The Mechanical Performance and

Histomorphological Structure of the

Descending Aorta in Hyperthyroidism

Konstantinos G. Moulakakis, MD,*,† Dimitrios P. Sokolis, PhD,* Despina N. Perrea,PhD,† Theodosios Dosios, MD, FACS, FETCS,† Ismene Dontas, DVM, PhD,†Maria V. Poulakou, BSc,† Constantinos A. Dimitriou, MSc,* George Sandris, BSc,* andPanayotis E. Karayannacos, MD, PhD,* Athens, Greece

Thyroid hormones decrease systemic vascular resistance by directly affecting vascular smooth muscle

relaxation. There is limited literature about their effect on the mechanical performance of the aortic wall.

Therefore, the authors determined the influence of hyperthyroidism on the mechanical properties and

histomorphological structure of the descending thoracic aorta in rats. Severe hyperthyroidism was

induced in 20 male Wistar rats by administering L-thyroxine (T4) in their drinking water for 8 weeks; age-

matched normal euthyroid rats acted as controls. Animals were sacrificed, and the mechanical and

histomorphometrical characteristics of the descending thoracic aorta were studied. The aortic wall of

hyperthyroid rats was stiffer than that of euthyroid animals at the upper physiologic levels of stress or

strain (p < 0.05) but less stiff at the lower physiologic and lower levels (p < 0.05). The aorta of hyper-

thyroid animals compared with that of euthyroid ones showed an increase of the internal and external

diameters (p < 0.05), the media area (p < 0.05), the number of smooth muscle cell nuclei (p < 0.05), and

the collagen density (p < 0.05) and a decrease in the elastin laminae thickness (p < 0.001) and elastin

density (p < 0.001). In hyperthyroid rats, the aortic wall was stiffer at the upper physiologic and higher

levels of stress and strain. These changes correlated with microstructural changes of the aortic wall. The

coexistence of hyperthyroidism with disease states or clinical conditions that predispose to increased

arterial pressure may be associated with increased arterial stiffness and have undesirable consequences

on the mechanical performance of the thoracic aorta and hemodynamic homeostasis. These changes

could lead to an increased risk for developing vascular complications.

It is well known that thyroid hormones increaseperipheral oxygen consumption and substrate

requirements and have direct and indirecteffects on the heart and the vascular system. Thebasis for these effects is the ability of thyroid hor-mones to exert cellular action on cardiacmyocytes, as well as vascular smooth musclecells and endothelium,1,2 resulting in an increasein cardiac output and myocardial contractility,tachycardia, and widened pulse pressure and adecrease in systemic vascular resistance.1-4

Although thyroid hormones decrease sys-temic vascular resistance in small- and medium-size arteries by directly affecting vascularsmooth muscle relaxation, there is limited liter-ature about their effect on the mechanicalproperties and the histomorphological structure

Angiology 58:343–352, June/July 2007From the *Center for Experimental Surgery, Foundation ofBiomedical Research, Academy of Athens, Greece, and the†Laboratory for Experimental Surgery and Surgical Research,School of Medicine, University of Athens, Greece. This workwas supported in part by a grant from the Center forExperimental Surgery, Foundation of Biomedical Research,Academy of Athens, Greece.Correspondence: Konstantinos G. Moulakakis, MD, 132 AndreaPapandreou St, Glyfada, 16561, GreeceE-mail: konmoulakakis@yahoo.grDOI: 10.1177/0003319707301759©2007 Sage Publications

343Angiology Volume 58, Number 3, 2007

of the aorta.5-7 A few investigators, using non-invasive techniques, attempted to assess centralarterial stiffness in thyrotoxic patients and tocorrelate changes in arterial distensibility withan increased impact on cardiovascular risk.5-7

However, these noninvasive techniques mea-sure the average stiffness at various locations ofthe arterial tree, presenting the general status ofthe cardiovascular system.

The aim of the present study was to deter-mine the in vitro mechanical properties of theaorta in experimentally induced hyperthy-roidism in rats and to correlate these changeswith the histomorphology of the aortic wall.

Materials and Methods

Animals

Male Wistar rats (n = 40), weighing 250 to300 g, were used in this study. The experimen-tal protocol was approved by the InstitutionalAnimal Care and Use Committee of the Schoolof Medicine, University of Athens, Greece, andthe Veterinary Directorate of the AthensPrefecture (permit number K/53/31.01.03). Allanimals received humane care in compliancewith the “Guide for the Care and Use ofLaboratory Animals,” published by the NationalInstitute of Health (NIH Publication No. 85-23,revised 1996), and were housed individually inthe animal facility in cages complying withEuropean Union standards, in a controlled envi-ronment with a temperature of 20 ± 2°C, rela-tive humidity 55%, central ventilation (15 airchanges/h), and an artificial light:dark cycle of12 h. The animals were allowed to have freeaccess to standard rat chow and water. After1 week of adaptation to the new environment,they were randomly divided into 2 groups: therats in the experimental group (n = 20) weretreated with L-thyroxine (T4), whereas those inthe control group (n = 20) were age matchedand housed under the same conditions as thoseof the experimental group.

