Moise et al.
methacin: studies of absorption and placental transfer. AM j OBSTET GYNECOL 1977;129:464-5. 10. Traeger A. Noschel H. ZaumseilJ. The pharmacokinetics of indomethacin in pregnant and parturient women and their newborn infants. Zentralbl Gynakol 1973;95:635-41. 11. Duggan DE, Hogans AF, Kwan KC, McMahon FG. The metabolism of indomethacin in man. J Pharmacal Exp Ther 1972;181:563-75. 12. Evans MA, Papazafiratou C, Bhat R, Vidyasagar D. In-
February 1990 Am J Obstet Gynecol
domethacin metabolism in isolated neonatal and fetal rabbit hepatocytes. Pediatr Res 1981;15:1406-10. 13. Thalji AA, Carr I, Yeh TF, Raval D, LukenJA, Pildes RS. Pharmacokinetics of intravenously administered indomethacin in premature infants. J Pediatr 1980;97:9951000. 14. Kirshon B, Moise Kj, Wasserstrum N, Ou C-N. Influence of short-term indomethacin theapy on fetal urine output production. Obstet Gynecol 1988;72:51-3.
Transcapillary fluid dynamics during ovarian stimulation for in vitro fertilization Anne Tollan, MD: Nicolai Holst, MD: Finn Forsdahl, MD: Hans Olav Fadnes, MD: PhD,b Pat 0ian, MD, PhD: and Jan M. Maltau, MD, PhD" Troms;, Norway Transcapillary fluid dynamics were studied in 10 women during ovarian stimulation for in vitro fertilization. The examinations were done on the first day of stimulation (day 3 of the menstrual cycle, mean serum estradiol concentration 0.2 nmoIlL), and the day before oocyte aspiration (day 10 to 12, mean serum estradiol concentration 6.8 nmoIlL). Interstitial colloid osmotic pressure was measured on the thorax at heart level by the "wick" method, and interstitial hydrostatic pressure by the "wick-in-needle" method. Plasma colloid osmotic pressure decreased (mean, 2.0 mm Hg; p < 0.002) and interstitial colloid osmotic pressure increased (mean, 1.0 mm Hg; p < 0.02) during hormonal stimulation. This implies a reduced transcapillary colloid osmotic gradient (plasma colloid osmotic pressure - interstitial colloid osmotic pressure), probably because of increased capillary permeability to plasma proteins. Hemoglobin and hematocrit were significantly reduced, and body weight and foot volume significantly increased. These results demonstrate that during ovarian stimulation there are both water retention and augmented filtration of fluid from the vascular to the interstitial compartment. This may be of significance for the pathophysiologic condition in the ovarian hyperstimulation syndrome. (AM J OBSTET GYNECOL 1990;162:554-8.)
Key words: Fluid dynamics, in vitro fertilization, hyperstimulation syndrome, estrogens During ovarian stimulation for in vitro fertilization serum estradiol levels exceed normal preovulatory values, and mastalgia, bloating, tiredness, swelling, and, more seldomly, edema are common complaints. l These symptoms may indicate changes in body water regulation. The ovarian hyperstimulation syndrome is a rare but serious complication to ovarian stimulation. In the most severe form this syndrome is characterized by massive
From the Department of Obstetncs and Gynecology" and the Department of Medicme.' University of Troms;. Dr. Tollan was a reCipient of a research fellowship from the Norwegian Research Council for Science and the Humanities. Received for publicatIOn October 7,1988; revzsedJuly 11, 1989; accepted September 22, 1989. Repnnt requests: Anne Tollan, MD, Department of Obstetncs and G.vnecology, UniverSity of Tromsl/J, N-9000 Tromsl/J. Norway. 611116902
554
cystic enlargement of the ovaries, edema, ascites, pleural effusion, hemoconcentration, and electrolyte imbalance. I. 2 In a less severe form only enlargement of the ovaries and fluid retention occurs. I. 2 Estrogen levels are found to be substantially elevated in these patients. l • 3 It is well known that estrogens in physiologic concentrations cause a slight sodium and water retention.' Estrogens are also known to reduce the total peripheral resistance. 5 . 6 An excessive estrogen level has been shown to increase the capillary permeability to plasma proteins in rats and this effect has been suggested as the main cause of the body fluid shift in severe ovarian hyperstimulation syndrome. 7 . 8 Fluid exchange (F) across the capillary wall is regulated by the Starling forces, and can be described by the Starling equation 9 : F = CFC < (Pc - Pi) - O'(COPp - COPi) >
Volume 162 Number 2
Fluid dynamics during ovarian stimulation
555
Table I. Changes in sex hormone concentrations, weight. hemoglobin. hematocrit. and foot volume in 10 women during ovarian stimulation for in vitro fertilization Indwuiual differences before and after stlmulaton
