Anti-C1q antibodies are associated with systemic lupus erythematosus global activity but not specifically with nephritis: A controlled study of 126 consecutive patients

Authors


Abstract

Objective

Several studies have shown that anti-C1q antibodies correlate with the occurrence and activity of nephritis in systemic lupus erythematosus (SLE). However, the significance of anti-C1q antibodies in SLE has not been fully characterized. The aim of this study was to investigate associations between anti-C1q antibodies and clinical and serologic parameters of SLE.

Methods

An enzyme-linked immunosorbent assay kit was used to measure anti-C1q antibodies in the sera of 126 consecutive patients with active SLE who were admitted to our university hospital from 2007 through 2009. Sera obtained from patients with high titers of anti-C1q antibodies at the initial evaluation (n = 20) were reevaluated following treatment. Control sera were obtained from patients with other autoimmune diseases and from normal healthy control subjects (n = 20 in each group). Associations between anti-C1q antibodies and clinical and serologic parameters of SLE were statistically analyzed.

Results

Anti-C1q antibodies were detected in the sera of 79 of 126 patients with SLE. The prevalence and titers of anti-C1q antibodies were significantly (P < 0.0001) higher in SLE patients than in patients with rheumatoid arthritis, patients with systemic sclerosis, and normal healthy control subjects. The prevalence and titers of anti-C1q antibodies were not significantly associated with active lupus nephritis (P = 0.462 and P = 0.366, respectively). Anti-C1q antibody titers were significantly correlated with SLE Disease Activity Index 2000 scores and the levels of anti–double-stranded DNA antibodies, C3, C4, CH50, and C1q (P < 0.0001 for all comparisons). Moreover, anti-C1q antibody titers significantly decreased as clinical disease was ameliorated following treatment (P = 0.00097).

Conclusion

These findings indicate that anti-C1q antibodies are associated with SLE global activity but not specifically with active lupus nephritis.

Systemic lupus erythematosus (SLE) is an autoimmune disease characterized by the presence of autoantibodies. The diagnosis of SLE can be difficult, because SLE is a great imitator of other diseases (1). Autoantibodies are clearly central to the pathogenesis of SLE, with different autoantibodies associated with different clinical features (2). Among these autoantibodies (>100), several have been associated with disease activity (1). Although anti–double-stranded DNA (anti-dsDNA) antibodies are the most extensively studied autoantibodies in SLE, others play a role in clinical manifestations such as autoimmune hemolytic anemia, thrombocytopenia, skin disease, and neonatal lupus (3).

C1q is a complex molecule consisting of a collagenous portion with globular heads, morphologically resembling a bundle of tulips. C1q is the first component of the classical pathway of complement activation, and its main function is to clear immune complexes (ICs) from tissues and self antigens generated during apoptosis (4). The hereditary deficiency of this component is a known risk factor for the development of SLE (4).

Anti-C1q antibodies were first described by Uwatoko et al in 1984 (5). Whereas complexed IgG mainly binds to the globular portion of C1q, anti-C1q antibodies bind to the collagenous portion, which apparently is the main immunogenic region of the molecule and its Fab fragments (6). Initially, anti-C1q antibodies were observed in patients with hypocomplementemic urticarial vasculitis syndrome and in patients with SLE (7). Anti-C1q antibodies were subsequently reported to be present in various other diseases (8). It has also been reported that anti-C1q antibodies correlate with the occurrence and activity of lupus nephritis, especially proliferative lupus nephritis (7, 9, 10). Many clinical studies showed a high negative predictive value (NPV) of anti-C1q antibodies for the occurrence of (proliferative) lupus nephritis, ranging up to 100% (9, 10). A pathogenic role for these antibodies in the development of lupus nephritis has been suggested (11). One study also showed that the prevalence of organ manifestations other than nephritis was the same in patients with and those without high titers of anti-C1q antibodies (12).

