8-Point Ordinal SCORE Scale Predicts SLAH Success

In a retrospective cohort, a clinical heuristic composed of eight binary variables helped identify which patients with medial temporal lobe epilepsy (MTLE) were best suited for stereotactic laser amygdalohippocampotomy (SLAH). Although identifying individual predictors of SLAH success has proven difficult, said authors led by Adam S. Dickey, MD, the novel scale accurately augurs post-procedural seizure freedom. The study was published online on July 12 in Annals of Clinical and Translational Neurology.
In devising the SCORE rubric, investigators combined classical descriptors of MTLE with additional measures to overcome statistical issues, including their small sample size. In a study published in Annals of Neurology in 1993, all 67 patients with MTLE were seizure free after an anterior temporal lobectomy (ATL).
The SCORE model awards 1 point for the presence of the following variables:
History of infantile febrile seizures
Onset age ≤ 16 years
Generalized tonic-clonic seizures rare or absent
No auditory, vertigo, or visual aura
Ipsilateral temporal interictal epileptiform discharges (IEDs)
Ipsilateral temporal ictal onset
Ipsilateral mesial temporal sclerosis (MTS) on an MRI
Ipsilateral temporal PET hypometabolism
Based on seizure-free percentages at various SCORE levels, investigators found that patients with a total SCORE ≥ 6 (77%) achieved good (Engel class I) outcomes. Conversely, patients with SCORE = 5 (54%) and SCORE ≤ 4 (31%) achieved fair and poor outcomes, respectively.
Comparable to Standard Open ATL
“Our 8-point ordinal score suggests that a patient with 6 or more concurrent data points has a 70%-80% chance of being seizure free, which is comparable to the historical seizure-free rate of a standard open anterior temporal lobectomy,” wrote Dickey and colleagues. Conventional ATL is a more invasive procedure that offers approximately a 10% better chance of seizure freedom than SLAH in all comers, they added, but with longer hospital stays and a greater likelihood of permanent cognitive sequelae.
Investigators reviewed charts of 101 consecutive patients who underwent SLAH between July 2011 and October 2019 at Emory University in Atlanta. All patients underwent comprehensive presurgical evaluation using standard tools ranging from long-term inpatient scalp video electroencephalographic monitoring to neuropsychological testing. Multiple sensitivity analyses yielded similar findings, although the SCORE model was not statistically superior to all alternative models analyzed.
MRI Alone Is Insufficient
Overall, authors said, their calculations highlight the relative value of MRI vs other modalities in epilepsy-related surgical decision-making, which can devolve into polarizing debates around whether a particular lesion accurately represents a patient’s epilepsy. “Given the highly focal nature of SLAH, which targets only the amygdala and hippocampus,” they wrote, “it is not surprising that patients with evidence of MTS have a higher chance of seizure freedom.”
In univariate analysis, MRI evidence of MTS and presence of unilateral IEDs were the only variables that showed statistical significance for predicting seizure freedom. IED presence also maintained significance in multivariate analysis. Although MTS modestly predicted better post-SLAH outcome, authors cautioned, “it is only 1 data point among many which should be considered.”
Aatif M. Husain, MD, professor of neurology and chief of the Division of Epilepsy, Sleep, and Clinical Neurophysiology at Duke University Medical Center in Durham, North Carolina, called the paper a prompt for epileptologists and neurosurgeons to “think not only of the MRI abnormality of MTS as the main marker of whether someone’s going to get better but also to look at the seven other things that potentially could have an impact. And in fact, unilateral epileptiform discharges seem to be equally important in predicting favorable outcomes as is MRI.” He was not involved with the study but was asked to comment.
Although general neurologists are unlikely to use SCORE, Husain said, its positive predictive value may help epileptologists and neurosurgeons in counseling patients referred from general neurology. “Whenever you have a conversation with the patient trying to decide optimal therapy,” he explained, “you want to give the patient, and to understand yourself, what the most likely outcome of this operation is. This tool helps you decide which patient approaches the outcome levels of standard ATL vs those that would be not just 10% less likely, but maybe 20%-30% less likely” to experience seizure freedom post-procedure.
Pragmatic Considerations
From a practical standpoint, Husain said, the SCORE tool incorporates procedures most patients being evaluated for epilepsy surgery already have undergone. However, he added, all eight variables are not always available.
“Many patients will not have a PET scan. If you have everything but that, then how does the score work?” Husain also questioned why all eight variables should carry equal weight. Using exclusively binary scores rather than gradations also limits the model’s utility, investigators allowed.
IEDs very rarely occur 100% unilaterally, added Husain. The distribution is usually around 90% on one side, 10% on the other. “It’s unclear from this paper whether 90-10 also works.”
Additionally, he said, the model omits some important variables, such as seizure semiology (eg, lip-smacking or posturing). “All of those observations carry weight and are valuable from a clinical perspective,” said Husain. Given such limitations, authors wrote, statistical tools must never replace clinical judgment.
Study authors were supported by the National Center for Advancing Translational Sciences and the National Institute of Neurological Disorders and Stroke, both part of the National Institutes of Health. Husain reported no relevant financial relationships.
John Jesitus is a Denver-based freelance medical writer and editor.
 
Send comments and news tips to [email protected].