Amniotic fluid embolism and leukotrienes

Amniotic fluid embolism and leukotrienes Masao Azegami, M.D., and Norimasa Mori, M.D.

Miyazaki, Japan Extracts of experimental amniotic fluid embolism from rabbit lungs showed the same biologic activities as leukotrienes and a peak on high-performance liquid chromatography at the same retention time as leukotriene D•. Acute death was prevented by administration of 5-lipoxygenase inhibitor. These results suggestthe possibility that leukotrienes contribute to the clinical and pathophysiologic features of amniotic fluid embolism. (AM J OBSTET GVNECOL 1986;155:1119-24.)

Key words: Amniotic fluid embolism, leukotrienes

Amniotic fluid embolism is one of the most catastrophic complications of obstetric patients. Meyer' first described the syndrome in 1926. In the following year, Warden 2 intravenously infused rabbits with autologous amniotic fluid, which resulted in convulsions and death in 30% of the animals. He suggested that large quantities of amniotic fluid escaping into the maternal circulation was a possible mechanism of eclampsia. In 1941 Steiner and Lushbaugh' reported on eight autopsied cases of amniotic fluid embolism and experimentally produced the disease in animals. Since then, many investigators have reproduced this syndrome in various species of experimehtal animals by intravenous infusion of autologous, homologous, and human amniotic fluid. The clinical features of the illness have been well documented. The symptoms of amniotic fluid embolism are dyspnea, acute shock, and bleeding tendency. The cardiovascular symptoms of amniotic fluid embolism resemble those of anaphylaxis and in fact may be caused by anaphylaxis itself. Although the etiologic mechanism of amniotic fluid escaping into the maternal circulation seems to be fully established, mediators that cause the clinical symptoms remain unclear. Slow-reacting substance, now characterized as a leukotriene, a newly dis" covered arachidonic metabolite, is released by various stimuli, including immunologic challenge and has long been considered an important mediator of immediate hypersensitivity reactions, such as bronchoconstriction in allergic asthma and anaphylactoid shock. We examined both the occurrence ofleukotrienes in the lungs From the Department ofObstetTics and Gynecology,Miyazaki Medical College. Supported by Grants-in-Aid for Scientific Research No. 57570607 and No. 59570714 from the Ministry of Education, Science and Culture, Japan. Received for publication January 23, 1986; revised June 5, 1986; acceptedJuly 17,1986. Reprint requests: Dr. Masao Azegami, Department of Obstetrics and Gynecology,Miyazaki Medical College, Kihara, Kiyotake-Cho, Miyazaki 889-16, Japan.

of experimental amniotic fluid embolism models as an involved mediator and the alleviating effects of lipoxygenase inhibitor.

Material and methods Production of experimental amniotic fluid embolism. Thirty-six female Japanese albino rabbits weighing from 2.23 to 3.42 kg were divided into five gr0UpS according to the substance injected: amniotic fluid, saline solution, AA861 (a selective 5-lipoxygenase inhibitor), supernatant, and precipitate (Table I). To cause experimental amniotic fluid embolism without fail, human amniotic fluid collected aseptically at the time of delivery was centrifuged at 3000 rpm for 15 minutes at 4° C, and then 2 ml of precipitate was resuspended in 20 ml of supernatant (adjusted amniotic fluid).' The adjusted amniotic fluid and 22 ml of saline solution were rapidly injected in ten seconds through the marginal ear vein of the amniotic fluid group and the saline group, respectively.' The rabbits in the supernatant group were injected with 22 ml of supernatant only, and the precipitate group was injected with 22 ml of resuspended precipitate (2 ml of precipitate in 20 ml of saline solution). Rabbits of the AA861 group were orally given AA861 (20 mg/kg), followed by injection of the adjusted amniotic fluid 1 hour later." AA861 (Takeda Chemical Industries Ltd., Osaka, Japan), which is insoluble in water, was emulsified in corn oil and administered orally through a stomach tube. Immediately after death or induction of death with diethyl ether 30 minutes after injection, lungs of the three groups were removed and finely diced. Lung tissue (2.4 gm) was stirred in 120 ml of 80% ethyl alcohol in saline solution (fraction A), and another 2.4 gm in 120 ml of 80% ethyl alcohol in distilled water (fraction B). After 60 min at 0° C the mixture was centrifuged at 5000 rpm for 15 minutes and the supernatant was filtered through gauze and rotary evaporated. The residue of fraction A was resuspended in saline for bioassay and that of fraction B in methanol for high-per-

1119

1120 Azegami and Mori

November 1986 Am J Obstet Gynecol

100 80 E E Q)

60

en

c

o ~ 40

Q)

0::

20

10-10 Concentration

(M)

Fig. 1. Log dose-response curve of leukotriene 0 4 , The abscissa indicates the final concentration of leukotriene 0 4 in the organ bath. The ordinate indicates the height of the concentration in the record.

