BACKGROUND. There is controversy and confusion regarding therapy for p
atients with ductal carcinoma in situ (DCIS) of the breast. The Van Nu
ys Prognostic Index (VNPI) was developed to aid in the complex treatme
nt selection process. METHODS. The VNPI combines three significant pre
dictors of local recurrence: tumor size, margin width, and pathologic
classification. Scores of 1 (best) to 3 (worst) were assigned for each
of the 3 predictors and then totaled to give an overall VNPI score ra
nging from 3 to 9. Three hundred thirty-three patients with pure DCIS
treated with breast preservation (195 by excision only and 138 by exci
sion plus radiation therapy) were studied with detection of local recu
rrence as the end point. RESULTS. There was no statistical difference
in the 8 year local recurrence free survival in patients with VNPI sco
res of 3 or 4, regardless of whether or not radiation therapy was used
(100% vs. 97%; P = not significant). Patients with VNPI scores of 5,
6, or 7 received a statistically significant 17% local recurrence free
survival benefit when treated with radiation therapy (85% vs. 68%; P
= 0.017). Patients with scores of 8 or 9, although showing the greates
t relative benefit from radiation therapy, experienced local recurrenc
e rates in excess of 60% at 8 years. CONCLUSIONS. DCIS patients with V
NPI scores of 3 or 4 can be considered for treatment with excision onl
y. Patients with intermediate scores (5, 6, or 7) show a 17% decrease
in local recurrence rates with radiation therapy. Patients with VNPI s
cores of 8 or 9 exhibit extremely high local recurrence rates, regardl
ess of irradiation, and should be considered for mastectomy. (C) 1996
American Cancer Society.