SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. PATIENTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgements
  8. REFERENCES

Objective

To describe the prevalence, characteristics, attribution, and clinical significance of neuropsychiatric (NP) events in an international inception cohort of systemic lupus erythematosus (SLE) patients.

Methods

The study was conducted by the Systemic Lupus International Collaborating Clinics (SLICC). Patients were enrolled within 15 months of fulfilling the American College of Rheumatology (ACR) SLE classification criteria. All NP events within a predefined enrollment window were identified using the ACR case definitions of 19 NP syndromes. Decision rules were derived to determine the proportion of NP disease attributable to SLE. Clinical significance was determined using the Short Form 36 (SF-36) Health Survey and the SLICC/ACR Damage Index (SDI).

Results

A total of 572 patients (88% female) were recruited, with a mean ± SD age of 35 ± 14 years. The mean ± SD disease duration was 5.2 ± 4.2 months. Within the enrollment window, 158 of 572 patients (28%) had at least 1 NP event. In total, there were 242 NP events that encompassed 15 of 19 NP syndromes. The proportion of NP events attributed to SLE varied from 19% to 38% using alternate attribution models and occurred in 6.1–11.7% of patients. Those with NP events, regardless of attribution, had lower scores on the SF-36 and higher SDI scores compared with patients with no NP events.

Conclusion

Twenty-eight percent of SLE patients experienced at least 1 NP event around the time of diagnosis of SLE, of which only a minority were attributed to SLE. Regardless of attribution, the occurrence of NP events was associated with reduced quality of life and increased organ damage.

Nervous system involvement as part of systemic lupus erythematosus (SLE) is well recognized, although the prevalence is highly variable among studies (1–7). A wide range of neuropsychiatric (NP) manifestations have been described, which span common features such as headache and mood disorders to rarer events such as psychosis (1–5). Although there is evidence to implicate primary immunopathogenic mechanisms in NPSLE, such as vasculopathy, autoantibodies and mediators of inflammation (8), the lack of specificity of most of the NP manifestations raises the possibility of alternative etiologies. This has important implications for the management and prognosis of individual SLE patients who present with NP events.

Differences in the reported prevalence of NP disease are likely due to a number of factors. Most studies have been performed by retrospective chart review in single academic centers and on established patient cohorts with variable disease duration. Differences in demographic characteristics, socioeconomic status, and selection bias among cohorts are additional potential confounders. Finally, the use of different classifications and definitions of NP disease and the failure to rigorously examine the attribution of NP events have been significant limitations.

In the present study, we have attempted to address these deficiencies by constituting an international, multicenter inception cohort of SLE patients. The prevalence of NP disease was determined using a standardized protocol based upon the American College of Rheumatology (ACR) nomenclature and case definitions of NPSLE (9). Decision rules were developed to determine the attribution of NP events to SLE or alternative etiologies. Our findings suggest that although NP events are common around the time of diagnosis of SLE and have a significant clinical impact, the majority of NP events are not directly attributable to lupus.

PATIENTS AND METHODS

  1. Top of page
  2. Abstract
  3. PATIENTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgements
  8. REFERENCES

Research study network.

The study was conducted by members of the Systemic Lupus International Collaborating Clinics (SLICC), which consists of 30 investigators at 27 international academic medical centers. Data were obtained prospectively on all patients presenting with a new diagnosis of SLE. All information was submitted to the coordinating center in Halifax, Nova Scotia, Canada, and entered into a centralized Access database. Appropriate procedures were instituted to ensure data quality, management, and security. Additional information on the same patients was collected concurrently as part of a study examining atherosclerosis in SLE and was submitted to the coordinating center for that study at the University of Toronto, Ontario, Canada. Electronic data transfer occurred between the Toronto and Halifax sites, and the merged data set was available for analysis. The study protocol was approved by the Capital Health Research Ethics Board in Halifax, Nova Scotia, Canada, and by the institutional research ethics review boards at each of the participating centers.

Patients.

