Gas chromatographic profiling and pattern recognition analysis of urinary organic acids from uterine myoma patients and cervical cancer patients

• Published in: Journal of chromatography. B, Biomedical sciences and applications Vol. 712; no. 1; pp. 11 - 22
• Main Authors: Kim, Kyoung-Rae, Park, Hyoung-Geun, Paik, Man-Jeoung, Ryu, Hee-Sug, Suk Oh, Kie, Myung, Seung-Woon, Liebich, Hartmut M
• Format: Journal Article
• Language: English
• Published: Amsterdam Elsevier B.V 07.08.1998; Elsevier Science
• Subjects: Adult; Aged; Biological and medical sciences; Carboxylic Acids - urine; Chromatography, Gas - methods; Discriminant Analysis; Female; Female genital diseases; Genital system. Mammary gland; Gynecology. Andrology. Obstetrics; Humans; Investigative techniques, diagnostic techniques (general aspects); Leiomyoma - urine; Medical sciences; Middle Aged; Organic acids; Pathology. Cytology. Biochemistry. Spectrometry. Miscellaneous investigative techniques; Tumors; Uterine Cervical Neoplasms - urine; Uterine Neoplasms - urine; Organic acids; Human; Urine; Biological fluid; Biochemical analysis; Coupled method; Biological marker; Tumoral marker; Uterine cervix; Malignant tumor; Female genital diseases; Gas chromatography; Uterus; Clinical biology; Uterine diseases; Female; Benign neoplasm; Mass spectrometry; Uterine cervix diseases
• ISSN: 0378-4347; 1387-2273
• DOI: 10.1016/S0378-4347(98)00155-8

An efficient organic acid profiling and pattern recognition method is described for the correlation between urinary organic acid profiles and uterine cervical cancer. After methoximation of keto acids in alkalinized urine samples, all free organic acids were recovered by a dual solid-phase extraction procedure, followed by conversion to tert.-butyldimethylsilyl derivatives for the profiling analysis by dual-capillary column gas chromatography (GC) with subsequent screening for acids by retention index ( I) library matching. A total of 50 organic acids were positively identified in urine samples (0.25 ml) from 12 uterine myoma (benign tumor group) and 14 uterine cervical cancer (malignant tumor group) patients studied. When the GC profiles were simplified to their corresponding organic acid I spectra in bar graphical form, characteristic patterns were obtained for each average of benign and malignant tumor groups. Stepwise discriminant analysis performed on the GC data selected 16 acids as the variables discriminating between the two groups. Canonical discriminant analysis applied to these 16 variables correctly classified 26 urine samples into two separate clusters according to tumor types in the canonical plot.