muscle contraction in anaerobic conditions. The lactic acid is converted to pyruvate, and then acetyl coenzyme A, at which time it enters the tricarboxylic acid cycle. After the tricarboxylic acid cycle, the polymers have been completely broken down to carbon dioxide and water and are excreted via the lungs. The rate of absorption through this pathway is dependent upon the starting molecular weight of the polymer, its crystallinity, and its porosity. 2 The nature of degradation is worth noting, as absorption occurs heterogeneously, faster at the center than at the surface of the implant. 7 In 1991, the first biodegradable polymerized implant for use in shoulder surgery was introduced, the PGA Suretac (Acufex Microsurgical, Mansfield, Mass). The Suretac is a molded tack that is expected to lose strength completely within 6 weeks and to absorb fully between 6 and 24 months. 5,10 Although PGA shoulder fixation devices were designed to be biocompatible, they have caused complications including hypertrophic fibrous encapsulation, sterile sinus tract formation, and osteolysis. Edwards et al 7 described 6 shoulders in 5 patients that displayed marked synovitis and nonspecific granulomatous reactions after labrum fixation with a PGA implant. Burkart et al 4 reported 4 cases of breakage and early loosening of the PGA Suretac device accompanied by massive foreign-body reaction 2 to 5 weeks after implantation. Some bioabsorbable implants use PLA, such as the PLLA Bankart Tack and the PLLA CuffTack. PLA is an asymmetric molecule that exists in both the crystalline l-form and the amorphous d-form. PLA biodegrades more slowly than PGA because it is highly hydrophobic and, therefore, less susceptible to the initial step of the biodegradative pathway, hydrolysis. The Bankart Tack in particular degrades more slowly than other PLA and PGA implants because it is self-reinforced. It is expected to maintain strength for up to 6 months and to reabsorb fully within 2 to 5 years. This tack is machined from a small block of polymer to its final form. Many authors report success with PLLA implants. 11,12,16 However, like PGA devices, PLA im-

120 Wilkerson et al J Shoulder Elbow Surg March/April 2003 plants have also been reported to have complications. 8,14 The slow degrading nature of PLLA and possible consequences associated with its degradation are underscored in the report by Böstman and Pihlajamäki 3 of a local inflammatory reaction to a PLLA screw used in the fixation of an ankle fracture more than 4 years after surgery. Copolymers of PLA have also been used in arthroscopic shoulder surgery. The TissueTak II implant is molded from the racemic combination of the two forms of PLA, crystalline l-pla and amorphous d-pla, called PLDLA. Athanasiou et al 2 highlighted that the minor differences in molecular arrangement between PLLA and PLDLA profoundly affects the biomechanical response of the body and the degradation characteristics of the implant. Research by Stahelin et al 13 confirms this theory, as they found that at 20 months postoperatively, there was no significant degradation of the PLLA interference screw, whereas the PLDLA screw had degraded completely by 10 months. In 2 of the 4 cases presented, the PLLA Bankart Tack was used during arthroscopic shoulder stabilization; the others used were the PLDLA TissueTak II and the PLLA CuffTack. The decision to use devices composed of PLA as opposed to PGA was primarily based on the slower biodegradation rate of PLA, which allows the devices to retain their mechanical properties for a longer period of time. In each case presented the shoulder pathology was treated with an arthroscopic repair with bioabsorbable tacks. All 4 patients had returned to athletics and reported a pain-free interval after surgery ranging from 3 to 10 months (mean, 7.5 months) before presenting with acute pain. All required further surgery to relieve their symptoms. At the time of repeat arthroscopy, the RTC tear had healed completely, as did 2 of the 3 labral tears. The third patient undergoing second-look arthroscopy had experienced a recurrent labral tear. Although 3 different types of PLA absorbable tacks were used, none of the 4 tacks had completely absorbed. The most striking example of the failure of the PLA implant to biodegrade properly and lose strength as the soft tissue healed was the CuffTack. Eleven months postoperatively, both removed CuffTacks retained their original appearance and physical properties. Consequences of this failed absorption include foreign-body reactions and the necessity of repeat surgery for implant removal, essentially nullifying the advantages of bioabsorbable devices over their metallic predecessors. Interestingly, each of the 3 labral repair devices failed near the head-shaft junction that was primarily exposed to soft tissue. We believe that this was because of the difference in microenvironments along the body of the tacks. It has been shown that bone and soft tissue degrade bioabsorbable polymers at different rates. Maitra et al 10 reported that biodegradation occurs more rapidly in soft tissue and occurs more quickly when the implant is exposed to load. Böstman and Pihlajamäki 3 explained that the rates of degradation are different because