|
|
||||||||
Sign In to gain access to subscriptions and/or personal tools. |
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
,*


From
Advanced Joint Care and Orthopedic Sports Medicine, Glendale, Arizona, and
Center of Orthopedic and Rehabilitation Excellence, West Jordan, Utah
* Address correspondence to Brent P. Hansen, DO, Advanced Joint Care and Orthopedic Sports Medicine, 4344 West Bell Road, Suite 102, Glendale, AZ 85308 (e-mail: orthoboy1do{at}yahoo.com).
| ABSTRACT |
|---|
|
|
|---|
Hypothesis: A cause of postarthroscopic glenohumeral chondrolysis will be associated with a specific factor (ie, implanted device, surgical technique, etc), and this factor can be identified by a review and comparison of cases seen in the senior authors office.
Study Design: Case series; Level of evidence, 4.
Methods: Analyze possible etiologic factors with imaging studies, demographics, history, and physical examinations of 10 patients (12 shoulders) with postarthroscopic glenohumeral chondrolysis, and then compare perisurgical information with a focused chart review and comparison with the rest of the 177 arthroscopic shoulder surgeries in the same period of time.
Results: There were 12 cases of postarthroscopic glenohumeral chondrolysis (all were the senior authors patients). Four common factors were identified, and only high-flow intra-articular pain pump catheters filled with bupivacaine and epinephrine were a new addition to years of shoulder surgery by the senior author; 177 shoulders underwent arthroscopy in the identified time frame, and only 19 shoulders, of 30 with capsular procedures, had intra-articular pain pump catheters filled with bupivacaine and epinephrine. Of these, 12 have been identified with chondrolysis.
Conclusion: Use of intra-articular pain pump catheters eluting bupivacaine with epinephrine appear highly associated with postarthroscopic glenohumeral chondrolysis.
Clinical Relevance: Intra-articular pain pump catheters, especially those eluting bupivacaine with epinephrine, should be avoided until further investigation.
Key Words: chondrolysis arthroscopy shoulder pain pump bupivacaine epinephrine
| INTRODUCTION |
|---|
|
|
|---|
The authors have encountered 12 cases of chondrolysis after arthroscopic shoulder stabilization procedures. The purpose of this retrospective case review is 2-fold: first, to report our experience with this condition, and, second, to attempt to discern its origin by looking at factors common to these cases to identify present and future patients at risk for developing this condition.
| METHODS |
|---|
|
|
|---|
|
|
|
|
| RESULTS |
|---|
|
|
|---|
The 12 shoulders with PAGCL were from the group of 30 arthroscopic shoulder stabilization procedures. They were identified based on our criteria of increased pain, stiffness, and crepitus (with or without decreased range of motion) and loss of the glenohumeral joint space as shown on radiographs. All 12 shoulders had been treated with arthroscopic stabilization for symptomatic instability. Table 3
shows the operative histories of the 12 affected patients.
|
The procedures included 5 capsular shifts, 6 anterior Bankart repairs, 3 posterior Bankart repairs, and 1 superior labral anterior posterior repair. Five shoulders had some combination of these procedures. Three shoulders had cartilage lesions; all were grade 2 or less at the index arthroscopic procedure (Table 2
).
All surgeries but 1 in the chondrolysis group were done with the patient in the lateral decubitus position. Of those shoulders with lateral decubitus positioning, traction time averaged 84 minutes (range, 26–150 minutes) with a range of 5 to 14 lb distal and 5 to 17 lb proximal. Glycine lavage was used in 10 involved shoulders (for improved conduction if there was a need for a thermal probe) and Ringers lactate in 2.
A thermal radiofrequency probe was used in 4 shoulders in the chondrolysis group to augment the capsular shift effect of a labral repair. In the nonchondrolysis group, thermal probes were used intra-articularly in 51 cases. Various sutures and anchors were used in these procedures without a consistent frequency between types (Table 2
). Anchors were made of absorbable polylactic acid (Depuy Mitek, Raynham, Mass) and poly(L-Lactide-co-D, L-Lactide) (Arthrex, Naples, Fla). Two patients had only suture placed (no anchor). One patient in the chondrolysis group had a broken anchor inserter tip noted on postoperative radiographs after the index procedure. This was retrieved at a second procedure after chondrolysis was diagnosed. The metal fragment measured 3 x 1 mm and was encapsulated within the synovium at the time of retrieval.
A preoperative and postoperative intra-articular injection was given in all 12 shoulders consisting of 25 mL of 0.25% bupivacaine HCL (BH) with epinephrine and 5 mg of morphine sulfate without preservative. Ketorolac 30 mg was added to the postoperative injection in 9 of the 12 shoulders that went on to chondrolysis.
All affected shoulders had postsurgical intra-articular pain pump catheters (Stryker Instruments, Kalamazoo, Mich) placed in the shoulder with a drip rate of 4.16 mL/h and filled with 250 mL of 0.25% BH with epinephrine. The patients were told to remove the pain pump catheter from their shoulders 2 days after surgery.
