In vivo measurement

of blood vessel

wall thickness GILBERT J. L’ITALIEN, IAIN AND WILLIAM M. ABBOTT

G. KIDSON,

JOSEPH

MEGERMAN,

Vascular Research Laboratory, Department of Surgery, Harvard Medical and Massachusetts General Hospital, Boston, Massachusetts 02114 L’ITALIEN, GILBERT J., IAIN G. KIDSON, JOSEPH MEGERMAN, AND WILLIAM M. ABBOTT. In vivo measurementof blood vesselwah thickness. Am. J. Physiol. 237(2): H265-H268, 1979 or Am. J. Physiol.: Heart Circ. Physiol. 6(2): H265-H268, 1979.-To understand the mechanical properties of arteries and vascular grafts, it is crucial that the wa.lI thickness in these vessels be known. Unfortunately, all available methods for measuring this parameter require the removal of the vessel, which precludes the study of such vessels as a function of time. A new radiographic technique for measuring the wall thickness of arteries and vascular grafts in vivo, utilizing contrast materials injected into the vessel lumen and applied to the outer surface of the vessel, is described. Radiographs are obtained with a portable X-ray machine and analyzed using a calibrated microscope. The technique has been successfully applied to the in vivo measurement of wall thickness in canine arteries, veins, and experimental vascular grafts. It is concluded that the method provides better than 95% accuracy in a variety of vessels and that it can be used to study changes in vascular grafts after their implantation into the arterial circulation. radiographic;

elastic properties;

arterial

School

Rovick (5) measured wall thickness angiographically in excised cylindrical specimens that were filled with mercury and coated externalIy with lead oxide. To determine thickness as a function of diameter, radiographs were taken at several intraluminal pressures or, utilizing the assumption of constant volume and uniform density, walI thickness was derived from a single measurement at the mean physiological pressure. Unfortunately, all of these techniques require excision of the vessel, which is known to produce significant dimensional alterations (9) and which precludes their subsequent study. The need to measure the thickness of vascular grafts as a function of time led us to adapt the method of Dobrin and Rovick, thereby enabling such measurements in vivo. The new technique has been applied to canine arteries, veins, and arterial grafts, as well as to uniform rigid tubes. The latter provide estimates of error that are independent of biological variability.

graft METHODS

APPLICATION OF ELASTIC THEORY to blood vessels depends on the accurate determination of the thickness of the vessel wall. This dimension partially determines the magnitude of the stress that is generated within the wall from which the elastic properties of the vessel itself are calculated (9). It is thought that the successful development of an arterial prosthesis depends in part on the proper matching of mechanical properties between the graft and host artery (7). In addition, changes in the elasticity of implanted vein grafts can provide information regarding their healing responses when subjected to the rigors of the arterial circulation (10). Because the thickness of such grafts undergoes significant changes after implantation, accurate thickness measurements become particularly important. Various techniques have been developed to measure the thickness of blood vessels. McDonald (8) calculated the thickness from the vessel’s weight in air and water, assuming a constant density and volume at various physiological pressures. Attinger (1) measured the thickness of unfixed, unstretched histological sections directly, thus avoiding the density assumption. Peterson et al. (11) used a similar method but fixed specimens at arterial pressure and length prior to measurement. Dobrin and A DETAILED

0363-6135/79/0000-0000$01.25

Copyright

0 1979 the American

Physiological

The mean physiological pressure is obtained, after the vessel of interest is exposed, by cannulating a distal branch with rigid polyethylene tubing (PE-190; OD, 0.067 in.) connected to a Statham P23dB pressure transducer. An X-ray film is prepared by opening a Kodak Dental X-ray envelope (Eastman-Kodak, Rochester, NY) trimmed to 2 cm2, inserting a piece of ultra-hard Kodachrome RC paper and resealing the envelope with lightproof adhesive tape. It is held in place beneath the exposed vessel by a special clamp (Fig. 1) designed to assure a perpendicular orientation between the film and the X-ray beam and a fixed distance (-2 cm) from the xray emitter. A narrow thin coating of lead oxide suspension is applied laterally to a l-cm length of the vessel and a reference rod of known, comparable dimension is placed beside it for the purpose of calibration. The vessel is then occluded proximally; flushed repeatedly with heparinized saline via the same distal branch used to monitor pressure, and then clamped distally. This flushing process assures the removal of any residual blood, thus avoiding filling defects at the vessel wall. The vessel is then filled with Renografin 60 contrast media (E. R. Squibb & Sons, Princeton NJ) while the intraluminal pressure is monitored. Enough radiographic Society

H265

Downloaded from www.physiology.org/journal/ajpheart by ${individualUser.givenNames} ${individualUser.surname} (128.111.121.042) on July 30, 2018. Copyright © 1979 American Physiological Society. All rights reserved.

