|
|
| |||||||||||||||||||||||||||||||
Background and Methods We describe four patients without major risk factors for human immunodeficiency virus (HIV) infection, each of whom presented with severe opportunistic infections and was found to have idiopathic CD4+ T-lymphocytopenia. We performed assays to detect the presence of retroviruses and undertook immunophenotyping of subgroups of peripheral-blood lymphocytes.
Results The opportunistic infections at presentation included Pneumocystis carinii pneumonia, cryptococcal meningitis (two patients, one with concurrent pulmonary tuberculosis), and histoplasma-induced brain abscess. During 10 to 68 months of observation, none of the four patients had evidence of infection with HIV type 1 or 2 or human T-cell lymphotropic virus type I or II on the basis of epidemiologic, serologic, or polymerase-chain-reaction studies or culture, nor was there any detectable reverse transcriptase activity. Although all the patients had severe, persistent CD4+ T-lymphocytopenia (range, 12 to 293 cells per cubic millimeter), the CD4+ cell count progressively declined in only one and was accompanied by multiple opportunistic infections. All four patients had significantly reduced numbers of circulating CD8+ T cells, natural killer cells, or B cells (or all three).
Conclusions These four patients had idiopathic CD4+ T-lymphocytopenia with opportunistic infections but no evidence of HIV infection. Instead of the progressive, selective depletion of CD4+ T cells characteristic of HIV infection, some patients with idiopathic immunodeficiency have stable CD4+ cell counts accompanied by reductions in the levels of several other lymphocyte subgroups.
We describe four patients who presented with opportunistic infections and idiopathic CD4+ T-lymphocytopenia, none of whom had risk factors for HIV infection. Patients 1, 2, 3, and 4 are also described in the accompanying article by Smith et al.3 as Patients 3, 18, 17, and 4, respectively.
Methods
The patients were independently referred to our facilities for evaluation of opportunistic infections and unexplained immunodeficiency. Laboratory studies were performed according to established methods in licensed laboratories at several institutions.
Lymphocyte immunophenotyping was performed by flow cytometry at several centers, with the use of monoclonal-antibody panels supplied by Becton Dickinson (San Jose, Calif.), Olympus Immunochemicals (Lake Success, N.Y.), or Coulter Immunology (Hialeah, Fla.), in accordance with recommended techniques4. Enzyme immunoassays and Western blotting directed against retroviral antigens or antibodies were performed according to established methods5.
Retroviral cultures were performed by Dr. Brigitte P. Griffith (Retrovirus Diagnostic Section, Viral Reference Laboratory, West Haven Veterans Affairs Medical Center, West Haven, Conn.), Dr. Richard Marlink (Department of Cancer Biology, Harvard School of Public Health, Boston), and Dr. Martin S. Hirsch (Infectious Disease Unit, Massachusetts General Hospital, Boston). In brief, peripheral-blood mononuclear cells (PBMCs) were cultured with normal donor PBMCs that had been stimulated for 24 to 72 hours with phytohemagglutinin. The cultures were then assayed twice a week for p24 antigen, gp120, or Abbott HIV-1 whole viral lysate (Abbott Laboratories, North Chicago, Ill.) and observed for cytopathic effects during the four weeks of culture6. An assay for magnesium-dependent reverse transcriptase activity in culture supernatants was performed during the third week of culture7.
The polymerase chain reaction (PCR) was used to test for HIV-1 in Patient 2 and was performed by Dr. J. Brooks Jackson (Department of Pathology, University Hospitals of Cleveland) with the gag primer SK38/39 and a radiolabeled SK19 probe, as previously described8. The PCR used to test for HIV-1 in Patients 1 and 4 was performed by Dr. C.-Y. Ou (CDC, Atlanta) with gag primers SK38/39 and SK145/150 and the probes gag-1, gag-2, gag-3, and gag-4 (Gen-Probe, San Diego, Calif.)9.
