Single time point quantitation of cerebral glucose metabolism by FDG-PET without arterial sampling

Cumming, Paul; Dias, André H.; Gormsen, Lars C.; Hansen, Allan K.; Alberts, Ian; Rominger, Axel; Munk, Ole L.; Sari, Hasan

doi:10.1186/s13550-023-01049-3

EJNMMI Research

Table 3 The population mean (SD) values for net influx of FDG (K_i; ml g⁻¹ min⁻¹) from blood to brain by standard IDIF method and four simplified methods

From: Single time point quantitation of cerebral glucose metabolism by FDG-PET without arterial sampling

	Aarhus (n = 52)		Bern (n = 24)
	GM	WM	GM	WM
Standard IDIF K_i (ml g⁻¹ min⁻¹) V_D (ml g⁻¹)	0.0350 ± 0.0086 0.55 ± 0.19	0.0143 ± 0.0099 0.51 ± 0.13	0.0345 ± 0.0091 0.55 ± 0.15	0.0153 ± 0.0047 0.45 ± 0.10
Method 1 K_i (ml g⁻¹ min⁻¹) %error	0.0336 ± 0.0077 − 3.5 ± 7.2% r = 0.93**	0.0119 ± 0.0026 − 16.7 ± 20.2% r = 0.44**	0.0354 ± 0.0099 + 2.5 ± 11.3% r = 0.86**	0.0135 ± 0.0040 − 11.6 ± 13.1% r = 0.94**
Method 2 K_i (ml g⁻¹ min⁻¹) %error	0.0338 ± 0.0090 − 3.7 ± 7.2% r = 0.71**	0.0119 ± 0.0031 − 16.7 ± 19.2% r = 0.58**	0.0333 ± 0.0093 − 3.4 ± 10.6% r = 0.86**	0.0127 ± 0.0037 − 16.7 ± 12.3% r = 0.94**
Method 3 K_i (ml g⁻¹ min⁻¹) %error	0.0352 ± 0.0118 + 0.7 ± 25.1% r = 0.61**	0.0123 ± 0.0044 − 13.9 ± 31% r = 0.33*	0.0346 ± 0.0098 + 0.4 ± 12% r = 0.88**	0.0132 ± 0.0040 − 13.6 ± 11.7% r = 0.97**
Method 4 K_i (ml g⁻¹ min⁻¹) %error	0.0348 ± 0.0090 − 0.7 ± 7.3% r = 0.68**	0.0124 ± 0.0032 − 13.2 ± 8.1% r = 0.36*	0.0346 ± 0.0098 + 0.2 ± 11.4% r = 0.86**	0.0132 ± 0.0039 − 13.6 ± 12.9% r = 0.94**

We present population mean (SD) values for net influx of FDG (K_i; ml g⁻¹ min⁻¹) from blood to the entire gray matter (GM) and white matter (WM) using data from Aarhus (n = 52) and Bern (n = 24), along with the error relative to the complete IDIF results, and the corresponding ordinate intercepts (V_D; ml g⁻¹). We tested four simplified methods based on pharmacokinetic findings from Table 2 with respect to their mean K_i values, the percentage error relative to standard IDIF values, calculated as [(Method-IDIF)/IDIF], where the IDIF method entails linear graphic analysis relative to the brain radioactivity concentrations measured during the interval 12–67 min. In all four simplified methods, the mean GM V_D estimate from the IDIF regression (0.55 ml g⁻¹) served to constrain the estimation of K_i. Method 1: From semi-logarithmic analysis of the FDG blood concentration during the interval 35–67 min, we calculated λ3 (min⁻¹), the fractional rate constant for the elimination of FDG from arterial blood. The AUC extrapolated to the time of first appearance of the radioactivity peak corresponded to a mean of 85% of the total measured AUC (Aarhus) or 84% (Bern). Using these correction factors, we estimated the complete AUC, and calculated the terminal theta_{(52–67 min)} relative to the terminal blood concentration (Ca_{(52–67 min)}. We then estimated K_i as a two point Patlak plot defined by gold standard population ordinate intercept (V_D; ml g⁻¹) and the tracer distribution volume calculated from the mean radioactivity concentrations measured in GM, WM, and blood during the final three frames (52–67 min) of the PET recording. Method 2: From the site-specific empirical relationships between AUC_(0–67) as a function of Ca_{(52–67 min)}, we similarly calculated K_i from Ca_{(52–67 min)} as a two point Patlak plot. Method 3: We similarly used the site-specific empirical relationships between theta_{(52–67 min)} and λ3 measured during the interval 35–67 min post injection. Method 4: We similarly used the site-specific empirical relationships between theta_{(52–67 min)} and subject age in the two populations. Significance of Pearson correlation coefficients: (*) p < 0.05; (**) p < 0.001

Back to article page