Abstract
We conducted a retrospective cohort study to evaluate factors influencing tissue culture positivity in patients with pyogenic vertebral osteomyelitis exposed to antibiotics before diagnosis. Tissue culture was positive in 48.3% (28/58) of the patients, and the median antibiotic-free period was 1.5 days (range, 0.7 to 5.7 days). In a multivariate analysis, a higher C-reactive protein (CRP) level (adjusted odds ratio [aOR], 1.18; 95% confidence interval, 1.07 to 1.29) and open surgical biopsy (aOR, 6.33; 95% confidence interval, 1.12 to 35.86) were associated with tissue culture positivity.
TEXT
If pyogenic vertebral osteomyelitis (PVO) is suspected, tissue cultures and/or blood cultures are recommended (1). Prior antibiotic exposure has been reported in some studies to have a negative effect on microbiologic diagnosis (2–5). In view of this, it is recommended that, for clinically stable patients, biopsy be delayed until at least 48 h after the last antibiotic has been administered (6). The aims of this study were to investigate the factors influencing tissue culture positivity in patients with PVO who were exposed to antibiotics prior to biopsy and to establish whether an antibiotic-free period improves culture positivity rates.
We conducted a retrospective cohort study at three university-affiliated teaching hospitals from May 2012 through February 2014. Patients with PVO who had been exposed to antibiotics during the 2 weeks before the acquisition of tissue culture specimens were investigated. Patients who were <18 years old and patients with infectious spondylitis caused by Mycobacterium tuberculosis or fungi were excluded. PVO was diagnosed when the causative microorganism was isolated from spinal or paraspinal tissues or if there were compatible clinical signs or symptoms and radiologic evidence of vertebral infection as described previously (7). The antibiotic-free period was defined as the time interval between the administration of the latest antibiotic and acquisition of specimens for tissue culture. If the antibiotic administered was not active on the microorganism eventually isolated, the patient was considered not to have been exposed to antibiotics. Tissue culture specimens were acquired by percutaneous biopsy or open surgical biopsy. The latter was performed during surgical treatment of the PVO. The Mann-Whitney U test was used to compare continuous variables, and the chi-square test was used to compare categorical variables. Multivariate analysis was performed with binary logistic regression to investigate the factors influencing tissue culture positivity. All P values were two tailed, and P < 0.05 was considered statistically significant.
During the study period, a total of 58 patients with PVO received antibiotics before the acquisition of tissue culture specimens. Blood cultures were done for 53 patients, and 30.2% (16/53) gave positive results. The tissue culture positivity rate was 48.3% (28/58). Tissue culture specimens of 48 patients were obtained by computed-tomography or fluoroscopy-guided percutaneous needle biopsy, whereas those of the other 10 patients were obtained during surgical debridement. The median number of culture specimens was 2 in the percutaneous biopsy group and 3.5 in the surgical debridement group (P = 0.66). Age, sex, the proportion of patients with abscesses, white blood cell (WBC) counts, and the duration of prior antibiotic exposure did not differ between patients with tissue culture-positive and -negative PVO, but CRP levels were significantly higher in the tissue culture-positive patients (17.2 versus 6.5 mg/dl, respectively) (Table 1). The median antibiotic-free period was significantly shorter in the patients with tissue culture-positive PVO (1.1 versus 3.3 days, respectively). The most common causative organism was Staphylococcus aureus (14 patients), followed by Streptococcus (5 patients), Escherichia coli (3 patients), etc. The most common antibiotics given prior to biopsy were third-generation cephalosporins (22 patients), followed by first-generation cephalosporins (18 patients) and fluoroquinolones (11 patients), in both groups. The tissue culture positivity rates were 62.5% (20/32), 30.0% (3/10), and 31.3% (5/16) in patients whose antibiotic-free periods were <48 h, 48 to 120 h, and ≥120 h, respectively (P = 0.029) (Table 2). In a multivariate logistic regression analysis, open surgical biopsy (adjusted odds ratio [aOR], 6.33; 95% confidence interval [CI], 1.12 to 35.86) and a higher CRP level (aOR, 1.18, 95% CI, 1.07 to 1.29) were significantly associated with tissue culture positivity (Table 3).