Induction of Hyperthyroidism

Severe hyperthyroidism was induced by chronicadministration of 0.0036% T4 (L-thyroxinesodium salt pentahydrate, Sigma Chemicals,

St. Louis, Missouri) in their drinking water for 8weeks. This solution was prepared every day.The daily consumption of water was measuredin each animal during the period of experiment.Although this therapeutic scheme requires a longperiod of treatment, it minimizes the large diur-nal variations in plasma hormone levels that takeplace when daily injections are used to maintainhyperthyroidism and avoids the stress related tothat procedure.8 We determined the dose of T4

in a pilot study and from the literature.9 Thisdose of T4 induced thyrotoxicosis, which wasconfirmed by serum hormonal determination aswell as by clinical and necropsy findings.

Clinical examination of the animals ofboth groups during the entire experimentalperiod included measurement of body weight(Navigator Balance, Ohaus Corporation, PineBrook, NJ), heart rate, rectal temperature, andevaluation of signs such as protracted diarrheaand hyperkinesis.

Serum total triiodothyronine (tT3), free tri-iodothyronine (fT3), and total thyroxine (tT4)were measured every 14 days during 8 weeksof treatment through to the day of sacrifice.Serum tT3 and fT3 concentrations were mea-sured by the microparticle enzyme immunoas-say (MEIA), and serum tT4 concentrations weredetermined by a fluorescence polarizationimmunoassay (FPIA). All of the above quanti-tative examinations were performed using anautoanalyzer IMX System (Abbott, DiagnosticsDivision, Abbott Park, Illinois).

Serum lipids were measured in animalsof both groups. Cholesterol and high-densitylipoprotein (HDL) cholesterol were determinedenzymatically by the CHOD-PAP method usinga commercially available kit (Biosis, Heraklion,Crete, Greece). Serum triglycerides were mea-sured by the enzymatic GPO-PAP method usinga commercially available kit (Biosis, Heraklion,Crete, Greece). Low-density lipoprotein (LDL)cholesterol was estimated by the Friedewaldformula.

Animal Sacrifice and Tissue Preparation

All animals of both groups were sacrificed 8weeks after the initiation of the study. Follow-ing sedation of the animal with a combinationof an intramuscular injection of ketamine (90mg/kg) and xylazine (5 mg/kg), the abdominalwall was entered through a midline incision,the inferior vena cava was dissected, and blood

Angiology Volume 58, Number 3, 2007344

was withdrawn with a heparinized syringe.Blood samples were centrifuged at 3000g for 15minutes, and the serum aliquots were stored at–80°C until analysis. Following the cessation ofrespiration and heart function, the thoracic cav-ity was opened through a median sternotomy.The heart was excised, washed from blood, andweighed. The descending aorta from the left sub-clavian artery to the diaphragm, with the sur-rounding loose periaortic tissue, was excisedwith extreme care to avoid damage to the aorticwall. A complete aortic ring of 2 to 3 mm inwidth, at the level of the second intercostalartery, was carefully excised. The aortic ring wasimmediately placed in formalin solution for 24hours, dehydrated in ethanol, and embedded inparaffin. By using a microtome (Leica RM 2125,Leica, Nussloch, Germany), 3 successive sectionsof 5 µm width were prepared from each specimen.These sections were stained with hematoxylin-eosin for nucleus staining, Verhoeff’s elastica forelastin, and Sirius red for collagen staining. Theremaining aortic segment was kept in saline solu-tion at room temperature for the mechanicalstudy, which took place 2 hours after the har-vesting of the aorta.

Aortic Wall Histomorphometry

Images of structural characteristics of the aorticwall were electronically acquired using lightmicroscopy and processed by the Sigma ScanPro 5 image analysis software (Aspire SoftwareInternational, Leesburg, Virginia), as previouslyreported.10-12 Measurements of the morphome-tric parameters of the aorta were performed,including the internal diameter, external diam-eter, media thickness, and media area. Theperimeters were measured directly using thesoftware by hand tracing the inner and outercircumference of the aortic media, defined bythe internal and external elastic laminae. Theimage analysis software automatically assessedthe media area when measuring the internaland the external perimeters. The effect ofshrinkage in these morphometric parametershas been considered. As described previously,shrinkage artifacts amounted on average to14% for fixation and 16% for dehydration andembedding in paraffin.13,14 The media elastinnetwork was analyzed in terms of the relativearea density of elastin, the number of lamellarunits, and the elastin laminae thickness. The

collagen network was analyzed in terms of therelative area density of collagen, whereas thesmooth muscle cell network was evaluated interms of the relative area density of smoothmuscle cells and the number of smooth musclecell nuclei in the media.

