Clinical characteristics and therapeutic outcomes of hematogenous vertebral osteomyelitis caused by methicillin-resistant Staphylococcus aureus

Journal of Infection (2013) 67, 556e564

www.elsevierhealth.com/journals/jinf

Clinical characteristics and therapeutic outcomes of hematogenous vertebral osteomyelitis caused by methicillin-resistant Staphylococcus aureus Ki-Ho Park a,b,c, Yong Pil Chong a,c, Sung-Han Kim a, Sang-Oh Lee a, Sang-Ho Choi a, Mi Suk Lee b, Jin-Yong Jeong c,d, Jun Hee Woo a, Yang Soo Kim a,c,* a

Department of Infectious Diseases, Asan Medical Center, University of Ulsan College of Medicine, Seoul, Republic of Korea b Division of Infectious Diseases, Department of Internal Medicine, Kyung Hee University Hospital, Kyung Hee University School of Medicine, Seoul, Republic of Korea c Center for Antimicrobial Resistance and Microbial Genetics, University of Ulsan, Seoul, Republic of Korea d Asan Institute for Life Sciences, Asan Medical Center, University of Ulsan College of Medicine, Seoul, Republic of Korea Accepted 25 July 2013 Available online 31 July 2013

KEYWORDS Vertebral osteomyelitis; Staphylococcus aureus; MRSA; Bacteremia; Spondylitis

Summary Objective: To evaluate the clinical characteristics and therapeutic outcomes of patients with hematogenous vertebral osteomyelitis (HVO) caused by methicillin-resistant Staphylococcus aureus (MRSA). Methods: We performed a cohort study of adult patients diagnosed with S. aureus HVO at a tertiary-care hospital over a 7-year period. Results: Of the 139 patients with S. aureus HVO, MRSA caused 62 (44.6%) cases. In multivariate analysis, compared with methicillin-susceptible S. aureus (MSSA), MRSA was associated with a higher risk of persistent bacteremia (
7 days) (8.40 fold; P < 0.001) and relapse (4.83 fold; P Z 0.03), and increased hospital stay (1.69 fold; P Z 0.001). Among the MRSA cases, relapse rates differed according to duration of antibiotics: 41.7% (4e6 weeks), 25.0% (6e8 weeks), and 5.6% (
8 weeks) (P Z 0.007). Bacteremia was more likely to persist for
7 days in patients with an initial vancomycin trough <15 mg/L than in those with an initial trough
15 mg/L (79.3% vs. 20.0%; P Z 0.001).

* Corresponding author. Department of Infectious Diseases, Asan Medical Center, University of Ulsan College of Medicine, 88, Olympic-ro 43-gil, Songpa-gu, Seoul 138-736, Republic of Korea. Tel.: þ82 2 3010 3303; fax: þ82 3010 6970. E-mail address: [email protected] (Y.S. Kim). 0163-4453/$36 ª 2013 The British Infection Association. Published by Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.jinf.2013.07.026

MRSA vertebral osteomyelitis

557 Conclusions: MRSA HVO was associated with more frequent persistent bacteremia (
7 days) and relapse, and longer hospital stay compared to MSSA HVO. Antibiotic therapy for
8 weeks and targeting a vancomycin trough of
15 mg/L may be benefit patients with MRSA HVO. ª 2013 The British Infection Association. Published by Elsevier Ltd. All rights reserved.

Introduction Hematogenous vertebral osteomyelitis has increased in recent years, probably associated with longer life expectancy, increasing prevalence of chronic disease, better diagnosis, and more frequent use of indwelling intravascular catheters and immunosuppressive therapy.1 It is associated with significant morbidity including prolonged antimicrobial therapy, risk of recurrence, and decreased functional status.2,3 Staphylococcus aureus is the leading cause of hematogenous vertebral osteomyelitis, with the majority of cases caused by methicillin-susceptible organisms (MSSA).4e7 Because the proportion of methicillin-resistant S. aureus strains (MRSA) has increased during the past decades,8,9 the relative importance of MRSA as a cause of hematogenous vertebral osteomyelitis has simultaneously increased in the healthcare and community setting.10e12 In spite of this, there are limited data on the clinical features, treatment, and outcomes of hematogenous vertebral osteomyelitis caused by MRSA. A large study performed before the dramatic increase of MRSA found that all cases were caused by MSSA.7 More recent studies included a small number of MRSA cases.10e12 We performed an observational cohort study to evaluate the clinical characteristics and therapeutic outcomes of hematogenous MRSA vertebral osteomyelitis compared to MSSA-infected cases. We also evaluated the molecular and microbiological characteristics of the MRSA isolates causing the infection.

Patients and methods Study design and setting This observational cohort study was conducted at the Asan Medical Center, Seoul, Republic of Korea. This 2700-bed university-affiliated teaching hospital provides both primary and tertiary care, and has an average of approximately 124,000 annual patient discharges and 2,000,000 outpatient visits. This study included all adult patients diagnosed with hematogenous S. aureus vertebral osteomyelitis from January 2005 to December 2011. Data from patients with bacteraemic S. aureus vertebral osteomyelitis were collected in a prospective cohort study of S. aureus bloodstream infections from August 2008 to December 2011.13e15 To increase the statistical power of comparison, additional data were included from the Asan Medical Center database. We identified all discharges with International Classification of Diseases, Tenth Revision, Clinical Modification (ICD-10-CM) codes for osteomyelitis of the vertebra (M46.2), pyogenic infection of intervertebral disc (M46.3), unspecified discitis (M46.4), other infective spondylopathy (M46.5), other specified inflammatory spondylopathy (M46.8), unspecified inflammatory spondylopathy

(M46.9), unspecified spondylopathy (M48.9), and epidural abscess (G06.1 and G06.2). Discharges containing these ICD-10-CM codes and cases of bacteraemic S. aureus vertebral osteomyelitis derived from our prospective cohort were reviewed to determine whether they met the study criteria defined below. The Asan Medical Center Institutional Review Board approved the study and waiver of informed consent (IRB number: 2012-0542). According to the hospital policy, patients with hematogenous vertebral osteomyelitis were admitted to the department of infectious diseases. All patients initially received intravenous antibiotics, and the decision to switch to oral route was left to the discretion of the attending physicians. The choice of oral antibiotics was based on the susceptibility of the isolated organisms to the available agents. The discontinuation of antibiotic therapy was decided by attending physician taking into account clinical resolution and a gradual decrease of inflammation indexes such as erythrocyte sedimentation rate (ESR) and C-reactive protein (CRP).

