Selective Thermal Stimulation Delays the Progression of Vasoconstriction During Body Cooling

Abstract

The objective of this study was to test the feasibility of selective thermal stimulation (STS) as a method to upregulate glabrous skin blood flow. STS is accomplished by mild surface heating along the spinal cord. Four healthy subjects were tested in this study. Each participated in a control experiment and an intervention experiment (STS). Both experiments included establishing a maximum level of vasodilation, considered unique to a subject on a test day, and then cooling to a maximum level of vasoconstriction. Perfusion was measured by a laser Doppler flow probe on the index fingertip. The percent of perfusion in the range of minimum to maximum was the primary outcome variable. The data were fit to a linear mixed effects model to determine if STS had a significant influence on perfusion during whole body cooling. STS had a statistically significant effect on perfusion and increased glabrous skin blood flow by 16.3% (P < 0.001, CI (13.1%, 19.5%)) as skin temperature was decreased. This study supports the theory that STS improves the heat exchanger efficiency of palmar and plantar surfaces by increasing the blood flow.

Introduction

The ability to adjust human core temperature, whether it be natural thermoregulation or medically induced, is imperative to sustain life. Compromised thermoregulation may be caused by Raynaud's phenomenon, iron-deficiency anemia, growth hormone deficiency, and diabetes [1–4]. Compromised thermoregulation can cause poor quality of life, poor health, and in extreme circumstances, even death. Medically induced core temperature changes occur unfavorably during anesthesia and intentionally during therapeutic hypothermia [5,6]. Therapeutic hypothermia provides neuroprotection for cardiac arrest patients, and evidence has shown benefits of therapeutic hypothermia for patients of stroke, traumatic brain injury, and subarachnoid hemorrhage [7]. Ambulatory thermoregulatory mechanism suppression is often treated by pharmaceuticals.

Perioperative hypothermia increases the risk of surgical site infection, increases the risk of postsurgical adverse myocardial events, and significantly increases perioperative blood loss. The most widely used device for establishing or maintaining perioperative normothermia is a forced air circulating blanket [8]. The efficacy of warming is dependent on the surface area coverage [9] and the ability of surface heating to be transmitted to the body core. Noninvasive methods of establishing therapeutic hypothermia are also dependent on surface area coverage, while invasive methods carry risk of infection [10].

A literature and patent search have not revealed any noninvasive medical device that recruits the body's native pathways and mechanisms of thermoregulatory heat transfer to manipulate core temperature. The operative thermoregulatory system includes sensors, a controller, and effectors. The sensors are thermoreceptors distributed in cutaneous tissue, deep abdominal and thoracic tissue, spinal canal, and brain [5]. The sensors located in the deep tissues can be represented by core temperature, whereas the cutaneous thermoreceptors are represented by mean skin surface temperature. These temperatures are weighted to represent whole body temperature. Previous studies have shown the controller favors the core temperature input over the shell [11–13]. The thermoregulatory controller, found in the preoptic anterior hypothalamus, processes the body temperature and elicits a thermoregulatory response as necessary [14,15]. The effector mechanisms include vasoconstriction of glabrous skin arteriovenous anastomoses (AVAs) followed by shivering when cold thermoreceptors are triggered and vasodilation of AVAs followed by sweating when warm thermoreceptors are triggered [16].

We have discovered the ability to influence thermoregulatory control function via simple, safe heating on the skin surface along the spine. The technique has been designated selective thermal stimulation (STS). It may be the solution to effective temperature manipulation with a relatively small footprint and low risk to the subject. STS is a process that works cooperatively with the native function of the thermoregulatory system rather than fighting against it as is the case with surface coverage heating and cooling to alter the core temperature. STS involves heating the skin adjacent to the spinal cord to target the localized thermoreceptors [17]. This intervention elicits the response of removing the vasoconstriction signal to the AVAs in glabrous skin (such as the palms and soles), thereby increasing blood flow to these large heat exchange vessels [16]. A heat sink or source can then be applied to the glabrous skin to alter the temperature of blood before it returns to the core. The glabrous skin then functions as a convective heat exchanger to either add or remove heat from the core using the body's effective natural process of thermoregulation. Invasive animal studies have shown that glabrous skin blood flow is upregulated by means of heating the spinal cord [18–21].

