Implications for burn shock resuscitation of a new in vivohuman vascular microdosing technique (microdialysis) fordermal administration of noradrenaline

Anders Samuelsson a,e, Simon Farnebo c,d, Beatrice Magnusson b,d, Chris Anderson b,d,Erik Tesselaar a,d, Erik Zettersten d, Folke Sjoberg a,c,d,*aDepartment of Anaesthesia and Intensive care, County Council of Ostergotland, Linkoping, SwedenbDepartment of Dermatology, County Council of Ostergotland, Linkoping, SwedencDepartment of Hand and Plastic Surgery and Burns, County Council of Ostergotland, Linkoping, SwedendDepartment of Clinical and Experimental Medicine, Faculty of Health Sciences, Linkoping University, 581 85 Linkoping, SwedeneDepartment of Medicine and Health Sciences, Division of Drug Research/Anesthesiology, Faculty of Health Sciences, Linkoping University,

581 85 Linkoping, Sweden

b u r n s 3 8 ( 2 0 1 2 ) 9 7 5 – 9 8 3

a r t i c l e i n f o

Article history:

Accepted 26 May 2012

Keywords:

Burn resuscitation

Burn shock

Glucose

Glucose homeostasis

Insuline resistance

Lactate

Microdialysis

Puruvate

Skin

Tissue blood flow

Tissue ischemia

Wound healing

a b s t r a c t

Introduction: Skin has a large dynamic capacity for alterations in blood flow, and is therefore

often used for recruitment of blood during states of hypoperfusion such as during burn

shock resuscitation. However, little is known about the blood flow and metabolic conse-

quences seen in the dermis secondary to the use vasoactive drugs (i.e. noradrenaline) for

circulatory support. The aims of this study were therefore: to develop an in vivo, human

microdosing model based on dermal microdialysis; and in this model to investigate effects

on blood flow and metabolism by local application of noradrenaline and nitroglycerin by the

microdialysis system simulating drug induced circulatory support.

Method: Nine healthy volunteers had microdialysis catheters placed intradermally in the

volar surface of the lower arm. The catheters were perfused with noradrenaline 3 or

30 mmol/L and after an equilibrium period all catheters were perfused with nitroglycerine

(2.2 mmol/L). Dermal blood flow was measured by the urea clearance technique and by laser

Doppler imaging. Simultaneously changes in dermal glucose, lactate, and pyruvate con-

centrations were recorded.

Results: Noradrenaline and nitroglycerine delivered to the dermis by the microdialysis

probes induced large time- and dose-dependent changes in all variables. We particularly

noted that tissue glucose concentrations responded rapidly to hypoperfusion but remained

higher than zero. Furthermore, vasoconstriction remained after the noradrenaline admin-

istration implicating vasospasm and an attenuated dermal autoregulatory capacity. The

changes in glucose and lactate by vasoconstriction (noradrenaline) remained until vasodi-

latation was actively induced by nitroglycerine.

Available online at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/burns

* Corresponding author at: Department of Clinical and Experimental Medicine, Faculty of Health Sciences, Linkoping University, 581 85Linkoping, Sweden. Tel.: +46 70 5571820.

E-mail address: folke.sjoberg@liu.se (F. Sjoberg).URL: http://www.hu.liu.se/ike/forskning/brannskador/sjoberg-folke?l=sv

0305-4179/$36.00 # 2012 Elsevier Ltd and ISBI. All rights reserved.http://dx.doi.org/10.1016/j.burns.2012.05.012

Conclusion: These findings, i.e., compromised dermal blood flow and metabolism are

particularly interesting from the burn shock resuscitation perspective where noradrena-

line is commonly used for circulatory support. The importance and clinical value of the

results obtained in this in vivo dermal model in healthy volunteers needs to be further

explored in burn-injured patients.

# 2012 Elsevier Ltd and ISBI. All rights reserved.

b u r n s 3 8 ( 2 0 1 2 ) 9 7 5 – 9 8 3976

1. Introduction

In burn injury, the skin is often the most important target for

the injury and maintaining skin homeostasis is very important

not to compromise the skin further, leading to larger and

deeper burns. In burn shock resuscitation, systemic vascular

resistance is often low and in order to increase blood pressure

and cardiac output, inotropes and most commonly vasocon-

strictor drugs are often applied. In theory such use may

compromise skin blood flow and indirectly its metabolism, as

the skin is a major tissue having a vasculature densely

innervated by alpha-adrenergic receptors. Such drug effects

may thus jeopardize the viability of the skin and most

importantly in the burn injury perspective, the dermis. This

may then in theory lead to an increased burn wound.

In burns it is accepted that the skin, and particularly the

wound, is the ‘‘motor’’ of various inflammatory, metabolic,

and circulatory changes [1,2]. We have shown in a previous

study, in which we used microdialysis in both the injured and

uninjured dermis of burned patients, that there is local

acidosis and a persistently attenuated autoregulation of blood

flow during fluid resuscitation. These changes were in parallel

to an altered local glucose homeostasis, possibly due to

cytopathic hypoxia [3,4]. Such changes could be further

aggravated when using vasoconstrictive drugs both under

normal conditions and in the intensive care unit. Despite that

these adrenergic drugs may be assumed to affect both blood

flow and skin tissue metabolism, the effects on dermal blood

flow and metabolism of commonly used vasoactive drugs are

largely unknown. Noradrenaline is most often used and

several studies have indicated that it has direct effects on cell

metabolism such as the inhibition of energy metabolism in

human mononuclear cells [5], depressed hepatocellular

function, and induced production of superoxide radicals with

impairment of mitochondrial respiration [6].

