A Girl with Autoimmune Cytopenias, Nonmalignant Lymphadenopathy, and Recurrent Infections

We describe a girl, now 9 years of age, with chronic idiopathic thrombocytopenic purpura, persistent nonmalignant lymphadenopathy, splenomegaly, recurrent infections, and autoimmune hemolytic anemia. Her symptoms partly fit the definitions of both autoimmune lymphoproliferative syndrome (ALPS) and common variable immunodeficiency disorders (CVIDs). Genetic analysis showed no abnormalities in the ALPS-genes FAS, FASLG, and CASP10. The CVID-associated TACI gene showed a homozygous polymorphism (Pro251Leu), which is found also in healthy controls.


Introduction
Acute idiopathic thrombocytopenic purpura (ITP) is a wellknown clinical entity in children. Generally, in children the disease is self-limiting and easily distinguished from a hematological malignancy, even without investigating the bone marrow [1]. However, the case becomes more complicated when the ITP becomes chronic and accompanying profound lymphadenopathy develops. We describe the diagnostic dilemma in a girl with these problems, who with time also developed recurrent respiratory infections, suffered from a prolonged episode of intractable diarrhea, a severe episode of varicella zoster infection and autoimmune hemolytic anemia.

Patient
The girl, now 9 years of age, is the second child of healthy nonconsanguineous Caucasian parents. She was born after an uncomplicated pregnancy and delivery and showed normal growth and development. Her family history reveals allergy on the paternal side and autoimmune disease and malignancies on the maternal side. At the age of 14 months, she developed ITP and showed a partial slow recovery after 3 days of high-dose intravenous immunoglobulins (IVIGs) followed by prednisolone. One year later, she suffered a relapse during a mild parainfluenza type 3 infection and treatment with 3 days of high-dose IVIG was started again. A few days after receiving this second course of high-dose IVIG she developed cervical, axillary and inguinal lymphadenopathy and enlarged tonsils: this lymphadenopathy never resolved. There was no hepatomegaly, splenomegaly, or mediastinal or abdominal lymph node enlargement at that time. Blood tests showed a mild normocytic anemia (Hb 6.4 mmol/L, MCV 77 fL) and granulocytopenia (0.7-1.0 × 10 9 /L) and large unstained cells in the hematology analyzer (8%; 0.4 × 10 9 /L). A bone marrow aspirate and biopsy showed some atypical lymphocytes and specific maturational disturbances, 2 Case Reports in Immunology but no malignancy. Bone marrow immunophenotyping was normal. FAS-mediated apoptosis of T-lymphoblasts was normal (two separate tests in two different laboratories). She suffered from recurrent upper respiratory tract infections and two pneumonias in the following years and a prolonged episode of intractable diarrhea. The infection frequency improved on cotrimoxazole prophylaxis. The enlarged tonsils, cervical, axillary and inguinal lymphadenopathy and variable amounts of atypical lymphocytes in her differential remained, splenomegaly developed as well. The parents increasingly felt that she got tired more easily than other children of her age. At the age of 5 years, an adenoidectomy was performed; this procedure was combined with an excision of an inguinal lymph node for histological examination. At the age of 7 years, an episode of severe varicella zoster infection occurred which was treated successfully with 1 week of intravenous aciclovir. Five months later, she acutely developed autoimmune hemolytic anemia, which initially responded well to another course of high-dose IVIG and prednisone, but she relapsed when the prednisone was slowly tapered and stopped. Unexpectedly, her chronic thrombocytopenia improved upon this treatment. At the age of 8 years, she developed pulmonary problems with dyspnea after an (probably viral) airway infection; high resolution CT scan showed a granulomatous lymphocytic interstitial lung disease (GLILD), a form of pulmonary lymphoproliferative disease. Therapy with mycophenolic acid was started, after which she showed a slow but nearly complete pulmonary recovery; she is still slightly dyspneic upon exertion. She also developed a uveitis, which was treated with prednisolone eye drops. Now, at the age of 9 years, she is relatively stable on mycophenolic acid; stem cell transplantation is being considered.