All patients fulfilled the ACR classification criteria for SLE (10) and provided written informed consent. The date of diagnosis was defined as the time when these cumulative criteria were first recognized. Enrollment in the study was encouraged as close as possible to the date of diagnosis, but was permitted for up to 15 months following the diagnosis. Among the demographic variables that were obtained were age, sex, ethnicity, and education. Lupus-related variables included the ACR classification criteria for SLE (10), history of medication use, the SLE Disease Activity Index (SLEDAI) (11), and the SLICC/ACR Damage Index (SDI) (12) in patients whose disease duration was ≥6 months. Laboratory variables included hematologic, serum and urine chemistry, and immunologic variables required for the generation of SLEDAI and SDI scores. Health-related quality of life was measured using the Short Form 36 (SF-36) Health Survey (13).

NP events.

An enrollment window was defined within which all NP events, some of which are inherently evanescent, were captured. To ensure inclusion of NP events that may have been a component of the presentation of lupus but which occurred prior to the time the ACR classification criteria were met, the enrollment window extended from 6 months prior to the date of diagnosis of SLE up to the enrollment date. Because the latter could occur up to 15 months following the diagnosis of SLE, the maximum duration of the enrollment window was 21 months. The specific NP events that were identified within this time frame were based upon the ACR nomenclature and case definitions for 19 NP syndromes described in SLE (9). Since 4 of these 19 syndromes have subcategories, there were a total of 31 separate NP manifestations.

Screening for all NP syndromes was performed primarily by clinical evaluation, and subsequent investigations were performed only if clinically warranted. In order to further improve the consistency of data collection, a checklist of NP symptoms was distributed to each of the participating sites for use during patient encounters. In the majority of cases, the diagnosis of cognitive impairment was made on the basis of clinical assessment rather than on formal neuropsychological testing. The 8 cognitive domains that were assessed were simple attention, complex attention, memory, visual-spatial processing, language, reasoning/problem solving, psychomotor speed, and executive functions.

The occurrence of all NP events within the enrollment window was identified and additional information was recorded. The specific information depended upon the type of NP event and was guided by the ACR glossary for the 19 NP syndromes (9). This included a list of potential etiologic factors other than SLE that were identified for exclusion or recognized as an “association,” acknowledging that in some situations it is not possible to be definitive about attribution. Collectively, these “exclusions” and “associations” were referred to as “non-SLE factors” and were used in part to determine the eventual attribution of NP events. Patients could have more than 1 type of NP event, but repeated episodes of the same NP event occurring within the enrollment window were recorded only once. In the latter case, the time of the first episode was taken as the date of onset of the NP event.

Factors in NP attribution.

Participating centers were asked to report all NP events regardless of etiology and, in particular, no NP events were excluded because an individual investigator felt that these were not attributable to SLE. Decision rules were derived to determine the attribution of NP events that occurred within the enrollment window. Factors that were considered included onset of NP event(s) prior to the enrollment window, presence of concurrent non-SLE factor(s) that were identified as part of the ACR definitions of each NP syndrome and considered to be a likely cause or significant contributor to the event, and the occurrence of “minor” NP events as defined by Ainiala et al (1), who have previously reported that such events occur with high frequency in normal population controls. These latter NP manifestations include all headaches, anxiety, mild depression (i.e., all mood disorders that fail to meet the criteria for “major depressive-like episodes”), mild cognitive impairment (deficits in <3 of the 8 specified cognitive domains), and polyneuropathy without electrophysiologic confirmation. These decision rules were used to determine the attribution of NP events to SLE. Thus, the onset of an NP event prior to the enrollment window, the identification of at least 1 non-SLE factor, or the occurrence of a “minor” NP event as defined by Ainiala et al (1) classified the NP event as not attributable to SLE.

Statistical analysis.

Values are expressed as the mean ± SD, unless otherwise indicated. The relationship between the occurrence of ≥1 NP event and sex, ethnicity, age at diagnosis of SLE, education level, SF-36 scores, and SDI scores was measured by logistic regression analysis with stratification for academic center and adjustment for length of observation within the enrollment window. Separate analyses were performed for the following predefined comparisons: patients with no NP events versus patients with any NP event, patients with no NP events versus patients with NP events attributed to SLE, and patients with NP events attributed to SLE versus patients with NP events not attributed to SLE. A sensitivity analysis was performed to evaluate the impact of decision rules on attribution of NP events.

