Applying Genetics and Public Health Strategies to Primary Immunodeficiency Diseases
November 8-9, 2001 ~ Atlanta, Georgia
Prepared by: Office of Genomics and Disease Prevention, Centers for Disease Control and Prevention Department of Health and Human Services
The participants met in breakout groups to devise short- and long-term steps for the development of a public health approach to PI diseases. Representatives from each breakout session presented their reports to the full group. The deliberations and conclusions of the breakout groups are summarized below.
The group determined that the main goals of surveillance for PI diseases are to define the prevalence of the disorders, reduce morbidity and mortality, and link patients to care and interventions (and perhaps research). Several PI diseases are appropriate for prioritization for public health assessment: profound T-cell defects because of their high mortality in the absence of intervention; antibody deficiencies because of the large numbers of persons affected and the high burden of morbidity; and CGD because of the existence of an established IDF data set.
The group identified a need for collaborative studies to define the natural history of PI diseases, research on hyper-IgM syndrome, and assistance to IDF for analysis of existing data on patients who are already in care. Resources needed to support this research include model centers, perhaps using the NIH network, and funding (including funds for specimen collection, chart abstraction, and form development). Potential partners are all persons who provide care to or have an interest in patients with PI diseases. These include professional organizations (e.g., American Academy of Allergy, Asthma, and Immunology (AAAI), Clinical Immunology Society (CIS), American College of Allergy and Immunology (ACAI), AAP, American Medical Association (AMA), National Association of Pediatric Nurse Associates and Practitioners (NAPNAP), American College of Obstetrics and Gynecology (ACOG), IDF, state and local health departments, payers, advocacy groups, and data sources (e.g., hospital-based registries, hospital discharge data sets, bone marrow transplantation registries, VSD).
To determine factors that affect the effectiveness of known treatment modalities, the group advocated research on impediments to accessing treatment and case management. Required resources are special studies, clinical trials, and case managers, supported by partnerships with NIH, HRSA/Medicaid/HMOs/payers, advocacy groups, state/local health departments, protective services, and professional organizations (AAAI, ACAI, CIS, AAP, AMA).
The group agreed on three immediate needs: (1) establish an advisory group charged with guidance on PI surveillance and case definitions, (2) assist IDF in the analysis of data on patients in care, and (3) encourage collaborations modeled after efforts related to hyper-IgM syndrome.
Suggested surveillance activities for the short-term include piloting surveillance systems in a limited number of large states (e.g., California, New York, Texas) for a limited number of PI diseases (as defined by the advisory group). Possibilities for surveillance are SCID, antibody deficiencies, and CGD. Surveillance should be linked to existing databases (e.g., VSD, Medicaid, hospital discharge data, specialized clinics, IDF registry, laboratory-based reporting). Longer-term activities are to expand surveillance beyond the pilot states, if feasible, and to conduct special studies through model PI centers. These might include longitudinal spectrum-of-disease studies, studies of biospecimen collection, collaborations with NIAID-funded network sites to collect and catalog genetic tests, clinical trials, and evaluation of genotype and phenotype relations.
There was limited agreement among the group on a tentative list of PI diseases to target for early recognition. Targeting five diseases--SCID, XLA, CVID, CGD, WAS--would capture 50 to 60 percent of patients and also pick up other disorders.
The next step is to select the target audiences and adjust the early-recognition tool for each audience. Important target audiences include primary care practitioners (pediatricians, internists, family practitioners, nurse practitioners, physician assistants), specialists (ENT, allergy, immunology, gastroenterology, pulmonology, infectious disease), patients and families, laboratories, and policymakers. The group reached no agreement on the best tool for each audience but recommended that working groups be created to construct a system of tools. These might include lists of warning signs, a scoring system, symptoms questionnaires, or a bulletin of alerts.
The group agreed on the importance of testing these tools to confirm that they can identify children with PI diseases but do not over identify and overburden the system. The discussion yielded no concrete methods for testing the tools, but one suggestion was to apply potential tools retrospectively to existing data on child health (e.g., vaccine registries, managed-care databases). Partnerships with CDC, managed-care organizations, foundations, and registries will probably be required to test any newly developed tools. Participants emphasized that research and action can occur simultaneously. It is important to continue to use what is currently available (e.g., 10 warning signs) until evaluation can be completed. Once data are available, the next step is to report on the effectiveness of the tools to the original target audiences and amend the tools as indicated. The long-term effect of this process is dissemination of the tools and behavior change among the target audiences.
The group agreed that the most promising candidate for potential development of a newborn screening protocol is SCID, defined phenotypically as profound T-cell deficiency, regardless of genotype or syndrome. This was chosen for two main reasons: (1) extremely high mortality if not detected early, and (2) availability of treatment that can substantially reduce mortality with early identification. The prevalence is not well established and can be reliably determined only through pilot testing of screening on a population basis.