The actual occurrence of anti-C1q antibodies in patients with active lupus nephritis remains controversial (9). Gunnarsson et al observed anti-C1q antibodies in only 11 of 18 patients with proliferative lupus nephritis (13). Heidenreich et al reported that tests for anti-C1q antibodies and ICs performed worse than tests for anti-dsDNA antibodies for the differentiation of lupus nephritis from other forms of glomerulonephritis (14). In their randomized controlled trial for proliferative lupus nephritis, Grootscholten et al showed that the prevalence of anti-C1q antibodies was 65%, and that renal flares were not preceded by increases in the titer of anti-C1q antibodies (15). Marto et al did not observe differences in the prevalence or levels of anti-C1q antibodies when comparing proliferative and nonproliferative forms of nephritis and suggested that results of other studies of a small number of patients—particularly those with nonproliferative nephritis—do not hold true in a large patient cohort (4). Most previous studies were retrospective (patients were not consecutively or randomly sampled) (4, 10, 12, 16) and frequently were small in size (9, 17). Thus, the significance of anti-C1q antibodies in SLE and lupus nephritis has not been fully characterized, and testing for these antibodies still does not have a defined place in routine clinical practice (4). The aim of the current study was to investigate associations between anti-C1q antibodies and clinical and serologic parameters of SLE in a larger controlled study of consecutive patients.

PATIENTS AND METHODS

Patients.

We performed a case–control study of patients who were treated for active SLE at the Tokyo Women's Medical University hospital from 2007 through 2009. A total of 126 consecutive patients with active SLE were identified using the Tokyo Women's Medical University SLE Database, and sera were obtained from these patients. All of these patients fulfilled at least 4 of the American College of Rheumatology (ACR) revised criteria for SLE (18, 19). At our institution, patients suspected of having SLE or those with newly diagnosed SLE were typically admitted for systemic evaluation regardless of the severity of disease and were eligible for inclusion in the study. Patients in whom SLE was previously diagnosed and who experienced disease flares were also included. Of the 135 patients admitted during the study period, 9 were excluded from the study because of the unavailability of samples or the lack of informed consent.

At the time of admission to the hospital, each patient completed a standardized medical history, including medication use, and was given a physical examination. Serology profiling for each patient was performed using the standard immunoassays described below. Serum samples that were obtained prospectively for another study were acquired before initiation or reinforcement of treatment, either at the time of admission or at the time of renal biopsy (when performed), and stored at −80°C. Treatment with corticosteroids or immunosuppressive drugs was instituted following completion of these evaluations. Sera from patients with high titers of anti-C1q antibodies (>125 units/ml) at the initial evaluation were reevaluated after disease was ameliorated by treatment (n = 20); these reevaluation samples were collected cross-sectionally in April or May, 2010. The mean ± SD interval between the time of initial evaluation and reevaluation was 645 ± 342 days. Control sera were obtained from age- and sex-matched normal healthy control subjects and from patients with rheumatoid arthritis (RA) or systemic sclerosis (SSc), diagnosed using standard criteria (20, 21). This study was approved by the ethics committee of our institution, and the principles of the Helsinki Declaration were followed throughout the study. Informed consent was obtained from all participants.

Data collection.

The information obtained from the medical records of the patients included demographic data such as age at the time of initiation of treatment, the clinical manifestations of SLE, and laboratory values. SLE disease activity was measured using the SLE Disease Activity Index 2000 (SLEDAI-2K) (22). Each SLE-related feature was defined according to the revised ACR criteria for SLE (18, 19) or, if the feature was not included in the criteria, the SLEDAI-2K (22). For example, active nephritis was defined as persistent proteinuria of >0.5 gm/day (or >3+ if quantification was not performed) or by the presence of cellular casts in urine. Leukopenia was defined as a decrease in the number of white blood cells to <4,000/mm3. Central nervous system (CNS) lupus was defined and diagnosed according to the standardized ACR nomenclature and case definitions for neuropsychiatric lupus syndromes (23). We included only 12 CNS syndromes in the present study because of the substantial differences in anatomy, function, and clinical characteristics between the central and peripheral nervous systems (24, 25). Anemia was defined as a decrease in the concentration of hemoglobin to <10.0 gm/dl due to any cause. Urticarial vasculitis was defined both clinically, as persistent urticarial skin lesions, and microscopically, as leukocytoclastic vasculitis (26). In some cases, renal biopsies were performed, and the histologic findings were classified according to the International Society of Nephrology/Renal Pathology Society (ISN/RPS) criteria (27).

Measurement of complement and anti-dsDNA antibodies.

Levels of C3 and C4 were measured by turbidimetric immunoassay. CH50 was measured using a modification of the method of Mayer. Levels of serum ICs were measured by a C1q solid-phase enzyme immunoassay. Levels of serum C1q complement component were measured by rate nephelometry. Levels of serum anti-dsDNA antibodies were measured using the Farr radioimmunoassay.

