The role of plasma oxidant-antioxidant status in survival after breast cancer surgery was investigated in a cohort of patients (n ؍ 363) hospitalized in Southern France between 1989 and 1992. The median follow-up was 8 years after surgery for histologically confirmed breast cancer. Plasma analyses were performed after diagnosis and before surgery and adjuvant therapy. We found an inverse relationship between plasma lipoperoxides (MDA) and tumor size at diagnosis, together with higher lipoperoxide levels in node-negative tumors than in node-positive ones (TNM). The longitudinal approach revealed an increased risk of recurrence for patients with plasma lipoperoxides in the highest tertile of the sample (RR ؍ 2.1, 95%CI 1.1-4.0). In addition, the risk of recurrence increased (RR ؍ 1.7, 95%CI 1.0 -3.0), after adjustment for the known prognostic factors (TNM), for patients with plasma lipid-adjusted vitamin E levels of over 22 mol/l. The risk of breast cancer death was twice as great for patients with plasma lipid-adjusted vitamin E levels above this value. Excesses of plasma lipoperoxides and vitamin E appear to be factors in poor prognosis for breast cancer-specific survival (OVS) and disease-free survival (DFS), respectively, independent of tumor characteristics at diagnosis. Several hypotheses are advanced to explain the possible role of plasma vitamin E as a factor in poor prognosis for survival. © 2002 Wiley-Liss, Inc.
Key words: breast cancer; survival; lipoperoxides; vitamin EA previous cross-sectional study that we had conducted 1 showed a statistically significant higher plasma vitamin E level in breast cancer patients than in controls. In line with these data, another of our studies 2 had shown that lipid peroxidation measured by malondialdehyde formation was lower in patients than in controls. Moreover, it was found that plasma antioxidants and lipid peroxidation were inversely associated with tumor size. 3 These results were completed by cross-sectional studies 4,5 of other cancer sites, which also showed that the risk of tumor growth increased 3-fold in the class of cancer patients in which plasma lipoperoxides were below 0.3 mol/l.These findings suggested that the oxidant-antioxidant status of patients could influence tumor growth and proliferative activity and subsequently overall survival and risk of recurrence after surgery. This hypothesis has also been investigated in other studies of both animal 6,7 and human tumors. 8 The present longitudinal study was therefore conducted to investigate the plasma concentrations of lipoperoxides, vitamin E (an antioxidant lipophile) and glutathione peroxidase (an antioxidant enzyme) as relevant prognostic factors for risk of death or recurrence for breast cancer patients.
MATERIAL AND METHODS
Patients and tumor characteristicsThe present study was carried out at the Val d'Aurelle-Paul