RESULTS

  1. Top of page
  2. Abstract
  3. PATIENTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgements
  8. REFERENCES

Demographic characteristics of the patients.

A total of 572 patients were recruited in 21 centers between October 1999 and March 2005. The median number of patients enrolled in each center was 19 (range 4–83). The patients were predominantly female, with a mean ± SD age of 35 ± 14 years, and a wide ethnic distribution, although the patients were predominantly white (Table 1). Forty-four percent of the patients were single and 63% had a college education. At enrollment, the mean ± SD disease duration was only 5.2 ± 4.2 months, despite the opportunity to recruit patients up to 15 months following the diagnosis of SLE. The average number of ACR classification criteria met was 4.9 ± 1.1, and the prevalence of individual criteria reflected an unselected patient population. The mean SLEDAI and SDI scores revealed moderate global disease activity and minimal cumulative organ damage, respectively. Therapy at the time of enrollment reflected the usual range of lupus medications, such as corticosteroids, antimalarials, immunosuppressants, acetylsalicylic acid, and warfarin in addition to antidepressants (11%), anticonvulsants (5%), and antipsychotic medications (1%).

Table 1. Demographic and clinical characteristics of the SLE patients*
  • *

    SLE = systemic lupus erythematosus; ACR = American College of Rheumatology; SLEDAI = SLE Disease Activity Index; SLICC = Systemic Lupus International Collaborating Clinics; ASA = acetylsalicylic acid.

No. of patients572
Sex, % female/% male88/12
Age, mean ± SD years35 ± 14
Ethnicity, % 
 White 52
 Hispanic 16
 Asian 16
 African American 13
 Other  3
Single/married/other, %44/41/15
Postsecondary education, % 63
Disease duration, mean ± SD months5.2 ± 4.2
ACR criteria fulfilled, mean ± SD no.4.9 ± 1.1
Cumulative ACR manifestations, % 
 Malar rash 37
 Discoid rash 12
 Photosensitivity 40
 Oral/nasopharyngeal ulcers 38
 Serositis 27
 Arthritis 74
 Renal disorder 29
 Neurologic disorder  5
 Hematologic disorder 61
 Immunologic disorder 76
 Antinuclear antibody 95
SLEDAI, mean ± SD score5.8 ± 5.6
SLICC/ACR Damage Index, mean ± SD score0.39 ± 0.82
Medications, % 
 Corticosteroids 68
 Antimalarials 59
 Immunosuppressants 37
 ASA 15
 Antidepressants 11
 Anticonvulsants  5
 Warfarin  3
 Antipsychotics  1

NP manifestations.

Within the enrollment window, 158 of 572 patients (28%) had at least 1 NP event and 54 of 572 (9.4%) had ≥2 events, with a maximum of 6 events. There were a total of 242 NP events, encompassing 15 of the 19 NP syndromes (Table 2): headache (38.8%), mood disorders (12.4%), cerebrovascular disease (7.9%), seizure disorders (7.9%), anxiety disorder (7.4%), cognitive dysfunction (5.4%), acute confusional state (5.0%), mononeuropathy (3.7%), polyneuropathy (3.3%), psychosis (2.9%), cranial neuropathy (2.1%), aseptic meningitis (1.2%), myelopathy (0.8%), movement disorder (0.8%), and autonomic disorder (0.4%). In patients with cognitive dysfunction, all cognitive domains were involved to varying degrees (simple attention 39%, complex attention 77%, memory 77%, visual-spatial processing 31%, language 23%, reasoning/problem solving 69%, psychomotor speed 31%, and executive functions 54%). There were no patients with Guillain-Barré syndrome, demyelinating syndrome, myasthenia gravis, or plexopathy.

Table 2. Characteristics of neuropsychiatric syndromes in systemic lupus erythematosus patients
 NP events (n = 242)*NP events with subcategories (n = 266)
  • *

    Values are the number (%) of neuropsychiatric (NP) events.

