Activity of GR30921X (NSC 382057) and GR63178A (NSC D611615) in human ovarian cancer lines

Compounds of a new group of pentacyclic pyrroloquinones have been shown to have activity against several experimental tumour systems. The first product to be considered for clinical evaluation was GR30921X (NSC 382057) (Figure 1). The drug appeared to be inactive against L1210 and P388 leukaemias, but active in a range of solid tumours, including mouse sarcoma 180, rat hepatoma D23, human colon HT29 and human mammary MX-1 xenografts in athymic nude mice (Fenton et al., 1985). Its mechanism of action, however, is undetermined. For clinical use the drug had to be formulated as a micro-crystalline suspension in low concentrations of polyethylene glycol 300 and propylene glycol (PEG/PG) to allow i.v. administration. In 1985 phase I trials started in a number of European cancer centers, but were discontinued with the introduction of the water soluble analogue, GR63178A. GR63178A (NSC D611615) (Figure 1) has a similar preclinical efficacy profile to that of GR30921X including activity against mouse adenocarcinoma MAC 30/T, mouse colon 38 and the human lung xenograft LX-1 (Fenton et al., 1989). Its toxicity profile in animals was shown to be relatively free of side effects and in particular, there was no evidence of bone-marrow suppression in animal models. Presently, the drug is in phase I trials in British and Dutch cancer centres (Cassidy et al., 1989; Eccles et al., 1989, Verweij et al., 1989). Secondary screens with a disease-oriented approach have been developed which utilise a series of human tumour lines, derived from the same tumour type and grown in nude mice. These screens may add important information on the potential clinical activity and the differential capacity of promising anti-cancer compounds (Winograd et al., 1987; Boven et al., 1988). Concurrent with phase I clinical trials, we have investigated GR30921X and GR63178A for their efficacy in a panel of human ovarian cancer lines to analyse their potential activity in ovarian cancer in the clinical situation. Female NMRI/Cpb nude mice (Harlan Cpb, Zeist, Netherlands) maintained under sterile conditions were used (at the age of 8-10 weeks), 2-3 mm diameter tumour fragments being implanted in both flanks. The tumour lines studied originated from various ovarian cancer subtypes (Table I), and had their own sensitivity patterns to conventional cytostatic drugs (Boven, 1988). GR30921X (suspended in PEG/PG 1 mgml-') and GR63178A (dissolved in 5% dextrose 1-10mgml-') were kindly provided by Glaxo Group Research Limited (Greenford, Middlesex, UK). Tumours were measured weekly in three dimensions and the volume was calculated by the equation, length x width x thickness x 0.5. Treatment was started at the time tumours reached a mean volume between 50 and 150mm3. In each experiment tumour-bearing mice were randomised to give at least six animals in the treatment and control group. Drugs were