ww.sciencedirect.com

j o u r n a l o f s u r g i c a l r e s e a r c h 1 8 6 ( 2 0 1 4 ) 7 3e8 0

Available online at w

ScienceDirect

journal homepage: www.JournalofSurgicalResearch.com

Effect of collagen and elastin content on the burst pressure ofhuman blood vessel seals formed with a bipolar tissue sealingsystem

Cassandra A. Latimer, MS,a,* Meghan Nelson, BS,a,b Camille M. Moore, MS,c

and Kimberly E. Martin, MSCSa,d

aDepartment of Research and Development, Covidien Surgical Solutions, Boulder, ColoradobDepartment of Biological Engineering, Massachusetts Institute of Technology, Cambridge, MassachusettscDepartment of Biostatistics and Informatics, University of Colorado Denver, Aurora, ColoradodColorado Clinical and Translational Sciences Institute, University of Colorado Denver, Denver, Colorado

a r t i c l e i n f o

Article history:

Received 3 June 2013

Received in revised form

22 July 2013

Accepted 5 August 2013

Available online 27 August 2013

Keywords:

Collagen

Elastin

Human blood vessel

Burst pressure

Bipolar vessel sealing

* Corresponding author. Department of ReseTel: þ1 303 581 7063; fax: þ1 303 516 6718.

E-mail address: Cassandra.Latimer@covid0022-4804/$ e see front matter ª 2014 Elsevhttp://dx.doi.org/10.1016/j.jss.2013.08.003

a b s t r a c t

Background: Bipolar devices are routinely used to seal blood vessels instead of sutures and

clips. Recent work examining the impact of vascular proteins on bipolar seal performance

found that collagen and elastin (CE) content within porcine arteries was a significant

predictor of a vessel’s burst pressure (VBPr). This study examined seal performance across

a range of human blood vessels to investigate whether a similar relationship existed. In

addition, we compared VBPr and CE content between porcine and human blood vessels.

Our primary hypothesis is that higher collagen-to-elastin ratio will predict higher VBPr in

human vasculature.

Methods: In six cadavers, 185 blood vessels from nine anatomic locations were sealed using

a bipolar electrosurgical system. A linear mixed model framework was used to evaluate the

impact of vessel diameter and CE content on VBPr.

Results: The effect of CE ratio on VBPr is modified by vessel size, with CE ratio having larger

influence on VBPr in smaller diameter vessels. Seal burst pressure of vessels 2e5 mm in

diameter was significantly associated with their CE content. Comparison of average VBPr

between species revealed porcine carotid and iliac arteries (440e670 mmHg) to be the best

vessel types for predicting the seal strength of most human blood vessels (420e570 mmHg)

examined.

Conclusions: CE content significantly modified the seal strength of small to medium sized

blood vessels but had limited impact on vessels >5 mm.

ª 2014 Elsevier Inc. All rights reserved.

1. Introduction evaluations have been performed to assess the seal strength,

Bipolar electrosurgical devices are routinely used in open

and laparoscopic surgical procedures to provide hemostasis

to dissected tissue structures and blood vessels. Multiple

arch and Development, C

ien.com (C.A. Latimer).ier Inc. All rights reserved

quantified through burst pressure, of different bipolar tissue

sealing systems; however, there are significant deviations in

reportedmeasurements [1e5]. In general, porcine arteries and

veins are used as amodel for human blood vessels due to their

ovidien Surgical Solutions, 5920 Longbow Dr., Boulder, CO, 80301.

.

j o u r n a l o f s u r g i c a l r e s e a r c h 1 8 6 ( 2 0 1 4 ) 7 3e8 074

anatomic and physiological similarities. Some variation in the

reported seal strength arises from the test method [6] or types

of bipolar devices used, but when these factors are controlled,

considerable differences remain. Further examination of

published data found that the type of porcine blood vessel

tested to evaluate seal strength differed by study and in some

cases within a study.

