Parity-related molecular signatures and breast cancer subtypes by estrogen receptor status

Table 4 Associations between parity and gene expression enrichment of PVLSI parity signature in cancer-adjacent and tumor tissues from PBCS patients by estrogen receptor tumor type ^a

	Cancer-adjacent normal tissue, ER+		Tumor tissue, ER+		Cancer-adjacent normal tissue, ER-		Tumor tissue, ER-
GSEA ES^b	0.45		0.50		0.37		0.31
GSEA FWER^b	12% (sig)		15% (sig)		21%		44%
Creighton analysis^c	Negative	Positive	Negative	Positive	Negative	Positive	Negative	Positive
Nulliparous	16 (62%)	10 (38%)	15 (58%)	11 (42%)	1 (25%)	3 (75%)	1 (25%)	3 (75%)
Parous	46 (51%)	44 (49%)	50 (56%)	40 (44%)	12 (43%)	16 (57%)	8 (29%)	20 (71%)
OR (95% CI)^d	1.53 (0.63 to 3.73)		1.91 (0.45 to 2.6)		0.44 (0.04 to 4.82)		0.83 (0.07 to 9.25)
P-value^e	0.35		0.85		0.63		1.00

^aCI, 95% confidence interval; ER, Estrogen receptor; ES, Enrichment score; FWER, Family-wise error rate; GSEA, Gene set enrichment analysis; OR, Odds ratio; PBCS, Polish Breast Cancer Case–Control Study; PVLSI, Samples obtained from normal breast tissue of 130 women who underwent reduction mammoplasty surgery at Baystate Medical Center in Springfield, MA, USA, and banked at the Pioneer Valley Life Sciences Institute; sig, Significant. ^bIn GSEA analysis, ES indicates the GSEA enrichment score and FWER the corresponding family-wise error rate, with results considered significant (sig) if ≤ 15%. ^cIn Creighton analysis, “Positive” and “Negative” indicate enriched and nonenriched expression of parity signature. ^dIn Creighton analysis, odds ratios (ORs) and 95% confidence intervals (CIs) were obtained by unconditional logistic regression analysis with observed parity as the outcome variable and nulliparity as the reference. ^eIn Creighton analysis, P-values were obtained by χ² test or Fisher’s exact test if the expected cell count was less than five, and values were considered significant (sig) if ≤ 0.05.

ISSN: 1465-542X