Meaningful Results in a Jiffy – A PERC of Multicenter Collaborations

Commentary

How Should Children With West Syndrome Be Efficiently and Accurately Investigated? Results from the National Infantile Spasms Consortium.

Wirrell EC, Shellhaas RA, Joshi C, Keator C, Kumar S, Mitchell WG; Pediatric Epilepsy Research Consortium. Collaborators: Berg A, Brooks-Kayal A, Coryell J, Dlugos D, Gaillard WD, Goodkin H, Grinspan Z, Jansen L, Knupp K, Kossoff E, Hartman AL, Joshi S, Loddenkemper T, Millichap JJ, Mytinger J, Nickels K, Nordli D, Childrens L, Ryan N, Sanchez-Fernandez I, Sullivan J, Valencia I, Wong-Kisiel L, Wusthoff C, Yozawitz E. Epilepsia 2015;56:617–625.

OBJECTIVE: To prospectively evaluate the etiology of new-onset infantile spasms and evaluate the yield of genetic and metabolic investigations in those without obvious cause after initial clinical evaluation and magnetic resonance imaging (MRI). METHODS: Twenty-one U.S. pediatric epilepsy centers prospectively enrolled infants with newly diagnosed West syndrome in a central database. Etiology and investigations performed within 3 months of diagnosis were documented. RESULTS: From June 2012 to June 2014, a total of 251 infants were enrolled (53% male). A cause was identified in 161 (64.4%) of 250 cases (genetic, 14.4%; genetic-structural, 10.0%; structural-congenital, 10.8%; structural-acquired, 22.4%; metabolic, 4.8%; and infectious, 2.0%). An obvious cause was found after initial clinical assessment (history and physical examination) and/or MRI in 138 of 161, whereas further genetic and metabolic studies were revealing in another 23 cases. Of 112 subjects without an obvious cause after initial evaluation and MRI, 81 (72.3%) had undergone genetic testing, which showed a causal abnormality in 23.5% and a variant of unknown significance in 14.8%. Although metabolic studies were done in the majority (serum, 79.5%; urine, 69.6%; and cerebrospinal fluid [CSF], 38.4%), these revealed an etiology in only five cases (4.5%). No correlation was found between type of health insurance (public vs. private) and either genetic or metabolic testing. SIGNIFICANCE: Clinical evaluation and MRI provide a specific diagnosis in 55% of children presenting with West syndrome. We propose that a cost-effective workup for those without obvious cause after initial clinical evaluation and MRI includes an array comparative genomic hybridization (aCGH) followed by an epilepsy gene panel if the microarray is not definitive, serum lactate, serum amino acids, and urine organic acids.

Response to Treatment in a Prospective National Infantile Spasms Cohort.

Knupp KG, Coryell J, Nickels KC, Ryan N, Leister E, Loddenkemper T, Grinspan Z, Hartman AL, Kossoff EH, Gaillard WD, Mytinger JR, Joshi S, Shellhaas RA, Sullivan J, Dlugos D, Hamikawa L, Berg AT, Millichap J, Nordli DR Jr, Wirrell E; Pediatric Epilepsy Research Consortium. Ann Neurol 2016;79(3):475–484.

OBJECTIVE: Infantile spasms are seizures associated with a severe epileptic encephalopathy presenting in the first 2 years of life, and optimal treatment continues to be debated. This study evaluates early and sustained response to initial treatments and addresses both clinical remission and electrographic resolution of hypsarrhythmia. Secondarily, it assesses whether response to treatment differs by etiology or developmental status. METHODS: The National Infantile Spasms Consortium established a multicenter, prospective database enrolling infants with new diagnosis of infantile spasms. Children were considered responders if there was clinical remission and resolution of hypsarrhythmia that was sustained at 3 months after first treatment initiation. Standard treatments of adrenocorticotropic hormone (ACTH), oral corticosteroids, and vigabatrin were considered individually, and all other nonstandard therapies were analyzed collectively. Developmental status and etiology were assessed. We compared response rates by treatment group using chi-square tests and multivariate logistic regression models. RESULTS: Two hundred thirty infants were enrolled from 22 centers. Overall, 46% of children receiving standard therapy responded, compared to only 9% who responded to nonstandard therapy (p < 0.001). Fifty-five percent of infants receiving ACTH as initial treatment responded, compared to 39% for oral corticosteroids, 36% for vigabatrin, and 9% for other (p < 0.001). Neither etiology nor development significantly modified the response pattern by treatment group. INTERPRETATION: Response rate varies by treatment choice. Standard therapies should be considered as initial treatment for infantile spasms, including those with impaired development or known structural or genetic/metabolic etiology. ACTH appeared to be more effective than other standard therapies.

As the old adage goes, there's no “I” in team—but there is a “me”—and that is too often the way epilepsy research has been carried out through the years. Frequently, the “me first” style has put individual accomplishments ahead of widespread collaboration, allowing the research to suffer and possibly delaying progress towards improved epilepsy care. While significant findings arise from single-center research, the investigation of rare disorders and the comparison of multiple approaches in evaluation and treatment require large quantities of patients, often more than a single center will collect in a reasonable timeframe. For years, the Children's Oncology Group has been a model of collaborative research, uniting multiple pediatric cancer centers in the common goal of evaluating cancer treatments and achieving a cure. While several multicenter collaborations have existed for epilepsy research, they infrequently survive beyond single studies and rarely have they focused solely on the investigation of pediatric epilepsy.