The aim of this study was to develop a human in vivo

dermal model in healthy volunteers for investigations of local

blood flow and metabolic effects (changes in dermal glucose,

lactate, and pyruvate concentrations assessed by microdia-

lysis) induced by adrenergic vasoconstriction (noradrenaline

given locally through the microdialysis catheter) followed by

nitrous oxide (NO)-mediated dilatation by nitroglycerine (also

provided by the microdialysis catheter). The microdialysis

system was also used to assess dermal blood flow (retro-

dialysis of urea [7] and ethanol [8]). Changes in blood flow in

the dermis were also measured by laser Doppler imaging.

Our underlying hypotheses were that increasing dermal

vasoconstriction might be accomplished by dermal infusion of

therapeutic concentrations of noradrenaline through the

microdialysis system. This leads to reduced dermal blood

flow and ischemia with concomitant metabolic consequences.

The vasoconstrictive effects on both flow and metabolism may

be restored by NO-mediated vasodilatation if not preceded by

a vascular autoregulatory escape induced by the ischemia

itself. Theoretically the established doses of noradrenaline

that were suggested by previous investigators seemed large

[9], so a tenth of that dose was also studied. Importantly, and

to validate the actual dose of the drug being delivered to the

tissue, we also measured the amount of noradrenaline

retained in the catheter system.

2. Subjects and methods

2.1. Subjects

Nine healthy, non-smoking volunteers (6 men and 3 women),

mean (SD) age of 28 (6) years, participated in the study after

giving informed consent.

Exclusion criteria were dermatological problems, allergies,

cardiovascular disease, or taking prescribed drugs. Subjects

were asked to refrain from drinking any substance containing

caffeine after midnight of prior to the day of the experiment.

The subjects were sitting with the arms at the level of the heart

throughout the experiments, which were conducted in an air-

conditioned room at 22–23 8C. The subjects acclimatized for

30 min before the experiments started. The study design

conformed to the declaration of Helsinki and was approved by

the Regional Ethics Committee for Human Research at

Linkoping University, Linkoping, Sweden.

2.2. Microdialysis technique

The microdialysis system consisted of CMA/107 pumps and

CMA/70 microdialysis catheters with a shaft length of 60 mm

and a membrane length of 10 mm (CMA, Microdialysis AB,

Stockholm, Sweden). The membrane of the probe had an outer

diameter of 0.6 mm and a molecular cut-off of 20 kDa. The

perfusion fluid (perfusate) was sterile Ringer acetate. Urea and

ethanol were added to the perfusate as blood flow markers

(APL, Umea, Sweden). Before each catheter was inserted, the

skin was cleaned with chlorhexidine ethanol (5 mg/mL,

Fresenius Kabi, Uppsala, Sweden) and anesthetized with a

0.2 mL intradermal injection of lidocain (10 mg/mL, AstraZe-

neca AB, Sweden).

In four subjects, two microdialysis catheters and in the

other five subjects, three catheters were inserted. The

catheters were inserted 3 cm apart in the ventral dermis of

the forearm. A venous cannula (Venflon Pro 1.2 mm � 32 mm,

Becton Dickinson AB, Helsingborg, Sweden) was used as an

introducer, and was inserted intradermally. The catheter was

then inserted through the guide, after which the guide was

Ta

ble

1–

Sch

em

ati

co

verv

iew

of

the

pro

toco

l.N

rep

rese

nts

nu

mb

er

of

cath

ete

rs.

Inse

rtio

nre

cov

ery

(A)

No

rad

ren

ali

ne

(B)

Reco

very

(C)

Nit

rogl

yce

rin

(D)

Reco

very

(E)

Tim

e90

min

60

min

60

min

60

min

20

min

Perf

usa

tegro

up

1

(N=

7)

Rin

ger’

sa

ceta

teR

inger’

sa

ceta

te+

no

rad

ren

ali

ne

30

mm

ol/

LR

inger’

sa

ceta

teR

inger’

sa

ceta

te+

nit

rogly

ceri

n

2.2

mm

ol/

L

Rin

ger’

sa

ceta

te

Perf

usa

tegro

up

2

(N=

9)

Rin

ger’

sa

ceta

te,

ure

a,

eth

an

ol

Rin

ger’

sa

ceta

te,

ure

a,

eth

an

ol

+n

ora

dre

na

lin

e30

mm

ol/

L

Rin

ger’

sa

ceta

te,

ure

a,

eth

an

ol

+

no

rad

ren

ali

ne

30

mm

ol/

L

Rin

ger’

sa

ceta

te,

ure

a,

eth

an

ol

+n

itro

gly

ceri

n2.2

mm

ol/

L

Rin

ger’

sa

ceta

te,

ure

a,

eth

an

ol

Perf

usa

tegro

up

3

(N=

7)