Material and Methods
Upon her first presentation at the pediatric immunology clinic in 's-Hertogenbosch at 5 years of age, extensive investigations were performed (Table 1). Four-color immunophenotyping was performed as previously described [2,3]. One representative paraffin-embedded tissue block of both lymph node (18 mm) and adenoid (20 mm) was selected and immunohistochemical stainings were performed ( Table 2). Routine protocols for Benchmark XT Ventana were used. Kappa, lambda, and EBV were assessed using routine in situ hybridisation technique (Kappa, ISH, Ventana; lambda, ISH, Ventana; EBER). Genomic DNA was isolated from peripheral blood granulocytes using the autopure kit (Qiagen, Venlo, the Netherlands). Exon-specific M13-tagged primers were used for amplification by PCR of all coding exons including flanking regions from the genes FAS (NCBI NM 00043), FASLG (NCBI NM 000639), CASP10 (NCBI NM 032974) and TACI (TNFRSF13B, NCBI NM 012452) followed by fluorescent sequencing (Applied Biosystems BigDye Terminator v1.1 Applied Biosystems). Table 1 shows the lymphocyte subpopulations in comparison with age-matched reference values from our laboratory [2].

Results
T-lymphocytes were slightly low to just normal, naive helper-T-lymphocytes and recent thymic emigrants were clearly decreased; double-negative TCRαβ + T-lymphocytes (cells without expression of CD4 or CD8 coreceptors) were normal to just increased in subsequent experiments, performed during episodes without immunosuppressive treatment.
Histological examination of lymphoid tissue revealed retention of the architectural features with a combined mild follicular hyperplasia and mild paracortical T zone expansion. In one lymph node a single granuloma was observed without necrosis or eosinophilia. Ziehl-Neelsen and other stainings for microorganisms were negative. The CD20 and CD79 positive B-cells were restricted to the normal Bcell areas. The follicular centres were highlighted by CD21 and CD23 B-cells, which stained the follicular dendritic cell meshwork. T-lymphocytes (CD2 + , CD3 + ) showed mild expansion in the paracortical T zone. BCL2 was negative and the proliferation in MIB1 was mainly restricted to the centrocytes and centroblasts in the reactive follicular centres. Kappa and lambda revealed a polyclonal plasma cell population. ALK staining was negative, there was no pathological CD30 staining. There were no signs of Rosai Dorfman; EBV was negative. Thus, histological examination showed abnormalities, but did not point to a specific diagnosis.
Analysis of the FAS, FASLG, and CASP10 genes revealed no abnormalities. Direct fluorescent sequencing of the TACI gene showed two polymorphisms, a homozygote polymorphism (p.Pro251Leu), and a heterozygote silent polymorphism.

