Despite what seems like a new, high-quality study being published on the topic every week or so, orthopaedic surgeons still have an extremely hard time determining whether a prosthetic hip or knee is infected or not. We have an array of available tests and the relatively easy-to-follow criteria for a periprosthetic joint infection (PJI) from the Musculoskeletal Infection Society (MSIS), but a large number of these patients still fall into the gray zone of “possibly infected.” This predicament is especially thorny in patients who received antibiotics just prior to the diagnostic workup, which interferes with the accuracy of many tests for PJI.
In the April 17, 2019 issue of The Journal, Shahi et al. remind orthopaedic surgeons about a valuable tool that can be used in this scenario. Their retrospective study looked at 121 patients who had undergone revision hip or knee arthroplasty due to an MSIS criteria-confirmed periprosthetic infection. Shahi et al. sought to determine which diagnostic tests were least affected by prior antibiotic administration. The authors found that erythrocyte sedimentation rate (ESR), C-reactive protein (CRP) level, synovial white blood cell (WBC) count, and polymorphonuclear neutrophil (PMN) percentage were all significantly lower in the 32% of patients who had received antibiotics within 2 weeks of those tests, compared with the 68% who did not receive antibiotics. The only test that was found not to be significantly affected by the prior admission of antibiotics was the urine-based leukocyte esterase strip test.
Considering the ease and rapidity with which a leukocyte esterase test can be performed and evaluated (at a patient’s bedside, with immediate results), its low cost, and the fact that it is included in the MSIS criteria, these findings are very important and useful. While we would prefer that patients with a possibly infected total hip or knee not receive antibiotics prior to their diagnostic workup, previous antibiotic exposure remains a relatively common scenario. The findings from this study can assist us in those difficult cases, and they add further evidence to support the value and reliability of the easy-to-perform leukocyte esterase test.
Chad A. Krueger, MD
JBJS Deputy Editor for Social Media
A study by Miller et al. in the February 20, 2019 issue of JBJS provides preclinical proof of concept that antibiotic-loaded coatings on orthopaedic implants could eventually reduce the incidence of implant-associated infection.
The researchers used in vivo bioluminescence imaging (BLI) and ex vivo analysis of colony-forming units (CFUs) to show the efficacy of an implant coating that released linezolid-rifampin over a 7-day period. Through a parapatellar arthrotomy, researchers reamed the femoral canal of 12 rabbits and inoculated the canals with a bioluminescent strain of MRSA. They then inserted a surgical grade titanium peg into each canal. All of the pegs were coated with a nanofiber coating; 6 of the pegs were loaded with the antibiotic coating and 6 were not.
Implants coated without antibiotics were associated with significantly increased in vivo BLI signals and significantly increased knee width, relative to implants with the antibiotic-releasing coatings. The animals were killed on day 7, and ex vivo analysis of CFUs isolated from soft tissue, bone, and implant specimens showed significantly increased CFUs in the specimens without the antibiotic-releasing coating, while CFUs were undetectable in the implants with antibiotics.
This larger-animal model to assess bacterial burden employed a clinically used orthopaedic implant and replicated a medial parapatellar arthrotomy in humans. According to the authors, the coating used is “highly versatile, and the polymers or drug concentrations could be modified for more rapid or longer release.” This rabbit model should be amenable to studying additional antibiotic-releasing strategies for possible translation to clinical research in humans.
OrthoBuzz occasionally receives posts from guest bloggers. In response to a recent New England Journal of Medicine study, the following commentary comes from Daniel Leas, MD and Joseph R. Hsu, MD.
Deep infections continue to be one of the most resource-intensive problems that orthopaedic surgeons face. Long-standing dogma has favored 6 or more weeks of intravenous (IV) antibiotics, resulting in increased healthcare costs during both the inpatient and outpatient treatment periods.
To explore the possibility of utilizing targeted oral antibiotics as an alternative, effective treatment for musculoskeletal infections, the OVIVA (Oral versus Intravenous Antibiotics) multicenter research collaboration conducted a prospective, randomized controlled trial. A total of 1,054 patients with deep musculoskeletal infections were randomized to oral or IV arms for 6 weeks of antibiotic treatment and followed for 1 year to determine treatment efficacy. The primary end point was treatment failure within 1 year, defined as the presence of predefined clinical symptoms of deep infection, microbiologic evidence of continued infection, or histologic presence of microorganisms or inflammatory tissue. Secondary outcomes included catheter-associated complications, discontinuation of therapy, and Clostridium difficile diarrhea.
