INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

THE ASSESSMENT IN HUMANS of the compliance of the venous system in situ is a rather difficult task. Indeed, strain-gauge plethysmography, the standard method adopted to determine vascular compliance does not allow the distinction to be made between venous and arterial blood volume shifts for a given pressure change. Measurements of vascular compliance reflect, in fact, the contribution of both compartments, even though in classic strain-gauge plethysmography the data are attributable to the dilation of only the venous tree.

A method allowing determination of leg vein compliance in healthy subjects in normal and in special environmental conditions such as bed rest and microgravity, as well as in the case of vascular disease, would be undoubtedly of physiological and clinical interest (3, 9, 10). Particularly useful would be a method, that, at variance with strain-gauge plethysmography, could allow measurement of the compliance of the superficial venous tree (9), being therefore potentially applicable to the detection of specific pathological aspects of microcirculation.

The purpose of the present study is to develop a novel method of measuring, noninvasively in humans, the compliance of the microvasculature of the calf, particularly of its superficial venous compartment, by near-infrared spectroscopy (NIRS). This method is complementary to plethysmography.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Subjects. The experiments were conducted in 10 (2 women) healthy subjects aged 39.7 ± 8.2 (SD) yr (body wt 63.6 ± 13.0 kg, height 172.5 ± 7.8 cm). They were aware of the purpose of the study and of the related risks. The protocol was approved by the ethical committee of the Medical School of the University of Geneva.

NIRS measurements. Tissue deoxyhemoglobin (Hb) and oxyhemoglobin (HbO₂) concentration changes ([Hb] and [HbO₂], respectively, in µM) were obtained by a NIRO-500 continuous-wave photometer (Hamamatsu Photonics, Hamamatsu City, Japan). This device operates at four wavelengths: 773, 828, 853, and 914 nm. Optical densities for the four wavelengths were acquired with a sampling time of 2 s. Changes in [HbO₂] and [Hb] were then calculated from the experimental optical density values by means of dedicated NIRO-500 software (1). The differential pathlength factor value adopted for the calf muscles was five (5), and it represents a "mean" value of a group of healthy subjects. Total tissue hemoglobin concentration changes were expressed as [Hb]_tot = [HbO₂] + [Hb]. It is usually assumed that near-infrared light penetrates ~1.5 cm when the source-detector spacing is 3-4 cm (8) and detects mainly small vessels (7), i.e., arterioles, capillaries, and venules.

1

1

(1)

Protocol. The subject was placed on a tilt table, secured in a supine position, and invited to relax during the whole protocol. [HbO₂] and [Hb] were continously monitored during head-up tilt from  = 10 to 0, 15, 30, 45, 60, and 75° and then back to 10, 75, and 10° for 5 (baseline), 2, 2, 2, 2, 2, 2, 10, 15, and 5 min, respectively (see Fig. 1). The transmitting and receiving optodes of the photometer were placed on the medial line of the upper one-third of the right gastrocnemius medialis. The optodes were positioned 3 cm apart. NIRS data were acquired during the whole protocol.

View larger version (17K): [in this window] [in a new window]   Fig. 1.   A: typical time course of change in total tissue hemoglobin concentration ([Hb]tot) during tilt protocol. B: position of tilt table (0° corresponds to horizontal position).

Pressure measurements. The intravascular pressure (P; mmHg) at the site of the NIRS measurements was taken as equivalent to the pressure of the blood column length (l) extending from the intersection of the midaxillary line with the horizontal plane through the upper border of the fourth chondromanubrial junction to the center of the optodes multiplied by sine () (6), i.e.

(2)

where  = 1,060 kg/m³ is the mass density of blood (4), and g is the acceleration of gravity.

Vascular compliance. Vascular compliance (expressed as constant K; ml · l¹ · mmHg¹) was calculated by dividing the blood volume changes by the corresponding P variation, i.e.

(3)

where K_Hbtot is total vascular compliance.

