Magnetization transfer contrast imaging of hepatic neoplasms.

A method of performing magnetization transfer contrast (MTC) using a pulse sequence based on a series of on-resonance binomial pulses preceding a conventional spin-echo sequence has been recently described. We investigated this technique in the evaluation of circumscribed hepatic neoplasms on a 0.5 T imager. Conventional spin-echo imaging was performed in 18 patients with hepatic neoplasms, 15 with primary or metastatic hepatic neoplasms, and 3 with cavernous hemangiomas. The T1-weighted sequence was repeated following MTC saturation in 7 patients and the proton density sequence was repeated following MTC saturation in 13 patients (both were used in two patients). The mean lesion-to-liver contrast for neoplasms worsened with T1-MTC (0.87 vs. 1.04), but improved slightly with proton density-MTC (1.10 vs. 1.22) when compared to the corresponding non-MTC spin-echo images. However, these differences were not statistically significant (p = .19 and p = .16, respectively). The limited data on cavernous hemangiomas does not suggest these have consistently different MTC imaging characteristics from other hepatic neoplasms. In conclusion, MTC imaging using a binomial saturation pulse did not offer significant improvement in contrast between hepatic neoplasms and normal liver. Refinements of the binomial pulse sequence will be necessary if a consistent improvement in lesion sensitivity is to be achieved. Evaluation of its role in increasing diagnostic specificity will require further investigation.