Blood vessels from different anatomic locations have

varying viscoelastic properties depending on their functional

role [7,8]. The mechanical properties of blood vessels may be

further influenced by genetics, age, lifestyle, and disease state

[9e13]. Two primary components of vessel walls that have an

important effect on the elasticity of blood vessels are elastin

and collagen [14e16]. Prior work by this group examined the

role of these structural proteins in porcine blood vessels and

their influence on bipolar vessel seal strength [17]. A signifi-

cant association between the ratio of collagen and elastin

content (CE ratio) and seal strength as defined by vessel burst

pressure (VBPr) was found when controlling for vessel diam-

eter; specifically vessels with larger CE ratios demonstrated

greater seal strength. Conversely, no association was detected

between vessel diameter and seal strength when controlling

for CE ratio.

Given the limited published data measuring the seal

strength of human blood vessels [18] and the lack of data

comparing seal strength measurements between human and

porcine blood vessels, it remains unclear whether porcine

arteries are the best model for predicting the strength of

human vessel seals. The objective of this investigation was to

determine the relationship between human cadaver VBPr and

CE ratio and evaluate the suitability of porcine arteries as

amodel for human blood vessels. Our primary hypothesis was

in human cadaver blood vessels, higher CE ratio is associated

with increased VBPr and the strength of this relationship

depends on vessel size. In keeping with Barlow’s formula

(VBPr ¼ 2 � Strength � Wall Thickness/Diameter), we further

hypothesized that the ratio of vessel wall thickness-to-

diameter (WTD ratio) may also contribute to VBPr variability

and impact the relationship between CE ratio and VBPr.

Finally, we qualitatively evaluated the suitability of various

porcine arteries as a model for predicting the bipolar seal

strength of human vessels by comparing the average VBPr of

human cadaver blood vessels observed in this study to those

observed in prior studies of porcine arteries [17].

2. Methods

2.1. Human vessel collection and analysis of VBPr

Our research was performed on six cadaveric subjects with

donor consent obtained through Science Care (Phoenix, AZ).

One female and five male subjects ranging in age (40e61 y),

body mass index (BMI; 14e28), smoking status (yes, 2; no, 4),

and who expired from various causes were evaluated.

Cadavers were stored for postmortem at 4�C until the time of

dissection. Vessel harvesting from cadavers was performed

within 3e11 d after death. Cadavers were dissected and nine

types of vessels (carotid artery, deep femoral artery, femoral

artery, iliac artery, inferior mesenteric artery, pulmonary

artery, pulmonary vein, renal artery, and splenic artery) were

evaluated. Bipolar tissue sealing systems are regularly used to

seal the inferior mesenteric artery, splenic artery, and

pulmonary arteries and veins in surgery. Although clinical

sealing of carotid, femoral, renal, and iliac arteries would

rarely occur, these vessels were included to compare with

previously published data on the porcine vessel testing model

and provide a range of vessel diameters and CE ratios to

further elucidate the relationship between size, CE, and VBPr.

Vessels segments were carefully dissected from

surrounding connective and fatty tissue and their diameters

were measured using white cotton string and a disposable

ruler. Following diameter measurement, vessels were sealed

in situ using a bipolar vessel-sealing system (LigaSure Atlas;

Covidien, Boulder, CO). All seals were made with the standard

two-bar setting on the ForceTriad generator system (Covidien)

and one seal cycle per vessel was performed. After the vessel

was sealed, the knife blade incorporated within the LigaSure

Atlas device was activated creating two sealed segments.

Before the dissection, one side of the seal was randomly

selected for either burst test and collagen and elastin (CE)

quantification or histologic examination. Sealed vessel

segments including at least a 1-cm margin from the sealed

tissue were removed from the cadaver for burst testing and

histologic processing.

Sealed VBPr was determined using previously described

methods [2,17]. Briefly, a blunt tip cannula was inserted into

the open vessel lumen and an iris was clamped around the

vessel to contain infused water within the vessel lumen.

Deionized water was injected into the vessel at a rate of 100

mL/min until the seal burst. Burst pressure was recorded

using a pressure meter (Fluke; Everett, WA). The maximum

VBPr was recorded for each vessel tested.