View larger version (19K): [in this window] [in a new window]   Fig. 2.   Change in total blood ([Hb]tot b; A) and blood deoxyhemoglobin ([Hb]b) and oxyhemoglobin ([HbO2]b) concentration (B), expressed in ml · l1 · mmHg1, as a function of hydrostatic pressure (P). Each point corresponds to mean of 10 subjects measured at steady state at a given P.

Statistics. At this stage of the study, the subject sample (n = 10) is considered to be representative of a healthy adult population. Some measurements were occasionally repeated in some of the subjects. Trial-to-trial intraindividual reproducibility of NIRS data shows only a variation of approximately ±1 µM. The average changes for the group were considered for statistical analysis. Data are reported as means ± SD. Student's paired t-test was adopted for comparison between different experimental conditions when applicable. Significance level was set at P < 0.05.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

A typical time course of [Hb]_tot after changes in the tilt-table angle are shown in Fig. 1. No significant difference was found between [Hb]_tot values obtained at 75° when reached by incremental angle changes (i.e., in 15° steps, starting from 10°; Fig. 1B, left) or by a single step (from 10 to 75°; Fig. 1B, right). The three reference levels (at 10°) were not statistically different. These results imply that for each step (15°), 2 min were enough to reach steady levels in all subjects.

Average [HbO₂]_{tot b} values as a function of P are shown in Fig. 2A. [HbO₂]_{tot b} and P appear to be linearly correlated (P < 0.05, r = 0.99). The slope of the regression corresponds by definition (see Eq. 3) to K_Hbtot = 0.086 ± 0.005 ml · l¹ · mmHg¹. The choice of a linear model, even though satisfactorily fitting the experimental data, is not based on physiological grounds but is only adopted to estimate vascular compliance.

[HbO₂]_b and [Hb]_b curves as a function of P are shown in Fig. 2B. The first three points of the curves (P = 16, 0, and 24 mmHg, respectively) are not statistically different. As a consequence, the linearity of [HbO₂]_{tot b} = [Hb]_b + [HbO₂]_b implies that [HbO₂]_b and [Hb]_b must also vary linearly. Thus K_Hb = K_HbO2 = K_Hbtot/2 = 0.043 ml · l¹ · mmHg¹ for P between 16 and 24 mmHg. From Fig. 2 it may be seen that the slope of the [HbO₂]_b curve decreases abruptly at P = 24 mmHg. The behavior of [Hb]_b for P  24 mmHg may also be considered as linear (P < 0.05, r = 0.99) with the slope K_Hb = 0.079 ± 0.008 ml · l¹ · mmHg¹. It follows that K_HbO2 = K_Hbtot  K_Hb = 0.007 ml · l¹ · mmHg¹.

Figure 3 describes the time course of the mean values of [HbO₂]_b (A) and [Hb]_b (B) in all subjects when the table was tilted from 75 to 10° (after incremental changes in the angle). Dotted lines correspond to the SD.

View larger version (14K): [in this window] [in a new window]   Fig. 3.   Change in tissue oxyhemoglobin ([HbO2]; A) and deoxyhemoglobin concentration ([Hb]; B) time course during 1-step tilt-table-angle variation from 75 to 10°. Dotted lines, SD (n = 10).

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

NIRS allows the identification of oxygenated and deoxygenated blood components in the region of interest. The Hb content in the arterial section of the vascular bed is usually <3% of Hb_tot and may, therefore, be considered negligible for the purpose of the present discussion. Thus [Hb]_b should represent blood volume changes occurring exclusively in the venous compartment (VC). On the other hand, [HbO₂]_b could account for blood volume changes occurring in the arterial compartment (AC) as well as in the capillaries and in the VC. The present results indicate that the fractional contribution of [HbO₂]_b and [Hb]_b to K_Hbtot (see Eq. 3) depends on the initial P value. In fact, for blood pressures P  24 mmHg, K_Hb appears to be equal to K_HbO2, whereas for P  24 mmHg, K_Hb values are much greater (up to ~12 times) than K_HbO2.