Mechanical Study of the Aortic Wall

Longitudinal strips of fixed dimensions wereobtained from the remaining aortic specimensafter excision of the histological sections.Mechanical testing of the aortic strips wascarried out on an automatic uniaxial tensile-testing device (Vitrodyne V1000 UniversalTester, Liveco Inc, Burlington, Vermont), as pre-viously described.11,12,15,16 The strips were heldin the grips of the apparatus, with the lowergrip connected to a fixed support and the upperone attached to the actuator. To minimize vis-coelastic phenomena, preconditioning, consist-ing of 10 successive cycles with constant finallevels of deformation, was performed, and thenthe strips were subjected to another cycle ofmechanical testing during which the data ofload versus deformed length were continu-ously recorded on the computer connected tothe tensile-testing device.

Owing to their nonuniform width and thick-ness, 4 equidistant measurements were takenfor each strip, and the average value of thosewas used. The deformed length of the strips andthe tensile load applied during mechanical testingwere recorded. These data were converted intostress-strain, elastic modulus-stress, and elasticmodulus-strain curves that were divided into 3parts, referring to low, physiologic, and highstresses, consistent with the biphasic nature ofthe aortic wall under passive conditions (seeappendix).

Statistical Analysis

Results are expressed as mean ± standard error ofmean (SEM). Microcal Origin v.6.1 for Windowsapplication (OriginLab Corp, Northampton,Massachusetts) was used for all analyses. Theresults of the experimental group were comparedwith those of the control group. The 2-tailedunpaired Student t test was used to estimate dif-ferences in aortic composition, morphometric, andregression parameters between groups. A p value<0.05 was considered significant.

Moulakakis Descending Aorta in Hyperthyroidism 345

Results

Biological Variables and SerumExamination

Three rats from the hyperthyroid group died dueto thyrotoxicosis at 46, 48, and 50 days, respec-tively, after the initiation of treatment with T4,whereas 1 rat from the control group died duringblood sampling. Thus, 17 animals from thehyperthyroid group and 19 animals from thecontrol group were included in the mechanicaland histomorphometric study of the aortic wall.

Daily measurements showed a gradualincrease in rectal temperature from the seventh

day of treatment in the hyperthyroid group,whereas euthyroid animals did not showchanges. On the fifth week, rectal temperaturewas significantly higher in the hyperthyroid(39.0 ± 0.4°C) than in the euthyroid (37.0 ±0.3°C) group (p < 0.001). The blood values ofthyroid hormones showed no differencebetween the animals of both groups at thebeginning of the study. The animals thatreceived T4 became markedly thyrotoxic by day15 of treatment, as manifested by serum thy-roid hormone levels, which remained highlyelevated until the time of sacrifice (Table I).These values were significantly different(p < 0.05) from those of the control group.

Angiology Volume 58, Number 3, 2007346

Table I. Thyroid hormone values and clinical and necropsy findings in both groups, 2 and 8 weeks after thebeginning of the study.

14th Day After Initiation of Experiment

Hyperthyroid (n = 20) Euthyroid (n = 20) p

Total T4, ng/mL 343.5 ± 22.4 39.3 ± 6.3 <0.001

Total T3, ng/mL 3.0 ± 0.6 0.34 ± 0.02 <0.001

Free T3, pg/mL 15.7 ± 6.4 0.57 ± 0.2 <0.001

Temperature, range °C 38.2-39.1 36.9-37.4 <0.05

Body weight, g 260 ± 14 330 ± 18 <0.001

56th Day After Initiation of Experiment (Sacrifice)

Hyperthyroid (n = 17) Euthyroid (n = 19) p

Total T4, ng/mL 478.2 ± 27.6 41.1 ± 6.9 <0.001

Total T3, ng/mL 3.2 ± 0.2 0.38 ± 0.03 <0.001

Free T3, pg/mL 20.4 ± 1.1 0.63 ± 0.06 <0.001

Cholesterol, mg/dL 49.4 ± 1.9 97.3 ± 4.2 <0.001

Triglycerides, mg/dL 85.0 ± 10.3 67.3 ± 3.3 0.319

LDL, mg/dL 21.2 ± 2.9 41.0 ± 3.2 <0.001

HDL, mg/dL 32.2 ± 1.2 42.7 ± 2.2 <0.001

Temperature, range °C 39.2-39.8 37.0-37.5 <0.05

Body weight, g 228 ± 12 448 ± 23 <0.001

Heart weight, g 1.62 ± 0.1 1.03 ± 0.08 <0.001

HR, beats/min 420 ± 10 360 ± 8 <0.001

LVW/BW 2.8 ± 0.1 1.8 ± 0.1 <0.001

Values are expressed as mean ± SEM. n = number of rats. LDL, low-density lipoprotein; HDL, high-density lipoprotein;LVW/BW = left ventricular weight/body weight ratio; HR = heart rate.

