The Application Of CA199and NSE In The Diagnosis,Treatment And Follow-up Of Intracranial Epidermoid Cyst

Background:Serum tumor biomarkers have been widely applied in the clinics for the diagnosis, prognostic evaluation and follow-up of various cancer categories including lung cancer, liver cancer, prostate cancer, etc. Serum tumor biomarkers for intracranial tumors were very rare. The preoperative diagnosis and postoperative follow-up rely solely on the imaging techniques such as CT and MRI. For lesions with similar imaging characteristics, physicians have to refer to stereotactic biopsy or exploratory craniotomy for final pathological diagnosis. It has been reported that serum carbohydrate antigen199(CA199) level of intracranial epidermoid cyst (EC) and dermoid cyst (DC) was mildly to moderately elevated preoperatively and declined to normal range after surgical resection. Therefore, serum tumor biomarkers, especially CA199, might be a possible candidate for the the diagnosis, treatment and follow-up of intracranial EC. Here we aim to evaluate the expression and discover the potential clinical application of common serum tumor biomarkers in this specific disease entity by retropectively exploring the patient data in Second Affliated Hopspital of Zhejiang University during the past5years.Materials and methods:We analyzed the expression of common serum tumor biomarkers, namely CA199, a feto-protein (AFP), carcinoembryonic antigen (CEA), carbohydrate antigen125(CA125), neuron specific enolase (NSE), squamous cell carcinoma (SCC) and β human chorionic gonadotropin (P-HCG), in both EC group recruited from Second Affiliated Hopspital of Zhejiang University during the past5years, and healthy control group. For those with significantly different expression, receiver operating characterisitics (ROC) curves were constructed to calculate areas under curves (AUCs) for each single tumor biomarker, together with combined markers after multi-variant fitting by logistic regression analysis. The optimum cutoff value was calculated by Youden index. The possible influencing factors for tumor biomarker expression, gross total resection (GTR) vs. subtotal resecton (STR) and recurrence were analysed by uni-variate analysis and multi-variate analysis.Result:A total of228ECs were treated in our hospital during the last5years. Preoperative serum tumor biomarkers and surgical resection were available in only34cases. The control group was matched with the EC group by age and gender. It was found that serum CA199level was significantly higher in the EC group (418.0±2080.7U/ml vs.7.3±4.3U/ml,p<0.001),while NSE level was significantly lower in the EC group (6.7±5.0ng/ml vs.9.7±4.7ng/ml,p=0.020). Although there were three cases with positive SCC expression, no significant difference existed between the two groups (0.9±0.4ng/ml vs.0.8±0.2ng/ml, p=0.104). AUC for CA199was0.791(95%CI0.669-0.913), with cutoff value of13.15U/ml (sensitivity73.5%, specificity94.1%); AUC for NSE was0.693(95%CI0.558-0.828), with cutoff value of4.35ng/ml (sensitivity51.6%, specificity90.3%); AUC for SCC was0.595(95%CI0.448-0.742), with cutoff value of1.15ng/ml (sensitivity20.0%, specificity93.3%); AUC for combination of CA199and NSE was0.874(95%CI0.773-0.976, sensitivity86.7%, specificity90.0%); Further addition of SCC didn’t improve the diagnostic efficiency. Tumor sizes between CA199positive and CA199negative groups were significantly different (86.87±79.45cm3vs.14.89±8.43cm3,p=0.049) and linear regression analysis found a positive correlation between the two indexes with a correlation coefficient of0.837(p<0.01).Although CA199level in the GTR group was lower than the STR group (24.0±44.3vs.74.4±97.5U/ml, p=0.564), and NSE higher than the STR group (6.8±5.1vs.6.0±4.9ng/ml, p=0.684), there existed no significant difference. The only factor that might affect surgical resection was tumor size (uni-variant analysis p=0.084, multi-variant analysi p=0.061). Similarly, although CA199level was higher in the recurrence group (72.5±68.4vs.56.6±97.8U/ml, p=0.750), and NSE lower in the recurrence group (5.4±4.5vs.6.3±5.5ng/ml,p=0.916), there existed no significant difference. Only tumor size (uni-variant analysis p=0.040, multi-variant analysi p=0.232) and follow-up time (uni-variant analysis p=0.007, multi-variant analysi p=0.128) might be related to recurrence.Conclusion:Our data indicated that for serum CA199level higher than13.15U/ml, NSE lower than4.35ng/ml, after excluding the possibility of gastrointestinal system tumor, lung cancer, inflammation and other related diseases, the existence of intracranial epidermoid cyst should be considered. There was no correlation between CA199and NSE, and combined determination can improve diagnostic accuracy. Positive correlation existed between tumor size and CA199, but not NSE. Although CA199level was higher and NSE was lower in cases with STR or recurrence, neither achieved statistical significance. Up to now, our study was the first to systematically analyze the clinical value of common tumor biomarkers in the diagnosis, treatment and follow-up of intracranial EC. Our research framework, statistical method and study results shed light on further fundamental research and clinical trials in this field.