- Email: [email protected]
Ascites dynamics in cirrhosis
Journal of Hepatology, 1992; 16:369-375 ©1992 ElsevierScientificPublishers Ireland Ltd. All rights reserved. 0168-8278/92/$05.00
369
HEPAT 01116
Ascites dynamics in cirrhosis Proposal and validation of a methylene blue dilution test A. Milani, A.M. Ciammella, C. Degen, M. Siciliano and L. Rossi lstituto di Patologia Medica, Unioersit~ Cattolica S. Cuore, Rome, Italy
(Received28 December 1991)
In order to investigate ascites dynamics, we examined a simple compartmental model based on the analysis of the peritoneal clearance rate of methylene blue in 58 patients with cirrhosis and ascites. After abdominal injection of 10-50 mg methylene blue, the ascitic concentration of the dye progressively decreased, following an exponential trend. The dye distribution volume (7.1+0.6 1, mean + SE) and its peritoneal clearance (87.6 + 5.0 ml/min) were determined by mathematical analysis. The accuracy of volume determinations was controlled in 5 subjects by total paracenteses. In 6 patients, methylene blue clearances were also compared with free-water peritoneal clearances, estimated by a deuterium oxide dilution technique. The decrease in the ascitic concentrations of both tracers followed a corresponding exponential decay in all patients, and the parameters of ascites dynamics determined by the two techniques (peritoneal volumes and clearance values) showed no statistical difference. We therefore suggest the use of the methylene blue dilution test to estimate free-water transperitoneal dynamics, which may be useful in the evaluation and control of cirrhotic patients with ascites. K e y words." Cirrhosis; Ascites turnover; Compartmental model; Dilution method
Ascitic fluid in cirrhosis constitutes a continuously circulating pool, the actual volume of which results from the steady state between the formation and reabsorption flows (1). However, the mechanisms and rates of ascites reabsorption are difficult to assess (1-15), and not yet completely understood. In this study we examined a simple dilution method to investigate ascites turnover. The procedure is based on a compartmental analysis of the peritoneal clearance rate of a suitable dye, following its injection into the ascitic fluid. After preliminary tests with different substances, we chose methylene blue, (MB) as a convenient tracer. MB is a low-molecular-weight dye (MW = 335.9), hydrosoluble, inexpensive and easily measurable by photometrical techniques. This substance diffuses freely along transmembrane concentration gradients, with no energydependent carrier. Because of its physico-chemical characteristics, the dye is meant to follow passively the water flows across the biological membranes. Therefore, it may
be considered a suitable tracer for the investigation of free-water transperitoneal movements. The aim of this study was to test the MB dilution method as a simple procedure for the quantitative evaluation of ascites dynamics in patients with cirrhosis. MB dilution test results were verified by comparing the peritoneal clearance of MB with that of deuterium oxide (DO), which allows a direct estimation of free-water peritoneal clearance.
Materials and Methods Theory
In the mono-compartmental model used in our study, the peritoneal cavity is the compartment into which the tracer substance is injected. The tracer is supposed to diffuse passively into the plasma and the extracellular fluids along the transperitoneal concentration gradients,
Correspondence to: Dr. AlessandroMilani,Istitutodi Patologia Medica, Universit~CattolicaS. Cuore, L.go.A. Gemelli8, 00168 Rome,Italy.
370
with eventual urinary excretion. After a MB bolus injection, the intraperitoneal concentrations (C) of the dye should therefore decrease progressively. The mono-compartmental theory assumes that the fraction of MB leaving the peritoneum per time unit is constant. Accordingly, after uniform and rapid peritoneal mixing of the dye, C decrement is exponential (C--a'eU), with a linear trend in a semilogarithmic plot. The slope k (also referred to as 'fractional turnover rate' of the dye) and the y-axis intercept a of the curve may be estimated. The slope k may also be expressed as its reciprocal, MB peritoneal half-life (MBt/2 = - l n
2/k). The intercept a (zero-time extrapolation of C) represents the initial theoretical concentration of the dye, ideally obtainable if MB should immediately and homogeneously diffuse into the whole compartment at the very instant of the infusion, prior to loss from any route. The theoretical value a allows therefore an estimation of the dye distribution volume V (MB injected dose/a) and MB peritoneal clearance per minute (V.k).
Subjects We studied 62 patients with cirrhosis and ascites (44 males and 18 females, aged between 39 and 78) consecutively admitted over a period of 30 months to the Department of Medicine, Policlinico A. Gemelli of Rome. The test was completed successfully in 58 patients, since in 4 subjects the tapping of an adequate number of samples for MB test parameter estimations was prevented, due to an insufficient amount of ascites. Diagnoses were established by histological (16 patients) or clinical and laboratory (42 patients) criteria. Cirrhotic patients were etiologically grouped into posthepatitic (type B, 7 patients; type C, I1 patients), alcoholic (27 patients) and cryptogenetic (13 patients). The investigated subjects belonged to Child-Pugh Class B (12 patients) or C (46 patients) respectively (16). According to the Declaration of Heisinki, detailed written consent was obtained from all patients, and the protocol was approved by the local Ethical Committee. None of the patients underwent previous paracenteses within the last 4 weeks. The NaC1 diet intake was 75 mEq/d, and diuretics were discontinued 5 or more days before the test.
Experimental procedure Paracenteses were carried out with a 18-gauge needle. Small ascites samples were tapped for routine analysis, standard dilution curve and blank specimens. Subsequently, 10-50 mg of sterile 1:200 MB water solution (J. Monico Lab., Mestre, Italy) were injected as a bolus
A. M I L A N I et a[.
into the peritoneal cavity, with repeated aspirations and re-injections to ensure mixing of the dye. Individual doses were chosen on the basis of a subjective estimate of the ascitic volume, in order to obtain initial MB peritoneal concentrations of about 3-5 mg/l (10-15 /~mol/1). Higher doses may cause abdominal tenderness and should therefore be avoided. Patients were allowed to rest for about 20 min, to achieve a homogeneous peritoneal distribution of the dye; 2-3 ml ascites were then periodically withdrawn through the same needle over the following 60 min, at intervals of about 5-10 min. The samples were centrifuged (10 min at 1000 x g), and the supernatant absorbance (660 nm) was determined using basal ascites as blank. The actual dye concentrations were calculated using a standard dilution curve (MB in basal ascites at known concentrations ranging from 0.5 to 4 mg/1). All determinations were carried out twice; the data were then analyzed according to the above-mentioned compartmental model. In 5 voluntary patients (3 males and 2 females), ascitic samples were simultaneously tapped from the injection needle and from a second abdominal site located 20 or more cm apart. The purpose of the procedure was to verify the adequacy of dye mixing. In the same subjects, the actual ascites volumes were directly measured performing a total paracentesis immediately after the MB test conclusion (administering 3 g i.v. of albumin per litre of ascites drained). These volume determinations were then compared with the volume estimates obtained by means of MB dilution test.
