Between 2003 and 2010, we enrolled patients who underwent single-level PVP for painful OVCF with IVC in our hospital. This study was approved by our institutional review board, and all patients received written informed consent for undergoing PVP prior to participation in this study.

Painful osteoporotic vertebral pseudarthrosis was diagnosed by the presence of IVC and tenderness at the same level. All preoperative plain radiographic, computed tomography (CT), and MRI data were evaluated for diagnosing IVC. On plain radiography and CT, an IVC was defined as a linear well-demarcated focus of intervertebral fluid or gas attenuation [7]. On MRI, an IVC was defined as a linear well-demarcated focus of T2 prolongation similar to that of cerebrospinal fluid. Low signal intensity on T1- and T2-weighted images, which is characteristic of gas, was also considered to indicate an IVC [18, 19]. On dynamic contrast-enhanced MRI, IVC was illustrated as a noncontrast area, and vertebrae around the IVC were enhanced with contrast medium [20] (Fig. 1).

Fig. (1). On radiographs and CT, IVC was defined as a linear well-demarcated focus of intervertebral fluid or gas attenuation (a: lateral radiographs, b: CT sagittal). On MRI, IVC was defined as a linear well-demarcated focus of T2 prolongation similar to that of adjacent cerebrospinal fluid. Signal void on T2- and T1-weighted images, which is characteristic of gas, was also considered an IVC (c: fat suppressed T2WI sagittal, d: T2WI sagittal). On dynamic contrast-enhanced MRI, IVC was shown as a non-contrast area and vertebra around the IVC was enhanced with contrast medium (e: sagittal).

Fig. (2). Before PMMA injection, we performed a “cavitygram” of the IVC using nonionic contrast agent to exclude needle placement within the basivertebral venous complex, and measured the capacity of IVC (a: before cavitygram, b: after cavitygram).

Fig. (3). Radiographic parameters: (a) “local kyphotic angle” (α°) measured as the angle between the lower and upper endplates of the uninvolved vertebrae adjusted cephalic and caudal to the fractured vertebra on lateral radiography with the patient in a sitting position, (b) “percent spinal canal compromise” calculated by dividing the area of intrusion by total spinal canal area multiplied by 100. The total canal area is outlined by the solid line; the area of the retropulsed vertebral wall is demarcated by the dotted line. Areas of the spinal canal and retropulsed posterior wall are calculated from the total number of pixels per cross-section area (pixel/mm2), (c) intravertebral instability of the affected vertebra, measured as the difference between local kyphotic angle on lateral radiography with the patient in a sitting position (α°) and that on sagittal reformatted CT in a supine position (β°); α°-β°.

Fig. (4). On CT before PVP, IVC was seen (a). On CT two years after PVP, bony bridge was seen between the fractured vertebra and the adjacent vertebra (b).

The selection criteria for PVP for treating OVCF with IVC were as follows: 1) sufficient back pain (visual analog scale [VAS] ≥ 4) refractory to standard medical treatment consisting of bed rest, analgesics, and/or external back bracing for at least 3 months; and 2) radiographic evidence of single-level OVCF with IVC with the presence of point tenderness on manual palpation. The exclusion criteria included the following: 1) spinal cancer, active infection, or uncorrectable bleeding diatheses; 2) inability to provide informed consent; and 3) a likelihood of noncompliance with direct follow-up; and 4) neurological deficit such as leg pain and motor weakness caused by neural compression or intravertebral instability of the affected vertebra.

PVP was performed by experienced spine surgeons. After general anesthesia, patients were carefully positioned in a prone position with extended posture on a radiolucent four-poster spinal frame. Next, 14-gauge bone needles (Ossiris; Hakko, Nagano, Japan) were inserted into the IVC through a bilateral transpedicular approach with direct biplane observation using a couple of fluoroscopes. The IVC was a confluent reservoir that was specifically targeted for polymethylmethacrylate (PMMA) (Osteobond; Zimmer, Warsaw, IN) injection. Before the PMMA injection, we performed a “cavitygram” of the IVC using nonionic contrast agent (Omnipaque 300; Nycomed, Princeton, NJ) to exclude needle placement within the basivertebral venous complex [21-23] and measured the capacity of the IVC via an injection of contrast medium from one side of the needle while the other side of the needle cannula was opened (Fig. 2). The residual contrast material was washed out with normal saline to clear the IVC adequately to visualize PMMA. Barium-opacified PMMA of the same volume as the capacity of the IVC was gently injected using 2-mL syringes. PMMA was injected via a one-sided needle with low pressure until the IVC was filled. PMMA assumed the shape of the IVC without evidence of extravasation into surrounding bone marrow space, paravertebral soft tissues, or epidural space. The procedure was terminated when the IVC was filled with PMMA. On the next day after PVP, CT was performed to determine whether extravertebral PMMA leakage had occurred. An orthopedic surgeon not involved in the procedure reviewed the CT images independently and reached a consensus for each case.

The following three radiographic parameters were assessed before PVP, 1 month after PVP and at final follow-up : 1) local kyphotic angle, measured as the angle between the lower and upper endplates of the uninvolved vertebrae adjusted cephalically and caudally to the fractured vertebra on lateral radiography with the patient in a sitting position; 2) percent spinal canal compromise on CT, calculated by dividing the area of intrusion by the total spinal canal area and multiplying the resulting value by 100; and 3) intravertebral instability of the affected vertebra, measured as the difference between the local kyphotic angle on lateral radiography with the patient in a sitting position and that on sagittal reconstructed CT with the patient in a supine position (Fig. 3).