Rin

ger’

sa

ceta

te,

ure

a,

eth

an

ol

Rin

ger’

sa

ceta

te,

ure

a,

eth

an

ol

+n

ora

dre

na

lin

e30

mm

ol/

L

Rin

ger’

sa

ceta

te,

ure

a,

eth

an

ol

+

no

rad

ren

ali

ne

30

mm

ol/

L

Rin

ger’

sa

ceta

te,

ure

a,

eth

an

ol

+n

itro

gly

ceri

n2.2

mm

ol/

L

Rin

ger’

sa

ceta

te,

ure

a,

eth

an

ol

Dia

lysa

tea

na

lysi

sG

luco

se,

lact

ate

,p

yru

va

te,

ure

a,

eth

an

ol,

no

rad

ren

ali

ne

Blo

od

perf

usi

on

aLa

ser

Do

pp

ler

perf

usi

on

ima

gin

g(1

sca

n/m

in)

aB

loo

dp

erf

usi

on

(LD

PI)

wa

so

nly

mea

sure

din

gro

up

1.

b u r n s 3 8 ( 2 0 1 2 ) 9 7 5 – 9 8 3 977

withdrawn. The intradermal position of the catheters was

confirmed by ultrasound measurements (Dermascan A,

Sonotron AB, Sweden), mean depth (SD) 0.78 (0.23) mm,

consistent with several other investigations [10–12].

2.3. Study design

The drug delivery protocol consisted of 5 phases (A–E), lasting

290 min in total. Throughout the protocol, the perfusate flow

was set to 2 mL/min and dialysate samples were collected into

prelabeled and capped microvials at 10-min intervals. The

protocol is visualized in Table 1.

2.3.1. Phase A (90 min)The microdialysis catheters (n = 23) were allowed to equili-

brate for 90 min, while being perfused with Ringer’s acetate. In

16 catheters, urea and ethanol were added for blood flow

measurements.

2.3.2. Phase B (60 min)In 16 catheters, noradrenaline (30 mmol/L) was added to the

perfusate. In 9 of these, urea and ethanol were also added to

the perfusate for blood flow measurements. Seven catheters

were perfused with noradrenaline (3 mmol/L) in Ringer’s

acetate and with the addition of urea and ethanol. A 5-min

flush was applied to clear the system for air bubbles. This was

followed by 55 min of drug delivery.

2.3.3. Phase C (60 min)The perfusate was changed to Ringer’s acetate, without the

addition of vasoactive drugs. In 16 catheters, urea and ethanol

were added for blood flow measurements. A 5-min flush was

applied to clear the system for air bubbles. This was followed

by a 55-min recovery period.

2.3.4. Phase D (60 min)The perfusate was changed to nitroglycerine (NTG, 2.2 mmol/

L), AstraZeneca AB, Sweden). In 16 catheters, urea and ethanol

were added for blood flow measurements and catheters were

flushed for 5 min, followed by an additional 55 min of NTG

infusion.

2.3.5. Phase E (20 min)The perfusate was changed to Ringer’s acetate, without the

addition of vasoactive drugs. In 16 catheters, urea and ethanol

were added for blood flow measurements. A 5-min flush was

applied to clear the system for air bubbles. This was followed

by a 15-min recovery period.

2.4. Laser Doppler perfusion imaging

A laser Doppler perfusion imaging technique (PIM II, LISCA

Development AB, Linkoping, Sweden) was used to measure

blood flow in the area surrounding 7 catheters perfused with

noradrenaline (30 mmol/L), and throughout the experiment.

The laser Doppler system contains a low power He–Ne laser

(1 mW, 632 nm), in which the beam was moved by a step

motor device, which provided the scanning procedure over the

surface of the skin. Doppler shifts in the backscattered light

were detected and processed to generate an output signal,

b u r n s 3 8 ( 2 0 1 2 ) 9 7 5 – 9 8 3978

which is linearly proportional to tissue perfusion by blood in

the upper 200–300 mm of the skin.

The head of the scanner was positioned 16 cm above the

skin surface and set to scan an area 3 cm � 3 cm at each

experimental site and on each occasion. Each image format

consisted of 64 � 64 measurement sites (medium resolution,

high scan speed) with a distance of about 1 mm between each

measurement point. The approximate time required for such

an image to be recorded was about 1 min, and images were

made with 5-min intervals.

The skin area overlying the microdialysis catheters was

scanned. Data were analyzed using the manufacturer’s

software (LDPIWin version 2.3, Patch Test Analysis 1.3). The

mean perfusion was calculated within a selected region of

interest of 1.0 cm � 0.5 cm corresponding to the area of skin

around the tip of the catheter.

The biological zero signal was recorded at the end of the

experiment by a temporary occlusion (2 min) of the arterial

circulation to the limb by a blood pressure cuff, and was

subtracted from the perfusion values.

2.5. Analysis of metabolites

The contents of the microvials were directly analyzed for

glucose, lactate, pyruvate and urea with a CMA 600 Micro-

dialysis analyzer (CMA Microdialysis AB, Solna, Sweden) using

enzymatic reagents and colorimetric assays [13]. Urea con-

centrations were calculated from the rate of breakdown of

nicotinamide adenine dinucleotide (NADH). Concentrations of

glucose, lactate, and pyruvate were assayed by glucose-, L-

lactate-, and pyruvate-oxidase techniques, respectively.

Reagents were obtained from CMA Microdialysis AB (Solna,

Sweden). The analyzer has an imprecision of <3% for urea,

<5% for glucose, <6% for lactate and pyruvate. The linearity is

>95% for urea, glucose, lactate and pyruvate.