Deuterium oxide study Deuterium is a stable, non-radioactive, high-density isotope of hydrogen. Its oxide (DO or 'heavy water', density 1.108 at 25 °C) is an ideal tracer to study freewater movements, since it is fully hydrosoluble and completely assimilable to the water itself from a metabolic and chemical point of view. To compare the MB and DO estimates of peritoneal water clearance, we performed contemporaneous MB and DO dilution tests in 6 consecutive cirrhotic patients (5 males and l female, aged between 53 and 72 years). An MB dilution test was carried out according to the above-mentioned procedure, by simultaneously injecting 200 g of sterile DO (Sigma Chemical Co., St. Louis, MO, USA; cat. no. D-5159). The DO content in serial ascites samples can be measured by sophisticated and expensive massspectrometry techniques. The DO percentage in biological fluids may also be determined in a cheaper and easier
ASCITES DYNAMICS IN CIRRHOSIS way (17) using a densimetric approach ('falling drop' technique), since the relative densities of the samples are fairly proportional to their DO concentrations. Even though the densimetric method is less precise than mass spectrometry, duplicate determinations indicate mass spectrometry and densimetric methods to be accurate to within 0.2-0.5% (17). The 'falling drop' apparatus consists of an 80-cm glass tube, immersed in a 28 °C thermostated bath, and filled with 4-orthofluorotoluol (Merck-Schuchardt, art. no. 821802), an organic fluid immiscible with water, with a density of about 0.995 at 28 °C. A 20-td drop of the ascites sample is released onto the surface of the liquid by a micropipette, so that it progressively falls towards the bottom of the tube, due to its higher density. The rate of fall of the drop between two mires marked 60 cm apart on the glass tube was measured by means of 0.01 s timing. To minimize any possible errors caused by small differences in drop volumes, the mean timing of several drops was calculated. The mean falling times were then converted into actual DO concentrations, referring to a standard curve (DO in basal ascites). The decrement curves of DO ascitic concentrations were analyzed using the monocompartmental model des~:ribed above.
Statistical analysis Non-normally-distributed data were subjected to logarithmic transformation. Comparison between MB and DO results and that between MB-estimated and actual ascites volumes were performed according to the procedure described by Bland and Altman (18), plotting the differences between the parameters against their mean. The statistical differences between parametric data were tested by one-way analysis of variance.
371 TABLE 1 Methylene blue dilution test results (58 subjects) Parameter Mean SEM Range Ascites volume (I} 7.1a 0.7-20.2 MB half-life (min) 69.3~ II-608 MB clearance (ml/min) 87.6 4.99 14-187 'Non-normally distributed data: median instead of mean is reported.
approached those from the injection site, thus indicating adequate abdominal mixing. Both volume and fractional turnover data (and therefore MB half-lives) showed a non-normal distribution, and so required a log transformation. Neither MB distribution volumes nor clearance values showed significant differences as regards age, sex or cirrhosis etiology. There was no correlation between MB dilution test parameters and routine liver laboratory tests (albumin, cholesterol, prothrombin time, AST, ALT, alkaline phosphatase, bilirubin, ";-glutamyltranspeptidase), or plasma and urinary Na +, K +, creatinine and aidosterone values. Simultaneous MB and DO dilution tests were performed in 6 subjects, without any appreciable sideeffects. The decrease in ascitic concentrations of both tracers followed a similar exponential decay. Figures I and 2 show the differences between MB and DO estimates (volume and clearance) against their mean, as well as the bias and confidence interval according to Bland and Altman (18). Log transformation was used for volume data (Fig. I), which were non-normally distributed. The accuracy of M B-estimated volumes in comparison with actual ascites volumes was examined in 5 patients who underwent total paracentesis (Table 2, bottom, and Fig. 3).
Results Discussion The M B dilution test was well tolerated by all patients. MB transperitoneal reabsorption was confirmed by the excretion of greenish urine in the hours following the test. MB dilution test results are summarized in Table !. The decrease in MB ascitic concentrations corresponds to a monoexponential function, approaching a linear trend in a semilogarithmic plot within 20-25 min from MB injection. During this period (two-site simultaneous paracentesis, Table 2), MB concentrations in the samples drawn from the non-injections site progressively
On the basis of our results, the MB dilution test may be regarded as a simple and safe procedure to estimate free-water peritoneal turnover in cirrhosis. A number of studies in the literature concern the evaluation of ascites turnover through the analysis of peritoneal exchange of protein-bound substances (bromosulfonphthalein, p-aminohippurate, radiolabelled albumin, etc.). However, dilution techniques which use protein-bound indicators can only estimate the ascitic free-water turnover indirectly, even though a fair corre-
A. MILANI et al.
372 TABLE 2 Results of methylene blue (MB) test (double-site paracentesis) in 5 subjects' MB ascitic concentrations (mg/ml)
Min after MB injection 5 I0
Subject 1 _b
Subject 2 _b
3.2 (2.6)
10.2 (6.8)
15 2.9 (2.8) 8.6 (7.4) 20 2.5 (2.6) 7.0 (6.7) 25 2.2 (2.3) 5.9 (5.7) 30 2.1 (2.1) 5.5 (5.4) 35 2.0 (2.2) 5.2 (5.1) 40 1.9 (2.0) 4.8 (4.9) 45 1.8 (1.8) 4.3 (4.4) 50 1.7 (1.7) 4.2 (4.2) 55 b 3.8 (3.9) Volume (1) 9.0 (8.9) 2.9 (3.1) Error 0.43 (0.37) 0.08 (0.08) Clearance (ml/min) 126 (121) 43 (40) Error 17.7 (15.6) 3.1 (2.8) Total paracentesis ( 1) 8.5 2.5 ~Results of samples drained from the second site are given in parentheses. bMissing values.
