Abstract |
Introduction
Neuropsychiatric systemic lupus erythematosus (NPSLE) poses a diagnostic and therapeutic
challenge. An American College of Rheumatology (ACR) research committee has published a set of
case definitions for 19 NPSLE syndromes, in an effort to homogenize terminology for research and
clinical practice purposes. These case definitions involve both the central and the peripheral nervous
system, are categorized into focal and diffuse and have a wide heterogeneity that ranges from overt
manifestations such as stroke, seizures and psychosis, to headache or cognitive dysfunction.
Attribution of neuropsychiatric events to lupus warrants a thorough investigation and exclusion of
alternative causes. Diagnostic workup and treatment decisions are typically performed on a patientby-
patient basis and often necessitate the involvement of multiple medical specialties. In an effort to
homogenize the management of patients with NPSLE, a EULAR task force has issued a set of
recommendations addressing diagnostic and therapeutic issues, using a combination of evidencebased
approach and expert consensus. A validation or comparison of these recommendations with
routine clinical practice has not been performed.
One particular neuropsychiatric manifestation included in the ACR nomenclature for NPSLE is
demyelinating syndrome (termed lupoid sclerosis in the past). However, distinction of this entity
from frank multiple sclerosis (MS) is not clear, given the recent advances in MS diagnostics, which
aim to increase sensitivity in diagnosing the disease.
Aims of the Thesis
For the purpose of this Thesis, we performed a comprehensive study of NPSLE in two European
centres (with the cooperation of a EULAR scholar, Dr. Cristina Pamfil from «Iuliu Hatieganu»
University of Pharmacy and Medicine, Cluj-Napoca, Romania). More specifically we:
• analyzed demographic, clinical and neuroimaging data from all “primary” NPSLE cases from
Heraklion and Cluj
• compared routine clinical practice against the EULAR recommendations for NPSLE to unveil
potential pitfalls and limitations
• evaluated treatment options and long-term outcome of NPSLE - analyzed in more detail
patients that received cyclophosphamide (CYC) for severe neuropsychiatric manifestations, using a
structured approach to assess response
• identified SLE patients with clinical and neuroradiological features of demyelination and
classified them as SLE-associated demyelinating syndrome or coexistence of SLE with frank MS, by
diagnostic criteria.
SUMMARY
6
Patients and methods
Two national tertiary referral centres for patients with SLE and suspected neuropsychiatric
involvement, Heraklion, Greece and Cluj, Romania participated in the study.
• For the characterization of the NPSLE cohort, SLE patients with confirmed “primary”
neuropsychiatric involvement were selected by retrospective chart review from all lupus cases over
the last 15 years. All patients fulfilled at least four of the revised American College of Rheumatology
(ACR) classification criteria for SLE at the time of NPSLE diagnosis and had undergone regular
follow-up in each centre. For each neuropsychiatric manifestation included, we recorded all
diagnostic procedures the patients underwent and the therapies they received. We then compared the
diagnostic and therapeutic decisions applied, against the EULAR recommendations for NPSLE (both
the general ones and those specific to the event).
• To assess the efficacy and safety profile of CYC in NPSLE, we identified “primary” NPSLE
cases that received CYC for their neuropsychiatric syndrome and documented all variables relating
to dosing, route of administration and cumulative dose, outcome and duration of follow-up, as well
as occurrence of serious adverse events.
• For the characterization of patients with SLE and demyelinating features, we scrutinized our
NPSLE cohorts and also utilized data from the independently established cohort of MS in the
Neurology Clinic of the University Hospital of Heraklion, to identify potential patients with features
of both diseases. Identified cases were followed up with combined rheumatologic/neurologic
evaluation on a regular basis at 3–6 month intervals, depending on disease activity. We also
reviewed the English language literature using the PubMed database from 1966 to January 2013 with
the following index terms: “multiple sclerosis” OR “myelitis” OR “myelopathy” OR
“demyelination” AND “SLE” OR “lupus” (terms present in title or abstract).
Results
• Characterization of the NPSLE cohort and comparison of usual clinical care with the EULAR
recommendations: We identified 94 patients who experienced a total of 123 lupus-related
neuropsychiatric events. Approximately 35% of events occurred within the first year after SLE
diagnosis. Most prevalent events were cerebrovascular diseases (CVD) (n=21, 17.1%), cognitive
dysfunction (n=18, 14.6%), intractable lupus headache and mood disorder (n=12 each, 9.8%).
Brain MRI was performed in 75 neuropsychiatric events (61.0%). In 21 of them (28.0%), MRI was
considered normal; in the remaining cases, the most common finding was non-specific
periventricular white matter hyperintensities (WMHIs, 40.8%), followed by cerebral infarcts
(21.1%). Treatment included steroids (either initiation or escalation of previous dose) in 89 cases
(72.4%) and immunosuppressives in 73 cases (59.3%). Antithrombotic therapy was administered in
41 neuropsychiatric events, most commonly in ischemic CVD.