2.6. Assay of noradrenaline

A high performance liquid chromatography system consisting

of a P680 HPLC pump with automated sample injector ASI-100

(Dionex GMBH, Idstein, Germany) and an electrochemical

detector (DECADE, Antec Leyden, Zoeterwoude, The

Netherlands) were used. The analytical column was an

Aquasil C18 250 mm � 4.6 mm, particle size 5 mm, with a

preceding matched guard column Aquasil C18

10 mm � 4 mm � 5 mm (Keystone Scientific, Bellefonte, PA,

USA). The temperature of the column was set at 23 8C with an

integrated oven (Dionex GMBH, Idstein, Germany).

The mobile phase consisted of sodium 1-heptane-sulpho-

nate (1 mmol), citric acid monohydrate (0.1 M), disodium-

EDTA (0.05 mmol), and 5% acetonitrile; pH was adjusted to 2.7

with sodium hydroxide (1 M) before the acetonitrile was

added. The flow rate was set at 1.0 mL/min, the runtime was

set at 15 min, and the detector at +750 mV (nA range) against

the silver/silver chloride reference electrode. The volume of

injection was 10 mL for both standards and samples.

Chromatograms were measured using Chromeleon soft-

ware from Dionex GMBH. Quantitation was achieved by

comparison of peak area generated from the standard curve.

The detection limit was 0.3 mmol/L.

2.7. Assay of ethanol concentration

The alcohol dehydrogenase method was used to measure the

ethanol concentration, as it is optimized for min samples and

low concentrations of ethanol. The concentration of the

reaction product NADH is proportional to the concentrations

of ethanol in the calibrators and samples. Absorbance of

NADH was measured at a wavelength of 334 nm in 96-hole

microtitre plates in a spectrophotometer (Mullikan1 Spec-

trum, Thermo Labsystems, Vantaa, Finland), and a nonlinear

calibration curve was used to evaluate the data. The lowest

concentrations detected as significantly different from the

blanks in 20 mL samples was 0.025 mmol/L. Coefficients of

variation within or between assays were 4.1% and 6.4%,

respectively, in the measurement range of these samples.

2.8. Data analysis

Four probes were excluded from the analysis because they

were damaged or because they did not reach a stable

equilibrium at the end of Phase A.

Data consisted of concentrations in the dialysate of

ethanol, urea, glucose, lactate, and noradrenaline collected

at 10-min intervals during phases B–E. A total of 460 dialysate

samples were collected and analyzed (20 for each catheter).

For glucose, 21 data points were missing and for lactate, 14

data points were missing due to problems with the analysis.

For pyruvate, 143 data points were missing due to problems

with the assay, which made it impossible to gain meaningful

statistics. Therefore, pyruvate was not further analyzed. To

reduce the anticipated intersubject differences, data were

normalized by subtracting the concentrations at baseline

(mean of the measured values at the final 20 min of phase A).

Two-way repeated measures analysis of variance (ANOVA)

were used to test the effects of time (phases B and D) and

concentration of NA (phase B) on the changes in dialysate

concentrations or changes in perfusion.

Glucose and lactate were correlated with perfusion and

urea measurements and correlations were analyzed using

Pearson’s correlation coefficient. Only the data points during

the period when NA or NTG were delivered (phases B and D)

were used for the correlation analyses.

Data are presented as mean � SEM. Probabilities of less

than 0.05 were accepted as significant. All statistical analyses

were made with the aid of GraphPad Prism version 5.02 for

Windows (GraphPad Software, San Diego California USA,

www.graphpad.com).

3. Results

There was an unacceptable variation in ethanol concentra-

tions, which did not correlate with other changes. Therefore,

we did not further analyze ethanol data.

3.1. Glucose

The concentration of glucose in the dialysate decreased

notably, already after 20 min of the NA infusion (Fig. 1). When

NA infusion was stopped (phase C), glucose concentrations

180120600

0.0

0.5

1.0

time (minutes)

chan

ge

in [

lact

ate]

(m

mol/

L)

NA NTG

3 mm ol/ L NA

30 mm ol/ L NA

Fig. 2 – Mean (SEM) change in the concentration of lactate in

the dialysate during perfusion of the microdialysis

catheters with NA and NTG. For the catheters perfused

with a high concentration of NA, the increase in the lactate

concentration was larger and the return to baseline was

delayed compared with the catheters perfused with a low

concentration of NA ( p < 0.001).

180120600-3

-2

-1

0

1

2

time (minutes)

chan

ge

in [

glu

:lac

] (m

mol/

L)

NA NTG

3 mm ol/ L NA

30 mm ol/ L NA

Fig. 3 – Mean (SEM) change in the glucose:lactate ratio in

the dialysate during perfusion of the microdialysis

catheters with NA and NTG.

180120600-1.0

-0.5

0.0

0.5

1.0

time (minutes)

chan

ge

in [

glu

cose

] (m

mol/

L)

NA

30 mm ol/ L NA

3 mm ol/ L NA

NTG

Fig. 1 – Mean (SEM) change in glucose concentration in the

dialysate during perfusion of the microdialysis catheters

with NA and NTG ( p < 0.0001). The increase in glucose

concentration during infusion of NTG was later for the

catheters perfused with the higher concentration of NA

(ANOVA, p = 0.001).

b u r n s 3 8 ( 2 0 1 2 ) 9 7 5 – 9 8 3 979

continued to be decreased and it did not recover until NTG was

administered (phase D). NTG resulted in a rapid and large

increase in the concentration of glucose in the dialysate. This

increase was delayed for the catheters perfused with the high

concentration of NA (ANOVA, p = 0.001). During recovery

phase (phase E), the glucose concentration did not return to

baseline for the catheters that were perfused with a high

concentration of NA. For the catheters perfused with a low

concentration of NA, glucose concentrations in the dialysate

decreased rapidly during phase E.

