Background: In order to make treatment decisions physicians should have knowledge about the relations between patient characteristics and drug disposition. Dose, route of administration, gender, age and renal function are reported to influence the serum concentrations of morphine, morphine-6-glucurnide (M6G) and morphine-3-glucuronide (M3G) during chronic treatment of cancer pain. These factors, however, are not evaluated in studies with a sample size sufficient to explore predictive factors.
Methods: Three hundred consecutive morphine users admitted because of a malignant disease were recruited. The relations of serum concentrations of morphine, M6G and M3G to patient characteristics (gender, age, weight, renal function, liver function, dose, route of administration) were explored, and regression analysis performed to investigate whether these characteristics predicted serum concentrations obtained during routine clinical drug monitoring.
Results: Morphine dose was associated with serum concentrations of morphine (r = 0.69), M6G (r = 0.76) and M3G (r = 0.76). Oral morphine resulted in higher dose-adjusted M6G and M3G serum concentrations compared with s.c. morphine. Creatinine serum concentrations correlated with serum concentrations of M6G and M3G. Dose and route of administration predicted morphine serum concentrations, while dose and renal function predicted M6G and M3G serum concentrations. Age was an additional factor predicting M3G concentrations. Dose was the only factor that explained a clinically significant part of the observed variability.
Conclusion: Patient characteristics predict only minor parts of the variability of morphine, M3G and M6G serum concentrations observed during routine clinical drug-monitoring in cancer patients.