High signal in the adenohypophysis on T1-weighted images presumably due to manganese deposits in patients on long-term parenteral nutrition.

Hypermanganesaemia is reported in patients on long-term parenteral nutrition. Deposition of manganese, giving high signal on T1-weighted images, may involve the basal ganglia. MRI in nine patients (mean age 51 years, range 31-75 years) on long-term parenteral nutrition (mean duration 30 months, range 6-126 months), demonstrated high signal in the anterior pituitary gland on T1-weighted sagittal and coronal images. The gland appeared normal on T2-weighted images. Signal intensity in the basal ganglia on T1-weighted images was increased in all patients. Endocrine assessment showed no significant abnormality. Neurological examination showed a mild parkinsonian movement disorder in one patient. Hypermanganaesemia was present in all nine (1.3-2.8 micromol/l, mean 1.87 micromol/l). The high signal in the anterior pituitary gland was probably related to deposition of paramagnetic substances, especially manganese.

[Spinal cord MRI in multiple sclerosis: comparison of turbo-spin echo sequences and turbo-MRI]. / IRM médullaire dans la sclérose en plaques: comparaison des séquences turbo-spin écho et turbo-IRM.

UNLABELLED: To compare T2-weighted turbo spin-echo (turboSE) and turbo inversion recovery magnitude (turboIRM) sequences in the detection of multiple sclerosis of the spinal cord. METHODS: 28 patients with clinically confirmed multiple sclerosis and signs of cervical cord involvement were examined on a 1.5 tesla MR system. Sagittal T2-weighted turboSE sequences were followed by sagittal turboIRM sequences. 24 patients also underwent a cranial study. RESULTS: Using turboIRM sequence, all the patients showed at least one spinal cord plaque with a total of 91 plaques. Using turboSE sequence, 61 lesions were identified in 25 patients. Spinal plaques were seen best on turboIRM sequences with a higher contrast between the lesion and the normal spinal cord. In 10 patients with cord plaques in both sequences, the cranial study was normal or non conclusive. CONCLUSIONS: The turboIRM sequences detect more cord lesions and with a better contrast than the turboSE sequence. This results confirm those described in other studies and demonstrate the inadequacy of the fast spin echo sequences. Assessment of spinal plaques is particularly important when MR findings of the brain are negative or non conclusive.