Clinical analysis of patients with new-onset untreated systemic lupus erythematosus complicated with cytomegalovirus infection

We investigated the clinical characteristics and short-term prognosis of patients with new-onset, untreated systemic lupus erythematosus (SLE) complicated with active cytomegalovirus (CMV) infection. Of the 86 patients enrolled in this study, 60 had active CMV infection. The proportions of patients with fever (65.0% versus 38.5%, P = 0.041), hematologic system involvement (80.0% versus 57.7%, P = 0.032), kidney involvement (50.0% versus 26.9%, P = 0.047), anti-double-stranded-DNA antibody positivity (51.7% versus 26.9%, P = 0.034), and severe disease activity (SLEDAI-2000 ≥ 15 points; 58.3% versus 30.8%, P = 0.019) were significantly higher, whereas serum C3 complement levels (0.48 ± 0.22 versus 0.60 ± 0.25, P = 0.032) were significantly lower in the CMV+ group than in the CMV- group, respectively. Initial SLE treatment, clinical features, laboratory data, and SLE treatment did not differ at 3-and 6-month follow-ups. In patients with CMV+, there were no significant differences in clinical features, laboratory data, SLE treatment, and laboratory index at 3-month follow-up between patients with or without antiviral treatment. These findings provided important insights into the characteristics of new-onset SLE complicated with active CMV infection and showed that antiviral treatment may have affect short-term prognosis.


Introduction
The etiology and pathogenesis of systemic lupus erythematosus (SLE) is unclear. According to recent studies, chronic infections are associated with SLE 1 . Human cytomegalovirus (HCMV, also known as CMV) is a pathogen that mostly infects the immunocompromised population 2 and is closely related to the occurrence and development of SLE 3,4 . Antibodies against CMV are commonly detected in patients with SLE. Previous studies reported a high incidence of CMV infection in patients with SLE being treated with steroids and/or immunosuppressants.

Comparison of initial SLE treatment between CMV+ and CMV-groups
The results of the initial treatment analysis for the two SLE groups are shown in Table 4. The differences in the rates of glucocorticoid pulse therapy (26.7% versus 11.5%, respectively; P = 0.120), high-dose glucocorticoid (45.0% versus 34.6%, respectively; P = 0.369), and glucocorticoids combined with immunosuppressants (50.0% versus 34.6%, respectively; P = 0.188) between the CMV+ and CMV-groups were not significant.

Follow-up results
At the 3-month follow-up, 77 patients were evaluated, among which 23 were in the CMVgroup, and 54 were in the CMV+ group. Ten patients (18.5%, 10/54) in the CMV+ group did not meet the SLE classification criteria. Differences in the percentages of increased serum complement C3 levels compared with initial serum complement C3 levels (55.4% versus 47.6%, respectively; P = 0.544) as well as liver transaminitis (12.5% versus 28.6%, respectively; P = 0.182) between the CMV+ and CMV-groups were not significant. For clinical manifestations, the difference in the rate of fever (5.4% versus 19.1%, respectively; P = 0.157) between the CMV+ and CMV-groups was not significant. In terms of SLE treatment, the CMV+ and CMV-groups showed no significant differences when a small dose of glucocorticoids (69.6% versus 90.5%, respectively; P = 0.059) and glucocorticoids combined with immunosuppressive agents (35.7% versus 28.6%, respectively; P = 0.555) were used.

Results of antiviral and nonantiviral treatments
Patients with new-onset SLE in the CMV+ group were divided into antiviral and nonantiviral treatment groups. The clinical manifestations, laboratory results, initial treatment, medication, and decreased SLEDAI-2000 scores of the two groups were statistically analyzed. As shown in Table 5, the differences in clinical manifestations between the antiviral+ and antiviralgroups, including the incidence of fever (69.2% versus 63.8%, respectively; P = 0.974), incidence of liver transaminitis (38.5% versus 57.5%, respectively; P = 0.225), and SLEDAI-2000 scores (16.23 ± 5.11 versus 16.23 ± 7.15, respectively; P = 0.998), were not significant.
At the 3-month follow-up, the differences in the rate of SLEDAI-2000 score less than or equal to 4 points (33.3% versus 56.8%, respectively; P = 0.149) and the PGA scores evaluated by doctors (0.92 ± 0.79 versus 0.66 ± 0.86, respectively; P = 0.355) between the antiviral treatment group and non-antiviral treatment group were not significant. In terms of SLE medication, the differences in the rates of low-dose hormone (58.3% versus 72.7%, respectively; P = 0.544) and hormone combined with immunosuppressant (50.0% versus 31.8%, respectively; P = 0.196) between the antiviral+ and antiviral-groups were not significant.

