Search

Background: Response to cisplatin-therapy is assumed to be related to the formation of platinum (Pt)-DNA adducts. Measurement of these adducts prior to therapy could be of value to improve cisplatin based cancer therapy. Materials and methods: We determined Pt-GG and Pt-AG adduct levels by use of 32P-postlabeling after ex vivo cisplatin treatment of fragments of head and neck squamous cell carcinoma (HNSCC) xenografts (five lines), and of tumor biopsies from patients with HNSCC (n = 8) and testicular cancer (n = 8). Results: Adduct levels in fragments (3 x 3 x 3 mm) exposed to 10 to 80 μM cisplatin for one hour, showed positive correlations with the in vivo response to cisplatin treatment (P < 0.05), as well as with the xenograft adduct levels observed after in vivo cisplatin treatment (P < 0.02). After an additional five-hour drug-free incubation period the correlations were absent. When patient tumor fragments were exposed ex vivo to 80 μM cisplatin for one hour, adduct levels were similar in HNSCC and testicular cancer. Persistence of adducts was observed for testicular cancer in the additional drug-free period. The adduct levels in the samples of two HNSCC patients who received cisplatin chemotherapy were in line with the hypothesis that higher adduct levels are associated with a better response. Conclusion: Our preliminary results show that analysis of DNA adducts following ex vivo drug treatment is a feasible approach towards a predictive assay, which warrants further investigation.

Total platinum contents and cisplatin–DNA adduct levels were determined in vivo in xenografted tumour tissues in mice and in vitro in cultured tumour cells of head and neck squamous cell carcinoma (HNSCC), and correlated with sensitivity to cisplatin. In vivo, a panel of five HNSCC tumour lines growing as xenografts in nude mice was used. In vitro, the panel consisted of five HNSCC cell lines, of which four had an in vivo equivalent. Sensitivity to cisplatin varied three- to sevenfold among cell lines and tumours respectively. However, the ranking of the sensitivities of the tumour lines (in vivo), also after reinjection of the cultured tumour cells, did not coincide with that of the corresponding cell lines, which showed that cell culture systems are not representative for the in vivo situation. Both in vitro and in vivo, however, significant correlations were found between total platinum levels, measured by atomic absorption spectrophotometry (AAS), and tumour response to cisplatin therapy at all time points tested. The levels of the two major cisplatin–DNA adduct types were determined by a recently developed and improved32P post-labelling assay at various time points after cisplatin treatment. Evidence is presented that the platinum–AG adduct, in which platinum is bound to guanine and an adjacent adenine, may be the cytotoxic lesion because a significant correlation was found between the platinum–AG levels and the sensitivities in our panel of HNSCC, in vitro as well as in vivo. This correlation with the platinum–AG levels was established at 1 h (in vitro) and 3 h (in vivo) after the start of the cisplatin treatment, which emphasizes the importance of early sampling. © 1999 Cancer Research Campaign