July 2010
Feature Story

William Check, PhD

Those who worked in clinical immunology 30 years ago knew only three types of primary lymphocyte deficiencies: DiGeorge syndrome with defective T cell production, B cell defects, and combined T and B immunodeficiencies. To recognize an immunodeficiency, they looked for infections with unusual organisms and a family history.

Genetic knowledge has brought dramatic change. “More and more, mutations in specific genes are being identified in immunodeficiency syndromes,” says Irene Check, PhD, D(ABMLI), director of the Division of Clinical Pathology in the NorthShore University HealthSystem, Evanston, Ill., and clinical professor of pathology, University of Chicago. She points to tyrosine kinase mutations in Bruton’s agammaglobulinemia, STAT3 mutations in hyper-IgE (Job) syndrome, DOCK8 mutations in an autosomal-recessive form of hyper-IgE syndrome, and adenosine deaminase. Finding these mutations leads directly to genetic tests and thus enhances diagnosis.

Rebecca Buckley, MD, professor of pediatrics and immunology at Duke University School of Medicine, says there are now more than 150 distinct primary immunodeficiency diseases. “And the causative genes have been identified for 80 percent of them,” she says. “That’s an enormous advance. And it all occurred since 1993. This is an exploding field.”

In addition to genetics, Dr. Check points to advances in the understanding of immune function, such as the relation between specific infections and specific immune defects. For instance, patients who are susceptible to atypical mycobacterial infections have defects in an interferon gamma (IFNγ)/interleukin 12 (IL-12) pathway. “Our clinicians are actually asking for this type of testing,” Dr. Check says.

Vijaya Nagabhushanam, MD, PhD, assistant medical director of the Advanced Diagnostic Laboratories at National Jewish Health in Denver, performs tests for defects in the IFNγ/IL-12 pathway. “Better clinical recognition and diagnostic testing have helped uncover an increasing number of immune defects,” Dr. Nagabhushanam says. She and her colleagues are focusing now on developing functional flow cytometry assays for cytokine receptors. “Expression of receptors on the cell surface might appear normal on immunophenotyping, but might not function normally,” she says.

Some progress has also been made in identifying genetic defects in common variable immunodeficiency, or CVID, one of the primary immunodeficiency diseases most commonly encountered in clinical practice. There are clinical tests for a few of them, though tests for others remain in the research arena since these defects are very rare, says Roshini Abraham, PhD, consultant in the Division of Clinical Biochemistry and Immunology and director of the cellular and molecular immunology laboratory at the Mayo Clinic (Park MA, et al. Lancet. 2008;372:489–502). Insights into the molecular basis of immune dysfunction in CVID are likely to be helpful in improving patient management and lowering mortality, says Dr. Abraham, who is also an associate professor of laboratory medicine/pathology and of medicine, Mayo Clinic College of Medicine.

If the advances in understanding primary immunodeficiency diseases are already notable, they’ll soon become even more so. “The biggest take-home point with respect to diagnosis of primary immunodeficiencies is how rapidly it is going to change in the next five years,” says Kathleen E. Sullivan, MD, PhD, chief of the Division of Allergy and Immunology, Wallace chair of pediatrics, and director of the immunology clinic at The Children’s Hospital of Philadelphia. Dr. Sullivan cites two advances discussed at the May 2010 North American Primary Immune Deficiency National Conference: new molecular technology for PID and newborn screening for severe combined immunodeficiency disease, or SCID.

“Current testing for immunodeficiencies is expensive and time-consuming,” Dr. Sullivan says. Looking for a mutated gene in a child with SCID costs between $3,000 and $5,000 per gene tested. “Often we need to test multiple genes. And it can take weeks to get a result. This is frustrating to families,” Dr. Sullivan says. “SNP arrays and whole-genome sequencing are also expensive now, but they offer the promise of diagnosing currently unknown immunodeficiencies and doing it faster.” She predicts that the cost of whole-genome sequencing, which has already decreased greatly, will decline further as more clinical labs begin to do it.

The second advance, newborn screening for SCID, will transform clinical immunology practice, she says. Only two states—Wisconsin and Massachusetts—now include SCID in their newborn screening panels. However, on May 21 Kathleen Sebelius, secretary of the Department of Health and Human Services, added SCID to the Uniform Screening Panel, a development Dr. Sullivan calls “tremendously exciting.”

