Osteoporotic vertebral compression fractures (OVCFs), one of the most common and severe complications of osteoporosis, are becoming more common among the aging population. Vertebral augmentation was first applied by American doctors in patients with OVCFs in 1993 and obtained satisfactory efficacy. Since then, vertebral augmentation techniques, including percutaneous vertebroplasty and posterior kyphoplasty, have been widely used in the treatment of OVCFs and have received international recognition. Several studies show that vertebral augmentation improves clinical outcomes because it can achieve instant pain relief and stabilization of the fractured vertebral body (1–4). Nevertheless, several reports have indicated that vertebral augmentation may negatively affect adjacent segments (5, 6). Previous studies have mainly focused on the influence of vertebral augmentation on the adjacent vertebral body, as cement injection has been reported to increase the risk of adjacent vertebral fractures. Recently, the impact of vertebral augmentation on adjacent intervertebral discs, which has rarely been discussed, has been increasingly concerning. However, several studies that have attempted to determine whether vertebral augmentation can aggravate the degeneration of adjacent intervertebral discs have reached contradictory conclusions. Qian et al. (7) first conducted a prospective study and discovered a significantly higher incidence of adjacent disc degeneration above the fractured vertebrae in the cement augmentation group than in the control group. These findings were confirmed by retrospective studies by Lu et al. (8) and Wang et al. (9). However, a long-term follow-up study by König et al. (10), come to a different conclusion that vertebral cement augmentation through kyphoplasty had no significant influence on disc degeneration.

Cement distribution pattern is reportedly related to the occurrence of adjacent vertebrae fracture (11–14). These studies indicated that the cement distribution pattern may have an impact on the regional biomechanical stress. Moreover, cement distributed along the endplates can interfere with the endplate nutritional pathway and induce intervertebral disc degeneration in porcine and rabbit models (15, 16). Thus, the distribution of cement, which might have an impact on adjacent discs by altering both the regional biomechanical stress and the endplate nutritional pathway, should not be ignored when discussing the influence of vertebral augmentation on adjacent disc degeneration. Neglecting cement distribution patterns can lead to biased conclusions. However, to our knowledge, the influence of the cement distribution pattern on adjacent intervertebral discs has never been discussed. This study aimed to evaluate the effect of bone cement distribution on adjacent disc degeneration after vertebral augmentation in osteoporotic vertebral compression fractures.

A total of 184 patients including 51males and 133 females were finally enrolled in this study. The mean follow-up period was 28.6 months (ranging from 24 to 49 months). There were 73 patients in group A, 62 patients in group B, 28 patients in group C, and 21 patients in group D. No significant differences were found among the groups with respect to age, sex ratio, BMI, number of smokers, fracture segment, or follow-up time (Table 1).

At the final follow-up, 102 (55.4%) patients were found to have degeneration of the adjacent disc, of whom 48 (65.8%) were in group A, 32 (51.6%) in group B, 10 (35.7%) in group C, and 12 (57.1%) in group D. Forty-eight (26.1%) patients had control disc degeneration, of whom 22 (30.1%) were in group A, 15 (24.2%) in group B, 6 (21.4%) in group C, and 5 (23.8%) in group D. The number of DDC cases in the adjacent discs was significantly higher than in the control discs in groups A (p < 0.001), B (p = 0.002), and D (p = 0.028), whereas the difference in group C was not statistically significant (p = 0.237). The number of DDC cases in the control discs was not significantly different among the four groups (p = 0.775). However, the number of DDC cases in the adjacent discs, was significantly different among the four groups (p = 0.047). The number of DDC cases was significantly higher in group A than that in group C (p = 0.06). The difference in the DDC number in the adjacent discs between groups A and B (p = 0.096), groups A and D (p = 0.469), groups B and C (p = 0.162), groups B and D (p = 0.661), and groups C and D (p = 0.136) was not statistically significant (Table 2).

The adjacent disc heights decreased significantly in groups A, B, and D (p < 0.001, p < 0.001, and p = 0.012, respectively), but did not decrease significantly in group C (p = 0.079). The follow-up control disc heights were not significantly different from the preoperative control disc heights in all of the four groups (p = 0.089, p = 0.138, p = 0.269, and p = 0.301, respectively). No statistical difference was detected among the four groups with respect to the preoperative adjacent disc height, follow-up adjacent disc height, preoperative control disc height, nor follow-up control disc height (p = 0.346, p = 0.519, p = 0.496, p = 0.218, respectively) (Table 3).