This study tested the hypothesis that we can noninvasively suppress the onset and progression of vasoconstriction by applying STS while dynamically lowering the body temperature.

Methods

Subject Selection.

Exclusion criteria were applied for subjects who met the criteria for contraindications listed in Table 1.

Table 1.

Exclusion criteria for subject selection

Exclusion criteria
< 80 lbs
Known or suspected obstructive disease of the gastrointestinal (GI) tract, including but not limited to diverticulitis, and inflammatory bowel disease
Exhibiting or having a history of disorders or impairment of the gag reflex
Previous GI surgery
Felinization of the esophagus
Might undergo nuclear magnetic resonance/magnetic resonance imaging scanning during the period that the CorTemp™ Core Body Temperature Sensor is within the body
Hypomotility disorders of the GI tract, including but not limited to ileus
Installed cardiac pacemaker or other implanted electromedical device
Impaired circulation
Impaired sensation
Hypersensitivity to cold, such as Raynaud's phenomenon, cold urticaria, cryoglobulinemia, and paroxysmal cold hemoglobulinuria
Diabetes
Angina pectoris or other sever cardiac diseases
Arthritic conditions
Open wound after 48 h
Pheochromocytoma
History of cold illness
Pregnancy
Drug use
Psychological factors
Smoking
Certain medications, like, vasodilators and calcium antagonists, sympathetic and ganglion blocking agents

This study was approved by the Institutional Review Board of The University of Texas at Austin prior to the start of experiments. All subjects provided voluntary written informed consent to participate.

Instrumentation.

Each subject was given a CorTemp™ (HQ, Inc., Palmetto, FL) pill at least 1 day prior to the experiment. They were instructed to take the pill between 2 and 12 h prior to the start of a trial. The sensor was synced to the data recorder the morning of the experiment. Core temperature was recorded every 10 s.

Subjects were clothed in a short sleeved polyester woven shirt and shorts. Subjects then donned a water perfused jacket and pants (Med-Eng, Ottawa, ON). The water perfused garment was connected to a heating and cooling circulating water bath (Cole-Parmer, Vernon Hills, IL). A resistance heater, F030200C8 (Watlow, St. Louis, MO) was placed between the water perfused jacket and the polyester shirt. The resistance heater measured 50.8 cm by 7.6 cm. Insulation was affixed to the resistance heater between the heater and water perfused jacket interface.

Subjects were instrumented with laser Doppler perfusion probe, VP1T (Moor Instruments, Inc., Wilmington, DE) on the index fingertip. The probe was connected to a Moor perfusion probe module, moorVMS-LDF2. Four thermocouples (made in house) were placed on the skin adjacent to the spine. Thermocouples (Omega Engineering, Inc., Norwalk, CT) were also placed in line with the garment tubes at the jacket inlet and outlet and the pants inlet and outlet to measure the net enthalpy change in the circulating water as it moved through the suit. Thermocouples and perfusion probe modules were connected to thermocouple and voltage modules, NI-9213 and NI-9205, respectively (National Instruments, Austin, TX). The modules were in an NI chassis, cDAQ-9178m, which was connected via USB to a desktop, Inspiron 530 (Dell, Round Rock, TX).

Skin temperatures were measured initially using thermocouples. However, due to the local thermal nonuniformity of the water perfused garment owing to the discrete pattern of the flow tubing, and the relative location of the point sensors with respect to a water tube in the garment, the thermocouples did not yield representative results. As skin temperature is an input to body temperature and body temperature is an input to the thermoregulatory controller, it is imperative to collect accurate and precise data in order to determine the body temperature at which thermoregulatory responses occur. So, two-dimensional resistance temperature detectors, described in a future publication, were designed, calibrated, and validated with the capability of measuring the temperature averaged over areas with dimensions as large as 300 cm². These sensors allowed for skin temperature to be calculated from the outlet temperature of the pants. In a secondary set of two experiments, the sensors were placed on subjects before donning the water perfused garment. Water was run through the garment tubes at a water bath temperature ranging from 10 °C to 40 °C. Water was increased and decreased so time constants for cooling and warming could be detected. Mean skin temperature was calculated by a derivation of the Hardy/Dubois 7 mean skin temperature calculation [22]. Equation (1) is the derivation of the formula. The differences include removing a fraction of the 35% allocated to trunk temperature. This fraction is equal to the surface area of the heater divided by the surface area of the body. This fraction is added back but represents the contribution of the skin temperature adjacent to the resistance heater. Also, the 5% of body surface area that is allocated to hands is split equally between hand and dorsal hand due to potential extreme differences in glabrous and nonglabrous skin temperatures. Similarly, the 7% allocated to feet is split equally between plantar and dorsal foot

