Basic physiochemical and rheological properties of detergent sclerosants

Introduction

Sodium tetradecyl sulphate (STS) and polidocanol (POL) are sclerosing agents used to permeabilise endothelial membranes with the ultimate aim of removing the intimal lining of target vessels and inducing endovascular occlusion. Prior to this study, little was published on the basic physiochemical or rheological parameters that influence the clinical activity of these agents. In this study, we investigated properties such as surface tension, density and viscosity. The surface tension of a fluid is the cohesive force at the surface of a substance that creates a surface film. The effect on surface tension determines the minimum concentration at which detergents possess maximal activity as sclerosants. The density of a substance is the mass of the substance per unit of volume. The viscosity of a fluid is the resistance of a fluid to shear. The density and viscosity influence the flow and distribution of sclerosants in the target vessels and the degree of mixing with the intra-vascular blood. These fundamental physiochemical properties may influence the clinical activity and in particular the potential for recanalization, treatment failure and complications secondary to the embolic ascent of foam-generated bubbles to the cranial circulation.

Both STS and POL are surface-active agents (surfactants). Surfactants are amphiphilic compounds that possess a polar hydrophilic head and a non-polar hydrophobic tail and therefore have a remarkable chemical resemblance to membrane phospholipids. Surfactants reduce the surface tension and ultimately solubilise cellular membranes. These agents are routinely used in membrane biochemistry to solubilise lipids. Surfactants are subdivided into four groups based on the chemical structure of the hydrophilic head, namely anionic, cationic, non-ionic and Zwitterionic (both negatively and positively charged heads). STS is an anionic and POL is a non-ionic surfactant.

Surfactants have limited solubility in polar solvents such as water. At low concentrations, they disperse in the solvent as monomers but as the concentration of surfactant increases, individual monomers aggregate to form micelles. Beyond a particular concentration specific for each surfactant, the critical micellar concentration (CMC), all additional surfactant added to the solution will aggregate in micelles. Due to their Amphiphilic nature, surfactants concentrate at gas/liquid or liquid/liquid interfaces and have the effect of reducing the surface (interfacial) tension. Increasing the concentration of surfactant has the effect of reducing the surface tension further until the CMC of the surfactant is reached. Beyond the CMC, any additional surfactant in the solution has no further effect in reducing the surface tension.

The interfacial activity of surfactants gives rise to a wide range of surface chemistry functions such as emulsifying, solubilising and foaming. Aqueous foams are dispersions of gas in liquid stabilized by surfactant adsorbed at the gas-liquid interface. These fluids are considered viscoelastic, possessing both the viscousness of liquids and the elastic properties of plastics. ¹

The flow dynamics of foams depend on a number of key factors and in particular their viscosity. Viscosity is an inherent property that determines the fluid’s resistance to flow and is affected by factors such as density (mass per unit volume) and shear rate (the velocity gradient across the diameter of the vessel). ² When viscosity remains constant irrespective of the shear rate, the fluid is termed a ‘Newtonian’ fluid and when it changes with the shear rate, the fluid is termed ‘Non-Newtonian’. Non-Newtonian fluids may be ‘shear thinning’, where the viscosity decreases at higher shear rates or ‘shear thickening’, where viscosity increases at higher shear rates. Water is a Newtonian fluid while blood and aqueous foams are examples of Non-Newtonian shear thinning fluids.^3,4

In this study, we aimed to determine some of the most fundamental physiochemical and rheological properties of detergent sclerosants such as their surface tension, CMC, density and viscosity. We compared these properties with that of the embolic agent ethanol and also investigated the influence of sclerosant type, format, concentration and foam composition.

Methods

Results

Density

Liquid agents

The density of NS, blood and ethanol were determined to be 1007 kg/m³, 1070 kg/m³ and 677 kg/m³, respectively (Figure 1). The density values for liquid sclerosants were higher than ethanol but lower than NS and blood (Figure 1). Density decreased with increasing concentrations of sclerosants declining to approximately 987 kg/m³ (p < 0.05) at 3% concentration for both agents (Figure 2). OPEN IN VIEWER

Figure 1.