Table I. Concentration of leukotriene in experimental groups

Group

Body weight (kg)

Survival time (min; sec)

Weight of lung (gm)

Concentration of bioassay

Leukotriene HPLC (ng/g)

Amniotic fluid

2.76 ± 0.25

5.49 ± 9.38

13.0 ± 1.7

42.4 ± 28.5

24.0 ± i9.2

3.00 ± 0.29

15.1 ± 2.9

NO

NO

2.95 ± 0.11

13.2 ± 1.5

NO

NO

3.08 ± 0.29

14.8 ± 0.8

NO

ND

14.2 ± l.l

20.0 ± 17.0

12.4 ± 10.2

(n = 10)

AA861 (n = 10)

Saline solution (n = 10)

Supernatant (n = 3)

Precipitate

3.23 ± 0.12

3.44 ± 1.37

(n = 3)

HPLC = High-performance liquid chromatography. NO = Not detected.

formance liquid chromatography." Other pieces oflung were fixed in 4% formaldehyde for histologic examination. Bioassay. Hartley strain guinea pigs weighing from 300 to 400 gm were put to death by a blow to the head. A 3 cm strip of ileum was excised about 10 ern from the ileocecal junction and suspended in a 10 ml organ bath of Tyrode's solution at 30° C. The solution contained 1 urnol/L of atropine (Sigma Chemical Co., St. Louis) and 1 umol/L of methapyrilene (Sigma) and was aerated with 95% oxygen and 5% carbon dioxide. The ileum strips were maintained under resting tension of 0.5 gm and allowed to equilibrate for at least 1 hour during which the bathing fluids were changed several times." Responses were recorded with an isotonic trans-

ducer (TD-1l2S, Nihon Kohden Kogyo Co., Ltd., Tokyo) and displayed by a pen recorder (Recticorder RJG3028S, Nihon Kohden Kogyo). A selective leukotriene antagonist, FPL 55712 (Fisons Pharmaceutical Co., Leicestershire, England), was used to reverse positive contractions.' The dose-response curve of synthetic leukotriene D4 is shown in Fig. 1. High-performance liquid chromatography. Reversed-phase high-performance liquid chromatography (tC-3A, Shimadzu Ltd., Kyoto, japan) was performed on a column (4.6 x 250 mm) packed with TSK gel ODS-120T (5 JLm particle, Toyo Soda Manufacturing Co., Ltd., Tokyo). The conditions for analysis were as follows: the solvent system consisted of acetonitrile/ methyl alcohol/water/acetic acid (300: 100: 420: 0.8,

Amniotic fluid embolism and leukotrienes

Volume 155 Number 5

1121

50 LTC 4 ~40

E E

-wlJ .c 01

'..-1

LTD 4

(J)

.c

~

20

rtl (J)

0.,

10

o

40

60 80 Dose (ng)

100

120

Fig. 2. Dose peak height curve of leukotrienes C, (LTC,) and D, (LTD,). TSK gel ODS-120T column, 250 x 4.6 mm inside diameter; mobile phase, acetonitrile/methyl alcohol/water/acetic acid (300: 100: 420: 0.8, vol/vol); flow rate, 1.0 ml/min, ultraviolet detector at 280 nrn.

vol/vol) (pH 5.6, adjusted with ammonium hydroxide), Ten microliters of the sample was injected into highperformance liquid chromatography, and eluents were continuously monitored at an absorbance of 280 nm with an ultraviolet detector (SPD-2A, Shimadzu). Solvents were eluted at the speed of 1 ml/min at room temperature." Synthetic leukotrienes C, and D, donated by Takeda Chemical Industries Ltd. (Osaka, Japan) were used as standards. The dose peak height cuves of leukotrienes C, and D, are shown in Fig. 2. Results

