Steroid induced psychosis in systemic lupus erythematosus: a possible role of serum albumin level
- Department of Autoimmune Diseases, Institut Clínic d'Infeccions i Immunologia, Hospital Clínic, Institut d'Investigacions Biomédiques August Pi i Sunyer (IDIBAPS), University of Barcelona, Barcelona, Catalonia, Spain
- Correspondence to:
Dr R Cervera, Servei de Malalties Autoimmunes, Hospital Clínic, Villarroel 170, 08036-Barcelona, Catalonia, Spain;
- Accepted 10 December 2001
Steroids may have diverse and sometimes severe adverse effects in the short and long term.1 We present three patients with systemic lupus erythematosus (SLE)2 and steroid induced psychosis (table 1), emphasising the importance that serum albumin levels may have on the development of this complication.
Patient No 1 is a 20 year old woman with SLE diagnosed five years ago, in whom a serum albumin level of 24 g/l and proteinuria of 3.2 g/l were detected in routine tests. Diffuse proliferative lupus nephritis was diagnosed by renal biopsy and she was treated with one pulse of cyclophosphamide (500 mg) and oral prednisone (60 mg/day). Three days later she developed anxiety, insomnia, euphoria, verbosity, grandiosity, and megalomaniac ideas. She was treated with oral risperidone (2 mg/12 h), oral clonazepam (0.5 mg/12 h), and the prednisone dosage was progressively tapered. Over the next 15 days she experienced a fluctuating but progressive improvement until she became psychiatrically asymptomatic. Five years previously, when she was first diagnosed as having SLE, she had been treated with oral prednisone (60 mg/day) but had not had any psychiatric symptoms. At that time, however, she had serum albumin levels of 33 g/l without proteinuria.
Patient No 2, a 21 year old woman who was diagnosed as having SLE, with cutaneous, articular and renal involvement, started treatment with oral prednisone (30 mg/day). At that time she had a serum albumin level of 30.2 g/l and proteinuria of 2.37 g/day. Renal biopsy was refused by the patient. Over the following days and in a progressive manner, she developed mania, with euphoria, disinhibition as well as ideas of grandiosity. All these manifestations disappeared under psychiatric supervision and when steroids were discontinued.
Patient No 3 is a 36 year old woman who had been diagnosed as having SLE with diffuse proliferative lupus nephritis seven years before. She had started oral prednisone (60 mg/day), without adverse psychiatric effects. At that time, she had serum total protein level of 66 g/l and a serum albumin level of 43 g/l. She currently presented with an articular “flare” of her lupus, and started treatment with oral prednisone (30 mg/day). Her total protein level was 59 g/l, the serum albumin level fell to 29 g/l with proteinuria of 1.2 g/l. Seven days later, she developed anxiety, insomnia, and hyperactivity that disappeared over the following few days after reduction of the dose of prednisone to 2.5 mg/day.
Steroid induced psychiatric disturbances appear in 3–6% of the patients who are treated with these drugs.3–6 The differential diagnosis with lupus psychosis7 is difficult. In case of doubt, some authors advocate increasing the dose of steroids and awaiting a clinical response over the next days. Others advocate rapid tapering and stopping steroids in order to eliminate a drug induced adverse event.
Our three patients developed psychiatric symptoms while receiving steroids for lupus nephritis with hypoalbuminaemia and significative proteinuria. Moreover, it seems relevant that two of them (patients 1 and 3) had not developed psychiatric manifestations when they had previously been receiving similar doses of steroids, but had higher serum albumin levels at that time. Lewis et al found a 37% incidence of adverse effects of steroids in those patients with serum albumin <25 g/l but of only 15% in those with higher serum albumin.8 The explanation for these findings may be that corticosteroid binding globulin does not bind to synthetic steroids, whose transport depends on serum albumin which, by contrast, presents low affinity but a great capacity to transport the steroids because of its high plasma concentration. Steroids are biologically inactive when bound to albumin. Therefore, the free (and active) fraction of steroids is higher in patients with low plasma albumin levels, and this will expose the patient to more adverse effects.
Interestingly, there is a lower incidence of psychosis in other groups of patients treated with steroids (for example, those with chronic obstructive pulmonary disease).9 Therefore, those patients whose disease causes low levels of serum proteins (as in those with SLE) will be more predisposed to have adverse effects of steroids.
F López-Medrano is currently at Servicio de Medicina Interna, Hospital Universitario 12 de Octubre, Madrid, Spain.