L’ITALIEN,

H266

FIG. 1. Schematic diagram film: C: point of attachment blood vessel or graft.

of fiim to X-ray

KIDSON,

MEGERMAN,

AND

ABBOTT

holder: A: set screws; B: X-ray machine; D: reference rod; E:

dye is injected to produce the mean physiological pressure and the film is exposed with a portable X-ray machine for 3 s at 15 mA and 72 kV. The film is then developed on a Kodak X-omat processor. Each film is mounted on a glass slide and subsequently read by three separate observers on a x10-x25 stereoscopic microscope adapted with an eyepiece that has a metric scale etched on its surface, providing a resolution of 0.01 mm. At least four measurements per observer are obtained within the length of the lead oxide coating. Each reading is corrected for magnification due to both the X-ray and the microscope by the observed enlargement of the reference rod diameter. The measured thickness is then taken as the mean of 12 individual readings. RESULTS

Examples of radiographs obtained with the various vessels are shown in Fig. 2, and a schematic diagram of a typical radiograph is given in Fig. 3. Uniform rigid tubes made of polyethylene and Teflon, with known wall dimensions, were initially measured by two variations of the method and the results, including error estimates, are shown in Table 1. Wall thickness was defined as the distance from either the inner edge of the lead oxide layer to the luminal surface, as delineated by the contrast medium, or from the outer edge of the lead oxide to the same inner surface of the tube. In both cases, less than a 5% error (standard error of the mean of 12 observations x 100, divided by that mean) was obtained. However, the ratio of measured to nominal thickness was nearer unity when the width of the lead oxide layer was included. This procedure was therefore adopted for all subsequent studies. The results of in vivo measurements of the wall thickness of arteries, veins, and various arterial grafts are shown in Table 2. Also included are in vitro measure-

FIG.

femoral

2. Typical radiographs of vessels artery; C: femoral vein; D: PTFE;

studied: A: latex E: Dacron.

tubing;

B:

ments on latex rubber tubing, which served as a model of a distensible but uniform vessel. It is clear that the precision with which the thickness of a vessel can be measured depends on the degree to which the surface of the vessel can be defined. That is, smooth vessels such as artery .vein, rubber, and PTFE yield an average SEM 5 0.008 mm, which represents the minimum error attainable by the radiographic procedure. The more irregular Dacron and vein grafts (after 2 wk implantation) yield twice this value, but since these grafts are relatively thick, the percent error incurred when measuring their thickness remains below 5%. The radiographic method has been used in vivo in several hundred vessels in chronic experimental animals without adverse reactions. No difference was found in the patency rates for 160 arterial grafts in which thickness measurements were or were not taken and such measurements are now routinely made in this laboratory upon implantation of experimental grafts. DISCUSSION

The importance of thickness in the assessment of vessel elasticity is evident from equations that describe the mechanics of a tube subjected to a distending pressure. For a thin-walled cylinder, the circumferential stress (tension) is given by

T=PxR/t

Downloaded from www.physiology.org/journal/ajpheart by ${individualUser.givenNames} ${individualUser.surname} (128.111.121.042) on July 30, 2018. Copyright © 1979 American Physiological Society. All rights reserved.

IN

VIVO

MEASUREMENT

OF

BLOOD

VESSEL

WALL

HZ67

THICKNESS REFERENCE

. .

.

.

;

,.‘:...:‘..: _.,.,‘.: . y:; . .,.: ._....._. ‘.:. ‘.:. : .“.I. _.. .... .;:. .’ .:. ,, (.‘I. :‘. . ‘. : : ‘. ‘. ‘,’._ .. _’ .;..‘._. :. ‘_... _.. ., . .. . . .. .... .I,.; .

. ..

,.,

._

._

., . .

. . .

.,

;_.

