User login
Bacteroides Fragilis Vertebral Osteomyelitis and Discitis: “Back” to Susceptibility Testing
Acute pyogenic vertebral osteomyelitis is often due to hematogenous spread of aerobic bacteria.1-4 Conversely, only 0.5% of anaerobic bacteremias lead to osteomyelitis.5 Anaerobic osteomyelitis typically results from the contiguous spread of polymicrobial infections through breaks in the gut mucosal barrier and involves the vertebral bodies in only 2% to 5% of cases.5,6 Although Bacteroides fragilis (B fragilis) is the most common anaerobic pathogen cultivated from blood, accounting for about half of all anaerobic blood isolates, it seldom leads to osteomyelitis.1,2,7-11 We report an uncommon case of B fragilis bacteremia and vertebral osteomyelitis confounded by uncertainties in anaerobic identification and susceptibilities.
Case Presentation
A healthy-appearing male aged 55 years presented to the Naval Medical Center Portsmouth (NMCP) with subacute low back pain and fevers of 103 °F for > 3 weeks. While traveling 4 weeks prior, he completed a course of oseltamivir for influenza B infection; afterward, he was diagnosed with community-acquired pneumonia and treated with a dose of ceftriaxone and a 7-day course of doxycycline. The patient presented to the same facility a week later for low back pain and nonresolving respiratory symptoms, and his therapy was changed to azithromycin, cefuroxime, prednisone, and inhalers. Additionally, after being treated for influenza, he developed constipation and hematochezia for which he did not seek care. The hematochezia was similar to a previous episode from an anal fissure 1 year prior that resolved with stool softeners. When he was finally seen at NMCP after 3 weeks of worsening back pain and fevers, lumbosacral magnetic resonance imaging (MRI) demonstrated vertebral osteomyelitis and discitis at L4-L5 and admitted to the hospital (Figure 1).
After a fluoroscopy-guided biopsy of the L4 vertebral body on hospital day 1, the patient was started on cefepime and vancomycin. The biopsy sample was inoculated onto solid media (blood agar, chocolate agar, and MacConkey agar) and incubated at 36 °C for 24 hours in a 5% CO2 atmosphere, as well as onto Shaedler agar with vitamin K and chopped meat glucose broth and incubated at 36 °C for 48 hours under anaerobic conditions. Metronidazole was added and vancomycin discontinued after 2 anaerobic blood culture vials obtained on hospital day 1, incubated in a Becton Dickinson BACTEC FX automated system, which demonstrated Gram-negative bacilli after 48 hours. The blood culture isolates demonstrated a > 99% probability of being identified as ß-lactamase positive Prevotella loescheii using Thermo Fischer Scientific RapID ANA II biochemical testing. Nitrocefinase discs were used to detect ß-lactamase activity.
The biopsy demonstrated nongranulomatous focal areas of necrotic bone and neutrophilia in a hematopoietic background consistent with acute osteomyelitis (Figure 2); on hospital day 4, ß-lactamase positive B fragilis grew from the bone culture. Additionally, 1 anaerobic vial from a surveillance blood culture set that was obtained on hospital day 3 grew ß-lactamasepositive B fragilis using the same identification methods. With these results he was thought to have a polymicrobial infection (B fragilis and Prevotella loescheii [P loescheii]) from a suspected bowel source based on his hematochezia and history of anal fissure. No aerobic, Gram-negative enterobacteriaceae were isolated, but he had previously been on cefuroxime, which has potential activity against these organisms, for ≥ 2 weeks prior to hospitalization and cultures. He was discharged on moxifloxacin and metronidazole pending final culture results, including requested anaerobic susceptibility testing.
At 1-week follow-up, both aerobic and anaerobic vials from surveillance blood cultures remained negative for any microbes, so antibiotics were deescalated to moxifloxacin monotherapy. However, after 3 days the patient was readmitted for increasing C-reactive protein (CRP) levels and intractable back pain with worsening bilateral radiculopathy. A repeat MRI demonstrated interval disease progression with near obliteration of the L4-L5 disc space and hyperenhancement of the prevertebral soft tissues and adjacent psoas musculature without focal rim-enhancing fluid collection (Figure 3). After repeat L4 biopsy, metronidazole was restarted and ertapenem added for enterobacteriaceae coverage, given the known B fragilis and potential suppression from previous cephalosporin therapy; moxifloxacin was discontinued. L4 biopsy cultures showed no growth, and CRP levels trended down from 154.2 mg/L (start of first admission) to 42.4 mg/L (start of second admission) to 14.9 mg/L (day of discharge) (reference range, 5-9.9 mg/L). He was discharged on ertapenem and metronidazole. He completed a 6-week course without further complication.
During antibiotic therapy he had an unremarkable colonoscopy, CRP normalized to 2.6 mg/L (reference range, 0-4.9 mg/L), and he underwent successful L4-L5 transforaminal lumbar interbody fusion 2 weeks after finishing antibiotics.
We retroactively sent both P loescheii isolates and the 1 B fragilis isolate that grew from the surveillance blood culture to the Multidrug-resistant Organism Repository and Surveillance Network (MRSN) at the Walter Reed Army Institute of Research for identification confirmation and susceptibility analysis. Whole genome sequencing with single nucleotide polymorphism (SNP)-based analysis revealed all isolates were 100% identical and consistent with B fragilis and not P loescheii, based on clustering around other B fragilis sequences found in the National Center for Biotechnology Information (NCBI) Genbank database (Figure 4). All isolates carried the antibiotic resistance genes— cepA, sul(2), tetQ— encoding for possible resistance to cephalosporins, sulphonamides, and tetracyclines, respectively; as well as a point mutation in the gyrA gene (Ser82Phe). None of the isolates carried the nim gene, and screening for the 3 subtypes of B fragilis enterotoxin gene (bft-1, bft-2, bft-3) was negative. Eventual susceptibility testing at the Mayo Clinic several months after the conclusion of the case indicated that the B fragilis isolate was sensitive to piperacillin-tazobactam, ertapenem, clindamycin, and metronidazole; however, testing was not performed against moxifloxacin.