Subject 3 5.3 (3.5)
Subject 4 2.7 (0.8)
Subject 5 8.7 (3.5)
5.0 (4.2)
2.5 (1.2)
8.0 (5.2)
4.9 (4.0) 4.8 (4.1) 4.5 (4.4) 4.4 (4.2) 4.3 (4.1) 4.1 (4.0) 3.9 (3.9) 3.8 (3.7) _b 7.3 (7.8) 0.15 (0.14) 53 (50) 4.8 (4.4) 7.3
2.4 (1.6) 2.3 (1.9) 2.2 (2.0) 2.1 (2.0) 2.1 (2.0) 2.0 (1.9) _b 1.9 (1.8) 1.9 (1.7) 16.2 (16.9) 0.36 (0.60) 81 (95) 10.7 (17.9) 18.0
6.0 (4.6) 4.8 (4.7) 4.3 (4.2) 3.6 (3.8) 3.3 (3.3) 3.1 (3.0) 2.7 (2.8) 2.5 (2.3) 2.2 (2.3) 4.9 (4.9) 0.20 (0.20) 105 (105) 9.4 (9.5) 4.6
,m i nm -s"
.4
Mean
3
o
2
E o
.1
>
2
SO
.16
"'
..........................................................
Jo
+
(.9
Mean
.03
0
~-.1
¸
~
1
_.2
g o'1
5 -.10
Mean
-
2
SO
t~-,3
cm-.4 I
i
io Average d l s t r l b u t l o n volume by MB and DO ( l l t r e s )
I00
Fig. I. Difference against mean for distribution volumes determined by simultaneous methylene blue and DO dilution test (log-transformed data). spondence between the peritoneal reabsorption rates of fluids and proteins has been observed (3,4,9,10,19,20). O n the contrary, M B is not b o u n d to plasma proteins in ascites, but remains in free solution, following the transperitoneal water movements passively. The differences in water solubility may explain why MB peritoneal disappearance rates are slightly higher (mean MB half-life 69 rain, Table 1) than those reported for other possible protein-bound tracers (7). Moreover, the observed clearance values (14-187 ml/min, i.e. 8-11 l/h, Table 1) fully correspond to the ascitic free-water turnover reported in tritium experimental studies (6). Therefore, MB seems to be particularly suitable for the direct investigation of transperitoneal free-water flows.
In our study, the actual decrease in MB intraperitoneal concentrations was always very close to that predicted on the basis of the theoretical monoexponential function. The occurrence of h o m o g e n e o u s dye mixing and the absence of significant abdominal s u b c o m p a r t m e n t s are further confirmed by the results of double-site simultaneous paracentesis (Table 2). These data indicate that MB undergoes passive peritoneal reabsorption and that the adopted m o n o c o m p a r t m e n t a l model represents a proper method of investigation. The suitability of MB as a tracer for the study of transperitoneai free-water flows is definitely proven by the similarity between the results of simultaneous MB and D O dilution tests (Figs. 1 and 2). Although the
ASCITES D Y N A M I C S IN C I R R H O S I S
373
6O .............................................................. I
M ~ ' a ~ " ; ~ s b ....
40
rn
29 20
• __.,_. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Mean .............
5d
0 c
-20
~
c (D
-19
Mean - 2 SO -40
...............................................................................
-60 i
0
20
i
t
I
I
.
I
I
i
i
40 60 80 I00 120 i40 160 180 Average water clearance by M8 end DO (ml/mln)
I
.
200
Fig. 2. Difference against mean ~ r p e r i t o n e a l c l e a r a n c e v a l u e s d e t e r m i n e d bysimultaneous methylene blue and DO dilutiontest.
.e4
eD 0
.2o .is
"~ .lo i
Mean + 2 SO
.05 . . . . . . . . . . . . . . . . . . . . . . . .~. . . . . . . . . . . . . . . . . . . . .
0 ...........
-- . . . . . .
:" . . . . . . . . . . . . . . . . . . . . M . .e an. . . . . . . . . . .
~
-~ . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . ..~. . ~
[
.03
01~" _L_ .01 u~
...........
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
•
o'~
Mean - 2 SO
,--,- .05 o
~-.lo
~-.I5 © o
~-~o f_
.
l rm
,
,
,
,
.
.
i
.
.
iO Average volume by HB and total peracontesls
.
.
.
.
.
i
I00 (lltres)
Fig. 3. Difference against mean for ascites volumes determined by methylene blue dilution test and total paracentesis (log-transformed data).
latter is not an established method, DO may be considered as a reference tracer for the study of water movements, since it is completely assimilable to water itself from a metabolic and chemical point of view, actually representing a 'tracer water'. DO ascitic concentrations were titrated using a poorly sensitive densimetric method. However, since DO is completely safe, we were able to make use of high doses of this tracer (200 g). In this way, peritoneal D O concentrations were achieved far beyond the limits of accuracy of the employed densimetric method, thus making it possible to avoid using expensive mass spectrometry techniques. The similarity between MB and D O peritoneal reab-
sorption rates was verified in a small group of ascitic patients. However, the wide range of values of the investigated parameters in the 6 subjects studied (volumes 2.9-19.3 1; half-life 25.1-547.4 min; clearance 9.2-175.6 ml/min) allowed the MB test to be validated in a variety of clinical conditions, so that the sample can be considered as representative of general population variance. Despite the limited data available, a convincing correlation was also observed between MB-estimated and total paracentesis volumes (Table 2, bottom and Fig. 3). The MB dilution test provides a quantitative estimate of several parameters of ascitic dynamics (MB distribution volume, slope k, MB half-life and peritoneal clear-