Lymphocyte proliferation was performed as described,10 with the use of PBMCs isolated by Ficoll-diatrizoate sodium (Hypaque) sedimentation. The PBMCs were incubated for four days with phytohemagglutinin and for seven days with tetanus toxoid. The results are expressed as a percentage of the proliferation in three simultaneously assayed samples of PBMCs from age-matched control subjects.
Case Reports
Patient 1
In October 1990, Patient 1, a 65-year-old white heterosexual male smoker with alcoholic cardiomyopathy and multiple recurrent cutaneous basal-cell and squamous-cell carcinomas, was found to have a nodular infiltrate on a preoperative chest radiograph. At presentation he was asymptomatic, except for his usual dyspnea on exertion. An open-lung biopsy in January 1991 revealed an atypical, granulomatous form of Pneumocystis carinii pneumonia, accompanied by pneumocysts invading the lymphatic system and small arteries. Stains and cultures revealed no evidence of cancer, fungi, or acid-fast bacilli. The radiographic abnormalities resolved after 21 days of treatment with trimethoprim-sulfamethoxazole. The patient continued to take the medication for prophylaxis against P. carinii pneumonia. At the most recent follow-up visit, 20 months after diagnosis, he had recurrent oral thrush, malaise, and night sweats and had recently been confirmed to have bacteremia due to Mycobacterium avium complex and a positive buffy-coat blood culture for cytomegalovirus without evidence of retinitis.
Patient 2
Patient 2 was a 40-year-old heterosexual white man with a remote history of alcohol abuse who was seen in August 1989 with a two-month history of headaches, confusion, dizziness, and ataxia. Computed tomography (CT) of the head revealed obstructive hydrocephalus. A lumbar puncture yielded no white cells, a glucose level of 54 mg per deciliter (3.0 mmol per liter), and a protein concentration of 25 mg per deciliter. Staining with India ink revealed encapsulated yeasts, and Cryptococcus neoformans grew from cultures of both cerebrospinal fluid and blood. The cryptococcal antigen titer of cerebrospinal fluid was 1:8192. Treatment with amphotericin B and flucytosine was started, and a ventriculoperitoneal shunt was placed. The patient's response to therapy was slow, and he was sent home in December 1989 after having received a total dose of 3.5 g of amphotericin B. His medication was changed to fluconazole (200 mg orally daily) on discharge.
The patient relapsed five months later, with headache, a serum cryptococcal antigen titer of 1:16,384, and a cerebrospinal fluid titer of 1:256. He was treated with fluconazole (400 mg daily), and the ventricular drain was removed, with improvement.
In September 1990 the patient was found to have optic atrophy and a serum cryptococcal antigen titer of 1:128. He was readmitted and treated with a total dose of 1.5 g of amphotericin B. At the most recent follow-up visit, 37 months after his initial presentation, he had had no further relapses while taking prophylactic oral fluconazole, and except for an episode of severe colitis due to Clostridium difficile, he was well.
Patient 3
Patient 3 was a 68-year-old Asian woman with non-insulin-dependent diabetes mellitus who was treated for presumed bacterial pneumonia in October 1991 and was subsequently sent home without a substantial change in her chest x-ray film. Five days later, she was readmitted with persistent fever, cough, and shortness of breath. Her evaluation, however, suggested acute cholecystitis, prompting cholecystectomy, yet her febrile course continued and she began to have headaches. CT of the head demonstrated communicating hydrocephalus, and a lumbar puncture yielded 68 leukocytes per cubic millimeter (95 percent of which were lymphocytes), a protein concentration of 172 mg per deciliter, and a glucose level of 49 mg per deciliter (2.72 mmol per liter). The serum glucose level, measured simultaneously, was 172 mg per deciliter (9.55 mmol per liter). Staining with India ink showed encapsulated yeasts, and the cryptococcal antigen titer of cerebrospinal fluid exceeded 1:16,384. Cultures of blood, cerebrospinal fluid, and urine grew C. neoformans. Treatment with amphotericin B and flucytosine was started, but flucytosine was soon discontinued because of concurrent autoimmune hemolytic anemia. The patient received a total dose of 2 g of amphotericin B before treatment was changed to oral fluconazole; she continued to take fluconazole for maintenance prophylaxis. Her hospital course was further complicated by culture-proved pulmonary tuberculosis, cytomegalovirus viremia, and multidermatomal herpes zoster. Her symptoms and radiographic abnormalities resolved after treatment with isoniazid, rifampin, pyrazinamide, and acyclovir. At the most recent follow-up visit, 10 months after presentation, she was continuing to regain her strength, although she still had anemia.