Group mean values Before stimulatwn
Estradiol (nmoUL) Progesterone (nmoUL) Weight (kg) Hemoglobin (gm/dl) Hematocrit (%) Volumetry (ml) Right foot Left foot
I
After stzmuiatlOn
Mean
I
SD
0.2 1.7 61.3 13.4 39.5
6.8 8.2 62.1 12.8 37.3
6.6 6.5 0.8 -0.6 -2.2
3.81 8.12 0.74 0.78 2.62
1201 1170
1256 1236
55.5 56.0
77.33 71.91
where CFC, the capillary filtration coefficient, represents the volume of net filtrate formed in 100 gm of tissue per minute for each millimeter of mercury rise in net transcapilIary filtration pressure. Pc and Pi are the hydrostatic pressures in the capillaries and interstitium, respectively; IT is the reflection coefficient for plasma proteins. In subcutaneous tissue IT is between 0.9 and 1.0. indicating that the capillaries in this vascular bed are not completely impermeable to plasma proteins. 10 COPp is the plasma colloid osmotic pressure and COPi is interstitial colloid osmotic pressure. The "wick" method II and the "wick-in-needle" method 12 have made it feasible to measure interstitial colloid osmotic pressure and interstitial hydrostatic pressure, respectively under clinical conditions associated with disturbed fluid balance. 13. 14 In these studies it has been shown that measurements of the tissue forces of interstitial hydrostatic pressure and interstitial colloid osmotic pressure are essential to understand transcapillary fluid dynamics. In normal pregnant women 13 and in patients with the nephrotic syndrome. I i reduced plasma colloid osmotic pressure is compensated for by a reduction in interstitial colloid osmotic pressure, serving both as an important edemapreventing and plasma volume-preserving factor. In other conditions, for example severe preeclampsia. this mechanism fails and can explain why many of these patients have edema and reduced plasma volume. 15 In a previous study we found that transcapillary fluid balance changed from the follicular to the luteal phase in normal menstrual cycles. 16 A significant reduction in both plasma colloid osmotic pressure and interstitial colloid osmotic pressure as well as a slight weight gain from follicular to luteal phase was observed. This study describes effects of supraphysiologic concentrations of endogenous estrogens on the transcapillary fluid dynamics in women who undergo ovarian stimulation with human menopausal gonadotropin (Humegon, Organon, Oss, The Netherlands) for in vitro fertilization.
I
Stgnificance
P= P= P= P= P=
0.0004 0.027 0.010 0.030 0.026
P = 0.026
Material and methods Ten women, 25 to 35 years old, who had been accepted for in vitro fertilization because of tubal damage. volunteered for the study after informed consent had been obtained. Protocol was approved by the human research ethics committee. None of the patients had a history of hypertension, endocrine dysfunction. or renal disease, and all had regular menstrual cycles. The women received 150 IU human menopausal gonadotropin from day 3 of the menstrual cycle until appropriate response, evaluated by daily serum estradiol measurements and ultrasonographic examinations. Ovulation was induced by an injection of 9000 IU of human chorionic gonadotropin (Physex, Leo, Copenhagen). and oocyte aspiration was performed approximately 35 hours later. The examinations were done before the first human menopausal gonadotropin injection (day 3 of the menstrual cycle), and the day after ovulation induction with human chorionic gonadotropin (cycle day 10 to 12), on both occasions between 12 noon and 2 PM, with the patient in the supine position. Blood samples were drawn from an antecubital vein without stasis. Hemoglobin concentration and hematocrit were measured by a Coulter Counter analyzer (Coulter Electronics, Hialeah, Fl.). Plasma estradiol and progesterone concentrations were determined by radioimmunoassay methods. Weight was recorded on the same scale with the patient wearing standard clothing. Foot volume was measured by foot volumetry.17 Each foot was placed in a separate steel chamber filled with temperate water and the spillover was measured. Interstitial colloid osmotic pressure was determined as described by Noddeland" and 0ian et al.l3 15 Six nylon wicks were sewn subcutaneously on the side of the thorax at heart level. The length of the wicks was 50 to 60 mm. The skin was anesthetized with 0.1 ml of lidocaine (20 mg/ml. without epinephrine). After 1 hour the wicks were removed and the wick fluid was isolated by centrifugation in mineral oil. The colloid
556
Tollan at al.