2.2. Histologic analysis of vessel structure and CEcontent

Selected sealed vessel segments not subjected to burst testing

were used for histologic analysis. After excision from the

cadaver, the vessel samples were placed in 10% phosphate-

buffered formalin for a minimum of 48 h before undergoing

standard histologic processing. Samples were shipped to an

independent histology laboratory (Premier Laboratory, LLC

Boulder, CO) for sectioning, staining, and imaging. Histologic

structure stains, hematoxylin and eosin and a modified

Mason’s trichrome stain [19], were used to qualitatively

examine the seal area, vessel structure, and CE content. Vessel

wall thickness measurements were performed on modified

Mason’s trichromeestained samples.

2.3. Quantification of vessel CE content

After burst testing, vessel sampleswere placed into cryo tubes

and stored in dry ice until they could be transferred to a�80�Cfreezer. Tissue sectionswere dissected from tissue adjacent to

the seal. Care was taken during dissection to ensure all vessel

layers were included in samples. Thawed tissue sections were

weighed and transferred to microcentrifuge tubes for CE

quantification. Total collagen was determined from tissue

hydroxyproline content using the method described in

j o u r n a l o f s u r g i c a l r e s e a r c h 1 8 6 ( 2 0 1 4 ) 7 3e8 0 75

previous work [17]. Total elastin content, including soluble

and insoluble elastin, was quantified using a commercially

available kit (Fastin Elastin Kit; Biocolor Ltd., Belfast, UK).

Following the kit instructions, duplicate measurements were

made for each tissue sample and the final elastin concentra-

tion was expressed in mg elastin/mg tissue.

2.4. Data analysis

Differences in vessel diameter, CE ratio, and VBPr between

vessel groups were examined using a KruskaleWallis test due

to nonnormal distribution of the data. A P value <0.05 was

used to determine statistical significance. Additionally, using

95% median confidence intervals (CIs), significant pairwise

differences between vessel groups were evaluated.

To address our first hypothesis that CE ratio would predict

VBPr, VBPr was modeled using a linear mixed model frame-

work to account for the correlation due to repeated

measurements taken on the same donors. VBPr was natural

log transformed to satisfy the assumptions of linear mixed

models. The primary explanatory variables included in the

model were CE ratio and vessel diameter. An interaction

between these variables was included to allow the effect of CE

ratio on burst pressure to differ by vessel size. In addition, age,

BMI, and smoking status were included as covariates in the

model to control for donor-specific factors. A random inter-

cept was used to account for between donor variability and for

correlation betweenmeasurements made on the same donor.

Sex was not included as a covariate in the model as only one

donor was female, completely confounding the effect of sex

and other sources of between subject variability, so that the

impact of sex could not be estimated.

A similar linear mixed model framework was used to

address our second hypothesis that WTD ratio could also

influence burst pressure. Due to the smaller sample size (N ¼75) available for this analysis and the large number of poten-

tial covariates, a backwards selection with a P to stay of 0.1

was used to develop a parsimonious model. In the model

selection process, interactions were removed from the model

before their associated main effects. Predictors considered in

themodel selection were CE ratio, vessel diameter, CE ratio by

vessel diameter interaction, WTD ratio, and CE ratio by WTD

ratio interaction, age, BMI, and smoking status.

All analyses were performed in R with the “nlme” package.

Sensitivity analyses were performed to determine the influ-

ence of the data from the cadaver with a BMI of 14 on the

results.

3. Results

3.1. Sample collection and exclusion criteria

A total of 223 vessel samples were obtained from the six

cadaveric subjects in the nine previously listed vessel groups.

To be included in the CE ratio and VBPr analysis, sampleswere

required to have both a burst pressure measurement and CE

quantification; 185 samples met these criteria. The excluded

samples did not have burst test measurements due to

technical difficulties during burst testing, such as equipment

failure or the sample length being too short.

For theWTD ratio and VBPr investigation, 75 samples were

available for analysis. Wall thickness measurements were

performed on samples collected for histology. Because

histology was only performed on selected samples, the

sample size for this analysis was smaller than the CE ratio and

VBPr analysis.