Patient 4
In January 1987, Patient 4, a 31-year-old heterosexual white man with asthma, was seen because of frequent upper respiratory tract infections and intermittent hemoptysis. An evaluation revealed splenomegaly, lymphocytopenia (800 lymphocytes per cubic millimeter), and numerous calcified granulomas in the spleen, hilum, and mediastinum. A histoplasmin skin test was positive (resulting in a 12-mm area of induration), and a tuberculin test with intermediate-strength purified protein derivative was negative. This constellation of signs and symptoms was thought to be consistent with histoplasmosis with mediastinal granuloma.
In February 1991, the patient presented with a severe headache, transient diplopia, and fever. A lumbar puncture revealed 154 leukocytes per cubic millimeter (63 percent of which were lymphocytes), a glucose level of 52 mg per deciliter (2.89 mmol per liter), and a protein concentration of 62 mg per deciliter. Tests for cryptococcal antigen were negative, as were the results of staining with India ink and cerebrospinal fluid cultures; however, the pleocytosis persisted. Serologic analyses were negative for multiple infectious agents, except for a serum histoplasma complement-fixation test (performed at the CDC), which was positive at a titer of 1:8.
The patient's headaches continued, and transient dysphasia developed. In April 1991 magnetic resonance imaging revealed a lesion in the basal ganglia with central clearing, consistent with the formation of an abscess. On the basis of the patient's clinical course and the positive serologic results, a presumptive diagnosis of central nervous system histoplasmosis was made, and he was treated with amphotericin B. The headaches and cerebrospinal fluid abnormalities resolved, with a reduction in the size of the lesion on subsequent magnetic resonance imaging. After nine weeks of amphotericin B, treatment was changed to itraconazole, which he now takes for long-term maintenance. Nineteen months after the episode of meningitis and 68 months after his initial evaluation, the patient was active and well.
Results
Epidemiologic Characteristics
None of the patients reported traditional risk factors for HIV infection, including multiple or same-sex sexual partners, injection-drug use, or transfusion, although Patient 1 reported contact with prostitutes while in China in 1945 and in New York City between 1972 and 1975. There were no apparent epidemiologic links between patients. Patients 1, 2, and 3 had each lived in the Far East for at least one year and had been sexually active while there: Patient 1 had been stationed in China and the South Pacific during World War II; Patient 2 was a Vietnam veteran who had been exposed to Agent Orange; Patient 3 was born in southern China and had lived in Burma and Macao before moving to the United States in 1979. Autoimmune hemolytic anemia developed in Patient 3 during her acute illness, but there was no history of immunosuppressive, autoimmune, or unusual infectious disorders in any of the other patients or their family members. All family members and sexual partners were reportedly healthy.
Evaluation for Retroviral Infection
Evaluations for HIV-1 infection were negative in all patients (Table 1), including multiple enzyme immunoassays, Western blotting, and p24 antigen tests. The initial Western blot in Patient 3 was indeterminate (the gp120/160 band was positive; all other bands were negative, including the gp41 band), but the results of a second study were entirely negative. All four patients had negative retroviral cultures with reverse transcriptase assays, negative tests for antibodies to human T-cell lymphotropic virus type I or type II, and negative enzyme immunoassays for HIV-2. The results of the PCR for HIV-1 were negative in Patients 1, 2, and 4; it was not performed in Patient 3. The sexual partner and the adult daughter of Patient 1 and the spouse of Patient 4 all had negative enzyme immunoassays for HIV-1, HIV-2, and p24 antigen, as well as normal CD4+ cell counts.
