Idiopathic CD4 lymphocytopenia: a case of missing, wandering or ineffective T cells

Idiopathic CD4 lymphocytopenia (ICL) is a presumed heterogenous syndrome with key element low CD4 T-cell counts (below 300/mm3) without evidence of HIV infection or other known immunodeficiency. The etiology, pathogenesis, and management of ICL remain poorly understood and inadequately defined. The clinical presentation can range from serious opportunistic infections to incidentally diagnosed asymptomatic individuals. Cryptococcal and non-tuberculous mycobacterial infections and progressive multifocal leukoencephalopathy are the most significant presenting infections, although the spectrum of opportunistic diseases can be similar to that in patients with lymphopenia and HIV infection. Malignancy is common and related to opportunistic pathogens with an oncogenic potential. Autoimmune diseases are also seen in ICL with an increased incidence. The etiology of ICL is unknown. Mechanisms implicated in CD4 reduction may include decreased production, increased destruction, and tissue sequestration. New distinct genetic defects have been identified in certain patients with ICL, supporting the hypothesis of the lack of a common etiology in this syndrome. The management of ICL is focused on the treatment of opportunistic infections, appropriate prophylactic antibiotics, and close monitoring. In selected patients with life-threatening infections or profound immunodeficiency, strategies to increase T-cell counts or enhance immune function could be considered and have included interleukin-2, interferon-gamma, interleukin-7, and hematopoietic stem cell transplantation. The prognosis is influenced by the accompanying opportunistic infections and may be affected by publication bias of severe cases with unfavorable outcomes. As newer laboratory investigation techniques are being developed and targeted experimental treatments become available, our comprehension and prognosis of this rare syndrome could be significantly improved.


Defi nition
ICL is defi ned by a documented absolute CD4 T lymphocyte count of less than 300 cells per cubic millimeter or of less than 20% of total T cells on more than one occasion, usually 2 to 3 months apart, without evidence of HIV infec tion or any defi ned immunodefi ciency or therapy associated with depressed levels of CD4 T cells.

Pathogenesis
ICL is a condition of unknown etiology and is considered a heterogenous syndrome possibly encompassing diff erent disorders sharing the common feature of reduced circulating CD4 T-cell counts. In that respect, it is no surprise that no uniform theory for the pathogenesis of ICL has been formalized. In this review, we attempt to summarize the most important reports of possible pathogenetic mechanisms and immuno logic abnormalities in ICL.
Th e failure of CD4 T-cell homeostasis in ICL could be attributed to decreased production, increased destruction, tissue sequestration, or any combination of these. Evidence supporting each of these explanations has been reported, although in some studies it is not possible to decipher whether the fi ndings relate to the cause of ICL or are consequences of the lymphopenia itself. Finally, three recent distinct genetic abnormalities have been identifi ed, although familial ICL cases were not seen in our cohort and have otherwise been very rare [1,2].
In terms of decreased CD4 T-cell production, decreased bone marrow clonogenic capability with diminished stem cell precursors contributing to CD4 depletion has been described in three relevant studies [3]. Reduction of p56 (Lck) kinase activity in ICL patients' T cells compared with healthy control donors was observed in another study, emphasizing the critical role of this kinase in the maintenance of the peripheral CD4 T-cell subset [4]. Disturbed thymic T-cell maturation was fi nally implicated in ICL pathogenesis in another study [5].
In terms of increased CD4 T-cell destruction, one of the initial reports dealing with ICL pathogenesis demonstrated enhanced apoptotic depletion of CD4 T cells [6], possibly associated with overexpression of Fas and Fas ligand [7]. In our large cohort of ICL cases, we described increased activation and cycling of CD4 T cells, which were inversely corre lated with CD4 T-cell numbers, suggesting a compen satory response to lymphopenia [8]. In another report, we demonstrated that CD4 cycling was strongly asso ciated with levels of plasma lipopolysaccharide [9], implying a possible association of microbial products with lymphopenia-induced proliferation. High serum interleukin 7 (IL-7) levels as well as an inverse correlation of CD4 count and IL-7 levels have been described in ICL [3,10,11]. Th e high IL-7 levels possibly represent cytokine accumulation [12] due to decreased IL-7 receptor alpha chain expression on T cells (CD127) [8,9]. Th is is supported by the fact that IL-7 levels do not correlate with T-cell cycling and inversely correlate with in vitro CD4 T-cell responses to IL-7 as measured by STAT-5 phosphorylation [13]. Specifi cally, lower levels of STAT-5 phosphorylation after IL-7 stimulation were observed in both CD4 and CD8 T cells from ICL patients compared with controls. In addition, destabilization of p27 kip1 , a critical step for IL-7-induced T-cell cycling, was decreased in ICL patients compared with controls after in vitro IL-7 stimulation. Th ese data suggest that high serum IL-7 levels do not necessarily represent a com pen satory response and that the blunted T-cell response to IL-7 may be further accentuating T-cell apoptosis and lymphopenia. Autoantibodies to CD4 T cells have also been reported in ICL [14] but with no clear insight into the implications of this fi nding.
