surgical pathology of the ascending aorta: a clinicopathologic study of 513 cases

Surgical Pathology of the Ascending Aorta:A Clinicopathologic Study of 513 Cases

James L. Homme, MD,* Marie-Christine Aubry, MD,w William D. Edwards, MD,wStephanie M. Bagniewski, BS,z Vernon Shane Pankratz, PhD,z Catherine A. Kral, MD,*

and Henry D. Tazelaar, MDy

Abstract: Only 2 comprehensive surgical series, published in

1977 and 1983, have evaluated clinicopathologic features of the

ascending aorta. Retrospective review of medical records and

microscopic slides was performed on 513 consecutive patients

with surgical resection of ascending aortic tissue (1985 to 1999).

Patients were 2 to 89 years old (mean 59 y), and 303 (59%) were

men. Aneurysm or dissection was the indication for surgery in

479. Aortic valves were also excised in 360. Systemic hyperten-

sion was present in 279, inherited connective tissue disease

(ICTD) in 67, arteritis in 33, and acquired connective tissue

disease in 16. Microscopy showed cystic medial degeneration

(CMD) in 209, aortic dissection (AD) in 109 (with CMD in 56),

normal media in 90, aortitis in 57 (with CMD in 14), and other

findings in 48. The most significant, independent risk factor of

CMD and AD was ICTD (confidence interval=7.61 and 2.26,

respectively). Systemic hypertension was more common in

patients with AD than without (P=0.0202). Normal media

was the most common histologic finding associated with

bicuspid aortic valve (P<0.0001). Among 57 patients with

aortitis (giant cell in 39), ages ranged from 16 to 85 years (mean

64 y), and 42 (74%) were women; only 8 had Takayasu arteritis,

and 11 had temporal or systemic arteritis. In surgically resected

ascending aorta, the 3 most common histologic findings were

CMD, AD, and normal media. ICTD, systemic hypertension,

and bicuspid aortic valve were common comorbid findings.

Giant cell aortitis occured predominantly in women, usually

without systemic disease.

Key Words: aortic aneurysm, aortic dissection, aortitis, cystic

medial degeneration

(Am J Surg Pathol 2006;30:1159–1168)

Replacement of the ascending aorta is currently themost commonly performed surgical procedure on the

thoracic aorta.40 At our institution, an increasing numberof cases with giant cell aortitis or histologically normalaortas were noticed in the past 10 years, and no similarobservation could be identified in the literature. Indeed,most descriptions of the pathology of the ascending aorta,on the basis of autopsy findings before 1965, haveemphasized syphilitic aortitis and mycotic aneurysms.7,16

Subsequently, only 2 large studies published between1975 and 1985 have evaluated in detail the microscopicfeatures of surgically resected ascending aortic tissues.Both sets of authors reported cystic medial degeneration(CMD) as the most common finding, with a relatively lowincidence of aortitis and no description of normalmedia.20,28

No comprehensive surgical pathology study of theascending aorta has been reported in the past 20 years.Instead, most recent studies have focused on specificdiseases, such as dissection or aortitis, rather than thewhole spectrum of disorders.

With the above considerations in mind, the currentinvestigation was undertaken to review the clinicopatho-logic features of a large number of consecutively excisedsurgical specimens of ascending aorta from a singleinstitution.

MATERIALS AND METHODS

Study GroupInstitutional surgical and pathology databases were

searched for patients who had undergone ascendingaortic surgery between January 01, 1985 and December31, 1999. Histologic slides were requested from theinstitutional tissue registry. If slides were unavailable,gross specimens were obtained, and tissues were pro-cessed for microscopy. The study group consisted of 513patients, excluding 10 cases with unavailable gross andmicroscopic material.

Clinical DataFor each patient, medical charts were reviewed

retrospectively and the following data recorded: age atoperation, gender, race, and history of arteritis, inheritedor acquired connective tissue disease (ACTD), inheritedor acquired cardiovascular disease, and previous cardiacCopyright r 2006 by Lippincott Williams & Wilkins

From the *Mayo Medical School; wDepartment of Laboratory Medicineand Pathology; zDivision of Biostatistics, Mayo Clinic, Rochester,MN; and yDepartment of Laboratory Medicine and Pathology,Mayo Clinic, Arizona.

