- Email: [email protected]
CNS biomarkers: Potential from a regulatory perspective
European Neuropsychopharmacology (2015) 25, 1003–1009
www.elsevier.com/locate/euroneuro
CNS biomarkers: Potential from a regulatory perspective Case study – Focus in low hippocampus volume as a biomarker measured by MRI Maria Isaaca,n,1, Christine Gispen-de Wiedb a
European Medicines Agency (EMA), United Kingdom Medicines Evaluation Board (MEB), The Netherlands
b
Received 6 October 2014; accepted 22 March 2015
KEYWORDS
Abstract
Biomarkers; Qualification advice; Qualification opinion; EMA; FDA
Our objectives are to describe the procedure for qualification advice and opinion from EU regulators on the use of novel methodologies in drug development, the key stakeholders involved and the evidence requirements for qualification opinion. We present a case study of the request from the Coalition Against Major Disease (CAMD) Consortium of the Critical Path (C-Path) Institute for EU regulators' qualification opinion on the use of low hippocampal volume as a biomarker for population enrichment in clinical trials of novel drugs in Alzheimer's disease (AD). We discuss the main concerns from the regulators, data analysis requests and guidance during the qualification. EU regulators concluded that low hippocampal volume, measured by vMRI and considered as a dichotomized variable (low volume or not), appears to help enriching recruitment into clinical trials aimed at studying drugs that potentially slow the progression of the pre-dementia stage of AD. The biomarker qualification procedure is a dynamic process in which pharmaceutical companies and research consortia can submit further data to update the qualifications and improve the predictive value of the biomarkers. & 2015 Elsevier B.V. and ECNP. All rights reserved.
n
Corresponding author. Tel.: +44 36607153; fax: +44 36607040. E-mail address: [email protected] (M. Isaac). 1 http://www.ema.europa.eu. http://dx.doi.org/10.1016/j.euroneuro.2015.03.013 0924-977X/& 2015 Elsevier B.V. and ECNP. All rights reserved.
1004
1.
M. Isaac, C. Gispen-de Wied
Introduction
The European Medicines Agency (EMA) coordinates the activity of the National Competent Authorities in the 28 Member States of the European Union in the framework of centralised evaluation procedures of medicinal products. Its work is underpinned by several Committees and Working Parties of the Agency, supported by the European Experts' Network including over 4900 nominated experts. For the purpose of the qualification procedure, the key agents involved are the Committee for Human Medicinal Products (CHMP), its standing Scientific Advice Working Party (SAWP), the CNS Working Party (CNSWP), the Biostatistics Working Party (BSWP), Modelling and Simulation Working Group (MSWG). The objective of the qualification of novel methodologies procedure can be a way of streamlining the marketing authorisation procedure (Figure 1). In operational terms scientific advice requests on the acceptability of the use can follow two parallel work streams, either through the ‘regular’ product- and indication-specific scientific advice (SA) or protocol assistance procedures with fixed timelines of 40 or 70 days (the latter including a discussion meeting between the SAWP and the applicant, e.g. in case of major disagreements or if clarifications are needed), and the ‘dedicated’ Qualification advice and opinion procedures, with a focus on a particular technology and eventually a broader scope which may span across several indications or products (EMA, EMA guidance for companies requesting SA or PA, EMA Qualification of Novel Methodologies). Although in general terms the main focus of the advice given by the SAWP rests on methodology, and pre-assessment of data intended to support a marketing authorisation application of a drug is formally not within the scope of the regular SA procedure, within the qualification procedure assessment of data has a key role. Due to the complexity of qualification procedures, requiring increased dialogue with the applicant, the timelines and proceedings are more flexible, often involving several ad hoc face-to-face meetings. A further important difference is that with the applicant's agreement, the outcome of the qualification opinion, and in the near future also of the qualification advice (letter of support) can be made public, in contrast to the outcome of the regular SA, which remains confidential at least until the product has received marketing authorisation for the concerned indication.
Figure 1
Dialogue with regulators timelines.
Letter of support: Based on the qualification advice a letter of support may be proposed by EMA as an option, when the novel methodology under evaluation cannot yet be qualified but is shown to be promising based on preliminary data. This letter is a positive regulatory feedback, including a high level summary of the novel methodology, context of use, available data and on-going/future investigations, meant to be made publicly available on the EMA website subject to sponsor's agreement. The objective of the letter of support is to encourage the efforts for data sharing and facilitate studies towards qualification for the novel methodology under evaluation. The scope of the biomarker qualification procedure in preclinical development involves pharmacological screening, understanding the mechanism of action, predictions of activity and safety, pharmacokinetic (PK) and pharmacodynamic (PD) modelling and toxicogenomics. Clinical development involves a verification of the mechanism of action, together with characterisation of the dose response effect, proof of concept in the therapeutic context, enriching the study population, considering surrogate endpoints and maintaining the paramount of safety by early detection of safety signals. Drug utilisation involves optimisation of the target population and guidance of the relevant treatment regimen. A CHMP qualification opinion on the acceptability of specific use of, say, a biomarker in a research and development context (whether clinical or non-clinical) is based on the assessment of data and is not specific to the individual product. The process involves a qualification team, peer review, public consultation and publication. CHMP qualification advice on future protocols and methods for further method development towards qualification is based on the evaluation of the scientific rationale and on preliminary data submitted. This process is confidential. The aim is that SAWP and CHMP are involved early in strategy design, with a commitment to evaluate data from agreed studies and to provide continuing opinion. The scope remains the focus on acceptability of specific use of the proposed technology or biomarker, developed for a specific intended use in the context of pharmaceutical research and development. The logistics of the qualification team include a coordinator, part of SAWP or CHMP, who facilitates a minimum of four experts from various disciplines to evaluate the therapeutic areas involved, the context of the intended use (for example, non-clinical safety testing or translational research), the relevant technology platform (proteomics, genomics, imaging) and statistics. A project manager is supplied by EMA. Again, the
Figure 2
Qualification procedure timelines.