“The survival rate of patients with SCID is so much higher if patients are transplanted early,” she explains. Transplantation in the first month has a 98 percent survival rate, in contrast to 50 percent for transplantation at six to eight months. “Identifying these children in the first week or so of life has the potential to save many lives, as well as to improve quality of life,” she says.

Screening for SCID in newborns is done by quantitating T cell receptor excision circles, or TRECs, a test devised by Jennifer M. Puck, MD, professor of pediatrics at the University of California, San Francisco, when she was working at the National Institutes of Health (Chan K, Puck JM. J Allergy Clin Immunol. 2005;115:391–398). “I didn’t discover excision circles,” Dr. Puck explains. “Other people found them in T and B cells, which are the only cells in the body that cut up genes and splice them back together, creating a small, free-floating circular piece of DNA. My idea was to take advantage of excision circles made by T cells for this test. If babies are not making T cells for any reason, they will not have TRECs.”

Classification of primary immunodeficiency diseases is based on the nature of the immune defect, says Luigi D. Notarangelo, MD, professor of pediatrics and pathology and Jeffrey Modell chair of pediatric immunology research at Harvard Medical School (Notarangelo LD. J Allergy Clin Immunol. 2010;125:S182–194). In SCID, the most severe primary immunodeficiency disease, both T and B cells are defective or absent, which can cause death in the first year unless the patient receives hematopoietic stem cell transplantation or other immunerestoring treatment.

Second in severity and more common are conditions in which antibody is absent, Dr. Notarangelo says. In some of these the mutated gene is well defined, including Bruton’s agammaglobulinemia, the first such condition described, in 1952, and the first for which a responsible gene—Bruton’s tyrosine kinase (BTK), on the X chromosome—was identified, in 1993. This condition is called X-linked agammaglobulinemia, or XLA. Before the introduction of intravenous immune globulin (IVIG), XLA was often fatal.

Common variable immunodeficiency is a more common antibody deficiency, Dr. Notarangelo says, occurring in one in 10,000 to one in 20,000 births. Typically, there is failure of antibody production. So far only a few well-defined genetic defects have been discovered that account for about 10 percent of all cases of CVID.

Also included among primary immunodeficiency diseases are disorders in which immune deficiency is one component of a pleiomorphic presentation. In Wiskott-Aldrich syndrome, for instance, patients also have a reduced number of platelets (leading to hemorrhages), are prone to eczema, and are at higher risk for malignancies.

In the immunology clinic that Dr. Sullivan directs, the two most frequent primary immunodeficiency diseases are deletion of a segment of chromosome 22 (900 patients), which is a typical genetic diagnosis for DiGeorge syndrome; and CVID (200 patients). DiGeorge syndrome is another pleiomorphic condition, in which the major features are developmental anomalies of the face and heart, among others, giving rise to its other name, velocardiofacial syndrome. Development of the thymus can be affected, resulting in T cell deficiency. In about one percent of cases the thymus is absent, Dr. Sullivan says, which mandates thymic transplant. Genetic diagnosis of DiGeorge syndrome “was always at the forefront” of technology, Dr. Sullivan says. It was formerly done with FISH; more recently SNP arrays are used.

She describes the identification of CVID as “not particularly good,” owing to criteria that are less than exact. Its hallmark is low Ig levels and nonresponse to vaccines. However, Dr. Sullivan calls a “normal” Ig level in children “a moving target.” CVID is presumed to have a genetic basis, she says, and several family studies are trying to pinpoint genes. So far different candidate genes have been found in different families.

Dr. Abraham’s laboratory at the Mayo Clinic performs some of the relevant testing for the evaluation of CVID patients. At present, she estimates that the four genes that have been identified so far probably account for not more than 15 percent to 20 percent of CVID patients. “We don’t know the underlying basis in the vast majority of cases,” she says. “Perhaps the etiology of CVID is monogenic in a subset of patients and oligogenic in others.” In addition to very low levels of IgG and either IgA or IgM plus impaired or absent antibody response to vaccine antigens, a third criterion for CVID is recurrent sinopulmonary infections, such as otitis media and lower and upper respiratory infections. “These patients may have very heterogeneous presentations,” Dr. Abraham says. Symptoms can include gastrointestinal complications; autoimmunity, especially autoimmune cytopenias; and increased susceptibility to malignancies. Laboratory workup includes measuring serum immune globulins and functional immune responses to protein and polysaccharide vaccines, as well as cellular defects in adaptive immunity. “About one-fourth of these patients can have T cell defects and 10 percent have been reported as having natural killer [NK] cell defects,” Dr. Abraham says.