Of the 1,054 patients enrolled, 909 patients were included in the final analysis. Treatment failure occurred in 14.6% of patients treated with IV antibiotics and 13.2% of patients in the oral-therapy group. This -1.4% difference indicated noninferiority based on the predetermined 7.5% noninferiority margin. Secondary outcomes between the groups differed only in catheter-related complications being more common in the IV group (9.4% vs 1.0% in the oral group).
These findings and conclusions should challenge us to re-evaluate the basis for extended IV antibiotics to treat complex musculoskeletal infections, and to consider a greater role for oral antibiotics for such infections. Further study of this question focused on patients with retained hardware is warranted.
Daniel P. Leas, MD is a PGY-5 orthopaedic resident at Carolinas Medical Center.
Joseph R. Hsu, MD is a Professor of Orthopaedic Trauma and Vice Chair of Quality at the Atrium Health Musculoskeletal Institute.
Periprosthetic joint infections (PJIs) create a significant burden for patients, surgeons, and healthcare systems. That is why so much research has gone into how best to optimize certain patients preoperatively—such as those with obesity, diabetes, or kidney disease—to decrease the risk of these potentially catastrophic complications. Still, it is not always possible or feasible to optimize every “high-risk” patient who would benefit from a total hip or knee replacement, and therefore many such patients undergo surgery with an increased risk of infection. In such cases, surgeons need additional strategies to decrease PJI risk.
In the December 19, 2018 issue of JBJS, Inabathula et al. investigate whether providing high-risk total joint arthroplasty (TJA) patients with extended postoperative oral antibiotics decreased the risk of PJI within the first 90 days after surgery. In their retrospective cohort study, the authors examined >2,100 total hip and knee replacements performed at a single suburban academic hospital. The patients in 68% of these cases had at least one risk factor for infection. Among those high-risk patients, about half received 7 days of an oral postoperative antibiotic, while the others received only the standard 24 hours of postoperative intravenous (IV) antibiotics.
Relative to those who received IV antibiotics only, those who received extended oral antibiotics experienced an 81% reduction in infection for total knee arthroplasties and a 74% reduction in infection for total hip arthroplasties. I was stunned by such large reductions in infection rates obtained simply by adding an oral antibiotic twice a day for 7 days. Most arthroplasty surgeons go to great lengths to decrease the risk of joint infection by percentages much less than that.
While further investigations are needed and legitimate concerns exist regarding the propagation of antimicrobial-resistant organisms from medical antibiotic misuse, these data are very exciting. I agree with Monti Khatod, MD, who, in his commentary on this study, says that “care pathways that aim to improve modifiable risk factors should not be seen as obsolete based on the findings of this paper.” Furthermore, the study itself is at risk for treatment and selection biases that could greatly influence its outcomes. Nevertheless, getting a successful result in these patients is challenging and, if validated with further data, this research may help surgeons obtain better outcomes when treating high-risk patients in need of hip or knee replacements.
Chad A. Krueger, MD
JBJS Deputy Editor for Social Media
This post comes from Fred Nelson, MD, an orthopaedic surgeon in the Department of Orthopedics at Henry Ford Hospital and a clinical associate professor at Wayne State Medical School. Some of Dr. Nelson’s tips go out weekly to more than 3,000 members of the Orthopaedic Research Society (ORS), and all are distributed to more than 30 orthopaedic residency programs. Those not sent to the ORS are periodically reposted in OrthoBuzz with the permission of Dr. Nelson.
A bioﬁlm is a complex combination of extracellular carbohydrates, proteins, lipids, and one or more species of bacteria that may adhere to an orthopaedic implant and surrounding tissue (see related OrthoBuzz post). Staphylococci bacteria are believed to account for more than 50% of all bioﬁlm infections of medical devices.
Researchers recently summarized what we know about the biofilm formation process.1 In the attachment phase, free-floating bacteria attach to a prosthetic surface via proteins. Extracellular DNA from autolysis add to the mix. Then begins the irreversible attachment phase, during which the initial bacteria are incarcerated while more free-floating bacteria are added. During this phase, autoinducers are expressed, which serve as inter- and intrabacterial signals.