Financial Support: In part by the Mayo Medical School, Mayo Collegeof Medicine.

Platform Presentation: At the annual meeting of the United States-Canadian Academy of Pathology, Atlanta, GA, March 2001.

Reprints: Marie-Christine Aubry, MD, Mayo Clinic, 200 First StreetS.W., Rochester, MN 55905 (e-mail: [email protected]).

ORIGINAL ARTICLE

Am J Surg Pathol � Volume 30, Number 9, September 2006 1159

or aortic surgery. The indication for surgical resection ofthe ascending aorta was categorized as aneurysm ordissection, infection, trauma, or other. Aortic valvestructure and function were obtained from echocardio-graphic studies. Maximum ascending aortic diameter hasbeen determined by echocardiography, thoracic com-puted tomography, or magnetic resonance imaging.

Microscopic DataFor each case, at least one block of ascending aortic

tissue had been prepared and stained with hematoxylinand eosin (H&E) and Verhoeff-van Gieson (VVG). Thenumber of aortic pieces and the total length of all aorticsegments were recorded for each case. For those showingmedial lymphoplasmacytic infiltrates, slides stained withH&E were prepared from 3 additional levels andexamined for giant cells. If identified, cases werereclassified as giant cell aortitis.

Presence or absence of each of the following medialprocesses was evaluated for each patient: CMD, laminar

medial necrosis (LMN), intramedial dissection, aortitis,no diagnostic abnormality, and other. CMD was definedas fragmentation and/or loss of elastic fibers associatedwith or without pools of glycoproteins (Fig. 1).23 Inpatients Z50 years old, CMD was considered presentonly if >5% of the medial thickness was involved. LMNwas defined as coagulative necrosis of medial smoothmuscle cells, resulting in loss of nuclei and collapse of theelastic lamellae (Fig. 2). Aortic dissection (AD) wasconsidered recent if the false lumen contained freshthrombus, acute inflammation, or granulation tissue, andas healed if it contained a fibrotic or atheroscleroticneointima (Fig. 3).

Aortitis, if present, was characterized as giant cell,lymphoplasmacytic, infectious, or healed. Giant cellaortitis was diagnosed when 1 or more giant cells wereidentified within medial inflammation (Fig. 4A). Co-existence of giant cell aortitis and CMD in similarproportions was classified as combined disease (CMD-GCA). For a diagnosis of lymphoplasmacytic aortitis, nogiant cells or granulomas could be present, even afterevaluation of the additional levels. Aortitis was furthercategorized as chronic if medial inflammation wasaccompanied by fibrosis. Healed aortitis was defined aspatchy areas of medial replacement fibrosis without activeinflammation (Fig. 4B).

With the exception of AD, the length of thespecimen involved in each process was measured andevaluated as a percentage of overall length. Transmuralextent of disease was designated as mild, moderate, orsevere, when 0% to 25%, 26% to 75%, or >75%respectively, of the medial thickness was involved. Inaddition, the overall extent of disease was expressedsemiquantitatively as the percentage of abnormal tissue.

For each case, the intima was evaluated for thepresence of atherosclerosis or a nonatheroscleroticfibroproliferative process. Atherosclerosis was furtherqualified as fibrotic, ulcerative, or calcific.

FIGURE 1. CMD. A, Pools of glycosaminoglycans with 25%transmural extent. B, Associated loss of elastic fibers. (A, H&E100� ; B, VVG 100� .)

FIGURE 2. Laminal medial necrosis. Linear region character-ized by nuclear loss and coagulative necrosis of smooth musclecells (H&E 100� ).

Homme et al Am J Surg Pathol � Volume 30, Number 9, September 2006

1160 r 2006 Lippincott Williams & Wilkins

Statistical AnalysisSummaries of the data were obtained for the

variables of interest. These summaries consisted of countsand percentages for nominal entities, and means andstandard deviations (mean±SD), as well as medians andranges when appropriate, for quantitative measurements.Differences in percentages were tested using Pearson w2

statistic, and differences in mean values were primarily

tested using 2-sample t tests. When the means ofquantitative variables were compared between more than2 groups, analysis of variance methods were used in placeof a t test.

