- Email: [email protected]
How do revised shoulders that are culture positive for Propionibacterium differ from those that are not?
J Shoulder Elbow Surg (2015) 24, 1427-1432
www.elsevier.com/locate/ymse
How do revised shoulders that are culture positive for Propionibacterium differ from those that are not? Clifford Hou, MD, Akash Gupta, MD, Michael Chen, BS, Frederick A. Matsen III, MD* Department of Orthopaedics and Sports Medicine, University of Washington, Clyde Hill, WA, USA Background: Many shoulder arthroplasties revised for pain, stiffness, or component loosening are culture positive for Propionibacterium. The culture results are unknown until days or weeks after surgery, too late to inform intraoperative surgical decisions and immediate postsurgical antibiotic treatment. The objective of this study was to identify preoperative and intraoperative characteristics that may alert surgeons to an increased likelihood of positive cultures. Methods: We analyzed the records of 132 shoulders that underwent surgical revision of a shoulder arthroplasty, 66 of which became culture positive for Propionibacterium and 66 did not. Results: Propionibacterium-positive and Propionibacterium-negative shoulders were similar with respect to many characteristics; however, Propionibacterium-negative shoulders were revised sooner after the index procedure and were significantly more likely to be female, to have sustained a fall, to have instability, and to have rotator cuff deficiency. Intraoperatively, Propionibacterium-positive shoulders demonstrated more glenoid erosions, glenoid osteolysis, glenoid loosening, and a higher incidence of a soft tissue membrane between the humeral component and humeral endosteum. Shoulders culture positive for Propionibacterium were more likely to be culture positive for another bacteria. Conclusions: Although Propionibacterium-positive and Propionibacterium-negative shoulders have many similarities, factors such as male gender, delayed presentation, glenoid osteolysis and loosening, humeral membrane, and the absence of instability or cuff failure should arouse suspicion of Propionibacterium and suggest the need for deep cultures and consideration of aggressive surgical and medical treatment. Level of evidence: Level III, Retrospective Cohort Design, Treatment Study. Ó 2015 Journal of Shoulder and Elbow Surgery Board of Trustees. Keywords: Propionibacterium; shoulder arthroplasty; culture positive
This prospective study was approved by the University of Washington Human Subjects Review Committee (Approval #43575). *Reprint requests: Frederick A. Matsen III, MD, Shoulder and Elbow Surgery, Department of Orthopaedics and Sports Medicine, University of Washington Medical Center, 1959 NE Pacific St, Box 356500, Seattle, WA 98195-6500, USA. E-mail address: [email protected] (F.A. Matsen III).
Shoulder arthroplasty is being performed with increasing frequency,10 and the frequency of revision surgery for failed arthroplasty is rising commensurately.2 A substantial number of these revisions are related to periprosthetic infection, some of which are clinically apparent,21,22 and others present more subtly, without the classical signs and laboratory findings.12,14,16,19 A recent study showed that approximately 40% of shoulder arthroplasties revised for stiffness, pain, or
1058-2746/$ - see front matter Ó 2015 Journal of Shoulder and Elbow Surgery Board of Trustees. http://dx.doi.org/10.1016/j.jse.2015.01.003
1428 prosthetic loosening were associated with positive intraoperative cultures for Propionibacterium.19 Propionibacterium have been recovered from arthroplasty revisions performed as many as 14 years after the index arthroplasty procedure.14 Propionibacterium are skin flora found in dermal sebaceous glands and hair bulbs, especially over the chest and back of men.9,13,15,17 Propionibacterium are usually subtle in presentation, rarely producing the evidence of acute infection that arouses clinical suspicion of infection.3,5-7,11,19,22 Even if cultures are obtained, the presence of Propionibacterium may go undetected unless special specimen harvest and culturing methods are used and unless specimens are observed for 2 or more weeks.4,12,13 When a revision is performed for a failed arthroplasty, it is important for the surgeon to anticipate the risk of Propionibacterium in that persistence of these organisms in the wound may predispose the revision to failure. Because the clinical and surgical findings are rarely conclusive, decisions about prosthesis exchange and antibiotic treatment must be made before the culture results become definitive days or weeks after surgery. The purpose of this study was to seek preoperative clinical, radiographic, and intraoperative findings associated with positive Propionibacterium cultures from the wound at the time of revision surgery. The presence or absence of these findings should help guide surgeons in their choice of initial surgical and antibiotic management before the final culture results are known.
Materials and methods To test the hypothesis that there were significant differences in the characteristics of Propionibacterium-positive and Propionibacterium-negative revision arthroplasties, we conducted a retrospective cohort study of the 190 revision shoulder arthroplasties performed between April 2009 and September 26, 2013, by a single surgeon (F.A.M.) at our medical center. None of these shoulders were revised because of a preoperative diagnosis of infection or because of failure obviously resulting from trauma. Shoulders with incomplete clinical records and radiographs and those for which preoperative antibiotics were not held before harvesting intraoperative cultures were excluded. None of these patients had preoperative laboratory marker values indicative of infection. None had preoperative aspiration or synovial biopsy. The start date for this study was selected as the date by which our protocol for the harvesting and culturing of intraoperative specimens at revision shoulder arthroplasty had become standardized, April 2009. A prior report19 presented 193 shoulders revised for pain, stiffness, or apparently aseptic loosening between May 2006 and January 2011. The current study included 73 shoulders from that group and 117 additional shoulders revised after January 2011. Patient demographics, range of motion, body mass index, indications for revision surgery, preoperative Simple Shoulder Test scores, intraoperative findings, and microbiology reports were obtained from the medical record. The immediate preoperative radiographic findings were interpreted by a resident orthopedic surgeon who was blinded to the clinical data, intraoperative findings, and culture results.
