The Toxicity of Cytarabine

Abstract

The principal toxicity of standard induction regimens for acute non-lymphocytic leukaemia (ANLL) [including cytarabine (ARA-C) 100 mg/m² for 7 days plus an anthracycline] is myelotoxicity, leading to death in at least 25% of cases during induction in non-selected patients. The complete remission rate is less than 35% in patients over 65 years of age, due in part to an age-related increase of myelotoxicity. The other important adverse effect of standard-dose cytarabine is gastrointestinal toxicity, especially oral mucositis, diarrhoea, intestinal ulceration, ileus and subsequent Gram-negative septicaemia. Idiosyncratic reactions like exanthema, fever and elevation of hepatic enzymes are relatively frequent, but do not represent therapeutic problems. Intermittent high-dose cytarabine (3 g/m² in 8 to 12 doses) is extremely myelosuppressive. Similarly, the gastrointestinal toxicity is formidable and dose-limiting. Severe, and sometimes irreversible, cerebellar/cerebral toxicity in 5 to 15% of courses of treatment limits the peak dose of cytarabine. The pathogenesis, prophylactic and therapeutic measures are unknown. These major toxicities are age-related and prohibitive to the use of high-dose cytarabine therapy in patients older than 55 to 60 years. Subacute noncardiogenic pulmonary oedema occurs in some patients, with an incidence of about 20%, and seems to have an intriguing coincidence with precedent streptococcal septicaemia; high-dose systemic steroids may be beneficial. Corneal toxicity is very frequent in high-dose cytarabine therapy but is always reversible. It is largely preventable with prophylactic steroid or 2-deoxycytidine eyedrops. Fever, exanthema and hepatic toxicity have an incidence similar to that in standard dosage. The maximum tolerable cumulated dose of cytarabine is significantly lower when the agent is administered as a continuous infusion, due to myelosuppression and gastrointestinal toxicity. Conversely, continuous infusion may be less neurotoxic. The antileukaemic effect of continuous infusion high-dose cytarabine is less well established. The only significant toxicity of low-dose cytarabine is myelosuppression. Given the generally poor condition of leukaemia patients, low-dose cytarabine therapy is well tolerated, although occasional cases of diarrhoea, reversible cerebellar symptoms, peritoneal and pericardial reactions, and ocular toxicity have been reported. Continuous infusion may be more toxic than the usual intermittent dosage. It is concluded that the toxicity of the standard induction regimen for ANLL is acceptable in patients younger than 60 to 65 years with no concurrent disease. Low dose cytarabine is tolerable for virtually all ANLL patients, but the overall therapeutic efficacy still needs to be defined and compared to standard therapy in the relevant age groups. High-dose cytarabine is valuable, even though it is extremely toxic, in relapsed and refractory acute leukaemia in patients younger than 55 to 60 years. It is not recommended as induction therapy, due to its toxicity. The regimen seems to be slightly more tolerable as consolidation therapy but cases of severe toxicity with lethal outcome cannot be avoided; doses of 0.5 to 1.0 g/m² may have comparably favourable pharmacological properties with less toxicity, but need more clinical documentation.