In the presence of an adequate quorum of bacteria, the maturation phase begins, during which the bacterial population cohesively shifts from replication to expression of virulence factors such as secretion systems, toxins, or bioﬁlm formation. A mature biofilm is immune-resistant, although bacterial replication decreases. In the dispersal phase bacteria become planktonic again, potentially available to repeat the process.
Once a biofilm has formed, antibiotic administration becomes problematic because of the toxicity of the high doses needed to treat biofilm colonies. An underlying challenge with pharmacologic intervention is the variety of quorum-sensing communication pathways between bacterial species. The authors suggest that a future biofilm-fighting strategy may be to force bacteria into bioﬁlm-forming behavior before they reach the necessary critical density to become virulent, although this notion remains unexplored. Researchers are investigating other possible strategies to disrupt the quorum-sensing communication among bacteria that enable them to behave as a “social” group.
- Mooney JA, Pridgen EM, Manasherob R, Suh G, Blackwell HE, Barron AE, Bollyky PL, Goodman SB, Amanatullah DF. Periprosthetic bacterial biofilm and quorum sensing. J Orthop Res. 2018 Sep;36(9):2331-2339. doi: 10.1002/jor.24019. Epub 2018 May 24.
When it comes to preventing infections associated with orthopaedic procedures, the question of which antibiotic to use is only one of several concerns. How and where to administer antibiotics is another relevant question, not only in terms of infection-fighting effectiveness but also in terms of combatting the proliferation of antibiotic-resistant microbes.
In the September 19, 2018 issue of The Journal of Bone & Joint Surgery, Sweet et al. report on findings from a study in rats that compared the infection-prevention efficacy of intravenous (IV) cefazolin (n = 20) and IV vancomycin (n = 20) with local application of 4 antimicrobials—vancomycin powder (n = 20), cefazolin powder (n = 20), tobramycin powder (n = 20), and dilute Betadine lavage (n = 20).
The researchers induced infection by surgically implanting a polytetrafluoroethylene vascular graft near each rat’s thoracic spine and inoculating it with methicillin-sensitive Staphylococcus aureus (MSSA). After 7 days, all of the rats in each of the IV cefazolin, IV vancomycin, and Betadine lavage groups had grossly positive cultures for MSSA, “with bacterial colonies too numerous to count.” Ninety percent of the rats in the local cefazolin-powder group also had positive cultures, but the infection rates with vancomycin and tobramycin powder were much lower than those with the other four approaches (p <0.000001).
In addition to the main “disclaimer” about this study (namely, that its findings cannot be extrapolated to clinical practice in humans), the authors caution that “the effect of locally applied antibiotics on the emergence of resistant organisms is unknown,” while citing evidence that systemic administration of antibiotics is “associated with the emergence of resistant organisms at an alarming rate.”
Sweet et al. say they plan to follow up this study with a similar model to investigate the efficacy of local antimicrobials against the more problematic methicillin-resistant Staphylococcus aureus (MRSA)—and they suggest further that “clinical studies should be considered to determine the relative clinical efficacy of local versus systemic antibiotics for surgical infection prophylaxis in humans.”
The number of articles published each year in orthopaedics that evaluate infections seems to approach, if not exceed, 1,000. Yet, despite all of these publications, consensus statements, and guidelines, we seem to have very few concrete recommendations about which every surgeon will say, “This is what needs to be done.” So we send out samples, run cultures, sonicate implants, and sometimes even perform DNA sequencing, and then we mix the data with selected recommendations and intuition to make our final treatment decisions. Foolproof? No, but it is the best we can do in many situations.
The article by Mijuskovic et al. in the September 5, 2018 edition of The Journal helps simplify this type of decision making in the setting of residual osteomyelitis after toe or forefoot amputation. The authors evaluated 51 consecutive patients with gangrene and/or infection who underwent either digit or partial foot amputations. They found that, after surgery, 41% of the patients without histological evidence of osteomyelitis (which the authors considered the reference, “true positive” analysis) had a positive culture from the same sample. In addition, only 12 patients (24%) had both positive histological findings and positive cultures, the criteria set forth by the Infectious Disease Society of America for the definitive diagnosis of osteomyelitis.
