Cement-augmented endplate-penetrating pedicle screws for osteoporotic vertebral collapse: a technical note

Article information

Asian Spine J. 2026;.asj.2025.0820

Publication date (electronic) : 2026 July 8

doi : https://doi.org/10.31616/asj.2025.0820

Masahiro Kawanishi, Hidekazu Tanaka, Yutaka Ito, Naokado Ikeda, Kunio Yokoyama, Makoto Yamada, Akira Sugie, Daiji Ichihashi

Department of Neurosurgery, Takeda General Hospital, Kyoto, Japan

Corresponding author: Naokado Ikeda, Department of Neurosurgery, Takeda General Hospital 28-1 Ishidamori Minamimachi, Fushimi-ku, Kyoto City, Kyoto 601-1495, Japan, Tel: +81-75-572-6331, Fax: +81-75-571-8877, E-mail: naokado.ikeda@ompu.ac.jp

Received 2025 December 11; Revised 2026 January 16; Accepted 2026 February 12.

Abstract

Introduction

A variety of procedures have been proposed for the optimal surgical management of a retropulsed posterior wall fragment from an osteoporotic vertebral fracture (OVF) protruding into the spinal canal and compressing the neural elements; however, it remains a matter of serious deliberation [1,2]. Pedicle screw (PS) pullout is a frequent concern in short-segment posterior fixation in elderly patients with OVFs. To prevent this complication, several techniques have been developed, including endplate-penetrating (pedicular transvertebral) screw fixation [3] and fenestrated hollow screws that allow polymethyl methacrylate (PMMA) injection through the screw lumen to enhance screw anchorage [4]. However, even with penetrating endplate screws (PES), problems such as screw loosening and pullout are not completely resolved [5].

To address this, we devised a simple technique to combine PES with PMMA augmentation. In this technique, the tip of a hollow PS is intentionally filled with disc tissue to act as a biological “plug” before cement injection to enhance screw fixation while preventing cement leakage into the intervertebral disc space. Herein, we present two illustrative cases detailing the surgical technique and outcomes. This technical report does not establish a new standard technique, but rather describes a preliminary salvage approach applicable in select cases.

Technical Notes

Surgical technique

Case 1

A 73-year-old woman developed severe bilateral leg pain and progressive paraparesis approximately 3 months after sustaining an L1 compression fracture; eventually, she was unable to stand. Magnetic resonance imaging (MRI) of the spine showed an L1 posterior wall fragment retropulsed into the spinal canal, compressing the conus medullaris (Fig. 1A). We performed a partial laminectomy of L1 followed by short-segment posterior fixation from T12 to L2, including the fractured vertebra (Fig. 1B). PSs at the T12 level were directed caudally to penetrate the T12 inferior endplate, and PSs at L2 were directed cranially to penetrate the L2 superior endplate. At each level, a guidewire was advanced transpedicularly to penetrate the bony endplate of the same vertebra (Fig. 2). Hollow PSs were then inserted over the guidewires.

Fig. 1

(A) Sagittal magnetic resonance imaging (MRI) showing collapse of the L1 vertebral body with a retropulsed posterior wall fragment compressing the conus medullaris. (B) Lateral radiograph showing the T12–L2 construct with endplate-penetrating screws at T12 and L2 and short screws at the collapsed L1 vertebra, all connected by rods. (C) Postoperative MRI confirming adequate decompression of the spinal cord. (D) Plain radiographs 1 year after surgery showing maintained sagittal alignment.

Fig. 2

(A, B) Direction of guidewire insertion. The guidewire is advanced transpedicularly to penetrate the bony endplate of the same vertebra.

Once the screw tip had slightly penetrated the endplate (Fig. 3A), the guidewire was removed (Fig. 3B), and the screw was advanced further so that intervertebral disc tissue entered the screw lumen and “plugged” the tip (Fig. 3C). Before cement injection, a guidewire was reinserted into the hollow screw lumen to confirm that the plug was firmly seated at the screw tip (Fig. 3D). Lateral fluoroscopy was used to confirm that the guidewire tip stopped within the vertebral body and did not cross the endplate.

