Table 1.

Predictors of invasive components in patients with preoperatively diagnosed ductal carcinoma in situ

	Total no. of patients	No. of underestimates of invasive cancer (%)	Predictors of invasive component
Jackman et al. (14)	1 326	183 (14%)	Mammographic mass (1.9 times as much as Calcification only, P < 0.001) Diagnosis by core-needle biopsy (1.9 times as much as VAB , P < 0.001)
Renshaw et al. (15)	91	17 (19%)	Comedo DCIS with cribriform/papillary pattern (OR: not shown , P = 0.002)
Hoorntje et al. (16)	255	41 (16%)	NG 3 DCIS (OR = 2.9, 95% CI 1.0–7.8) Periductal inflammation in core biopsies (OR = 3.3, 95% CI 1.3–8.7)
Yen et al. (17)	398	80 (20%)	Age ≤ 55y (OR = 2.19, 95% CI 1.11–4.32) Mammographic size ≥4.0 cm (OR = 2.92, 95% CI 1.51–5.66) HG 3 (OR = 3.06, 95% CI 1.49–6.30) Diagnosis by core-needle biopsy (Comparison with open biopsy, OR = 3.76, 95% CI 1.46–9.63)
Mittendorf et al. (18)	85	7 (20%)	Diagnosis by core-needle biopsy (Comparison with open biopsy, OR: not shown)
Wilkie et al. (19)	675	66 (10%)	HG 3 (OR: not shown , P = 0.003) Mammographic mass (OR = 2, 95% CI 1.08–4.65)
Goyal et al. (20)	587	220 (38%)	Clinically palpable mass (OR = 5.09, 95% CI 3.06–8.48) Mammographic mass (OR = 7.37, 95% CI 3.27–16.64)
Huo et al. (21)	200	41 (20.5%)	Mass lesion on imaging (OR = 2.48, 95% CI 1.1–5.62) Lesion size ≥1.5 cm (OR = 3.15, 95% CI 1.44–6.88)
Miyake et al. (22)	103	37 (35.9%)	Palpable lesion (OR = 4.091, 95% CI 1.399–11.959) MRI size ≥2.0 cm (OR = 4.506, 95% CI 1.322–15.358)
Park et al. (23)	86	27 (31.4%)	Palpable mass or nipple discharge (43% vs 22%; P = 0.04) Number of core specimens < 5 (45% vs 20%; P = 0.011) Mammographic size ≥2.5 cm (44% vs 14%; P = 0.022)US size ≥3.2 cm (55% vs 24%; P = 0.009)MRI size ≥3.0 cm (48% vs 19%; P = 0.004)

DCIS, ductal carcinoma in situ; VAB, vacuum-assisted biopsy; CNB, core needle biopsy.