Yao, Jie J. MD¹; Jurgensmeier, Kevin BS¹; Woodhead, Benjamin M. DO¹; Whitson, Anastasia J. BSPH¹; Pottinger, Paul S. MD¹; Matsen, Frederick A. III MD¹; Hsu, Jason E. MD¹

Department of Orthopaedic Surgery and Sports Medicine (J.J.Y., B.M.W., A.J.W., F.A.M., and J.E.H.), Division of Allergy and Infectious Diseases, Department of Medicine (P.S.P.), and School of Medicine (K.J.), University of Washington, Seattle, Washington

The Journal of Bone & Joint Surgery. February 2020.
DOI: 10.2106/JBJS.19.00846

Summary by Arjun Singh, BS

When performing revision shoulder arthroplasty, it is necessary to determine the presence of peri-prosthetic joint infection in order to determine the appropriate antibiotic protocol. Yet, culture results of intraoperative specimens may take weeks to be finalized, as in the case of the most common cause of infection after shoulder surgery, Cutibacterium acnes.^1,2  As a result, surgeons must choose an antibiotic protocol solely based on preoperative and intraoperative observations.³ Currently, there are no standardized guidelines for postoperative antibiotic therapy after revision shoulder arthroplasty.⁴ At the authors institution, patients are placed on either intravenous (IV) or oral antibiotics following revision arthroplasty, depending on the level of suspicion as to whether or not intraoperative samples will confirm periprosthetic joint infection.⁵ This is defined as identifying 2 or more positive cultures of the same species.⁶

This study was a retrospective review of 175 revision shoulder arthroplasties that met the study inclusion and exclusion criteria. Patients who opted out of the recommended protocol were excluded, as were patients with obvious infection. 62 patients were placed in the high suspicion group, and 113 in the low suspicion group. High suspicion patients were started on IV antibiotics, which were discontinued at 21 days if cultures were negative, but continued for 6 weeks if cultures were positive. Low suspicion patients were given oral antibiotics for 21 days, and this was either discontinued or patients were switched to 6 weeks of IV antibiotics followed by 3 months of oral antibiotics if cultures returned positive.

Factors found to be significantly associated with the initiation of IV antibiotics on univariate analysis include male sex (p < 0.001; 79% vs 49%), history of infection in the shoulder (p <0.001; 14% vs. 4%), intraoperative humeral loosening (p=0.003; 34% vs. 15%), and membrane formation (p<0.001; 73% vs. 33%). On the basis of preoperative and intraoperative characteristics, surgeons successfully anticipated culture results in 75% of cases, and modification of antibiotic therapy was required in 25%. Modification from oral to IV antibiotics occurred significantly more often in male patients (p < 0.001). Adverse effects from antibiotic administration occurred in 19% of patients. The rate of complication was significantly lower in the group of patients treated with oral antibiotics  discontinued at 3 weeks when compared to the other 3 groups (p < 0.001; 6% vs. 23-40%). Of the 92 patients with a PICC line, 4% (n=4) developed a venous thromboembolism.

The higher conversion rate from IV to oral antibiotics for male patients is consistent with the known characteristics of C. acnes and the male cutaneous biome.⁷ This higher rate may suggest that male sex is an underestimated risk factor for periprosthetic joint infection. The authors concluded that complications associated with antibiotic administration are not infrequent and are more common in patients for whom an initial protocol of IV antibiotics were employed. They also concluded that certain preoperative and intraoperative characteristics include male sex, history of infection, and membrane formation were predictive of surgeon selection of antibiotic protocol and allowed for the anticipation of culture results in 75% of cases. The main limitations of this study were: 1) this was a retrospective review at one referral center, 2) the indications for the protocols used and the antibiotic protocols themselves were unique to this institution, 3) the relative effectiveness of the two antibiotic protocols in minimizing the risk of recurrent infection after surgical revision was not reported, and 4) other diagnostic considerations such as synovial markers were not employed.

References:

1. Levy, P. Y. et al. Propionibacterium acnes Postoperative Shoulder Arthritis: An Emerging Clinical Entity . Clin. Infect. Dis. (2008). doi:10.1086/588477
2. Dodson, C. C. et al. Propionibacterium acnes infection after shoulder arthroplasty: a diagnostic challenge. J. shoulder Elb. Surg. 19, 303–307 (2010).
3. Updegrove, G. F., Armstrong, A. D. & Kim, H. M. M. Preoperative and intraoperative infection workup in apparently aseptic revision shoulder arthroplasty. Journal of Shoulder and Elbow Surgery (2015). doi:10.1016/j.jse.2014.10.005
4. Cuff, D. J. et al. The treatment of deep shoulder infection and glenohumeral instability with debridement, reverse shoulder arthroplasty and post-operative antibiotics. J. Bone Jt. Surg. – Ser. B (2008). doi:10.1302/0301-620X.90B3.19408
5. Hsu, J. E., Gorbaty, J. D., Whitney, I. J. & Matsen, F. A. Single-stage revision is effective for failed shoulder arthroplasty with positive cultures for propionibacterium. Journal of Bone and Joint Surgery – American Volume (2016). doi:10.2106/JBJS.16.00149
6. Pottinger, P. et al. Prognostic Factors for Bacterial Cultures Positive for < J. Bone &amp (2012). doi:10.2106/jbjs.k.00861
7. Falconer, T. M. et al. Contamination of the surgical field with propionibacterium acnes in primary shoulder arthroplasty. J. Bone Jt. Surg. – Am. Vol. (2016). doi:10.2106/JBJS.15.01133