A multicenter, descriptive epidemiologic survey of the clinical features of spinal metastatic disease in China.

OBJECTIVE
Spinal metastases have unique epidemiological features and treatment methods. Unfortunately, the relative scarcity of spinal metastases has limited the widespread development of descriptive epidemiological studies, especially in Asian countries. The purpose of this study was to describe the epidemiological characteristics of patients with metastatic spinal tumors in China between 2007 and 2019.


METHODS
From January 2007 and July 2019, data on patients with spinal metastases were collected from five cancer centers in China, and demographic characteristics, primary tumor types, segments and numbers of vertebral lesions, disease-related scores, and treatment methods were reviewed.


RESULTS
A total of 2228 patients with spinal metastases were reviewed in this study, including 1279 male patients and 949 female patients, and the male to female ratio was 1.35: 1. More than half of patients developed metastatic diseases between the ages of 50 years and 69 years (64%). Overall, lung cancer (824 cases, 37%) was the most common primary tumor type and the most common level of spinal involvement was multi-level of metastases (860 cases, 39%). 705 patients (32%) had undergone surgical treatments, 1028 patients (46%) had undergone radiotherapy for metastatic vertebrae, and 855 patients (38%) had received systemic treatments. The age, primary tumor type, number of involved vertebrae, Frankel grade, and spinal instability neoplastic score would affect the surgical decision-making.


DISCUSSION
This study provides insight into the epidemiological characteristics of spinal metastasis and health care service utilization in spinal metastasis patients in China.


ABBREVIATIONS
ICD-10: International Classification of Disease, Tenth Revision; VAS: Visual Analogue Scale; KPS: Karnofsky Performance Status; SINS: Spinal Instability Neoplastic Score; SOSG: Spine Oncology Study Group; MRI: Magnetic Resonance Imaging; CT: Computed Tomography; ECT: Emission Computed Tomography.

Due to the special anatomical characteristics of the spine, spinal metastases have different epidemiological characteristics and treatment methods compared with metastatic tumors in other anatomic sites. Unfortunately, the relative rarity of spinal metastases has limited the widespread development of descriptive epidemiological studies, especially in Asian countries. Several populationbased morbidity and epidemiological surveys have been reported in Western countries [7][8][9]. Zaikova et al. [7] described the population-based incidence of spinal metastases which required treatment, including patients with or without spinal cord compression, and summarized the neurological function of these patients. The first nationwide epidemiological survey of spinal tumors (including metastatic tumors) in Asia was performed by Sohn and colleagues [10]. They analyzed the age and gender differences, annual incidence, location of primary tumors, medical costs, and hospital stays associated with spinal tumors in Korea. However, there are few descriptive epidemiological reports on spinal metastases in China, which is undoubtedly unfavorable for providing information to health care institutions and promoting research on spinal metastases.
To identify the epidemiological characteristics of spinal metastases in China, we conducted a descriptive epidemiological survey in five cancer centers throughout the country. By documenting the gender and age, type of primary tumor, site of bone metastases, neurological status, and treatment methods, we expected to provide epidemiological evidence for spinal metastases in China.

Sources of Data
The patients in this survey were consecutively recruited from five cancer centers between January

Recorded Indicators
The medical records were systematically reviewed and the following indicators were recorded: (1) demographic characteristics: age and gender; (2) primary tumor types; (3) segments and numbers of vertebral lesions; (4) disease-related scores: Frankel grade, visual analogue scale (VAS) score, karnofsky performance status (KPS) score and spinal instability neoplastic score (SINS); (5) interventions: surgical treatments and non-invasive treatments.
The Frankel grade is divided into five levels to assess the persistence of sensory and motor functions below the level of injury in cancer patients [11]. The VAS score is used to assess the subjective pain intensity of patients, and is also one of the commonly used indicators for the evaluation of the efficacy after treatment [12]. Assessing the stability of spine requires comprehensive evaluation of imaging and clinical manifestations. The SINS scoring system has a high sensitivity and specificity for potential unstable spinal lesions, which can help doctors to identify patients who may have spinal instability or deformity [13]. The KPS score can be used to evaluate the general condition of cancer patients. Patients with higher scores can tolerate the side effects of treatments to the greatest extent, and therefore may receive thorough treatments [14].
The management of spinal metastases requires multidisciplinary collaboration, of which surgery and radiotherapy are the main treatment methods. Surgical treatment can be divided into open surgery and minimally invasive surgery. Open surgery includes palliative surgery and radical surgery.
Palliative surgery is aimed at alleviating the symptoms of patients and improving the quality of life of patients without pursuing the thoroughness of tumor resection, including laminectomy and vertebra corpectomy. Radical surgery is based on the principle of complete resection of tumor lesions. The goal is to extend the survival time of cancer patients as much as possible, including piecemeal resection and total en bloc spondylectomy. Minimally invasive surgery has the advantages of rapid postoperative recovery, low incidence of complication, and no delay in adjuvant treatment. It is suitable for patients with poor general conditions, including percutaneous vertebroplasty, radiofrequency ablation, and percutaneous pedicle screw fixation.