C. Hou et al. In each of the included cases, intraoperative cultures were obtained before antibiotics were administered, were handled carefully to avoid contamination, and were cultured in broth and on aerobic and anaerobic media for a minimum of 3 weeks, as previously described.4,13,19 If any of the specimens from a shoulder grew Propionibacterium, the shoulder was included in the Propionibacterium-positive group. The Propionibacterium-positive group and the Propionibacterium-negative group were compared using univariate analysis with unpaired t tests for means, z tests for proportions, and c2 analyses for proportions of parametric data to identify statistically significant differences (P < .05.)
Results During the study period 190 shoulders were available for review. Of these, 18 Propionibacterium-positive shoulders and 40 Propionibacterium-negative shoulders failed to meet the inclusion criteria (Fig. 1), leaving 66 Propionibacterium-positive and 66 Propionibacterium-negative shoulders for analysis. Many of the characteristics of the Propionibacteriumpositive shoulders were not significantly different from those of the Propionibacterium-negative shoulders, including average age, shoulder dominance, number of prior procedures, diagnosis before index arthroplasty, body mass index, tobacco usage, depression, prerevision Simple Shoulder Test score, stiffness, presence of clinical signs of infection, and presence of pain (Table I). Propionibacterium-positive shoulders were almost twice as likely to be male (P < .01). The time from initial arthroplasty to the revision was more than 4 years for the Propionibacterium-positive shoulders and less than 3 years for the Propionibacterium-negative shoulders (P < .01). Diabetes was almost 3 times more common among the Propionibacterium-negative shoulders (P < .005). A history of a fall was only found in the Propionibacterium-negative shoulders (P < .007). Propionibacterium-negative shoulders were more likely to be revised for instability (P < .03) or cuff failure (P < .02). The percentage of reverse total shoulder arthroplasties was higher in the Propionibacteriumpositive group (9.1% vs 1.5%, P ¼ 0.05). Preoperative radiographs of Propionibacterium-positive shoulders were approximately twice as likely to show glenoid erosion (P < .02), osteolysis (P < .04), and glenoid component loosening (P < .05; Table II and Fig. 2). At surgery, Propionibacterium-positive shoulders were 3 times as likely to have cloudy joint fluid (P < .04), 50% more likely to have a humeral membrane (P < .01), and substantially more likely to have glenoid osteolysis (P < .003), glenoid fracture (P < .03), and glenoid malposition (P < .008; Table III). The number of specimens submitted for culture were greater for the Propionibacterium-positive shoulders (6.6 2.4 vs 4.8 3.0, P < .0001). At least 1 additional bacterial species was found in 56% of Propionibacterium-positive shoulders, whereas only
Propionibacterium in shoulder arthroplasty revision surgery
Figure 1 Table I
1429
Diagram shows the distribution of the shoulders between our 2 cohorts and exclusion from analysis based on exclusion criteria. Shoulder characteristics, comorbidities, and presenting shoulder function and symptoms
Preoperative clinical data Age, y Sex Male Female Right shoulder Left shoulder Months from original surgery to re-presentation No. of prior procedures Body mass index, kg/m2 Diabetes History of tobacco use Current tobacco use Major depressive disorder Flexion arc,o Simple Shoulder Test score at re-presentation History of a traumatic event Revised for Stiffness Instability Infectious signs Systemic Local Rotator cuff failure Revised for pain
Propionibacteriumþ patients (n ¼ 66)
Propionibacterium– patients (n ¼ 66)
Mean SD or No.
Mean SD or No.
60.1 12.3
61.7 12.1
48 18 37 29 52.0 53.4
25 41 41 25 30.5 40.8
2.48 28.4 7 12 6 13 89.8 2.72
2.24 5.2
48.4 2.67
2.41 28.6 20 24 1 17 67.7 2.05
1.75 5.9
54.4 2.66
P value
<.01)
<.01y
<.005z
<.02y
0
7
<.007z
38 3
45 11
<.03z
2 1 0 63
2 3 10 57
<.02z
SD, standard deviation. ) By c2 test. y By unpaired t test. z By z test for proportions.
17.4% of the Propionibacterium-negative shoulders had polymicrobial cultures.
Discussion Positive cultures for Propionibacterium are commonly reported from specimens harvested at revision shoulder
arthroplasty.3,5,7,19 Although these positive cultures were initially dismissed as contaminants not requiring treatment,7 it now appears that Propionibacterium can contribute to shoulder pain, stiffness, and component loosening without the classical evidence of a periprosthetic infection.3,5-7,11,19,22 For this reason, it is important to manage revision arthroplasties with the recognition that
1430 Table II
C. Hou et al. Summary of preoperative radiographic findings
Preoperative radiographic findings Glenoid Wear Loosening Osteolysis Fracture Malposition Tuberosity failure Humeral Malposition Loosening Osteolysis Acromial fracture Heterotopic ossification )
Propionibacteriumþ patients (n ¼ 66)
Propionibacterium– patients (n ¼ 66)
(No.)
(No.)
