*4.1. Retrovirus Infection (Other Than HTLV-1 Infection) and SS*

Representative studies on the involvement of retroviruses in SS are listed in Table 2.



**Table 2.** *Cont.*


CMV: cytomegalovirus, ELS; ectopic lymphoid structures, HIV: human immunodeficiency, HRV: human retrovirus, PCR; polymerase chain reaction, SGD: salivary gland disease, SGs; salivary glands, SLS: Sjögren's-like syndrome, TCR: T cell receptor.

The possibility of retrovirus infections other than HTLV-1 infection in individuals with SS was a research focus in the early 1990s. In 1989 it was reported that patients with HIV-associated salivary gland disease (HIV-SGD) showed a high frequency of parotid enlargement, sicca symptoms, and a low CD4/8 ratio in LSG biopsy specimens [75]. Interestingly, no anti-Ro/SS-A or La/SS-B antibodies were detected in the enrolled patients. Garry and colleagues described a human intracisternal A-type retroviral particle that was closely associated with HIV which they detected in lymphoid cells that were exposed to homogenates of salivary gland tissue from SS patients [76]. In investigations of the presence of serum antibodies of retrovirus, sera from 30% of healthy subjects had positive reactivity toward p24 (gag) but not gp41 or gp120 (env), with low reactivity to Ro/SS-A and La/SS-B [77,78]. Itescu et al. reported cases of HIV-positive patients demonstrating parotid enlargement, sicca symptoms, and CD8 T-lymphocyte dominance with HLA-DR5 [79], and the term 'diffuse infiltrative lymphocytosis syndrome (DILS)' was proposed [80].

The diagnostic criteria of DILS are as follows: (1) HIV infection with positive serology, (2) bilateral salivary gland enlargement or xerostomia, (3) persistence of signs or symptoms for ≥6 months, and (4) histologic confirmation of salivary or lacrimal gland lymphocytic infiltration without granulomatosis or neoplastic involvement. All four of these criteria must be met for the diagnosis of DILS. A difference in the T-cell receptor (TCR) beta-chains was also revealed between SS patients and subjects with DILS [81]. In a cohort study in Greece, SS-like syndrome (SLS) was confirmed in 7.79% of HIV-positive subjects, in whom CD8<sup>+</sup> T cell-dominant infiltration was observed in salivary gland biopsy specimens without anti-Ro/SS-A, La/SS-B antibody [82]. An assessment of the prevalence of DILS at an HIV outpatient clinic in the U.S. revealed that 3% of the HIV-seropositive patients (15/523) had DILS with enlargement of the parotid gland and sicca symptoms [83]. In their study of HIV-associated antigen, Yamano et al. observed that 33% of the sera and 47% of the salivary gland specimens from SS patients reacted with p24 (gag) of HIV, although target genes for HIV were not detected in either type of samples from SS patients [84]. Their findings suggest the involvement not of HIV but of an unknown retrovirus in SS.

The presence of human retrovirus-5 (HRV-5) in salivary glands was examined by nested PCR in 55 SS patients and 37 non-SS subjects, and no association of HRV-5 with SS was observed [85]. The estimated prevalence of SLS among 131 HIV patients under an intensive anti-retroviral strategy, i.e., highly active anti-retroviral therapy (HAART), was 7.8% [86]. As of 2020, the number of publications concerning a relationship between SS and HIV has decreased, probably because of the development of HAART and the 2002 description of HIV infection as well as hepatitis C infection in the exclusion criteria in the classification of SS provided by the American-European Consensus Group (AECG) [87].

## *4.2. Clinical and Epidemiological Findings of HTLV-1 Infection in SS*

In parallel with investigations of involvement of the HIV in Sjögren's syndrome, the engagement of HTLV-1 in SS has been researched. In this section, we summarize the clinical and epidemiological studies of the association between HTLV-1 infection and SS. HTLV-1, a member of the type C delta-retrovirus family [88], is a causative agent of adult T-cell leukemia (ATL), which is a hematological neoplastic disorder first described by Takatsuki et al. in 1977 [89]. After HTLV-1 was recognized as causative virus for ATL, the whole base sequence [90] and the function of Tax protein [91] were determined. HTLV-1 is also a causative agent for the tropical spastic paraparesis/HTLV-1-associated myelopathy (HAM) identified in the Caribbean basin and the southeast area of Japan; patients with HAM exhibit slowly progressive neurological symptoms such as spastic gait and bladder and rectal disturbance [92,93]. HTLV-1-associated uveitis [94] is also observed in HTLV-1 carriers, and the frequent complication of Grave's disease is reported [95]. Broncho-alveolitis is an example of organ involvement in HTLV-1 infection as a pulmonary complication [96]. As transmission pathways of HTLV-1, maternal infection with various levels of HTLV-1 proviral DNA during pregnancy was reported [97]. Representative publications about the detection of genes and proteins in HTLV-1-associated SS are listed in Table 3.


