Significance of Chronic Tachycardia in Systemic Lupus Erythematosus

Authors


Abstract

Objective

A significant subset of systemic lupus erythematosus (SLE) patients exhibit chronic tachycardia (CT) of unknown significance. We postulated that CT is a marker of lupus activity and severity.

Methods

A cross-sectional database at the University of Chicago recorded disease activity, damage, disease manifestations, pain, anxiety, and physical function (PF). CT was defined as a heart rate of ≥95 beats per minute in at least 3 out of 4 sequential visits. Demographic, disease-specific, and self-reported symptoms were compared between groups with and without tachycardia.

Results

Of the 243 subjects analyzed, 14.8% had CT. On univariate analysis, CT was associated with younger age at the time of enrollment (P = 0.004), number of hospitalizations adjusted for years of SLE (P = 0.001), current prednisone dose (P < 0.0001), history of serositis (P = 0.03), anxiety score (P = 0.004), and poor PF (P = 0.0017). All domains of the Short Form 36 (SF-36) health survey correlated strongly with CT, but on multivariate regression this correlation appeared to be driven by poor PF. On multivariate regression, the Systemic Lupus Erythematosus Disease Activity Index score (P = 0.03), younger age (P = 0.04), and poor PF by the SF-36 domain (P = 0.006) were independently correlated with CT, and anxiety trait and hemoglobin both trended closely to significant association (P = 0.08 for both).

Conclusion

CT is prevalent in SLE and is a clinically relevant physical finding. It implies greater lupus activity and physical frailty. Univariate association with serositis raises the possibility of subclinical serositis or pancarditis. Further study to elucidate the cardiopulmonary status of SLE patients with unexplained CT is planned.

Introduction

Systemic lupus erythematosus (SLE) is a chronic autoimmune disorder that manifests in a variety of ways. Sinus tachycardia is more prevalent in those with SLE when compared with the general population (1–3). This was first noted by Hejtmancik et al in 1964, despite adjustment for other causes of tachycardia (1). Early studies suggested that sinus tachycardia may be a manifestation of disease activity (4, 5). In 1987, Mandell noted tachycardia in SLE patients during disease flares, which resolved with treatment (4). In 1994, Guzmán et al discovered a correlation between tachycardia and Mexican Systemic Lupus Erythematosus Disease Activity Index (SLEDAI) scores (5). However, the sample size was small and there has been no subsequent work on this topic.

Tachycardia may, at times, represent occult myocarditis or pericarditis (4). Mandell suggested that, based on echocardiographic evidence of pericarditis and myocarditis, this organ inflammation may often be subclinical in SLE (4). Similarly, Doherty and Siegel calculated a clinical and autopsy incidence of pericarditis to be 25% and 62%, respectively (6). Myocarditis was clinically evident in less than 10%, but had an autopsy prevalence of 40% (6).

In addition to intermittent tachycardia, we have noticed chronic tachycardia (CT) among our ambulatory SLE patient population. The significance of this finding is not known. Because of frequent comorbidities in SLE, the finding of CT may precipitate an extensive medical evaluation. The differential diagnosis of chronic sinus tachycardia is broad, including severe anemia, cardiovascular disease, thyroid disorders, hypoxemia, anxiety, pain, and medications. Better understanding of CT in SLE would inform our clinical assessment of these patients. Conversely, the presence of CT may inform the physician about clinical status and/or prognosis of the SLE (2). Herein, we hypothesized that CT correlates positively with disease activity and severity.

Patients and methods

Patients fulfilling American College of Rheumatology (ACR) criteria for the classification of SLE (7) were enrolled during ambulatory visits at the University of Chicago rheumatology clinic in a cross-sectional lupus clinical database from June 2008 through October 2011. Near total enrollment (>95%) of patients with SLE receiving longitudinal lupus care by the authors was achieved during the study enrollment interval.

Definitions.

CT was defined by a resting heart rate of ≥95 beats per minute during 3 out of 4 consecutive office visits. The 4 office visits spanned from the day of enrollment to the previous 3 visits. The cutoff of 95 beats per minute is based on the distribution of heart rates in normal populations (8). Patients were considered to have CT despite a normal heart rate if they were explicitly under treatment for chronic sinus tachycardia. However, use of drugs that affect the heart rate for other indications was not exclusionary. Patients with known hyperthyroidism were excluded. The pulse was measured, using a mechanized blood pressure cuff, when the patient was brought to the clinic room. CT was then compared with 4 major variables: disease activity, disease severity, cardiac disease, and psychosocial measures. Each of these 4 major categories contained specific variables.

Variables.

