• HIV infection;
  • Parotid enlargement;
  • Sjögren's syndrome;


  1. Top of page
  2. Abstract
  7. Acknowledgements


To describe the changing clinical spectrum of patients with diffuse infiltrative lymphocytosis syndrome (DILS) after the introduction of highly active antiretroviral treatment (HAART), and to carry out HLA class II oligotyping in these patients.


A retrospective chart review of patients with DILS who were referred to an outpatient facility for human immunodeficiency virus (HIV)–positive individuals between 1994 and 2003 was performed. DILS was diagnosed as suggested by previous criteria. Demographic features and relevant clinical, laboratory, and radiologic data were recorded and results analyzed.


A total of 129 patients with DILS were identified. Of them, 56 (43%) were African American, 41 (32%) were white, and 32 (25%) were Hispanic. Parotid gland swelling appeared to be the sine qua non of DILS. Twenty-seven percent of patients had opportunistic infections. The status of 103 patients was available as of December 2003: 26 (25%) had died, of which only 6 (6%) succumbed to opportunistic infections. The prevalence of DILS had significantly decreased in the post-HAART era (1998 onwards) compared with that of the pre-HAART period (P < 0.000001). The prevalence of lymphocytic interstitial pneumonitis had also dropped significantly following introduction of HAART therapy (P = 0.015). A higher frequency of certain HLA class II alleles (DRB1) was found in African Americans with DILS compared with those with HIV without DILS (P = 0.006).


The epidemiology, clinical presentation, and certain extraglandular manifestations of DILS have changed, concomitant with the introduction of HAART, further suggesting that DILS is an antigen (viral)-driven response and the primary treatment for it is anti-HIV therapy.


  1. Top of page
  2. Abstract
  7. Acknowledgements

The diffuse infiltrative lymphocytosis syndrome (DILS) is a disorder in patients with human immunodeficiency virus type 1 (HIV-1) that is characterized by salivary and lacrimal glandular swelling and sicca symptoms of varying intensity, frequently accompanied by persistent circulating and visceral CD8-positive lymphocytic infiltration (1). Although cases of a Sjögren's syndrome–like illness in HIV-1–positive patients were reported as early as 1985, DILS was first defined as a discrete disease entity in these patients by Itescu et al in 1989 (2). At that time HIV was a rapidly fatal illness, and zidovudine, the first effective antiretroviral agent, had only recently been introduced. Therefore the early clinical descriptions of DILS came from an era when many patients had yet to experience effective HIV treatment. Data from that time suggested that patients with DILS tended to be African Americans (3), commonly had numerous extraglandular manifestations such as lymphocytic pneumonitis (1, 4, 5), and had a slower time of progression to opportunistic infections and death (1, 4, 5).

Much has changed since those times. Highly active antiretroviral treatment (HAART) has changed the quality and length of life for individuals with HIV, and at the same time has brought new rheumatic diseases as complications (osteonecrosis due to protease inhibitors, complications of lipodystrophy, etc). It is entirely plausible that the introduction of HAART has changed the prevalence, natural history, and spectrum of clinical manifestations of DILS. The current study tests that hypothesis and presents epidemiologic, clinical, and immunogenetic data on a large multiethnic cohort of patients with DILS followed longitudinally during the time when these new treatments were introduced.


  1. Top of page
  2. Abstract
  7. Acknowledgements


Patients in this study were referred to the rheumatology clinic at the Thomas Street Clinic, an outpatient facility for indigent HIV-1–positive individuals administered within the Harris County Hospital District (comprising the city of Houston, TX and its environs), for evaluation of either parotid gland enlargement or sicca symptoms between July 1994 and December 2003. Approximately 652 patients who were HIV positive were evaluated and followed in the Thomas Street Clinic during this period. The criteria proposed by Itescu and Winchester (4) for DILS require that the patient be HIV-1 seropositive by enzyme-linked immunosorbent assay and Western blot analysis, have a bilateral salivary gland enlargement or xerostomia persisting for >6 months, and have histologic confirmation of salivary or lacrimal gland lymphocytic infiltration in the absence of granulomatous or neoplastic involvement. All of our patients in the study were seen at least once including the baseline visit and had either serial followup or December 2003 status available. Demographic and risk factor information, lymphocyte subsets (CD4, CD8, CD4:CD8 ratio), HIV viral load (measured either as the number of copies of RNA per milliliter of serum or their corresponding log10 values) analysis, HIV stage (6) at DILS diagnosis, duration of HIV, duration of DILS, time from HIV diagnosis to DILS diagnosis, information about the types of HIV medications, chest radiography, serum creatine kinase levels, complete blood count, serum electrolytes, biochemistry profile, viral hepatitis serology, and results of visceral biopsies were obtained by systematic chart review.

