Magnetic resonance imaging (MRI) is one of medical images used by doctors for diagnosing diseases. MRI shows higher quality in displaying soft tissues and tumors. Pharmacokinetic dynamic gadolinium-enhanced MR imaging and functional MR imaging (fMRI) were used in this dissertation. Dynamic MR images are obtained using fast spin-echo sequences at consecutive time after the injection of gadolinium-diethylene-triamine penta-acetic (Gd-DTPA) acid. A pharmacokinetic model analyzes time-signal intensity curves of suspected lesions. Functional MR imaging produces images of activated brain regions by detecting the indirect effects of neuronal activity on local blood volume, flow, and oxygen saturation. Thus it is a promising tool for further understanding the relationships between brain structure, function, and pathology. Because of patients' movement during imaging, serially acquired MR images do not correspond in the same pixel position. Therefore, matching corresponding points from MR images is one of fundamental tasks in this dissertation. Least-squares estimation is a standard method for parameter estimation. However, outliers (due to non-Gaussian noise, lesion evolution, motion-related artifacts, etc.) may exist and thus may cause the motion parameter estimation result to deteriorate. In this dissertation, we describe two robust estimation algorithms for the registration of serially acquired MR images. The first estimation algorithm is based on the Newton method and uses the Tukey's biweight objective function. The second estimation algorithm is based on the Levenberg-Marquardt technique and uses a skipped mean objective function. The robust M-estimators can suppress the effects of the outliers by scaling down their error magnitudes or completely rejecting outliers using a weighting function. Experimental results show the accuracy of the proposed robust estimation algorithms is within subpixel. MR imaging has been used to evaluate nasal papilloma. However, postoperative MR imaging of nasal papilloma becomes more complicated because repair with granulation and fibrosis occurs after surgery. Therefore, it is possible to misclassify recurrences as postoperative changes or to misclassify postoperative changes as recurrences. Recently, dynamic gadolinium-enhanced MR imaging with pharmacokinetic analysis has been successfully used to identify the post-treatment recurrence or postoperative changes in rectal and cervical carcinoma. Nasopharyngeal carcinoma (NPC) comprising malignant tumors is a disease more common in Asia than in other parts of the world. Hence, in this dissertation, we evaluate the feasibility of dynamic gadolinium-enhanced MR imaging with pharmacokinetic analysis in detecting NPC and distinguishing recurrent nasal papilloma from postoperative changes (fibrosis or granulation tissue). In this dissertation, a new approach to differentiate recurrent nasal papilloma from postoperative changes using pharmacokinetic dynamic gadolinium-enhanced MR imaging and robust motion correction is presented. First, a robust estimation technique is incorporated into nonlinear minimization method to robustly register dynamic gadolinium-enhanced MR images. Next, user roughly selects the region of interest (ROI) and an active contour technique is used to extract a more precise ROI. Then, the relative signal increase (RSI) is calculated. We use a three-parameter mathematical model for pharmacokinetic analysis. The pharmacokinetic parameters A (enhancement amplitude) and Tc (tissue distribution time) are calculated by a nonlinear least-squares fitting technique. The calculated A and Tc are used to characterize tissue. Pharmacokinetic analysis shows that recurrent nasal papilloma has faster tissue distribution time (Tc, 41 versus 88 seconds) and higher enhancement amplitude (A, 2.4 versus 1.2 arbitrary units) than do postoperative changes. A cut-off of 65 seconds for tissue distribution time and 1.6 units for enhancement amplitude yields an accuracy of 100% for differentiating recurrent nasal papilloma from postoperative changes. Though the above methods obtained good results, finding the region of interest (ROI) was done in a semi-automatic manner. For diagnosing NPC and improve the drawback, a system that automatically detects and labels NPC with dynamic gadolinium-enhanced MR imaging is presented. This system is a multistage process, involving motion correction, gadolinium-enhanced MR data quantitative evaluation, rough segmentation, and rough segmentation refinement. Three approaches, a relative signal increase method, a slope method and a relative signal change method, are proposed for the quantitative evaluation of gadolinium-enhanced MR data. After the quantitative evaluation, a rough NPC outline is determined. Morphological operations are applied to refine the rough segmentation into a final mask. The NPC detection results obtained using the proposed methods had a rating of 85% in match percent compared with these lesions identified by an experienced radiologist. However, the proposed methods can identify the NPC regions quickly and effectively. In this dissertation, the proposed methods provide significant improvement in correcting the motion-related artifacts and can enhance the detection of real brain activation and provide a fast, valuable diagnostic tool for detecting NPC and differentiating recurrent nasal papilloma from postoperative changes.

