Follow-up of intracranial aneurysms treated with detachable coils: comparison of plain radiographs, 3D time-of-flight MRA and digital subtraction angiography.

All patients with aneurysms treated with Guglielmi detachable coils (GDC) are undergo angiography to assess long-term stability of aneurysm exclusion or to show recurrence of the aneurysm sac, which may require further treatment. We prospectively compared the plain-film appearance of the coil-mass, 3D time-of-flight MR angiography (TOF MRA) and digital subtraction angiography (DSA) for the detection of aneurysm recanalisation during follow-up. We studied 60 patients with 74 intracranial aneurysms treated with Guglielmi detachable coils. We used the unsubtracted image of the angiograms performed at the completion of any embolisation procedure and at follow-up as the plain radiographs. Recanalisation was considered if loosening, compaction or reorientation of the coil mass was apparent. TOF MRA was performed to assess the presence and size of a neck remnant. DSA was regarded as the definitive investigation. Comparison of the techniques showed good agreement as regards aneurysm recanalisation. MRA was more accurate than plain radiography and could replace DSA for long term follow- up. The initial follow-up examination should, however, include both modalities. In cases of contraindications or limitations to MRA, the interval between follow-up angiographic examinations could be increased if there is no change in the plain-film coil-mass appearances.

Prospective evaluation of time-of-flight MR angiography in the follow-up of intracranial saccular aneurysms treated with Guglielmi detachable coils.

PURPOSE: The purpose of our study was to prospectively evaluate 3D time-of-flight (TOF) MR angiography (MRA) in the follow-up of 27 intracranial aneurysms treated with Guglielmi detachable coils (GDCs). METHOD: From February 1997 to June 1998, 26 patients with 27 aneurysms were included in this prospective study. Aneurysms were located in the anterior circulation in 23 cases and in the posterior circulation in 4 cases. All patients underwent 3D TOF MRA and digital subtraction angiography (DSA) in the same week within 4 months after aneurysmal treatment with GDCs. No clinical events occurred during the follow-up. We analyzed residual flow within the coil mass and within the aneurysmal neck and the patency of the parent and adjacent arteries on MRA and DSA. MRA analysis was based upon MIPPED and source images. DSA was our gold standard. RESULTS: In all cases, the quality of MRA was good enough to be informative. In aneurysmal analysis, the sensitivity, specificity, positive predictive value, and negative predictive value of MRA were, respectively, 80, 100, 100, and 96% to diagnose residual flow within the coil mass (one false-negative case) and 83, 100, 100, and 95.5% to diagnose residual flow within the aneurysmal neck (one false-negative case). In arterial analysis, sensitivity and positive predictive value of MRA were 89 and 100% to diagnose patency of the parent artery (three false-negative cases) and 83 and 100% to diagnose patency of adjacent arteries (seven false-negative cases). CONCLUSION: In the follow-up of intracranial aneurysms treated with GDCs, 3D TOF MRA could be used as a screening test to select patients that should undergo DSA and thus could improve patient follow-up in terms of risk-benefit.

Failure to demonstrate a direct caroticocavernous fistula due to arterialisation of the ipsilateral cavernous sinus.

We report a woman in whom a small direct caroticocavernous fistula (DCCF) was revealed after successful transarterial occlusion of a contralateral fistula which drained into both cavernous sinuses. We underline that a second smaller fistula can be masked by a contralateral larger one due to the lack of pressure gradient between the internal carotid artery (ICA) and arterialised blood in the ipsilateral cavernous sinus. We suggest that bilateral ICA angiograms should be performed before and after occlusion of all DCCF.

[Comparison of three fat suppression sequences for the detection of vertebral detection. Turbo STIR, phase contrast gradient-echo, and MISTEC-Chopper after gadolinium injection]. / Comparaison de trois séquences de suppression de graisse pour la détection des métastases vertébrales: Turbo STIR, écho de gradient en opposition de phase et MISSTEC-Chopper après injection de Gadolinium.