3.2. Histologic analysis of vessel structure and CEexpression

A qualitative difference in the amount of collagen (blue)

between different blood vessels was observed, with the

splenic arteries displaying less collagen staining than the

other vessels. Elastin content (black) did not appear to vary

considerably between vessel types. Within the seal structure,

vessel layers were compressed and most of the boundaries

were not visible, but staining indicating CE remained, Figure 1.

3.3. Descriptive statistics

Mean and standard deviations for VBPr, CE ratio, collagen

content, elastin content, and number of samples per vessel

group are summarized in Table 1. A comparison of vessel

diameter between vessel types using a KruskaleWallis test

resulted in statistically significant differences (P < 0.001)

between groups. Vessel groups that do not share the same

letter have significantly different diameter sizes (95% median

CIs), Figure 2. Performing a KruskaleWallis median test on the

difference in median VBPr between vessel types also resulted

in a significant difference (P ¼ 0.001). Comparison tests con-

ducted among vessel groups using the 95% median CIs

showed a significant difference in median VBPr between all

vessel groups, with the exception of the pulmonary artery

group and the splenic artery group. Carrying out the same

analysis on CE ratio resulted in a significant difference (P <

0.001). Deep femoral arteries had a significantly larger CE ratio

than carotid, splenic, and pulmonary arteries; femoral arteries

had significantly larger CE ratio than carotid and pulmonary

arteries.

3.4. Hypothesis I: CE ratio and diameter as predictorsof VBPr

Controlling for age, BMI, and smoking status, the effect of CE

ratio on burst pressure was significantly modified by vessel

size (P¼ 0.01). As vessel size increases, the effect of CE ratio on

burst pressure decreases, Figure 3A. For example, for 2-mm

diameter vessels, a 0.1 unit increase in CE ratio would result in

a 5.9% increase in burst pressure (95% CI: 2.3% to 9.7%

increase, P ¼ 0.001); however, for a 6 mm vessel, the same

increase in CE ratio would result in only a 0.8% increase in

burst pressure (95% CI: 0.9% reduction to 2.7% increase, P ¼0.4). The effect of CE ratio for a range of vessel diameters was

modeled and is shown in Table 2. Although the impact of

vessel diameter on burst pressure became stronger with

increasing CE ratio, size was not a significant predictor of

burst pressure for vessels with typical CE ratios between

Fig. 1 e Blood vessel structure with a modified trichrome stain following sealing with a bipolar tissue sealing device. Blue

stained fibers (collagen) and black stained fibers (elastin) vary by vessel type. Ca[ carotid artery; DFe[ deep femoral artery;

Fe [ femoral artery; Il [ Iliac artery; IMA [ inferior mesenteric artery; PuV [ pulmonary vein; PuA [ pulmonary artery; Re

[ renal artery; Sp [ splenic artery. (Color version of figure is available online.)

Table 1 e VBPr, CE ratio, collagen concentration, elastin concentration, and number of samples by vessel group.

Vessel group VBPr (mmHg) CE ratio (w/w) Collagen (mg/mg) Elastin (mg/mg) n

Deep femoral artery 529 (249) 1.6 (0.6) 70.2 (19.5) 45.6 (13.0) 14

Femoral artery 428 (206) 1.3 (0.5) 59.1 (16.5) 48.5 (12.5) 60

Carotid artery 423 (131) 0.8 (0.3) 46.2 (13.6) 58.2 (11.1) 22

Iliac artery 459 (251) 1.1 (0.4) 64.5 (26.1) 56.3 (16.1) 12

IMA 564 (271) 1.1 (0.3) 54.9 (18.8) 54.0 (15.3) 7

Renal artery 480 (226) 1.0 (0.3) 59.0 (14.8) 60.1 (15.9) 26

Splenic artery 170 (50) 0.7 (0.3) 34.8 (16.0) 51.2 (5.6) 5

Pulmonary artery 320 (173) 0.8 (0.2) 48.9 (15.1) 65.6 (11.4) 20

Pulmonary vein 353 (183) 1.0 (0.3) 62.2 (20.1) 63.8 (11.4) 19

IMA ¼ inferior mesenteric artery; w/w ¼ weight/weight.

