Association of antinucleosome antibodies with disease flare in serologically active clinically quiescent patients with systemic lupus erythematosus

Authors


Abstract

Objective

To identify the prevalence of serologically active clinically quiescent (SACQ) patients in a cohort of 290 patients with systemic lupus erythematosus (SLE). We investigated if the presence of anti–double-stranded DNA (anti-dsDNA) or antinucleosome (anti-NCS) antibodies during the SACQ period was associated with future flares.

Methods

SACQ patients defined as clinically inactive for 6 months (global British Isles Lupus Activity Group index [BILAG] scores <6) and serologically active (anti-dsDNA antibodies >50 units/ml on at least 2 occasions by enzyme-linked immunosorbent assay [ELISA]) were identified. Patient sera collected during the defined SACQ period were also tested for anti-NCS antibodies (ELISA). We retrospectively reviewed patient clinical details and episodes of flare using the BILAG activity index.

Results

Twenty-seven (9%) patients were SACQ. Seventeen (81%) patients experienced a flare (total of 91 flares, up to 12 flares per person) in the next 5 years. Median duration to first flare was 15 months (range 2–46). Time to first flare after SACQ period was significantly correlated with the presence of anti-NCS (P = 0.0012), high anti-NCS antibody titers (P = 0.0006), and anti-dsDNA titers 5 times above the normal limit (P = 0.02). Patients with higher absolute anti-NCS antibody titers showed a significant correlation with the number of flares (r = 0.57, P = 0.007).

Conclusion

A minority of patients with SLE are SACQ. The majority of these patients experience a flare in the next 5 years and close followup is recommended. Anti-NCS antibodies may be a better predictor than anti-dsDNA antibodies for future flares.

INTRODUCTION

Systemic lupus erythematosus (SLE) is an autoimmune rheumatic disease with great diversity of clinical manifestations. More than 100 different autoantibodies have been identified in the sera of patients with SLE (1). The most prominent of these autoantibodies, anti–double-stranded DNA (anti-dsDNA) antibodies, are present in 40–80% of patients with SLE. For more than 35 years, the anti-dsDNA antibody has been utilized as a marker for disease activity, especially in renal disease. According to some reports, the increase of anti-dsDNA titer may indicate impending flare, supporting the idea that these antibodies have a pathogenic role (2, 3). Recently, newer assays detecting antinucleosome (anti-NCS) antibodies have shown promise in assessing disease activity, especially when anti-dsDNA antibodies are negative (4). IgG3 subclass anti-NCS antibodies have been shown to be associated with lupus nephritis and flare (5).

In 1979, Gladman and colleagues reported that 14 (8%) of 180 of their patients with SLE exhibited active serologic markers in the absence of clinical activity, and these patients were described as being serologically active clinically quiescent (SACQ) (6). Of the 11 patients followed up for more than 10 years, 7 patients remained clinically inactive without the need for any immunosuppressive agents (7). Walz LeBlanc et al observed that almost half of the 74 patients with SLE who had an SACQ period experienced a flare within a year using the Systemic Lupus Erythematosus Disease Activity Index (SLEDAI) global activity score, with no predictive factors identified during and before the SACQ period (8). The prevalence of anti-NCS antibodies in this group of patients is unknown. It is also not known if anti-NCS antibodies can help predict flares.

We aimed to confirm and extend the original study over a 5-year period using the British Isles Lupus Activity Group (BILAG) index (9), which distinguishes disease activity into 8 organs or systems. In addition, we determined whether measuring anti-NCS antibodies and anti-dsDNA antibodies during these SACQ periods better predicts disease outcome.

PATIENTS AND METHODS

Patients.

SACQ patients were identified from a large, single-center cohort of patients with SLE attending the Middlesex/University College London clinic from January 1, 1999 to June 30, 1999. A total of 290 patients with lupus were regularly followed up at that time. All patients fulfilled at least 4 of the revised American College of Rheumatology (ACR) criteria for the classification of SLE (10). At each clinic visit, which occurred on average every 3 months, disease activity was assessed using the BILAG index (9). To be included in the study, patients had to have been followed up subsequently for 5 years and had to have SACQ disease. SACQ was defined as a mean global BILAG score <6 and anti-dsDNA antibody titers above the normal level on at least 2 occasions during the 6-month period.

Laboratory analysis.