In addition to these simple summaries and compar-isons, multivariate models were built to determine whichvariables were independently associated with AD and theseverity of CMD. For AD, a backward stepwise selectionlogistic regression model was obtained by eliminating allvariables with a multivariate-adjusted P value greaterthan 0.10. For CMD, again a backward stepwiseregression model was built by sequentially eliminatingvariables with a multivariate-adjusted P value greaterthan 0.10. In this instance, a logistic regression modelappropriate for an ordered response was used.

All tests of significance were 2-sided, and P valuesless than 0.05 were considered to be statistically sig-nificant. Analyses were performed using SAS and JMP(Cary, NC).

RESULTS

General Clinical FeaturesThe study group included 303 men (59%) and 210

women, with mean age of 59 years (Tables 1 and 2).Surgical indications were aortic aneurysm or dissection in479 patients (93%). Among the 457 patients for whommeasurements were available, the maximum diameter ofthe ascending aorta ranged from 2.0 to 17.5 cm, with amean of 5.9 cm and a SD of 1.5 cm. Concomitant surgerywas performed on 417 patients (81%), mostly men(P=0.0035), and included aortic valve surgery in 360patients (70%), mainly for moderate to severe aorticregurgitation in 252 (70%).

Overall, 460 patients (90%) had one or more formsof acquired cardiovascular disease, including systemichypertension in 279 and atherosclerotic coronary arterydisease in 129. Patients with acquired cardiovasculardisease were, on average, older (60 y, SD=16.5) thanthose without (mean±SD=51±18, P=0.0003). Con-genital heart disease affected 155 patients (30%). Bicuspidaortic valve (BAV) was the most frequent anomalyand affected 127 patients, 17 of whom also hadother coexistent congenital cardiovascular lesions.BAV affected relatively younger patients (mean±SD=53±17 y) and 80% were men (P<0.0001) (Fig. 5).It was associated more frequently with histologicallynormal media, in contrast to patients without BAV(40% vs. 10%, P<0.0001).

Inherited connective tissue disease (ICTD) affected67 patients (13%). It was confirmed genetically as Marfansyndrome in 32 and as Ehlers-Danlos in 2 and wasclinically suspected but not genetically confirmed inanother 33 patients. The average age of these patientswas 40 years (SD=15), which was significantly youngerthan those without ICTD (P<0.0001). Sixty patients(90%) also had CMD or AD, which was significantlyhigher than the 63% prevalence observed in those withoutICTD (P<0.0001).

FIGURE 3. Healed dissection. The dissection, involving theouter third of the media, contains fibrosis, which correspondsto the linear disruption of the media (VVG 20� ).

FIGURE 4. Giant cell aortitis. A, Active aortitis with severalmultinucleated giant cells present within the medial inflam-mation. B, Healed aortitis characterized by patchy medialfibrosis without significant inflammation. (A, H&E 100� ; B,VVG 100� .)

Am J Surg Pathol � Volume 30, Number 9, September 2006 Surgical Pathology of Ascending Aorta

r 2006 Lippincott Williams & Wilkins 1161

ACTD occurred in 16 patients (3%), 8 men and 8women, and included 9 with rheumatoid arthritis, 6 withspondyloarthropathy (3 psoriatic arthritis, 2 ankylosingspondylitis, and 1 Reiter syndrome), and 1 with systemiclupus erythematosus and Raynaud disease. Men were

diagnosed with ACTD at significantly younger ages(52±19 y) than women (71±6.0 y, P=0.018).

History of arteritis existed in 33 (6%), was moreoften seen in women (P<0.0001) and in patients withaortitis and CMD-GCA (P<0.0001), and included

TABLE 1. Clinical Summary of Study Group of 513 Patients

Women Men Total

N=210 N=303 N=513 P

Age (y)Mean (SD) 61.4 (18.5) 57.6 (16.0) 59.2 (17.2) 0.0001

Median 68.0 61.0 64.0Range 2.0 to 87.0 5.0 to 89.0 2.0 to 89.0

Concomitant surgery 158 (75.2%) 259 (85.5%) 417 (81.3%) 0.0035

AV surgery 130 (61.9%) 230 (75.9%) 360 (70.2%) 0.0007

CABG 29 (13.8%) 70 (23.1%) 99 (19.3%) 0.0087

ACVD 191 (91.0%) 269 (88.8%) 460 (89.7%) 0.4264HTN 125 (59.5%) 154 (50.8%) 279 (54.4%) 0.0518CAD 44 (21%) 85 (28.1%) 129 (25.1%) 0.0683CHF 36 (17.1%) 29 (9.6%) 65 (12.7%) 0.0112