Disease activity was assessed by the SLEDAI score (9), hemoglobin, anti–double-stranded DNA (anti-dsDNA) titers, complement levels (C3 and C4), history of serositis, number of hospitalizations per year of SLE duration, and the patient's dose of prednisone at the time of enrollment. Disease severity was assessed by the Systemic Lupus International Collaborating Clinics/ACR Damage Index (SDI) (10), renal involvement by ACR criteria, and highest prednisone dose ever used. Cardiac disease was assessed by the presence of atherosclerotic disease and/or cardiomyopathy. Atherosclerotic disease was defined as a history of coronary artery disease, peripheral vascular disease, cerebrovascular disease, and/or transient ischemic accident. Cardiomyopathy was defined as congestive heart failure, myocarditis, or pulmonary hypertension. Psychosocial parameters were assessed by a measure of trait anxiety (State-Trait Anxiety Inventory; STAI) (11), the patient's perception of their health (Medical Outcomes Study Short Form 36 [SF-36] health survey) (12), and by a 100-mm visual analog pain scale characterizing pain over the week prior to database enrollment.

At the time of study enrollment, patients were examined, interviewed, and their charts reviewed to determine heart rates, demographics, ACR criteria for SLE (7), medical history, medication usage, and hospitalization history, as well as SLEDAI (9) and SDI (10) scores. Patients completed the self-reported health outcomes SF-36 survey (12) and the STAI (11) within a week of enrollment. Retrospectively, paper charts were reviewed to confirm the accuracy of the primary outcome measure of CT, by reviewing heart rates in the subjects' clinic notes from the time of enrollment and adjacent appointments. Laboratory data were generally obtained on the day of enrollment or, if not available, the nearest available laboratory result was used.

Statistical analysis.

Data were entered into a Microsoft Access database. Statistical analysis was performed using Stata, version 11.2. Descriptive statistics included frequencies as well as mean ± SD values or median (interquartile range) values of variables based on data characteristics.

The concordance between CT and sex, race, and several other categorical variables representing disease activity and severity was assessed by chi-square analyses and Fisher's exact tests. CT and all other continuous variables were compared using t-tests and Wilcoxon's rank sum tests. In all cases, a 2-sided P value was reported using an alpha value of 0.05 to define significance.

A multivariate logistic regression analysis was then performed with the outcome of CT. The model was based on a forward stepwise selection, retaining covariates with a P value ≤0.10. The model was run sequentially adding age at enrollment, age at diagnosis of SLE, STAI score, the presence of serositis, prednisone dose at the time of enrollment, cardiomyopathy, hemoglobin, number of hospitalizations corrected for SLE duration, SLEDAI score, and physical function (PF) score from the SF-36.

All patients in this study signed a written informed consent form, and the study was approved by the Institutional Review Board of the University of Chicago.

Results

A total of 245 subjects were enrolled from June 2008 through October 2011 in the outpatient rheumatology clinic. Two subjects did not have sufficient data to determine whether or not they met our criteria for CT. Thus, the study group consisted of 243 subjects. Clinical and demographic data are presented in Table 1. Females represented 92.9% of the cohort. African Americans were the largest ethnic group, making up 60.9% of the sample. The mean ages at enrollment and at the time of SLE diagnosis were 46.2 years and 31.3 years, respectively. The mean heart rate was 84.9 beats per minute. The mean SLEDAI score and median SDI score in this ambulatory population were 4.1 and 1.0, respectively. The subjects in the study had a median of 0.25 hospitalizations per year of SLE duration. Serositis had been present in 48.6% of the population. Cardiomyopathy was present in 15.2% of subjects, atherosclerotic disease in 18.1%, and renal disease in 38.0%.

Table 1. Baseline characteristics of study group (n = 243)*
CharacteristicNo.Value
  • *

    SLE = systemic lupus erythematosus; IQR = interquartile range; SLEDAI = Systemic Lupus Erythematosus Disease Activity Index; SDI = Systemic Lupus International Collaborating Clinics/American College of Rheumatology Damage Index.

  • Congestive heart failure, pulmonary hypertension, or myocarditis.

  • Coronary artery disease, peripheral or cerebrovascular disease, or transient ischemic accident.

Age, mean ± SD years23846.3 ± 13.3
SLE duration, median (IQR) years24312 (8–20)
Heart rate, mean ± SD beats per minute24284.9 ± 14.7
SLEDAI, mean ± SD2334.1 ± 3.8
SDI, median (IQR)2371 (0–3)
Hospitalizations/year of SLE, median (IQR)1930.25 (0.12–0.6)
Female, % 92.9
Ethnicity, %233 
 White30.5
 African American61.37
 Hispanic4.7
 Asian/Pacific Islander3.4
Chronic tachycardia, %24314.8
Cardiomyopathy, %24315.2
Renal disease, %24338.3
Atherosclerotic disease, %24318.1

Age, race, blood pressure, and SLE duration.

