Vertebral Osteomyelitis (VO) is an infectious disease of the vertebral body which can involve the intervertebral disc space.1 The clinical manifestations, laboratory markers and radiological findings may be sometimes unspecific, thus its diagnosis turns frequently into a challenge.2,3 Therefore, current ISDA (Infectious Diseases Society of America) guidelines suggest that a proper evaluation of a patient should include a thorough neurological examination, baseline CRP and ESR determination, a spine imaging technique (preferred MRI) and bacterial blood cultures. Moreover it is recommended, if possible, to withhold the antibiotic therapy until a microbiological diagnosis could be established. However, if blood cultures do not yield the causative microorganism, an image-guided aspiration biopsy is encouraged.3 The most frequent microorganism involved is Staphylococcus aureus in more than half of the cases but other causes could be seen such as other bacteria species or even fungi or mycobacteria.4,5

Biopsy material could be obtained via open procedure (open biopsy) or by percutaneous access (usually assisted by CT or fluoroscope). There are several studies reporting both techniques yield, which had shown a wide range of results. Some of them had shown low positivity rates with CT guided biopsy (30–50%) and higher with open biopsy,6–8 however other groups highlighted the importance of withholding antibiotics prior to the procedure,9–12 fact that needs to be further explored because other studies have failed in finding any relationship between the antibiotic exposure and the technique result. Regardless, though open biopsy seems to be more reliable, percutaneous procedures such as CT guided biopsy are preferred1 because they are safer, less invasive, require shorter hospital stays and in most cases, general anesthesia is not required.13

Since there is still controversy around this topic,6 the aim of this study was to analyze which are the main factors that could influence in the result of a CT guided biopsy in vertebral osteomyelitis patients.

In this study including VO patients, the focus has been placed on the biopsy that is needed in many patients, in an attempt of elucidating the causative microorganism so as to provide a targeted antibiotherapy, and thus, reducing side effects and improving the efficacy.1,7 In particular CT guided needle biopsy had been analyzed in a population with demographic and clinical comorbidities distribution similar to other studies.2,6,7 The population of the study had been split in two groups attending to the result of the culture of the sample obtained from de biopsy.

In this cohort a higher presence of immunosuppressed patients could be noted, due to the characteristics of the center, which is a university hospital. The results of that subset of patients have been displayed separately. In general, the distribution of cardiovascular risk factors and other clinical variables has been fair between both groups. Although there are some differences on the radiological variables (more prevalence of epidural abscess and cord compression) and also, the acute phase markers were slightly greater on the positive result group. In any case, extravertebral extension could not act as a factor of poor prognosis.4 This fact would be explored on further studies but seems to be logical due to a presumed more aggressiveness of the infection on immunocompromised host. Moreover, other studies found correlation between higher CRP value and greater chance of a positive result.7–9

The signs and symptoms collected seems to be in agreement with other series3,4,7 with the expected high prevalence of back pain, presented in almost all patients and fever in just 1 out 3 patients, approximately in both groups. As seen in other works, delay in diagnosis is still a problem in VO patients2,4 who usually waits at least a month since the beginning of the pain, until diagnosis. Additionally, in some cases this period is increased dramatically. Further strategies should be set up in order to reduce this term and prevent future complications.14

Infectious endocarditis is a frequent association with VO,15,16 but in this study were scarcely represented because in those patients the causative microorganism is more frequent detected through the blood cultures.

Acute phase markers (CRP and ESR) were elevated in both groups. Numerically, values were higher in biopsy positive groups, as seen in other studies7–9 but statistical signification was not reach. Radiological variables were comparable in both groups. A slight difference was noted in cord compression (higher in negative group) probably related to a greater presence of patients with acute paraparesis at diagnosis in this group. This result might be linked to the higher presence of further surgical approach detected in the negative culture group.

As it has been pointed out in several studies delay in diagnosis is probably the main issue in VO.2,17 Moreover, since antibiotic withholding is widely recommended,1 these biopsies should be performed as soon as possible. In this study a median delay of 6 days was detected on both groups, with few differences in the interquartile range (Table 2). Marschall et al. achieved a 65% of positivity with a median delay of 3 days.10 Better mean values in immunosuppressed patients have been noted, but overall, this should be improved and analyzed (for instance, creating a specific circuit).

Microorganisms isolated had a distribution comparable to another studies.2,18 Antibiotic exposure was registered in half of the cases on the positive group and in the 58% of the negative group. The higher prevalence of patients with acute paraparesis at diagnosis could be a reasonable cause. In addition, more antibiotic exposure is presumed to reduce the culture yield,11 but it was not detected on our study. This is a matter of controversy, but there are other several studies that concluded the same.12,17 A possible explanation of the lack or the small influence of antibiotic exposure on the culture result could be that intervertebral space is avascular. A high correlation between blood cultures and biopsy had been noted, so as it is had been previously recommended on the guidelines and other works,1 CT-guided biopsy could be avoided in most cases.

Overall positivity (56%) was on the average of the studies published which vary from 21% to 89%.4,7–9,19–21 The biopsy result was not modified by prior CRP values, fever as clinical presentation nor antibiotic exposure. The longer the length of painful period before diagnosis was related to a lower chance of obtaining a positive result on culture. This result could be explained by a greater aggressiveness of pyogenic organisms that perhaps congregate in the lesser time span instead of non-pyogenic agents and seems logical that the delay in diagnosis makes a negative impact on the technique reliability. Other studies has been performed in order to detect predictive factors of positivity, such as a recent study found that using larger diameter biopsy needle increases the likelihood of causative microorganism.19 Other factor detected related to higher chance of positivity has been the attenuation value on CT scan.20

In order to point out the limitations of the study, since this is a retrospective study, clinical cases may be missed during the search process. Moreover, patients were admitted to different wards, and that hence on variability in patient management. Concerning the procedure, adding more variables would provide more information such as areas specifically punctured, specific antimicrobial treatment received prior to the procedure or number of samples taken at each procedure. During statistical analysis some difficulties were found, such a huge variability in days of antimicrobial therapy received prior the procedure, finally solved dichotomizing in “exposure: yes/no” and reducing the statistical power. The inclusion of mycobacteria and fungi might influence the clinical picture and the yield of the CT-guided biopsy.

In conclusion our study displays an acceptable reliability of CT guided needle biopsy in VO patients, even in cases under antibiotic treatment. The presence of fever or elevated CRP values did not predict a positive culture in our study. A positive blood culture would provide a microbiological diagnosis in most cases. Delay in diagnosis could impact negatively in culture yield. Therefore, improving diagnostic strategies in general and shortening the span from VO diagnosis to the procedure performance in particular, may enhance not only the culture yield reliability but also the patient prognosis.

None funding source was required.

None.