Made available in DSpace on 2014-06-11T19:29:05Z (GMT). No. of bitstreams: 0 Previous issue date: 2009-12-18Bitstream added on 2014-06-13T20:18:51Z : No. of bitstreams: 1 simoni_rf_me_botfm.pdf: 286447 bytes, checksum: 285c09bd4daeaa3c8d21ed54be946f3b (MD5) / Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES) / A constante de equilíbrio entre o plasma e o sítio efetor (ke0) é utilizada pelos modelos farmacocinéticos para prever a concentração do fármaco em seu local de ação (Ce). Seria interessante que a Ce de propofol fosse semelhante na perda e na recuperação da consciência. O objetivo desse estudo foi avaliar o desempenho clínico de duas diferentes ke0 (rápida = 1,21 min-1 e lenta = 0,26 min-1) com relação à concentração de propofol prevista em seu local de ação durante a perda e a recuperação da consciência usando o modelo farmacocinético de Marsh. Material e Método – Participaram desse estudo 20 voluntários adultos, sadios e do sexo masculino. Em todos os voluntários foi administrado propofol em regime de infusão alvo-controlada modelo farmacocinético de Marsh ke0 rápida e em outra oportunidade foi usado, o mesmo modelo farmacocinético com a ke0 lenta. Inicialmente, o propofol foi infundido em concentração-alvo plasmática de 3,0 μg.mL-1. A perda da consciência e recuperação da consciência foi baseada na resposta ao estímulo verbal. A concentração de propofol prevista em seu local de ação foi anotada no momento da perda e recuperação da consciência. Resultados - Na perda e recuperação da consciência, a concentração média de propofol prevista em seu local de ação pela ke0 rápida foi diferente (3,64 ± 0,78 e 1,47 ± 0,29 μg.mL-1, respectivamente, p < 0,0001), enquanto que com a ke0 lenta a concentração média de propofol prevista em seu local de ação foi semelhante (2,20 ± 0,70 e 2,13 ± 0,43 μg.mL-1, respectivamente, p = 0,5425). Conclusão - Do ponto de vista clínico, a ke0 lenta (0,26 min-1) incorporada ao modelo farmacocinético de Marsh apresentou melhor desempenho que a ke0 rápida (1,21 min-1), uma vez que a concentração de propofol prevista em seu local de ação na perda e recuperação da consciência foi... / The ke0 can be defined as the proportional variation of the gradient of concentration between the plasma and the effect-site in relation to the unit of time. Theoretically, the higher the value of the ke0, the faster the drug enters the effectsite. Therefore, drugs with short T½ke0 have high ke0s and fast onset of action. The aim of this study was to assess the clinical performance of two different ke0s (fast and slow) in terms of propofol effect-site concentration (Ce) during the loss and recovery of consciousness, using Marsh's pharmacokinetic model. Method: Twenty healthy male adult volunteers participated in this study. Propofol was first administered to the individual volunteer using Marsh's pharmacokinetic targetcontrolled infusion model with ke0 of 1.21 min-1 and, on another opportunity, with the same pharmacokinetic model but ke0 of 0.26 min-1. Propofol was infused in plasma target-concentration of 3.0 μg.mL-1. Loss and recovery of consciousness was defined as response of the volunteer to verbal stimulus. The Ce was registered at the moments of loss and recovery of consciousness. Results: At loss and recovery of consciousness, propofol Ce means predicted by the fast ke0 were different (3.64 ± 0.78 and 1.47 ± 0.29 μg.mL-1, respectively, p < 0.0001), whereas with the slow ke0 the predicted Ce means were similar (2.20 ± 0.70 and 2.13 ± 0.43 μg.mL-1, respectively, p = 0.5425). Conclusion: It can be concluded that slow ke0 (0.26 min-1) incorporated into Marsh's pharmacokinetic model showed better clinical performance than fast ke0 (1.21 min-1), since the predicted effect-site concentrations of propofol at loss and recovery of consciousness were similar. Key words: Intravenous anesthesia: propofol, pharmacokinetic model; Monitoring: bispectral index.

