Detecting colorectal cancer by 1H magnetic resonance spectroscopy of fecal extracts.

Colorectal cancer is one of the most common cancers in the western world. Its early detection has been found to improve the prognosis of the patient, providing a wide window of opportunity for successful therapeutic interventions. However, current diagnostic techniques all have some limitations; there is a need to develop a better technique for routine screening purposes. We present a new methodology based on magnetic resonance spectroscopy of fecal extracts for the non-invasive detection of colorectal cancer. Five hundred twenty-three human subjects (412 with no colonic neoplasia and 111 with colorectal cancer, who were scheduled for colonoscopy or surgery) were recruited to donate a single sample of stool. One-dimensional (1)H magnetic resonance spectroscopy (MRS) experiments were performed on the supernatant of aqueous dispersions of the stool samples. Using a statistical classification strategy, several multivariate classifiers were developed. Applying the preprocessing, feature selection and classifier development stages of the Statistical Classification Strategy led to approximately 87% average balanced sensitivity and specificity for both training and monitoring sets, improving to approximately 92% when only crisp results, i.e. class assignment probabilities > or =75%, are considered. These results indicate that (1)H magnetic resonance spectroscopy of human fecal extracts, combined with appropriate data analysis methodology, has the potential to detect colorectal neoplasia accurately and reliably, and could be a useful addition to the current screening tools.

Relation of lipoprotein subclasses as measured by proton nuclear magnetic resonance spectroscopy to coronary artery disease.

Although each of the major lipoprotein fractions is composed of various subclasses that may differ in atherogenicity, the importance of this heterogeneity has been difficult to ascertain owing to the labor-intensive nature of subclass measurement methods. We have recently developed a procedure, using proton nuclear magnetic resonance (NMR) spectroscopy, to simultaneously quantify levels of subclasses of very low density (VLDL), low density (LDL), and high density (HDL) lipoproteins; subclass distributions determined with this method agree well with those derived by gradient gel electrophoresis. The objective of the current study of 158 men was to examine whether NMR-derived lipoprotein subclass levels improve the prediction of arteriographically documented coronary artery disease (CAD) when levels of lipids and lipoproteins are known. We found that a global measure of CAD severity was positively associated with levels of large VLDL and small HDL particles and inversely associated with intermediate size HDL particles; these associations were independent of age and standard lipid measurements. At comparable lipid and lipoprotein levels, for example, men with relatively high (higher than the median) levels of either small HDL or large VLDL particles were three to four times more likely to have extensive CAD than were the other men; the 27 men with high levels of both large VLDL and small HDL were 15 times more likely to have extensive CAD than were men with low levels. In contrast, adjustment for levels of triglycerides or HDL cholesterol greatly reduced the relation of small LDL particles to CAD. These findings suggest that large VLDL and small HDL particles may play important roles in the development of occlusive disease and that their measurement, which is not possible with routine lipid testing, may lead to more accurate risk assessment.

Development of a proton nuclear magnetic resonance spectroscopic method for determining plasma lipoprotein concentrations and subspecies distributions from a single, rapid measurement.

We are developing a method for quantifying plasma lipoproteins by proton nuclear magnetic resonance (NMR) spectroscopy that offers advantages over existing clinical methods. We showed that the major lipoproteins have distinct NMR properties sufficient to permit their concentrations to be extracted from a computer lineshape analysis of the plasma lipid methyl resonance envelope (Clin Chem 1991; 37:377-86). We have now discovered that the spectra of the subspecies within each lipoprotein class are different enough to influence the composite spectrum of that class and hence the spectrum of whole plasma. By including spectra representative of these subspecies as additional components in the lineshape-fitting analysis, a complete concentration profile of very-low-density, low-density (LDL), and high-density (HDL) lipoproteins plus the subspecies distributions within these classes can be simultaneously generated. A pilot study of 30 plasma samples of widely varied composition demonstrated good agreement between NMR-derived values and lipoprotein lipid concentrations and LDL and HDL subspecies distributions determined by gradient-gel electrophoresis.

Relationships between the proton nuclear magnetic resonance properties of plasma lipoproteins and cancer.

We conducted a comprehensive investigation of the origin of nuclear magnetic resonance (NMR) lineshape variability of plasma lipids among healthy individuals and those with cancer. The methyl and methylene resonances of lipid in human plasma, whose linewidths have been reported to correlate with the presence of malignancy, are composed of the overlapping resonances of "mobile" protons from the major lipoproteins (very-low-, low-, and high-density lipoproteins). We tested two hypotheses for the origin of the narrower plasma linewidths observed for cancer patients: (a) malignancy-associated differences in the spectral properties (chemical shift, lineshape) of one or more of the lipoproteins, and (b) differences in the fraction of lipoprotein lipid giving rise to detectable NMR signal. Analysis of the concentrations of lipoprotein lipid and of 500 MHz NMR spectra of the lipoprotein constituents in greater than 100 plasma samples failed to provide support for either hypothesis. Although linewidths were found to be significantly narrower for the cancer group, the difference is entirely attributable to differences in the concentrations of the lipoproteins.

Quantification of plasma lipoproteins by proton nuclear magnetic resonance spectroscopy.

A new analytical procedure for quantifying plasma lipoproteins by proton nuclear magnetic resonance (NMR) spectroscopy has been developed that potentially offers significant advantages over existing clinical methods used for assessing risk of coronary heart disease. Analysis of a single spectrum of a nonfasting plasma sample, acquired simply and rapidly at moderate magnetic field strength (250 MHz), yields a complete profile of lipoprotein concentrations: chylomicrons and very-low-, low-, and high-density lipoproteins. The method is based on curve-fitting (spectral deconvolution) of the plasma methyl lipid resonance envelope, the amplitude and shape of which depend directly on the amplitudes of the superimposed methyl resonances of the lipoprotein components. A linear least-squares curve-fitting algorithm was developed to efficiently extract the signal amplitudes (concentrations) of the lipoproteins from the plasma spectrum. These signal amplitudes correlate well with lipoprotein concentrations determined by triglyceride and cholesterol measurements.