In situ carmustine wafers containing 1,3-bis(2-chloroethyl)-1-nitrosourea (BCNU) are commonly used for the treatment of recurrent glioblastoma to overcome the brain-blood barrier. In theory, this chemotherapy diffuses into the adjacent parenchyma and the excipient degrades in maximum 8 weeks but no clinical data confirms this evolution, because patients are rarely operated again. A 75-year-old patient was operated twice for recurrent glioblastoma, and a carmustine wafer was implanted during the second surgery. Eleven months later, a third surgery was performed, revealing unexpected incomplete degradation of the wafer. 1H-Nuclear Magnetic Resonance was performed to compare this wafer to pure BCNU and to an unused copolymer wafer. In the used wafer, peaks corresponding to hydrophobic units of the excipient were no longer noticeable, whereas peaks of the hydrophilic units and traces of BCNU were still present. These surprising results could be related to the formation of a hydrophobic membrane around the wafer, thus interfering with the expected diffusion and degradation processes. The clinical benefit of carmustine wafers in addition to the standard radio-chemotherapy remains limited, and in vivo behavior of this treatment is not completely elucidated yet. We found that the wafer may remain after several months. Alternative strategies to deal with the blood–brain barrier, such as drug-loaded liposomes or ultrasound-opening, must be explored to offer larger drug diffusion or allow repetitive delivery.
Conclusion
Despite exclusive in situ drug delivery during 5 days between surgery and standard radio-chemotherapy, car- mustine shows limited efficiency. Our report shows that it could be related to an uncontrolled local tolerance and degradation of the copolymer. This finding underlines the importance of new studies concerning the long-term in vivo degradation of copolymer wafer in human. Moreover, alternative strategies to deal with the blood-brain barrier, such as drug-loaded liposomes or opening using ultrasound [4] must be considered. They could offer larger drug dif- fusion and repetitive delivery compared to that observed with copolymer wafers [9].