Radiosurgery for Drug-Resistant Epilepsies: State of the Art, Results, and Perspectives



Fig. 58.1
Hypothalamic hamartomas topological classification. The exact location of the lesion in relation to the interpeduncular fossa and the walls of the third ventricle correlates with the extent of excision required, the seizure control, and the complication rate. Although it may be an exceptional observation, Type V (pedunculated) tend not to have neurological symptoms (no epilepsy, no cognitive deterioration, and no behavioral disturbances). They may present with precocious puberty or be symptom-free. Types I, II, III, and IV may cause seizures in many cases, as well as mental retardation, behavioral abnormalities, and precocious puberty. Type VI HH are frequently found in patients with especially severe clinical presentations. We consider radiosurgery as a first-line treatment for small Type I, II, III, or IV HH. From Regis J, Hayashi M, Eupierre LP, Villeneuve N, Bartolomei F, Brue T, Chauvel P. Gamma Knife surgery for epilepsy related to hypothalamic hamartomas. Acta Neurochir Suppl 91:33–50, 2004; Used with permission



In Type II (when the lesion is small and mainly in the third ventricle) radiosurgery is certainly the safer alternative. Even though the endoscopic and transcallosal interforniceal approaches have been proposed, the risks of short-term memory worsening, endocrinological disturbance (hyperphagia with obesity, low thyroxine, sodium metabolism disturbance), and thalamic or thalamocapsular infarcts have been reported also by the more enthusiastic and skillful neurosurgeons. However in cases of very severe repeated status epilepticus we propose as a salvage surgery either a transcallosal interforniceal approach or an endoscopic approach (depending on the width of third ventricle). In an emergency situation if the lesion is small and the third large, the endoscopic approach is chosen.

In Type III (lesion located essentially in the floor) the extremely close relationship between the mammillary body, the fornix, and the lesion is clearly leading us to prefer GKS. We speculate that sessile hypothalamic hamartomas have always more or less an extension in the hypothalamus close to the mammillary body. Thus when a lesion is classified as a Type II, it means that the lesion appears on the MR like mainly located in the third ventricle but is likely to have a root in the hypothalamus. The same assumption is made for Type III.

In Type IV (the lesion sessile in the cistern) a disconnection can be discussed (pterional approach with or without orbitozygomatic osteotomy). However, if the lesion is small, GKS can be recommended due to its safety and to its capability to reach at the same time also the small associated part of the lesion in the hypothalamus itself, frequently visible on the high resolution MR. In Delalande experience only two patients among 14 are seizure free after a single disconnection through a pterional approach [49]. Consequently, we do use this approach in case of lesion too large for GKS as a first step of a staged approach. In most circumstances, the patient is improved but not seizure free after the first surgical step and GKS is organized at 3 months as a second step of the treatment.

Type V (pediculate) are rarely epileptic and can be easily cured by radiosurgery or disconnection through a pterional approach. In case of severe epilepsy the second therapeutic modality will allow certainly a faster seizure cessation. However a distant extension of the HH in the hypothalamus close to the mammillary bodies must be cautiously searched on high resolution MR and its discovery will eventually lead to prefer GKS allowing to treat both parts of the lesion, especially in cases in which if the cisternal component is small.

Type VI (giant) do not represent good indications for first intention radiosurgery, as in nearly all the cases a combination of several therapeutic modalities should be utilized. Even if GKS does not seem to be suitable when the lesion is large, radiosurgical disconnection can be considered (targeting only the superior part in the hypothalamus and/or the third ventricle leaving untreated all the lesion lower than the floor) but has been systematically disappointing. In our opinion, this strategy may cause loss of a precious time for the child to be treated effectively. Consequently, we do not advocate for such a strategy. When microsurgical resection has left a small remnant in the third ventricle and a still active epilepsy, re-operation can be envisaged by GKS.

Two major questions remain. First, we know that complete treatment or resection of the lesion is not always mandatory [5052] but we do not know how to predict in an individual patient the amount (and mapping) of the HH that must be treated in order to obtain a complete antiepileptic effect. Secondly, we know that these patients frequently present with an electro-clinical semiology suggesting involvement of the temporal or frontal lobe and which can mimic a secondary epileptogenesis phenomenon [37, 49]. In our experience, some of these patients can be completely cured by the isolated treatment of the HH, while in others, a partial result is obtained, with residual seizures despite a significant overall psychiatric and cognitive improvement. In this second group, it is tempting to propose that such a secondary epileptogenic area accounts for the partial failure.