Results are expressed as mean (SD) where applicable.

j o u r n a l o f s u r g i c a l r e s e a r c h 1 8 6 ( 2 0 1 4 ) 7 3e8 076

Fig. 2 e Blood vessel diameter was significantly different

between human vessels evaluated. Vessel groups that do

not share the same letter (A, B, C, D or E) have significantly

different (P-value < 0.05) diameter sizes. Results are

expressed as mean ± standard error.

j o u r n a l o f s u r g i c a l r e s e a r c h 1 8 6 ( 2 0 1 4 ) 7 3e8 0 77

0.8 and 1.2. These results were robust to sensitivity analyses

excluding the subject with low BMI.

3.5. Hypothesis II: effects of WTD ratio on VBPr

After model selection, the following variables remained in the

model: WTD ratio, CE ratio, a WTD ratio by CE ratio interac-

tion, and age. Controlling for age, the effect of CE ratio on burst

Fig. 3 e The effect of CE ratio on VBPr decreases as vessel diame

versus Ln VBPR (mmHg) for vessels 2e5 mm and 5e10 mm in d

(mmHg) for vessels with WTD ratios of 0.06e0.15 and 0.15e0.29

pressure is significantly modified by a vessel’s WTD ratio (P ¼0.046). As the WTD ratio increases, the impact of CE ratio on

burst pressure decreases, Figure 3B. For example, for a vessel

with a relatively lowWTD ratio of 0.1, burst pressure increases

by 6.6% for each 0.1 unit increase in CE ratio (95% CI: 1.4% to

12.2% increase, P ¼ 0.01). For vessels with higher WTD ratio of

0.2, this effect is attenuated, with burst pressure increasing by

only 1.1% for each 0.1 unit increase in CE ratio (95% CI: 1.4%

decrease to 3.6% increase, P ¼ 0.4). In this study, the mean

WTD ratio was 0.15, standard deviation ¼ 0.05. For vessels in

this range, burst pressure increases by 3.8% for each 0.1 unit

increase in CE ratio (95% CI: 0.7% to 7.0% increase, P ¼ 0.02),

Table 3. Again, these results were robust to sensitivity anal-

yses excluding data from the cadaver with low BMI.

4. Discussion

The objective of our investigation was to evaluate the rela-

tionship between the structural proteins CE and seal strength

in human blood vessels. In terms of immediate utility, the aim

was to use this information to determine which type(s) of

porcine arteries were the most clinically relevant models for

human blood vessels. This data may also drive improvements

in device design or in generator algorithms. Our primary

hypothesiswas that in human cadaver vessels, higher CE ratio

is associated with increased VBPr and that the strength of this

association depends on vessel size. We found a complex

ter and WTD ratios increase. (A) Regression plot of CE ratio

iameter. (B) Regression plot of CE ratio versus Ln VBPR

.

Table 2 e Estimated percent change in VBPr for a 0.1 increase in CE ratio for a range of vessel diameters.

Vesseldiameter(mm)

Percent change inVBPr for a 0.1

increasein CE ratio (%)

Lower 95%confidencelimit (%)

Upper 95%confidencelimit (%)

P value

2 5.9 2.3 9.7 0.001

3 4.6 1.9 7.4 <0.001

4 3.3 1.4 5.3 <0.001

5 2.0 0.5 3.5 0.008

6 0.8 �0.9 2.4 0.4

j o u r n a l o f s u r g i c a l r e s e a r c h 1 8 6 ( 2 0 1 4 ) 7 3e8 078

relationship between CE ratio, vessel diameter, and VBPr in

cadaver vessels, with CE ratio having a greater influence on

VBPr for smaller vessels compared with larger vessels. Similar

to previous porcine studies, VBPr was found to increase with

CE ratio, although this effect was not statistically significant

for vessels >5 mm in diameter. Previous porcine studies did

not investigate the potential for differing effects of CE ratio on

VBPr by vessel size because these studies were restricted to

vessels in the most clinically relevant range (1e5 mm).

Although the devices are cleared for vessels up to 7 mm in

diameter, sealing vessels of this size rarely occurs without

a clip or suture backup in clinical practice due to surgeon

concern for the outcome.

