I. Balslev et al., FLOW CYTOMETRIC DNA-PLOIDY DEFINES PATIENTS WITH POOR-PROGNOSIS IN NODE-NEGATIVE BREAST-CANCER, International journal of cancer, 56(1), 1994, pp. 16-25
Flow cytometric DNA analysis was performed on fine-needle aspirates fr
om frozen tumour biopsies from 421 node-negative, non-adjuvantly-treat
ed breast-cancer patients with a median observation time of 6.75 years
. Among premenopausal patients (n = 175), those having at leas one DNA
''hypoploid'' sub-population defined as DNA index (DI) <0.96 or 1.44
less than or equal to DI < 1.92 (n = 81) were characterized by early r
ecurrences (log-rank p = 0.05, Wilcoxon p = 0.007), poor overall survi
val (OS) (p < 0.001) and poor survival after recurrence (p < 0.001). I
n the postmenopausal group (n = 246), there were no significant differ
ences among 7 different DI classes regarding either recurrence-free su
rvival (RFS) or OS. S-phase fraction (SPF), divided into quartiles, pr
edicted OS in premenopausal patients only (p = 0.02). Conventional mul
tivariate Cox analysis of OS in the premenopausal group revealed hypop
loidy to be the only independent prognostic factor involving a relativ
e risk (RR) of 22.8. Age less than or equal to 40 years was of margina
l significance, whereas SPF, histological grade (WHO), oestrogen and p
rogesterone receptor (PgR) content, tumour size and number of lymph no
des removed were excluded from the model. Application of the conventio
nal Cox model to the premenopausal group regarding RFS was found inapp
ropriate due to lack of proportionality of the hazards of hypoploidy,
SPF and histological grade. However, introduction of time-dependent co
-variates using 2 years as cut-off level showed hypoploidy with a RR o
f 3.52 and age less than or equal to 40 years with a RR of 3.28 to be
independent prognostic factors. In the postmenopausal group, the conve
ntional Cox model identified the number of lymph nodes removed to be t
he only independent prognostic factor regarding RFS as well as OS, whe
reas SPF <9% (lowest quartile) was of marginal significance in RFS ana
lysis. Hypoploidy was correlated to high SPF, low PGR content of low d
ifferentiation, indicating that hypoploid tumours proliferate rapidly
and hormone-independently. These patients may therefore benefit from a
djuvant chemotherapy administered while tumour burden and risk of drug
resistance are still low. (C) 1994 Wiley-Liss, Inc.