Background

The spine is the most common anatomic site for osteoporotic fractures. The incidence of osteoporotic vertebral compression fractures in the aging population is increasing worldwide and becoming a major healthcare issue [1, 2]. Unlike peripheral fractures, osteoporotic vertebral fractures (OVF) often cause nonspecific symptoms and the risk of further fractures increases by a factor of up to 7.3 [3,4,5]. Symptomatic OVFs were previously treated conservatively to reduce pain through immobilization [6]. Now, however, percutaneous vertebroplasty (PVP) is widely used as a minimally invasive procedure to treat pain from vertebral fractures with different pathologies—osteoporotic, traumatic, or neoplastic [7,8,9,10,11]. Many studies [12, 13] have addressed the indications for PVP, its success, and its complications, but rarely within the geriatric subgroup and considering other factors such as cardiovascular comorbidities.

The aim of this study was to evaluate the safety and efficacy of PVP in geriatric patients with cardiovascular comorbidities at our center.

Methods

Patient selection and inclusion criteria

In this retrospective single-center cohort study, we analyzed data on all patients with OVF aged ≥ 65 years who were surgically treated with PVP at our spine center between January 2017 and December 2020. Patients’ clinical information was registered and documented, including age, sex, BMI, duration of surgery, ASA score, fracture location, number of fractured vertebrae, kyphosis angle, history of cardiovascular comorbidities, surgery-related complications, and in-hospital complications.

The inclusion criteria were subacute osteoporotic fracture. The indication for PVP was increasing treatment-refractory pain after osteoporotic fracture. The diagnosis of OVF was performed according to the Osteoporosis International algorithm (2020) [14]. Radiological imaging [15] of the spine was undertaken before surgery with CT and MRI with short tau inversion recovery (STIR) sequences for edema at the vertebra (Fig. 1).

Fig. 1
figure 1

Indication for surgery. 79 y/o female with osteoporotic L1 vertebral fracture. A MRI STIR sagittal sequence, B T2-sagittal sequence, C CT scan, and D axial sequence of L1 without spinal stenosis

Full size image

Postoperative surgery-related and in-hospital complications were defined as any adverse events occurring within 30 days of the initial surgery.

Patients underwent standardized preoperative clinical, MRI, and CT examinations. VAS scores and the kyphosis angle were recorded both pre- and post-operatively. A CT scan was routinely performed after the PVP surgery. Clinical follow-up examinations were undertaken 1 month after surgery. For more comprehensive clinical and radiological data, we evaluated the EQ-5D, a measure of health-related quality of life from the EuroQol Group, and the compression index of fractured vertebra (using the Orbis® OpenMed health information system) [16, 17].

Patients with tumor fractures, other pathologies such as traumatic fractures, and acute OVF less than 3 weeks old were excluded; we also excluded all patients whose history of comorbidities was unknown.

Sugita et al. [18] described a prognostic classification system for osteoporotic vertebral fractures with 5 subtypes. Our cohort with fracture subtypes 2–4 was treated with PVP, and patients with subtype 5 were treated with augmented dorsal fixation with screws and dorsal decompression because of spinal stenosis und a high kyphosis angle.

The study was approved by the local ethics committee of our institution (reference no. 067/21). Patient approval to publish data is not required, because these data are collected as standard in our clinic, and therefore consent for retrospective data is not necessary. This retrospective data analysis complies with the applicable local and international guidelines for data protection and carrying out scientific studies. This manuscript does not contain any patient-sensitive data or images that require approval for publication.

This cohort study has been reported in line with the PROCESS Guideline [19].

