Odom-Maryon T L, Williams L E, Chai A, Lopatin G, Liu A, Wong Y C, Chou J, Clarke K G, Raubitschek A A
Department of Biostatistics, City of Hope National Medical Center, Duarte, California 91010, USA.
J Nucl Med. 1997 Dec;38(12):1959-66.
The objective of this article was to model pharmacokinetic data from clinical diagnostic studies involving the 111In-labeled monoclonal antibody (MAb) chimeric T84.66, against carcinoembryonic antigen. Model-derived results based on the 111In-MAb blood, urine and digital imaging data were used to predict 90Y-MAb absorbed radiation doses and to guide treatment planning for future therapy trials. Fifteen patients with at least one carcinoembryonic antigen-positive lesion were evaluated. We report the kinetic parameter estimates and absorbed 111In-MAb dose and projected 90Y-MAb doses for each patient as well as describe our approach and rationale for modeling an extensive set of pharmacokinetic data.
The ADAPT II software package was used to create three- and five-compartment models of uptake against time in the patient population. The "best-fit" model was identified using ordinary least squares. Areas under the curve were calculated using the modeled curves and input into MIRDOSE3 to estimate absorbed radiation doses for each patient.
A five-compartment model best described the liver, whole body, blood and urine data for a subcohort of nine patients with digital imaging data. A three-compartment model best described the blood and urine data for all 15 clinical patients accrued in the clinical trial. For the subcohort, the largest projected 90Y-MAb doses were delivered to the liver (mean, 24.78 rad/mCi; range, 15.02-37.07 rad/mCi), with red marrow estimates on the order of 3.32 rad/mCi (range, 1.24-5.55) of 90Y. Corresponding estimates for the 111In-MAb were 3.18 (range, 2.09-4.43) and 0.55 (range, 0.34-0.74), respectively.
The three- and five-compartment models presented here were successfully used to represent the blood, urine and imaging data. This was evidenced by the small standard errors for the kinetic parameter estimates and R2 values close to 1. As planned future therapeutic trials will involve stem cell support to alleviate hematological toxicities, the development of an approach for estimating doses to other major organs is crucial.
本文的目的是对涉及针对癌胚抗原的111铟标记单克隆抗体(MAb)嵌合T84.66的临床诊断研究中的药代动力学数据进行建模。基于111铟-MAb血液、尿液和数字成像数据的模型推导结果用于预测90钇-MAb吸收的辐射剂量,并指导未来治疗试验的治疗计划。对15例至少有一个癌胚抗原阳性病变的患者进行了评估。我们报告了每位患者的动力学参数估计值、111铟-MAb吸收剂量和预测的90钇-MAb剂量,并描述了我们对大量药代动力学数据进行建模的方法和原理。
使用ADAPT II软件包为患者群体创建摄取随时间变化的三室和五室模型。使用普通最小二乘法确定“最佳拟合”模型。使用建模曲线计算曲线下面积,并输入MIRDOSE3以估计每位患者的吸收辐射剂量。
五室模型最能描述9例有数字成像数据的亚组患者的肝脏、全身、血液和尿液数据。三室模型最能描述临床试验中累积的所有15例临床患者的血液和尿液数据。对于该亚组,预测的90钇-MAb最大剂量输送到肝脏(平均值,24.78拉德/毫居里;范围,15.02 - 37.07拉德/毫居里),红骨髓估计值约为90钇的3.32拉德/毫居里(范围,1.24 - 5.55)。111铟-MAb的相应估计值分别为3.18(范围,2.09 - 4.43)和0.55(范围,0.34 - 0.74)。
本文提出的三室和五室模型成功用于表示血液、尿液和成像数据。动力学参数估计值的小标准误差和接近1的R2值证明了这一点。由于计划中的未来治疗试验将涉及干细胞支持以减轻血液学毒性,开发一种估计其他主要器官剂量的方法至关重要。