  • Values are the number (%) of NP events and subcategories of events.

Headache94 (38.8)111
 Migraine 58 (52)
 Tension 36 (32)
 Cluster 3 (3)
 Pseudotumor cerebri 2 (2)
 Nonspecific 12 (11)
Mood disorders30 (12.4) 33
 Major depression 22 (67)
 Depressive features 7 (21)
 Manic features 0 (0)
 Mixed features 4 (12)
Cerebrovascular disease19 (7.9) 21
 Stroke 13 (62)
 Transient ischemic attack 4 (19)
 Multifocal disease 2 (9)
 Subarachnoid 1 (5)
 Sinus thrombosis 1 (5)
Seizure disorder19 (7.9) 21
 Generalized 16 (76)
 Partial 5 (24)
Anxiety disorder18 (7.4) 18
Cognitive dysfunction13 (5.4) 13
Acute confusional state12 (5.0) 12
Mononeuropathy9 (3.7)  9
Polyneuropathy8 (3.3)  8
Psychosis7 (2.9)  7
Cranial neuropathy5 (2.1)  5
Aseptic meningitis3 (1.2)  3
Myelopathy2 (0.8)  2
Movement disorder2 (0.8)  2
Autonomic disorder1 (0.4)  1
Guillain-Barré syndrome0  0
Demyelinating syndrome0  0
Myasthenia gravis0  0
Plexopathy0  0

Prior to the enrollment window, 138 of 572 patients (24%) had at least 1 NP event, and 48 of 572 patients (8.4%) had ≥2 events, with a maximum of 4 events. There were a total of 202 NP events encompassing 13 NP syndromes: headache (45%), mood disorders (23%), anxiety disorder (8%), seizure disorders (7%), cerebrovascular disease (5%), cognitive dysfunction (5%), cranial neuropathy (2%), acute confusional state (2%), aseptic meningitis (1%), demyelinating syndrome (1%), polyneuropathy (1%), mononeuropathy (0.5%), and myasthenia gravis (0.5%). The time of onset of NP events prior to the enrollment window is illustrated in Figure 1. There were no patients with Guillain-Barré syndrome, autonomic disorder, movement disorder, myelopathy, plexopathy, or psychosis.

thumbnail image

Figure 1. Time of onset of neuropsychiatric (NP) events prior to the enrollment window.

Download figure to PowerPoint

Attribution of NP events.

The individual decision rules for determining attribution of the NP events revealed that 100 of the 242 NP events (41.3%) that occurred within the enrollment window had their onset prior to this predefined time frame. The mean ± SD duration between the first occurrence of these NP events and the diagnosis of SLE was 9.0 ± 9.4 years. The presence of non-SLE factors that contributed to the occurrence of NP events was identified in 76 of 242 episodes (31.4%). Of these, ≥1 “exclusion factor” was identified in 20 events, indicating that SLE was not the cause. In the remaining 56 events, only “association factors” were present, suggesting that SLE was partly responsible for the NP event. Also, 127 of 242 NP events (52.5%) were in the “minor” NP category previously identified by Ainiala et al (1). Combining all 3 decision rules for attribution (Table 3 and Figure 2) indicated that 196 of 242 NP events (81%) were deemed not to be due to SLE (model A). Thus, in the total cohort of 572 patients, 35 (6.1%) had 46 NP events that were directly attributed to SLE.

Table 3. NP events attributed to SLE using 2 attribution models*
NP manifestationEvents, no. (%)
Model AModel B
  • *

    The attribution of neuropsychiatric (NP) events to systemic lupus erythematosus (SLE) or other causes was determined using models of greater stringency (model A) or lesser stringency (model B). In model A, the onset of NP events prior to the enrollment window, the identification of non-SLE factors that contributed to or were responsible for the NP event, and the occurrence of a “minor” NP event as defined by Ainiala et al (1) were each considered to indicate that the NP event was not attributed to SLE. In model B, the onset of events >10 years before the diagnosis of SLE, the identification of non-SLE factors that were responsible for the NP event (“exclusion factors”), and the occurrence of a “minor” NP event as defined by Ainiala et al (1) were each considered to indicate that the NP event was not attributed to SLE.

