Recently, a combination of intra-operative acute haemodilution and deliberate hypotension to avoid the costs and hazards associated with homologous blood transfusions has been introduced [1,2].
Trimethaphan , nitroglycerin , and sodium nitroprusside  have been used to achieve deliberate hypotension during total hip arthroplasty. All have some disadvantages  such as rebound hypertension, difficulty in controlling blood pressure, increased intracranial pressure and cyanide intoxication. As a hypotensive agent, prostaglandin E1 (PGE1) has some advantageous effects, e.g. positive inotropic action, increased renal blood flow and increased diuresis [7,8]. It has been used safely to induce hypotension, thereby reducing blood loss and provide a dry surgical field during general anaesthesia for mastectomy , total hip replacement  or cerebral aneurysm surgery . Since the compensatory mechanisms of haemodilution, i.e. increased flow and tissue O2 extraction are based on increased cardiac output during normovolaemia , PGE1 may be an appropriate hypotensive drug during haemodilution.
It is known that both PGE1 and synthetic plasma volume expanders for acute haemodilution cause a number of coagulation disturbances, such as inhibition of platelet aggregation , reduction in platelet adhesiveness  and structural alterations in fibrin clots . We also know that these disturbances have minor clinical significance. However, it remains unclear whether the combined technique of PGE1-induced hypotension and acute haemodilution with synthetic plasma volume expanders could cause further deterioration of haemostatic function. This study was carried out to investigate this question.
The subjects were 40 ASA physical status I or II patients without hypertension, ischaemic heart disease, cerebral infarction, hepatic disease or renal disease, aged 50–73 years and weighing 43–71 kg, who were scheduled for total hip arthroplasty. Patients with preoperative coagulopathy suspected by clinical history or pre-operative blood coagulation tests (platelet count = PLT; <10 × 104 mm−3, prothrombin time = PT; <85%, and activated partial thromboplastin time = aPTT; >35 s) were excluded. The protocol was approved by the hospital's institutional human investigation committee, and written informed consent was obtained from each patient.
Premedication consisted of atropine 0.5 mg and hydroxyzine 1 mg kg−1 given intramuscularly (i.m.) 1 h before scheduled time of surgery. Anaesthesia was induced with intravenous (i.v.) thiamylal 4–5 mg kg−1 and fentanyl 2 μg kg−1, and maintained with 60% nitrous oxide in oxygen supplemented with 0.6–1.0% end-tidal isoflurane. Tracheal intubation was facilitated with i.v. vecuronium 0.1 mg kg−1. Intravenous fentanyl 1–2 μg kg−1 and vecuronium 0.05 mg kg−1 were injected i.v. during surgery. Ventilation was controlled to maintain end-tidal CO2 tension at 35 mmHg (Nelcor N-1000). A radial artery catheter was inserted for continuous monitoring of arterial pressure and to obtain blood samples. Lactated Ringer's (LR) solution containing 5% glucose was infused 15 mL kg−1 before surgery. The infusion of LR solution was continued at a rate of 8 mL kg−1 h−1 during surgery. In addition, LR was given at a rate of three times the blood loss. Rectal temperature was maintained between 35 and 37°C by using a circulating water blanket and by adjusting room temperature.
Prior to surgery, patients were randomly divided into four groups of 10. Group A received no induced hypotension or acute haemodilution, group B received induced hypotension alone, group C received acute haemodilution alone and group D received hypotension combined with acute haemodilution. In groups A and B, autologous blood of 600–800 mL was obtained from patients on the 20 th day prior to surgery and on the 14th day before surgery, and stored at 4°C in a refrigerator. The haemodilution in groups C and D was carried out after induction of anaesthesia by drawing ≈ 1000 mL of blood, and same amount of 6% hydroxyethyl starch (HES, average molecular weight 70000-6% in isotonic saline) was immediately infused. Induced hypotension was induced with PGE1 in groups B and D, and mean arterial blood pressure was maintained at ≈55 mmHg for 80 min during surgery. The autologous blood was transfused after the end of the study in all groups. The volume of blood loss was estimated during surgery by weighing swabs and by measuring suction drainage, and after operation by measuring blood collected from the wound drainage. Surgical procedure was standardized; the same surgeon performed all surgery, and the same anaesthesiologist was responsible for peri- and post-operative care and follow-up.