Dr. Abraham also offers complete workup for XLA. Defining criteria include profound decrease in all three major isotypes of gamma globulin, absent or drastically reduced circulating B cells, and recurrent sinopulmonary or enteroviral infections. In addition to tests for the immune defects, Dr. Abraham performs flow cytometry for BTK protein and full genetic sequencing and known mutation analysis for the BTK gene. “There are over 600 known mutations in the BTK database,” Dr. Abraham says. Since her laboratory started doing BTK sequencing in fall 2008, it has identified 21 patients with BTK mutations, nine of which were novel.

“We haven’t stopped gaining knowledge in XLA patients,” Dr. Abraham says. Current thinking in the literature is that BTK is also important in innate immunity and that BTK’s role in innate immunity may explain the susceptibility to enterovirus meningitis in XLA patients. “We didn’t even suspect this 10 years ago,” Dr. Abraham says. “We are still digging and looking for defects in the B cell differentiation pathway.”

Dr. Abraham emphasizes the importance of geno­type/phenotype correlations in primary immunodeficiency diseases, citing XLA as a prime example. She and her clinical colleagues at the Mayo Clinic recently saw a 65-year-old man who had had infections all his life. He had one brother who had died at age eight and three surviving brothers who also had frequent infections. “Yet no one had ever given him a diagnosis, and no one had ever treated him with IVIG,” Dr. Abraham says. “By doing the appropriate tests, including genetic analysis, we were able to identify the man’s problem as XLA and initiate IVIG therapy.”

Over the past year Dr. Nagabhushanam has been developing functional flow cytometry tests for patients who are susceptible to mycobacterial infections, such as atypical mycobacterial infections, leprosy, and M. avium, and who present with extensive disease. “We think these patients might have defects in receptors for IL-12 or interferon gamma, which are parts of well-defined pathways,” she says. “These receptors may not be expressed or they may be expressed but defective.” Dr. Nagabhushanam has generated assays to look at signals coming through these receptors. “So far we haven’t found cases where that’s the problem,” she says. “We have found a few cases with diminished activity, but not a genetic defect.” Functional defects could be causative or they could be secondary to immune repression by the infection. “At this point we are still researching what we can do with these assays,” Dr. Nagabhushanam says. She is now collecting a larger cohort to analyze.

She has also developed a flow cytometric assay for measuring naive T cells bearing TRECs using CD31. “This marker correlates very well with naive T cells emerging from the thymus,” Dr. Nagabhushanam says.

Another area she and colleagues have been focusing on is the innate immune system. They are looking at cell-surface expressed, pattern recognition molecules called Toll-like receptors, or TLRs, which are essential for recognizing pathogens. To that end, they have developed a functional test that evaluates whether all nine TLRs are acting correctly. (The assay measures cytokines produced following TLR stimulation with their cognate ligands.) “We have detected TLR defects in a couple of cases that correlate with the clinical picture, but we have not been able to follow these up with gene sequencing, so we are not ready yet to call these gene defects,” she says. Even so, she adds, “We’re pretty excited that we have seen diminished responses that correlate clinically.” For instance, in a patient who was ill with herpes simplex virus infection, there was no response to TLR3, one of three TLRs that recognize viruses.

Whole-genome sequencing is the future of testing for primary immunodeficiency diseases. Being able to look across the entire genome will revolutionize treatment, Dr. Sullivan says. “Our immunology clinic [at The Children’s Hospital of Philadelphia] is one of the largest in the country. Probably 80 percent of our patients under treatment for immune deficiencies do not have a genetic diagnosis,” she says. “We can’t talk to the parents about recurrence risk or talk to the kids about adult life. That’s a significant lapse in our specialty.” In practice, Dr. Sullivan foresees patients with recurrent infections undergoing a typical battery of immune function tests. “At that point we will go directly to sequencing.”

