Chronic inflammatory demyelinating polyneuropathy (CIDP) is such a remarkably heterogeneous disorder – one with diverse presentations, a highly variable disease course, and differing, often unpredictable treatment responses – that neurologist-researchers around the world have established research registries in recent years to prospectively study patients over time.
These registries – some of which have biobanks – have made some headway. Published papers, including one study that describes the frequency and clinical course of patients with atypical CIDP variants, have recently come from the Italian CIDP database, for instance.
The problem is, the registries are themselves a heterogeneous collection of databases. They all collect slightly different sets of data and have slightly different inclusion criteria, and “they don’t talk with each other,” said Jeffrey Allen, MD, of the department of neurology at the University of Minnesota, Minneapolis. And even in large countries, the registries don’t include sufficient numbers of patients to address some of the most pressing questions.
To really understand the rare immune-mediated disease, we “need a worldwide database, so that existing databases can talk with each other, and so that countries that don’t have registries, like the United States, have a place to [register] their patients,” he said.
This vision is nearing reality, with development of an international registry – a central, worldwide database for the collection of standardized data about the natural history of CIDP and long-term treatment outcomes – almost complete. Filip Eftimov, MD, PhD, who is chairing the registry’s steering committee from his home in the department of neurology at the Academic Medical Center in Amsterdam, anticipates that “beginning in early 2020, we will be able to enroll the first patients.”
Among the goals of the new registry, named INCbase: To improve diagnostic criteria, to better understand underlying immunologic problems and pathogenesis of the disease, and to identify clinical or biological variables that can be used to predict treatment response and monitor disease activity.
“We don’t want a registry that validates what we think we already know – we want to learn from the registry what we don’t know,” said Dr. Allen, who sits on the INCbase steering committee. To start, “we want to better understand the full spectrum of what CIDP is and what it isn’t, so that we can diagnose it early and avoid misdiagnosing patients with CIDP when they actually have something else.”
“We don’t want a registry that validates what we think we already know – we want to learn from the registry what we don’t know,” said Dr. Allen, who sits on the INCbase steering committee. To start, “we want to better understand the full spectrum of what CIDP is and what it isn’t, so that we can diagnose it early and avoid misdiagnosing patients with CIDP when they actually have something else.”
Over time, the identification of biomarkers should “help us to customize treatment protocols,” he said. “Therapy and treatment for CIDP is moving toward personalization. We want to know which therapy is best for an individual at each point in time.”
Two tiers of data collection
Three years ago, in 2016, Dr. Eftimov and five other physician-scientists involved in national CIDP registries got together at a meeting of the Inflammatory Neuropathy Consortium (INC) of the Peripheral Nerve Society in Glasgow for a debate entitled “Why my database is better than yours.”
The debate shifted from questions about who had the best database to questions about the underlying objective of such a debate. Wasn’t a central, mutually formed database a more worthwhile goal than “mine-versus-yours” comparisons? Two problems drove the discussion, said Dr. Eftimov. “One, we were scattered, in terms of everyone [doing things differently], and two, there were a lot of people who wanted to start a database or join one, and didn’t know which database to choose.”
The neurologists decided to organize a workshop to get consensus on criteria and standards for an international registry and biobank – on what kinds of data and biomaterials should be collected, for instance, and how the registry should be governed and run. The workshop, held in conjunction with a 2017 meeting of the European Neuromuscular Center, drew 24 people from 13 countries, including Dr. Allen. Out of the workshop – as well as follow-up task force work – came the plan for INCbase to become a “flexible” international registry that complements ongoing databases rather than replacing them and that includes two main tiers of data collection, each with a different level of complexity.
“We wanted something in between a top-down registry [that’s very strict and very costly, with all participants collecting exactly the same data] and one with very ‘free-like’ registration of [patients] in which everyone collects data they believe is necessary or have time to do,” Dr. Eftimov said. While the bureaucracy of the first can cause weariness over time, the looseness of the second can lead to “dirty data,” he said.
INCbase, named after the Inflammatory Neuropathy Consortium that hosted the database debate, will allow participating centers to use either a simple core research module or a more extensive research module for follow-up, the latter of which includes more data points and outcomes measures and encourages more frequent follow up. Use of disease-specific outcomes like the Inflammatory Rasch-built Overall Disability Scale (I-RODS) and quantitative grip strength are included in both tiers (see figure).
Requirements for the baseline visit also are the same in both modules and include various parameters to accurately describe demographics, clinical phenotype, and findings of diagnostic tests. At minimum, the registry calls for follow-up visits every 6 months, for at least 2 years.