$$T_{\mathrm{ms}}=0.07\mathrm{*}{T}_{\mathrm{he}}+(0.35-\left(\frac{{\mathrm{SA}}_{\mathrm{rh}}}{{\mathrm{SA}}_{\mathrm{bo}}}\right))\mathrm{*}{T}_{\mathrm{tr}}+0.14\mathrm{*}{T}_{\mathrm{ar}}+0.025\mathrm{*}{T}_{\mathrm{dh}}+0.025\mathrm{*}{T}_{\mathrm{ph}}+0.19\mathrm{*}{T}_{\mathrm{th}}+0.13\mathrm{*}{T}_{\mathrm{le}}+0.035\mathrm{*}{T}_{\mathrm{df}}+0.035\mathrm{*}{T}_{\mathrm{pf}}+\left(\frac{{\mathrm{SA}}_{\mathrm{he}}}{{\mathrm{SA}}_{\mathrm{bo}}}\right)\mathrm{*}{T}_{\mathrm{ba}}$$ (1)

where T_ms is the mean skin temperature (°C), T_he is the head temperature (°C), SA_rh is the resistance heater surface area (cm²), SA_bo is the body surface area calculated by the Mosteller formula (cm²), T_tr is the trunk temperature (°C), T_ar is the arm temperature (°C), T_dh is the dorsal hand temperature (°C), T_ph is the palmar hand temperature (°C), T_th is the thigh temperature (°C), T_le is the leg temperature (°C), T_df is the dorsal foot temperature (°C), T_pf is the plantar foot temperature (°C), and T_ba = temperature of the back adjacent to the spine (°C); this is a constant temperature of 40 °C for the STS experiments after STS has been initiated.

Protocol.

Each subject participated in a control (no STS) and an STS intervention experiment while subjected to identical whole body cooling protocols. The order in which intervention or control was applied was randomized by order of availability. The first two subjects that were available participated in a control experiment first followed by an STS experiment. The last two subjects participated in experiments in the opposite order. At least 7 days separated the control and the STS trials for each subject. The experiments were conducted in a temperature-controlled room (22 °C).

Each subject was positioned in the supine position for the duration of the experiment while wearing the full body water perfused garment that extended to the neck, wrists, and ankles. Water was circulated through the garment from a regulated temperature bath with an immersion pump.

The instrumentation was used to collect body temperature and perfusion data while subjects transitioned from a state of being fully vasodilated to fully vasoconstricted. The protocol for the water bath temperature is shown in Fig. 1. The first 5 min of the experiment served as a brief acclimation period. Data were collected while the water bath was isolated from the garment. The baseline state of thermoregulation (vasoconstricted, thermally neutral, or vasodilated), represented by glabrous skin blood flow, relates to the internal energy stored in the body as well as a multiplicity of environmental factors, history of physical activity, a subject's unique pattern of thermoregulatory function, and the time of day in the circadian cycle [23]. In order to eliminate the variations in energy storage due to time of day, control and STS experiments were run at early to midmorning. Daily variations in internal energy storage are due to the weather, recent eating history, and exercise regimen. The body may hold internal energy stores in excess of what is required for maintenance of a thermoneutral state. The baseline internal energy affects how much heat may be added to the system before eliciting a sweating response. Sweating was a desirable behavior neither in the context of this study nor one of interest. Because of these various factors that can influence the outcome, all subjects had to be brought to a common state of thermoregulation prior to initiation of a stress protocol. For this purpose, following the baseline period, 20 °C water was circulated through the garment for 30 min or until full vasoconstriction was observed for 5 min defined by a laser Doppler velocimeter reading <100 PU, indicating that excess internal energy was removed. If vasoconstriction did not occur, the water bath temperature was decreased to 15 °C for 30 additional minutes or until full vasoconstriction was observed for 5 min. Again, if vasoconstriction did not occur, water bath temperature was further decreased 10 °C until full vasoconstriction was observed for 5 min. In all experiments, vasoconstriction was observed at or by 10 °C and no further decrease was necessary.

Fig. 1.

