Arthroplasty for Postfracture Sequelae and Failed Open Reduction and Internal Fixation

Arthroplasty for Postfracture Sequelae and Failed Open Reduction and Internal Fixation

Rick F. Papandrea, MD

Christopher M. Kilian, MD


Proximal humerus fractures continue to be an enigma for the shoulder surgeon. Advanced arthroscopic and percutaneous fixation of fractures are technically demanding techniques and have yet to show superiority to other options. Despite advances in fixation, including locking and anatomically shaped plates, open reduction and internal fixation has, in some studies, demonstrated equivalence to nonoperative treatment.1,2,3,4

Primary operative treatment of proximal humerus fractures with arthroplasty is an important treatment option. Hemiarthroplasty (HA) had been the mainstay for prosthetic replacement, especially in three- and four-part fractures, and still has a role.5,6 Reverse total shoulder arthroplasty (RTSA) has recently shown promise for acute treatment of complex fractures but is not without its complications.5,6,7,8,9,10 Acute treatment of proximal humerus fractures with arthroplasty is covered in Chapters 28 and 29. Surgeons treating proximal humeral fractures understand the equivalence of fixation and nonoperative treatment is not because both give excellent outcomes. On the contrary, the current treatment options for proximal humerus fractures provide surgeons, and more importantly patients, with less than optimal outcomes, often not exceeding nonoperative management.1,4,11,12

Although some studies demonstrate worse outcomes with delayed arthroplasty in proximal humerus fractures, other studies have demonstrated equivalency in outcomes for delayed treatment of fractures with RTSA.13,14 This equivalence in delay, combined with the equivalence in some studies of intervention with nonoperative treatment, gives credence to a thoughtful approach for intervention of acute proximal humerus fractures, especially in the elderly.

Following proximal humerus fracture, patients should maximize the use of rest, anti-inflammatories, physical therapy, and corticosteroid injections (if appropriate), in order to maximize recovery of their premorbid shoulder function. Patients should recognize that following fracture healing their range of motion most likely will be decreased compared to the uninvolved side. However, often the degree of loss is accommodated very well. Mild to moderate malunions of the proximal humerus can be tolerated. However, when range of motion is less than 120° of forward elevation and 30° of external rotation, it has been suggested that osteotomies should be considered to address the mechanical block.15 After adequate time, many patients with a proximal humerus fracture have reasonable return of function, and additional treatment is no longer needed.3,16,17

While there may be a small amount of continued improvement from 6 to 12 months after fracture, it tends to be insignificant, with most of the perceived improvement in this period coming from accommodation and acceptance by the patient of their new level of function. If this new “normal” is accompanied with significant pain or dysfunction, reconstruction should be considered. This chapter will address shoulder arthroplasty for postfracture sequelae (PFS), including failed open reduction and internal fixation (ORIF). Revision of initial arthroplasty for fracture is covered in Chapter 30. Delayed arthroplasty treatment can consist of HA, anatomic total shoulder replacement (ATSA), or RTSA.


Despite a growing body of evidence suggesting that ORIF of proximal humeral fractures may not result in improvement compared with nonoperative treatment, fixation is still often performed. As surgeons, we are accustomed to restoration of function following anatomic fracture fixation. Unfortunately, this is not always the case with proximal humerus fractures. Arthritis from hardware complications, nonunions, malunions, and osteonecrosis (ON) are all potential sequelae from fracture fixation in the proximal humerus, which may be treated with arthroplasty. Shannon et al has shown a higher complication rate following RTSA for failed ORIF compared to primary RTSA for fracture.18 However, the revision rate and clinical outcomes were similar. Similarly, Nowak et al found the reoperation rate following arthroplasty for failed ORIF to be significantly higher compared to primary arthroplasty. The higher rate of revision for arthroplasty after failed ORIF has also been documented by the Danish shoulder arthroplasty registry.19 Thus, the decision to undergo primary ORIF should be made with
caution a full understanding of the outcomes and the risk of complications and reoperations.

Multiple classifications for PFS exist, but none are comprehensive.20,21,22,23 Despite this, the Boileau classification is both simple and practical, providing the surgeon with some guidance for treatment (FIGURE 31.1).21,22

Boileau et al. originally classified nonunion and malunion depending on whether the injury was intracapsular (category 1, type 1 and 2) or extracapsular (category 2, type 3 and 4). Type 1 fractures are characterized by humeral head ON or impaction, and type 2 are chronic dislocations or fracture-dislocations. They recommended ATSA without greater tuberosity osteotomy for these injuries. Category 2 includes the sequelae of extracapsular/disimpacted fractures. These include type 3 fractures with nonunion of the surgical neck and type 4 fractures characterized by severe malunion of the tuberosities. They recommended RTSA for these fracture sequelae. (See text below for current recommendations, especially for type 3 and 4.)

The Boileau classification has been utilized in multiple studies but has not undergone a formal, peer-reviewed validation process for inter- and intra-observer reliability.21,24,25,26,27,28,29,30 PFS of proximal humerus fractures includes pathology that is not specified in any of the four types in Boileau classification. Additionally, advances in prosthetic design and surgeon experience have provided additional options for treatment.