Considering that the compliance of the arteriolar and of the capillary compartments of the vascular tree is very low, it may be assumed that the latter are not affected when the subject is tilted at P  24 mmHg. This is tantamount to assuming that, on tilt, the change in blood volume occurring in the muscle region of interest takes place only within the VC. Should this be the case, the identity of K_Hb and K_HbO2 in the range 0-24 mmHg implies that venous blood saturation must be

(4)

Indeed, human femoral venous blood saturation at rest ranges from 47 to 52% (11, 12). Thus the above assumption appears to be supported by experimental findings, and the calculated compliances should be representative of the VC only. The same assumption is usually made when vascular compliance is measured by classic strain-gauge plethysmography (9), even though the vascular district involved in the latter is different as it includes mainly deep veins (2).

For P values  24 mmHg, the data shown in Fig. 2A indicate that [Hb]_tot increases linearly as for P  24 mmHg. However, the two components of [Hb]_tot show a divergence (see Fig. 2B). On the basis of observations made in the previous paragraphs, the latter change may be explained only by a reduction of blood flow in the muscle, possibly mediated by arteriolar vasoconstriction. In fact, assuming a constant O₂ consumption by the muscles throughout the tilt maneuver, a decrease in blood flow would lead to a progressive increase of [Hb]_b in VC (higher O₂ extraction), therefore to a tendency of the [HbO₂]_b curve to level off (Fig. 2B). The above-described adaptations are also supported by the dynamic measurements shown in Fig. 3. In fact, when the table is tilted from 75 to 10°, the time course of [Hb]_b is an indicator of the shrinking of VC. [HbO₂]_b also returns to the reference level after a transitory dip. Such a pattern can be explained 1) initially, by a rapid washout of the oxygenated component from the vascular tree; and 2) later, by blood flow recovery after release of arteriolar tone and the consequent increase in blood O₂ saturation. Thus, for P values 24 mmHg, K_Hbtot depicts a "pure" venous compliance provided there are no AC volume changes. Unfortunately, the latter cannot be demonstrated experimentally by the present approach. In any case, the level at which the two curves in Fig. 2B diverge represents the threshold of the vascular (arteriolar) response.

Considering that near-infrared light penetrates ~1.5 cm and detects mainly small vessels, i.e., arterioles, capillaries, and venules, (see MATERIALS AND METHODS), the assessment of K_HbO2, K_Hbtot, and K_Hb might provide useful information about microcirculation in various experimental and pathological conditions. The estimate of K_HbO2, K_Hbtot, and K_Hb could be slightly improved by 100% O₂ breathing that eliminates the ~3% arterial Hb. On the other hand, P, instead of being calculated, can be measured invasively (9). However, in either case, the conclusions of the present study are essentially unchanged.

The high scatter of the experimental data shown in Figs. 2 and 3 is mainly due to interindividual physiological variability and differences in subcutaneous adipose tissue (in the range 0.4-0.9 cm by skinfold caliper), as discussed elsewhere (1). In fact, as indicated in Statistics, trial-to-trial intraindividual reproducibility of NIRS data shows only a variation of approximately ±1 µM. Future developments of a method allowing a correction for the adipose tissue thickness, and the utilization of the individual differential pathlength factor (see the beginning of MATERIALS AND METHODS), should further improve the precision of the present approach for the assessment of K_Hb and K_HbO2 and reduce the scatter of the data.

In conclusion, NIRS allows the assessment of the superficial venous compliance of the lower limbs at P levels 24 mmHg and of the characteristics of the vasomotor response of the arteriolar system during a tilt maneuver. The above information cannot be obtained by strain-gauge plethysmography, which is mainly adopted for the assessment of deep-vein compliance. Therefore, the present method may only be considered complementary to classic strain-gauge plethysmography, and a direct comparison between data obtained with the two techniques would be essentially meaningless.