Seizure disorder11 (23.9)17 (18.3)
Cerebrovascular disease7 (15.2)18 (19.4)
Mononeuropathy6 (13)9 (9.7)
Acute confusional state5 (10.9)9 (9.7)
Cranial neuropathy3 (6.5)3 (3.2)
Myelopathy2 (4.3)2 (2.2)
Polyneuropathy3 (6.5)4 (4.3)
Aseptic meningitis2 (4.3)2 (2.2)
Mood disorders2 (4.3)12 (12.9)
Psychosis2 (4.3)6 (6.5)
Movement disorder1 (2.2)1 (1.1)
Autonomic disorder1 (2.2)1 (1.1)
Cognitive dysfunction1 (2.2)9 (9.7)
Total46 (19)93 (38)
thumbnail image

Figure 2. Attribution of neuropsychiatric (NP) events to systemic lupus erythematosus (SLE) or other causes using models of greater stringency (model A) or lesser stringency (model B). In model A, the onset of NP events prior to the enrollment window, the identification of non-SLE factors that contributed to or were responsible for the NP event, and the occurrence of a “minor” NP event as defined by Ainiala et al (1) were each considered to indicate that the NP event was not attributed to SLE. In model B, the onset of NP events >10 years before the diagnosis of SLE, the identification of non-SLE factors that were responsible for the NP event (“exclusion factors”), and the occurrence of a “minor” NP event as defined by Ainiala et al (1) were each considered to indicate that the NP event was not attributed to SLE.

Download figure to PowerPoint

A sensitivity analysis was performed by making the following modifications to 2 of the 3 individual decision rules: all NP events whose onset was within 10 years prior to the diagnosis of SLE were attributed to SLE, in view of previous work suggesting that clinical manifestations of SLE may precede the diagnosis of lupus by several years (14–16); and NP events for which only “association factors” but not “exclusion factors” were identified were attributed to SLE. In this modified attribution model (model B), 149 of 242 NP events (61.6%) were deemed not to be due to SLE. In the total cohort of 572 patients, 67 (11.7%) had 93 NP events that were directly attributed to SLE (Table 3 and Figure 2).

Thus, depending upon which set of composite decision rules was used, the proportion of NP events attributed to SLE varied from 19% to 38% and affected 6.1–11.7% of patients. This variability emphasizes the impact of altering the stringency of the decision rules dealing with the attribution of NP events to SLE or alternative etiologies. In both attribution models, a total of 13 NP syndromes were represented, with seizure disorders, cerebrovascular disease, acute confusional states, and neuropathies being the most common.

Demographic variables.

Academic centers participating in the study were assembled into geographic locations (Canada, US/Mexico, Europe, and Asia). Of the 150, 230, 141, and 51 patients in the 4 regions, respectively, 50 (33%), 51(22%), 48 (34%), and 9 (18%) had NP events within the enrollment window. Results of a global test for regional differences were significant (P = 0.009). For NP events attributed to SLE according to model A, the numbers of events were 10 (7%), 20 (9%), 15 (11%), and 1 (2%) (P = 0.18), and for NP events attributed to SLE according to model B, they were 20 (13%), 37 (16%), 33 (23%), and 3 (6%) (P = 0.12). Significance test results were similar after adjustment for length of observation within the enrollment window. All logistic regression analyses reported were subsequently stratified for geographic region and length of observation within the enrollment window.

There was no association (P > 0.10) between the occurrence of NP events (overall or attributed to SLE under models A and B) within the enrollment window and patient sex, age at diagnosis, ethnicity (white, Hispanic, Asian, African American, other), or educational status (postsecondary education or not).

Health-related quality of life (HRQOL), disease activity, organ damage, and NP events.