Surgical procedure

The surgical procedure generally involved anesthesia, and all the procedures followed the same process. Patients were placed in a prone position. We routinely used a radiolucent table with a gel mattress. A posterior surgical approach was taken at the level of the segments to be operated on. Prior to the skull incision, an X-ray was performed to verify the region of interest by 2D fluoroscopy. Jamshidi needles for later cement application were inserted by using the fluoroscopy in anterior posterior and lateral projections. After the pedicle was perforated, a contrast medium was injected to confirm that the pedicle had not been breached und to verify the correct localization of the Jamshidi needles. Finally, the cement was applied (bi-pedicular approach). The volume of cement injected was usually around 1.5–2 mL/side, but this was not mandatory. To eliminate as much as possible bias due to the skill or experience of the surgeon, operations were performed by only three neurosurgeons in our department.

Postoperative management

Patients received early postoperative mobilization with physiotherapy. A postoperative CT scan was conducted on the day of surgery [20].

Statistics

All data were analyzed using IBM® SPSS® Statistics V22.0 (IBM, Chicago, Illinois, USA). Quantitative, normally distributed data are presented as mean values ± standard deviation (SD), while non-parametric data are summarized by median values [first quartile–third quartile]. In the case of categorical variables, data are given as numbers and percentages. After normality testing via the Shapiro–Wilk test, continuous normally distributed data were compared using t-tests, while the Mann–Whitney U test was used for non-parametric data. Nominal data were tested between groups using Fisher’s exact test and in case of multinomial data with a Chi-square test. Correlation analysis was carried out using Spearman’s correlation coefficient. The Wilcoxon test was used to compare VAS profiles. p values < 0.05 were considered statistically significant.

Results

Patient characteristics

A total of 49 patients aged ≥ 65 years with a total of 88 osteoporotic vertebral fractures were surgically treated with PVP at our center. Table 1 shows the baseline data.

Table 1 Baseline data
Full size table

Postoperative complications

The most relevant postoperative complication was cement extravasation (intra- or para-spinal) with no new neurological postoperative deficits, which occurred in seven out of 49 patients (14%) (Fig. 2). There was one case of pulmonary cement embolism (2%).

Fig. 2
figure 2

Postoperative CT scan showing complications consisting of cement extravasation (intraspinal) without neurological deficit

Full size image

Further postoperative in-hospital adverse events consisted of urinary tract infections (6%) and pneumonia (2%). One patient died 2 days after the operation due to a myocardial infarction with cardiopulmonary resuscitation (Table 1).

Patient-related factors influencing postoperative complications

Influence of clinical admission status

Patients suffering from postoperative complications following PVP exhibited similar values for ASA, BMI, and age; there were no significant variations. Female patients exhibited postoperative complications significantly more often than males: while 7 out of 27 female patients suffered from adverse postoperative events, there were no complications in the male group (Table 2).

Table 2 Patient-related factors influencing complications after PV
Full size table

Influence of patients’ comorbidities, duration of surgery, level/localization of OVF, and length of stay

Patients with arterial hypertension, coronary artery disease, atrial fibrillation, diabetes mellitus, or a history of pulmonary embolism or deep vein thrombosis were not significantly affected by postoperative complications (Table 2). Furthermore, patients with more than one OVF (Fig. 3) did not exhibit elevated levels or localization of postoperative complications compared to patients with a single OVF (p = 0.834) (Table 2). There was no significant difference in the duration of surgery, or length of stay.

Fig. 3
figure 3

Vertebroplasty of thoracic vertebral fracture (T12), OF type 1. A–C CT scan before surgery (sagittal, axial and coronal), D–F after surgery

Full size image

Effect of VAS score, EQ-5D, kyphosis angle, and compression index

The patients in our group recorded a significant improvement of at least 97% in their VAS scores, and this improvement was maintained at the 4 weeks clinical follow-up; additionally, the patients reported a better quality of life. Seven patients were lost to follow-up.

A moderate kyphosis angle reduction of at least one degree was also seen in the post-surgery CT scans. After vertebroplasty, the compression index improved by 3–10% (Fig. 4).