Enzyme-linked immunosorbent assay (ELISA) for anti-C1q antibodies.

Anti-C1q antibodies were measured using a solid-phase ELISA kit (Bulhmann Laboratories) according to the manufacturer's protocol. Briefly, calibrators, control sera, and patient sera (stored samples) were diluted in high salt buffer (0.5M NaCl) to avoid false-positive results by binding of ICs (28) and were incubated with human C1q adsorbed onto microtiter wells. After washing, horseradish peroxidase–labeled anti-human IgG conjugate was added, followed by a second washing step and the addition of tetramethylbenzidine substrate. The intensity of the blue color developed was in proportion to the amount of anti-C1q antibodies bound in the initial step. The reaction was terminated by the addition of 0.25M H2SO4. The absorbance was measured in a microtiter plate reader (Bio-Rad) at a wavelength of 450 nm and converted into units (units/ml) by plotting against the autoantibody titer of the calibrators/standards given by the manufacturer. The assay range was 1.0–400 units/ml. The cutoff value suggested by the manufacturer (15 units/ml) was obtained by testing the samples from 220 normal healthy blood donors using the same assay procedure. In some analyses in the present study, an increased cutoff value of 40 units/ml, as proposed by Trendelenburg et al, was used in order to achieve results comparable with those of previous studies (9). Sera from 20 patients with high titers of anti-C1q antibodies (>125 units/ml) at the initial evaluation were reevaluated when disease improved with treatment. Sera from patients with RA, patients with SSc, and normal healthy control subjects were also tested (n = 20 in each control group).

Statistical analysis.

Associations between anti-C1q antibodies and clinical and serologic parameters of SLE were analyzed using Fisher's exact test for categorical variables and the Mann-Whitney U test for continuous variables for 2-group comparisons. Comparisons of 3 or more groups were performed using the Kruskal-Wallis test, and Steel's multiple comparison test for continuous variables and the chi-square test for categorical variables. The relationships between anti-C1q antibody levels and other continuous variables were analyzed using Spearman's rank correlation. The anti-C1q antibody levels before and after treatment were compared using Wilcoxon's signed rank test. P values less than 0.05 were considered significant. All tests were 2-tailed. The sensitivity, specificity, positive predictive value (PPV), and NPV of anti-C1q antibodies were also calculated. All analyses were performed using JMP statistical software (SAS Institute).

RESULTS

Demographic characteristics of the patients with SLE.

Of the 126 patients with SLE enrolled in the present study, 123 were women, and 3 were men. The patients ranged in age from 17 years to 77 years (median age 37 years). With the exception of 2 Chinese women, all of the patients were Japanese. Twenty-one patients had clinically active nephritis. Renal biopsies were performed to confirm lupus nephritis by histopathology in 20 of these patients. Because of complications, a renal biopsy was not performed in 1 patient. Of the 20 patients who underwent biopsy, 1 patient had class I, 1 patient had class II, 3 patients had class III, 5 patients had class IV, and 10 patients had class V lupus nephritis, according to the abbreviated version of the ISN/RPS classification system (27); combined classes III/V and IV/V were considered as class III and IV, respectively. Contrary to previous reports, the prevalence and titers of anti-dsDNA antibodies in these 21 patients were not significantly associated with active lupus nephritis (P = 0.974 and P = 0.628, respectively).

Higher prevalence and titers of anti-C1q antibodies in patients with active SLE compared with patients with RA, patients with SSc, and normal healthy control subjects.

Using the cutoff value of 15 units/ml, as recommended by the manufacturer of the ELISA kit, the prevalence of anti-C1q antibodies in patients with active SLE (79 of 126) was significantly higher than that in normal health control subjects (2 of 20), patients with RA (2 of 20), and patients with SSc (3 of 20) (P < 0.0001 for all). When samples from patients with SLE and normal healthy control subjects were compared, the sensitivity, specificity, PPV, and NPV for the diagnosis of SLE were 63%, 90%, 98%, and 28%, respectively. Using the higher cutoff value described by Trendelenburg et al (40 units/ml) (9), the prevalence of anti-C1q antibodies in patients with SLE (51 of 126) remained significantly higher than that in normal healthy control subjects (1 of 20), patients with RA (1 of 20), and patients with SSc (2 of 20) (P = 0.002, P = 0.002, and P = 0.011, respectively). Titers for anti-C1q antibodies were significantly higher in patients with SLE than in normal healthy control subjects or patients with RA or patients with SSc (the median values were 21.9 units/ml, 2.9 units/ml, 4.4 units/ml, and 2.5 units/ml, respectively; P < 0.0001 for all) (Figure 1).