Of the 243 subjects, 36 (14.8%) had CT. CT was significantly more common at a younger age at enrollment (40.4 years versus 47.3 years; P = 0.004) and in those subjects with an earlier age at diagnosis (27.2 years versus 32.1 years; P = 0.04). However, SLE disease duration, race, and blood pressure did not differ between those with and without CT (Table 2).

Table 2. Characteristics of groups with and without chronic tachycardia (CT)*
VariableCT groupNon-CT groupP
  • *

    Values are the mean ± SD or median (interquartile range) unless indicated otherwise. SLE = systemic lupus erythematosus; SLEDAI = Systemic Lupus Erythematosus Disease Activity Index; SDI = Systemic Lupus International Collaborating Clinics/American College of Rheumatology Damage Index; anti-dsDNA = anti–double-stranded DNA; SF-36 = Short Form 36 health survey; PCS = physical component summary.

Age at diagnosis, years27.2 ± 14.832.1 ± 12.60.04
Age at enrollment, years40.4 ± 15.047.3 ± 12.70.004
Systolic blood pressure, mm Hg122.3 ± 3.3125.5 ± 17.50.31
SLE duration, years12.8 ± 9.515.0 ± 9.70.21
SLEDAI5.2 ± 4.43.9 ± 3.70.09
SDI1 (1–3)1 (0–3)0.56
Hemoglobin, gm/dl11.7 ± 2.312.3 ± 1.70.07
No. of hospitalizations/year of SLE0.7 (0.3–1.0)0.2 (0.11–0.5)0.001
Anti-dsDNA antibody, titer10 (0–60)10 (0–40)0.62
C3 complement, mg/dl109.6 ± 39.9105 ± 28.60.52
C4 complement, mg/dl25.2 ± 16.723.1 ± 12.40.38
Prednisone at enrollment, mg/day15.1 ± 14.27.4 ± 13.8< 0.0001
Serositis, ever %66.745.90.03
Renal involvement, ever %36.138.60.85
Cardiomyopathy, %2513.50.08
Atherosclerotic disease, %16.718.41.00
Visual analog pain scale, mm49.6 ± 32.233.6 ± 28.50.06
State-Trait Anxiety Inventory46.8 ± 8.839.6 ± 10.70.004
SF-36 PCS29.9 ± 10.539.9 ± 11.70.0017

SLE activity measures.

Higher current doses of prednisone correlated with CT (15.1 versus 7.4 mg; P < 0.0001). History of serositis was also more frequent in those subjects with CT (66.7% versus 45.9%; P = 0.03). There was a trend towards higher SLEDAI scores (5.2 versus 3.9; P = 0.09) and lower hemoglobin values (11.7 versus 12.3 gm/dl; P = 0.07) with respect to CT. Other markers of disease activity (anti-dsDNA, complement levels, erythrocyte sedimentation rate, and C-reactive protein level) did not differ between groups (Table 2).

SLE severity measures and cardiac disease.

The number of hospitalizations per year of SLE duration was significantly more frequent in subjects with CT (0.7 versus 0.2 hospitalizations per year of SLE; P = 0.001). Cardiomyopathy was more common in the CT group, but did not reach statistical significance (25.0% versus 13.5%; P = 0.08). SDI scores, highest prednisone dose ever, history of renal disease, and atherosclerotic disease did not correlate with CT (Table 2).

Psychosocial/self-reported measures.

STAI scores were higher in subjects with CT, indicating a higher level of anxiety trait (46.8 versus 39.6; P = 0.004) (Table 2). Pain levels were elevated in those with CT, approaching significance (49.6 versus 33.6; P = 0.06). Lower SF-36 physical component summary scores, representing poorer overall patient-reported PF, were strongly associated with CT (29.9 versus 39.9; P = 0.0017).

Multivariate analysis.

Forward stepwise regressions were performed on CT with covariates of interest (Table 3). Age at SLE diagnosis, STAI score, history of serositis, prednisone dose, cardiomyopathy, hemoglobin, number of hospitalizations per year of SLE, SLEDAI score, and the SF-36 PF domain were sequential covariates. Age at diagnosis of SLE, serositis, current prednisone dose, cardiomyopathy, and number of hospitalizations were no longer significant and were dropped from the regression. Age at enrollment (odds ratio [OR] 0.92, 95% confidence interval [95% CI] 0.85–1.00, P = 0.04); SLEDAI scores (OR 1.26, 95% CI 1.02–1.55, P = 0.03); and PF scores (OR 0.87, 95% CI 0.79–0.96, P = 0.006) remained statistically significant. There were trends toward significance with STAI scores (OR 1.09, 95% CI 0.99–1.19, P = 0.08) and hemoglobin (OR 0.62, 95% CI 0.36–1.05, P = 0.08).

