Evaluation of hemolysis in microcatheter directed blood infusion at different flow rates for transarterial salvage reperfusion: In-vitro study

Abstract

Background:

Microcatheter directed blood reperfusion is an endovascular salvage option for acute cerebral artery occlusions. It has not been investigated whether this technique may be associated with hemolysis.

Objective:

Analysis of hemolysis during blood infusion through different microcatheters and infusion rates to assess related risks.

Methods:

Four microcatheters with different inner diameters were perfused with blood samples at three infusion rates. Hemolytic markers including lactate-dehydrogenase (LDH) and haptoglobin were analyzed. Samples before and after blood infusion were compared using Student’s t-test. Flow-related degree of hemolysis was analyzed with regression analysis. Resulting shear stress was calculated and correlated with LDH and haptoglobin.

Results:

Significant increase of LDH and decrease of haptoglobin was found after blood reperfusion through small microcatheters at progressive flow rates ( $p < 0.05$ ). No hemolysis was found with larger diameter microcatheters at all flow rates ( $p > 0.05$ ). Correlation between shear stress, LDH and haptoglobin was $r = 0.86$ and $r = 0.75$ , respectively.

Conclusions:

Progressive hemolysis occurs during blood perfusion of small lumen microcatheters at increasing flow rates. This phenomenon may be related to turbulent flow, exposure time and increased shear stress. Larger microcatheters did not induce hemolysis and may be the preferred choice for stroke reperfusion.

Keywords

Stroke ischemia hemolysis catheterization blood reperfusion

1. Introduction

Acute cerebral artery occlusions may be associated with immediate life threatening conditions [1]. Particularly reperfusion of acute stroke is time critical to prevent irreversible damage [2]. Various medical and endovascular options including vacuum assisted thrombectomy and stent retrievers are available to achieve arterial revascularization. Recent clinical trials have demonstrated excellent results for endovascular stroke management [3–9]. With endovascular techniques, initial recanalization across an acute cerebral artery obstruction may be achieved rapidly, whereas removal of the underlying obstruction takes 30–45 min [7,9,10]. In order to ‘buy time’ for preservation of viable brain tissue – particularly in challenging situations – where initial vacuum- or stent retriever assisted revascularization fails, reperfusion with oxygenated blood beyond arterial occlusions was suggested as an adjunct neuroprotective strategy. Once rapid tissue reperfusion has been achieved with temporary blood reperfusion, definitive endovascular reconstruction to resolve the underlying arterial occlusion may be commenced on a less time-critical basis [2,10,11].

Limited clinical case series have stated that microcatheter directed blood reperfusion of acute arterial occlusions is safe [2,10,11]. However, it has never been investigated whether this technique may be associated with mechanical blood trauma and hemolysis. Even low level hemolysis is associated with release of free hemoglobin and related antioxidant action [12]. Additional general blood cell damage such as platelet activation and white blood cell dysfunction may result. Scavenging of nitric oxide [13], damage to the glycocalyx and endothelial cells, and impairment of the vascular smooth muscle tone may result from hemolysis [14]. If hemolysis occurred in acute stroke management, further ischemia or hemorrhage might develop within the ischemic tissue. Consequently this study was performed to investigate whether hemolysis occurs after blood reperfusion with different microcatheters and infusion rates in an in-vitro model.

2. Methods

This in-vitro study was approved by the institutional review board (HREC–H0014127). Experiments were conducted on four different microcatheters at infusion rates of 1-, 2- and 3 ml/s.

For each of the 4 microcatheters 90 ml venous blood were drawn from a single test person to reduce variability. 90 IU Heparin, dissolved in 0.9 ml of 0.9% saline solution were added to the blood and mixed gently. 20 ml of blood were directly transferred into Vacutainer test tubes to obtain two baseline biochemical and hematological measurements.

The remaining 70 ml of heparinized blood was transferred into the 150 ml syringe of an angiographic power injector (Avidia, Imaxeon, Rydalmere, Sydney NSW 2116, Australia) at 37°C. The power injector was connected to the hub of the microcatheters without connection tubing to reduce dead space. All microcatheters had straight-tip configurations and were 150 cm long with one exception (Penumbra 5MAX, 132 cm length). Inner luminal diameters were different for all microcatheters (Table 1).

Table 1
Specification of microcatheter details investigated for blood reperfusion

Microcatheter Manufacturer Lumen diameter (mm) Length (cm) Pressure limit (kPa)

Excelsior SL-10 Stryker Neurovascular, Fremont, CA 94538, USA 0.42 150 N/A

Vasco+21 Balt Extrusion, 95160 Montmorency, France 0.6 150 N/A

Renegade HI-FLO Boston Scientific, Marlborough, MA 01752, USA 0.69 150 5520

Penumbra 5MAX Penumbra, Alameda, CA 94502, USA 1.52 132 N/A

Microcatheter	Manufacturer	Lumen diameter (mm)	Length (cm)	Pressure limit (kPa)
Excelsior SL-10	Stryker Neurovascular, Fremont, CA 94538, USA	0.42	150	N/A
Vasco+21	Balt Extrusion, 95160 Montmorency, France	0.6	150	N/A
Renegade HI-FLO	Boston Scientific, Marlborough, MA 01752, USA	0.69	150	5520
Penumbra 5MAX	Penumbra, Alameda, CA 94502, USA	1.52	132	N/A

Notes: Except for the Renegade HI-FLO microcatheter, injection pressure limits were not provided by the manufacturers.