All patients had previously made significant gains in motion after surgery and appeared to be recovering as expected initially, after a period of immobilization lasting 4 weeks. They were noted in the chart review to have progressed with motion and function with physical therapy during postoperative weeks 5 to 12 and subsequently were prescribed a home/gym exercise program. Postoperative clinic notes indicated good clinical and functional return of glenohumeral stability in all patients. All affected patients then were seen with complaints of new onset of pain, stiffness, increased pain with motion, and crepitus within the first year after surgery (range, 3–12 months). Each patient had radiographic evidence of glenohumeral joint space narrowing at the time of diagnosis. Active motion was limited in nearly all patients because of pain and was significantly decreased when compared with the office visits before the onset of their symptoms. Although active motion often decreased because of patient guarding, most patients could obtain full passive range of motion at each follow-up office examination; crepitus was a common feature. Mean onset of symptoms from the time of the index procedure was 4.3 months (range, 3–13 months). Mean age at surgery was 28.9 years (range, 16–47 years).
Rheumatology and serology study results were normal in all chondrolysis patients. Needle aspiration findings of all affected shoulders were negative for an infectious process in all but 1 case. One shoulder culture finding was positive for Propionibacterium sp, which grew on 1 of 3 agar plates and was believed to be a contaminant by the hospitals microbiology department. No further evidence of infection developed subsequently in any patient, and none were treated with antibiotics other than perioperatively. Operative specimens from those patients returning to surgery to treat their chondrolysis included excised humeral heads in 3 patients, which demonstrated cystic changes with giant cells, lymphocytes, and eosinophils, without necrotic bone, and were read by the pathology department as end-stage osteoarthritis. No evidence of osteonecrosis was noted on either microscopy or MRI (Table 1
).
At the time of this data collection, 10 patients (10 shoulders, 83%) had been returned to the operating room for at least 1 additional procedure after diagnosis of PAGCL (Table 3
).
In addition, further surgery is contemplated in 3 shoulders, 2 having failed to improve by 6 months postoperatively after debridement and microfracture. At last follow-up (range, 5–25 months), all patients continued to have some complaints of disability and pain despite partial improvement after these treatments.
On chart review of the 152 patients treated in this same time frame, another 104 patients were treated with the same preoperative and postoperative injection and pain pump techniques described above but with the catheters placed extra-articularly (superficial to the rotator cuff tendons). None of these patients experienced chondrolysis based on our chart review. Of the 30 arthroscopic stabilization procedures, 19 shoulders in 17 patients were given intra-articular pain pump catheters, of which 12 have developed chondrolysis thus far (63%). An additional 13 patients were treated with arthroscopic stabilization but no pain pump catheters, and none of these patients have since developed symptoms related to PAGCL, as confirmed by continued clinical follow-up through the first postoperative year.
| DISCUSSION |
|---|
|
|
|---|
Thermal radiofrequency energy was used in 4 of 13 cases presented here. Although suspected as an etiologic factor by some authors,9,10,13,19,20 it does not appear to be clearly proven to be the only factor in these cases.
Loose metallic hardware (a broken inserter tip from a suture anchor) was seen in only 1 of our cases. There is limited available literature suggesting a secondary, nonmechanical, immunogenic process that may further erode the hyaline cartilage when a mechanical process, such as pin protrusion, has occurred.2,27 The metallic fragment was incorporated within the synovium of this patient and is not suspected to have contributed to chondrolysis in this patient or series.
The use of various suture and anchor types without a consistent pattern seems to preclude their implication as an etiologic factor (Table 2
).
No known association of chondrolysis with traction type or time was identified in the literature, nor could we implicate those factors in this series. Glycine lavage was used in all but 2 cases in this series. Ringers lactate was used in the other 2 cases. There is no known adverse effect of Ringers lactate5 or glycine lavage on articular cartilage. In 1987, Bert4 looked at glycine lavage when used with electrocautery and found no evidence of synovial damage or histopathology when chondrocyte function was evaluated with safranin O staining. Glycine has been associated with hyponatremia7 and transient blindness3 after arthroscopy. The senior author has used glycine in the vast majority of shoulder and knee procedures he has performed over the past 18 years and has had no prior identifiable cases of chondrolysis. Therefore, we do not believe that traction type, time, or fluids were contributory.
Aside from the same surgeon performing all of the procedures on the patients (C.L.B.), the 1 consistent factor in every patient with chondrolysis is that all of the patients received intra-articular pain pump catheters that eluted 0.25% bupivacaine and epinephrine, and all were arthroscopic capsular procedures. The bupivacaine used in these patients was from several different vendors used by the hospital where the procedures were done.
