Issue: April 2014
April 01, 2014
14 min read

Panel discusses use of reverse total shoulder arthroplasty for proximal humeral fractures

Issue: April 2014
You've successfully added to your alerts. You will receive an email when new content is published.

Click Here to Manage Email Alerts

We were unable to process your request. Please try again later. If you continue to have this issue please contact

Management of displaced three- and four-part proximal humeral fractures continues to challenge surgeons. Advocates of nonoperative management, percutaneous pinning, open reduction and internal fixation and hemiarthroplasty all support their technique in certain clinical scenarios. Historically, hemiarthroplasty was considered the “gold standard” for displaced four-part proximal humeral fractures in the elderly patient. While pain relief was reliable, active motion was less predictable and sometimes catastrophic if the tuberosities failed to heal. Reverse total shoulder replacement has emerged as a viable option in the elderly patient with complex displaced four-part proximal humeral fractures.

I would like to thank our esteemed panel, which includes international and national leaders in the field of shoulder arthroplasty, all of whom have extensive experience with reverse total shoulder replacement.

William N. Levine, MD

Roundtable Participants

  • William N. Levine
  • Moderator

  • William N. Levine, MD
  • New Tork
  • Pascal Boileu
  • Pascal Boileau, MD
  • Paris
  • Mark A. Frankle
  • Mark A. Frankle, MD
  • Tampa
  • Christian Gerber
  • Christian Gerber, MD, FRCSed
  • Zurich
  • Guido Marra
  • Guido Marra, MS
  • Chicago
  • Gilles Walch, MD
  • Lyon, France
  • Gerald Williams
  • Gerald R. Williams Jr., MD
  • Philadelphia

William N. Levine, MD: What are your indications for performing a humeral head replacement (HHR) with tuberosity repair for proximal humeral fractures?

Mark A. Frankle, MD: Much of the decision-making comes down to patient age and the quality of the greater tuberosity. Thin or comminuted greater tuberosity fractures in older patients have lower healing rates. In a younger patient with a good quality greater tuberosity, I will use a hemiarthroplasty when the fracture is not amenable to fixation. This has become a rare clinical scenario and I have not performed this in the past 5 years.

Christian Gerber, MD, FRCSed: My indications for using HHR with tuberosity repair are in proximal humerus fractures with an avascular head that cannot be reduced anatomically and stabilized sufficiently to allow undisturbed healing. Prerequisites include: intact glenoid, intact cuff, good bone quality and a mental state that allows for collaboration in the postoperative course.

Guido Marra, MD: The choice of performing hemiarthroplasty with tuberosity repair in patients with complex proximal humeral fractures is based on four factors: fracture morphology; bone quality; patient’s physiologic age; and the patient’s functional requirements.

During the past 2 decades, there has been a significant change in our understanding of the pathoanatomic variants of complex humeral fracture amenable to fracture fixation. Our improved knowledge has led to a refinement of fractures managed with hemiarthroplasty. In my practice, three-part fractures and valgus impacted four-part fractures will generally undergo fracture fixation. True four-part fractures and four-part fracture dislocations will undergo shoulder arthroplasty (hemiarthroplasty vs. reverse total shoulder arthroplasty [RTSA]) based on the additional criteria.

Patient physiologic age and functional requirements are intertwined and are an important consideration in today’s society due to the variability of health and activity level in patients who sustain these fractures. In considering these factors, one must take into account associated medical comorbidities, which impact a patient’s functional needs or effectively decrease their functional requirements. While patient function after hemiarthroplasty is less predictable than RTSA, the ceiling of a well-functioning hemiarthroplasty is higher than a well-functioning RTSA. With this in mind, I reserve hemiarthroplasty for patients with higher functional demands and with fewer medical comorbidities.

Finally, when bone quality prevents adequate greater tuberosity reconstruction while performing an open reduction and internal fixation of a complex proximal humeral fracture then conversion to a hemiarthroplasty with tuberosity reconstruction is performed.

Gilles Walch, MD, and Pascal Boileau, MD: Our indications are patients with fracture dislocation (with complete head fragment dislocation) between 50 years and 65 years.

Gerald R. Williams, Jr., MD: My indications for HHR include a motivated patient willing to do the extensive rehabilitation who is younger than 70 years (or older if physiologically young) with a three-part fracture that is not fixable or four-part fracture without pre-existing cuff tear and a normal glenoid. In properly selected patients in whom the tuberosities heal in an anatomic or near anatomic position, the results are comparable to RTSA and may even be better with regard to internal and external rotation.