Air was completely purged from the injector syringe and microcatheters were completely filled with blood from the syringe prior to infusion. 10 ml aliquots were injected through each microcatheter at flow rates of 1-, 2- and 3 ml/s. Experiments were repeated twice for each infusion rate and microcatheter according to sample size analysis. To prevent pressure related microcatheter disruption, injection pressure was limited to 5171 kPa on the power injector with an initial continuous rise of 0.3 s for all microcatheters and infusion rates. Injected samples were collected in 10 ml syringes and subsequently divided into Vacutainers for immediate biochemical analysis. Injection duration and volumes were recorded. The following biochemical parameters were analyzed:

Electrolytes: sodium (Na⁺) and potassium (K⁺) (mmol/l),

Biochemistry: lactate-dehydrogenase (U/l); haptoglobin (g/l),

Hematology: red blood count (RBC; 10⁹/l), hemoglobin (Hb; g/l); hematocrit (Hct; %), mean corpuscular volume (MCV; fl), mean corpuscular hemoglobin (MCH; pg), mean corpuscular hemoglobin concentration (MCHC; g/l), platelet count (Plt; 10⁹/l),

Coagulation: international normalized ratio (INR; %), activated partial thromboplastin time (APTT; s), fibrinogen (Fib; g/l).

Means and standard deviations were calculated for baseline and post-perfusion measurement parameters for all microcatheters and flow rates. A paired Student’s t-test was performed for significance between baseline-values and corresponding samples after blood infusion for each microcatheter and infusion rate. Significant difference was accepted when $p ⩽ 0.05$ . Linear regression analysis was performed between LDH and haptoglobin levels for all test samples. Manufacturer provided specifications of all microcatheters were used to calculate the radius of inner- and outer catheter diameters, inner surface area, catheter dead space, inner circumference and inner area diameters. Infusion rates and flow velocities (m/s) were calculated correspondingly.

Whole human blood was considered a non-Newtonian power-law fluid with a density of 1050 kg/m³ [14]. Consequently the viscosity of blood is a decreasing function of shear rate [15] and obeys power law characteristics: $\begin{matrix} τ = K {\dot{γ}}^{n} . \end{matrix}$ The apparent viscosity of blood was subsequently be expressed as: $\begin{matrix} μ = K {\dot{γ}}^{n - 1}, \end{matrix}$ where τ is shear stress (Pa), K is the power law coefficient (kg/m·s^2 − n), $\dot{γ}$ is shear rate (s⁻¹) and n (dimensionless) is the power law exponent. For blood as a shear thinning fluid, $n < 1$ [16].

K and n were determined by non-linear regression of experimental data published by Castellini et al. [15], as 3.23E−03 and 0.567, respectively.

A modified Reynold’s number as derived by Metzner and Reed for time-independent non-Newtonian fluids was used to determine flow regime [16]: $\begin{matrix} R_{MR} = \frac{D^{n} V^{2 - n} ρ}{K^{'} 8^{n - 1}}, where K^{'} satisfies \frac{D Δ P}{4 L} = K^{'} {(\frac{8 V}{D})}^{n} . \end{matrix}$ D is diameter (m), V is average velocity (m/s), ρ is fluid density (kg/m³) and L is tube length (m). The transition to turbulent flow begins at $R_{MR}$ in the range of 2000 to 2500 [16].

Shear rate at the catheter wall (laminar region) was calculated using the Rabinowitsch–Mooney relationship [16]: $\begin{matrix} {\dot{γ}}_{ω} = \frac{8 V}{D} (\frac{1 + 3 n^{'}}{4 n^{'}}), \end{matrix}$ where $n^{'} = n$ for power law fluids.

For turbulent flow of non-Newtonian fluids obeying power law characteristics (as is the case with blood), the Dodge and Metzner chart was used to determine the friction factor in the equation provided by Wilson and Thomas [16]: $\begin{matrix} \frac{1}{\sqrt{f}} = \frac{1}{\sqrt{f_{N}}} + 8.2 \frac{1 - n}{1 + n} + 1.77 ln (\frac{1 + n}{2}), \end{matrix}$ where $f_{N}$ is the friction factor for Newtonian fluid evaluated at $R = D V ρ / μ_{eff}$ and $μ_{eff}$ is given by [16]: $\begin{matrix} μ_{eff} = K {(\frac{3 n + 1}{4 n})}^{n - 1} {(\frac{8 V}{D})}^{n - 1} . \end{matrix}$ Shear stress at the pipe wall was then given by: $\begin{matrix} τ_{ω} = f ρ V^{2} / 2, \end{matrix}$ where f is the Fanning friction factor [16].

$f_{N}$ was be calculated using the Churchill equation [16]: $\begin{matrix} \frac{1}{\sqrt{f_{N}}} = - 4 log [\frac{0.27 ϵ}{D} + {(\frac{7}{R})}^{0.9}] valid for R > 4000 . \end{matrix}$ In laminar flow, f is independent of $ϵ / D$ . In turbulent flow, the friction factor for rough pipe follows the smooth tube curve for a range of Reynold’s numbers (hydraulically smooth flow).

ϵ is a measure of surface roughness, and for smooth tubes the ratio of $ϵ / D$ can be as low as 1E−06 [16]. While the surface roughness of the different microcatheters was unknown, they were assumed to be (equally) smooth. Velocities, modified Reynold’s numbers and values of shear stress were calculated according to these equations.

Linear regression analysis was performed between mean shear stress, LDH- and haptoglobin measurements. Correlation coefficients (r) were estimated. Sample size calculation for two-sided analysis was $n = 2$ ( $α = 0.05$ , $β = 0.1$ , $ς = 0.05$ , $δ_{0} = 0.5$ , $power = 1 - β = 0.8$ ).

3. Results

Except for lactate-dehydrogenase and haptoglobin, coagulation, electrolyte and hematologic parameters were unchanged before and after blood reperfusion at all infusion rates for the 4 microcatheters ( $p > 0.05$ , t-test, Table 2).