Amniotic fluid group. The animals displayed acute dyspnea, convulsions, urinary incontinence and shock, resembling characteristic features for an anaphylactoid reaction. All rabbits except one of this group died within 4 minutes after the amniotic fluid injection, with an average survival time for nine animals of two minutes, 41 seconds. Lungs were edematous, with particulate matter being found in the pulmonary capillaries and arteries (Fig. 3, A). Extracts from lungs of this group caused a slow- and long-lasting contraction of guinea pig ileum. The contraction was reversed with a low concentration of FPL 55712 (10- 6 M), a selective slow-reacting substance antagonist, as shown in Fig. 4. The contractile activity of the lung extracts averaged 42 ± 8.5 ng/gm, similar to the activity caused by leukotriene D,. High-performance liquid chromatography of the lung extracts all indicated leukotriene D, at a retention time of 16.7 minutes (Fig. 5). However, a peak

cochromatographing with synthetic leukotriene C, was not observed. The values of the peak corresponding to snythetic leukotriene D, were 2.5 to 62.3 ng/gm with a mean value of 24.0 ng/gm. AA861 group. Some animals showed the same symptoms seen in the amniotic group, but no animal died. Lungs showed the same histologic findings as the amniotic group, with edema and intravascular debris. Leukotriene-like activities on guinea pig ileum and highperformance liquid chromatography peak were not detected in the lung extracts. Saline solution group. No animals showed any symptoms or died within 30 minutes. Histologically, the lungs were slightly edematous but intravascular debris was not observed (Fig. 3, B). Extract of the lungs showed neither biologic activity nor high-performance liquid chromatography peak. Supernatant group. Rabbits showed slight tachypnea after injection but none of this group died within 30 minutes. The lungs were slightly edematous, which was the same finding as that of the saline solution group. Leukotriene-like activity was not observed in this group. Precipitate group. The same symptoms as those of the amniotic fluid group were observed in this group. All died within 4 minutes, with the same pathologic findings as those of the amniotic fluid group. Leukotriene-like activity in the extract of the lungs was shown not only by bioassay but also by high-performance liquid chromatography.

November 1986

1122 Azegami and Mori

Am J Obst et Gynecol

Fig. 3. A, Lung of amniotic fluid gro up. Embo li in sma ll dilated arteries (arrows). (x 200.) B , Lung of saline solution gro u p. No remarkable path ologic cha nge. (x 200. )

Wash

Wa sh

J

Sample 11

5 min

I

Sample 8

Fig. 4. Responses of guinea pig ileum to sta nda rd LT D, and samples. The ileum was suspended in 10 ml of organ bath in th e presence of atropine (10 urnol /L) and methapyrilene (10 umol/L) . FPL 55712 was added to the bath buffer afte r maximal concentration.

Comment The main symptoms in amniotic fluid emb olism are d yspne a, sho ck, and bleeding tendency. Amniotic fluid embolism ha s been regarded as different fr om "common pulmonary embolism." Though cardiovascular symptoms in amniotic fluid embolism are partially caused by mechanical vascular ob struction , additiona l factors are required to explain all the sym ptoms of the disease. Various animal studies have suggested other factors such as proteolytic enzymes, histamine, serotonin, placental extract, and prostaglandins as mediators involved in de veloping symptoms. In 1968 MacMillan" injected human amniotic fluid containing 220 to 880 cells per millimeter of squamous cells into rabbits. The severity of the embolic symptoms was re-

lated to the number of cells given. However, from th e number of cells required to cause death in rabbit, the y estim ated the volume of amniotic fluid nece ssary for women to develop amniotic fluid embolism to be 7 L, which was unacceptable from a clinical point of view and suggested certain chemical mediator(s) involved in the mechanism of amniotic fluid embolism. In 1965 Attwood and Downing" observed a marked alteration of circulatory dynamics in dogs by injection of filtered meconium suspensions into the left atrium. They postulated the presence in amniotic fluid of a vasoacti ve substance such as tripsinlike proteolytic enz yme or substances such as histamine and serotonin. In 1970 Dutta et al. 10 proved the effectiveness of the antihistaminic drug Phenergan in alleviating shock in rabbit experi-

