· JAMA 1996 · randomized controlled trial · n=596

Effects of hormone replacement therapy on endometrial histology in postmenopausal women. The Postmenopausal Estrogen/Progestin Interventions (PEPI) Trial. The Writing Group for the PEPI Trial.

Cited 604 times in the scientific literature.

Level 2 - randomized trial

Multicenter, randomized, double-masked, placebo-controlled trial

PubMed 8569016 · doi:10.1001/jama.1996.03530290040035 · record verified 2026-08-30

What was done

In the 3-year multicenter, randomized, double-masked, placebo-controlled Postmenopausal Estrogen/Progestin Interventions (PEPI) trial, 596 postmenopausal women aged 45 to 64 years were assigned to five 28-day cyclic regimens: placebo, 0.625 mg/d conjugated equine estrogens (CEE) alone, CEE plus cyclic medroxyprogesterone acetate (MPA; 10 mg/d for 12 days), CEE plus continuous MPA (2.5 mg/d), or CEE plus cyclic micronized progesterone (MP; 200 mg/d for 12 days). Endometrial histology was evaluated at baseline, annually, and during unscheduled visits via biopsy, curettage, or hysterectomy, and analyzed by intention to treat.

What was found

Women receiving CEE alone had significantly higher rates of simple (27.7% vs 0.8%), complex (22.7% vs 0.8%), and atypical hyperplasia (11.8% vs 0%) compared to placebo (P < .001 for all). Hyperplasia rates across the three CEE plus progestin regimens were similar to placebo (P = .16). CEE-only participants also required more unscheduled biopsies (66.4% vs 8.4%; P < .001) and curettages (17.6% vs 0.8%; P < .001) than placebo, while surgical rates in combination groups were similar to placebo (P = .38). Following study medication discontinuation, 34 of 36 women (94%) with complex or atypical hyperplasia reverted to normal histology after progestin therapy. One endometrial adenocarcinoma occurred in the placebo group.

Why it matters

This trial established that unopposed estrogen therapy at 0.625 mg/day substantially increases the risk of endometrial hyperplasia and proved that adding cyclic MPA, continuous MPA, or cyclic micronized progesterone effectively prevents these hyperplastic changes.

Limits

The study was limited to 3 years of follow-up, leaving long-term invasive cancer risk beyond this timeframe unmeasured. It tested only one estrogen dose (0.625 mg/d CEE) and specific progestin regimens, so results cannot be directly generalized to lower estrogen doses, transdermal administration, or other progestogen types.

Cited by