SPAMM-tagging MR sequence assesses endovascular repair

A cine phase-contrast MR sequence coupled with a spatial magnetization modulation sequence is proving to be a potent combination for evaluating aortic endoleaks, according to Stanford University researchers.

Spatial magnetization modulation, or SPAMM-tagging, applies a triple prescan sequence that essentially superimposes a grid, tags a tissue, and allows qualitative assessment on global and regional aneurysm wall motion. Phase-contrast MR imaging provides determination of endoleak direction as well as characterization of endoleaks as high or low velocity.

"Using both phase-contrast MR and SPAMM-tagging, qualitatively, we have observed correlations between endoleaks and aneurysm sac wall motion," said Dr. Jeffrey C. Hellinger, an assistant professor of radiology at Stanford.

To assess wall motion, a region of interest is drawn on the periphery of the aneurysm sac on the SPAMM image and correlated with the phase-contrast image. The percentage of area change determines the presence of an endoleak. One example showed an area change of 3.9%, which correlated with the presence of an endoleak, Hellinger said on Saturday at the European Congress of Radiology.

The Stanford group evaluated 38 patients (seven thoracic, 31 abdominal) using this protocol. Eight have undergone follow-up, resulting in 46 exams. Most patients were imaged in the immediate postoperative course, the rest after 12 months.

Eighteen cases had endoleaks, mostly type 2; 16 of these had CT angiography correlation and 100% agreement with the endoleaks; 17 had ultrasound correlation. The acoustic window limited five ultrasound exams, three failed to reveal endoleaks, and one suggested an endoleak but was negative by both MR angiography and digital subtraction angiography.

All MR sequences - T1-weighted, angiographic, phase-contrast, and SPAMM-tagging - were technically adequate and interpretable. Flow analysis data, despite the presence of a stent-graft, were also technically adequate.

The sequences are fairly standard, Hellinger said, except the MRA repetition time is 3 to 4 msec and the echo time is less than 1 msec. SPAMM-tagging was applied with a steady-state free precession sequence with one slight modification: the flip angle was about 45°. The other additional parameter was a superior to inferior saturation band. The additional sequences added about 30 minutes to a 40-minute examination.

Phase-contrast MR was applied with multiple velocity encoding values to acquire in-plane and through-plane velocities for the aneurysm sac lumen and wall. For flow and endoleaks, the phase-contrast sequence was acquired with higher in-plane (25 cm/sec) and lower through-plane (150 cm/sec) velocity encoding values.

For wall motion, a phase-contrast sequence of low in-plane velocity encoding values (5 cm/sec) and high through-plane velocity encoding values (800 cm/sec) was used to improve velocity resolution and minimize flow artifacts.

Hellinger concluded that cine phase-contrast MR and SPAMM-tagging are promising techniques for aneurysm surveillance following endovascular repair. Although CTA is considered the gold standard for this application, it's a static three-phase acquisition that looks only at morphology and can be limited by streak and motion artifact. Ultrasound conveys both morphologic and physiologic information but is operator-dependent and has a limited window.

"The additional information regarding aneurysm sac wall motion gained from combining the two applications might afford an opportunity to impact patient therapy and outcome, as well as stent-graft design," he said.

Future research will try to determine with MR imaging whether an endoleak is type 1 or 2.