Laboratory analyses.

All of these patients' sera were tested for antinuclear antibodies (ANA) by indirect immunofluorescence using HEp-2 cells as antigen substrate (Antibodies Inc, Davis, CA), and for anti-Ro/SSA and anti-La/SSB antibodies by immunodiffusion. Serologic analyses for hepatitis B (hepatitis B surface antigen [HBsAg] and IgM–hepatitis B core antigen [HBcAg]), hepatitis C (anti–hepatitis C virus [anti-HCV]), and HIV-1 were performed in all cases. HLA–DRB1, DQA1, and DQB1 alleles were ascertained using standard oligotyping methods (7).

Minor salivary gland biopsy.

Minor salivary gland biopsy samples were obtained in all patients prior to 1998, and the diagnosis of DILS was established when a focus score of ≥2 by the criteria of Daniels (8) was determined.

Radionuclide scintigraphy using 67Ga.

Gallium scanning was performed ∼72 hours after intravenous injection of 5 mCi (185 MBq) of 67Ga citrate, and a positive scan was determined by a score of ≥2 (9).

Statistical analysis.

Stepwise Cox regression analysis was applied to some of these data to analyze their statistical significance. Using MINITAB 14 statistical software (Minitab, State College, PA), chi-square test, Kruskal-Wallis test, generalized linear model, and correlation coefficients were performed to assess the strength of these data and analyze their significance.


  1. Top of page
  2. Abstract
  7. Acknowledgements


Of the 652 consecutive patients with HIV referred for evaluation at the Thomas Street Rheumatology Clinic between July 1994 and December 2003, 129 patients with DILS were identified based on the criteria of Itescu and Winchester (4). DILS was diagnosed in 111 of the 129 patients during or before 1998, and the remaining 18 were diagnosed between January 1, 1999 and December 31, 2003. Of 306 new referrals to the rheumatology clinic at the Thomas Street Clinic between 1994 and 1997 (HAART therapy was introduced on a widespread basis at the Thomas Street Clinic at the end of 1997), 103 had DILS compared with only 26 of 346 new referrals between January 1, 1998 and December 31, 2003 (P < 0.000001, odds ratio [OR] 4.8, 95% confidence interval [95% CI] 2.86–8.13) (Table 1, Figure 1). Therefore, a much lower prevalence of DILS in new clinic referrals was seen since the introduction of HAART. Diagnoses of concomitant consecutive patients without DILS are recorded in Table 2.

Table 1. Clinical features of 3 cohorts of patients with diffuse infiltrative lymphocytosis syndrome*
SiteItescu and Winchester (4) pre-HAART (1990)Houston clinic pre-HAART (1993–1997)Houston clinic post-HAART (1998–2003)P
  • *

    Values are the number unless otherwise indicated. HAART = highly active antiretroviral treatment; NS = not significant.

  • Biopsy proven.

  • 32 months of followup.

  • §

    60 months of followup.

No. of patients1710326 
Incidence (per 100 population)40.8 
African Americans124313NS
Parotid gland enlargement176520NS
Lymphocytic interstitial pneumonitis103230.015
VIIth nerve palsy362NS
Aseptic meningitis143NS
Lymphocytic gastritis123NS
Lymphocytic hepatitis244NS
Opportunistic infections123§12§NS
thumbnail image

Figure 1. New cases of diffuse infiltrative lymphocytosis syndrome (DILS) seen at the Thomas Street Clinic 1994–2003. HAART = highly active antiretroviral treatment.

Download figure to PowerPoint

Table 2. Diagnoses of new patients referred to the rheumatology clinic at the Thomas Street Clinic from February 1994 through January 2004*
  • *

    Although the rheumatology services have been available at the Thomas Street Clinic since January 1991, systematic recording of new patient visits only began in February 1994 until the present. More than 1 diagnosis may be present for any patient. Values are the percentage unless otherwise indicated. DILS = diffuse infiltrative lymphocytosis syndrome; NOS = not otherwise specified; HIV = human immunodeficiency virus.

  • Including patients seen through the end of January 2004.