Water bath temperature profile for the duration of the experiments. The dashed line indicates when the STS heater would be energized for an STS intervention experiments. During the first 5 min, room temperature is shown, as the water bath is isolated from the subject.

After the initial period of vasoconstriction, subjects were warmed as the water bath was adjusted to 40 °C and held for 30 min to allow subjects to reach their personal maximum level of vasodilation on the trial day. If the experiment included STS, after the initial 15 min of warming a resistance heater was set to hold a temperature of 40 °C. The heater remained energized for the duration of the experiment. Following 30 total minutes of warming, water bath temperature was decreased at 1 °C/min until a state of maximum vasoconstriction was observed for 5 min.

Measurements

Figure 2 shows the results of a linear fit of water perfused pants outlet temperature and mean skin temperature. Figure 2(a) shows the result of the fit using the experimental data. This fit will be used to calculate mean skin temperature during the control experiments. Figure 2(b) shows the same except that back temperature is replaced by a constant 40 °C. This fit will be used to calculate mean skin temperature during the STS experiments. Due to the strong relationship between water perfused pants outlet temperature and mean skin temperature, mean skin temperature is back calculated for the primary set of experiments using inline water perfused pants outlet temperature.

Fig. 2.

Skin temperature is fit to the outlet temperature of the water perfused pants. (a) Skin temperature is calculated with T_ba derived from a two-dimensional resistance temperature detectors. This represents the relationship between outlet pants temperature and skin temperature during a control experiment. (b) Skin temperature is calculated with T_ba = 40 °C. This represents the relationship between outlet pants temperature and skin temperature during an STS experiment.

A LabView program (National Instruments) was created to collect raw data from the voltage and thermocouple modules at a sampling rate of 10 Hz. The same program passed the maximum temperature from the thermocouples adjacent to the spine to a PID controller adjusted to weighting factors of 2, 0.7, 0, respectively. The output from the proportional-integral-derivative controller (PID) was digitally wired to an NI voltage output module, USB-6009. The analog output from the device controlled the AC side of a solid state relay, CWD2410S (Crydom, San Diego, CA) that was connected in series to a variable autotransformer, SC-20M (PHC Enterprise, Inc., Torrance, CA) and the resistance heater.

Data Analysis

A Python code was used to extract the start time and duration of the experiment from the LabVIEW LVM file [24]. The start time and duration were used to extract the time and core temperature data for the experimental period from the core temperature CSV file. The extracted data were saved as time and temperature text files.

All data were processed in matlab (version R2017b, MathWorks, Natick, MA). Perfusion noise was suppressed with a 5 min running median filter. The period of vasoconstriction was found by running a slope analysis on the perfusion vector. The start of vasoconstriction was found by either the time at which the rate of perfusion decreased to less than −1 PU/min (arbitrarily chosen) or the commencement of water bath cooling, whichever occurs last. The end of the vasoconstriction period was defined as the time when the slope of perfusion increases above −2 PU/min (arbitrarily chosen).

Because hypothesis of the study is that STS inhibits the vasoconstriction response causing higher perfusion to be maintained at lower body temperatures during continuous cooling, when comparing body temperatures at fully vasodilated or fully vasoconstricted, a one-tailed t-test is applied to the data.

Perfusion data were extracted and presented as the percentage of perfusion between fully vasodilated and fully vasoconstricted at 2 min intervals while the glabrous skin is actively vasoconstricting. The extracted data additionally included subject number, intervention (control or STS), relative time, skin temperature, and core temperature.

Subject number and treatment were converted to categorical variables, and the data were fit to a linear mixed effects model. The outcome variable is percent of perfusion between fully vasoconstricted and fully vasodilated. The perfusion at which vasoconstriction begins is considered fully vasodilated for subject trial and set to 100%. The perfusion at which active vasoconstriction ceases is considered fully vasoconstricted for subject trial and set to 0%. This normalization defined a full range of vasoactivity for each subject trial day to provide a basis for comparing responses among the individual data sets. The fixed effects were intervention, time, skin temperature, core temperature, and a skin temperature and time interaction term. The random effect was subject.

The final model was determined by eliminating variables from the full model one at a time and comparing the nested model and the larger model with a likelihood ratio test. The likelihood ratio test tested the null hypothesis that the models return the same results. The reduced model remained if the null was accepted. The full model was reduced by the removing the random effect to determine if it was significant. After the random effect was tested with the likelihood ratio test, each insignificant fixed effect was removed individually and compared with the model it was removed from.

