Incidence of ovarian failure in systemic lupus erythematosus after treatment with pulse cyclophosphamide.

OBJECTIVE To investigate the incidence of ovarian failure after pulse cyclophosphamide treatment in systemic lupus erythematosus (SLE) and to compare this with two control groups: SLE patients treated with azathioprine, and a healthy age matched population. METHODS All women patients with SLE treated with pulse cyclophosphamide in our department were identified and questioned concerning menstrual history. All the hospital notes were reviewed and details recorded on dose of cyclophosphamide, duration of treatment, side effects and lowest pretreatment neutrophil and leucocyte counts during the course of treatment. Disease controls were recruited from our department and healthy controls from the local family health services authority (FHSA) register. RESULTS Incidence of ovarian failure in the premenopausal cyclophosphamide treated group was 54% and the incidence of premature menopause (occurring before age 40 years) was 41%. Increasing age at start of treatment showed a linear trend with incidence of ovarian failure (p = 0.01). Using logistic regression, increasing duration of treatment was related to incidence of ovarian failure (p = 0.047 in those treated age 35 years or younger). An association between the lowest neutrophil count throughout the treatment period, when taken immediately before each planned cyclophosphamide pulse, and the incidence of ovarian failure was also demonstrated (p = 0.04 in those treated before age 40 years). CONCLUSION Ovarian failure--in particular, premature failure after treatment with pulse cyclophosphamide--is common. Factors associated with increased risk include greater age at start of treatment, longer period of treatment, and greater degree of marrow suppression as assessed by the neutrophil count immediately before each planned cyclophosphamide pulse.