374
A. MILANI et al.
ance). However, all information is given by only two parameters (volume and peritoneal clearance), the others being simple intermediate results or mathematical derivatives. An accurate determination of the dye distribution volume (i.e. the actual ascites volume) is useful in the evaluation and follow-up of patients with ascites, since conventional clinical estimates of ascitic volume are affected by the possible presence of gaseous distension (abdominal girth) or peripheral oedema (total body weight). However, plain volume determinations investigate a 'static' feature of ascitic fluid, and provide no further information than sonographic volume estimation techniques. The parameters related to the dye disappearance rate (slope k or MB half-life) seem to be of greater interest. The MB fractional turnover rate (k) represents the percentage dye-concentration decrease per minute in the infusion compartment, due to MB passive peritoneal reabsorption. However, this value has no intrinsic significance, since it depends on the amount of ascites volume. Actually, MB fractional turnover rate (k) and MB distribution volume do not represent two independent physiological parameters: in large-volume ascites, abdominal dilation cannot result in a parallel adjustment of the peritoneal exchange surface. In fact, the volume of hollow objects varies with the third power of their linear dimensions, while their surface varies only with the square. Thus, high ascites volumes may not have corresponding exchange surfaces, and result in higher MB peritoneal half-lives. The value of MB clearance, defined as the actual volume of ascites cleared per minute, has an intrinsic physiological significance, integrating the information provided by the values of volume and fractional turnover rate. MB clearance occurs due to the reabsorption of a volume of MB-containing water, which is balanced by
the influx of the same quantity of clear water, so that the total ascites volume remains in a steady-state equilibrium. MB clearance measures this exchange, providing a quantitative evaluation of free-water turnover through the peritoneal membrane. Its value is not related to the plain dimensions of the peritoneal exchange surface, but involves a dynamic concept which depends on many factors such as peritoneal vascularization, exchange membrane thickness, lymphatic drainage, intraabdominal pressure, etc. On this basis, the MB dilution test explores a peritoneum-plasma exchange aptitude, which cannot be assessed by means of the usual laboratory parameters since it is substantially unrelated to liver functional status. The suggested procedure provides a safe quantitative in vivo estimation of free-water peritoneal turnover, which was previously possible only in anesthetized animals under rigorous experimental conditions (10,20). This method may be especially useful in classifying different functional grades of ascites severity, in following up the clinical course of patients with ascites, or in testing and carefully monitoring the effects of diuretic treatment. In conclusion, a simple MB dilution test performed during a routine paracentesis may contribute to a more complete clinical picture of the patient with cirrhosis and ascites.
References
6 Prentice TC, Siri W, Joiner EE. Quantitative studies of ascitic fluid circulation with tritium-labeled water. Am J Med 1952; 12: 668-73. 7 Baker L, Puestow RC, Kruger S, Last JH. Estimation of ascitic fluid volumes. J Lab Clin Med 1952; 39: 30-5. 8 Shear L, Ching S, Gabudza GJ. Compartmentalization of ascites and edema in patients with hepatic cirrhosis. N Engl J Med 1970: 282: 1391-6. 9 Luttvack EM, Fabian RP, Mordochovich D. Role of peritoneal absorption in ascites. Surg Gynecol Obstet 1975; 141: 693-8. 10 Zink J, Greenway CV. Control of ascites absorption in anesthetized cats: effects of intraperitoneal pressure, protein, and furosemidc diuresis. Gastroenterology 1977; 73:1119-24. I1 Zink J, Greenway CV. lntraperitoneal pressure in formation and reabsorption of ascites in cats. Am J Physiol 1977; 233: 185-90. 12 Kaplan WD, Davis MA, Uren RF, Wisotsky T, La Tegola M. A model for the radionuclide measurement of ascitic fluid volumes J Nucl Med 1978; 19: 1138-41.
I McKee FW, Wilt WG, Hyatt RE, Whipple GH. The circulation of ascitic fluid. J Exp Med 1950; 91:115-20. 2 Hahn PF, Miller LL, Robscheit-Robbins F, Bale WF, Whipple GH. Peritoneal absorption; red cells labeled by radio iron hemoglobin move promptly from peritoneal cavity into circulation. J Exp Med 1944; 80: 77-82. 3 Courtice FC, Steinbeck AW. The rate of absorption of heparinized plasma and of 0.9% NaC1 from the peritoneal cavity of the rabbit and guinea pig. Aust J Exp Biol Med Sci 1950; 28: 171-82. 4 Courtice FC, Steinbeck AW. The lymphatic drainage of plasma from the peritoneal cavity of the cat. Aust J Exp Biol 1951; 29: 45-58. 5 McKee FW, Yuile CL, Lamson BC, Whipple GH. Albumin and globulin circulation in experimental ascites. J Exp Med 1952; 95: 161-9.
Acknowledgements Part of this work was supported by grants from the Ministero della Publica Istruzione 60% a.a. 1987, and of the Consiglio Nazionale delle Ricerche (no. 88.02099.04). The authors wish to thank Ms. Denise Schembri-Wismayer for her gracious and helpful assistance in revising the English text.
ASCITES DYNAMICS 1N CIRRHOSIS 13 Lill SR, Parsons RH, Buhac I. Permeability of the diaphragm and fluid reabsorption from the peritoneal cavity in the rat. Gastroenterology 1979; 76: 997-1001. 14 Lopez-Novoa JM, Rengel MA, Hernando L. Dynamics of ascites formation in rats with experimental cirrhosis. Am J Physiol 1980; 238: F353-7. 15 Rector WG, Ibarra F. Observations on the mechanism and location of ascites reabsorption in man. Am J Gastroenterol 1987; 82: 342-6. 16 Pugh RNH, Murray-Lion IM, Dawson JL, Pietroni MC, Williams R. Transection of the oesophagus for bleeding oesophageal varices. Br J Surg 1973; 60: 646-9.
375 17 Schloerb PR, Friis-Hansen BJ, Edelman IS, Solomon AK, Moore FD. The measurement of total body water in the human subject by deuterium oxide dilution. J Clin Invest 1950; 29: 1296-312. 18 Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1986; i: 307-10. 19 Kruger S, Greve DW, Schueler FW. The absorption of fluid from the peritoneal cavity. Arch int Pharmacodyn Ther 1962; 137: 173-8. 20 McKay T, Zink J, Greenway CV. Relative rates of absorption of fluid and protein from the peritoneal cavity in cats. Lymphology 1978; I1: 106-10.