|
The leukocyte counts were within normal limits or slightly low in all four patients, but all patients had lymphopenia (Table 2). The CD4+ T-lymphocyte counts were consistently depressed in all patients, ranging from 12 to 293 cells per cubic millimeter on multiple determinations (representing 5 to 37 percent of total lymphocytes) (Table 2), when measured with monoclonal antibodies reacting with different epitopes of the CD4 molecule. Similarly, the ratios of CD4+ cells to CD8+ cells were uniformly less than 1. The absolute numbers of CD8+ T lymphocytes were more variable but were usually below the normal 95 percent reference ranges reported for healthy adults11,12,13,14 in all but Patient 2. Figure 1 depicts the CD4+ and CD8+ cell counts during 10 to 68 months of observation.
|
|
All but Patient 4 had anergy to multiple recall antigens during the acute illness. In vitro lymphoproliferation was depressed in both patients in whom it was examined. As compared with control subjects, the responses to mitogen (phytohemagglutinin) and to recall antigen (tetanus toxoid) were 10 percent and 0 percent, respectively, in Patient 2 and 44 percent and 28 percent in Patient 4. Immunoglobulin levels were normal in three patients; the total levels were also normal in Patient 4, but the IgG2 level was somewhat depressed at 53 mg per deciliter (normal range, 117 to 747 mg per deciliter) and the level of IgG3 was 21 mg per deciliter (normal range, 41 to 129 mg per deciliter). There has been no evidence of a malignant condition. Restriction-fragment-length analysis of DNA extracted from a skin-biopsy sample from Patient 1 showed no evidence of gene rearrangements or monoclonality in cutaneous B or T cells.
Discussion
These four patients have CD4+ T-lymphocytopenic immunodeficiency of unknown cause. Several lines of evidence strongly suggest that none are infected with HIV-1 or HIV-2. First, in contrast to the patients in the earliest reports of the acquired immunodeficiency syndrome (AIDS),15,16,17 none of our patients reported any of the major risk factors traditionally associated with AIDS, except Patient 1, whose last contact with prostitutes occurred several years before AIDS was first reported. Second, extensive searches for evidence of HIV-1 and HIV-2 infection were all negative. Third, in addition to CD4+ T-lymphocytopenia, all our patients had depletion of B cells, CD8+ T cells, or natural killer cells, or of all three types of cells, and three patients did not have the progressive depletion of CD4+ T lymphocytes that is characteristic of advanced HIV infection. Finally, long-term follow-up of three of these patients suggests that their course may be more benign than that associated with AIDS.
Cases of unexplained CD4+ T-lymphocytopenia are undoubtedly more common than previously appreciated, remaining undetected because CD4+ cell counts have rarely been determined in patients without HIV infection and because methods of measuring these cells have only been generally available in the past five to six years. The CDC recently described a preliminary investigation of 14 patients with idiopathic CD4+ T-lymphocytopenia (including our Patients 1, 3, and 4)1,2. Twenty-one other patients from seven countries have been described since 19831,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34. Nine of these patients either did not meet the CDC criteria for the syndrome because they had other disorders or were receiving therapy that could cause immunosuppression or did not have any AIDS-defining illnesses1,28,29,30,31,32,33,34. Several patients reported high-risk behavior for HIV infection and had illnesses entirely compatible with HIV infection, and may thus represent a group of patients who have undetected infection with HIV-1, HIV-2, or as yet undiscovered retroviruses.
Suppression of cell-mediated immunity by infectious agents other than HIV has long been recognized,35,36 yet the full role of lymphocyte subgroups in both conventional and opportunistic infections is incompletely described. A variety of acute and chronic infections may be associated with the depletion of CD4+ T cells, which is usually transient and accompanied by CD8+ T-lymphocytosis11,37,38. Depressed cell-mediated immune function can occur as a sequela of chronic fungal infection39,40,41. Payan et al. described six patients with chronic complications of histoplasmosis who had low CD4+ cell counts (mean, 179 cells per cubic millimeter) and low ratios of CD4+ cells to CD8+ cells (mean, 0.40)42. CD4+ T-lymphocytopenia has also been associated with ultraviolet radiation,43,44 the formation of antilymphocyte autoantibodies,45 and the use of certain medications, including corticosteroids and chemotherapeutic agents46.