Inclusion and exclusion criteria Adult patients (
16 years of age) who presented with hematogenous S. aureus vertebral osteomyelitis or those in whom vertebral osteomyelitis developed during treatment of S. aureus bacteremia were included. Hematogenous S. aureus vertebral osteomyelitis was defined as radiographic evidence of vertebral osteomyelitis with microbiologic demonstration of S. aureus either from the site of infection itself (e.g., abscess, intervertebral disc, or vertebral bone) or the blood. Cases were excluded if there was a non-hematogenous source of vertebral infection, which included (1) penetrating trauma, (2) previously-placed hardware, (3) laminectomy within 1 year prior to the diagnosis of vertebral osteomyelitis, or (4) the presence of a stage 3e4 decubitus ulcer at the time of diagnosis. Other reasons for exclusion were incomplete medical records or polymicrobial infection. Cases were required to have received at least 4 weeks of antibiotics and to have follow-up for at least 12 months after completing all antibiotic treatments.

Data collection Electronic medical records were reviewed for demographic information, underlying illness/conditions, presumed source of infection, concurrent metastatic infections, diagnostic work-up, clinical presentation, laboratory and radiologic data, medical and surgical treatments, and clinical outcomes.

Definitions Hospital-onset vertebral osteomyelitis cases were defined in patients demonstrating no signs or symptoms of infection on admission.7 Outcomes were evaluated

558 by in-hospital mortality, persistent bacteraemia, length of hospital stay, relapse, and sequelae. Duration of bacteraemia was defined as the number of days between the first and the last positive blood cultures.16,17 Persistent bacteraemia was defined as S. aureus bacteraemia for
7 days. Relapse was defined as the development of recurrent symptoms and positive culture after apparent clinical resolution and completion of antibiotic treatment. Sequelae were defined as persisting signs and symptoms related to osteomyelitis lasting 12 months or longer after completing treatment, including paresis necessitating wheelchair-use, bladder/bowel incontinence, and/or unresolved pain that required analgesic therapy. Vancomycin-induced nephrotoxicity was assessed, with nephrotoxicity was defined as a minimum of two or three consecutive documented increases in serum creatinine (defined as an increase of 0.5 mg/dL or
50% increase from, whichever is greater) in the absence of an alternative explanation.18

Laboratory and molecular tests The minimum inhibitory concentration (MIC) of vancomycin was determined using the Etest (AB Biodisk, Solna, Sweden) according to the manufacturer’s instructions. MRSA isolates were assessed by the population analysis profilingarea under the curve (PAP-AUC) method described by Wootton et al.19 An isolate was identified as heteroresistant vancomycin-intermediate S. aureus (hVISA) if the ratio of the AUC of the test isolate to the reference strain (Mu3; ATCC 700698) was
0.9. Multiplex PCR was performed to determine SCCmec type for MRSA isolates, as described previously.20 Multilocus sequence typing (MLST) was done as described previously.21,22 agr dysfunction was determined by examining d-hemolysin expression on blood agar plates using S. aureus RN4220, as described previously.23

Statistical analysis All statistical analyses were performed using SPSS for Windows, version 14.0 (SPSS, Inc., Chicago, IL, USA). Categorical variables were compared by the chi-square or Fisher’s exact test. Continuous variables were compared using Student’s t-test or the ManneWhitney U test. Univariate analysis of the risk factors for five outcomes studied was performed using regression models that examined the individual variable of interest. In-hospital mortality, persistent bacteremia (
7 days), relapse, and sequelae were analyzed with logistic regression. Length of hospital stay was log-transformed to achieve a normal distribution and analyzed with linear regression models. The coefficients were converted to the measures of effect using an exponential transformation; the measures of effect are referred to as the multiplicative effect (ME). Five separate multivariate analyses were performed for the five outcomes using the same regression methods described above. Variables with P values <0.1 on the univariate analysis were included in the corresponding multivariate analysis. All statistical tests were two tailed, and a P value of 0.05 was considered significant.

K.-H. Park et al.

Results A total of 172 patients with hematogenous S. aureus vertebral osteomyelitis were identified during the study period. Of the 172 cases, 33 were excluded because of a followup of <12 months (n Z 26), incomplete medical records (n Z 4), polymicrobial infection (n Z 2), or antibiotic therapy of <4 weeks (n Z 1). Finally, 139 patients met the eligible criteria and were included in the analysis.

Patient demographics and characteristics Demographic and baseline characteristics of the patients are summarized in Table 1. The median age was 65 years (interquartile range [IQR], 55e71 years), and 79 (56.8%) were male. Seventy-seven cases (55.4%) were caused by MSSA and 62 (44.6%) by MRSA. Patients infected with MRSA were more likely to be older and to have diabetes or a malignancy, and their infections were more frequently of hospital-onset. Based on clinical and microbiological evaluation, a potential portal of entry for hematogenous S. aureus vertebral osteomyelitis was identified in 61 patients (43.9%). Intravenous venous catheters were more likely to be the origin in MRSA than in MSSA cases (14/30 [46.7%] vs. 7/31 [22.6], P Z 0.048). Frequency of other concurrent metastatic foci was similar between both groups.

Diagnosis and clinical presentations Bacteremia was documented in 115 patients (82.7%). Microbiological confirmation in the remaining 24 patients was made by needle biopsy in 16 (11.5%) and by open biopsy in 8 (5.8%). MRIs, CT scans, and bone scans were obtained for diagnosis in 131 (94.2%), 5 (3.6%), and 3 (2.2%) patients, respectively. Fourteen infections (10.1%) were cervical, 33 (23.7%) were thoracic, and 107 (77.0%) were lumbosacral. Fifty-six patients (40.3%) had >2 infected vertebral bodies. Seventy-two patients (51.8%) had epidural involvement, and 101 patients (72.7%) had paravertebral and/or psoas involvement. No differences in clinical, laboratory, and radiologic presentation between patients infected with MSSA and MRSA were found (Table 2).