Levine: What is your age limit for performing a RTSA for a displaced three- or four-part proximal humeral fracture?


Frankle: There is no real age limit, as physiologic age and ability for the greater tuberosity to heal are the critical variables to consider. Comorbidities and appearance of bone are more influential in the decision-making process. Decision-making is often made preoperatively. Rarely, in younger patients, intraoperative decision-making can occur. If fixation is inadequate to allow for motion without displacement, or if adequate reduction cannot be achieved, then RTSA can be used as a salvage procedure in these specific circumstances.

Gerber: There is no real age limit. A mental disability or other factors not compatible with patient collaboration in the postoperative period would define an indication for RTSA. Bone quality is, however, rarely poor enough to determine in favor of RTSA in patients younger than 65 years. On the other hand, for patients older than 75 years, I favor RTSA and the results are typically outstanding in this cohort.

Marra: I have no specific age limit, rather I look at the patient’s functional requirements and expectations when considering options for fracture management. For higher functional demand patients, the preferred treatment strategy would be to perform fracture fixation. However, if fracture morphology prevents adequate fixation due to technical or biologic consideration, then I select hemiarthroplasty in the higher demand patient and RSTA for the low demand to sedentary patient population. In a younger age group with high energy injuries which preclude fracture fixation, RTSA can only be considered in the patient who understands the long-term consequences and the restriction it would impose on their functional requirements after implantation. This is a minority subset of these patients.

Walch and Boileau: There is no age limit. RTSA is indicated if there is no other satisfactory option.

Williams: Although patients of the same age can vary considerably from a physiologic perspective, I generally do not use RTSA for patients younger than 70 years, except in select circumstances (physiologically older, not motivated for rehab for hemiarthroplasty or pre-existing cuff tear).

Levine: What are your ‘top five pearls’ to ensure success of RTSA?

Frankle: Here are my top seven pearls:

Match the implant to patient size. Use the humeral head and glenoid sizes to select the proper implants for each patient. My goal is to deviate as little as possible from the patient’s normal anatomical relationships.

Use the calcar to restore humeral height. Rest the medial portion of the humeral shell on the fractured calcar.

Use the patient’s forearm to achieve proper version. I prefer 30° of retroversion to the forearm and the system I prefer has a sighting guide, which allows simplicity in positioning the implant.

To avoid over-lengthening of the arm, place the glenosphere in the center of the glenoid. Also use the patient’s native glenoid version to dictate the version of the glenosphere.

Place all greater tuberosity sutures before implanting the stem. Ensure that the sutures have been passed through the humerus and stem prior to implanting the stem. This facilitates suture management. In addition, when repairing the tuberosities, start by securing the greater tuberosity to the stem in its anatomical position.

Use the black and tan technique. When cementing the stem, compress morselized bone from the humeral head into the humeral shaft, proximal to the cement, before the stem is implanted. This will create a bone interface between cement and the tuberosities, helping avoid cement induced bone necrosis around the area of tuberosity healing.

Use fluoroscopy. Confirm there are no unaccounted bony fragments visible. This can help to avoid heterotopic ossification, a troublesome complication following RTSA for fracture.

Gerber: My top five pearls are:

Height of the stem and size and position of the glenoid prosthesis should ensure arm lengthening of approximately 2 cm to 2.5 cm.


Rotation of the humeral prosthesis should be no more than 20° of retrotorsion for easier healing of greater tuberosity.

In large patients, larger diameter glenospheres are preferred to ensure stability.

The supraspinatus and infraspinatus should not be resected but left on the greater tuberosity, which is repaired to the stem and the shaft using direct and circumferential sutures.

Reaming of the glenoid should be avoided. Removing the cartilage on the glenoid should suffice not to weaken the subchondral plate.

Marra: My top five pearls are:

Patient education. This is particularly true in patients who sustain complex proximal humerus fractures. Providing the patient with a detailed understanding of the surgical procedure and how RTSA works empowers them to assist in their rehabilitation and to protect their replacement long term. Unlike patients with arthritis who do extensive research prior to physician consultation, fracture patients are less informed about RTSA and its inherent benefits and downsides. In the absence of surgeon-directed education, patients will turn to online information, which often reflects the use of RTSA in arthritis.

