This article is more than 5 years old. Information may no longer be current.
Shoulder replacements are an evolving technology
Click Here to Manage Email Alerts
It is an honor to serve as moderator for this Round Table discussion on shoulder arthroplasty. I have invited international leaders in the field to participate, all of whom have been instrumental not only in teaching a multitude of new surgeons their techniques but also in developing new implants that have made a dramatic impact on our care for patients requiring shoulder arthroplasty. We hope you enjoy this discussion.
William N. Levine, MD
Moderator
| ||||||||
William N. Levine, MD
Stephen A. Copeland, FRCS
Gilles Walch, MD
Evan L. Flatow, MD 
Gerald R. Williams Jr., MDWilliam N. Levine, MD: Shoulder arthroplasty has evolved dramatically over the last decade or so. Please give your perspective on how you have changed your management and/or decision-making during this time?
Stephen A. Copeland, FRCS: I have been doing surface replacement for the last 20 years so it has not made a dramatic difference to my practice within the last 10. However, during the last 10 years I have definitely gone more towards hemiarthroplasty than total arthroplasty, because of the results.
In the beginning, as in hip and knee arthroplasty, I wanted to do total replacement. It seemed obvious to do a total shoulder replacement. However in some patients, a total was not possible, ie, those with major cuff tears, or rheumatoid arthritis (RA) with gross erosion on the glenoid side. In long-term follow-ups of these, hemi-arthroplasties appeared to do extraordinarily well and, in fact, as we started to do more hemi than total, there was no statistically significant difference between the two.
The results of surface hemiarthroplasty appear to be better than stemmed hemiarthroplasty. In hemi surface replacement, it is very difficult to get the geometry wrong. The head is correctly placed on the original humeral head and, hence, normal anatomy is mimicked from rotation, varus and offset.
Because surface replacement is used for arthritis and not for fracture, the principles behind it are different in that we try to make up bony loss by erosion through the prosthesis’ thickness, whereas for fractures the head is removed and the same thickness of bone replaced. With a stemmed hemiarthroplasty, the eroded head is removed and the same thickness of head is put in, which leaves the geometry distorted. This could explain the better results of surface hemi than stemmed hemi. Also in the longer-term, glenoid component wear in the totals was becoming a problem after 10 years, which corresponded to other series. Hence if we are trying to avoid revision surgery, not performing a glenoid replacement with present designs and materials seems like a good idea.
Glenoid failure appears to be a materials problem; better materials are needed and better designs of glenoid are required and I am sure this will be the major thrust of development over the next 10 years.
The introduction of the reverse-type prosthesis has not changed my practice radically, but I do use it on rare occasions. Because of the high complication rate and poor longer-term results with the present designs of reverse prosthesis, I only use this for cuff tear arthropathy in patients 75 years old or older. It is now in wide use in Europe in some revision and fracture situations, but the results are poor and a revision reverse prosthesis is an unrewarding procedure. However, there is no doubt that with the constrained geometry imposed by the reverse prosthesis, the potential for regaining good range of motion in these dysfunctional shoulders has been excellent.
We have, therefore, developed a different type of reverse prosthesis (Verso) to overcome the mechanical problems of present designs and look forward to obtaining long-term follow-up data in the future.
Evan L. Flatow, MD: Since I started practice 20 years ago, the most dramatic change has been the development of reverse total shoulder arthroplasty (TSA). This has given us an option we did not previously have, allowing us to offer functional improvement to patients with pseudo-paralysis and pain from cuff-deficient arthritis and after failed prior surgery. However it has also raised the stakes, as there are much higher rates of complications, especially infection, after reverse arthroplasty.
My decision-making in cuff-intact osteoarthritis (OA) has not changed much. Better sizing of implants and instruments have improved the precision of arthroplasty, but our earlier functional results are not really dramatically better than in the days of the Neer prosthesis, at least at specialized centers. However modern instruments and implants have made this procedure more predictable for the occasional shoulder surgeon, and the better restoration of anatomy and mechanics may help improve durability. For example, current cement pressurization techniques have virtually eliminated early glenoid lucent lines, which we thought were part of the procedure before.
There has also been a revolution in anesthesia, as interscalene regional blocks have made the need for endotracheal intubation rare except in revision cases. Finally the need for replacement in proximal humeral fractures has decreased, as minimally-invasive, percutaneous techniques and locking plates have expanded the role of internal fixation. However when humeral head replacement is needed, newer fracture-specific prostheses, better techniques to restore height, and better materials have made results, especially tuberosity healing, more predictable.
