Spinal biopsy is important to obtain culture and histopathological diagnosis in spine infection and tumors. To date, there is no prospective randomized study comparing both techniques. Therefore, the aim of this prospective randomized trial was to evaluate the accuracy, safety and diagnostic outcome of both these techniques.A prospective randomized trial was performed in 60 patients divided equally into fluoroscopic and CT guided spinal biopsy groups. Transpedicular approach was done with 8G core biopsy needle. Specimens were sent for histopathological examinations and cultures. Diagnosis were made based on biopsy results, clinical criteria and disease progression during 6 months follow up. Radiation expose to patients and doctors were measured with optically stimulated luminescence dosimeters (OSLDs).

This prospective randomized study was carried out in a single institution from November 2016 to June 2018.

Patients were recruited by three spine surgeons from the same institution after eligibility assessment. Sixty consecutive patients were randomized equally into fluoroscopic and CT guided spinal biopsy groups. Randomization procedure was performed using a computer generated random number sequence on www.random.org. Separate individuals completed the random allocation sequence, patient enrollment and outcome assessment. Outcome assessors were blinded to group assignment.

Data collection Clinical data collected were: patient demographics, comorbidities, infective parameters, tumor markers, Magnetic Resonance Imaging (MRI) findings and pre-biopsy pain score at rest using visual analogue scale (VAS). Intraoperative data included radiation exposure dosage, biopsy duration and complications encountered. Following biopsy, we recorded the biopsy result, 6-hourly pain score (VAS) at for 48 hours, additional analgesia requirement and complications.

Radiation measurement Commercial optically stimulated luminescence dosimeters, OSLDs (NanoDotTM Dosimeter, Landauer Inc, Glenwood, IL) were used to measure radiation dose. Ten OSLDs were placed on patients and doctors prior to all biopsies over selective areas representing radiosensitive organs. Placements of the OSLDs were done by a single observer. OSLDs were processed using MicroStar OSLD reader system (Landauer, Inc. Glenwood, IL). Each OSLD was processed 3 times during each reading to obtain a mean value to minimize error. The absorbed doses were measured in centigrays (cGy). Bleaching/zeroing of the OSLDs were done using optical annealing system containing fluorescent lamps (Osram lumilux, 24 W, 280 ~780 nm) in between each use of OSLDs.

Biopsy technique Transpedicular approach with an 8-gauge T-Lok™ bone marrow biopsy needle (Argon Medical Devices, Athens, TX, USA) was used in both arms. 10cc 2% lignocaine was infiltrated prior to biopsy. Biopsy needle was tapped gently with mallet through the pedicle and into the lesion under imaging guidance. Trocar was then withdrawn once it reached the affected area. The cannula was rotated several times to disengage sample from surrounding tissue. Then, aspiration of samples was done using 10-ml syringe prior to retraction of the cannula. Tissue samples were retrieved from the cannula with the use of stylet.

Samples collected were sent to laboratories for histopathological examination (HPE), culture and sensitivity (bacteriology, mycology and acid-fast bacilli), staining (Gram and Ziehl-Neelsen) and polymerase chain reaction (PCR) test for Mycobacterium tuberculosis.

1. Fluoroscopy Guided Spinal Biopsy Arm: Fluoroscopy guided spinal biopsies were done by spine surgeons in the operation theatre using C-Arm fluoroscopy device (OECFluorostar7900Compact - GE®). C-arm image intensifier was positioned until fluoroscopic beam was collinear with the corresponding vertebral body on AP view prior to insertion of biopsy needle.
2. CT guided Spinal Biopsy Arm: CT guided spinal biopsies were done by radiologist under guidance of 128-slice CT scanner (SOMATOM Definition AS+, Siemens Medical Solutions USA, Inc.). Preliminary axial CT scanning was done in bone window (window width, WW:2500; window level, WL:500) at the pre-selected levels decided from previous imaging to plan the needle access. Biopsy was then performed under sequential CT-fluoroscopy scan.

Post biopsy care Antimicrobials were commenced upon completion of biopsy for patients who were diagnosed to have spine infection. Disease progression was assessed based on clinical symptoms, ESR, CRP and/or MRI repeated during 6 months follow up for patients who were not treated for spinal infection.

For suspected spinal tumor cases, negative biopsies that were reported to have inadequate sample, were repeated with the other method or other investigation modalities will be resorted accordingly. Patients with negative biopsy that was reported to have adequate sample were follow up for 6 months for disease progression.

Sample size Sample size was determined based on accuracy of biopsy as outcome variable, in which the accuracy rate was 0.89. It was estimated using G*Power software (version 3.1.9.2) based on 2 proportions formula (Fisher's exact test). With alpha level of 0.05, 95% confidence interval and power of study 0.8, a minimum number of 52 patients (26 patients for each arm) was required for this study.

Statistical analysis Data were analyzed with SPSS Inc., Chicago, IL (SPSS v 23). Continuous data were reported as mean ± standard deviation and compared using t test. Chi-squared testing was used for between-groups comparisons. Fisher-exact test was used for the comparison of diagnostic accuracy between fluoroscopic and CT guided spinal biopsy. A p-value of <0.05 was considered significant.

Inclusion Criteria:

1. Age more than 18;
2. Patients who presented with clinical symptoms and radiological features suggestive of infection or malignancy;
3. MRI reported as probable infection or tumour.

Exclusion Criteria:

1. Cervical spine lesion;
2. Significant neurological deficit that required urgent surgery;
3. Partially treated spinal infections;
4. Vascular tumour;
5. Infection over skin puncture site;
6. Pregnancy;
7. Bleeding diathesis;
8. Medically unfit for spinal biopsy.

Separate individuals completed the random allocation sequence, patient enrollment and outcome assessment. Outcome assessors were blinded to group assignment.