Age-related changes in microvascular permeability: a significant factor in the susceptibility of children to shock?

Transcription

1 Clinical Science (2000) 98, (Printed in Great Britain) 211 R A P I D C O M M U N I C A T I O N Age-related changes in microvascular permeability: a significant factor in the susceptibility of children to shock? J. GAMBLE*, D. BETHELL, N. P. J. DAY, P. P. LOC, N. H. PHU, I. B. GARTSIDE, J. F. FARRAR and N. J. WHITE *School of Sport and Exercise Sciences, Birmingham University, Birmingham, U.K., Nuffield Department of Clinical Medicine, Oxford, U.K., Centre for Tropical Diseases, Ho Chi Minh City, Vietnam, Department of Physiology, Charing Cross and Westminister Medical School, London, U.K., and Wellcome Trust Clinical Research Unit, Ho Chi Minh City, Vietnam A B S T R A C T During studies of the pathogenesis of dengue shock syndrome, a condition largely confined to childhood and characterized by a systemic increase in vascular permeability, we observed that healthy controls, age-matched to children with dengue shock syndrome, gave high values of filtration capacity (K f ), a factor describing vascular permeability. We hypothesized that K f might be age dependent. Calf K f was studied in 89 healthy Vietnamese subjects aged 5 to 77 years. The K f was highest in the youngest children [7.53 ( ) K f U; median (range); where the units of K f, K f U ml min ml 1 mmhg 1 ]. Values were 3- to 4-fold lower towards the end of the second decade [4.69 ( ) K f U]. Young mammals are known to have a larger microvascular surface area per unit volume of skeletal muscle than adults. During development the proportion of developing vessels is greater. Moreover, the novel microvessels are known to be more permeable to water and plasma proteins than when mature. These factors may explain why children more readily develop hypovolaemic shock than adults in dengue haemorrhagic fever and other conditions characterized by increased microvascular permeability. INTRODUCTION Microvascular permeability is altered in severe systemic infections and may lead to microcirculatory dysfunction and shock. Increased vascular permeability causes hypovolaemia and shock in dengue haemorrhagic fever, a condition largely confined to children [1]. In the course of studying the pathogenesis of this condition, it became apparent that there were very few data on capillary permeability in normal healthy children for comparison. Venous congestion plethysmography (VCP) provides a non-invasive assessment of microvascular permeability parameters in human calves [2]. The small cumulative congestion cuff pressure (P cuff ) step protocol, used for VCP, assesses changes at the whole microvascular interface, rather than some functional fraction of it [3]. The increases in P cuff during VCP protocol affect neither limb arterial inflow [4] nor interstitial fluid pressure [5]. Preliminary data from studies on healthy Vietnamese children (aged 5 to12 years) indicated that they had much higher values for filtration capacity (K f ), an index of microvascular hydraulic permeability, than those of Key words: vascular permeability, filtration capacity, children, adults, age, plethysmography. Abbreviations: K, filtration capacity; K U, K units; VCP, venous congestion plethysmography; P, congestion cuff pressure; f f f cuff V, venous filling; J, fluid filtration; Pv, isovolumetric venous congestion cuff pressure. a v i Correspondence: Dr John Gamble, 26 Sweetcroft Lane, Hillingdon, Middlesex UB10 9LD ( john.gamble virgin.net).

