We studied the correlation of S-phase fraction (SPF) with clinical out
come in 127 pre- or perimenopausal patients with breast cancers treate
d by neoadjuvant chemotherapy from October 1986 to June 1990. When the
patients were analysed using the median value of the SPF as a thresho
ld, there was a small but non-significant difference in favour of low
SPF tumours for metastasis-free survival. SPF was the only parameter p
redicting overall survival in multivariate analysis (P < 0.002) which
included T, N, histopathological grade and steroid hormone receptors.
The results of metastasis-free survival contrasted with previous analy
ses with shorter follow-up, so we tested the time-dependent influence
of SPF on prognosis. It was thus shown that SPF significantly predicts
metastasis-free survival only during the first 30 months, whereas the
relative risk of cancer-related death according to SPF remains signif
icant for 56 months. In order to find an explanation for the differenc
e in predictivity between metastasis-free survival and overall surviva
l, we studied the post-relapse survival. Significantly shorter surviva
l (median 12 months) was associated with tumours presenting pre-treatm
ent high SPF values, compared to the low SPF group for which 60% of th
e patients were still alive after 30 months of metastasis phase (P = 0
.002). Our current results, in a homogeneous series with a median foll
ow-up of over 5 years, emphasise the importance of proliferation-relat
ed parameters for breast cancer management. (C) 1997 Elsevier Science
Ltd.