CNS biomarkers: Potential from a regulatory perspective case study team members are selected according to expertise not Member State. EMA provides scientific, administrative and logistic support to the process and there is extensive networking within the agency. CHMP is involved before and during the procedure, underlining the commitment to evaluate the data obtained from the agreed studies. The timeline and the qualification team meetings can be adjusted case by case. In practice, such timelines can be as flexible as necessary. An example of the qualification procedure timeline is shown in Figure 2. Applicants can be private sector consortia, networks, public-private partnerships, learned societies, pharmaceutical industry and academia. Fee reductions can be granted for paediatric and orphan conditions, as well as to small and medium-sized enterprises (SME), and on a case-by-case basis the EMA Executive Director may grant a discretionary fee reduction based on grounds of public health benefit. EMA strongly recommends pre-submission meetings. These are important to facilitate communication then and later, and early help with amount and type of data to be submitted, questions to be answered and scope can be dealt with early. This can help the applicant to clarify its strategy and direct accordingly. Experience teaches that this usually improves outcomes. A pre-submission meeting can also help the applicant decide when and whether formally to start the procedure. An applicant can decide when the procedure starts and can also ask for the procedure to be suspended at any time (a ‘clock stop’). Again, at the applicant's request, meetings and teleconferences can be organised at any time in the procedure, as can additional data be presented. If a positive opinion is to be issued, the applicant is consulted on the amount and type of data to be made publicly available. There can be as many discussion meetings during the procedure as necessary, providing the applicant to justify its position and introduce new, significant data. A discussion meeting can also be used to influence the eventual outcome by presenting new relevant data or expert views of opinion leaders, justifying the applicant's strategy. The scientific administrators at EMA facilitate the process by providing scientific procedural and administrative support, informal teleconferences and meetings, help in organisation, expert meetings, out-of-the-procedure meetings and scientific discussion, providing a single contact point, able to radiate to other parts of EMA. There is also the possibility of parallel Food and Drugs Administration (FDA) and EMA qualification advice. This is voluntary at the sponsor's request, and involves questions on product development being put to both FDA and EMA, followed by discussions between the two Agencies and joint discussion with the sponsor. Each Agency, however, will issue separate responses to a sponsor's questions according to their individual procedures. The possible benefits of such an approach involve increased dialogue between Agencies and the sponsor from the early stages of development, together with the exchange of views and sharing of expertise and best practice. Moreover, parallel advice can optimise and facilitate global development, meeting the requirements of both Agencies (FDA, 2014). Optimising a request for qualification advice or opinion involves various practical considerations by the applicant, including the analytical platform, biological matrices and the timing for evaluation. Concerning biomarkers, the characterisation of the relationship between the biomarker response
1005
with normal biological processes or pathogenic processes, or pharmacological responses to a therapeutic intervention and the expected quantitative response threshold is necessary, together with the intended use of the biomarker in drug development and the clinic, especially deciding whether the use is to be exploratory or confirmatory, as well as its position in the preclinical, translational, toxicity, inclusion criteria for studies, stratification, enrichment, drug monitoring, primary and secondary endpoints, initiation of treatment or change of intervention nexus. Dossier in data can include, but not be limited to, protocols, study reports and raw data, all of which can establish the use of a defined methodology for a specific purpose in drug development. It is important that the range of opinion is represented so that the amount of data presented should not be over selective. There is a balance, however, between too little data and too much. The case needs to be presented in concise form in a separate document (the Briefing Document) with major points and data summaries in place. The main document must use references. Comprehensive data that are available at the time of submission should be presented, even if not controlled. As far as questions to be asked are concerned, here, as in life, the clearer the question, the clearer the answer. Questions that are too broad tend not to provide helpful answers. It is particularly important to ask about suspected pitfalls in the development, as well as choice of scope and various aspects of study design, such as outcomes, the study population and the relevant statistics. Qualification advice is peer reviewed and adopted by CHMP (or the relevant standing committee, if, say, advanced therapies or paediatrics are involved) and a competitive fee structure for the advice is also supplied, especially for orphan drugs. It is important to distinguish between CHMP qualification advice and CHMP qualification. Advice on future protocols and methods for further method development or its qualification is based on the evaluation of the scientific rational and on preliminary data. A qualification opinion on the acceptability of a specific use of the proposed method (such as a biomarker) and in research and development context is based on the assessment of submitted data. The immediate benefits of qualification advice include early identification of strengths and potential pitfalls and the opportunity to discuss them with leading experts in the field. This mostly involves a continuing dialogue, rather than a single interaction, so that issues can be revisited and agreement on what is required for ultimate regulatory acceptance of a new method can be discussed. In the long term, qualification advice speeds up the time to regulatory acceptance of novel approaches, as well as time to new marketing authorisation applications. This is an obvious potential saving in resources.
1.1.
Qualification advice
To date, qualification advice has been sought mainly for CNS products for conditions such as Alzheimer's disease, schizophrenia and Down's syndrome (17/29), with other qualification advice being sought from other disciplines, including nephrology, oncology, gastroenterology, chest medicine,
1006
M. Isaac, C. Gispen-de Wied
musculoskeletal medicine, cardiology and hepatology. The procedure flow is shown in Figure 1. To ensure flexibility, the product flow is adjusted on a case by case basis, with the timelines present for guidance, rather than as a prescription. The process can be suspended at any time at the applicant's request and meetings and teleconferences can be organised to discuss emergent issues. This individual tailoring is similar to that discussed above.
1.2.
Qualification opinion
A CHMP qualification opinion on the acceptability of a specific use of the proposed method, such as use of a biomarker, in a research and development context (non-clinical or clinical) is based on the assessment of submitted data. A qualification team, peer review and public consultation, followed by public publication, are involved and the context of use of the proposed method is clarified. As for qualification advice, private sector consortia, networks, public-private partnerships, learned societies, the pharmaceutical industry and academia can apply, with a vision of speeding up and optimising drug development and utilisation, thus improving public health. The opinion is not product specific and is based on the assessment of data. The initiation, consultation and qualification opinion itself are confidential until the qualification opinion recommendation is made public for market authorisation. The EMA procedure for qualification opinions involves discussion meetings and further assessment of data followed by a production of a document by SAWP with the working party's qualification opinion submitted to CHMP. CHMP will put the document out to peer review, after which it will be presented by an administrator or the chairman of SAWP during the CHMP plenary. This leads to the CHMP opinion, public consultation, feedback and a final opinion. As before, the process is adjusted case by case. This paper is a summary of the regulatory procedural thinking focus in a case. EMA published a qualification opinion on CSF biomarkers (see: Procedure EMEA/H/SAB/012/1/QA/2010; Isaac et al., 2011) and in October, EMA (2011) published a qualification opinion of low hippocampus volume (see Procedure EMA/ CHMP/SAWP/809208/2011; Hill et al., 2014; EMA, 2012, 2013).
2. Case study – qualification opinion of low hippocampal volume (atrophy) by MRI for use in clinical trials for regulatory purpose – in predementia stage of Alzheimer's disease (Figure 4) The data in support of this request for a qualification opinion on low HC (HC) volume were derived from an extensive literature search. The HC atrophy would be measured by MRI and proposed as a biomarker in predementia stage of Alzheimer's disease. CAMD with the help of consortium members conducted a systematic literature review (Higgins and Green, 2011) with the specific goal of identifying longitudinal studies that evaluated MRI-assessed HC atrophy as a biomarker in predicting conversion to AD dementia from a memory impaired state at baseline, either amnestic MCI or MCI determined by
criteria outlined in each specific manuscript. The literature search focused on data for those individuals who started out with cognitive impairment and progressed to AD dementia. Traditionally, cognitive measures have been used as determinants of likelihood of progression to dementia, usually AD dementia. As the Dubois approach suggests, the addition of supportive biomarker(s) could increase the accuracy of the prediction of progression from a prodromal AD state to AD dementia. The measurement tools used to determine the cognitive impairment varied among the studies reviewed. No attempt was made here to classify and weigh the specific methodologies used across the studies.
3.
Qualitative analysis
A systematic review was conducted of the published literature of the performance of HC volume, as measured by vMRI (volumetric magnetic resonance imaging), in predicting progression from a memory-impaired status to AD dementia. The literature search objectives were: to determine if baseline measurement of low HC volume (atrophy) by MRI in patients with episodic memory deficit (Dubois criteria) or MCI is a useful means of predicting those patients likely to evolve to AD. Specifically, CAMD incorporated here the identified literature that reports on the use of MRI-based volumetric measures of HC volume as a predictor of conversion from memory impairment status to AD type dementia.
4.
Quantitative analysis
Results: Volumetric MRI in Predicting Progression to AD. The focus of the accepted studies was the baseline HC atrophy in MCI subjects measured using vMRI with the investigative question regarding the ability of vMRI measurement of HC volume to predict progression to AD. Most studies were published in the last five years. The follow-up duration of all studies ranged from 18 months (Fellgiebel et al., 2006) to 108 months (Desikan et al., 2008). HC volume measurements were reported with various methods: as a total volume of both hippocampi, or as separate measurements for left and right hippocampi. The context of use specifies that the MRI based HC volume be measured at a single time point (baseline) and appropriate for subject inclusion into a clinical trial. Results: Hazard Ratios, Sensitivity, and Specificity of vMRI Biomarkers. The opinion to qualify was that low HC volume (atrophy) by MRI in patients with episodic memory deficit (Dubois criteria) was a useful mean of predicting those individuals likely to evolve to AD dementia during the course of an AD clinical trial. Specific data requested by EMA. Upon examination of all the data above EMA posed several questions to CAMD, the most relevant of them are briefly outlined below along with the answers provided by the Consortium. EMA requested that accuracy data should be represented in a classical ROC analysis and the information summarised from ADNI data (sROC model or average operating point along with criticism about its representativeness). CAMD agreed that the ROC analysis was a key evaluation parameter in the data prediction. However, the
CNS biomarkers: Potential from a regulatory perspective case study data included in the reported studies were insufficient to perform an AUC analysis on even a subset of the studies. Instead, the Consortium performed a de novo analysis on HC volume interpretation of MRI images generated from the ADNI study with the objective to determine whether HC volume quantification method significantly impacted on the prediction of conversion. The results of such review across the literature suggested that the HC volume quantification method (the one heterogeneous variable across the studies analysed here de novo) most likely did not impact on the interpretation of biomarker results. The CAMD was also asked to present the results of this de novo analysis based on the ADNI study with well-standardized MCI patient data, from four distinct HC volume analysis methodologies. In this case the variable being addressed was the volume reconstruction and its interpretation. The subject definition and image acquisition methods were standardized as described in the ADNI protocol. The de novo analysis presented the results in the requested ROC format. Additionally, the EMA requested the CAMD to provide the predictive value of other different biomarkers for pre-dementia stage of AD (Dubois et al., 2010). CAMD confirmed that several biomarkers are currently being used for the purpose of enrichment of clinical trials in pre-dementia stage of AD, including biochemical assessment of CSF Aβ42, tau and p-tau levels (see Isaac et al. (2011)), PET imaging of amyloid, metabolic activity (FDGPET), cerebral perfusion (SPECT) and structural neuroimaging. Each biomarker has advantages and disadvantages in terms of ease of use, cost and invasiveness to patients. Relative to the other biomarkers, vMRI was less invasive and had wider applicability, although was limited by the cost, availability and interoperability of the Neuroimaging machines and their software elaborated data across different centres and in different time. It seems that at present, the field was not yet at a stage to employ combinatorial imaging biomarker measurements beyond research purposes as highlighted in a recent webinar focused on early AD biomarkers (see http://www.alzforum.org/res/for/journal/ detail.asp?liveID=192). The EMA requested further information to compare the performance of CSF and HC volume in predicting conversion to AD. CAMD performed an additional analysis of the ADNI dataset and reported the results of progression at the two year time point. Comparison of the AUC values of baseline vMRI of hippocampus and CSF analytes (Aβ; Phospho-τ; Total τ; single and combined) indicated that vMRI in this dataset is as good as, or slightly better at, predicting conversion to AD than CSF analytes; specifically, a greater ROC AUC was observed for MRI than CSF biomarkers in this comparison. The analysis was a comparison of AUC values of baseline vMRI data predicting conversion to AD to AUC values of CSF analytes predicting conversion to AD in 2 years. The EMA was also concerned on the quantitative measures of the hippocampus in comparison to other brain regions in AD brain. Clinical Utility: The conclusions on the final qualification opinion was as follows: Low hippocampal volume, as measured by MRI and considered as a dichotomized variable (low volume or not), appears to help enriching recruitment into clinical trials aimed
1007
at studying drugs potentially slowing the progress/conversion to AD dementia of the included subjects. Low hippocampal volume might be considered a marker of progression to dementia in subjects with cognitive deficit compatible with predementia stage of AD (Dubois, 2010). However, neither the actual value of low hippocampal volume to accurately predict rate of such progression to dementia in the referred subjects nor the relative value of other biomarkers have been reported. “As currently planned in the current opinion subjects might be included in the studies based on clinical criteria and low hippocampal volume biomarker (if positive). The CHMP has given a previous positive opinion in the predementia stage of Alzheimer's disease: cerebro–spinal fluid related biomarkers for drugs affecting amyloid burden. This may lead first to a heterogeneous population and, moreover, it will not be possible to explore the relationship among them. Although not required form a regulatory perspective, the concomitant assessment of the two biomarkers in predementia stage of AD would be of great value.” “The process of measurement of low hippocampal volume is also complex. To obtain reliable results implies the standardisation of all steps (at least imaging acquisition protocol, imaging reconstruction/analysis methods, timing to conversion, etc). International guidelines have been produced. These guidelines must be enforced.”
5.
Discussion
Accepting the value of these biomarkers in “enriching” recruitment is, probably, less demanding than assessing its value in other potential uses such as diagnostic and thus, less accuracy in the prediction is required than to include a particular individual into a diagnostic category. The qualification of biomarkers to “enrich” recruitment into clinical trials aimed at studying drugs potentially slowing the progress/conversion to (AD) dementia of the included subjects has been used in drug development mainly in phase I and 2. The biomarkers qualification procedure is targeting recruitment in phase 3 studies. EU regulators felt that identifying subjects at higher risk of developing AD dementia (as intended in this procedure) maybe serve useful purposes even in the absence of effective treatments for the disease. Enroling “non-enriched” samples (basing inclusion only on the cognitive deficit) could mean that fewer subjects would convert during the duration of the trial. Impractically large numbers of subjects and/or duration of follow-up would be required and the trials would be unfeasible or inefficient. Other biomarkers to “enrich” recruitment into this type of clinical trials are already known (e.g. some CSF analytes). The purpose of this “qualification” procedure was to assess whether low HC volume as measured by MRI and considered as a dichotomized variable (low volume or not) could be considered as an enrich population for predementia stage of AD clinical trials. The potential value of the proposed marker in other settings as a criterion for the diagnosis of a disease such as AD in a particular subject or the usefulness of repeated measurements to assess the effect of therapeutic interventions – as a marker of efficacy – were not considered.
