Conjugated Estrogens Reduce Endothelial Prostacyclin Production and Fail to Reduce Postbypass Blood Loss* Robert W Hull, M.D.; James A. Hasbargen, M.D.; Ste1Jiwn Fall, M.D., F.C.C.P.; am/ ThonUJS P. ()'Barr; Ph.D.
Intravenous conjugated estrogens correct bleeding times and reduce bleeding in uremia, gastrointestinal telangiectasias, and liver disease. One study found a similar benefit in patients undergoing open heart surgery. The mechanism by which conjugated estrogens improve bleeding times is unknown. We report on the effect of estrogens on endothelial prostacyclin production and bleeding in coronary bypass surgery. In a randomized, double-blind trial, 16 male patients undergoing elective coronary artery bypass surgery received four daily infusions of conjugated estrogens
Q ne study has suggested that preoperative high-
dose intravenous conjugated estrogens reduce blood loss in open heart surgery, but no mechanism was established.• Since this original study, conjugated estrogens have been shown to correct abnormal bleeding times and hemorrhage in uremic patients.2.3 Estrogen preparations have also been used in bleeding due to gastrointestinal telangiectasias, 4 ·H and recently, in the bleeding diathesis of liver disease. 9 The mechanism, however, by which conjugated estrogens improve bleeding times still remains unknown. Conjugated estrogens do not appear to affect the coagulation system, circulating levels of von Willebraud's factor, von Willebrand's factor multimers, in vitro platelet adhesion and aggregation, or serum thromboxane B2 levels.'1 The improvement in bleeding times during high-dose conjugated estrogen therapy, therefore, may be due to an effect on the vascular endothelium, perhaps reducing prostacyclin production. Studies addressing estrogen's eflect on endothelial prostacyclin production have yielded conflicting results. Much of this disparity is probably due to differences in the animal models used and the estrogen preparations administered. Ire, decided to reevaluate the effect of preoperative high-dose intravenous conjugated estrogens on blood loss in patients undergoing coronary artery bypass surgery. We focused on the eRect of conjugated estrogens on endothelial prostacyclin production. METIIODS
Stutly Dt•sign
\\'to> l"ndncted a randomized, douhle-hlind study in whil'h patients n•ct•ivt'd t•ither conjugated estrogens (0.6 mw'kg IPremarin]) or placeho. Our protollowing estrogen administration, ami hlood loss was recorded into the sixth day. The course of estrogen administration was chosen hased on the work of Livio et al' which identified the peak estrogen elfel'l on days .5 through 7 of therap): All antiplatelet agents except dipyridamole were withheld li>r at least seven days preoperative(): It was felt that dipyridamole could not ethically he withheld due to its henefit in redul'ing platelet Blood Loss in Cardiopulmonary Surgery (Hull et a/)
activation during t>xlrat·nfJ>ort>al cireulatinn. Its nst' was eontrnllt>d, hm\'t'Vt'r, sinl't' both groups rt'et•ivt>d idt>nlil'al rt>ginlt'ns of this drug. Aspirin was not adminislert>d in tilt' first 24 hours postnperativt'ly. Prt>opt>raliw lt>mplalt' blt>t>ding times, prothrombin timt>s, and aetivatt>d partial thromboplastin tinws wt'rt' rt't11nlt>d. Extral'OfJ>ort>al cireulatinn time was nott>d. Twt>nty-fs was ret11rdt>d, and hlnod prnduet rt>quin•ments Wt'rt' tld.
fhtil'nts The population was eomprist>d nf men whn Wt'rt' undt>rgning t>lt>ctive enrnnary arlt>ry bypass surgery with tlw li>llowing t>xdusinns: ( l) antit11agnlation; (2) history nf a bleeding disnrdt>r nr almnrmal prt>npt>rativt' t11agulation sert't'n nr blt•t>ding time; (3) history of a thrombntie disorder; (4) ht>patie nr rt'nal insullidt>ney whieh was defint>d as a history of prior disease nr abnormal sereening livt>r-associated enzymes, blood nrea nitrogen or St'rum ereatinint'; (5) malignaney; (6) enrrent smoking; (7) unstable angina or ld"t main d saline solution for transport. Basal prostaeydin prcKinetion, as determined by the formation nf 6PGF,., was nwasnrt>d in an attempt In asst>ss tht' aetna! ill viL'O pnKinction rates of proshtt·ydin. Although stimulated prostaeydin produdinn nlt'asnres the eapadty of tissnt's to produee prostaeydin, it may not rt'llt>l'l tht' ill tliVo rate nf prnduetinn."' Basal 6PCF,. prcKiudinn was mt'asurt>d by radioimmunoassay as prt>vinusly deseribed." Spedmens were plaeed in a series nf baths t11nlaining mid Krebs-Hingt>r-bil'arbnnatt' bulft>r with 1.5 mwml glumse. This slowed prnstaeydin synlht>lase aetivity and diluted any 6PGF,. that had lwt>n fnlitial tissne, the veins wert' slil't>d into I In -2 mm widt' rings and again plaeer I h prior to liSt'. Tlw rings Wt'rt' tht>n transft>rrt>d to silit•mized glass vials r 10 minutes on il't', the pll was rt>adjustt>d to 7.4 with IN NaOll. Tlw 6PGF,., was assayt>d by standard HIA proeednrt's using "ll-6PCF and antist'ra to 6PGF, •. This assay was St'nsitiw tn 10 PW 0. 7-ml tnlw. Cross rt>adivity was 0 pt>reent In thrnmboxane B, and 10 pert·ent In prostaglandin F, ami prostaglandin E,. Tht>st' lattt'r twn prostaglandins art' prt>st>nl in vt'r) small levt>ls in \'t'nnus tissnt' when eompart>d to the levels of 6PGF, ... lh lintht>r rednet' tlw dmnal'livity. samplt>s Wt'rt' dilutt>d prior to assay. Vahtt's wert> t'Xprt>ssed in tt>nns of pil•>grams nf 6PGF, .. released per milligram weight of tissut•.
