Although abolishment of ovarian cyclicity by the use of a long-acting
GnRH agonist (
GnRH-a) provides effective treatment for
premenstrual syndrome (PMS), its use is limited by sequellae of the resultant hypoestrogenism. In this study the effects of
estrogen/
progestin replacement on the symptomatic improvement afforded by
GnRH-a were evaluated in eight women with severe PMS. The 8-month study design included 2 months of control, 2 months of
GnRH-a alone, and 4 further months in which the exogenous
steroids were replaced in randomized, double blind, placebo-controlled cross-over fashion using 1 month each of 1)
conjugated equine estrogen (CEE) on days 1-25, 2) 10 mg
medroxyprogesterone acetate (MPA) on days 16-25, 3) CEE (days 1-25) plus MPA (days 16-25), and 4) placebo alone. Mood and physical symptoms were measured daily on a valid and reliable instrument, the Calendar of Premenstrual Experiences. As expected, administration of
GnRH-a alone resulted in a 75% improvement in luteal phase symptom scores (17.8 +/- 4.8 vs. 4.2 +/- 1.6; P less than 0.01). Combined sequential administration of CEE and MPA in addition to
GnRH-a was effective in maintaining the reduced symptom scores seen after
GnRH-a alone and was superior to the addition of CEE alone, MPA alone, or placebo. This combination of CEE and MPA resulted in a 60% improvement (P less than 0.05) compared to the luteal phase of control months in both behavioral (14.1 +/- 3.9 vs. 4.2 +/- 0.8) and total (17.8 +/- 4.8 vs. 6.5 +/- 1.8) symptoms. We conclude that the undesirable consequence of ovarian
steroid deficiency in the treatment of PMS by
GnRH-a can be overcome by the addition of sequential
estrogen and
progestogen replacements without significantly reducing the effectiveness of
GnRH-a in this disorder.