SESSION III. LABORATORY ISSUES
This website is archived for historical purposes and is no longer being maintained or updated.
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
Speakers discussed current diagnostic tests for PI diseases and addressed the role of genetic tests in clinical practice. They also considered three essential criteria for evaluating diagnostic and screening tests:
Analytic validity is the ability of a laboratory test or clinical screening tool to accurately and reliably measure the item of interest. This includes a test's sensitivity and specificity, quality control (test reliability), and robustness (test variability). An example for asymptomatic screening is performance of a complete blood count and manual differential on a cord blood sample to measure the absolute lymphocyte count. How well does this test measure the lymphocyte count? Is the result repeatable? An example for symptomatic screening is patient reporting of the number and type of infections. How accurate and reliable are these reports? Do they agree with the medical record?
Clinical validity is the ability of a test to detect or predict the disease of interest. This includes the clinical sensitivity of the test (detection rate), clinical specificity of the test, prevalence of the disease, positive and negative predictive values, and penetrance of the disease. For example, for asymptomatic screening, what proportion of persons with agammaglobulinemia will have a detectable mutation in the gene for BTK (clinical sensitivity)? What proportion of normal persons do not have a mutation in the gene for BTK (clinical specificity)? What is the prevalence of agammaglobulinemia in the population? What is the prevalence of mutations in the BTK gene in the population? If there is a detectable mutation in the BTK gene, what is the likelihood that a person will have aggamaglobulinemia? Does the type of mutation in the BTK gene (genotype) correlate with the severity of the disease (phenotype)? Do all persons with a similar mutation in the BTK gene develop disease, at the same age, with the same severity? For symptomatic screening, what proportion of persons with a specific set of symptoms will have PI? What proportion of persons with common variable immunodeficiency develop pneumonia (clinical sensitivity)?
Clinical utility encompasses all of the elements to be considered in the evaluation of a test before its introduction into routine practice. These include considerations of the timing of use of the test, the availability of effective interventions, quality assurance procedures, assessment of real world performance, health risks associated with testing, economic evaluation of cost-benefit ratio for testing, and availability of facilities and resources to manage all aspects of the service. For example, how does detection of PI in a newborn, or early in the course of the disease, affect the outcome? Is it cost effective to diagnose PI early? Is diagnosis at birth associated with any risks for the patient or the family? Do we have the ability to correctly diagnose and treat all patients identified by screening or early recognition?
Dr. Haddow outlined a method for systematically evaluating screening tests (Wald and Cuckle. Brit J Obstet Gynecol 1989). Despite the lack of a universally accepted definition of medical screening, general agreement exists on three characteristics. First, screening is a process of selection to identify persons who are at a sufficiently high risk for a specific disorder to warrant further investigation or direct preventive action; it is usually a preliminary process to diagnostic testing and, if required, preventive action. Second, screening is systematically offered to a population of persons who have not sought medical attention for symptoms of the disease for which screening is being conducted; it is usually initiated by medical authorities and not by patient request. Third, the purpose of screening is to benefit the persons being screened. Although a screening test typically consists of a laboratory measurement, it may involve only a simple question, such as asking a pregnant woman her age.
The proposed process for evaluating a screening test builds upon Wald and Cuckle's method and begins with a description of the medical disorder that is the focus of screening. The disorder should be well-defined, distinct from the screening test, and sufficiently serious in terms of morbidity and mortality to warrant testing. Dr. Haddow emphasized that the medical disorder should be defined not in terms of the screening result (e.g., hypercholesterolemia) but rather the actual clinical condition (e.g., risk for coronary artery disease).
The state of knowledge about the prevalence of the disorder in the population being tested can then be documented. Prevalence is defined as the number or proportion of existing cases of a specified condition in a population. Prevalence is a key factor in determining test performance. The more uncommon the disorder in the population tested, the more the system for identifying it is stressed.
The test being considered for screening purposes is then examined in some detail, including whether it is used alone or as part of a series and whether detection and false-positive rates have been established for the medical disorder being sought. Constructing a flow diagram that applies the detection rate, false-positive rate, and prevalence of the disorder can help clarify overall test performance. Such a diagram might start with a population of 100,000, divided into those with and without the disorder, and end with the number of cases detected and the number of
false-positive results. The evaluation also includes a determination of therapeutic intervention(s) to be initiated for persons who test positive and of the efficacy of the intervention(s).