Discussion
In recent decades, many studies have examined the roles of viral infection in the pathogenesis of SLE, also known as the viral hypothesis 5 . CMV belongs to the herpes virus family and is thought to trigger SLE, with CMV DNA and high levels of serum anti-CMV IgM detected in patients with initial symptoms 6 or disease exacerbation 7 . However, whether CMV infection triggers SLE or occurs simultaneously with or after SLE onset remains unclear. Newkirk et al. showed that the U1 small nuclear ribonucleoprotein (U1 snRNP) antibody is related to CMV antibody in the serum of patients with SLE 8 . An earlier study by Mohamed et al. suggested that the positive rate of anti-HCMV antibodies is higher in patients with SLE than in healthy individuals 9 . Stratta et al. found that HCMV infection is inseparable from the "vascular" manifestations of SLE and is related to SLE with nephrotic syndrome 10 . Nawata et al. also confirmed that HCMV infection can lead to the occurrence of SLE 11 . These conclusions suggest that the onset of SLE is induced by active CMV infection.
Our study showed that the incidence of active CMV infection in patients with new-onset SLE was 69.8%, which was relatively high. A review of previous studies revealed differences in the reported proportion of patients with SLE and CMV infection. The reasons for these discrepancies may be related to the definition of active CMV infection, material obtained, and selection of samples. Takizawa et al. 12 used the CMV pp65 antigen to detect active CMV infection; they found that the highest incidence of CMV infection was observed in patients with SLE (n = 51) and that 75% of CMV-infected patients presented with CMV disease.
Rozenblyum 13 improved CMV etiology examination using urine and bronchoalveolar lavage fluid and found that the active CMV infection rate in patients with SLE was 1.04%. Through neutralization of the RF factors that may interfere with CMV detection, Su 14  to CMV infection 20 . CMV may increase the risk of rejection mediated by human immunity and chronic graft-versus-host disease, following allogeneic grafts 21 . In addition, increased generation of hypergammaglobulinemia, cryoglobulinemia, and autoantibodies is observed on monocytes induced by CMV and post-reperfusion syndrome 22,23 . Anti-dsDNA antibodies are SLE-specific antibodies that act as markers of SLE disease activities 24 . Previous research has suggested that after obtaining pp65 immunity, C57BL/6 mice show a significant rise in anti-dsDNA antibody levels; therefore, CMV may be related to the generation of anti-dsDNA 25 .
CMV infection may induce SLE-related morbidity or cause variations in the illness state.
According to previous research 26 , U1 snRNP, an important target antigen for autoimmune diseases, is involved in the process of mRNA maturity as a type of small nuclear ribonucleoprotein. In the hematologic system of patients with SLE, anti-CMV antibodies are associated with anti-U1 snRNP antibodies. However, in the serum of patients with other connective tissue diseases, anti-CMV does not associate with U1 snRNP. Thus, CMV is an important element that induces patients with SLE to generate an autoimmune reaction.
Consistent with these results, we found that the incidence of anti-dsDNA antibody-positivity in the CMV+ group was higher than that in the CMV-group.
There may be several reasons for the higher proportion of hematologic system involvement in the CMV+ group than that in the CMV-group in this study. First, like most herpes viruses, CMV has latent-activated biological characteristics. Once CMV invades the human body, most people will have latent infection without symptoms. When active CMV infection occurs, the virus first replicates in peripheral blood mononuclear cells and then invades the phagocytic cells of other organs 27 ; therefore, the hematologic system is first involved when CMV replication is active. Second, a crossreaction occurs between the anti-platelet and anti-CMV antibodies, resulting in thrombocytopenia. Shigemura proposed that different fever levels cannot discriminate between CMV infection and SLE; however, thrombocytopenia is mainly present in patients infected with CMV 28  Moreover, previous studies 32,33 found no correlation between CMV IgM and SLEDAI scores.
Although the above-mentioned 10 patients with SLE may have achieved clinical remission, the number of patients enrolled in this study was small, making it difficult to adequately distinguish between them; a larger sample size is needed for verification, along with laboratory parameters or further pathological examinations, to confirm this assumption.
As stated, patients in the CMV+ group were further divided into two groups: those receiving antiviral therapy and those not receiving antiviral therapy. At follow-up, the results obtained for the antiviral+ and antiviral-groups in this study were consistent with those of previous studies 34 . Moreover, antiviral therapy has been shown to be nonessential for treating active CMV infection complicated with incipient SLE. In addition, with the remission of SLE, the CMV virus marker can give a negative result without intervention 35 . It is thought that the death rate of patients with SLE and active CMV infection is not high, regardless of high morbidity 36 ; therefore, antiviral therapy is not necessary. With remission of SLE, the CMV infection marker can spontaneously change from positive to negative 37 . On study 38 showed that inflammation itself was important for the persistence of CMV infection and that controlling inflammation could facilitate active CMV infection. During follow-up visits with the group to which antiviral therapy was administered and the group that did not receive antiviral therapy, no significant differences in clinical symptoms and laboratory results were identified. This indicates that controlling inflammation with steroid and immunosuppressant drugs is more important than antiviral treatment for controlling both SLE and CMV infection.
This was a single-center study. Considering the limited time and small number of samples, we could not fully summarize the clinical characteristics of the combination of new-onset SLE with active CMV infection; therefore, further investigations are needed to verify our results.