2 212 J. Gamble and others adults. This suggested that microvascular permeability might change with age. We assessed the relationship between age and K f in a population of healthy Vietnamese subjects, aged 5 to 75 years. METHODS AND PROCEDURES Methods Vascular and microvascular parameters were derived from changes in calf circumference. These were measured using the mercury-in-silastic strain gauge system ( MSG ) and P cuff protocol described in detail previously [2]. Procedures Healthy Vietnamese adults and children were recruited provided they or their attendant relatives gave fully informed consent. Each subject was questioned and examined to confirm the absence of metabolic and vascular disease, including varicose veins. They had been asked to refrain from smoking, or drinking caffeine or alcohol containing beverages for a minimum of 4 h before the study. The subjects rested supine for 15 min, with the mid point of the calf to be studied supported at the level of the right atrium, which was assumed to be one-third of the distance down from the sternal angle to the posterior surface when supine. Once supine, arterial blood pressure was measured using the Riva Rochi method, and all subjects were found to be normotensive. A mercury-in-silastic strain gauge was fixed around the calf at a marked site of measured circumference and stretched to a standard tension, so that both increases and decreases in circumference were encompassed within the linear part of the strain-gauge range. A congestion cuff, placed around the ipsilateral thigh, was inflated by an automated pump and the pressure (P cuff ) sensed by a transducer connected to the cuff. The P cuff never exceeded the subjects mean diastolic pressure. The strain gauge was calibrated at the start of each study [2]. Both skin surface temperature, measured at a site close to the strain gauge, and room temperature, which was kept constant within the range of C by a temperature control unit, were recorded during the course of the study. The strain-gauge and cuff-pressure signals were amplified and sampled by an analogue-to-digital converter (PC26AT, Amplicon Liveline Ltd, U.K.) and saved for subsequent off-line analysis. Protocol The calf microvascular parameters and values of blood flow were obtained by assessing the change in tissue volume, derived from alterations in calf circumference, Figure 1 Calf volume response (top panel) of an 11 year old male child to a step increase in P cuff (lower panel) from 45 to 54 mmhg The upper trace of the top panel reflects the whole volume response and the superimposed dotted line, the regression slope (J v ) fitted to the last 150 sec of the response. The lower trace of the top panel shows the volume response after the regression slope (J v ) has been subtracted. The volume V a reflects the compliance volume change in response to this pressure step. that were obtained in response to increases in P cuff (Figure 1). These were obtained, in each study, by applying a series of five to seven small (8 10 mmhg) cumulative pressure steps. Once the ambient venous pressure in the limb is exceeded, each additional pressure increment causes a change in limb volume that is attributable to venous filling (V a ), and when the congestion cuff pressure exceeds the value of the isovolumetric venous congestion cuff pressure (Pv i ), the steady-state change in volume reflects fluid filtration (J v ; Figure 1). Blood flow was assessed from the initial slope of the volume response to a single 10-sec duration, 60-mmHg-step pressure change [4]. Computer-based analysis enables differentiation between the volume and filtration responses [2]. The microvascular K f is determined by linear regression, based on P cuff and J v co-ordinates obtained at pressures above those causing a measurable increase in J v (Figure 2, upper panel). The slope of this relationship is the microvascular K f (units, K f U ml min 100 ml mmhg ) [2]. The intercept of the line on the abscissa, that is where J v 0, represents Pv i (Figure 2, upper panel). This is the P cuff that has to be exceeded to induce net J v at the level of the strain gauge. The value Pv i reflects the effective colloid osmotic pressure, σπ c (see eqn. 1), at the microvascular interface [6]. The Pv i may be less than the isovolumetric pressure at the microvascular interface, the discrepancy depending