**Table 3.** Detection of HTLV-1-related genes and proteins in sera or SGs from patients with Sjögren's syndrome.

AA: amino acid, EM: electron microscopy, HAM: HTLV-1-associated myelopathy, IF: immunofluorescence, IB: immunoblot, IHC: immunohistochemistry, ISH: in situ hybridization, LSG: labial salivary gland, PBMC: peripheral blood mononuclear cell, PCR: polymerase chain reaction, RA: rheumatoid arthritis, RT: reverse transcription, SG: salivary gland, SGEC: salivary gland epithelial cell, SSCP: single-strand confirmation polymorphism, TCR; T-cell receptor.

Although several case reports regarding SS associated with HTLV-1 infection are available, it was initially reported that HTLV-1 carriers with SS showed high frequencies of extra-glandular manifestations including uveitis and myopathy [112]. It was subsequently demonstrated that in Nagasaki, Japan (an endemic area for HTLV-1), the prevalence of HTLV-1 in SS patients was significantly higher than that of 27,284 blood donors [113]. In addition, a serum antibody toward HTLV-1 in HTLV-1-seropositive SS patients was similar to that from patients with HTLV-1-associated myelopathy (HAM). In a study conducted in Chile, among 48 patients with HAM, 29.1% of the patients who lacked autoantibodies including anti-Ro/SS-A antibody or rheumatoid factor (RF) showed chronic dacryosialadenitis without being classified as having SS [110]. We observed that 60% of a series of HAM patients fulfilled the preliminary European Community criteria for SS with positive mononuclear cell infiltration in LSGs [114], and these HAM patients with SS showed low frequencies of anti-Ro/SS-A antibody, anti-nuclear antibody, and RF compared to those observed in typical SS.

The high prevalence of SS in HAM patients was also observed in a later investigation in which chronic sialadenitis and exocrine dysfunction were detected by Saxon test and Schirmer's test, respectively [115]. Magnetic resonance imaging (MRI) revealed that 92% of HAM patients with SS lacked typical characteristics usually found in SS, although the salivary flow rate of these patients was similar to that of SS patients without HAM [99]. We subsequently documented a low frequency of abnormal findings in HTLV-1-seropositive patients with SS compared to those of HTLV-1-seronegative patients with SS [100], although the aforementioned MRI findings showed much clearer imaging differences than those in sialography. The relationship between abnormality in such images and the pathogenesis of SS is discussed later in this review. With reference to our previous study [88] that showed a high frequency of myopathy in HTLV-1-seropositive SS, Cruz et al. [101] noted that their study conducted in Brazil revealed that 38% of HTLV-1-seropositve subjects had fibromyalgia, and their HTLV-1-seropositive subjects had a high prevalence of rheumatic diseases. We found that among atomic-bomb survivors in Japan, the prevalence of HTLV-1 infection in SS patients was higher than that in non-SS subjects [102].

We also re-evaluated the complications of SS experienced by patients with HAM 18 years after we reported the association [116], because the classification criteria were revised as AECG criteria [87] and the aging of HTLV-1-seropositive subjects with the changes in the composition of the population in Japan was established [117,118]. As a result, 38.5% of our patients with HAM were classified as having SS based on the AECG criteria, and the prevalence of anti-Ro/SS-A antibody was significantly lower than that in HTLV-1 carriers complicated with SS. A cross-sectional study performed in Brazil demonstrated that in a group of 272 HTLV-1-seropositive participants at an HTLV-1 clinic, 59 (21.7%) patients (none of whom had HAM) showed sicca symptoms without anti-Ro/SS-A, La/SS-B antibodies [119]. That study also revealed that proinflammatory cytokines were elevated in HTLV-1-seropositive subjects with sicca symptoms regardless of the presence/absence of SS, suggesting that HTLV-1 itself has the potential to induce nonspecific inflammation in salivary glands regardless of the presence of autoimmune disease.

Vale et al. [104] examined the association between HTLV-1 infection and the presence of SS in 129 HTLV-1-seropositive subjects in Brazil; they observed xerostomia and xerophthalmia in 35.7% and 13.95% of the subjects, respectively. In contrast, these HTLV-1-seropositive subjects showed only 0.77% anti-La/SS-B antibody. Regarding pulmonary involvement in SS, we detected an increased frequency of airway diseases in HTLV-1-seropositive SS patients [120], indicating an influence of HTLV-1 infection as a systemic manifestation.