7
We found overall satisfactory concordance rates between usual care and the EULAR
recommendations, with level of agreement reaching 68.7% for diagnostic work-up and 62.7% for
treatment decisions. In a post-hoc analysis, we did not observe statistically significant differences in
terms of agreement with the EULAR recommendations, when neuropsychiatric events were stratified
according to the time period (prior to or after 2010, year of publication of the EULAR
recommendations) they occurred. Despite this good concordance, we identified a number of issues
such as overutilization of brain MRI (42.9% of neuropsychiatric events with no such
recommendation), suboptimal evaluation for cognitive dysfunction (less than 30% of patients
underwent formal neurocognitive assessment) and frequent use of immunosuppressives in CVD
disease (52.4% of cases received immunosuppression in addition to antiplatelets/anticoagulants).
• Efficacy and safety of CYC for NPSLE: CYC was administered in 50 neuropsychiatric events.
Most frequent indications were psychosis (12 cases), polyneuropathy (6 cases), and cerebrovascular
disease, seizure disorder and cranial neuropathy (5 cases). CYC was mainly administered as monthly
pulses (median number: 8.0, median cumulative dose: 7.2 gr). Cases were followed for a median of
46.5 months. At last follow-up, partial or complete response of NPSLE was observed in 84% of
events; 10% had stable disease, whereas the remaining 6% failed to improve or worsened and were
rescued with rituximab. Relapses were observed in six events (12%) at median 8 months after initial
response. No malignancies were observed, yet there were three cases of severe infections.
Amenorrhea was recorded in three patients, who had not received gonadal protection.
• Characterization of SLE patients with demyelinating features: Our cohort of NPSLE patients
included patients with myelopathy and optic neuropathy, however no patients qualified for the ACR
definition of SLE-associated demyelinating syndrome. On the contrary, scrutinization of both SLE
and MS cohorts identified nine patients who fulfilled the diagnostic criteria for both SLE and MS.
This corresponded to prevalence rate 1.0-1.2% in each cohort. (SLE and MS). Initial presentation of
MS included spinal symptoms in seven patients. All patients had features of mild SLE with
predominantly cutaneous, mucosal and musculoskeletal manifestations. Accordingly, therapeutic
decisions were mainly guided by the severity of the neurological syndrome. During median followup
of 4 years, three patients remained stable and the remaining experienced gradual deterioration in
their neurological status. SLE remained quiescent in all patients while on standard
immunomodulatory MS therapy.
The systematic literature search identified detailed reports of nine cases of SLE and MS coexistence.
Unlike our patients who carried a mild SLE phenotype, cases from the literature tended to have more
severe SLE, with three patients having at least one major manifestation including CNS, renal, and
severe hematologic disease.
Conclusions
We characterized the cohorts of NPSLE patients in two European experienced centres and attempted
to juxtapose real-life management of SLE patients with neuropsychiatric manifestations with the
EULAR recommendations, and identify areas that may require additional attention. Notably, the time
period of our study predominantly included events that occurred before the publication of the
EULAR recommendations in 2010. In this regard, the overall good concordance rates between usual
care and the recommendations and the absence of a significant difference in this concordance
between events occurring prior and after publication of the recommendations is a reassuring
observation, as the management of NPSLE has traditionally been based on expert opinion.
Nevertheless, despite good concordance between EULAR recommendations for NPSLE and usual
clinical practice, we identified a number of issues such as overutilization of brain MRI, suboptimal
evaluation for cognitive dysfunction and frequent use of immunosuppressives in cerebrovascular
disease that need to be further investigated.
Regarding the efficacy of CYC in NPSLE, our observations confirm the efficacy of pulse CYC inthis
situation, since more than 80% of events demonstrated at least moderate improvement from their
baseline status during the follow-up period. Our finding of higher response rates in cases where CYC
was given as 1st line therapy could imply that intense immunosuppression is more efficacious if
instituted early in NPSLE. Not withstanding the retrospective nature of our data, in cases wherein
gonadotoxicity is not a major concern, pulse CYC should not be withheld in severe NPSLE.
Finally, we did not find cases of “demyelinating syndrome” in our cohort of NPSLE patients.
Instead, using our hospital-based SLE and MS cohorts at the University of Crete, we identified
patients with SLE who additionally fulfilled the diagnostic criteria for MS and described the
prevalence, diagnosis, treatment, and prognosis of cases that have both diseases. We found that
coexistence of the two disorders reaches an estimated point prevalence of about 1% among patients
with SLE or MS. The combination of these findings suggests that, given the high sensitivity of new
diagnostic criteria for MS, the concept of “MS-like” syndrome in SLE may need to be reevaluated,
since it may actually represent overlap of two diseases. Patients with SLE-MS overlap in our
experience tend to have mild SLE without major extra-CNS organ involvement, which does not
require intensive immunosuppressive treatment. MS tends to follow a relapsing-remitting course
(frequent relapses), yet with minimal accumulation of disability and its clinical severity dictates the
choice of immunomodulating agents.
|