3 Age and vascular permeability 213 Statistical Analysis Student s t test was used to compare normally distributed paired or unpaired data. The Wilcoxon Signed Rank Test was used to test two sets of observations made on the same subjects, but on different occasions. The Mann Whitney Rank Sum Test was used for comparison of non-parametric data. Multiple comparisons of nonnormally distributed data were made using the Kruskal Wallis test. For normally distributed data, ANOVA was used in conjunction with the Student Newman Keuls All Pair-wise Multiple Comparison Procedure. All these tests, including the linear and non-linear regression analyses, were performed using SigmaStat (Jandel Scientific, Erkrath, GbmH). Curve fitting was carried out using Jandel Scientific s non-linear regression routine on the data pairs. The Jandel routine uses the Marquardt Levenberg algorithm to calculate parameters that minimize the sum of the squares of differences between the model-derived and real values. Ethical approval for these studies was obtained from the Ethics committee of the Centre for Tropical Diseases, Ho Chi Minh City, Vietnam. RESULTS The studies were performed on 89 healthy Vietnamese volunteers, comprising 47 adults aged 37 (19 77) years [median (range)], of which 29 were female, and 42 Figure 2 The relationship between J v and P cuff (top panel) and the vascular compliance (lower panel) of the subject in Figure 1 (Top panel) It can be seen that the relationship between v J and P cuff was linear (K f ) at pressures greater than approx mmhg, which represents the Pv i in this subject. The value of K f in this subject was K f U. (Lower panel) There is no volume change (V a ) until the P cuff exceeds the ambient venous pressure in the calf. Above this pressure the relationship between a V and pressure is curvilinear. A curve fit of this relationship to the abscissa, when a V 0, represents the noninvasive assessment of calf venous pressure which, in this subject, was 5.0 mmhg. The closed circles represent the data points admitted to the regression analysis. upon the values of post-capillary resistance and blood flow. When the values of V a are plotted against the corresponding values of P cuff, they give a curvilinear relationship that is a function of the vascular and surrounding tissue compliance (Figure 2, lower panel). The intercept on the P cuff axis, where V a 0, is obtained by extrapolation of the relationship between the data points above the pressure at which the first discernible response is obtained. The intercept on the P cuff axis, at V a 0, gives a non-invasive estimate of the venous pressure at the level of the strain gauge [7]. Figure 3 K f data obtained from 89 healthy Vietnamese volunteers aged 5 to 77 years Each data point represents either the value from a single study (n 30) or the average of two studies performed 14 days apart (n 59)., male subjects;, female subjects. The solid line represents a third-order polynomial fit for these data and the dotted lines the 95% confidence limits for that fit. More information on the fit is given in the Methods and procedures section.

4 214 J. Gamble and others Table 1 Data from Vietnamese control subjects grouped on the basis of sex and age The age groups were 5 to 18 years ( 18 years) and greater than 18 years ( 18 years). The data are given as either means S.D. or medians (range), depending upon the normality of their distribution. Group n K f (K f U) P v i (mmhg) Blood flow (ml min ml 1 ) Venous pressure (mmhg) Males 18 years ( ) 7.3 ( ) Females 18 years ( ) 7.0 ( ) Males 18 years ( ) 7.3 ( ) Females 18 years ( ) 7.0 ( ) children aged 10 (5 17) years, of which 18 were female. The males and females in the two groups were matched for age. Intra-individual repeatability Studies on 49 of the subjects (20 males), ages ranging from 5 to 49 years, were repeated after a 14 day interval. The median values of K f (range) for the two groups were 6.29 ( ) and 6.54 ( ) K f U respectively (P 0.84, paired t test). These observations are compatible with previous assessments of the coefficient of variation on repeated assessments of K f, which gave values of 10.6% [3] and 7.7% [8]. The averaged value from each data pair was used in the present paper. K f Figure 3 shows the line of best fit for the third-order polynomial equation, K f ( ) ( age) ( age ) ( age ). There was a marked decline in K f in the first two decades of life. When the values were grouped on the basis of age, subjects less than 18 years of age had a median value of 7.53 ( ) K f U compared with the value of 4.69 ( ) (P 0.001) in subjects over 18 years. There were no differences in K f between males and females in the 18 year old group, whereas in adults the K f values in females were significantly higher than those of the sex-matched males (P 0.01; Table 1). Pv i and arterial blood flow exhibited similar age dependent trends. There were no age or sex dependant differences in either calf isovolumetric venous pressure, or venous pressure. The median calf blood flow in the younger ( 18 years) group was significantly higher than that in the adults [5.16 ( ) compared with 3.01 ( ) ml min 100 ml (P 0.01)]. There were no sex-dependent differences in