Discussion
In the era of growing antibiotic resistance patterns, antimicrobial stewardship programs recommend interventions to improve antimicrobial use through targeted narrow- spectrum antibiotics.12 The Clinical and Laboratory Standards Institute (CLSI) maintains guidelines on the major indications for anaerobic antimicrobial susceptibility testing (AST) to help direct narrow-targeted antimicrobial therapy. However, in a 2008 practice survey Goldstein and colleagues reported that less than half of US hospitals performed anaerobic AST, and only 21% of these facilities did it in-house, while the remainder sent out their isolates for testing.11-14 The CLSI major indications for AST include situations in which the selection of agents is important because of the (1) known resistance of a particular species; (2) confirmation of appropriate therapy for severe infections or for those that may require long-term therapy; (3) persistence of infection despite adequate treatment with an appropriate therapeutic regimen; and (4) difficulty in making empirical decisions based on precedent.14 Additionally, isolates from brain abscess, endocarditis, osteomyelitis, joint infection, infection of prosthetic devices or vascular grafts, bacteremia, and normally sterile body sites (unless contamination suspected) should be tested.14
Because of the lack of anaerobic AST, health care providers must base empiric treatment on reported sensitivities from the medical literature. Empiric selection of antimicrobials for anaerobic infections is made even more challenging by the increased rates of resistance reported in the literature, leading to recommendations to increase susceptibility testing to guide therapy.13,15,16 Empiric therapy of deep-seated anaerobic infections may lead to use of inactive agents or overly broad-spectrum antibiotics. Current antimicrobial stewardship initiatives recognize the importance of narrow-spectrum antibiotics to minimize risk of adverse events and selective pressure for antimicrobial resistance.
Although we attempted to confirm the identification of the anaerobic isolates via commercially available methods, it was not until we performed genetic testing that we were able verify the isolates as B fragilis. Furthermore, earlier susceptibility testing would have allowed for more narrow-targeted antimicrobial therapy and could have potentially prevented our patient’s readmission and use of ertapenem, despite its > 98% susceptibility rates against B fragilis.13,17
All of the B fragilis isolates carried the cepA gene, which is a cephalosporinase that encodes for resistance to cephalosporins and aminopenicillins but not to ß-lactam ß-lactamase inhibitor combinations.13 Although not a substitution for susceptibility analysis, genetic testing showed that all of the isolates carried a nonsynonymous mutation from serine to a phenylalanine at amino acid position 82 (S82F) in the gyrA gene. The S82F mutation has been implicated in fluoroquinolone resistance, via inhibition of substrate–target recognition and binding between fluoroquinolones and the target topoisomerase protein,18 and may potentially explain why our patient clinically worsened while on moxifloxacin monotherapy. Although moxifloxacin susceptibility was not performed, susceptibility rates remain highly variable, ranging from 50% to 70% for B fragilis.13,15,16
It is important to note that the metronidazole the patient received during his first hospital admission could have sterilized the vertebral body without completely eradicating the microbe; thus could explain his clinical worsening while on moxifloxacin monotherapy despite no growth from the repeat biopsy culture. Our rationale for initially continuing moxifloxacin was based on its excellent bioavailability and bone penetration properties. Additionally, of the fluoroquinolones it has the most reliable anaerobic activity and is the only one recommended as monotherapy for complicated intraabdominal infections.19 However, guidelines recommend avoiding its use in patients who have received a fluoroquinolone in the past 90 days or at institutions with high rates of resistance. At our institution Escherichia coli has a > 90% susceptibility rate to fluoroquinolones. Given this rate and our concern that the patient had a polymicrobial infection, we felt that moxifloxacin would provide appropriate anaerobic and aerobic coverage, especially since he had no previous fluoroquinolone exposure.
Additionally, none of the isolates carried the nim or bft toxin genes. Although the nim gene is associated with metronidazole resistance,its presence does not invariably result in resistant strains of B fragilis; in fact, metronidazole resistance is relatively uncommon, with the majority of B fragilis showing < 1% resistance, based on CLSI breakpoints (≥ 32 mg/L).13,20,21 However, one recent epidemiologic study on anaerobic wound isolates from Iraq and Afghanistan casualties found that 12% (2/17) of B fragilis isolates were resistant to metronidazole.15 Given the improvement of the patient’s symptoms while on metronidazole, it is likely that the B fragilis was susceptible. Nevertheless, susceptibility testing with minimum inhibitory concentrations is necessary to verify this result. Also, although enterotoxigenic strains of B fragilis have been associated with bloodstream infections, our patient’s isolates lacked the 3 subtypes of B fragilis enterotoxin gene.22
Conclusions
We report a case of B fragilis bacteremia and vertebral osteomyelitis complicated by challenges in anaerobic identification and sensitivities that led to brief use of a possibly inactive antimicrobial and the subsequent use of carbapenem therapy, which may have been avoided if susceptibility testing were more readily available. This case led to changes in our hospital’s processing of anaerobic isolates to include susceptibility testing on request.
Acknowledgments
We thank Keith Thompson, MD (staff pathologist, Naval Medical Center Portsmouth Virginia), for providing the pathology images from the initial vertebral biopsy, and Dr. Kate Hinkle (director, Multidrug-Resistant Organism Repository and Surveillance Network, Silver Spring, Maryland ) for providing the whole genome sequencing results from the B fragilis isolates.
1. Zimmerli W. Vertebral osteomyelitis. N Eng J Med. 2010;362(11):1022-1029.
2. Chazan B, Strahilevitz J, Millgram MA, Kaufmann S, Raz R. Bacteroides fragilis vertebral osteomyelitis secondary to anal dilatation. Spine (Phila PA 1976). 2001;26(16):E377-E378.
3. Kierzkowska M, Pedzisz P, Babiak I, et al. Orthopedic infections caused by obligatory anaerobic Gram-negative rods: report of two cases. Med Microbiol Immunol. 2017;206(5):363-366.