TABLE 1.
Demographic and clinical characteristics of patients with PVO who received antibiotics before the acquisition of tissue culture specimens
| Patient parameter | Total (n = 58) | Culture positive (n = 28) | Culture negative (n = 30) | P value |
|---|---|---|---|---|
| Age (yr)a | 66.1 ± 14.3 | 64.6 ± 13.7 | 67.4 ± 14.8 | 0.375 |
| No. (%) of males | 24 (41.4) | 15 (53.6) | 9 (30.0) | 0.069 |
| No. (%) with underlying disease/treatment | ||||
| Hypertension | 16 (27.6) | 8 (28.6) | 8 (26.7) | 0.871 |
| Diabetes mellitus | 13 (22.4) | 7 (25.0) | 6 (20.0) | 0.648 |
| Liver cirrhosis | 2 (3.4) | 1 (3.6) | 1 (3.3) | 1.000b |
| Spinal procedure in previous yr | 14 (24.1) | 6 (21.4) | 8 (26.7) | 0.641 |
| Spinal surgery in previous yr | 7 (12.3) | 4 (14.3) | 3 (10.3) | 0.706b |
| No. (%) with: | ||||
| Feverc | 20 (34.5) | 10 (35.7) | 10 (33.3) | 0.849 |
| Spine area pain | 56 (96.9) | 26 (92.9) | 30 (100) | 0.229b |
| Motor weaknessd | 10 (17.2) | 5 (17.9) | 5 (16.7) | 1.000b |
| Sepsise | 19 (32.8) | 12 (42.9) | 7 (23.3) | 0.113 |
| No. of WBC/mm3f | 10,050 (8,215–12,405) | 10,480 (8,742–14,350) | 9,450 (7,950–11,350) | 0.141 |
| CRP level (mg/dl)f | 11.0 (5.0–18.8) | 17.2 (11.0–24.4) | 6.5 (1.7–11.3) | <0.001 |
| No. (%) with paravertebral abscesses | 13 (22.4) | 7 (25.0) | 6 (20.0) | 0.648 |
| No. (%) with epidural abscesses | 28 (48.3) | 13 (46.4) | 15 (50.0) | 0.786 |
| No. (%) with psoas abscesses | 19 (32.8) | 12 (42.9) | 7 (23.3) | 0.113 |
| No. (%) with following vertebral area involved: | ||||
| C-spine | 1 (1.7) | 1 (3.6) | 0 | 0.579 |
| T-spine | 7 (12.1) | 5 (17.9) | 2 (6.7) | |
| TL-spine | 4 (6.9) | 2 (7.1) | 2 (6.7) | |
| L-spine | 32 (55.2) | 14 (50.0) | 18 (60.0) | |
| LS-spine | 13 (22.4) | 6 (21.4) | 7 (23.3) | |
| S-spine | 1 (1.7) | 0 | 1 (3.3) | |
| No. (%) with following no. of vertebral bodies involved: | ||||
| 1 | 8 (13.8) | 3 (10.7) | 5 (16.7) | 0.757 |
| 2 | 35 (60.3) | 17 (60.7) | 18 (60.0) | |
| 3 | 10 (17.2) | 8 (28.6) | 2 (6.7) | |
| ≥4 | 5 (8.6) | 0 | 5 (16.7) | |
| Duration (days) of prior antibiotic exposuref | 7.0 (3.0–15.0) | 4.5 (2.3–14.3) | 7.5 (4.0–15.5) | 0.086 |
| Antibiotic-free period (days)f | 1.5 (0.7–5.7) | 1.1 (0.2–3.3) | 3.3 (1.2–7.6) | 0.007 |
| No. (%) undergoing biopsy by: | ||||
| Needle | 48 (82.8) | 21 (75.0) | 27 (90.0) | 0.173b |
| Open surgery | 10 (17.2) | 7 (25.0) | 3 (10.0) |
Mean ± standard deviation.