1008
M. Isaac, C. Gispen-de Wied
The relative value to low HC volume (to accurately predict rate of conversion within the recruited population) was in general only theoretically discussed. The data on which CAMD based their request for the biomarker to be accepted as qualified derive from a systematic review conducted after searching the literature for longitudinal studies evaluating MRI-assessed HC “atrophy”. However this was controversially used as a synonymous to low HC volume in predicting conversion to AD dementia from a baseline memory impaired state. The CAMD conclusions were mainly obtained via a “voting” procedure (“[…] the majority of studies report that […]”) but although it could be accepted that a true metaanalysis would, probably, had been unfeasible given the heterogeneity of the studies, further attempts to obtaining more rigorous and homogenous estimates was requested by the regulators. Also some discussion with CAMD representatives was needed, both to clarify some aspects of the systematic review and its internal and external validity and to explore whether a more in depth analysis of the retrieved data could justify a more precise statement than simply accepting the vague view that using low HC volume as a biomarker would “somewhat” enrich recruitment into clinical trials within the considered context. Other questions discussed during the qualification process were centred on how to standardise the criteria to define “low HC volume” for a particular method of image analysis, considering variability within-centres and within-subjects. The Consortium was advised to address and avoid potential biases on the definition of normal values of HC volume. To validate further the final cohort of patients to be included in clinical trials for pre-dementia stage AD companies should clarified the role of any other clinical condition and/or concomitant treatment or potential interactions, which may affect the HC volume. The regulators were concern that the patients identified by these biomarkers could define a different population from that in the previous qualification opinion (Isaac M. et al., 2001), implying that the MRI HC volume will bring additional heterogeneity in predementia studies. The EU regulators stated that subjects might be included in the studies based on clinical criteria and low HC volume biomarker. It's noteworthy to remind that EMA has very recently given a positive opinion in the predementia stage of Alzheimer's disease on CSF related biomarkers for drugs
Figure 4
Mixed methods.
affecting amyloid burden. The use of either one of the criteria discussed in these two qualification opinions (i.e. CSF-biomarkers and MRI HC volume) would generate, at first, two different and even most likely heterogeneous populations unless companies explore any relationship among them. The qualification procedure even with all its limitations and doubts must be shared in order to include more data in biomarkers and beginning to add numerical criteria to the drug development to decrease heterogeneity. The present qualification is only fit for purpose of regulatory clinical trials and we hope that it will facilitate information access to and from companies and research networks to develop new drugs for treating in this devastating disease. Several other qualifications opinions have been published at the EMA website in relation to biomarkers in Alzheimer's disease. Several qualifications advices in relation with people with Autism Spectrum Disorders, Down syndrome, Bipolar, Depression and Schizophrenia have been given already (Figure 3). In conclusion, the creation of a development road map for innovative medicinal products can be thought of as following scientific advice and/or qualification advice and opinion (Figure 1). Qualification opinion has a broader scope and may concern several indications or products fit for pharmaceutical medicine/academics/government, where a rapid update of translational pharmaceutical science for a global medicines development can be formulated (Figure 4). What we mean with the term, ‘qualification’, is the official regulatory opinion on the specific use of the proposed method (e.g. use of a biomarker) in a research and development (R&D) and/or clinical context after the review of the presented data. It can be a way of streamlining the marketing authorization procedure Finally, if the applicant agrees, the qualification is made public if a positive opinion is issued. The opinion will be incorporated in the future regulatory guidelines (EMA/ CHMP/617734/2013).
Role of the funding source Figure 3 Number of procedures.
We declare no funding support this paper.
CNS biomarkers: Potential from a regulatory perspective case study
Contributors We declare our Contribution.
Conflict of interest We declare our conflict of interest as employees of EMA & MEB.
Disclaimer The views expressed in this article are the personal views of the authors and may not be understood or quoted as being made on behalf of EMA or MBE.
Acknowledgements All these data used for the qualification has been based on the kindness of patients and careers who were voluntarily participating in these trials; they deserve our unreserved admiration. We recognise the work done by the SAWP, CHMP and from the CAMD Consortium. We are grateful that CAMD allowed the publication of their analysis requested by the EMA.
References Concept paper on need for revision of the guideline on medicinal products for the treatment of Alzheimer's disease and other dementias EMA/CHMP/617734/2013. Desikan, RS, Fischl, B, Cabral, HJ, et al., 2008. MRI measures of temporoparietal regions show differential rates of atrophy during prodromal AD. Neurology 71 (11), 819–825. Dubois, B, Feldman, HH, Jacova, C, et al., 2010. Revising the definition of Alzheimer's disease: a new lexicon. Lancet Neurol. 9, 1118–1127. EMA, Qualification opinion of low hippocampal volume (atrophy) by MRI for use in clinical trials for regulatory purpose – in predementia stage of Alzheimer's disease EMA/CHMP/SAWP/ 809208/2011. EMA guidance for companies requesting SA or PA. 〈http://www.emea. europa.eu/pdfs/human/sciadvice/426001en.pdf〉,〈http://www.emea. europa.eu/pdfs/human/sciadvice/426001en.pdf〉 and. EMA Qualification of Novel Methodologies. 〈http://www.emea.europa.eu/ pdfs/human/biomarkers/7289408en.pdf〉, 〈http://www.emea.europa.
1009
eu/pdfs/human/biomarkers/7289408en〉, 〈http://www.emea.europa. eu/pdfs/human/biomarkers/7〉. EMA, 2013. Qualification opinion of a novel data driven model of disease progression and trial evaluation in mild and moderate Alzheimer's disease 〈http://www.ema.europa.eu/docs/en_GB/ document_library/Regulatory_and_procedural_guideline/2013/ 10/WC500151309.pdf〉. EMA, 2012.Qualification opinion of Alzheimer's disease novel methodologies/biomarkers for the use of CSF AB 1-42 and t-tau and/ or PET-amyloid imaging (positive/ negative) as biomarkers for enrichment, for use in regulatory clinical trials in mild and moderate Alzheimer's disease. 〈http://www.ema.europa.eu/ docs/en_GB/document_library/Regulatory_and_procedural_gui deline/2012/04/WC500125019.pdf〉. EMA, 2011. Qualification opinion of Alzheimer's disease novel methodologies/biomarkers for the use of CSF AB 1-42 and ttau signature and/or PET-amyloid imaging (positive/ negative) as a biomarkers for enrichment, for use in regulatory clinical trials – in mild and moderate of Alzheimer's. 〈http://www.ema. europa.eu/docs/en_GB/document_library/Regulatory_and_pro cedural_guideline/2011/12/WC500118365.pdf〉. Fellgiebel, A, Dellani, PR, Greverus, D, Scheurich, A, Stoeter, P, Muller, MJ., 2006. Predicting conversion to dementia in mild cognitive impairment by volumetric and diffusivity measurements of the hippocampus. Psychiatry Res. 146 (3), 283–287. FDA, 2014. 〈http://www.fda.gov/downloads/Drugs/GuidanceCom plianceRegulatoryInformation/Guidances/UCM230597.pdf〉. Hill, D., Schwarz, A., Isaac M., Pani, L., Vamvakas, S., Hemmings R., Carrillo, M.,2014. CAMD/EMA Biomarker Qualification of Hippocampal Volume for Enrichment of Clinical Trials in Predementia Stages of Alzheimer's Disease. Alzheimer's & dementia: the journal of the Alzheimer's Association, 10 (4):421–429.e3, http://dx.doi.org/10.1016/j.jalz.2013.07.003. Higgins, J., & Green, S., 2011 March. Cochrane Handbook for Systematic Reviews of Interventions Version 5.1.0. Retrieved June 9, 2011, from The Cochrane Collaboration. 〈http://www.mrc-bsu.cam.ac. uk/cochrane/handbook/http://www.mrc-bsu.cam.ac.uk/ cochrane/handbook/〉. Isaac, M, Vamvakas, S, Abadie, E, Jonsson, B, Gispen, C, Pani, L., 2011. Qualification opinion of novel methodologies in the predementia stage of Alzheimer's disease: cerebro-spinal-fluid related biomarkers for drugs affecting amyloid burden–regulatory considerations by European Medicines Agency focusing in improving benefit/risk in regulatory trials Nov. Eur. Neuropsycho-pharmacol. 21 (11), 781–788 Epub 2011 Sep 7.