Orientador: Pedro Thadeu Galvão Vianna / Banca: Eliana Marisa Ganem / Banca: Gilberto Denucci / Resumo: A constante de equilíbrio entre o plasma e o sítio efetor (ke0) é utilizada pelos modelos farmacocinéticos para prever a concentração do fármaco em seu local de ação (Ce). Seria interessante que a Ce de propofol fosse semelhante na perda e na recuperação da consciência. O objetivo desse estudo foi avaliar o desempenho clínico de duas diferentes ke0 (rápida = 1,21 min-1 e lenta = 0,26 min-1) com relação à concentração de propofol prevista em seu local de ação durante a perda e a recuperação da consciência usando o modelo farmacocinético de Marsh. Material e Método - Participaram desse estudo 20 voluntários adultos, sadios e do sexo masculino. Em todos os voluntários foi administrado propofol em regime de infusão alvo-controlada modelo farmacocinético de Marsh ke0 rápida e em outra oportunidade foi usado, o mesmo modelo farmacocinético com a ke0 lenta. Inicialmente, o propofol foi infundido em concentração-alvo plasmática de 3,0 μg.mL-1. A perda da consciência e recuperação da consciência foi baseada na resposta ao estímulo verbal. A concentração de propofol prevista em seu local de ação foi anotada no momento da perda e recuperação da consciência. Resultados - Na perda e recuperação da consciência, a concentração média de propofol prevista em seu local de ação pela ke0 rápida foi diferente (3,64 ± 0,78 e 1,47 ± 0,29 μg.mL-1, respectivamente, p < 0,0001), enquanto que com a ke0 lenta a concentração média de propofol prevista em seu local de ação foi semelhante (2,20 ± 0,70 e 2,13 ± 0,43 μg.mL-1, respectivamente, p = 0,5425). Conclusão - Do ponto de vista clínico, a ke0 lenta (0,26 min-1) incorporada ao modelo farmacocinético de Marsh apresentou melhor desempenho que a ke0 rápida (1,21 min-1), uma vez que a concentração de propofol prevista em seu local de ação na perda e recuperação da consciência foi... (Resumo completo, clicar acesso eletrônico abaixo) / Abstract: The ke0 can be defined as the proportional variation of the gradient of concentration between the plasma and the effect-site in relation to the unit of time. Theoretically, the higher the value of the ke0, the faster the drug enters the effectsite. Therefore, drugs with short T½ke0 have high ke0s and fast onset of action. The aim of this study was to assess the clinical performance of two different ke0s (fast and slow) in terms of propofol effect-site concentration (Ce) during the loss and recovery of consciousness, using Marsh's pharmacokinetic model. Method: Twenty healthy male adult volunteers participated in this study. Propofol was first administered to the individual volunteer using Marsh's pharmacokinetic targetcontrolled infusion model with ke0 of 1.21 min-1 and, on another opportunity, with the same pharmacokinetic model but ke0 of 0.26 min-1. Propofol was infused in plasma target-concentration of 3.0 μg.mL-1. Loss and recovery of consciousness was defined as response of the volunteer to verbal stimulus. The Ce was registered at the moments of loss and recovery of consciousness. Results: At loss and recovery of consciousness, propofol Ce means predicted by the fast ke0 were different (3.64 ± 0.78 and 1.47 ± 0.29 μg.mL-1, respectively, p < 0.0001), whereas with the slow ke0 the predicted Ce means were similar (2.20 ± 0.70 and 2.13 ± 0.43 μg.mL-1, respectively, p = 0.5425). Conclusion: It can be concluded that slow ke0 (0.26 min-1) incorporated into Marsh's pharmacokinetic model showed better clinical performance than fast ke0 (1.21 min-1), since the predicted effect-site concentrations of propofol at loss and recovery of consciousness were similar. Key words: Intravenous anesthesia: propofol, pharmacokinetic model; Monitoring: bispectral index. / Mestre

No description available.