Fig. 3

(A) A hollow pedicle screw is inserted over a guidewire. When the screw tip slightly penetrates the bony endplate, the guidewire is removed (red arrow). (B) The now hollow screw is advanced further along the same trajectory (red arrow). (C) As the screw is advanced, intervertebral disc tissue enters the screw lumen, and the screw tip becomes effectively “plugged” by disc material (red ellipse). (D) Under lateral fluoroscopic guidance, a guidewire is reinserted into the lumen of the hollow screw. The surgeon confirms that the wire tip stops within the vertebral body (green asterisk) without crossing the endplate, indicating that the screw tip is firmly plugged before cement injection. (E) Polymethyl methacrylate is injected into the vertebral body through the side holes of the pedicle screw after its tip has been plugged.

In our intraoperative experience, the plug usually contains disc tissue as well as some bony endplate component; this is evident when the guidewire meets clear resistance and does not easily advance beyond the screw tip even when gentle pressure is applied. Under fluoroscopic guidance, PMMA is injected only after it has reached a paste-like, highly viscous “doughy” state, similar to that commonly used for fenestrated PSs, i.e., when cement extruded from the cannula tip does not drip by gravity. Under standard operating-room temperatures, this typically occurs approximately 10–15 minutes after mixing. For this technique, we kept the injection volume smaller than usual, with an upper limit of about 1.0 mL per screw; however, this limit was not absolute. Cement injection was performed slowly under continuous fluoroscopy and was immediately stopped in case of suspected extravasation. Because the screw tips were close to the posterior vertebral wall, special attention was paid to avoiding leakage into the basivertebral vein region (Fig. 3E). The caudal PSs were treated in the same manner.

The collapsed L1 vertebral body was filled with hydroxyapatite and instrumented with short PSs, and rods were connected after cement polymerization (Fig. 1B). Postoperatively, paraparesis and leg pain improved rapidly, and the patient could walk independently at discharge. Postoperative MRI confirmed adequate decompression of the spinal cord (Fig. 1C). Radiographs obtained 1 year after surgery showed good sagittal alignment with the cement mass confined within the vertebral body and no posterior leakage (Fig. 1D).

Case 2

A patient with L2 osteoporotic collapse presenting with paraparesis was treated with L2 decompression and short-segment posterior fixation from L1 to L3 using the disc-plug and cement-augmentation technique described above (Fig. 4A). Because the distance between the posterior wall of L3 and the caudal screw tip was very short, only a small amount of cement was injected at L3 with meticulous control of injection rate and viscosity (Fig. 4B). This patient also had a favorable clinical course, with independent ambulation regained and sagittal alignment maintained at the 6-month follow-up (Fig. 4C, D).

Fig. 4

(A) Sagittal magnetic resonance imaging (MRI) showing L2 vertebral body collapse with a retropulsed fragment compressing the conus medullaris. (B) Lateral radiograph/schematic showing L1–L3 short-segment fixation in case 2, with limited cement injection at L3 because of the short distance to the posterior wall. (C) Postoperative MRI confirming sufficient decompression of the spinal cord. (D) Plain radiographs 6 months after surgery showing maintained sagittal alignment.

Discussion

Many patients with OVFs, where a retropulsed posterior wall fragment protrudes into the spinal canal and directly compresses neural tissue, or where instability of the fractured vertebra leads to neurological deficits, are elderly and have severe osteoporosis [1,2]. Consequently, the optimal surgical approach in this population remains a matter of debate. Anterior surgery is often difficult because of the patient’s age and poor bone quality; therefore, posterior indirect decompression and posterior fusion are generally preferred. However, in such cases, because of a kyphotic deformity at the fracture level, decompression alone is frequently insufficient to improve neurological symptoms, necessitating posterior fixation [6]. Although short-segment fixation is preferred to preserve motion segments, PS pullout remains a major concern in osteoporotic vertebrae.