Back pain was measured using a VAS score of 0-10, with 0 indicating no pain and 10 indicating the maximum imaginable pain [24]. Physical disability was measured using the Oswestry Disability Index (ODI) on a scale of 0-100%, with higher scores indicating greater disability [25]. The improvement rate was calculated using the following formula: ([baseline score − postoperative score]/baseline score) × 100. Patients were followed up directly and periodically after PVP. The mean follow-up period was 28 months (range, 6-55 months). Orthopedic surgeons not involved in the treatment performed the follow-up and clinical examinations to assess the patients’ functional status. The VAS score and ODI questionnaire o was self-administered and collected before PVP, 1 month after PVP and at final follow-up by a medical secretary to avoid interviewer bias.

Intra- and post-operative complications such as anesthesia- and general condition-related complications, surgical procedure-related complications, and cement leakage were evaluated. Risk factors for vertebral fractures adjacent to the augmented vertebra following PVP were also evaluated.

Clinical characteristics, radiographic parameters, and procedural outcomes were analyzed using the Wilcoxon signed-rank test, Mann-Whitney U-test, and/or Bonferroni-Dunn post hoc test. The rates of procedure-related complications were estimated with 95% confidence intervals (CIs). Risk factors for vertebral fractures adjacent to the augmented vertebra following PVP were analyzed using multivariate logistic regression models. Statistical significance was defined at the level of p < 0.05 for a two-sided hypothesis. Data are presented as the mean ± standard deviation. Analyses were performed using SPSS version 16.0 (SPSS, Chicago, IL).

Two hundred ninety-one patients with painful OVCF with IVC (207 women and 84 men; mean age, 77 years) who underwent PVP at our institution were included in this study. Patients who died without operative complications (15), those with severe dementia (16), and those lost to follow-up (5) were excluded from the analysis, leaving 255 patients in this study. Patients were aged 61-93 years (mean 77) at the time of examination. OVCFs were detected from T7 to L5, and they occurred most frequently at the thoracolumbar junction (84%). There were 232 patients (91%) with primary osteoporosis and 23 patients (9%) with steroid-induced osteoporosis. The mean number of OVCFs was 1.2 in patients with primary osteoporosis and 1.4 in those with steroid-induced osteoporosis, and there were no significant differences in these findings. The mean duration from the onset of acute back pain to PVP was 29 ± 9.1 weeks (Table 1).

The average local kyphotic angle was 10.5° before PVP, 6.2° 1 month after PVP, and 8.1° at the final follow-up. The average local kyphotic angles 1 month after PVP and at the final follow-up were significantly smaller than that before PVP (p < 0.001). The average percentage spinal canal cross-sectional area of compromise was 7.9% before PVP, 17.1% 1 month after PVP, and 15.2% at the final follow-up. There was no statistical significance in the average percentage spinal canal cross-sectional area among the time points. The average intravertebral instability was 6.1° before PVP, 1.7° 1 month after PVP, and 0.8° at the final follow-up. The average intravertebral instability 1 month after PVP and that at the final follow-up was significantly less than that before PVP (p < 0.001) (Table 2). In total, 62 of 106 (58.5%) patients who received the reconstructed CT 2 years after PVP exhibited bony bridges between the fractured vertebra and the lower or upper vertebrae on sagittal reconstructed CT 2 years after PVP (Fig. 4).

The mean VAS values were 8.4 before PVP, 2.3 1 month after PVP, and 3.9 at the final follow-up. The improvement rates according to the VAS were 72.6% 1 month after PVP and 53.6% at the final follow-up. All patients displayed statistically significant improvements in the VAS after PVP (p < 0.001). The mean ODI values were 60.0% before PVP, 33.1% 1 month after PVP, and 35.4% at the final follow-up. The recovery rates according to the ODI were 45.1% 1 month after PVP and 40.0% at the final follow-up. All patients exhibited statistically significant improvement in the ODI after PVP (p < 0.001). The numbers of patients with no back pain (VAS = 0) were 73 (28.6%) 1 month after PVP and 46 (18%) at the final follow-up (Table 3).

Cement leakage was observed in 60 of 255 patients (23.5%). Small amounts of cement leakage occurred in the epidural space (3 of 255 patients, 1.2%), perivertebral soft tissue (12 of 255, 4.7%), and intervertebral disc space (45 of 255, 17%) (Table 4). There were no neurological or systemic complications due to cement leakage. Vertebral fracture adjacent to the augmented vertebra was noted in 62 of 255 patients (24.3%). The risk factors for vertebral fracture adjacent to the augmented vertebra were evaluated. We analyzed age, sex, bone density, injected PMMA volume, local kyphotic angle before surgery, and PMMA leakage into the disc space as risk factors for vertebral fracture. The risk factors for new fracture were the injected PMMA volume (odds ratio [OR] = 1.1, 95% CI = 1.0-1.3, p = 0.007), local kyphotic angle before surgery (OR = 1.1, 95% CI = 1.0-1.2, p = 0.017), and age (OR = 2.0, 95% CI = 1.2-3.4, p = 0.013) (Table 5).