Use intraoperative imaging to assist in tuberosity reconstruction. The use of an intraoperative image intensifier assists in tuberosity positioning and ensures precise reduction which is a critical component of postoperative function. While tuberosity position is more important in hemiarthroplasty than RTSA, I use an image intensifier in both scenarios to confirm my reductions prior to committing my fixation.

Obtain radiographs early and often. When managing patients postoperatively I obtain radiographs at 1 week, 2 weeks, 4 weeks and 6 weeks to confirm maintenance of the tuberosity reconstruction. Tuberosity position has been identified as important to functional outcomes, so it needs to be carefully monitored in the early postoperative period.

Beware of patients using the operative arm to push or pull while standing up. This is part of patient education, as many patients fail to recognize that, while they may be compliant in their use of a sling, the sling does not prevent them from detrimentally weight-bearing on the operative extremity. This is frequently seen when a patient is getting up from a seated position, as using the operative arm is instinctive.

Communicate with the therapist. Making contact with the therapist involved in the rehabilitation of fracture patients is important as the distinction of the indication for arthroplasty (fracture) must be communicated. Verbal communication of limitations minimizes the rehabilitation risks and assists in continued patient education in the early postoperative period.

Walch and Boileau: Our top five pearls:

Superior transdeltoid approach (with or without osteotomy of the anterior acromion) makes it easier to manage the tuberosities and place the glenoid implant.

Baseplate must be flush to the inferior rim and use a 42-mm inferior eccentric sphere in men and a 36-mm inferior eccentric sphere in women to avoid scapular notching.

Height of the prosthesis is determined by reduction of the greater tuberosity fragment on the diaphysis and the prosthesis: the prosthesis must be at the level or slightly above the reduced greater tuberosity. Retroversion is determined with alignment of the outer jig with the forearm equal to 20° of retroversion.

Separate fixation of greater tuberosity and lesser tuberosity, starting with the greater tuberosity. Greater tuberosity reduction on the lateral part of the prosthesis with the arm in external rotation. Use a standardized suturing technique with six sutures — “Sixtec” (four horizontal cerclages and two vertical tension-band sutures). Use looped sutures with sliding-locking knot.

No more use of an internal rotation sling but instead use a neutral rotation brace. The patient should do immediate pendulum exercises and have no formal rehabilitation with physiotherapist before 4 weeks.

Williams: My top 5 pearls are:

Restore arm length. Avoid shortening as well as over-lengthening.

For tuberosity reduction and healing, use a similar horizontal and vertical suture pattern as used in hemiarthroplasty.

Identify and protect the axillary nerve.

Use minimal cement. This can be accomplished with distal cement and proximal bone ingrowth.

Use bone graft tuberosities.


Levine: You are all involved with prosthetic design teams. Please outline the prosthetic design features you believe are important to increase the success of RTSA in the management of proximal humeral fractures.

Frankle: Tuberosity healing has shown to be important in optimizing outcomes for RTSA in the acute fracture setting. Prosthetic designs that help to position the greater tuberosity in a more anatomic position are important. For this reason, I use an implant with an anatomic neck shaft angle (135°) and a lateralized glenosphere. This combination allows reduction of the greater tuberosity to a more anatomic position. Reconstruction of the glenoid-tuberosity offset and acromial-tuberosity distance places minimal tension on the tuberosity fragments.

Most patients who undergo RTSA for fracture are women who tend to have smaller humeral heads and glenoids. Because of this, I avoid larger diameter spheres and I place the glenopshere in the center of the glenoid to avoid over-tensioning the shoulder. Inferior glenosphere placement and larger spheres both increase acromial tuberosity distance. One of the devastating complications that can occur with over-lengthening of the arm is brachial plexopathy. This can occur when a large implant is placed in a small patient or when the arm is excessively lengthened. Awareness of patients’ humeral head and glenoid sizes can be helpful to select appropriately sized implants. The prosthesis I use for fracture treatment has fins, suture holes and a textured surface around the area of tuberosity contact, all which aid in tuberosity repair and enhance tuberosity healing.

Gerber: It should be possible to fix the tuberosities to the stem and the prosthesis in a near anatomical fashion. As opposed to hemiarthroplasty, a substantial metaphyseal volume is not desired. The tuberosities do heal more frequently due to medialization of the stem relative to the scapula. The shoulder should not be forced to go through excessive excursions, however, to allow at least a tenodesis effect of external rotators.