James W. Verbsky, MD, PhD, assistant professor of pediatrics and microbiology and molecular genetics at the Medical College of Wisconsin, told CAP TODAY about one case in which whole-genome sequencing has already proved valuable. “In some children with recurrent infections or other immune problems, we don’t know whether a transplant would help,” Dr. Verbsky says. “In these cases we have begun doing whole-genome sequencing. In one child we found a very rare gene defect.” The child had an unusual course, with a presentation featuring inflammatory bowel disease but also recurrent infections. After following the child for several years, Dr. Verbsky and his colleagues did whole-genome sequencing and found a mutation in a gene called XIAP that was discovered a few years ago to cause X-linked lymphoproliferative syndrome type 2, or XLP-2 (Rigaud S, et al. Nature. 2006; 444:110–114). “This presentation is not typical of XLP-2, and only with whole-genome sequencing was this diagnosis made.” With this information, Dr. Verbsky felt that hematopoietic stem cell transplantation was justified since this is the accepted treatment for XLP-2.

“Genetic diagnosis can help guide how we treat these kids,” Dr. Verbsky says. “The child with XLP spent most of his life in the hospital. After more than three years of looking for a cause of his symptoms, with sequencing it took a few months to define therapy.”

At the opposite end of the spectrum from selective use of the expensive sequencing technology is the relatively inexpensive public health measure, newborn screening. The recommendation for states to implement this test followed a review in 2009 by the Secretary’s Advisory Committee on Genetics, Health, and Society, for which Dr. Puck wrote the nominating proposal. The committee said the idea was a good one but noted the lack of evidence that it works. “In the following year Wisconsin reported that they had found a baby who had a primary immunodeficiency related to SCID who needed a transplant to survive,” Dr. Puck says (Routes JM, et al. JAMA. 2009;302:2465–2470). When the proposal was reviewed again early this year, it passed unanimously. HHS secretary Sebelius said on May 21 that she concurred and would be interested in following the outcome of SCID newborn screening as more programs become established. “The idea of tracking outcomes on a national level is also new,” Dr. Puck says. Pilot programs are being sponsored in Texas, New York, and California.

“Now comes the tricky part,” Dr. Puck says. “Implementation falls on states but no money is being provided.” It must be argued that newborn screening for SCID is cost-effective overall, yet there are no reliable data to support this contention. “Everyone agrees that in the absence of universal screening, the only way to detect babies born with SCID—outside of a family history, which most babies don’t have—is for babies to start getting infections,” Dr. Puck says. “And that’s what you want to avoid. Anyway, all babies get infections. How are pediatricians supposed to recognize in the first year of life that a baby has too many infections and has an immune problem? By the time the pediatrician recognizes it, that baby might have a fatal infection.” Or the baby could get Pneumocystis pneumonia and go on a ventilator for a week or two. “If that’s the way you make the diagnosis of SCID, you may already have a million-dollar bill,” Dr. Puck says. “SCID is so rare that most pediatricians and family doctors will never see a case. Only newborn screening can bail you out.”

The basis of the TREC assay is that T and B cells cut and re-join gene segments to create diversity in their receptors for recognizing antigens. “What happens to the pieces that are excised?” Dr. Puck asks. “Are they left on the cutting room floor? No. They are turned into circles because their ends are recognized by the enzymes that stitch genes back together.” Dr. Puck devised a method to extract DNA from the dried blood spots already obtained for all babies and perform the PCR reaction previously described by Daniel Douek, MD, PhD, that amplifies the junction of the circles (Douek DC, et al. Nature. 1998;396:690–695). If the PCR reaction is positive, then that person has T cells that are making circles. TRECs can be measured on the dried blood spots routinely collected from babies. “We don’t need separate blood samples since DNA circles are quite stable,” Dr. Puck says. She validated the test using leftover dried blood spots: DNA from SCID babies showed no TRECs. What is so elegant about the TREC assay is that it is a DNA test but not a gene test, so it works for all genotypes. “This is important because SCID is caused by at least 16 or 17 genes that we know and more that we don’t know,” Dr. Puck says.