Conclusions
Patients with new-onset SLE and active CMV infection were more likely to exhibit fever, hematologic system involvement, kidney damage, decreased serum complement C3, and positive anti-dsDNA antibodies than those without active CMV infection. In addition, in the CMV+ group, the SLEDAI-2000 score showed a higher proportion of severe disease activity (58.3%). Active CMV infection may be involved in the pathogenesis or exacerbation of SLE; however, antiviral therapy may have no significant effect on the treatment and short-term prognosis of SLE. This study provides data supporting the role of active CMV infection in the pathogenesis of SLE.

Definition of active CMV infection
Active CMV infection was defined as the presence of any of the following: CMV DNA in peripheral blood (> 5 × 10 2 , copy number/mL); CMV pp65 antigen in the blood (cell number/2 × 10 5 leukocytes); CMV IgM antibody (> 30 U/mL) + CMV IgG antibody (> 1 U/mL). The titer of serum antibody CMV IgG, obtained by double testing, was at least 4-fold higher than that of the baseline data or both CMV IgM and CMV IgG were positive, indicating recent active infection. Patients were divided into two groups according to their active CMV infection status: those with active CMV infection were placed in the CMV+ group, and those with no active CMV infection were placed in the CMV-group.

Laboratory tests
Laboratory tests mainly included routine blood, biochemical, and urine tests. Other parameters included related antibodies, such as anti-dsDNA antibodies, anti-smooth muscle cell (anti-sm) antibodies, and immunoglobulins.

Treatment for active CMV infection
Patients with new-onset untreated SLE and CMV infection were further divided into antiviral and non-antiviral therapy groups according to their symptoms, laboratory examination results, clinician judgment, and patient wishes. The antiviral therapy group, in addition to treatment for SLE, was also intravenously administered ganciclovir as an antiviral treatment for 14 days (5 mg/kg, q12h), after which patients continued to take ganciclovir for 14 days (1 g tid). Thus, the course of treatment lasted for 28 days.

Follow-up
Patients were followed up as outpatients for 3 and 6 months after treatment. Changes in the SLEDAI-2000 score, medication, clinical symptoms, and laboratory results were analyzed.
According to the clinical symptoms and laboratory results, the PGA score (0, 1, 2-2.5, 3) was assessed by the same rheumatologist. The results were evaluated and recorded during the follow-up visits and classified as remission, low disease activity, moderate disease activity, and severe disease activity.

Statistical analysis
Statistical analyses were performed using the SPSS software version 21.0 (SPSS, Inc., Chicago, IL, USA). Count results were analyzed by the chi-square or Fisher's exact tests; measurement data were expressed as means ± standard deviations, and t-tests were used for analyses. Results with P values less than 0.05 were statistically significant.  a Fever corresponds to axillary body temperature > 37.3°C; b hematological system involvement refers to white blood cells < 3 .5 × 10 9 /L, hemoglobin < 115 g/L, and platelets < 125 × 10 9 /L.
Any one of the three items that meet the above criteria represents hematological system involvement; c respiratory system involvement includes pleural effusion, pleurisy, and interstitial lung disease; d kidney involvement refers to urinary protein > 0.5/24 h or +++, which needs to be excluded due to insufficient intake or drugs; e digestive system involvement includes liver function (elevated liver enzymes such as ALT and AST) and mesenteric vasculitis; f neurological complications need to exclude drugs and metabolic disorders, recent occurrence of convulsions, disorientation, and cerebrovascular accidents; g rheumatoid factor