The bioavailability of a drug can be described as the fraction of an orally administered dose that reaches the systemic circulation and is often limited by first-pass metabolism in the gut and the liver. It is important to have knowledge about these processes since the systemic blood drug concentration is tightly connected to the effect of the drug. The general aim of this project was to quantitatively examine the role of the intestine in relation to the liver in first-pass metabolism of orally administered drugs. The first-pass metabolism of verapamil and raloxifene was investigated in detail with in vivo, in vitro and simulation studies, using the pig as an experimental model. The intestine contributed to the same extent as the liver to first-pass metabolism of R/S-verapamil in vivo in pigs. The S-isomer of verapamil was found in lower plasma concentrations compared to the R-isomer after oral dosing. The in vitro metabolism of verapamil in pig and human liver showed interspecies similarity and indicated equal intrinsic clearance for R- and S-verapamil. Through physiologically based pharmacokinetic modeling the stereoselectivity was explained by a combination of several processes, including enantioselective plasma protein binding, blood-to-plasma partition, and gut and liver tissue distribution. For raloxifene the intestine was the dominating organ in first-pass glucuronidation in vivo in pigs. Furthermore, the raloxifene concentration entering the intestine or the dose administered in the gut did not influence the plasma PK of raloxifene and indicated that the intestinal metabolism was not saturable with clinical relevant doses. For both verapamil and raloxifene, a time-dependent hepatic metabolism was noted with major consequences to the pharmacokinetic of the drugs. This project has pointed out the importance of intestinal metabolism in the overall first-pass extraction of drugs and indicates that intestinal metabolism should be considered and evaluated early in drug development.

Ertapenem is an antibiotic commonly used to treat a broad spectrum of infections and is part of a broader class of antibiotics called carbapenems. Unlike other carbapenems, ertapenem has a longer half-life and thus only has to be administered once a day. Previously, a physiologically-based pharmacokinetic (PBPK) model was developed to investigate the uptake, distribution, and elimination of ertapenem following a single one gram dose in normal height, normal weight males. Due to the absorption properties of ertapenem, the amount of fat in the body can influence how the drug binds, how quickly the drug passes through the body, and thus how effective the drug might be. Thus, we have revised the model so that it is applicable to males and females of differing body mass index (BMI). Simulations were performed to consider the distribution of the antibiotic in males and females with varying body mass indexes. These results could help to determine if there is a need for altered dosing regimens in the future.

Ertapenem is an antibiotic commonly used to treat a broad spectrum of infections, which is part of a broader class of antibiotics called carbapenem. Unlike other carbapenems, ertapenem has a longer half-life and thus only has to be administered once a day. A physiologically-based pharmacokinetic (PBPK) model was developed to investigate the uptake, distribution, and elimination of ertapenem following a single one gram dose. PBPK modeling incorporates known physiological parameters such as body weight, organ volumes, and blood ow rates in particular tissues. Furthermore, ertapenem is highly bound in human blood plasma; therefore, nonlinear binding is incorporated in the model since only the free portion of the drug can saturate tissues and, hence, is the only portion of the drug considered to be medicinally effective. Parameters in the model were estimated using a least squares inverse problem formulation with published data for blood concentrations of ertapenem for normal height, normal weight males. Finally, an uncertainty analysis of the parameter estimation and model predictions is presented.

