Biophysical markers of the peripheral vasoconstriction response to pain in sickle cell disease

Abstract

Painful vaso-occlusive crisis (VOC), a complication of sickle cell disease (SCD), occurs when sickled red blood cells obstruct flow in the microvasculature. We postulated that exaggerated sympathetically mediated vasoconstriction, endothelial dysfunction and the synergistic interaction between these two factors act together to reduce microvascular flow, promoting regional vaso-occlusions, setting the stage for VOC. We previously found that SCD subjects had stronger vasoconstriction response to pulses of heat-induced pain compared to controls but the relative degrees to which autonomic dysregulation, peripheral vascular dysfunction and their interaction are present in SCD remain unknown. In the present study, we employed a mathematical model to decompose the total vasoconstriction response to pain into: 1) the neurogenic component, 2) the vascular response to blood pressure, 3) respiratory coupling and 4) neurogenic-vascular interaction. The model allowed us to quantify the contribution of each component to the total vasoconstriction response. The most salient features of the components were extracted to represent biophysical markers of autonomic and vascular impairment in SCD and controls. These markers provide a means of phenotyping severity of disease in sickle-cell anemia that is based more on underlying physiology than on genotype. The marker of the vascular component (BMv) showed stronger contribution to vasoconstriction in SCD than controls (p = 0.0409), suggesting a dominant myogenic response in the SCD subjects as a consequence of endothelial dysfunction. The marker of neurogenic-vascular interaction (BMn-v) revealed that the interaction reinforced vasoconstriction in SCD but produced vasodilatory response in controls (p = 0.0167). This marked difference in BMn-v suggests that it is the most sensitive marker for quantifying combined alterations in autonomic and vascular function in SCD in response to heat-induced pain.

Fig 1. Schematic diagram of the model of peripheral vasoconstriction response to heat-induced pain.

Functional relationships between changes in temperature (ΔTemp), blood pressure (ΔMAP) and respiration (ΔV_T) as inputs and changes in finger blood volume (ΔFBV) as the output. ε_FBV denotes extraneous fluctuations on FBV that are not captured by the model.

Fig 2

Representative data recorded during a test procedure in (a) a non-SCD subject and (b) a SCD subject. The top row shows changes in temperature (ΔTemp, °C). 0°C indicates no heat was delivered. Row 2 shows changes in beat-averaged blood pressure (MAP, mmHg). Row 3 shows changes in tidal volume (ΔV_T, L). The bottom row shows corresponding changes in finger blood volume (ΔFBV, %).

Fig 3. FBV response partitioned into contributions from each model component.

Predicted ΔFBV by each model component (thick lines) plotted on top of the measured ΔFBV (thin lines) during a test procedure in (a) a non-SCD subject and (b) a SCD subject, showing the relative contribution from each component on the total ΔFBV. The bottom row shows the sum of the predicted ΔFBV by all model components (thick line).

Fig 4. Standardized FBV responses to a pain pulse and derived biophysical marker BM_n.

(a) h_THM, standardized FBV responses to a pain pulse (mean ± SEM) in SCD (N = 22, thick line) and non-SCD (N = 23, thin line); (b) biophysical marker BM_n capturing the initial neurogenic vasoconstriction caused by the direct effect of pain. Error bars show mean ± SEM. SCD had slightly stronger vasoconstriction than non-SCD (not statistically significant).

Fig 5. Standardized FBV responses to a blood pressure pulse and derived biophysical marker BM_v.

(a) h_BPC, standardized FBV responses to a blood pressure increase (mean ± SEM) of SCD (N = 22, thick line) and non-SCD (N = 23, thin line); (b) biophysical marker BM_v reflecting vasoconstriction as a result of blood pressure increase. Error bars show mean ± SEM. SCD had significantly stronger vasoconstriction than non-SCD after adjusting for age and sex (p = 0.0409).

Fig 6. FBV responses of the interaction between pain and blood pressure and derived biophysical marker BM_n-v.

(a) FBV_BPCTHM, responses of the interaction between the simulated Temp and MAP pulses (mean ± SEM) in SCD (N = 22, thick line) and non-SCD (N = 23, thin line); (b) biophysical marker BM_n-v reflecting total change in FBV as a result of the interaction between the two inputs. Error bars show mean ± SEM. SCD vasoconstricted while non-SCD vasodilated (p = 0.0167).