Th e low CD4 T-cell numbers and higher proportion of memory-activated CD4 T cells may lead to restricted pathogen recognition in ICL. Alpha/beta and gamma/ delta T-cell repertoires of three patients with ICL were reportedly highly restricted, giving rise to mono-/ oligoclonal T-cell expansions [15]. In addition, defective production of interferon-gamma (IFN-γ) and tumor necrosis factor (TNF) was seen in two patients with ICL and cryptococcal meningitis [16], showing that functional T-cell defects can be associated with presenting infections.
T-cell traffi cking and tissue distribution had not been previously studied in ICL. In an immunodefi ciency syndrome characterized by warts, hypogammaglobulinemia, infections, and myelokathexis (WHIM), mutations of the chemokine receptor CXCR4 leading to neutropenia and immunodefi ciency were identifi ed [17]. In a recent study that focused on chemokine receptor CXCR4 [18], patients with ICL had very low to undetectable levels of surface CXCR4 expression with intracellular accumulation of CXCR4 and its ligand, CXCL12, in CD4 T cells in comparison with healthy controls. Impaired CXCR4 expression was seen exclusively in T cells, both naïve and memory subsets, and was restored after overnight rest. In chemotaxis assays, it was shown that T cells from patients with ICL had impaired chemotactic responses to CXCL12 and normal responses to CXCL8. Th ere was also a slower reemergence of CXCR4 after ligand binding and internalization. In vivo IL-2 administration seemed to restore CXCR4 expression and responses to CXCL12 in three out of four patients treated. Th is observation has not yet been reproduced by other groups and should be further explored. To date, our group has not found ICL patients with absent CXCR4 expression on T cells (unpublished observations).
A genetic etiology in patients with ICL was identifi ed in three recent studies. Th e fi rst study [19] described two novel missense mutations in RAG1 gene related to a mild clinical ICL phenotype in a girl. RAG1 or RAG2 mutations have been responsible for the majority of patients with severe combined immunodefi ciency of the T-B-SCID phenotype as well as for the Omenn syndrome [19]. It is consequently concluded that RAG mutations can be responsible for clinical presentations spanning from severe immunodefi ciencies to an almost normal phenotype, as seen in this patient with ICL.
Th e second study [1] implicated mutations in the magnesium transporter gene, MAGT1, to an X-linked primary human immunodefi ciency with impaired thymic production of CD4 T cells and circulating T cells in two brothers with ICL. Th is novel syndrome, XMEN (X-linked immunodefi ciency with magnesium defect, Epstein virus infection and neoplasia), revealed a role for Mg + as an intracellular second messenger.
Th e third and most recent study [20] described a heterozygous dominant negative missense mutation of the signaling adaptor protein Uncoordinated 119 (Unc119) in a patient with ICL. Th e mutation disrupts the Unc119-lymphocyte specifi c kinase (Lck) interaction that is necessary for the T-cell response to T-cell receptor (TCR) stimulation. In fact, there is impairment in both localization and enzymatic activation of Lck. Th e patient was a 32-year-old woman with ICL and recurrent otitis media, shingles, widespread fungal nail infection, oral herpetic lesions, and bronchiolitis obliterans organizing pneumonia.