No description available.

Existing therapeutic options for the alleviation of nicotine addiction have been largely ineffective at stemming the tide of tobacco use. Immunopharmacotherapy, or vaccination, is a promising, alternate therapy that is currently being explored. Results from previous studies indicate that nicotine vaccines (NVs) are effective in subjects that achieve high drug-specific antibody titers, though overall efficacy has not been observed. Consequently, improvement of these vaccines is necessary before they can achieve approval for human use. In this report, three separate approaches towards NV potentiation are explored. The first approach applied physiologically-based pharmacokinetic (PBPK) modeling to better assess NV potential. Rat and human physiological and pharmacological parameters were obtained from literature and used to construct compartmentalized models for nicotine and cotinine distribution. These models were then calibrated and validated using data obtained from literature. The final models verified the therapeutic potential of the NV concept, identified four key parameters associated with vaccine success, and established correlates for success that could be used to evaluate future NVs prior to clinical trials. In the second approach, conjugate NV scaffoldings were engineered by using wild-type (WT) and chimeric human papilloma (HPV) 16 L1 protein virus-like particles (VLPs). The chimeric protein was created by removing the last 34 C-terminal residues from the WT protein and then incorporating a multi-epitope insert that could universally target major histocompatibility complex (MHC) class II molecules. The proteins were subsequently expressed in E. coli and purified using a multi-step process. Comparisons between the separation outcomes revealed that the insert was able to modulate individual process outcomes and improve overall yield without inhibiting VLP assembly. In the third approach, commonly used carrier proteins were computationally mined for their MHC class II epitope content using human leukocyte antigen (HLA) population frequency data and MHC epitope prediction software. The most immunogenic epitopes were concatenated with interspacing cathepsin cleavage sequences and the resulting protein was re-evaluated using the earlier methods. This work represents the first ever in silico design of chimeric antigens that could potentially target all of the major HLA DQ and HLA DR allotypes found in humans. / Doctor of Philosophy / Existing treatment options for addressing nicotine addiction have been largely ineffective at preventing tobacco use. Vaccination, on the other hand, is a promising, alternate treatment option that is currently being explored. Previous studies have shown that nicotine vaccines (NVs) are effective in the subjects that respond well to the vaccine. Effectiveness in the majority of vaccine recipients, however, has not been observed. Consequently, improvement of these vaccines is necessary before they can be used in humans. In this report, three separate approaches for improving NV effectiveness are explored. The first approach applied physiologically-based pharmacokinetic (PBPK) modeling to better assess NV potential. Parameters were obtained from literature and used to construct models that could predict NV effectiveness in rats and humans. These models were then calibrated and validated using data obtained from literature. The final models verified that NVs could work if certain conditions were met, identified four key parameters associated with vaccine success, and allowed for estimation of NV efficacy prior to their evaluation in humans. In the second approach, protein carriers for conjugate NVs were constructed using the human papilloma (HPV) 16 L1 protein. This protein is known for its ability to form virus-like particles (VLPs). Both a modified and an unmodified (wild-type) protein were constructed. The modified HPV 16 L1 protein was created by replacing the last 34 C-terminal amino acids with a polypeptide insert that could enhance the immunogenicity of the vaccine. The modified and unmodified proteins were then expressed in E. coli and purified. Results indicated that the insert was able to modulate individual process outcomes and improve overall process yield without preventing VLP assembly. In the third approach, commonly used carrier proteins were computationally mined for their MHC class II epitope content using human gene frequency data and MHC epitope prediction software. The epitopes that were predicted to be the most immunogenic were linked together with interspacing protease recognition sequences and the immunogenicity of the resulting protein was re-evaluated using the prediction software. This work represents the first computational design of antigens that could potentially allow a vaccine to be effective in a large portion of human population regardless of the genetic variability.

