Osteoporotic vertebral fractures are often effectively managed using conservative methods. However, complications such as nonunion, neurologic deficits, and persistent high-grade kyphosis can occur in some cases [1]. Vertebroplasty and kyphoplasty have been reported to be effective and safe treatment options for patients with osteoporotic vertebral fractures [2]. These methods are less invasive and generally associated with fewer complications, but adjacent vertebral fractures (AVFs) remain a commonly observed issue [3].

Meta-analyses of randomized controlled trials (RCTs) have shown no significant difference in the incidence of AVFs following vertebroplasty or kyphoplasty as compared with conservative treatment [4–7]. However, several retrospective studies reported a higher incidence of AVFs after these procedures [8,9]. Although various risk factors for AVF have been identified, the reasons for the inconsistency in the reported incidence of AVF remain unclear. However, patient selection is one possible explanation. In many retrospective studies, vertebroplasty and kyphoplasty were performed on patients for whom conservative treatment had failed [9]. The condition of the fractured vertebrae—including local kyphosis, vertebral instability, and endplate damage—is a known risk factor for AVF and may deteriorate over time following injury [10,11]. Consequently, the high incidence of AVFs reported in retrospective studies might be related to the delayed intervention and deterioration of the vertebral condition as a result of initial conservative management. We hypothesized that early surgical intervention could reduce the incidence of AVFs. To investigate this, we compared the AVF incidence rates in patients who underwent balloon kyphoplasty (BKP) within 30 days of onset of fracture to those treated more than 30 days after the fracture.

The Ethics Committee of Shimonoseki City Hospital approved this retrospective study (approval no., 2024SCHEC-37). Because of the study’s retrospective nature, an opt out approach was used to obtain informed consent, in accordance with the ethical guidelines of Shimonoseki City Hospital. All participants were informed of their right to opt out of the study. Between 2012 and 2023, 230 patients at our institution underwent BKP for osteoporotic vertebral fractures, excluding cases of pathologic fractures. There were no cases in which BKP alone was performed on patients with preexisting neurologic symptoms. In addition, among patients who underwent BKP, no new neurologic symptoms emerged postoperatively. We excluded 41 patients who were followed for less than 3 months, leaving 189 patients for retrospective analysis. BKP was performed under general anesthesia, with patients typically mobilizing the day after the procedure while wearing a soft corset. The duration of corset use varied based on the patient’s condition. The brace-wearing period was generally set to 3 months after surgery. We assessed the incidence of AVF at 3 months post-BKP. We identified new AVFs based on sitting-to-supine instability, a 15% loss of height, or changes in bone signal intensity on magnetic resonance imaging (MRI).

The analyzed risk factors included age, sex, fracture site, history of previous vertebral fractures, bone mineral density, kyphosis angle of the fractured vertebra, endplate damage, signal changes in the fractured vertebral body on MRI, vertebral instability, and time from vertebral fracture to BKP. Bone mineral density was measured using dual-energy X-ray absorptiometry scans of the hip. We calculated the kyphosis angle of the fractured vertebra from the endplate angles with the patient in the sitting or standing positions. We determined vertebral instability as the difference in the kyphosis angle between the sitting and supine positions (Fig. 1A). The corrected angle was defined as the change in the kyphosis angle before and after BKP in the sitting position (Fig. 1B). Computed tomography (CT) was used to confirm endplate damage. CT performed 1 week after BKP confirmed cement leakage into the disc space. We categorized MRI signal changes into two groups based on the presence or absence of T2-high signal changes in the fractured vertebral body [12]. We defined the date of injury as the first day the patient-reported pain, regardless of when the fracture was confirmed by imaging. For group comparisons, patients without pain awareness within 30 days and with unclear pain onset were categorized as having an injury more than 30 days prior.

Table 1 summarizes the preoperative demographic data. The study included 48 men and 141 women, with a mean age of 81±6 years. The fracture sites were at the thoracolumbar junction (T12 or L1) in 129 patients and elsewhere in 60 patients. The average bone mineral density was 0.66±0.14 g/cm², local kyphosis was 17°±8°, and vertebral instability was 10°±5°. The average time from fracture to intervention was 70±57 days; the exact date of injury was unclear in 25 cases. At 3 months post-BKP, 55 patients had AVFs (29% incidence).

Table 2 compares the demographics of patients treated within 30 days versus more than 30 days after onset of fracture. Patients aged 80 years were more likely to undergo early BKP (35/44 [80%] versus 80/145 [55%], p=0.005). Fewer patients with vertebral clefts received early BKP (8/44 [18%] versus 85/145 [59%], p<0.001). Table 3 outlines the outcomes of BKP and incidence of AVF, showing no significant difference between the early and delayed BKP groups (9/44 [21%] versus 46/145 [32%], p=0.15).

The results of the univariate analysis showed significant differences in vertebral instability (p=0.046) and endplate damage (p=0.01) but not in the timing of the intervention (p=0.15) (Table 4). In the multivariate analysis, we observed no significant difference in the timing of intervention, with an odds ratio of 1.65 (95% confidence interval [CI], 0.66 to 4.10) for patients treated more than 30 days after the fracture. However, we observed a significant difference for vertebral instability, for which an odds ratio of 2.73 (95% CI, 1.25–5.97) was noted for instability of 10° or greater (Table 5).

