Short and long-term effects on survival in breast cancer patients treated by primary chemotherapy: an updated analysis of a randomized trial

A potential advantage of primary over adjuvant chemotherapy in breast cance r survival had been proposed on theoretical grounds. In 1994, early results of the S6-trial comparing primary chemotherapy vs. adjuvant chemotherapy f or operable breast cancer in 390 premenopausal patients had shown significa nt improvement in survival of the primary chemotherapy arm (p=0.04). An upd ated analysis conducted in 1995 showed the disappearance of this difference between the two arms (p=0.18). In the present analysis, we investigated the potential short and long-term benefits attributable to primary chemotherapy by applying weighted logrank tests designed to assess specifically these effects. Results were compared to those obtained with the classical logrank test. At a median follow-up of 105 months, a significant short-term survival bene fit (p=0.02) in favor of the primary chemotherapy has been shown. However, no long-term survival benefit (p=0.36) could be documented. The classical l ogrank test had revealed no significant difference (p=0.24) between the two groups but the proportional hazard assumption being rejected (p=0.04), the efficiency of this test can be questioned. Results using the present analysis suggested that primary chemotherapy dela yed early death rates, without significantly modifying long-term event rate s. It emphasizes that a short-term effect which is not necessarily associat ed with a long-term benefit may be seen at an early evaluation and disappea r later on.