Determination of the in vitro limit of detection for pulmonary surfactant using proton magnetic resonance spectroscopy at 1.5T

Purpose To establish the lowest concentration of a solution of dipalmitoyl phosphatidylcholine which can be detected in vitro by proton magnetic resonance spectroscopy at 1.5T. Background Pulmonary surfactant is a phospholipid and protein mixture which is normally excreted by the fetal lungs from around 23 weeks gestation and plays a vital role after birth in reducing surface tension at the alveolar tissue-air interface. Lecithin (consisting of dipalmitoyl and other disaturated phosphatidylcholines) is the major active component of pulmonary surfactant. In cases where there is poor development of fetal lungs (pulmonary hypoplasia), or where there is a good chance a baby will be delivered prematurely, the ability to assess how well the lungs are likely to function after birth is of value for counselling parents and also in the planning of perinatal care. Fetal magnetic resonance (MR) imaging can be used to investigate lung volumes and thoracic anatomy in many causes of pulmonary hypoplasia, however there is at present no established method for quantitative MR assessment of fetal lung fluid composition. This study examines the technical feasibility of measuring the concentration of lecithin by proton magnetic resonance spectroscopy (H-MRS) using a 1.5T clinical imaging system. Methods and materials A 12.8mM solution of lecithin (Product P5911: D-L-α-dipalmitoyl phosphatidylcholine, Sigma-Aldrich Company Ltd, Gillingham UK) was prepared using chloroform as a solvent (Product 100776B: Chloroform AnalaR, VWR International Ltd, Lutterworth UK). Initial attempts to dissolve the compound in both water and saline were unsuccessful, providing only a transient suspension which was unsuitable for magnetic resonance spectroscopy. Further serial dilutions of lecithin solution were prepared at 9.7mM, 5.5mM, 5mM and 2.9mM concentrations, along with a blank reference sample of chloroform without lecithin. Point resolved spectroscopy (PRESS) sequences were used with an 8-channel head coil on a 1.5T clinical imaging system (GE HDx Signa: GE Healthcare, Slough UK). A range of sequence parameters were used to acquire spectra for each of the samples as follows: TE 144ms; TR 1500ms; single cubic voxels of linear dimension 1cm, 1.25cm and 1.5cm; and either 128, 256 or 512 spectral acquisitions. Signal-to-noise ratio (SNR) was measured for the characteristic peak attributed to the choline group of the lecithin, and these measurements were used to estimate a lower limit of detection for lecithin in chloroform solution. Results The largest spectral peak seen in all samples was attributed to chloroform and was assigned a chemical shift of 7.25ppm (see figure 1). Several other much smaller peaks were also seen in the blank chloroform sample however, and these were attributed to the presence of 1% ethanol used to stabilise laboratory-grade chloroform solvent. Samples containing lecithin displayed a characteristic spectral peak corresponding to the choline moiety at 3.2ppm and also a larger peak at 1.3ppm. The latter peak was attributed to (CH2)n groups within the saturated fatty acid side chains of lecithin. This peak was not used in subsequent analysis as this region of in vivo spectra is often difficult to interpret due to the presence of other broad non-specific lipid macromolecule peaks. The results of forty individual measurements of SNR for the choline peak of lecithin are shown in figure 2.