February 1990 Am J Obstet Gynecol
Table II. Changes in plasma colloid osmotic pressures, interstitial colloid osmotic pressure on thorax, transcapillary colloid osmotic gradient, and interstitial hydrostatic pressure on thorax in mm Hg in 10 women during ovarian stimulation for in vitro fertilization lndimdual differences before and after stimulatIOn
Group mean values Before stnnulatwll
COPp COPi COPp - COPi Pi
27.3 13.6 13.6 - 3.0
I
After stimulation
25.3 14.6 10.6 -2.5
Mean
-2.0 1.0 -3 .U 0.5
I
SD
1.9 2 .0 3. 1 2.4
I
Szgnificance
P = 0.002
P = 0 .02* P = 0.025 NS
COP, Plasma colloid osmotic pressure ; COPt, interstitial colloid osmotic pressure; COPp - COPz, transcapillary colloid osmotic gradient; PI, interstitial hydrostatic pressure. *Analysis of variance showed significant interaction for the change during stimulation between persons.
osmotic pressure was measured by an osmometer designed for small fluid samples (5 ILl) with a membrane impermeable to proteins with molecular weights above 30,000 daltons. Plasma colloid osmotic pressure was determined by the same colloid osmometer. Interstitial fluid hydrostatic pressure was measured by the wick-in-needle technique described by Fadnes et al. 12 and 0ian et al. 13 A sterile cannula with an outer diameter of 0.6 mm was filled with loose bundles of nylon filaments. The cannula was provided with a 4 mm long side hole 5 mm from the tip, and connected to the pressure transducer by a tube filled with physiologic saline solution. After calibration the needle was inserted subcutaneously at the level of the implanted wicks, and the hydrostatic pressure was recorded. Daily individual ratings of tiredness, mastalgia, feeling of swelling, edema, and discomfort in the pelvic region were performed late in the evening by each woman on a visual analog scale from 0 to 10. Statistical methods. Differences in single measurement variables during the stimulation period were evaluated by paired l test and the replicated measurements of plasma and interstitial colloid osmotic pressu res were made by analysis of variance. Values of p < 0.05 were considered significant. Results
During ovarian stimulation the mean serum level of estradiol increased from 0.2 to 6.8 nmol/L. Serum progesterone concentrations increased from 1.7 to 8.2 nmollL (Table I). Eight women gained weight, one remained unchanged, and one lost weight. The average weight gain for the whole group was 0.8 kg (p < 0.01). Foot volume increased, and hemoglobin and hematocrit decreased significantly during stimulation (Table I). Changes in colloid osmotic and hydrostatic pressures are presented in Table II. During ovarian stimulation
plasma colloid osmotic pressure decreased 2.0 mm Hg, whereas interstitial colloid osmotic pressure increased 1.0 mm Hg. Both changes were statistically significant. There was a significant interaction (p = 0.0 I) for interstitial colloid osmotic pressure indicating that the increase is inconsistent. The colloid osmotic gradient (plasma colloid osmotic pressure - interstitial colloid osmotic pressure) across the capillary wall was significantly and consistently reduced. Individual changes in colloid osmotic gradient (plasma colloid osmotic pressure - interstitial colloid osmotic pressure) during stimulation are presented in Fig. 1. The colloid osmotic measurements were unsuccessful in one woman after ovulation induction because of technical problems with the transducer. The increase in interstitial hydrostatic pressure was not significant. The daily ratings demonstrated only modest discomfort during the stimulation period. None of the women developed ovarian hyperstimulation syndrome after ovulation induction. Comment
Ovarian stimulation for in vitro fertilization is characterized by substantially increased serum levels of estradiol compared with the natural menstrual cycle. In this study we found significant effects on the transcapiIIary fluid balance during this treatment. These effects differ from those observed during normal menstrual cycles" and are most likely caused by the high estradiol concentrations. However, other hormones that may influence fluid balance during ovarian stimulation cannot be excluded. Thus prolactin has been shown to increase significantly during this treatment,'· and prolactin may have a role in the water and electrolyte balance.'9 The significant reduction in plasma colloid osmotic pressure, hemoglobin, and hematocrit may well be explained by simple dilution, indicating that estrogens are responsible for renal fluid retention and a rise in plasma volume. According to the Starling equation, a reduction
Fluid dynamics during ovarian stimulation
Volume 162 Number 2