The advantages of the VCA include its minimally invasive nature, good biomechanical support, and satisfactory clinical outcomes. Regarding the adverse effects of vertebral augmentation, researchers have mainly focused on intraoperative complications, such as cement extravasation, and the postoperative negative effects of cement injection on adjacent vertebral bodies (6, 19). The influence of vertebral augmentation on adjacent intervertebral discs has rarely been discussed in previous studies. A prospective study of 97 patients conducted by Qian et al. (7) compared the grade of disc degeneration above the fractured vertebra on MR between the PVP t and conservative treatment groups. They discovered that the incidence of degeneration of the adjacent intervertebral disc above the fractured vertebra was significantly higher in the vertebral augmentation group than in the control group (52.5% vs. 29.0%). Similar results were indicated in retrospective studies conducted by Pachowsky et al. (20), Lu et al. (8), and Wang et al. (9). However, a long-term follow-up study by König et al. (10) found that vertebral cement augmentation using PKP had no significant effect on disc degeneration. As König's study included discs with preoperative Pfirrmann grade ≥3, the endplate nutrition could have already been affected and could cause bias in evaluating the actual effect of cement augmentation. Li et al. (17) suggested that adjacent disc degeneration might be attributable to the endplate injury of the fractured vertebrae, which was ignored in the above studies. The design defects of the aforementioned studies might be the reason for the discordance in the results. Furthermore, previous studies did not consider one key factor—cement distribution—that might have an impact on adjacent discs by altering both the nutrient supply to the discs and local biomechanics. To determine the effect of bone cement distribution on adjacent intervertebral disc degeneration after vertebral augmentation, we used strict inclusion criteria to reduce potential bias. In our study, we included patients only with upper endplate injuries and the Pfirrmann score of the adjacent discs ≤2. Additionally, endplate injuries are often accompanied by disc injuries, which have been proven to have an important impact on disc degeneration (21) and could therefore influence the evaluation of the actual effect of cement augmentation. Therefore, the influence of disc injuries should be considered. Therefore, patients with severe adjacent disc injury were excluded, and the DDC definition excluded deterioration from grade 1 to grade 2 Pfirrmann scores to further diminish the bias induced by the potentially inaccurate evaluation of the preoperative Pfirrmann score of injured discs on MR.

Several patient characteristics, such as age, sex, body weight, and smoking, have been reported as risk factors for intervertebral disc degeneration (22, 23). In this study, these risk factors were evaluated to improve the reliability of our conclusions. None of these characteristics showed statistically significant difference among the four groups. In the present study, the incidence of degeneration in control discs was not significantly different among the four groups. The number of DDC cases in adjacent discs was significantly higher than that in control discs in groups A, B, and D, but the difference between adjacent and control discs in group C was not statistically significant. These results suggested less severe degenerative changes in the adjacent discs in group C. Further analysis showed a significant difference in the incidence of adjacent disc degeneration among groups A, B, C, and D. Compared with group C, patients in group A had a significantly higher number of DDC cases. Patients in groups B and D also had a higher number of DDC cases than those in group C, although the difference was not statistically significant. These results indicate that vertebral augmentation has an impact on adjacent disc degeneration, which is largely influenced by bone cement distribution. Cement distribution proximal to the injured endplate could accelerate the adjacent disc degeneration, and cement in contact with both the cranial and distal endplates might have a greater adverse impact on adjacent discs. However, cement distributed distally to the injured endplate may have no significant influence on disc degeneration. The alteration in the adjacent disc height further supports this conclusion. The adjacent disc heights decreased significantly at follow-up compared with baseline in groups A, B, and D, whereas the decrease in group C was not statistically significant. However, the control discs, did not exhibit a significant decrease in disc heights. These results were consistent with those of previous reports (9, 17), and further proved the fewer degenerative changes in the adjacent discs in group C. However, the adjacent disc heights at follow-up were not significantly different among the four groups. This might be attributed to the comparatively short follow-up time, as the adjacent disc degeneration was still in the early stages and the decrease in disc height was minor. Similarly, in a previous report with comparable follow-up time (2 years), the height loss of adjacent disc in the vertebral augmentation group was not significantly different from that in the control group (7). Disc height is not strongly related to Pfirrmann scores from grade 1 to 4, and only significantly differentiate between grade 4 and 5 Pfirrmann scores (24, 25). However, no disc degenerated to grade 5 at final follow-up in this study, that might make the disc height change less significant. Significant disc height loss and high-grade disc degeneration were observed in some studies with long-term follow-up (ranging from 8 to 17 years) (26–28). Thus, prolonging the follow-up time may be helpful to observe the differences of disc height change among the four groups.

