Previous studies have evaluated the implant complications and long-term kyphosis correction outcomes after the treatment of TLBFs with SS (single-level cephalad and caudal), IS (two levels cephalad and one-level caudal to the fracture level), and LS (two-level cephalad and caudal) posterior fixations. The advantages of the traditional posterior SS fixation include a small incision wound, short surgical duration, and early mobilization. However, a high implant complication rate in SS fixations has been widely reported, with an incidence of up to 57%. This may be due to untreated vertebral instability and poor resistance to the anterior compressive force caused by cephalad and caudal single-level pedicle screw fixations without further support [17–19]. The addition of one or two more intermediate pedicle screws at the level of the fracture has been found to lower the rates of progressive kyphosis and implant failure in traditional SS posterior fixations, and the similar postoperative spinal stability were presented between one or two more intermediate pedicle screws. [20, 21]. Zhang et al. found hybrid balloon kyphoplasty with SS posterior fixation to produce better clinical and radiological outcomes than SS posterior fixation with two additional intermediate pedicle screws [22]. Cho et al. found TLBF treatment with combined SS posterior fixation and PMMA vertebroplasty to provide immediate vertebral stability and prevented any incidence of posterior instrumentation failure at 2-year follow-ups [23].
Cement augmentation treatments, including vertebroplasty and kyphoplasty, alone or with posterior fixation, can be used in the treatment of burst fractures. A porcine biomechanical study found that the vertebral stiffness of burst fractures treated with vertebroplasty with PMMA increased from 47.5–107.8% postoperatively [24]. A cadaveric study showed reduced bending of pedicle screws, with a 59% decrease in flexion and a 38% decrease in extension. Initial stiffness in the flexion-extension movement increased by 40%, following SS posterior fixation with additional cement augmentation of the fractured vertebra [25]. A finite element study used cement augmentation with SS fixation for the treatment of TLBFs. The results showed a stiffer construct and lower von Mises stress on the pedicle screws and rods compared with other types of SS fixation, including one-level cephalad and caudal fixations without cement augmentation and one-level cephalad and caudal fixations with additional index level screw fixation with and without cement augmentation [12]. In this study, we performed balloon kyphoplasty or SJ implantation to correct spinal kyphosis, followed by injection of radiopaque low-viscosity PMMA bone cement. Kyphoplasty, which is distinct from vertebroplasty, facilitates the low-pressure injection of PMMA cement, reducing the incidence of cement leakage [26]. It also produces better immediate post-op kyphosis correction and less progressive kyphotic degeneration than vertebroplasty and nonsurgical treatments [27]. A delayed post-op vertebral collapse was seen in a 61-year-old male patient with an A4 type burst fracture at the L1 level treated with SpineJack kyphoplasty and low-viscosity PMMA (Fig. 1). This may have resulted from a dislodged pedicle screw and rod. Furthermore, a 21-year-old male patient with TLBFs at the L1 level was treated with balloon kyphoplasty, low-viscosity PMMA cement, and SS fixation. The patient presented with allodynia and muscle grade 4 in the right dermatome below level L2 due to cement leakage. There was right-sided lateral recess stenosis on the L1–2 level, which was treated with right-sided extensive laminectomy surgery on levels L1–2 to remove extravasated cement. The high incidence of cortical defects in TLBFs has been attributed to bone cement leakage [28]. Improvement in the components of conventional high-viscosity PMMA bone cements is necessary to prevent cement leakage [29].
IS and LS pedicle screw fixation techniques are used to lengthen the arm level of the implants and both can improve spinal stability. Moreover, they have lower implant complication rates than SS pedicle screw fixation techniques without anterior reconstructions in the treatment of TLBFs [30]. Finite studies have found high stiffness and low range of motion during extension, bending, and torsion in LS pedicle screw fixations compared with SS posterior fixations for the treatment of TLBFs [12, 31, 32]. IS posterior fixations for TLBFs have advantages over LS fixations, including greater resistance to anterior compressive forces and less progressive kyphosis due to the two cephalad level fixations. Moreover, the one-level caudal fixation in IS preserves lumbar motion [33, 34]. However, caudal one-level pedicle screw fixations may have poor resistance to compressive force acting anteriorly. This possibility was supported by our finding that this approach had the highest implant complication rate.
Few studies have assessed and compared the surgical risks, implant complication rate, and clinical efficacy of HSS, IS, and LS pedicle screw fixations in the treatment of TLBFs. In this study, HSS pedicle screw fixation required shorter surgical durations and resulted in less intraoperative blood loss than IS and LS posterior fixations. A previous meta-analysis found no significant differences in implant-related complications between short-, intermediate-, and LS posterior fixations [7]. However, in the present study, the HSS group had the lowest implant complication rate (13.33%), followed by the LS group (37.5%), and the IS group (58.33%). Aly found no significant differences in the postoperative kyphosis correction and progressive kyphosis outcomes between short and LS fixation groups [7]. In our study, the HSS pedicle screw fixation technique produced better immediate kyphosis correction outcomes and maintained the corrected alignment with the least progressive kyphosis at a mean follow-up of 1 year.
The current study had several limitations. First, patients with previous malignancy, spine surgery, and more than one-level fractured vertebra were excluded from the study. We only compared three different surgical techniques in this study, so the HSS technique was not compared to other surgical techniques, such as the one-level cephalad, two-level caudal posterior fixation, and short-segment fixation with additional index level screw fixation. Therefore, our sample was relatively small, reducing the validity and reliability of our results and preventing more specific comparisons between subsets. For example, most patients in all three groups had suffered mild neurological deficits with grade D on the American Spinal Injury Association Impairment Scale. Hence, we could not perform a statistical evaluation of the relationship between the severity of neurologic deficits and surgical blood loss and duration. Second, we only compared three different surgical techniques in this study, so the HSS technique was not compared to other surgical techniques, such as one-level cephalad, two-level caudal posterior fixation, and short-segment fixation with additional index level screw fixation. Third, some neurosurgeons found the HSS technique preferable in the treatment of single-level TLBFs, and the allocation of patients to undergo the three different surgical techniques was based on patient evaluations and neurosurgeon preference rather than sequential selection. Thus, sampling was not random due to the retrospective study design. Significant differences in surgical decisions, volume of blood loss, and surgical durations between the 13 neurosurgeons in this hospital occurred due to differences in their surgical techniques, experience, and aptitude. This may have introduced further bias. Fourth, we did not investigate differences in the efficacy of the HSS technique between TLBFs of different types or severity or between cases with different neurological deficits. Fifth, strict standing lateral thoracolumbar radiographic examinations were not performed, which led to a significant bias in the measurement of Cobb angles at different stages. Sixth, we only recorded and analyzed clinical radiological findings. Additional neurological evaluation tools, such as the visual analog scale and the Oswestry disability index, could have been used to better determine the surgical technique likely to produce the best radiological and functional outcome for each patient. Further research should address these limitations.