Figure 1.

Anti-C1q antibody (Ab) levels in sera obtained from 126 consecutive patients with active systemic lupus erythematosus (SLE), 20 control patients with rheumatoid arthritis (RA), 20 control patients with systemic sclerosis (SSc), and 20 normal healthy controls (NHCs), as measured using an enzyme-linked immunosorbent assay (ELISA) kit. The broken line indicates the cutoff value (15 units/ml) recommended by the manufacturer of the ELISA kit. The titers of anti-C1q antibodies were significantly higher in patients with SLE than in normal healthy controls, patients with RA, and patients with SSc (P< 0.0001 for each comparison).

Associations between SLE-related clinical features and prevalence and titers of anti-C1q antibodies.

Anti-C1q antibodies were detected in the sera of patients showing various clinical manifestations of active SLE (Table 1). The prevalence of anti-C1q antibodies was significantly higher in SLE patients with leukopenia (P = 0.029) or hypocomplementemia (P < 0.0001) and in those who were positive for anti-dsDNA antibodies (P < 0.0001) and ICs (P < 0.0001) than in those without these features. Contrary to previous reports, the prevalence of anti-C1q antibodies was not significantly associated with active nephritis (P = 0.462), although the NPV of anti-C1q antibodies for active nephritis was as high as 87%, in accordance with previous reports (7, 9).

Table 1. Associations between systemic lupus erythematosus–related clinical features and incidence of anti-C1q antibodies*
Clinical featureAnti-C1q antibodyPSensitivity, %Specificity, %PPV, %NPV, %
Positive (n = 79)Negative (n = 47)
  • *

    Values are the number (%). P values were determined by Fisher's exact test. PPV = positive predictive value; NPV = negative predictive value; CNS = central nervous system; anti-dsDNA = anti–double-stranded DNA; ANAs = antinuclear antibodies.

Malar rash/discoid rash31 (39)13 (28)0.24770413972
Urticarial vasculitis2 (3)0 (0)0.529100383100
Alopecia6 (8)3 (6)1.0006738894
Oral or nasal ulcers2 (3)1 (2)1.0006737398
Serositis5 (6)3 (6)1.0006337694
Arthritis24 (30)10 (21)0.30571403079
Active nephritis15 (19)6 (13)0.46271391987
CNS lupus3 (4)2 (4)1.0006037496
Vasculitis2 (3)2 (4)0.6295037396
Myositis2 (3)0 (0)0.529100383100
Fever >38°C15 (19)7 (15)0.63468381985
Thrombocytopenia7 (9)3 (6)0.7437038994
Leukopenia15 (19)2 (4)0.02988411996
Anemia21 (27)7 (15)0.18375412785
Positive anti-dsDNA antibodies62 (78)14 (30)<0.000182667870
Hypocomplementemia62 (78)20 (43)<0.000176617857
Positive immune complex38 (48)1 (2)<0.000197534898
Positive ANAs65 (94)35 (83)0.07665629417

The titer of anti-C1q antibodies was significantly higher in SLE patients with leukopenia (P = 0.046), anemia (P = 0.027), or hypocomplementemia (P < 0.0001) and in those who were positive for anti-dsDNA antibodies (P < 0.0001) and ICs (P < 0.0001) than in those without these features (Table 2). Also in disagreement with previous reports, we observed that the titer of anti-C1q antibodies was not significantly associated with active nephritis (P = 0.366). Even when active nephritis was subcategorized into “proliferative lupus nephritis” (class III and IV according to the ISN/RPS criteria) (4) and other categories, we observed no significant difference in the prevalence of anti-C1q antibodies in patients with proliferative lupus nephritis and those without active nephritis (7 of 8 patients and 64 of 105 patients, respectively; P = 0.257). Interestingly, in the 2 SLE patients with urticarial vasculitis, the prevalence and titers of anti-C1q antibodies were very high (Tables 1 and 2, respectively), although the number of these patients was too small to reach statistical significance.