If, at this stage, introduction of the test appears feasible, then issues such as costs and benefits can be explored, along with practical problems that might accompany introduction. The cost-benefit analysis includes both financial and medical costs and benefits. The medical aspect of the analysis takes into account morbidity and/or mortality that might result from diagnostic or therapeutic procedures and weighs these costs against the benefits. Practical problems with implementation include program management, access to the population, obtaining and transporting samples, and quality assurance issues.
When the preliminary process suggests that a test is feasible to introduce, then a more detailed, second-level assessment can be carried out. One such process that Dr. Haddow is developing and testing with colleagues from CDC is the ACCE evaluation process for genetic tests, which is derived from the four components of evaluation: Analytic validity (ability of the test to accurately and reliably measure the item of interest), Clinical validity (ability of the test to detect or predict the disorder/disease of interest), Clinical utility (encompasses all of the elements to be considered in the evaluation of a test before its introduction into routine practice), and Ethical, legal, and social issues associated with the test. This second level is performed not only for validation but also to identify problems and remedies aimed at improving the test's performance in practice.
Early diagnosis is possible for most PI diseases, is essential for optimal therapy, and may be life saving. Available diagnostic tools include clinical examination, cellular testing, immunologic testing, metabolic testing, molecular testing, and genetic analysis.
Because primary immune defects are often inherited, a family or patient history may be one of the best aids to diagnosis. The family or patient history is also important in identifying infections that are persistent or recurrent, difficult to treat, or caused by unusual microbes. The age of onset may also help in diagnosis. Physical examination can identify chronic anatomic changes secondary to infections and characteristic physical findings. Clinical presentations can vary, however, from patient to patient, even for mutations of the same gene.
Most laboratory tests used by informed clinicians are available through commercial or hospital laboratories. Specimens required for analysis include anticoagulated whole blood, serum, cultured T and B cells, lymphoid tissue, cell- or tissue-derived genomic DNA, mRNA, and cell extracts. Tests to confirm or establish a diagnosis of PI disease in each of the four categories of defects include the following:
Assessment of humoral immunity
- Measurement of serum immunoglobulin and IgG subclass levels.
- Quantification of antigen-specific antibody responses.
- Isotype switching, amplification, immunologic memory, V-beta rearrangement.
- B-cell enumeration: identification of pre-B cells, mature B cells with or without B27 on the surface, plasma cells.
For example, a specific bacteriophage (a bacterial virus that is not infective for humans) is a potent, safe, and useful antigen; it allows measurement of antigen clearance and primary and secondary immune responses.
Assessment of cellular immunity
- Absolute lymphocyte count.
- Evaluation of surface-marker subsets for T cells (CD3, CD4, CD8) and natural killer cells (CD16).
- Expression of B cells (CD19, 20), major histocompatibility complex class II (MHC II).
- In vitro lymphocyte proliferation test.
- Cell activation markers (CD40 ligand [CD40L]).
- Lymphokine tests (protein, receptors, mRNA, lymphokine function).
The latter group of tests are all specific, cumbersome, and not yet suitable for screening.
Assessment of phagocytic cell function
- Polymorphonuclear cell counts.
- Nitroblue tetrazolium (NBT) dye reduction test; quantitative flow cytometry (dihydrorhodamine).
- Measurement of phagocytic response using chemiluminescence.
- Bactericidal assays.
- Assessment of the integrity of the inflammatory response by Rebuck skin window techniques.
- In vitro analysis of the inflammatory response by measurement of chemotaxis (Boyden chamber).
- Adhesion molecule assays (e.g., CD18); neutrophil rolling.
Assessment of complement components and function
- Functional and immunochemical assays that measure total complement components of the classical and alternative activation pathways (CH50, AH50).
- DNA analysis.
In addition, immunologic tests have been made specific for particular gene products and can be used to establish a specific genetic diagnosis. These gene products include: Bruton's tyrosine kinase (BTK) in platelets; CD40L expression by activated T cells; common gamma chain expression; Wiskott-Aldrich syndrome protein (WASP) by Western blot or direct staining of lymphocytes; interferon gamma receptors, MHC II, FAS expression; adenosine deaminase in red blood cells; and 22q11 deletion.