To consider an orthopedic intervention, one must obtain a relevant history, physical examination, imaging studies, and any special testing needed. Fractures affect more than the skeletal structures, and the reconstruction of the proximal humerus post fracture requires attention to more than just the bone. When evaluating the patient who is having difficulties after proximal humerus fracture treatment, one must evaluate the soft tissue envelope and consider its status and function before finalizing a treatment plan. Additionally, pain, nerve deficits, and rotator cuff functional status are all important factors, which must be considered when discussing the potential for reconstruction.


The detailed physical examination of the shoulder will not be reviewed here. The aspects of the physical examination that are critical to patient education and shared decision-making for PFS will be emphasized.

Both active and passive range of motion should be assessed in the PFS patient. Deficits should be pointed out to patients, with further discussion as to perceived potential for change with intervention. This will provide patients with realistic expectations following any proposed surgical procedure. Previous surgical approaches should be noted in the context of additional surgery. These incisions can be utilized, partially incorporated into an approach, or ignored due to the excellent vascularity about the shoulder.

Regardless of the arthroplasty option being considered, a functional deltoid is necessary. Examination should first include visual inspection of the deltoid to evaluate for atrophy. The shoulder should be examined with direct comparison to the contralateral side.

Assessment for disuse atrophy should be made while the examiner inspects the anterior, lateral, and posterior heads of the deltoid. If previous surgery has been performed, any relationship of atrophy to previous incisions should be noted. If a surgical approach included a deltoid split, it is possible that the portion anterior to the split could be denervated with selective atrophy noted.

During the palpation portion of the physical examination, the deltoid should be examined for contraction with use. Even a poorly functioning shoulder can typically be examined for function of all three portions of the deltoid. The examiner can place one hand over the deltoid while using the other hand to support the forearm (FIGURE 31.2). From this position, the patient may extend, abduct, and flex to assess the posterior, middle, and anterior deltoid, respectively. The examiner should also assess sensation over the lateral deltoid, although that is not always reliable to confirm axillary nerve
function. When there is a question or concern for an axillary nerve injury, electrodiagnostic studies should be obtained.

If there is concern about the status of the deltoid after the physical examination, additional imaging such as ultrasound or magnetic resonance imaging (MRI) may be considered to visualize the muscle itself. Nerve conduction studies (NCS) and electromyograms (EMGs) can be performed to assess muscle and nerve function. Even though these tests may provide more information about the deltoid, the surgeon should consider treatment options other than arthroplasty if the deltoid demonstrates severe dysfunction upon inspection and physical examination.31,32,33

A patient with tuberosity malunion or nonunion will likely show deficits during the physical examination. The tuberosity position and examination findings help determine possible future interventions. Patients with greater tuberosity malunion or nonunion will likely show weakness on resisted external rotation with the arm at the side. Additionally, an external rotation lag sign may be present with severe malposition or nonunion of the greater tuberosity. These patients will usually lack active external rotation, and passive external rotation may be limited by the blocking effect of the greater tuberosity malposition. Those with lesser tuberosity malunion generally show weakness with internal rotation based upon assessment by the belly-press maneuver, bear hug test, and lift-off test. These findings may be difficult to distinguish clinically from rotator cuff compromise.

Depending on a patient’s ability to compensate with their deltoid, the loss of the posterior or posterior-superior rotator cuff may also cause a loss of active external rotation. If this is severe, there can be significant difficulties getting the hand to the mouth with the arm at the side, as neutral rotation cannot be maintained and the hand “falls” to the abdomen. If possible, a patient may accommodate this by hyperabducting the shoulder, such that the rotation “bottoms out” with gravity and the hand can reach the mouth. This posturing forms the position of the “hornblower sign.” Even with this accommodation, patients find it difficult to eat, especially with a spoon, as the utensil cannot be held horizontal.


The standard radiographic evaluation of the shoulder includes a Grashey view, scapular Y, and an axillary view. These images will provide reasonable visualization of most proximal humeral anatomy including malunions and nonunions. The glenoid can also be visualized although less well.

If surgical intervention is being considered, a three-dimensional (3D) computed tomography (CT) scan of the shoulder is very important. This study allows for a more thorough evaluation of the fracture fragments, as well as the position of the tuberosities and articular surface. It also provides some information of the rotator cuff musculature and often provides a good indication of rotator cuff tendon status. A standard CT scan may be performed without 3D reconstructions. However, the 3D reconstruction provides a more thorough understanding of the relevant anatomy. If metallic hardware is in place, Metal Artifact Reduction Software (MARS) is essential for proper visualization of the shoulder. This algorithm is sometimes both hardware and software dependent and cannot be done after the study is acquired. It is important that the ordering physician understand these constraints so that MARS is ordered and utilized when needed to avoid the need for a second study with the associated additional cost and radiation exposure.