Gilles Walch, MD: I have gradually changed my practice toward an increased use of the reverse prosthesis. (see graph) These changes were motivated by the poor results I observed in some indications with nonconstrained shoulder arthroplasties. I regularly perform (either individually or through multicenter projects) follow-up studies of all patients following shoulder arthroplasty. From these retrospective studies, I learned that the results of nonconstrained prostheses were not satisfactory: for any kind of degenerative or inflammatory joint diseases when the cuff was not functioning well; in cases of static humeral instabilities, primary or secondary arthritis with large or massive cuff tears or superior subluxation; cuff tear arthropathy; and RA.
In addition, I now recommend reverse TSA in primary glenohumeral (GH) arthritis in three specific circumstances: 1) static posterior subluxation with bi-concave glenoid (I have observed failures of noncontrained prostheses performed concomitantly with glenoid reconstruction); 2) fatty infiltration of infraspinatus and/or subscapularis < stage 2; and 3) associated ruptures of both the supra- and infraspinatus. Also the functional results of nonconstrained prostheses were poor for sequelae of fracture or post-traumatic arthritis in cases of nonunion or severe malunion of the greater tuberosity, and I switched to the reverse prosthesis in those cases if the patient was older than 70 years. For the same reasons, three- or four-part fractures in the elderly patients with poor bone quality and very thin greater tuberosities were progressively treated with a reverse prosthesis > 80 years old at the beginning and now >75 years.
I also use the reverse prosthesis for tumors necessitating tuberosity resection to achieve better function. This is only for a return to gentle activities of daily living: no sports, no strenuous activities. The patients are told that if they overuse the prosthesis, a resection arthroplasty or a standard hemiarthroplasty would be the only solution left, and results of these salvage procedures would be the same as if we started with these options.
Reverse arthroplasty has also gradually become my first option for revision of humeral head replacement when the cuff was not functioning well. In cases of revision TSA, when a glenoid reconstruction with bone graft was necessary, the reverse is also the best solution because the metallic noncemented base plate allows a safer and better reconstruction than a cemented polyethylene component; the base plate acts as a plate to stabilize the graft even if the cuff is still there.
In the beginning, I summarized my indications for a reverse when the nonconstrained prosthesis was not able to give satisfactory results. The good results I have witnessed over the years have motivated me to gradually expand the indications, as in cases of arthritis with severe static posterior instability and severe glenoid wear (biconcave glenoid). On the other hand, the fears I had about the long-term results prevent me from using the reverse in younger patients or when there was another satisfactory option. We have learned from our long-term studies (Favard, Guerry, Molé), that the functional results of the reverse do deteriorate, after 7 to 10 years, and this is why we do not propose using it in patients younger than 70 years old, especially in post-traumatic arthritis.
Gerald R. Williams Jr., MD: Several factors have influenced my management and decision-making in shoulder arthroplasty over the last 10-15 years. These include decreasing patient age, increasing patient activity level, glenoid component durability, postoperative subscapularis dysfunction, and dissatisfaction with functional outcomes following hemiarthroplasty in certain patients with cuff deficiency.
|
Potential shoulder arthroplasty patients seem to be getting progressively younger while older patients are remaining active longer. Our implants, therefore, are required to function for longer periods of time under potentially greater stresses. Moreover, glenoid component durability remains a potentially limiting factor, despite the fact that glenoid resurfacing has been associated with better pain relief than hemiarthroplasty alone in most reported series. Therefore, I have moved toward using glenoid components whenever the glenoid surface is pathologic, the cuff is intact, and adequate glenoid bone exists. As a general rule, however, patients who are under 50 years old or involved in vigorous activity such as weightlifting or heavy manual labor do not receive a polyethylene glenoid component.
Biologic glenoid resurfacing has become increasingly popular in patients thought to be too young or active for a polyethylene glenoid component. However, biologic glenoid resurfacing (with or without humeral hemiarthroplasty) can be more technically demanding than prosthetic glenoid resurfacing. In addition, commonly used biologic surfaces such as meniscal allograft, dermal allograft, and others likely share some of the same durability issues as polyethylene.
Humeral hemiarthroplasty with a concentric glenoid, good range of motion, and an intact cuff may provide good results for 10-15 years. Therefore, I use biologic glenoid resurfacing in patients with glenoid disease who are too young to reach their 50s with a hemiarthroplasty alone. Concentricity of the native glenoid surface remains an important prerequisite for a good result, even in biologic resurfacing.