Table 2. Associations between SLE-related clinical features and titers of anti-C1q antibodies
FeaturePresentAbsentP
  • * Values are the median units/ml of anti-C1q antibodies in patients with or without the clinical feature of systemic lupus erythematosus (SLE). CNS = central nervous system; anti-dsDNA = anti–double-stranded DNA; ANAs = antinuclear antibodies.

  • By Mann-Whitney U test.

Malar rash/discoid rash38.721.30.085
Urticarial vasculitis152.721.80.160
Alopecia18.122.20.751
Oral or nasal ulcers400.021.90.313
Serositis36.821.90.916
Arthritis23.821.80.357
Active nephritis29.821.80.366
CNS lupus15.822.00.508
Vasculitis13.622.00.310
Myositis22.621.90.907
Fever >38°C49.321.70.175
Thrombocytopenia85.921.90.189
Leukopenia48.120.80.046
Anemia46.620.80.027
Positive anti-dsDNA antibodies54.29.9<0.0001
Hypocomplementemia39.99.9<0.0001
Positive immune complex165.715.8<0.0001
Positive ANAs23.813.40.077

Correlations between anti-C1q antibody titers and markers of disease activity.

When assessing the correlation between anti-C1q antibody titers and an established index and known serologic markers of disease activity, the titers of anti-C1q antibodies were correlated with the SLEDAI-2K, anti-dsDNA antibodies, C3, C4, CH50, and IC by C1q assay (Figures 2A, B, C, D, E, and F, respectively). For all comparisons, there were significant positive or negative correlations (all P < 0.0001).

Figure 2.

Correlations between anti-C1q antibody (Ab) titers and the Systemic Lupus Erythematosus Disease Activity Index 2000 (SLEDAI) score (A), anti–double-stranded DNA (anti-dsDNA) antibodies (B), C3 (C), C4 (D), CH50 (E), and immune complexes (ICs) (F), as measured in 126 consecutive patients with active SLE. For all comparisons, significant positive or negative correlations (P < 0.0001) were observed.

Correlation between anti-C1q antibody titers and levels of C1q.

When anti-C1q antibody titers were compared with the levels of C1q in patients with SLE, a significant negative correlation was observed (P < 0.0001), as shown in Figure 3. The fact that whether or not the level of C1q decreased beyond its normal range (8.8–15.2 mg/dl) in each patient was also significantly associated with the presence of anti-C1q positivity (>15 units/ml), as determined using Fisher's exact test for categorical variables (P = 0.0010).

Figure 3.

Significant negative correlation (P < 0.0001) between anti-C1q antibody (Ab) titers and the levels of C1q measured in 126 consecutive patients with active systemic lupus erythematosus.

Anti-C1q antibody titers before and after treatment in patients with SLE who had high titers of anti-C1q antibodies at the initial evaluation.

In a subgroup of 20 patients with SLE who had high titers of anti-C1q antibodies (>125 units/ml) at the initial evaluation, retesting showed that anti-C1q antibody titers decreased significantly in accordance with disease amelioration following treatment (median decrease 84% [range −147% to 99%; P = 0.00097]) (Figure 4). The SLEDAI-2K scores in these 20 patients also decreased (median decrease 66% [range 25–100%], P < 0.0001). Four of these 20 patients initially had active lupus nephritis. In the only patient in whom the titer of anti-C1q antibodies increased (from 162 units/ml to 400 units/ml), a new rash had developed, and anti-dsDNA antibody positivity and hypocomplementemia persisted at the time of reevaluation (615 days later), although other SLE-related manifestations had improved. In the only patient in whom anti-C1q antibody titers did not change (400 units/ml before and after treatment), anti-dsDNA antibody positivity and hypocomplementemia persisted, although other SLE-related manifestations improved. In the latter patient, the anti-C1q antibody titers exceeded the upper limit of the assay; thus, there may have been some difference between the actual titers before and after treatment.

Figure 4.

Anti-C1q antibody (Ab) titers before and after treatment in 20 patients with active systemic lupus erythematosus (SLE) who had high titers of anti-C1q antibodies (>125 units/ml) at the initial evaluation. The broken line indicates the cutoff value (15 units/ml) recommended by the manufacturer of the enzyme-linked immunosorbent assay kit. The anti-C1q antibody titers decreased significantly in accordance with clinical amelioration of disease following treatment (median decrease 84% [range −147% to 99%]; P = 0.00097).