Table 2
Summary of means for baseline and post-microcatheter infusion measurements including standard deviation for electrolytes, hematology and coagulation parameters

Excelsior SL-10 Vasco+21 Renegade HI-FLO Penumbra 5MAX

Baseline 1 m/s 2 ml/s 3 ml/s Baseline 1 m/s 2 ml/s 3 ml/s Baseline 1 m/s 2 ml/s 3 ml/s Baseline 1 m/s 2 ml/s 3 ml/s

Na⁺ (mmol/l) 138 ± 1 138 ± 0.7 138 ± 0.5 132 ± 5.7 138 ± 1.4 140 ± 0.5 138 ± 2.1 137 ± 0.7 137 ± 1.4 138 ± 0.7 138 ± 0.7 138 ± 0.7 137 ± 0.3 138 ± 0.7 138 ± 0.7 137 ± 0.7

K⁺ (mmol/l) 4.3 ± 0.1 4.5 ± 0.2 4.6 ± 0.1 4.25 ± 0.1 4.3 ± 0.1 4.4 ± 0.1 4.5 ± 0.1 4.7 ± 0.1 4.6 ± 0.1 4.8 ± 0.1 4.8 ± 0.1 5 ± 0.1 4.6 ± 0.3 4.7 ± 0.1 4.7 ± 0.1 4.7 ± 0.1

RBC (10⁹/l) 4.80 ± 0.03 4.72 ± 0.02 4.86 ± 0.01 4.84 ± 0.1 4.80 ± 0.04 4.72 ± 0.09 4.76 ± 0.04 4.78 ± 0.09 4.75 ± 0.03 5.00 ± 0.18 5.01 ± 0.19 5.01 ± 0.06 4.95 ± 0.17 5.19 ± 0.06 5.18 ± 0.12 5.26 ± 0.14

Hb (g/l) 147 ± 0 146 ± 0.7 150 ± 1 151 ± 2.1 147 ± 0 146 ± 1.5 148 ± 0.7 148 ± 2.8 146 ± 1.4 153 ± 6.4 154 ± 5.7 155 ± 2.1 152 ± 7.1 160 ± 3.5 160 ± 4.2 162 ± 5

Hct (%) 43.1 ± 0.2 42.6 ± 0.3 46.2 ± 2.1 43.6 ± 1.1 43.1 ± 0.2 42.3 ± 0.8 42.6 ± 1.4 42.8 ± 1.4 42.3 ± 1 43.9 ± 1.5 44.4 ± 2.1 44.6 ± 0.6 44.3 ± 1 46.3 ± 0.5 46.2 ± 1.1 47.9 ± 1.2

MCV (fL) 89.8 ± 0.2 90.4 ± 0.2 90.5 ± 0.2 90 ± 0.4 89.8 ± 0.2 89.6 ± 0 89.5 ± 0.1 89.6 ± 0.1 89.2 ± 1.6 88 ± 0.1 88.6 ± 0.7 89.1 ± 0 89.5 ± 0.2 89.2 ± 0.1 89.3 ± 0.1 89.2 ± 0.1

MCH (pg) 30.6 ± 0.2 30.9 ± 0 30.9 ± 0.1 31.1 ± 0.1 30.6 ± 0.3 30.9 ± 0.3 31 ± 0.1 31 ± 0 30.8 ± 0.1 30.6 ± 0.2 30.8 ± 0.1 30.9 ± 0.1 30.7 ± 0.4 30.8 ± 0.4 30.9 ± 0.1 30.8 ± 0.1

MCHC (g/l) 341 ± 2 342 ± 1 341 ± 1 346 ± 4 341 ± 3 344 ± 3 347 ± 2 346 ± 0 345 ± 7 347 ± 3 348 ± 4 347 ± 1 342 ± 6 345 ± 4 347 ± 1 345 ± 2

Plt (10⁹/l) 273 ± 8 305 ± 0 268 ± 13 276 ± 12 273 ± 11 264 ± 0 248 ± 22 261 ± 9 276 ± 11 260 ± 15 246 ± 18 229 ± 3 275 ± 9 255 ± 10 250 ± 20 249 ± 8

INR (%) 1.35 ± 0.11 1.28 ± 0.00 1.28 ± 0.02 1.20 ± 0.06 1.35 ± 0.15 1.42 ± 0.03 1.36 ± 0.09 1.40 ± 0.01 1.28 ± 0.03 1.06 ± 0.01 1.08 ± 0.05 1.10 ± 0.08 1.30 ± 0.22 1.21 ± 0.04 1.19 ± 0.00 1.15 ± 0.00

APTT (s) 154 ± 109 219 ± 8 210 ± 4 222 ± 37 154 ± 84 201 ± 0 227 ± 37 268 ± 95 260 ± 2 242 ± 57 245 ± 52 260 ± 35 107 ± 88 211 ± 2 196 ± 4 193 ± 2

Fib (g/l) 3.32 ± 0.05 3.43 ± 0.05 3.35 ± 0.28 3.33 ± 0.26 3.32 ± 0.07 3.31 ± 0.07 3.2 ± 0.38 3.2 ± 0 3.52 ± 0.01 3.54 ± 0.06 3.46 ± 0 3.4 ± 0.02 3.61 ± 0.33 3.63 ± 0.06 3.52 ± 0.08 3.63 ± 0.02

