Therapeutic effect of percutaneous kyphoplasty combined with anti-osteoporosis drug on postmenopausal women with osteoporotic vertebral compression fracture and analysis of postoperative bone cement leakage risk factors: a retrospective cohort study

A total of 112 OVCF patients at an average age of 51.1 ± 3.2 ranging from 45 years to 55 years who were treated in Guangdong Academy of Medical Sciences, Guangdong Provincial People's Hospital from January 2015 to June 2017 were collected. The compressed vertebrae of all patients showed high signal on MRI-T2 weighted image, and X-ray examination indicated the presence of OVCF. This study was approved by the Medical Ethics Committee of Guangdong Academy of Medical Sciences, Guangdong Provincial People's Hospital, and subjects gave informed consent to the work.

Inclusive criteria include the following: postmenopausal women, the site of compressed vertebrae with low back pain, osteoporosis indicated by pathological examination, complete clinical data, and PKP performed for the first time.

Exclusive criteria include the following: patients with leukemia, hemophilia, immune disease, and idiopathic thrombocytopenia; patients complicated with vertebral tumors; patients with nerve root dysfunction; patients with mental disorder; a history of allergies to zoledronic acid; patients who received conservative treatment; and patients complicated with osteoporosis due to other related diseases such as spinal tuberculosis.

Vitamin D2 Lin Pu Gai Pian (a compound preparation including calcium hydrophosphate, calcium gluconate, and vitamin D2) was purchased from the Beijing Huarun Double-Crane Pharmaceutical Co., Ltd., China; zoledronic acid was obtained from the Sinopharm Group Guorui Pharmaceutical Co., Ltd., Anhui, China.

Patients were divided into control group (n = 52) and observation group (n = 60) according to their therapeutic regimens. All patients were treated with PKP. During the surgical procedure, the sites of the injured vertebrae and pedicle of the vertebral arch were determined under the guidance of a C-arm X-ray machine. The patient was injected with antibiotics 30 min before surgery. The patient was placed in a prone position, the abdomen suspended, anesthetized (general anesthesia) by tracheal intubation, sterilized, and the drapes spread. The needle insertion direction was adjusted according to the patient's vertebral compression condition, keeping the needle insertion path in the center of the compressed vertebrae. The puncture needle was inserted into the 1/3 site in front of the lesion of the vertebrae through the pedicle of the vertebral arch, and the puncture site was confirmed by a C-arm X-ray machine. Balloon dilatation was performed on both sides of the vertebrae after a successful puncture. Bone cement of 2.0–2.5 mL was injected to each side of the vertebrae according to patient's condition. After the errorless frontal and lateral film was confirmed by a C-arm X-ray machine, the T-shaped cannula was repeatedly rotated. The cannula was taken out after the bone cement was fixed (stirring for 15 min). The incision was sutured. During the postoperative management, all patients lay on the back after the surgery and were allowed to turn over 4 h later. Patients were allowed to move around gradually under the protection of thoracolumbosacral orthosis 3–5 days after the surgery. Patients in the observation group were administrated with vitamin D2 Lin Pu Gai Pian (2 tablets/time, 3 times/day, for 3 months) plus zoledronic acid (4 mg/time, intravenous drip of 100 mL diluted normal saline with the drip duration greater than 15 min, only administration once) on the 2^nd postoperative day; patients in the control group were treated with vitamin D2 Lin Pu Gai Pian only. The patient was reviewed by X-ray film 3 months after the treatment.

Interleukin-6 (IL-6), tumor necrosis factor-α (TNF-α), serum procollagen type I N-terminal propeptide (PINP), and serum bone gla protein (BGP) ELISA kits were obtained from the Shanghai Meilian Biotechnology Co., Ltd., China.

IL-6, TNF-α, PINP, and BGP were detected using ELISA kits, which were performed according to the manufacturer's instructions. Three duplicate microwells were set, and the experiment was repeated three times. Bone mineral density (BMD) of the injured vertebrae was measured by X-ray absorption method.

For the main outcome measurements, the curative effect after treatment was observed and evaluated by the Common Diagnosis and Classification Methods in Orthopedics and Functional Outcome Scale which were issued by Liu [15]. Changes in IL-6, TNF-α, PINP, BGP, and BMD before treatment and 1 week after treatment were observed. Vertebral height ratio of injured vertebrae 3 months after surgery and the Cobb angle were calculated. Risk factors affecting clinical efficacy were analyzed by multivariate logistic regression analysis.

For the secondary outcome measurements, pain degree before and 1 week after treatment was evaluated using visual analogue scale/score (VAS) with a total score of 10; the higher the score, the more severe the pain. Dysfunction before and 1 week after treatment was evaluated using Oswestry disability index (ODI) with a total score of 50; the higher the score, the more severe the dysfunction. The condition of postoperative bone cement leakage was observed. Risk factors causing postoperative bone cement leakage were analyzed by multivariate logistic regression analysis.