Our initial results indicate that GKS is as effective as microsurgical resection and much safer [53]. GKS also avoids the vascular risk related to radiofrequency lesioning or stimulation. With transcallosal interforniceal approach, both cognitive (long-term memory impairment) and severe endocrinological complications (23 % long-term appetite stimulation and major weight gain) have been reported [54, 55]. Stabell et al. have reported with endoscopy a serious deterioration of memory and reading skills associated with a permanent oculomotor paresis [56]. Using interstitial implants (Iodine 125 seeds), Schulze-Bonhage et al. [57] reported in a series of 24 patients, 5 patients developing a symptomatic edema; 4 patients having a weight gain of more than 5 kg, which was severe in 2; and a persistent decline of episodic memory in 2 patients. None of these complications have been observed after Gamma Knife surgery. The main disadvantage of radiosurgery is its delayed action. Longer follow-up is mandatory for proper evaluation of the role of GKS. Results are faster and more complete in patients with smaller lesions inside the III ventricle (Stage II). The early effect on subclinical EEG discharges appears to play a major role in the dramatic benefit to sleep quality, behavior, and cognitive-developmental improvement. Gamma Knife surgery can safely lead to the reversal of the epileptic encephalopathy [38, 40, 41, 58].

Due to the very poor clinical prognosis of the majority of these patients with HH and the invasiveness of microsurgical resection, GK can now be considered the first-line intervention for small middle size HH associated with epilepsy, as it can lead to dramatic improvements to the future of these young patients. The role of secondary epileptogenesis or of widespread cortical dysgenesis in these patients needs to be better evaluated and understood, in order to optimize patient selection and define the best treatment period.



Mesial Temporal Lobe Epilepsy


The first Gamma Knife surgery operations for MTLE were performed in Marseille in March 1993. As far as no similar experience was available at this time in the literature, we were obliged to base our technical choices on hypothesis and experience of radiosurgery for other pathological conditions. Four patients were treated with different technical strategies (dose, volume, target definition). The delayed huge radiological changes observed some months after radiosurgery [59] led us to stop such treatment and follow these first four patients. Due to the clinical safety of the procedure in these patients and the gradual disappearance of the acute MR changes after some months, we treated several new series of patients under strict prospective controlled trial conditions (with ethical committee approval). The treatment for the following 16 patients was based upon that of the first patient who had a successful outcome (as opposed to the three others who had partial or no effect). This “classic planning” was based on the use of two 18 mm shots, covering a volume of around 7 cc at the 50 % isodose (24 Gy), and has turned out to produce a high rate of seizure cessation [60, 61]. For epileptological reasons, as well as for safety reasons, the targeting was very much centered on the parahippocampal cortex and spared a significant part of the amygdaloid complex and hippocampus. The refinement of the Gamma Knife surgery technique, and the desire to find a dose which would create less transient acute MR changes, led us to reduce the dose from 24 Gy to 20 and 18 Gy at the margin. However, this brought about a significant decrease in the rate of seizure cessation. We have reviewed the long-term follow-up of our first 15 patients operated by GKS for MTLE at the state of the art (24 Gy). The mean follow-up was 8 years and at the last follow-up 73 % were seizure free. These long-term results are comparing favorably to microsurgical one. No permanent neurological deficit was reported out of a visual field deficit in nine patients [11]. After microsurgery for MTLE on the dominant side, a verbal memory deficit is typically observed in 30–50 % of the patients [62, 63]. This is of special importance to note that none of our patients have observed a neuropsychological worsening (using the evaluation published by Clusman et al.) and specifically no verbal memory decline [4, 5, 7, 11]. This finding of our four prospective trials has been confirmed by the US prospective trial [64].