. .

. ... _’: .,.,

.

,:.

. . .

..

;,

‘...‘,‘....,

: : .. ._

..:....: :::,:..‘,..:, .,. .:..._... ..:. ‘_ ; .‘;. ..::; .,: ‘. ‘...::‘.:. .,. ,;...I...,’._ ‘... _, . . ‘,.... ..‘...:..

-.::,::

,....

.

. .._

1. Estimated

errors for stiff tubing Polyethylene*

~~ Nominal thickness,t mm Measured thickness, mm =fr SEM To inner edge of lead oxide Percent error* Measured/nominal To outer edge of lead oxide: Percent error* Measured/nominal * Becton, cations.

2. Estimated

TABLE

Vessel

Latex rubber Artery Vein Vein grafts after 2 wk PTFE$ Dacron8 * SEM n vessels. vessels. MedicaIs,

0.203

0.301 * 0.013 4.3 0.82 0.392 & 0.007 1.8 1.06

0.170 * 0.006 3.5 0.84 0.221 A 0.008 3.6 1.09

t Manufacturer’s of same) X 100.

errors for compliant Range of thickness, mm

n

12

25 15

6 14 16

Teflon*

0.368

Dickinson, Rutherford, NJ. $ (SEM of 12 ob servations/mean

specifi-

vessels Avg SEM, mm*

Avg Percent Error?

0.316-0.391 0.174-0.474 0.132-0.330 0.282-0.611

0.005 0.008 0.008 0.019

1.60 2.61 4.16 4.01

0.615-0.793 0.588-0.860

0.008 0.019

1.19 2.59

obtained from 12 observations per vessel, then averaged for t Percent error as described in Table 1, averaged for n $ W. L. Gore & Associates, Newark, DE. 8 Meadox Inc., Oakland, NJ.

where P, R, and t are the intraluminal pressure and the vessel radius and thickness, respectively. The elasticity of nonbiological (linear) materials can be adequately represented by Young’s modulus, which is defined as the ratio of a given stress to the strain (AR/Ro) that is incurred upon the application of that stress. In the case of blood vessels however, the lack of a convenient reference radius (Ro) has led to the use of an “incremental modulus of elasticity” (4), defined as E

ni C

FIG.

.,’

-‘LEAD

TABLE

ROD

=- ATxR AR

which approximates the slope of the stress-strain curve. It is this parameter that describes the inherent elastic properties of the material from which a vessel is made, regardless of the vessel’s dimensions, and which can be most readily addressed in the manufacture of prosthetic arterial grafts. Because errors in t contribute to the cumulative error in T (2) and because the modulus of elasticity is obtained by a process of differentiation which enhances sensitivity to such errors, the need for accuracy in measuring t (as well as the other variables) becomes clear. Difficulties in assessing vessel thickness had led to the use of two

VESSEL LUMEN (filled with contrast VESSEL

graphic

3. Schematic technique.

diagram

of radio-

media)

WALL OXIDE

LAYER

parameters that reflect the elastic properties of the vessel as a whole, without needing to consider its thickness. The “pressure strain elastic modulus” or “elastance” was defined by Peterson et al. (11) as R X AP/A.R, and the “compliance” is simply its reciprocal. These parameters do provide useful indices of the behavior of an intact vessel, but they do not provide information on the inherent elasticity of the vessel material. That is, compliance information obtained for a specificvessel cannot be readily translated to vessels of different dimensions. The in vitro technique we have described to measure wall thickness no longer restricts the description of a vessel’s elastic properties to the use of elastance and compliance. The radiographic procedure is subject to several sources of error. These include ambiguities in the wall boundary as defined by the observer, actual variations in graft thickness along its length, magnification errors, and filling defects. It is apparent that the combined errors in rigid tubes (Table 1) and biological vessels (Table 2) containing well-defined external and internal borders is well below 5% (SEM x lOO/mean). Grafts demonstrating less well-defined surfaces, such as Dacron and implanted vein, also yield reliable results because the SEM remains small relative to the greater thickness of these vessels. Furthermore, a slight reduction in error may be obtained by coating the external wall on diametrically opposing sides, thereby doubling the sample size with little effort. The degree of magnification of the image varies directly with its distance from the X-ray source. The apparatus described in Fig. 1 serves to minimize this distance (