Statistical Analysis All statistil·al analyses ust>d a nonpairt>d Student's t-tt>st statistie with results prest'nted as the mean ::t SEl\.1.
ited proper handling of the vein sample. After exclusions, the estrogen and control groups were each comprised of eight patients and were without significant differences with respect to age and preoperative hemostatic screens. The mean age was 57.5 ± 2 ..5 years in the estrogen group vs 54.9 ± 3.4 years in the control group. Preoperative hemostatic screens yieldt•d mean template bleeding times of 4. 7 ± 0.4 min and .5.0 ± 0.3 min, prothmmbin times of 10.4 ± 0.2 s ami 10.8 ± 0.1 s, and activated partial thromboplastin times of 27.0 ± 1.1 sand 27.1 ± 0. 7 sin the estmgen and control groups, respectively. All patients were receiving similar antianginal regimens, and none was receiving additional medications known to affect prostaglandin production. There were no differences between gmups with respect to the length of extracorporeal circulation with estrogen vs control values of 100 ± 29 min and 98 ± 45 min, respectively. Endothelial prostacydin production as assayed by 6PGF 1a levels was significantly lower in the estrogen group when compared to control gmup, as seen in Figure l. The mean 6PGF 1., levels in tlw estrogen and control groups were 457 ± 45 and 685 ± 94 pWmg vein, respectively (p
RESULTS
ENDOTHELIAL 6PGF1
:1
E
0 0 0 0
I~o 00
P
Intravenous conjugated estrogens correct bleeding times and reduce bleeding in uremia, gastrointestinal telangiectasias, and liver disease. One study found a similar benefit in patients undergoing open heart surgery. The mechanism by which conjugated estrogens improve bleeding times is unknown. We report on the effect of estrogens on endothelial prostacyclin production and bleeding in coronary bypass surgery. In a randomized, double-blind trial, 16 male patients undergoing elective coronary artery bypass surgery received four daily infusions of conjugated estrogens
Q ne study has suggested that preoperative high-
dose intravenous conjugated estrogens reduce blood loss in open heart surgery, but no mechanism was established.• Since this original study, conjugated estrogens have been shown to correct abnormal bleeding times and hemorrhage in uremic patients.2.3 Estrogen preparations have also been used in bleeding due to gastrointestinal telangiectasias, 4 ·H and recently, in the bleeding diathesis of liver disease. 9 The mechanism, however, by which conjugated estrogens improve bleeding times still remains unknown. Conjugated estrogens do not appear to affect the coagulation system, circulating levels of von Willebraud's factor, von Willebrand's factor multimers, in vitro platelet adhesion and aggregation, or serum thromboxane B2 levels.'1 The improvement in bleeding times during high-dose conjugated estrogen therapy, therefore, may be due to an effect on the vascular endothelium, perhaps reducing prostacyclin production. Studies addressing estrogen's eflect on endothelial prostacyclin production have yielded conflicting results. Much of this disparity is probably due to differences in the animal models used and the estrogen preparations administered. Ire, decided to reevaluate the effect of preoperative high-dose intravenous conjugated estrogens on blood loss in patients undergoing coronary artery bypass surgery. We focused on the eRect of conjugated estrogens on endothelial prostacyclin production. METIIODS
Stutly Dt•sign
\\'to> l"ndncted a randomized, douhle-hlind study in whil'h patients n•ct•ivt'd t•ither conjugated estrogens (0.6 mw'kg IPremarin]) or placeho. Our protollowing estrogen administration, ami hlood loss was recorded into the sixth day. The course of estrogen administration was chosen hased on the work of Livio et al' which identified the peak estrogen elfel'l on days .5 through 7 of therap): All antiplatelet agents except dipyridamole were withheld li>r at least seven days preoperative(): It was felt that dipyridamole could not ethically he withheld due to its henefit in redul'ing platelet Blood Loss in Cardiopulmonary Surgery (Hull et a/)
activation during t>xlrat·nfJ>ort>al cireulatinn. Its nst' was eontrnllt>d, hm\'t'Vt'r, sinl't' both groups rt'et•ivt>d idt>nlil'al rt>ginlt'ns of this drug. Aspirin was not adminislert>d in tilt' first 24 hours postnperativt'ly. Prt>opt>raliw lt>mplalt' blt>t>ding times, prothrombin timt>s, and aetivatt>d partial thromboplastin tinws wt'rt' rt't11nlt>d. Extral'OfJ>ort>al cireulatinn time was nott>d. Twt>nty-fs was ret11rdt>d, and hlnod prnduet rt>quin•ments Wt'rt' tld.