3.2. Lactate

Lactate concentrations in the dialysate increased when the

catheters were perfused with NA (Fig. 2). This increase

continued for 30 min into phase C, followed by a plateau

(with 30 mmol/L NA) or decrease (with 3 mmol/L NA). During

perfusion of the catheters with NTG, lactate concentrations

decreased steadily and returned to baseline after the

perfusion with NTG was stopped (phase E). With the

catheters perfused with a high concentration of NA, lactate

increased again after the NTG infusion was stopped, at the

end of phase E.

For the catheters perfused with a high concentration of NA,

the increase in the lactate concentration was larger and the

return to baseline was delayed compared with the catheters

perfused with a low concentration of NA ( p < 0.001).

3.3. Glucose:lactate

After 20 min, the glucose:lactate ratio decreased and remained

lower than baseline during the recovery phase (C). The size of

the decrease was not dependent on the NA concentration

(Fig. 3).

During infusion of NTG, the glucose:lactate ratio rapidly

returned towards baseline. In the catheters that had been

perfused with the higher concentration of NA, baseline

conditions were restored after 60 min, whereas in the

catheters perfused with the lower NA concentration, the

glucose:lactate ratio returned to baseline already after 20 min

of NTG infusion. After the NTG infusion was stopped,

glucose:lactate concentration stabilized in the catheters

perfused with high concentration of NA, while a sharp drop

was seen in the catheters perfused with the low concentration

of NA.

3.4. Urea clearance

The urea clearance, defined as the concentration of urea in the

dialysate, increased during the infusion of NA and remained

increased during phase C, indicating a decrease in blood flow

around the catheter (Fig. 4). Perfusing the catheters with NTG

0 60 12 0 180

0.07

3

30

0.3

time (minutes)

chan

ge

in [

nora

dre

nal

ine]

(m

mol/

L)

NA NTG

30 mm ol/ L NA

3 mm ol/ L NA

Fig. 6 – Mean (SEM) concentration of noradrenaline in the

dialysate during the experiment. A mean reverse recovery

of 64% was obtained for the catheters perfused with

30 mmol/L NA and a mean reverse recovery of 88% was

obtained for the catheters perfused with 3 mmol/L NA.

180120600-4

-2

0

2

4

time (minutes)

chan

ge

in [

ure

a] (

mm

ol/

L)

NA NTG

3 mm ol/ L NA

30 mm ol/ L NA

Fig. 4 – Mean (SEM) change in the concentration of urea in

the dialysate during perfusion of the catheters with NA

and NTG. There was no significant difference in urea

clearance between the catheters perfused with high and

low concentrations of NA.

b u r n s 3 8 ( 2 0 1 2 ) 9 7 5 – 9 8 3980

(phase D) resulted in an immediate decrease in urea,

suggestive of local increase in blood flow, which reversed

during the recovery phase (phase E). There was no significant

difference in urea clearance between the catheters perfused

with high and low concentrations of NA.

3.5. Laser Doppler perfusion imaging

Baseline perfusion was 0.75 PU (Fig. 5). During NA infusion,

perfusion decreased, and a significant difference from

baseline was seen at the end of phase B (0.60 (0.04) PU,

p < 0.05). This decreased perfusion level was sustained

throughout phase C. When catheters were perfused with

NTG, a rapid and large increase in perfusion was observed,

which continued during phase E. A maximum perfusion level

of 1.9 (0.3) PU was reached at phase E.

0 60 12 0 180

0.5

1.0

1.5

2.0

time (minutes)

per

fusi

on (

PU

)

NA NTG

30 mm ol/ L NA

Fig. 5 – Mean (SEM) absolute change in perfusion units

measured by laser Doppler during infusion of NA and

NTG.

3.6. Noradrenaline

The mean concentration of NA in the dialysate during phase B

was 10.6 (0.9) mmol/L for the catheters perfused with

30 mmol/L NA (Fig. 6). This implies a reverse recovery of

64%. The catheters perfused with 3 mmol/L NA had a mean

concentration of NA in the dialysate of 0.36 (0.11) mmol/L,

corresponding to a reverse recovery of 88%.

Immediately after changing to a perfusate without NA

(phase C), concentrations of NA in the dialysate showed a

rapid and large decrease to 0.5 mmol/L for the catheters

perfused with 30 mmol/L NA, and to concentrations below the

detection limit of 0.07 mmol/L for the catheters perfused with

3 mmol/L NA.

3.7. Correlations between metabolic markers and ureaclearance

There was a significant correlation between the change in urea

(dermal blood flow) and the change in lactate and glucose:-

lactate ratio during noradrenaline infusion, whereas there

was a significant correlation between the change in urea

(dermal blood flow) and the change in glucose and glucose:-

lactate ratio during nitroglycerine infusion. The correlation

coefficients and p-values are presented in Table 2.

Table 2 – Individual correlations between urea andmetabolite data during pharmacological interventionsphase B and D, respectively.

Noradrenaline (B) Nitroglycerin (D)

Pearson’s r p Pearson’s r p

Glucose �0.30 0.2 �0.88 <0.01

Lactate 0.80 <0.01 0.35 0.2

Glucose:lactate ratio �0.63 0.03 �0.81 <0.01

b u r n s 3 8 ( 2 0 1 2 ) 9 7 5 – 9 8 3 981

4. Discussion

Several new and interesting findings are presented in this

study that may have future implications for the use of

circulatory support in critical care and especially burn critical

care.

