Drug Concentration Errors Related to Sampling from Silastic Catheters

To the Editor: Constant infusion of some antitumor agents appears to improve their efficacy and/or diminish adverse effects associated with their administration. Silastic catheters placed for long-term central access facilitate this type of therapy. Also, they provide a means to obtain blood with minimal trauma for the patient. During treatment it is sometimes useful to withdraw samples for assay of drug blood levels. When doing so, we became concerned about unusually high 5-fluorouracil (5-FU) values in several patients, causing us to measure drug concentration in samples taken simultaneously from the catheter and a peripheral vein. Catheter samples showed seven to 80 times greater concentration of 5-FU. In one individual who had been off infusion for three days, a small 5-FU concentration persisted in the catheter sample whereas the peripheral vein value was zero as expected. Subsequently, we performed an in vitro study. A Broviac catheter (volume, 0.6 mL) was perfused for five days at 2.0 mL per hour with a solution containing 500 mg 5-FU made up to 60 mL with 0.9% saline. The catheter then was flushed with 30-mL water to remove all fluid containing 5-FU. Ensuing sampling and flushing was done in accordance with Table 1. The last milliliter of flush solution was analyzed in each instance by high-pressure liquid chromatography technology. Since all of the 5-FU infusion solution had been removed, the presence of drug in the water wash indicates that 5-FU was passing into the catheter lumen from the silastic wall. We believe it is important to recognize that other drugs may act in the same way. The use of silastic catheters for both infusion and sampling may lead to an overestimation of drug concentration even though sufficient blood is withdrawn before taking the definitive sample. Please see the PDF for Table.