The timetable of events after radiosurgery and the follow-up are quite standardized. Patients are informed that delayed efficacy of radiosurgery is its main drawback. Typically, the frequency of the seizures is not modified significantly for the first few months. Thereafter, there is a rapid and dramatic increase in auras for some days or weeks and then the seizures disappear. Usually the peak in seizure cessation is observed around the 8th–18th month with a clear variability in the delay in onset. In one patient, this occurred 26 months after GK radiosurgery. We usually consider a delay of 2 years as a minimum for post-radiosurgery follow-up. In the absence of initial radiological changes or clinical benefit, the recommendation is to wait for the onset of the MRI changes and their subsequent disappearance. All our patients had the same pattern of MR changes regardless of marginal dose (18–24 Gy) and treatment volume (5–8.5 cc). However, the degree of these changes and their delay of onset varied according to the dose delivered to the margin, the volume treated, and the individual patient. In order to allow an optimal evaluation, we recommend that subsequent microsurgery not be considered before the third year after radiosurgery. Similarly, we believe that a patient who undergoes a corticectomy before the onset of the MR changes has occurred cannot be assumed to have failed radiosurgical treatment. Of course, before consideration of any further surgery, the question of the reason for the failure needs to be addressed. After reviewing files of patients treated for MTLE with radiosurgery, it was sometimes possible to identify likely causes of failure, such as:

1

Poor patient selection (e.g., patients with epilepsy involving more than the MTL structures)

 

2

Patients with the diagnosis of “treatment failure” (<3 years) who had been operated upon too early after radiosurgery [65]

 

3

Targeting of the amygdala and hippocampus (which is not in our opinion the optimal target in terms of safety and efficacy) instead of parahippocampal cortex [66]

 

4

Insufficient dosage [6567]

 

Our current strategy of treatment is based on our first series of MTLE patients who were strictly selected and treated systematically with a very simple but very reproducible dose planning strategy [4, 6]. The identification of putative improvements in the methodology requires a systematic analysis of the influence of the technical data from our experience and from the literature on the outcome of those patients.


The “Technical” Questions



The Dose Issue


The first targets used in functional GK radiosurgery (capsulotomy, thalamotomy of VIM or the centromedianum, pallidotomy) were treated using high dose (300–150 Gy) delivered in very small volumes (3–5 mm in diameter) [16]. The goal was to destroy a predefined very small anatomical structure with stereotactic precision. Quite a significant variability in the delay and amplitude of the MR changes has been reported with fixed regimen of doses [28, 68]. Barcia Salorio et al. have presented several times a small and heterogeneous group of patients treated with different kinds of devices and dosage regimens [69]. Apparently some of those patients had no expanding lesion and were treated with very large volumes and very low dosage (around 10 Gy). Based on this experience, several teams have made the assumption that very low doses, as low as 10–20 Gy at the margin, should be as effective as the 24 Gy protocol (at the margin) that we used for our first series of patients with MTLE [4]. A cautious examination of the last proceeding of Barcia Salorio et al. shows that the individual information concerning the dose at the margin, the volume, and the topography of the epileptogenic zone are not provided. Moreover, among the 11 patients reported, the real rate of seizure cessation is apparently only 36 % (4/11), which is much lower than what we would expect with resection in MTLE [30]. In a heterogeneous group of 176 patients, Yang et al. confirmed that only a very low rate of seizure control is achieved when low doses (from 9 to 13 Gy at the margin) are used [67].

The experience of the radiosurgical treatment of HH indicates that 18 Gy at the margin appears to be a threshold in terms of probability of seizure cessation [38]. In this group of patient (36 cases), only one showed MR changes. The majority of the AVM cases with worsening of the epilepsy were treated with a range of doses between 15 and 18 Gy. Similarly, poor results have been reported by Cmelak et al. in one case of MTLE treated with Linac-based radiosurgery, with 15 Gy at the 60 % isodose line, who underwent surgical resection 1 year later. In this case, the authors first observed a slight improvement followed by an obvious worsening [65]. A recent de-escalation study has allowed us to demonstrate poorer results in patients receiving doses of 18 or 20 Gy at the margin as compared to 24 Gy [59, 70]. Due to the rate of seizure cessation that is achievable by conventional resection, a radiosurgical strategy associated with a much lower rate of seizure cessation appears unacceptable. Fractionated stereotactically guided radiotherapy has been demonstrated to fail systematically in controlling seizures. Among 12 patients treated by Grabenbauer et al., none have achieved seizure cessation [71, 72]; only seizure reduction was obtained in this series.