While most facilities completing a CT scan will create selected 3D images for the study, it is more valuable for the surgeon to be able to choose the perspective and “drive” the 3D reconstructions themselves. This can be done on some workstations but often is difficult and/or requires the surgeon to be at the main facility where the study was done. A simpler and yet more powerful way to select the perspective of the 3D images is for the surgeon to perform the reconstruction on their own. All CT scanners can export the raw axial data in Digital Imaging and Communication in Medicine (DICOM) format. The highest quality (thinnest cut) data should be obtained by the surgeon and can be manipulated in a DICOM viewer. If there is data with MARS, it should be utilized. Open-source software for both PCs (Radiant: and Mac (Horos: are available. The DICOM data can be uploaded to these programs, and 3D reconstructions can be created and manipulated. Experienced users will be able to subtract bone, change perspective, and make measurements on their personal computers.

MRI is not always needed, even when surgery is planned, but may be considered to better evaluate for ON and rotator cuff tendon and muscle compromise.

Nuclear medicine studies may be considered but typically do not add to the treatment plan. Three-phase bone scans will show nonunion and degenerative changes postoperatively, but these are typically well understood from standard radiographs and CT scans. Indolent infections are not readily diagnosed with a three-phase bone scan, and the addition of indium labeling of the white blood cells does not add enough specificity or sensitivity to make it valuable.34


Routine laboratory studies prior to arthroplasty reconstruction are typically performed at the surgeon’s discretion. Modern arthroplasty techniques, especially the use of tranexamic acid, have nearly eliminated the need for
transfusions during or after even extensive arthroplasty reconstructions. Despite this, it is typically advisable to have a preoperative complete blood count (hemoglobin/hematocrit), in case blood loss exceeds expectations.

When previous fracture surgery is being evaluated for reconstruction with a shoulder arthroplasty, one does need to consider that there may be an indolent infection. Laboratory evaluation with an erythrocyte sedimentation rate (ESR), C-reactive protein (CRP), and white blood count (WBC) with differential will provide a baseline but typically are normal, especially when an indolent infection is present.34

The most common postoperative infections in the shoulder are from Cutibacterium acnes (approximately 39%) and are notoriously difficult to diagnose, especially preoperatively.34,35,36 As noted, standard laboratory tests are typically normal. Aspiration of the joint may show high specificity and positive predictive value (PPV) but low specificity and negative predictive value (NPV). Thus, a positive result may be quite useful in guiding treatment, but a negative result does not rule out infection.37

Cultures from aspiration need to be held for at least 14 days on aerobic, anaerobic, and broth media.35 Recent investigations into IL-6 and leukocyte esterase for synovial samples have yet to yield readily available reliable testing.38,39 Alpha defensin measurement (Synovasure—Zimmer/Biomet, Warsaw IN) is commercially available but has been met with mixed literature reviews.40

Next-generation sequencing (NGS), in which the bacterial genome is identified, has been reported for diagnostic use in periprosthetic joint infection (PJI).41,42,43,44 While this technique is intellectually appealing, its clinical use is still evolving. When NGS is performed in native shoulders during arthroplasty, more bacteria are identified compared to culture results.45 This may be due to dead bacteria, or there may be a normal bacterial load in synovial joints that is yet to be understood. The study of the microbiome of specific parts of the body, including synovial joints, has demonstrated that few places in the body are truly sterile. Some investigators believe that there may be a role of bacteria as a causative agent in arthritis.46

The first concern to address if one is considering arthroplasty for PFS reconstruction is that of infection. When surgical treatment of proximal humeral fractures fails, one must always consider the possibility of an underlying infection. If infection has been ruled out, or infection has been treated as outlined below, then one can proceed to reconstruction based on the structural pathology, with specific attention to the soft tissue envelope of the proximal humerus.

Many postoperative infections about the shoulder are clinically clear, especially when they occur early. Indolent infections are much more difficult to diagnose. Therefore, it is important to have a high index of suspicion when evaluating failed proximal humerus fracture treatments. The rate of low-grade infection after fracture fixation in the shoulder may be higher than after fracture fixation of hips and knees.47

A painful or dysfunctional shoulder after treatment for fracture which is obviously infected needs to be aggressively treated. The more common and difficult presentation is the failed fracture fixation or primary arthroplasty for fracture with an indolent infection.

The shoulder which is clearly infected will typically have pain, erythema, swelling, and induration. The laboratory workup will often be notable for increased CRP, ESR, and a high WBC count with increased neutrophils. Drainage and/or sinus tracts may be present. Aspiration will usually yield grossly purulent fluid that can be analyzed for cell count and differential. Clinical scenarios like this are often infected with a staph species or a gram-negative bacterium.

An obvious infection in a posttraumatic proximal humerus fracture which will undergo eventual arthroplasty must first undergo aggressive debridement including hardware removal. A decision has to be made as to a single-stage reconstruction with immediate implantation or a multistage reconstruction with delayed final implantation.36,47,48,49,50 When delayed implantation is chosen, a polymethyl methacrylate (PMMA) spacer is placed to maintain the resection space, afford some support/stability, and provide antibiotics via elution.

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

Jun 23, 2022 | Posted by in ORTHOPEDIC | Comments Off on Arthroplasty for Postfracture Sequelae and Failed Open Reduction and Internal Fixation
Premium Wordpress Themes by UFO Themes