Uterine bleeding is associated with platelet counts less than 20,000/μl in women with severe aplastic anemia

Abstract

OBJECTIVE: To assess the association between uterine bleeding and platelet count in women with severe aplastic anemia (SAA). DESIGN: Single institution retrospective study of SAA patients treated at NIH from 1989-2006. MATERIALS AND METHODS: We reviewed medical records and referring physician letters for 94 female SAA patients aged 12-60 years during the first 6 months of treatment under NIH immunosuppression protocols. Baseline platelet count, presence of and treatment for uterine and non-uterine bleeding events were recorded. Information on heavy bleeding episodes including treatment, treatment response, and platelet counts were assessed. Data were analyzed using Fisher's exact test and Wilcoxon rank-sum tests as appropriate. RESULTS: The analysis included 88 women with SAA. Of 6 excluded patients, 3 did not achieve menarche and 3 had prior hysterectomy. The median age was 28 years (13-56 years). Of the 71 (80.7%) women with known bleeding over 6 months, 54 (76.1%) women had uterine bleeding with 16 (22.5%) having only uterine and 38 (53.5%) having both uterine and non-uterine bleeding. The entry median platelet count was 20,000/μl (1,000-83,000/μl). Those with uterine bleeding had lower median platelet counts at study entry than those without (7,000/μl vs 24,500/μl; p<.05). Of those with platelet counts <20,000/μl at entry, 33.3% (14/42) had uterine bleeding compared to only 6.2% (3/45) in those with platelet counts ≥20,000/μl (p<.05). Six women (6.8%) developed platelet alloimmunization. Of 39 episodes of heavy uterine bleeding in 34 patients over the 6-month period, 29 had 1 and 5 had 2 episodes. Of 26 instances of uterine bleeding with platelet data, the median platelet count was 18,500/μl (1,000-88,000/μl). The 3 most common hormones for treatment/prevention of uterine bleeding were birth control pills (BCP) (46.5%), oral progestins (18.6%) and progestins + BCP or estrogen (8%). The overall success of the first hormonal management in treating/preventing uterine bleeding was 73%. Six (6.8%) women had surgery, including hysterectomy (n=4) and endometrial ablation (n=2), to control uterine bleeding. CONCLUSIONS: Uterine bleeding is associated with platelet count below 20,000/μl in women with SAA. Careful history of uterine bleeding should be taken from these patients. Although most patients received hormones in addition to platelet transfusions, the optimal management of a thrombocytopenic patient with significant uterine bleeding is unclear.