	Excelsior SL-10	Vasco+21	Renegade HI-FLO	Penumbra 5MAX
Na⁺ (mmol/l)	138 ± 1	138 ± 0.7	138 ± 0.5	132 ± 5.7	138 ± 1.4	140 ± 0.5	138 ± 2.1	137 ± 0.7	137 ± 1.4	138 ± 0.7	138 ± 0.7	138 ± 0.7	137 ± 0.3	138 ± 0.7	138 ± 0.7	137 ± 0.7
K⁺ (mmol/l)	4.3 ± 0.1	4.5 ± 0.2	4.6 ± 0.1	4.25 ± 0.1	4.3 ± 0.1	4.4 ± 0.1	4.5 ± 0.1	4.7 ± 0.1	4.6 ± 0.1	4.8 ± 0.1	4.8 ± 0.1	5 ± 0.1	4.6 ± 0.3	4.7 ± 0.1	4.7 ± 0.1	4.7 ± 0.1
RBC (10⁹/l)	4.80 ± 0.03	4.72 ± 0.02	4.86 ± 0.01	4.84 ± 0.1	4.80 ± 0.04	4.72 ± 0.09	4.76 ± 0.04	4.78 ± 0.09	4.75 ± 0.03	5.00 ± 0.18	5.01 ± 0.19	5.01 ± 0.06	4.95 ± 0.17	5.19 ± 0.06	5.18 ± 0.12	5.26 ± 0.14
Hb (g/l)	147 ± 0	146 ± 0.7	150 ± 1	151 ± 2.1	147 ± 0	146 ± 1.5	148 ± 0.7	148 ± 2.8	146 ± 1.4	153 ± 6.4	154 ± 5.7	155 ± 2.1	152 ± 7.1	160 ± 3.5	160 ± 4.2	162 ± 5
Hct (%)	43.1 ± 0.2	42.6 ± 0.3	46.2 ± 2.1	43.6 ± 1.1	43.1 ± 0.2	42.3 ± 0.8	42.6 ± 1.4	42.8 ± 1.4	42.3 ± 1	43.9 ± 1.5	44.4 ± 2.1	44.6 ± 0.6	44.3 ± 1	46.3 ± 0.5	46.2 ± 1.1	47.9 ± 1.2
MCV (fL)	89.8 ± 0.2	90.4 ± 0.2	90.5 ± 0.2	90 ± 0.4	89.8 ± 0.2	89.6 ± 0	89.5 ± 0.1	89.6 ± 0.1	89.2 ± 1.6	88 ± 0.1	88.6 ± 0.7	89.1 ± 0	89.5 ± 0.2	89.2 ± 0.1	89.3 ± 0.1	89.2 ± 0.1
MCH (pg)	30.6 ± 0.2	30.9 ± 0	30.9 ± 0.1	31.1 ± 0.1	30.6 ± 0.3	30.9 ± 0.3	31 ± 0.1	31 ± 0	30.8 ± 0.1	30.6 ± 0.2	30.8 ± 0.1	30.9 ± 0.1	30.7 ± 0.4	30.8 ± 0.4	30.9 ± 0.1	30.8 ± 0.1
MCHC (g/l)	341 ± 2	342 ± 1	341 ± 1	346 ± 4	341 ± 3	344 ± 3	347 ± 2	346 ± 0	345 ± 7	347 ± 3	348 ± 4	347 ± 1	342 ± 6	345 ± 4	347 ± 1	345 ± 2
Plt (10⁹/l)	273 ± 8	305 ± 0	268 ± 13	276 ± 12	273 ± 11	264 ± 0	248 ± 22	261 ± 9	276 ± 11	260 ± 15	246 ± 18	229 ± 3	275 ± 9	255 ± 10	250 ± 20	249 ± 8
INR (%)	1.35 ± 0.11	1.28 ± 0.00	1.28 ± 0.02	1.20 ± 0.06	1.35 ± 0.15	1.42 ± 0.03	1.36 ± 0.09	1.40 ± 0.01	1.28 ± 0.03	1.06 ± 0.01	1.08 ± 0.05	1.10 ± 0.08	1.30 ± 0.22	1.21 ± 0.04	1.19 ± 0.00	1.15 ± 0.00
APTT (s)	154 ± 109	219 ± 8	210 ± 4	222 ± 37	154 ± 84	201 ± 0	227 ± 37	268 ± 95	260 ± 2	242 ± 57	245 ± 52	260 ± 35	107 ± 88	211 ± 2	196 ± 4	193 ± 2
Fib (g/l)	3.32 ± 0.05	3.43 ± 0.05	3.35 ± 0.28	3.33 ± 0.26	3.32 ± 0.07	3.31 ± 0.07	3.2 ± 0.38	3.2 ± 0	3.52 ± 0.01	3.54 ± 0.06	3.46 ± 0	3.4 ± 0.02	3.61 ± 0.33	3.63 ± 0.06	3.52 ± 0.08	3.63 ± 0.02

Notes: None of the evaluated microcatheters and flow rates have demonstrated significant differences among baseline and test samples ( $p > 0.05$ , t-test). Blood samples were obtained from a single test person to reduce variability. Consequently standard deviations are generally small.

3.1. Lactate-dehydrogenase (LDH)

Mean baseline LDH concentration from all experiments was 163 ± 23 U/l. Significant increase of LDH was found after perfusion of the Excelsior SL-10 and Vasco+21 microcatheters at all three infusion rates ( $p < 0.05$ , t-test; Table 3).

Table 3
Means and standard deviations of LDH- and haptoglobin concentrations before and after blood perfusion through the four microcatheters at flow rates of 1-, 2- and 3 ml/s

Infusion rate (ml/s) Excelsior SL-10 Vasco+21 Renegade HI-FLO Penumbra 5MAX

LDH (U/l) Haptoglobin (g/l) LDH (U/l) Haptoglobin (g/l) LDH (U/l) Haptoglobin (g/l) LDH (U/l) Haptoglobin (g/l)

Baseline 155 ± 3 1.89 ± 0.02 157 ± 2 1.84 ± 0.01 189 ± 45 1.64 ± 0.14 154 ± 6 1.82 ± 0.02

1 ml/s 249 ± 40 (p<0.05) 1.54 ± 0.01 (p<0.005) 165 ± 1 (p<0.05) 1.79 ± 0.01 (p<0.05) 173 ± 6 (p>0.05) 1.80 ± 0.01 (p>0.05) 150 ± 10 (p>0.05) 1.81 ± 0.01 (p>0.05)

2 ml/s 251 ± 8 (p<0.005) 1.53 ± 0.01 (p<0.005) 170 ± 2 (p<0.05) 1.78 ± 0.01 (p<0.05) 179 ± 4 (p>0.05) 1.80 ± 0.04 (p>0.05) 151 ± 1 (p>0.05) 1.81 ± 0.01 (p>0.05)

3 ml/s 292 ± 2 (p<0.005) 1.47 ± 0.05 (p<0.005) 182 ± 9 (p<0.05) 1.79 ± 0.01 (p<0.05) 204 ± 35 (p>0.05) 1.78 ± 0.06 (p>0.05) 152 ± 0 (p>0.05) 1.80 ± 0 (p>0.05)

