Copyright © 2009 John Wiley & Sons, Ltd. Biomed. Chromatogr. 2010; 24: 544–549 Research Article Received 8 July 2009, Accepted 21 July 2009 Published online in Wiley Interscience: 8 September 2009 (www.interscience.wiley.com) DOI 10.1002/bmc.1324 Infuence of number of calibration standards within a defned range on pharmacokinetic disposition—case studies with omeprazole and clopidogrel carboxylic acid Henry John Baptist D’Souza 1 , Balakrishna Pai 1 , Anil Kumar 1 , Radha Shekar 1 , Nuggehally R. Srinivas 2 and Fjalar Kristjansson 1 * ABSTRACT: While the practice of using a smaller number of non-zero standards (typically seven to eight) has not been enter- tained in routine bioanalytical work, it is important to innovate and be pragmatic about minimizing the number of calibration standards to promote cost-efective and speedy assessment. In this exercise, two important compounds, omeprazole and clopidogrel carboxylic acid, were considered. Additionally, both analytes ofered a 1000-fold calibration curve range with eight non-zero standards to permit a systematic evaluation. Accordingly various scenarios of post-hoc analysis of the calibra- tion data were formulated which included step-wise reduction of the number of calibration standards from a maximum of n = 8 to a minimum of n = 3. In all the scenarios evaluated in this exercise, a calibration curve was reconstructed and both quality control samples and in vivo pharmacokinetics were calculated in each instance. Based on the data generated in this exercise, a minimum of three non-zero calibration standards were adequate to predict the quality control samples with the predefned accuracy and precision estimates for both omeprazole and clopidogrel carboxylic acid. Additionally, the in vivo pharmacoki- netic characterization of the chosen compounds was not hampered by the reduction of calibration standards (from n = 8 to n = 3). Hence, consideration for reducing number of calibration standards in bioanalytical work may provide a viable alternative in several situations such as formulation screening strategies, routine therapeutic drug monitoring and sparse sample analy- ses. Copyright  2009 John Wiley & Sons, Ltd. Keywords: calibration curve; regression; omeprazole; clopidogrel carboxylic acid; pharmacokinetics * Correspondence to: F. Kristjansson, Lotus Labs Pvt. Ltd, ‘Lotus House’, 7 Miller Tank Bed Area, Jasma Bhavan Road, Vasanthnagar, Bangalore 560 052, India. E-mail: falark@lotuslabs.com 1 Lotus Labs Pvt. Ltd, ‘Lotus House’, 7 Miller Tank Bed Area, Jasma Bhavan Road, Vasanthnagar, Bangalore 560 052, India 2 Suramus Biopharm, Integrated Drug Development, No. 77, 10th Cross, 29th Main, J. P. Nagar I Phase, Bangalore 560 078, India Abbreviations used: ANDA, abbreviated new drug application; AUC 0–t , area under the plasma concentration vs time curve up to time = t; AUC 0–∞ , area under the plasma concentration vs time curve up to time = infnity; C max , peak plasma concentration; CYP, cytochrome P450 isozyme; EM, extensive metabo- lizer; IS, internal standard; HQC, high quality control; LLQ, lower limit of quan- titation; LQC, low quality control; MTD, maximum tolerated dose; MQC, medium quality control; NCE, new chemical entity; NDA, new drug applica- tion; PM, poor metabolizer; QC, quality control sample; TDM, therapeutic drug monitoring; t 1/2 , elimination half-life; ULQ, upper limit of quantitation. Introduction The pharmacokinetic characterization of novel chemical entities (NCEs) in preclinical and clinical development is achieved by employing non-zero standards (typically n = 8) to cover a well- defned range of concentrations. As part of the assay validation, it is customary to defne the lower bound (denoted as lower limit of quantitation, LLQ) and the upper bound (denoted as upper limit of quantitation, ULQ) of the calibration curves. While the lower bound is greatly infuenced by factors such as method of sample extrac- tion (protein precipitation, solid-phase extraction or liquid–liquid extraction), detection platform and inherent attributes of the analyte etc., the upper bound is generally dictated by the satura- tion of the instrument’s dynamic response. It is not uncommon to construct calibration curves that cover a more than 100-fold range in order to characterize the pharmacokinetics from starting low human doses to the anticipated maximum tolerated dose (MTD) in frst-in-man clinical protocols and also for application in bioequiva- lence/pharmacokinetic studies (Vijaya Bharathi et al., 2009; Zhang and Chen, 2009; Tang et al., 2009; Zeng et al., 2009; Jiang et al., 2009; Arnold et al., 2008; Handy et al., 2008; Minkin et al., 2008; Jain et al., 2008; Xue et al., 2007; Zeng et al., 2007; Xu et al., 2007; Shen et al., 2004; Upreti et al., 2003). Additionally, the existence of a 100-fold range may be useful when dealing with NCEs that are substrates for polymorphic cytochrome P450 (CYP) isozymes wherein log- orders of diferences in the plasma concentrations of the parent and/or metabolite are expected to occur between the extensive metabolizer and poor metabolizer phenotypes (Preskorn et al., 2009; Shao et al., 2009; Davies et al., 2008; Stamer et al., 2007; Shilbayeh and Tutunji, 2006; Inomata et al., 2005). Rationale It is important that common yardsticks, unbiased assessment and rigor need to be strictly adhered during assay validation of any analyte(s) and/or its associated metabolite(s) since it will be used 544