Blood sampling and data collection
Measurements included PLT, PT, aPTT and plasma concentration of fibrinogen (FIB) for indices of the coagulation system, and plasma antithrombin-III (AT-III) activity, plasma plasminogen (PLG) activity and serum fibrin degradation products (FDP) for indices of the fibrinolysis system. PLT and haematocrit values were measured by blood cells counter analyser (Sysmex NE-6000, Toa Iyo Denshi Corp, Kobe, Japan). PT, aPTT and FIB values were measured by a coagulation system analyser (Coag-Stat Super BC-2230, Kyoto Daiichi Kagaku Corp, Kyoto, Japan). AT-III and PLG activities were measured by chromogenic peptide substrates method  and FDP by latex clumping method  using a latex photometric immuno assay system (LPIA 100, DIA-IATRON, Mitsubishi Kasei, Tokyo).
Measurements of PLT, PT, aPTT, AT-III, FIB, FDP, PLG and haematocrit values were made after induction of anaesthesia (T1), before starting surgery (T2) (before inducing hypotension = groups B and D, after haemodilution = group C), 80 min after starting surgery (T3), (80 min after starting hypotension, groups B and D) and 60 min after the end of surgery (T4). Sampling times and their relation to the various phases of the investigation were similar in all groups. Blood samples were measured immediately.
The data were expressed as mean ± SD. Analysis of variance and Scheffe's test were performed for statistical analysis of haemodynamic and haemostatic data. Unpaired Student's t-test and analysis of variance were used for statistical analysis of the other data. A P-value less than 0.05 was considered statistically significant.
The four age groups were similar in anthropometric characteristics (sex, age and weight), operative time, total dosage of PGE1, operative blood loss and haemorrhagic risk (Table 1).
The changes in haemodynamic and haemostatic parameters in group A, group B, group C and group D are shown in Tables 2–5, respectively. Mean arterial blood pressure was maintained at ≈55 mmHg for 80 min during PGE1-induced hypotension in groups B and D, and at ≈100 mmHg for 80 min during surgery in groups A and C. After acute haemodilution in groups C and D, haematocrit values showed a significant decrease, and final haematocrit values were 23 ± 2% in groups C and D. Minimum haematocrit values during surgery were 31% in group A and 32% in group B, respectively. There was no significant difference among the groups in total amount of i.v. LR solution given during surgery.
In group A, there was no change in PLT, PT, aPTT, FIB or AT-III during surgery, but there was a significant decrease in PLG (−27 ± 8%, P<0.05) and a significant increase in FDP (+200 ± 33%, P<0.01) 60 min after the end of surgery. In group B, PGE1-induced hypotension alone caused no significant change in PLT, PT, aPTT, FIB or AT-III during surgery, but there was a significant decrease in PLG (−16 ± 3%, P<0.05) and a significant increase in FDP (+300 ± 48%, P<0.01) at 60 min after the end of surgery. In group C, haemodilution alone caused significant decreases in PLT (−43 ± 9%, P<0.01), PT (+21 ± 5%, P<0.05), FIB (−33 ± 12%, P<0.05) AT-III (−21 ± 6%, P<0.05) or PLG (−27 ± 8%, P<0.05), and a significant increase in aPTT (+26 ± 8%, P<0.05) during surgery, but there was a significant decrease in PLG (−30 ± 10%, P<0.05) and a significant increase in FDP (+300 ± 47%, P<0.01) at 60 min after the end of surgery. In group D, haemodilution caused significant decreases in PLT (−43 ± 2%, P<0.005), PT (+21 ± 8%, P<0.05), FIB (−33 ± 9%, P<0.05), AT-III (−21 ± 5%, P<0.05), or PLG (−27 ± 7%, P<0.05), and a significant increase in aPTT (+26 ± 6%, P<0.05). The addition of PGE1-induced hypotension did not cause any further change in these values, but there was a significant decrease in PLG (−34 ± 10%, P<0.05) and in FDP (+280 ± 39%, P<0.01) at 60 min after the end of operation.