OBJECTIVES: Assess three fat suppression sequences used to search for spinal metastases: TurboSTIR, phase contrast gradient-echo, and MISTEC-Chopper after gadolinium injection. MATERIAL AND METHODS: A prospective study was conducted in 10 patients with primary neoplasia. MIR sequences acquired (1 Tesla) were TurboSTIR, T1 spin-echo with and without gadolinium injection, phase contrast gradient-echo and M-Chop after gadolinium injection. Signal intensity in normal bone marrow, metastatic tissue, and subcutaneous fat as well as background noise was measured. Signal-to-noise (S/N) ratio was determined. Lesion borders, artefacts, and extent of detected lesions were determined quantitatively. Bone marrow signal intensity was also recorded. RESULTS: S/N ratio was best with gradient-echo which identified well the borders of lesions within the hemopoietic marrow. For lesions located in high-fat marrow (as in post-radiation marrow), the high intensity signal of the lesion confounded with the fat signal. TurboSTIR gave effective fat signal suppression and was particularly useful for yellow marrow, less so for red marrow. This technique confounded cell proliferation with perilesional edema (enlarging lesion extention). In one case, this sequence did not detect a small lesion visible with the two other sequences. This sequence was sensitive to artefacts (especially vascular artefacts) which can produce false nodular images. M-Chop gave good suppression of vertebral fat tissue (better for yellow marrow) but subjective detection of lesions was more difficult. CONCLUSION: The phase contrast gradient-echo sequence after gadlinium injection appeared to be the best sequence excepting cases of post-trauma (radiotherapy or chemotherapy) fat transformation of the marrow where the TurboSTIR sequence could be preferred.

[MRI diagnosis of sinus cavernous invasion by pituitary adenomas]. / Diagnostic en IRM de l'envahissement du sinus caverneux par les adénomes hypophysaires.

PURPOSE: To evaluate the preoperative MRI criteria of a sinus cavernous invasion by a pituitary adenoma. MATERIAL AND METHODS: Study of 102 cavernous sinuses among 51 patients who had had a surgical cure of pituitary adenoma. Thirteen patients had a surgical invasion of the cavernous sinus. RESULTS: A certain number of signs eliminated cavernous sinus invasion. The best means consisted in not crossing the intercarotid line (Sensitivity-Se = 100%, Specificity-Sp = 85% and Negative Predictive Value-NPV = 100%). The others means implied: not going past twelve o'clock on the internal carotid artery-ICA (NPV = 97.1%), symmetrical size of the cavernous sinus (NPV = 92.5%), non-convexity of the lateral wall (NPV = 90.2%), visualization of at least two venous groups of the laterosellar space (NPV = 90.2%) and finally, non-displacement of the ICA (NPV = 89.2%). The best criteria for diagnosis were passing by the intra and supracavernous ICA lateral tangent (Se = 84.6%, Sp = 95%) and the percentage of ICA encasement by the adenoma exceeding 25% (Se = 92.3%, Sp = 85%). CONCLUSION: Except the total encasement of the intracavernous ICA, the cavernous sinus can be invaded when the lateral tangent of the supra and the intracavernous ICA is crossed, and also when the percentage of ICA encasement exceeds 25%.

Pregnancy after intravenous pulsatile gonadotropin-releasing hormone in a hyperprolactinaemic woman resistant to treatment with dopamine agonists.

It is difficult to achieve pregnancy in hyperprolactinaemic patients in whom prolactin inhibiting agents are ineffective. Medical treatment with bromocriptine, lisuride and the new agent CV 205-502 (quinagolide) was unsuccessful in normalizing hyperprolactinaemia in a 28 year-old woman to treat anovulatory infertility. Repeated Magnetic Resonance Imaging (MRI) was normal, with no images suggestive of prolactin adenoma. A live child was born after pulsatile GnRH treatment despite persistently elevated prolactin levels; normal MRI and decreased prolactin levels were observed after pregnancy. In summary, successful pregnancy can be obtained with pulsatile GnRH treatment in women resistant to old and new medical treatments of hyperprolactinaemia.

Diffusion imaging with a multi-echo MISSTEC sequence.

An imaging method is presented to measure the water-diffusion coefficient. The sequence (MISSTEC) uses the simultaneous acquisition of a spin echo and several stimulated echoes with the same intensity except for diffusion weighting. The optimal number of stimulated echoes was calculated to minimize the diffusion coefficient error (D). D values obtained in vitro and in vivo were in good agreement with those from the spin-echo sequence (IntraVoxel Incoherent Motion [IVIM] method). The total acquisition time is half that of the classic IVIM method.

[Use of a stimulated echo sequence in the MRI study of the brain and spine]. / Application de la séquence d'écho stimulé dans l'exploration IRM de l'encéphale et du rachis.

We describe in this paper how the STEAM sequence can be an efficient tool to obtain images free of flow artifacts in anatomical situation where the spin echo failed. The simplest way to eliminate flow artifacts is to exploit the dephasing induced by motion in magnetic field gradients and to reduce to zero the signal from moving tissues. This can be achieve by increasing the time elapsed between the spin excitation and the signal observed. Because of T2 relaxation, such an increase results in a signal decrease when the spin echo sequence is used. The STEAM sequence has the unique property that the time elapsed between observation and excitation can be increased without change in T2 value and so allows a good suppression of signals from the moving spins with short TE. Our results demonstrate that, although the stimulated echo intensity is only half that of a spin echo taken at the same read out time, the advantages of STEAM imaging can compensate for this partial loss in signal to noise in some particular clinical situations. The influence of mixing time on contrast has been evaluated using thoracic spine imaging and it has been shown that contrast between spine and CSF can be significantly improved (+ 60%) when TM is increased (from 17 ms to 50 ms). In the same time, the contrast between spine and fat issue decreases (40%). This last effect facilitates the adjustment of contrast window. Suppression of motion artifacts has first been evaluated with thoracic spine imaging, using a whole body coil. Suppression of artifacts was better than that obtained with a flow compensated spin echo sequence, especially in the case of kyphotic patients when a presaturation band was inefficient. In a second step suppression of motion artifacts has been evaluated from posterior fossa examination after injection of a paramagnetic contrast agent. The images obtained with the stimulated echo sequence show a dramatic reduction of signal from blood in the lateral sinus, and therefore an increase of quality by elimination of motion artifacts.

[MRI and computer-assisted tomography in Kufs disease. Apropos of a familial form]. / IRM et scanographie de la maladie de Kufs. A propos d'une forme familiale.

Kufs disease is the adult form of ceroid neurolipofuscinosis, and an uncommon cause of degenerative nervous system disease affecting young adults. We present here 4 cases of family form revealed by a demential syndrome. In all 4 patients MRI showed diffuse cortical atrophy predominant in the parietal regions. In 3 of these 4 patients MRI also exhibited a low signal in T2-weighted sequences on the putamens. There was no abnormality of the white matter. Diagnosis was made by cerebral biopsy in one case and by rectal biopsy in all other cases. Although the MRI images are not specific, they must be used when the diagnosis of Kufs disease is suspected in young patients with demential syndrome.

[Primary intramedullary melanoma. Apropos of a case]. / Mélanome intra-médullaire primitif. A propos d'un cas.

Primary melanoma is a rare spinal tumour first reported by Hirschberg in 1906. Since then, only 34 cases have been reported. When present here a new case of primary intramedullary thoracic melanoma developed in a 64-year old male patient. MRI showed a paramagnetic signal with reactive cysts. Macroscopy and histology confirmed the diagnosis. Spinal cord melanoma is presumed to be primary when no other melanoma is found outside the CNS. The tumour is often located in the middle or lower thoracic cord, may be intra- or extra-medullary and leptomeningeal or extradural. It frequently progresses slowly. MRI is the essential examination as it demonstrates a lesion with paramagnetic properties. Its image is not specific and may correspond to other pigmented tumours (meningeal melanocytoma, melanotic schwannoma), to a lipoma or to a vascular or tumoral haemorrhagic lesion. Treatment is uncertain, but surgery is frequently associated with radiation. Postoperative follow-up aims at detecting a local regrowth of the tumour or leptomeningeal dissemination which affects the prognosis.

[Dural fistula of the cavernous sinus. Clinical and angiographic aspects. Results of particulate intravascular treatment]. / Fistules durales de la loge caverneuse. Aspects cliniques et angiographiques. Résultats du traitement endovasculaire au moyen de particules.

Thirteen patients with dural fistula of the cavernous sinus were studied by angiography. Nine of them presented with ophthalmic symptoms (chemosis and oculomotor disorders caused by the fistula). In one patient the lesion was discovered by chance, and 2 other patients had consulted for a disabling tinnitus. Eight patients accepted to the treated by the endovascular route for embolization of the maxillary artery, using particles. Embolization was unilateral in 4 cases and bilateral in 4 other cases. All embolizations were followed by serial control angiography immediately performed. All subjects were seen again as out-patients at 3 month, and 5 of them accepted a control angiography. Three patients were then regarded as clinically and anatomically cured. Two patients with incomplete clinical and angiographic results had a second embolization which resulted in clinical and anatomical cure at a 4-month control examination. These 8 patients were re-examined clinically after one month of treatment and found to be symptomless. Only one complication (transient oedema of the face) was noted. Dural fistulae are lesions that are most probably acquired by alteration of the physiological dural arteriovenous shunts occurring soon after venous thrombosis. Their course is capricious, and they sometimes heal spontaneously. However, the cavernous sinus location with its repercussion on the eye usually requires treatment. This treatment is initially endovascular; surgery and multifascicular irradiation being reserved for failures. Particle embolization of maxillary arteries is a simple and efficient procedure which must be used initially. If it proves insufficient, embolization of other arterial feeders (but it is often more dangerous) or the venous route can be tried.

T1 mapping from spin echo and stimulated echoes.