Hyperthyroid animals had an increased heartrate (420 ± 10 vs 360 ± 8 beats/min, p < 0.05),protracted diarrhea, and decreased final bodyweight (228 ± 12 vs 448 ± 23 g, p < 0.001)(Table I). Thyrotoxic status was confirmed bypathologic findings, such as left ventricularhypertrophy indicated by greater left ventricu-lar weight/body weight ratio (2.8 ± 0.1 vs 1.8 ±0.1, p < 0.001).

In the hyperthyroid group, serum total cho-lesterol (97 ± 4 vs 49 ± 2 mg/dL, p < 0.001),LDL cholesterol (41 ± 3 vs 21 ± 3 mg/dL, p <0.001), and HDL cholesterol levels (43 ± 2 vs32 ± 1 mg/dL, p < 0.001) were significantlydecreased compared with control rats (Table I).

Morphometric and StructuralCharacteristics of the DescendingThoracic Aorta

Morphometric parameters and media compositionof the descending thoracic aorta are listed in TableII, and representative histological sections areshown in Figure 1. The aortas of the hyperthyroidanimals in comparison to euthyroid ones showedan increase in internal diameter (p < 0.05), exter-nal diameter (p < 0.05), media area (p < 0.05),relative area density of collagen (p < 0.05), and

number of smooth muscle cell nuclei in the media(p < 0.05) and a decrease in the relative area den-sity of elastin (p < 0.001) and the elastin laminaethickness (p < 0.001). No difference was foundin the relative area density of smooth musclecells of the aortic wall between the animals ofthe 2 groups.

Mechanical Properties

Stress-strain curves of the thoracic aorta arepresented (Figure 2). Parts I and II of the curvefor hyperthyroid animals are displaced down-wards in comparison with those for euthyroidanimals, whereas the opposite is true for partIII. The elastic modulus-stress curves and theelastic modulus-strain curves showed an identi-cal graphical representation. Parts I and II ofthe curves for hyperthyroid animals were dis-placed downwards, but part III initiated atstrain levels referring to part II of the curve foreuthyroid animals (Figures 3 and 4).

The regression parameters for hyperthyroidand euthyroid animals are listed in Table III.Parameter k of part I increased (p < 0.001) andparameter q decreased (p < 0.001) in hyper-thyroid rats. Regarding the parameters of partII, intercept a decreased (p < 0.001) and slopeb increased (p < 0.05) in hyperthyroid animals.In part III, intercept c increased (p < 0.05) and

Moulakakis Descending Aorta in Hyperthyroidism 347

Figure 2. Stress-strain curves of the thoracic aortafor hyperthyroid (Hyper) and euthyroid (Eut) animals.Vertical bars denote ± SEM. The left panel refers to thelow (part I) and physiologic (part II) strain parts andthe right panel to the high (III) strain part of the curves.

Figure 1. Representative histological sections of thethoracic aortic wall from a euthyroid (Eut) animal and ahyperthyroid (Hyper) animal, stained with hematoxylin-eosin (left panel), Verhoeff’s elastica (middle panel),and Sirius red (right panel) (×40). Note the decrease inthe relative area density of elastin and the increase inthat of collagen, as well as the increase in the numberof smooth muscle cell nuclei in the aortic wall of thehyperthyroid animal.

Angiology Volume 58, Number 3, 2007348

Table II. Morphometry and composition of the aortic media in rats of the euthyroid and hyperthyroid groups.

Morphometric Parameters Hyperthyroid (n = 17) Euthyroid (n = 19) p

ID, mm 2.69 ± 0.10 2.37 ± 0.05 <0.05

ED, mm 3.02 ± 0.12 2.66 ± 0.06 <0.05

MA, mm2 1.47 ± 0.09 1.19 ± 0.07 <0.05

MT, µm 164.24 ± 12.59 165.74 ± 9.41 0.926

MT/ID, % 6.07 ± 0.29 6.93 ± 0.16 <0.05

Aortic Composition Hyperthyroid (n = 17) Euthyroid (n = 19) p

ELT, µm 3.3 ± 0.2 4.4 ± 0.1 <0.001

nSMCN 101.3 ± 6.1 75.0 ± 6.0 <0.05

nLU 8.3 ± 0.2 9.1 ± 0.4 0.084

El, % 16.1 ± 1.1 25.2 ± 1.1 <0.001

Col, % 30.5 ± 2.1 23.3 ± 1.8 <0.05

SMC, % 53.4 ± 1.9 51.5 ± 1.7 0.465

Values are expressed as mean ± SEM. n = number of rats. ID, internal diameter; ED, external diameter; MA, media area;MT, media thickness; ELT, elastin laminae thickness; nSMCN, number of smooth muscle cell nuclei; nLU, number oflamellar units; El, relative area density of elastin; Col, relative area density of collagen; SMC, relative area density ofsmooth muscle.