  • Where DILS could not be confirmed either due to lymphocytic interstitial pneumonitis (LIP) biopsy or, when LIP biopsy not done, negative 67Ga scanning, or where patient never returned for LIP biopsy or 67Ga scanning.

  • §

    Bursitis, tendonitis, or fibromyalgia.

  • Reactive arthritis, psoriatic arthritis, ankylosing spondylitis, or undifferentiated spondylarthritis.

  • #

    Including 2 patients with Wegener's granulomatosis, 1 with isolated central nervous system angiitis, 1 with polyarteritis nodosa, 1 with Henoch-Schönlein purpura, and 1 with cutaneous vasculitis.

Number of new patients80736680626252567253
Parotid gland enlargement without DILS (DILS not proven)4053888414
Sicca symptoms without DILS1020022200
Tuberculous adenitis (scrofula)0020002200
Soft tissue rheumatism§9262531313133344040
Low back pain (NOS)101061355813310
Carpal tunnel syndrome00000510738
HIV arthritis6433554476
Avascular necrosis1004004212
Bone or joint infection0551044400
Systemic lupus erythematosus0110200200
Rheumatoid arthritis0000000002

Clinical and laboratory features.

The mean ± SD age at DILS diagnosis was 39.6 ± 8.0 years (range 23–59 years), and the mean ± SD followup length from DILS diagnosis date until last contact date was 46.9 ± 21.6 months. The mean ± SD HIV disease duration (from diagnosis until last contact date) was 64.7 ± 42.5 months. The mean ± SD duration from HIV diagnosis to DILS diagnosis was 59.55 ± 43.8 months, and the mean duration of DILS-related symptoms (salivary glandular swelling and/or sicca symptoms) when first seen at the clinic was 9.05 ± 2.3 months.

Bilateral parotid enlargement was present on initial visit in 114 (88%) patients, and unilateral swelling was present in 11%. Five patients presented with only sicca symptoms without any salivary gland swelling, and all 5 were hepatitis C antibody negative.

Eleven (9%) patients presented with dyspnea on moderate to severe exertion and diffuse infiltrates on chest roentgenograph, and were found to have lymphocytic interstitial pneumonitis (LIP) established either by bronchoscopic biopsy (in 9 cases) or open lung biopsy (in 2 cases) showing CD8 lymphocytic infiltration. Compared with the pre-HAART period, it was observed that the prevalence of LIP was significantly decreased in our patients with DILS following the introduction of HAART therapy (P = 0.015, OR 4.36, 95% CI 1.12–24.74).

Of 19 patients with abnormal liver function test results, 8 had a lymphocytic hepatitis attributed to DILS. An elevated creatine phosphokinase was seen in 21 (17%) patients. Ten patients had renal involvement, of which 3 had lymphocytic interstitial nephritis without glomerular involvement on renal biopsies leading to renal insufficiency. Three patients had type IV renal tubular acidosis. Only 3 of the 129 patients with DILS had lymphoma, 2 shortly after presentation and the third 6 years after DILS diagnosis.

All patients' sera tested negative for anti-Ro/SSA and anti-La/SSB antibodies, and only 7 patients were positive for ANA by indirect immunofluorescence, all of whom had titers <1:320 dilution. One patient with positive ANA had associated autoimmune hepatitis, and 2 others with positive ANA had concurrent hepatitis C infection. Fifteen patients were found to have antibodies against hepatitis B (HBsAg or IgM-HBcAg positive) and 10 patients had antibodies against hepatitis C (anti-HCV positive) by the time of their initial rheumatology clinic visit.

By 1998, it was noted that patients referred to our clinic for parotid enlargement who were taking protease inhibitors had a significantly decreased frequency of positive minor salivary gland biopsy results (only 4 of 18 consecutive patients) compared with those not taking this group of drugs (85 out of 107 patients prior to 1997; P < 0.000001; OR 6.07, 95% CI 3.43–11.57). Subsequently, 67Ga scintigraphy became the primary diagnostic tool for patients receiving protease inhibitors.

Analysis using general linear model demonstrated strong correlations between parotid swelling and pain symptoms (as assessed by pain visual analog scale [VAS]; P = 0.009), parotid enlargement and the presence or extent of sicca symptoms (as assessed by sicca VAS; P = 0.018 for right parotid and P = 0.014 for left parotid gland), and pain and sicca symptoms (P < 0.001).

HLA analyses.