Molecular and microbiological characteristics of MRSA isolates causing hematogenous vertebral osteomyelitis Of the 54 cases of bacteraemic MRSA vertebral osteomyelitis, 40 blood isolates (74.1%) were available for molecular and microbiological testing. Three MRSA strains were responsible for 90% of the MRSA cases: ST72-SCCmecIV (n Z 18), ST5-SCCmecII (n Z 14), and ST239-SCCmecIII (n Z 4). ST72-SCCmecIV was responsible for 70.8% of the community-onset cases and 12.5% of the hospital-onset cases. Twenty isolates (50%) were agr dysfunctional, which was higher in ST5-SCCmecII (92.9% [13/14]) than in ST72SCCmecIV (22.2% [4/18]) (P < 0.001). Thirty-four isolates (85%) had vancomycin MICs of
1.5 mg/L and 11 isolates (27.5%) had hVISA phenotype. Resistance rates against several other antibiotics were as followed: clindamycin

MRSA vertebral osteomyelitis Table 1 MRSA.

559

Demographic and baseline characteristics of patients with hematogenous vertebral osteomyelitis caused by MSSA and

Variable

All study patients (n Z 139)

MSSA (n Z 77)

MRSA (n Z 62)

Age, median years (IQR) Male gender Hospital-onset infection Transfer from outside hospital Underlying illness/conditions Diabetes mellitus Malignancy Liver cirrhosis Rheumatic disease End stage renal disease Immunosuppression Remote history of back surgery Presumed source of the infection Infected vascular access Skin and subcutaneous tissues Endocarditis Postoperative wounds Othera Unknown Other concurrent metastatic infectionb

65 79 32 36

(55e71) (56.8) (23.0) (25.9)

63 42 10 22

(53e69) (54.5) (13.0) (28.6)

67 37 22 14

(61e74) (59.7) (35.5) (22.6)

44 16 12 10 8 15 12

(31.7) (11.5) (8.6) (7.2) (5.8) (10.8) (8.6)

17 5 7 6 4 7 6

(22.1) (6.5) (9.1) (7.8) (5.2) (9.1) (7.8)

27 11 5 4 4 8 6

(43.5) (17.7) (8.1) (6.5) (6.5) (12.9) (9.7)

0.007 0.04 0.83 >0.99 >0.99 0.47 0.69

21 18 8 8 6 78 14

(15.1) (12.9) (5.8) (5.8) (4.3) (56.1) (10.1)

7 13 3 4 4 46 10

(9.1) (16.9) (3.9) (5.2) (5.2) (59.7) (13.0)

14 5 5 4 2 32 4

(22.6) (8.1) (8.1) (6.5) (3.2) (51.6) (6.5)

0.03 0.12 0.47 >0.99 0.69 0.34 0.20

P 0.04 0.54 0.002 0.42

Data are no. (%) of patients, unless otherwise indicated. IQR, interquartile range; MRSA, methicillin-resistant S. aureus; MSSA, methicillin-susceptible S. aureus. a Included urinary tract (n Z 3) and lungs (n Z 3). b Included septic arthritis (n Z 5), soft tissue abscess (n Z 5), septic pulmonary emboli (n Z 3), and cerebral emboli (n Z 1).

(55.0%), ciprofloxacin (52.5%), gentamicin (47.5%), fusidic acid (25.0%), trimethoprime-sulfamethoxazole (12.5%), and rifampin (7.5%).

in those without epidural involvement (33/72 [45.8%] vs. 7/66 [10.4%]; P < 0.001) and in patients with neurologic deficit than in those without neurologic deficit (18/35 [51.4%] vs. 22/104 [21.2%]; P Z 0.001).

Treatment Therapeutic outcomes All patients received intravenous antibiotics during their initial treatment. The most frequently prescribed intravenous antibiotics were nafcillin (43/77 [55.8%]) and cefazolin (31/77 [40.3%]) in the MSSA group, and vancomycin (60/62 [96.8%]) in the MRSA group. The median duration of intravenous antibiotic therapy was 50 days (IQR, 34e64) and was significantly longer in the MRSA than in the MSSA group (55 vs. 46 days; P Z 0.005). Oral antibiotics were prescribed more frequently after completion of intravenous therapy in the MSSA than the MRSA group (37/77 [48.1%] vs. 18/62 [29.0%]; P Z 0.02). The most frequently prescribed oral antibiotics were cephradine (13/37 [37.1%]) and a fluoroquinolone plus rifampin (10/37 [28.6%]) in the MSSA group and a fluoroquinolone plus rifampin in the MRSA group (6/18 [33.3%]). The difference in median duration of total antibiotic therapy between the MSSA and MRSA groups was not significant (58 vs. 72 days; P Z 0.16). The antibiotic treatments are summarized in Supplementary Table 1. In addition to antibiotic therapy, surgical debridement was performed in 23 (29.9%) MSSA-infected patients and 17 (27.4%) MRSA-infected patients (P Z 0.75). CT-guided aspiration drainage procedure was performed in 22 (28.6%) and 16 (25.8%) patients, respectively (P Z 0.72). Surgery was more frequent in patients with epidural involvement than

The outcomes of MSSA and MRSA infections are shown in Table 2. In-hospital mortality was similar in patients with MSSA and MRSA infection (10.4% vs. 8.1%; P Z 0.64). Significant univariate risk factors for in-hospital mortality were liver cirrhosis and malignancy (Table 3). Multivariate analysis indicated that liver cirrhosis was the only independent risk factor for in-hospital mortality (OR, 17.14; 95% CI, 4.38e67.08; P < 0.001). Duration of bacteremia was longer in MRSA-infected patients than MSSA-infected patients (mean 10.1 vs. 3.1 days; P < 0.001). Persistent bacteremia (
7 days) was more common in MRSA than in MSSA cases (59.3% vs. 14.8%; P < 0.001). Significant univariate for persistent bacteremia were MRSA and diabetes mellitus (Table 3). Multivariate analysis indicated that MRSA was the only independent risk factor for persistent bacteremia (OR, 8.40; 95% CI, 3.45e20.50; P < 0.001). Length of hospital stay was longer in MRSA-infected patients than in MSSA-infected patients (median 69 vs. 52 days; P Z 0.001) (Table 2). Significant univariate risk factors for pronged length of hospital were MRSA, diabetes mellitus, end stage renal disease, and surgical debridement (Table 3). Multivariate analysis indicated that MRSA (ME, 1.40; 95% CI, 1.16e1.69; P Z 0.001), end stage renal

560 Table 2

K.-H. Park et al. Clinical features and outcomes of hematogenous vertebral osteomyelitis caused by MSSA and MRSA.