www.elsevier.com/locate/euroneuro
CNS biomarkers: Potential from a regulatory perspective Case study – Focus in low hippocampus volume as a biomarker measured by MRI Maria Isaaca,n,1, Christine Gispen-de Wiedb a
European Medicines Agency (EMA), United Kingdom Medicines Evaluation Board (MEB), The Netherlands
b
Received 6 October 2014; accepted 22 March 2015
KEYWORDS
Abstract
Biomarkers; Qualification advice; Qualification opinion; EMA; FDA
Our objectives are to describe the procedure for qualification advice and opinion from EU regulators on the use of novel methodologies in drug development, the key stakeholders involved and the evidence requirements for qualification opinion. We present a case study of the request from the Coalition Against Major Disease (CAMD) Consortium of the Critical Path (C-Path) Institute for EU regulators' qualification opinion on the use of low hippocampal volume as a biomarker for population enrichment in clinical trials of novel drugs in Alzheimer's disease (AD). We discuss the main concerns from the regulators, data analysis requests and guidance during the qualification. EU regulators concluded that low hippocampal volume, measured by vMRI and considered as a dichotomized variable (low volume or not), appears to help enriching recruitment into clinical trials aimed at studying drugs that potentially slow the progression of the pre-dementia stage of AD. The biomarker qualification procedure is a dynamic process in which pharmaceutical companies and research consortia can submit further data to update the qualifications and improve the predictive value of the biomarkers. & 2015 Elsevier B.V. and ECNP. All rights reserved.
n
Corresponding author. Tel.: +44 36607153; fax: +44 36607040. E-mail address: [email protected] (M. Isaac). 1 http://www.ema.europa.eu. http://dx.doi.org/10.1016/j.euroneuro.2015.03.013 0924-977X/& 2015 Elsevier B.V. and ECNP. All rights reserved.
1004
1.
M. Isaac, C. Gispen-de Wied
Introduction
The European Medicines Agency (EMA) coordinates the activity of the National Competent Authorities in the 28 Member States of the European Union in the framework of centralised evaluation procedures of medicinal products. Its work is underpinned by several Committees and Working Parties of the Agency, supported by the European Experts' Network including over 4900 nominated experts. For the purpose of the qualification procedure, the key agents involved are the Committee for Human Medicinal Products (CHMP), its standing Scientific Advice Working Party (SAWP), the CNS Working Party (CNSWP), the Biostatistics Working Party (BSWP), Modelling and Simulation Working Group (MSWG). The objective of the qualification of novel methodologies procedure can be a way of streamlining the marketing authorisation procedure (Figure 1). In operational terms scientific advice requests on the acceptability of the use can follow two parallel work streams, either through the ‘regular’ product- and indication-specific scientific advice (SA) or protocol assistance procedures with fixed timelines of 40 or 70 days (the latter including a discussion meeting between the SAWP and the applicant, e.g. in case of major disagreements or if clarifications are needed), and the ‘dedicated’ Qualification advice and opinion procedures, with a focus on a particular technology and eventually a broader scope which may span across several indications or products (EMA, EMA guidance for companies requesting SA or PA, EMA Qualification of Novel Methodologies). Although in general terms the main focus of the advice given by the SAWP rests on methodology, and pre-assessment of data intended to support a marketing authorisation application of a drug is formally not within the scope of the regular SA procedure, within the qualification procedure assessment of data has a key role. Due to the complexity of qualification procedures, requiring increased dialogue with the applicant, the timelines and proceedings are more flexible, often involving several ad hoc face-to-face meetings. A further important difference is that with the applicant's agreement, the outcome of the qualification opinion, and in the near future also of the qualification advice (letter of support) can be made public, in contrast to the outcome of the regular SA, which remains confidential at least until the product has received marketing authorisation for the concerned indication.
Figure 1
Dialogue with regulators timelines.
Letter of support: Based on the qualification advice a letter of support may be proposed by EMA as an option, when the novel methodology under evaluation cannot yet be qualified but is shown to be promising based on preliminary data. This letter is a positive regulatory feedback, including a high level summary of the novel methodology, context of use, available data and on-going/future investigations, meant to be made publicly available on the EMA website subject to sponsor's agreement. The objective of the letter of support is to encourage the efforts for data sharing and facilitate studies towards qualification for the novel methodology under evaluation. The scope of the biomarker qualification procedure in preclinical development involves pharmacological screening, understanding the mechanism of action, predictions of activity and safety, pharmacokinetic (PK) and pharmacodynamic (PD) modelling and toxicogenomics. Clinical development involves a verification of the mechanism of action, together with characterisation of the dose response effect, proof of concept in the therapeutic context, enriching the study population, considering surrogate endpoints and maintaining the paramount of safety by early detection of safety signals. Drug utilisation involves optimisation of the target population and guidance of the relevant treatment regimen. A CHMP qualification opinion on the acceptability of specific use of, say, a biomarker in a research and development context (whether clinical or non-clinical) is based on the assessment of data and is not specific to the individual product. The process involves a qualification team, peer review, public consultation and publication. CHMP qualification advice on future protocols and methods for further method development towards qualification is based on the evaluation of the scientific rationale and on preliminary data submitted. This process is confidential. The aim is that SAWP and CHMP are involved early in strategy design, with a commitment to evaluate data from agreed studies and to provide continuing opinion. The scope remains the focus on acceptability of specific use of the proposed technology or biomarker, developed for a specific intended use in the context of pharmaceutical research and development. The logistics of the qualification team include a coordinator, part of SAWP or CHMP, who facilitates a minimum of four experts from various disciplines to evaluate the therapeutic areas involved, the context of the intended use (for example, non-clinical safety testing or translational research), the relevant technology platform (proteomics, genomics, imaging) and statistics. A project manager is supplied by EMA. Again, the
Figure 2
Qualification procedure timelines.
