LAUSR

In this issue, Ganschow et al1 describe their multicenter experience with 56 pediatric liver transplant recipients who were converted within the first 6 months from standard immunosuppression with calcineurin inhibitors (CNI) to a combination of everolimus, an mTOR inhibitor, and reduceddose CNI. Recruitment was stopped prematurely due to an 8.9% incidence of posttransplant lymphoproliferative disease (PTLD). Further, adverse events occurred in all children, included a variety of bacterial and opportunistic viral infections in 89%, and led to discontinuation of protocol drugs in 44.6% children. Two children each experienced treated or untreated biopsyproven acute cellular rejection. This report is reminiscent of others describing primary immunosuppression of adult and pediatric renal transplant recipients with CNI and sirolimus, the first mTOR inhibitor to be used clinically.2,3 In pivotal trials in adult recipients, protocol discontinuation rates approached 50%, and side effects, which also included nephrotoxicity, accounted for the larger fraction of these discontinuations.2 Among 274 children with renal transplantation, PTLD occurred in 19 (6.9%), and the authors advocated against using such combinations.3 To avoid complications seen in these previous studies, Ganschow et al appropriately used several criteria to exclude atrisk subjects and mandated CNIdose reduction upon starting everolimus. This approach produced early improvement in glomerular function, from T0 to week 1, likely because renal hemodynamics improved with CNIdose reductions. These improvements do not appear to have been sustained as the study progressed, suggesting that perhaps synergistic nephrotoxicity between mTOR inhibitor and CNIs had not been eliminated. Whether these events also contributed to discontinuations is not clear. A predominantly infectious toxicity profile emerged. Despite blood level monitoring, PTLD occurred in six children, with mean (SD, range) tacrolimus blood levels 6 ng/mL (0.77, 5.19.4) and everolimus levels 4.4 ng/mL (0.39, 3.25.5). Overall, CNI blood levels continued to exceed preestablished limits for at least a month after starting everolimus in a third of children receiving tacrolimus and in twothirds of those receiving cyclosporine. Ganschow et al appropriately implicate overimmunosuppression as the basis for side effects. Poor outcomes with yet another combination of CNI and mTOR inhibitor are not surprising given the unpredictable consequences of synergy between these two classes of drugs. Ganschow et al are to be complimented for a thorough review of their results. Pglycoprotein, which mediates export of both these drugs, is inhibited by each drug, but to a greater extent by CNI, in immunocompetent and other types of cells.4 This may explain why the halflife of mTOR inhibitors can increase over time, and why improvements in renal function were not sustained.5,6 Paradoxically, several studies show that tacrolimus bioavailability decreases in the presence of mTOR inhibitors.6,7 Whether this potential interaction compounded the effects of CNIdose reduction during conversion to everolimus and led to rejection in two children within 75 and 111 days after starting everolimus is not certain. Detailed pharmacokinetic studies were not performed. In hindsight, pharmacodynamic modeling of a cyclosporine and sirolimus combination may explain the recurring problems with combinations of CNI and mTOR inhibitors, in which each drug is titrated to fixed dose or concentration targets, especially among children.8 This previous study modeled the inhibitory effects of several concentration mixtures of the two drugs, on expression of CD54 or intercellular adhesion molecule1 (ICAM1). Expression was measured on mitogenstimulated B lymphocytes from five normal human subjects. Using spline models, the various mean concentration mixtures at which CD54 expression was inhibited halfmaximally (EC50) were plotted as a curve (Figure 1). This EC50 curve was used to identify a mean concentration mixture which achieved halfmaximal inhibition of CD54 at minimal cyclosporine exposure. This EC50 concentration mixture was nearly identical to that associated with optimal rejection