Patients' sera were obtained during the SACQ period and stored at −20°C. All samples were tested for anti-dsDNA antibodies (normal <50 units/ml) by enzyme-linked immunosorbent assay (ELISA; Shield Diagnostics, Dundee, UK) and Crithidia luciliae immunofluorescent assay. Anti-NCS antibodies were tested by ELISA. For this assay, nucleosomes were prepared from Jurkat cells. Nuclei were extracted using a fine-tissue homogenizer, and chromatin was digested using a micrococcal nuclease. Integrity of the product was confirmed by running aliquots on an agarose gel following phenol–chloroform extraction. This demonstrated the characteristic oligonucleosome ladder pattern. The concentration of dsDNA in the nucleosome sample derived by measuring the absorbance at 260 nm using a spectrophotometer was ∼1 mg/ml.

The nucleosome preparation was diluted at 1:1,000 in phosphate buffered saline (PBS) and was coated on half of a Nunc MaxiSorp plate (the test half; Nunc, Naperville, IL). The other half was coated with PBS alone (the control half). The plates were washed with PBS–Tween (PBST) and then blocked with 2% bovine serum albumin. After further washing, serum samples were loaded in duplicate on both the test and control sides of the plate, and were incubated for 1 hour at 37°C. The plates were washed again with PBST. Bound antibody was detected by adding goat anti-human IgG alkaline phosphatase conjugate (Sigma A3150; Poole Dorset, UK) and incubating for 1 hour at 37°C. Substrate was added and optical density (OD) at 405 nm was read. The true OD for each sample was calculated as OD in test well minus OD in control well to exclude effects of background nonspecific binding.

The cutoff level of positive anti-NCS titers was standardized with a control group. The control group comprised 30 healthy individuals (6 men and 24 women). The threshold for positive anti-NCS was defined as 3 SDs above the mean value of 0.018 OD/405 nm in the control group. The patients were subdivided into groups of low versus high titers of anti-NCS antibodies. High anti-NCS titers were defined as >0.134 OD/405 nm (twice the threshold value for positive anti-NCS antibodies).

Data analysis.

The patients' charts were retrospectively reviewed to determine whether new episodes of flare had occurred during the 5-year followup period. A flare was defined as the appearance of a new BILAG A or B from a previous record of BILAG C, D, or E in any organ system. The number of flares and organ systems involved were noted. Patients who experienced >5 flares in the followup period were considered to have had multiple flares. The time to first flare after SACQ period was calculated. Patient demographics, clinical features of disease, treatment during SACQ, and immunosuppressive agent use (from diagnosis until end of followup period) were identified.

Statistical analysis.

Statistical analysis using the PRISM software program (Graphpad, San Diego, CA) was performed to look for associations of the presence of anti-dsDNA antibodies (by ELISA and Crithidia luciliae assays) or anti-NCS antibodies with future flare after the SACQ period. SACQ patient characteristics were compared using chi-square test. Fisher's exact test was used to compare patients who experienced flare and those who did not. Kaplan-Meier curves were generated to analyze time to first flare after SACQ period in relation to anti-dsDNA or anti-NCS antibody positivity and titer levels. The association between absolute anti-dsDNA and anti-NCS antibody titers and number of flares was analyzed using regression analysis with Spearman's correlation coefficient. A P value less than 0.05 was regarded as statistically significant.

RESULTS

Of the total SLE cohort, 27 patients (9%) were SACQ. Six patients were excluded from the study due to death, loss to followup, and transfer to another health facility. We analyzed data on the remaining 21 SACQ patients in detail. Patient demographics, clinical manifestations, and immunosuppressive drug use are shown in Table 1. The group was predominantly white (57%) with a mean age of 45 years. Most patients had had SLE for many years (mean duration 16 years). These demographics were similar to the total SLE cohort. Steroid use was low (mean dose 5 mg, range 3–10 mg) during the SACQ period. Two patients (patients 3 and 13) had their immunosuppressive agent dose reduced and discontinued during this time.

Table 1. Patients' demographic and clinical characteristics*
Patient no.SexEthnicityAgeDisease duration, yearsMain clinical featuresImmunosuppressives/ antimalarial useSLE medications during SACQ period
  • *

    SLE = systemic lupus erythematosus; SACQ = serologically active clinically quiescent; HCQ = hydoxychloroquine; CNS = central nervous system; AZA = azathioprine; CYC = cyclophosphamide; pred = prednisolone; MTX = methotrexate; CSA = cyclosporin A; NSAID = nonsteroidal antiinflammatory drug; APLS = antiphospholipid syndrome; NIL = no medications.