25 16 17 7 7 1
23 12 4 1 0 4
5 8 6 5 2
5 13 6 1 1
P value
<.003) <.03) <.008)
By z test for proportions.
Figure 2 Glenoid osteolysis and loosening are demonstrated in (A) Grashey and (B) axillary lateral radiographs of a patient with pain and stiffness. Ten intraoperative cultures were sent for this patient, of which 6 were positive for Propionibacterium.
substantial Propionibacterium may be present in the surgical wound, and if so, may persist in an antibiotic-resistant biofilm1,20 unless prosthesis exchange and prompt and appropriate antibiotic therapy are implemented. Because these therapeutic steps are expensive and not without their own risks, surgeons and patients can benefit from guidance about which revision shoulder arthroplasties are most likely to be culture positive for Propionibacterium. Our study shows that there are significant differences between revision shoulder arthroplasties that are Propionibacterium positive and those that are Propionibacterium negative that can be observed before and at
the time of revision shoulder surgery. Factors such as male sex, delayed presentation, lack of instability or cuff failure, and radiographically and intraoperatively identified glenoid bony changes should raise suspicion for the presence of Propionibacterium. Shoulders with these characteristics may merit harvest of multiple specimens for Propionibacterium culture and consideration for prosthesis exchange and immediate, vigorous antibiotic therapy until the results of the cultures are known. The present study suggests that Propionibacteriumpositive revision shoulder arthroplasties may be associated
Propionibacterium in shoulder arthroplasty revision surgery Table III
1431
Summary of intraoperative findings
Intraoperative findings Glenoid Erosions Osteolysis Loosening Fracture Malposition Tuberosity failure Rotator cuff failure Humeral Malposition Loosening Osteolysis Fracture Broken component Acromial Fracture Surgical membrane Surgical cloudy fluid Patients with polymicrobial cultures No. of cultures
Propionibacteriumþ patients (n ¼ 66)
Propionibacterium– patients (n ¼ 66)
Mean SD or No.
Mean SD or No.
27 12 17 1 0 2 8
14 4 8 1 1 4 14
4 12 8 0 1 0 60 12 37 6.62 2.35
4 8 8 2 1 1 44 4 4 4.80 2.95
P value
.02) .04) .05)
<.01) <.04) <4 108) <.00002y
SD, standard deviation. ) By z test for proportions. y By unpaired t test.
with a different mechanism of failure than revised arthroplasties that are negative for Propionibacterium. Although the latter are more likely to have traumatic events, rotator cuff deficiency, and instability in female patients, the former are more commonly associated with delayed bone resorption leading to component loosening in male patients. This is particularly evident around the glenoid where eccentric loading and the rocking horse phenomenon8,18,23 may potentiate the effects of bone loss. Persistent Propionibacterium may contribute to the reported 67% recurrence rate of glenoid loosening in apparently aseptic failed arthroplasties with loose glenoid components that are revised with reinsertion of another glenoid component.3 Our results should be interpreted in light of certain limitations. First, the number of deep cultures were not the same for each case: an average of 6.6 2.4 specimens were obtained in the shoulders that proved to be Propionibacterium positive, and 4.8 2.9 specimens were obtained in shoulders that proved to be Propionibacterium negative. This difference may have been due to an increased level of surgical suspicion in the shoulders that proved Propionibacterium positive. The rate of positive cultures per shoulder is recognized to increase with the number of specimens submitted.19 Second, our criterion for a Propionibacterium-positive shoulder was at least 1 positive specimen. We recognize that other investigators might use a different threshold.
Third, this was a retrospective study; as a result, the medical records for some of the risk factors may be incomplete. Fourth, this is not a study of periprosthetic shoulder infections but rather a study comparing Propionibacteriumpositive and Propionibacterium-negative shoulder revision surgeries. Finally, 68 of the original 190 cases (18 Propionibacterium positive and 50 Propionibacterium negative) were excluded because of failure to meet our inclusion criterion.
Conclusions This study supports the observation that a substantial number of revision shoulder arthroplasties are culture positive for Propionibacterium despite the absence of the characteristic clinical signs of infection. Our results demonstrate that certain clinical, radiographic, and intraoperative characteristics are significantly associated with positive cultures for Propionibacterium. Shoulders with these findings may merit consideration for the harvest of multiple cultures and vigorous surgical and antibiotic treatment to prevent persistence of this organism. Further research will be required to make specific recommendations on clinical management.
1432
Disclaimer The authors, their immediate families, and any research foundations with which they are affiliated have not received any financial payments or other benefits from any commercial entity related to the subject of this article.