Experimental studies on small animals have demonstrated the antiepileptic effect of radiosurgery [26, 28, 73], the dose dependence of this effect [26, 27, 28, 74], and the possibility of obtaining clear antiepileptic effect without macroscopic necrosis using certain doses [27]. Of course, the rat models of epilepsy are far from being good models of human MTLE. However, taking into account the huge difference in volume of the target, it is intriguing to notice that according to our clinical experience in humans, a similar maximum dose range of 40–50 Gy is currently the range of dose providing the optimal safety-efficacy ratio.


The Target Definition


When the target is a lesion that is precisely defined radiologically, the question of the selection of the marginal dose can be quite easily addressed by correlating safety-efficacy individual outcome to the marginal dose (Fig. 58.2). This can be refined based upon stratification according to volume, location, age, etc. However, in patients presenting with MTLE, this process is invalid for two reasons. Firstly, there is no consensus regarding the requirement for extent of mesial temporal lobe resection. Secondarily, the concept of MTLE syndrome with a stable extent of the epileptogenic zone and surgical target is increasingly the topic of debate [75, 76].

A83315_2_En_58_Fig2_HTML.jpg


Fig. 58.2
Dosimetry in a case of a typical right mesial temporal lobe epilepsy (24 Gy at the 50 %)

The volume (in association with marginal dose) is well known to be a major determinant of the tissue effect, as shown in integrated risk/dose volume formulae [77]. In the first series of patients that we treated, this marginal isodose volume (or prescription isodose volume) was approximately 7 cc (range 5–8,5).

An attempt to correlate dose/volume and the effect on seizures and on the MR changes (as evaluated by volume of the contrast enhancement ring, extent of the high T2 signal, and the importance of the mass effect) has been published recently [70]. In this study, we found, not surprisingly, that the higher the dose and the volume, the higher the risk of having more severe MR changes, but also the higher the chance of achieving seizure cessation. However, these data have limited value. Hence, more precise identification of those structures of the mesial temporal lobe which need to be “covered” by the radiosurgical treatment may allow more selective, but just as efficacious, dose planning strategies, in spite of smaller prescription isodose volumes.

There is growing evidence to support the organization of the epileptogenic zone in networks, meaning that several different and possibly distant structures are discharging simultaneously at the onset of the electro-clinical seizure. This kind of organization explains why the risk of failure is so high when a simple topectomy (without preoperative investigations) is performed in severe drug-resistant epilepsies associated with a benign lesion [78]. This has been also reported in MTLE [75, 76]. Certain nuclei of the amygdaloid complex, the head body and tail of the hippocampus, the perirhinal, entorhinal (EC) and parahippocampal cortices may be associated with the genesis of the seizures. The role of the EC cortex in epilepsy is supported by experimental studies in animal [79, 80]. The EC is considered to be the amplifier of the “amygdalohippocampal epileptic system.” The pattern of the associated structures, including that of the structure playing the leader role, can vary significantly from one patient to another [75, 76]. There is a subgroup of patients who have clonic discharges and the involvement of the EC, amygdala, and head of the hippocampus, with a clear leader role of the EC. Wieser et al. have analyzed the postoperative MR images of patients operated by Yasargil (amygdalohippocampectomy) and were able to correlate the quality of the resection of each substructure of the mesial temporal lobe area and the outcome with respect to seizures [81]. Only the quality of the removal of the anterior parahippocampal cortex was correlated strongly with a higher chance of seizure cessation [81]. We tried to perform a similar study in patients treated with GK radiosurgery [70]. We defined and manually drew the limits of subregions on the stereotactic images of all these patients. The amygdala, the head, the body, and the tail of the hippocampus were first delineated. The white matter, the parahippocampal cortex, and the cortex of the anterior wall of the collateral fissure were then separately drawn and divided into four sectors in the rostrocaudal axis, corresponding to the amygdala, the head, the body, and the tail of the hippocampus [70].


