We compared the prognostic value of quality of life (QL) scores in the adju
vant setting and after relapse in two randomized trials of the Internationa
l Breast Cancer Study Group (IBCSG). More than 2000 premenopausal and postm
enopausal patients with node-positive breast cancer participating in random
ized trials comparing adjuvant therapies completed QL assessments for physi
cal wellbeing (PWB), mood, appetite and coping at study entry, at months 3
and 18 if they remained relapse-free, and at I month and at 6 months after
relapse. Cox regression models were used to test the relationship between Q
L scores and disease-free survival (DFS) in the adjuvant setting, or overal
l survival in the case of post-relapse QL measurement. All models were stra
tified by language/country group and included other factors related to QL a
nd/or outcome. DFS was not significantly predicted by QL scores at baseline
or month 18, or by changes in QL score between baseline and months 3 or 18
. In contrast, after relapse, QL scores were predictive for subsequent over
all survival One month after relapse, better mood (P = 0.04) was associated
with longer survival in premenopausal patients, and better appetite (P = 0
.005) in postmenopausal patients. Six months after relapse, PWB (P = 0.03)
and appetite (P = 0.03) were predictive for survival in premenopausal, and
PWB ( P < 0.0001), mood (P = 0.002), appetite (P = 0.0001) and coping (P =
0.0001) in postmenopausal patients.
Any prognostic significance of QL scores in the adjuvant setting is minimal
or obscured by chemotherapy effects, but there is strong prognostic signif
icance after disease relapse. The contrast suggests that patient perception
of the severity of underlying illness may determine reported QL scores. (C
) 2001 Harcourt Publishers Ltd.