Bupivacaine is similar to lidocaine but is slightly more lipid soluble and binds protein to a greater extent.18 It is typically medically administered in the operating room as methyl paraben–free BH to avoid the potential allergic reaction possible from methyl paraben, a potent antiseptic preservative.24 According to the package insert, BH has a half-life of about 2.7 hours in adults and 8.1 hours in neonates; these are increased when epinephrine is added (owing to capillary constriction of epinephrine and delayed local resorption). The pH of methyl paraben–free BH with epinephrine is between 3.3 and 5.5. Stomach acid can be higher than pH 4.0 in conditions such as achlorhydria.11 Although BH is considered a "weak" acid, it can lean toward the acidic side of "weak" acids at pH 3.3. Epinephrine may have played a role in this series. It has been shown to increase neurotoxicity when added to lidocaine.12
Nole et al25 looked at 35SO4 incorporation into articular cartilage after intra-articular bupivacaine injection in pigs and dogs. Recovery of 35SO4 incorporation at 1 to 3 days was seen, and therefore no contraindication to intra-articular bupivacaine was found. Dogan et al8 looked at the histopathologic affect of intra-articular bupivacaine in rabbits and found articular inflammatory cell infiltration and synovial hypertrophy and hyperplasia when specimens were examined microscopically on days 1, 2, and 10 after the initial injection. These results were compared with those of a saline control group in a blinded fashion. Day 10 showed the most significant inflammatory changes from bupivacaine. These studies looked at a single injection of bupivacaine without epinephrine. No one has yet examined and reported, to the authors knowledge, on the effects of a constant infusion of BH, either at 2.08 mL/h (as in the pain pump catheters available and used by the surgeon before this series) or at a 4.16 mL/h rate (as used in the current series), for 24 to 72 hours on hyaline cartilage in vivo. This rate is eluted within the glenohumeral joint, which is a confined space with a limited capacity. This space is reduced further after capsulorrhaphy procedures.
Bupivacaine with epinephrine contains sodium metabisulfite and can cause an allergic reaction in the sulfa-allergic patient (no patients had known sulfa allergies). Possibly there is a secondary link to inflammation and chondrolysis. It is known that inflammatory processes have been implicated in destructive arthritis, as this has been experimentally evaluated.21,29 Ketorolac injections have been shown to induce an inflammatory response in both the synovium and hyaline cartilage of animals.17 We doubt it played a significant role in these chondrolysis patients as it was only used in 9 of the 12 shoulders, and there are no cases of chondrolysis in the 177 comparison patients, of whom 82 had this Ketorolac added to their postoperative injections.
When glucose is added to bupivacaine in regional anesthesia, it appears that less volume is needed to obtain a desired anesthetic affect, and speed of onset is increased.3,7,30 A hyperosmolar effect (that may irritate nerves with spinal anesthesia) may have occurred with bupivacaine or in combination with the glycine used in all but 2 patients.7,28 This may induce a cartilage dehydration, which can lead to chondrocyte degradation as with the effect of dimethyl sulfoxide on the water content of a chondrocyte matrix.23
We did not evaluate fibronectin release, although it is implicated in chondrocyte cellular breakdown by proteoglycan suppression.1,14–16,22 Cartilage destruction is likely a combination of mechanical and chemical factors.
Most patients initial symptoms of pain and progressive motion loss occurred by 3 to 5 months after surgery. Radiographic breakdown was typically seen by 5 or 6 months. It is suspected by the authors that the insult to the articular cartilage that led to chondrolysis occurred in the intraoperative and perioperative period. The onset of symptoms and signs is suspected to have been delayed for a few months because of postoperative protected mobilization and lack of significant compressive forces seen by the damaged hyaline cartilage. The hyaline cartilage was thus relatively protected until after 3 months after surgery, at which time increasing forces from more aggressive activities and resistance exercises caused mechanical degradation of the compromised surfaces. Whatever the mechanism may be, the senior author has not seen chondrolysis in arthroscopic shoulder patients before the use of this high-flow pain pump catheter with bupivacaine and epinephrine. It should be noted that 27 of the 177 patients had the high-flow pain pump catheters used in the bursal side, and none have had reported chondrolysis to date.
We believe this information is important to disseminate to the orthopaedic community at large. A recent publication from Chu et al6 showed that 0.5% bupivacaine was indeed cytotoxic to bovine cartilage.
The incidence of chondrolysis in this series is startlingly high, with 12 shoulders out of 177 total shoulder surgeries (6.8%). Of those who received arthroscopic capsular stabilization with postoperative intra-articular pain pump catheters with bupivacaine and epinephrine, 63% developed chondrolysis. The authors have had no other known patients with chondrolysis before or since the dates of this study.
| CONCLUSION |
|---|
|
|
|---|
| FOOTNOTES |
|---|
| REFERENCES |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
J. L. Dragoo, T. Korotkova, R. Kanwar, and B. Wood The Effect of Local Anesthetics Administered Via Pain Pump on Chondrocyte Viability Am. J. Sports Med., August 1, 2008; 36(8): 1484 - 1488. [Abstract] [Full Text] [PDF] |
||||
![]() |
C. R. Chu, N. J. Izzo, C. H. Coyle, N. E. Papas, and A. Logar The in vitro effects of bupivacaine on articular chondrocytes J Bone Joint Surg Br, June 1, 2008; 90-B(6): 814 - 820. [Abstract] [Full Text] [PDF] |
||||
![]() |
P. E. Greis, A. LeGrand, and R. T. Burks Bilateral Shoulder Chondrolysis Following Arthroscopy. A Report of Two Cases J. Bone Joint Surg. Am., June 1, 2008; 90(6): 1338 - 1344. [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | CONTACT US | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |