Angela Schulz, MD
University Medical Center Hamburg-Eppendorf (UKE)
WP03 Epidemiology and Natural History Study
- To collect, as thoroughly as possible, patients with different NCL forms in all participating countries including India in order to measure the incidence and prevalence of each type of NCL
- To establish an NCL patient registry of long-lasting function to enable both retrospective and prospective patient data collection and to precisely describe the clinical course and its variability in the different forms of NCL
- To correlate the genotype of patients with their phenotype by linking clinical and genetic mutation data
- To establish a tool for evaluation of future experimental therapy studies, as well as current palliative therapies
Collection of patients with different NCL forms in all participating countries to measure incidence and prevalence of each type of NCL
The first aim of the epidemiology task was to establish prevalence figures in the participating countries (Germany, Italy, Finland, UK and India). As the precision of these figures clearly depends on i) the awareness of NCL diseases, ii) existing medical infrastructure, and iii) diagnostic capabilities in the respective areas, the DEM-CHILD consortium has worked towards optimising diagnostic capabilities in the participating countries by taking several steps:
i) Increasing awareness of NCL diseases
Multiple actions have been undertaken in order to increase the awareness of NCL diseases across different medical specialties such as paediatricians, child neurologists, neurologists, ophthalmologists, and geneticists. Numerous members of the DEM-CHILD consortium have given talks about the DEM-CHILD project and NCL diseases both at national and international scientific meetings addressing physicians from different medical specialties (see list of dissemination activities). In addition, six DEM-CHILD teaching workshops have been organised in several countries including India. In addition, several review articles on the clinical presentation of NCL diseases have been published by DEM-CHILD partners during the reporting period. Moreover, the project website www.dem-child.eu provides up-to-date information and useful links for clinicians who would like to learn more about these diseases.
ii) Development and distribution of a diagnostic algorithm
A diagnostic algorithm for NCL diseases has been developed and published in collaboration with all NCL clinical experts in the consortium as part of a review on clinical presentation of NCL diseases (Schulz A. et al, BBA 2013). A flyer illustrating both, this diagnostic algorithm for NCL diseases as well as providing information where to send diagnostic samples, has been designed and distributed not only during the respective DEM-CHILD teaching seminaries but also at various international congresses (see list of dissemination activities). The flyer can also be downloaded for free from the DEM-CHILD website.
iii) Optimization of NCL-diagnostic workflow
Several steps have been taken in order to optimise the diagnostic workflow for NCL diseases:
a) 24h expert advice: Experts from the NCL specialty clinic of co-ordinator partner 1 can be reached 24 h/ 7 days by phone (+49-40-74105-6391) and discuss patient cases with suspected NCL in order to be recommended on diagnostic procedures to be performed.
b) Centralised sample database: Once NCL disease is clearly suspected in a patient, physicians are advised to refer their patient’s samples (DBS, DNA; fibroblasts etc.) for diagnosis to the project. Information regarding where and under which conditions diagnostic samples should be sent can be found on the project’s website. The diagnostic workflow developed by the DEM-CHILD consortium recommends that all diagnostic samples should first be sent to partner 01, where each sample receives a unique identifying code and the respective patient’s information such as age at disease onset, consanguinity etc. is entered into a sample database under this code. The project-specific patient sample database was developed by partner 01 in order to code and track samples referred to the DEM-CHILD project. The database is online accessible and password protected. Geneticists and physicians who have referred samples to the project will receive their samples’ unique codes as well as a username and password to login into the sample database and to track their patient’s diagnostic results.
The sample database currently holds information on 363 samples from patients with clinically clearly suspected NCL disease (DNA n=278; DBS n=60; skin biopsies / fibroblasts n=25).
iv) Collaboration with NCL family associations
Throughout the entire reporting period, the DEM-CHILD consortium has closely collaborated with the national NCL family associations in the participating countries: Clinical experts have given talks at the annual meeting of the Italian, German and UK family associations meetings. The above mentioned 24 h expert advice phone number is offered not only to referring physicians but also to families affected by NCL disease. As part of this expert advice, newly diagnosed patient have been provided with contact information of the respective national family association. Families without a clear genetic diagnosis have been referred to the consortium by the family associations. Moreover, numbers of total patients with NCL disease have been regularly exchanged between the consortium and the family associations in order to establish prevalence figures.
By implementing the step described above throughout the entire project, prevalence figures for CLN1, CLN2, CLN3, CLN5, CLN6, CLN7, CLN8, CLN10 have been successfully obtained for Germany, UK, Italy and Finland.
However, one major challenge was to obtain reliable prevalence figures for India:
Due to the late joining of our Indian partner in month 13 of the project and due to a restriction that only 10% of patient material was allowed to leave India and to be entered into the project, the number of samples (n=63) received during the entire project is too small to be used for any prevalence studies. The 10% restriction rule had been enforced by the Indian Council of Medical Research (ICMR) after the project started and was a pre-condition that partner 05 got permission from the ICMR to sign the DEM-CHILD grant agreement and be part of the consortium. The new restriction rule for India had been included in Amendment No. 1 of the DoW and accepted by the European Commission.
As already pointed out under WP02, as per amendment No.2 of the DoW the work programme to obtain incidence figures for specific NCL forms (CLN1, CLN2, CLN10) has been changed from screening of anonymously collected dry blood spot samples derived from newborn screening programs to targeted population screening in selected areas.
Due to the mentioned challenges developing MS-based enzyme assays for CLN1, CLN2, and CLN10 and the necessity to include a new partner into the consortium, the new MS-based triplex enzyme assay has only been available since end of month 35 of the project. Therefore statistical numbers from targeted population screening are still under way.
Despite these problems, the project has well prepared the basis for targeted population screening by clearly defining the target patient groups: These groups comprise patients who present with “risk factors” for CLN1, CLN2 or CLN10 disease, respectively. These “risk factors” are early, sometimes pre-clinical symptoms which have been shown to occur with high frequency in the respective NCL disease. Risk factors have been identified by extensive analysis of natural history data in the patient database.
In conclusion, studies on epidemiology of NCL diseases in the participating countries have been successful especially with regard to the establishment of up-to-date prevalence figures. As expected, the generation of incidence figures has been a challenge but solutions have been generated.
Establishment of an NCL patient registry for both retrospective and prospective patient data collection
Development of an NCL patient registry has been performed by the DEM-CHILD Patient Database Consortium. To summarise, the DEM-CHILD NCL patient database has the following main features:
- Online accessible (in compliance with international data safety rules)
- Patient management software (to avoid inadvertent double or multiple entries of the same patient)
- Statistical analysis features (to allow immediate comparison of result of one patient to other patients)
- Database charter (ensuring rules for transparency, data ownership and publication)
The type of data in the database is differentiated into “static” and “dynamic” data.
The intended patient recruitment rates were 200 NCL patients in total for this project. All patients had to have a clear genetically verified diagnosis of NCL disease. Due to the low number of diagnosed patients in India, data collection was restricted for this country. Still the intended recruitment rates have been well achieved both for retrospective data analysis (n=220; Table 2) and for prospective data analysis (n= 202; Table 3). Moreover, the database consortium has expanded beyond the DEM-CHILD partners with currently 12 countries participating allowing the database to be the world largest collection of NCL patient data to date (for further details please refer to section 1.3).
Type of data collected
Data collected in the DEM-CHILD database are separated into “static” and “dynamic” data. Static data are e.g. genetic diagnosis, age at first symptoms etc. They have in common that they do not change over time. Dynamic data are related to the patient’s age and change with the progression of the disease. Examples for dynamic data are clinical scoring data.