369
HEPAT 01116
Ascites dynamics in cirrhosis Proposal and validation of a methylene blue dilution test A. Milani, A.M. Ciammella, C. Degen, M. Siciliano and L. Rossi lstituto di Patologia Medica, Unioersit~ Cattolica S. Cuore, Rome, Italy
(Received28 December 1991)
In order to investigate ascites dynamics, we examined a simple compartmental model based on the analysis of the peritoneal clearance rate of methylene blue in 58 patients with cirrhosis and ascites. After abdominal injection of 10-50 mg methylene blue, the ascitic concentration of the dye progressively decreased, following an exponential trend. The dye distribution volume (7.1+0.6 1, mean + SE) and its peritoneal clearance (87.6 + 5.0 ml/min) were determined by mathematical analysis. The accuracy of volume determinations was controlled in 5 subjects by total paracenteses. In 6 patients, methylene blue clearances were also compared with free-water peritoneal clearances, estimated by a deuterium oxide dilution technique. The decrease in the ascitic concentrations of both tracers followed a corresponding exponential decay in all patients, and the parameters of ascites dynamics determined by the two techniques (peritoneal volumes and clearance values) showed no statistical difference. We therefore suggest the use of the methylene blue dilution test to estimate free-water transperitoneal dynamics, which may be useful in the evaluation and control of cirrhotic patients with ascites. K e y words." Cirrhosis; Ascites turnover; Compartmental model; Dilution method
Ascitic fluid in cirrhosis constitutes a continuously circulating pool, the actual volume of which results from the steady state between the formation and reabsorption flows (1). However, the mechanisms and rates of ascites reabsorption are difficult to assess (1-15), and not yet completely understood. In this study we examined a simple dilution method to investigate ascites turnover. The procedure is based on a compartmental analysis of the peritoneal clearance rate of a suitable dye, following its injection into the ascitic fluid. After preliminary tests with different substances, we chose methylene blue, (MB) as a convenient tracer. MB is a low-molecular-weight dye (MW = 335.9), hydrosoluble, inexpensive and easily measurable by photometrical techniques. This substance diffuses freely along transmembrane concentration gradients, with no energydependent carrier. Because of its physico-chemical characteristics, the dye is meant to follow passively the water flows across the biological membranes. Therefore, it may
be considered a suitable tracer for the investigation of free-water transperitoneal movements. The aim of this study was to test the MB dilution method as a simple procedure for the quantitative evaluation of ascites dynamics in patients with cirrhosis. MB dilution test results were verified by comparing the peritoneal clearance of MB with that of deuterium oxide (DO), which allows a direct estimation of free-water peritoneal clearance.
Materials and Methods Theory
In the mono-compartmental model used in our study, the peritoneal cavity is the compartment into which the tracer substance is injected. The tracer is supposed to diffuse passively into the plasma and the extracellular fluids along the transperitoneal concentration gradients,
Correspondence to: Dr. AlessandroMilani,Istitutodi Patologia Medica, Universit~CattolicaS. Cuore, L.go.A. Gemelli8, 00168 Rome,Italy.
370
with eventual urinary excretion. After a MB bolus injection, the intraperitoneal concentrations (C) of the dye should therefore decrease progressively. The mono-compartmental theory assumes that the fraction of MB leaving the peritoneum per time unit is constant. Accordingly, after uniform and rapid peritoneal mixing of the dye, C decrement is exponential (C--a'eU), with a linear trend in a semilogarithmic plot. The slope k (also referred to as 'fractional turnover rate' of the dye) and the y-axis intercept a of the curve may be estimated. The slope k may also be expressed as its reciprocal, MB peritoneal half-life (MBt/2 = - l n
2/k). The intercept a (zero-time extrapolation of C) represents the initial theoretical concentration of the dye, ideally obtainable if MB should immediately and homogeneously diffuse into the whole compartment at the very instant of the infusion, prior to loss from any route. The theoretical value a allows therefore an estimation of the dye distribution volume V (MB injected dose/a) and MB peritoneal clearance per minute (V.k).
Subjects We studied 62 patients with cirrhosis and ascites (44 males and 18 females, aged between 39 and 78) consecutively admitted over a period of 30 months to the Department of Medicine, Policlinico A. Gemelli of Rome. The test was completed successfully in 58 patients, since in 4 subjects the tapping of an adequate number of samples for MB test parameter estimations was prevented, due to an insufficient amount of ascites. Diagnoses were established by histological (16 patients) or clinical and laboratory (42 patients) criteria. Cirrhotic patients were etiologically grouped into posthepatitic (type B, 7 patients; type C, I1 patients), alcoholic (27 patients) and cryptogenetic (13 patients). The investigated subjects belonged to Child-Pugh Class B (12 patients) or C (46 patients) respectively (16). According to the Declaration of Heisinki, detailed written consent was obtained from all patients, and the protocol was approved by the local Ethical Committee. None of the patients underwent previous paracenteses within the last 4 weeks. The NaC1 diet intake was 75 mEq/d, and diuretics were discontinued 5 or more days before the test.
Experimental procedure Paracenteses were carried out with a 18-gauge needle. Small ascites samples were tapped for routine analysis, standard dilution curve and blank specimens. Subsequently, 10-50 mg of sterile 1:200 MB water solution (J. Monico Lab., Mestre, Italy) were injected as a bolus
A. M I L A N I et a[.
into the peritoneal cavity, with repeated aspirations and re-injections to ensure mixing of the dye. Individual doses were chosen on the basis of a subjective estimate of the ascitic volume, in order to obtain initial MB peritoneal concentrations of about 3-5 mg/l (10-15 /~mol/1). Higher doses may cause abdominal tenderness and should therefore be avoided. Patients were allowed to rest for about 20 min, to achieve a homogeneous peritoneal distribution of the dye; 2-3 ml ascites were then periodically withdrawn through the same needle over the following 60 min, at intervals of about 5-10 min. The samples were centrifuged (10 min at 1000 x g), and the supernatant absorbance (660 nm) was determined using basal ascites as blank. The actual dye concentrations were calculated using a standard dilution curve (MB in basal ascites at known concentrations ranging from 0.5 to 4 mg/1). All determinations were carried out twice; the data were then analyzed according to the above-mentioned compartmental model. In 5 voluntary patients (3 males and 2 females), ascitic samples were simultaneously tapped from the injection needle and from a second abdominal site located 20 or more cm apart. The purpose of the procedure was to verify the adequacy of dye mixing. In the same subjects, the actual ascites volumes were directly measured performing a total paracentesis immediately after the MB test conclusion (administering 3 g i.v. of albumin per litre of ascites drained). These volume determinations were then compared with the volume estimates obtained by means of MB dilution test.
