Abstract
Temporal Lobe Surgery in Germany from 1988 to 2008: Diverse Trends in Etiological Subgroups.
Helmstaedter C, May TW, von Lehe M, Pfaefflin M, Ebner A, Pannek HW, Elger CE, H. Stefan H, Schramm J. European Journal of Neurology 2014;21:827–834.
BACKGROUND AND PURPOSE: In the epilepsy community, there is talk that the number of classical patients with early onset temporal lobe epilepsy (TLE) and Ammon's horn sclerosis (AHS) is decreasing. This is counterintuitive, considering the success story of epilepsy surgery, improved diagnostic methods and the current recommendation of early admission to surgery. In order to recognize trends, the development of temporal lobe surgery over 20 years in three major German epilepsy centers was reviewed. METHODS: Age at surgery and duration of epilepsy, which was differentiated according to histopathology (AHS, developmental, tumor, vascular), year of surgery and center, were evaluated in a cohort of 2812 patients from three German epilepsy centers who underwent temporal lobe surgery between 1988 and 2008. The analysis was carried out for the pooled cohort as well as for each center separately. RESULTS: Of all patients, 52% showed AHS. Compared with other pathologies, the AHS group had the earliest epilepsy onset and the longest duration of epilepsy. Across five time epochs, the diagnosis of AHS increased in the first epoch, remaining constant thereafter. Contrary to the trends in other pathologies, in the AHS group the mean age of patients at surgery increased by 7 years and the duration of epilepsy until surgery increased by 5 years. This trend could be replicated in all three centers. As initially hypothesized for all groups, age and duration of epilepsy in other pathology groups remained constant or indicated earlier submission to surgery. CONCLUSIONS: During the first few years studied, most probably due to progress in brain imaging, the proportion of patients with AHS increased. However, despite stable numbers over time, and contrary to the trends in other pathology groups, age and duration of epilepsy in mesial TLE with AHS (mTLE + AHS) increased over time. This supports the hypothesis of a decreasing incidence of AHS. This trend is discussed with respect to disease-modifying factors which have changed the incidence of classical mTLE + AHS or, alternatively, to recent developments in antiepileptic drug treatment, the appraisal of surgery and economic incentives for treatment options other than surgery.
Commentary
Mesial temporal lobe epilepsy (mTLE) is the most prevalent form of focal-onset epilepsy. Although poorly understood, a majority of patients diagnosed with mTLE are expected to develop hippocampal sclerosis (HS) as evidenced by either mesial temporal atrophy and or signal change on MRI T2/FLAIR sequences (1). In addition, subgroups of mTLE patients will develop hippocampal changes not evident by visual inspection on standard high-resolution MRI studies (2). It is expected that a majority of patients with mTLE-HS will develop pharmacological resistance. In HS patients, data show that anterior temporal lobectomies (ATL) that include the anteromesial portion of the hippocampal formation provide a much greater likelihood of seizure freedom than does medical therapy (3). As a result, mTLE with HS is recognized as the most common indication for resective surgery for the treatment of pharmaco-resistant epilepsy. Surgical evaluations for ATL have increased since the 1990s worldwide; however, ATL corrected for population growth have remained relatively stable (4). Furthermore, some studies even suggest that patients with early-onset mTLE-HS are decreasing. On the surface, such findings contradict the utility of developing sophisticated diagnostic methods and improved surgical techniques, producing successful long-term outcomes for pharmacoresistant mTLE.
Helmstaedter et al. presented a retrospective multicenter study that reviewed the development of temporal lobe surgery over a 20-year period (1988–2008) at three major epilepsy centers in Germany. This period marked the beginning of modern ATL in Germany since the end of World War II. The authors discovered that despite stable numbers of patients with mTLE-HS over time—and contrary to the trends in other identified pathology groups—age and duration of epilepsy in mTLE-HS increased over time. They concluded that these data supported the notion of a decreasing incidence of mTLE-HS. The authors discussed possible explanations for this trend including 1) disease-modifying factors that have changed the incidence of classical TLE-HS, 2) relatively recent developments in antiepileptic drug (AED) treatment, and 3) the economic incentives for treatment options other than surgery. Inherent in the first two of these explanations is the assumption that the natural history of the disorder does not always follow a relentless course of pharmacological resistance.
The authors identified a cohort of 2,812 patients, with 52% classified as having HS. The operated patients represented 85% of all patients with mTLE completing a presurgical evaluation at the three study centers. The other groups identified included tumors, gliosis, focal cortical dysplasia (FCD), vascular malformations, and no structural findings on neuroimaging. Of note, the FCD group demonstrated the earliest onset of epilepsy. This latter finding is intuitive in that earlier identification of FCD is due, in large part, to improved neuroimaging technologies and techniques. The authors concluded that the stable rate of mTLE patients with HS during recent years did not indicate a decrease in the absolute number of patients with mTLE-HS undergoing surgery as adjusted for population growth in Germany. The group of mTLE-HS patients highlighted one of the major findings of Helmstaedter et al. in that it was the only group where age at the time of surgery was continuously increasing over time. Importantly, this group additionally showed an increasingly longer duration of epilepsy prior to surgical intervention. These findings are counterintuitive when compared with the FCD patient group.
The authors also explained the downward trend of ATL beginning in 2004 at the three centers as possibly representative of a movement away from these large centers to smaller regional hospitals in the country. However, they admittedly did not include these smaller centers in their study. This hypothesis was echoed by other studies for Level 4 U.S. epilepsy centers (4). Due to the retrospective and multicenter strategy of the study, a limited set of common data was available. Furthermore, the authors admitted that they did not take into consideration those patients with mTLE who were not considered resective surgery candidates.
To better understand the relative plateau effect of ATL worldwide, studies must consider a number of variables. One such variable includes a medical therapy comparison group not undergoing ATL. For example, Tseng et al. (7) examined the rate of long-term seizure control in patients with mTLE from 1998–2013 at a large U.S. center who did not undergo ATL. Among such patients, a quarter may have sustained seizure freedom with medical therapy alone. Surgical and nonsurgical groups in this retrospective study shared similar overall demographics and risk factors, suggesting that the surgical group may have been self-selected. Among other variables, epidemiological data must include smaller facilities blossoming in the United States and other developed countries that perform presurgical evaluations and possess surgical capabilities for performing ATL.
Furthermore, very few population-based studies exist that followed the natural history of medically treated TLE with HS irrespective of age. Available data originating from large well-equipped surgery centers suggest that medically treated TLE without surgical intervention is highly variable, often with unpredictable transient remissions and low rates of seizure freedom (5, 6). The International Classification of HS in mTLE as proposed by a recent International League Against Epilepsy task force report should also be discussed in terms of identifiable and hippocampal subfield compartmentalized HS and/or gliosis, or apparent lack of changes found on MRI datasets as assessed by visual inspection (2). The natural history of patients medically treated without surgical intervention should be followed in each of these histopathological categories and then correlated with neuroimaging characteristics potentially not shared across all categories. Moreover, hippocampal formation volumetry techniques of high-resolution MRI datasets should be consistently utilized rather than visual inspection alone.
Helmstaedter et al. considered the idea of disease-modifying factors that include availability of an increased selection of AEDs since 1988. Although etiology and failure of first and second drugs continue to be the most important predictors for treatment prognosis, these data should be evaluated in the context of delaying surgery as a therapy alternative, simply due to more AED treatment options. Overall, potential disease modifying factors are attractive but currently lack population-based epidemiological studies.
Lastly, the chronic effects of neuromodulation—currently deployed worldwide as vagal nerve stimulation therapy and FDA approved in 1997—is arguably a well-known disease modifying therapy, with more than 100,000 implants worldwide. Neuromodulation is well-known to demonstrate delayed positive modifications to unstable pathological neural circuits. Although this therapy is not a first-line surgical treatment for mTLE, neuromodulation must be factored into the possible disease-modifying factors for the focal-onset epilepsies that include TLE.
With further research, we can better understand the relative plateau effect of ATL in patients with mTLE and the factors determining the natural history of mTLE with and without HS. Little prospective information is available for the prognosis of individual subtypes of TLE in tandem with hippocampal-related changes. Such work should include large population-based studies compared across countries and healthcare economies. Clarity of this plateau effect will follow if these trends also include potential disease modifying variables, such as long-term clinical outcome comparisons between medical therapy alone and surgical intervention.