DISCUSSION

In our study of a large cohort of 126 consecutive patients with SLE, we observed that anti-C1q antibodies are associated with SLE global activity. However, in contrast to the results of previous, predominantly retrospective studies (4, 10, 12, 16) that frequently involved fewer subjects (9, 17), our results demonstrated that anti-C1q antibodies are not specifically associated with active lupus nephritis. We also showed that anti-C1q antibody titers significantly decreased as the patients' condition improved with clinical treatment. To our knowledge, this study is the first to investigate associations between anti-C1q antibodies and clinical and serologic parameters of SLE using a large prospective or consecutive cohort of patients with SLE.

This study demonstrated that the presence of anti-C1q antibodies is significantly associated with SLE. When serum samples from patients with SLE and those from normal healthy control subjects were compared, the specificity of the presence of anti-C1q antibodies for the diagnosis of SLE was as high as 90% and that for the PPV was as high as 98%. The prevalence of these antibodies in patients with RA and patients with SSc was as low as that in normal healthy control subjects. However, because anti-C1q antibodies have been observed to occur in many disease conditions (8), their presence cannot be said to be specific for any one disease (6).

Anti-C1q antibodies were detected in the sera of patients with active SLE who had a variety of manifestations. Leukopenia was shown to be significantly associated with anti-C1q antibodies. Interestingly, Armstrong et al also observed a correlation between the presence of anti-C1q antibodies and hematologic disease as well as renal disease in patients with SLE (29). However, another study demonstrated that the incidence of organ manifestations other than nephritis was the same in patients with and those without high titers of anti-C1q antibodies (12). As in earlier studies (4, 17), we observed good correlations between anti-C1q antibodies and recognized markers of SLE activity: the SLEDAI, anti-dsDNA antibodies, C3, C4, CH50, and IC by C1q assay. Our findings indicate that anti-C1q antibodies are associated with SLE global activity but not specifically with active lupus nephritis. Although the main function of C1q is to clear ICs from tissues and self antigens generated during apoptosis (4), C1q has other biologic functions and could play roles as both facilitating and inhibiting/protective factors (30). As one example, Lood et al recently reported a novel function for C1q in the regulation of IC-induced production of interferon-α and other cytokines by plasmacytoid dendritic cells (31). It appears that anti-C1q antibodies may have several different pathogenic roles in SLE (6, 32) and may account for various clinical manifestations of the disease.

Unlike previous studies, our study did not reveal a specific association between anti-C1q antibodies and active lupus nephritis. We speculate that the discrepancy between previous findings and our results may arise from differences in the patient populations. Most previous studies were retrospective (not consecutively or randomly sampled) (4, 10, 12, 16) and were frequently small in size (9, 17). In addition, most previous studies were led by departments/divisions of nephrology and/or focused on nephritis, usually severe; this could lead to sampling bias, because patients with severe lupus nephritis might be more likely to be included in such studies than in population samples of all SLE patients. Therefore, these earlier studies may not adequately reflect the full spectrum of SLE.

In contrast, our cohort is more likely to accurately represent the population of lupus patients commonly seen at clinics/hospitals specializing in rheumatology. Although many of the previous studies have shown a high NPV of anti-C1q antibodies for the occurrence of severe lupus nephritis, ranging up to 100%, it is worth noting that not all studies demonstrated such a definite result (6, 8). Several explanations for the variation among studies have been proposed, including the following: across studies, the timing of anti-C1q antibody testing in relation to the renal flare was not uniform, different definitions of active lupus nephritis and positive test results were used, and assays were not standardized (32). Our study avoided these issues because 1) serum samples were always obtained before initiation or reinforcement of treatment, on admission or at the time of renal biopsies, when performed, 2) active lupus nephritis was defined consistently using the most standardized method, and 3) we used the same commercial kit as was used in previous studies (9, 17).

In many previously reported cases, the levels of anti-C1q antibodies increased prior to the exacerbation of lupus nephritis (6). Thus, in some of the patients in our study who did not have active lupus nephritis, overt lupus nephritis may have developed over time if treatment had not been initiated. In fact, the number of patients with clinically overt nephritis in this study was only a small subset of the entire group (21 of 126 patients). Furthermore, some of our SLE patients who did not fulfill the criteria for active lupus nephritis did have renal biopsies, which revealed class II, III, IV, or V lupus nephritis. Our early clinical intervention could partly account for this discrepancy.