Variable

All study patients (n Z 139)

MSSA (n Z 77)

MRSA (n Z 62)

Back pain Body temperature >38  C Neurologic deficit Time to diagnosis, median days (IQR) Laboratory data WBC count, K/mm3, median (IQR) C-reactive protein, mg/L, median (IQR) ESR, mm/h, median (IQR)a Positive blood cultures Radiologic data Involvement of >2 vertebral bodies Involvement of cervical spine Involvement of thoracic spine Involvement of lumbosacral spine Epidural involvementb Paravertebral and/or psoas involvementb Outcomes In-hospital mortality Duration of bacteremia, daysc Median (IQR) Mean  S.D. Persistent bacteremia (
7 days)c Length of hospital stay, median days (IQR) Relapsed Sequelaee

127 (91.4) 93 (66.9) 35 (25.2) 30 (9e56)

72 51 18 32

55 42 17 27

118 (80e178) 144 (73e238) 89 (61e113) 115/134 (85.8)

(93.5) (66.2) (23.4) (11e57)

P

(88.7) (67.7) (27.4) (7e57)

0.32 0.85 0.59 0.20

111 (80e173) 143 (65e250) 91 (61e119) 61/73 (83.6)

138 (82e188) 145 (71e233) 85 (61e99) 54/61 (88.5)

0.27 0.85 0.35 0.41

56 (40.3) 14 (10.1) 33 (23.7) 107 (77.0) 72 (51.8) 101 (72.7)

31 (40.3) 8 (10.4) 19 (24.7) 60 (77.9) 43 (55.8) 52 (67.5)

25 (40.3) 6 (9.7) 14 (22.6) 47 (75.8) 29 (46.8) 49 (79.0)

0.99 0.89 0.77 0.77 0.29 0.13

13/139 (9.4)

8/77 (10.4)

5/62 (8.1)

0.64

3 (0e10) 6.4  8.5 41 (35.7) 56 (42e83) 12/125 (9.6) 39/111 (35.1)

0 (0e6) 3.1  5.2 9 (14.8) 52 (36e68) 3/69 (4.3) 20/62 (32.3)

9 (0e16) 10.1  9.9 32 (59.3) 69 (45e108) 9/56 (16.1) 19/49 (38.8)

<0.001 <0.001 <0.001 0.001 0.03 0.48

Data are no. (%) of patients, unless otherwise indicated. ESR, erythrocyte sediment rate; IQR, interquartile range; MRSA, methicillinresistant S. aureus; MSSA, methicillin-susceptible S. aureus; WBC, white blood cell. a Performed in 110 patients (65 in MSSA group and 45 in MRSA group). b Either phlegmon or abscess. c Evaluated in 115 bacteraemic patients. d Evaluated in 125 patients who completed an initial course of antibiotic therapy. e Evaluated in 111 patients who survived for at least 12 months after completing of antibiotic therapy.

disease (ME, 1.60; 95% CI, 1.07e2.40; P Z 0.02), and surgical debridement (ME, 1.29; 95% CI, 1.04e1.58; P Z 0.02) were independent risk factors for prolonged length of stay. Of the 139 study patients, 14 died before completing antimicrobial therapy for their infection (8 in the MSSA group and 6 in the MRSA group). The remaining 125 patients completed at least 4 weeks of antibiotic therapy. Of those 125 patients, 12 (9.6%) experienced relapses within 12 weeks after completing antibiotic therapy; all patients with relapse presented with bacteremia. At the time of relapse, none had clinical evidence of endocarditis, and the echocardiogram was normal in the nine patients tested. More frequent relapse were observed among MRSA than MSSA cases (16.1% vs. 4.3%; P Z 0.03) (Table 2). In univariate analysis, MRSA and <8 weeks of antibiotics therapy were associated with relapse (Table 3). In multivariate analysis, MRSA (OR, 4.83; 95% CI, 1.20e19.44; P Z 0.03) and <8 weeks of antibiotics therapy (OR, 4.21; 95% CI, 1.15e15.44; P Z 0.03) remained significant risk factors for relapse. Among the MRSA cases, relapse rates were lower in patients with a longer duration of antibiotic therapy: 41.7% (4e6 weeks), 25.0% (6e8 weeks), and 5.6% (
8 weeks) (P Z 0.007). However, this association was not significant for patients with MSSA infections: 6.7% (<6 weeks),

0% (6e8 weeks), and 4.9% (
8 weeks) (P > 0.99) (Fig. 1A). Similar associations were observed for relapse rates and duration of intravenous antibiotic therapy (Fig. 1B). Of the 125 patients who completed antibiotic therapy, 14 died within 12 months. The 111 patients who survived for at least 12 months were evaluable for sequelae. In univariate analysis, age of 65 years or older and epidural involvement were significant risk factors for sequelae (Table 3). In multivariate analysis, age of 65 years or older (OR, 4.82; 95% CI, 1.93e12.03; P Z 0.001) and epidural involvement (OR, 3.36; 95% CI, 1.37e8.25; P Z 0.008) remained significant risk factors for sequelae.

