Session: Slide Session: Human Pharmacodynamics of Effect and Toxicity
Monday, October 27, 2008: 12:00 AM
Room: Constitution B (Grand Hyatt)
Background: Data suggest that higher V doses increase N. No study has examined the con-time profile of V in pts to solidify the linkage between V exposure & N. This study examined relationship between V AUC0-24h & incidence of N in infected pts. Methods: Study period: 1/05-1/06. Inclusion criteria: 1) ≥ 18 yrs, 2) non-neutropenic, 3) V for > 48 h, 4) ≥ 1 V level, 5) baseline Scr <2 mg/dL, 6) non-cystic fibrosis pts, 7) no IV contrast dye within 7 d, & 8) no vasopressors at start of V. Demographics, co-morbidities, and tx data were collected. N was defined as an increase in Scr of 0.5mg/dl or 50%, whichever was greater, post V initiation. V PK parameters were estimated by the MAP Bayesian procedure in ADAPT II. After the Bayesian step, the estimated V SCL was use to calculate AUC0-24h (mg*h/L) for each pt. Logistic regression (LR) was used to establish the relationship between V AUC0-24h and N. CART was used to identify the range of V AUCs associated with an increased probability of N. Results: 182 pts met the study criteria. After the Bayesian step, the estimated mean (SD) values for CLslope and CLintercept were: 0.03 (0.02) L*min/h*ml and 0.34 (0.22) L/hr, respectively. In LR, a significant relationship (p = 0.02) between V AUC0-24h and N was observed (figure). In CART, AUC0-24hr > 2260 was associated with N incidence 53.4%. Conclusions: A significant relationship between V AUC and N was observed.