Mei W. Baker, MD, FACMG, science director of the newborn screening program in the Wisconsin State Laboratory of Hygiene, oversees Wisconsin’s SCID screening test. “We really did quite a bit of homework before setting up the program,” she says. “The tough part was handling 300 samples per day. Because we are in an environment for newborn screening, we deal with high throughput all the time.” Dr. Baker, who is assistant professor of pediatrics in the School of Medicine and Public Health at the University of Wisconsin-Madison, says SCID newborn screening is the first newborn screening program to use DNA as the primary screening test. In newborn screening for cystic fibrosis, the mutation panel follows an immunoassay.

Dr. Baker points out that the TREC assay doesn’t detect any specific gene defect. “A beauty of this test is that it doesn’t detect genomic DNA, but episomal DNA excised from the genome,” she says. “That allows us to avoid detecting carriers.”

During the first year of the program they screened 71,000 newborns. “The test performed very well,” Dr. Baker says. “It has a very low screening positive rate.” Several infants were diagnosed with T-cell lymphopenia. “If you stick to the traditional definition of SCID, we detected one case. However, we detected more than one case with clinically significant immunodeficiency.”

Infants with a positive TREC screen get confirmatory testing in the laboratories of Dr. Verbsky and Jack Routes, MD, professor of pediatrics and microbiology and medical genetics, Medical College of Wisconsin. Follow-up tests consist of flow cytometry to evaluate T, B, and NK cell counts. “The screen is designed to find children with very low T cells who have a risk of dying in the first year of life,” Dr. Verbsky says. “All forms of SCID have low T cells, which is detected by the TREC assay. In addition, we also pick up kids with low numbers of T cells but not low enough to be considered SCIDs, but who may still have very severe immune deficiencies.” Since the program was started, four babies were detected who did not meet the definition for SCID but who had dangerously low T cells. “A fifth, who was born before the screen was instituted, was picked up because her brother was positive,” Dr. Verbsky says. It can be difficult to determine the best treatment option in children detected by the TREC assay but with T cell numbers not classically considered SCID. Drs. Verbsky and Routes plan to use whole-genome sequencing on these cases to help guide therapy.

To emphasize the power of the TREC screen, Dr. Verbsky notes that the first child to be detected as needing stem cell transplantation in the Wisconsin program did not meet the classic definition for SCID. “It was kind of confusing,” he says. “This baby did not have the absent T cells characteristic of SCID, but he was getting infections at a very young age. We discovered he had a de novo mutation in the Rac2 gene, which was described only once before in a child with neutrophil problems” (Williams DA, et al. Blood. 2000;96:1646–1654; Ambruso DR, et al. Proc Natl Acad Sci USA. 2000;97:4654–4659). Children with SCID usually have intact neutrophils, yet the baby found in the Wisconsin program was acting as though he had a neutrophil problem. At the same time he was found because he failed the TREC screen. A possible explanation arises from the fact that Rac2 is necessary for chemotaxis, which explains why a defect in Rac2 affects neutrophils. Perhaps Rac2 affects T cells the same way. “Very likely the child failed the screen because his T cells couldn’t move,” Dr. Verbsky speculates. In any case, the combination of deficiencies in both neutrophils and T cells created a very bad immunodeficiency. “We think he would have developed the kinds of infections babies get with SCIDs but even earlier,” Dr. Verbsky says. The child received a cord blood transplant and is now doing well off all medication. “This de novo mutation was picked up out of 70,000 babies,” Dr. Verbsky points out, “so the test is very sensitive. It’s not going to miss much.” In addition, he notes, this case demonstrates that this assay can also detect severe immune deficiencies not typically thought to be SCID.

While the TREC assay is a good test, Duke’s Dr. Buckley is concerned that most state labs are not geared up to use DNA-based technology. “I have been advocating for the past 13 years using the white blood cell count and manual differential on cord blood to detect immunodeficiencies in newborns,” she says. Seventy percent of circulating leukocytes are T cells. In SCID there are no T cells. “If a person is missing 70 percent of their white cells, they are bound to be lymphopenic,” Dr. Buckley says. “On babies born here because of a positive family history of SCID or early infant death, we do a stat automated blood count and differential to look for lymphopenia on cord blood. If we find it, we do stat flow cytometry for T, B, and NK cells. We pick up affected infants within several hours after birth that way and transplant within a few days after birth.”

