JLE

Epileptic Disorders

MENU

7 tesla T2*-weighted MRI as a tool to improve detection of focal cortical dysplasia Volume 18, numéro 3, September 2016

Illustrations


  • Figure 1A

  • Figure 1B

  • Figure 2

  • Figure 3

Tableaux

Focal cortical dysplasia (FCD) is one of the most common underlying pathologies in patients who undergo surgery for refractory epilepsy. A substantial percentage of these lesions is not visible using routine 1.5 tesla (T) or 3T MRI; 37% for FCD type I and 15% for type FCD type II in a review of surgical series from 2000 to 2008, and 36% and 2%, respectively, in a 2000-2007 UCLA series of 97 patients(Lerner et al., 2009). No visible lesion on MRI is a predictor of poor surgical outcome and necessitates additional diagnostic studies which may include invasive intracranial electrode registration(Téllez-Zenteno et al., 2010). Advances in imaging techniques are expected to improve the detection of epileptogenic lesions.

Susceptibility-based contrast MRI sequences are sensitive to the susceptibility effect created by the paramagnetic property of deoxyhaemoglobin in blood, and thus augment the visibility of the cerebral venous microvasculature (Haacke et al., 2009). Sequences such as T2* and susceptibility-weighted imaging (SWI) have proven useful in the detection and characterization of a variety of cerebral vascular abnormalities (e.g. Sturge-Weber syndrome and cavernous malformations), as well as haemorrhagic, calcified or iron-containing lesions(Mittal et al., 2009). Reports on susceptibility-based contrast sequences in epilepsy patients are limited and mostly describe the benefits for detection and characterization of calcified lesions, and not for the detection of otherwise inconspicuous lesions, such as FCD (Saini et al., 2009).In this series, we illustrate the potential of the T2* MRI sequence at 7 tesla (T) in assisting the detection of FCD.

Methods

We retrospectively selected six consecutive patients from the Dutch Epilepsy Surgery Program with refractory epilepsy due to histopathologically-confirmed FCD,for whom T2* images at 7T were acquired pre-surgically between November 2008 and November 2014. These patients either had lesions suggestive of FCD on 3T (n=2) or had normal 3T MRI, with semiology, electroencephalogram, and telemetry findings strongly suggesting a structural focal abnormality (n=4). 3T MRI was deemed normal if no indications for structural abnormalities were identified on assessment by expert neuroradiologists and during review by the Dutch Epilepsy Surgery Program board, which includes neurosurgeons, neurologists, and neuroradiologists with expertise in the field of epilepsy. 3T MRI-negative patients underwent additional pre-surgical evaluation, including 7T MRI and PET, SPECT, or MEG, in order to localize the epileptogenic zone. In addition, 7T MRI was used to further characterize the lesions for the two subjects with lesions on 3T.

The 7T MRI protocol included (conventional) 3D FLAIR, 3D double inversion recovery, 3D T1 and 3D T2-weighted sequences, and a previously described T2*-weighted sequence(Zwanenburg et al., 2011). Parameters used for the gradient echo T2* MRI were: isotropic 0.5-mm resolution, echo time of 27 ms, flip angle of 24̊, repetition time of 57-93 ms (shortest possible), with EPI and flow compensation. For two patients, scans were acquired with 0.6-mm resolution (isotropic), echo time of 20 ms, flip angle of 20̊, and repetition time of 25-26 ms. Images were acquired on a Phillips 7T system (Philips Healthcare, Cleveland, OH, USA) with a volume transmit and 16- or 32-channel receive head coil (Nova Medical, Wilmington, MA, USA).

7T T2* images were reviewed for abnormalities co-localizing with epileptogenic zones. To aid visual detection of hypointense structures, minimum intensity projections of T2* sequences were constructed with 5-mm slab thickness. Following resective surgery, tissues were histopathologically classified according to ILAE guidelines(Blümcke et al., 2011).

Informed consent was obtained for performing 7T MRI and the use of patient data. The report complies with the declaration of Helsinki.

In addition, we reviewed 7T T2* images of eight control subjects (four male and four female; age: 27±4) (previously published by Zwanenburg et al. [2011]).

Case series

Four of six patients had normal 3T MRI findings; in two of these, conventional 7T MRI sequences (T1, T2, double inversion recovery, and FLAIR) revealed a lesion suggestive of FCD. In two other patients, abnormalities were seen both on 3T and 7T images. All six patients underwent resective surgery and had histopathological confirmation of FCD (ILAE type Ib in Patient II, type IIa in Patient V and VI, type IIb in Patient I and IV, and mild malformation of cortical development type 2 in Patient III). In four of these patients (Patients I and II [figure 1A]andPatients III and IV [figure 1B]),7T T2* showed areas containing marked hypointensities with a branched, partly tortuous configuration, and a signal compatible with venous blood, suggestive of increased venous vasculature in the sulci neighbouring the malformed cortex.

Because the described T2* signal changes appeared to be located in the leptomeningeal tissue or subarachnoid space overlying the malformed cortex, we looked for a histopathological correlate in the leptomeninges of the operated patients. Unfortunately, due to the locations of the malformations, in the lower part of a sulcus and interhemispheric, respectively, en-bloc resection with intact leptomeningeal structures was only possible in one patient (Patient III). This lesion was classified as mMCD type 2, with fibrotic and thick meninges containing prominent vascular structures (figure 2).