The effect of concentration and liquid-plus-air fraction (LAF) on foam sclerosant density. Density was measured for (a) STS and (b) POL foam for a range of LAFs (1 + 2 to 1 + 8) and concentrations. Sclerosant foam density was compared with that of blood (), saline () and ethanol (). Results are presented as mean values ± SEM (n > 3).

OPEN IN VIEWER

Figure 2.

The effect of concentration on liquid sclerosant density. Density was measured for liquid STS () and POL (). Results are presented as mean values ± SEM (n = 3).

Foam sclerosants

Foam sclerosants were 3- to 5-folds lighter than liquid sclerosants (Figure 2). Wet foams (1 + 2 and 1 + 3) were denser than dry foams. Foam density was not influenced by sclerosant concentration.

Surface tension and CMC

There was a progressive fall in surface tension with increasing concentrations of sclerosants until the CMC was reached (Figure 3). The CMC value for STS was graphically determined to be 0.075% when diluted in NS and 0.200% in water. The CMC value for POL diluted in NS or water was approximately 0.002%. OPEN IN VIEWER

Figure 3.

The effect of sclerosant concentration and diluent on surface tension. Sclerosant liquid surface tension was measured using a tensiometer for sodium tetradecyl sulphate (STS) dilutions in normal saline (NS, ) or water () and for polidocanol (POL) dilutions in NS () or water (). The critical micellar concentration (CMC) for STS was 0.075% in NS and 0.200% in water. POL CMC in NS or water was approximately 0.002% (n ≥ 3).

Viscosity

Liquid agents

STS viscosity was comparable with POL while ethanol was more viscous than liquid detergents (Figure 4). Both sclerosants were more viscous when diluted in water than in NS (Figure 5). OPEN IN VIEWER

Figure 4.

Sclerosant liquid and foam viscosity in different diluents. Liquid and foam sclerosant viscosities were obtained for (a) STS (red) and (b) POL (blue) using a Rheometer for a range of concentrations and liquid-plus-air fractions (LAF). Viscosity values for foam sclerosants are compared with those of blood (), saline () and ethanol (). Results are presented as mean values ± SEM (n > 3).

OPEN IN VIEWER

Figure 5.

Non-Newtonian flow mechanics of sclerosing agents and ethanol. The viscosity as a function of shear rate was determined for (a) embolic agent ethanol and compared with liquid sclerosants (b) sodium tetradecyl sulphate (STS) and (c) polidocanol (POL) diluted in water (n = 3, ▪ and •) or normal saline (n = 3, □ and ○) and 3% foam sclerosants (LAF 1 + 2 to 1 + 8) (d) STS and (e) POL. As the foam viscosity measurements were 10,000-fold higher than liquid, foam viscosity is shown in logarithmic scale and liquid viscosity is shown in linear scale. Results are presented as mean values ± SEM (n ≥ 3).

Foam sclerosants

Foam sclerosants were significantly more viscous than liquid agents by at least four orders of magnitude (10,000-fold) (Figure 4). STS foam viscosity was not influenced by changes in concentration or LAFs. By contrast, POL foam viscosity increased at higher sclerosant concentrations and LAFs.

Variations with shear rate

All liquid agents demonstrated a fall in viscosity at lower shear rates (<10 s^–1) (Figure 5). Similarly, foam viscosity decreased with an increase in shear rate. Due to breakage and dissipation of bubbles, foam viscosity could not be determined at shear rates > 1 s^–1.

Discussion

This study aimed to determine the basic physiochemical and rheological properties of detergent sclerosants and to investigate the role of sclerosant type, concentration, format (liquid vs. foam) and LAF. Liquid sclerosant density was influenced by the sclerosant concentration while the foam density was influenced by the air fraction. The CMC of POL was determined to be 0.002% and not influenced by the solvent while the CMC of liquid STS diluted in NS was found to be lower (0.075%) than that diluted in water (0.200%). Foam sclerosants were found to be approximately 10,000-fold more viscous than liquid agents at very low shear rates but decreased at higher shear rates. While STS foam viscosity was not affected by concentration or LAF, POL foam viscosity increased at higher sclerosant concentrations and LAF.