In the present study, moderate haemodilution, with a haematocrit of 23–24%, was produced by a controlled haemorrhage of 1000 mL and simultaneous substitution with 6% hydroxyethyl starch (HES), and the mean arterial blood pressure was lowered by prostaglandin E1 (PGE1) to 55 mmHg. The degree of haemodilution and hypotension was adopted because these measures have been effectively used to reduce surgical blood loss [2,5]. The duration of hypotension was about 80 min, which was the necessary time for total hip arthroplasty.
The indices of coagulation and fibrinolysis system showed no significant change during PGE1-induced hypotension alone. Acute haemodilution with HES alone caused significant changes in the indices of coagulation system, and AT-III and PLG activities. These changes would be because of the dilutional coagulopathy. However, these parameters remained within acceptable limits, and they were not of sufficient magnitude to cause any clinical disturbance. PLG activity and FDP showed significant increases after surgery in all groups, and probably caused by the surgical procedure. Because the stored autologous blood was transfused after the end of the study, its activation would not influence these results. The significant decreases of PLG in haemodilution groups probably related to the influence of haemodilution. Addition of PGE1-induced hypotension to the haemodilution did not cause further change in the values of the blood coagulation-fibrinolysis parameters and the operative blood loss compared with acute haemodilution alone. Thus, the combination of acute haemodilution and PGE1-induced hypotension is well tolerated by the blood coagulation-fibrinolysis system.
Colloids have the advantage of intravascular retention  and can maintain the stability of haemodynamics for a long period, but may also cause haemostatic defects. All artificial colloid solutions affect platelet function and dilute coagulation factors, compromising haemostasis in a dose-dependent way . Lewis et al. reported that haemodilution with dextrans or HES caused a marked impairment of platelet adhesiveness. Kobori et al. reported that HES suppresses platelet aggregation according to the infusion volume and causes a coagulopathy, i.e. decreases in PT, AT-III and FIB, and a increase in aPTT. Severe coagulopathy has been reported in a patient given 2.66 g kg−1 of HES in connection with a major orthopaedic operation. The bleeding ceased after about 20 h, but the coagulation factors did not become normal until 4 days post-operatively . Karanko  mentioned that in order to avoid haemostatic defects macromolecules can be given only up to an amount of 1–1.5 g kg−1.
Hines  demonstrated that trimethaphan provides control of arterial blood pressure with preservation of platelet function. Sodium nitroprusside and nitroglycerin, in clinically relevant dosages, significantly alter haemostatic mechanisms and inhibit platelet function [22,23]. PGE1 has a suppressive effect on platelet aggregation induced by adenosine diphosphate (ADP), collagen, adrenaline and serotonin . Shio et al. reported that PGE1, one of the strongest inhibitors of platelet aggregation, inhibits ADP-induced swelling and subsequent aggregation of rat platelets and stimulates the recovery process from the swollen state without affecting the metabolism of added ADP. Ashby  reported that in a model of prostaglandin regulation of intact platelet adenylate cyclase receptors, PGE1 binds to a stimulatory receptor, leading to rapid activation of adenylate cyclase. It also binds very tightly to an inhibitory receptor, activation of which results in slow inhibition of adenylate cyclase. Further phosphodiesterase activity may vary with prostaglandin concentration in a rapid fashion in the same model . Boullin et al. concluded that ADP included platelet aggregation caused by binding to specific receptors probably located on the plasma membrane, and that PGE1 inhibited this effect by interfering with the ADP binding. In spite of these findings, small doses (0.05 to 0.1 μg kg−1 min−1) of i.v. PGE1 were found to have no discernible effect on human platelet aggregation. Tanaka et al. reported that platelet aggregation was not suppressed during clinical PGE1-induced hypotension, and Nomura et al. demonstrated that the coagulation-fibrinolysis system also showed no significant changes in thrombelastography during PGE1-induced hypotension at a clinical dosage.
These results suggest that acute haemodilution with HES causes a slight coagulopathy, which is not enhanced by PGE1-induced hypotension.