Marra: When considering the design of a RTSA for fracture, its parameters should be cognizant of the unique environment seen in proximal humeral fractures. Humeral implant design should allow for a biologic environment that promotes fracture healing. This can be accomplished with three strategies: 1) minimize metallic interfaces, which can inhibit fracture healing; 2) use surfaces with biologic ingrowth properties or surface roughness; and 3) provide conduits for fracture fixation.

When considering the first parameter, humeral implant for reverse arthroplasty has traditionally had a large proximal geometry with smooth metallic surfaces. This is detrimental in the context of fracture fixation as the geometry of classic proximal humeral reverse components often required bone removal to adequately reconstruct the fractured tuberosities. Bone removal not only minimized areas for healing, but also risked devascularization of the tuberosities. In addition, low friction surfaces and spherical geometry increase micromotion at fracture sites following fractures fixation, increasing the likelihood of nonunion or loss of tuberosity fixation. High friction surfaces and high friction surfaces with the capability of ingrowth or on-growth are ideal in the context of fracture fixation, as they minimize micromotion at the fracture interfaces. Finally, the humeral implant needs to serve as an effective conduit for fracture fixation.

Walch and Boileau: A low profile prosthesis allows anatomical placement of the greater tuberosity and additional bone graft harvested from the humeral head. The prosthetic design should include a smooth and polished medial neck to avoid breakage of suture cerclages.

In addition, we favor the use of instrumentation that allows accurate prosthetic placement with respect to height and retroversion.

Williams: Even with RTSA, tuberosity reduction and healing are important. Therefore, some of the most important design features relate to ease in reduction and fixation of the tuberosities. These features include proximal stem shape that allows accurate placement of the tuberosities and bone grafting, more than one epiphyseal size, and a surface treatment that allows or encourages bone ingrowth. In addition, to minimize instability, intraoperative flexibility with regard to varying arm length (spacers, multiple liner thicknesses) is important. Cementless fixation is attractive when possible. Therefore, stems that are designed to press fit into the metaphysis, preferably with proximal porous coating, can be helpful. Glenoid spheres with multiple radii of curvature are essential.


Levine: Please outline your postoperative rehabilitation program for RTSA following proximal humeral fractures. Does the protocol differ dependent on the etiology? Does rehabilitation differ if you perform a HHR with tuberosity repair?

Frankle: The rehabilitation protocol is similar to primary RTSA. Shoulder motion is restricted to pendulum exercises for the initial 6 weeks postoperatively. At 6 weeks, the patients are instructed to initiate active-assisted shoulder range of motion (ROM), advancing motion as tolerated. This is done as a home-based program without the supervision of a therapist.

Gerber: The rehabilitation is extremely simple. Patients are put in a sling and start to use their arm with pendulum exercises or small movements with the arm at the side (reading newspaper, turning pages or using hand with elbow on a table). They wear a sling at night for 6 weeks and during the day for 4 weeks. After 4 weeks, they use their arm at their discretion without any weights. Humeral head replacement with reconstructions of tuberosities is treated with postoperative immobilization in a neutral position in a sling for 6 weeks.

Marra: Rehabilitation for RTSA following fracture in the initial phase focuses on passive motion while protecting the tuberosity reconstruction. In general, I begin therapy for passive motion within safe limits determined intra-operatively, typically 90° of forward flexion and external rotation to 20° to 30°. I restrict internal rotation stretching for 6 weeks. No strengthening or assistive terminal stretching performed in this initial phase. Radiographs are obtained at 1 week, 2 weeks, 4 weeks and 6 weeks to monitor the tuberosity reconstruction. With confirmation of tuberosity union, immobilization is discontinued at 6 weeks. Active-assisted and active motion is initiated. At 8 weeks, a strengthening program is begun.

My rehabilitation program for RTSA for fracture is different than for patients who undergo RTSA for arthritis, as it needs to be respectful of the fracture fixation of the tuberosities. Immobilization is longer in the fracture patient and early follow-up occurs more frequently.

Rehabilitation for humeral head arthroplasty is focused on obtaining passive shoulder flexibility in the initial 6 weeks. My follow-up for hemiarthroplasty patient is similar. I will be more aggressive in increasing limits of therapy based on the clinical and radiographic follow-up.

Walch and Boileau: Our rehabilitation program is to protect tuberosities repair. This involves:

  • Use a neutral rotation brace for 3 weeks to 6 weeks as tolerated.
  • Avoid active ROM for 6 weeks.
  • Start passive ROM at day 2 with pendulum exercises, self-mobilization and hydrotherapy.
  • For the tuberosity repair’s protection, the rehabilitation is a little slower after fracture than after elective surgery.