4. McHenry M, Easley K, Locker G. Vertebral osteomyelitis: long-term outcome for 253 patients from 7 Cleveland-area hospitals. Clin Infect Dis. 2002;34(10):1342-1350.
5. Raff MJ, Melo JC. Anaerobic osteomyelitis. Medicine (Baltimore).1978;57(1):83-103.
6. Lewis R, Sutter V, Finegold S. Bone infections involving anaerobic bacteria. Medicine (Baltimore). 1978;57(1):279-305.
7. Brook I. The role of anaerobic bacteria in bacteremia. Anaerobe. 2010;16(3):183-189.
8. Lassmann B, Gustafson DR, Wood CM, Rosenblatt JE. Reemergence of anaerobic bacteremia. Clin Infect Dis. 2007;44(7):895-900.
9. Lazarovitch T, Freimann S, Shapira G, Blank H. Decrease in anaerobe-related bacteraemias and increase in Bacteroides species isolation rate from 1998 to 2007: a retrospective study. Anaerobe. 2010;16(3):201-205.
10. Keukeleire S, Wybo I, Naessens A, et al. Anaerobic bacteraemia: a 10-year retrospective epidemiological survey. Anaerobe. 2016;39:54-59.
11. Goldstein EJC, Citron DM, Goldman PJ, Goldman RJ. National hospital survey of anaerobic culture and susceptibility methods: III. Anaerobe. 2008;14(2):68-72.
12. Barlam TF, Cosgrove SE, Abbo LM, et al. Implementing an antibiotic stewardship program: Guidelines by the Infectious Diseases Society of America and the Society for Healthcare Epidemiology of America. Clin Infect Dis. 2016;62(10):e51-e77.
13. Schuetz AN. Antimicrobial resistance and susceptibility testing of anaerobic bacteria. Antimicr Resist. 2014;59(5):698-705.
14. Clinical and Laboratory Standards Institute. M11-A8: Methods for Antimicrobial Susceptibility Testing of Anaerobic Bacteria; Approved Standard. 8th ed. Wayne, PA: Clinical and Laboratory Standards Institute; 2012.
15. White B, Mende K, Weintrob A, et al; Infectious Disease Clinical Research Program Trauma Infectious Disease Outcome Study Group. Epidemiology and antimicrobial susceptibilities of wound isolates of obligate anaerobes from combat casualties. Diagn Mircrobiol Infect Dis. 2016;84(2):144-150.
16. Hastey CJ, Boyd H, Schuetz AN, et al; Ad Hoc Working Group on Antimicrobial Susceptibility Testing of Anaerobic Bacteria of CLSI. Changes in the antibiotic susceptibility of anaerobic bacteria from 2007-2009 to 2010-2012 based on the CLSI methodology. Anaerobe. 2016;42:27-30.
17. Brook I, Wexler HM, Goldstein EJC. Antianaerobic antimicrobials: spectrum and susceptibility testing. Clin Microbiol Rev. 2013;26(3):526-546.
18. Pumbwe L, Wareham D, Aduse-Opoku J, Brazier JS, Wexler HM. Genetic analysis of mechanisms of multidrug resistance in a clinical isolate of Bacteroides fragilis. Clin Microbiol Infect. 2007;13(2):183-189.
19. Solomkin J, Mazuski J, Bradley J, et al. Diagnosis and management of complicated intra-abdominal infection in adults and children: guidelines by the Surgical Infection Society and the Infectious Diseases Society of America. Clin Infect Dis. 2010;50(2):133-164.
20. Breuil J, Dublanchet A, Truffaut N, Sebald M. Transferable 5-nitroimidazole resistance Bacteroides fragilis group. Plasmid. 1989;21(2):151-154.
21. Nagy E, Urbán E, Nord CE; ESCMID Study Group on Antimicrobial Resistance in Anaerobic Bacteria. Antimicrobial susceptibility of Bacteroides fragilis group isolates in Europe: 20 years of experience. Clin Microbiol Infect. 2011;17(3):371-379.
22. Avila-Campos M, Liu C, Song Y, Rowlinson M-C, Finegold SM. Determination of bft gene subtypes in Bacteroides fragilis clinical isolates. J Clin Microbiol. 2007;45(4):1336-1338.
Acute pyogenic vertebral osteomyelitis is often due to hematogenous spread of aerobic bacteria.1-4 Conversely, only 0.5% of anaerobic bacteremias lead to osteomyelitis.5 Anaerobic osteomyelitis typically results from the contiguous spread of polymicrobial infections through breaks in the gut mucosal barrier and involves the vertebral bodies in only 2% to 5% of cases.5,6 Although Bacteroides fragilis (B fragilis) is the most common anaerobic pathogen cultivated from blood, accounting for about half of all anaerobic blood isolates, it seldom leads to osteomyelitis.1,2,7-11 We report an uncommon case of B fragilis bacteremia and vertebral osteomyelitis confounded by uncertainties in anaerobic identification and susceptibilities.
Case Presentation
A healthy-appearing male aged 55 years presented to the Naval Medical Center Portsmouth (NMCP) with subacute low back pain and fevers of 103 °F for > 3 weeks. While traveling 4 weeks prior, he completed a course of oseltamivir for influenza B infection; afterward, he was diagnosed with community-acquired pneumonia and treated with a dose of ceftriaxone and a 7-day course of doxycycline. The patient presented to the same facility a week later for low back pain and nonresolving respiratory symptoms, and his therapy was changed to azithromycin, cefuroxime, prednisone, and inhalers. Additionally, after being treated for influenza, he developed constipation and hematochezia for which he did not seek care. The hematochezia was similar to a previous episode from an anal fissure 1 year prior that resolved with stool softeners. When he was finally seen at NMCP after 3 weeks of worsening back pain and fevers, lumbosacral magnetic resonance imaging (MRI) demonstrated vertebral osteomyelitis and discitis at L4-L5 and admitted to the hospital (Figure 1).