1FWhite6719Arthralgia/arthritis, rash, oral ulcers, alopeciaHCQHCQ
2FOther4225Arthralgia/arthritis, rash, CNS, renalAZA, CYC, pred, HCQHCQ, pred
3FAfrican Caribbean4712Rash, oral ulcers, alopecia, serositis, renalAZA, predAZA, pred
4FWhite4411Arthralgia/arthritis, rash, serositisPred, HCQHCQ
5FWhite4510Arthralgia/arthritis, rashHCQHCQ
6FAfrican Caribbean5027Arthralgia/arthritis, rash, mouth ulcers, serositis, CNSMTX, CYC, CSA, predPred
7FAsian3015Arthralgia/arthritis, mouth ulcers, rash, serositis, CNSAZA, CYC, pred, HCQNSAID
8FOther5824Arthralgia/arthritis, rash, pleuritis, renalAZA, predAZA, pred
9FWhite5324Arthralgia/arthritis, rashPred, HCQHCQ
10FWhite5329Arthralgia/arthritis, rash, mouth ulcers, renalAZA, pred, HCQAZA, HCQ, pred
11FWhite5715Arthralgia/arthritis, rash, CNS, renalAZA, pred, HCQAZA, pred
12FAsian3517Rash, serositis, renalCYC, AZA, MTX, pred, HCQAZA, pred
13FAsian3611Arthralgia/arthritis, rash, CNSAZA, MTX, CYC, pred, HCQMTX, HCQ, NSAID
14FWhite408Arthralgia/arthritis, rash, serositisPred, HCQHCQ, pred
15FAfrican Caribbean518Arthralgia/arthritis, renalAZA, CYC, pred, HCQAZA, pred
16FWhite5719Rash, renalAZA, CYC, pred, HCQAZA, pred
17FWhite3114Arthralgia/arthritis, rash, alopecia, mouth ulcers, APLSAZA, predAZA, warfarin
18FWhite3916Renal, CNS, APLSAZA,CYC, predAZA, pred
19FWhite5510Arthralgia/arthritisAZA, pred, HCQAZA, HCQ, pred
20FAsian309Arthralgia/arthritis, CNSAZA, pred, HCQAZA, HCQ, pred
21MWhite3111Rash, alopecia, renalAZA, pred, HCQNIL

There was a higher prevalence of arthritis/arthralgia during the entire disease duration in the total SLE cohort compared with the SACQ group (96% versus 76%; P ≤ 0.001). All other clinical features were similar.

The majority of patients (17 patients [81%]) experienced flare in the 5 years following the SACQ period. Eleven (65%) patients experienced flare within the first year after the SACQ period, with a mean time to first flare of 15 months (range 2–46).

There were 91 flares observed, with up to 12 flares per patient. There were 12 A flares in 6 patients, and half of these flares (6 of 12) were neurologic. We observed 79 B flares in 17 patients. The B flares were mainly musculoskeletal (24%), but all other organ systems were also affected. Seven patients (41%) had multiple flares.

Seventeen patients (81%) tested positive for anti-NCS antibodies. These patients had a significantly shorter time to first flare after the SACQ period compared with patients without anti-NCS antibodies (P = 0.0012) (Figure 1). In addition, higher anti-NCS titers correlated with a shorter time to first flare (P = 0.0006) (Figure 2). In comparison with anti-dsDNA antibody titers, only patients who had anti-dsDNA antibody levels 5 times above the normal limit (>250 units/ml) showed a significant shorter time to first flare after the SACQ period (P = 0.02).

Figure 1.

Association of antinucleosome (anti-NCS) antibodies with time to first flare after serologically active clinically quiescent period.

Figure 2.

Association of high-level antinucleosome (anti-NCS) antibodies with time to first flare after serologically active clinically quiescent period.

Patients with a higher absolute anti-NCS antibody titer had a significant correlation with the number of flares observed after the SACQ period (r = 0.57, P = 0.007) (Figure 3). This association was not seen with anti-dsDNA antibody titers (r = 0.13, P = 0.58). We then attempted to determine if anti-dsDNA and anti-NCS antibodies measured during the SACQ period could be associated with future flares by comparing patients who subsequently experienced flare in the next 5 years (81%) versus those who did not experience flare (19%). Using Fisher's exact analysis, we did not identify any significant serologic profile during the SACQ period with future flares. Similarly, we did not find any specific antibody association in patients who had multiple flares.

Figure 3.