Discussion
The girl we describe poses a diagnostic dilemma: there are several clinical entities that can be considered. Firstly, she undoubtedly fits the diagnosis of "Evans syndrome"which is defined by the presence of at least two autoimmune cytopenias-since she suffers from chronic idiopathic thrombocytopenic purpura and autoimmune hemolytic anemia [4]. However, this is only a descriptive diagnosis that does not encompass all her features. "Evans syndrome" is increasingly being associated with specific diseases, such as autoimmune lymphoproliferative syndrome (ALPS) and common variable immunodeficiency (CVID) and, mainly in adults, lymphoproliferative disorders [4][5][6][7][8].
Since she also suffers from persistent nonmalignant lymphadenopathy and splenomegaly, ALPS is an option for this girl. ALPS is a disorder of lymphocyte apoptosis leading to chronic nonmalignant lymphoproliferation. Affected individuals often suffer from autoimmune cytopenia, splenomegaly and hepatomegaly [9]. Laboratory findings include hypergammaglobulinemia and expansion of a unique population of circulating T-lymphocytes, referred to as TCRαβ + double negative T-cells which owe their name to the fact that they do not express CD4 or CD8 coreceptors [10]. These T-cells respond poorly to antigens. The genetic deficit in most patients is a mutation in the FAS gene which encodes a cell surface receptor which, upon stimulation, induces programmed cell death [11]. The diagnostic criteria for ALPS have recently been revised. The two required  Case Reports in Immunology    criteria for the diagnosis of ALPS are (1) chronic (>6 months), nonmalignant, noninfectious, lymphadenopathy, or/and splenomegaly, (2) elevated CD3 + TCRαβ + CD4 − CD8 − double-negative T-cells [10]. There are also four secondary accessory criteria; for a definite ALPS diagnosis a patient has to meet both required criteria and one of the primary accessory criteria. The diagnosis of ALPS is probable when two required criteria and any one of the secondary accessory criteria is present [10]. According to these criteria, our patient does not have ALPS, since she does not have consistently elevated CD3 + TCRαβ + CD4 − CD8 − double-negative T-cells. There are several well-defined ALPS-related disorders. Of these, RALD (RAS-associated autoimmune lymphoprolipherative disease) could be an option, but our girl shows no characteristic features of this disorder such as elevations in cells of myeloid origin [10].
The girl's recurrent infections could point to possible CVID. This disease is characterized by recurrent infections and hypogammaglobulinemia. Additional clinical manifestations vary, but can include autoimmunity, splenomegaly and nonmalignant lymphoproliferation [9]. CVID is a heterogeneous group of disorders; the age of onset can be in childhood, adolescence, or even adult life. Most patients have no molecular diagnosis as yet; cases can be sporadic, or familial. There is a high incidence of hematological malignancies in CVID [12]. For a definite diagnosis, all of the following criteria need to be present: (1) onset of immunodeficiency at greater than 2-4 years of age, (2) absent isohemagglutinins and/or poor response to vaccines and (3) exclusion of defined causes of hypogammaglobulinemia. Criteria for the diagnosis of probable CVID are a marked decrease of IgG (at least 2 SD below the mean for age) and a marked decrease in at least one of the isotypes IgM or IgA and for possible CVID a marked decrease (at least 2 SD below the mean for age) in one of the major isotypes (IgM, IgG, and IgA) [11]. Our patient, however, only suffers from IgA deficiency and shows reasonable responses to vaccines. The girl, therefore, only partly fits ALPS as well as CVID definitions (Table 3) and the diagnosis remains obscure. Even extensive genetic analysis did not help in this case. The ALPSgenes FAS, FASLG and CASP10 revealed no abnormalities; the CVID-associated TACI-gene showed one homozygous polymorphism (Pro251Leu). This homozygous polymorphism has been identified in 6 out of 912 healthy controls from Sweden, similar to what is seen in CVID or IgAdeficient patients (Personal communication U. Salzer, Centre of Chronic Immunodeficiency, University Medical Centre, Freiburg, Germany). Moreover, recent work showed that the terminal intracellular part of TACI including Pro251Leu is dispensable for TACI signaling, which also supports the view that the Pro251Leu variation is not harmful [13]. Now and then, pediatric hematologists and immunologists are confronted with a patient like this without a clear diagnosis where manifestations overlap between these two adjacent fields and the clinical course is different from what is usually seen. As also shown by this girl, not all cases of autoimmune thrombocytopenia in children are self-limiting, nor all persistent profound lymphadenopathies malignant. These "hematological" features can be the first manifestation of a primary immunodeficiency syndrome. Unfortunately, this girl also illustrates that despite recent successes in further classifying primary immunodeficiencies, there are still children with clinical and laboratory features which linger between several diagnostic entities. It is important to share the medical history of these children with other specialists in the field, so that combination of experience may enhance further classification of these diseases in the future.