After a fluoroscopy-guided biopsy of the L4 vertebral body on hospital day 1, the patient was started on cefepime and vancomycin. The biopsy sample was inoculated onto solid media (blood agar, chocolate agar, and MacConkey agar) and incubated at 36 °C for 24 hours in a 5% CO2 atmosphere, as well as onto Shaedler agar with vitamin K and chopped meat glucose broth and incubated at 36 °C for 48 hours under anaerobic conditions. Metronidazole was added and vancomycin discontinued after 2 anaerobic blood culture vials obtained on hospital day 1, incubated in a Becton Dickinson BACTEC FX automated system, which demonstrated Gram-negative bacilli after 48 hours. The blood culture isolates demonstrated a > 99% probability of being identified as ß-lactamase positive Prevotella loescheii using Thermo Fischer Scientific RapID ANA II biochemical testing. Nitrocefinase discs were used to detect ß-lactamase activity.
The biopsy demonstrated nongranulomatous focal areas of necrotic bone and neutrophilia in a hematopoietic background consistent with acute osteomyelitis (Figure 2); on hospital day 4, ß-lactamase positive B fragilis grew from the bone culture. Additionally, 1 anaerobic vial from a surveillance blood culture set that was obtained on hospital day 3 grew ß-lactamasepositive B fragilis using the same identification methods. With these results he was thought to have a polymicrobial infection (B fragilis and Prevotella loescheii [P loescheii]) from a suspected bowel source based on his hematochezia and history of anal fissure. No aerobic, Gram-negative enterobacteriaceae were isolated, but he had previously been on cefuroxime, which has potential activity against these organisms, for ≥ 2 weeks prior to hospitalization and cultures. He was discharged on moxifloxacin and metronidazole pending final culture results, including requested anaerobic susceptibility testing.
At 1-week follow-up, both aerobic and anaerobic vials from surveillance blood cultures remained negative for any microbes, so antibiotics were deescalated to moxifloxacin monotherapy. However, after 3 days the patient was readmitted for increasing C-reactive protein (CRP) levels and intractable back pain with worsening bilateral radiculopathy. A repeat MRI demonstrated interval disease progression with near obliteration of the L4-L5 disc space and hyperenhancement of the prevertebral soft tissues and adjacent psoas musculature without focal rim-enhancing fluid collection (Figure 3). After repeat L4 biopsy, metronidazole was restarted and ertapenem added for enterobacteriaceae coverage, given the known B fragilis and potential suppression from previous cephalosporin therapy; moxifloxacin was discontinued. L4 biopsy cultures showed no growth, and CRP levels trended down from 154.2 mg/L (start of first admission) to 42.4 mg/L (start of second admission) to 14.9 mg/L (day of discharge) (reference range, 5-9.9 mg/L). He was discharged on ertapenem and metronidazole. He completed a 6-week course without further complication.
During antibiotic therapy he had an unremarkable colonoscopy, CRP normalized to 2.6 mg/L (reference range, 0-4.9 mg/L), and he underwent successful L4-L5 transforaminal lumbar interbody fusion 2 weeks after finishing antibiotics.
We retroactively sent both P loescheii isolates and the 1 B fragilis isolate that grew from the surveillance blood culture to the Multidrug-resistant Organism Repository and Surveillance Network (MRSN) at the Walter Reed Army Institute of Research for identification confirmation and susceptibility analysis. Whole genome sequencing with single nucleotide polymorphism (SNP)-based analysis revealed all isolates were 100% identical and consistent with B fragilis and not P loescheii, based on clustering around other B fragilis sequences found in the National Center for Biotechnology Information (NCBI) Genbank database (Figure 4). All isolates carried the antibiotic resistance genes— cepA, sul(2), tetQ— encoding for possible resistance to cephalosporins, sulphonamides, and tetracyclines, respectively; as well as a point mutation in the gyrA gene (Ser82Phe). None of the isolates carried the nim gene, and screening for the 3 subtypes of B fragilis enterotoxin gene (bft-1, bft-2, bft-3) was negative. Eventual susceptibility testing at the Mayo Clinic several months after the conclusion of the case indicated that the B fragilis isolate was sensitive to piperacillin-tazobactam, ertapenem, clindamycin, and metronidazole; however, testing was not performed against moxifloxacin.
Discussion
In the era of growing antibiotic resistance patterns, antimicrobial stewardship programs recommend interventions to improve antimicrobial use through targeted narrow- spectrum antibiotics.12 The Clinical and Laboratory Standards Institute (CLSI) maintains guidelines on the major indications for anaerobic antimicrobial susceptibility testing (AST) to help direct narrow-targeted antimicrobial therapy. However, in a 2008 practice survey Goldstein and colleagues reported that less than half of US hospitals performed anaerobic AST, and only 21% of these facilities did it in-house, while the remainder sent out their isolates for testing.11-14 The CLSI major indications for AST include situations in which the selection of agents is important because of the (1) known resistance of a particular species; (2) confirmation of appropriate therapy for severe infections or for those that may require long-term therapy; (3) persistence of infection despite adequate treatment with an appropriate therapeutic regimen; and (4) difficulty in making empirical decisions based on precedent.14 Additionally, isolates from brain abscess, endocarditis, osteomyelitis, joint infection, infection of prosthetic devices or vascular grafts, bacteremia, and normally sterile body sites (unless contamination suspected) should be tested.14
Because of the lack of anaerobic AST, health care providers must base empiric treatment on reported sensitivities from the medical literature. Empiric selection of antimicrobials for anaerobic infections is made even more challenging by the increased rates of resistance reported in the literature, leading to recommendations to increase susceptibility testing to guide therapy.13,15,16 Empiric therapy of deep-seated anaerobic infections may lead to use of inactive agents or overly broad-spectrum antibiotics. Current antimicrobial stewardship initiatives recognize the importance of narrow-spectrum antibiotics to minimize risk of adverse events and selective pressure for antimicrobial resistance.