Correlation of antinucleosome (anti-NCS) antibody titers with number of flares. OD = optical density.

All 6 patients with A flares were positive for anti-NCS antibodies and demonstrated high titers during the SACQ period. This finding was also observed in the 4 patients who had renal flares. Of the 6 patients with neurologic flares, 83% had high anti-NCS antibody levels.

DISCUSSION

For many years, the anti-dsDNA antibody assay has been regarded as the serologic gold standard in the diagnosis and assessment of disease activity in patients with SLE. In many cases, although not all cases, there is serologic and clinical concordance (11, 12). This study using the BILAG activity index to measure disease activity confirms the findings of Gladman et al that a minority of patients with SLE exhibit discordance of clinical and serologic activity (6). However, the SACQ patients described in the study by Gladman et al differ from our study population in several aspects. First, the patients who were regarded as serologically active in their study (n = 14) did not have persistently raised anti-dsDNA antibodies during the entire study duration. Second, the majority of their patients were not receiving any immunosuppressive agents (16 of 21 were taking these agents in our study). Lastly, the criteria used for the diagnosis of SLE in their patients were the preliminary American Rheumatism Association criteria (13) instead of the revised ACR criteria, which did not exist at that time.

The characteristics of SACQ patients in our study were similar to those of the SLE cohort as a whole except for a higher prevalence of arthralgia/arthritis reported in the SACQ patients. Our study demonstrated that the majority (81%) of the SACQ patients experienced flare in the subsequent 5 years after the SACQ period. This percentage is much higher than the 36% (4 of 11) recorded after a 14-year followup in the study by Walz LeBlanc et al (7). None of their patients who experienced a flare had renal or central nervous system involvement. In their study, a flare was based on clinical findings; no validated disease activity tool was used.

It is difficult to draw any direct comparisons with the study by Gladman et al (6) due to the different population groups studied and because the defined period of clinical quiescence was different in both studies. Other factors such as ethnic variation may also account for the differences seen. However, patient ethnic groups were not identified in the study population of Gladman et al (6).

Most (65%) of the SACQ patients in our study experienced a flare within the first year after the SACQ period, and half of the A flares (6 of 12) involved the neurologic system. The majority of patients in our cohort had a long duration of disease (mean duration 16 years). It would be interesting to determine in the future if patients who are SACQ early in the disease develop more flares later in the disease.

Anti-NCS antibody titers have been found to be associated with disease activity measured by the SLEDAI, particularly with renal flares (14). There are emerging data that nucleosomes play a fundamental role as the driving autoantigen in SLE (15). In the present study, all of the SACQ patients with renal flares had high anti-NCS antibody titers. Interestingly, 83% of patients with neurologic flares demonstrated positive nucleosome binding.

One of the limitations of this study was the difficulty in assessing if anti-NCS antibodies were specifically associated with a particular organ system flare, because an individual patient could have various organ systems involved with a flare. All of the SACQ patients who had A flares had high anti-NCS antibody levels. It is unknown if there is a correlation of higher anti-NCS antibody levels with severity of flares.

This study demonstrated that patients with anti-NCS antibodies had a shorter time to first flare after the SACQ period, especially if they had high titers. The only factor that was significantly associated with multiple episodes of flares was a higher absolute anti-NCS antibody titer. These results suggest that measuring anti-NCS antibodies may be a better predictor than measuring anti-dsDNA antibodies to identify the SACQ patients who will experience a flare earlier and proceed to have multiple flares. Animal models have shown that anti-NCS antibodies appear before the occurrence of anti-dsDNA antibodies in lupus mice, supporting the role of the nucleosome as a trigger factor in the disease (16).

Using Fisher's exact statistical analysis, we did not identify any serologic factors, including measuring anti-dsDNA antibodies with Crithidia luciliae immunofluorescent assay (data not shown), that aid in delineating SACQ patients who will experience flare in the future. This finding may be confounded by the small number of patients who did not experience a flare.

In conclusion, the clinician needs to be aware that a minority of patients with SLE demonstrate discordance of serologic and clinical activity. The majority of these patients will have a flare in the next 5 years and close followup is recommended, especially in the first year. There may be a role for measuring anti-NCS antibodies in addition to anti-dsDNA antibodies during this period to help identify patients who are more likely to have an earlier flare and multiple flares.

Acknowledgements

The authors would like to thank Professor Jo Berden and Dr. Johan van der Vlag (Nijmegen, The Netherlands) for their help in setting up the antinucleosome assay.

Ancillary