In summary, hypointensities on T2* were identified in four of six consecutive patients with histologically-confirmed FCD, for whom 7T T2* images where available. Table 1 lists clinical and imaging characteristics of all six patients.

On 7T T2* images and corresponding minimum intensity projections of eight healthy volunteers, symmetric venous leptomeningeal vascular structures were clearly visible but the phenomenon described above was not identified(figure 3 for 7T T2* images of a healthy subject).

Discussion

We describe the observation of 7T T2* hypointensities overlying pathologically-confirmed (n=4) epileptogenic malformations of cortical development. We propose that these signal abnormalities reveal a pathological change in the leptomeningeal venous vasculature. The location in sulci and on the cortical surface, the configuration, and a T2* signal compatible with venous blood all support this hypothesis. The origin of the observed T2* hypointensities could be either, or a combination, of the following:

  • (1) Increased blood volume may be the result of increased vascular diameter or density in the context of a developmental anomaly of the leptomeningeal vasculature, parallel to the developmental malformation of the underlying cortex. The leptomeningeal vascular network is formed from a gestational age of eight weeks, starting as the pial capillary anastomotic plexus which produces penetrating vessels into the cortex (Marín-Padilla, 2012). It is conceivable that, in addition to immaturity and abnormal migration of neurons, the vascular network may show analogous abnormalities in dysplastic cortex, resulting in abnormal configuration of superficial cortical vasculature.
  • (2) Paroxysmal increased metabolic demand associated with epileptic activity(la Fougère et al., 2009) may lead to reactive vascular changes in the form of hypervascularisation. An increase in venous vascular diameter or density results in a greater deoxygenated blood volume per voxel and concomitant T2* changes.
  • (3) A focal increase in deoxyhaemoglobin may be explained by altered hemodynamics. Other studies have shown that interictal hypometabolism was associated with decreased perfusion on arterial spin labelling MRI(Pendse et al., 2010), while cortical activation is generally linked to positive blood oxygen-dependent signal changes caused by an increase in oxygenated blood(Logothetis et al., 2001). Decreased perfusion in epileptogenic lesions may lead to an interictal increase in venous deoxyhaemoglobin, visible as T2* hypointensity.

The clinical relevance of the described phenomenon is that it might be indicative of underlying epileptogenic cortex and even be distinctive in cases where conventional MRI sequences reveal no abnormalities.

Two additional patients with refractory epilepsy evaluated in our centre had clinical and electrophysiological characteristics indicative of a structural seizure focus, but no supportive findings on 3 and 7T MRI. Retrospective review of 7T MRI T2* images showed hypointensities that might have provided clues for localizing the seizure focus (data not shown). Lack of concordance between other ancillary diagnostic results prevented surgical treatment.

Comparison of T2*-weighted sequences acquired at 7T and at lower field strength was not possible since this sequence is not routinely performed with 1.5 or 3T MRI for epilepsy patients in our centre. Fine vascular structures should be more detectable on 7T systems because of the higher spatial resolution and the positive relationship between field strength and T2* effects (Haacke et al., 2009). Because of the type of pathology and the as yet experimental nature of 7T MRI, it is difficult to obtain T2* images in large groups of patients in order to assess the presence of the described phenomenon.

An important limitation is that we report a qualitative analysis of the T2* signal change. Furthermore, histopathological validation of changes in leptomeningeal tissues is limited, because leptomeninges could only be assessed in one surgical specimen. While the impression of increased vasculature on T2* images is congruent with the leptomeningeal histological findings (thick and fibrosed meninges containing large caliber vessels), this may be a non-specific finding in patients with epilepsy. To elucidate the histopathological substrate for this potential marker of subtle epileptogenic lesions and assess its clinical relevance, studies on larger groups of patients are needed, including histopathological analysis of leptomeningeal vascular structures.

Despite its limitations, this case series suggests that it may be worthwhile to further explore the possibilities of using T2* sequences at 7T, and possibly also at lower field strength, to improve the detection of FCD in patients with refractory epilepsy.

Conclusion

We report a novel, 7T MR T2* finding co-localizing with FCD. This hypointense signal may indicate an increase in vascular prominence in the leptomeningeal vascular network overlying the dysplastic cortex. Adding T2*-weighted sequences to the 7T MRI protocol may aid in the detection of FCDs and guide effective epilepsy surgery, while obviating the need for additional (invasive) diagnostic tests.

Supplementary data

Summary didactic slides are available on the www.epilepticdisorders.com website.

Acknowledgements and disclosures

This study was supported by a grant from the Dutch Epilepsy Foundation (NEF 12.12). The authors thank Fredy Visser for technical support.

The data in this report was previously presented at the 11th European Congress on Epileptology, Stockholm, 29th June-3rd July, 2014 (Focal cortical dysplasia on 7 tesla susceptibility weighted magnetic resonance imaging. Veersema T.J., Ferrier C.H., van Eijsden P., Gosselaar P.H., Hendrikse J., Spliet W.G.M., Aronica E., Braun K.P.J.)