We also hypothesized that in keeping with Barlow’s

formula, the WTD ratio may also interact with other variables

to contribute to VBPr variability. To investigate the role of

additional vessel wall structures, we examined the impact of

total vessel wall thickness on the relationship of CE ratio and

VBPr. The wall thickness parameter was chosen to provide

a general estimation of the impact of all vessel wall structures,

not just CE, on VBPr. Analysis of the data subset showed that

as the WTD ratio increased, the effect of CE ratio on VBPr was

reduced. On average for a WTD of 0.15, for each 0.1 unit

increase in CE ratio VBPr increased by 3.8%. Given the varia-

tion in CE ratio across the vascular tree, this translates into

meaningful differences in VBPr between vessel types, Table 2.

Histologically, vessel walls of large diameter vessels

exhibited thicker smooth muscle layers than small diameter

vessels. Smooth muscle may have an impact on either the

ability of the device to seal the vessel or on large vessel wall

distensibility. If the smooth muscle layer is thick enough,

additional jaw pressure, seal cycle time, or power may be

required to approximate the opposing vessel walls and

denature the vascular constituents into an amalgam.

Table 3 e Estimated percent change in VBPr for a 0.1 increase

WTD Percent change inVBPr for a 0.1

increasein CE ratio (%)

Lowconlim

0.1 6.6

0.12 5.5

0.15 3.8

0.18 2.1 �0.2 1.1 �

Furthermore, thick smooth muscle layers may affect the

mechanics of the bursting process by altering the overall

distensibility of the vessel. Including this parameter in future

studies may assist with identifying a more accurate VBPr

model and may lead to minor alterations in the device which

could lead to improvements in seal quality in these types of

vessels.

Bipolar seal strength was significantly different between

human blood vessel types evaluated. The splenic artery,

pulmonary artery, and pulmonary vein groups had lower

average burst pressures than the other tested vessel types.

Pulmonary blood vessels were anticipated to have lower burst

pressures due to the range (4e30 mmHg) of normal physio-

logical blood pressure exposure [20]. Limited data have been

published on the physiological blood pressure of the splenic

artery; one author reported the proximal back pressure of 15

splenic artery stumps to be 48.0 � 9.8 mmHg [21]. Using this

measurement, the average splenic artery burst pressure (170

� 50 mmHg, Table 1) measured in cadaver tissue was at least

three times the reported back pressure. The remaining human

vessel types evaluated had an average VBPr within 400e600

mmHg, Table 1. Comparing these average VBPrs with young

porcine artery average VBPr, porcine iliac (440 mmHg) and

carotid (670mmHg) [17] arteries appeared to be the best vessel

models for predicting bipolar seal strength of most of the

tested human blood vessels. Additionally, porcine femoral

(270 mmHg) [17] and cadaver splenic arteries had similar

average VBPr. Porcine renal arteries had a significantly higher

average VBPr (1030 mmHg) [17] than all cadaver vessel groups

tested.

Our study did have some limitations. A key difference

between the porcine and cadaver investigations was the state

of the tissue evaluated: living porcine tissue versus cadaver

tissue. Clinical results in a living patient may not be identical

in CE ratio for a range of WTD.

er 95%fidenceit (%)

Upper 95%confidencelimit (%)

P value

1.4 12.2 0.01

1.2 10.0 0.01

0.7 7.0 0.02

0.3 4.6 0.1

1.4 3.6 0.4

j o u r n a l o f s u r g i c a l r e s e a r c h 1 8 6 ( 2 0 1 4 ) 7 3e8 0 79

to results in the cadaver model. We determined that using

human vessels excised from a surgical procedure would not

be feasible due to the limited number of excised vessels

collected and the difficulty performing the burst pressure tests

with these types of samples. Published data comparing the

mechanical properties of cadaver, excised, and intact human

blood vessels indicated that the viscoelastic, biochemical, and

functional properties were similar between vessels among

collection types [22e24]. As a result, we determined cadaver

vessels would be an appropriate model for our study because

a range of vessel sizes and types could be collected from

a single subject and the mechanical properties would be

comparable with living blood vessels.