The aim of this study was to evaluate whether performing BKP within 30 days of vertebral fracture reduces the incidence of AVFs. Our findings show that early BKP does not significantly lower the rates of AVF. Instead, the occurrence of AVFs appears to be influenced primarily by the instability of the fractured vertebral body rather than the timing of the intervention.

In our previous retrospective study, we found that the incidence of AVF was greater in the BKP group than in the conservative treatment group, with most AVFs occurring within 2 months after BKP. The incidence rate of AVF at 1 year was 23.2% for BKP and 7.3% for conservative treatment [9], with the latter aligning closely with the results of a prior cohort study [13]. In contrast, meta-analyses of RCTs reported an AVF rate of approximately 8% after BKP [4–7]. Notably, in our earlier study, the patients who received BKP were those who had not improved with conservative treatment, and none had undergone BKP within 30 days of their initial injury. Thus, differences in patient backgrounds may account for the variations in the incidence of AVF compared with RCTs. Based on the hypothesis that early BKP might reduce the occurrence of AVF, we conducted the present study at a different institution. Our results indicate that performing BKP early after injury does not prevent AVF, and the occurrence of AVF is primarily influenced by the instability of the fractured vertebral body.

One report indicated that BKP performed within 30 days of injury can suppress AVFs [14,15]. In that report, the authors also observed significant differences in vertebral instability between the early and late intervention groups (with the late group having large vertebral instability), which suggests the results may reflect variations in vertebral instability. In contrast, in our investigation, we found almost no difference in vertebral instability between the early and late groups regarding patient backgrounds. At our institution, early BKP is indicated only for cases with severe pain or identifiable poor prognostic factors on MRI (diffuse low- or high-intensity area in fractured vertebrae on T2-weighted images) [16,17]. This could explain the observed lack of difference in instability between the two groups, as patients with poor prognostic factors often exhibit greater vertebral instability [18]. Furthermore, the higher proportion of elderly patients in the early intervention group may have contributed to these findings. We also believe that the relatively high incidence of AVFs in this study was influenced by these factors.

Vertebral instability can intensify over time [18], and early intervention may reduce the AVF risk if performed before the instability worsens. Our study suggests that early intervention itself does not suppress AVFs. With regard to multivariate analysis, because there may be a correlation between instability and the timing of surgery, we also conducted analyses excluding each factor, but the results remained consistent, with only instability showing a significant difference (data not shown). Vertebral instability was not measured in the RCTs, and it is unclear which patient groups were included; the differences in the incidence of AVFs were likely due to variations in patient backgrounds and the condition of the fractured vertebrae. We believe that AVFs are complications of BKP, and for cases with significant instability, combining BKP with fixation of the upper and lower vertebrae may be a potential solution. Although conservative treatment is the standard approach for osteoporotic vertebral fractures, recent reports have advocated for early vertebroplasty [14,15]. One argument for conducting early intervention is that it can help prevent AVFs and yield good outcomes. However, the findings of our report indicate that early intervention does not always lead to favorable outcomes, and we believe that it should be considered primarily for cases resistant to conservative treatment or vertebral fractures with poor prognostic factors.

One limitation of this study is its retrospective nature, which means there may be unmeasured factors, such as global alignment. Due to fracture pain, global sagittal balance is often difficult to measure, and even when it can be measured, its accuracy for acute vertebral fractures is questionable. In addition, postoperative therapies (e.g., for osteoporosis) and the duration of corset wear have not been consistent across patients. Finally, although we categorized the patients based on a 30-day threshold, we observed no significant difference, although we noted a trend. It is possible that a significant difference would emerge if the sample size were increased. In addition, if the timing of the intervention was shifted to an earlier window, such as within 14 days, the results could differ.

Our study suggests that early BKP does not significantly reduce the occurrence of AVFs. Instead, the primary factor influencing the development of AVF appears to be the instability of the fractured vertebral body rather than intervention timing. Given the complex relationship between instability and surgical timing, our findings underscore the importance of evaluating each patient’s individual condition—particularly the degree of vertebral instability—when determining the timing for BKP. For patients with significant vertebral instability, early intervention may still be considered to prevent further deterioration of the vertebral alignment. However, it should not be regarded as a definitive strategy for the prevention of AVFs. Clinical decision-making should include a comprehensive assessment of fracture instability, patient-reported pain levels, and other identifiable poor prognostic factors. Guidelines for managing cases with high instability should incorporate detailed risk assessments, including advanced imaging modalities to better evaluate vertebral alignment and stability. In addition, for patients with severe instability, combining BKP with supplemental stabilization measures, such as fixation of adjacent vertebrae, may be necessary to reduce the risk of AVFs. Future prospective studies with larger sample sizes, standardized assessment methods, and consideration of other potential confounding factors—such as global spinal alignment and osteoporosis management—are needed to refine the guidelines for BKP in the context of vertebral fractures with varying degrees of instability.

The findings of this study suggest that performing BKP early after an osteoporotic vertebral fracture does not significantly reduce the incidence of AVFs. The primary factor influencing the development of AVFs appears to be vertebral instability rather than the timing of the intervention. Therefore, in the management of osteoporotic vertebral fractures, the clinical focus should prioritize the assessment and management of vertebral instability over the timing of surgical intervention.