Quantification of 18F-Fluorocholine Kinetics in Patients with Prostate Cancer

Eline E. Verwer; Daniela E. Oprea-Lager; Alfons J.M. van den Eertwegh; Reindert J.A. van Moorselaar; Albert D. Windhorst; Lothar A. Schwarte; N. Harry Hendrikse; Robert C. Schuit; Otto S. Hoekstra; Adriaan A. Lammertsma; Ronald Boellaard

doi:10.2967/jnumed.114.148007

Article Figures & Data

Figures

Tables
Additional Files

Download figure
Open in new tab
Download powerpoint
FIGURE 1.
Typical example of PET/CT images acquired from patient diagnosed with PCa. Shown are low-dose CT (gray scale) fused with early PET image acquired from 35 to 40 s after injection displaying blood pool (A) and averaged image over 25–40 min after ¹⁸F-fluoromethylcholine injection (B) (color).
Download figure
Open in new tab
Download powerpoint
FIGURE 2.
Two typical examples of measured whole-blood time–activity curves for patient imaged over thorax region (A) and patient imaged over abdominal region (B). Lines = calibrated blood sampler data; dashed lines (image-derived) = aortic arch; dash-dot lines (image-derived) = abdominal descending aorta; triangles = manual arterial blood samples; squares = manual venous blood samples.
Download figure
Open in new tab
Download powerpoint
FIGURE 3.
(A) Typical measured time–activity curves for patient shown in Figure 1. ◱ = liver; ○ = lesion; ♢ = muscle; △ = fat tissue. (B) Nonlinear regression fits to lesion time–activity curve displayed in A, using various compartment models.
Download figure
Open in new tab
Download powerpoint
FIGURE 4.
K₁ obtained using 1T1k+V_B with calibrated (A) and noncalibrated (B) IDIF-derived, compared with BSIF method. Symbols indicate IDIF origin: ▲ = aortic arch; ◱ = descending aorta; + = femoral artery.
Download figure
Open in new tab
Download powerpoint
FIGURE 5.
Manual blood sample data as function of time: mean parent fractions (△ = arterial, ◱ = venous) (A) and ratio of venous to arterial parent fractions (B). Error bars represent ± SD.
Download figure
Open in new tab
Download powerpoint
FIGURE 6.
Correlation between simplified uptake measures and K₁ (1T1k+V_B). (A) SUV normalized to body weight (SUV_BW), (B) LMr and lesion activity concentration in kBq at 30–40 min after injection divided by AUC of parent plasma (C), and whole-blood activity concentration (D) in MBq over 0–30 min (SUV_AUC,PP and SUV_AUC,WB, respectively).

View popup

TABLE 1

Comparing Scan Durations

			Linear regression analysis^*
Parameter (model)	Scan duration (min)	Robustness (%)	Slope	Intercept	R²	ICC^*
K₁ (1T1k+V_B)	5	100	0.98	0.05	0.82	0.84
	10	100	0.98	0.03	0.94	0.95
	15	100	0.99	0.02	0.97	0.98
	20	100	1.01	0.01	0.99	0.99
	30	100	1.00	0.00	1.00	1.00
K_i (2T3k+V_B)	5	60	0.82	0.09	0.59	0.75
	10	65	0.93	0.03	0.97	0.98
	15	75	1.07	−0.02	0.94	0.97
	20	75	1.05	−0.01	0.90	0.95
	30	80	1.03	−0.01	1.00	1.00

↵* Results for comparison of quantification parameters from short scan durations to those derived from full 40-min dynamic scan.

View popup
TABLE 2
Comparing 1T1k+V_B with 2T3k+V_B
Linear regression analysis^*
Scan duration (min) Slope Intercept R² ICC^*
5 0.98 0.09 0.69 0.65
10 0.98 0.07 0.82 0.79
15 1.01 0.05 0.88 0.85
20 1.03 0.03 0.92 0.89
30 1.04 0.02 0.95 0.93
40 1.04 0.02 0.96 0.94
↵* Comparison of K₁ derived with 1T1k+V_B for various scan durations to K_i derived with 2T3k+V_B from full 40-min dynamic scan.

View popup

TABLE 3

Comparison of Simplified Parameters to K₁ Resulting from Full Kinetic Modeling with 1T1k+V_B

	Linear regression analysis
Simplified parameter	Slope	Intercept	R²
SUV_BW	13.53	1.18	0.30
SUV_BSA	315.39	27.53	0.33
SUV_LBM	9.02	0.86	0.32
SUV_BMI	4.65	0.20	0.34
SUV_IBW	10.96	1.28	0.31
LMr	22.54	−0.36	0.50
LBr	30.38	0.32	0.44
SUV_AUC,WB	5.88	−0.02	0.65
SUV_AUC,PP	14.73	−0.20	0.92

SUV_BW = SUV normalized to body weight; SUV_BSA = SUV normalized to body surface area; SUV_LBM = SUV normalized to lean body mass; SUV_BMI = SUV normalized to normalized body mass index; SUV_IBW = SUV normalized to ideal body weight.