Firstly, microdosing of vasoactive drugs at a fixed rate and

concentration through the microdialysis probe, to the

dermal layer in the skin of healthy volunteers induces

reproducible vascular and metabolic time-dependent re-

sponse patterns. Secondly, although previous microdialysis

studies have claimed that the dose used in this study

(30 mmol/L) is in the physiological range, the one-tenth of

that dose still induced appreciable and remaining vasocon-

striction in the dermis. Thirdly, dermis exposed to these

noradrenaline doses seems to lack mechanisms to induce

autoregulatory vasodilatation by normal intrinsic pathways.

This resulted in a sustained vasoconstriction with clear

signs of hypoperfusion and local ischemia with significant

metabolic disturbances. Fourthly, the data presented sug-

gest that the present human microvascular dermal model

may be used to assess local microvascular dose-response

effects of vasoactive substances. This needs to be further

extended in patients to better understand the effects both of

the burn injury and the drugs given for circulatory support.

As the dermal glucose concentrations during the later

ischemic period did not reach zero levels, the data suggest

that during hypoperfusion and/or ischemia that has been

induced by noradrenaline, a defect in dermal cellular glucose

uptake may be present. This is possibly the result of dermal

depletion of energy, or insulin, or both. Fifth, in the present

dermal model urea clearance by retrodialysis seems to be

superior to the laser Doppler technique in registering blood

flow effects in the dermis during vasoconstriction. A

significant advantage of the urea clearance technique is

that it provides a dermal tissue blood flow estimate in the

same tissue volume as were the metabolites are sampled.

This simplifies the experimental design as it reduces the

need to incorporate another technique to measure dermal

blood flow. Unfortunately, the variation in the ethanol data

was unacceptable in this study, indicating that this tech-

nique is non-operational at microdialysis perfusion rates of

2 mL/min in the dermis. This contradicts the findings in e.g.,

skeletal muscle tissue.

4.1. Dose

There have been many experiments applying the microdia-

lysis technique in skin and using noradrenaline to study

effects on drug recovery and effects on skin blood flow [14].

The dose of noradrenaline that we chose for this study is well

established and is thought to be in the physiologic and

therapeutic range [14–16]. However, we found long-lasting

effects, including ischemia (lactate accumulation) and re-

duced glucose uptake, which implies that the dose was high

and that providing the drug to dermis with reduced blood flow

might lead to a deposition of noradrenaline in the tissue. This

was the reason why we also examined the effects of a lower

dose (one tenth). However, this dose was also found to have

significant dermal effect that comprised both hypoperfusion

and ischemia as indicated by increases in lactate.

4.2. Autoregulatory escape – dermal protection

Given the short half-life of noradrenaline, we expected a fast

release of the vasoconstriction when we changed the perfu-

sion fluid from noradrenaline to buffer solution. The micro-

circulation is regulated by intrinsic metabolic systems that

balance sympathetic tone to maintain oxygen tension above

critical values [17]. These mechanisms protect other vascular

beds such as skeletal muscle and intestine during high

sympathetic tone, and during exposure to pharmacological

vasoconstriction [17,18]. Our results in this human dermal

model suggest that these protective mechanisms, at the doses

examined seem to be absent or dysfunctional in the dermis.

This may have implications for the dermis when exposed to

vasoconstrictive drugs during e.g., burn shock resuscitation

where the dermis thereby may be hypo perfused or made

ischemic leading to burn wound progression and an increased

burn wound depth.

4.3. Metabolic effects

Glucose concentrations in the dermis changed inversely to

lactate during the noradrenaline administration to the

dermis. This finding using microdialysis is consistent with

previous investigations in non-insulin-dependent diabetic

patients where it has been shown that tissue glucose uptake

during hypoperfusion and ischemia is dependent on local

insulin delivery and the integrity of the energy dependent

insulin receptor. As dermal hypoperfusion and ischemia

leads to local depletion of energy and insulin, it mediates

activation of other pathways of glucose turnover that

are insulin independent [19]. Our finding is interesting, as

the mechanism of stress-induced peripheral insulin resis-

tance still is under debate and it suggests that the present

model may be useful in this perspective for future

investigations.

During dermal exposure to nitroglycerine, changes in

both urea clearance and laser Doppler showed a pronounced

hyperemic response with increased dermal blood flow

during which initially interstitial dermal glucose concentra-

tions was well above zero. This may indicate remaining

glucose uptake impairment, possibly due to a sustained

microcirculatory hypoperfusion, ischemia and/or a reperfu-

sion disturbance. The late dermal interstitial glucose

increase, well above baseline and present despite anticipat-

ed unchanged dermal blood flow levels, is consistent with a

peripheral insulin resistance, such as seen in the dermis of

burn patients [3]. Such a delivery, i.e. exceeding metabolic

capacity is also claimed in cytopathic hypoxia. Similar signs

of cellular metabolic defects have not been recorded in other

microdialysis models of tissue hypoperfusion or ischemia

such as secondary to the use of tourniquets or in skin flap

surgery [20,21]. This suggests that it may be related to the

effects of noradrenaline in the present model. Both nor-

adrenaline [5,6] and nitric oxide [22] have been associated

with impaired mitochondrial function and altered energy

metabolism.