Among these 125 patients, 85 underwent HLA class II oligotyping by the standard methods described above. HLA–DRB1*1301 was found in 12 white and 11 African American patients with DILS, whereas HLA–DRB1*11, specifically the DRB1*1102 allele, was found in 16 patients, all of them African American. Other HLA–DRB1 alleles, specifically DRB1*0701(DR7), DRB1*1302(DR6), and DRB1*1201(DR5), were seen in 11, 8, and 5 patients, respectively. The susceptibility to DILS has been traced to the ILEDE (isoleucine-leucine-glutamic acid-aspartic acid-glutamic acid) epitope, encoded by HLA–DRB1*1102, *1301, *1302, and *1304 located in the third diversity region of the outermost domain of the HLA–DR molecule (10). We could not confirm this association; however, we did find an association of the ILED (isoleucine-leucine-glutamic acid-aspartic acid) epitope, encoded by the above HLA–DRB1 alleles plus HLA–DRB1*0701, *0803, *1201, and *1303, with DILS in African Americans (compared with African American HIV-positive controls [P = 0.006, OR 6.2] or African American HIV-negative controls [P = 0.04, OR 2.5]) (see Table 3). However, similar associations were not observed in the white and Hispanic patients with DILS.

Table 3. Frequencies of HLA–DRB1 alleles encoding the ILEDE and ILED epitopes in the third diversity region of the outermost domain in DILS patients from 3 ethnic groups*
GroupnILEDE epitopeILED epitope
  • *

    Values are the percentage unless otherwise indicated. ILEDE = isoleucine-leucine-glutamic acid-aspartic acid-glutamic acid; ILED = isoleucine-leucine-glutamic acid-aspartic acid; DILS = diffuse infiltrative lymphocytosis syndrome; HIV = human immunodeficiency virus.

  • The ILEDE epitope is encoded by HLA–DRB1*1102, *1301, *1302, and *1304. HLA–DRB1*0103 and *0402, also encoding this epitope, were not found in any of the patients in this study.

  • The ILED epitope is encoded by the above plus HLA–DRB1*0701, *0803, *1201, and *1303.

  • §

    P = 0.006, odds ratio 6.2 compared with African American HIV-positive controls.

  • P = 0.04, odds ratio 2.5 compared with African American HIV-negative controls.

African American DILS patients374976
African American HIV-positive controls183333§
African American HIV-negative controls583455
White DILS patients212952
White HIV-positive controls231330
White HIV-negative controls2003149
Hispanic DILS patients172929
Hispanic HIV-positive patients71414
Hispanic HIV-negative controls1051833


Ninety-one patients (73%) were receiving antiretroviral therapy at the time of DILS diagnosis. Of these, 87 (70%) were taking ≥1 nucleoside reverse transcriptase inhibitors, 11 (9%) were taking non–nucleoside reverse transcriptase inhibitors, and 38 (31%) were taking protease inhibitors. Of the 470 patients without DILS, other parotid gland swelling, or sicca symptoms for whom current medications were recorded at initial visit (medications were not recorded for 76 additional patients), 334 (71%) were taking antiretroviral agents at the time of their first rheumatology clinic visit. Of these, 326 (69%) were taking ≥1 nucleoside reverse transcriptase inhibitors and 173 (37%) were taking protease inhibitors. Ninety-six of the 226 patients seen were taking no antiretroviral medications. We then examined antiretroviral treatment in patients with DILS versus those without DILS prior to 1997 (the year HAART was started in general usage at the Thomas Street Clinic) compared with treatment after 1997. Of those with DILS initially seen before 1997, 31 (51%) were taking nucleoside reverse transcriptase inhibitors and 2 (3%) were taking protease inhibitors, compared with 60 (45%) and 4 (3%), respectively, of those 156 patients without DILS, parotid gland enlargement, or sicca symptoms (31 did not have medications listed). Of those 24 patients with DILS initially seen after 1997, 19 (74%) were taking antiretroviral treatment at the time of DILS diagnosis; all 19 were taking reverse transcriptase inhibitors and 15 (63%) were taking protease inhibitors. Of those 305 new patients without DILS, parotid gland enlargement, or sicca symptoms (of whom 33 did not have medications listed), 206 (76%) were taking reverse transcriptase inhibitors and 15 (63%) protease inhibitors. None of the differences between the DILS group and non-DILS group in the respective periods were significant.

Outcome and prognosis.

