8.3.1 Anesthesia and Patient Positioning

We favor general anesthesia with the use of an interscalene block for pain management with hypotensive measures to decrease the risk of bleeding. Additionally, we use tranexamic acid intravenously or at the end of the procedure, depending on patients’ characteristics.

We place our patients in the beach chair position. The scapula must be accessible, and putting the patient very lateral on the table or using a specific shoulder surgical table can achieve this. We use a static adjustable stand to place the arm, but pneumatic specific arm holders or a simple Mayo stand can be used. Changing the arm position during the procedure as well as the height or inclination of the surgical table might be helpful.

8.3.2 Approach

In case of revision surgery, most patients will benefit from a long deltopectoral approach. Ideally, the incision will be placed lateral to the conjoined tendon, but we generally try to incorporate the previous skin incision. If sinus tracts are present, the whole sinus tract is removed. Staining the sinus tract with methylene blue might be helpful to determine the extent of the tract. A long deltopectoral approach benefits from access to the distal insertion of the deltoid and humerus and can be extended anterolaterally to expose the whole humerus if necessary. In cases of distal extension, it is safer to dissect and control the radial nerve at the beginning of the procedure. A proximal extension can be performed up to the clavicle for resection of the distal clavicle, which can serve as bony autograft if needed.

We tend to leave the cephalic vein in the proximal part of the interval so it can guide the dissection, should revision surgery be necessary. Dissecting the cephalic vein and protecting it are probably best for the patient to avoid hand swelling. Finding the deltopectoral interval can be difficult in revision cases. To be most efficient in cases where the vein is not found, or the deltopectoral interval is not evident, one can palpate the acromioclavicular joint and slide medially 2–3 cm, which marks the medial end of the deltoid muscle and then progress distally toward the deltoid insertion. Where to place the vein is controversial. Our preference is to place the vein medial in primary cases, which requires ligation of the vessels feeding the deltoid. In revision cases our preference is to place the vein laterally just in case there is a need to extend the dissection distally so it does not get in the way. A Hohman retractor is placed over the coracoid and under the distal part of the deltoid to provide initial tension for further subdeltoid dissection.

Subdeltoid release of adhesions is usual, and we perform this by placing the arm in abduction to decrease the tension of the deltoid, and we then place the arm in progressive internal rotation until we have freed all adhesions. We must be careful to protect the axillary nerve at the lower part of the inner deltoid as we are dissecting these adhesions.

By using a long deltopectoral approach, we can perform a limited dissection of the distal insertion of the deltoid if we need access to the proximal part of the humerus, which is rather safe and does not compromise postoperative function. If we feel the deltoid is very scarred and rigid and can suffer during the approach, it is probably safer to extend the approach proximally and detach the deltoid from the clavicle (anteromedial approach) (Fig. 8.1) [4]. The detachment is performed from medial to lateral and requires periosteal dissection to preserve the deltoid fascia. On occasion releasing only the first 2 cm of this extensive approach is enough to decrease tension in the deltoid. A secure transosseous fixation and use of an abduction brace postoperatively are recommended to achieve healing of the deltoid. Other reasons to use this extended deltoid include the realization of a frail deltoid, to minimize the stress of an osteopenic humerus during the approach, and to gain better access to expose the glenoid or the rotator cuff.

The next step is to perform subcoracoid dissection, as it is frequent to find adhesions that can limit the range of motion in external rotation. Careful dissection is necessary as excessive traction might injure the axillary nerve or the brachial plexus and extension medial to the coracoid are not advised.

How to proceed from here depends on the cause of failure and our treatment strategy. If we are planning a revision to an anatomical total shoulder, delicate handling of the rotator cuff is warranted. We will then perform a tenotomy of the subscapularis. If the long head of the biceps is still present, we tenodese it at the proximal part of the groove so it can serve as an adjunct for later repair of the subscapularis. Performing an adequate release of the subscapularis is important for tension-free repair although it is compromised. Releasing adhesions between the posterior part of the subscapularis and the anterior glenoid and capsule is critical not only to obtain the maximum range of motion but to gain control of the anterior wall of the glenoid vault which can be useful for intraoperative orientation for glenoid instrumentation.

We then go on to release the proximal humerus until we can perform a safe external rotation of the arthroplasty to gain access to the head of the arthroplasty. In cases of pre-existing stiffness, external rotation is performed with caution; otherwise we can produce a fracture. If the approach to the humerus is compromised, performing an anteromedial approach is justified. Most systems have instruments to disengage the humeral head from the stem, but in case of finding a non-modular stem, we attempt to disimpact it from underneath with the use of a bolt and a mallet, which has to be done with care because there is a risk of fracturing the greater tuberosity. Once the head is removed, we gain more access to complete the capsulotomy in the humerus. At this point, there is enough space to work around the glenoid to achieve the necessary capsulotomy.