Clinical features
Th e diagnosis of ICL is typically suspected when an opportunistic infection is identifi ed in an otherwise healthy individual. Th e literature is dominated by such cases and is constantly expanding. Th e majority of cases report opportunistic infections -most notably cryptococcal or mycobacterial disease and progressive multifocal leukoencephalopathy (PML) -that are normally seen in HIV-infected patients. Th e detailed listing of all reported infections exceeds the scope and purpose of this review and has been summarized eff ectively elsewhere [21,22]. It is possible that the apparent presenting clinical spectrum of ICL is infl uenced by publication bias of more serious infections and adverse outcomes. In that respect, our longitudinal study of 39 patients with ICL [8] off ers some insight into the incidence of infections and other complications (namely autoimmune disease and malignancy) during a median follow-up period of 49 months. In this study, the three most common infections at presentation were crypto coccal, genital human papilloma virus (HPV), and non-tuberculous mycobacterial infections. As described extensively in the available literature [23], cryptococcal infections were mainly, but not exclusively, pulmonary or central nervous system disease. Human papilloma virus infections were, notably, treatment-resistant or locally spread with or without cervical intraepithelial neoplasia. Most non-tuberculous mycobacterial infections were due to Mycobacterium avium complex (MAC). Other signifi cant infections at presentation or during follow-up were Pneumocystis jiroveci pneumonia (PCP), PML, and Epstein-Barr virus (EBV)-related lymphoproliferative disease leading to Bcell lymphoma. Most importantly, though, there was a minority of ICL patients who were asymptomatic and remained as such for a long period of time with no infections or other disease. It is reasonable to hypothesize, as a result, that there is a subpopulation within ICL with a more favorable outcome that may be underreported in the literature. Th is has been suggested in the past [24], although asymptomatic CD4 lympho cyto penia may be extremely rare or transient [25]. Other notable infections described are varicella zoster virus infection, histoplasmosis, Candida, toxoplasmosis, asper gillosis, cytomegalovirus, and Leishmania. Recently, a number of cases of ICL-related PML have been reported, signifying the potential gravity of the ICL diagnosis given the ominous prognosis of this infection and the absence of validated anti-viral therapy targeting John Cunning ham (JC) virus [26]. A special consideration is warranted for the cases of ICL and tuberculosis since untreated tuberculosis can lower CD4 T-cell counts and eff ective tuberculosis treatment usually leads to restoration of CD4 lymphocytopenia [27].
Malignancy is also common in ICL, mainly lymphoma, and specifi cally EBV-related B-cell [8], primary eff usion, and central nervous system lymphoma. Other possible malignancies include Kaposi's sarcoma and cervical or perineal neoplasias in the setting of long-term HPV infections [8]. In summary, patients with ICL and HIVinfected individuals with comparable CD4 T-cell counts appear to have a similar pattern of susceptibility to opportunistic pathogens (manifesting as infections or neoplasms).

Diagnosis and diff erential diagnosis
Although establishing the presence of lymphocytopenia in an HIV-uninfected patient with an opportunistic infection is relatively straightforward, determining that the patient has ICL is not. Th e diff erential diagnosis of ICL is large (Table 1) [22] and this diagnosis remains one of exclusion and requires an extensive immunologic, hematologic, rheumatologic, and infectious disease workup as well as follow-up testing to confi rm the persistence of lymphocytopenia. As stated in the syndrome defi nition, other forms of immunodefi ciency have to be excluded and the provisional ICL diagnosis has to be laboratory-confi rmed at least twice during a period of 1 to 3 months. A comprehensive immunologic work-up should be applied in order to exclude HIV infection, lymphoma, autoimmune diseases, other forms of immuno defi ciency (such as common variable immunodefi ciency), and sarcoidosis ( Table 1). As mentioned above with tuberculosis, special attention should also be paid to the possibility that the presenting infection or treatment may be the cause of lymphocytopenia.
Th e relationship between ICL and autoimmune disease is a complicated one. Although lymphocytopenia is a known consequence of some autoimmune diseases such as systemic lupus erythematosus and Sjögren's syndrome [8,28], patients with ICL appear to be at increased risk for autoimmune diseases possibly due to a break of tolerance (lack of self-recognition) that can occur in the setting of lymphopenia-induced T-cell proliferation lymphopenia [29,30]. Autoimmune diagnoses described among patients with ICL include antiphospholipid antibody syndrome, psoriasis, autoimmune hemolytic anemia, ulcerative colitis, Grave's disease, vitiligo, autoimmune thyroiditis, Behçet's disease, and vasculitis. Th e diagnosis of ICL, when autoimmune disease is present, can therefore be challenging since lymphocytopenia might be a feature of autoimmunity itself or ICL may be the underlying cause of autoimmunity. Complicating the assessment further may be the possibility of transient lymphocytopenia related to immunosuppressive therapies.
Lymphocytopenia in the setting of Sjögren's syndrome needs to be addressed separately. Soon after ICL was fi rst described, it was noted that patients with Sjögren's syndrome have an increased incidence of low CD4 T-cell numbers which fulfi lled ICL criteria but also correlated with anti-SSA antibody positivity [31] and predisposed these patients to lymphoma [28,32]. In a key study designed to identify predictors of lymphoma in a primary Sjögren's syndrome population, CD4 lymphocytopenia and a low CD4/CD8 ratio were strongly associated with non-Hodgkin's lymphoma [33].
