Use of Conjugated Estrogens in Life-Threatening Gastrointestinal Bleeding in Hemodialysis Patients—A Review

Case

A 67-year-old lady with a medical history of hypertension, end-stage renal disease on hemodialysis, polycystic kidney disease, and multiple medical problems initially presented to an outside hospital following a fall. She was diagnosed with fracture of left hip and underwent surgery. Her postoperative period was complicated by gram-negative bacteremia. Six days after admission, she developed melena and hemoglobin dropped to 6 g/dL. Esophagogastroduodenoscopy (EGD) revealed gastric ectasia. She continued to have melena and hematemesis and had a total of 5 endoscopies and multiple transfusions prior to transfer. At our institution, 2 days after the transfer (24 days after her initial admission), a capsule endoscopy was nondiagnostic for any evidence of bleeding. On the day after the capsule endoscopy, she developed melena and her hemoglobin dropped to 6.7 g/dL despite 6 units of packed red blood cells (PRBCs). The patient was started on octreotide drip, intravenous proton pump inhibitor, desmopressin, and epsilon amino caproic acid. An emergent EGD showed normal upper and middle third of the esophagus, and mild esophagitis seen in the lower third of the esophagus. Blood was found in the stomach and fundus and an ulcer was seen in the stomach. There were areas of angioectasia and angiodysplasia (AD) found in the antrum and empiric norepinephrine intragastric injections were given. Over the next 4 days, she continued to have intermittent bleeding requiring 3 more units of PRBC’s. In addition, she also received multiple transfusions with fresh frozen plasma, cryoprecipitate, and platelets. A repeat EGD showed blood in the fundus and multiple ulcers in the fundus of the stomach. At this time, empiric left gastric artery embolization was performed. Despite the procedure, after 1 day, the patient had another significant GI bleed requiring blood transfusions and a repeat EGD. This again showed active upper GI bleed from the stomach. At this time, it was decided to start the patient on conjugated estrogens 25 mg intravenous for 5 days. Within 2 days of starting conjugated estrogens, patient hemoglobin stabilized and she did not require any further endoscopies. Her hemoglobin improved to 9.7 g/dL and discharged to a nursing home. The patients bleed did recur a few months later; however, conjugated estrogens were discontinued at her nursing home.

Discussion

The rationale for the use of hormonal therapy in the treatment of bleeding from vascular malformations was provided by observations that epistaxis due to HHT decreased during pregnancy and worsened in the postmenopausal period, ² several days before menstruation, or after oophorectomy. ¹ Since then, hormonal therapy has been reported in several case reports, series, and a few uncontrolled and controlled studies of GI bleeding. All these publications report either an arrest of bleeding or an increase in hemoglobin for the duration of follow-up and an important reduction of transfusion requirements. The successes have been mainly anecdotal and none of the trials have been on patients with renal failure and with a life-threatening GI bleeding.

To date, there are seven^3–9 reported trials on the efficacy of conjugated estrogens in patients with renal sufficiency and bleeding. In addition, Bronner et al ¹⁰ used either estrogen alone or estrogen in combination with progestrone. Also, there are 5 reported clinical trials on the use of combined (estrogen and progestrone) therapy.^11–15 Van Cutsem et al ¹⁴ published the first crossover study in 1990. However, this study was somewhat extended to another publication ¹⁵ and thus reducing the comparisons among only 4 different studies. We have summarized them in Table 1. All these studies are heterogeneous in type of patients included, method of study, associated with or without additional therapy, inclusion and exclusion criteria, and so on, and hence, it is difficult to compare results across these studies. The studies have included both men and women and none of them have reported a difference in response based on this. None of the trials have included patients with life-threatening GI bleeding with renal failure. Our patient is unique in this regard and also that this life-threatening bleed, which did not respond to any other modality of treatment, responded to hormonal therapy and the patient survived a long and complicated hospital stay. The exact mechanism of action of hormones is unknown. However, a number of theories have been proposed. ² These include stabilization of fragile vessels through increased keratinization of surrounding squamous epithelium, improvement of possible coagulation abnormalities, shortening of bleeding time, reduction in leaking from fragile capillaries, improved vasoconstrictive effect of vasopressin and noradrenaline ¹⁶ , and reduction of mesenteric blood flow through increased stasis. In patients with HHT, estrogen and progesterone receptors have been detected in nasal and skin telangiectatic lesions. Estrogen-receptor binding and improved endothelial integrity have been observed in patients with HHT. Hormones may also decrease vascular endothelial growth factor. ¹⁷ There is no evidence, however, that the GI ADs will regress with this therapy.

OPEN IN VIEWER

Table 1.

Summary of Clinical Trials Reported to Date on the Use of Hormonal Therapy for GI Bleeding.