Our protocol is the same after HHR with tuberosity repair.

Williams: In general, I have found rehabilitation to be less important with RTSA for fracture than hemiarthroplasty for fracture. Following RTSA for fracture, the patient is kept in a sling with only elbow, wrist and hand motion for 10 days to 2 weeks. At that time, pendulum exercises, passive external rotation to 30° and passive supine elevation to 130° are instituted — usually as a home-based program. At 6 weeks, an overhead pulley and active ROM below shoulder height is instituted — again usually with a home-based program. At 12 weeks, full active ROM is encouraged and a home-based theraband strengthening program is added. At 6 months, the patient is released to activities as tolerated. Obviously, this progression is dependent on progressive radiographic union without tuberosity displacement.

With most other diagnoses, the rehabilitation protocol following RTSA is advanced more rapidly. I typically keep the patient in a sling for 10 days to 14 days at which time I encourage them to use their arm for daily activities below shoulder height and begin pendulum or table gliding exercises. At 4 weeks to 6 weeks I add an overhead pulley and allow the patient to use their arm for all daily activities — even those overhead. If they have less than 120° of passive elevation, supervised therapy is added. At 3 months, I allow patients to begin doing all activities as tolerated.


In my experience, rehabilitation following hemiarthroplasty for fracture is more important and requires dedication on the part of the patient and often requires 1 month to 2 months of formalized therapy in an outpatient setting. This requires secure and anatomic fixation of the tuberosities. Pendulum exercises, passive external rotation to 30° and supine passive elevation to 130° are started in the hospital and continued at home by the patient. At 10 days to 2 weeks, the range of supine passive elevation is increased to 160° and passive external rotation is increased to 40°. If the patient has been diligent with the exercises and successful in maintaining the ROM from hospital discharge, these exercises are continued at home. If not, therapy is moved to a supervised, outpatient setting. At 6 weeks postoperatively, manual stretching and an overhead pulley are added and most patients are moved to a supervised setting at this time. At 12 weeks, active and active assisted ROM along with theraband strengthening exercises are added. At 4 months to 6 months postoperatively, patients are typically moved to a home-based program. They will continue to improve for 12 months to 15 months postoperatively. Again, all of these progressions are dependent on progressive tuberosity healing without displacement.

For more information:
Pascal Boileau, MD, can be reached at Hopital de l’Archet, Department of Orthopaedic Surgery BP 79, 151 Rte de St Antoine de Ginestiere, 06202 Nice, France.
Mark A. Frankle, MD, can be reached at Florida Orthopaedic Institute, 13020 Telecom Parkway North, Tampa, FL 33637.
Christian Gerber, MD, FRCSed, can be reached at Balgrist University Hospital, Department of Orthopedic Surgery, Forchstrasse 340, CH 8008 Zurich, Switzerland; email:
William N. Levine, MD, can be reached at Center for Shoulder, Elbow and Sports Medicine, Columbia University, 622 W. 168th St. PH1117, New York, NY 10032; email:
Guido Marra, MD, can be reached at Northwestern University Feinberg School of Medicine, 676 N. St. Clair St., Suite 1350, Chicago, IL 60611.
Gilles Walch, MD, can be reached at Centre Orthopedique Santy, 24 Avenue Paul Santy, 69008 Lyon, France.
Gerald R. Williams, Jr. MD, can be reached at The Rothman Institute, 925 Chestnut St., 5th Floor, Philadelphia, PA 19107.
Disclosures: Boileau receives royalties from Tornier, is a paid consultant for Smith & Nephew and receives financial or material support from Mitek; Frankle receives royalties from Tornier, is a paid consultant for Smith & Nephew and receives financial or material support from Mitek; Gerber receives royalties and is on the speakers bureau for Zimmer, is a paid consultant for Storz, and received research support from Medacta; Levine is an unpaid consultant and on the design team for Zimmer’s Next Generation Shoulder system; Marra is a consultant for Zimmer and on the board for the American Shoulder and Elbow Surgeons; Walch receives royalties from Tornier and receives financial or material support from Imascap and Tornier; Williams receives royalties from, is on the speakers bureau and is a paid consultant for DePuy and IMD, is an unpaid consultant for Checkpoint, IMDS, has stock or stock options in CrossCurrent Business Analytics, Force Therapeutics, ForMD, In Vivo Therapeutics and OBERD, and receives research support from DePuy, Synthasome and Tornier.