The mean ± SD HIV disease duration (from diagnosis until last contact date) was 94.7 ± 52.5 months. Twenty-six (25%) of 103 patients whose status was available had died by December 2003. The cause of death could be ascertained in 18 of these patients (Table 4). Only 6 (6%) patients died due to opportunistic infections such as Pneumocystis jiroveci pneumonia, cytomegalovirus, or Mycobacterium avium-intracellulare. On stepwise Cox regression analysis of those 129 patients, neither age, sex, ethnicity, HIV stage at DILS diagnosis, time from HIV diagnosis until DILS diagnosis, CD4 count, nor medication taken at DILS diagnosis (reverse transcriptase inhibitors, protease inhibitors) exerted any independent influence on DILS outcome. Only the HIV disease duration since diagnosis (P < 0.0001) independently predicted survival in these DILS patients.

Table 4. List of the different causes of death in patients with DILS*
Cause of deathNo. of patients
  • *

    DILS = diffuse infiltrative lymphocytosis syndrome; PCP = Pneumocystis jiroveci pneumonia; MAI = Mycobacterium avimumintracellulare; KS = Kaposi sarcoma; CMV = cytomegalovirus; GI = gastrointestinal.

Cardiopulmonary failure3
Cardiopulmonary failure with sepsis1
Sudden cardiac death2
PCP with hepatic failure1
MAI, KS, and CMV1
MAI with rupture of the myocardium1
PCP and MAI1
Renal failure with pneumonia1
Renal failure with pancytopenia1
GI perforation with upper GI bleeding1
Liver carcinoma1
Non–small cell lung carcinoma1
Propoxyphene overdose1


  1. Top of page
  2. Abstract
  7. Acknowledgements

The following data emerged from this study: the frequency of DILS appears to be diminishing remarkably with HAART therapy; with the introduction of HAART, the diagnosis of DILS by classical techniques (i.e., minor salivary gland biopsy) may be confounded; the frequencies of certain extraglandular manifestations of DILS, specifically LIP, have become less common; and predictors of mortality in patients with DILS are probably not different than in patients without DILS.

Previously we defined the prevalence of DILS at 3–4% of HIV-positive outpatients (where parotid gland enlargement was used as the criterion for screening) (3) and at 7.8% among HIV-positive Greek patients (where presence of sicca symptoms was used as the criterion) (11). These prevalence data were found before the widespread use of HAART therapy. In contrast, using histologic (and not clinical) criteria, the prevalence of DILS among 30 African Cameroonian adults with HIV (who might not have as ready access to antiretroviral agents) has recently been found to be 48% (12). The data from this study as well as recent data from Greece (13) suggest that the frequency of DILS is diminishing, most likely reflecting the effect of antiretroviral treatment, and further suggesting this to be an antigen-driven response (possibly viral antigen).

The lower frequency of some of the extraglandular symptoms is probably also due to HAART as mentioned before. The frequency of LIP was described to be as high as 25–50% in early series of patients with DILS (4, 5). With the introduction of HAART, this complication is now occurring less frequently.

Studies of major histocompatibility complex allele frequencies in patients with DILS have suggested a hereditary predisposition. Susceptibility to DILS has been previously shown to be associated with specific HLA–DR11 and DR13(DR6) serologic specificities in African American individuals and with HLA–DR13(DR6) only in white individuals (5, 10). Our data do not confirm this finding, although a similar amino acid sequence in the third diversity region of the HLA–DRB1 molecule was implicated (ILED in our study versus ILEDE in the studies of Itescu et al [10]).

DILS is a close phenotypic mimic of Sjögren's syndrome in terms of sicca symptoms, salivary glandular enlargement, gland histology, and propensity to develop non-Hodgkin's lymphoma. DILS differs from Sjögren's syndrome by the very frequent occurrence of extraglandular sites of lymphocytic infiltration, by the scarcity of serum autoantibodies, by the nature of infiltrating lymphocytes (CD4 in Sjögren's syndrome and CD8 in DILS), and by the different HLA associations (Table 5). Interestingly, viruses have been implicated in the etiology of Sjögren's syndrome (14). Expression of antigen reactive with a monoclonal antibody to human T lymphotropic virus type 1 (HTLV-1) P19 was found in the salivary glands of patients with Sjögren's syndrome (15). Also, Epstein-Barr virus (EBV) infection is abnormally regulated in Sjögren's syndrome and HIV (16, 17), as evidenced by the isolation of both EBV-1 and EBV-2 DNA from the tear film of patients affected with these 2 diseases in combination (17). Recently, patients with anti-Ro–positive DILS have been reported (18), although autoantibody negativity is characteristic. Studies looking for other viruses with an operative role in DILS, however, have been unrewarding to date (19), suggesting a role for the HIV virus itself. Emphasizing the diversity and changing spectrum of underlying disease processes, the sicca syndrome and CD8 lymphocyte tissue infiltration are also seen as common manifestations of chronic graft-versus-host disease (20) and HTLV-1 infection (21) (Table 5). Moreover, hepatitis C can present with sicca symptoms, parotid gland enlargement, certain extraglandular features that can mimic DILS (22), and a milder disease outcome when associated with HLA–DR5(DR11) (4).

