To investigate whether using 3 Tesla (T) instead of 1.5T modifies the data obtained from first-pass perfusion in relation to the quantitative values of cerebral blood volume (CBV) and estimation of micro-vascular leakage (MVL). To describe the differences in data in the setting of neuro-oncology cases and propose explanations based on the discrepancies.

In total, 21 patients presenting an intracranial intra-axial space-occupying lesion underwent two MRI explorations, one at 1.5T and another at 3T, including a first-pass perfusion sequence using sequence parameters, defined by the manufacturer Philips. Using a gamma variate analysis, the ratio of cerebral blood volume (rCBV) in tumor, peritumoral, and normal appearing areas was first assessed. After a global analysis, a subgroup analysis was conducted according to the rCBV value measured at 1.5T. Lastly, MVL was assessed based on the signal intensity recorded above baseline after the passage of the contrast medium.

At 3T, compared to 1.5T data that are currently the reference, rCBV was constantly and significantly over-evaluated (P=0.0041 for all tumors), while MVL was constantly and significantly under-evaluated (P<0.0001 for all tumors).

The increase in magnetic field strength along with the associated modifications in sequence parameters led to variations in rCBV and MVL when measured using first-pass perfusion. In some cases, such as lymphomas, there was a loss of diagnostic information. It therefore appears necessary to optimize the acquisition parameters to allow for radiologic semiology to become relevant again.