Although we attempted to confirm the identification of the anaerobic isolates via commercially available methods, it was not until we performed genetic testing that we were able verify the isolates as B fragilis. Furthermore, earlier susceptibility testing would have allowed for more narrow-targeted antimicrobial therapy and could have potentially prevented our patient’s readmission and use of ertapenem, despite its > 98% susceptibility rates against B fragilis.13,17
All of the B fragilis isolates carried the cepA gene, which is a cephalosporinase that encodes for resistance to cephalosporins and aminopenicillins but not to ß-lactam ß-lactamase inhibitor combinations.13 Although not a substitution for susceptibility analysis, genetic testing showed that all of the isolates carried a nonsynonymous mutation from serine to a phenylalanine at amino acid position 82 (S82F) in the gyrA gene. The S82F mutation has been implicated in fluoroquinolone resistance, via inhibition of substrate–target recognition and binding between fluoroquinolones and the target topoisomerase protein,18 and may potentially explain why our patient clinically worsened while on moxifloxacin monotherapy. Although moxifloxacin susceptibility was not performed, susceptibility rates remain highly variable, ranging from 50% to 70% for B fragilis.13,15,16
It is important to note that the metronidazole the patient received during his first hospital admission could have sterilized the vertebral body without completely eradicating the microbe; thus could explain his clinical worsening while on moxifloxacin monotherapy despite no growth from the repeat biopsy culture. Our rationale for initially continuing moxifloxacin was based on its excellent bioavailability and bone penetration properties. Additionally, of the fluoroquinolones it has the most reliable anaerobic activity and is the only one recommended as monotherapy for complicated intraabdominal infections.19 However, guidelines recommend avoiding its use in patients who have received a fluoroquinolone in the past 90 days or at institutions with high rates of resistance. At our institution Escherichia coli has a > 90% susceptibility rate to fluoroquinolones. Given this rate and our concern that the patient had a polymicrobial infection, we felt that moxifloxacin would provide appropriate anaerobic and aerobic coverage, especially since he had no previous fluoroquinolone exposure.
Additionally, none of the isolates carried the nim or bft toxin genes. Although the nim gene is associated with metronidazole resistance,its presence does not invariably result in resistant strains of B fragilis; in fact, metronidazole resistance is relatively uncommon, with the majority of B fragilis showing < 1% resistance, based on CLSI breakpoints (≥ 32 mg/L).13,20,21 However, one recent epidemiologic study on anaerobic wound isolates from Iraq and Afghanistan casualties found that 12% (2/17) of B fragilis isolates were resistant to metronidazole.15 Given the improvement of the patient’s symptoms while on metronidazole, it is likely that the B fragilis was susceptible. Nevertheless, susceptibility testing with minimum inhibitory concentrations is necessary to verify this result. Also, although enterotoxigenic strains of B fragilis have been associated with bloodstream infections, our patient’s isolates lacked the 3 subtypes of B fragilis enterotoxin gene.22
Conclusions
We report a case of B fragilis bacteremia and vertebral osteomyelitis complicated by challenges in anaerobic identification and sensitivities that led to brief use of a possibly inactive antimicrobial and the subsequent use of carbapenem therapy, which may have been avoided if susceptibility testing were more readily available. This case led to changes in our hospital’s processing of anaerobic isolates to include susceptibility testing on request.
Acknowledgments
We thank Keith Thompson, MD (staff pathologist, Naval Medical Center Portsmouth Virginia), for providing the pathology images from the initial vertebral biopsy, and Dr. Kate Hinkle (director, Multidrug-Resistant Organism Repository and Surveillance Network, Silver Spring, Maryland ) for providing the whole genome sequencing results from the B fragilis isolates.
Acute pyogenic vertebral osteomyelitis is often due to hematogenous spread of aerobic bacteria.1-4 Conversely, only 0.5% of anaerobic bacteremias lead to osteomyelitis.5 Anaerobic osteomyelitis typically results from the contiguous spread of polymicrobial infections through breaks in the gut mucosal barrier and involves the vertebral bodies in only 2% to 5% of cases.5,6 Although Bacteroides fragilis (B fragilis) is the most common anaerobic pathogen cultivated from blood, accounting for about half of all anaerobic blood isolates, it seldom leads to osteomyelitis.1,2,7-11 We report an uncommon case of B fragilis bacteremia and vertebral osteomyelitis confounded by uncertainties in anaerobic identification and susceptibilities.
Case Presentation
A healthy-appearing male aged 55 years presented to the Naval Medical Center Portsmouth (NMCP) with subacute low back pain and fevers of 103 °F for > 3 weeks. While traveling 4 weeks prior, he completed a course of oseltamivir for influenza B infection; afterward, he was diagnosed with community-acquired pneumonia and treated with a dose of ceftriaxone and a 7-day course of doxycycline. The patient presented to the same facility a week later for low back pain and nonresolving respiratory symptoms, and his therapy was changed to azithromycin, cefuroxime, prednisone, and inhalers. Additionally, after being treated for influenza, he developed constipation and hematochezia for which he did not seek care. The hematochezia was similar to a previous episode from an anal fissure 1 year prior that resolved with stool softeners. When he was finally seen at NMCP after 3 weeks of worsening back pain and fevers, lumbosacral magnetic resonance imaging (MRI) demonstrated vertebral osteomyelitis and discitis at L4-L5 and admitted to the hospital (Figure 1).
After a fluoroscopy-guided biopsy of the L4 vertebral body on hospital day 1, the patient was started on cefepime and vancomycin. The biopsy sample was inoculated onto solid media (blood agar, chocolate agar, and MacConkey agar) and incubated at 36 °C for 24 hours in a 5% CO2 atmosphere, as well as onto Shaedler agar with vitamin K and chopped meat glucose broth and incubated at 36 °C for 48 hours under anaerobic conditions. Metronidazole was added and vancomycin discontinued after 2 anaerobic blood culture vials obtained on hospital day 1, incubated in a Becton Dickinson BACTEC FX automated system, which demonstrated Gram-negative bacilli after 48 hours. The blood culture isolates demonstrated a > 99% probability of being identified as ß-lactamase positive Prevotella loescheii using Thermo Fischer Scientific RapID ANA II biochemical testing. Nitrocefinase discs were used to detect ß-lactamase activity.