CNS biomarkers: Potential from a regulatory perspective case study team members are selected according to expertise not Member State. EMA provides scientific, administrative and logistic support to the process and there is extensive networking within the agency. CHMP is involved before and during the procedure, underlining the commitment to evaluate the data obtained from the agreed studies. The timeline and the qualification team meetings can be adjusted case by case. In practice, such timelines can be as flexible as necessary. An example of the qualification procedure timeline is shown in Figure 2. Applicants can be private sector consortia, networks, public-private partnerships, learned societies, pharmaceutical industry and academia. Fee reductions can be granted for paediatric and orphan conditions, as well as to small and medium-sized enterprises (SME), and on a case-by-case basis the EMA Executive Director may grant a discretionary fee reduction based on grounds of public health benefit. EMA strongly recommends pre-submission meetings. These are important to facilitate communication then and later, and early help with amount and type of data to be submitted, questions to be answered and scope can be dealt with early. This can help the applicant to clarify its strategy and direct accordingly. Experience teaches that this usually improves outcomes. A pre-submission meeting can also help the applicant decide when and whether formally to start the procedure. An applicant can decide when the procedure starts and can also ask for the procedure to be suspended at any time (a ‘clock stop’). Again, at the applicant's request, meetings and teleconferences can be organised at any time in the procedure, as can additional data be presented. If a positive opinion is to be issued, the applicant is consulted on the amount and type of data to be made publicly available. There can be as many discussion meetings during the procedure as necessary, providing the applicant to justify its position and introduce new, significant data. A discussion meeting can also be used to influence the eventual outcome by presenting new relevant data or expert views of opinion leaders, justifying the applicant's strategy. The scientific administrators at EMA facilitate the process by providing scientific procedural and administrative support, informal teleconferences and meetings, help in organisation, expert meetings, out-of-the-procedure meetings and scientific discussion, providing a single contact point, able to radiate to other parts of EMA. There is also the possibility of parallel Food and Drugs Administration (FDA) and EMA qualification advice. This is voluntary at the sponsor's request, and involves questions on product development being put to both FDA and EMA, followed by discussions between the two Agencies and joint discussion with the sponsor. Each Agency, however, will issue separate responses to a sponsor's questions according to their individual procedures. The possible benefits of such an approach involve increased dialogue between Agencies and the sponsor from the early stages of development, together with the exchange of views and sharing of expertise and best practice. Moreover, parallel advice can optimise and facilitate global development, meeting the requirements of both Agencies (FDA, 2014). Optimising a request for qualification advice or opinion involves various practical considerations by the applicant, including the analytical platform, biological matrices and the timing for evaluation. Concerning biomarkers, the characterisation of the relationship between the biomarker response
1005
with normal biological processes or pathogenic processes, or pharmacological responses to a therapeutic intervention and the expected quantitative response threshold is necessary, together with the intended use of the biomarker in drug development and the clinic, especially deciding whether the use is to be exploratory or confirmatory, as well as its position in the preclinical, translational, toxicity, inclusion criteria for studies, stratification, enrichment, drug monitoring, primary and secondary endpoints, initiation of treatment or change of intervention nexus. Dossier in data can include, but not be limited to, protocols, study reports and raw data, all of which can establish the use of a defined methodology for a specific purpose in drug development. It is important that the range of opinion is represented so that the amount of data presented should not be over selective. There is a balance, however, between too little data and too much. The case needs to be presented in concise form in a separate document (the Briefing Document) with major points and data summaries in place. The main document must use references. Comprehensive data that are available at the time of submission should be presented, even if not controlled. As far as questions to be asked are concerned, here, as in life, the clearer the question, the clearer the answer. Questions that are too broad tend not to provide helpful answers. It is particularly important to ask about suspected pitfalls in the development, as well as choice of scope and various aspects of study design, such as outcomes, the study population and the relevant statistics. Qualification advice is peer reviewed and adopted by CHMP (or the relevant standing committee, if, say, advanced therapies or paediatrics are involved) and a competitive fee structure for the advice is also supplied, especially for orphan drugs. It is important to distinguish between CHMP qualification advice and CHMP qualification. Advice on future protocols and methods for further method development or its qualification is based on the evaluation of the scientific rational and on preliminary data. A qualification opinion on the acceptability of a specific use of the proposed method (such as a biomarker) and in research and development context is based on the assessment of submitted data. The immediate benefits of qualification advice include early identification of strengths and potential pitfalls and the opportunity to discuss them with leading experts in the field. This mostly involves a continuing dialogue, rather than a single interaction, so that issues can be revisited and agreement on what is required for ultimate regulatory acceptance of a new method can be discussed. In the long term, qualification advice speeds up the time to regulatory acceptance of novel approaches, as well as time to new marketing authorisation applications. This is an obvious potential saving in resources.
1.1.
Qualification advice
To date, qualification advice has been sought mainly for CNS products for conditions such as Alzheimer's disease, schizophrenia and Down's syndrome (17/29), with other qualification advice being sought from other disciplines, including nephrology, oncology, gastroenterology, chest medicine,
1006
M. Isaac, C. Gispen-de Wied
musculoskeletal medicine, cardiology and hepatology. The procedure flow is shown in Figure 1. To ensure flexibility, the product flow is adjusted on a case by case basis, with the timelines present for guidance, rather than as a prescription. The process can be suspended at any time at the applicant's request and meetings and teleconferences can be organised to discuss emergent issues. This individual tailoring is similar to that discussed above.
1.2.
Qualification opinion
A CHMP qualification opinion on the acceptability of a specific use of the proposed method, such as use of a biomarker, in a research and development context (non-clinical or clinical) is based on the assessment of submitted data. A qualification team, peer review and public consultation, followed by public publication, are involved and the context of use of the proposed method is clarified. As for qualification advice, private sector consortia, networks, public-private partnerships, learned societies, the pharmaceutical industry and academia can apply, with a vision of speeding up and optimising drug development and utilisation, thus improving public health. The opinion is not product specific and is based on the assessment of data. The initiation, consultation and qualification opinion itself are confidential until the qualification opinion recommendation is made public for market authorisation. The EMA procedure for qualification opinions involves discussion meetings and further assessment of data followed by a production of a document by SAWP with the working party's qualification opinion submitted to CHMP. CHMP will put the document out to peer review, after which it will be presented by an administrator or the chairman of SAWP during the CHMP plenary. This leads to the CHMP opinion, public consultation, feedback and a final opinion. As before, the process is adjusted case by case. This paper is a summary of the regulatory procedural thinking focus in a case. EMA published a qualification opinion on CSF biomarkers (see: Procedure EMEA/H/SAB/012/1/QA/2010; Isaac et al., 2011) and in October, EMA (2011) published a qualification opinion of low hippocampus volume (see Procedure EMA/ CHMP/SAWP/809208/2011; Hill et al., 2014; EMA, 2012, 2013).
2. Case study – qualification opinion of low hippocampal volume (atrophy) by MRI for use in clinical trials for regulatory purpose – in predementia stage of Alzheimer's disease (Figure 4) The data in support of this request for a qualification opinion on low HC (HC) volume were derived from an extensive literature search. The HC atrophy would be measured by MRI and proposed as a biomarker in predementia stage of Alzheimer's disease. CAMD with the help of consortium members conducted a systematic literature review (Higgins and Green, 2011) with the specific goal of identifying longitudinal studies that evaluated MRI-assessed HC atrophy as a biomarker in predicting conversion to AD dementia from a memory impaired state at baseline, either amnestic MCI or MCI determined by
criteria outlined in each specific manuscript. The literature search focused on data for those individuals who started out with cognitive impairment and progressed to AD dementia. Traditionally, cognitive measures have been used as determinants of likelihood of progression to dementia, usually AD dementia. As the Dubois approach suggests, the addition of supportive biomarker(s) could increase the accuracy of the prediction of progression from a prodromal AD state to AD dementia. The measurement tools used to determine the cognitive impairment varied among the studies reviewed. No attempt was made here to classify and weigh the specific methodologies used across the studies.
3.
Qualitative analysis
A systematic review was conducted of the published literature of the performance of HC volume, as measured by vMRI (volumetric magnetic resonance imaging), in predicting progression from a memory-impaired status to AD dementia. The literature search objectives were: to determine if baseline measurement of low HC volume (atrophy) by MRI in patients with episodic memory deficit (Dubois criteria) or MCI is a useful means of predicting those patients likely to evolve to AD. Specifically, CAMD incorporated here the identified literature that reports on the use of MRI-based volumetric measures of HC volume as a predictor of conversion from memory impairment status to AD type dementia.
4.
