Link between cyclosporin exposure in tissues and graft versus host disease in paediatric bone marrow transplantation: analysis by a PBPK model
C. Gérard (1), N. Bleyzac (2), P. Girard (1), B. Tranchand (1), G. Freyer (1), Y. Bertrand (2), M. Tod (1).
(1) EA 3738 Ciblage Thérapeutique en Oncologie, Faculté de médecine Lyon Sud, Oullins, France. (2) Institut d’Hématologie et d’Oncologie Pédiatrique, Lyon, France.
Objectives: In bone marrow transplantation (BMT), cyclosporin is used to prevent the graft versus host disease (GVHD). A retrospective analysis of clinical data showed different patterns of cyclosporin anti-GVHD effect depending on its mode of administration. Non-linearity in cyclosporin distribution may be an explanation. The objective of this study was to link the occurrence of GVHD to cyclosporin exposure in blood, target organs of GVHD (skin, liver and intestine) and also bone marrow and thymus (effect of cyclosporin on T lymphocytes).
Method: Using a PBPK model of cyclosporin disposition in children (see companion abstract), AUC in blood and organs were calculated for 61 paediatrics patients (31 with intermittent 2 h every 12 h infusions and 30 with continuous infusion) undergoing BMT. The influence of cyclosporin exposure on the probabilities of GVHD (grade > 0) and GVHD in skin, liver and intestine (score > 0) were assessed by binary logistic regression (SPSS V.17 software).
Results: Severe GVHD (grade III and IV) and cutaneous GVHD were significantly more frequent after continuous infusion than after intermittent infusions (17 vs 0 % and 89 vs 59 % respectively). Mean time to GVHD was 13 (intermittent) and 18 days (continuous). Mean cyclosporin dose at day 1 was 3.71 mg/kg (intermittent) and 3.55 mg/kg (continuous). There was no link between blood cyclosporin AUC and the occurrence of GVHD. There were significant links between AUCs in bone (including bone marrow) or thymus at the beginning of the treatment (0-24h) and the occurrence of GVHD (p < 0.05 in each case). The ratio of mean AUCs (no GVHD / GVHD) was 1.26 for bone and 1.13 for thymus. AUC in skin and intestine (0-12h) were a significant covariate of cutaneous GVHD and intestinal GVHD, respectively. The ratio of mean AUCs (no GVHD / GVHD) was 1.23 for skin and 1.21 for intestine. Cyclosporin infusion duration was not a significant covariate of the occurrence of GVHD once AUCs were taken into account.
Conclusions: In BMT, cyclosporin exposure in blood on day 1 was not a predictor of GVHD occurrence contrary to exposures in bone and thymus, which show the interest of the PBPK model. Differences in the pattern of GVHD depend only on differences in drug exposure in target tissues and not of duration of infusion per se. Non-linearity in tissue distribution explain these differences. The strong influence of cyclosporin exposure at day 1 confirms the importance of decreasing the activation of T-cells from the graft.