Finally, another signifi cant association that may be overlooked is between lymphocytopenia and sarcoi dosis. Th ere have been reported cases of the concomitant existence of the two conditions in the past [34], and there is also an increased incidence of opportunistic infections in sarcoidosis (cryptococcosis in particular), especially in corticosteroid-treated patients. Recently, it has been shown that the low CD4 T-cell count in certain patients with sarcoidosis may relate to disease severity [35] and may represent a distinct phenotype with a favorable response to anti-TNF-α treatment [36]. Since it is unclear whether sarcoidosis may coexist with ICL or lymphopenia may be merely a manifestation of severe sarcoidosis, the diagnosis of ICL should be questioned when sarcoidosis is confi rmed. Our position is that CD4 lymphopenia in the presence of an established diagnosis (sarcoidosis or Sjögren's syndrome, for example) could be viewed as an extreme phenotype of the underlying condition and not necessarily as coexistence of ICL with another condition.
Th ere is no consensus for diagnostic investigation of possible ICL cases. Th e diagnostic work-up focused on excluding alternative diagnoses should be dictated by the presenting symptoms, and special attention should be given to signs of rheumatologic, infectious, immunologic, or hematologic abnormality. Th e investi ga tion should typically consist of lymphocyte fl ow cyto metry, including CD4, CD8, natural killer, and B-cell subsets as well as immunoglobulin levels, HIV 1-2 serology and PCR, HTLV 1 and 2 testing, EBV and cyto megalovirus PCRs, and an autoantibody panel. Th e presenting opportunistic infection or suggestive signs or symptoms should dictate the infectious disease work-up. Appropriate imaging by computed tomography or positron emission tomography and bone marrow biopsy should be performed if lymphoma is suspected. If clinical and laboratory features of WHIM are observed, T cells should be assessed for CXCR4 expression. Genetic testing should be performed when familial clustering is observed or when the clinical features suggest any of the described genetic syndromes.
Functional T-cell assays looking at cytokine production or proliferation in response to antigen or TCR repertoire analysis are of interest to better understand pathogenesis but have not been clearly linked to clinical prognosis or diagnosis.

Treatment
Th e treatment of ICL is infl uenced by the initial clinical presentation. Monitoring of asymptomatic ICL individuals should be regular during the immediate followup period after diagnosis since most opportunistic infections have been described in the fi rst months of follow-up. Although there is no consensus about how often these patients should be followed, a 3-to 6-month interval may suffi ce in most cases. After the fi rst few years of follow-up, less frequent evaluation could be appropriate. Special attention should be paid in certain populations with a possible adverse outcome. Low CD8 T-cell count (below 180/mm 3 ) has been associated with opportunistic infection-related death [23]. Th erefore, this subset of ICL individuals might benefi t from evaluations that are more frequent. Screening for lymphoma or HPVrelated cervical or anal dysplasia or a combination of these may also be appropriate.
Although it is clear that opportunistic infections in patients with ICL should be treated vigorously with continuing secondary prophylaxis, there is insuffi cient evidence to guide primary infection prophylaxis. In a median follow-up of 49.5 months in our cohort, infection was diagnosed in 15 (41.6%) of the patients and severe opportunistic infections were diagnosed in four (13.8%) of these 15 [8]. Interestingly, only one episode of PCP and two of MAC were diagnosed during a cumulative 164 patient-years of follow-up. Our natural history study precluded any fi rm conclusions about who should be off ered prophylaxis but does seem to demonstrate that the risk of MAC and PCP infections is much greater in untreated HIV-infected individuals with AIDS. Although HIV infection and ICL diff er substantially in etiology, progression, and prognosis, the literature on prophylaxis of opportunistic diseases in HIV infection provides the best available data with respect to opportunistic infection prophylaxis in ICL. Th us, patients with ICL may reasonably be off ered infection prophylaxis that is based on established guidelines for patients with HIV/AIDS. In selected patients with a history of life-threatening infections or profound immunodefi ciency or persistent opportunistic infections, strategies to increase the circulating CD4 lymphocytes could be considered. It is possible that there is publication bias toward reports of successful outcomes of immune-based therapies in this setting. Th erapeutic options used to increase CD4 cells or improve immune function (or both) in ICL include the use of IL-2, IFN-γ therapy, IL-7, and hematopoietic stem cell transplantation. Th erapy with IL-2 was fi rst used in the form of subcutaneous polyethylene glycol-IL-2 for 5.5 years in a woman with chronic severe mycobacterial disease; she had signifi cant clinical