Skip to main content
. 2006 Oct 24;95(10):1410–1414. doi: 10.1038/sj.bjc.6603444

Table 2. Univariate analysis of predictors of DCIS recurrence at 5 years using the log rank test.

Clinical factor Non-recurrent (n=94) Recurrent (n=35) P-value
HER2 status (%) (n=125)
 Positive (⩾2) 52 (57.8) 29 (82.9)  
 Negative (<2) 38 (42.2) 6 (17.1) 0.01
       
Ki67 score (%) (n=125)
 <12.2 (median) 52 (57.8) 10 (28.6)  
 ⩾12.2 38 (42.2) 25 (71.4) 0.006
       
Activated c-Src (%) (n=129)
 Low 23 (24.5) 2 (5.7)  
 Moderate 38 (40.4) 15 (42.9)  
 High 33 (35.1) 18 (51.4) 0.011
       
ER status (%) (n=126)
 Positive 61 (67.0) 21 (60.0)  
 Negative 30 (33.0) 14 (40.0) 0.524
       
Tumour grade (%) (n=128)
 Low 9 (9.7) 0 (0.0)  
 Intermediate 34 (36.5) 4 (11.4)  
 High 50 (53.8) 31 (88.6) 0.0003
       
Margins (%) (n=129)      
 Involved (<1 mm) 78 (83.0) 19 (54.3)  
 Clear (⩾1 mm) 16 (17.0) 16 (45.7) 0.001
       
Patient age (years) (n=129) (%)
 <50 6 (6.4) 8 (22.9)  
 ⩾50 88 (93.6) 27 (77.1) 0.019

HER2 positivity, high epithelial proliferation, high levels of total cellular activated c-Src, high tumour nuclear grade, involved margins and young age at presentation were associated with a higher disease recurrence at 5 years.