Retrospective analysis of the use of cyclosporine in myasthenia gravis

Patients and methods.

We reviewed the records of all patients with MG followed in the DUMC MG clinic who took cyclosporine between 1987 and 1999. MG was diagnosed on clinical, electrophysiologic, and immunologic criteria. The severity and distribution of weakness at the time cyclosporine was begun was expressed using the MG Foundation of America (MGFA) Clinical Classification (table 1). The initial dosage was 5 mg/kg/day divided into two doses taken 12 hours apart. The dose was initially adjusted to maintain a trough level between 100 and 150 ng/L, and was subsequently adjusted on the basis of serum creatinine levels, clinical improvement, or both. After clinical improvement was maximum, the dose was reduced to the minimum necessary to maintain maximum improvement. The clinical change after beginning cyclosporine was expressed using the MGFA Post-Intervention Status.6 We performed single fiber electromyography (SFEMG) and acetylcholine receptor (AChR) antibody titers were measured at the Mayo Medical Laboratory, Rochester, MN.

Results.

Among 386 patients seen from 1987 until January 1999, 75 took cyclosporine (table 2). Sixteen were excluded from this analysis because of incomplete medical records (2), lack of regular follow-up (12), or discontinuation of therapy before 6 months for financial reasons (2). Two patients were unable to tolerate cyclosporine: a 70-year-old woman had worsening of hypertension, headaches, and tremors, and a 44-year-old man had new onset migraine headaches.

Fifty-seven took cyclosporine for 6 months to 10 years (mean 3.5 years) beginning from 8 weeks to 38 years (mean 4 years) after onset of MG symptoms.

Clinical improvement was noted as early as 1 week after beginning cyclosporine and in most cases between 4 and 12 weeks. Patients without bulbar weakness reached the best clinical improvement faster (median 5 months) than those with bulbar weakness (median 8 months). No significant correlation was found between the duration of disease or age and the time to best response. The two patients with purely ocular MG reached the best response at 4 and 8 weeks, respectively. In 24 patients (44%), the cyclosporine dose was decreased to 3 mg/kg/day or less without worsening.

Jitter was measured by SFEMG in the forearm extensor digitorum communis muscle before and at least 6 months after beginning cyclosporine in 17 patients. The average jitter value before cyclosporine treatment was 139 μsec and after treatment was 60 μsec. In all cases jitter fell more then 10% from the pretreatment value.7

AChR binding antibody levels were measured before beginning cyclosporine and after the best clinical response had been achieved in nine patients: the mean antibody level fell from 15.4 to 6.6 nm/L (57%).

Discussion.

This was a retrospective analysis of the clinical course and side effects in a large and heterogeneous group of patients with MG taking cyclosporine with long-term follow-up. Treatment with cyclosporine was followed by improvement in virtually all patients regardless of age, gender, or severity or duration of disease. Improvement began soon after beginning therapy and became maximal after a median of 7 months. This is longer than the 3.6 months previously reported and most likely reflects longer follow-up.1 Improvement occurred in patients who had failed to improve with azathioprine, suggesting that cyclosporine is an alternative for patients with MG who need long-term immunosuppression. In 85% of patients taking prednisone when cyclosporine was begun, the prednisone dose was subsequently significantly decreased or discontinued without worsening of MG. Renal toxicity and hypertension are the major factors limiting the use of cyclosporine. Thus, it is suitable only for compliant patients who have regular follow-up and frequent renal function tests. The most severe renal toxicity was seen in patients over 60 who had taken cyclosporine for several years. In such patients it is therefore crucial to reduce the dose as much as possible after the best clinical response has been reached. In 44% of our patients the cyclosporine dose was reduced considerably from the starting dose without compromising the clinical improvement. These observations suggest that cyclosporine may be effective at lower doses than are used in transplant patients or in the original MG trials once the best clinical response has been reached.

Four patients developed skin cancer. The role of cyclosporine in skin cancer is difficult to establish, but this observation emphasizes the need for frequent skin surveillance and aggressive sunscreen protection in patients taking cyclosporine. Two patients had other malignancies—one uterine carcinoma and one lymphoma. Recent studies suggest that cyclosporine can promote cancer progression in vitro and in vivo by production of tumor growth factor–β, independent of its effect on the host immune system.8 This raises concerns about the already recognized potential of cyclosporine to increase the incidence of cancer. This risk must be balanced carefully against the need to treat a potentially disabling or life-threatening disease such as MG.