Table III. Effect of two months hyperthyroidism on regression parameters of the aortic stress-strain relationshipin rats.

Regression Parameters Hyperthyroid (n = 17) Euthyroid (n = 19) p

k, kPa 75.789 ± 4.134 59.740 ± 3.630 <0.001

q 0.221 ± 0.020 0.391 ± 0.026 <0.001

a, kPa 70.416 ± 5.889 108.075 ± 5.990 <0.001

b 1.552 ± 0.076 1.199 ± 0.083 <0.05

c, kPa –958.996 ± 164.685 –2015.557 ± 229.351 <0.05

d 17.390 ± 1.218 21.649 ± 1.274 <0.05

TI, kPa 29.836 ± 1.258 28.833 ± 0.885 0.509

εI 0.243 ± 0.003 0.210 ± 0.006 <0.001

TII, kPa 90.661 ± 3.531 108.556 ± 4.415 <0.05

εII 0.617 ± 0.009 0.650 ± 0.007 <0.05

Values are expressed as mean ± SEM. n = number of rats. k, q = parameters of part I; a, b = intercept, slope of part II; c, d =intercept, slope of part III; TI, εI = stress, strain at the first transition point; TII, εII = stress, strain at the second transition point.

slope d decreased (p < 0.05) in hyperthyroidanimals. Concerning the first transition point,this occurred in hyperthyroid rats at signifi-cantly higher levels of strain (p < 0.001),whereas the second transition point occurred atsignificantly lower (p < 0.05) levels of stressand strain.

These findings indicate that the aortic wallsof hyperthyroid animals were significantlystiffer than those of euthyroid animals at theupper physiologic levels of stress and strain butsignificantly less stiff at the lower physiologicand lower levels.

Discussion

Overt hyperthyroidism induces a hyperdynamiccardiovascular state characterized by high cardiacoutput with low systemic vascular resistance, afaster heart rate, enhanced left ventricular systolicand diastolic function, and increased prevalenceof supraventricular tachyarrhythmias.1,2 Also,hyperthyroidism induces biochemical and struc-tural changes in the myocardial tissue that canlead to progressive organ dysfunction. In theheart, this state has been implicated as the pri-mary cause of cardiomegaly, congestive heart fail-ure, and decreased exercise tolerance.17,18

However, the effects of this state on the structureand the performance of the arterial wall have notbeen extensively investigated.

This study revealed a bipolar variability ofthe mechanical behavior of the aorta in hyper-thyroid rats. Cumulative elastic modulus-strainand elastic modulus-stress curves demonstrateda different behavior in relation to the stress andstrain levels applied. The aortic wall of hyper-thyroid animals was significantly stiffer thanthat of euthyroid ones at the upper physiologiclevels of stress but significantly less stiff at thelower physiologic and lower levels. Moreover,the aortic wall was significantly stiffer at theupper physiologic and higher levels of strain butsignificantly less stiff at the lower physiologicand lower levels. Differences in mechanical per-formance between the 2 groups were found,indicating that the vessel wall was stiffer inhyperthyroid animals. Based on the in situstrain of the descending thoracic aorta reportedin our previous study,16 physiologic endovas-cular pressures are included in part II of thestress-strain curves, whereas parts I and III cor-respond to low and high endovascular pres-sures, respectively.

Light microscopy examination disclosed a sig-nificant increase in internal diameter, externaldiameter, media area, area density of collagen,and number of smooth muscle cell nuclei inthe aortic segments of hyperthyroid animals.

Moulakakis Descending Aorta in Hyperthyroidism 349

Figure 4. Elastic modulus-strain curves of thethoracic aorta for hyperthyroid (Hyper) and euthyroid(Eut) animals. Vertical bars denote ± SEM. The leftpanel refers to the low (part I) and physiologic (II)strain parts and the right panel to the high (III) strainpart of the curves.

Figure 3. Elastic modulus-stress curves of thethoracic aorta for hyperthyroid (Hyper) and euthyroid(Eut) animals. Horizontal and vertical bars denote ±SEM. The left panel refers to the low (part I) andphysiologic (II) stress parts and the right panel to thehigh (III) stress part of the curves.