References 1. Bayston R, Ashraf W, Barker-Davies R, Tucker E, Clement R, Clayton J, et al. Biofilm formation by Propionibacterium acnes on biomaterials in vitro and in vivo: impact on diagnosis and treatment. J Biomed Mater Res A 2007;81:705-9. 2. Bohsali KI, Wirth MA, Rockwood CA. Complications of total shoulder arthroplasty. J Bone Joint Surg Am 2006;88:2279-92. 3. Bonnevialle N, Melis B, Neyton L, Favard L, Mole D, Walch G, et al. Aseptic glenoid loosening or failure in total shoulder arthroplasty: revision with glenoid reimplantation. J Shoulder Elbow Surg 2013;22: 745-51. http://dx.doi.org/10.1016/j.jse.2012.08.009 4. Butler-Wu SM, Burns EM, Pottinger PS, Magaret AS, Rakeman JL, Matsen FA 3rd, et al. Optimization of periprosthetic culture for diagnosis of Propionibacterium acnes prosthetic joint infection. J Clin Microbiol 2011;49:2490-5. http://dx.doi.org/10.1128/JCM.00450-11 5. Cil A, Veillette CJ, Sanchez-Sotelo J, Sperling JW, Schleck C, Cofield RH. Revision of the humeral component for aseptic loosening in arthroplasty of the shoulder. J Bone Joint Surg Br 2009;91:75-81. http://dx.doi.org/10.1302/0301-620X.91B1.21094 6. Dodson CC, Craig EV, Cordasco FA, Dines DM, Dines JS, Dicarlo E, et al. Propionibacterium acnes infection after shoulder arthroplasty: a diagnostic challenge. J Shoulder Elbow Surg 2010;19:303-7. http://dx. doi.org/10.1016/j.jse.2009.07.065 7. Foruria AM, Fox TJ, Sperling JW, Cofield RH. Clinical meaning of unexpected positive cultures (UPC) in revision shoulder arthroplasty. J Shoulder Elbow Surg 2013;22:620-7. http://dx.doi.org/10.1016/j.jse. 2012.07.017 8. Franklin JL, Barrett WP, Jackins SE, Matsen FA 3rd. Glenoid loosening in total shoulder arthroplasty. Association with rotator cuff deficiency. J Arthroplasty 1988;3:39-46. 9. Grice EA, Segre JA. The skin microbiome. Nat Rev Microbiol 2011;9: 244-53. http://dx.doi.org/10.1038/nrmicro2537 10. Kim SH, Wise BL, Zhang Y, Szabo RM. Increasing incidence of shoulder arthroplasty in the United States. J Bone Joint Surg Am 2011; 93:2249-54. http://dx.doi.org/10.2106/JBJS.J.01994
C. Hou et al. 11. Klimkiewicz JJ, Iannotti JP, Rubash HE, Shanbhag AS. Aseptic loosening of the humeral component in total shoulder arthroplasty. J Shoulder Elbow Surg 1998;7:422-6. 12. Levy PY, Fenollar F, Stein A, Borrione F, Cohen E, Lebail B, et al. Propionibacterium acnes postoperative shoulder arthritis: an emerging clinical entity. Clin Infect Dis 2008;46:1884-6. http://dx.doi.org/10. 1086/588477 13. Matsen FA III, Butler-Wu S, Carofino BC, Jette JL, Bertelsen A, Bumgarner R. Origin of Propionibacterium in surgical wounds and evidence-based approach for culturing Propionibacterium from surgical sites. J Bone Joint Surg Am 2013;95:e1811-7. http://dx.doi.org/ 10.2106/JBJS.L.01733 14. McGoldrick E, McElvany MD, Butler-Wu S, Pottinger PS, Matsen FA 3rd. Substantial cultures of Propionibacterium can be found in apparently aseptic shoulders revised three years or more after the index arthroplasty. J Shoulder Elbow Surg 2015;24:31-5. http://dx. doi.org/10.1016/j.jse.2014.05.008 15. McLorinan GC, Glenn JV, McMullan MG, Patrick S. Propionibacterium acnes wound contimination at the time of spinal surgery. Clin Orthop Relat Res 2005;437:67-73. 16. Millett PJ, Yen YM, Price CS, Horan MP, van der Meijden OA, Elser F. Propionibacterium acnes infection as an occult cause of postoperative shoulder pain: a case series. Clin Orthop Relat Res 2011; 469:2824-30. http://dx.doi.org/10.1007/s11999-011-1767-4 17. Nakatsuji T, Chiang H-I, Jiang SB, Nagarajan H, Zengler K, Gallo RL. The microbiome extends to subepidermal components of normal skin. Nat Commun 2013;4:1431. http://dx.doi.org/10.1038/ncomms2441 18. Patel RJ, Wright TM, Gao Y. Load transfer after cemented total shoulder arthroplasty. J Shoulder Elbow Surg 2014;23:1553-62. http:// dx.doi.org/10.1016/j.jse.2014.01.038 19. Pottinger PS, Butler-Wu S, Neradilek MB, Merritt A, Bertelsen A, Jette JL, et al. Prognostic factors for bacterial cultures positive for Propionibacterium acnes and other organisms in a large series of revision shoulder arthroplasties performed for stiffness, pain, or loosening. J Bone Joint Surg Am 2012;94:2075-83. http://dx.doi.org/ 10.2106/JBJS.K.00861 20. Ramage G, Tunney MM, Patrick S, Gorman SP, Nixon JR. Formation of Propionibacterium acnes biofilms on orthopaedic biomaterials and their susceptibility to antimicrobials. Biomaterials 2003;24:3221-7. 21. Singh JA, Sperling JW, Schleck C, Harmsen WS, Cofield RH. Periprosthetic infections after shoulder hemiarthroplasty. J Shoulder Elbow Surg 2012;21:1304-9. http://dx.doi.org/10.1016/j.jse.2011.08. 067 22. Singh JA, Sperling JW, Schleck C, Harmsen WS, Cofield RH. Periprosthetic infections after total shoulder arthroplasty: a 33-year perspective. J Shoulder Elbow Surg 2012;21:1534-41. http://dx.doi. org/10.1016/j.jse.2012.01.006 23. Wirth MA, Rockwood CA. Complications of total shoulderreplacement arthroplasty. J Bone Joint Surg Am 1996;87:603-16.