b u r n s 3 8 ( 2 0 1 2 ) 9 7 5 – 9 8 3982

4.4. Methodological and model considerations

Dermal blood flow as assessed by the urea clearance technique

showed considerable changes over time during the pharma-

cological interventions, and the close correlation with the

alterations in the concentrations of metabolites, lactate, and

glucose:lactate ratio, strongly supports the idea that urea

clearance adequately reflected the relative change in dermal

blood flow. This was further supported by the close correlation

to the laser Doppler results. However, the continuing change

(increase in dialysate recovery) in urea over time during

vasoconstriction indicates that dermal blood flow decreased

further, even after laser Doppler technique had lost its

sensitivity to detect perfusion decreases. This assumed

decline in dermal perfusion was further supported by the

continuing changes in concentrations of metabolic markers

(mainly glucose and lactate). The low sensitivity of laser

Doppler to detect skin vasoconstriction is previously known

[23]. Urea clearance in this dermal model was operational at

low perfusion velocities (0.3 mL/min) and this is in contrast to

the most commonly used retrodialysis-based blood flow

technique which uses clearance of ethanol [8]. The ethanol

technique was developed for skeletal muscle applications and

in that setting it is dependent on a dialysis perfusion velocity

in the level of 2 mL/min. We are not aware of any publications,

which have applied the ethanol clearance technique to skin.

Although ethanol is considered the gold standard for micro-

dialysis-based assessment of local blood flow, it did not work

in the present dermal model where blood flow is significantly

less than in skeletal muscle. For the urea technique it is an

advantage to be operational at low perfusion velocities as it

also permits sampling of substances with low interstitial

concentrations and dialysate recoveries, such as cytokines

[24,25]. These features make the urea clearance technique

potentially interesting for further clinical use [26].

4.5. Tissue monitoring

Our findings of local tissue changes in glucose homeostasis,

which are not reflected systemically during either dermal

hypoperfusion, ischemia or reperfusion injury, also highlight

the fact that tissue glucose concentrations may not be fully

mirrored by systemic changes. This emphasizes that the use

of local glucose monitoring has to be interpreted with caution

[27,28], and this has previously also been previously debated

[29]. The discrepancy between local tissue-specific, in the

present case dermal, and systemic glucose responses to

hypoperfusion, ischemia and reperfusion injury suggests

not only a future value for tissue monitoring, but also a better

understanding of local tissue glucose homeostasis, particu-

larly as this may be important in the debate about different

regimens and outcomes using intensive insulin treatment and

particularly in the burn injured [30,31].

4.6. Clinical perspective

From a burn critical care perspective this human in vivo

dermal model using microdialysis has the potential to be of

future value for investigations of dermal effects of pharmaco-

logical burn shock circulatory support measures in patients.

Furthermore, it may also be applied in other situations of

circulatory failure in the burn unit, such as during sepsis, and

where it may be assumed that e.g., newly transplanted skin

may be at a risk due to dermal hypoperfusion secondary to the

use of potent vasoconstrictors [32].

5. Conclusion

Noradrenaline administration by microdialysis induced re-

producible and dose-dependent hypoperfusion and ischemia

of the dermis in healthy volunteers. We particularly noted that

dermal glucose concentrations responded rapidly to hypo-

perfusion but remained higher than zero indicative of an

energy-dependent deficiency in cellular uptake. Furthermore,

vasoconstriction remained after the noradrenaline perfusion

stopped implicating vasospasm and an attenuated dermal

autoregulatory capacity. These findings, i.e., compromised

dermal blood flow and metabolism are particularly interesting

from the burn shock resuscitation perspective where nor-

adrenaline is used for circulatory support. The clinical value

and the importance of these results obtained in this in vivo

dermal model in healthy volunteers needs to be further

explored in burn-injured patients.

Conflict of interest

All authors state that there are no conflict of interest.

r e f e r e n c e s

[1] Arturson G. Forty years in burns research – the postburninflammatory response. Burns 2000;26:599–604.

[2] Rosenthal SR. Pharmacologically active and lethalsubstances from skin. Arch Environ Health 1965;11:465–76.

[3] Samuelsson A, Steinvall I, Sjoberg F. Microdialysis showsmetabolic effects in skin during fluid resuscitation in burn-injured patients. Chem Commun 2006;10:R172.

[4] Samuelsson A, Sjoberg F. The autors reply. Crit Care Med2007;35:1445–6.

[5] Lunemann JD, Buttgereit F, Tripmacher R, Baerwald CG,Burmester GR, Krause A. Norepinephrine inhibits energymetabolism of human peripheral blood mononuclear cellsvia adrenergic receptors. Biosci Rep 2001;21:627–35.

[6] Rump AF, Klaus W. Evidence for norepinephrinecardiotoxicity mediated by superoxide anion radicals inisolated rabbit hearts. Naunyn-Schmiedeberg’s ArchPharmacol 1994;349:295–300.

[7] Farnebo S, Samuelsson A, Henriksson J, Karlander LE,Sjoberg F. Urea clearance: a new method to register localchanges in blood flow in rat skeletal muscle basedon microdialysis. Clin Physiol Funct Imaging 2010;30:57–63.

[8] Hickner RC, Rosdahl H, Borg I, Ungerstedt U, Jorfeldt L,Henriksson J. Ethanol may be used with the microdialysistechnique to monitor blood flow changes in skeletalmuscle: dialysate glucose concentration is blood-flow-dependent. Acta Physiol Scand 1991;143:355–6.