Patient Selection


Whang (without having first performed specific preoperative epileptological work-up) treated patients with epilepsy associated with slowly growing lesions and observed seizure cessation in only 38 % (12/31) of the patients [82]. This kind of observation emphasizes the importance of preoperative definition of the extent of the epileptic zone and of its relationship with the lesion [78, 83]. In our institution, the philosophy is to adapt the investigations for each individual case. In some patients, the electro-clinical data, the structural and functional imaging, and the neuropsychological examination are sufficiently concordant for surgery of the temporal lobe to be proposed without depth electrode recording. In other cases, the level of evidence for MTLE is judged insufficient, and a stereoelectroencephalographic (SEEG) study is performed. The strategy of SEEG implantation is based on the primary hypothesis (mesial epileptogenic zone) and alternative hypotheses (early involvement of the temporal pole, lateral cortex, basal cortex, insular cortex, or other cortical areas). The goal of these studies is to record the patient’s habitual seizures, in order to establish the temporospatial pattern of involvement of the cortical structures during these seizures. Clearly in these patients, the high resolution of depth electrode recording allows fine tailoring of surgical resection, according to the precise temporospatial course of the seizures. The main limitation of radiosurgery is that of size of the target (prescription isodose volume). The radiosurgical treatment of MTLE is certainly the most selective surgical therapy for this group of patients. The requirements for precision and accuracy in the definition of the epileptogenic zone are consequently higher. Furthermore, if depth electrode investigation enables demonstration of a particular subtype of MTLE, this can lead to tailoring of the treatment volume and frequently allows this to be reduced.


The Potential Concerns


The risk of long-term complications must always be cautiously scrutinized in functional neurosurgery. Radiotherapy is most frequently used in the brain for short-term life-threatening pathologies. The use of radiotherapy in young patients with benign disease, such as pituitary adenomas or craniopharyngiomas, has been associated with a significant rate of cognitive decline [61, 68] and tumor genesis [84] including some carcinogenesis [85]. If the risk of radiation-induced tumor was similar with radiosurgery, we should have by now already observed numerous cases. However, such reported cases [8688] are extremely rare and frequently fail to meet the classical criteria by which tumors are deemed to be “radiation induced” [89]. In fact it is considered that, if this risk exists, it is likely to be around 1/10,000 which is far lower than the mortality risk associated with temporal lobectomy [60, 9093].

Epilepsy is a life-threatening condition. The risk of sudden unexplained death in epileptic patients (SUDEP) is higher than in the general population [94, 95]. This risk is higher in patients treated with more than two antiepileptic drugs and IQ lower than 70 (as independent factors). Because seizure cessation after surgery reduces the mortality risk to that of the general population [95], microsurgical resection of the epileptogenic zone may confer a benefit in terms of the possibility of immediate seizure cessation and therefore reduced mortality risk, as compared to the more delayed benefits of radiosurgical treatment. Our patients are systematically informed about this disadvantage of radiosurgery.


What Are the Current Indications?


The demonstrated advantages of radiosurgery are the comfort of the procedure, the absence of general anesthesia, the absence of surgical complications and mortality, the very short hospital stay, and the immediate return to the previous level of functioning and employment. In MTLE the potential sparing of memory function is still a matter of debate and needs to be established using comparative studies. There is also a requirement for further demonstration of long-term efficacy and safety of radiosurgery. Worldwide, microsurgical corticectomies for MTLE are proving to be very satisfactory due to the rarity of surgical complications and a high rate of seizure freedom. In our experience, the most important selection parameters are the demonstration of the purely mesial location of the epileptogenic zone, as well as clear understanding by the patient of the advantages, disadvantages, and limitations. Best candidates are young patients, with middle severity epilepsy (working, socially well inserted), with a high level of functioning (able to understand well the limits and constraints of radiosurgery), a quite high risk of memory deficit with microsurgery (MTLE on the dominant side with few or no atrophy, few deficit of the verbal memory preoperatively), and potentially huge social and professional consequences in case of postoperative memory deficit [2, 13, 96]. One other very good indication in our experience is that of patients with proven MTLE but previous failure of microsurgery, supposedly due to insufficient posterior extent of the resection.

Only gold members can continue reading. Log In or Register to continue

Stay updated, free articles. Join our Telegram channel

Jun 2, 2017 | Posted by in GENERAL RADIOLOGY | Comments Off on Radiosurgery for Drug-Resistant Epilepsies: State of the Art, Results, and Perspectives

Full access? Get Clinical Tree

Get Clinical Tree app for offline access