Statistical Analysis
Continuous variables were described as mean ± standard deviation (normal distribution) or median (skewed distribution), and categorical variables were described as frequency or percentage. The Student t-test and chi-square test were used to identify any statistical differences between the means and proportions among the groups. Univariate logistic regression and multivariate logistic regression were used to identify potential factors affecting treatment methods. The variables significant at the p < 0.15 level were included in the multivariate logistic regression analysis to screen out independent predictors. All of the statistical analyses were performed with IBM SPSS Statistics 22.0 (IBM, Armonk, NY, USA) and P < 0.05 (two-sided) was considered statistically significant.

Patient Demographics
A total of 2228 patients with spinal metastases were reviewed in this study, including 1279 male patients and 949 female patients, and the male to female ratio was 1.35: 1. The average age of onset time of spinal metastases in the general population was 58.6 ± 11.8 years (range 13-92 years), with a median age of 59.0 years. More than half of patients developed metastatic diseases between the age of 50 years and 69 years (63.47%). In terms of subgroups, the mean age of male patients was 59.3 ± 12.3 years (range 14-92 years), with the median age was 60.0 years, and the mean age of female patients was 57.6 ± 11.2 years (range 13-91 years), with the median age was 58.0 years. Using the mean value for analysis, the onset time of female patients was earlier than that of male patients, and the difference was statistically significant (P = 0.001). Further analysis showed that the proportion of elderly patients (over 60 years) in the male population was greater than that of female patients (53.17% vs 47.63%). Similarly, the difference between the proportions was statistically significant (P = 0.012). The distribution of gender and age of patients with spinal metastases was described in Table 1 and Fig. 1.

Primary Tumor Type
A study conducted in Korea showed that the six most common primary tumor sites for spinal metastases were lung, liver and biliary tract, breast, colon, stomach and prostate [17]. When subgroup analysis was performed by age, the incidences of lung cancer, liver cancer, colon cancer and gastric cancer were the highest in the 70-79 years age group; in contrast, the incidence of breast cancer was the highest in the age group of 50-59 years age group, and the incidence of prostate cancer was the highest in the over 80 years age group. An epidemiological survey based on a population of spinal metastases in South-East Norway showed that prostate cancer, lung cancer and breast cancer accounted for 66% of all patients, followed by myeloma, lower gastrointestinal tumor, kidney cancer and tumors of unknown origin [7]. Another study comparing trends in surgical treatments of spinal metastases across two decades and three continents noted that colon cancer, liver cancer and lung cancer accounted for a higher proportion in Asian countries, while breast cancer, prostate cancer and myeloma had a lower proportion [18]. The authors believed that the difference was largely dependent on the different high-prevalence cancer types and different early screening programs in different parts of the world.
In the current study, the most common primary tumor was lung cancer, followed by tumors of unknown origin, breast cancer, kidney cancer, and gastrointestinal tumors. For male patients, lung cancer was the most common type, followed by tumors of unknown origin, kidney cancer, prostate cancer, and liver cancer. For female patients, lung cancer was the most common type, followed by breast cancer, tumors of unknown origin, myeloma, and gastrointestinal tumors. This result is partially different from previous literature, especially those based on the population of western countries.
On the one hand, this difference may reflect the significant differences in the incidence of various malignant tumors in different countries, regions and ethnicities. Lung cancer and breast cancer are the most common cancers and the leading causes of cancer death among male and female patients in less developed countries, respectively. However, in more developed countries, prostate cancer is the most common cancer among male patients and lung cancer is the leading cause of cancer death among female patients [19]. In addition, the incidence of liver cancer in China is more than three times than that in North America and 10 times than that of some European countries; while the incidence of breast cancer in China is lower than that in North America [18].
On the other hand, the proportion of tumors of unknown origin in our study reached 16.5%, and the proportion reported in the previous literature were 1.8%-10% [7,18], which was largely determined by China's national conditions. In some economically underdeveloped areas, some patients who had