Reference	Study Design	Number of Patients	Intervention	Results
Liu et al 3	Prospective, single center	6	CE 5-50 mg in divided doses every 12 hours	Reduced bleeding time in 5 patients and normalized in 4 patients after 2-5 days of treatment
Livio et al 4	Randomized, double blind, placebo controlled, crossover	6	CE 0.6 mg/kg IV for 5 days or placebo	Reduced bleeding time in all estrogen-treated patients within 6 hours
Bronner et al 10	Prospective, single center	7 (3 men and 4 women)	Norethynodrel/mestranol at varied doses (n = 6); ethinyl estradiol (n = 1)	Cessation of clinical bleeding during treatment. Reduction in transfusion requirements during treatment
Vigano et al 5	Prospective, controlled, single center	15 (10 men and 5 women)	CE 0.3 mg/kg IV, n = 5; CE 0.6 mg/kg IV, n = 10	Lower dose did not reduce bleeding time in the 5 patients; 4-5 daily infusions of 0.6 mg/kg IV are needed to prolong reduction of bleeding time
Heistinger et al 6	Randomized, double blind, placebo controlled, crossover, single center	7 (4 men and 3 women)	CE 0.6 mg/kg/d IV for 5 days	Significant reduction of bleeding time on days 7 and 14. Minimal effect at 21 days, no effect at day 28
Shemin et al 7	Group 1: prospective, single center; group 2: prospective, randomized, placebo controlled, single center	Group 1: 4; group 2: 10	Group 1: CE 50 mg orally until bleeding time normalized or for 9 days; group 2: CE 50 mg orally or placebo until bleeding time normalized or for 9 days	Group 1: cessation of bleeding in all patients, bleeding time normalized in 2 and decreased by 50% in 1; group 2: bleeding time normalized in 3 of 5 patients, bleeding time decreased by <50% in remaining 2 patients
Sloand and Schiff 8	Prospective, single center	6 (4 with active bleeding; 2 men and 4 women)	17β-estradiol 50 µg/24 hours transdermally (n = 3); 17β-estradiol 100 µg/24 hours transdermally (n = 3)	Reduced blood transfusion requirements. Improved bleeding time in all patients
Heunisch et al 9	Case report	One	CE 10 mg/d IV for 7 days	Hemoglobin level remained stable and requirement for blood products stopped
Van Cutsem et al14,15	Double blinded, crossover	26 (13 in control and 13 in treatment arm)	0.05 mg ethinyloestradiol and 1 mg norethisterone orally daily	A very significant drop in transfusion requirements was reported
Lewis et al 11	Controlled	64 (34 in control and 30 in treatment arm)	Control arm—5-10 mg norethynodrel either with mestranol, 0.075-0.15 mg (24 patients) or with CE 0.625 mg (6 patients)	The mean monthly transfusion requirement of packed red blood cells in the treated group was not significantly different from that found before therapy or from that of the control group
Barkin and Ross 13	Crossover, controlled	43 (14 men and 29 women)	Initially treated with 5 mg norethynodrel and 75 µg of mestranol orally. This was changed to 1 mg norethindrone and 0.05 mg of mestranol, orally twice a day	Combination hormonal therapy was shown to stop rebleeding in patients with occult GI bleeding of obscure origin
Junquera et al 12	Double blinded, randomized control trial	68 (35 in control and 33 in treatment arm; about 50% were men)	Treatment arm—ethinylestradiol (0.01 mg) plus norethisterone (2 mg) orally daily; control arm—placebo	Continuous estrogen–progestogen treatment is not useful in the prevention of rebleeding from GI angiodysplasias
Lamba et al, this paper	Case report	One	Conjugated estrogens 25 mg IV for 5 days	Hemoglobin stabilized and she did not require any further endoscopies

Abbreviations: GI, gastrointestinal; CE, conjugated estrogen; IV, intravenous.

Galbusera et al ¹⁸ recently published guidelines for management of bleeding in dialysis patients. They recommend the use of conjugated estrogens by intravenous infusion in a cumulative dose of 3 mg/kg as daily divided doses (ie, 0.6 mg/kg for 5 consecutive days). This should be utilized in persistent chronic bleeding. There are no guidelines about their use in an acute life-threatening bleed like our case. At this time, therefore, there are no clear guidelines on this issue. Furthermore, there are a number of relative and absolute contraindications on the use of these hormones. In addition, there are several side effects that may limit use of these hormones. These include an increased risk of stroke, deep vein thrombosis, pulmonary embolism, and myocardial infarction. Hormonal therapy is also associated with an increased risk of endometrial and breast cancer, dementia, gall bladder disease requiring surgery, hypercalcemia, and angioedema. Contraindications include undiagnosed abnormal genital bleeding, known, suspected, or history of estrogen-dependent neoplasia except in appropriately selected patients being treated for metastatic disease, active deep vein thrombosis, pulmonary embolism, or a history of these conditions, active or recent arterial thromboembolic disease, liver dysfunction or disease, known thrombophilic disorders, and known or suspected pregnancy.

To conclude, estrogen therapy is currently reserved for selected patients, in whom frequent recurrent bleeding occurs and endoscopy or surgery is ineffective or inapplicable due to the extent or distribution of the lesions or existence of comorbid disorders. Our patient had a total of 8 upper GI endoscopies with unsuccessful use of epinephrine injections, cautery, and argon photocoagulation. She also had extensive comorbidities, including renal failure, hypertension, and prior bowel surgery, making her ineligible for gastrectomy. However, after initiation of estrogen therapy, her bleeding stopped, and she was eventually discharged from the hospital. Estrogen therapy can thus be used for emergency and short-term control of bleeding with more definitive therapies to follow. We, therefore, urge other physicians to be aware of the use of hormonal therapy in similar patients.