Deuterium oxide study Deuterium is a stable, non-radioactive, high-density isotope of hydrogen. Its oxide (DO or 'heavy water', density 1.108 at 25 °C) is an ideal tracer to study freewater movements, since it is fully hydrosoluble and completely assimilable to the water itself from a metabolic and chemical point of view. To compare the MB and DO estimates of peritoneal water clearance, we performed contemporaneous MB and DO dilution tests in 6 consecutive cirrhotic patients (5 males and l female, aged between 53 and 72 years). An MB dilution test was carried out according to the above-mentioned procedure, by simultaneously injecting 200 g of sterile DO (Sigma Chemical Co., St. Louis, MO, USA; cat. no. D-5159). The DO content in serial ascites samples can be measured by sophisticated and expensive massspectrometry techniques. The DO percentage in biological fluids may also be determined in a cheaper and easier
ASCITES DYNAMICS IN CIRRHOSIS way (17) using a densimetric approach ('falling drop' technique), since the relative densities of the samples are fairly proportional to their DO concentrations. Even though the densimetric method is less precise than mass spectrometry, duplicate determinations indicate mass spectrometry and densimetric methods to be accurate to within 0.2-0.5% (17). The 'falling drop' apparatus consists of an 80-cm glass tube, immersed in a 28 °C thermostated bath, and filled with 4-orthofluorotoluol (Merck-Schuchardt, art. no. 821802), an organic fluid immiscible with water, with a density of about 0.995 at 28 °C. A 20-td drop of the ascites sample is released onto the surface of the liquid by a micropipette, so that it progressively falls towards the bottom of the tube, due to its higher density. The rate of fall of the drop between two mires marked 60 cm apart on the glass tube was measured by means of 0.01 s timing. To minimize any possible errors caused by small differences in drop volumes, the mean timing of several drops was calculated. The mean falling times were then converted into actual DO concentrations, referring to a standard curve (DO in basal ascites). The decrement curves of DO ascitic concentrations were analyzed using the monocompartmental model des~:ribed above.
Statistical analysis Non-normally-distributed data were subjected to logarithmic transformation. Comparison between MB and DO results and that between MB-estimated and actual ascites volumes were performed according to the procedure described by Bland and Altman (18), plotting the differences between the parameters against their mean. The statistical differences between parametric data were tested by one-way analysis of variance.
371 TABLE 1 Methylene blue dilution test results (58 subjects) Parameter Mean SEM Range Ascites volume (I} 7.1a 0.7-20.2 MB half-life (min) 69.3~ II-608 MB clearance (ml/min) 87.6 4.99 14-187 'Non-normally distributed data: median instead of mean is reported.
approached those from the injection site, thus indicating adequate abdominal mixing. Both volume and fractional turnover data (and therefore MB half-lives) showed a non-normal distribution, and so required a log transformation. Neither MB distribution volumes nor clearance values showed significant differences as regards age, sex or cirrhosis etiology. There was no correlation between MB dilution test parameters and routine liver laboratory tests (albumin, cholesterol, prothrombin time, AST, ALT, alkaline phosphatase, bilirubin, ";-glutamyltranspeptidase), or plasma and urinary Na +, K +, creatinine and aidosterone values. Simultaneous MB and DO dilution tests were performed in 6 subjects, without any appreciable sideeffects. The decrease in ascitic concentrations of both tracers followed a similar exponential decay. Figures I and 2 show the differences between MB and DO estimates (volume and clearance) against their mean, as well as the bias and confidence interval according to Bland and Altman (18). Log transformation was used for volume data (Fig. I), which were non-normally distributed. The accuracy of M B-estimated volumes in comparison with actual ascites volumes was examined in 5 patients who underwent total paracentesis (Table 2, bottom, and Fig. 3).
Results Discussion The M B dilution test was well tolerated by all patients. MB transperitoneal reabsorption was confirmed by the excretion of greenish urine in the hours following the test. MB dilution test results are summarized in Table !. The decrease in MB ascitic concentrations corresponds to a monoexponential function, approaching a linear trend in a semilogarithmic plot within 20-25 min from MB injection. During this period (two-site simultaneous paracentesis, Table 2), MB concentrations in the samples drawn from the non-injections site progressively
On the basis of our results, the MB dilution test may be regarded as a simple and safe procedure to estimate free-water peritoneal turnover in cirrhosis. A number of studies in the literature concern the evaluation of ascites turnover through the analysis of peritoneal exchange of protein-bound substances (bromosulfonphthalein, p-aminohippurate, radiolabelled albumin, etc.). However, dilution techniques which use protein-bound indicators can only estimate the ascitic free-water turnover indirectly, even though a fair corre-
A. MILANI et al.
372 TABLE 2 Results of methylene blue (MB) test (double-site paracentesis) in 5 subjects' MB ascitic concentrations (mg/ml)
Min after MB injection 5 I0
Subject 1 _b
Subject 2 _b
3.2 (2.6)
10.2 (6.8)
15 2.9 (2.8) 8.6 (7.4) 20 2.5 (2.6) 7.0 (6.7) 25 2.2 (2.3) 5.9 (5.7) 30 2.1 (2.1) 5.5 (5.4) 35 2.0 (2.2) 5.2 (5.1) 40 1.9 (2.0) 4.8 (4.9) 45 1.8 (1.8) 4.3 (4.4) 50 1.7 (1.7) 4.2 (4.2) 55 b 3.8 (3.9) Volume (1) 9.0 (8.9) 2.9 (3.1) Error 0.43 (0.37) 0.08 (0.08) Clearance (ml/min) 126 (121) 43 (40) Error 17.7 (15.6) 3.1 (2.8) Total paracentesis ( 1) 8.5 2.5 ~Results of samples drained from the second site are given in parentheses. bMissing values.