The prevalence of subscapularis dysfunction following shoulder arthroplasty is higher than expected and recently has been shown to correlate with worse results. The development of subscapularis deficiency is obviously multifactorial. However, one common factor is partial or complete subscapularis repair failure.
Subscapularis deficiency is less common when a lesser tuberosity osteotomy is used to reflect the subscapularis than when the tendon itself is divided and repaired. Subscapularis lengthening or advancement is limited with the osteotomy approach. However, unless the subscapularis has been shortened in a previous surgical procedure or loss of external rotation is extreme, perimuscular and peritendinous releases are usually sufficient to restore adequate subscapularis length. Therefore, I perform a lesser tuberosity osteotomy to reflect the subscapularis unless external rotation is less than zero or the subscapularis attachment to the lesser tuberosity is not completely intact. When an osteotomy is not used, it is important to repair the tendon securely and to protect it postoperatively.
Humeral hemiarthroplasty provides predictable pain relief in most patients with glenohumeral arthritis associated with irreparable cuff deficiency. However, even when appropriate tendon transfers are added, hemiarthroplasty results in poor function in many of these patients, especially when anterosuperior dislocation is present. The reverse arthroplasty designed by Paul Grammont in the mid- to late- ’80s provides reliable increase in function and decrease in pain in this subset of patients. In fact, in some patients the postoperative results are spectacular.
The reverse arthroplasty has become an important tool in my armamentarium. However, our enthusiasm for this design should be tempered by its increased complication and reoperation rates compared to anatomically designed implants. As a general rule, reverse arthroplasty in my practice is reserved for patients older than 70 years with sedentary activity levels. It should also be remembered that the reverse arthroplasty is only approved for use in patients with irreparable cuff deficiency and arthritis.
Levine: Drs. Copeland, Flatow and Williams – Dr. Walch has indicated that he would recommend a reverse TSA in a 60-year-old patient with a biconcave (B-2 Walch classification) glenoid and posterior humeral subluxation. Would you agree?
Copeland: Absolutely not! I would not do a reverse in a 60-year-old patient. I do not consider straightforward OA with an intact cuff as an indication for reverse at any age.
Flatow: I would not. I would do a “conventional” TSR such a patient.
Williams: I would not, in part because it is an off-label use. I am sure Gilles is referring to the severe cases. I personally would perform a bone graft and anatomic component given the current state of implants available, etc. I think the reason Gilles is saying this is because of the relative ease of bone grafting using the reverse.
 |  |
 | Pre-operative anteroposterior radiograph (1a) of a 73-year-old woman with recurrent glenohumeral bloody effusions, humeral head flattening, glenohumeral arthritis and an irreparable cuff tear. Prior to surgery, she was unable to raise her arm more than 45°. Post-operative anteroposterior radiographs (1b, 1c) reveal a reverse arthroplasty in place. Clinical evaluation at 6 months postoperatively revealed restoration of overhead elevation. Images: Flatow EL |
Levine: Please describe your routine preoperative evaluation for patients whom you are scheduling arthroplasty surgery?
Copeland: The routine preoperative evaluation obviously depends on the indication, ie, OA, RA, post instability, avascular necrosis, etc. But in general terms, the indication is pain and the clinical evaluation includes full range of motion, assessment of strength of the rotator cuff, deltoid and complete neurovascular examination of the upper limb. I rely heavily on plain radiographs and do not do other imaging studies routinely. However, if there is any doubt about the geometry of joint or adequacy of bone, then CT scan, can be very helpful for evaluating the remaining bone stock. If they have not compensated for cuff loss and they are very elderly, I would be thinking in terms of a reverse prosthesis. If they have had previous surgery, I will work them up to exclude a possible infection.
Flatow: It is important to understand the onset of pain; it may be gradual as in OA, or have come on after a fracture or trauma, which may suggest more scarring and surgical difficulty. The location of the pain may be helpful, as not all radiographic arthritis is symptomatic.
Deep anterior and/or posterior shoulder pain is consistent with glenohumeral disease, but an unusual pattern may suggest another diagnosis, for example a pattern of posterior trapezius pain radiating down the back of the arm to the hand along with paresthesias may raise the possibility that a cervical condition is involved.
