Abrams (21) presented a thorough review of the important concerns related to clinical application of shoulder thermal capsulorraphy. Major issues include inconsistent tissue shrinkage, capsular ablation, over constraint, and neurologic complications. Anderson et al. (22) reported a 14% failure rate after thermal shoulder stabilization, with failure noted on average of 6 months after surgery. Failure was associated with multiple dislocations and prior surgery, with a tendency (inadequate statistical power) for decreased results with multidirectional instability and contact sports.

Early reports of thermal capsulorrhaphy tended to describe encouraging results. Lyons et al. (23) described a 12% unsatisfactory rate in 27 shoulders treated with laser shrinkage for multidirectional instability with a minimum follow-up of 2 years. Fitzgerald et al. (24) followed 30 of 33 consecutive multidirectional instability patients treated with radiofrequency thermal capsulorrhaphy in a military setting, with an average follow-up of 36 months. There were 3 excellent, 20 good, and 7 poor results, and 76% returned to active duty.

More recent reports tend to be more pessimistic about the clinical outcomes of thermal shrinkage. D’Alessandro et al. (25) followed 84 shoulders after thermal capsulorrhaphy performed with a monopolar radiofrequency device; most surgeries were done with the capsular “painting” technique. With an average follow-up of 38 months, the overall success rate was 63%, and the most favorable outcomes were seen in the patients with a traumatic dislocation. The worst outcomes were in patients with multidirectional instability with only a 55% success rate. Of note, these patients did well for the first year after which the clinical results declined. The most common complication was axillary nerve dysaethesias, seen in 14% of the patients. Patients who underwent open revision had a success rate of only 50% with the secondary procedure. In contrast, Savoie and Field (26) reported satisfactory results in 28 of 30 multidirectional instability patients treated with thermal capsulorrhaphy and adjunctive interval closure followed for an average of 26 months.

Levy et al. (27) noted a 36.1% failure rate at a mean follow-up of 40 months after laser capsular shrinkage, compared to a 23.7% failure rate at a mean follow-up of 23 months after shrinkage with a radiofrequency device. Noonan et al. (28) followed 59 patients after laser thermal treatment (some patients also had a suture anchor repair of a labral lesion); they observed a 4% failure rate after treatment of unidirectional instability compared to a 39% failure rate with multidirectional laxity. This report is concerning because unidirectional laxity is the best condition for treatment with current suture anchor and plication methods, whereas multidirectional laxity continues to be associated with the least predictable outcomes. In a very discouraging report, Miniaci and McBirnie (29) describe clinical failure in 9/19 instability patients treated by thermal techniques, with stiffness observed in 5 patients and neurologic symptoms noted in 4 patients.

Overhead athletes may be a specific subset of patients who may benefit from limited capsular shrinkage. Dugas and Andrews (30) reported on 170 overhead athletes treated by standard shoulder arthroscopy and debridement with addition of capsular shrinkage and a well-defined postoperative rehabilitation protocol. They noted that 81% returned to competition, and felt that, in their hands, addition of thermal capsular shrinkage resulted in about a 20% improvement in the rate of return to play. Follow-up averaged 30 months in this study. In a prior report by the same authors (31), however, heat shrinkage was not felt to improve the rate of return to play in baseball players, although the authors thought that treatment of internal impingement with thermal capsulorrhaphy might decrease the re-injury rate after return to sports.

Another clinical situation that may be reasonable for application of thermal tissue modification involves augmentation in cases of labral repair, such as SLAP repair or Bankart repair (32,33,34) utilized thermal capsular treatment as an adjunct to suture anchor stabilization in 42 patients with recurrent traumatic anterior instability. Recurrent instability occurred in 3 patients (7%). Although these authors report that thermal treatment was an important adjunct, it should be emphasized that comparable results are achievable with suture anchor placation methods alone. The selection criteria for adjunctive thermal treatment are not well-defined, and use of this approach as during suture anchor stabilization is at best left to the subjective judgment of the surgeon at the time of arthroscopic evaluation of capsular laxity.

Failure of treatment is typically thought of as recurrent instability, with a subset of patients requiring a subsequent stabilization procedure. Destruction of the capsule by thermal treatment can be extremely difficult to treat, and this condition may require tissue transfer or grafting procedures to reconstruct the deficient capsule (35,36). McFarland et al. (37) did not observe a specific histologic feature that could explain capsular deficiency in 7 patients who underwent biopsy after failed thermal treatment. Denuded synovium was observed in all of these patients, and 71% had changes in the collagen including “hyalinization.”

However, postoperative stiffness can also occur, and this complication may require additional surgical intervention as well (36). Gerber and Warner (3) believe that the major complication of thermal capsulorrhaphy is postoperative stiffness, with an incidence between 1% and 10% in their review.

Axillary nerve injury is an infrequent but important complication that has been associated with thermal capsular shrinkage (3,38,36). The axillary nerve is vulnerable due to its proximity to the capsule in the axillary recess, and heat application should be well-controlled in this region to minimize the risk of neurologic injury (39,40). McCarty et al. (41) noted a significant increase in temperature along the course of the axillary nerve in 11 of 15 cadaveric specimens during thermal application to the capsule; the branch to the teres minor appeared to be most vulnerable during this procedure. In a clinical follow-up study, however, shoulder proprioception returned to baseline when evaluated 6 months to 24 months after thermal treatment in twenty patients (42). These results suggest that normal neurologic function can be anticipated unless there is a catastrophic injury along the axillary nerve’s ramifications.

Rotator cuff repair sutures can be conveniently separated into monofilament or braided sutures and absorbable or nonabsorbable sutures (Table 20-2). An advantage of monofilament suture is the relative ease of tissue passing, particularly with devices that utilize hollow needles, because the inherent stiffness of the suture facilitates “pushing” or “wheeling” the suture within the tool; however, that stiffness comes with a price in terms of knot tying, because knots don’t lay
down easily with monofilaments, and they have a greater tendency for knot slippage (62). In contrast, braided sutures tend to be quite pliable and are generally relatively strong compared to similar size monofilaments. The introduction of antegrade suture passing tools has made management of braided sutures a bit more convenient (these instruments grab the suture on a needle prior to tendon puncture).

Absorbable sutures are advantageous in the long term because they minimize residual knots that can cause mechanical impingement/irritation in the subacromial space. However, mechanical integrity must be maintained at least until the strength of the healing response is sufficient to withstand physiologic loads. Rapid hydrolysis can lead to an inflammatory response, which in some cases can be confused with an early postoperative infection. In the case of a rotator cuff repair, it is unlikely that the bone-tendon interface is particularly strong before 6 weeks, and it probably takes 12 weeks or more for the bone tendon junction to partially mature. Bourne et al. (63) noted significant mechanical degradation between 2 and 6 weeks after in-vivo placement with various resorbable suture materials (Table 20-2). Similar findings were noted by Greenwald et al. (64), with the important observation the PDS suture was only about 10% to 15% as strong after 6 weeks compared to time zero in-vivo. Boileau et al. (65) recently reported a 71% healing rate with arthroscopic cuff repair using predominantly PDS sutures, suggesting that sufficient early mechanical integrity was maintained for cuff healing using this resorbable suture construct. In another recent study, Boehm et al. (66) found no significant difference in clinical outcome or re-tear rate (about 20%) after cuff repair with either Ethibond or PDS suture. At this time, there is no clearly optimal suture material for rotator cuff repair. The basic principle, however, holds that we must stabilize the cuff with suture material that is strong enough and lasts long enough to provide mechanical stability until adequate tissue healing is achieved.