Non‐parametric pharmacokinetic method for assessment of the variance of lifespan distribution of blood cells

Backround The description of lifespan distribution of blood cells (RBC, platelets) is commonly limited to its mean that can be calculated from the initial slope of the survival curve. The standard deviation (SD) can be determined from the area under survival curve. Similar descriptors (slope, area) are used in non‐compartmental analysis of pharmacokinetic data. These methods were adopted to assess the SD of lifespan distribution for RBC and platelets. Methods Linear and log‐linear methods of interpolation of cell survival data were applied. The initial slope and area under interpolated curves were calculated according to known formulas. The methods were validated on hypothetical data generated by Monte Carlo simulations. The SD of platelet (3 healthy and 4 idiopathic thrombocytopenic purpura (ITP) patients) and RBC (8 healthy and 4 congenital non‐spherocytic hemolytic anemia (NSHA) patients) were determined. All data were obtained from published results. Results Both methods yielded acceptable precision and bias that increased with the level of noise in survival data. The mean (SD) RBC lifespan for normal subjects were 115 (30) days and for NSHA patients 29 (27) days. The mean (SD) platelet lifespan for normal subjects were 7.3 (6.1) days and for ITP patients 1.8 (1.5) days. Conclusions The SD of blood cell lifespan distribution can be adequately determined from survival data using model‐independent methods. The blood diseases affect not only the mean but also the SD of blood cell lifespan. Clinical Pharmacology & Therapeutics (2005) 77 , P91–P91; doi: 10.1016/j.clpt.2004.12.241