www.elsevier.com/locate/ymse
How do revised shoulders that are culture positive for Propionibacterium differ from those that are not? Clifford Hou, MD, Akash Gupta, MD, Michael Chen, BS, Frederick A. Matsen III, MD* Department of Orthopaedics and Sports Medicine, University of Washington, Clyde Hill, WA, USA Background: Many shoulder arthroplasties revised for pain, stiffness, or component loosening are culture positive for Propionibacterium. The culture results are unknown until days or weeks after surgery, too late to inform intraoperative surgical decisions and immediate postsurgical antibiotic treatment. The objective of this study was to identify preoperative and intraoperative characteristics that may alert surgeons to an increased likelihood of positive cultures. Methods: We analyzed the records of 132 shoulders that underwent surgical revision of a shoulder arthroplasty, 66 of which became culture positive for Propionibacterium and 66 did not. Results: Propionibacterium-positive and Propionibacterium-negative shoulders were similar with respect to many characteristics; however, Propionibacterium-negative shoulders were revised sooner after the index procedure and were significantly more likely to be female, to have sustained a fall, to have instability, and to have rotator cuff deficiency. Intraoperatively, Propionibacterium-positive shoulders demonstrated more glenoid erosions, glenoid osteolysis, glenoid loosening, and a higher incidence of a soft tissue membrane between the humeral component and humeral endosteum. Shoulders culture positive for Propionibacterium were more likely to be culture positive for another bacteria. Conclusions: Although Propionibacterium-positive and Propionibacterium-negative shoulders have many similarities, factors such as male gender, delayed presentation, glenoid osteolysis and loosening, humeral membrane, and the absence of instability or cuff failure should arouse suspicion of Propionibacterium and suggest the need for deep cultures and consideration of aggressive surgical and medical treatment. Level of evidence: Level III, Retrospective Cohort Design, Treatment Study. Ó 2015 Journal of Shoulder and Elbow Surgery Board of Trustees. Keywords: Propionibacterium; shoulder arthroplasty; culture positive
This prospective study was approved by the University of Washington Human Subjects Review Committee (Approval #43575). *Reprint requests: Frederick A. Matsen III, MD, Shoulder and Elbow Surgery, Department of Orthopaedics and Sports Medicine, University of Washington Medical Center, 1959 NE Pacific St, Box 356500, Seattle, WA 98195-6500, USA. E-mail address: [email protected] (F.A. Matsen III).
Shoulder arthroplasty is being performed with increasing frequency,10 and the frequency of revision surgery for failed arthroplasty is rising commensurately.2 A substantial number of these revisions are related to periprosthetic infection, some of which are clinically apparent,21,22 and others present more subtly, without the classical signs and laboratory findings.12,14,16,19 A recent study showed that approximately 40% of shoulder arthroplasties revised for stiffness, pain, or
1058-2746/$ - see front matter Ó 2015 Journal of Shoulder and Elbow Surgery Board of Trustees. http://dx.doi.org/10.1016/j.jse.2015.01.003
1428 prosthetic loosening were associated with positive intraoperative cultures for Propionibacterium.19 Propionibacterium have been recovered from arthroplasty revisions performed as many as 14 years after the index arthroplasty procedure.14 Propionibacterium are skin flora found in dermal sebaceous glands and hair bulbs, especially over the chest and back of men.9,13,15,17 Propionibacterium are usually subtle in presentation, rarely producing the evidence of acute infection that arouses clinical suspicion of infection.3,5-7,11,19,22 Even if cultures are obtained, the presence of Propionibacterium may go undetected unless special specimen harvest and culturing methods are used and unless specimens are observed for 2 or more weeks.4,12,13 When a revision is performed for a failed arthroplasty, it is important for the surgeon to anticipate the risk of Propionibacterium in that persistence of these organisms in the wound may predispose the revision to failure. Because the clinical and surgical findings are rarely conclusive, decisions about prosthesis exchange and antibiotic treatment must be made before the culture results become definitive days or weeks after surgery. The purpose of this study was to seek preoperative clinical, radiographic, and intraoperative findings associated with positive Propionibacterium cultures from the wound at the time of revision surgery. The presence or absence of these findings should help guide surgeons in their choice of initial surgical and antibiotic management before the final culture results are known.
Materials and methods To test the hypothesis that there were significant differences in the characteristics of Propionibacterium-positive and Propionibacterium-negative revision arthroplasties, we conducted a retrospective cohort study of the 190 revision shoulder arthroplasties performed between April 2009 and September 26, 2013, by a single surgeon (F.A.M.) at our medical center. None of these shoulders were revised because of a preoperative diagnosis of infection or because of failure obviously resulting from trauma. Shoulders with incomplete clinical records and radiographs and those for which preoperative antibiotics were not held before harvesting intraoperative cultures were excluded. None of these patients had preoperative laboratory marker values indicative of infection. None had preoperative aspiration or synovial biopsy. The start date for this study was selected as the date by which our protocol for the harvesting and culturing of intraoperative specimens at revision shoulder arthroplasty had become standardized, April 2009. A prior report19 presented 193 shoulders revised for pain, stiffness, or apparently aseptic loosening between May 2006 and January 2011. The current study included 73 shoulders from that group and 117 additional shoulders revised after January 2011. Patient demographics, range of motion, body mass index, indications for revision surgery, preoperative Simple Shoulder Test scores, intraoperative findings, and microbiology reports were obtained from the medical record. The immediate preoperative radiographic findings were interpreted by a resident orthopedic surgeon who was blinded to the clinical data, intraoperative findings, and culture results.