The biopsy demonstrated nongranulomatous focal areas of necrotic bone and neutrophilia in a hematopoietic background consistent with acute osteomyelitis (Figure 2); on hospital day 4, ß-lactamase positive B fragilis grew from the bone culture. Additionally, 1 anaerobic vial from a surveillance blood culture set that was obtained on hospital day 3 grew ß-lactamasepositive B fragilis using the same identification methods. With these results he was thought to have a polymicrobial infection (B fragilis and Prevotella loescheii [P loescheii]) from a suspected bowel source based on his hematochezia and history of anal fissure. No aerobic, Gram-negative enterobacteriaceae were isolated, but he had previously been on cefuroxime, which has potential activity against these organisms, for ≥ 2 weeks prior to hospitalization and cultures. He was discharged on moxifloxacin and metronidazole pending final culture results, including requested anaerobic susceptibility testing.
At 1-week follow-up, both aerobic and anaerobic vials from surveillance blood cultures remained negative for any microbes, so antibiotics were deescalated to moxifloxacin monotherapy. However, after 3 days the patient was readmitted for increasing C-reactive protein (CRP) levels and intractable back pain with worsening bilateral radiculopathy. A repeat MRI demonstrated interval disease progression with near obliteration of the L4-L5 disc space and hyperenhancement of the prevertebral soft tissues and adjacent psoas musculature without focal rim-enhancing fluid collection (Figure 3). After repeat L4 biopsy, metronidazole was restarted and ertapenem added for enterobacteriaceae coverage, given the known B fragilis and potential suppression from previous cephalosporin therapy; moxifloxacin was discontinued. L4 biopsy cultures showed no growth, and CRP levels trended down from 154.2 mg/L (start of first admission) to 42.4 mg/L (start of second admission) to 14.9 mg/L (day of discharge) (reference range, 5-9.9 mg/L). He was discharged on ertapenem and metronidazole. He completed a 6-week course without further complication.
During antibiotic therapy he had an unremarkable colonoscopy, CRP normalized to 2.6 mg/L (reference range, 0-4.9 mg/L), and he underwent successful L4-L5 transforaminal lumbar interbody fusion 2 weeks after finishing antibiotics.
We retroactively sent both P loescheii isolates and the 1 B fragilis isolate that grew from the surveillance blood culture to the Multidrug-resistant Organism Repository and Surveillance Network (MRSN) at the Walter Reed Army Institute of Research for identification confirmation and susceptibility analysis. Whole genome sequencing with single nucleotide polymorphism (SNP)-based analysis revealed all isolates were 100% identical and consistent with B fragilis and not P loescheii, based on clustering around other B fragilis sequences found in the National Center for Biotechnology Information (NCBI) Genbank database (Figure 4). All isolates carried the antibiotic resistance genes— cepA, sul(2), tetQ— encoding for possible resistance to cephalosporins, sulphonamides, and tetracyclines, respectively; as well as a point mutation in the gyrA gene (Ser82Phe). None of the isolates carried the nim gene, and screening for the 3 subtypes of B fragilis enterotoxin gene (bft-1, bft-2, bft-3) was negative. Eventual susceptibility testing at the Mayo Clinic several months after the conclusion of the case indicated that the B fragilis isolate was sensitive to piperacillin-tazobactam, ertapenem, clindamycin, and metronidazole; however, testing was not performed against moxifloxacin.
Discussion
In the era of growing antibiotic resistance patterns, antimicrobial stewardship programs recommend interventions to improve antimicrobial use through targeted narrow- spectrum antibiotics.12 The Clinical and Laboratory Standards Institute (CLSI) maintains guidelines on the major indications for anaerobic antimicrobial susceptibility testing (AST) to help direct narrow-targeted antimicrobial therapy. However, in a 2008 practice survey Goldstein and colleagues reported that less than half of US hospitals performed anaerobic AST, and only 21% of these facilities did it in-house, while the remainder sent out their isolates for testing.11-14 The CLSI major indications for AST include situations in which the selection of agents is important because of the (1) known resistance of a particular species; (2) confirmation of appropriate therapy for severe infections or for those that may require long-term therapy; (3) persistence of infection despite adequate treatment with an appropriate therapeutic regimen; and (4) difficulty in making empirical decisions based on precedent.14 Additionally, isolates from brain abscess, endocarditis, osteomyelitis, joint infection, infection of prosthetic devices or vascular grafts, bacteremia, and normally sterile body sites (unless contamination suspected) should be tested.14
Because of the lack of anaerobic AST, health care providers must base empiric treatment on reported sensitivities from the medical literature. Empiric selection of antimicrobials for anaerobic infections is made even more challenging by the increased rates of resistance reported in the literature, leading to recommendations to increase susceptibility testing to guide therapy.13,15,16 Empiric therapy of deep-seated anaerobic infections may lead to use of inactive agents or overly broad-spectrum antibiotics. Current antimicrobial stewardship initiatives recognize the importance of narrow-spectrum antibiotics to minimize risk of adverse events and selective pressure for antimicrobial resistance.
Although we attempted to confirm the identification of the anaerobic isolates via commercially available methods, it was not until we performed genetic testing that we were able verify the isolates as B fragilis. Furthermore, earlier susceptibility testing would have allowed for more narrow-targeted antimicrobial therapy and could have potentially prevented our patient’s readmission and use of ertapenem, despite its > 98% susceptibility rates against B fragilis.13,17
All of the B fragilis isolates carried the cepA gene, which is a cephalosporinase that encodes for resistance to cephalosporins and aminopenicillins but not to ß-lactam ß-lactamase inhibitor combinations.13 Although not a substitution for susceptibility analysis, genetic testing showed that all of the isolates carried a nonsynonymous mutation from serine to a phenylalanine at amino acid position 82 (S82F) in the gyrA gene. The S82F mutation has been implicated in fluoroquinolone resistance, via inhibition of substrate–target recognition and binding between fluoroquinolones and the target topoisomerase protein,18 and may potentially explain why our patient clinically worsened while on moxifloxacin monotherapy. Although moxifloxacin susceptibility was not performed, susceptibility rates remain highly variable, ranging from 50% to 70% for B fragilis.13,15,16
It is important to note that the metronidazole the patient received during his first hospital admission could have sterilized the vertebral body without completely eradicating the microbe; thus could explain his clinical worsening while on moxifloxacin monotherapy despite no growth from the repeat biopsy culture. Our rationale for initially continuing moxifloxacin was based on its excellent bioavailability and bone penetration properties. Additionally, of the fluoroquinolones it has the most reliable anaerobic activity and is the only one recommended as monotherapy for complicated intraabdominal infections.19 However, guidelines recommend avoiding its use in patients who have received a fluoroquinolone in the past 90 days or at institutions with high rates of resistance. At our institution Escherichia coli has a > 90% susceptibility rate to fluoroquinolones. Given this rate and our concern that the patient had a polymicrobial infection, we felt that moxifloxacin would provide appropriate anaerobic and aerobic coverage, especially since he had no previous fluoroquinolone exposure.