calf blood flow in either group (Table 1). DISCUSSION The pioneering work of Landis in 1927 [9] highlighted the relationship between capillary pressure, surface area and the movement of fluid across the microvascular endothelium. Using the principle originally put forward by Starling [10], Landis described his experimental data with the equation: J v A Lp[P c P t ) σ(π c π t )] (1) where J v A is the rate of fluid movement per unit area of vessel wall (A), P and π represent the pressures attributable to the hydrostatic and colloid osmotic forces respectively. The subscripts c and t represent the capillary and interstitial compartments. Lp and σ are the microvascular permeability coefficients for water and plasma proteins respectively. Pappenheimer and Soto-Rivera [6] showed that the J v characteristics of animal hind limbs paralleled those described by Landis [9]. However, these observations were empirical, since the total surface area available for exchange in the hind limb tissue was unknown. Early studies of the effect of P cuff on J v in the calves of healthy adults established that the relationship, defined by the K f, was linear [2]. We suggested that the value K f approximated the product of Landis s single capillary model value of Lp and the surface area (A) of the microvasculature under examination. After invoking a peripheral vasoconstrictor response by the imposition of a passive head-up tilt, it was the rate of blood flow through the limb, and not the surface area available for filtration, that altered [3,4]. These data suggested that, in healthy adult males, differences in K f reflected differences in the total microvascular density, rather than the proportion of vessels being perfused. In a long-term study of healthy young adult males, an individual s K f value was shown to be relatively constant giving a coefficient of variation of % [3], although marked inter-individual differences between age- and sex-matched subjects were noted. Some of the variations could be attributed to the changes in K f that have been observed in young women during the menstrual cycle, irrespective of their oral contraceptive status [11]. Careful examination of human lower limb biopsy material has provided good estimates of capillary surface area [12], and the way in which capillary density changes as a result of sports training [13,14]. Thus, in the healthy adult male population, it seems probable that differences in K f reflect differences in microvascular density.

5 Age and vascular permeability 215 In the current study the greatest vascular permeability was found in the youngest subjects. These high K f values might reflect the relatively increased permeability of growing capillaries and changes in microvascular density associated with growth. Newly grown capillaries, in rabbit ear chambers, have a much higher blood flow and greater permeability to the albumin-bound dye, Evan s Blue, than do mature capillaries [15]. Similar observations were made using a cremaster muscle preparation [16]. Smaje et al. [17] have shown that, in these novel microvessels, both blood flow and albumin efflux could be reduced by the application of histamine H2 antagonists, whereas histamine H1 antagonists had little effect. Since, in mature vessels, histamine H1 and H2 receptors are normally associated with the control of permeability and blood flow respectively, these data suggest that either the receptor expression or function changes with maturity. Whilst there is a considerable amount of information on capillary density in the calf skeletal muscle of human adults [13], there is little information on that of children. Studies in rodent models suggest that the number of muscle fibres remains constant from birth to maturity and that increase in muscle girth can be accounted for by increases in the cross-sectional area of the individual muscle fibres. Moreover, as a result of these studies, it was concluded that increases in some of the volume occupied by muscular elements occurred at the expense of interstitial tissue [18]. Further studies, in rodents, showed that the capillary density per unit area of muscle decreased as a hyperbolic function of muscle fibre area during maturation [19]; studies in developing dogs showed that the function was linear [20]. It should be stressed that all of the studies cited relied on morphometric analysis of age-related changes in skeletal muscle elements alone. It is possible that the ratio of total volume encompassed within the calf circumference, relative to the subcutis, bone, skin and muscle components, changed during development, as did their relative capillary densities and, therefore, their possible contribution to the observed changes. However, skeletal muscle represents a large percentage of the total body mass. If