Amniotic fluid embolism and leukotrienes 1123

Volume 155 Number 5

ments with human amniotic fluid and proposed histamined-related anaphylactoid reaction as the possible mechanism in amniotic fluid embolism. In 1974 Spence and Mason" infused autologous, cell-free placental extract into rabbits with results of rapid death within 2 minutes, and autopsy proved the presence of blood clot in the heart and pulmonary vessels. In 1972 Kitzmiller and Lucas" produced acute, transient hypotension, bradycardia, and elevated central venous pressure in cats upon injection of amniotic fluid from women in labor, but not from that of women before the onset of labor. Also similar symptoms were observed in cats by injection of 10 f,Lg/kg of prostaglandin F2 " , which is present in the amniotic fluid only of women in labor. There have been no reports concerning leukotrienes in amniotic fluid embolism. In 1938 Feldberg and Kellaway" first described a substance, which they named slow-reacting substance, that caused guinea pig ileum to contract with a slower and more sustained mode of action than did histamine. Two years later Kellaway and Trethewie reported that slow-reacting substance was also released from tissue sensitized with specific antigen. Slow-reacting substance was subsequently shown to be one of several pharmacologic mediators in human asthma. In 1977 Jakschik et al. 14 demonstrated that slow-reacting substance was derived directly from arachidonic acid through the lipoxygenase pathway. Thereafter Murphy et al. I5 identified slow-reacting substance as a cysteine-containing derivative of 5-hydroxy-7,9,11,14-icosatetraenoic acid and introduced the name, "leukotriene." Slow-reacting substance has been shown to consist of leukotrienes C4 , D., and E•. It has been proved that leukotrienes are released from various tissues such as lungs, coronary and pulmonary arteries with various stimuli. Guinea pig lung releases leukotrienes during anaphylaxis, human lung by repeated freezing and thawing, perfused cat paw by the histamine releaser, compound 48/80 and human leukocyte and rat peritoneal cells by calcium ionophore A23187. The cell elements and stimuli that triggered the production of leukotrienes in our studies remain to be determined. Leukotrienes exert marked biologic effects on ileum with contraction, on trachea with constriction in vitro and airway with resistance increase in vivo, and on skin with hyperpermeability and systemic hypotension. In in vivo experiments, 10 nmol/kg of leukotriene C. and 3 nmol/kg of leukotriene D. administered in rabbit from the left jugular vein caused hypotension, but 100 nmol/kg of leukotrienes C. and D. had negligible effects on airway resistance." In various animals including rabbit a marked constriction of coronary arteries has been reported after intracardial administration of leukotrienes C., D., and E•. It has

i

i

20

10 H

o

10 20 Retention time (min)

Fig. 5. Chromatogram of LTC., LTD., and sample. Peak I had the same retention time as that of LTD•. been definitely shown by our study that the precipitate of human amniotic fluid contains certain substances that produce leukotriene in rabbit lung and that suppression of leukotriene production can overcome experimental amniotic fluid embolism. The results of our study strongly suggested that leukotrienes may play an important role as a mediator in amniotic fluid embolism. We express deep appreciation to Drs. Y. Maki, Y. Ashida, and S. Terao, Central Research Division, Takeda Chemical Industries Ltd., for technical advice. We also thank Prof. M. Katori and Dr. A. Ueno, Department of Pharmacology, Kitasato V niversity of Medicine, for valuable suggestions. REFERENCES 1. MeyerJR. Embolia pulmonar amnio caseosa. Brazil-Medico 1926;2:301-3. 2. Warden MR. Amniotic fluid as a possible factor in the etiology of eclampsia. AM J OBSTET GYNECOL 1927; 14:292-300. 3. Steiner PE, Lushbaugh Cc. Maternal pulmonary embolism by amniotic fluid.JAMA 1941;117:1245-345. 4. Koba H. Experimental studies on the amniotic fluid embolism and afibrinogenemia. Obstet Gynaecol Japonica 1959;11:673-82. 5. Ashida Y, SaijoT, Kuriki H, Makino H, Terao S, Maki Y. Pharmacological profile of AA861, a 5-lipoxygenase inhibitor. Prostaglandins 1983;26:955-72. 6. Katori M, Ueno A, Tanaka K. Roles of leukotrienes in non-allergic and allergic inflammatory models. Prostaglandins 1984;28:617-20. 7. AugsteinJ, Farmer JB, Lee TB, Sheard P, Tattersall ML. Selective inhibitor of slow reacting substance of anaphylaxis. Nature [New Bioi] 1973;245:215-7. 8. MacMillan D. Experimental amniotic fluid infusion.J Obstet Gynaecol Br Commonw 1968;75:849-52.