Quantitative analysis
Results: Volumetric MRI in Predicting Progression to AD. The focus of the accepted studies was the baseline HC atrophy in MCI subjects measured using vMRI with the investigative question regarding the ability of vMRI measurement of HC volume to predict progression to AD. Most studies were published in the last five years. The follow-up duration of all studies ranged from 18 months (Fellgiebel et al., 2006) to 108 months (Desikan et al., 2008). HC volume measurements were reported with various methods: as a total volume of both hippocampi, or as separate measurements for left and right hippocampi. The context of use specifies that the MRI based HC volume be measured at a single time point (baseline) and appropriate for subject inclusion into a clinical trial. Results: Hazard Ratios, Sensitivity, and Specificity of vMRI Biomarkers. The opinion to qualify was that low HC volume (atrophy) by MRI in patients with episodic memory deficit (Dubois criteria) was a useful mean of predicting those individuals likely to evolve to AD dementia during the course of an AD clinical trial. Specific data requested by EMA. Upon examination of all the data above EMA posed several questions to CAMD, the most relevant of them are briefly outlined below along with the answers provided by the Consortium. EMA requested that accuracy data should be represented in a classical ROC analysis and the information summarised from ADNI data (sROC model or average operating point along with criticism about its representativeness). CAMD agreed that the ROC analysis was a key evaluation parameter in the data prediction. However, the
CNS biomarkers: Potential from a regulatory perspective case study data included in the reported studies were insufficient to perform an AUC analysis on even a subset of the studies. Instead, the Consortium performed a de novo analysis on HC volume interpretation of MRI images generated from the ADNI study with the objective to determine whether HC volume quantification method significantly impacted on the prediction of conversion. The results of such review across the literature suggested that the HC volume quantification method (the one heterogeneous variable across the studies analysed here de novo) most likely did not impact on the interpretation of biomarker results. The CAMD was also asked to present the results of this de novo analysis based on the ADNI study with well-standardized MCI patient data, from four distinct HC volume analysis methodologies. In this case the variable being addressed was the volume reconstruction and its interpretation. The subject definition and image acquisition methods were standardized as described in the ADNI protocol. The de novo analysis presented the results in the requested ROC format. Additionally, the EMA requested the CAMD to provide the predictive value of other different biomarkers for pre-dementia stage of AD (Dubois et al., 2010). CAMD confirmed that several biomarkers are currently being used for the purpose of enrichment of clinical trials in pre-dementia stage of AD, including biochemical assessment of CSF Aβ42, tau and p-tau levels (see Isaac et al. (2011)), PET imaging of amyloid, metabolic activity (FDGPET), cerebral perfusion (SPECT) and structural neuroimaging. Each biomarker has advantages and disadvantages in terms of ease of use, cost and invasiveness to patients. Relative to the other biomarkers, vMRI was less invasive and had wider applicability, although was limited by the cost, availability and interoperability of the Neuroimaging machines and their software elaborated data across different centres and in different time. It seems that at present, the field was not yet at a stage to employ combinatorial imaging biomarker measurements beyond research purposes as highlighted in a recent webinar focused on early AD biomarkers (see http://www.alzforum.org/res/for/journal/ detail.asp?liveID=192). The EMA requested further information to compare the performance of CSF and HC volume in predicting conversion to AD. CAMD performed an additional analysis of the ADNI dataset and reported the results of progression at the two year time point. Comparison of the AUC values of baseline vMRI of hippocampus and CSF analytes (Aβ; Phospho-τ; Total τ; single and combined) indicated that vMRI in this dataset is as good as, or slightly better at, predicting conversion to AD than CSF analytes; specifically, a greater ROC AUC was observed for MRI than CSF biomarkers in this comparison. The analysis was a comparison of AUC values of baseline vMRI data predicting conversion to AD to AUC values of CSF analytes predicting conversion to AD in 2 years. The EMA was also concerned on the quantitative measures of the hippocampus in comparison to other brain regions in AD brain. Clinical Utility: The conclusions on the final qualification opinion was as follows: Low hippocampal volume, as measured by MRI and considered as a dichotomized variable (low volume or not), appears to help enriching recruitment into clinical trials aimed
1007
at studying drugs potentially slowing the progress/conversion to AD dementia of the included subjects. Low hippocampal volume might be considered a marker of progression to dementia in subjects with cognitive deficit compatible with predementia stage of AD (Dubois, 2010). However, neither the actual value of low hippocampal volume to accurately predict rate of such progression to dementia in the referred subjects nor the relative value of other biomarkers have been reported. “As currently planned in the current opinion subjects might be included in the studies based on clinical criteria and low hippocampal volume biomarker (if positive). The CHMP has given a previous positive opinion in the predementia stage of Alzheimer's disease: cerebro–spinal fluid related biomarkers for drugs affecting amyloid burden. This may lead first to a heterogeneous population and, moreover, it will not be possible to explore the relationship among them. Although not required form a regulatory perspective, the concomitant assessment of the two biomarkers in predementia stage of AD would be of great value.” “The process of measurement of low hippocampal volume is also complex. To obtain reliable results implies the standardisation of all steps (at least imaging acquisition protocol, imaging reconstruction/analysis methods, timing to conversion, etc). International guidelines have been produced. These guidelines must be enforced.”
5.
Discussion
Accepting the value of these biomarkers in “enriching” recruitment is, probably, less demanding than assessing its value in other potential uses such as diagnostic and thus, less accuracy in the prediction is required than to include a particular individual into a diagnostic category. The qualification of biomarkers to “enrich” recruitment into clinical trials aimed at studying drugs potentially slowing the progress/conversion to (AD) dementia of the included subjects has been used in drug development mainly in phase I and 2. The biomarkers qualification procedure is targeting recruitment in phase 3 studies. EU regulators felt that identifying subjects at higher risk of developing AD dementia (as intended in this procedure) maybe serve useful purposes even in the absence of effective treatments for the disease. Enroling “non-enriched” samples (basing inclusion only on the cognitive deficit) could mean that fewer subjects would convert during the duration of the trial. Impractically large numbers of subjects and/or duration of follow-up would be required and the trials would be unfeasible or inefficient. Other biomarkers to “enrich” recruitment into this type of clinical trials are already known (e.g. some CSF analytes). The purpose of this “qualification” procedure was to assess whether low HC volume as measured by MRI and considered as a dichotomized variable (low volume or not) could be considered as an enrich population for predementia stage of AD clinical trials. The potential value of the proposed marker in other settings as a criterion for the diagnosis of a disease such as AD in a particular subject or the usefulness of repeated measurements to assess the effect of therapeutic interventions – as a marker of efficacy – were not considered.
