Purpose: The disappointing results of chemotherapy in glioblastoma might be
caused by the choices of agents, which mostly include nitrosurea. We compa
red the in vitro efficacy of chemotherapeutic agents, developing a method t
o summarize published data. Method: Between 1966 and 1995 chemotherapy in g
lioma cells was reported in 1643 articles. Efficacy was mostly described by
the drug concentration that killed 50% of the cells (LC50). It was measure
d with various cell-culture techniques, of which a colorimetric test [3(4,5
-dimethylthiazol-2-yl)-2,5-diphenyltrazolium bromide] was mostly used. We c
alculated factors from these data to transform results to LC50 values as if
the colorimetric test was used in all of them. This allowed data from all
publications to be summarized in a new type of meta analysis. Results: The
most important agents and the average LC50 values (mg/l) were actinomycin-D
0.042, vincristine 0.075, mitoxantrone 0.12, vinblastine 0.21, doxorubicin
0.29, diaziquone 0.76, cisplatin 1.1, methotrexate 1.1, cytasine arabinosi
de 1.59, 5-flurouracil 2.33, bleomycin 18.6, carboplatin 29.8, carmustine 3
7.0, nimustine 48.9, and lomustine 76.7. The most resistent cell was SNB56,
followed with increasing sensitivity by SF128 and A172, and primary cultur
es P497, SF210, U87MG, SF126, 9L, P540, U251MG, HU62, C6. The complete list
of original data is available upon request. Conclusion: The efficacy of ni
trosourea in vitro is low.