Results

The experiments were performed on four healthy nonsmokers (three male) with an average age (±SD) of 19.5 ± 0.6 years and body mass index (±SD) of 22.3 ± 4.3 kg/m².

Ambient temperature (±SD) was maintained at 22.2 ± 0.2 °C. The average maximum perfusion value (±SD) within the first 15 min of warming was 313 ± 122 PU and 350 ± 107 PU for the control and STS experiments, respectively. These were not significantly different (P = 0.73). This indicates that a common thermal state was achieved prior to the intervention.

Figure 3 shows the skin temperature and percent of perfusion from fully vasoconstricted to fully vasodilated. For each subject, STS caused an elevated perfusion during vasoconstriction due to decreasing skin temperature. The skin temperatures at the onset of vasoconstriction (fully vasodilated) (±SD) were 34.7 ± 0.1 °C and 33.5 ± 2.2 °C for control and STS experiments, respectively. These initial skin temperatures are not significantly different (P = 0.18).

Fig. 3.

Perfusion is plotted as a function of skin temperature during active vasoconstriction. 0% perfusion represents minimum perfusion for that subject on that day and 100% perfusion represents maximum perfusion for that subject on that day. Each subject is represented by the same symbol for their control and STS experiment.

Figure 4 shows the difference in skin temperatures for control and STS for each subject when the fully vasoconstricted state was reached after cooling the skin surface. The average final skin temperatures (±SD) were 29.4 ± 2.0 °C and 26.3 ± 1.7 °C for control and STS experiments, respectively. In all cases, STS caused perfusion to remain at an elevated active state to a significantly lower skin temperature during progressive cooling (P = 0.005). Figure 4 also shows the levels of elevated perfusion during the STS experiment when each subject's mean skin temperature was equivalent to the final skin temperature reached at full vasoconstriction in the control experiment. At equivalent mean skin temperatures, the perfusions (±SD) for the control experiment and STS experiment were 0 ± 0% (0 denoting maximum vasoconstriction) and 23.1 ± 14.5%, respectively. In other words, at a skin temperature that caused full vasoconstriction in the absence of STS, perfusion remained elevated with STS (P = 0.02). The fact that STS maintains a significant AVA blood flow under conditions when it would normally be fully shut down means that STS has created an added capacity for convection of heat between the body core and surface. This additional capacity can be recruited for therapeutic benefit to add heat to the core under conditions for which warming is needed.

Fig. 4.

Difference in skin temperature between the control and STS experiments at the final maximum vasoconstriction for each subject is shown in blue (control—STS). Yellow bars show the remaining active perfusion (as percent of the difference between minimum and maximum values) during the STS experiment when the mean skin temperature during cooling reaches the value of the skin temperature at maximum vasoconstriction for the control experiment.

The first linear mixed effects model used to fit the percent of perfusion data is shown in Eq. (2) in Wilkinson notation. The time and skin temperature interaction term was chosen because, as shown in Fig. 3, skin temperature appears to have a larger effect on perfusion as time (and vasoconstriction) progresses

$$\mathrm{perfusion}\hspace{0.17em}\sim 1+\mathrm{time}\mathrm{*}\mathrm{skin}\hspace{0.17em}\mathrm{temperature}+\mathrm{core}\hspace{0.17em}\mathrm{temperature}+\mathrm{intervention}+\left(1\hspace{0.17em}\right|\hspace{0.17em}\mathrm{subject})$$ (2)

Second, we excluded the subject to determine if the random effect was necessary. The model, Eq. (3), was shown to produce significantly different results by the likelihood ratio test that compared Eq. (3) to Eq. (2) (χ²(1) = 63, P < 0.001), indicating that the random effect is necessary.

$$\mathrm{perfusion}\hspace{0.17em}\sim 1+\mathrm{time}\mathrm{*}\mathrm{skin}\hspace{0.17em}\mathrm{temperature}+\mathrm{core}\hspace{0.17em}\mathrm{temperature}+\mathrm{intervention}$$ (3)

