Quantitative assessment of Gd-DTPA contrast agent from signal enhancement: an in-vitro study

Quantitative determination of in-vivo gadolinium diethylenetriamine-pentaacid (Gd-DTPA) concentration is attractive in various studies involving perfusion, tracer kinetics and permeability constants. Using a 1.5 T clinical system and a 7 T small-bore system, we evaluated a method for absolute determination of Gd-DTPA concentrations in plasma solutions. Different solutions of Gd-DTPA and (99m)Tc-DTPA were mixed in human plasma and concentrations in the range of 0-5.0 mmol/l (1.5 T system) or 0-3.0 mmol/l (7 T system) of Gd-DTPA were divided into thirteen tubes. All MRI measurements were carried out using conventional sequences (SE, FLASH and GRASS). The MR measured intensity was converted to Gd-DTPA concentration by mathematical interpretation of the sequences. All MRI sequences showed, that the measured concentrations of Gd-DTPA revealed a slight non-linear difference compared with the calculated Gd-DTPA concentrations determined by the plasma (99m)Tc-DTPA using gamma counting. This non-linearity was most pronounced at high Gd-DTPA concentrations, suggesting that the discrepancy could be a result of an increased plasma relaxivity at higher concentrations. Adjustment of measured Gd-DTPA concentration was therefore performed using a selected power function, A[Gd-DTPA](a), which yielded the best linear relationship. Regression analysis showed that the scaling constant (A) varied from 0.11 to 97.45 and the power constant (a) varied from 0.83 to 1.6. Based on these constants, the MRI measured concentrations of Gd-DTPA did not differ from the calculated concentrations of Gd-DTPA obtained from reference measurements of (99m)Tc-DTPA. In the 1.5 T system, a linear relationship (r(2) > or = 0.95) was demonstrated in the range of 0-5.0 mmol/l Gd-DTPA, and in the 7 T system, a linear relationship (r(2) > or = 0.92) was demonstrated in the range of 0-3.0 mmol/l Gd-DTPA. Additionally, the effect of signal-to-noise on measured concentrations of Gd-DTPA was simulated using MR data of the mixed solutions of Gd-DTPA in plasma and the analytical expression of the pulse sequences. The simulations showed that the concentrations were most sensitive to noise in the GRASS sequence. In conclusion, this study demonstrates a novel approach to quantify accurately the Gd-DTPA concentration directly from MRI signal data using different routine sequences.