Gd-EOB-DTPA-enhanced MR guidance in thermal ablation of liver malignancies

Abstract

Objective: To evaluate the potency of Gd-EOB-DTPA to support hepatic catheter placement in laser ablation procedures by quantifying time-dependent delineation effects for instrumentation and target tumor within liver parenchyma. Monitoring potential influence on online MR thermometry during the ablation procedure is a secondary aim.

Materials and methods: 30 cases of MR-guided laser ablation were performed after i.v. bolus injection of gadoxetic acid (0.025 mmol/Kg Gd-EOB-DTPA; Bayer Healthcare, Berlin, Germany). T1-weighted GRE sequences were used for applicator guidance (FLASH 3D) in the catheter placement phase and for therapy monitoring (FLASH 2D) in the therapy phase. SNR and consecutive CNR values were measured for elements of interest plotted over time both for catheter placement and therapy phase and compared with a non-contrast control group of 19 earlier cases. Statistical analysis was realized using the paired Wilcoxon test.

Results: Sustainable signal elevation of liver parenchyma in the contrast-enhanced group was sufficient to silhouette both target tumor and applicator against the liver. Differences in time dependent CNR alteration were highly significant between contrast-enhanced and non-contrast interventions for parenchyma and target on the one hand (p = 0.020) and parenchyma and instrument on the other hand (p = 0.002). Effects lasted for the whole procedure (monitoring up to 60 min) and were specific for the contrast-enhanced group. Contrasting maxima were seen after median 30 (applicator) and 38 (tumor) minutes, in the potential core time of a multineedle procedure. Contrast influence on T1 thermometry for real-time monitoring of thermal impact was not significant (p = 0.068-0.715).

Conclusion: Results strongly support anticipated promotive effects of Gd-EOB-DTPA for MR-guided percutaneous liver interventions by proving and quantifying the delineating effects for therapy-relevant elements in the procedure. Time benefit, cost effectiveness and oncologic outcome of the described beneficiary effects will have to be part of further investigations.

Figure 1. Comparative signal alteration for liver parenchyma and target tumors in group 1 (Gd-EOB-DTPA) and group 2 (non-contrast).

Signal-to-noise ratios (SNR, delta values) are displayed as a function of time after bolus injection of Gd-EOB-DTPA (0.025 mmol/Kg i.v.) in group 1 and start time in group 2. Standardized curves are compensating 5^th grade polynomial functions. A rapid contrast uptake in group 1 (G1) leads to an enduring signal elevation in liver parenchyma (P) during catheter placement phase (–▪–). Values in group 2 (G2) are oscillating around zero as non-contrast signals are not significantly altered (–•–). Contrast-induced tumor (T) signal increase appears in the early extracellular phase (<10 minutes), while the curve (–□–) is approximating zero and non-contrast values (–○–) thereafter. As a consequence the graph displays a favorable gap between P^G1 and T^G1 on the one hand and P^G1 and both P^G2 + T^G2 on the other hand.

Figure 2. Liver-to-tumor contrast in dependence of Gd-EOB-DTPA application on the one hand and absence or presence of early phase tumor enhancement on the other hand.

Contrast-to-noise ratios (CNR, delta values) are displayed in the same manor as SNR curves in figure 1. While sharing the favorable increase in contrast between parenchyma (P) and target tumor (T) at late phase, enhancing (Tpos) and non-enhancing (Tneg) tumors in group 1 (G1) show different curve progressions in the early phase after bolus contrast. Reversion from negative to positive delta CNR values (–♦–) and traversing an early equilibrium is characteristic for enhancing tumors (Tpos^G1) in the given setting.

Figure 3. Liver-to-tumor contrast promotes planning and placement of applicators.

Dynamic imaging (breath-hold T1-weighted GRE FLASH 3D sequences) of both non-enhancing Tneg^G1 (A) and enhancing Tpos^G1 (B) during planning and catheter placement phase in two different cases. Column A shows a sustainable high-low contrast between liver parenchyma and target tumor (arrow), whereas column B shows a transient low-high contrast between parenchyma and tumor (arrow). Advantageous target outlining at late phase initiates catheter (arrow heads) approach in both cases.

Figure 4. Tissue-to-instrument contrast alteration in dependence of Gd-EOB-DTPA application.

The graph shows a superior potency of cellular Gd-EOB-DTPA uptake to silhouette a metallic applicator (A) against a parenchymal (P) background (–▪–), as compared with non-contrast image guiding (–•–) in group 2 (G2).

Figure 5. Liver-to-instrument contrast visualizes the procedure.

Dynamic imaging (breath-hold T1-weighted GRE FLASH 3D sequences) confirms the advantageous properties of late phase Gd-EOB-DTPA uptake showing cases of non-contrast (group 2, A-D) and contrast-enhanced guidance (group 1, E-H). Four different cases each, demonstrate multiple applicator (arrow heads) placement approaching different target tumors (arrows). Image G shows a fluid collection (▪) consistent with protective saline injection towards the stomach and prior to the ablation.

Figure 6. Comparative signal and contrast alterations for liver parenchyma and thermal impact zone in group 1 (Gd-EOB-DTPA) and group 2 (non-contrast) during therapy phase.

Ratios (SNR+CNR, delta values) are displayed as a function of time after technically starting an ablation cycle in group 1 (Gd-EOB-DTPA) and group 2 (non-contrast). Standardized curves are compensating 4^th grade polynomial functions. Graphs show homogeneous decrease of SNR values within the thermal impact or ablation zones (AZ), consistent with temperature-induced T1 signal drop, in both groups 1+2 (G1, G2). At the same time CNR values (parenchyma P vs. AZ) increase for both groups in a parallel manor. It seams, thermometry monitoring is not influenced by Gd-EOB-DTPA application.

Figure 7. MR thermometry at late phase contrast-enhanced imaging.

Thermometric MR imaging (T1 FLASH 2D) delineates the development of a T1-hypointense zone of thermal impact (arrowheads show boundaries) encountering a 2 cm hepatic metastasis (double-headed arrow). Notice that metallic applicator mandrins have been replaced by non-visible laser fibers.