in plasma colloid osmotic pressure will result in a rise in net filtration pressure with loss of fluid from the vascular to the interstitial compartment, if not compensated for by a reduction in interstitial colloid osmotic pressure. Such a compensatory reduction of interstitial colloid osmotic pressure is included in the physiologic adjustments that take place during normal menstrual cycles, from follicular to luteal phase,!6 in normal pregnancy, from first to third trimester,13 and in the nephrotic syndrome," and the colloid osmotic gradiem (plasma colloid osmotic pressure - interstitial colloid osmotic pressure) remains unchanged . The reduction in interstitial colloid osmotic pressure in these physiologic situations is an important edemapreventing factor, counteracting an increase in interstitial fluid volume and thereby maintaining plasma volume. During ovarian stimulation, however, a significant reduction in plasma colloid osmotic pressure was observed without any decrease in interstitial colloid osmotic pressure; in fact, a significant increase was observed . Consequently, the plasma-interstitial colloid osmotic gradient is reduced, and this favors transport of fluid from the vascular compartment to the interstitium. On the side of the thorax, interstitial hydrostatic pressure did not change significantly. This may be because of small changes in the interstitial fluid volume in that area but also may reflect a high compliance in the subcutaneous tissue that permits large quantities of fluid to transude from the capillaries without any significant increase in interstitial hydrostatic pressure. However, there is indirect evidence of increased interstitial volume as demonstrated by the rise in foot volume and weight gain. This shows that the interstitial compensatory mechanisms are insufficient during ovarian stimulation. The objection may be made that interstitial colloid osmotic pressure and interstitial hydrostatic pressure were measured on the thorax, whereas the change in interstitial fluid volume was demonstrated as a rise in foot volume. When transcapillary fluid balance is studied in man, variations in capillary pressure caused by onhostasis, which normally causes a lower interstitial colloid osmotic pressure at the ankle than on thorax,"" must be considered. On the basis of measurements of interstitial colloid osmotic pressure"" and protein concentration in lymph,"1 it takes more than 48 hours of bed rest to achieve steady-state interstitial colloid osmotic pressure in the lower limbs. Changes in the interstitial fluid caused by changes in body posture are avoided by measuring interstitial colloid osmotic pressure and interstitial hydrostatic pressure on the thorax at heart level. Although the observed decrease in the transcapillary colloid osmotic pressure gradient on the thorax will favor net capillary filtration in this region, a rise in interstitial fluid volume will be
557
mmHg
18 16
a::
0
()
a.
0...
0
14 12
()
10 8
l'
• Before stimulation
After stimulation
Fig. 1. Individual changes In colloid osmotic gradient (plasma colloid osmotic pressure - interstitial colloid osmotic pressure) in millimeters of mercury in nine women during ovarian stimulation.
more marked in the lower limbs, where edema usually first appears in ambulatory patients, because of orthostasis. Because interstitial colloid osmotic pressul'e and probably interstitial volume are increased the day after ovulation induction, the interstitial protein mass must also be increased. This may be caused by one or more of the following factors: increased capillary permeability to plasma proteins, reduced lymph flow, or increased protein synthesis. In this study plasma volume was apparently increased. This suggests that either an increase in lymph flow, a decrease in capillary pressure, or both occurred to compensate for the decrease in the colloid osmotic plasma-interstitial gradient. Reductions in plasma colloid osmotic pressure and hypoproteinemia, have been shown to be associated with increases in lymph flow."" An increase in capillary permeability and a decrease in the colloid osmotic gradient also would be expected to cause steady-state increase in both capillary filtration and lymph flow. The third factor, increased protein synthesis, would be expected to increase both plasma and interstitial colloid osmotic pressure unless capillary permeability was also increased. Because plasma colloid osmotic pressure was reduced and interstitial colloid osmotic pressure was increased, this supports increased capillary permeability. Thus in light of all the data presented, an increase in cap-
558 Tollan et al.
illary protein permeability is the most likely mechanism. In the ovarian hyperstimulation syndrome serum estradiol levels are usually higher than during ovarian stimulation for in vitro fertilization. It is therefore tempting to speculate whether the effects on transcapillary fluid balance, demonstrated in this study, may be exaggerated in this syndrome. A further decrease in the transcapillary colloid osmotic gradient could lead to increased interstitial fluid volume (edema, pleural effusion, ascites) and hemoconcentration, all characteristic features of the ovarian hyperstimulation syndrome. In conclusion, this study shows a reduction in the colloid osmotic gradient (plasma colloid osmotic pressure - interstitial colloid osmotic pressure) in women who have ovarian stimulation for in vitro fertilization. The most likely mechanism is increased capillary permeability to plasma proteins. According to the Starling equation, if these changes are not compensated for, they will result in increased interstitial fluid volume and possibly edema formation. This may be of significance in the pathophysiologic features of the ovarian hyperstimulation syndrome.