The CD4+ T-lymphocyte counts in all our patients remained persistently depressed during up to 68 months of observation. However, although the CD4+ cell counts declined in Patient 1, they increased in Patient 3, from 20 to as high as 224 per cubic millimeter during 10 months of recuperation (Figure 1). Moreover, all the patients had quantitative deficiencies of other lymphocyte subgroups, including total T cells and natural killer cells (four patients each) and B cells and CD8+ T cells (three patients each) (Table 2). Thus, in contrast to advanced HIV infection, which is characterized by progressive depletion of CD4+ T cells,47 some patients with idiopathic depletion of CD4+ T cells have profound disturbances of other lymphocyte subgroups, suggesting disparate immunopathogenic mechanisms for these syndromes. Such patients may represent some of the cases of unexplained opportunistic infection that have occurred at a background level over several decades36 and are only now, with the assistance of lymphocyte immunophenotyping, coming to light.
The optimal treatment of patients with idiopathic CD4+ T-lymphocytopenia remains to be defined. CD4+ cell counts provide an effective surrogate marker for clinical disease progression in HIV-infected patients48,49. Prophylaxis against pneumocystosis is recommended for those with CD4+ cell counts below 200 per cubic millimeter,50 and prophylactic regimens with activity against other AIDS-associated opportunistic pathogens are being studied. Moreover, when certain infections (e.g., cryptococcosis) occur in the setting of AIDS, lifelong suppressive therapy is indicated51. It is unclear whether similar principles apply to patients with idiopathic CD4+ T-lymphocytopenia. The occurrence of multiple opportunistic infections and relapses in some of our patients suggests that an aggressive approach to diagnosis, treatment, and prophylaxis is needed. It is clear that idiopathic CD4+ T-lymphocytopenia should be included in the differential diagnosis of unexplained opportunistic infection.
Supported in part by a grant (AI-25780) from the National Institutes of Health.
We are indebted to Dr. Douglas Golenbock and Dr. Richard Moscicki for performing additional flow-cytometric studies, to Dr. Thomas Spira for valuable comments, to Dr. Donald Craven for reviewing the manuscript, and to Maria Tetzaguic and Julia Gunn, R.N., M.P.H., for their assistance in the preparation of the manuscript.
Source Information
From the Evans Memorial Department of Clinical Research and the Department of Medicine, Thorndike Memorial Laboratory and the Maxwell Finland Laboratory for Infectious Diseases, Boston City Hospital and University Hospital, Boston University School of Medicine, Boston (R.A.D., G.M.A., A.M.S., S.M.L.); the Infectious Disease Section, Department of Medicine, Dartmouth-Hitchcock Medical Center, Lebanon, N.H. (C.F.v.R.); and Cleveland Veterans Affairs Medical Center and Case Western Reserve University, Cleveland (Z.T.). Presented at the 32nd Interscience Conference on Antimicrobial Agents and Chemotherapy, Anaheim, Calif., October 11-14, 1992.
Address reprint requests to Dr. Duncan at Thorndike Bldg. 311, Boston City Hospital, 818 Harrison Ave., Boston, MA 02118.
References
CD2-, 3+, 4-, 8- lymphocytes in histoplasmosis-associated immunodeficiency. Int Arch Allergy Appl Immunol 1989;90:213-218. [Medline]
| |||||||||||||||||||||||||||||||
This article has been cited by other articles:
HOME | SUBSCRIBE | SEARCH | CURRENT ISSUE | PAST ISSUES | COLLECTIONS | PRIVACY | TERMS OF USE | HELP | beta.nejm.org Comments and questions? Please contact us. The New England Journal of Medicine is owned, published, and copyrighted © 2009 Massachusetts Medical Society. All rights reserved. |
Previous