The density values for STS and POL were similar and lower than that of water. The hydrophobic non-polar tails of both detergent molecules contain long hydrocarbon chains. The relatively smaller water molecules can pack together more tightly and hence water has a higher density compared with detergents. As expected, the density of detergents decreased at higher concentrations due to the increased number of detergent molecules present. This fall in density was more rapid at concentrations below the CMC, demonstrating a steep slope. At concentrations above the CMC, the decline in density was more gradual. This is due to the tighter packing of detergent micelles compared with monomers at such concentrations.^2,10

The density of foam sclerosants was significantly less than that of liquid agents. The major determinant in foam density was the LAF, while the concentration of sclerosant had little influence. In a previous study, ¹¹ the density of sclerosant foam was determined to be between 160–200 kg/m³ for STS and 180–240 kg/m³ for POL. This is in line with our measurements of 197 ± 2.9 kg/m³ for STS and 200 ± 6.4 kg/m³ for POL foams (3%, 1 + 4). By comparison, blood density was determined to be 1070 kg/m³ and liquid sclerosants to be 987 kg/m³ for both agents at 3%. Hence, foam sclerosants were much lighter per unit volume compared with liquid sclerosants, which in turn were slightly lighter than blood. In a recent study, ⁶ we reported on the formation of three relatively distinct horizontal layers of foam, liquid sclerosant and blood following the injection of a wet foam preparation in an experimental model of a superficial vein. Given its lighter density, foam ascends to the top of the vessel and advances over blood. In time, liquid sclerosant drains down due to gravity and remains sandwiched between foam and blood at the foam front. ⁶ Clinically, the presence of a distinct blood layer at the bottom wall of the target vessel would predispose to non-closure and recanalization. Hence, care should be taken to ensure exposure of the bottom wall of the target vessels to adequate volumes of the injected sclerosant.

Despite the clinical use of detergent sclerosants for many decades, the CMC values for these agents were previously not published. In this study, we determined the CMC value for the stock solution of STS to be 100-fold higher than that of POL (0.2% vs. 0.002%). This is due to a number of reasons. Firstly, ionic surfactants have higher CMC values compared to non-ionic surfactants.^8,12,13 Furthermore, the hydrophobic tail length influences the CMC, with longer tail lengths resulting in lower CMC values. ¹⁴ POL is a non-ionic detergent and has a much longer tail compared to STS and hence a much lower CMC. Given the higher CMC values for STS, this sclerosant may be considered less efficient than POL in reducing surface tension and achieving maximal surfactant activity. However, given the smaller molecular weight of STS, there are more surfactant molecules present per unit volume of STS compared with POL for the same sclerosant concentration. Therefore, target endothelial membranes are exposed to more STS molecules per unit volume compared with POL. This partly explains the increased potency of STS as a membrane solubilising agent compared with POL.

CMC of surfactants is significantly affected by the addition of buffer additives. ⁶ High salt concentrations strongly decrease the CMC of ionic (anionic or cationic) surfactants and the CMC of anionic surfactants, in particular, is significantly decreased when these agents are diluted in saline compared with water.^8,15 Here, we investigated whether dilutions in NS (sodium chloride 0.9% w/v) had a different CMC to those supplied by the manufacturers prepared in water. While POL had the same CMC value in water and saline, STS showed a much lower CMC in saline (0.075%) compared with water (0.2%). It may not be surprising that the lowest concentration of STS manufactured by STD Pharm (preparation in water) is 0.2%, ¹⁶ which is the CMC of this agent in water. Clinically, most practitioners dilute the stock solutions in NS. Based on our present findings, the choice of diluent would make no difference for POL. However, when STS is diluted in NS, concentrations as low as 0.075% would still maintain maximal surfactant activity.

