Surgery for focal cortical dysplasia

Three decades since its comprehensive and elegant description (Taylor et al., 1971), focal cortical dysplasia (FCD) remains an enigmatic condition. Its true incidence is not well known, as even high-resolution MRI studies may not enable its detection. In surgical series, which are inevitably pre-selected, FCD is the most common developmental pathology identified. Undoubtedly, FCD may cause severe refractory epilepsy including epilepsia partialis continua or generalized status epilepticus that can be directly life-threatening. Despite the attention focused on FCD because of this not infrequent malign presentation, much of its biology remains obscure, from its molecular pathology to its natural history. For example, we do not know why paediatric and adult presentations may differ; we do not know if the newly comprehended cellular dynamics of the postnatal human brain, including neoneurogenesis, constitute an important component of FCD biology; and fundamentally we do not know why FCD causes epilepsy or looks so bizarre under the microscope. When antiepileptic drugs fail to bring about complete seizure freedom in FCD, surgical resection of the FCD may hold the prospect of at least some relief for some patients. Practising epileptologists will all have seen individual patients who, following focal resection of the FCD, have done well or even become seizure free, and whose quality of life has improved even if seizure freedom has not been achieved. A previous review found that in comparison with the benchmark of patients having therapeutic resective surgery for epilepsy caused by hippocampal sclerosis, post-surgical outcome for FCD is less good, with ∼45% of reported patients becoming seizure free overall (Sisodiya, 2000); nevertheless, this figure is likely to be considerably higher than the chance of achieving seizure freedom with antiepileptic drugs alone (Semah et al., 1998). Completeness of resection of the dysplastic tissue is felt to be one critical factor in determining outcome after surgery. Recent studies suggest that in fact outcome after surgical resection may be better still (Chassoux et al., 2000; Tassi et al., 2002; Kral et al., 2003; Cohen-Gadol et al., 2004). Seizure-free rates of up to 92% have been reported in specific, usually small, subgroupings (Cohen-Gadol et al., 2004). It remains unclear if this improvement in fact simply reflects better patient selection, or represents a real advance in patient management in terms of preoperative investigations, and operative techniques. A further important issue is the quality of data presentation in the literature. Recent surgical series have incorporated advances made in the classification of FCD, which in turn attempt to encompass more recent data on FCD biology. Detailed study of FCD suggests that there are distinct histological subtypes. According to the most comprehensive recent classification (Palmini and Luders, 2002), cortical dysplasia can be categorized in order of increasing cytological disruption as ‘mild malformation of cortical development’ (mMCD), or FCD type 1a (isolated architectural abnormalities), 1b (with additional ‘immature’ or giant neurons), 2a (with additional dysmorphic neurons) and 2b (with additional balloon cells). The true clinico-pathological bases and implications of such histological subtyping remain to be understood, but clearly not all patients with FCD manifest the same type or severity of epilepsy. In this issue, Fauser and colleagues report surgical outcome data on patients of a wide range of ages in one of the largest FCD series from a single surgical centre. For those patients for whom data were available, documentation is excellent, with standard outcome and histological classification schemes employed. Outcome is analysed with respect to histological subtype. The authors find that histological subtype did have a bearing on outcome, with a better outcome for patients with less ‘severe’ histopathology (i.e. mMCD or FCD types 1a or 1b). Problems, however, persist. The proportion of patients in any given category actually followed-up for any length of time was small: at best, 25% of patients seen at 6 months had follow-up data at 48 months. Though for patients for whom data were available, outcome class ‘remained almost constant’, this can only apply for those patients for whom data were available. In addition, the actual numbers of patients in any one histological subtype were small, limiting the analyses that could be undertaken and the conclusions that could be drawn. We do not know what influence the degree of resection had, or what impact on quality of life surgery produced. These are not problems by any means unique to this study; indeed, this study provides better data than most and is an important contribution to the field. However, it serves to highlight what remains to be done. For example, from this paper and the literature overall, it is still not possible to determine conclusively the impact of histopathology, dual pathology, multilobar pathology or FCD location (temporal or extratemporal). Even if histopathology does influence outcome, we still need reliable preoperative means of predicting this histopathology. Therefore, even from these authors' careful study, and from other recent data (some of which conflict with Fauser et al.'s findings), we still cannot determine which patients with FCD will benefit most or, indeed, at all, from resective surgery. We have no means of identifying such patients prospectively pre-surgically. Perhaps even more significantly, we cannot predict post-surgical quality of life, which for FCD at least need not parallel seizure outcome, especially for children (Duchowny et al., 1996). Whilst some limitations in current reports are superable, others are not. The most important limitation is that of patient numbers: Fauser et al. studied 67 patients operated on over 60 months and, worldwide, very few centres have the surgical throughput to be able to answer the important remaining questions. Sizeable, adequately powered, standardized multicentre or international studies are commonplace in other medical disciplines, are published for neurological conditions, including for the drug treatment of epilepsy, and at least one such study is ongoing for the surgical treatment of epilepsy due to hippocampal sclerosis (http://www.erset.org). Whilst it may not be possible to standardize pre-surgical approaches to FCD across international centres, for example in terms of use of invasive EEG monitoring, it should nevertheless prove possible to agree follow-up parameters (e.g. postoperative MRI), and to standardize histological analysis and control for degree of resection. Additionally, accepted outcome and quality-of-life scales should be widely adopted. A large, prospective, standardized and quality-assured study of surgery in FCD should be feasible. In turn, this should enable us to establish how predictable outcome is, and which available pre-surgical characteristics contribute to outcome prediction: this should then hone our pre-surgical selection. At the moment, there are few concrete predictors of outcome, though completeness of resection and histopathology are probably important. We need better understanding of FCD biology. In the absence of good animal models of FCD, resected human surgical tissue is an invaluable resource. Its multimodal study at histological, molecular, developmental, genetic and network levels may not only reveal to us information critical to improved treatment and, possibly, prevention of FCD, but may also cast light on broader processes of neurodevelopment, neurophysiology and neurodegeneration (Bothwell and Giniger, 2000). One intriguing aspect is the ontogeny and role of the balloon cell, a grossly malformed cell, the presence of which may or may not alter visibility on MRI but whose presence appears from some series to augur a better outcome even when otherwise ominous dysplastic neurons are found. Could this phenomenon be exploited to generate a new treatment option, capitalizing on intercellular regulation? In times to come, when, for example, we understand that FCD is the end result of a variety of different processes, mediated by a pattern of changes in gene expression causing neuronal dysmorphology and accelerated local neurogenesis, we may well be able to reverse FCD epileptogenicity by coordinated manipulation of activity of a number of key genes, even if these are not always causative. However, until then, surgery still offers for some individuals the chance of significant improvement in quality of life, and should at least be considered for patients with refractory epilepsy caused by FCD. Our short-term aim must be to improve outcome from such surgical treatment. We would seem to have the means: do we have the will?