Periprosthetic joint infections (PJI) of the shoulder can be difficult to diagnose and difficult to treat. The most commonly infecting organisms are those that are commensal inhabitants of normal skin: Cutibacterium and coagulase negative Staphylococcus.

The clinical onset of a shoulder PJI can range from a "stealth" presentation (with the insidious onset of unexplained pain and stiffness months or years after the arthroplasty (see this link)) to an "obvious" presentation with swelling, erythema, component loosening and wound drainage (see this link).

The stealth presentation of a shoulder PJI

With stealth presentations, the surgeon cannot be sure at the time of surgery whether or not the pain and stiffness are due to infection or to some other cause - the final diagnosis must wait until cultures of deep specimens harvested at the time of revision are finalized several weeks after surgery. In that situation, a single stage revision - vigorous debridement, irrigation, and prosthesis exchange followed by antibiotic therapy until the culture results are finalized - provides an effective approach as demonstrated in Single-Stage Revision Is Effective for Failed Shoulder Arthroplasty with Positive Cultures for Propionibacterium. The effectiveness of this approach is illustrated in the case presented below,

A 49 year old athletic man presented with osteoarthritis of the right shoulder, a Simple Shoulder Test score of 6 out of 12, and a type B2 glenoid with biconcavity, retroversion and posterior decentering.

Because of his active lifestyle he elected a ream and run arthroplasty. His six week postoperative films show centering of the humeral prosthesis in a concave glenoid.

Following this procedure, the patient was able to return to regular competitive swimming at the master’s level.

Five years after his index arthroplasty, he returned to the office with complaints of the insidious onset of pain and stiffness without an identified explanation (what we refer to is the "Honeymoon" phenomenon). His Simple Shoulder Test score was 3 out of 12. His CBC, sedimentation rate and C-reactive protein were all within normal limits. His radiographs showed glenoid erosion.

Six years after his index arthroplasty he had a single stage revision. At the time of his revision he was 55 years of age, ASA 2, BMI 24. The preoperative Propionibacterium cultures of his unprepared skin over the area of the prior incision were strongly positive with a EpidermalSpPV of 2. His intraoperative findings included synovitis. His revision was followed by six weeks of intravenous antibiotic treatment (Ceftriaxone) and 6 months of Augmentin. At 3 weeks after surgery his eight deep intraoperative cultures were strongly positive with a total Propionibacterium score (ShPS) of 7.4.

One year after his revision arthroplasty his Simple Shoulder Test score was 11 out of 12. His radiographs showed showed a centered humeral head with radiographic joint space between the humeral head and the glenoid.

He has returned to his regular swimming program

The obvious presentation of a shoulder PJI

In the obvious presentation of a shoulder PJI the shoulder typically shows signs of tenderness, swelling, drainage and component loosening along with laboratory evidence of inflammation in tests of serum and joint fluid and cultures positive for virulent organisms.
In this situation and in cases of failed single stage revision, surgeons often prefer a two-stage revision, with prothesis removal, debridement, irrigation and insertion of a methyl methacrylate spacer containing up to 5 grams of heat-resistant antibiotics, such as vancomycin, tobramycin and/or genatamycin in 40 grams of methyl methacrylate bone cement. The rationale for the use of these spacers is two-fold: (1) to provide ongoing local elution of antibiotics during the postoperative period and (2) to preserve the joint space for later re-implantation of a definitive arthroplasty; this is the "two-stage" approach to revision. The authors of Antibiotic Spacers for Shoulder Periprosthetic Joint Infection conducted an extensive review of the use of spacers to treat obvious periprosthetic infections of the shoulder. The sense of the literature gained from multiple relatively small case series is that the two stage approach is often effective in eradication of infection, but that the resulting shoulder comfort and function is only modest. The complications of antibiotic spacers include bone erosion, fracture of the spacer, loosening and rotation of the spacer, and humeral shaft fracture.

These authors prefer to manage obvious shoulder periprosthetic infections by fabricating shoulder spacers intraoperatively with the use of commercially available molds allowing for the creation of press-fit stems and the ability to control the antibiotic choice and concentration. After spacer implantation, patients are treated with intravenous antibiotics for 6 weeks based on culture-specific sensitivities. They then check ESR and CRP levels and do a shoulder aspiration once intravenous antibiotics have been discontinued for a minimum of 2 weeks. If serum markers remain elevated or if cultures are positive, then repeat irrigation and débridement is performed with insertion of a new spacer. If serum markers return to normal levels and aspirate cultures are negative, they proceed to the second stage of revision (insertion of a definitive arthroplasty implant) at 10 to 12 weeks after the placement of the antibiotic spacer. Five sets of intraoperative cultures are obtained at the time of reimplantation. Patients are started on oral antibiotics until these cultures finalize. If cultures are negative, then oral antibiotics are discontinued. If positive, intravenous antibiotics are restarted for a 6-week duration followed by possible lifetime suppression with oral antibiotic therapy.

It is noted that some patients elect to retain the methyl methacrylate spacer, rather than proceeding with the second of the two stage revision as shown in the example below. Spacer retention carries the risk of glenoid erosion.