Circulatie bij leverziekten : bepaling van het hartminuutvolume aan het ziekbed

SUMMARY It has been common knowledge for more than a century that changes in the structures of the vascular bed can occur in diseases of the liver. An extensive venous collateral circulation can develop in Laennec's cirrhosis of the liver and in the case of extrahepatic portal block, the most serious clinica1 complication arising from the much-feared rupture of oesophageal varices. Hepatic circulatory changes are understandable in these circumstances, and have been demonstrated by many investigators. Research carried out in the past decade, however, indicates the probability that general as well as local circulatory changes occur. Chapter 2 summarizes possible circulatory and respiratory changes in the case of diffuse lesion of the liver parenchyma, particularly Laennec's cirrhosis. They include a slight fa11 in arterial oxygen saturation, chronic hyperventilation with a concomitant fa11 in arterial pCO2, increased total blood volume and cardiac output, and decreased peripheral vascular resistance as a result of the occurrence of numerous arteriovenous anastomoses. The two last mentioned phenomena are of paramount pertinence to this study. The cardiac output appears to be increased in about one-third of patients with cirrhosis of the liver (68,3); the spread of the values found exceeds that in normal subjects. The increase is believed to be a direct effect of the functioning of arteriovenous anastomoses, and to be entirely determined by the condition of the liver parenchyma. The presence of ascites does not affect the cardiac output (67) and there is no correlation between the arterial oxygen saturation' and the increase in cardiac output 3. The functioning arteriovenous anastomoses in the digestive tract and spleen are believed a factor in the occurrence of the hyperkinetic pattern of circulation (97), which is associated with increased peripheral circulation and decreased peripheral vascular resistance. There is considerable similarity between this circulatory syndrome and wet beri-beri (Shoshin)'*. The two clinica1 conditions are also similar in the response to pitressin (1,62,78,106) and (nor-)adrenaline (71) 105). Shorr's hypothetical VDM (vasodilating material) (109)was held responsible for the opening up of arteriovenous anastomoses(l08). It is impossible to exclude the existence, in cases of liver cirrhosis, of a type of heart disease specific of this condition, characterized by segmental fibrosis of myocardial fibres (75).But this segmental fibrosis is more likely to originate from fat emboli resulting from ruptured fatty cysts in het liver sinusoids (51). Chapter 3 briefly reviews methods of determining the liver circulation. This can be measured either indirectly by means of extraction methods (l8,30,1O4,120,25) and dye dilution techniques 91, or directly with the aid of electromagnetic flowmeters (33)The extraction method using indocyanine green as indicator, is best suited to clinica1 purposes (92). The values found in this way in normal subjects average 1600 ml/min, while those in patients with cirrhosis of the liver are considerably lower (Tables 1 and 2). A brief review of the development of the indicator dilution method of determining the cardiac output, is followed by a discussion of the principle of the method and its various applications (Chapter 4). Mention is also made of factors governing the form and size of the indicator dilution curve, requirements to be met by dyes used for this purpose, and the calibration of dye dilution curves. Of the many dyes used in the indicator dilution technique, patent blue V and indocyanine green (Fox green, cardiogreenB) proved to be most suitable (129). Sparling's dynamic calibration procedure was found to be the method of choice among many calibration procedures (114)In Sparling's procedure the patient's dye dilution curve is compared with a dilution curve obtained following injection of a known smal1 dose of dye into the withdrawal system. Chapter 5 deals with a description of apparatus and method used in bedside determination of the cardiac output. To ensure a rapid and accurate bedside procedure, a few changes were made in the reflection method with dynamic calibration procedure as developed by Zijlstra, Mook and Sparling. The dye was injected into a peripheral vein, initially in the conventional way but subsequently with the aid of an injection apparatus especially designed for the dye dilution technique by Ten Hoor (129). The complete arrangement is shown in Fig. 7; 5-6 cardiac output determinations can be carried out within 10 minutes. For calculation of the cardiac output from dye dilution curves, a measure for the area under the curve was obtained by cutting out and weighing the area between the curve and the base line, after transferring the curve to suitable homogeneous paper and separating the primary curve from the recirculation peak by logarithmic extrapolation of the descending limb of the primary curve. Chapter 6 presents a study of factors possibly exerting an unfavourable influence on the accuracy of the method. These factors include the binding of the dye to the erythrocytes, the influence of the Hb concentration and the decrease in light reflection caused by dye, adhesion of the dye to tubing walls and the influence of some blood constituents on patent blue V and indocyanine green. Although some of these factors were not without influence under experimental conditions, the accuracy of the method as now applied is not appreciably affected. The accuracy of the method was further studied in four different circulation models (Chapter 7). It was experimentally demonstrated that the amount of dye used in the calibration procedure is of definite influence on the method's accuracy. In view of the nonlinear relationship between dye concentration and decrease in the amount of light reflected, the amplitude of the calibration peak should be of about the Same size as the height of the dilution curve in a given patient. Table 19 shows that values calculated from dye dilution curves are in good agreement with the flow as measured with the aid of a graded glass and a stopwatch. Another impression of the method's accuracy was gained by studying the reproducibility of cardiac output values determined in patients (Chapter 8). The material was divided into three groups. Group a (Table 21, Fig. 14a) includes measurements made after conventional dye injection not followed by flushing. The standard deviation as calculated from 82 duplicate determinations, was 1.1 l/min, which corresponds with a coefficient of variation of 12.7%. Group b (Table 22, Fig. 14b) concerns determinations made after flushing dye into the patient with the special injection device. The 71 duplicate determinations gave a standard deviation of 0.58 l/min, which corresponds with a coefficient of variation of 6.6%. After yet another technica1 improvement in the calibration system (group c, Table 22, Fig. 14c), the standard deviation as calculated from 81 duplicate determinations, diminished to 0.40 l/min, corresponding with a coefficient of variation of 5.1%. The spread of the cardiac output values determined did not differ significantly from that of the matching stroke volume values (Table 24). Selecting patients without clinically manifest pulmonary or cardiovascular disease or other diseases strongly affecting the cardiovascular system, we found a normal resting cardiac output of 7.0 l/min, with a standard deviation of 0.76 l/min - the standard deviation of the mean (_S__) being 0.18 llmin. This resting cardiac Vn output value agrees well.with the first peak of 6.5 l/min in the frequency distribution of the cardiac output of al1 patients examined (Fig. 15). In one-third of patients with chronic liver disease, the resting cardiac output exceeded 9.0 l/min. A notable feature was that resting cardiac output values in patients with infectiouc hepatitis were found in the lower part of the normal spread (cf. Table 25 and Fig. 15). In view of this finding, and because parenchymal liver disease does not produce hypertension either in the literature or in this study, and also because peripheral vasoconstriction is demonstrable in acute infectious hepatitis, the hypothesis is forwarded that the liver injury exerts a direct influence, positively and negatively, on peripheral vascular resistance, and that the blood pressure is kept nearly constant by variations in the cardiac output resulting from vasopressoreceptor action. Zie: Summary