C. Hou et al. In each of the included cases, intraoperative cultures were obtained before antibiotics were administered, were handled carefully to avoid contamination, and were cultured in broth and on aerobic and anaerobic media for a minimum of 3 weeks, as previously described.4,13,19 If any of the specimens from a shoulder grew Propionibacterium, the shoulder was included in the Propionibacterium-positive group. The Propionibacterium-positive group and the Propionibacterium-negative group were compared using univariate analysis with unpaired t tests for means, z tests for proportions, and c2 analyses for proportions of parametric data to identify statistically significant differences (P < .05.)
Results During the study period 190 shoulders were available for review. Of these, 18 Propionibacterium-positive shoulders and 40 Propionibacterium-negative shoulders failed to meet the inclusion criteria (Fig. 1), leaving 66 Propionibacterium-positive and 66 Propionibacterium-negative shoulders for analysis. Many of the characteristics of the Propionibacteriumpositive shoulders were not significantly different from those of the Propionibacterium-negative shoulders, including average age, shoulder dominance, number of prior procedures, diagnosis before index arthroplasty, body mass index, tobacco usage, depression, prerevision Simple Shoulder Test score, stiffness, presence of clinical signs of infection, and presence of pain (Table I). Propionibacterium-positive shoulders were almost twice as likely to be male (P < .01). The time from initial arthroplasty to the revision was more than 4 years for the Propionibacterium-positive shoulders and less than 3 years for the Propionibacterium-negative shoulders (P < .01). Diabetes was almost 3 times more common among the Propionibacterium-negative shoulders (P < .005). A history of a fall was only found in the Propionibacterium-negative shoulders (P < .007). Propionibacterium-negative shoulders were more likely to be revised for instability (P < .03) or cuff failure (P < .02). The percentage of reverse total shoulder arthroplasties was higher in the Propionibacteriumpositive group (9.1% vs 1.5%, P ¼ 0.05). Preoperative radiographs of Propionibacterium-positive shoulders were approximately twice as likely to show glenoid erosion (P < .02), osteolysis (P < .04), and glenoid component loosening (P < .05; Table II and Fig. 2). At surgery, Propionibacterium-positive shoulders were 3 times as likely to have cloudy joint fluid (P < .04), 50% more likely to have a humeral membrane (P < .01), and substantially more likely to have glenoid osteolysis (P < .003), glenoid fracture (P < .03), and glenoid malposition (P < .008; Table III). The number of specimens submitted for culture were greater for the Propionibacterium-positive shoulders (6.6 2.4 vs 4.8 3.0, P < .0001). At least 1 additional bacterial species was found in 56% of Propionibacterium-positive shoulders, whereas only
Propionibacterium in shoulder arthroplasty revision surgery
Figure 1 Table I
1429
Diagram shows the distribution of the shoulders between our 2 cohorts and exclusion from analysis based on exclusion criteria. Shoulder characteristics, comorbidities, and presenting shoulder function and symptoms
Preoperative clinical data Age, y Sex Male Female Right shoulder Left shoulder Months from original surgery to re-presentation No. of prior procedures Body mass index, kg/m2 Diabetes History of tobacco use Current tobacco use Major depressive disorder Flexion arc,o Simple Shoulder Test score at re-presentation History of a traumatic event Revised for Stiffness Instability Infectious signs Systemic Local Rotator cuff failure Revised for pain
Propionibacteriumþ patients (n ¼ 66)
Propionibacterium– patients (n ¼ 66)
Mean SD or No.
Mean SD or No.
60.1 12.3
61.7 12.1
48 18 37 29 52.0 53.4
25 41 41 25 30.5 40.8
2.48 28.4 7 12 6 13 89.8 2.72
2.24 5.2
48.4 2.67
2.41 28.6 20 24 1 17 67.7 2.05
1.75 5.9
54.4 2.66
P value
<.01)
<.01y
<.005z
<.02y
0
7
<.007z
38 3
45 11
<.03z
2 1 0 63
2 3 10 57
<.02z
SD, standard deviation. ) By c2 test. y By unpaired t test. z By z test for proportions.
17.4% of the Propionibacterium-negative shoulders had polymicrobial cultures.
Discussion Positive cultures for Propionibacterium are commonly reported from specimens harvested at revision shoulder
arthroplasty.3,5,7,19 Although these positive cultures were initially dismissed as contaminants not requiring treatment,7 it now appears that Propionibacterium can contribute to shoulder pain, stiffness, and component loosening without the classical evidence of a periprosthetic infection.3,5-7,11,19,22 For this reason, it is important to manage revision arthroplasties with the recognition that
1430 Table II
C. Hou et al. Summary of preoperative radiographic findings
Preoperative radiographic findings Glenoid Wear Loosening Osteolysis Fracture Malposition Tuberosity failure Humeral Malposition Loosening Osteolysis Acromial fracture Heterotopic ossification )
Propionibacteriumþ patients (n ¼ 66)
Propionibacterium– patients (n ¼ 66)
(No.)
(No.)
25 16 17 7 7 1
23 12 4 1 0 4
5 8 6 5 2
5 13 6 1 1
P value
<.003) <.03) <.008)
By z test for proportions.