Self-reported HRQOL at study enrollment was compared among patients with NP events attributed to SLE, patients with NP events attributed to non-SLE etiologies, and patients with no NP events. SF-36 scores were available for 361 patients. Regardless of attribution, patients with NP events had consistently lower mean scores (P < 0.05) on all of the subscales of the SF-36, indicating a lower HRQOL (Figure 3). The differences in the physical and mental SF-36 composite scores were also significant (P < 0.05 and P < 0.001, respectively) (Figure 3). The results were unchanged after additional adjustment for age at diagnosis, race, education, and SLEDAI score. There were no significant differences in scores between patients with NP events attributed to SLE or to other causes.

thumbnail image

Figure 3. Differences in health-related quality of life (HRQOL) in patients with versus those without neuropsychiatric (NP) events. Top, Differences in HRQOL as indicated by subscale scores. Bottom, Differences in HRQOL as indicated by physical and mental composite scores. Patients with NP events, regardless of attribution, had consistently lower scores, indicating poorer HRQOL. Values are the mean and SEM. SF-36 = Short Form 36 Health Survey. ∗ = P < 0.05; ∗∗ = P < 0.01; ∗∗∗ = P < 0.001.

Download figure to PowerPoint

The association between SLEDAI scores and NP disease was not significant in the 519 patients for whom a SLEDAI score was available. However, SDI scores available for the 236 patients with disease duration of >6 months were significantly higher in patients with NP events compared with those without (mean ± SD score 0.65 ± 1.10 versus 0.29 ± 0.66; P = 0.004) and were also higher in patients with NP events attributed to SLE compared with non-SLE etiologies, using attribution model A (1.29 ± 1.57 versus 0.42 ± 0.77; P < 0.001) or model B (1.06 ± 1.37 versus 0.24 ± 0.50; P < 0.001).

Because the SDI includes NP variables, the analysis was repeated following removal of NP variables from the index. As expected, this led to a reduction in effect sizes. Thus, the difference in modified SDI scores in patients with versus those without any NP event was 0.40 ± 0.83 versus 0.28 ± 0.64 (P = 0.20), and the differences between scores in patients with NP events attributed to SLE compared with scores in patients without NP events were 0.71 ± 0.99 versus 0.28 ± 0.64 (P = 0.05) (model A) and 0.63 ± 1.04 versus 0.28 ± 0.64 (P = 0.03) (model B). Due to the small number of patients (n = 65) for whom SDI scores were available, there were no significant differences between modified SDI scores in patients with NP events attributed to SLE compared with patients with NP events attributed to other causes, using either model A (0.71 ± 0.99 versus 0.29 ± 0.74; P = 0.47) or model B (0.63 ± 1.04 versus 0.18 ± 0.46; P = 0.25).

DISCUSSION

  1. Top of page
  2. Abstract
  3. PATIENTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgements
  8. REFERENCES

Although NP events are well recognized in SLE, there is uncertainty surrounding several issues, including the true prevalence of events, their attribution to SLE versus other etiologies, and their clinical significance. In the present study, we characterized all 242 NP events that occurred in 158 patients (28%) within a predefined enrollment window in an international, multicenter, multiethnic inception cohort of 572 patients. The proportion of NP events attributed to SLE varied between 19% and 38%, depending upon the decision rules for attribution. Regardless of attribution, patients with NP events had significantly lower scores on the SF-36 and higher scores for cumulative organ damage, indicating the negative impact of NP disease in SLE.

Previous studies of NP-SLE have shown a wide prevalence of NP events, varying from 37% to 95% (1–7). This variability is due in part to methodologic differences between studies. For example, until the publication of the ACR nomenclature and case definitions for NPSLE in 1999 (9), there were no universally accepted definitions for NP events, and various ad hoc classifications were used in most studies. Restriction of studies to single centers, inherent selection bias in established lupus cohorts, and variable disease duration were additional potential confounders. The correct attribution of NP events to SLE or to other causes is particularly challenging in view of the lack of diagnostic gold standards for the majority of NP events. Thus, it is not surprising that methodologic differences in determining attribution have also contributed to the lack of consensus in the literature.