fhtil'nts The population was eomprist>d nf men whn Wt'rt' undt>rgning t>lt>ctive enrnnary arlt>ry bypass surgery with tlw li>llowing t>xdusinns: ( l) antit11agnlation; (2) history nf a bleeding disnrdt>r nr almnrmal prt>npt>rativt' t11agulation sert't'n nr blt•t>ding time; (3) history of a thrombntie disorder; (4) ht>patie nr rt'nal insullidt>ney whieh was defint>d as a history of prior disease nr abnormal sereening livt>r-associated enzymes, blood nrea nitrogen or St'rum ereatinint'; (5) malignaney; (6) enrrent smoking; (7) unstable angina or ld"t main d saline solution for transport. Basal prostaeydin prcKinetion, as determined by the formation nf 6PGF,., was nwasnrt>d in an attempt In asst>ss tht' aetna! ill viL'O pnKinction rates of proshtt·ydin. Although stimulated prostaeydin produdinn nlt'asnres the eapadty of tissnt's to produee prostaeydin, it may not rt'llt>l'l tht' ill tliVo rate nf prnduetinn."' Basal 6PCF,. prcKiudinn was mt'asurt>d by radioimmunoassay as prt>vinusly deseribed." Spedmens were plaeed in a series nf baths t11nlaining mid Krebs-Hingt>r-bil'arbnnatt' bulft>r with 1.5 mwml glumse. This slowed prnstaeydin synlht>lase aetivity and diluted any 6PGF,. that had lwt>n fnlitial tissne, the veins wert' slil't>d into I In -2 mm widt' rings and again plaeer I h prior to liSt'. Tlw rings Wt'rt' tht>n transft>rrt>d to silit•mized glass vials r 10 minutes on il't', the pll was rt>adjustt>d to 7.4 with IN NaOll. Tlw 6PGF,., was assayt>d by standard HIA proeednrt's using "ll-6PCF and antist'ra to 6PGF, •. This assay was St'nsitiw tn 10 PW 0. 7-ml tnlw. Cross rt>adivity was 0 pt>reent In thrnmboxane B, and 10 pert·ent In prostaglandin F, ami prostaglandin E,. Tht>st' lattt'r twn prostaglandins art' prt>st>nl in vt'r) small levt>ls in \'t'nnus tissnt' when eompart>d to the levels of 6PGF, ... lh lintht>r rednet' tlw dmnal'livity. samplt>s Wt'rt' dilutt>d prior to assay. Vahtt's wert> t'Xprt>ssed in tt>nns of pil•>grams nf 6PGF, .. released per milligram weight of tissut•.
Statistical Analysis All statistil·al analyses ust>d a nonpairt>d Student's t-tt>st statistie with results prest'nted as the mean ::t SEl\.1.
ited proper handling of the vein sample. After exclusions, the estrogen and control groups were each comprised of eight patients and were without significant differences with respect to age and preoperative hemostatic screens. The mean age was 57.5 ± 2 ..5 years in the estrogen group vs 54.9 ± 3.4 years in the control group. Preoperative hemostatic screens yieldt•d mean template bleeding times of 4. 7 ± 0.4 min and .5.0 ± 0.3 min, prothmmbin times of 10.4 ± 0.2 s ami 10.8 ± 0.1 s, and activated partial thromboplastin times of 27.0 ± 1.1 sand 27.1 ± 0. 7 sin the estmgen and control groups, respectively. All patients were receiving similar antianginal regimens, and none was receiving additional medications known to affect prostaglandin production. There were no differences between gmups with respect to the length of extracorporeal circulation with estrogen vs control values of 100 ± 29 min and 98 ± 45 min, respectively. Endothelial prostacydin production as assayed by 6PGF 1a levels was significantly lower in the estrogen group when compared to control gmup, as seen in Figure l. The mean 6PGF 1., levels in tlw estrogen and control groups were 457 ± 45 and 685 ± 94 pWmg vein, respectively (p
RESULTS
ENDOTHELIAL 6PGF1
:1
E
0 0 0 0
I~o 00
P
