WASHINGTON – The benefit of implanting a responsive brain stimulator for the control of refractory epilepsy may be grossly underestimated without relying on an objective measure of baseline seizure activity rather than patient reports, according to a study presented at the annual meeting of the American Epilepsy Society.
In a retrospective evaluation at one center, the efficacy of the Responsive Neurostimulation System (RNS) came nowhere near that observed in the pivotal clinical trial until objective measures of seizure activity were analyzed, reported Michael Young, DO, a neurophysiology fellow in the department of neurology at the University of California, Irvine (UCI).
In this study, investigators evaluated seizure frequency in the first 2 months after RNS implantation with the ECoG component of the RNS device. They assessed change in seizure frequency relative to this baseline at 3, 6, and 12 months, and also compared the reduction in seizures against the patient self-report of baseline seizure activity.
The differences were large. On patient report, the reduction in seizure activity at month 3 was just 10%, compared with 85% when measured on ECoG.
“Our results with the RNS compare favorably to the pivotal trial only when using the ECoG seizure frequency baseline. The reason for this discrepancy is due to underreporting of seizures by patients and consequently a falsely low seizure frequency,” Dr. Young explained at the meeting.
The RNS system has been implanted for refractory focal or partial seizures in adult patients at UCI since 2015. The device is indicated for adjunctive use in patients not adequately controlled on at least two antiepileptic medications. Twelve patients have been treated, but two were excluded from this analysis because they had surgical resection at the time of the RNS implantation and one because of an infection related to the implantation.
In general, patients treated at UCI had characteristics similar to those in the pivotal trial, which was published more than 3 years ago (Epilepsia. 2014;55:432-41). In that 191-patient trial, the reduction in seizure frequency at the end of 5 months of blinded analysis with RNS was 37.9% versus 17.3% for a sham procedure. Progressive further reductions in seizure activity were observed during an extended open-label follow-up.
In the UCI analysis, the mean reduction in seizure frequency at 12 months was 56% relative to the patient-reported baseline but 78% on the basis of the ECoG analysis. Although only four of the nine patients have 12 or more months of follow-up, three were considered to be responders to RNS whether evaluated in relation to the patient-reported baseline seizure activity or in relation to ECoG. The responder rate at 3 months on the basis of patient-reported baseline activity, however, was only 56%, compared with 100% based on ECoG.
“The big issue is underreporting of seizures by patients,” Dr. Young explained. He cited numerous other studies demonstrating the same phenomenon. He noted that noncompliance is only one reason patients underreport. In many cases, patients are simply unaware of seizure activity.
Based on these data, “we think ECoG may be a more objective way to track patient response to RNS,” Dr. Young said. He acknowledged that the number of patients limits this study and suggested that larger studies are needed to confirm the findings.
Dr. Young reported having no potential conflicts of interest related to this topic.
SOURCE: Young M et al. AES abstract 3.109.