Initial vancomycin trough level and clinical outcome Initial vancomycin trough levels were measured at the steady state in 42 (70.0%) of the 60 MRSA-infected patients who received vancomycin for
48 h. The median time to the measurements of initial vancomycin trough levels was 3 days (IQR, 2e4 days). Bacteremia was more likely to persist for
7 days in patients with initial trough levels of <15 mg/ L than in those with initial trough levels of
15 mg/L (23/29 [79.3%] vs. 2/10 [20.0%]; P Z 0.001). Median duration of

Univariate analysis of clinical and microbiological risk factors associated with outcomes of Staphylococcus aureus hematogenous vertebral osteomyelitis.

Variable

Age
65 years Male gender Hospital-onset infection MRSA Diabetes mellitus Malignancy Liver cirrhosis End stage renal disease Endocarditis Other concurrent metastatic foci Involvement of >2 vertebral bodies Epidural involvementc Paravertebral and/or psoas involvementc Surgical debridement Cases of undrained abscess Positive blood culturesd <8 weeks of intravenous antibiotic therapy <8 weeks of all antibiotic therapy
2 weeks of concurrent rifampin therapy agr dysfucntione Vancomycin MIC
1.5 mg/L by Eteste hVISAe

In-hospital mortality(n Z 139)

Persistent bacteremiaa (n Z 115)

Length of stay (n Z 139)

Relapse (n Z 125)

OR (95% CI)

P

OR (95% CI)

P

ME (95% CI)

P

OR (95% CI)

0.73 (0.23e2.30) 0.62 (0.20e1.96) 2.29 (0.69e7.58) 0.76 (0.24e2.44) 0.96 (0.28e3.29) 4.22 (1.13e15.80) 17.14 (4.38e67.08) 1.42 (0.16e12.50) 1.42 (0.16e12.50) 1.73 (0.34e8.73)

0.59 0.42 0.17 0.64 0.94 0.03 <0.001 0.75 0.75 0.51

1.83 1.07 1.06 8.40 2.33 1.24 0.16 5.00 3.29 0.78

0.13 0.87 0.89 <0.001 0.04 0.71 0.09 0.06 0.12 0.70

1.15 1.17 1.13 1.40 1.28 0.80 0.84 1.56 0.91 0.92

0.16 0.14 0.33 0.001 0.02 0.17 0.33 0.04 0.66 0.63

2.85 0.96 2.07 4.21 1.11 0.84 NCb 4.32 NCb 2.06

1.30 (0.41e4.10)

0.65

0.83 (0.39e1.80)

0.64

0.96 (0.78e1.18)

0.69

1.08 (0.32e3.61)

0.90 1.98 (0.90e4.38)

0.09

0.55 (0.17e1.78) 1.28 (0.33e4.94)

0.32 0.72

0.60 (0.28e1.29) 1.50 (0.62e3.67)

0.19 0.37

1.07 (0.88e1.31) 0.95 (0.76e1.19)

0.48 0.67

0.59 (0.18e1.96) 0.33 (0.10e1.10)

0.39 2.52 (1.11e5.72) 0.07 0.60 (0.26e1.39)

0.03 0.23

0.19 (0.02e1.48) 2.81 (0.88e8.92) NCb NA

0.11 0.08 0.21 NA

NA NA NA NA

NA NA NA NA

1.25 (1.00e1.55) 0.87 (0.70e1.08) 1.23 (0.92e1.66) NA

0.05 0.20 0.16 NA

0.18 0.45 0.94 3.83

0.11 0.31 0.93 0.09

(0.80e4.11) (0.53e3.08) (0.54e4.94) (0.40e1.92)

0.16 0.60 0.39 0.74

NA

NA

NA

NA

NA

NA

3.65 (1.04e12.88)

0.04 2.08 (0.93e4.68)

0.08

NCb

0.13

NA

NA

NA

NA

NCb

0.12 0.55 (0.20e1.51)

0.24

0.30 (0.03e3.15) NCb

0.31 >0.99

0.88 (0.24e3.19) 0.36 (0.04e3.55)

0.84 0.38

1.06 (0.71e1.59) 0.99 (0.57e1.74)

0.76 0.98

7.39 (0.79e69.36) 0.40 (0.06e2.81)

0.08 0.42 (0.09e1.92) 0.36 1.88 (0.17e20.61)

0.26 0.61

0.87 (0.08e9.34)

0.91

2.00 (0.44e9.18)

0.37

1.26 (0.81e1.97)

0.29

1.05 (0.17e6.55)

0.96 0.22 (0.02e2.10)

0.19

(0.84e4.00) (0.49e2.33) (0.45e2.52) (3.45e20.50) (1.05e5.19) (0.41e3.76) (0.02e1.30) (0.93e27.04) (0.74e14.53) (0.23e2.71)

(0.94e1.41) (0.95e1.43) (0.89e1.43) (1.15e1.70) (1.04e1.59) (0.59e1.10) (0.59e1.20) (1.02e2.39) (0.59e1.40) (0.66e1.29)

Sequelae (n Z 111) P

(0.73e11.06) (0.29e3.22) (0.57e7.49) (1.08e16.40) (0.32e3.95) (0.10e7.16)

0.13 0.95 0.27 0.04 0.87 0.88 >0.99 (0.74e25.19) 0.10 >0.99 (0.40e10.74) 0.39

(0.02e1.44) (0.09e2.14) (0.19e4.65) (0.80e18.27)

OR (95% CI)

P

3.84 0.74 0.48 1.33 1.14 NCb NCb 1.42 0.92 0.34

0.002 0.46 0.18 0.48 0.76 0.60 >0.99 0.66 0.92 0.17

1.81 1.27 1.63 0.87

(1.63e9.04) (0.34e1.63) (0.16e1.41) (0.61e2.91) (0.49e2.62)

(0.30e6.68) (0.16e5.26) (0.07e1.61)

561

hVISA, heteroresistant vancomycin-intermediate S. aureus; IQR, interquartile range; ME, multiplicative effect; MRSA, methicillin-susceptible S. aureus; NA, not applicable; NC, not calculated. a Defined as S. aureus bacteremia for
7 days. b Odds ratio could not be calculated because of zero cell. c Either phlegmon or abscess. d Determined in 134 patients for in-hospital mortality, in 134 for length of stay, in 120 for relapse, and in 106 for sequelae. e Determined in 40 MRSA isolates for in-hospital mortality, in 39 for persistent bacteremia, in 40 for length of stay, in 36 for relapse, and in 30 for sequelae.