PES fixation or pedicular transvertebral screw fixation was first described by Abdu et al. [3]. It has since evolved into the single or double-endplate penetrating screw (SEPS/DEPS) technique, in which the screw penetrates the endplates of adjacent vertebrae [7]. This technique involves intentional puncture of the intervertebral disc, and therefore long-term adjacent segment degeneration and instability are potential concerns. However, from a biomechanical standpoint, disc violation and local cement augmentation may increase stiffness at the treated level and shift motion and load to adjacent segments, potentially predisposing to adjacent segments to degeneration or junctional fractures, especially within a short fusion construct. As summarized in the Table 1, previous reports have shown acceptable short- to mid-term outcomes [5,7–9]; however, at present, this technique should probably be limited to cases in which conventional PS fixation is expected to fail, such as very elderly patients with osteoporotic vertebrae in whom intraoperative loosening is anticipated. A previous report noted that at the 1-year follow-up, cases treated with the one-above/one-below short fusion using PESs without cement for OVFs showed radiolucent zones around screws (n=3/11; 27.3%), screw loosening or nonunion (n=1/11; 9.1%), and adjacent vertebral fracture (n=1/11; 9.1%) [5]. Thus, even when PESs were used, problems such as screw loosening and pullout were not completely resolved.

Table 1

Fixation using endplate-penetrating screws: summary of four studies

Cement augmentation through fenestrated PSs is a widely used technique to increase pullout strength in osteoporotic vertebrae [4,10]. However, when PMMA is simply injected through a PES, the cement can leak from the screw tip into the intervertebral disc space, potentially increasing the risk of adjacent vertebral fractures and accelerating disc degeneration [11]. In addition, once cement has leaked into the disc, it becomes difficult to create a sufficient cement mass within the vertebral body, and the expected improvement in PS fixation may not be achieved.

To address these issues, we used to intentionally draw the disc tissue into the hollow PS tip to act as a “plug” before PMMA injection. This maneuver helped prevent intradiscal cement leakage from the screw tip while allowing cement to distribute from the side holes of the PS into the cancellous bone of the vertebral body. As a result, the contact area between the vertebral body and the PS was increased, and stronger fixation could be obtained. In both illustrative cases, the cement was injected in a highly viscous state under continuous fluoroscopic monitoring, and injection was immediately stopped when the cement front approached the posterior wall, which may have ensured the absence of clinically relevant leakage [4,10].

Nevertheless, great caution is always required regarding cement leakage outside the vertebral body. In particular, when the distance between the screw tip and the posterior vertebral wall is small, cement leakage in this region is difficult to detect on lateral fluoroscopy. In such cases, the injection volume should be kept to a minimum as a safety measure, as shown in Case 2.

This report has several limitations. First, there were only two patients, with a short follow-up and no control group; hence, we could not provide definitive data on long-term safety, durability, or superiority over conventional cement-augmented screws. Second, no biomechanical testing or animal model was used to validate the behavior of the disc-tissue “plug,” the cement distribution, and its effectiveness in markedly degenerated or nearly empty discs. Intraoperatively, resistance faced by the guidewire at the screw tip suggested a mechanical barrier of disc and endplate tissue (Fig. 3D), but this cannot substitute for formal biomechanical evidence. Third, the long-term effects of intentional disc puncture and localized cement augmentation in terms of adjacent disc degeneration, fracture risk, and implant stress within a short-fusion construct are unknown. Notably, the two cases presented here, as well as previous case series reporting the short- to mid-term outcomes of cementless endplate-penetrating short fusions, have not shown screw breakage or catastrophic construct failure. Accordingly, this technique is not presented as a superior method, but as a preliminary salvage option when PES alone are judged insufficient and conventional cement injection would likely cause intradiscal leakage. Wider clinical application of this technique can only be recommended after adequate biomechanical and longer-term clinical studies have been conducted using larger patient cohorts.

Article information Continued