Serum sclerostin levels in renal cell carcinoma patients with bone metastases

Sclerostin has been proposed as a potent inhibitor of bone formation. Sclerostin antibodies are under clinical development to treat osteoporosis and metastatic bone disease. Serum sclerostin level is elevated in multiple myeloma, an osteolytic malignancy, where it might serve as predictive marker for the use of sclerostin-directed antibodies. As renal cell carcinoma (RCC) patients often present with osteolytic metastases, we aimed to investigate serum sclerostin levels in RCC patients. Our study included 53 RCC patients (19 with bone metastases, 25 with visceral metastases and 9 with localized disease) and 53 age- and gender-matched non-osteoporotic controls. Frozen serum samples were subjected to sclerostin quantitative sandwich ELISA. The mean serum sclerostin levels of RCC patients and controls were 45.8 pmol/l and 45.1 pmol/l, respectively (p = 0.86). Analysis of variance showed no difference between the subgroups of RCC patients with regard to visceral or bone metastases or localized disease (p = 0.22). There was no significant association between eGFR (estimated glomerular filtration rate) and serum sclerostin levels in RCC patients (r = 0.05; p = 0.74) and controls (r = 0.06; p = 0.68). Our results indicate that serum sclerostin levels appear not to be a valuable biomarker to assess the occurrence of bone metastases in RCC patients.

Scientific RepoRts | 6:33551 | DOI: 10.1038/srep33551 severity of bone disease and circulating levels of serum sclerostin has been found, suggesting that MM-cells utilize sclerostin in the formation of osteolytic lesions 17,18 .
Renal cell carcinoma (RCC) accounts for approximately 4% of all malignancies in Western Europe and the USA; the age-adjusted incidence rate lies between 6.2-8.4/100,000/year, and the male to female ratio is 3:2 19,20 . It occurs less frequently in developing countries 21 . Tobacco smoking and overweight have been identified as major etiological factors 21 . The 5-year survival rate increased significantly to 75% during the last 30 years 19,20 . About 30% of the RCC patients develop bone metastases 22 which are osteolytic and highly vascularized, as well as chemoand radio-resistant. RCC was the fourth most common cancer presenting with symptomatic bone metastases in our own patient collectives, accounting for seven percent of patients treated for spinal metastases 23,24 . Since there are no reports to date on serum sclerostin levels in RCC, we systematically evaluated them in different groups of RCC patients: patients with localized disease, bone or visceral metastases and compared them to serum sclerostin levels in non-osteoporotic controls.

Methods
This case-control study included 53 patients with histologically confirmed RCC treated at the Division of Oncology, Medical University of Graz between 1986 and 2012. All included RCC patients with bone or visceral metastases had a minimum follow-up of 12 months if alive with disease. For the patients with localized disease a minimum follow-up of 4 years was required for inclusion in this study. Clinico-pathological data were retrieved from medical records at the Division of Oncology. Bone metastases were confirmed either by histology, or characteristic magnetic resonance imaging (MRI), computed tomography (CT), or bone scan with multiple lesions; visceral metastases were also confirmed by imaging techniques (MRI or CT). All patients were routinely followed at the Division of Oncology, Medical University of Graz, with examinations at regular intervals (three-month intervals for the first two years after diagnosis, then six-month intervals for years two to five, 12-month intervals for years 6-10). The serum samples were taken within 90 days after diagnosis of primary disease or bone or visceral metastasis, respectively.
The controls were matched for age and gender as recent studies have demonstrated that sclerostin levels are age dependent 25,26 . Controls were either blood donors (age under 65 years) or non-osteoporotic patients with osteoarthritis (OA) (age over 65 years), as no difference in sclerostin levels has been reported between patients with OA and healthy controls 27 . The controls with OA were treated at the Department of Orthopedics and Orthopedic Surgery, Medical University of Graz and the clinical and demographic data were obtained from their medical records.
Exclusion criteria for the RCC patients were primary diagnosis of RCC before 1986, further malignant disease other than RCC and osteoporosis. Exclusion criteria for the controls were osteoporosis, malignant disease and chronic kidney disease. Furthermore, all patients and controls with concomitant use of drugs potentially interfering with serum sclerostin levels, such as glucocorticoids, bisphosphonates and denosumab [28][29][30][31][32] , were excluded from our study. Renal function was estimated using serum creatinine and the CKD-EPI creatinine equation 33 . As reduced renal function is common in RCC patients, usually following tumor nephrectomy, we analyzed the correlation between eGFR (estimated glomerular filtration rate) and serum sclerostin levels. The participants had no dietary restrictions. The samples used for this research project were kindly provided by Biobank Graz (RCC patients and patients with OA) and the Division of Endocrinology and Diabetology, Department of Internal Medicine, Medical University of Graz (blood donors).
Sclerostin levels in blood serum samples (stored at − 70 °C) were determined with a quantitative sandwich ELISA (Sclerostin ELISA, Biomedica, Vienna, Austria) according to the standard protocol provided by the manufacturer. Intra-and interassay coefficients of variation were < 7% and < 10%, respectively. All samples were assayed in duplicate.
This study and its experimental protocol were approved by the ethics committee of the Medical University of Graz (EK-Nr: 18-271 ex 06/07, amendment 2, of 22 July, 2013) IRB00002556 (Institutional Review Board Registry) and the methods and all experiments were performed in accordance with relevant guidelines and regulations. Informed consent was obtained from all subjects before sample-taking. Statistical analysis. The data were first analyzed with descriptive statistics, using mean and standard deviation for continuous data like age and laboratory parameters. Categorical data are described by absolute frequencies. The differences in serum sclerostin levels between RCC patients and controls were analyzed with the t-test and ANOVA. Pearson's correlation coefficient was calculated for eGFR and serum sclerostin values. All analyses were performed with the statistical software package IBM SPSS Statistics version 23. A two-sided p-value less than 5% was considered significant.