MRSA vertebral osteomyelitis

Table 3

562

K.-H. Park et al.

A)

B) 4-6 weeks 6-8 weeks ≥8 weeks

<6 weeks 6-8 weeks ≥8 weeks

Figure 1 Relapse rates of hematogenous vertebral osteomyelitis caused by MSSA and MRSA related to the duration of all antibiotic therapy (A) and intravenous antibiotic therapy (B).

hospital stay was longer in patients with initial trough levels of <15 mg/L than in those with initial trough levels of
15 mg/L (84 vs.53; P Z 0.053). There were no significant differences between the initial trough level and other clinical outcomes including mortality, relapse, and sequelae (Table 4).

Discussion In this study, the rate of MRSA among hematogenous S. aureus vertebral osteomyelitis was much higher than the previously reported 0e18.9% rate,3e6 but was comparable to 40.0e57.1% rate reported in more recent studies.11,24,25 Increasing rates of MRSA infection among hematogenous S. aureus vertebral osteomyelitis may be explained by the overall increase of MRSA in hospital and community settings. Interestingly, ST72-MRSA-SCCmecIV was responsible for more than two-thirds of community-onset MRSA vertebral osteomyelitis. ST72-MRSA-SCCmecIV has increased in community and hospital settings in our country in recent years.26,27 Our data suggest that ST72-MRSA-SCCmecIV may have a possible role in the increasing rates of hematogenous MRSA vertebral osteomyelitis.

Table 4

We found that MRSA was independently associated with persistent bacteremia, increased length of hospital stay, and relapse. This may be explained by limited bactericidal activity and bone penetration of vancomycin.28e30 Vancomycin therapy has been associated with higher recurrence rates following treatment of MSSA osteomyelitis than b-lactam antibiotics.31 Some vancomycin failures may result from MRSA with reduced vancomycin susceptibility32 and unrecognized heterogeneously resistant S. aureus (hVISA), which is not readily detected by standard clinical laboratory methods.33 However, our data showed that increased vancomycin MIC and a hVISA phenotype were not associated with the clinical outcomes of MRSA osteomyelitis. In addition, our study showed an increase in length of hospital stay due to MRSA (increased 1.69 fold) similar to that from a study that examined S. aureus bacteremia.34 That study noted a 1.29-fold increase in length of hospital stay among patients with S. aureus bacteremia due to MRSA compared with MSSA. There are limited data on the optimal duration of hematogenous vertebral osteomyelitis caused by MRSA. A previous small study of 29 patients with hematogenous S. aureus vertebral osteomyelitis (23 MSSA and 6 MRSA), found higher relapse rates in patients who received <8 weeks of

Outcomes of hematogenous MRSA osteomyelitis according to initial vancomycin trough levels.

Variable

Persistent bacteremia (
7 days) Vancomycin-induced nephrotoxicitya In-hospital mortality Length of hospital stay, median days (IQR) Relapseb Sequelaec

Initial vancomycin trough

P

<15 mg/L


15 mg/L

23/29 (79.3)

2/10 (20.0)

4/27 (14.8)

2/11 (18.2)

>0.99

3/30 (10.0) 84 (62e125)

1/12 (8.3) 53 (42e99)

>0.99 0.053

2/26 (7.7) 5/22 (22.7)

3/11 (27.3) 5/9 (55.6)

0.14 0.11

Data are no. (%) of patients, unless otherwise indicated. IQR, interquartile range; MRSA, methicillin-resistant S. aureus. a Denominator reflect exclusion of patient with end-stage renal disease from analysis of nephrotoxicity. b Evaluated in 37 patients who completed an initial course of antibiotic therapy. c Evaluated in 31 patients who survived for at least 12 months after completing of antibiotic therapy.

0.001

MRSA vertebral osteomyelitis intravenous antibiotic therapy.10 We found that a prolonged course of all antibiotic therapy (>8 weeks) was independently associated with lower relapse risk. This association was evident in MRSA-infected patients, but not in MSSAinfected patients. The relapse rate was much higher in the MRSA group than in the MSSA group despite a trend toward to the longer total duration of antibiotic therapy in MRSA group than MSSA group. Therefore, our data suggest that a patient with MRSA vertebral osteomyelitis should receive more prolonged antibiotic therapy (for at least 8 weeks) by either intravenous or oral route than those with MSSA vertebral osteomyelitis. Oral bactericidal drugs with excellent bioavailability, such as a fluoroquinolone, rifampin, and trimethoprim-sulfamethoxazole may allow for the possibility of an early switch to the oral route.35e38 However, because three fourth of MRSA cases was treated with only intravenous antibiotic therapy, we could not evaluate whether an early switch to the oral route was effective to treat these infection. Further studies should be performed to evaluate this issue. To improve tissue penetration and minimize the selection of resistant strains, a vancomycin trough level of 15e20 mg/L is recommended for serious MRSA infections.18,39 Although previous studies failed to find an association between a trough of
15 mg/L and improved outcomes,40e43 a recent study involving 320 MRSA bacteremia patients reported a lower rate of vancomycin failure in patients with an initial vancomycin trough of 15e20 mg/L.44 Clinical data supporting higher trough targets are limited in patients with osteomyelitis. In our patients, an initial vancomycin trough level of <15 mg/L was associated with an increased frequency of persistent bacteremia and longer hospital stay. Therefore, our data suggest that targeting a vancomycin trough of 15e20 mg/L may benefit patients with MRSA osteomyelitis. This study has limitations. First, initial vancomycin trough concentrations were not measured in one-third of those who received it with their initial treatment. However, we believe that this did not affect the results because the baseline clinical characteristics and the therapeutic approach were similar for all patients whether or not an initial vancomycin trough was measured. Second, the microbiological and molecular characteristics of isolates causing hematogenous MRSA vertebral osteomyelitis could not be evaluated in one-third of the cases. In the remaining two-thirds of MRSA isolates, however, we clearly documented that a substantial proportion of the infections was caused by ST72-MRSA-SCCmecIV. In summary, we found that hematogenous vertebral osteomyelitis caused by MRSA was associated with a longer duration of bacteremia, longer hospital stay, and more frequent relapse. Our results also indicate that antibiotic therapy for at least 8 weeks and targeting a vancomycin trough of >15 mg/L may be beneficial for MRSA vertebral osteomyelitis.