The registry does not include any minimum requirements or standard protocols for nerve conduction studies, Dr. Eftimov noted. “There is tremendous variation in practice on the neurophysiology protocols used and sometimes there are differences in interpretation. And, in most cases, neurophysiology data has to be put manually into the database, which is very time consuming,” he said.
From a research perspective, moreover, the “value of all this data is rather limited as it’s difficult to compare ‘raw data’ from different centers. … Protocols are hardly standardized and different laboratory specific normative values are used to interpret the results,” Dr. Eftimov explained. “Right now, there’s no reliable way to control [for this].”
In the future, it may be possible to automatize the transfer of data from the EMG machine to INCbase, he said. In the meantime, “investigators will indicate whether and how diagnosis fulfills the neurophysiologic criteria by using the categories used in the EFNS/PNS [European Federation of Neurological Societies/Peripheral Nerve Society] guideline.”
In the United States, Dr. Allen is leading the effort to select and engage a group of American medical centers – no more than 10 or so to start with – to enroll patients in INCbase. Most if not all of the participating centers will have the Centers of Excellence designation of the GBS/CIDP [Guillain-Barré] Foundation International, he said.
Dr. Eftimov said he expects current databases and registries outside of the United States to continue to exist. “Every center or participating registry will [essentially] be a sponsor of INCbase and remains the owner of their data and biomaterials,” he said. “What we’re doing is setting up the infrastructure to allow everyone to [participate] in some way.”
Diagnosis and biomarkers
INCbase’s extended research module is an adaptation of the Netherlands’ own 4-year-old registry and prospective study, called ICOS (the International Chronic Inflammatory Demyelinating Polyneuropathy Outcome Study). Initiated in 2015 at three Dutch academic medical centers (with a 1-year pilot phase) with the idea of eventually growing internationally, ICOS drew upon the basic design of the 7-year-old IGOS (International GBS Outcome Study), which now includes more than 1,800 patients in 22 countries, including the United States, and which has documented regional variations and current treatment practices in recently published papers. ICOS has enrolled more than 210 patients thus far – at least half of whom have newly diagnosed CIDP – for a follow-up of at least 2 years. (Patients who enrolled initially are currently at 4 years of follow-up.)
A key component of the ICOS has been its broad inclusion criteria, said Bart C. Jacobs, MD, PhD, who serves on the steering committees for ICOS and INCbase and who leads IGOS. “We’re still not sure of the diagnostic boundaries of this syndrome, so we need to have an extensive group of patients [so we can] use the data we collect to [improve] the diagnostic criteria and identify subgroups.” (Broad inclusion criteria that center on there being a clinical suspicion of the disease have been similarly developed for INCbase.)
A study coauthored in 2015 by Dr. Allen shed important light on diagnostic difficulties, Dr. Jacobs said. The retrospective study of 59 consecutive patients referred with a diagnosis of CIDP found that almost half (47%) of these patients did not meet the 2010 EFNS/PNS diagnostic criteria.
Diagnostic errors were magnified in clinically atypical patients and involved an overreliance on patient-reported perception of treatment benefit, liberal electrophysiologic interpretation of demyelination, and an overstated importance placed on mild or moderate cytoalbuminologic dissociation, Dr. Allen and his coinvestigator reported.
The first study results from ICOS, not yet published, show a median time for diagnosis of 11 months, which Dr. Jacobs called remarkable. “For a large number of patients, the diagnosis is delayed for more than a year,” he said. “This is important because it’s treatable. We want to prevent the process of axonal degeneration that may be irreversible.”
Ongoing analyses are attempting to tease apart diagnostic difficulties. “We’re now evaluating all the contributions of the clinical presentation, the diagnostic value of the CSF [cerebrospinal fluid] examination and nerve conduction studies, the evaluation of the treatment response, and the atypical variance of CIDP, to get a better grip on what the pitfalls are and what’s causing the diagnostic delay,” said Dr. Jacobs of the department of neurology and immunology at Erasmus Medical Center in Rotterdam, the Netherlands.
“It will also be very interesting to see how those patients who don’t strictly fulfill the EFNS/PNS criteria or who have an atypical form respond to current treatment over a long period of time,” he said, adding that he’s also interested in examining the pharmacokinetics of intravenous immunoglobulin to better understand the large variation in individual IVIg elimination rates.