Table 3. Multivariate analysis model of chronic tachycardia status*
VariableOR95% CIP
  • *

    OR = odds ratio; 95% CI = 95% confidence interval; SLEDAI = Systemic Lupus Erythematosus Disease Activity Index; STAI = State-Trait Anxiety Index; SF-36 = Short Form 36 health survey.

SLEDAI1.261.02–1.550.03
Age at enrollment, years0.920.85–1.000.04
STAI1.090.99–1.190.08
Hemoglobin, gm/dl0.620.36–1.050.08
SF-36 physical function0.870.79–0.960.006

Discussion

In this study, we examined CT for the first time in a population of ambulatory SLE patients. We found that CT, defined as ≥95 beats per minute, occurred in approximately 15% of our population. This was a similar frequency to that of atherosclerotic disease in our cohort. However, CT did not correlate with atherosclerosis. There was a trend toward correlation with the cardiomyopathy variable (P = 0.08). A history of serositis was a univariate correlate of CT (P = 0.03), but renal involvement or overall damage score was not. Demographically, younger patients were more likely to have CT, but CT was unrelated to SLE duration.

Multiple disease activity measures correlated with CT on univariate analysis, only to become nonsignificant on multivariate regression adjusting for current SLEDAI score. The SLEDAI score trended toward significant correlation on univariate analysis (P = 0.09) and became significant on multivariate analysis (P = 0.03). Current prednisone dose strongly correlated on univariate analysis (P < 0.0001), but became nonsignificant on multivariate regression. The number of hospitalizations per year of SLE duration was chosen as an activity measure to reflect clinical instability of the patient. On univariate analysis, this hospitalization variable correlated strongly with CT (P = 0.001), but on multivariate regression this association disappeared. Similarly, serositis history correlated with CT only on univariate analysis. Surprisingly, standard laboratory measures of disease activity such as complement and dsDNA titer did not correlate with CT.

Among self-reported symptoms, pain level during the week prior to enrollment did trend toward correlation with CT on univariate analysis (P = 0.06), but did not persist on multivariate regression. In contrast, chronic anxiety symptoms associated significantly with CT on univariate analysis (P = 0.004) and trended toward significance on multivariate analysis (P = 0.08). The PF domain of the SF-36 was still strongly associated with CT in the multivariate regression, even after correction for disease activity, age, anxiety, and hemoglobin (Table 3). This suggests that physical frailty independently correlates with CT, rather than simply reflecting current disease activity.

While we are the first to study CT in the lupus patient, our findings generally support those of Guzmán et al, who assessed acute tachycardia (5). In that study, the authors measured resting heart rate and Mexican SLEDAI scores during each office visit, and tachycardia was strongly predictive of active disease (8). While Guzmán et al found heart rates to correlate with acute disease activity, we feel that CT may reflect sustained disease activity. However, because this is a cross-sectional study, which is vulnerable to survivorship bias, we cannot yet say if the marker of CT indicates a distinct refractory disease group.

In the setting of new-onset tachycardia, the differential diagnosis is broad and includes volume depletion, thyroid dyscrasias, intrinsic cardiac pathology, and severe anemia. When the tachycardia is longstanding, however, these conditions are unlikely. The pathophysiology of CT in SLE is not well understood, but one potential explanation is autonomic dysfunction. This dysfunction tends to disproportionately affect the parasympathetic nervous system, hypothetically increasing likelihood of high sympathetic tone (3, 13, 14).

In summary, this is the first study to examine CT in SLE. The most significant findings were an association with disease activity, impaired PF, and young age. While it is limited by cross-sectional design, the data were uniformly collected and the primary variable (CT) independently confirmed by chart review. Based on the association with disease activity and poor PF, one might suspect that, in a prospectively designed study, the variable CT may be a clinically useful prognostic marker. Other limitations included our inability to detect subclinical serositis or myocarditis as a cause for the CT. However, we have shown that CT does correlate with active SLE and should increase suspicion for subclinical pericardial or myocardial inflammation if no other causes are apparent. Advanced cardiac imaging techniques such as cardiac magnetic resonance imaging or positron emission tomography scans may be useful in future studies of CT.

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 submitted for publication. Dr. Utset 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. Utset, Ward.

Acquisition of data. Utset, Ward, Thompson, Green.

Analysis and interpretation of data. Utset, Ward.

ADDITIONAL DISCLOSURE

Dr. Utset is a consultant for Genentech, Human Genome Sciences/GlaxoSmithKline, and Anthera Pharmaceuticals, but received no consulting fees or grants related to this project. Genentech, Human Genome Sciences/GlaxoSmithKline, and Anthera Pharmaceuticals had no financial interest in this project and had no input in the design, content, data collection, or analysis, and had no role in the writing or approval of this article, with all opinions and conclusions expressed herein those of the authors.

Ancillary