The patient population in the current study was broadly representative of SLE, with an appropriate age and sex distribution, and diverse clinical manifestations. Recruitment of patients to a disease inception cohort using an established, international research network ensured ethnic diversity and short disease duration, thereby minimizing the potential confounding effects of chronic disease and long-term medication use. Using the ACR nomenclature and case definitions for NP syndromes (9) permitted a standardized approach across centers. As expected, there was a wide array of NP events, with representation in 15 of the 19 NP categories. Despite the short disease duration, 28% of patients had at least 1 NP event, and multiple NP events were observed in almost 10% of patients.

Because previous studies of NPSLE have usually excluded those NP events that were considered not to be due to lupus, it has not been possible to examine the clinical impact of all NP events in SLE patients regardless of attribution. Therefore, a specific aim of our study was to document all NP events occurring around the time of diagnosis of SLE, in order to examine the attribution of events in a systematic manner and to assess the overall impact on HRQOL.

One factor in determining attribution is the time of onset of the NP event in relation to the diagnosis of SLE. Thus, an interval from 6 months prior to the diagnosis of SLE up to study enrollment was chosen in order to capture all NP events surrounding the diagnosis of SLE that were potentially attributable to the disease. A less stringent approach included all NP events that occurred within 10 years prior to the diagnosis of SLE. This interval was selected in view of previous studies that examined the accrual of clinical manifestations of SLE (16) and the onset of NP events (15) prior to the diagnosis of SLE. Arbuckle et al (14) found that lupus autoantibodies and, less frequently, clinical manifestations of SLE may occur up to 10 years prior to the diagnosis of the disease. When considering nonlupus causes for NP events, the ACR case definitions (9) provide a mechanism for the systematic identification of factors other than SLE that may be contributing to the etiology. Depending upon the strength of the etiologic association between these factors and NP syndromes, they are regarded as alternative diagnoses (“exclusions”) or contributors (“associations”) to the event.

Finally, we incorporated the recommendations of Ainiala et al (1), whose study of SLE patients and randomly matched healthy population controls suggested that headaches, anxiety, mild depression, mild cognitive impairment, and peripheral neuropathies lacking confirmation by electrophysiologic studies may not be primary NP manifestations of SLE due to their high frequency in normal population controls. For example, 80% of their SLE patients and 28% of their normal controls were impaired in at least 1 of 8 cognitive domains. However, these prevalence rates fell to 24% and 4%, respectively, when mild cognitive impairment was excluded. This decision rule for cognitive impairment in SLE patients is similar to that used in a previous study (17), with comparable findings for the prevalence of cognitive impairment in SLE patients and controls. A sensitivity analysis performed in the current study assessing all 19 NP syndromes indicated the profound effect, in terms of frequency, of the use of different decision rules for timing of the onset of the NP event and identifying alternative diagnoses. Thus, depending upon which decision rules were used, 19–38% of all NP events were attributed to SLE and affected 6.1–11.7% of patients.

Previous studies have examined the morbidity and mortality associated with NPSLE in established lupus cohorts. One report described significantly lower scores on a generic self-report measure of HRQOL in patients with NP events, regardless of attribution; lower scores were not seen in patients with a history of renal disease in the same patient cohort (3). In another study, Jonsen et al (18) reported a higher frequency of disability in SLE patients with NP disease compared with patients without NP events and the general population. In contrast, cognitive impairment has not always been associated with excess morbidity (19–21). Although some studies have shown increased mortality in patients with NP disease (22–26), others have not (7, 15, 27, 28).

The present study is the first to assess the clinical impact of NP events in a large, international inception cohort. There were 2 associations with NP events, which indicated a negative clinical impact. First, patients with NP events had significantly lower scores on the SF-36 subscales and summary scores, which, consistent with a previous study (3), was independent of the attribution of the event. Second, the occurrence of NP events was associated with increased organ damage, albeit less impressive when neurologic variables were excluded. Because damage scores were available only in patients with a minimum disease duration of 6 months, which is a requirement for computing the SDI, this association will require confirmation in a larger sample. Nevertheless, collectively these results indicate that even within a few months of diagnosis, nervous system disease is associated with a significant adverse clinical impact.