Another issue with the cadaver model is the exposure of

vessel seals to fluid pressure. Vessel seals created during

in vivo porcine laboratories were immediately exposed to

blood and seals could be judged on pass or fail criteria. Seals

performed on cadaver tissuewere only evaluated during burst

pressure tests. If fluid pressure in situ does influence seal

quality, we were not able to determine its impact in our study.

To limit the effect of tissue degradation on blood vessel

protein quantification and mechanical function, cadavers

were stored at 4�C and we attempted to harvest and test blood

vessels within a week after death. One donor was tested 11

d after death but protein content and VBPr were not signifi-

cantly different between this cadaver and the cadavers within

our donor criteria. Therefore, vessels obtained from all

subjects were combined into one data set.

The evaluations performed in both the porcine and cadaver

tissue studies used the LigaSure vessel sealing system (Covi-

dien). This system uses a proprietary closed-loop algorithm,

whichuses tissue impedance todetermineenergydelivery rate

and seal completion. Consequently, the conclusions from

these studies may have been influenced by the tissue sealing

system and may not entirely be applicable to other bipolar

technologies. However, because vessels from various

anatomical beds have different mechanical properties, the

circumferential force acting on any vessel seal would still be

influenced by vessel structure itself. Thus, vessel structure

likely has an effect on all energy sealed vessels but the extent

may depend on the type of bipolar device used. Further eluci-

dating the impactof tissuestructureonsealperformancecould

lead to the development of more reliable tissue sealing tech-

nologies that can tailor seals to tissue type. If it isnotpossible to

improve bipolar sealing performance by altering mechanical

and energy delivery algorithms alone, incorporating an exog-

enous substrate such as collagen could assist with sealing

problematic tissue or vessels with thin walls such as the

pulmonary artery and pulmonary vein. At the least, tissue

structure information could be used by vessel sealing systems

to determine the likelihood of seal success and could warn

users when incomplete seals are more likely to occur.

New bipolar vessel sealing technologies are increasingly

being approved for clinical use.Hemostatic efficacy is themost

important function of bipolar tissue sealers. Standardizing the

way these devices are evaluated [6] and having an appropriate

model to determine seal performance is critical. Most of the

youngporcineartery typesandcadaverbloodvesselsexhibited

comparable seal performance; in particular, porcine carotid

and iliac arteries and the majority of cadaver vessel groups

evaluatedhad similar averageVBPrs. In contrast, porcine renal

arteries were the outlier group between studies, exhibiting

much larger VBPrs than all other porcine and cadaver vessel

groups examined. In both comparison studies between tech-

nologies and in testing to gain regulatory clearance, care

should be taken to ensure vessel types are clinically relevant

and uniformly distributed between groups, so conclusions are

not influenced by the vessel model.

Acknowledgment

The authors thank Jaime Kean (PhD), Kimberly Krugman (MS),

Clayton Ramey (DVM), and Behzad Elizeh (MS) for technical

assistance.

r e f e r e n c e s

[1] Person B, Vivas DA, Ruiz D, Talcott M, Coad JE, Wexner SD.Comparisonof four energy-basedvascular sealing and cuttinginstruments: a porcine model. Surg Endosc 2008;22:534.

[2] Newcomb WL, Hope WW, Schmelzer TM, et al. Comparisonof blood vessel sealing among new electrosurgical andultrasonic devices. Surg Endosc 2009;23:90.

[3] Katsuno G, Nagakari K, Fukunaga M. Comparison of twodifferent energy-based vascular sealing systems for thehemostasis of various types of arteries: a porcine model-evaluation of LigaSure ForceTriad. J Laparoendosc Adv SurgTech A 2010;20:747.

[4] Rothmund R, Kraemer B, Neis F, et al. Efficacy and safety ofthe novel electrosurgical vessel sealing and cuttinginstrument BiCision((R)). Surg Endosc 2012;26:3334.

[5] Milsom J, Trencheva K, Monette S, et al. Evaluation of thesafety, efficacy, and versatility of a new surgical energydevice (THUNDERBEAT) in comparison with Harmonic ACE,LigaSure V, and EnSeal devices in a porcine model. JLaparoendosc Adv Surg Tech A 2012;22:378.

