Virtually every critically ill patient in an intensive care unit (ICU) receives potent psychoactive medications such as sedatives and opiate analgesics to relieve pain and discomfort. During critical illness, older patients are often given the same high doses of these sedatives and analgesics that younger patients receive. There are virtually no data about the relationships between dose or concentration of these drugs and clinical outcomes to allow ICU teams to titrate therapy appropriately. We have documented that sedatives and analgesics contribute adversely to the risk of delirium and long-term cognitive impairment (LTCI) if exposure is not optimized. Older patients are particularly prone to the risks posed by sedative and analgesic exposure. A driving unmet need of our ongoing NIA-sponsored R01 called the BRAIN-ICU study is to investigate the complicated pharmacokinetic (PK) and pharmacodynamic (PD) relationships of these medications in ICU patients by developing the best statistical methods in order to define optimal therapy. Our preliminary work has shown us that there are statistical challenges that must be overcome. First, only a small number of blood samples can be drawn for measurement of drug concentrations for patient safety. With such sparse sampling, mean plasma concentration and area under the plasma concentration curve do not provide a good measure of drug exposure. Second, we need to avoid biased estimates of PK parameters that might result from artifactual outlying values due to inadvertent blood sampling from a vein or line contaminated with infused drug as occurs in the real life complex ICU setting. Thus, we will develop and use a novel and robust population PK and PK/PD model that can accommodate sparse sampling and the presence of artifactual outliers, and allow us to examine the associations between exposure to sedative and analgesic medications and adverse outcomes related to drug exposure such as delirium and LTCI. In addition, it will identify patient characteristics altering kinetics, and subpopulations with different rates of metabolism, which would support a larger scale of pharmacogenetic study. Therefore, this work will provide a basis for intervention to optimize doses of potent medications in order to reduce acute and long-term brain dysfunction and will have far reaching implications, ultimately leading to individualized therapy with sedatives and analgesics.
There is increasing evidence in older (and younger) patients that sedatives and analgesics, medications nearly universally provided to critically ill patients in Intensive Care Units (ICUs), are leading risk factors for delirium and long-term cognitive impairment (LTCI) if patients'exposure is not optimized. We propose to develop, for the first time, a novel population PK and PK/PD model that can accommodate both the presence of patients'inter-individual differences in metabolism due to variations in organ function and genetic factors and artifactual outliers in drug levels through samples obtained in a large NIA-sponsored observational study. Using the novel method, this important study will examine the associations between exposure to sedative and analgesic medications and outcomes due to drug exposure such as development of delirium and LTCI that are so profoundly disabling older patients during critical illness and preventing their complete recovery.
|Choi, Leena; Crainiceanu, Ciprian M; Caffo, Brian S (2013) Practical recommendations for population PK studies with sampling time errors. Eur J Clin Pharmacol 69:2055-64|
|Kohli, Utkarsh; Pandharipande, Pratik; Muszkat, Mordechai et al. (2012) CYP2A6 genetic variation and dexmedetomidine disposition. Eur J Clin Pharmacol 68:937-42|
|Birdwell, Kelly A; Grady, Ben; Choi, Leena et al. (2012) The use of a DNA biobank linked to electronic medical records to characterize pharmacogenomic predictors of tacrolimus dose requirement in kidney transplant recipients. Pharmacogenet Genomics 22:32-42|
|Choi, Leena; Caffo, Brian S; Kohli, Utkarsh et al. (2011) A Bayesian hierarchical nonlinear mixture model in the presence of artifactual outliers in a population pharmacokinetic study. J Pharmacokinet Pharmacodyn 38:613-36|