[9] Leis S, Drenkhahn S, Schick C, Arnolt C, Schmelz M,Birklein F, et al. Catecholamine release in human skin—amicrodialysis study. Eur Neurol 2004;188:86–93.

b u r n s 3 8 ( 2 0 1 2 ) 9 7 5 – 9 8 3 983

[10] Krogstad AL, Jansson PA, Gisslen P, Lonnroth P.Microdialysis methodology for the measurement of dermalinterstitial fluid in humans. Br J Dermatol 1996;134:1005–12.

[11] Anderson C, Andersson T, Wardell K. Changes in skincirculation after insertion of a microdialysis probevisualized by laser Doppler perfusion imaging. J InvestDermatol 1994;102:807–11.

[12] Petersen LJ, Kristensen JK, Bulow J. Microdialysis of theinterstitial water space in human skin in vivo: quantitativemeasurement of cutaneous glucose concentrations. J InvestDermatol 1992;99:357–60.

[13] Lloyd B, Burrin J, Smythe P, Alberti KG. Enzymicfluorometric continuous-flow assays for blood glucose,lactate, pyruvate, alanine, glycerol, and 3-hydroxybutyrate.Clin Chem 1978;24:1724–9.

[14] Clough GF, Boutsiouki P, Church MK, Michel CC. Effects ofblood flow on the in vivo recovery of a small diffusiblemolecule by microdialysis in human skin. J Pharmacol ExpTher 2002;302:681–6.

[15] Boutsiouki P, Clough GF. Modulation of microvascularfunction following low-dose exposure to theorganophosphorous compound malathion in human skinin vivo. J Appl Physiol 2004;97:1091–7.

[16] Wilson TE, Monahan KD, Short DS, Ray CA. Effect of age oncutaneous vasoconstrictor responses to norepinephrine inhumans. Am J Physiol Regul Integr Comp Physiol2004;287:R1230–4.

[17] Vallet B. Endothelial cell dysfunction and abnormal tissueperfusion. Crit Care Med 2002;30:229–34.

[18] Nygren A, Thoren A, Ricksten S-E. Vasopressors andintestinal mucosal perfusion after cardiac surgery:norepinephrine vs. phenylephrine. Crit Care Med2006;34:722–9.

[19] Niklasson M, Holmang A, Sjostrand M, Strindberg L,Lonnroth P. Muscle glucose uptake is effectively activatedby ischemia in type 2 diabetic subjects. Diabetes2000;49:1178–85.

[20] Ostman B, Michaelsson K, Rahme H, Hillered L. Tourniquet-induced ischemia and reperfusion in human skeletalmuscle. Clin Orthop Relat Res 2004;260–5.

[21] Setala LP, Korvenoja EM-L, Harma MA, Alhava EM, UusaroAV, Tenhunen JJ. Glucose, lactate, and pyruvate response in

an experimental model of microvascular flap ischemia andreperfusion: a microdialysis study. Microsurgery2004;24:223–31.

[22] Protti A, Singer M. Bench-to-bedside review: potentialstrategies to protect or reverse mitochondrial dysfunctionin sepsis-induced organ failure. Crit Care 2006;10. 228-.

[23] Golster H, Thulesius O, Nilsson G, Sjoberg F. Heterogeneousblood flow response in the foot on dependency, assessed bylaser Doppler perfusion imaging. Acta Physiol Scand1997;159:101–6.

[24] Farnebo S, Lars-Erik K, Ingrid S, Sjogren F, Folke S.Continuous assessment of concentrations of cytokines inexperimental injuries of the extremity. Int J Clin Exp Med2009;2:354–62.

[25] Sjogren F, Svensson C, Anderson C. Technical prerequisitesfor in vivo microdialysis determination of interleukin-6 inhuman dermis. Br J Dermatol 2002;146:375–82.

[26] Farnebo S, Zettersten EK, Samuelsson A, Tesselaar E,Sjoberg F. Assessment of blood flow changes in human skinby microdialysis urea clearance. Microcirculation2011;18:198–204.

[27] Beardsall K, Ogilvy-Stuart AL, Ahluwalia J, Thompson M,Dunger DB. The continuous glucose monitoring sensor inneonatal intensive care. Arch Dis Child Fetal Neonatal Ed2005;90:307–10.

[28] De Block C, Manuel-Y-Keenoy B, Van Gaal L, Rogiers P.Intensive insulin therapy in the intensive care unit:assessment by continuous glucose monitoring. DiabetesCare 2006;29:1750–6.

[29] Klonoff DC. Subcutaneous continuous glucose monitoringin severe burn patients. Crit Care Med 2007;35:1445–6.

[30] Finfer S, Chittock DR, Su SY-S, Blair D, Foster D, Dhingra V,et al. Intensive versus conventional glucose control incritically ill patients. N Engl J Med 2009;360:1283–97.

[31] van den Berghe G, Wouters P, Weekers F, Verwaest C,Bruyninckx F, Schetz M, et al. Intensive insulin therapy incritically ill patients. N Engl J Med 2001;345:1359–67.

[32] Silverman HJ, Penaranda R, Orens JB, Lee NH. Impairedbeta-adrenergic receptor stimulation of cyclic adenosinemonophosphate in human septic shock: association withmyocardial hyporesponsiveness to catecholamines. CritCare Med 1993;21:31–9.