Additionally, none of the isolates carried the nim or bft toxin genes. Although the nim gene is associated with metronidazole resistance,its presence does not invariably result in resistant strains of B fragilis; in fact, metronidazole resistance is relatively uncommon, with the majority of B fragilis showing < 1% resistance, based on CLSI breakpoints (≥ 32 mg/L).13,20,21 However, one recent epidemiologic study on anaerobic wound isolates from Iraq and Afghanistan casualties found that 12% (2/17) of B fragilis isolates were resistant to metronidazole.15 Given the improvement of the patient’s symptoms while on metronidazole, it is likely that the B fragilis was susceptible. Nevertheless, susceptibility testing with minimum inhibitory concentrations is necessary to verify this result. Also, although enterotoxigenic strains of B fragilis have been associated with bloodstream infections, our patient’s isolates lacked the 3 subtypes of B fragilis enterotoxin gene.22
Conclusions
We report a case of B fragilis bacteremia and vertebral osteomyelitis complicated by challenges in anaerobic identification and sensitivities that led to brief use of a possibly inactive antimicrobial and the subsequent use of carbapenem therapy, which may have been avoided if susceptibility testing were more readily available. This case led to changes in our hospital’s processing of anaerobic isolates to include susceptibility testing on request.
Acknowledgments
We thank Keith Thompson, MD (staff pathologist, Naval Medical Center Portsmouth Virginia), for providing the pathology images from the initial vertebral biopsy, and Dr. Kate Hinkle (director, Multidrug-Resistant Organism Repository and Surveillance Network, Silver Spring, Maryland ) for providing the whole genome sequencing results from the B fragilis isolates.
1. Zimmerli W. Vertebral osteomyelitis. N Eng J Med. 2010;362(11):1022-1029.
2. Chazan B, Strahilevitz J, Millgram MA, Kaufmann S, Raz R. Bacteroides fragilis vertebral osteomyelitis secondary to anal dilatation. Spine (Phila PA 1976). 2001;26(16):E377-E378.
3. Kierzkowska M, Pedzisz P, Babiak I, et al. Orthopedic infections caused by obligatory anaerobic Gram-negative rods: report of two cases. Med Microbiol Immunol. 2017;206(5):363-366.
4. McHenry M, Easley K, Locker G. Vertebral osteomyelitis: long-term outcome for 253 patients from 7 Cleveland-area hospitals. Clin Infect Dis. 2002;34(10):1342-1350.
5. Raff MJ, Melo JC. Anaerobic osteomyelitis. Medicine (Baltimore).1978;57(1):83-103.
6. Lewis R, Sutter V, Finegold S. Bone infections involving anaerobic bacteria. Medicine (Baltimore). 1978;57(1):279-305.
7. Brook I. The role of anaerobic bacteria in bacteremia. Anaerobe. 2010;16(3):183-189.
8. Lassmann B, Gustafson DR, Wood CM, Rosenblatt JE. Reemergence of anaerobic bacteremia. Clin Infect Dis. 2007;44(7):895-900.
9. Lazarovitch T, Freimann S, Shapira G, Blank H. Decrease in anaerobe-related bacteraemias and increase in Bacteroides species isolation rate from 1998 to 2007: a retrospective study. Anaerobe. 2010;16(3):201-205.
10. Keukeleire S, Wybo I, Naessens A, et al. Anaerobic bacteraemia: a 10-year retrospective epidemiological survey. Anaerobe. 2016;39:54-59.
11. Goldstein EJC, Citron DM, Goldman PJ, Goldman RJ. National hospital survey of anaerobic culture and susceptibility methods: III. Anaerobe. 2008;14(2):68-72.
12. Barlam TF, Cosgrove SE, Abbo LM, et al. Implementing an antibiotic stewardship program: Guidelines by the Infectious Diseases Society of America and the Society for Healthcare Epidemiology of America. Clin Infect Dis. 2016;62(10):e51-e77.
13. Schuetz AN. Antimicrobial resistance and susceptibility testing of anaerobic bacteria. Antimicr Resist. 2014;59(5):698-705.
14. Clinical and Laboratory Standards Institute. M11-A8: Methods for Antimicrobial Susceptibility Testing of Anaerobic Bacteria; Approved Standard. 8th ed. Wayne, PA: Clinical and Laboratory Standards Institute; 2012.
15. White B, Mende K, Weintrob A, et al; Infectious Disease Clinical Research Program Trauma Infectious Disease Outcome Study Group. Epidemiology and antimicrobial susceptibilities of wound isolates of obligate anaerobes from combat casualties. Diagn Mircrobiol Infect Dis. 2016;84(2):144-150.
16. Hastey CJ, Boyd H, Schuetz AN, et al; Ad Hoc Working Group on Antimicrobial Susceptibility Testing of Anaerobic Bacteria of CLSI. Changes in the antibiotic susceptibility of anaerobic bacteria from 2007-2009 to 2010-2012 based on the CLSI methodology. Anaerobe. 2016;42:27-30.
17. Brook I, Wexler HM, Goldstein EJC. Antianaerobic antimicrobials: spectrum and susceptibility testing. Clin Microbiol Rev. 2013;26(3):526-546.