Figure 2 Glenoid osteolysis and loosening are demonstrated in (A) Grashey and (B) axillary lateral radiographs of a patient with pain and stiffness. Ten intraoperative cultures were sent for this patient, of which 6 were positive for Propionibacterium.
substantial Propionibacterium may be present in the surgical wound, and if so, may persist in an antibiotic-resistant biofilm1,20 unless prosthesis exchange and prompt and appropriate antibiotic therapy are implemented. Because these therapeutic steps are expensive and not without their own risks, surgeons and patients can benefit from guidance about which revision shoulder arthroplasties are most likely to be culture positive for Propionibacterium. Our study shows that there are significant differences between revision shoulder arthroplasties that are Propionibacterium positive and those that are Propionibacterium negative that can be observed before and at
the time of revision shoulder surgery. Factors such as male sex, delayed presentation, lack of instability or cuff failure, and radiographically and intraoperatively identified glenoid bony changes should raise suspicion for the presence of Propionibacterium. Shoulders with these characteristics may merit harvest of multiple specimens for Propionibacterium culture and consideration for prosthesis exchange and immediate, vigorous antibiotic therapy until the results of the cultures are known. The present study suggests that Propionibacteriumpositive revision shoulder arthroplasties may be associated
Propionibacterium in shoulder arthroplasty revision surgery Table III
1431
Summary of intraoperative findings
Intraoperative findings Glenoid Erosions Osteolysis Loosening Fracture Malposition Tuberosity failure Rotator cuff failure Humeral Malposition Loosening Osteolysis Fracture Broken component Acromial Fracture Surgical membrane Surgical cloudy fluid Patients with polymicrobial cultures No. of cultures
Propionibacteriumþ patients (n ¼ 66)
Propionibacterium– patients (n ¼ 66)
Mean SD or No.
Mean SD or No.
27 12 17 1 0 2 8
14 4 8 1 1 4 14
4 12 8 0 1 0 60 12 37 6.62 2.35
4 8 8 2 1 1 44 4 4 4.80 2.95
P value
.02) .04) .05)
<.01) <.04) <4 108) <.00002y
SD, standard deviation. ) By z test for proportions. y By unpaired t test.
with a different mechanism of failure than revised arthroplasties that are negative for Propionibacterium. Although the latter are more likely to have traumatic events, rotator cuff deficiency, and instability in female patients, the former are more commonly associated with delayed bone resorption leading to component loosening in male patients. This is particularly evident around the glenoid where eccentric loading and the rocking horse phenomenon8,18,23 may potentiate the effects of bone loss. Persistent Propionibacterium may contribute to the reported 67% recurrence rate of glenoid loosening in apparently aseptic failed arthroplasties with loose glenoid components that are revised with reinsertion of another glenoid component.3 Our results should be interpreted in light of certain limitations. First, the number of deep cultures were not the same for each case: an average of 6.6 2.4 specimens were obtained in the shoulders that proved to be Propionibacterium positive, and 4.8 2.9 specimens were obtained in shoulders that proved to be Propionibacterium negative. This difference may have been due to an increased level of surgical suspicion in the shoulders that proved Propionibacterium positive. The rate of positive cultures per shoulder is recognized to increase with the number of specimens submitted.19 Second, our criterion for a Propionibacterium-positive shoulder was at least 1 positive specimen. We recognize that other investigators might use a different threshold.
Third, this was a retrospective study; as a result, the medical records for some of the risk factors may be incomplete. Fourth, this is not a study of periprosthetic shoulder infections but rather a study comparing Propionibacteriumpositive and Propionibacterium-negative shoulder revision surgeries. Finally, 68 of the original 190 cases (18 Propionibacterium positive and 50 Propionibacterium negative) were excluded because of failure to meet our inclusion criterion.
Conclusions This study supports the observation that a substantial number of revision shoulder arthroplasties are culture positive for Propionibacterium despite the absence of the characteristic clinical signs of infection. Our results demonstrate that certain clinical, radiographic, and intraoperative characteristics are significantly associated with positive cultures for Propionibacterium. Shoulders with these findings may merit consideration for the harvest of multiple cultures and vigorous surgical and antibiotic treatment to prevent persistence of this organism. Further research will be required to make specific recommendations on clinical management.
1432
Disclaimer The authors, their immediate families, and any research foundations with which they are affiliated have not received any financial payments or other benefits from any commercial entity related to the subject of this article.