Spinal Involvement
The segments and numbers of vertebral lesions in the current study were similar to those in the previous literature [7,15,20]. In our study, the most common level of spinal involvement was multilevel metastasis, followed by thoracic vertebrae, lumbar vertebrae, sacral vertebrae, and cervical vertebrae. For 58.98% of patients, the number of involved vertebrae was less than 3.
Bollen and colleagues found that the most common level of spinal involvement was multi-level metastasis, followed by thoracic and lumbar vertebrae, and patients with less than 3 involved vertebrae accounted for 49.57% [15]. In another study conducted by Zaikova and colleagues, based on available radiological records, 83% of patients had multiple lesions, 15% of patients had a single lesion, and 2% of patients were unknown [7]. And they found a statistically significant correlation between the number of metastatic lesions and the type of primary cancer: patients with prostate cancer had the lowest rate of single metastasis (4%), and patients with renal cancer and lymphoma had the highest rate of single metastasis (31% and 47%). In the further subgroup analysis of the surgical population, we found that patients with good neurological function (Frankel E) were three times more likely to undergo minimally invasive surgery rather than open surgery, compared with patients with neurologic deficits. This is a relatively novel conclusion. We speculate that patients with pure pain and good neurological function tend to receive minimally invasive percutaneous vertebroplasty, and those with neurological dysfunction have to choose open decompression surgery, because the minimally invasive pedicle screw fixation combined with minimally invasive decompression in our institutions is relatively rare.

Discussion
The current study identified 2228 patients with spinal metastases, including 1279 male patients and years, accounting for 63.47% of the whole population. The mean age of female patients was less than that of male patients, and the difference was statistically significant. Lung cancer was the most common primary tumor type for male and female patients, followed by tumors of unknown origin.
When the number of involved vertebrae was less than 3, the lumbar vertebrae were the most common level, and when the number was greater than 3, the multi-level metastases were the most common level. About half of the patients experienced varying degrees of neurological deficits, 80% of the patients experienced potential or actual spinal instability, and 77% of the patients occurred moderate to severe pain. In terms of therapeutic interventions, 31.6% of the patients underwent surgical treatments, 46.1% of the patients underwent radiotherapy, and 38.4% of the patients underwent systemic treatments. Younger patients with tumors of moderate to slow growth speed, less than 3 involved vertebrae, poor neurological function and spinal instability were more likely to receive surgical treatments, while, patients with normal neurological function were more likely to select minimally invasive surgery rather than open surgery.

Patient Demographics
In a large retrospective cohort study of 1043 patients with spinal metastases, the baseline data showed that 542 male patients (52%) and 501 female patients (48%) were included in the study, with a mean age of 64.8 years [15]. Another retrospective study of 544 patients with spinal metastases treated with radiation therapy was performed in Japan, Mizumoto and colleagues [16] reported that there were 287 male patients (52.8%) and 200 female patients (47.2%), with a mean age of 63 years.
Several studies on the epidemiological investigation of spinal metastases also showed that the incidences were higher in male patients than in female patients. In addition, the incidences of spinal metastases increased with age and the proportions of patients in the 60-69 years age group were the highest [10,17].
In the current study, male patients accounted for approximately 57% of the whole population, with a male to female ratio of 1.35: 1. The mean age of patients at the time of diagnosis was 58.6 years, and the peak age of onset was 50 to 69 years. Male patients had a larger proportion of elderly patients (over 60 years) compared with female patients (53.2% vs 47.6%), and the difference was statistically significant. The male to female ratio and distribution of age in our study are similar to previous literature. The proportion of male patients is slightly larger than that of female patients, and the mean age at diagnosis is about 60 years. Although there is a statistically significant difference in the age at onset between male patients and female patients, the difference is only 1.7 years, and we think it has little clinical significance. Statistically significant values may be due to the large sample size of the current study.

Limitations
We acknowledge that this study has some limitations. First, due to the retrospective design of this study, selection bias and recall bias are unavoidable. With that in mind, we try to be as inclusive as possible. The study included all consecutive cases that were recorded by specific personnel, and we also tracked cases that were referred through other hospitals or remained in other wards. Second, due to the limited information in our database, we cannot analyze the proportion and grade of spinal cord compression, so we cannot describe the cause of surgical operations, such as emergency spinal cord compression or progressive disease. In the end, the preferred treatment methods for different cancer centers may be different, and the selection criteria for surgical patients were not uniform, which may skew the results. In addition, these data were collected over the span of more than a decade. During this period, the diagnosis and treatment methods of spinal metastases had changed significantly.

Conclusions
The current study provides a relatively detailed epidemiological analysis of patients with spinal metastases in China, and contribute to better understand age and gender differences, primary tumor    multi-level of metastases: involve two or more segments of cervical vertebra, thoracic vertebra, lumbar vertebra, and sacral vertebra at the same time; sacral vertebrae: due to inconsistency in radiographic records of different medical centers, the sacral vertebrae and caudal vertebra were defined as one vertebra