Central Nervous System Disease in Systemic Lupus Erythematosus Therapy and Prognosis

JOHN S. SERGENT, M.D.* MICHAEL D. LOCKSHIN, M.D. MARK S. KLEMPNER, M.D.+ BENJAMIN A. LIPSKY. M.D.t New York, New York

From the Rheumatic Disease Service, The Hospital for Special Surgery, affiliated with the New York Hospital-Cornell University Medical Center and Cornell University Medical College, New York, New York 1002 1. This study was supported in part by the John A. Hartford Foundation, by U.S. Public Health Service Grant AM 05683, and by the Hospital for Special Surgery Vasculitis Fund. Requests for reprints should be addressed to Dr. Michael D. Lockshin, The Hospital for Special Surgery, 535 East 70th Street, New York, New York 10021. Manuscript accepted June 27, 1974. Postdoctoral Fellow, Arthritis Foundation. + Present address: Department of Medicine, Massachusetts . General Hospital, 32 Frult Street, Boston, Massachusetts 02114. * Present address: Department of Medicine, Unlverslty of Washington Medical Center, 1959 NE Pacific, Seattle, Washington 98 195. l

644

May 1975

The effect of corticosteroid therapy in 28 patients with 52 episodes of neuropsychiatric disease in systemic lupus erythematosus (SLE) was evaluated. Categories of organic central nervous system disease were seizures (eight patients), organic brain syndromes (nine patients), aseptic meningitis (four patients) and a variety of focal neuroiogic findings (seven patlents). Fourteen patients had 15 episodes of functional psychosis without other evldence of neurologic disease. Although there was a general correlation between clinical and serologic evidence of active SLE and the development of organic neuroiqic disease, there was no evidence that therapy with very large doses of corticosteroids was beneficial. Of the deaths in this series, two were’ due to probable active SLE involving the central nervous system whereas five were attrlbutable to compllcations of therapy. The long-term morbidity, likewise, was high in the patients who received large doses of corticosterolds. In all, 12 patients had major complications of corticosteroid therapy. Functional psychosis was usually precipitated by corticosterold therapy and responded to a reduction In steroid dosage and administration of psychotropic drugs. Central nervous system dysfunction is a common, difficult and frequently emergent problem in systemic lupus erythematosus (SLE) [ 1,2]. Although the clinical features have been well described [ 1,3], and several recent studies have helped to define the pathology and pathogenesis [3-61, the therapeutic implications of central nervous system disease in SLE remain unclear. Therapy for isolated manifestations of central nervous system disease in SLE has been suggested on the basis of anecdotal experience [7-g], and Bennett et al. [7] and Dubois [lo] have recommended the use of large doses of corticosteroids for central nervous system disease. However, previous studies have not detailed the response to therapy or the long-term prognosis of patients who have had episodes of central nervous system dysfunction. We review the experience of a rheumatic disease service with psychiatric and organic central nervous system dysfunction in patients with SLE. The responses of the patients to various modes of therapy and their present status were evaluated. Our experience reveals a diversity of clinical manifestations, a high immediate mortality, long-range neurologic morbidity (only partly dependent on other system involvement) and unimpressive evidence of the effi-

The American Journal of Medlclne

Volume 58

CENTRAL NERVOUS SYSTEM DISEASE IN SYSTEMIC LUPUS ERYTHEMATOSUS-SEROENT

ET AL.

cacy of treatment with large doses of corticosteroids. These data may serve as guidelines with which future therapeutic regimens may be compared.

nent, their use has been noted. For convenience, all corticosteroid doses have been converted to equivalent doses of prednisone.

METHODS

RESULTS

Patient Selection and Follow-Up. Patients with typical SLE manifesting abnormaiiiies of central nervous system dysfunction and actively followed since June 1970 by the Rheumatic Disease Croup at Cornell University Medical College have been prospectively noted; two additional patients seen by us elsewhere before 1970 and currently followed by us are also included. All but one has been seen by one of us (M.D.L.). Because of a diversity of referral practices, the denominator from which these patients have been drawn is unknown. In 1973, charts of these patients were reviewed and details of illness, laboratory findings, therapy and current neurologic status noted. All but two of the living patients are currently being followed or have been reexamined. The patient population is our total experience with central nervous system disease in SLE during this time with two exceptions: one of infectious meningitis and one of diphenylhydantoin-induced SLE.

Patients and Episodes.