As interesting as the main findings of the study are, some of the “minor” results are even more curious. The decision regarding which patients received antibiotics after amputation seemed largely arbitrary, with 10 of the 14 patients who had a positive histological result not receiving any postoperative antibiotics. (Five of those patients ended up needing a secondary procedure.) In addition, because of the need for decalcification prior to analysis, the median time to receiving histological results was almost a week. Based on the findings in this study, in many instances patients are sent home or to a rehabilitation facility with antibiotics based only on the results of a potentially “false-positive” culture.
The authors conclude that their results “cast doubt on the strategy of relying solely on culture of bone biopsy specimens when deciding whether antibiotic treatment for osteomyelitis is necessary after toe or forefoot amputation.” But this paper also highlights the fact that we are still looking for definitive answers about which data to use and which to disregard when it comes to the detection and treatment of post-amputation osteomyelitis. We surgeons decide on which side to err, and we need to appreciate all three facets—data, guidelines, and patient factors—when discussing treatment options with patients.
Chad A. Krueger, MD
JBJS Deputy Editor for Social Media
Every month, JBJS publishes a Specialty Update—a review of the most pertinent and impactful studies published in the orthopaedic literature during the previous year in 13 subspecialties. Click here for a collection of all OrthoBuzz Specialty Update summaries.
This month, Arvind Nana, MD, and his co-authors of the July 18, 2018 Specialty Update on Musculoskeletal Infection, selected the most clinically compelling findings from among the more than 130 studies summarized in the Specialty Update.
CDC Guidelines on Surgical Site Infection (SSI) Prevention
–The most provocative recommendation in the CDC SSI Prevention Guidelines,1 released in 2017, was to encourage administration of parenteral antimicrobial prophylaxis prior to surgery so that a bactericidal concentration is established in the tissue and serum when the incision is made. Postoperatively, the CDC recommended that antimicrobial prophylaxis not be administered in clean and clean-contaminated procedures after incision closure, even if a drain is present.
Treating Periprosthetic Infection
–When performing debridement to treat a periprosthetic joint infection, dilute methylene blue (0.1%) applied to the tissue prior to debridement (with removal of excess dye) may help surgeons visualize devitalized tissue (biofilm) that should be debrided at the time of infection.2,3
–Two Level-I studies showed that specific wound-closure techniques can improve incisional perfusion. This was seen in the setting of ankle fracture with the Allgower-Donati suture technique4 and in elective total knee arthroplasty with a running subcuticular closure5.
–Two studies reported on the microbiological impact of locally applied vancomycin powder.6,7 For patients who developed infections after surgery despite the application of vancomycin powder, a greater frequency of gram-negative organisms was identified, highlighting the importance of obtaining specimens for culture.
–Atypical hand infections caused by Mycobacterium tuberculosis, non-tuberculous mycobacterium, and fungi are uncommon, making high-level clinical trials unrealistic. But these atypical infections are frequent enough that multiple cases are reported, drawing attention to the need for awareness of their clinical presentation,8 even in immunocompetent patients,9 and the need for understanding that cultures should be sent when suspicion is high, even if there is purulence consistent with a typical bacterial infection.10
- Berrios-Torres SI, Umscheid CA, Bratzler DW, Leas B, Stone EC, Kelz RR, Reinke CE, Morgan S, Solomkin JS, Mazuski JE, Dellinger EP, Itani KMF, Berbari EF, Segreti J, Parvizi J, Blanchard J, Allen G, Kluytmans JAJW, Donlan R, Schecter WP; Healthcare Infection Control Practices Advisory Committee. Centers for Disease Control and Prevention guideline for the prevention of surgical site infection, 2017. JAMA Surg. 2017 Aug 1;152(8):784-91.
- Shaw JD, Miller S, Plourde A, Shaw DL, Wustrack R, Hansen EN. Methylene blue guided debridement as an intraoperative adjunct for the surgical treatment of periprosthetic joint infection. J Arthroplasty. 2017 Dec;32(12):3718-23.
- Parry JA, Karau MJ, Kakar S, Hanssen AD, Patel R, Abdel MP. Disclosing agents for the intraoperative identification of biofilms on orthopedic implants. J Arthroplasty. 2017 Aug;32(8):2501-4.