It is vital to obtain prior treatment records, especially the reports of any prior surgery. I always question the patient about any signs of infection in that shoulder or elsewhere in the body when the symptoms began, as one can never be too careful in searching for indolent infection. Finally it is vital to determine if the patient’s chief complaint is pain or loss of function. They often say both, but usually one is more important. While both are typically improved in OA, there are many situations (especially revision cases) where functional gains may be difficult to predict, and patient expectations must be reasonable.
I also try to assess patient activity level and the demands of their occupation, especially as to whether they are willing and able to restrict heavy, repetitive and impact use which might accelerate implant failure.
On physical examination I document motion loss, and look for weakness or active lag signs that might demonstrate cuff loss or nerve damage. It is important, especially in those having undergone prior surgery, to look for erythema, swelling, or enlarged lymph nodes that might suggest indolent infection. We always obtain a true AP and lateral X-ray of the scapula and an axillary view. While I used to say that a CT-scan best demonstrated bony morphology, current MR-technology is such that glenoid bone loss and version changes can be seen and excellent information about the cuff status and muscle quality is provided. It can be helpful to get a baseline scan to monitor any progressive glenoid bone loss for patients who want to know if there is any harm in delaying replacement. In revision situations, we may obtain infection work-ups (ESR, CBC and CRP blood tests along with an indium-WBC scan or other such study) or EMGs, but these are not routinely performed.
Walch: All of my patients are evaluated carefully preoperatively and we obtain a constant score and document active range of motion. I usually do not perform a reverse if the patient has normal or near-normal active forward elevation, even though it is painful.
I obtain plain radiographs on all patients including AP views in external, neutral and internal rotation and an axillary view all performed under fluoroscopic control. We also obtain an arthroCT in all patients. I look for the quality of the rotator cuff (muscle and tendon), glenoid morphology, and static anterior or posterior instability of the humeral head.
Williams: In primary shoulder arthroplasty, important aspects of preoperative evaluation include history, physical examination, plain radiographs, and three-dimensional studies. Important historical findings include the severity and duration of symptoms, the types of treatments that have been tried and failed, any prior surgical procedures, the presence of comorbidities, prior deep vein thrombosis or pulmonary embolus, prior infection, patient age and activity level, and patient postoperative expectations.
Important physical findings include the degree of passive range of motion loss (especially external rotation), the presence of rotator cuff weakness, evidence of dynamic (eg, anterosuperior escape) or static (eg, posterior subluxation) instability, and neurovascular integrity. Rotator cuff weakness can often be inferred by positive lag signs or by weakness to manual resistance. However, in cases of severe stiffness, rotator cuff evaluation can be very difficult. Special attention should be afforded to the acromioclavicular joint, as symptomatic acromioclavicular joint arthropathy can spoil an otherwise outstanding result.
Plain radiographs should be obtained on all potential shoulder arthroplasty patients. Routine radiographs used in our practice include anteroposterior views in the scapular plane in internal and external rotation and an axillary view. These radiographs can help confirm the diagnosis of GH arthritis, suggest rotator cuff deficiency if superior migration is present, and roughly gauge the degree of bone deformity on both the humeral and glenoid sides. The axillary view is especially helpful in identifying posterior glenoid wear and subluxation associated with osteoarthritis. Radiographs may also be used for prosthetic templating.
I obtain a three-dimensional study on almost all patients undergoing shoulder arthroplasty. Possible exceptions include patients with presumably normal glenoids (ie, avascular necrosis) and excellent axillary views demonstrating no posterior wear. In patients with OA or other patient populations thought to have normal or reparable rotator cuffs, a CT scan is best. In patients with RA, an MRI will show rotator cuff thinning or tearing, as well as bone deformity. Contrast is not routinely used. Either of these three-dimensional modalities allows accurate quantification of bone deformity, which can be useful in preoperative planning.
Levine: How would you manage a 42-year-old right hand dominant (RHD) male who has end-stage degenerative joint disease (DJD) with a Walch A2 (concentric, flat) glenoid? Would you ever consider arthroscopic treatment in this type patient? Would you consider hemiarthroplasty? If you perform interpositional biologic glenoid arthroplasty, what is your implant of choice?
|
Copeland: For this patient my procedure of choice would obviously depend on the severity of the arthrosis. If this was minimal, I would attempt arthroscopic debridement and abrasion of the glenoid. If arthritis and symptoms were more severe, I would choose surface replacement hemiarthroplasty. I would avoid using a glenoid in a patient of 42 years old as this prosthesis probably needs to last 40 years and none of the glenoids at present are showing signs of lasting that long. When doing the hemiarthroplasty, I do not do a biological interposition but instead perform a microfracture on the glenoid side. I do not perform concentric reaming as I believe this increases pain for the patient and encourages erosion and medialization.
