Long-term biomonitoring of breast cancer patients under adjuvant chemotherapy: the comet assay as a possible predictive factor

Most chemotherapy treatments induce DNA damage in the exposed patients. Using the comet assay and peripheral blood mononuclear cells (PBMC), we have quantified this induced DNA damage and studied its relationship with GSTM1 and GSTT1 polymorphisms, and clinical parameters. For this purpose, 29 Caucasian women, breast cancer patients under CMF or CEF adjuvant chemotherapy were included in the study. The clinical parameters considered were (i) therapies side effects, like haematological and biochemical toxicities, (ii) prognostic and predictive factors, like hormonal receptor expression, tumour differentiation degree, sickness stage, and nodal status, and (iii) the effectiveness of the chemotherapy measured as five years relapse probability. The results were also related to the confounding factor age. Comet assay results indicate that 13 patients were characterised by absence of induced DNA strand breaks, and 16 patients presented induced DNA strand breaks along the treatment. Relationships between comet variables and clinical parameters, found with principal component analysis, correlations, one-way ANOVA and multivariate logistic regression analyses revealed that: (1) baseline levels of DNA damage are related to GSTM1 genotype and to hormonal receptor expression; (2) GSTM1 genotype also influences comet results after chemotherapy, as it does the AST level; (3) the tail moment values of the cycle 6.1 and the sickness stage might predict cancer relapse at five years: for the Stage, OR = 13.8 (IIB versus I+IIA), 95% CI 0.80-238.97, and for 6.1 cycle TM, OR = 1.3, 95%, CI 0.97-1.79, with a potential model (10* Stage (I-IIA = 0, IIB = 1) + 6.1 cycle), that has a good predictive capacity, with an area under ROC curve of 0.872 (CI 0.62-1.00). To our knowledge, this is the first time such a predictive value is found for the comet assay. Nevertheless, before the comet assay could be used as a tool for oncologists, this relationship should be confirmed in more patients, and problems of standardisation and data interpretation should be solved.