Screening And Quantitative Analysis Of Volatile Markers In Breath Of Ovarian Carcinoma Patients

Objective Quantitative analysis of the concentration of volatile organic compounds inthe breath of ovarian carcinoma patients, benign ovarian tumors patients and healthycontrols for the purpose of screen out the volatile markers for ovarian carcinoma. Toevaluate the diagnostic values of volatile markers for ovarian carcinoma. Finally, toestablish the breath diagnostic function models of ovarian carcinoma.Methods The samples were collected from47ovarian carcinoma patients (including39the epithelial ovarian cancer,2germ cell ovarian cancer,6borderline ovarian cancer),86benign ovarian tumor patients (31the epithelial ovarian tumor,20the matureteratoma,25ovarian cyst) and48healthy controls with3L Tedlar bag, meanwhile theambient air were collected for reference. VOCs were extracted with SPEM andanalyzed by Gas chromatography (GC)/Mass spectrometry (MS). Compare theconcentrations of volatile organic compounds among the three group (ovarian cancerpatients, the benign ovarian tumors and the healthy controls) and screen out the volatilemarkers for ovarian carcinoma. Mann-Whitney U Test was used to compare thedifferent groups. The receiver operator characteristic curves (ROC) of volatile markerswere made and were used to evaluate the diagnostic abilities. Besides, the diagnosticfunctions of ovarian carcinoma were established with Fisher discriminate analysis, andthe diagnostic abilities of the functions were assessed with a leave-one-out classificationprocedure.Results Three kinds of volatile markers were screened from ovarian carcinoma. Theyare styrene, nonanal and decanal and were significant different among the ovariancarcinoma group, the benign ovarian tumors and healthy. The styrene, nonanal and decanal all were significant different when compare the ovarian cancer and healthycontrol (p <0.05). The sensitivity and specificity of styrene for diagnostic epithelialovarian cancer were respectively90%and52%, the nonanal were82%and60%, thedecanal were82%and54%. The three kinds of VOCs were significant differentbetween the ovarian carcinoma (the epithelial original, the germ original and theborderline ovarian tumor) and the healthy. When compared the epithelial ovarian cancerand the epithelial benign ovarian tumor, nonanal and decanal were also significantdifferent (p <0.05) and styrene, nonanal and decanal all were significant differentbetween the ovarian carcinoma and the benign mass. Only styrene was found significantdifferent between the epithelial benign ovarian tumor and the health group and novolatile marker was significant different between the early stage and the advancedovarian cancer. The sensitivity and specificity of diagnostic function model combinedby styrene, nonanal and decanal for epithelial ovarian cancer were respective69.2%and73.4%, and the sensitivity and specificity69.2%and70.9%by the cross-validation. Thesensitivity and specificity of diagnostic function model combined by styrene, nonanaland decanal for ovarian carcinoma were respective70.2%and70.8%, and thesensitivity and specificity were66.0%and68.8%by the cross-validation.Conclusions Styrene, nonanal and decanal may be considered as the potential volatilemarkers for diagnosis of ovarian carcinoma. The concentration of styrene and nonanalincreased gradually among the healthy control, the benign ovarian tumors and theovarian carcinoma group. The two kinds of volatile markers may be associated with thebenign ovarian tumors and carcinogenesis. The concentration of volatile markers has norelation with the ovarian cancer stages, which could be used for the early diagnosis ofovarian cancer. The breath diagnosis function model Combined with three kinds ofvolatile markers can distinguish between ovarian carcinoma and non-ovariancarcinoma.