Fisher's exact test.
Body temperature of ≥38.0°C.
Motor power, ≤grade 4 (full range of movement but resisted by examiner) (8).
Two or more of the following: body temperature of >38°C or <36°C, heart rate of >90/min, respiratory rate of >20/min or partial CO2 pressure of <32 mm Hg, WBC count of >12,000/mm3 or <4,000/mm3 or >10% bands.
Median (interquartile range).
TABLE 2.
Characteristics and positivity rates of tissue and/or blood cultures according to antibiotic-free periods in patients with PVO
| Parameter | Antibiotic-free period of: |
P value | ||
|---|---|---|---|---|
| <48 h (n = 32) | ≥48 to <120 h (n = 10) | ≥120 h (n = 16) | ||
| Mean age (yr) ± SD | 63.4 ± 14.3 | 66.5 ± 15.3 | 71.1 ± 13.1 | 0.209 |
| No. (%) of males | 16 (50.0) | 5 (50.0) | 3 (18.8) | 0.052 |
| No. (%) with fever | 12 (37.5) | 4 (40.0) | 4 (25.0) | 0.433 |
| No. (%) with motor weakness | 4 (12.5) | 2 (20.0) | 4 (25.0) | 0.273 |
| No. (%) with sepsis | 13 (40.6) | 4 (40.0) | 2 (12.5) | 0.065 |
| No. (%) with lumbar spine involvement | 27 (84.4) | 9 (90.0) | 14 (87.5) | 0.730 |
| No. (%) with >3 vertebral bodies involved | 9 (28.1) | 3 (30.0) | 3 (18.8) | 0.523 |
| No. (%) with abscesses | 22 (68.8) | 9 (90.0) | 11 (68.8) | 0.844 |
| No. of WBC/mm3a | 10,480 (9,158–14,375) | 10,950 (8,463–13,828) | 7,625 (6,718–10,025) | 0.023b |
| CRP level (mg/dl)a | 13.6 (6.8–19.3) | 15.0 (3.8–21.0) | 5.9 (1.6–11.3) | 0.089b |
| Prior antibiotic exposure duration (days)a | 4.0 (3.0–8.8) | 5.0 (2.8–12.8) | 17.0 (7.0–33.5) | <0.001b |
| No. (%) with cephalosporin as prior antibiotic | 25 (78.1) | 4 (40.0) | 14 (87.5) | 0.768 |
| No. (%) with quinolone as prior antibiotic | 5 (15.6) | 1 (10.0) | 5 (31.2) | 0.245 |
| No. (%) with glycopeptide as prior antibiotic | 3 (9.4) | 1 (10.0) | 2 (12.5) | 0.747 |
| No. (%) with open surgical biopsy | 8 (25.0) | 2 (20.0) | 0 | 0.037 |
| No. of tissue cultures positive/total (%) | 20/32 (62.5) | 3/10 (30.0) | 5/16 (31.3) | 0.029 |
| No. of blood culture positive/total (%) | 12/29 (41.4) | 2/10 (20.0) | 2/14 (14.3) | 0.058 |
Median (interquartile range).
P value calculated by one-way analysis of variance.
TABLE 3.