There are a number of limitations to the present study. First, restriction of NP syndromes to the 19 identified in the ACR case definitions (9) could potentially have excluded other forms of nervous system disease. However, this did not emerge as an issue during the execution of the study. In fact, 4 of the 19 NP syndromes were never identified in this relatively large inception cohort. A few patients presenting with very severe NP manifestations may have been excluded due to their inability to provide informed consent or because they died prior to study enrollment. It is difficult to compare the lower overall frequency of NP events attributed to SLE in the present study with previous and usually higher estimates of NPSLE, since the latter studies were not usually performed around the time of inception of SLE and did not use the same rigor to exclude other causes of NP disease.

Second, formal neuropsychological assessments were not performed routinely on all patients, largely for logistical reasons. There is currently no validated, universally accepted screening tool for cognitive impairment in SLE. The Modified Mini–Mental Status examination is not sufficiently sensitive and, hence, eliciting a history of cognitive difficulties is the initial step in the detection of cognitive impairment (29). If formal neuropsychological assessment had been included in the study protocol, this would likely have increased the prevalence of cognitive impairment identified in our cohort, although the additional impairment identified would be subtle and subclinical in the majority of cases. However, several cross-sectional and longitudinal studies have indicated that such deficits do not adversely affect HRQOL (19–21) or lead to long-term, clinically significant neurologic sequelae (19, 30–32). Also, given that formal neuropsychological assessments are not part of routine followup, we believe our protocol reflects clinical practice.

Finally, the study did not include structural or functional neuroimaging in all patients, although specific diagnostic tests were performed as required to meet the criteria for certain individual NP syndromes. Again, this approach mirrors clinical practice, and a requirement for more stringent testing, which would likely have resulted in lower enrollment and selection bias, was avoided.

In summary, we identified NP syndromes in 28% of patients in an international inception cohort of SLE patients. Of these NP events, up to 38% were attributed to SLE. Regardless of attribution, the occurrence of NP events is associated with reduced HRQOL and increased organ damage. Further followup will determine the evolution of these events over time and whether they continue to be associated with adverse clinical outcomes following the institution of lupus-specific therapies.

AUTHOR CONTRIBUTIONS

  1. Top of page
  2. Abstract
  3. PATIENTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgements
  8. REFERENCES

Dr. Hanly 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 design. Drs. Hanly, Urowitz, Sanchez-Guerrero, Bae, Gordon, Isenberg, Merrill, Dooley, Gladman, Fortin, Steinsson, Khamashta, Van Vollenhoven, and Farewell.

Acquisition of data. Drs. Hanly, Urowitz, Sanchez-Guerrero, Bae, Gordon, Wallace, Isenberg, Alarcón, Clarke, Bernatsky, Merrill, Petri, Dooley, Gladman, Fortin, Steinsson, Bruce, Manzi, Khamashta, Zoma, Aranow, Ginzler, Van Vollenhoven, Font, Sturfelt, Nived, Ramsey-Goldman, and Kalunian.

Analysis and interpretation of data. Drs. Hanly, Gordon, Isenberg, Alarcón, Bernatsky, Merrill, Gladman, Manzi, Van Vollenhoven, Ramsey-Goldman, Thompson, and Farewell.

Manuscript preparation. Drs. Hanly, Sanchez-Guerrero, Gordon, Wallace, Isenberg, Alarcón, Bernatsky, Merrill, Bruce, Manzi, Aranow, Ginzler, Van Vollenhoven, Ramsey-Goldman, and Farewell.

Statistical analysis. Dr. Hanly, Ms Thompson, and Dr. Farewell.

Database design and creation. Ms Douglas.

Acknowledgements

  1. Top of page
  2. Abstract
  3. PATIENTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgements
  8. REFERENCES

We are grateful for the generous donation of our patients' time and the dedication of all the research coordinators and research assistants in the SLICC network to the completion of this work.

REFERENCES

  1. Top of page
  2. Abstract
  3. PATIENTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgements
  8. REFERENCES