Infusion rate (ml/s)	Excelsior SL-10	Vasco+21	Renegade HI-FLO	Penumbra 5MAX
Baseline	155 ± 3	1.89 ± 0.02	157 ± 2	1.84 ± 0.01	189 ± 45	1.64 ± 0.14	154 ± 6	1.82 ± 0.02
1 ml/s	249 ± 40 (p<0.05)	1.54 ± 0.01 (p<0.005)	165 ± 1 (p<0.05)	1.79 ± 0.01 (p<0.05)	173 ± 6 (p>0.05)	1.80 ± 0.01 (p>0.05)	150 ± 10 (p>0.05)	1.81 ± 0.01 (p>0.05)
2 ml/s	251 ± 8 (p<0.005)	1.53 ± 0.01 (p<0.005)	170 ± 2 (p<0.05)	1.78 ± 0.01 (p<0.05)	179 ± 4 (p>0.05)	1.80 ± 0.04 (p>0.05)	151 ± 1 (p>0.05)	1.81 ± 0.01 (p>0.05)
3 ml/s	292 ± 2 (p<0.005)	1.47 ± 0.05 (p<0.005)	182 ± 9 (p<0.05)	1.79 ± 0.01 (p<0.05)	204 ± 35 (p>0.05)	1.78 ± 0.06 (p>0.05)	152 ± 0 (p>0.05)	1.80 ± 0 (p>0.05)

Notes: These differences were significant ( $p < 0.05$ , t-test) or highly significant ( $p < 0.005$ , t-test) at all flow rates for the Excelsior SL-10 and Vasco+21 microcatheters. No significant differences between baseline and post-infusion measurements were found for the Renegade HI-FLO and Penumbra 5 MAX catheters ( $p > 0.05$ , t-test).

Linear regression analysis showed progressive increase of LDH with higher flow rates for the Excelsior SL-10 and Vasco+21 microcatheters. Correlation between flow rates and LDH levels was $r = 0.92 and 0.99$ for these microcatheters, respectively. The Renegade HI-FLO and Penumbra 5MAX microcatheters did not show differences of LDH concentrations before and after blood reperfusion at all infusion rates ( $p > 0.05$ , t-test, Table 3). Correlation between LDH concentration and flow rates was $r = 0.8$ for the Renegade HI-FLO and $r = 0.3$ for the Penumbra 5MAX.

3.2. Haptoglobin

Mean baseline haptoglobin concentration was 1.80 ± 0.07 g/l. Significant reduction of haptoglobin was found after blood infusion through the Excelsior SL-10 and Vasco+21 microcatheters at all three infusion rates ( $p < 0.05$ , t-test; Table 3). Regression analysis demonstrated decreased haptoglobin concentrations with increasing infusion rates for the Excelsior SL-10 and Vasco+21 microcatheters. Correlation between flow rates and haptoglobin levels was $r = - 0.87 and - 0.72$ , respectively for these microcatheters. Haptoglobin concentration remained unchanged before and after blood infusion through the Renegade HI-FLO and Penumbra 5MAX catheters at all three flow rates ( $p > 0.05$ , t-test, Table 3). Correlation between haptoglobin levels and flow rates was $r = - 0.7$ for the Renegade HI-FLO and $r = - 1$ for the Penumbra 5MAX.

3.3. LDH/haptoglobin relationship

Regression analysis between LDH and haptoglobin for all samples before and after blood perfusion and for all catheters and infusion rates demonstrated a negative relationship ( $r = - 0.93$ , Fig. 1). For the Excelsior SL-10 and Vasco+21 microcatheters a significant ( $p < 0.05$ , t-test) or highly significant ( $p < 0.005$ , t-test) increase of the LDH/haptoglobin ratio was found. The LDH/haptoglobin ratio remained unchanged for the Renegade HI-FLO and Penumbra 5MAX catheters at all flow rates ( $p > 0.05$ , t-test).

Fig. 1.

Linear regression between mean haptoglobin and LDH levels for all samples before and after blood perfusion of all microcatheters and infusion rates, demonstrating a negative relationship between LDH and haptoglobin ( $r = - 0.93$ ).

3.4. Infusion volume and time

At the selected uniform pressure limit of 5171 kPa, the full selected blood volume of 10 ml was delivered in expected time for the Vasco+21, Renegade HI-FLO and Penumbra 5MAX microcatheters at flow rates of 1 ml/s (10 s infusion time), 2 ml/s (5 s infusion time) and 3 ml/s (3.3 s infusion time).

The preset pressure limit was exceeded with the Excelsior SL-10 microcatheter during blood infusion at rates of 2- and 3 ml/s. The power injector aborted blood infusion when 6 ml of blood had been infused at a rate of 2 ml/s, and when 4 ml were infused at a flow rate of 3 ml/s. Subsequently infusion was repeated immediately to obtain sufficient sample volumes for biochemical analysis. The Excelsior SL-10 microcatheter reliably delivered 10 ml of blood within 10 s at an infusion rate of 1 ml/s without reaching the pressure limit of 5171 kPa. No microcatheter or hub connector rupture occurred during the experiments.

3.5. Flow velocities, modified Reynold’s numbers and shear stress

Flow velocities in microcatheters ranged between 0.6 and 21.6 m/s and modified Reynold’s numbers between 7753 and 718,917, which means that for all experiments, the flow regime was turbulent [17]. Resulting shear stress ranged from 0.9 to 657 Pa. Resulting velocities, modified Reynold’s numbers and shear stress for each microcatheter/flow rate combination are summarized in Table 4. Positive correlation was found between LDH levels and shear stress ( $r = 0.86$ ), whereas correlation was negative between haptoglobin concentration and shear stress ( $r = - 0.75$ ). Figure 2 shows that below shear stress of 100 Pa, no significant change in LDH levels or haptoglobin concentration is found. Above shear stress of 100 Pa however, the changing concentrations indicate progressive hemolysis of blood cells.

