Table 4.

Estimated odds ratios and 95% confidence intervals of high grade hematological toxicities associated with GSTP1 genotypes in SWOG 8897

Genotype	# with high grade vs with low grade toxicity	Unadjusted OR (95% CI)	Unadjusted P-value	Adjusted OR (95% CI)†	Adjusted P-value†	FPRP§
Neutropenia (grade 3 and 4 vs grade ≤2)
AA	138/51	1.00		1.00
AG/GG	136/80	0.63 (0.41–0.96)	0.03	0.63 (0.41–0.97)	0.04	0.31
Ordinal model‡	-	0.72 (0.50–1.04)	0.08	0.72 (0.50–1.05)	0.08	0.47
Leucopenia (grade 3 and 4 vs grade ≤2)
AA	130/59	1.00		1.00
AG/GG	123/93	0.60 (0.40–0.90)	0.01	0.59 (0.39–0.89)	0.01	0.10
Ordinal model‡	-	0.59 (0.41–0.86)	0.006	0.58 (0.39–0.84)	0.004	0.04
Grade 4 hematological toxicity (yes vs no)
AA	97/92	1.00		1.00
AG/GG	94/122	0.73 (0.49–1.08)	0.12	0.72 (0.48–1.07)	0.11	0.55
Ordinal model‡	-	0.71 (0.49–1.04)	0.08	0.70 (0.48–1.03)	0.07	0.44

^†Footnote: Factors adjusted in the multivariable models include age, menopausal status, chemotherapy (CAF vs. CMF), tamoxifen treatment, and time from surgery to randomization.

^‡In ordinal models, it was assumed there is a linear relationship between toxicity grades as well as between counts of grade 4 toxicity.

^§FRRP, false positive report probability was calculated for an assumed prior probability of associations with toxicities of 0.1. An FPRP below 0.5 was deemed as a suggestive association and an FPRP below 0.2 was deemed as a more definitive association.