ICOS is collecting blood samples at entry and additional serum samples at consecutive visits, as well as CSF and material from nerve biopsies when a lumbar puncture or biopsy is performed as part of routine diagnostic workup. (The registry has a “substantial number” of CSF samples, Dr. Jacobs noted.) In the effort to investigate potential biomarkers, serum samples will be tested for neurofilament light chain (NfL), which has been shown in preliminary research to reflect disease activity in patients with CIDP by detecting ongoing neuroaxonal damage.
ICOS is also planning to measure antibodies against the contactin-1 protein (CNTN1) and contactin-associated protein 1 (CASPR1), Dr. Jacobs said. Detection of these autoantibodies was reported several years ago by Luis Querol, MD, PhD, of the Hospital de la Santa Creu and Universitat Autonoma de Barcelona in a subset of patients with CIDP, and “it has been shown that these patients do not respond to IVIg very well,” Dr. Jacobs said.
ICOS researchers also will investigate the genome whenever possible “to see if there are risk factors that help us better understand the pathogenesis,” he noted.
INCbase has similarly established itself to be able to investigate potential biomarkers and search for new ones. “We’ll have very well-structured patient data at [consistent time-points], and very clean metrics about disease activity and treatment response,” said Dr. Eftimov. “Secondly, we’ll have biomaterials. For some [potential] biomarkers, you need large numbers of samples to discover whether something is useful or not, and for others, you need fewer.”
He said he hopes that in a few years, registry research will produce “a panel of serum biomarkers – a kind of menu” that could be selected from, perhaps based on identified subgroups, to help personalize treatment and monitor treatment response. “Right now we have nothing [to definitively use as biomarkers],” Dr. Eftimov said. “But there are lot of ideas about [potential] serum biomarkers.”
Dr. Eftimov hopes that INCbase participants will collect biomaterials to the extent possible, with at least serum collected at baseline (existing registries have taken variable approaches). However, for various logistical, financial, and other reasons, the INCbase steering committee decided not to set strict requirements for biobanking. “Everyone can choose whether or not to do any biobanking,” he said.
“We want to know when patients respond to various treatments, and when they don’t. We know to know when relapses occur, or perhaps when a patient enters a state of remission and no longer needs immunotherapy. We also want to learn how treatment responses are best measured. We want to make sure that data we collect is meaningful,” he said.
A versatile database
Dr. Eftimov emphasizes that INCbase is versatile. Participating centers will be able to adjust the database to meet their research interests – to add questionnaires on diet, for instance – and can extract data at the center level to share with patients and/or use in patient charts. “We’ve been designing this to avoid double-entry [of data into a patient chart and then in a research database]. That’s been one of the most frustrating parts of being a researcher and a doctor at the same time,” he said. “Ideally, you’ll use [INCbase] as your patient chart.”
Patients themselves can have access to the database prior to their medical center visits in order to fill out various patient-reported outcomes surveys, he noted.
INCbase and ICOS are currently connected in both objectives and infrastructure, and in the future, data from ICOS will be compatible with data from INCbase. Efforts will be made to allow other large registries, such as the Italian CIDP database, to combine data with INCbase, Dr. Eftimov said.
INCbase has some support from the pharmaceutical industry – i.e. CSL Behring, Shire, Terumo BCT, and Kedrion – both for overall development and for the funding of particular study proposals made by INCbase investigators (e.g., treatment satisfaction with subcutaneous immunoglobulin therapy). But Dr. Eftimov stressed that the international registry is “not a pharma-driven project.”
Centers will be able to share data and will be able to opt in or out of broader research analyses and common projects approved by the INCbase steering committee.
In the United States, one of the biggest challenges with participation in INCbase will be “to make sure we capture all the key moments in patients’ outcomes [throughout follow-up],” Dr. Allen said.
“We want to know when patients respond to various treatments, and when they don’t. We know to know when relapses occur, or perhaps when a patient enters a state of remission and no longer needs immunotherapy. We also want to learn how treatment responses are best measured. We want to make sure that data we collect is meaningful,” he said.
Dr. Allen disclosed that he is a consultant and advisory board member for CSL Behring, Biotest, Alexion, Akcea, and Argenx. Dr. Eftimov disclosed that he has investigator-initiated grants for INCbase with the companies listed in the story; all funding is payed to his institution. Dr. Jacob disclosed that his CIDP research is sponsored by the Netherlands Organisation for Health Research and Development, Princess Beatrix Muscle Disease Fund, GBS/CIDP Foundation International, Grifols, CSL-Behring, Annexon Biosciences, and Hansa Biopharma.