capillary density and permeability patterns are the same in children as in some other mammalian species, this could account for the age-dependent susceptibility to shock, increased morbidity and mortality in both serious infection [1] and trauma [21]. This study was prompted by investigation into the pathophysiology of dengue haemorrhagic fever. Hypovolaemic shock is a characteristic feature of severe dengue haemorrhagic fever in children but not in adults. Severe dengue haemorrhagic fever is particularly likely in a second infection with a different virus serotype to that which caused the primary infection. Dengue shock syndrome results from a sudden generalized increase in microvascular permeability. The pathogenesis of this is not known. One reason why children, but generally not adults, develop shock may be that their baseline microvascular permeability is considerably greater than that of adults. They may therefore have less microvascular reserve with which to accommodate extraneous factors affecting vascular permeability. ACKNOWLEDGEMENTS We thank the hospital leaders and the staff at the Centre for Tropical Diseases for their assistance with this study. The study was funded by the Wellcome Trust of Great Britain. REFERENCES 1 Halstead, S. B. (1988) Pathogenesis of dengue: challenges to molecular biology. Science 239, Gamble, J., Gartside, I. B. and Christ, F. (1993) A reassessment of mercury in silastic strain gauge plethysmography for microvascular permeability assessment in man. J. Physiol. 464, Gamble, J., Christ, F. and Gartside, I. B. (1997) The effect of passive tilting on microvascular parameters in the human calf: a strain gauge plethysmography study. J. Physiol. 498, Gamble, J., Christ, F. and Gartside, I. B. (1998) Human calf precapillary resistance decreases in response to small cumulative increases in venous congestion pressure. J. Physiol. 507, Christ, F., Dellian, M., Goetz, A. E., Gamble, J. and Messmer, K. (1997) Changes in subcutaneous interstitial fluid pressure, tissue oxygenation, and skin red cell flux during venous congestion plethysmography in men. Microcirculation 4, Pappenheimer, J. R. and Soto-Rivera, A. (1948) Effective osmotic pressure of the plasma proteins and other quantities associated with the capillary circulation in the hind limbs of cats and dogs. Am. J. Physiol. 152, Christ, F., Gamble, J., Baschnegger, H. and Gartside, I. B. (1997) Relationship between venous pressure and tissue volume during venous congestion plethysmography in man. J. Physiol. 503, Jaap, A. J., Shore, A. C., Gartside, I. B., Gamble, J. and Tooke, J. E. (1993) Increased microvascular fluid permeability in young Type I (insulin dependent) diabetic patients. Diabetologia 36, Landis, E. M. (1927) Microinjection studies of capillary permeability. II. The relation between capillary pressure and the rate at which fluid passes through the walls of single capillaries. Am. J. Physiol. 82, Starling, E. H. (1896) On the adsorption of fluids from connective tissue spaces. J. Physiol. 19, Gamble, J. and Gartside, I. B. (1994) Six, weekly strain gauge plethysmography (MSG) measurements of hydraulic conductance in the calves of young women on oral contraceptives, and age matched men. Br. J. Clin. Pharmacol. 37, 481P 12 Clough, G. (1987) A quantitative study of the exchange microvasculature of muscles from the human foot and hand. Int. J. Microcirc. Clin. Exp. 6, Andersen, P. and Kroese, A. J. (1978) Capillary supply in soleus and gastrocnemius muscles of man. Pflugers Arch. 375, Kiens, B., Essen Gustavsson, B., Christensen, N. J. and Saltin, B. (1993) Skeletal muscle substrate utilization during submaximal exercise in man: effect of endurance training. J. Physiol. 469,

6 216 J. Gamble and others 15 Abell, R. B. (1946) The permeability of blood capillary sprouts and newly formed blood capillaries as compared to that of older blood capillaries. Am. J. Physiol. 147, Schoefl, G. I. (1963) Studies on inflammation. III. Growing capillaries: their structure and permeability. Virchows Arch. Pathol. Anat. 337, Smaje, L. H., Noor, N. and Clough, G. F. (1988) Changing sensitivity to HI and H2 receptor antagonists in the growing vasculature. Adv. Exp. Med. Biol. 242, Ontell, M. and Dunn, R. F. (1978) Neonatal growth: a quantitative study. Am. J. Anat. 152, Sillau, A. H. and Banchero, N. (1977) Effect of maturation on capillary density, fibre size and composition in rat skeletal muscle. Proc. Soc. Exp. Biol. Med. 154, Aquin, L. and Banchero, N. (1981) The cytoarchitecture and capillary supply in the skeletal muscle of growing dogs. J. Anat. 132, Neuspiel, D. R. and Kuller, L. H. (1985) Sudden and unexpected natural death in childhood and adolescence. J. Am. Med. Assoc. 254, Received 5 October 1999/19 November 1999; accepted 8 December 1999