Azegami and Mori

9. Attwood RD, Downing SE. Experimental amniotic fluid and meconium embolism. Surg Gynecol Obstet 1965; 120:255-62. 10. Dutta D, Bhargava KC, Chakravarti RN, Dhall SR. Therapeutic studies in experimental amniotic fluid embolism in rabbits. AM] OBSTET GYNECOL 1974;106:1201-8. 11. Spence MR, Mason KG. Experimental amniotic fluid in rabbits. AM] OBSTET GYNECOL 1974;119:1073-8. 12. Kitzmiller ]L, Lucas WE. Studies on a model of amniotic fluid embolism. Obstet Gynecol 1927;39:626-7. 13. Feldberg W, Kellaway CR. Liberation of histamine and formation of Iysocithin-like substances by cobra venom. ] PhysioI1938;94:187-226.

November 1986 Am J Obstet Gynecol

14. ]akschik BA, Falkenhein S, Parker CWo Precursor role of arachidonic acid in release of slow reacting substance from rat basophilic leukemia cell. Proc Natl Acad Sci USA 1977;74:4577-81. 15. Murphy RC, Hammarstrom S, Samuelsson B. Leukotriene C: a slow reacting substance from murine mastcytoma cells. Proc Natl Acad Sci USA 1979;76:4275-9. 16. Ueno A, Tanaka K, Katori M. Possible involvement of thromboxane in bronchoconstrictive and hypertensive effects of LTC 4 and LTD 4 in guinea pigs. Prostaglandins 1982;23:865-80.

Enhancement of colonization of Ureaplasma urealyticum in the mouse genital tract by estrogen treatment Tsuyoshi Iwasaka, M.D., Toshio Wada, D.V.M., and Hajime Sugimori, M.D. Saga and Fukuoka, Japan Experimental colonization by Ureaplasma urealyticum in the vagina of Balb/c mice was greatly enhanced by administration of estrogen, whereas it was not affected by treatment with progesterone. Ureaplasma was recovered from almost half of the estrogen-treated animals 7 days after inoculation, and the organism persisted in all of these animals until 21 days after inoculation; on the other hand, the organism was isolated in less than 20% of the untreated, progesterone-treated, or danazol-treated mice 7 days after inoculation and was no longer isolated in any of these groups 21 days after inoculation. At least a hundredfold more organisms were recovered from the estrogen-treated mice as compared to the untreated mice 14 days after inoculation and thereafter. U. urealyticum was also isolated from the uterine cavity and salpinx in approximately half of the mice in which the organism was isolated from the vagina. (AM J OSSTET GYNECOL 1986;155:1124-7.)

Key words: Ureaplasma urealyticum, estrogen, mouse, genital tract Ureaplasma urealyticum has been isolated from the genital tract in women of all ages (newborn to postmenopausal).' Most of the previous reports have suggested that sexual activity is the most important factor for acquisition of U. urealyticum in the female genital tract.":' On the other hand, some investigators have suggested the possible role of hormonal condition on colonization rates of this organism.>" Our clinical studies have also suggested that the serum estrogen or progesterone level is an important factor for the proliferation and the persistence of this organism in the female genital tract." This idea is based on the observation that prepubertal girls, puerperal women, and post-

From the Department of Obstetrics and Gynecology, Saga Medical School, and the Laboratory of Animal Experiments, Kyushu University School of Medicine. Received for publication April 4, 1986; revised june 30, 1986; acceptedjuly 12, 1986. Reprint requests: Tsuyoshi 1wasaka, M.D., Department of Obstetrics and Gynecology, Saga Medical School, Sanbonsugi, Nabeshima, Saga 840-01,japan. 1124

menopausal women have a low colonization rate of this organism, whereas pregnant women showed the highest incidence of U. urealyticum colonizations. The present study was undertaken to confirm the significance of estrogen, progesterone, or both on colonization and persistence of U. urealyticum in the genital tract with use of the mouse model, in which natural U. urealyticum colonizations are not prevalent.

Material and methods Mice. Balb/c female mice, 6 weeks old and bred in the specific pathogen-free unit, were used. Each animal was checked for mycoplasmas in the genital tract by culture procedures before use in any of the experiments. Hormonal treatment. One hundred female mice were divided into five groups. In group I, estradiol (Progynon-Depot: Schering, Berlin, West Germany) was injected subcutaneously (0.5 mg in 0.05 ml). In group 2, hydroxyprogesterone (proge-Depot: Mochidia Pharmaceutical Co., Tokyo, Japan) was injected