We present an imaging method to obtain a map of the spin-lattice relaxation time. Images were acquired with the same spatial resolution and in the same time as for a regular spin-echo acquisition. The sequence was based on the simultaneous acquisition of a spin echo and several stimulated echoes with the same intensity except for T1 weighting which increases with the interval between the excitation pulse and the readout pulse. T1 values obtained on phantoms were compared to those from the inversion-recovery method and show the accuracy (2%) and the precision (5%) of the method. T1 images of the brain of a healthy volunteer are presented and demonstrate the ability of the method to obtain T1 mapping in vivo in 12 min and without susceptibility artifacts. In vivo and in vitro results were compared to those obtained by a TOMROP sequence in the same acquisition time.

[Spinal cord compression due to thoracic spinal chondroma. A study of 2 case and a review of the literature]. / Compression médullaire par chondrome rachidien thoracique. Etude de 2 cas et revue de la littérature.

We report two cases of thoracic spine chondroma in two girls aged 14 and 12 years. The lesion was revealed by signs of spinal cord compression and affected T1 and T7 respectively. Chondroma is a benign bone tumour, fairly frequent in the extremities but very rare in the spine. It accounts for 2.8% of benign bone tumours and 12% of all bone tumors. 3.6% to 4% of chondroma are located in the spine. Out of 51 cases reported in the literature 21 had been revealed by spinal cord compression, with an age range of 5 to 76 years. In all cases of spinal cord compression magnetic resonance imaging (MRI) is necessary as first-line examination, but CT scans are still useful in the diagnosis and post-operative follow-up of spinal chondroma because of the osteo-cartilaginous nature of these tumors. The purpose of this study was to report two new cases of spinal cord compression by thoracic spine chondroma and to review the literature.

[The value of magnetic resonance imaging in the diagnosis of spinal cord hemangioblastoma. Apropos of 12 cases]. / Intérêt de l'imagerie par résonance magnétique dans le diagnostic des hémangioblastomes médullaires. A propos de 12 observations.

This study concerned a series of 12 patients, 4 of whom had Von Hippel-Lindau disease. Six of these patients were explored by myelography, 6 by spinal cord angiography, 8 by CT scan with contrast injection and 12 by MRI, with gadolinium injection in 8. MRI proved to be the choice examination for the diagnosis of spinal cord tumor, but gadolinium injection was necessary since it made it possible to detect the tumoral bud and its intense enhancement. The absence of gadolinium injection led us to an erroneous initial diagnosis of syringomyelia in two patients and glioma in one. Sagittal sections made it easier to evaluate the tumoral extension in patients with evidence or suspicion of Von Hippel-Lindau disease. Arteriography was indicated, as it provided a preoperative map and diagnosed punctiform lesions.

Contribution of MR-angiography to the diagnosis and the therapeutic indication and follow-up of intracranial aneurysms.

The purpose of this study was to evaluate MR-Angiography (MRA) in comparison with digital angiography in the diagnosis, therapeutic indication and therapeutic follow-up of cerebral aneurysms. Out of 27 patients explored by angiography and MRA, 17 had aneurysm, 5 had an aneurysm excluded by a detachable balloon, and 5 without aneurysm were used as controls. MRA, performed by the paradoxical enhancement technique, required a 3 DFT acquisition with short TR and TE (40-8 ms) in gradient echo, with a 15 degrees flip angle. Analysis of angiograms and partitions was compared with that of angiography. Sixteen of the 17 aneurysms were found by a study of both angiograms and partitions. The result was doubtful in 1 patient. No false-negative result was recorded. Spasm was always detected, but the neck of the aneurysm was never sufficiently well defined. The quality of angiogram depends on the type of balloon used in the patients treated: only the balloon without metallic index did not induce an artefact. Thus, MRA seems to be useful in the detection of cerebral aneurysms, and the absence of artefact induced by detachable balloons without metallic index could allow a therapeutic follow-up. But for the moment, arteriography remains the reference for pretherapeutic evaluation.

Acquisition of spin echo and stimulated echo by a single sequence: application to MRI of diffusion.

A new method is described to measure the restricted diffusion coefficient with magnetic resonance imaging. The two images necessary to calculate the diffusion image are obtained with a simultaneous acquisition of a spin-echo and a stimulated echo, and so, in half the time needed by usual spin-echo or stimulated echo method. A different diffusion contrast is created on each echo. A map of an estimate of the diffusion coefficient and an estimation of T1 value are obtained with only one experiment. The accuracy of the method has been evaluated on phantom and results are in agreement with values found in previous papers and with measurements performed with a usual spin-echo method. Furthermore, in vivo measurements have shown that this method can be used without electrocardiogram triggering.

Sacral pseudomeningocele and Marfan's disease. One case.

The authors report a case of sacral pseudomeningocele discovered by chance in a 29-year old male patient presenting with Marfan's disease. In this disease, such abnormalities are common and due to excessive fragility. of the dura mater. CT and MRI provide an accurate morphological analysis and a complete evaluation of lumbosacral osteomeningeal abnormalities.