Table 4
Infusion rate related blood flow velocities, modified Reynold’s numbers and shear stress within the 4 microcatheters

Microcatheter Velocity (m/s) Modified Reynold’s number Shear stress (Pa)

1 ml/s 2 ml/s 3 ml/s 1 ml/s 2 ml/s 3 ml/s 1 ml/s 2 ml/s 3 ml/s

Excelsior SL-10 7.2 14.4^* 21.6^* 148,980 402,161 718,917 88 312 657

Vasco+21 3.4 7.1 10.6 65,640 177,190 316,751 24 86 180

Renegade HI-FLO 2.7 5.3 8 47,609 128,516 229,740 15 52 109

Penumbra 5MAX 0.6 1.1 1.7 7753 20,930 37,415 0.9 3.2 6.6

Microcatheter	Velocity (m/s)	Modified Reynold’s number	Shear stress (Pa)
Excelsior SL-10	7.2	14.4^*	21.6^*	148,980	402,161	718,917	88	312	657
Vasco+21	3.4	7.1	10.6	65,640	177,190	316,751	24	86	180
Renegade HI-FLO	2.7	5.3	8	47,609	128,516	229,740	15	52	109
Penumbra 5MAX	0.6	1.1	1.7	7753	20,930	37,415	0.9	3.2	6.6

indicates that the predefined pressure limit was exceeded.

Fig. 2.

Stacked line graph with markers for LDH (U/l) and haptoglobin (g/l) concentrations in relation to shear stress. Below 100 Pa, no significant change in LDH levels or haptoglobin concentration is evident. Above 100 Pa however, the changing concentrations indicate progressive hemolysis of blood cells, until degradation is complete. Correlation between shear stress, LDH and haptoglobin was $r = 0.96$ and $r = - 0.75$ , respectively.

4. Discussion

The only effective treatment for arterial occlusions, particularly in acute stroke is rapid recanalization of the occluded vessel by restoring physiological blood flow [1,3]. Time to recanalization is critical to prevent irreversible deficits [2,10,11]. Previous studies have demonstrated that recanalization frequently occurs too late, resulting in infarcts with lifelong disability or death [18]. In stroke settings it has recently been demonstrated that endovascular transarterial procedures are associated with high rates of arterial recanalization and superior to conservative management approaches [4–9]. However, definitive recanalization with mechanical thrombectomy can be challenging in certain clinical situations, resulting in delayed reperfusion of penumbral tissue [2].

To reduce the critical reperfusion time in challenging situations, an adjunct blood reperfusion technique has been proposed [2,10,11]. Heparinised arterial blood from an aortic or iliac artery sheath is reinfused through a microcatheter positioned with its tip distal to the arterial occlusion. Once blood supply to the ischemic territory has been temporarily reconstituted, the underlying thromboembolic occlusion may subsequently be treated by means of thrombolysis, angioplasty or mechanical thrombectomy. Such a microcatheter reperfusion technique may consequently be used when initial vacuum-, respectively, stent retriever assisted thrombectomy fails in order to ‘buy more time’ for definite and sustainable treatment of the underlying condition. Previous publications have suggested that such microcatheter directed perfusion of ischemic brain territories with oxygenated blood through arterial occlusions is clinically feasible and safe [2,10,11]. However, this technique may be associated with significant blood trauma and hemolysis, which may result in hemorrhage or progressive ischemia within the affected brain tissue. Such effects have not been investigated previously.

4.1. Markers of hemolysis

Changes in LDH and serum haptoglobin levels are considered the most sensitive general tests to determine mechanical hemolysis [19]. Consequently this study was focused on investigation of these two basic biochemical markers. The correlation found between LDH and haptoglobin ( $r = - 0.93$ , Fig. 1), confirms that these markers are suitable for assessment of hemolysis.

4.2. Effects of mechanical hemolysis

It has previously been revealed that blood trauma may result from non-physiological flow conditions such as elevated shear forces, turbulence, cavitation, prolonged contact and collision between blood cells and foreign surfaces [14]. These factors are associated with a variety of damage mechanisms: overstretching or fragmentation of erythrocytes causing release of free hemoglobin, activation or dysfunction of platelets and leukocytes, increased concentrations of inflammatory mediators and complement activation. Even low levels of hemolysis have been shown to increase RBC aggregation at low shear conditions [20]. Consequently even low level hemolysis is considered an important warning sign of other potential blood cell damage such as platelet activation, white blood cell dysfunction and other serious complications such as scavenging of nitric oxide, damage to glycocalyx and endothelial cells, and impairment of the vascular smooth muscle tone [14]. In consequence, hemolysis associated with microcatheter directed blood reperfusion may promote platelet aggregation, hypercoagulation and intravascular thrombosis [14]. Beside complete cellular destruction, damage can also be induced to the RBC membrane on a sublethal level, resulting in decreased deformability and increased aggregability, again resulting in an increased risk of intravascular thrombosis. This phenomenon however may be reversible to a certain extent depending on duration and intensity of damage [21]. A decrease in cell deformability results in a reduced capacity for RBC’s to enter small capillaries and a reduced contact of the cell surface with the surrounding vessel wall, compromising microcirculation and oxygen delivery to surrounding tissue [22].

When RBC membranes have been overstretched or fragmented by non-physiological blood flow conditions, plasma free hemoglobin (PfHb) and heme enter the circulation. PfHb is cleared by the Hb scavengers, haptoglobin and CD163. Additionally LDH is released when red blood cells are destroyed [23]. The resulting products from hemolysis together with interleukin-10 release, induced by the Hb clearing mechanism exert an antioxidant, anticoagulant, and vasodilating effect, thus compensating for the adverse effects caused by PfHb and heme [19]. However, on saturation of free Hb scavengers, PfHb binds nitric oxide (NO) derived from the vascular endothelium in a fast and irreversible way [24]. NO depletion then results in elevated vascular resistance, increased thrombin formation, fibrin deposition, platelet activation and -aggregation [25]. In summary PfHb causes endothelial dysfunction and vasomotor instability by reducing nitric oxide bioavailability [13].

