Table 2.

Menstrual resumption rate by treatment group

Outcomes	EC-wP		EP-wP		Estimated differencea, % (95% CI)	P b
	No./total	% (95% CI)	No./total	% (95% CI)
Menstrual resumption at 12 moc
Intention-to-treat analysis	126/261	48.3 (42.2 to 54.3)	164/260	63.1 (57.2 to 68.9)	14.8 (6.4 to 23.2)	<.001d
Sensitivity analysis	126/233	54.1 (47.7 to 60.5)	164/237	69.2 (63.3 to 75.1)	15.1 (6.4 to 23.8)	<.001e
Pregnancy outcomes
Considered pregnancy at enrollment	34/113	30.1 (21.6 to 38.5)	33/115	28.7 (20.4 to 37.0)	−1.4 (−13.2 to 10.4)	.82
Attempted pregnancy within 48 mo	11/113	9.7 (4.3 to 15.2)	19/115	17.4 (10.5 to 24.3)	7.7 (−1.2 to 16.5)	.09
Achieved pregnancy within 48-month	3/113	2.7 (0 to 5.6)	11/115	9.6 (4.2 to 14.9)	6.9 (0.7 to 13.0)	.03

^aThe estimated treatment difference was calculated, and the between-group difference was tested by the stratified Miettinen and Nurminen method. CI = confidence interval; EC-wP = epirubicin/cyclophosphamide followed by weekly paclitaxel; EP-wP = epirubicin/paclitaxel followed by weekly paclitaxel.

^bP values were based on χ² test and were 2-sided, if not specified.

^cResumption time is calculated from the last dose of adjuvant chemotherapy. In the intention-to-treat analysis, patients with no results of menstrual resumption, because of loss to follow-up, early intervention of ovarian suppression, or any recurrence event if it occurred first, were treated as nonresumed. The cases with no results of menstrual resumption were excluded in the sensitivity analysis.

^dThe rate of menstrual resumption at 12 months after chemotherapy is the coprimary endpoint, and the statistical significance level is .01. The P value was based on the stratified Miettinen and Nurminen method and was 2-sided.

^eIn the sensitivity analysis, comparisons of the resumption rate between groups used the stratified Mantel-Haenszel testand the P value was 2-sided.