A 66-year-old woman with osteoporosis presented with 2 months of severe pain in the lower back, pelvic region, and buttock. The T score of bone mineral density was –3.5. The pain was exacerbated by sitting or standing, and was relieved by assuming the supine position.

Before the sacroplasty, the pain on walking was rated as 8–9/10 on the visual analog scale (VAS) of 0 to 10. Magnetic resonance imaging (MRI) revealed fractures of the S1 sacral body and both alae (Fig. 1A).

The patient had no specific significant trauma history. SIFs are stress fractures that result from normal stress on abnormally weak bones and are most frequently observed in the setting of osteoporosis.

The absolute bed rest and analgesic medication over the 3-week period did not achieve a satisfactory reduction of the severe disabling pain. We discussed with the patient the potential risks and benefits of transiliac sacroplasty and those of conservative medical management. After serious discussion with the patient and her family regarding the procedure, the patient decided to undergo transiliac sacroplasty.

The procedure was performed under conscious sedation with midazolam and fentanyl, and continuous monitoring. A single-shot antibiotic (cefazolin 1 g intravenously) was routinely administered immediately before the intervention to provide coverage against skin flora. The patient was placed in the prone position under the MSCT scanner. MSCT in an axial slice orientation with multiplanar reformations was then performed before the operation.

Preoperative MSCT (axial spiral with a layer thickness of 5 mm and coronal and sagittal reformations with a reconstruction interval of 2 mm) of the pelvis was performed, and the entry point was marked on the skin. The entry point was established by tilting the MSCT scan gantry to find a safe pathway that does not cross the anatomical landmarks such as the neural foramen and cortical margins of the sacrum. The skin entry points for the alar fractures were also checked (Fig. 1B). After establishment of the skin entry point by using a CT scan, the usual preparation was performed and the respective approach angle was checked. Skin anesthesia was induced, and the muscle and periosteum were anesthetized using 1% lidocaine.

Under MSCT guidance using the thin-slice technique, an 11-G percutaneous vertebroplasty (PVP) needle (Hyun Medical, Korea) was introduced to the iliac bone via the transiliac approach from lateral to medial via the trajectory previously determined on the CT scanner [8]. We confirmed the placement of the needle tip within the sacral vertebral body without compromising the sacral canal and sacral neural foramens (Fig. 2). After removing the inner needle, a polymethylmethacrylate (PMMA) mixture (CMW1, DePuy International Ltd., UK) was injected to the sacrum under MSCT (Fig. 3) by using 1-ml syringes. The cement, which was in a toothpaste consistency, was injected with the needle directed laterally to encourage the flow of the cement away from the sacral foramen. The goal of the PMMA injection was to fill the fracture site and surrounding bone without extravasation of the cement. Cement filling was performed discontinuously under MSCT guidance. PMMA cement was injected in 0.5-ml aliquots, and the total volume of cement injected to the sacral body fracture site was 2.5 ml. The sacral alar fractures were treated after PMMA injection to the S1 sacral body.

An 11-G PVP needle was then advanced into the respective fracture zone in the sacrum from a dorsal to a ventral direction (short axis), and cement filling was performed discontinuously (0.5 ml) under MSCT guidance. Furthermore, 4.1 (left) and 3.7 ml (right) of cement was injected via each cannula (Fig. 4A). The patient was then allowed to walk 2 hours after the procedure.

The pain intensity in the buttock area was determined using the VAS before the intervention, 2 days after the intervention, and 1 month after the intervention. The VAS pain score was 8 before the intervention, and a significant reduction in pain to a VAS score of 3 was observed on the second postoperative day, which stabilized after 1 month to a VAS score of 4. No complication occurred related to the procedure, such as iliac fracture, at the 1-month follow-up. Transiliac sacroplasty was technically feasible under the MSCT-guided technique. The MSCT scan showed a central distribution of the cement in the sacrum and a good distribution over the area of the sacral fracture (Fig. 4).

A 73-year-old woman with a history of osteoporosis was admitted after a slip injury. Fifteen years ago, she underwent lumbar posterior fusion at L3/L4 with pedicle screw fixation and PVP at L1 at another hospital. She reported severe coccygeal and buttock pains limiting her motions. A subjective pain assessment revealed an 8/10 pain score, which was not controlled by analgesic drug administration and injection therapy. The pain had been preventing her from sitting and walking normally. An MRI scan demonstrated sacral S1 and S2 body fractures without an alar fracture (Fig. 5A). We performed MSCT-guided percutaneous sacroplasty at the S1 and S2 sacral bodies via the transiliac approach by using the same method. A preoperative MSCT scan of the pelvis was performed, and the skin entry points at S1 and S2 were marked. The entry point was established by tilting the MSCT scan gantry to find a safe pathway that does not cross the anatomical landmarks such as the neural foramen and cortical margins of the sacrum (Figs. 5B and 5C). The amount of cement required depended on the extent of the fracture and PMMA leakage. Furthermore, 2.0 ml of PMMA cement was introduced at the S1 sacral body (Fig. 6) and 1.6 ml at the S2 sacral body (Figs. 7A and 7B). The post-procedural CT and radiography revealed the optimal filling of the lesions with no evidence of PMMA extravasation into the central canal or sacral foramina (Fig. 7C). The VAS pain score significantly decreased from 9 preoperatively to 3 on the second postoperative day and 3 in the first month after the intervention, and the fracture of the patient was successfully mobilized.