Results
Our study included 9 patients with localized disease, 19 patients with bone metastases and 25 patients with visceral metastases only (no bone metastases). The male to female ratio was approximately 2:1, the mean age 62 years (range 31 to 89). At last follow-up (July 11, 2016) 70% of all included RCC patients had died, in detail: 22% of the RCC patients with localized disease and 84% of the RCC patients with metastases (visceral or bone metastases). Table 1 summarizes the clinical characteristics and laboratory data of the RCC patients and controls.

Table 2. Serum sclerostin levels in RCC patients with visceral metastases, bone metastases and localized
disease and age and gender-adjusted controls. # Comparison of RCC patients and controls using t-Test, § using ANOVA.

Discussion
In recent reports, it has been hypothesized that sclerostin might be a possible target for therapeutic agents in prostate and breast cancer patients with bone metastases 11,12,16 . Bone metastases can be osteoblastic or osteolytic 16 ; both may lead to acute and chronic pain, but with osteolytic metastases there is also a higher risk of pathologic fracture 11 . Besides the common anti-resorptive agents such as bisphosphonates and the RANKL-inhibitor Denosumab, new treatment options with stimulators of bone formation are under intensive investigation 11 , in the hope of achieving a better bone metabolism balance. A sclerostin-directed antibody as a treatment option for bone metastases would address such an osteoanabolic approach. Since RCC patients frequently develop osteolytic bone metastases 22 and many of them require orthopedic-oncological surgery for pathological or impending fractures, we examined whether serum sclerostin levels in RCC patients are different from healthy controls and whether osteolytic bone metastases are associated with higher sclerostin serum levels. In this study, we found neither significant differences between serum sclerostin levels in RCC patients and controls, nor different sclerostin levels in RCC patients with bone versus non-bone metastases. Interestingly, the opposite has been reported for MM, which also presents with osteolytic lesions. In this entity elevated serum sclerostin levels and even a correlation between sclerostin expression in MM-cells and severity of disease, have been found 17 . Elevated serum sclerostin levels were also reported in prostate cancer, but without an analysis of bone metastases 34 . The   role of sclerostin in cancer and SOSTDC1 (sclerostin domain containing 1)-expression in cancerous tissue was addressed in several studies. A downregulation of SOSTDC1 was found in breast cancer and RCC tissues 35,36 .
Additionally, immunochemistry showed lower sclerostin expression in prostate cancer than in normal prostate tissue 37 . It is likewise reported that prostate and breast cancer cells produce DKK-1 38,39 which, like sclerostin, inhibits Lrp5 (activator of the Wnt/ß-catenin pathway). An in vitro three dimensional model of metastastic bone formation has been developed to study microenvironmental interactions with breast cancer cells 40 . This model showed a reduction in osteoblastic tissue thickness and an increase in osteoclastogenesis in the presence of breast cancer cells 41 . Even though these models do not yet provide detailed information on sclerostin, they seem to show promise for clarifying the underlying pathomechanism between sclerostin and metastatic bone disease. If and how sclerostin antibody will represent a therapy option in bone metastases in RCC patients is still unresolved. Our results imply that serum sclerostin levels are not suitable to detect bone metastases in RCC patients and that increased levels of sclerostin cannot be used as a potentially predictive marker for sclerostin-directed treatment strategies.
Our study population of RCC patients had lower eGFR-levels than the controls, mainly due to previous nephrectomy. Data on the correlation of serum sclerostin and chronic kidney disease (CDK) is scarce and controversially reported. Most of the published data report elevated serum sclerostin levels with declining kidney function [42][43][44] , but other studies found elevated renal elimination of sclerostin in CKD 45 and reduced serum sclerostin in children with CKD 46 . In our study, we found no significant association between eGFR and serum sclerostin levels in RCC patients and controls.
Recent reports demonstrated that elevated serum sclerostin levels correlate with high bone turnover 42 , increasing age 25,47,48 and high ratio of fat mass as well as bone mass 49,50 . On the other hand, serum sclerostin levels decrease with greater physical activity 51 . Results differ on changes in serum sclerostin levels in relation to osteoporosis. Ardawi et al. suggest that serum sclerostin levels predict the risk of osteoporosis related fractures (ORF), reporting a sevenfold higher risk for ORF for each standard deviation of the sclerostin level above the normal range 26 . This is in contrast to a study that compared serum sclerostin levels in healthy controls and/or patients with osteoporosis to patients with OA 27 where no significant difference was found (0.78 vs 0.71 vs 0.80 ng/ml) between these groups, as well as to the findings of Amrein et al. that serum sclerostin levels do not linearly correlate with fracture risk 52 . As the correlation of serum sclerostin levels to sclerostin expression in bone and cancer tissue, and the role of sclerostin in different context are not thoroughly clear and, similarly, the effects of CKD on serum sclerostin levels are uncertain, we conclude that serum sclerostin levels have to be interpreted with caution and are currently not a useful biomarker in this setting.
We acknowledge the limitations of this study, mostly arising from the retrospective design and the small number of cases, as well as the absence of data on physical activity, which might bias the results.

Conclusions
As we found no significant differences between the serum sclerostin levels in patients with RCC and controls, sclerostin serum level appears not to be a useful biomarker for assessing bone metastases in RCC patients. Furthermore, other approaches have to be explored to select RCC-patients, who might benefit from sclerostin antibodies as a further therapy option for bone metastases.