Subject 3 5.3 (3.5)
Subject 4 2.7 (0.8)
Subject 5 8.7 (3.5)
5.0 (4.2)
2.5 (1.2)
8.0 (5.2)
4.9 (4.0) 4.8 (4.1) 4.5 (4.4) 4.4 (4.2) 4.3 (4.1) 4.1 (4.0) 3.9 (3.9) 3.8 (3.7) _b 7.3 (7.8) 0.15 (0.14) 53 (50) 4.8 (4.4) 7.3
2.4 (1.6) 2.3 (1.9) 2.2 (2.0) 2.1 (2.0) 2.1 (2.0) 2.0 (1.9) _b 1.9 (1.8) 1.9 (1.7) 16.2 (16.9) 0.36 (0.60) 81 (95) 10.7 (17.9) 18.0
6.0 (4.6) 4.8 (4.7) 4.3 (4.2) 3.6 (3.8) 3.3 (3.3) 3.1 (3.0) 2.7 (2.8) 2.5 (2.3) 2.2 (2.3) 4.9 (4.9) 0.20 (0.20) 105 (105) 9.4 (9.5) 4.6
,m i nm -s"
.4
Mean
3
o
2
E o
.1
>
2
SO
.16
"'
..........................................................
Jo
+
(.9
Mean
.03
0
~-.1
¸
~
1
_.2
g o'1
5 -.10
Mean
-
2
SO
t~-,3
cm-.4 I
i
io Average d l s t r l b u t l o n volume by MB and DO ( l l t r e s )
I00
Fig. I. Difference against mean for distribution volumes determined by simultaneous methylene blue and DO dilution test (log-transformed data). spondence between the peritoneal reabsorption rates of fluids and proteins has been observed (3,4,9,10,19,20). O n the contrary, M B is not b o u n d to plasma proteins in ascites, but remains in free solution, following the transperitoneal water movements passively. The differences in water solubility may explain why MB peritoneal disappearance rates are slightly higher (mean MB half-life 69 rain, Table 1) than those reported for other possible protein-bound tracers (7). Moreover, the observed clearance values (14-187 ml/min, i.e. 8-11 l/h, Table 1) fully correspond to the ascitic free-water turnover reported in tritium experimental studies (6). Therefore, MB seems to be particularly suitable for the direct investigation of transperitoneal free-water flows.
In our study, the actual decrease in MB intraperitoneal concentrations was always very close to that predicted on the basis of the theoretical monoexponential function. The occurrence of h o m o g e n e o u s dye mixing and the absence of significant abdominal s u b c o m p a r t m e n t s are further confirmed by the results of double-site simultaneous paracentesis (Table 2). These data indicate that MB undergoes passive peritoneal reabsorption and that the adopted m o n o c o m p a r t m e n t a l model represents a proper method of investigation. The suitability of MB as a tracer for the study of transperitoneai free-water flows is definitely proven by the similarity between the results of simultaneous MB and D O dilution tests (Figs. 1 and 2). Although the
ASCITES D Y N A M I C S IN C I R R H O S I S
373
6O .............................................................. I
M ~ ' a ~ " ; ~ s b ....
40
rn
29 20
• __.,_. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Mean .............
5d
0 c
-20
~
c (D
-19
Mean - 2 SO -40
...............................................................................
-60 i
0
20
i
t
I
I
.
I
I
i
i
40 60 80 I00 120 i40 160 180 Average water clearance by M8 end DO (ml/mln)
I
.
200
Fig. 2. Difference against mean ~ r p e r i t o n e a l c l e a r a n c e v a l u e s d e t e r m i n e d bysimultaneous methylene blue and DO dilutiontest.
.e4
eD 0
.2o .is
"~ .lo i
Mean + 2 SO
.05 . . . . . . . . . . . . . . . . . . . . . . . .~. . . . . . . . . . . . . . . . . . . . .
0 ...........
-- . . . . . .
:" . . . . . . . . . . . . . . . . . . . . M . .e an. . . . . . . . . . .
~
-~ . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . ..~. . ~
[
.03
01~" _L_ .01 u~
...........
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
•
o'~
Mean - 2 SO
,--,- .05 o
~-.lo
~-.I5 © o
~-~o f_
.
l rm
,
,
,
,
.
.
i
.
.
iO Average volume by HB and total peracontesls
.
.
.
.
.
i
I00 (lltres)
Fig. 3. Difference against mean for ascites volumes determined by methylene blue dilution test and total paracentesis (log-transformed data).
latter is not an established method, DO may be considered as a reference tracer for the study of water movements, since it is completely assimilable to water itself from a metabolic and chemical point of view, actually representing a 'tracer water'. DO ascitic concentrations were titrated using a poorly sensitive densimetric method. However, since DO is completely safe, we were able to make use of high doses of this tracer (200 g). In this way, peritoneal D O concentrations were achieved far beyond the limits of accuracy of the employed densimetric method, thus making it possible to avoid using expensive mass spectrometry techniques. The similarity between MB and D O peritoneal reab-
sorption rates was verified in a small group of ascitic patients. However, the wide range of values of the investigated parameters in the 6 subjects studied (volumes 2.9-19.3 1; half-life 25.1-547.4 min; clearance 9.2-175.6 ml/min) allowed the MB test to be validated in a variety of clinical conditions, so that the sample can be considered as representative of general population variance. Despite the limited data available, a convincing correlation was also observed between MB-estimated and total paracentesis volumes (Table 2, bottom and Fig. 3). The MB dilution test provides a quantitative estimate of several parameters of ascitic dynamics (MB distribution volume, slope k, MB half-life and peritoneal clear-