18. Pumbwe L, Wareham D, Aduse-Opoku J, Brazier JS, Wexler HM. Genetic analysis of mechanisms of multidrug resistance in a clinical isolate of Bacteroides fragilis. Clin Microbiol Infect. 2007;13(2):183-189.
19. Solomkin J, Mazuski J, Bradley J, et al. Diagnosis and management of complicated intra-abdominal infection in adults and children: guidelines by the Surgical Infection Society and the Infectious Diseases Society of America. Clin Infect Dis. 2010;50(2):133-164.
20. Breuil J, Dublanchet A, Truffaut N, Sebald M. Transferable 5-nitroimidazole resistance Bacteroides fragilis group. Plasmid. 1989;21(2):151-154.
21. Nagy E, Urbán E, Nord CE; ESCMID Study Group on Antimicrobial Resistance in Anaerobic Bacteria. Antimicrobial susceptibility of Bacteroides fragilis group isolates in Europe: 20 years of experience. Clin Microbiol Infect. 2011;17(3):371-379.
22. Avila-Campos M, Liu C, Song Y, Rowlinson M-C, Finegold SM. Determination of bft gene subtypes in Bacteroides fragilis clinical isolates. J Clin Microbiol. 2007;45(4):1336-1338.
1. Zimmerli W. Vertebral osteomyelitis. N Eng J Med. 2010;362(11):1022-1029.
2. Chazan B, Strahilevitz J, Millgram MA, Kaufmann S, Raz R. Bacteroides fragilis vertebral osteomyelitis secondary to anal dilatation. Spine (Phila PA 1976). 2001;26(16):E377-E378.
3. Kierzkowska M, Pedzisz P, Babiak I, et al. Orthopedic infections caused by obligatory anaerobic Gram-negative rods: report of two cases. Med Microbiol Immunol. 2017;206(5):363-366.
4. McHenry M, Easley K, Locker G. Vertebral osteomyelitis: long-term outcome for 253 patients from 7 Cleveland-area hospitals. Clin Infect Dis. 2002;34(10):1342-1350.
5. Raff MJ, Melo JC. Anaerobic osteomyelitis. Medicine (Baltimore).1978;57(1):83-103.
6. Lewis R, Sutter V, Finegold S. Bone infections involving anaerobic bacteria. Medicine (Baltimore). 1978;57(1):279-305.
7. Brook I. The role of anaerobic bacteria in bacteremia. Anaerobe. 2010;16(3):183-189.
8. Lassmann B, Gustafson DR, Wood CM, Rosenblatt JE. Reemergence of anaerobic bacteremia. Clin Infect Dis. 2007;44(7):895-900.
9. Lazarovitch T, Freimann S, Shapira G, Blank H. Decrease in anaerobe-related bacteraemias and increase in Bacteroides species isolation rate from 1998 to 2007: a retrospective study. Anaerobe. 2010;16(3):201-205.
10. Keukeleire S, Wybo I, Naessens A, et al. Anaerobic bacteraemia: a 10-year retrospective epidemiological survey. Anaerobe. 2016;39:54-59.
11. Goldstein EJC, Citron DM, Goldman PJ, Goldman RJ. National hospital survey of anaerobic culture and susceptibility methods: III. Anaerobe. 2008;14(2):68-72.
12. Barlam TF, Cosgrove SE, Abbo LM, et al. Implementing an antibiotic stewardship program: Guidelines by the Infectious Diseases Society of America and the Society for Healthcare Epidemiology of America. Clin Infect Dis. 2016;62(10):e51-e77.
13. Schuetz AN. Antimicrobial resistance and susceptibility testing of anaerobic bacteria. Antimicr Resist. 2014;59(5):698-705.
14. Clinical and Laboratory Standards Institute. M11-A8: Methods for Antimicrobial Susceptibility Testing of Anaerobic Bacteria; Approved Standard. 8th ed. Wayne, PA: Clinical and Laboratory Standards Institute; 2012.
15. White B, Mende K, Weintrob A, et al; Infectious Disease Clinical Research Program Trauma Infectious Disease Outcome Study Group. Epidemiology and antimicrobial susceptibilities of wound isolates of obligate anaerobes from combat casualties. Diagn Mircrobiol Infect Dis. 2016;84(2):144-150.
16. Hastey CJ, Boyd H, Schuetz AN, et al; Ad Hoc Working Group on Antimicrobial Susceptibility Testing of Anaerobic Bacteria of CLSI. Changes in the antibiotic susceptibility of anaerobic bacteria from 2007-2009 to 2010-2012 based on the CLSI methodology. Anaerobe. 2016;42:27-30.
17. Brook I, Wexler HM, Goldstein EJC. Antianaerobic antimicrobials: spectrum and susceptibility testing. Clin Microbiol Rev. 2013;26(3):526-546.
18. Pumbwe L, Wareham D, Aduse-Opoku J, Brazier JS, Wexler HM. Genetic analysis of mechanisms of multidrug resistance in a clinical isolate of Bacteroides fragilis. Clin Microbiol Infect. 2007;13(2):183-189.
19. Solomkin J, Mazuski J, Bradley J, et al. Diagnosis and management of complicated intra-abdominal infection in adults and children: guidelines by the Surgical Infection Society and the Infectious Diseases Society of America. Clin Infect Dis. 2010;50(2):133-164.
20. Breuil J, Dublanchet A, Truffaut N, Sebald M. Transferable 5-nitroimidazole resistance Bacteroides fragilis group. Plasmid. 1989;21(2):151-154.
21. Nagy E, Urbán E, Nord CE; ESCMID Study Group on Antimicrobial Resistance in Anaerobic Bacteria. Antimicrobial susceptibility of Bacteroides fragilis group isolates in Europe: 20 years of experience. Clin Microbiol Infect. 2011;17(3):371-379.
22. Avila-Campos M, Liu C, Song Y, Rowlinson M-C, Finegold SM. Determination of bft gene subtypes in Bacteroides fragilis clinical isolates. J Clin Microbiol. 2007;45(4):1336-1338.