Thrombosis in the microcirculation Bingmei M Fu Department

Thrombosis in the microcirculation Bingmei M. Fu Department of Biomedical Engineering The City College of The City University of New York Microcirculation Lab

Thrombosis Normal blood flow Thrombosis Embolus 30µm large vessel microvessel

Risk Factors Vessel injury Flow Stasis Surgery Trauma Heart failure Altered coagulability Increased blood Prolonged immobilization coagulability Thrombosis

Thrombosis in microcirculation • Previous studies under conditions – flow retarded (Nicolaides et al. , 1972); – flow disturbed • secondary flow in branches (Chen et al. , 2004); – vessel injured/damaged by • electrical (Massad et al. , 1987; Wong et al. , 2000); • mechanical (oude Egbrink et al. , 1988); • biochemical (Begent et al. , 1970); • PDT, light/dye (Sato et al. , 1990; Sasaki et al. , 1996; Rucker et al. , 2002).

Question 1 Can thrombosis occur in non-injured but bent/stretched microvessels under non-disturbed laminar flow (Re ~0. 01) condition?

Question 2 What is the structural mechanism by which PDT induces thrombosis?

Experimental setup CCD camera

Experimental Design • Sprague-Dawley rat, 250 -300 g; • The mesentery gently arranged on the surface of a polished quartz pillar; • With a microvessel (postcapillary venule, 20 -50 µm) under observation, a roundedtip glass restraining micropipette used to bend/stretch the microvessel.

micropipette inlet Experimental Observations outlet • In 10 -60 min, thrombi formed in 19 out of 61 (~31%) sites in 28 noninjured bent/stretched microvessels; • all thrombi were initiated from the inner side of the curvature. (Liu et al, J. Biomech. 2008)

Results (Liu et al, J. Biomech. 2008)

What are the mechanical factors that initiate thrombosis in these non-injured bent/stretched microvessels?

Vessel Geometry A A θ B Ө 90 o A 180 o B B 2 r 2 a 2 b microvessel diameter 2 r = 25 µm (circular) equal perimeters of circular and elliptical cross sections (Liu et al, J. Biomech. 2008)

Numerical Methods • Fluent used to solve • Continuity Eq. • • • Navier-Stokes Eq. Element No. : • 610 x 103, 770 x 103 (90 o/180 o, circular); • 1. 48 x 106, 1. 52 x 106 (90 o/180 o, elliptical); µ = 2. 5 cp (Levenson et al. , 1990), ρ = 1050 kg/m 3; Outlet pressure = 10 cm. H 2 O, mean blood velocity = 1 mm/s, Reynolds No. ~ 0. 01; Convergence criteria: 10 -8 of residues.

Velocity Distributions Circular Elliptical (Liu et al, J. Biomech. 2008)

Shear Rate Distributions Circular A—A A—A A A max Elliptical max A (Liu et al, J. Biomech. 2008)

Shear Rate Distributions along the Curvature (Circular) straight 90 o 180 o (Liu et al, J. Biomech. 2008)

Shear Rate along the Curvature inlet outlet Circular inlet outlet Elliptical (Liu et al, J. Biomech. 2008)

Velocity along the Curvature inlet outlet Circular inlet outlet Elliptical (Liu et al, J. Biomech. 2008)

Pressure along the Curvature Circular Elliptical (Liu et al, J. Biomech. 2008)

Newtonian & non-Newtonian fluid Casson Model (Das et al. , 1998, 2007) α = 1. 621, β = 0. 627 (Das et al. , 1998); μp = 2. 5 c. P (Levenson et al. , 1990). (Liu et al, J. Biomech. 2008)

Summary • Thrombosis occurred in 19 out of 61 sites (31%) of 28 non-injured bent/stretched microvessels. Thrombi were initiated at the inner side of these microvessels. • Numerical simulation results showed higher shear stress/rate and higher shear stress/rate gradient at the inner sides of the bent/stretched microvessels, suggesting they were two mechanical factors that initiate thrombi.

Light-dye Treatment Light-dye treatment (Photodynamic Therapy, PDT): Use of a photosensitizer, activated by a laser of a specific wavelength, to treat tumor and other diseases in the presence of oxygen.

Advantages of PDT • Applied repeatedly at the same site; • Selective: photosensitizer can selectively accumulate in the tumor cells; • Harmless without light illumination; • Treatment for diseases that surgery is not possible (such as the upper bronchi, the structure cannot be removed surgically).