Multivariate logistic regression analysis of factors associated with tissue culture positivity
| Factor | Univariate analysis |
Multivariate analysisa |
||
|---|---|---|---|---|
| OR (95% CI) | P value | aOR (95% CI) | P value | |
| Sex | ||||
| Male | 1.00 | |||
| Female | 0.37 (0.13–1.09) | 0.072 | 0.286 | |
| Age of ≥65 yr | 0.89 (0.31–2.53) | 0.825 | 0.374 | |
| Fever | 1.11 (0.38–3.28) | 0.849 | 0.533 | |
| Motor weakness | 1.09 (0.28–4.25) | 0.905 | 0.990 | |
| Sepsis | 2.46 (0.80–7.63) | 0.118 | 0.405 | |
| L-spine involved | 0.51 (0.11–2.37) | 0.392 | 0.186 | |
| No. of involved vertebral bodies | ||||
| 1 or 2 levels | 1.00 | |||
| ≥3 levels | 1.31 (0.41–4.27) | 0.649 | 0.619 | |
| Duration of prior antibiotic exposure | 0.97 (0.92–1.02) | 0.176 | 0.783 | |
| Prior antibiotic exposureb | ||||
| Cephalosporin | 0.76 (0.23–2.47) | 0.649 | 0.842 | |
| Quinolone | 0.33 (0.08–1.40) | 0.133 | 0.476 | |
| Glycopeptide | 1.08 (0.20–5.85) | 0.929 | 0.905 | |
| Antibiotic-free period | 0.80 (0.66–0.96) | 0.015 | 0.407 | |
| Presence of abscess | 1.83 (0.56–5.96) | 0.314 | 0.550 | |
| Initial CRP levelc | 1.15 (1.06–1.25) | 0.001 | 1.18 (1.07–1.29) | <0.001 |
| Tissue biopsy method | ||||
| Needle | 1.00 | |||
| Open surgery | 3.00 (0.69–13.02) | 0.142 | 6.33 (1.12–35.86) | 0.037 |
The aORs were calculated with a stepwise forward selection logistic regression model.
Cephalosporin, glycopeptide, and quinolone were not simultaneously included in the multivariate model.
Because the CRP level and WBC count showed a significant correlation (Pearson correlation coefficient of 0.549, P < 0.001), we included only the CRP level in the multivariate analysis model.
In this study, a higher tissue culture positivity rate in patients with PVO who were exposed to antibiotics before the acquisition of tissue culture specimens was associated with a higher CRP level and open surgical biopsy but not with a longer antibiotic-free period. The use of a variety of antibiotics and classes of antibiotics, even broad-spectrum antibiotics such as fluoroquinolones, had no effect on culture positivity. It is well known that the yield of open surgical biopsy is higher than that of needle biopsy (9–11). Also, a higher CRP level occurs more frequently in patients with culture-positive PVO (3, 12, 13). One might expect that the longer the antibiotic-free period is, the higher the culture positivity rate is because more antibiotic would be removed from tissue and the negative effect of antibiotic exposure would be lessened. However, tissue culture positivity rates were significantly higher in patients whose antibiotic-free period was less than 48 h than in those for whom it was 48 to 120 h or >120 h. This could be because patients with shorter antibiotic-free periods underwent open surgical biopsy more frequently and had higher CRP levels than those with longer antibiotic-free periods, and the attending physicians were reluctant to postpone the acquisition of tissue culture specimens from patients with high CRP levels.
In this study, we found no association between the antibiotic-free period duration and a higher culture positivity rate. We suggest that obtaining culture specimens by the open surgical method instead of postponing their acquisition is desirable for patients who have received antibiotics during the preceding 2 weeks in order to increase culture positivity, especially when a needle biopsy does not reveal the causative organism, because the culture positivity rates associated with a surgical biopsy were higher (70%, 7/10) than those associated with a needle biopsy (43.8%, 21/48). Our study had several limitations. First, the number of patients investigated was small and the power of our analysis was limited. Second, the antibiotic-free period was relatively short (median, 1.5 days) and there remains the possibility that a period of a few weeks instead of a few days would improve culture positivity. A prospective observational cohort study investigating the effect of a longer antibiotic-free period on patients who are clinically stable, as defined by the absence of neurologic deficits without progression and with lower CRP levels, would be very helpful.
In conclusion, the tissue culture positivity rate of patients with PVO who were exposed to antibiotics was associated with higher CRP levels and open surgical biopsy; however, it was not improved by an antibiotic-free period of a few days.
ACKNOWLEDGMENTS
This work was supported by a grant from the research fund of the Seoul National University Hospital (0420120420).
We have no conflicts of interest to declare.