DEFtNlTtONS SLE. All patients but one fulfilled the American Rheumatism Association criteria for SLE [ 111; all had positive antinuclear antibody or anti-DNA tests and multisystem disease. Central Nervous System Dysfunction. This is defined as unequivocal evidence .on physical examination of neurologic deficit attributable to central nervous system disease or a change in personality sufficient to require treatment and/or hospitalization. “Organic” and “psychiatric” as used in this paper are not precise terms but are consistent with conventional use except that a psychiatric manifestation accompanied by any objective neurologic sign or laboratory abnormalit (cerebrospinal fluid, brain scan, angiogram) is arbitrarily classified as organic. Episode. An episode was defined as an organic or psychiatric event resulting in a qualitative therapeutic decision (increase, decrease or maintain steroid dose) for that event. If an organic episode was preceded or succeeded by a psychiatric episode, the episodes were considered separately. The outcome was noted separately for each episode. Recovery implied full return to the premorbid neuropsychiatric state of the patient. Any residual deficits or intervening complications were noted. OBSERVATIONS

AND ADJUSTMENT

OF THERAPY

All patients in this study were seen by the Rheumatic Disease Group on frequent rounds, and therapy was decided by group evaluation and discussion on an ad hoc basis. Therapy was decided on the basis of the entire clinical Status of the patient with due regard to hypertension, infection, diabetes, thrombocytopenia and other system involvement. Although our main purpose was to evaluate the effect of various corticosteroid regimens on central nervous system manifestations of SLE, other medications including immunosuppressive drugs, psychotropic medications and antihypertensives were freely used. When perti-

37 episodes

Twenty-eight patients had of organic central nervous system dis-

ease and 15 episodes of psychiatric disease (Tables I through VI). Seventeen patients were white, 9 black and 2 yellow: 17 were under 30 years of age. All were female. Mortality was greatest in whites over

30 years of age and least in blacks under 30 years of age. CLINICAL

MANIFESTATIONS

Seizures. Episodes of grand mal seizures occurred 10 times in 8 patients. AH these patients had clinical and laboratory evidence of active SLE at the time of their seizures, and five of the eight had coexistent hypertension, azotemia or both (Table I). Seven of the eight patients were treated with large doses of prednisone plus anticonvulsant medications.

In one patient (R.L.) seizures, followed by chorea, developed shortly after the administration of 60 mg/day of prednisone. Because of a prior similar response to prednisone, the dosage was gradually reduced, and the patient made an uneventful recovery. Four patients in this group died. One patient (C.T.) was in an apparent remission, without clinical signs of active disease, when status epilepticus developed as her initial organic central nervous system manifestation. By the time she reached the hospital, she had been anoxic for a prolonged period, and she never regained consciousness. Although she received prednisone in large doses, it was probably of no consequence.

The other three deaths in this group occurred in patients who were treated with 200 mg/day or more of prednisone; in all three, death was due to infection. In all four patients who survived, recovery was complete, without any neurologic residua. Psychosis 3es of prednisone followed the institution c in three patients. Thus, in the 10 episodes observed,

5 patients were treated with 200 mg/day of prednisone or more, and three of these patients died of complicating infections. Of the five treated with smaller doses of prednisone, four recovered: in the only patient who died (C.T.), death was probably not preventable after she arrived at the hospital. Patlent R.B. A 20 year old black woman entered the hospftal with a 6 month history of erythematous rash, myalgias, arthralgias and fever. Exudates in the fundus and pain on palpation of her muscles were noted on physical examination. She was also thought to be somewhat dull mentally, and bilateral extensor plantar responses were present. Laboratory studies showed mild anemia and thrombocytopenia; a test for antinuclear antibodies was May 1975

The

Am&can Journalof Medlclna Volume 58

645

CENTRAL NERVOUS SYSTEM DISEASE IN SYSTEMIC LUPUS ERYTHEMATOSUS-SERGENT

TABLE I

Patient

Seizures Possible Contributing Factors

R.B.

...

J.M.

...

C.T.

...

Serum Complement*

Anti-DNA Antibodiest (%)

122

- 63

ND

ND

Therapy of Episode

Prednisone, 40 mg/day

Prednisone, 200 mg/day

Died of infection

None

Prednisqne, 200 mg/day

Transient psychosis, then complete recovery

Prednisone, 10 mg/day

Prednisone, 75 mg/day

Died of irreversible brain damage after prolonged status epilepsy

~_____

6.4