February 1990 Am J Obstet Gynecol
7. 8. 9. 10. 11.
12. 13. 14. 15. 16. 17.
REFERENCES
1. Schenker JG, Weinstein D. Ovarian hyperstimulation syndrome. A current survey. Ferti! Steril 1978;30:255-68. 2. Thaler I, Yoffe N, Kaftory JK, Brandes JM. Treatment of ovarian hyperstimulation syndrome. The physiological basis for a modified approach. Ferti! Steril 1981;36: 110-3. 3. Haning RV, Austin CW, Carlson IH, Kuzma DL, Shapiro SS, Zweibel W]. Plasma estradiol is superior to ultrasound and urinary estradiol glucuronide as a predictor of ovarian hyperstimulation during induction of ovulation with menotropins. Fertil Steril 1983;40:31-6. 4. Guyton AC. Textbook of medical physiology. Philadelphia: WB Saunders, 1986. 5. Slater AJ, Gude N, Clarke IJ, Walters WA. Haemodynamic changes and left ventricular performance during highdose oestrogen administration to male transsexuals. Br J Obstet Gynaecol 1986;93:532-8. 6. Luotola H, Pyorala T, Lahteenmaki P, Toivanen]. Hae-
18.
19. 20. 21.
22.
modynamic and hormonal effects of short-term estradiol treatment in postmenopausal women. Maturitas 1979; 1: 287-94. Polishuk WZ, Schenker JG. Ovarian overstimulation syndrome. Fertil Steril 1969;20:443-50. Davis JS. Hormonal control of plasma and erythrocyte volumes of rat uterus. Am J Physiol 1960; 199:841-6. Starling EH. On the absorption of fluids from the connective tissue space. J Physiol (Lond) 1896;19:312-26. Aukland K, Nicolaysen G. Interstitial fluid volume: local regulatory mechanisms. Physiol Rev 1981;61:556-643. Noddeland H. Colloid osmotic pressure of human subcutaneous interstitial fluid sampled by nylon wicks: evaluation of the method. Scand J Clin Lab Invest 1982; 42:123-30. Fadnes HO, Reed RK, Aukland K. Interstitial fluid pressure in rats measured with a modified wick techniq'r,e, Microvasc Res 1977;14:27-36. 0ian P, Maltau JM, Noddeland H, Fadnes HO. Oedemapreventing mechanisms in subcutis in normal pregnant women. Br J Obstet Gynaecol 1985;92: 1113-9. Fadnes HO, Pape JF, Sundsfjord JA. A study on oedema mechanism in nephrotic syndrome. Scand J Clin Lab Invest 1986;46:533-8. 0ian P, Maltau JM, Noddeland H, Fadnes HO. Transcapillary fluid balance in pre-eclampsia. Br.J Obstet Gynaecol 1986;93:235-9. 0ian P, Tollan A, Noddeland H, Fadnes HO, MaltauJM. Transcapillary fluid dynamics during the menstrual cycle. AM J OBSTET GYNECOL 1987; 156:952-5. Nilsson S, Bjerknes Hagen G. Volume try in evaluation of swelling in the ankle and foot. J Oslo City Hosp 1981; 31:11-5. Holst N, Jenssen TG, Burhol PG, Haug E, Forsdahl F. Plasma gastrointestinal hormones during spontaneous and induced menstrual cycles. J Clin Endocrinol Metab 1989;68:1160-6. Nicol CS. Role of prolactin in water and electrolyte balance in vertebrates. In: Jaffe RB, ed. Prolactin. New York: Elsevier, 1981: 127-66. Noddeland H. Influence of body posture on transcapillary pressures in human subcutaneous tissue. ScandJ Clin Lab Invest 1982;42:131-8. Olszewski WL, Engeset A, J a:ger PM, Sokolowski I, Theodorsen L. Flow and composition of leg lymph in normal men during venous stasis, muscular activity and focal hyperthermia. Acta Physiol Scand 1977;99:149-55. Kramer GC, Harms BA, Bodai BI, Demling RH, Renkin EM. Mechanisms for redistribution of plasma proteins following acute protein depletion. AmJ PhysioI1982;243: H803-9.