improvement but a moderate and slow CD4 count increase from approximately 29/mm 3 before treatment to 100/mm 3 after 1 year of therapy [37]. A man with ICL and relapsing generalized herpes zoster infection was treated with IL-2 every other week for 5 months and had clinical improve ment and a CD4 count increase from 24 to 93/mm 3 . Th e patient developed gastric anaplastic large-cell mucosa-associated lymphoid tissue lymphoma 1 year after IL-2 treatment and subsequently died [22,38]. A 33-year-old woman who had ICL but who also had an IgA and IgG monoclonal gammopathy and renal disease received IL-2 and had a regression of gammopathy and a signifi cant CD4 increase from 40/mm 3 to constantly above 300/mm 3 after treatment [39]. A 16-year-old boy with ICL and recurrent cryptococcal meningitis was treated with recom binant IL-2 twice weekly and showed clinical improvement and a CD4 count increase from 202 to 262 and 323/mm 3 after 5 and 9 months of therapy, respectively [40]. All of the above cases are summarized in a report of a 39-year-old man who had a Myco bacterium avium intracellulare infection that was treated with IL-2 for 35 months and who was apparently cured while his CD4 count increased from 365 to 470/mm 3 at the end of therapy [41]. A 64-year-old man with recurrent disseminated Mycobacterium avium intracellu lare infection was also treated with IL-2 in three cycles of 5 consecutive days each every other month and had a favorable clinical outcome and CD4 number increases after each IL-2 cycle [42]. Finally, a recent report des cribed a series of four ICL patients who received several courses of recombinant IL-2 for up to 3 years [18]. Th ree patients responded with either sustained or transient CD4 count increases whereas one did not. Two ICL patients with disseminated MAC were treated with IL-2 at the National Institutes of Health but had no clinical improvement and both subsequently died (Michael Sneller, personal communication).
IFN-γ therapy has been off ered in selected ICL cases. Th ree patients with ICL were included in a study of IFN-γ in refractory disseminated non-tuberculous mycobacteria infection, and two of them had clinical improvement and CD4 count increases [43]. Two more patients with a potential benefi t from IFN-γ treatment have been documented in the literature [16,42], whereas others had no clear benefi t [37,38].
It has been shown that, in patients with idiopathic CD4 lymphopenia, an impaired IL-7 response to CD4 depletion may contribute to the impaired lymphocyte homeostasis [11]. As such, an increased availability of IL-7 could play a critical role in restoring T-cell homeostasis by enhancing eff ective proliferation and by improving T-cell survival. IL-7 has been reported to lead to increases in both CD4 and CD8 T cells in other lymphopenic settings at well-tolerated doses [44][45][46]. IL-7 immunotherapy is currently being investigated in an open-label, single-arm, phase I/IIa clinical trial of glycosylated recombinant human IL-7 (rhIL-7) in ICL patients at risk of disease progression [47]. Successful control of PML has been reported in a patient with ICL by using a combination of CMX001 (investigational oral drug for double-stranded DNA viral diseases, including JC) and IL-7 that led to stabilization of neurologic symptoms and CD4 increases [48].
Finally, it is worth mentioning two key studies of a complete restoration of CD4 counts to normal levels and apparent cure of ICL in individuals treated with nonmyeloablative allogeneic bone marrow transplantation for aplastic anemia or ICL or both [49,50]. Although these results are promising, this therapeutic option clearly should be reserved for a carefully selected population with severe disease in an appropriate clinical setting.

Prognosis
Th e prognosis of ICL is critically infl uenced by the opportunistic infections that may complicate the natural history of the syndrome. Low CD8 numbers (<180/mm 3 ) and the degree of CD4 T-cell activa tion as measured by HLA-DR expression have been associated with an adverse outcome [8], namely oppor tunistic infection-related death. Lymphocytopenia is typically not progressive, and in some instances CD4 numbers can return to normal levels in the follow-up period for reasons that are unclear. Finally, as chronic HPV infections may predispose patients to squamous cell carcinomas, appropriate cancer screening should be entertained.

Conclusions
Despite 20 years of investigation, ICL remains a syndrome of uncertain pathogenesis, prognosis, and treatment and can be a devastating diagnosis for some patients. Although a subset of patients with ICL may remain asymptomatic, others may go on to develop lifethreatening opportunistic infections or autoimmune conditions or severe HPV-related dysplasia and malignancy. Th e cause or causes of ICL remain unknown, although several distinct genetic mutations have been reported recently and appear to be responsible in some patients. Further study of the genetic basis and immunopathogenesis of this heterogenous condition is required so that targeted treatments for ICL patients at increased risk of signifi cant morbidity and mortality can be safely established.