Detection of multiple sclerosis from exhaled breath using bilayers of polycyclic aromatic hydrocarbons and single-wall carbon nanotubes

Abstract

A cross-reactive array of polycyclic aromatic hydrocarbons and single wall carbon nanotube bilayers was designed for the detection of volatile organic compounds (tentatively, hexanal and 5-methyl-undecane) that identify the presence of disease in the exhaled breath of patients with multiple sclerosis. The sensors showed excellent discrimination between hexanal, 5-methyl-undecane, and other confounding volatile organic compounds. Results obtained from a clinical study consisting of 51 volunteers showed that the sensors could discriminate between multiple sclerosis and healthy states from exhaled breath samples with 85.3% sensitivity, 70.6% specificity, and 80.4% accuracy. These results open new frontiers in the development of a fast, noninvasive, and inexpensive medical diagnostic tool for the detection and identification of multiple sclerosis. The results could serve also as a launching pad for the discrimination between different subphases or stages of multiple sclerosis as well as for the identification of multiple sclerosis patients who would respond well to immunotherapy.

Keywords: Multiple sclerosis; biomarker; breath; carbon nanotube; diagnosis; polycyclic aromatic hydrocarbon.

Figures

Figure 1

Discotic PAH derivatives employed in the study.

Figure 2

Field emission scanning electron microscopy (FE-SEM) images of the sensing films made of PAH derivatives on top of random networks of SWCNTs: (a) PAH-1/SWCNT; (b) PAH-2/SWCNT; (c) PAH-3/SWCNT; and (d) PAH-4/SWCNT. As seen in the figure, the PAH derivatives were assembled in different nano- or microstructures, such as nanoropes (PAH-1), microflowers (PAH-2), nanocables with a winding pattern orientation (PAH-3), or nanocables with quasi-uniform orientation (PAH-4).

Figure 3

Normalized change of sensor resistance at the middle of the exposure (S1) calculated from sensor responses to the 10 VOCs and water molecules analyzed at pa/po = 0.5 (see Methods section).

Figure 4

Nine-class DFA analysis performed with the features extracted from three PAH/SWCNT sensors (PAH-1/SWCNT, PAH-3/SWCNT, and PAH-4/SWCNT). A unique class was considered for the nonpolar VOCs because of the similar responses shown by the sensors to them (see Figure 3). For each analyte, three measurements were used for calibration (plotted using the symbols from figure’s legend) and one for testing (plotted using star symbols).