Fig. 4
figure 4

A, B Sagittal CT and MRI of osteoporotic thoracic vertebral fractures (T4 and T6); T4: OF type 3, T6: OF type 2. C–E CT scan (sagittal and axial) after PVP of both vertebral fractures without complications

Full size image

Discussion

Osteoporosis is one of the most common diseases in geriatric patients and the number of osteoporotic vertebral fractures is increasing [4, 5]. This fact confronts the treating surgeons with a medical and moral question as to how far vertebroplasty can be offered as a treatment option [6, 7]. Two studies with randomized controlled trials compared the pain-relief benefit of PVP and non-surgical management. These two studies, by Buchbinder et al. [12] and Kallmes et al. [13], came to opposing conclusions—Buchbinder found no benefit of PVP, whereas Kallmes described a trend toward a higher rate of clinically meaningful improvement in pain after PVP.

This study evaluates the usefulness of vertebroplasty in geriatric patients and in this cohort focuses specifically on postoperative complications and the influence of comorbidities on these complications.

Cement leakage is known to be the most common complication following PVP and has been reported in about 30% to 65% of patients with osteoporotic vertebral fractures [21, 22]. This finding has important clinical implications. Despite generally showing no clinical symptoms, there seems to be a considerable incidence of cement extravasation that can occasionally result in severe complications. Chew et al. reported rates of severe complications ranging from 2% to 11.5% resulting from cement leakage [23]. Cement leakage or extravasation is defined as the presence of extra-vertebral cement. Cement can leak into the spinal canal, the neural foramina, or another location in the body, leading to neurological complications such as paraplegia or root compression [24], and can also leak into the pulmonary arteries, causing a pulmonary embolism [25, 26]. We found reports in the literature of these complications ranging from 9% to 59.6% [21, 25, 27, 28]. At 14%, the cement leakage rates in our study are at the lower end of the reported rates. The rate of cement leakage and the low prevalence of clinically significant symptoms or complications in our series are probably due to technical and organizational reasons at our center. On the technical side, we injected a high-viscosity cement and applied as much cement volume as bone and resistance allowed. On the organizational side, we believe that, in addition to strict selection of suitable patients, the standardized approach to vertebroplasty by our experienced neurosurgeon is significant. Varied results are reported in the literature on the influence of the volume of cement injected on the clinical outcome. In the majority of these studies, the volume of cement is not demonstrably related to the quality of postoperative pain reduction, but to an increase in the rate of complications [29].

Several studies with different patient cohorts reported on the benefits of PVP in relation to VAS scores after surgery with low complication rates. Although these studies differed in their structure, including both prospective and retrospective cohorts, their results were along the lines of our data [30,31,32,33,34]. The change of kyphosis angle and the height of the affected vertebral body (compression index) did not appear to have brought about a significant improvement in clinical symptoms [35]. An injection of < 4 mL per vertebral body was sufficient in our patients. Taking into account the result of other studies mentioned above and the clinically high reduction of pain in our study, the authors consider a low volume of cement to be both safe and effective.

Given the high correlation between age, sex, and the distribution of OVF, it is not surprising that there are a number of studies which express the view that OVF are restricted to the female sex [1, 2, 36]. This was reflected in our cohort, which also had a higher mean age than usually described in the literature [33, 37]. We believe that our relatively low complication rates, despite our higher age range, suggest that it is possible to offer vertebroplasty for appropriate indications without increased risk, even at a more advanced age.

Except for sex, we found no factors that specifically correlated with the occurrence of postoperative complications—not even weight or cardiovascular comorbidities.

Our recommendation meets the criteria for a level of evidence 3, based on the Oxford Centre for Evidence-Based Medicine 2011 Levels of Evidence.

Limitations

We were able to achieve an acceptable degree of kyphosis angle correction, although we cannot judge on the basis of the short-term follow-up and the small patient cohort to what extent this had an influence on the overall result. This represents one of the main limitations of this retrospective study.

Conclusions

Our data demonstrate that vertebroplasty was a feasible and effective treatment for pain reduction in geriatric patients with osteoporotic fractures despite cardiovascular comorbidities. This treatment option offered rapid and significant pain reduction, with a reasonable level of risk.