Nonhistological diagnosis of human cerebral tumors by H-1 magnetic resonance spectroscopy and amino acid analysis

Citation
Jm. Roda et al., Nonhistological diagnosis of human cerebral tumors by H-1 magnetic resonance spectroscopy and amino acid analysis, CLIN CANC R, 6(10), 2000, pp. 3983-3993
Citations number
44
Categorie Soggetti
Oncology
Journal title
CLINICAL CANCER RESEARCH
ISSN journal
10780432 → ACNP
Volume
6
Issue
10
Year of publication
2000
Pages
3983 - 3993
Database
ISI
SICI code
1078-0432(200010)6:10<3983:NDOHCT>2.0.ZU;2-B
Abstract
We describe a multivariate analysis procedure to classify human cerebral tu mors nonhistologically in vitro, combining the use of H-1 magnetic resonanc e spectroscopy (MRS) with automatic amino acid analysis of biopsy extracts. Eighty-one biopsies were obtained surgically and classified histologically in eight classes: high-grade astrocytomas (class 1, n = 19), low-grade ast rocytomas (class 2, n 10), normal brain (class 3, n = 9), medulloblastomas (class 4, n = 4), meningiomas (class 5, it = 18), metastases (class 6, n = 8), neurinomas (class 7, n = 9), and oligodendrogliomas (class 8, n = 4). P erchloric acid extracts were prepared from every biopsy and analyzed by hig h resolution 1H MRS and automatic amino acid analysis by ionic exchange chr omatography. Intensities of 27 resonances and ratios of resonances were mea sured in the H-1 MRS spectra, and 17 amino acid concentrations were determi ned in the chromatograms. Linear discriminant analysis provided the most ad equate combination of these variables for binary classifications of a biops y between any two possible classes and in multiple choice comparisons, invo lving the eight possible classes considered. Correct diagnosis was obtained when the class selected by the computer matched the histological diagnosis . In binary comparisons, consideration of the amino acid profile increased the percentage of correct classifications, being always higher than 75% and reaching 100% in many cases. In multilateral comparisons, scores were: hig h-grade astrocytomas, 80%; low-grade astrocytomas, 74%; normal brain, 100%; medulloblastomas, 100%; meningiomas, 94.5%; metastases, 86%; neurinomas, 1 00%; and oligodendrogliomas, 75%. These results indicate that statistical m ultivariate procedures, combining H-1 MRS and amino acid analysis of tissue extracts, provide a valuable classifier for the nonhistological diagnosis of biopsies from brain tumors in vitro.