1008
M. Isaac, C. Gispen-de Wied
The relative value to low HC volume (to accurately predict rate of conversion within the recruited population) was in general only theoretically discussed. The data on which CAMD based their request for the biomarker to be accepted as qualified derive from a systematic review conducted after searching the literature for longitudinal studies evaluating MRI-assessed HC “atrophy”. However this was controversially used as a synonymous to low HC volume in predicting conversion to AD dementia from a baseline memory impaired state. The CAMD conclusions were mainly obtained via a “voting” procedure (“[…] the majority of studies report that […]”) but although it could be accepted that a true metaanalysis would, probably, had been unfeasible given the heterogeneity of the studies, further attempts to obtaining more rigorous and homogenous estimates was requested by the regulators. Also some discussion with CAMD representatives was needed, both to clarify some aspects of the systematic review and its internal and external validity and to explore whether a more in depth analysis of the retrieved data could justify a more precise statement than simply accepting the vague view that using low HC volume as a biomarker would “somewhat” enrich recruitment into clinical trials within the considered context. Other questions discussed during the qualification process were centred on how to standardise the criteria to define “low HC volume” for a particular method of image analysis, considering variability within-centres and within-subjects. The Consortium was advised to address and avoid potential biases on the definition of normal values of HC volume. To validate further the final cohort of patients to be included in clinical trials for pre-dementia stage AD companies should clarified the role of any other clinical condition and/or concomitant treatment or potential interactions, which may affect the HC volume. The regulators were concern that the patients identified by these biomarkers could define a different population from that in the previous qualification opinion (Isaac M. et al., 2001), implying that the MRI HC volume will bring additional heterogeneity in predementia studies. The EU regulators stated that subjects might be included in the studies based on clinical criteria and low HC volume biomarker. It's noteworthy to remind that EMA has very recently given a positive opinion in the predementia stage of Alzheimer's disease on CSF related biomarkers for drugs
Figure 4
Mixed methods.
affecting amyloid burden. The use of either one of the criteria discussed in these two qualification opinions (i.e. CSF-biomarkers and MRI HC volume) would generate, at first, two different and even most likely heterogeneous populations unless companies explore any relationship among them. The qualification procedure even with all its limitations and doubts must be shared in order to include more data in biomarkers and beginning to add numerical criteria to the drug development to decrease heterogeneity. The present qualification is only fit for purpose of regulatory clinical trials and we hope that it will facilitate information access to and from companies and research networks to develop new drugs for treating in this devastating disease. Several other qualifications opinions have been published at the EMA website in relation to biomarkers in Alzheimer's disease. Several qualifications advices in relation with people with Autism Spectrum Disorders, Down syndrome, Bipolar, Depression and Schizophrenia have been given already (Figure 3). In conclusion, the creation of a development road map for innovative medicinal products can be thought of as following scientific advice and/or qualification advice and opinion (Figure 1). Qualification opinion has a broader scope and may concern several indications or products fit for pharmaceutical medicine/academics/government, where a rapid update of translational pharmaceutical science for a global medicines development can be formulated (Figure 4). What we mean with the term, ‘qualification’, is the official regulatory opinion on the specific use of the proposed method (e.g. use of a biomarker) in a research and development (R&D) and/or clinical context after the review of the presented data. It can be a way of streamlining the marketing authorization procedure Finally, if the applicant agrees, the qualification is made public if a positive opinion is issued. The opinion will be incorporated in the future regulatory guidelines (EMA/ CHMP/617734/2013).
Role of the funding source Figure 3 Number of procedures.
We declare no funding support this paper.
CNS biomarkers: Potential from a regulatory perspective case study
Contributors We declare our Contribution.
Conflict of interest We declare our conflict of interest as employees of EMA & MEB.
Disclaimer The views expressed in this article are the personal views of the authors and may not be understood or quoted as being made on behalf of EMA or MBE.
Acknowledgements All these data used for the qualification has been based on the kindness of patients and careers who were voluntarily participating in these trials; they deserve our unreserved admiration. We recognise the work done by the SAWP, CHMP and from the CAMD Consortium. We are grateful that CAMD allowed the publication of their analysis requested by the EMA.
References Concept paper on need for revision of the guideline on medicinal products for the treatment of Alzheimer's disease and other dementias EMA/CHMP/617734/2013. Desikan, RS, Fischl, B, Cabral, HJ, et al., 2008. MRI measures of temporoparietal regions show differential rates of atrophy during prodromal AD. Neurology 71 (11), 819–825. Dubois, B, Feldman, HH, Jacova, C, et al., 2010. Revising the definition of Alzheimer's disease: a new lexicon. Lancet Neurol. 9, 1118–1127. EMA, Qualification opinion of low hippocampal volume (atrophy) by MRI for use in clinical trials for regulatory purpose – in predementia stage of Alzheimer's disease EMA/CHMP/SAWP/ 809208/2011. EMA guidance for companies requesting SA or PA. 〈http://www.emea. europa.eu/pdfs/human/sciadvice/426001en.pdf〉,〈http://www.emea. europa.eu/pdfs/human/sciadvice/426001en.pdf〉 and. EMA Qualification of Novel Methodologies. 〈http://www.emea.europa.eu/ pdfs/human/biomarkers/7289408en.pdf〉, 〈http://www.emea.europa.
1009
eu/pdfs/human/biomarkers/7289408en〉, 〈http://www.emea.europa. eu/pdfs/human/biomarkers/7〉. EMA, 2013. Qualification opinion of a novel data driven model of disease progression and trial evaluation in mild and moderate Alzheimer's disease 〈http://www.ema.europa.eu/docs/en_GB/ document_library/Regulatory_and_procedural_guideline/2013/ 10/WC500151309.pdf〉. EMA, 2012.Qualification opinion of Alzheimer's disease novel methodologies/biomarkers for the use of CSF AB 1-42 and t-tau and/ or PET-amyloid imaging (positive/ negative) as biomarkers for enrichment, for use in regulatory clinical trials in mild and moderate Alzheimer's disease. 〈http://www.ema.europa.eu/ docs/en_GB/document_library/Regulatory_and_procedural_gui deline/2012/04/WC500125019.pdf〉. EMA, 2011. Qualification opinion of Alzheimer's disease novel methodologies/biomarkers for the use of CSF AB 1-42 and ttau signature and/or PET-amyloid imaging (positive/ negative) as a biomarkers for enrichment, for use in regulatory clinical trials – in mild and moderate of Alzheimer's. 〈http://www.ema. europa.eu/docs/en_GB/document_library/Regulatory_and_pro cedural_guideline/2011/12/WC500118365.pdf〉. Fellgiebel, A, Dellani, PR, Greverus, D, Scheurich, A, Stoeter, P, Muller, MJ., 2006. Predicting conversion to dementia in mild cognitive impairment by volumetric and diffusivity measurements of the hippocampus. Psychiatry Res. 146 (3), 283–287. FDA, 2014. 〈http://www.fda.gov/downloads/Drugs/GuidanceCom plianceRegulatoryInformation/Guidances/UCM230597.pdf〉. Hill, D., Schwarz, A., Isaac M., Pani, L., Vamvakas, S., Hemmings R., Carrillo, M.,2014. CAMD/EMA Biomarker Qualification of Hippocampal Volume for Enrichment of Clinical Trials in Predementia Stages of Alzheimer's Disease. Alzheimer's & dementia: the journal of the Alzheimer's Association, 10 (4):421–429.e3, http://dx.doi.org/10.1016/j.jalz.2013.07.003. Higgins, J., & Green, S., 2011 March. Cochrane Handbook for Systematic Reviews of Interventions Version 5.1.0. Retrieved June 9, 2011, from The Cochrane Collaboration. 〈http://www.mrc-bsu.cam.ac. uk/cochrane/handbook/http://www.mrc-bsu.cam.ac.uk/ cochrane/handbook/〉. Isaac, M, Vamvakas, S, Abadie, E, Jonsson, B, Gispen, C, Pani, L., 2011. Qualification opinion of novel methodologies in the predementia stage of Alzheimer's disease: cerebro-spinal-fluid related biomarkers for drugs affecting amyloid burden–regulatory considerations by European Medicines Agency focusing in improving benefit/risk in regulatory trials Nov. Eur. Neuropsycho-pharmacol. 21 (11), 781–788 Epub 2011 Sep 7.