A significant decrease in the elastin laminaethickness and the elastin density was alsonoticed. These structural findings allow theinterpretation of the mechanical characteristicsof the aorta. It is known that the aortic wall isprimarily composed of elastin, collagen, andsmooth muscle, accounting for its structuralintegrity, metabolism, and mechanical proper-ties. Elastin and collagen determine the passivemechanical properties of the aortic wall,19

whereas smooth muscle cells are responsible forits active properties and production of extracel-lular matrix.20 We previously reported16 that atlow pressures, passive stiffness is determined bythe content and ultrastructure of elastin. At highpressures, it is determined by collagen and atphysiologic pressures by both elastin and colla-gen content. These findings are consistent withthe reports of other histological studies.13,19,21,22

The induced alterations in the mechanical prop-erties of the aortic wall, observed in the hyper-thyroid animals of the present study, may beattributed to structural changes of the aortic wall.

There has been an increased interest in therelationship between arterial distensibility andcardiovascular disease. It has been suggestedthat increased arterial stiffness is associatedwith the development of cardiovascular diseaseand that it may even predict its development atan early stage before vascular lesions or clini-cal symptoms become evident.23,24 Changes inarterial stiffness may contribute directly to thedevelopment of systemic hypertension.25-27

Similar findings showing a reduction in arterialdistensibility have been found in populations ofpatients suffering from diabetes mellitus,28,29

hypercholesterolemia,30 and end-stage renaldisease.31 Also, stiffness is correlated with ath-erosclerosis, probably through the effects ofcyclic stress on arterial wall histological alter-ations.32-34 Experimental studies in animalsreported a reduction in arterial distensibility tobe accompanied by accelerated atherosclerosis,possibly because arterial stiffening increases thetraumatic effect of intravascular hemodynam-ics on the arterial wall.35,36 Based on these find-ings, the stiffness of large vessels has beenrecognized as an important determinant of car-diovascular morbidity. Central arterial stiffnessin hyperthyroid patients has been investigatedin 3 studies. Obuobie et al,5 based on arterialpressure curves, reported that patients withuntreated thyrotoxicosis have a lowered central

arterial stiffness. Similar studies based on non-invasive techniques demonstrated a signifi-cantly increased stiffness in the common carotidarteries in hyperthyroid patients.6,7

As was previously reported, hypercholes-terolemia has been implicated as a determi-nant factor of arterial stiffness.30 Patients withhypercholesterolemia have a higher centralpulse pressure and stiffer blood vessels thanmatched controls, despite similar peripheralblood pressures.37 There are a number ofpotential mechanisms linking larger artery stiff-ening and plasma lipids, including atheroscle-rosis, changes in the elastic elements of thearterial wall, endothelial dysfunction, and inflam-mation.38 Aggressive lipid-lowering therapy,particularly with statins, leads to a reduction inarterial stiffness by the antioxidant and anti-inflammatory effects. Through these effects,statins probably contribute to the significantreduction in vascular events.39 In our study, thehypermetabolic state in hyperthyroid animalswas associated with hemodynamic and biologi-cal changes as well as a decrease in serum lipidlevels. The aortic wall of hyperthyroid animalswas significantly stiffer, although a decrease intotal cholesterol and LDL cholesterol levels wasobserved. However, there is some evidence thatHDL cholesterol decreases aortic stiffness.40

Therefore, it is difficult to interpret the overalleffect of the change in the lipid profile (fall intotal cholesterol, LDL cholesterol, and HDL cho-lesterol) on arterial stiffness. Based on theseobservations, it would seem that the histomor-phological changes that we observed can beattributed to the action of thyroid hormones onthe arterial wall and the induced hyperdynamiccardiovascular state.

The experimental method of tensile testingused in this study to establish the stress-strain,elastic modulus-stress, and elastic modulus-strain curves of the aortic wall has been previ-ously considered in detail.11,12,15,16 Applicationof this procedure presents a number of advan-tages over the in vivo methods. It measures themechanical properties of the tested materialover a wide range of stresses, beginning fromzero and including stresses that correspond tothe systolic-diastolic range of blood pressureduring each cardiac cycle. Furthermore, in con-trast to the in vivo methods, it is not subjectedto errors induced by the effects of ongoing anes-thesia and iatrogenic manipulation.

Angiology Volume 58, Number 3, 2007350

Conclusions

The thoracic aorta of thyrotoxic rats showedsignificant alterations in its mechanical proper-ties. It became stiffer at high levels of stress andstrain, including the upper physiologic range.These changes correlated with microstructuralchanges of the aortic wall, including a differentquantitative distribution of connective tissue fil-aments. In the clinical setting, the coexistenceof hyperthyroidism with disease states or clini-cal conditions that predispose to increased arte-rial pressure may be associated with increasedarterial stiffness and have undesirable conse-quences on the mechanical performance of thethoracic aorta and the hemodynamic home-ostasis, leading to an increased risk of develop-ing vascular complications.