SPSS 20.0 software package (Guangzhou Pomine Info. Tech. Co., Ltd., China) was used to analyze the data; the data were drawn using GraphPad Prism 7 (Cabit Information Technology Co., Ltd., Shanghai, China). The normal distribution of data was detected by K–S test; the enumeration data were expressed as the rate (%); the chi-squared test was used and shown as χ²; the ranked data were analyzed using the rank sum test, and expressed as Z. The measurement data that conformed to normal distribution and homogeneity of variance were expressed as mean ± SD; comparison between two groups was performed using independent sample t-test; comparison between before and after treatment in a group was carried out by paired t-test, and expressed as t. The data that did not conform to normal distribution and homogeneity of variance were analyzed by non-parametric test, and expressed as Z. Risk factors causing postoperative bone cement leakage were analyzed by multivariate logistic regression analysis. There was a significant difference at P < 0.05.

There were no significant differences in age, body mass index (BMI), marital status, past medical history, place of residence, degree of education, cause, site of injured vertebrae, and dosage of bone cement between the two groups (all P > 0.05, Table 1).

The detection found that there were no significant differences in TNF-α and IL-6 levels in serum, BMD, PINP, and BGP levels; vertebral height ratio of injured vertebrae; and Cobb angle as well as VAS and ODI scores before treatment between the two groups (all P > 0.05). TNF-α and IL-6 levels in serum, PINP level, Cobb angle as well as VAS and ODI scores after a 6-month treatment were significantly reduced in both groups compared with those before treatment, while BMD and BGP levels and vertebral height ratio of injured vertebrae were significantly increased (all P < 0.001). There were no differences in difference values of pre and posttreatment TNF-α and IL-6 levels between the two groups (all P > 0.05), but changes in difference values of BMD, PINP, and BGP levels; vertebral height ratio of injured vertebrae; and Cobb angle as well as VAS and ODI scores during the 6-month treatment were significantly lower in the control group than in the observation group (all P < 0.01, Tables 2, 3, 4, and 5, Figs. 1, 2, 3, and 4).

Evaluation on clinical efficacy after treatment showed that there were 32 patients with excellent efficacy, 26 patients with general efficacy, 2 patients with failed efficacy in the observation group, while 18 patients with excellent efficacy, 28 patients with general efficacy, and 6 patients with failed efficacy in the control group. The rank sum test showed a significant difference in clinical efficacy between the two groups (Z = −2.233, P = 0.026). The imageological feature of a patient showing pre and postoperative changes was shown in Fig. 5.

According to clinical efficacy after treatment, patients with excellent efficacy were divided into excellent efficacy group (n = 50), and patients with general efficacy and failed efficacy were divided into general efficacy group (n = 62). The results revealed that there were differences in the use of zoledronic acid, vertebral height ratios of injured vertebrae, Cobb angle, and ODI score (all P < 0.05, Table 6).

Multivariate logistic regression analysis of factors having differences by univariate analysis and age found that Cobb angle was not a risk factor affecting clinical efficacy, and the use of zoledronic acid (odds ratio (OR) = 4.600, 95% confidence interval (CI) 1.483~14.266), vertebral height ratio of injured vertebrae (OR = 0.916, 95% CI 0.863~0.971), and ODI (OR = 1.224, 95% CI 1.131~1.325) were independent risk factors affecting clinical efficacy (Tables 7 and 8).

There were 15 patients suffering from postoperative bone cement leakage in the control group and 10 patients suffering from postoperative bone cement leakage in the observation group. There was no significant difference in the incidence of postoperative bone cement leakage between the two groups (χ² = 2.383, P = 0.123).

According to postoperative bone cement leakage, patients were divided into leakage group (n = 25) and no leakage group (n = 87). The data of age, BMI, marital status, past medical history, place of residence, degree of education, causes, site of injured vertebrae, use of zoledronic acid, fracture acuteness, peripheral vertebrae wall damage, bone cement dosage, TNF-α, IL-6, BMD, PINP, BGP, vertebral height ratios of injured vertebrae, Cobb angle, VAS score, and ODI score were analyzed by univariate analysis. The results revealed that there were differences in peripheral vertebrae wall damage and bone cement dosage (both P < 0.05, Table 9).

Multivariate logistic regression analysis of factors that have differences by univariate analysis and age found that age was not a risk factor causing postoperative bone cement leakage (P > 0.05), and bone cement dosage (OR = 2.845, 95% CI 2.164~3.741), and peripheral vertebrae wall damage (OR = 3.388, 95% CI 1.966~5.840) were independent risk factors causing postoperative bone cement leakage (Tables 10 and 11).