Dr. Buckley recently reviewed the published literature on bone marrow transplantation from centers around the world with regard to B cell function (Buckley RH. J Allergy Clin Immunol. 2010;125:790–797). “My main finding was that pretransplantation conditioning with chemotherapy does not guarantee B cell function,” she says. In favor of conditioning regimens, Dr. Buckley found that the percentage of survivors with B-cell chimerism, function, or both was higher and the percentage requiring immunoglobulin replacement was lower at those centers that used pretransplantation conditioning. However, substantial numbers of patients required immunoglobulin replacement at all centers. More important, she says, “Independent of whether B cell function was present, mortality was horrible at centers that used chemotherapy or posttransplantation graft-versus-host disease prophylaxis. We advocate no pretransplantation chemotherapy and no posttransplantation graft-versus-host disease prophylaxis.”

Aggressive measures like pre-transplant chemotherapy and post-transplant immunosuppressive drugs to prevent GVHD are not necessary and are counterproductive to the goal of reconstituting the immune system, Dr. Buckley says. “Heme-onc doctors argue that unless you give chemotherapy you won’t get B cell function,” she says. “But these patients come in very ill. If you give them chemo, they get even more ill and some die. What kills most of them is the infection they come in with.” With Dr. Buckley’s protocol, survival is 94 percent for infants transplanted in the first 3.5 months of life and 70 percent for those transplanted after 3.5 months (Railey MD, Lokhnygina Y, Buckley RH. J Pediatr. 2009;155:834–840). “Mostly we use mothers as stem cell donors,” she says. “We bring them in for 24 hours, do the transplant, and send them to an apartment in town. We can do this because we don’t give chemotherapy or immunosuppressive drugs. Our approach is cheaper, safer, and faster.”

Advances in hematopoietic stem cell transplantation may come from a recently formed consortium. “For many years in the U.S., stem cell transplantation for immunodeficiencies has been done at single centers without major collaborative efforts,” Harvard’s Dr. Notarangelo says. “Now, for the first time, all major centers that care for patients with severe immunodeficiencies in the United States have formed a consortium and were awarded NIH funds.” The intent, he says, is to perform retrospective and prospective studies with creation of a national registry (similar to what has been done successfully in Europe), with the ultimate goal of developing more effective forms of treatment based on cell therapy for these disorders. “Collaboration is key to finding effective treatment. In this regard, we will collaborate with the major primary Immunodeficiency Network in the United States [USIDNET] and with the Center for International Blood and Marrow Transplant Research [CIBMTR].”

While stem cell transplantation works in many cases, an alternative solution to primary immunodeficiency disease may be gene therapy. At the recent North American PID Conference, Dr. Sullivan says there was a report showing that a lot of progress has been made with gene therapy. “There is still concern about malignancies, but there are some innovations that seem to have minimized that risk,” she says. Gene therapy protocols have been reopened for X-linked SCID, and are now ongoing for SCID due to adenosine deaminase (ADA) deficiency, Wiskott-Aldrich syndrome, and chronic granulomatous disease.

“In the U.S. there has been a long moratorium on trials of gene therapy with the exception of ADA-SCID,” Dr. Notarangelo says (Aiuti A, et al. N Engl J Med. 2009;360:447–458). Trials were stopped when several subjects in gene therapy protocols for X-linked SCID and chronic granulomatous disease developed leukemia or myelodysplasia, respectively. “Now a new gene therapy protocol for X-linked SCID has been approved.” Dr. Notarangelo is the principal investigator of this trial at Children’s Hospital in Boston. The trial will involve five centers in London, Paris, Boston, Cincinnati, and Los Angeles. “For the first time,” he says, “all of these five centers will be using exactly the same vector to introduce the gene of interest. This new vector has significant differences as compared to the one used in previous trials. To decrease the risk of malignancy, all potentially dangerous viral sequences were eliminated from the vector. All preclinical data indicate it is much safer,” he says. “Expression of the transgene is now driven by a human promoter.”

Gene therapy technology and disease identification are advancing in tandem. “We know a lot more about these diseases than we used to,” Dr. Puck says, “and we will learn a ton more when widespread genetic screening is in place. We will prospectively see infants who are not making T cells properly. In Wisconsin, the first baby they found was not one you would expect not to be making excision circles. We will find things that are brand new.” They will also have a chance, she says, to see how frequent these diseases are and how many babies die of infections without anyone even knowing they had a host defense problem. Only when such children are identified can they be treated.