1 Wong KC, Webster LR, Coleman LR, Coleman SS, Dunn HK. Hemodilution and induced hypotension for insertion of a Harrington rod in a Jehova's Witness patient. Clin Orthopaedics
2 Hur Su-R, Huizenga BA, Major M. Acute normovolemic hemodilution combined with hypotensive anesthesia and other techniques to avoid homologous transfusion in spinal fusion surgery. Spine
3 Amaranath L, Cascorbi HF, Singh-Amaranath AV, Frankmann DB. Relation of anesthesia to total hip replacement and control of operative blood loss. Anesth Analg
4 Fahmy NR. Nitroglycerin as a hypotensive drug during general anesthesia. Anesthesiology
5 Barbier-Böhm G, Desmonts JM, Couderc E, Moulin D, Prokocimer P, Olivier H. Comparative effects of induced hypotension and normovolaemic haemodilution on blood loss in total hip arthroplasty. Br J Anaesth
6 Miller ED. Deliberate hypotension. In: Miller RD, ed. Anesthesia,
2nd Edn, Vol. 3. New York: Churchill Livingstone, 1989; 1349–1367.
7 Nakano J, McCurdy JR. Cardiovascular effects of prostaglandin E1. J Pharmacol Exp Ther
8 Goto F, Otani E, Kato S, Fujita T. Prostaglandin E1 as a hypotensive drug during anaesthesia. Anaesthesia
9 Yukioka H, Asada K, Fujimori M, Shimazu A. Prostaglandin E1 as a hypotensive drug during general anesthesia for total hip replacement. J Clin Anesth
10 Abe K, Fujino Y, Demizu A et al.
The effect of prostaglandin E1 on local cerebral blood flow during cerebral-aneurysm clip ligation. Eur J Anaesthesiol
11 Tremper KK. Techniques and solutions to avoid homologous blood transfusion. ASA Annual Refresher Course Lectures,
vol. 214. Park Ridge, IL: American Society of Anesthesiologists, 1993: 1–7.
12 Kinlough-Rathboth RL, Packham NA, Mustard JF. The effect of Prostaglandin E1 on platelet function in vivo and in vitro. Br J Haematol
13 Lewis JH, Szeto ILF, Bayer WL, Takaori M, Safar P. Severe hemodilution with hydroxyethyl starch and dextran. Arch Surgery
14 Muzaffar TZ, Stalker AL, Bryce WAJ, Dhall DP. Structural alterations in fibrin clots with dextran. Thromb Diath Haemorrh
15 Friberger P, Knos M. Chromogenic Peptide Substrates,
Kakkar VV, ed. London: Churchill Livingstone, 1979: 128.
16 Allington MJ. Detection of fibrin (ogen) degradation products by a latex clumping method. Scand J Haemat
17 Stehling L. Autotransfusion and hemodilution. In: Miller RD, ed. Anesthesia,
3rd edn, vol. 2. New York: Churchill Livingstone, 1990: 1501–1511.
18 Karanko MS. Plasma volume substitution. Acta Anaesthesiol Scand
Suppl 89: 54–57.
19 Kobori M, Hosoyamada A, Satoh N, Kouno S, Gomi K. Effect of hemodilutional autotransfusion on coagulative-fibrinolytic dynamics and metabolic response. Masui
20 Lockwood DNJ, Bullen C, Marchin SJ. A severe coagulopathy following volume replacement with hydroxyethyl starch in a Jehova's Witness. Anaesthesia
21 Hines R. Preservation of platelet function during trimethaphan infusion. Anesthesiology
22 Hines R, Barash PG. Infusion of sodium nitroprusside induces platelet dysfunction in vitro. Anesthesiology
23 Lichtenal P, Rossi E, Louis G et al.
Dose related prolongation of the bleeding time intravenous nitroglycerin. Anesth Analg
24 Shio H. Plasse AM, Ramwell PW. Platelets swelling and prostaglandins. Microvas Res
25 Ashby B. Model of Prostaglandin-regulated cyclic AMP metabolism in intact platelets: Examination of time-dependent effects on adenylate cyclase and phosphodiesterase activities. Mol Pharmacol
26 Boullin DJ, Green AR, Price KS. The mechanism of adenosine diphosphate induced platelet aggregation, binding to platelet receptors and inhibition of binding aggregation by Prostaglandin E1. J Physiol
27 Tanaka M, Hosokawa T, Tanaka Y, Miyazaki M. The Platelet aggregation during hypotensive anesthesia using prostaglandin E1. Masui
28 Nomura Y. Effects of induced hypotensive anesthesia on the blood coagulation-fibrinolysis system measured by thromboelastography. Comparison between prostaglandin E1 and trimethaphan. Masui