4.3. Shear stress and modified Reynold’s number associated with blood reperfusion

The mechanisms of mechanical cell damage during catheter reperfusion have been described by different investigators [26,27], identifying positive and negative pressure, wall impact forces, blood non-endothelial surfaces, time of exposure and blood–air interference as responsible parameters for blood trauma. RBC’s seem to have a high tolerance to most of these forces, except for shear stress [28]. Physiologic intravascular shear stress ranges from 0.1–5 Pa, sublethal damage is reported to appear from 21–43 Pa, lethal damage to RBC occurs from 150 Pa, and above. The thresholds for platelet and leucocyte activation are, respectively, ∼10 and 7.5 Pa [29]. Beside shear stress, duration of exposure is another factor related to cellular damage [29]. One hour of exposure to wall shear stress exceeding ∼40 Pa was found to damage endothelial cells leading to their destruction [14].

It has been described that exposure time related to laminar shear stress may result in almost five orders of magnitude for hemolysis and shear stresses up to 500 Pa, where exposure times varied between 60 and 300 s [27]. In the present study however, exposure times varied only between 3.3 and 10 s, consequently limiting the effect of exposure time on results.

Table 4 shows the values for velocity, modified Reynold’s numbers, and shear stress, for all combinations of catheters and flow rates evaluated. According to these calculations, shear stress within the Excelsior SL-10 microcatheter ranged between 87 and 657 Pa. These forces would subsequently be considered within or close to the lethal range for RBC damage [29], reflected by significant hemolysis in all analyzed re-perfused blood samples with this microcatheter. The Vasco+21 microcatheter demonstrated a lower shear stress range between 24–180 Pa. These forces are within the reported range of sublethal–lethal RBC damage [29] and again significant hemolysis was found after blood reperfusion at all infusion rates. With the Renegade HI-FLO microcatheter shear stress was further reduced and ranged between 15–109 Pa. Despite exceeding the reported sublethal shear stress-level of RBC damage [29], no associated hemolysis was found in all corresponding samples after blood reperfusion. Shear stress within the Penumbra 5 MAX catheter remained below the sublethal level [29] at all infusion rates (0.9–6.6 Pa) and consequently no hemolysis was found.

Turbulent flow is associated with significantly increased hemolysis when compared with laminar flow at the same wall shear stress level [17]. Furthermore, in turbulent flow there are additional effects due to the interaction of the smallest sized eddies (Kolmogorov-sized eddies) with the red blood cells. Assuming dissipation of eddies occurs instead via cell–cell interactions mediated by the plasma, turbulent velocity fluctuations, normally ascribed to turbulent (Reynold’s) stresses, give rise to viscous shear stresses [30]. These effects were not investigated with our experimental design.

4.4. Study limitations

This in-vitro study is associated with certain limitations. In a clinical setting thromboembolic occlusions usually occur at branch bifurcations. Microcatheter directed whole blood reperfusion, however, is generally targeting a single branch only. Consequently other occluded branches may not benefit from reperfusion. In this experiment venous blood has been used to study effects on hemolytic markers. We have not investigated whether arterial blood samples may be affected from a loss of oxygen saturation, even though this appears unlikely. In this experiment the microcatheters were directly connected to the power injector without connecting tubing. In a clinical setting this approach would be impractical and the additional tubing length may impact the extent of hemolysis. The current assessment of haemolysis is incompletely defined by shear stress considerations alone. Additional factors, such as cell-wall interactions, roughness of the microcatheter luminal wall, surface charge, time of exposure, etc. may be relevant for hemolysis but were not assessed in this investigation. When using microcatheter assisted blood reperfusion in vivo, an increased postobstructive endovascular pressure related to high flow rates may potentially cause proximal dislodgement of thrombus with transfer of embolic material into different arterial branches. However, since cerebral autoregulation in ischemic stroke is associated with vasodilation, distal vascular resistance and pressure would be decreased [31]. This lack of vascular resistance and the persistent trans-occlusive pressure gradient should contribute to a decreased risk of proximal clot migration during whole blood reperfusion. Even though results of this study have not demonstrated coagulation disorders in vitro, the use of heparinized blood may be associated with an increased risk of hemorrhage, particularly in stroke cores where the blood brain barrier may have already been irreversibly damaged. Apart from such considerations, blood reperfusion may help to locally clear vaso- or neurotoxic substances that have accumulated within the ischemic territory and which may otherwise promote hemorrhagic complications [2]. Current revascularization techniques with stent retrievers frequently generate instant reperfusion channels [4,6–9] and the ADAPT technique seems to provide the fastest stroke recanalization time [5]. In consequence microcatheter directed blood reperfusion should be reserved for complex or challenging cases, where initial revascularization with stent retrievers of vacuum assisted clot extraction fails, in order not to delay definitive and efficiently proven treatment options.

5. Conclusions

Microcatheters with inner luminal diameters of less than 0.69 mm seem to induce excessive hemolysis during whole blood reperfusion. Apart from other factors, such as exposure time, the combination of microcatheter diameter and perfusion rate seems to determine the resulting shear stress, and associated hemolysis under turbulent flow conditions. This study indicates that significant hemolysis may occur at shear stress values greater than 100 Pa, which is in agreement with previous findings [28]. Microcatheters with diameters of 0.69 mm or more did not induce significant hemolysis at flow rates up to 3 ml/s. Depending on required flow rates, larger lumen microcatheters consequently may be the preferred choice for autologous blood reperfusion in acute stroke settings.

References

[1] Molina

Saver

. Extending reperfusion therapy for acute ischemic stroke – Emerging pharmacological, mechanical, and imaging strategies. Stroke. 2005;36:2311–2320.

[2] Ribo

Rubiera

Pagola

Rodriguez-Luna

Meler

Flores

, et al. Bringing forward reperfusion with oxygenated blood perfusion beyond arterial occlusion during endovascular procedures in patients with acute ischemic stroke. Am J Neuroradiol. 2010;31:1899–1902.

[3] Gralla

Brekenfeld

Mordasini

Schroth

. Mechanical thrombolysis and stenting in acute ischemic stroke. Stroke. 2012;43:280–285.