- Shannon SF, Houdek MT,Wyles CC, Yuan BJ, CrossWW3rd, Cass JR, Sems SA. Allgower-Donati versus vertical mattress suture technique impact on perfusion in ankle fracture surgery: a randomized clinical trial using intraoperative angiography. J Orthop Trauma. 2017 Feb;31(2):97-102.
- Wyles CC, Jacobson SR, Houdek MT, Larson DR, Taunton MJ, Sim FH, Sierra RJ, Trousdale RT. The Chitranjan Ranawat Award: running subcuticular closure enables the most robust perfusion after TKA: a randomized clinical trial. Clin Orthop Relat Res. 2016 Jan;474(1):47-56.
- Adogwa O, Elsamadicy AA, Sergesketter A, Vuong VD, Mehta AI, Vasquez RA, Cheng J, Bagley CA, Karikari IO. Prophylactic use of intraoperative vancomycin powder and postoperative infection: an analysis of microbiological patterns in 1200 consecutive surgical cases. J Neurosurg Spine. 2017 Sep;27(3):328-34. Epub 2017 Jun 30.
- Chotai S, Wright PW, Hale AT, Jones WA, McGirt MJ, Patt JC, Devin CJ. Does intrawound vancomycin application during spine surgery create vancomycin-resistant organism? Neurosurgery. 2017 May 1;80(5):746-53.
- Lopez M, Croley J, Murphy KD. Atypical mycobacterial infections of the upper extremity: becoming more atypical? Hand (N Y). 2017 Mar;12(2):188-92. Epub 2016 Jul.
- Sotello D, Garner HW, Heckman MG, Diehl NN, Murray PM, Alvarez S. Nontuberculous mycobacterial infections of the upper extremity: 15-year experience at a tertiary care medical center. J Hand Surg Am. 2017 Dec 6:S0363-5023(16)30908-X. Epub 2017 Dec 6
- Kazmers NH, Fryhofer GW, Gittings D, Bozentka DJ, Steinberg DR, Gray BL. Acute deep infections of the upper extremity: the utility of obtaining atypical cultures in the presence of purulence. J Hand Surg Am. 2017 Aug;42(8):663.e1-8. Epub 2017 May 25.
At any given time, a patient’s blood-glucose level is easy to measure. Beyond the standard pre/postoperative lab values, there are finger sticks, transdermal meters, and other modalities that make taking a patient’s glucose “snapshot” pretty straightforward. So why don’t we surgeons keep track of it more frequently before and after joint replacement, when, according to the prognostic study by Shohat et al. in the July 5, 2018 issue of JBJS, fluctuating glucose levels can have a critical impact on outcomes?
By retrospectively studying more than 5,000 patients who had undergone either total hip or total knee arthroplasty, the authors found that increased variability of glucose levels (measured by a coefficient of variation) was associated with increased risks of 90-day mortality, surgical-site infection, and periprosthetic joint infection. Specifically, the authors demonstrated that for every 10-percentage-point increase in the glycemic coefficient of variation, the risk of 90-day mortality increased by 26%, and the risk of periprosthetic or surgical-site infection increased by 20%. These are remarkable increases in extremely important outcome measures, and the associations held regardless of the patient’s mean glucose values prior to or after the surgery. In fact, some of the highest levels of glucose variability were found in patients who had well-controlled glucose levels preoperatively. Furthermore, as Charles Cornell, MD points out in a commentary on this study, “Glucose variability appears to affect surgical prognosis more than chronic hyperglycemia.”
These findings were surprising and a bit concerning. I don’t tend to order routine blood-glucose measurements postoperatively on patients who appear to be euglycemic based on preoperative testing. Yet, according to these data, maybe I should. Findings of high glucose variability postoperatively might now prompt me to consult with endocrine or perioperative medicine specialists or at least consider informing patients with fluctuating glucose levels that they may be at increased risk of serious postoperative complications.
Measuring a patient’s blood sugar is neither challenging nor prohibitively expensive. So why don’t we monitor it more closely? Probably because, until now, we have not had a compelling reason to do so with “low-risk” patients. What this study suggests is that our definition of a “low-risk” patient from a glycemic-control standpoint may be misinformed. And while further research needs to be performed to corroborate these findings, that is a pretty scary thought to digest.
Chad A. Krueger, MD
JBJS Deputy Editor for Social Media