O bussulfano (Bu) é um agente alquilante usado no condicionamento que precede o transplante de células-tronco hematopoiéticas (TCTH) em crianças. Sua farmacocinética (FC) apresenta uma grande variabilidade interindivíduo, que pode ser parcialmente explicada pelas variantes genéticas de GSTA1, gene da enzima glutationa S-transferase α1, crucial para o metabolismo do Bu. Vários métodos de predição da FC do Bu são usados para calcular sua dose, essencialmente com base na idade e peso do paciente. Até o momento, apenas um modelo adulto incorporou as variantes de GSTA1 no cálculo da sua dose do Bu. No presente trabalho, avaliou-se, inicialmente, o desempenho de métodos atualmente disponíveis em pediatria, em função das variantes genéticas de GSTA1. Foram avaliados os parâmetros de FC da primeira dose de 101 crianças e adolescentes submetidos a TCTH alogênico no CHU Sainte-Justine, Montreal, Canadá, após regime de condicionamento que incluía Bu intravenoso (BuCR, do inglês busulfan-containing regimen). Os haplótipos GSTA1 foram interpretados em pares (diplótipos) e depois classificados em três grupos com base nos seus diferentes potenciais de expressão enzimática. As AUCs (area under the curve) medidas e as AUCs calculadas a partir de doses de Bu preditas por 11 modelos diferentes foram classificadas de acordo com a sua capacidade para atingir a AUC-alvo (900 a 1.500 μM.min). Também foram calculados os erros de previsão do clearance do Bu. Após a primeira dose, as AUCs medidas atingiram a AUC-alvo em 38,7%. Os diplótipos de GSTA1 relacionados ao metabolismo lento (G3) e regimes contendo fludarabina (FluCR, do inglês fludarabine-containing regimen) foram os únicos fatores associados à AUC no alvo (OR 4,7, IC 95%, 1,1 - 19,8, p = 0,04 e OR 9,9, IC 95%, 1,6 - 61,7, p = 0,01, respectivamente). Utilizando os outros métodos para o cálculo da dose, a percentagem de AUC no alvo variou de 16% a 74%. G3 e FluCR foram, em alguns modelos, associados à AUC no alvo ou na faixa tóxica, enquanto que os metabolizadores rápidos (G1) foram por vezes associados a AUCs subterapêuticas. Essas associações foram confirmadas na análise de predição do clearance, em que os diplótipos da GSTA1 e o regime de condicionamento influenciaram significativamente a maioria dos erros de previsão dos métodos testados. Uma vez que GSTA1 mostrou influenciar significativamente os algoritmos disponíveis, pretendeu-se desenvolver um modelo de FC de população que incluísse variantes genéticas de GSTA1 como um fator no cálculo de dose do Bu. Para tanto, foram analisados os dados de concentração-tempo de 112 crianças e adolescentes que receberam um BuCR mieloablativo antes de 115 TCTH (autólogos e alogênicos), realizados também no CHU Sainte-Justine. Para a construção do modelo de FC de população, utilizou-se uma análise mista não linear. Sexo, doença de base (maligna vs. não maligna), idade pós-menstrual (PMA) ou idade cronológica, regime de condicionamento e diplótipos de GSTA1 foram avaliados como fatores potenciais. Um modelo de um compartimento com eliminação de primeira ordem foi o que melhor descreveu os dados disponíveis. Um fator de maturação do metabolismo de Bu (Fmat) e o peso elevado a exponencial alométrico teórico foram incluídos no modelo de base. A análise dos fatores revelou PMA (ΔOFV = -26,7, p = 2,3x10-7) e grupos de diplótipos de GSTA1 (ΔOFV = -11,7, p = 0,003) como fatores significativamente associados, respectivamente, ao volume e ao CL do Bu. Os CL dos metabolizadores rápidos (G1) foram preditos como sendo 7% mais elevados que os definidos como metabolizadores normais (G2), enquanto que os metabolizadores lentos (G3) foram descritos com CL 12% menor que os G2. Em conclusão, após se evidenciar que os métodos disponíveis para o cálculo de dose do Bu não são adequados para todos os grupos de diplótipos de GSTA1, propôs-se o primeiro algoritmo de cálculo de dose de Bu em