In Vivo Sickle Cell Blood Rheology in Humans with Ventral Tongue Microscopy

INTRODUCTION RBC sickling has been studied and visualized using models that approximate the in vivo human microcirculatory environment. Ex vivo microfluidic studies have characterized cell adherence, occlusion rates, cell deformability, and other rheological parameters, establishing biomarkers that reflect disease state, forms of SCD, and treatment efficacy. Intravital microscopy on SCD mouse models has been used to develop theories for vaso-occlusion mechanisms leading to therapeutic strategies. However, none of these models can accurately mimic the human microcirculatory environment. We have developed a label-free and non-invasive microscope to observe blood flow in vivo in humans that uses oblique back-illumination (OBM) to acquire high-resolution, high-speed (>100 frames per second) videos of blood cells flowing through superficial capillaries (<100 um deep) in the ventral tongue. Adhered sickle cells, occlusion mechanisms, and flow dynamics can be observed with subcellular detail, potentially allowing for better evaluation of therapy efficacy, personalized treatment planning, and identification of therapeutic targets enabled by a better understanding of SCD.

METHODS Ten subjects with SCD receiving chronic transfusions (either simple or manual partial exchange) from the Johns Hopkins Sickle Cell Center for Adults and ten healthy volunteers with no hematologic abnormalities and HbAA genotype were recruited under an IRB-approved protocol. For subjects with SCD, we took five 90-second videos of blood flow during two sessions, one before and the other after chronic transfusion therapy was complete. For healthy controls, we took ten 90-second videos. On average, each video contained 18 visible capillaries. Complete blood counts were also acquired at each imaging time point. Analysis of videos was performed by manually classifying each observed vessel into one of three categories: fast flowing; slow, intermittent, or partially occluded; or fully occluded based on the qualitative flow speed over time. The number of adhered sickle cells per observed vessel was also counted. T-tests were performed to compare healthy, pre- and post-transfusion data. Finally, since OBM can resolve cell contours by using phase contrast rather than absorption contrast, transparent cells and occlusion mechanisms were observed in detail.

RESULTS Among ten subjects with SCD, 9 had hemoglobin SS disease and 1 had Sβ^o. From before to after a transfusion, SCD patients demonstrated a significant increase in fast flowing vessels (50% before vs. 65% after, p=0.03), decrease in slow or partially occluded vessels (35% before vs. 29% after, p = 0.33), and significant decrease in fully occluded vessels (15% before vs. 6% after, p=0.001). The number of adhered sickle cells per observed vessel decreased from 13.5 to 6.9 cells (p=0.02) from before to after transfusions.

Compared to controls, before transfusion there were differences between the fraction of fast flowing vessels (77% controls vs. 50% pre, p=0.0006), slow vessels (20% controls vs. 35% pre, p=0.005) and fully occluded vessels (3% controls vs. 15% pre, p=0.0002). After transfusion, there was no significant difference in flow classifications between control and SCD vessels (fast flowing: 77% controls vs. 65% post, p=0.07; slow: 20% controls vs. 29% post, p=0.18; occluded: 3% controls vs. 6% post, p=0.07). Lastly, we directly visualized SCD occlusion mechanisms in vivo. In all observed occlusion formations, occlusions were initiated by adherence of a sickled RBC, followed by a buildup of other cells until flow stopped in the vessel. There were no white blood cells involved in occlusion initiation.

CONCLUSIONS Flow rankings and adhered cells in sublingual capillaries indicate that transfusions improve SCD blood flow to match healthy controls, reducing the frequency of occlusions and adhesion. These findings demonstrate the value of OBM in studying SCD. We have shown that in vivo occlusion mechanisms and biomarkers for aberrant SCD blood rheology, such as adhesion, flow dynamics, and occlusions, can be non-invasively observed in SCD patients. In addition to enhancing the understanding of SCD, this technique has the potential to improve the assessment of therapeutic efficacy in clinical trials, evaluation of patient responsiveness to therapy, and discovery of pathologic differences between heterogeneous SCD types for therapeutic target discovery.

Lanzkron:bluebird bio: Membership on an entity's Board of Directors or advisory committees; HRSA: Research Funding; Teva: Current holder of stock options in a privately-held company; Pfizer: Current holder of stock options in a privately-held company; Agios: Membership on an entity's Board of Directors or advisory committees; Takeda: Research Funding; Novo Nordisk: Membership on an entity's Board of Directors or advisory committees; PCORI: Research Funding; Novartis: Consultancy, Research Funding; Pfizer: Consultancy; Merck: Consultancy; Glycomimetics: Consultancy; CSL-Behring: Research Funding. Pecker:Alexion: Research Funding; Novartis: Research Funding; Affimmune: Research Funding; Novo Nordisk: Consultancy. Durr:PlenOptika: Current equity holder in private company; Intuitive Surgical: Research Funding; Olympus Corporation of the Americas: Research Funding.