References 1. Bayston R, Ashraf W, Barker-Davies R, Tucker E, Clement R, Clayton J, et al. Biofilm formation by Propionibacterium acnes on biomaterials in vitro and in vivo: impact on diagnosis and treatment. J Biomed Mater Res A 2007;81:705-9. 2. Bohsali KI, Wirth MA, Rockwood CA. Complications of total shoulder arthroplasty. J Bone Joint Surg Am 2006;88:2279-92. 3. Bonnevialle N, Melis B, Neyton L, Favard L, Mole D, Walch G, et al. Aseptic glenoid loosening or failure in total shoulder arthroplasty: revision with glenoid reimplantation. J Shoulder Elbow Surg 2013;22: 745-51. http://dx.doi.org/10.1016/j.jse.2012.08.009 4. Butler-Wu SM, Burns EM, Pottinger PS, Magaret AS, Rakeman JL, Matsen FA 3rd, et al. Optimization of periprosthetic culture for diagnosis of Propionibacterium acnes prosthetic joint infection. J Clin Microbiol 2011;49:2490-5. http://dx.doi.org/10.1128/JCM.00450-11 5. Cil A, Veillette CJ, Sanchez-Sotelo J, Sperling JW, Schleck C, Cofield RH. Revision of the humeral component for aseptic loosening in arthroplasty of the shoulder. J Bone Joint Surg Br 2009;91:75-81. http://dx.doi.org/10.1302/0301-620X.91B1.21094 6. Dodson CC, Craig EV, Cordasco FA, Dines DM, Dines JS, Dicarlo E, et al. Propionibacterium acnes infection after shoulder arthroplasty: a diagnostic challenge. J Shoulder Elbow Surg 2010;19:303-7. http://dx. doi.org/10.1016/j.jse.2009.07.065 7. Foruria AM, Fox TJ, Sperling JW, Cofield RH. Clinical meaning of unexpected positive cultures (UPC) in revision shoulder arthroplasty. J Shoulder Elbow Surg 2013;22:620-7. http://dx.doi.org/10.1016/j.jse. 2012.07.017 8. Franklin JL, Barrett WP, Jackins SE, Matsen FA 3rd. Glenoid loosening in total shoulder arthroplasty. Association with rotator cuff deficiency. J Arthroplasty 1988;3:39-46. 9. Grice EA, Segre JA. The skin microbiome. Nat Rev Microbiol 2011;9: 244-53. http://dx.doi.org/10.1038/nrmicro2537 10. Kim SH, Wise BL, Zhang Y, Szabo RM. Increasing incidence of shoulder arthroplasty in the United States. J Bone Joint Surg Am 2011; 93:2249-54. http://dx.doi.org/10.2106/JBJS.J.01994
C. Hou et al. 11. Klimkiewicz JJ, Iannotti JP, Rubash HE, Shanbhag AS. Aseptic loosening of the humeral component in total shoulder arthroplasty. J Shoulder Elbow Surg 1998;7:422-6. 12. Levy PY, Fenollar F, Stein A, Borrione F, Cohen E, Lebail B, et al. Propionibacterium acnes postoperative shoulder arthritis: an emerging clinical entity. Clin Infect Dis 2008;46:1884-6. http://dx.doi.org/10. 1086/588477 13. Matsen FA III, Butler-Wu S, Carofino BC, Jette JL, Bertelsen A, Bumgarner R. Origin of Propionibacterium in surgical wounds and evidence-based approach for culturing Propionibacterium from surgical sites. J Bone Joint Surg Am 2013;95:e1811-7. http://dx.doi.org/ 10.2106/JBJS.L.01733 14. McGoldrick E, McElvany MD, Butler-Wu S, Pottinger PS, Matsen FA 3rd. Substantial cultures of Propionibacterium can be found in apparently aseptic shoulders revised three years or more after the index arthroplasty. J Shoulder Elbow Surg 2015;24:31-5. http://dx. doi.org/10.1016/j.jse.2014.05.008 15. McLorinan GC, Glenn JV, McMullan MG, Patrick S. Propionibacterium acnes wound contimination at the time of spinal surgery. Clin Orthop Relat Res 2005;437:67-73. 16. Millett PJ, Yen YM, Price CS, Horan MP, van der Meijden OA, Elser F. Propionibacterium acnes infection as an occult cause of postoperative shoulder pain: a case series. Clin Orthop Relat Res 2011; 469:2824-30. http://dx.doi.org/10.1007/s11999-011-1767-4 17. Nakatsuji T, Chiang H-I, Jiang SB, Nagarajan H, Zengler K, Gallo RL. The microbiome extends to subepidermal components of normal skin. Nat Commun 2013;4:1431. http://dx.doi.org/10.1038/ncomms2441 18. Patel RJ, Wright TM, Gao Y. Load transfer after cemented total shoulder arthroplasty. J Shoulder Elbow Surg 2014;23:1553-62. http:// dx.doi.org/10.1016/j.jse.2014.01.038 19. Pottinger PS, Butler-Wu S, Neradilek MB, Merritt A, Bertelsen A, Jette JL, et al. Prognostic factors for bacterial cultures positive for Propionibacterium acnes and other organisms in a large series of revision shoulder arthroplasties performed for stiffness, pain, or loosening. J Bone Joint Surg Am 2012;94:2075-83. http://dx.doi.org/ 10.2106/JBJS.K.00861 20. Ramage G, Tunney MM, Patrick S, Gorman SP, Nixon JR. Formation of Propionibacterium acnes biofilms on orthopaedic biomaterials and their susceptibility to antimicrobials. Biomaterials 2003;24:3221-7. 21. Singh JA, Sperling JW, Schleck C, Harmsen WS, Cofield RH. Periprosthetic infections after shoulder hemiarthroplasty. J Shoulder Elbow Surg 2012;21:1304-9. http://dx.doi.org/10.1016/j.jse.2011.08. 067 22. Singh JA, Sperling JW, Schleck C, Harmsen WS, Cofield RH. Periprosthetic infections after total shoulder arthroplasty: a 33-year perspective. J Shoulder Elbow Surg 2012;21:1534-41. http://dx.doi. org/10.1016/j.jse.2012.01.006 23. Wirth MA, Rockwood CA. Complications of total shoulderreplacement arthroplasty. J Bone Joint Surg Am 1996;87:603-16.