Funding This study was supported by a grant of the Korean Health Technology R&D Project, Ministry of Health & Welfare, Republic of Korea (grant number: A120843).

563

Conflicts of interest There are no potential conflicts of interest for any authors.

Appendix A. Supplementary data Supplementary data related to this article can be found at http://dx.doi.org/10.1016/j.jinf.2013.07.026.

References 1. Gasbarrini AL, Bertoldi E, Mazzetti M, Fini L, Terzi S, Gonella F, et al. Clinical features, diagnostic and therapeutic approaches to haematogenous vertebral osteomyelitis. Eur Rev Med Pharmacol Sci 2005;9:53e66. 2. McHenry MC, Easley KA, Locker GA. Vertebral osteomyelitis: long-term outcome for 253 patients from 7 Cleveland-area hospitals. Clin Infect Dis 2002;34:1342e50. 3. Carragee EJ. Pyogenic vertebral osteomyelitis. J Bone Joint Surg Am 1997;79:874e80. 4. Nolla JM, Ariza J, Gomez-Vaquero C, Fiter J, Bermejo J, Valverde J, et al. Spontaneous pyogenic vertebral osteomyelitis in nondrug users. Semin Arthritis Rheum 2002;31:271e8. 5. Pigrau C, Almirante B, Flores X, Falco V, Rodriguez D, Gasser I, et al. Spontaneous pyogenic vertebral osteomyelitis and endocarditis: incidence, risk factors, and outcome. Am J Med 2005; 118:1287. 6. Hadjipavlou AG, Mader JT, Necessary JT, Muffoletto AJ. Hematogenous pyogenic spinal infections and their surgical management. Spine (Phila Pa 1976) 2000;25:1668e79. 7. Jensen AG, Espersen F, Skinhoj P, Frimodt-Moller N. Bacteremic Staphylococcus aureus spondylitis. Arch Intern Med 1998;158:509e17. 8. Fridkin SK, Hageman JC, Morrison M, Sanza LT, Como-Sabetti K, Jernigan JA, et al. Methicillin-resistant Staphylococcus aureus disease in three communities. N Engl J Med 2005;352: 1436e44. 9. Wyllie DH, Crook DW, Peto TE. Mortality after Staphylococcus aureus bacteraemia in two hospitals in Oxfordshire, 1997e2003: cohort study. BMJ 2006;333:281. 10. Priest DH, Peacock Jr JE. Hematogenous vertebral osteomyelitis due to Staphylococcus aureus in the adult: clinical features and therapeutic outcomes. South Med J 2005;98:854e62. 11. Livorsi DJ, Daver NG, Atmar RL, Shelburne SA, White Jr AC, Musher DM. Outcomes of treatment for hematogenous Staphylococcus aureus vertebral osteomyelitis in the MRSA ERA. J Infect 2008;57:128e31. 12. Al-Nammari SS, Lucas JD, Lam KS. Hematogenous methicillinresistant Staphylococcus aureus spondylodiscitis. Spine (Phila Pa 1976) 2007;32:2480e6. 13. Park HJ, Lee YM, Bang KM, Park SY, Moon SM, Park KH, et al. Clinical significance of Staphylococcus aureus bacteremia in patients with liver cirrhosis. Eur J Clin Microbiol Infect Dis 2012;31:3309e16. 14. Park SY, Park KH, Bang KM, Chong YP, Kim SH, Lee SO, et al. Clinical significance and outcome of polymicrobial Staphylococcus aureus bacteremia. J Infect 2012;65:119e27. 15. Park KH, Kim ES, Kim HS, Park SJ, Bang KM, Park HJ, et al. Comparison of the clinical features, bacterial genotypes and outcomes of patients with bacteraemia due to heteroresistant vancomycinintermediate Staphylococcus aureus and vancomycinsusceptible S. aureus. J Antimicrob Chemother 2012;67:1843e9.