Table 5. Comparison between DILS and certain other diseases that share some similarities*
VariablesSjögren's syndromeDILSHepatitis CHTLV-1Chronic GVHD
  • *

    DILS = diffuse infiltrative lymphocytosis syndrome; HTLV-1 = human T lymphotropic virus type 1; GVHD = graft-versus-host disease; RA = rheumatoid arthritis; ANA = antinuclear antibodies; SS = Sjögren's syndrome.

Sicca symptomsPresentPresentPresentPresentPresent
Glandular manifestationsModerate parotid enlargementModerate to severe parotid enlargementMild to moderate parotid enlargementMay be presentUsually absent
Extra glandular manifestationsMainly pulmonary, gastrointestinal, renal, and neurologic involvementMainly musculoskeletal, pulmonary, gastrointestinal, and neurologic involvementMainly gastrointestinal and musculoskeletal involvementMainly uveitis, pulmonary, and neurologic involvementMainly cutaneous, gastrointestinal, and pulmonary involvement
Infiltrating lymphocytic phenotypeCD4CD8CD4CD8CD8
AutoantibodiesHigh frequency RA factor, ANA, anti-Ro/SSA and anti-La/SSB Ro (rarely)Low frequency RA factor, ANA and anti-Ro (rarely)High frequency of RA factor and ANA, anti-Ro/SSA present, though less frequently than primary SSUnknownAbsent
HLA associationB8, DR2, DR3, and DR4B45, B49, B50, DR11(DR5), and DRw6DR-11(DR5)UnknownNone

These data and those from other recent studies suggest that the natural history of DILS is changing with the refinements in treatment of HIV infection. Although systematic data are lacking from populations where effective HIV treatment is not available, such as from Africa and Asia, the diminishing frequency of DILS itself as well as the extraglandular manifestation in our treated patients suggest that it is playing a less significant role than was previously believed.


  1. Top of page
  2. Abstract
  7. Acknowledgements

We would like to thank Thomas Cate, MD, Susan Miller, MD, MPH, Chris Lahart, MD, and Thomas P. Giordano, MD, MPH, the past and present medical directors of the Thomas Street Clinic, for their continued support of our efforts at the Thomas Street Clinic over the years. We would also like to thank the past and present rheumatology fellows at the University of Texas Health Science Center at Houston, especially Drs. Rosemarie Hirsch and Salahuddin Kazi, who have contributed so much to the care of these patients.