Static data do not change over time. They are mainly derived from retrospective data collections based on information from patient charts and detailed interviews with parents and caregivers. Static data collected in the database comprise the following categories: Gender, family history, perinatal history, psychomotor development, medical history, diagnostic summary, neurologic findings, experimental therapy studies. Information collected under the category diagnostic summary do not only comprise the genetic diagnosis, results of enzyme activity assays or electron microscopy but also whether and where (in which participating center) patient material is available such as DNA, fibroblast cell lines etc. This information creates a kind of “virtual biobank” which allows scientists to search for patient material with specific mutations for certain research project. Once a patient of whom material is available is identified, the scientist can contact the participating center and ask for permission to use the material for the respective research project. During the DEM-CHILD project, this “virtual biobank” has been well in use e.g. for the identification of specific CLN1 patient fibroblast cell lines for WP05. Table 2 provides an overview on numbers of patients with static data sets for each NCL form segregated by different ages of disease onset (infantile, late infantile, juvenile, adult) and shows that intended recruitment rate has been well achieved.
The static data sets represent the background for the analysis of the dynamic data. They allow the interpretation of different disease progression rate based on the genetic diagnosis and thereby lead to a better understanding of the genotype-phenotype correlation.
In addition, detailed analysis of the retrospective static data has led to the identification of early symptoms supporting early diagnosis.
Table 2: Number of patients with static data (from retrospective data analysis)
ii) Dynamic data
Dynamic data are related to the patient’s age at the time of examination. They will change as the disease progresses. Dynamic data collected in the database comprise the following: Clinical status, neurologic status, current medication, ophthalmologic exam, cardiologic exam, EEG, brain imaging (MRI/MRS). In addition, clinical scoring systems have been regularly used for each patient such as the late infantile NCL Scoring (Steinfeld et al. 2002), juvenile NCL scoring (Kohlschütter et al. 1988), juvenile UBDRS scoring (Marshall et al. 2005), GMFCS (Gross Motor Function Classification System), BFMF (Bimanual Fine Motor Function), and QoL (Quality of Life) Questionnaires. The intended schedule for regular follow-up examination had been every 6 months for patients with infantile and late infantile phenotype and every 6-12 months of patients with a juvenile phenotype. Table 3 provides an overview of the number of patients and exams collected per NCL form showing that the intended recruitment rate has been well achieved.
Table 3: Number of patients with dynamic data (from prospective data analysis)
Analysis of genotype-phenotype variability in each NCL form
The analysis of genotype-phenotype variability in NCLs is challenging due to the heterogeneity of these diseases with at least 13 different affected genes and multiple mutations identified in each gene.
Extensive analysis on genotype-phenotype variability has been performed based the data collected in the database. Prospective longitudinal data on clinical scoring data, MRI-based brain volumetric analysis etc. have been evaluated based on static retrospective data such as genetic diagnosis etc. The following are examples for such analyses:
Clinical Scoring in CLN2 disease
For the analysis of natural history of CLN2 disease, the late infantile NCL scoring system established by Steinfeld et al. 2002, has been used in a total of 62 patients with genetically diagnosed CLN2 disease. Scoring data have been collected longitudinally for all patients from birth to the age of 10-12 years. Disease progression was measured longitudinally by the sums of the 3-point motor and language subscales of the Steinfeld et al. score. Disease progression was very similar in the majority of patients. Slowly progressing patients were uncommon and mostly related to unusual genotypes.
(Schulz A et al. The Natural History of Late Infantile CLN2 Disease: Striking Homogeneity of Clinical Progression in Two Independently Obtained Large Clinical Cohorts. Manuscript in preparation).
Brain volumetric analysis in CLN2 disease
In addition to clinical scoring, we have used longitudinal MRI-based brain volumetric studies to study disease progression and genotype-phenotype correlations in CLN2 disease. In this study, twenty-one MRIs of eight patients (3 male; 5 female; mean age, 6.9 ± 2.5 years) with genetically confirmed CLN2 disease were performed. Results showed that longitudinal MRI volumetry of gray matter reveals a unique disease progression. Morever, volumetric data were more sensitive in describing disease progression in later stages.