564 16. Musta AC, Riederer K, Shemes S, Chase P, Jose J, Johnson LB, et al. Vancomycin MIC plus heteroresistance and outcome of methicillin-resistant Staphylococcus aureus bacteremia: trends over 11 years. J Clin Microbiol 2009;47:1640e4. 17. Khatib R, Johnson LB, Sharma M, Fakih MG, Ganga R, Riederer K. Persistent Staphylococcus aureus bacteremia: incidence and outcome trends over time. Scand J Infect Dis 2009; 41:4e9. 18. Rybak M, Lomaestro B, Rotschafer JC, Moellering Jr R, Craig W, Billeter M, et al. Therapeutic monitoring of vancomycin in adult patients: a consensus review of the American Society of Health-System Pharmacists, the Infectious Diseases Society of America, and the Society of Infectious Diseases Pharmacists. Am J Health Syst Pharm 2009;66:82e98. 19. Wootton M, Howe RA, Hillman R, Walsh TR, Bennett PM, MacGowan AP. A modified population analysis profile (PAP) method to detect hetero-resistance to vancomycin in Staphylococcus aureus in a UK hospital. J Antimicrob Chemother 2001;47:399e403. 20. Oliveira DC, de Lencastre H. Multiplex PCR strategy for rapid identification of structural types and variants of the mec element in methicillin-resistant Staphylococcus aureus. Antimicrob Agents Chemother 2002;46:2155e61. 21. Enright MC, Day NP, Davies CE, Peacock SJ, Spratt BG. Multilocus sequence typing for characterization of methicillinresistant and methicillin-susceptible clones of Staphylococcus aureus. J Clin Microbiol 2000;38:1008e15. 22. Fowler Jr VG, Nelson CL, McIntyre LM, Kreiswirth BN, Monk A, Archer GL, et al. Potential associations between hematogenous complications and bacterial genotype in Staphylococcus aureus infection. J Infect Dis 2007;196:738e47. 23. Sakoulas G, Eliopoulos GM, Moellering Jr RC, Wennersten C, Venkataraman L, Novick RP, et al. Accessory gene regulator (agr) locus in geographically diverse Staphylococcus aureus isolates with reduced susceptibility to vancomycin. Antimicrob Agents Chemother 2002;46:1492e502. 24. Bhavan KP, Marschall J, Olsen MA, Fraser VJ, Wright NM, Warren DK. The epidemiology of hematogenous vertebral osteomyelitis: a cohort study in a tertiary care hospital. BMC Infect Dis 2010;10:158. 25. Marschall J, Bhavan KP, Olsen MA, Fraser VJ, Wright NM, Warren DK. The impact of prebiopsy antibiotics on pathogen recovery in hematogenous vertebral osteomyelitis. Clin Infect Dis 2011;52:867e72. 26. Joo EJ, Chung DR, Ha YE, Park SY, Kim HA, Lim MH, et al. Clinical predictors of community-genotype ST72-methicillin-resistant Staphylococcus aureus-SCCmec type IV in patients with community-onset S. aureus infection. J Antimicrob Chemother 2012;67:1755e9. 27. Park SH, Park C, Yoo JH, Choi SM, Choi JH, Shin HH, et al. Emergence of community-associated methicillin-resistant Staphylococcus aureus strains as a cause of healthcare-associated bloodstream infections in Korea. Infect Control Hosp Epidemiol 2009;30:146e55. 28. Landersdorfer CB, Bulitta JB, Kinzig M, Holzgrabe U, Sorgel F. Penetration of antibacterials into bone: pharmacokinetic, pharmacodynamic and bioanalytical considerations. Clin Pharmacokinet 2009;48:89e124. 29. Graziani AL, Lawson LA, Gibson GA, Steinberg MA, MacGregor RR. Vancomycin concentrations in infected and noninfected human bone. Antimicrob Agents Chemother 1988;32:1320e2. 30. Webb LX, Holman J, de Araujo B, Zaccaro DJ, Gordon ES. Antibiotic resistance in staphylococci adherent to cortical bone. J Orthop Trauma 1994;8:28e33.

K.-H. Park et al. 31. Tice AD, Hoaglund PA, Shoultz DA. Risk factors and treatment outcomes in osteomyelitis. J Antimicrob Chemother 2003;51: 1261e8. 32. Howden BP, Davies JK, Johnson PD, Stinear TP, Grayson ML. Reduced vancomycin susceptibility in Staphylococcus aureus, including vancomycin-intermediate and heterogeneous vancomycin-intermediate strains: resistance mechanisms, laboratory detection, and clinical implications. Clin Microbiol Rev 2010;23:99e139. 33. van Hal SJ, Paterson DL. Systematic review and meta-analysis of the significance of heterogeneous vancomycinintermediate Staphylococcus aureus isolates. Antimicrob Agents Chemother 2011;55:405e10. 34. Cosgrove SE, Qi Y, Kaye KS, Harbarth S, Karchmer AW, Carmeli Y. The impact of methicillin resistance in Staphylococcus aureus bacteremia on patient outcomes: mortality, length of stay, and hospital charges. Infect Control Hosp Epidemiol 2005;26:166e74. 35. Euba G, Murillo O, Fernandez-Sabe N, Mascaro J, Cabo J, Perez A, et al. Long-term follow-up trial of oral rifampincotrimoxazole combination versus intravenous cloxacillin in treatment of chronic staphylococcal osteomyelitis. Antimicrob Agents Chemother 2009;53:2672e6. 36. Lora-Tamayo J, Euba G, Narvaez JA, Murillo O, Verdaguer R, Sobrino B, et al. Changing trends in the epidemiology of pyogenic vertebral osteomyelitis: the impact of cases with no microbiologic diagnosis. Semin Arthritis Rheum 2011;41: 247e55. 37. Viale P, Furlanut M, Scudeller L, Pavan F, Negri C, Crapis M, et al. Treatment of pyogenic (non-tuberculous) spondylodiscitis with tailored high-dose levofloxacin plus rifampicin. Int J Antimicrob Agents 2009;33:379e82. 38. Schrenzel J, Harbarth S, Schockmel G, Genne D, Bregenzer T, Flueckiger U, et al. A randomized clinical trial to compare fleroxacin-rifampicin with flucloxacillin or vancomycin for the treatment of staphylococcal infection. Clin Infect Dis 2004; 39:1285e92. 39. Liu C, Bayer A, Cosgrove SE, Daum RS, Fridkin SK, Gorwitz RJ, et al. Clinical practice guidelines by the Infectious Diseases Society of America for the treatment of methicillin-resistant Staphylococcus aureus infections in adults and children. Clin Infect Dis 2011;52:e18e55. 40. Jeffres MN, Isakow W, Doherty JA, McKinnon PS, Ritchie DJ, Micek ST, et al. Predictors of mortality for methicillinresistant Staphylococcus aureus health-care-associated pneumonia: specific evaluation of vancomycin pharmacokinetic indices. Chest 2006;130:947e55. 41. Lodise TP, Graves J, Evans A, Graffunder E, Helmecke M, Lomaestro BM, et al. Relationship between vancomycin MIC and failure among patients with methicillin-resistant Staphylococcus aureus bacteremia treated with vancomycin. Antimicrob Agents Chemother 2008;52:3315e20. 42. Hidayat LK, Hsu DI, Quist R, Shriner KA, Wong-Beringer A. Highdose vancomycin therapy for methicillin-resistant Staphylococcus aureus infections: efficacy and toxicity. Arch Intern Med 2006;166:2138e44. 43. Maor Y, Hagin M, Belausov N, Keller N, Ben-David D, Rahav G. Clinical features of heteroresistant vancomycin-intermediate Staphylococcus aureus bacteremia versus those of methicillin-resistant S. aureus bacteremia. J Infect Dis 2009; 199:619e24. 44. Kullar R, Davis SL, Levine DP, Rybak MJ. Impact of vancomycin exposure on outcomes in patients with methicillin-resistant Staphylococcus aureus bacteremia: support for consensus guidelines suggested targets. Clin Infect Dis 2011;52:975e81.