REFERENCES
- 1.Gouliouris T, Aliyu SH, Brown NM. 2010. Spondylodiscitis: update on diagnosis and management. J Antimicrob Chemother 65(Suppl 3):iii11–iii24. doi: 10.1093/jac/dkq303. [DOI] [PubMed] [Google Scholar]
- 2.de Lucas EM, Gonzalez Mandly A, Gutierrez A, Pellon R, Martin-Cuesta L, Izquierdo J, Sanchez E, Ruiz E, Quintana F. 2009. CT-guided fine-needle aspiration in vertebral osteomyelitis: true usefulness of a common practice. Clin Rheumatol 28:315–320. doi: 10.1007/s10067-008-1051-5. [DOI] [PubMed] [Google Scholar]
- 3.Kim CJ, Song KH, Park WB, Kim ES, Park SW, Kim HB, Oh MD, Kim NJ. 2012. Microbiologically and clinically diagnosed vertebral osteomyelitis: impact of prior antibiotic exposure. Antimicrob Agents Chemother 56:2122–2124. doi: 10.1128/AAC.05953-11. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Rankine JJ, Barron DA, Robinson P, Millner PA, Dickson RA. 2004. Therapeutic impact of percutaneous spinal biopsy in spinal infection. Postgrad Med J 80:607–609. doi: 10.1136/pgmj.2003.017863. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Hassoun A, Taur Y, Singer C. 2006. Evaluation of thin needle aspiration biopsy in the diagnosis and management of vertebral osteomyelitis (VO). Int J Infect Dis 10:486–487. doi: 10.1016/j.ijid.2006.02.011. [DOI] [PubMed] [Google Scholar]
- 6.Zimmerli W. 2010. Clinical practice. Vertebral osteomyelitis. N Engl J Med 362:1022–1029. doi: 10.1056/NEJMcp0910753. [DOI] [PubMed] [Google Scholar]
- 7.Colmenero JD, Jimenez-Mejias ME, Sanchez-Lora FJ, Reguera JM, Palomino-Nicas J, Martos F, Garcia de las Heras J, Pachon J. 1997. Pyogenic, tuberculous, and brucellar vertebral osteomyelitis: a descriptive and comparative study of 219 cases. Ann Rheum Dis 56:709–715. doi: 10.1136/ard.56.12.709. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.O'Brien M. 2010. Aids to the examination of the peripheral nervous system, 5th ed Saunders Elsevier, New York, NY. [Google Scholar]
- 9.Marschall J, Bhavan KP, Olsen MA, Fraser VJ, Wright NM, Warren DK. 2011. The impact of prebiopsy antibiotics on pathogen recovery in hematogenous vertebral osteomyelitis. Clin Infect Dis 52:867–872. doi: 10.1093/cid/cir062. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.McHenry MC, Easley KA, Locker GA. 2002. Vertebral osteomyelitis: long-term outcome for 253 patients from 7 Cleveland-area hospitals. Clin Infect Dis 34:1342–1350. doi: 10.1086/340102. [DOI] [PubMed] [Google Scholar]
- 11.Osenbach RK, Hitchon PW, Menezes AH. 1990. Diagnosis and management of pyogenic vertebral osteomyelitis in adults. Surg Neurol 33:266–275. doi: 10.1016/0090-3019(90)90047-S. [DOI] [PubMed] [Google Scholar]
- 12.Bhagat S, Mathieson C, Jandhyala R, Johnston R. 2007. Spondylodiscitis (disc space infection) associated with negative microbiological tests: comparison of outcome of suspected disc space infections to documented non-tuberculous pyogenic discitis. Br J Neurosurg 21:473–477. doi: 10.1080/02688690701546155. [DOI] [PubMed] [Google Scholar]
- 13.Lora-Tamayo J, Euba G, Narvaez JA, Murillo O, Verdaguer R, Sobrino B, Narvaez J, Nolla JM, Ariza J. 2011. Changing trends in the epidemiology of pyogenic vertebral osteomyelitis: the impact of cases with no microbiologic diagnosis. Semin Arthritis Rheum 41:247–255. doi: 10.1016/j.semarthrit.2011.04.002. [DOI] [PubMed] [Google Scholar]