Flatow: My experience with arthroscopy has been mixed and often disappointing, but if there are signs of other sources of pain beyond just the glenohumeral arthritis (eg, AC joint pain, cuff/bursal symptoms, loose bodies) or if there is stiffness out of proportion to the bone deformity (especially if there was a prior tight instability repair) then it may occasionally be of benefit.
Otherwise, I would either manage the patient nonoperatively or perform an arthroplasty. If the patient is a young 40 with a demanding job or athletic tendencies, I would consider a hemiarthroplasty with an interpositional biologic arthroplasty, usually a lateral meniscal allograft placed after reaming the glenoid to normal version and raw bone. I tend to use a resurfacing to avoid fractures at the tip of a stem and to save bone, but it does make glenoid exposure more difficult and the advantages may be more theoretical than real.
If the patient is more sedentary, I’d prefer to implant a total shoulder since the results are far more predictable. We have developed a polyethylene glenoid backed with a low-modulus ingrowth tantalum material which seems to preserve bone should revision become necessary. This is exactly the type of patient for whom it may be helpful, but longer follow-up is needed to evaluate this.
Walch: I would propose a hemiarthroplasty and do not consider any kind of biologic resurfacing. Biologic resurfacing does not prevent progressive erosion of the glenoid and medialization of the humeral head.
Williams: My decision is based on patient activity, the degree of stiffness, and the degree of bone deformity on the humeral side. If this patient was involved in weightlifting or heavy manual labor, I would consider arthroscopic management unless the humerus had large osteophytes or was severely flattened.
Without substantial humeral deformity, I would perform an arthroscopic debridement and subacromial bursectomy, as there is some evidence that subacromial bursectomy may improve the results of arthroscopic debridement. I would perform a capsular release if stiffness were substantial (< 30° of external rotation, <120° of elevation, <10° of internal rotation with the arm at 90° of elevation). I would not attempt to alter the concavity of the glenoid surface because I am not confident in my ability to make it perfectly concentric arthroscopically.
Alternatively, if the head had large osteophytes or was flattened, I would perform a humeral hemiarthroplasty with concentric reaming of the glenoid surface. If the glenoid surface was concave, rather than flat, I would likely perform a humeral resurfacing based on the theoretical possibility that one could preserve more bone for arthrodesis, should one be required.
Levine: Would your answers change to the above scenario if the patient is 52? How about 62?
Copeland: My answers would be the same for a patient of 52 and 62.
|
Image: Copeland SA |
Flatow: I would almost always lean toward total replacement as the age gets above 50, except in unusual cases. Above 60 I would prefer a standard cemented polyethylene glenoid, given the superb track record.
Walch: Same indication
Williams: At 52 and 62, I would likely recommend a total shoulder arthroplasty, unless he were a weightlifter or heavy manual laborer who could not or would not agree to give them up. Under those circumstances, I would do a standard hemiarthroplasty with concentric reaming of the glenoid surface.
Levine: How would you manage a 73-year-old RHD female with end-stage cuff tear arthropathy of the left shoulder who has active forward elevation of 150° but severe pain and “cannot live with it any longer?”
Copeland: This patient is obviously well-compensated and I would do a surface replacement arthroplasty. We now have an extended articular surface (EAS) prosthesis which would be ideal for this patient as a hemiarthroplasty. (See Figure 2.)
Flatow: I would of course first try a cortisone shot, and if an MRI showed an intact biceps I might consider an arthroscopic debridement with a biceps tenotomy. Once all other options are exhausted, an arthroplasty becomes the only option, and the choice is difficult in this scenario. In younger patients with good but painful elevation we still feel a hemiarthroplasty is good, and I would use an implant which can be converted to a reverse without the need for stem revision. At 73, I might consider a reverse arthroplasty even though she is not yet pseudoparalytic, to try to avoid the need for a second procedure.
Walch: If the patient still has good active forward elevation, I do not propose a reverse except if there is an anterosuperior cuff tear involving the subscapularis. In these patients, subcoracoid impingement is often observed and they have painful, slow active forward elevation. Pectoralis major transfer does not give satisfactory results and I do not know of any soft tissue procedure which is able to re-center the humeral head in the horizontal plate. If the subscapularis is present and functioning, I propose a debridement with biceps tenotomy or tenodesis for pain relief.
