1350733 - Accelerated T1 Reactivity Mapping Under Dobutamine Stress with Ferumoxytol-enhanced MRI: A Feasibility Study in Swine

Myocardial T1 reactivity, defined as the relative T1 change from rest to stress, has been proposed as a marker for detection of ischemic heart disease. SASHA-based T1 mapping provides higher accuracy than MOLLI, due to its robustness to heart rate, T2, and MT effects.^1-3 However, for T1 reactivity studies under exercise or dobutamine stress, the scan time needed for SASHA can be prohibitive. Recent work has shown the feasibility of accelerating SASHA by using a reduced number of T1-weighted (T1w) images when the SNR is boosted using denoising.⁴ An alternative can be to boost the signal by using low-dose ferumoxytol-enhanced (FE) protocols given the desirable properties of ferumoxytol for T1 reactivity studies.⁵ In this work, we investigated the feasibility of accelerated SASHA T1 mapping for FE dobutamine-stress T1 reactivity studies at 3T.

Methods: Based on prior work by Kellman et al.,⁶ we used a SASHA pulse sequence with optimized sampling scheme, i.e. fixed saturation recovery time, with 2-parameter fitting. We conducted phantom studies with ferumoxytol-doped agar (Fig 1a) to determine the optimum number, N, of T1w images to use for accelerated SASHA mapping with acceptable precision. We compared the precision of accelerated SASHA T1 maps for all feasible choices of N.  We used the coefficient of variation (CoV) defined as “standard deviation over mean T1” as a measure of precision⁶ and determined the lowest N (highest acceleration) that can achieve CoV < 1%. We conducted stress/rest FE studies in 6 healthy pigs under a wide range of dobutamine-induced stress heart rates (to mimic exercise stress) and performed T1 reactivity analysis (2 myocardial ROIs per animal) for conventional SASHA and our proposed accelerated approach (with retrospective subsampling of SASHA-acquired T1w images).


Results:

The FE phantom experiments showed that, to achieve CoV< 1%, we need a minimum of N=5 (Fig 1b), implying a 2-fold acceleration compared to conventional SASHA (which uses N=10). Fig 2 shows example SASHA maps for native rest and FE rest/stress scans in 2 representative animals. As highlighted with arrows, FE stress T1 values are lower than FE rest indicating increased intravascular space (higher myocardial blood volume) during dobutamine stress. Fig 3 compares the 2-fold accelerated SASHA maps vs. conventional SASHA in terms of rest/stress T1 maps in (a) and T1 reactivities in (b). Both the error maps in Fig 3(a) and the correlation plot in Fig 3(b) show excellent agreement.

Conclusion:

In this work, we showed the feasibility of accelerating SASHA T1 mapping in FE rest/stress T1 reactivity studies under dobutamine stress. Our initial results suggest that, in FE stress studies, the scan time can be accelerated without a noticeable deterioration in the accuracy of derived T1 maps or T1 reactivity values when compared to conventional SASHA. A 2-fold acceleration has the potential to improve the feasibility of SASHA-based T1 reactivity studies under dobutamine/exercise stress.