(Loebel U. et al. Longitudinal MRI Volumetry of Gray Matter in Neuronal Ceroid Lipofuscinosis Subtype CLN2 Reveals a Unique Uniformity of Disease Progression. Manuscript submitted).
Genotype-phenotype correlation in CLN3 disease
Similar to CLN2 disease, many different mutations are described for CLN3 disease (n=65). However, literature shows that the majority of patients (up to 80%) harbor a common 1 kb deletion (c.462-677del) which leads to the loss of exons 7 and 8. A total of 58 patients (89%) were homozygous for this mutation in the CLN3 patient cohort analysed from the DEM-CHILD database.
When using a clinical scoring system developed by Kohlschütter et al. for juvenile NCL in all CLN3 patients homozygous for the 1 kb deletion in our cohort, we could detect a rather high phenotype-genotype variability which supports the data previously published by Lebrun et al. When calculating the “Score of Relative Disease Severity” (method described by Lebrun et al. 2011), these patients segregated into three groups with slow (24%), average 36%) and rapid disease progression (40%) despite the fact that they harbor the same mutation in the CLN3 gene.
Clinical scoring in CLN5 disease
CLN5 disease is very heterogenic with currently at least 45 mutations identified in a total of 85 patients so far according to the DEM-CHILD mutation database. The phenotype observed in CLN5 patients predominantly from Italy was a mixture of late infantile / early juvenile. Therefore the current clinical scoring systems were not optimal in order to describe the disease in these patients,
Because of the domains which were observed to be affected at the disease onset, an ad hoc evaluation scale was developed by Simonati et al. which is modified from the late infantile Steinfeld et al. scoring. A total of 18 patients with CLN5 disease has been analysed with this new clinical scoring system.
(Simonati A. et al. Analysis of genotype-phenotype variability in Italian CLN5 patients. Manuscript in preparation)
Evaluation tool for (future) experimental studies
To date the DEM-CHILD NCL patient database contains the largest set of natural history data for all NCL forms worldwide which will continue to grow with the participation of the new database consortium members from a total of 12 countries. Results from the analysis longitudinal natural history data describe well that data derived from the DEM-CHILD NCL patient database represent a valid tool for the evaluation of current and future experimental therapy studies. In fact, longitudinal natural history data of 62 patients with CLN2 disease have been combined with a set of cross-sectional data from a total of 43 CLN2 patients seen at the Weill-Cornell-Medical-College by Dr. Ron Crystal. The level of loss of motor and language function over time was similar in both cohorts. In fact, these combined CLN2 natural history data are already used as control data in a phase I/II clinical trial on intraventricular enzyme replacement therapy in CLN2 disease.
Evaluation tool for current palliative therapies
As to date there is still no cure for any form of NCL, palliative treatment and supportive care are an important part of current treatment in all NCL forms.
Therefore information on current medication to treat the most severe symptoms in all NCL forms such as epilepsy, myoclonus and spasticity has been collected in the DEM-CHILD database including the respective positive and negative effects. First results have been published already in a review on the clinical presentation of NCL diseases (Schulz A. et al, BBA 2013) More detailed results are summarised in a review about this study in the Journal of Pediatric Epilepsy which is in press (Kohlschütter et al., J Pediatr. Epilepsy 2014). These data will help to improve palliative medical care for NCL patients.
In WP3, prevalence figures for the different NCL forms have been obtained in the participating countries. For incidence studies, a new approach has been implemented by defining criteria such as early symptoms for targeted population screening. An online NCL patient database has been established containing both retrospective and prospective data from over 200 NCL patient to date. Extensive analysis on genotype-phenotype variability has been performed based on longitudinal collection of natural history data. These datasets represent invaluable tools for the evaluation of experimental therapy studies and are already used as control data in clinical trials. Likewise, current palliative medical treatment options have been evaluated based on these data in order to improve palliative care in NCL patients.