374
A. MILANI et al.
ance). However, all information is given by only two parameters (volume and peritoneal clearance), the others being simple intermediate results or mathematical derivatives. An accurate determination of the dye distribution volume (i.e. the actual ascites volume) is useful in the evaluation and follow-up of patients with ascites, since conventional clinical estimates of ascitic volume are affected by the possible presence of gaseous distension (abdominal girth) or peripheral oedema (total body weight). However, plain volume determinations investigate a 'static' feature of ascitic fluid, and provide no further information than sonographic volume estimation techniques. The parameters related to the dye disappearance rate (slope k or MB half-life) seem to be of greater interest. The MB fractional turnover rate (k) represents the percentage dye-concentration decrease per minute in the infusion compartment, due to MB passive peritoneal reabsorption. However, this value has no intrinsic significance, since it depends on the amount of ascites volume. Actually, MB fractional turnover rate (k) and MB distribution volume do not represent two independent physiological parameters: in large-volume ascites, abdominal dilation cannot result in a parallel adjustment of the peritoneal exchange surface. In fact, the volume of hollow objects varies with the third power of their linear dimensions, while their surface varies only with the square. Thus, high ascites volumes may not have corresponding exchange surfaces, and result in higher MB peritoneal half-lives. The value of MB clearance, defined as the actual volume of ascites cleared per minute, has an intrinsic physiological significance, integrating the information provided by the values of volume and fractional turnover rate. MB clearance occurs due to the reabsorption of a volume of MB-containing water, which is balanced by
the influx of the same quantity of clear water, so that the total ascites volume remains in a steady-state equilibrium. MB clearance measures this exchange, providing a quantitative evaluation of free-water turnover through the peritoneal membrane. Its value is not related to the plain dimensions of the peritoneal exchange surface, but involves a dynamic concept which depends on many factors such as peritoneal vascularization, exchange membrane thickness, lymphatic drainage, intraabdominal pressure, etc. On this basis, the MB dilution test explores a peritoneum-plasma exchange aptitude, which cannot be assessed by means of the usual laboratory parameters since it is substantially unrelated to liver functional status. The suggested procedure provides a safe quantitative in vivo estimation of free-water peritoneal turnover, which was previously possible only in anesthetized animals under rigorous experimental conditions (10,20). This method may be especially useful in classifying different functional grades of ascites severity, in following up the clinical course of patients with ascites, or in testing and carefully monitoring the effects of diuretic treatment. In conclusion, a simple MB dilution test performed during a routine paracentesis may contribute to a more complete clinical picture of the patient with cirrhosis and ascites.
References
6 Prentice TC, Siri W, Joiner EE. Quantitative studies of ascitic fluid circulation with tritium-labeled water. Am J Med 1952; 12: 668-73. 7 Baker L, Puestow RC, Kruger S, Last JH. Estimation of ascitic fluid volumes. J Lab Clin Med 1952; 39: 30-5. 8 Shear L, Ching S, Gabudza GJ. Compartmentalization of ascites and edema in patients with hepatic cirrhosis. N Engl J Med 1970: 282: 1391-6. 9 Luttvack EM, Fabian RP, Mordochovich D. Role of peritoneal absorption in ascites. Surg Gynecol Obstet 1975; 141: 693-8. 10 Zink J, Greenway CV. Control of ascites absorption in anesthetized cats: effects of intraperitoneal pressure, protein, and furosemidc diuresis. Gastroenterology 1977; 73:1119-24. I1 Zink J, Greenway CV. lntraperitoneal pressure in formation and reabsorption of ascites in cats. Am J Physiol 1977; 233: 185-90. 12 Kaplan WD, Davis MA, Uren RF, Wisotsky T, La Tegola M. A model for the radionuclide measurement of ascitic fluid volumes J Nucl Med 1978; 19: 1138-41.
I McKee FW, Wilt WG, Hyatt RE, Whipple GH. The circulation of ascitic fluid. J Exp Med 1950; 91:115-20. 2 Hahn PF, Miller LL, Robscheit-Robbins F, Bale WF, Whipple GH. Peritoneal absorption; red cells labeled by radio iron hemoglobin move promptly from peritoneal cavity into circulation. J Exp Med 1944; 80: 77-82. 3 Courtice FC, Steinbeck AW. The rate of absorption of heparinized plasma and of 0.9% NaC1 from the peritoneal cavity of the rabbit and guinea pig. Aust J Exp Biol Med Sci 1950; 28: 171-82. 4 Courtice FC, Steinbeck AW. The lymphatic drainage of plasma from the peritoneal cavity of the cat. Aust J Exp Biol 1951; 29: 45-58. 5 McKee FW, Yuile CL, Lamson BC, Whipple GH. Albumin and globulin circulation in experimental ascites. J Exp Med 1952; 95: 161-9.
Acknowledgements Part of this work was supported by grants from the Ministero della Publica Istruzione 60% a.a. 1987, and of the Consiglio Nazionale delle Ricerche (no. 88.02099.04). The authors wish to thank Ms. Denise Schembri-Wismayer for her gracious and helpful assistance in revising the English text.
ASCITES DYNAMICS 1N CIRRHOSIS 13 Lill SR, Parsons RH, Buhac I. Permeability of the diaphragm and fluid reabsorption from the peritoneal cavity in the rat. Gastroenterology 1979; 76: 997-1001. 14 Lopez-Novoa JM, Rengel MA, Hernando L. Dynamics of ascites formation in rats with experimental cirrhosis. Am J Physiol 1980; 238: F353-7. 15 Rector WG, Ibarra F. Observations on the mechanism and location of ascites reabsorption in man. Am J Gastroenterol 1987; 82: 342-6. 16 Pugh RNH, Murray-Lion IM, Dawson JL, Pietroni MC, Williams R. Transection of the oesophagus for bleeding oesophageal varices. Br J Surg 1973; 60: 646-9.
375 17 Schloerb PR, Friis-Hansen BJ, Edelman IS, Solomon AK, Moore FD. The measurement of total body water in the human subject by deuterium oxide dilution. J Clin Invest 1950; 29: 1296-312. 18 Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1986; i: 307-10. 19 Kruger S, Greve DW, Schueler FW. The absorption of fluid from the peritoneal cavity. Arch int Pharmacodyn Ther 1962; 137: 173-8. 20 McKay T, Zink J, Greenway CV. Relative rates of absorption of fluid and protein from the peritoneal cavity in cats. Lymphology 1978; I1: 106-10.