However, the pathogenesis of disc degeneration remains unclear. Nevertheless, alteration of local biomechanics and impairment of nutrient supply have been suggested to accelerate adjacent disc degeneration after vertebral augmentation according to previous reports. Biomechanical studies have shown significant decreases in adjacent disc pressure after vertebral fractures (29–31). A finite-element study by Luo et al. (29) reported that vertebral injury reduced the intradiscal pressure at the affected level to an average of 47% of the pre-fracture values. These results are consistent with those of previous studies (30, 31). Bone cement augmentation can enhance strain on the intervertebral disc by increasing the elastic modulus of the injured endplate and further altering the local biomechanical profile. A finite element study by Polikeit et al. (32) showed that the maximum pressure increased by 16% above the treated level and by 13% below the treated level. Ananthakrishnan et al. (30) also reported an increased nuclear pressure after augmentation treatment. Similar results have been reported in other finite-element studies (31, 33, 34). Thus, cement distribution closer to the injured cranial endplate in groups A, B, and D may have made the endplates stiffer and induced a higher incidence of adjacent disc degeneration. Bone cement distributed distal to the injured endplate, as shown in group C, would accordingly exert less stress on the adjacent disc and therefore induce less disc degeneration. Additionally, clinical studies have suggested a correlation between disc degeneration and increased adjacent bone density (35, 36). Previous studies also reported that cement distribution close to the endplate is associated with a higher risk of adjacent vertebral fractures (13, 37). Sun et al. (13) reported that cement distribution proximal to the adjacent vertebral body (distributed at the disc level or endplate level) could increase the risk of post- vertebroplasty fractures in adjacent vertebral bodies. Similarly, Zhao et al. (37) found that the diffusion of bone cement into cracks in the injured endplate may increase the incidence of adjacent vertebral fractures. These results further explain the differences in the numbers of DDC cases induced by different cement distribution patterns. Furthermore, bone cement in group A contacted both the cranial and distal endplates could induce to higher elastic modulus as there were no “buffering bones”, which might exert greater stress on peripheral structures and thus induce to greater disc degeneration.

Impairment of the nutrient supply to the disc has been identified as another important factor affecting adjacent disc degeneration. As the largest inert structure, the intervertebral disc is largely dependent on endplate diffusion for nutrient supply (38), and can be impaired by cement augmentation by damaging the capillaries underneath the endplate. Signs of adjacent disc degeneration followed vertebral augmentation were reported (15, 16). However, conclusions from adult animal models (goats, sheep, and dogs) are contradictory (39–41). Because the vertebrae in these animal models were intact, these conclusions may not be applicable to fractured vertebrae. Our study indicated that a combination of vertebral fracture and cement distribution proximal to the injured endplate may be required to influence the nutrient supply and accelerate the degeneration, as the incidence of disc degeneration increased in groups A, B, and D, but not in group C.

This study had several limitations. First, the strict inclusion and exclusion criteria may have led to sampling bias and comparatively small sample sizes in the subgroups. Second, the distribution of bone cement was only evaluated on x-ray films, owing to the lack of postoperative CT images. Third, the severity of endplate injuries was not evaluated due to the lack of a proper rating system, which may have resulted in biased results. Fourth, the average follow-up time was relatively short, which may not have been sufficient to fully evaluate the degenerative condition of the adjacent discs. As no disc degenerated to grade 5 in this study, it might make the disc height change less significant.

The present study suggests that bone cement distribution influences adjacent disc degeneration after vertebral augmentation in osteoporotic vertebral compression fractures. Cement distribution proximal to the injured endplate can accelerate adjacent disc degeneration, and cement in contact with both the cranial and distal endplates can induce a higher incidence of adjacent disc degeneration.