APPENDIX

Stress (T) was calculated using the formula T =F/w0t0, where F is the tensile load exerted to theaortic strips, and w0 and t0 are the initial widthand thickness. Strain (ε) was calculated using theformula ε = λ − 1, where the longitudinal stretchratio (λ) is the strips’ deformed length at eachtensile load over their initial length at no load.Aortic stiffness was quantified in terms of theelastic modulus (M), which was calculated as thefirst derivative of stress over strain, M = dT/dε.

Plotting these data as elastic modulus versusstress, 1 nonlinear and 2 linear parts wereobtained that correspond to low, physiologic, andhigh pressures, respectively (Figures 2-4). Theelastic modulus-stress curves were expressed inmathematical terms as follows12,15,16:

M = kTq, 0 ≤ T ≤ TI, Part I,M = a + bT, TI ≤ T ≤ TII, Part II,M = c + dT, TII ≤ T ≤ Tf, Part III. (1)

The 3 parts were defined by determiningthe transition points among them. TI, TII, and Tf

denoted stress at the first transition point, stressat the second transition point, and maximumstress applied. The transition points, definingthe 3 parts, were manually selected as thosegiving the highest correlation coefficient uponthe power and bilinear regression of the elasticmodulus-stress data of parts I, II, and III.Together with the determination of the 3 parts,

parameters k and q of part I, intercept a andslope b of part II, and intercept c and slope d ofpart III were computed in terms of equation (1).These were used to regress the stress-strain datain terms of power and biexponential relations,as previously described.12,15,16

Acknowledgments

The authors thank Ms Anna Agapaki forthe excellence of the histological preparationsand also Chrysoula Dosiou, MD, Divisionof Endocrinology and Metabolism, StanfordUniversity, for her helpful comments on themanuscript. This work was supported in partby a grant from the Center for ExperimentalSurgery, Foundation of Biomedical Research,Academy of Athens, Greece.

REFERENCES

1. Klein I, Ojamaa K: Thyroid hormone and the cardio-vascular system. N Engl J Med 344:501-509, 2001.

2. Ojamaa K, Klemperer JD, Klein I: Acute effects of thy-roid hormone on vascular smooth muscle. Thyroid6:505-512, 1996.

3. Obuobie K, Smith J, Evans LM, et al: Increased cen-tral arterial stiffness in hypothyroidism. J ClinEndocrinol Metab 87:4662-4666, 2002.

4. Mizuma H, Murakami M, Mori M: Thyroid hormoneactivation in human vascular smooth muscle cells:expression of type II iodothyronine deiodinase. CircRes 88:313-318, 2001.

5. Obuobie K, Smith J, John R, et al: The effects of thy-rotoxicosis and its treatment on central arterial stiff-ness. Eur J Endocrinol 147:35-40, 2002.

6. Czarkowski M, Hilgertner L, Powalowski T, et al: Thestiffness of the common carotid artery in patientswith Graves’ disease. Int Angiol 21:152-157, 2002.

7. Inaba M, Henmi Y, Kumeda Y, et al: Increased stiffnessin common carotid artery in hyperthyroid Graves’ dis-ease patients. Biomed Pharmacother 56:241-246, 2002.

8. Connors JM, Hedge GA: Feedback regulation of thy-rotropin by thyroxine under physiological conditions.Am J Physiol 240:308-313, 1981.

9. Gredilla R, Barja G, Lopez-Torres M: Thyroidhormone-induced oxidative damage on lipids, glu-tathione and DNA in the mouse heart. Free Radic Res35:417-425, 2001.

10. Sokolis DP, Boudoulas H, Kavantzas NG, et al: Amorphometric study of the structural characteristics

Moulakakis Descending Aorta in Hyperthyroidism 351

of the aorta in pigs using an image analysis method.Anat Histol Embryol 31:21-30, 2002.

11. Angouras D, Sokolis DP, Dosios T, et al: Effect ofimpaired vasa vasorum flow on the structure andmechanics of the thoracic aorta: implications for thepathogenesis of aortic dissection. Eur J CardiothoracSurg 17:468-473, 2000.

12. Sokolis DP, Zarbis N, Dosios T, et al: Post-vagotomymechanical characteristics and structure of the thoracicaortic wall. Ann Biomed Eng 33:1504-1516, 2005.

13. Wolinsky H, Glagov S: Structural basis for the staticmechanical properties of the aortic media. Circ Res14:400-413, 1964.

14. Wolinsky H, Glagov S: A lamellar unit of aorticmedial structure and function in mammals. Circ Res20:99-111, 1967.

15. Sokolis DP, Boudoulas H, Karayannacos PE:Assessment of the aortic stress-strain relation inuniaxial tension. J Biomech 35:1213-1223, 2002.