Enter the Pediatric Epilepsy Research Consortium (PERC), a collaborative network of 38 U.S. pediatric epilepsy centers formed in 2011 to share data and carry out protocols related to the diagnosis, evaluation, and treatment of pediatric epilepsy. The multicenter structure allows for rapid investigation of rare disorders and comparison of various approaches to diagnosis and treatment. Individual centers or small research groups conceive the topics of investigation and member institutions contribute data, thus uniting the ingenuity and research interests of a single-center with the strength of the larger epilepsy community. PERC has already produced several papers (1) and numerous abstracts with ongoing or planned research in Lennox-Gastaut syndrome, myoclonic astatic epilepsy, electrical status epilepticus of slow-wave sleep, early onset epilepsy of childhood, and infantile spasms.

Two recent publications from the National Infantile Spasms Consortium (NISC) of PERC underscore the power of their collaborative approach. With an incidence of 2 to 5 per 10,000 live births (2), the investigation of infantile spasms lends itself to pooled data, as one center would take years to generate any meaningful cohort for study. Between 2012 and 2014, 22 centers prospectively collected 282 patients with new onset infantile spasms between the ages of 2 to 24 months. In their initial study, the investigators evaluated the approach to investigating etiology in new onset West syndrome within the first 3 months from diagnosis. Diagnostic testing was ordered at the discretion of the treating neurologist, with the goal to quantify what is being done in the “real world” and assess the value of individual tests, ultimately leading to more refined protocols. All patients had initial evaluation by a pediatric neurologist, and neuroimaging was available for 92%. Following initial investigation, 55% had an obvious cause of infantile spasms that included causative abnormality on MRI, dysmorphic features diagnostic of genetic conditions, or diagnosis of a condition highly associated with infantile spasms (i.e., tuberous sclerosis). It is not surprising that MRI revealed an etiology in half the patients and should be considered as a first-line test in evaluating new onset spasms. More importantly, their results support thorough exam and imaging before extensive metabolic/genetic testing, which can be quite costly and often may not be necessary.

Amongst children without obvious cause after initial evaluation, more underwent further serum, urine, and CSF metabolic testing, yet the yield was quite low (4.5%). When positive results were found, they were often in children with imaging abnormalities that suggested a metabolic disease. Therefore, targeted metabolic testing may be worthwhile in a subset of patients, but given the turnaround time for results of many metabolic tests, the low yield does not justify earlier testing unless the clinical suspicion is very high. Further genetic testing was performed in 56% of the patients (72% of those without cause at diagnosis). Testing was not standardized and included karyotype, array comparative genomic hybridization (aCGH), extensive epilepsy gene panels, and whole exome sequencing amongst other tests. Twenty-three percent of patients without cause at diagnosis had a causative abnormality identified on genetic testing, the majority using aCGH and epilepsy panels. This supports earlier use of genetic testing over metabolic testing in most cases. Twenty-eight percent of patients without cause did not have further genetic testing, though this may be related to the short duration of the study (3 months). As more extensive gene panels and whole exome sequencing were used only in a minority of patients, it is unclear how much higher the positive yield of genetic testing might be if these were standard practice. No doubt, evaluating the yield of this type testing in a more standardized approach will be in PERC's future.

In their second paper, the consortium characterized the response to treatment for new onset infantile spasms, including standard treatments of ACTH, oral steroids, and vigabatrin, as well as nonstandard therapies. The authors made a point to not only examine initial response to therapy but also included sustained response at 3 months. Again, choice of treatment and the regimen prescribed were at the discretion of the treating neurologist. The study revealed some treatment biases that are likely commonly practiced elsewhere: Children with a history of seizures prior to infantile spasm onset were more likely to receive nonstandard treatment over ACTH, children with structural causes were more likely to receive vigabatrin, and children with severe developmental delay at initial diagnosis were less likely to receive standard therapies. However, standard therapies were superior to nonstandard therapy, and the likelihood of response was more strongly associated with the choice of treatment than the etiology and presence of delays at onset. This finding alone unveils a potential deficiency in our approach to patient care based on bias and should serve as an impetus for process improvement.

While the study included a suggested dosing protocol for ACTH, it did not offer a protocol for other treatments used, and adherence to the dosing regimen was not required. Thus, the study was not designed to compare dosing regimens but instead compared treatment arms. Regardless, the fact that 22 different institutions could agree on a single recommended dosing protocol is an impressive step forward. ACTH was more often associated with higher rates of response compared to vigabatrin and oral steroids, and 3-month response rates also remained higher. Given the large number of patients recruited in such a short duration, it is likely that more standardized comparison of dosing regimens could be achieved by the group in time.

The initial research production of PERC has demonstrated the advantages of multicenter collaboration. Their early work characterizes the landscape of infantile spasm evaluation and treatment, providing a foundation to further investigate more refined protocols. Hopefully, soon will come the day when all patients with epilepsy are evaluated within a research protocol aimed at delineating the most cost-efficient and efficacious approach to care. It will only be through large-scale collaboration such as this, that we effectively and efficiently improve outcome.