Core temperature was not a significant variable from Eq. (2) (P = 0.66). This indicates that core temperature is not a significant input into the thermoregulatory controller when skin temperature is being regulated independently with the core temperature floating, as was the case for this study. Core temperature was excluded for the third model

$$\mathrm{perfusion}\hspace{0.17em}\sim 1+\mathrm{time}\mathrm{*}\mathrm{skin}\hspace{0.17em}\mathrm{temperature}+\mathrm{intervention}+\left(1\hspace{0.17em}\right|\hspace{0.17em}\mathrm{subject})$$ (4)

Equation (4) was compared to Eq. (2) with a likelihood ratio test to test the null hypothesis that the models would yield similar results. The null was accepted (χ²(1) = 0.18, P = 0.67). Equation (3) yielded an insignificant intercept, so it was reduced to Eq. (4)

$$\mathrm{perfusion}\hspace{0.17em}\sim \hspace{0.17em}\mathrm{time}\mathrm{*}\mathrm{skin}\hspace{0.17em}\mathrm{temperature}+\mathrm{intervention}+\left(1\hspace{0.17em}\right|\hspace{0.17em}\mathrm{subject})$$ (5)

The models from Eqs. (4) and (5) were compared with a likelihood ratio model which accepted the null hypothesis (χ²(1) = 1.3, P = 0.25). The model from Eq. (5) yielded all significant terms.

Parameter estimates are shown in Table 2. In the absence of other parameters, STS increases perfusion by 16.3% during cooling of the body. The random effect is strong, but there is considerable variation between subjects.

Table 2.

Parameter estimates, upper and lower 95% confidence limits, and probability values for fixed effects: intervention (STS), time, skin temperature, and time–skin temperature interaction term and standard deviation and upper and lower confidence limits of the intercept of the random effect, subject

Parameter	Estimate	Lower-95	Upper-95	P-value
Intervention	16.3	13.1	19.5	<0.001
Time (s)	0.10	0.08	0.12	0.07
Skin temp (°C)	2.6	2.4	2.9	<0.001
Time: skin temp (s: °C)	−5.5 × 10−3	−6.4 × 10−3	−4.7 × 10−3	<0.001
Subject (SD)	8.4	4.1	17.2

The model prediction plotted against skin temperature is shown in Fig. 5 along with the experimental data. The goodness of fit (R²) for the model is 0.95.

Fig. 5.

Experimental data and model predictions for each subject during the control and STS experiments are plotted against mean skin temperature

The residuals of Eq. (4) model are shown in Fig. 6. The normal distribution supports the validity of the model.

Fig. 6.

Histogram of the residuals of the final linear mixed effects model shows a normal distribution

Discussion

The goal of this study was to investigate the thermoregulatory response to STS. The data show that there was a suppression of active vasoconstriction as skin temperature was decreased, meaning that perfusion is relatively elevated at lower body temperatures, thereby enabling convective heat exchange with the body core under conditions for which it would not normally be possible. The data indicate that the targeted thermoreceptors along the spinal cord during STS sent a signal to the hypothalamus indicating the deep tissue was warm. The warm signal was added to the cold signal sent from the cutaneous thermoreceptors, and the thermoregulatory response was a reduced vasoconstriction effect.

The statistical analysis showed that core temperature did not have a significant input to the controller during active vasoconstriction when skin temperature is controlled independently. We suspect that the skin temperature manipulation is overwhelming the thermoregulatory controller inputs and causing core temperature to react as a response variable. This interpretation will need to be investigated further.

Potential clinical applications for STS include maintaining perioperative normothermia. After glabrous skin blood flow is upregulated with STS, heaters would be applied to the glabrous skin and the increased volume of blood flowing to the periphery would warm and recirculate back to the core, thus increasing core temperature. Another potential application is establishing therapeutic hypothermia for patients who have suffered cardiac arrest, traumatic brain injury, or stroke. Conversely to perioperative normothermia, coolers would be applied to the glabrous skin on the hands and feet after STS upregulated blood flow. Warm blood from the core would circulate to the periphery where it would lose heat to the coolers. That colder blood would recirculate back to the core, thus reducing the temperature.

A limitation of the study described includes the sample population being restricted to healthy young adults with no known medical conditions. In a clinical setting, there is a contrasting population that has endured some insult or injury to their body. Each condition would have to be investigated to determine the efficacy of STS with some known functional deficiency.

Funding Data

• Mercury Biomed, LLC (Cleveland, OH) (Grant No. 2R42GM119871-02).