[6] Krugman KA, Martin KE, Cosgriff N, Slakey DP. In search ofthe autologous clip: a case for experimental standardization.J Laparoendosc Adv Surg Tech A 2011;21:721.

[7] Bia D, Aguirre I, Zocalo Y, et al. [Regional differences inviscosity, elasticity and wall buffering function in systemicarteries: pulse wave analysis of the arterial pressure-diameter relationship]. Rev Esp Cardiol 2005;58:167.

[8] Valdez-Jasso D, Haider MA, Banks HT, et al. Analysis ofviscoelastic wall properties in ovine arteries. IEEE TransBiomed Eng 2009;56:210.

[9] Bjarnegard N, Lanne T. Arterial properties along the upperarm in humans: age-related effects and the consequence ofanatomical location. J Appl Physiol 2010;108:34.

[10] Simon A, Levenson J. Effect of hypertension onviscoelasticity of large arteries in humans. Curr HypertensRep 2001;3:74.

[11] London GM, Marchais SJ, Guerin AP. Arterial stiffness andfunction in end-stage renal disease. Adv Chronic Kidney Dis2004;11:202.

[12] Agrotis A. The genetic basis for altered blood vessel functionin disease: large artery stiffening. Vasc Health Risk Manag2005;1:333.

[13] Learoyd BM, Taylor MG. Alterations with age in theviscoelastic properties of human arterial walls. Circ Res 1966;18:278.

[14] Bank AJ, Kaiser DR. Arterial wall mechanics. In: Lanzer P,Topol EJ, editors. PanVascular Medicine: Integrated Clinical

j o u r n a l o f s u r g i c a l r e s e a r c h 1 8 6 ( 2 0 1 4 ) 7 3e8 080

Management. New York: Springer Verlag Berlin Heidelberg;2002. p. 151.

[15] Bank AJ, Wang H, Holte JE, Mullen K, Shammas R, Kubo SH.Contribution of collagen, elastin, and smooth muscle toin vivo human brachial artery wall stress and elasticmodulus. Circulation 1996;94:3263.

[16] Burton AC. Relation of structure to function of the tissues ofthe wall of blood vessels. Physiol Rev 1954;34:619.

[17] Sindram D, Martin K, Meadows JP, et al. Collagen-elastinratio predicts burst pressure of arterial seals created usinga bipolar vessel sealing device in a porcine model. SurgEndosc 2011;25:2604.

[18] Tsunezuka Y, Waseda R, Yachi T. Electrothermal bipolarvessel sealing device LigaSureV for pulmonary arteryligationdburst pressure and clinical experiences incomplete video-assisted thoracoscopic major lungresection for lung cancer. Interact Cardiovasc Thorac Surg2010;11:229.

[19] Goodfellow B, Mikat M. A stain for the simultaneousdemonstration of collagen, muscle, and elastin elements inmammalian tissue. Lab Med 1988;243.

[20] Klingensmith ME, Chen LE, Glasgow SC, Goers TA, Melby SJ.Table 30-1 Normal Hemodynamic Parameters. TheWashington manual of surgery. Fifth Edition. Philadelphia:Wolters Kluwer Health/Lippincott Williams & Wilkins;2008:512.

[21] Archie JP Jr. Splenic artery stump back pressure. Am Surg1992;58:504.

[22] Bia D, Zocalo Y, Pessana F, et al. [Viscoelastic and functionalsimilarities between native femoral arteries and fresh orcryopreserved arterial and venous homografts]. Rev EspCardiol 2006;59:679.

[23] Santana DB, Armentano RL, Zocalo Y, et al. Functionalproperties of fresh and cryopreserved carotid and femoralarteries, and of venous and synthetic grafts: comparisonwith arteries from normotensive and hypertensive patients.Cell Tissue Bank 2007;8:43.

[24] Holzapfel GA, Sommer G, Gasser CT, Regitnig P.Determination of layer-specific mechanical properties ofhuman coronary arteries with nonatherosclerotic intimalthickening and related constitutive modeling. Am J PhysiolHeart Circ Physiol 2005;289:H2048.