Thomas F Heston MD, Presenter: Nothing to Disclose

Richard L. Wahl MD, Abstract Co-Author: Research grant, General Electric Company License agreement, Naviscan PET Systems, Inc License agreement, GlaxoSmithKline plc License agreement, Cell Therapeutics, Inc Scientific Advisor, Cellectar, LLC Scientific Advisor, Actinium Pharmaceuticals, Inc Consultant, Cellectar, LLC Consultant, Nihon Medi-Physics Co, Ltd Stockholder, Threshold Pharmaceuticals, Inc

The measurement of tumor metabolism is commonly expressed in terms of its maximum standardized uptake value (SUV). However, the measured SUV over time can change because of non-metabolic variables such as time from tracer injection to scan time or recent diet. Lesion to hepatic SUV ratios have the possibility of removing some of this variation, but this requires that the baseline be highly reproducible. We hypothesized that due to attenuation correction issues, a central hepatic SUV would be different than a peripheral hepatic SUV, and that a standardized measurement must take this into account.

Two separate measurements of the average SUV of the liver were obtained. The first was obtained peripherally in the liver, near the abdominal wall but well within liver parenchyma. The second was obtained of the more proximal liver parenchyma in relation to the porta hepatis. SUV corrected for lean body mass was used. Spherical regions of interest (ROI) were utilized. Hounsfield Units (HU) were obtained simultaneously with the SUV's. A 2-tailed matched t-test was performed comparing the distal versus proximal SUV. Results reported in terms of mean +/- standard deviation.

Ninety two patients were evaluated. The proximal SUV was 1.37 +/- 0.29 and the peripheral SUV was 1.40 +/- 0.30 (p = 0.018). The 95% confidence interval for the difference was 0.006 to 0.60. The proximal HU was 54.0 +/- 12.2 and the peripheral HU 56.14 +/- 13.8 (p=0.001), with the 95% confidence interval of the difference 0.904 to 3.376.

The hepatic SUV is significantly higher statistically in peripheral compared to central regions. Although the difference is small, standardization of the hepatic SUV for research trials and clinical purposes should take this difference into consideration. This finding may be due to subtle differences in vascular and duct volumes throughout the liver as evidenced by the proximal vs distal HU values.

Standardization of a patient's baseline hepatic SUV requires that the measurement be obtained at a consistent location in relationship to the porta hepatis.

Heston, T, Wahl, R, Reliability of the FDG PET Liver Standardized Uptake Value Depends Upon Depth of Measurement.  Radiological Society of North America 2009 Scientific Assembly and Annual Meeting, November 29 - December 4, 2009 ,Chicago IL. http://archive.rsna.org/2009/8016734.html