the intraosseous location of the polymer will have a different mechanism for debris clearance, neoangiogenesis, and the replacement of implant with tissue versus the extraosseous location. Perhaps the proximal portion of the implant fragmented because there was a difference in absorption rates between microenvironments in bone, soft tissue, and the uncovered portion of the head. Furthermore, tensile stresses are focused at the area slightly below the head-shaft juncture. The increased absorption rate of the proximal region of the shaft exposed to soft tissue causes it to lose its material properties more quickly. This decrease in mechanical strength combined with the stress probably causes the implants to develop microfractures at this particular position along the tack. Athanasiou et al 2 reported that microfractures will enhance the absorption rate by increasing the surface area available for the hydrolytic degradation process. We presume that this results in the separation of the head fragment from the shaft. Because the head of the tack is the area least exposed to soft tissue, its absorption takes longer than any other position along the tack. Furthermore, the slow degrading nature of PLLA and PLDLA implants relative to their PGA counterparts allows the mechanical forces at the prominent tissue to capture portions to work for a longer period of time before the devices begin to biodegrade and attain viscoelastic properties, perhaps rendering PLA implants more susceptible to mechanical failure. In conclusion, we have identified potential complications when using surgical devices containing PLA. On the basis of our experience, we have limited our use of PLA implants relative to our use of PGA implants, with which we have had no significant complications. The possibility of implant breakage and loose fragments causing irritation in the joint should be included in the differential diagnosis when patients with an acute exacerbation in shoulder pain weeks to several months after the initial procedure are evaluated. This may be even more relevant in physically active patients participating in sports that require significant stresses across the shoulder. It is possible that the combination of increased demand across the shoulder joint with the increased absorption rate of the tacks in the soft tissue precipitated the fracture of the PLA absorbable implants. REFERENCES 1. Arciero RA, Taylor DC, Snyder RJ, Uhorchak JM. Arthroscopic bioabsorbable tack stabilization of initial anterior shoulder dislocations: a preliminary report. Arthroscopy 1995;11:410-7. 2. Athanasiou KA, Agrawal CM, Barber FA, Burkhart SS. Orthopae-

J Shoulder Elbow Surg Wilkerson et al 121 Volume 12, Number 2 dic applications for PLA-PGA biodegradable polymers. Arthroscopy 1998;14:726-37. 3. Böstman OM, Pihlajamäki HK. Late foreign-body reaction to an intraosseous bioabsorbable polylactic acid screw. J Bone Joint Surg Am 1998;80:1791-4. 4. Burkart A, Imhoff AB, Roscher E. Foreign-body reaction to the bioabsorbable Suretac device. Arthroscopy 2000;16:91-5. 5. Ciccone WJ, Motz C, Bentley C, Tasto JP. Bioabsorbable implants in orthopaedics: new developments and clinical applications. J Am Acad Orthop Surg 2001;9:280-8. 6. DeBerardino TM, Arciero RA, Taylor DC, Uhorchak JM. Prospective evaluation of arthroscopic stabilization of acute, initial anterior shoulder dislocations in young athletes. Am J Sports Med 2001;29:586-92. 7. Edwards DJ, Hoy G, Saies AD, Hayes MG. Adverse reactions to an absorbable shoulder fixation device. J Shoulder Elbow Surg 1994;3:230-3. 8. Friederichs MG, Greis PE, Burks RT. Pitfalls associated with fixation of osteochondritis dissecans fragments using bioabsorbable screws. Arthroscopy 2001;17:542-5. 9. Laurencin CT, Stephens S, Warren RF, Altchek DW. Arthroscopic Bankart repair using a degradable tack. A followup study using optimized indications. Clin Orthop 1996;332:132-7. 10. Maitra RS, Brand JC, Caborn DNM. Biodegradable implants. Sports Med Arthrosc Rev 1998;6:103-17. 11. McGuire DA, Barber FA, Elrod BF, Paulos LE. Bioabsorbable interference screws for graft fixation in anterior cruciate ligament reconstruction. Arthroscopy 1999;15:463-73. 12. McGuire DA, Barber FA, Milchgrub S, Wolchok JC. A postmortem examination of poly-l lactic acid interference screws 4 months after implantation during anterior cruciate ligament reconstruction. Arthroscopy 2001;17:988-92. 13. Stahelin AC, Weiler A, Rufenacht H, et al. Clinical degradation and biocompatibility of different bioabsorbable interference screws: a report of six cases. Arthroscopy 1997;13:238-44. 14. Takizawa T, Akizuki S, Horiuchi H, Yasukawa Y. Foreign body gonitis caused by a broken poly-l-lactic acid screw. Arthroscopy 1998;14:329-30. 15. Ticker JB, Lippe RJ, Barkin DE, Carroll MP. Infected suture anchors in the shoulder. Arthroscopy 1996;12:613-5. 16. Warden WH, Friedman R, Teresi LM, Jackson DW. Magnetic resonance imaging of bioabsorbable polylactic acid interference screws during the first 2 years after anterior cruciate ligament reconstruction. Arthroscopy 1999;15:474-80. 17. Warme WJ, Arciero RA, Savoie FH, Uhorchak JM, Walton M. Nonabsorbable versus absorbable suture anchors for open Bankart repair. Am J Sports Med 1999;27:742-6.