Williams: I would perform a hemiarthroplasty and would add an extended head if the anatomy did not require an offset humeral head or resection of any remaining cuff to insert it.
Levine: How would you manage the same patient who has 30° of active forward elevation?
Copeland: At 73 years old she is still rather young for a reverse prosthesis. The average life expectancy at 73 is 12 years in a European. There is no doubt in the short term this lady would do better in range of motion. Function would be better with a reverse prosthesis but with a high chance of failure within her life span. Depending on the patient`s activities, I may still do a surface replacement and accept reasonable pain relief but no elevation above shoulder height and limited function. If the patient was unstable, ie, had forward subluxation of the head on resisted external rotation, surface replacement is unlikely to be helpful and a reverse prosthesis would be indicated.
Flatow: I would do a reverse arthroplasty and not consider a hemiarthroplasty.
Walch: Reverse, of course! I would consider an associated latissimus dorsi transfer (as described by Pascale Boileau) if there is no active external rotation and complete disappearance of the teres minor.
Williams: I would perform a reverse arthroplasty, assuming her inability to raise the arm was thought to be the result of weakness or instability, not pain. I would inject her joint with a local anesthetic. If she could still not raise her arm above 90° or control it on the way down after I passively elevated it, I would do a reverse.
|
Image: Flatow EA |
Levine: Finally, you all have developed prostheses for different companies – please provide a situation when you might consider using one of the other prostheses from a different company?
Copeland: Having developed something specifically for arthritis of the shoulder, I still require a prosthesis for proximal humeral fractures and therefore would be happy to use any of the 3rd generation-type stemmed prostheses. As I am unhappy with present designs of reverse prostheses we are using, a new design for this indication. If the patient had a pure punched out defect in the humeral head this is a possible indication for the hemi cap partial humeral head replacement, although I haven`t had a patient yet with this specific indication.
Flatow: Although I helped to design the Bigliani-Flatow prosthesis, the Trabecular Metal prosthesis, and the Trabecular Metal Reverse (all Zimmer) (Figure 3), I will use other designs. For young patients, I will use a resurfacing, and have often used Steve Copeland’s design from Biomet as well as the CAP design from DePuy. For smaller lesions I think the Arthrosurface mini-resurfacing, designed in part by Tony Miniaci, is a good option. Finally, I do a good deal of revision surgery and if a stem is in reasonable position, I will often just exchange the head and consider a glenoid implant from that specific system.
Walch: When I perform revision surgery, I do not hesitate to use the original prosthesis to avoid difficult stem removal. In cases of focal defects of the humeral head, I consider using the Arthrosurface resurfacing prosthesis. Otherwise the company Tornier is able to provide any kind of prosthesis I need (including “custom made” versions for tumors).
Williams: Before answering this question, I would like to emphasize that the most important variable that the surgeon has control over is surgical technique. As Dr. Matsen is known to say, “The surgeon is the method.” Learning the techniques of soft-tissue handling, humeral and glenoid exposure, and adequate subscapularis repair is far more important than selecting the implant.
The 73-year-old patient that was mentioned above presents one potential situation where I would consider using a prosthesis that would not require removing the stem to convert from a hemiarthroplasty to a reverse.
Levine: I would like to thank all of you for participating in this stimulating virtual round table discussion and thank you for sharing your wealth of experience and knowledge in shoulder arthroplasty.
For more information:
- Stephen A. Copeland, FRCS, can be reached at Reading Shoulder Unit, Berkshire Independent Hospital Reading, Berkshire, England; +44 118 9028116; e-mail: stephen.copeland@btinternet.com.
- Evan A. Flatow, MD, can be reached at 5 E. 98th St., Box 1188, New York, NY 10029; 212-241-1663; e-mail: evan.flatow@msnyuhealth.org.
- William N. Levine, MD, can be reached at 622 W 168th St, PH-11, New York, NY 10032; 212-305-0762; e-mail: wnl1@columbia.edu.
- Gilles Walch, MD, can be reached at Centre Orthopédique SANTY, 24 Ave. P, Santy, Lyon France; +33 427 530 054; walch.gilles@wanadoo.fr.
- Gerald R. Williams Jr., MD, can be reached at 1 Cupp Pavilion-Presbyterian Hospital, 39th and Market Street, Philadelphia, PA 19104; 215-349-8734; fax: 215-614-0450; e-mail: grwjr@comcast.net.