Unfortunately, among the tools available for diagnosis of PI diseases, few are suitable for newborn screening of specific diseases. Possibilities might include screening for XLA by measurement of B cells in cord blood, BTK activity in platelets, or gene analysis; screening for hyper-IgM syndrome by measuring expression of CD40L; and screening for WAS by assessing platelet size, platelet number, and presence of WASP in cells.
A plethora of mutations across the spectrum of exons characterize genes for inherited immune deficiencies; no one mutation predominates. Therefore, screening for particular mutations is not practical. Genetic testing is used to establish a diagnosis, but no single testing modality is appropriate for all situations. Moreover, the state of the art is progressing so rapidly that tests quickly become obsolete.
The simplest diagnostic procedure is to obtain a family history. Families are often able to point out potentially affected members in past generations whose presentation and course were compatible with the suspected disorder. For example, the X chromosome harbors more recognized immunodeficiency genes than any other chromosome, and the frequency of occurrence of X-linked immunodeficiency diseases is disproportionately high. The X-linked inheritance pattern of multiple affected males related through maternal lines should therefore be sought in all patients presenting with immunodeficiency. In addition, several PI diseases (e.g., SCID, CGD) can be caused by defects in any of several different genes. Family history can be critical in differentiating the X-linked from autosomal forms. Unfortunately, however, a negative family history for immune disorders never helps to rule out a PI disease; often, the most common clinical history is no family history. Therefore, clinicians must recognize both the need to take a family history and the fact that a negative family history does not preclude a certain defect.
Molecular methods for mutation diagnosis are being applied to many genetic conditions. For PI diseases that have been mapped to a unique genomic region but not yet identified, linkage analysis can define which chromosome segments in a family carry mutated disease gene alleles, provided both affected and unaffected relatives are available for testing. Linkage analysis can also be useful when the defective gene is known but the specific mutation has not been determined. Studies of mRNA production are important in assessing the consequences of a mutation in DNA sequence but will not pick up all patients with any of the PI disorders.
X-inactivation testing has been useful in identifying female carriers of selected X-linked immunodeficiencies (e.g., X-linked SCID, XLA, WAS). Lyonization (random X inactivation) normally results in two distinct cell populations, each with one of the X chromosomes active. Female carriers can be identified because their mutation-bearing X chromosomes are not found as active chromosomes in cell lineages affected by the gene defect. Now that the genes for these diseases have been identified, however, specific mutation detection is a more accurate and appropriate diagnostic approach.
Single-strand conformation polymorphism (SSCP) analysis is the simplest technique for mutation screening. SSCP distinguishes between similar-sized DNA fragments according to the mobility of the single-stranded DNA under polyacrylamide gel electrophoresis. Although useful because of its simplicity, SSCP is not sensitive enough to rule out any disease-causing mutation in a given segment of DNA. Dideoxy fingerprinting achieves improved sensitivity over SSCP by using a dideoxy nucleotide mixture from a dideoxy terminator sequencing reaction instead of polymerase chain reaction (PCR) of an entire fragment to make single strands of DNA for conformation testing. DNA sequencing (determination of the actual nucleotide sequence of mutant DNA) is the gold standard for mutation detection. Because of technology advances spurred by the Human Genome Project, methodologies for performing and analyzing DNA sequences are rapidly advancing. Sequencing can be done on direct amplified DNA from affected patients, known or possible carriers, and prenatal samples.
Molecular diagnostic testing is generally undertaken to establish a diagnosis in an affected person. Prenatal testing is also possible for some PI diseases. Dr. Puck reported on a series of families followed with pregnancies at risk for X-linked SCID. Findings indicated that most couples wanted to have prenatal testing, but few affected pregnancies were terminated. In this disease, the improving outlook over the past two decades of bone marrow transplantation therapy has given families hope. Couples anticipating delivery of affected infants took advantage of the information gained through counseling and prenatal testing to learn about and obtain treatment.
- Page last reviewed: June 15, 2009 (archived document)
- Content source: