Good interobserver and intraobserver agreement in the evaluation of the new ILAE classification of focal cortical dysplasias.

PURPOSE: An International League Against Epilepsy (ILAE) consensus classification system for focal cortical dysplasias (FCDs) has been published in 2011 specifying clinicopathologic FCD variants. The aim of the present work was to microscopically assess interobserver agreement and intraobserver reproducibility for FCD categories among an international group of neuropathologists with different levels of experience and access to epilepsy surgery tissue. METHODS: Surgical FCD specimens covering a broad histopathology spectrum were retrieved from 22 patients with epilepsy. Three surgical nonepilepsy specimens served as controls. A total of 188 slides with routine or immunohistochemical stainings were digitalized with a slide scanner to allow Internet-based microscopy review. Nine experienced neuropathologists were invited to review these cases twice at a time gap of 3 months and different orders of case presentation. The 2011 ILAE FCD consensus classification served as instruction. Kappa analysis was calculated to estimate interobserver and intraobserver agreement levels. In a third evaluation round, 21 additional neuropathologists with different experience and access to epilepsy surgery reviewed the same case series. KEY FINDINGS: Interobserver agreement was good (κ = 0.6360), with 84% consensus of diagnoses during the first evaluation (21 of 25 cases). Kappa values increased to 0.6532 after reevaluation, and consensus was obtained in 24 (96%) of 25 cases. Overall intraobserver reproducibility was also good (κ = 0.7824, ranging from 0.4991 to 1.000). Fewest changes in the classification were made in the FCD type II group (2.2% of 225 original diagnoses), whereas the majority of changes occurred in FCD type III (13.7% of 225 original diagnoses). In the third evaluation round, interobserver agreement was reflected by the level of experience of each neuropathologist, with κ values ranging from moderate (0.5056; high level of experience >40 cases/year) to low (0.3265; low level of experience <10 cases/year). SIGNIFICANCE: Our study achieved a good and reliable interobserver agreement among the group of expert neuropathologists originally involved in the ILAE FCD consensus classification system. Intraobserver reproducibility in this group was even more robust. These results showed considerable improvement compared to a previous study evaluating the 2004 Palmini FCD classification. Agreement levels were lower in our second group of neuropathologists and were related to their level of access and experience with epilepsy surgery specimens. These results suggested that the more precise ILAE definition of FCD histopathology patterns improves operational procedures in the diagnosis of FCDs. On the other hand, microscopic assessment of FCD is a challenge and requires sustained experience and teaching. The virtual slide review system allowed testing of this hypothesis and reached a widespread group of participating colleagues from different centers all over the world. We propose to further use this tool as a teaching device and also to address other epilepsy-associated entities still difficult to classify such as hippocampal sclerosis, long-term epilepsy-associated tumors, or mild malformations of cortical development (mMCDs), which were not yet covered by current ILAE classification systems.

Progressive supranuclear palsy diagnosis and confounding features: report on 16 autopsied cases.

We evaluated 16 (15 men, 1 woman) autopsy-verified progressive supranuclear palsy (PSP) cases during 31 years (1969-2000) for clinical diagnosis and the course of the disease. The onset was gait difficulty or postural instability in 9 (56.3%), general motor slowing in 3 (18.8%), and tremor in 2. One case had onset with cognitive decline and 1 as hemidystonia. Four cases had supranuclear ophthalmoplegia (SNO) at the first assessment and were diagnosed as PSP. By last assessment, PSP diagnosis was made in 4 additional cases, but in 8 (50%) who never manifested ophthalmoplegia (mean 9.8 years after onset), PSP diagnosis was not made. Other manifestations included bulbar symptoms in 13 (81.3%), and cognitive impairment in 10 (62.5%) during the course of illness. Fifteen cases received levodopa, amantadine, anticholinergics, dopamine agonists, and selegiline in different combinations with symptomatic benefit in 9 of 15 (60%). Five had some improvement on levodopa alone and 3 showed more improvement when a dopamine agonist was added to levodopa. In general, the benefit was minimal and occurred only early in the course of illness. The mean age at onset was 63.7 (range, 53-85) years. Mean duration at PSP diagnosis was 4.8 (range, 2-11) years. Mean survival was 8.6 (range, 3-24) years and mean age at death was 72.3 (range, 60-89) years. When the different diagnostic criteria recommended in the literature were used, the accuracy of clinical diagnosis did not improve substantially.

Clinical-pathological study of levodopa complications.

We sought to determine the continued benefit and the pattern of motor complications of long-term levodopa treatment in Parkinson's disease. Patients were evaluated between 1968 and 1996. Only those who had an adequate levodopa trial and in whom autopsy revealed Lewy body Parkinson's disease were included. Total levodopa and mean daily dose were calculated in each case. Dyskinesia, wearing-off and on-off were collectively classified as motor adverse effects and reported as cumulative incidence. Forty-two patients (male, 30; female, 12) with mean 15.9 years of illness and 9.1 years follow-up received on average 500-mg levodopa daily over 9.8 years. Seventeen of 21 patients assessed during the last 18 months of life reported some motor benefit. Adverse effects were seen in 71.4% of patients. The most common was dyskinesia, in 61.9%; wearing-off in 35.7%; and on-off in 16.7% of patients. The earliest adverse effect was dyskinesia and the last to emerge was on-off. Isolated dyskinesia was seen in 35.7% and wearing-off in 7.1% of patients; 15.5% of patients developed dyskinesia after 2.6 years and 31% after 6.4 years on levodopa. We concluded that levodopa benefit declined and adverse effects increased with time. Dyskinesia was the earliest and the most common isolated adverse effect.