8.3.3 Glenoid Component

Failure of the glenoid implant due to loosening is the most frequent cause for revision of glenoid failure after total shoulder arthroplasty (TSA) and is typically associated with osteolysis and bone loss. Bone loss can be due to primary pre-existing asymmetric wear, aseptic loosening as just mentioned, and unintended bone loss when performing the explantation of the prior glenoid. As outlined by Antuña et al., these defects can be contained or uncontained and classified in terms of the quantity of the defect in mild, moderate, or severe [5]. In 17 out of 48 shoulders, the authors found associated instability. While radiologic lucent lines around the glenoid component are frequent, the definition of glenoid failure is somewhat equivocal. Most patients can have glenoid implant mobilization and be relatively asymptomatic. The decision to operate on a mobilized glenoid is based on clinical symptoms, radiologic signs, desires and volition of the patient, and possible procedures and outcomes.

If the soft tissue are in good condition, glenoid revision can be attempted with a new glenoid implant. Implanting a new glenoid component achieves better pain relief than not revising the implant [5]. However, associated bone loss may preclude insertion at the same surgical procedure and can be deferred until graft consolidation. Reaming the remaining glenoid to achieve a congruous surface and trying to center the humeral head on the glenoid are the goals of the procedure although bone grafting is preferable, as it rebuilds glenoid bone stock and can improve our chances to implant a new glenoid component.

If the surrounding tissues are not in perfect condition, revision to a reverse shoulder implant is probably preferable. In cases without glenoid loss, revision to a reverse shoulder arthroplasty is a straightforward procedure. We use a saw to cut the polyethylene in sectors, and we use a rongeur to take them out. We then use a reamer to prepare the glenoid surface, and the pegs can be overdrilled. Implantation of the metaglene is then carried out (Fig. 8.2).

If there is a bone defect of less than 2 mm and the shoulder offset is restored by the implant, no bone graft is needed. In cases of a peripheral glenoid defect with a stable implant and more than 50% contact in native bone, structural autograft or allograft can be used to reconstruct the defect. Cases with more than 50% of a peripheral defect may need a structural autograft form the iliac crest. In cases of a central defect with more than 30% contact, morselized autograft or allograft can be used to improve contact. If the contact of the implant with native bone is <30%, iliac crest autograft, distal clavicle, or allograft can be used [6]. The use of a reverse design produces significant forces on the glenoid, and the graft can be performed as a single or a staged procedure; first, the graft is fixed with screws, and on a second procedure after graft healing, the baseplate is implanted (Fig. 8.3) [7].

8.3.4 Stem Exchange

Stem exchange can be a very challenging part of the procedure. Platform systems may significantly reduce the time and burden of a revision stem procedure. While the rate of stem survival at 10 years is greater than 90% and the risk of loosening is circa 5%, the rate of stem exchange during surgery approaches 50% during revision shoulder arthroplasty [8]. The reasons for these are unclear, but version, height, and access to the glenoid can all precipitate the revision. The reasons for exchanging the stem must be compelling because it is related to an increased amount of intraoperative complications including cement extrusion, intraoperative humeral shaft fracture, and tuberosity fractures.

In a cemented stem, we extract the humeral stem by impacting it from distal to proximal into the collar. If a collar is absent, we can perform a notch with a high-speed burr in the medial aspect of the stem and can then vertically impact into this notch from south to north. A best-case scenario features the prostheses coming out with the whole mantle of cement. If the stem doesn’t come out, a longitudinal humerus split for the length of the prostheses is performed through the anterolateral cortex and the cement mantle typically reaching to the tip of the stem. Very carefully, osteotomes are used to crack this split open, and the stem is freed from the cement mantle and extracted carefully. However, when the stem is fully textured, it is probably safer to perform an anterior humeral window. The extent of the window starts proximally below the subscapularis insertion and extends to the end of the prostheses. A 1-cm-wide osteotomy including the insertion of the pectoralis major is favored. We prefer to drill holes at the corners to dissipate any energy created during the extraction to minimize the risk of fracture. A saw is used to connect the corners, and the window is elevated very carefully with the sequential use of osteotomes [9, 10].

When cement is retained, we must weigh the benefit of extracting all the cement and placing a longer stem, which requires time and effort and can compromise remaining stock against the possibility of implanting a shorter stem with a cement-in-cement technique. This technique has been proved safe and reliable at short follow-up in a group of patients at the Mayo Clinic [11].

Removing a well-fixed uncemented stem can be a grueling experience depending on the texturing and the extent of the texturing of the stem. More modern stems are porous coated only proximally, and this will facilitate removal. Otherwise, the need for an osteotomy dramatically increases. The first step is to disrupt the interface between the host bone and the prostheses while minimizing bone stock loss. We favor using a small burr and a saw or flexible osteotomies circumferentially in the proximal part of the humerus. We then try to extract the humerus with the aid of the stem’s specific extractor or a fish-mouth clamp and a mallet. If this is unsuccessful, a longitudinal osteotomy is performed with the use of a saw form the proximal part to the tip of the stem, and it is then pried opening carefully with the use of osteotomes (Fig. 8.4). After the stem has been extracted, the osteotomy is closed and secured with the aid of wire cerclage, and surgery is continued to insert another humeral component. Attention to correct version, height of the entire arm, and relationship with the greater tuberosity is advised to provide for the best outcome. Adequate reconstruction is based on the relative length of the deltoid to maximize stability and function.