pediatria baseado em farmacogenética. Seu uso pode contribuir para uma melhor previsibilidade da FC do Bu e, desta forma, melhor predizer a exposição de crianças e adolescentes à droga, de acordo com a capacidade metabólica de cada indivíduo. / Busulfan (Bu) is an alkylating agent used in the conditioning before hematopoietic stem cells transplantation (HSCT) in children. Its pharmacokinetics (PK) presents a great inter-individual variability, which can be partially explained by GSTA1 genetic variants, gene coding for the enzyme glutathione s-tranferase α1, crucial for Bu metabolism. Several methods of predicting PK are available and are used to calculate the Bu dose, based essentially on patients’ age and anthropometric characteristics. So far, a single adult model successfully incorporated this factor into the Bu dose calculation. In the present work, we initially evaluate the performance of the currently available guidelines across the different GSTA1 genetic variants. The PK parameters from the Bu first doses from 101 children and adolescents who have undergone allogenic SCT at the CHU Sainte-Justine, Montreal, Canada following a IV Bu-containing conditioning regimen (BuCR). GSTA1 haplotypes were interpreted in pairs (diplotypes) and then classified in 3 groups based on different potentials of enzyme expression. Measured AUCs and AUCs calculated from Bu doses predicted by 11 different models were classified according to their ability to achieve the AUC target (900 and 1500μM.min). Clearance prediction errors were also calculated. After the first dose, measured AUCs achieved the target in 38.7%. GSTA1 diplotypes groups related to poor Bu metabolism (G3) and fludarabine-containing regimens (FluCR) were the only factors associated with AUC within target (OR 4.7, 95% CI, 1.1 - 19.8, p=0.04 and OR 9.9, 95% CI, 1.6 - 61.7, p=0.01, respectively). Using other methods for dose calculation, percentage of AUCs within target varied from 16% to 74%. G3 and FluCR were, in some models, associated to AUC within the target and in the toxic range, whereas rapid-metabolizers (G1) were correlated with sub therapeutic AUCs. These associations were confirmed in clearance-prediction analysis, where GSTA1 diplotypes groups and conditioning regimen consistently influenced methods’ most prediction errors. Once GSTA1 status was demonstrated to influence significantly the available Bu dosing algorithms, we aimed to develop a population PK (PPK) model which included GSTA1 genetic variants as a covariate. For that, concentration-time data from 112 children and adolescents receiving IV Bu as a component of the conditioning regimen for 115 stem cell transplantations (autologous and allogenic) performed at CHU Sainte-Justine were analyzed. Non-linear mixed effects analysis was used to build a PPK model. Sex, baseline disease (malignant vs. non-malignant), post-menstrual age (PMA) or chronological age, conditioning regimen and GSTA1 diplotypes groups were evaluated as potential covariates. A one-compartment model with first-order elimination best described the data. A factor of Bu metabolism maturation (Fmat) and theoretical allometric scaling of weight were included in the base model. Covariate analysis revealed PMA (ΔOFV=-26.7, p=2.3x10-7) and GSTA1 diplotypes groups (ΔOFV=-11.7, p=0.003), as significant factors on volume and clearance (CL), respectively. CL of rapid metabolizers (G1) were predicted as being 7% higher and that of poor ones (G3) 12% lower than CL of those defined as normal metabolizers (G2). In conclusion, after evidencing that available Bu dosing methods are not suitable for all GSTA1 diplotypes groups, we have proposed the first pharmacogenomics-based dosing algorithm for Bu to be used in a pediatrics. Its use may contribute considerably to better predict Bu exposure in children and adolescents tailoring the dose according to individual metabolic capacity.