  1. Top of page
  2. Abstract
  7. Acknowledgements
  • 1
    Itescu S, Brancato LJ, Buxbaum J, Gregersen PK, Rizk CC, Croxson TS, et al. A diffuse infiltrative CD8 lymphocytosis syndrome in human immunodeficiency virus (HIV) infection: a host immune response associated with HLA-DR5. Ann Intern Med 1990; 112: 310.
  • 2
    Itescu S, Brancato LJ, Winchester R. A sicca syndrome in HIV infection: association with HLA-DR5 and CD8 lymphocytosis. Lancet 1989; 26: 4668.
  • 3
    Williams FM, Cohen PR, Jumshyd J, Reveille JD. Prevalence of the diffuse infiltrative lymphocytosis syndrome among human immunodeficiency virus type 1–positive outpatients. Arthritis Rheum 1998; 41: 8638.
  • 4
    Itescu S, Winchester R. Diffuse infiltrative lymphocytosis syndrome: a disorder occurring in human immunodeficiency virus-1 infection that may present as sicca syndrome. Rheum Dis Clin North Am 1992; 18: 68397.
  • 5
    Kazi S, Cohen PR, Williams F, Schempp R, Reveille JD. The diffuse infiltrative lymphocytosis syndrome: clinical and immunogenetic features in 35 patients. AIDS 1996; 10: 38591.
  • 6
    Centers for Disease Control and Prevention. 1993 revised classification system for HIV infection and expanded surveillance case definition for AIDS among adolescents and adults. MMWR Recomm Rep 1992; 41: 113.
  • 7
    Olsen ML, Arnett FC, Reveille JD. Contrasting molecular patterns of MHC class II alleles associated with the anti-Sm and anti-RNP precipitin autoantibodies in systemic lupus erythematosus. Arthritis Rheum 1993; 36: 94104.
  • 8
    Daniels TE. Labial salivary gland biopsy in Sjögren's syndrome: assessment as a diagnostic criterion in 362 suspected cases. Arthritis Rheum 1984; 27: 14756.
  • 9
    Rosenberg ZS, Joffe SA, Itescu S. Spectrum of salivary gland disease in HIV-infected patients: characterization with Ga-67 citrate imaging. Radiology 1992; 184: 7614.
  • 10
    Itescu S, Rose S, Dwyer E, Winchester R. Certain HLA-DR5 and -DR6 major histocompatibility complex class II alleles are associated with a CD8 lymphocytic host response to human immunodeficiency virus type 1 characterized by low lymphocyte viral strain heterogeneity and slow disease progression. Proc Natl Acad Sci U S A 1994; 91: 114726.
  • 11
    Kordossis T, Paikos S, Aroni K, Kitsanta P, Dimitrakopoulos A, Kavouklis E, et al. Prevalence of Sjögren's-like syndrome in a cohort of HIV-1-positive patients: descriptive pathology and immunopathology. Br J Rheumatol 1998; 37: 6915.
  • 12
    McArthur CP, Subtil-DeOliveira A, Palmer D, Fiorella RM, Gustafson S, Tira D, et al. Characteristics of salivary diffuse infiltrative lymphocytosis syndrome in West Africa. Arch Pathol Lab Med 2000; 124: 17739.
  • 13
    Panayiotakopoulos GD, Aroni K, Kyriaki D, Paikos S, Vouyioukas N, Vlachos A, et al. Paucity of Sjögren-like syndrome in a cohort of HIV-1-positive patients in the HAART era: part II. Rheumatology (Oxford) 2003; 42: 11647.
  • 14
    Talal N, Flescher E, Dang H. Evidence for possible retroviral involvement in autoimmune diseases. Ann Allergy 1992; 69: 2214.
  • 15
    Shattles WG, Brookes SM, Venables PJ, Clark DA, Maini RN. Expression of antigen reactive with a monoclonal antibody to HTLV-1 P19 in salivary glands in Sjögren's syndrome. Clin Exp Immunol 1992; 89: 4651.
  • 16
    Willoughby CE, Baker K, Kaye SB, Carey P, O'Donnell N, Field A, et al. Epstein-Barr virus (types 1 and 2) in the tear film in Sjögren's syndrome and HIV infection. J Med Virol 2002; 68: 37883.
  • 17
    Fox RI, Luppi M, Kang HI, Pisa P. Reactivation of Epstein-Barr virus in Sjögren's syndrome. Springer Semin Immunopathol 1991; 13: 21731.
  • 18
    Hansen A, Feist E, Hiepe F, Burmester GR, Scholze J. Diffuse infiltrative lymphocytosis syndrome in a patient with anti–52-kd Ro/SSA and human immunodeficiency virus type 1. Arthritis Rheum 1999; 42: 57880.
  • 19
    Lee-Wing MW, Hodge WG, Diaz-Mitoma F. Investigating a viral etiology for keratoconjunctivitis sicca among patients who are positive for human immunodeficiency virus. Cornea 1999; 18: 6714.
  • 20
    Anderson NG, Regillo C. Ocular manifestations of graft versus host disease. Curr Opin Ophthalmol 2004; 15: 5037.
  • 21
    Merle H, Cabre P, Olindo S, Merle S, Smadja D. Ocular lesions in 200 patients infected by the human T-cell lymphotropic virus type 1 in martinique (French West Indies). Am J Ophthalmol 2002; 134: 1905.
  • 22
    Ramos-Casals M, Loustaud-Ratti V, de Vita S, Zeher M, Bosch JA, Toussirot E, et al, and the SS-HCV Study Group. Sjogren syndrome associated with hepatitis C virus: a multicenter analysis of 137 cases. Medicine (Baltimore) 2005; 84: 819.