We investigated a number of physiochemical and rheological properties of sclerosant foams. Aqueous foams are mixtures of a gas phase, a liquid phase and a surfactant. In general, liquids are significantly less viscous than aqueous foams and the higher the gas content of the foam (dry foams), the higher the viscosity. In very wet foams, spherical bubbles float in the liquid and move easily. By contrast, dry foams have a tightly packed structure of polyhedral bubbles. With the increasing gas fraction, the foam cells become more polyhedral in structure and more difficult to move. ⁹ Hence, as a general rule, wet foams are less viscous while dry foams have a higher viscosity.^17,18 In this study, POL followed this general pattern and showed a higher viscosity at higher LAFs.

The viscoelastic properties of foams are also influenced by the concentration and the ionic nature of the surfactant. For non-ionic surfactants, an increase in surfactant concentration results in a decrease in surface tension, an increase in foam stability, a tightly packed structure and a higher viscosity. ¹⁹ Once again, POL followed this predictable pattern and demonstrated higher viscosity values at higher sclerosant concentrations. The positive correlation between POL concentration and foam viscosity, foam half-time ⁷ and bubble size ⁶ is shown in Figure 6. OPEN IN VIEWER

In contrast to POL, STS foam viscosity did not increase at higher LAFs or concentrations of the surfactant. This is most likely due to the anionic nature of STS. The adsorption of anionic surfactants on the liquid/gas interface produces a diffuse electrostatic double layer which generates repulsive forces within the liquid lamella separating the gas bubbles. With an increase in surfactant concentration, the charged surfaces repel any additional surfactant molecules. This repulsive force slows down the adsorption process and as a result the liquid lamellae in ionic surfactants drain at significantly slower rates compared with non-ionic surfactants.^15,20,21 Hence, despite the rise in surfactant concentration, the relative wet structure and a stable viscosity is maintained. As shown in Figure 6(a), foam viscosity, half-time or bubble size is not significantly influenced by STS concentration. Microscopic studies are underway to further investigate the effect of the ionic nature of the surfactant on the three-dimensional architecture of the sclerosant foams.

While viscosity is not a function of shear rate for Newtonian fluids, it decreases at high shear rates for Non-Newtonian shear thinning fluids. Here, we demonstrated that STS and POL in both liquid and foam formats as well as the embolic agent ethanol are shear thinning Non-Newtonian fluids. The measured viscosity values for STS and POL liquid and ethanol reached that of water (1 m Pa.s) at higher shear rates (>10 s^–1). This implies that at high shear rates, such as in a high-flow blood vessel, the viscosity of the liquid agent will be close to that of blood resulting in propagation away from the intended site of action. At low shear rates, the foam sclerosant viscosity was at least 10,000-fold higher than that of the liquid agents but still decreased at high shear rates. This is most likely due to the lack of foam stability and disruption of the foam structure at high shear rates. This implies that at low shear rates, such as in superficial veins, foams are much more preferable to liquid sclerosants as they are more likely to stay at the intended site of action and less likely to mix with the intra-vascular blood. By contrast, liquid sclerosants have a comparable viscosity to that of blood and are more likely to mix. In high shear rates, such as in an arterio-venous malformation (AVM), the viscosity of the sclerosant or embolic agent will be close to that of blood, resulting in propagation away from the intended site of action. It would be clinically prudent to reduce the velocity and volume flow in such vessels to benefit from the clinical effect of these agents at the intended site of action. This can be achieved by using methods such as infiltration of tumescent fluid in the perivascular space before the administration of the sclerosing or embolic agent.

This study had a number of limitations. Foams were generated in this study using room air as the foaming gas, filters and a modified Tessari technique. Other techniques and gases may yield different results and should be investigated in comparative studies. The effect of temperature on foam viscoelastic properties has been investigated by the authors in a recent study. ⁷ Future projects will explore the effect of the anionic nature of sclerosants on the structure of the foam bubbles,

In summary, sclerosant foams demonstrated lower densities but significantly higher viscosities compared with liquid sclerosants. The CMC of STS was significantly influenced by the choice of diluent while POL was not affected. POL (but not STS) foam viscosity progressively increased at higher LAF and sclerosant concentrations.