Evaluation of T-cell numbers was the proposed screening test. The group also agreed that T-cell enumeration can be reliably performed using cord blood samples with manual differentials. The drawbacks are logistical and (secondarily) financial. Dried blood spots collected through centralized state-based newborn screening programs have none of these drawbacks. The priority should therefore be the development of T-cell assays using dried blood spots. These assays can be validated by blinded comparison of results from tests on newborn dried blood spots and cord blood samples. The three potential types of dried blood spot T-cell assays are enzyme linked immuno-sorbent assay(ELISA), PCR, and tandem mass spectrometry. The introduction of tandem mass spectrometry into newborn screening programs substantially increases the number of disorders that can be detected from newborn dried blood spot specimens.
Once an assay has been developed and validated on the basis of reasonable numbers of blinded pairs of samples, the next step is to develop a test suitable for piloting in one or more newborn screening programs. HRSA would probably provide funds for pilot testing in one or two state newborn screening programs.
In summary, the group concluded that the following four steps should be followed to determine whether newborn screening for SCID is a realistic option:
Investigators should be encouraged to develop assays to count T cells in dried blood spot samples from newborns. Three potential types of assays have been identified.
Once assays have been developed, they need to be validated. The most logical validation would consist of blinded comparisons of T-cell counts calculated using the proposed assays from dried blood spots, with manual differential count from cord blood samples as the gold standard. The question of IRB approval needs to be investigated.
After one or more of the assays have been developed and validated in terms of analytic validity, they will be ready for pilot testing to determine clinical validity. Pilot testing will be performed by integrating the proposed assay into an existing newborn screening panel on an investigational basis with IRB approval (including documentation of informed consent as required). Pilot testing will also require demonstration of adequate follow-up capacity and ability to ensure access to treatment without financial barriers.
After pilot testing has demonstrated that newborn screening for T-cell lymphopenia can be performed with an extremely high degree of accuracy and with the ability to provide follow-up services and treatment to all affected children identified through screening, the next step is for a national-level body to officially recommend that states include this test in the standard newborn screening panel. Each state should have an advisory committee that would consider such a recommendation.
Dr. Anne Comeau provided some subsequent follow-up points to be considered in any further discussions of clinical and laboratory-based protocols for newborn screening for profound T-cell depletion, from the perspective of how newborn screening programs function:
Cost of assay--The typical "add-on" charge for laboratory-based services (i.e., no tracking, follow-up, consultation, or assurance of universality) for any one disorder is $2 to $5.
Format of assay for primary screen--All newborn screening laboratories should have the technical capabilities to run ELISA or fluorometric assays if they can be standardized on a blood spot. Other potential assays that require nationwide technical capabilities for running DNA- and MS/MS-based systems would suffer delays in nationwide implementation. Regionalizing this aspect of newborn screening outside of established newborn screening programs would not be optimal.
Absence of detectable analyte (e.g., T-cell marker) as it relates to specificity of primary assay--Consider an internal standard within the measurement.
Format of assay for secondary screen vs. confirmatory testing--This issue needs further discussion.
Treatment of confirmed cases--Access to local "centers of immunology excellence" would be invaluable in the work-up of many of the babies who screen positive. Other issues to consider include: (1) adequacy of quality-controlled flow cytometry facilities nationwide to provide confirmatory T-cell differentials to "rule in" babies for further clinical follow-up; (2) adequacy of facilities for diagnostic work-up and treatment; and (3) adequacy of feedback mechanisms from diagnostic centers to screening laboratories for continual optimization of screening algorithms.
Dr. José Cordero, Director, NCBDDD, CDC
Dr. Cordero thanked the organizers and participants and reviewed some of the meeting highlights. He applauded the group for their acknowledgment of the importance of accurate case definitions, the role of early diagnosis of PI diseases in preventing morbidity and mortality, and the need for population-based prevalence data as a tool for assessment. In terms of interventions, Dr. Cordero pointed out the need to demonstrate success with one example that can be used as a model for other diseases. The interventions discussed encompass several goals: helping children, educating practitioners, developing and maintaining awareness of PI diseases, and prompting and informing policies and policy changes. He urged the group to maintain its focus on assessment; as new trails are blazed in public health and genetics, assessment tools are essential for tracking and monitoring progress and documenting outcomes and successes. He closed by cautioning that business as usual is not acceptable. The need now is to move on, to do more, and to embark on the next steps. The public health framework that is being developed is important for PI diseases and also for other genetic disorders that can benefit from early diagnosis and opportunities for interventions to improve health outcomes.