The NPV of anti-C1q antibodies for active nephritis was as high as 87% in our study, in accordance with previous reports (4, 9, 10), while the PPV was only 19%. Because nephritis does not develop in many other diseases involving the presence of anti-C1q antibodies (8), indications are that anti-C1q antibodies are required but not solely sufficient for development of renal inflammation in SLE (11, 32). Flierman and Daha proposed a hypothetical model of lupus nephritis pathogenesis that involves 4 steps (11). First, circulating ICs containing nucleosomes deposit in the glomerulus by charge interactions with the glomerular basement membrane. Second, glomerular IC deposits bind C1q. Third, the exposure of neoantigens within the solid-phase C1q molecule serves as a focus, i.e., planted antigen, for circulating anti-C1q antibodies. Last, full activation of the classical pathway of the complement system leads to tissue injury mediated by the membrane attack complex as well as the influx of inflammatory cells, such as neutrophils. In addition, SLE patients who have anti-dsDNA antibodies but not anti-C1q antibodies may exhibit some degree of renal disease that may be only mild and/or effectively regulated via complement regulators.

The titers of anti-C1q antibodies in SLE patients showing a variety of clinical manifestations significantly decreased and, in some patients, decreased even beyond their cut-off levels (<15 units/ml) as the disease condition improved with treatment, as previously reported (9, 16). These findings also support the view that anti-C1q antibodies are associated with SLE global activity but not specifically with active lupus nephritis. It is possible that anti-C1q antibodies could be useful as a surrogate marker of SLE disease activity in patients positive for this antibody.

When anti-C1q antibody titers were compared with the levels of C1q in patients with SLE, the correlation was significant and negative; when the 2 parameters were measured together, they seemed to have a mirror-like pattern of serum concentrations, as previously reported (17, 33). These findings also suggested that the 2 parameters are potential markers of SLE global activity. We speculate that a decrease in the number of C1q antibodies in patients with active SLE is mainly attributable to excessive consumption of C1q by complement activation by anti-C1q antibodies and by the formation and tissue deposition of C1q–C1q antibody complexes (33–35). According to an alternative theory, anti-C1q antibodies are a mere product of this activation itself, which creates neoepitopes in the C1q molecule (17, 36).

Although our study was appreciably larger than the majority of previous studies dealing with anti-C1q antibodies and SLE, an insufficient number of patients with SLE still limited our ability to draw definitive conclusions about the significance of anti-C1q antibodies in SLE. Another potential weakness of our study arises from the variation in treatment protocols among the patients, reflecting different clinical presentations. Followup durations also varied among the patients, because reevaluation samples were collected at certain time points cross-sectionally. Because our study was not designed to validate anti-C1q antibodies as a biomarker for predicting lupus flares, as suggested previously (6), we are not able to comment on this issue, except to note that in most of our study patients, lupus flares of any form did not occur during followup. In addition, because almost all of the patients in the present study were Japanese, it is not clear whether anti-C1q antibodies have a different effect on lupus nephritis in patients with different ethnic backgrounds. However, several studies on patients of other ethnicities support our findings (29).

In conclusion, anti-C1q antibodies are associated with SLE global activity but not specifically with active lupus nephritis. The results of this study should caution clinicians against relying too much on anti-C1q antibodies as a diagnostic marker of lupus nephritis. Anti-C1q antibodies may, however, be useful as a surrogate marker for disease activity in patients with SLE in whom this antibody is present. Standardization of the measurement of anti-C1q antibodies and its increased use would be of benefit in the routine clinical diagnosis and treatment of SLE.

AUTHOR CONTRIBUTIONS

All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be published. Dr. Katsumata had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study conception and design.Katsumata, Kawaguchi, Yamanaka.

Acquisition of data.Katsumata, Miyake, Kawaguchi, Okamoto, Kawamoto, Gono, Baba.

Analysis and interpretation of data.Katsumata, Kawaguchi, Hara.

Acknowledgements

We thank Dr. Katsuji Nishimura (Department of Psychiatry, Tokyo Women's Medical University) for assisting with the psychological examinations and diagnoses.

Ancillary