X-RAY FLUORESCENCE MEASUREMENT OF SKIN IRON CONCENTRATIONS AS A SURROGATE MARKER FOR LIVER IRON CONCENTRATIONS IN CADAVERS

Abstract

This thesis explores the use of skin iron concentration, assessed using a portable x- ray fluorescence (PXRF) analyzer, as an indicator of liver iron concentration. Ultrasound images were acquired from four sites on human cadavers: thenar eminence (TE), between the metacarpophalangeal joint 1 and 2 (MJ), patella (P), and plantar arch (PA). Skin thickness was subsequently measured using boundary detection software applied to the ultrasound images. Skin iron concentrations, from the same sites, were obtained directly from 13 cadavers using a PXRF analyzer. As the iron concentrations determined using PXRF may have been impacted due to beam attenuation through the skin, these iron concentrations were subsequently corrected using the Beer-Lambert Law accounting for the skin thickness at each site. Huber XRF is a bulk tissue measurement system that was used to measure iron concentrations from the extracted samples of the four sites. The skin iron concentrations determined using both PXRF and Huber XRF were compared to evaluate the agreement between the two methods. As such, samples from the four sites were extracted and the iron concentrations of the extracted samples were measured using a Huber X-Ray Fluorescence (XRF) analyzer as the control. A statistically significant relationship with a p-value of <0.01 and a strong correlation with a R2 value of 0.87 was seen between the corrected skin iron concentrations determined using PXRF and the skin iron concentrations determined using Huber XRF analysis of the extracted samples. Thus, the skin iron concentrations obtained using PXRF are representative of the skin iron concentrations collected by Huber XRF. For further validation, liver samples were obtained from 6 of the 13 cadavers and the liver iron concentrations were assessed using Huber XRF. The non- corrected PXRF, corrected PXRF and the Huber XRF skin iron concentrations were then compared with the liver iron concentrations. The relationships between iron concentrations measured at the PA site and in the liver were found to be statistically significant with strong correlations (R2 = 0.94 for Huber XRF, R2 = 0.87 for non-corrected PXRF, and R2 = 0.95 for corrected PXRF). Therefore, the skin iron concentrations at the PA site determined by PXRF and determined by Huber XRF analysis of skin samples could be used in the future as a surrogate marker for the liver iron concentrations in humans.