[4] Berkhemer

Fransen

Beumer

van den Berg

Lingsma

Yoo

, et al. A randomized trial of intraarterial treatment for acute ischemic stroke. N Engl J Med. 2015;372:11–20.

[5] Turk

Frei

Fiorella

Mocco

Baxter

Siddiqui

, et al. ADAPT FAST study: A direct aspiration first pass technique for acute stroke thrombectomy. J NeuroIntervent Surg. 2014;6:260–264.

[6] Saver

Goyal

Bonafe

Diener

Levy

Pereira

, et al. Solitaire™ with the intention for thrombectomy as primary endovascular treatment for acute ischemic stroke (SWIFT PRIME) trial: Protocol for a randomized, controlled, multicenter study comparing the solitaire revascularization device with IV tPA with IV tPA alone in acute ischemic stroke. Int J Stroke. 2015;10:439–448.

[7] Goyal

Demchuk

Menon

Eesa

Rempel

Thornton

, et al. Randomized assessment of rapid endovascular treatment of ischemic stroke. N Engl J Med. 2015;12:1019–1030.

[8] Molina

Chamorro

Rovira

de Miquel

Serena

Roman

, et al. REVASCAT: A randomized trial of revascularization with SOLITAIRE FR device vs. best medical therapy in the treatment of acute stroke due to anterior circulation large vessel occlusion presenting within eight-hours of symptom onset. Int J Stroke. 2015;10:619–626.

[9] Campbell

Mitchell

Kleinig

Dewey

Churilov

Yassi

, et al. Endovascular therapy for ischemic stroke with perfusion-imaging selection. N Engl J Med. 2015;12:1009–1018.

10.

[10] Chan

Lai

Liu

. Autologous heparinized oxygenated blood reperfusion in acute ischaemic stroke caused by infective endocarditis: A case report. Acta Neurol Taiwan. 2011;20:267–271.

11.

[11] Ribó

Molina

Dinia

Alvarez-Sabin

Matas

. Buying time for recanalization in acute stroke: Arterial blood infusion beyond the occluding clot as a neuroprotective strategy. J Neuroimaging. 2009;19:188–190.

12.

[12] Schae

Buehler

Alayash

Belcher

Vercellotti

. Hemolysis and free hemoglobin revisited: Exploring hemoglobin and hemin scavengers as a novel class of therapeutic proteins. Blood. 2013;121:1276–1284.

13.

[13] Minneci

Deans

Zhi

Yuen

Star

Banks

, et al. Hemolysis-associated endothelial dysfunction mediated by accelerated NO inactivation by decompartmentalized oxyhemoglobin. J Clin Invest. 2005;115:3409–3417.

14.

[14] Baskurt

Hardeman

Rampling

Meiselman

. Handbook of hemorheology and hemodynamics. Amsterdam: IOS Press; 2007.

15.

[15] Castellini

Baskurt

Castellini

Meiselman

. Blood rheology in marine mammals. Front Physiol. 2010;1:146.

16.

[16] Green

Perry

. Perry’s chemical engineers’ handbook. 7th ed. Australia: McGraw-Hill; 1998.

17.

[17] Kamenava

Burgreen

Kono

Repko

Antaki

Umezu

. Effects of turbulent stresses upon mechanical hemolysis: Experimental and computational analysis. ASAIO J. 2004;50:418–423.

18.

[18] Ribo

Alvarez-Sabin

Montaner

Romero

Delgado

Rubiera

, et al. Temporal profile of recanalization after intravenous tissue plasminogen activator: Selecting patients for rescue reperfusion techniques. Stroke. 2006;37:1000–1004.

19.

[19] Mohammadzadeh

Jafari

Hasanpour

Sahandifar

Ghafari

Alaei

. Effects of pulsatile perfusion during cardiopulmonary bypass on biochemical markers and kidney function in patients undergoing cardiac surgeries. Am J Cardiovasc Dis. 2013;3:158–162.

20.

[20] Seiyama

Suyuki

Tateshi

Maeda

. Viscous properties of partially hemolized erythrocyte suspension. Japanese Society of Biorheology. Annual Meeting 1990. Biorheology. 1991;28:452.

21.

[21] Lee

Antaki

Kamenava

Dobbe

Hardeman

Ahn

, et al. Strain hardening of red blood cells by accumulated cyclic supraphysiological stress. Artif Organs. 2007;31:80–86.

22.

[22] Watanabe

Sakota

Ochuchi

Tatakani

. Deformability of red blood cells and its relation to blood trauma in rotary blood pumps. Artif Organs. 2007;31:352–358.

23.

[23] Vercaemst

. Hemolysis in cardiac surgery patients undergoing cardiopulmonary bypass: A review in search of a treatment algorithm. J Extra Corp Technol. 2008;40:257–267.

24.

[24] Tseng

Lin

Huang

Lin

Mao

. Antioxidant role of human haptoglobin. Proteomics. 2004;4:2221–2224.

25.

[25] Langlois

Delanghe

. Biological and clinical significance of haptoglobin polymorphism in humans. Clin Chem. 1996;42:1589–1600.

26.

[26] Mulholland

Massay

Shelton

. Investigation and quantification of the blood trauma caused by the combined dynamic forces experienced during cardiopulmonary bypass. Perfusion. 2001;16:385–394.

27.

[27] Arwatz

Smits

. A viscoelastic model of shear-induced hemolysis in laminar flow. Biorheology. 2013;50:45–55.

28.

[28] Wright

. Hemolysis during cardiopulmonary bypass. Update Perfusion. 2001;16:345–351.

29.

[29] Leverett

Hellums

Alfrey

Lynch

. Red blood cell damage by shear stress. Biophys J. 1972;12:257–273.

30.

[30] Antiga

Steinman

. Rethinking turbulence in blood. Biorheology. 2009;46:77–81.

31.

[31] Palomares

Cipolla

. Vascular protection following cerebral ischemia and reperfusion. J Neurol Neurophysiol. 2011;20:1–4.