PDT Induced Thrombosis • Thrombi induced by light/dye consist primarily of platelets and occasionally of leukocytes in venules (Rumbaut et al. , 2004). • The interaction between blood platelets and vessel wall plays an important role in thrombosis.

Surface Glycocalyx glycocalyx 150 nm (Squire et al. , 2001)

Molecular Composition of SGL (Tarbell and Pahakis, J. Intern. Med , 2006)

Glycocalyx Layer Damage • Light/dye increases the penetration of macromolecules in the endothelial surface glycocalyx of the vascular wall (Vink & Duling, 1996). • Disruption of the glycocalyx would result in adhesion of platelets and blood cells to the vessel wall (Mulivor & Lipowsky, 2002).

Hypotheses PDT disrupts the endothelial surface glycocalyx increase microvessel permeability to water and solutes platelet and blood cells binding to the endothelium and induces thrombi.

Experimental Setup Laser Spot CCD camera VCR Xenon Laser (495 nm)

Experimental Design • Sprague-Dawley rat, 250 -300 g; • Laser: Xenon laser, 495 nm, intensity 0. 37 and 0. 70 m. W/mm 2; • Na. F: 50 mg/kg body wt. , injected from the carotid artery; • With a microvessel (post-capillary venule, 20 -50 µm) under observation, Na. F was injected and the laser was turned on simultaneously.

Thrombosis by Light/dye 20µm

Thrombus Growth Rate 15. 5 ± 1. 8 29. 3 ± 2. 2 2. 5 3. 8 (Liu et al. , BMMB, 2010)

Technique of Lp Measurement (Curry, 1984) Marker Cell 2 r L 0 d. L/dt Micropipette Blocker

Lp Measurement 30µm

Lp Change under Light/Dye Treatment (Liu et al. , BMMB, 2010)

Early Change of Lp under Light/Dye Treatment * (Liu et al. , BMMB, 2010)

Results Lp baseline Lp PDT (mean±SE) 10 -7 cm/s/cm. H 2 O n=11 (mean±SE) 10 -7 cm/s/cm. H 2 O n = 11 0. 98± 0. 08 2. 07± 0. 21 (range: 0. 69 -1. 52) 1. 11 -3. 16) Lp PDT / Lpbaseline (mean±SE) 2. 2± 0. 2 (Liu et al. , BMMB, 2010)

Technique of P Measurement Dye side Washout side (d. I/dt)0 400 µm Measuring Window ∆Ifo 200 µm (Fu et al. , 2005) 10 seconds

P Measurement

P to albumin * (Liu et al. , BMMB, 2010)

Early Change of P under Light/dye Treatment * (Liu et al. , BMMB, 2010)

Results P baseline PPDT (mean±SE) 10 -6 cm/s n=7 PPDT / Pbaseline 0. 84± 0. 09 3. 46± 0. 53 4. 1± 0. 7 (mean±SE) (Liu et al. , BMMB, 2010)

Comparison of Permeability Change w. & w/o. Blood Cells Baseline Na. F + laser Lp Baseline Na. F + laser P to albumin (Liu et al. , BMMB, 2010)

What are the most likely structural mechanisms by which light-dye treatment induced microvascular hyperpermeability and thrombosis?

Model Geometry Y Junction strand 2 a Z Tissue side 2 D Surface glycocalyx Lumen side Δ X 2 B Y 2 d X O L 2 d 2 D Lumen side Lf L Lf (revised from Fu et al. , 1994)

Model Predictions (Liu et al. , BMMB, 2010)

Model Predictions and Exp. Results 0. 08

Model Predictions and Exp. Results 0. 14

Growth Rate vs. Permeability

Summary • Light/dye treatment with 0. 37 m. W/mm 2 induced thrombosis in 3. 8 min, complete occlusion at ~29 min. • This power gradually increased Lp and Palbumin to a plateau in 3 – 5 min by 2. 2 -fold and 4. 1 -fold respectively. • Our model predictions indicated that Lp and P increase under light/dye treatment was most likely due to 86% - 92% diminishment of the endothelial surface glycocalyx.

Summary • Increased Lp would increase the radial fluid flow and enable more platelets and leukocytes move towards the vessel wall. • Degradation of glycocalyx layer exposes endothelium and increases the binding of platelets and other blood cells to the endothelial cells, therefore induces thrombosis.

Acknowledgements Dr. Qin Liu David Mirc Min Zeng NSF CBET- 0133775 and 0754158 CUNY Graduate Fellowship Thank you!