Most typically bone loss will be minimal but will facilitate ease of extraction of the humeral implant. If there is a residual cement mantle, ideally we should remove the cement mantle, but this can be a time-consuming and challenging process, and there is the risk of producing humeral fractures. A hip cement extraction set is helpful in these situations. Some surgeons are fond of ultrasound to help remove cement, but there is a risk of injuring the radial nerve. If the bone-cement mantle is stable, one option is to roughen the cement mantle and cement a new humeral component. This “cement-in-cement technique” has proven successful at midterm follow-up. Wagner et al. reported on the use of this technique in revision surgery in 38 patients [11]. Specifically looking at the performance of the new implant, there was a new revision surgery in three patients at 3.7 years of follow-up. One patient was revised for humeral loosening, one patient for instability and glenoid loosening, and one for a periprosthetic humeral fracture, giving an overall implant revision-free survival at 2 years and 5 years of 95% and 91%, respectively. Additionally, there was one “at-risk” humeral component with moderate subsidence that was not revised at final follow-up and two additional grade 3 humeral lucencies without subsidence. The authors reported seven (18%) non-displaced greater tuberosity fractures that happened in every case during stem removal.

Bone loss is typically seen proximally and depending on the extent, and quality of residual bone loss can be managed by neglecting it with or without exchanging the stem or by reconstructing it with an allograft-prosthesis composite (APC) reconstruction or a tumor prostheses. Typically, defects of less than 5 cm with good-quality residual bone can be safely revised with a new humeral stem, while defects greater than 5 cm and bad-quality bone are revised with an APC technique.

Budge et al. in a series of 15 patients with an average proximal humeral bone loss of 38 mm (range, 26–72 mm) treated with modular implants without the use of allografts reported 87% of patient satisfaction with the procedure with improved function. They did not see loosening or subsidence at a minimum of 2 years and observed a fracture of a modular humeral stem, which inclined them to conclude that monobloc implants could decrease the risk of implant fracture [12]. Stephens et al. also recommended the use of a monobloc stem in cases of revision without the need for allograft in case of moderate humeral bone loss (mean, 3.3 cm) [13].

Some surgeons are using tumor implants for proximal humeral bony defects as it is a more straightforward and yet effective way of treatment [14]. Reconstruction of the humeral offset with a lateralized humerus design or the use of tuberosity augments may be useful to reduce the rate of implant instability due to the glenohumeral compressive forces provided by increased deltoid wrapping effect.

An APC technique is a technique in which a humeral stem is cemented in a matched proximal humerus allograft, and this construct is cemented into the native residual humerus [15–17]. This technique can be done with a step cut or with a simple cut, and we favor a plate fixation to add for stability and to improve bone integration. The decision to replicate the humeral length is based on the extent of humeral loss and also on the type of glenoid revision being performed. In most cases, a reverse glenoid configuration will be used, and depending on the distalization and the size of the glenosphere, one must fine-tune the predicted bone loss planned on the preoperative study. Typically, we will obtain marked bilateral radiographs, and we will measure the amount of bone loss and the level of healthy bone to plan the humeral cut. After performing the host humeral bone and performed the glenoid revision we perform a “shuck test” in which we pull downward from the elbow and gauge the amount of residual bone defect. Ideally, we will have information regarding the diameter of the diaphyseal humerus to match the native humeral width as much as possible. The humeral cut is a simple but essential step. It is important to prevent gapping when fixing the APC to increase host-allograft contact. All efforts are made in order to make the cuts of the residual humerus and APC as parallel as possible to increase the rate of bony union. We trial the APC-trial composite before cementing, and after we are satisfied with the length and stability, we plan plate fixation by predrilling the distal screws avoiding the trial stem.

The host humeral canal is then cemented, and the APC is introduced, and plate fixation is performed. Residual instability is managed by increasing humeral polyethylene liners and augmented humeral trays (Fig. 8.5). We favor using allografts with remnants of the cuff, and we will try to suture the remaining cuff of the patient to the allograft to improve stability and function.

Only gold members can continue reading. Log In or Register to continue

Share this:

• Click to share on Twitter (Opens in new window)
• Click to share on Facebook (Opens in new window)

Related

Related posts:

Total Elbow Arthroplasty: Complications and Results Total Shoulder Arthroplasty: Complications and Results Total Hip Arthroplasty: Complications and Results Total Shoulder Arthroplasty: Epidemiology and Causes Total Hip Arthroplasty: Surgical Technique Total Elbow Arthroplasty: Surgical Technique

Stay updated, free articles. Join our Telegram channel

Join