16. Sokolis DP, Kefaloyannis EM, Kouloukoussa M, et al:A structural basis for the aortic stress-strain relationin uniaxial tension. J Biomech 39:1651-1662, 2006.

17. Magner JA, Clark W, Allenby P: Congestive heart fail-ure and sudden death in a young woman with thyro-toxicosis. West J Med 149:86-91, 1988.

18. Kimura H, Kawagoe Y, Kaneko N, et al: Low effi-ciency of oxygen utilization during exercise in hyper-thyroidism. Chest 110:1264-1270, 1996.

19. Roach MR, Burton AC: The reason for the shape ofthe distensibility curves of arteries. Can J BiochemPhysiol 35:681-690, 1957.

20. Cox RH: Mechanics of canine iliac artery smoothmuscle in vitro. Am J Physiol 230:462-470, 1976.

21. Clark JM, Glagov S: Transmural organization of the arte-rial media: the lamellar unit revisited. Arteriosclerosis5:19-34, 1985.

22. Dobrin PB, Canfield TR: Elastase, collagenase, andthe biaxial elastic properties of dog carotid artery.Am J Physiol 247:124-131, 1984.

23. Van Bortel LM, Struijker-Boudier HA, Safar ME: Pulsepressure, arterial stiffness, and drug treatment ofhypertension. Hypertension 38:914-921, 2001.

24. Domanski M, Norman J, Wolz M, et al: Cardiovascu-lar risk assessment using pulse pressure in the firstnational health and nutrition examination survey(NHANES I). Hypertension 38:793-797, 2001.

25. Safar ME, London GM, Asmar R, et al: Recentadvances on large arteries in hypertension. Hyperten-sion 32:156-161, 1998.

26. Simons PC, Algra A, Bots ML, et al: Common carotidintima-media thickness and arterial stiffness: indica-tors of cardiovascular risk in high-risk patients. TheSMART Study (Second Manifestations of ARTerialdisease). Circulation 100:951-957, 1999.

27. Dart AM, Kingwell BA: Pulse pressure: a review ofmechanisms and clinical relevance. J Am Coll Cardiol37:975-984, 2001.

28. Giannattasio C, Failla M, Grappiolo A, et al:Progression of large artery structural and functional

alterations in Type I diabetes. Diabetologia 44:203-208, 2001.

29. Lehmann ED, Gosling RG, Sonksen PH: Arterial wallcompliance in diabetes. Diabet Med 9:114-119, 1992.

30. Giannattasio C, Mangoni AA, Failla M, et al: Impairedradial artery compliance in normotensive subjectswith familial hypercholesterolemia. Atherosclerosis124:249-260, 1996.

31. London GM, Marchais SJ, Safar ME, et al: Aortic andlarge artery compliance in end-stage renal failure.Kidney Int 37:137-142, 1990.

32. Farrar DJ, Bond MG, Riley WA, et al: Anatomic cor-relates of aortic pulse wave velocity and carotidartery elasticity during atherosclerosis progressionand regression in monkeys. Circulation 83:1754-1763, 1991.

33. Hirai T, Sasayama S, Kawasaki T, et al: Stiffness ofsystemic arteries in patients with myocardial infarc-tion: a noninvasive method to predict severity ofcoronary atherosclerosis. Circulation 80:78-86, 1989.Erratum in: Circulation 80:1946, 1989.

34. Laurent S, Boutouyrie P, Asmar R, et al: Aortic stiff-ness is an independent predictor of all-cause and car-diovascular mortality in hypertensive patients.Hypertension 37:1236-1241, 2001.

35. Booth RF, Martin JF, Honey AC, et al: Rapid developmentof atherosclerotic lesions in the rabbit carotid arteryinduced by perivascular manipulation. Atherosclerosis76:257-268, 1989.

36. Giannattasio C, Failla M, Emanuelli G, et al: Localeffects of atherosclerotic plaque on arterial distensi-bility. Hypertension 38:1177-1180, 2001.

37. Wilkinson IB, Prasad K, Hall IR, et al: Increased cen-tral pulse pressure and augmentation index in sub-jects with hypercholesterolemia. J Am Coll Cardiol39:1005-1011, 2002.

38. Wilkinson I, Cockcroft JR: Cholesterol, lipids andarterial stiffness. Adv Cardiol 44:261-277, 2007.

39. Tsiara S, Elisaf M, Mikhailidis DP: Early vascular ben-efits of statin therapy. Curr Med Res Opin 19:540-556, 2003.

40. Havlik RJ, Brock D, Lohman K, et al: High-densitylipoprotein cholesterol and vascular stiffness at base-line in the activity counseling trial. Am J Cardiol87:104-107, 2001.

Angiology Volume 58, Number 3, 2007352