MI 14 (6.6%) 32 (10.6%) 46 (9%) 0.1290Cong HD 41 (19.5%) 114 (37.6%) 155 (30.2%) <0.0001

BAV 25 (11.9%) 102 (33.7%) 127 (24.8%) <0.0001

Cong HD other 20 (9.5%) 24 (7.9%) 44 (8.6%) 0.5238ICTD 24 (11.4%) 43 (14.2%) 67 (13.1%) 0.3612ACTD 8 (3.8%) 8 (2.6%) 16 (3.1%) 0.4538Arteritis 28 (13.3%) 5 (1.7%) 33 (6.4%) <0.0001

ACVD indicates acquired cardiovascular disease; AV, aortic valve; CABG, coronary artery bypass graft; CAD, coronaryartery disease; Cong HD, congenital heart disease; CHF, congestive heart failure; HTN, hypertension; MI, myocardial infarct.

TABLE 2. Clinical Features Versus Primary Aortic Medial Process (513 Patients)

Clinical Feature CMD AD Normal Aortitis CMD-GCA Other Total P

PatientsNumber 209 109 90 43 14 48 513Percentage 40.7 21.2 17.5 8.4 2.8 9.4 100

Age (y)Mean (SD) 59.6 (17.5) 63.1 (14.1) 53.5 (16.5) 62.1 (19.3) 71.2 (7.6) 52.9 (19.1) 59.2 (17.2) <0.0001

Median 65.0 67.0 53.5 70.0 72.0 56.5 64.0Range 6.0 to 89.0 22.0 to 87.0 5.0 to 82.0 16.0 to 85.0 50.0 to 84.0 2.0 to 80.0 2.0 to 89.0

SexF (%) 79 (37.8) 40 (36.7) 31 (34.4) 31 (72.1) 11 (78.6) 18 (37.5) 210 (40.9) <0.0001

M (%) 130 (62.2) 69 (63.3) 59 (65.6) 12 (27.9) 3 (21.4) 30 (62.5) 303 (59.1)F:M ratio 1:1.7 1:1.7 1:1.9 2.6:1 3.7:1 1:1.7 1:1.4

Concomitant surgeryN (%) 182 (87.1) 61 (56) 87 (96.7) 38 (88.4) 6 (42.9) 43 (89.6) 417 (81.3) <0.0001

AV surgery 169 (80.9) 50 (45.9) 73 (81.1) 30 (69.8) 5 (35.7) 33 (68.8) 360 (70.2) <0.0001

CABG 34 (16.3) 13 (11.9) 24 (26.7) 16 (37.2) 2 (14.3) 10 (20.8) 99 (19.3) 0.0039

Medical historyACVD 181 (86.6) 94 (86.2) 86 (95.6) 39 (90.7) 12 (85.7) 48 (100) 460 (89.7) 0.001

Hypertension 118 (56.5) 70 (64.2) 33 (36.6) 26 (60.5) 9 (64.3) 23 (47.9) 279 (54.4) 0.0029

CAD 40 (19.1) 33 (30.3) 27 (30.0) 15 (34.9) 2 (14.3) 12 (25.0) 129 (25.1) 0.0842CHF 31 (14.8) 11 (10.1) 8 (8.9) 8 (18.6) 1 (7.1) 6 (12.5) 65 (12.7) 0.4856MI 15 (7.2) 17 (15.6) 9 (10.0) 3 (7.0) 0 (0) 2 (4.2) 46 (9.0) 0.0827

CHD 58 (27.8) 14 (12.8) 59 (65.6) 3 (7.0) 0 (0) 21 (43.8) 155 (30.2) <0.0001

Bicuspid AV 47 (22.5) 12 (11.0) 51 (56.7) 3 (7.0) 0 (0) 14 (29.2) 127 (24.8) <0.0001

Other CHD 18 (8.6) 4 (3.7) 14 (15.6) 0 (0) 0 (0) 8 (16.7) 44 (8.6) 0.004

CTDInherited 42 (20.1) 18 (16.5) 3 (3.3) 0 (0) 0 (0) 4 (8.3) 67 (13.1) 0.0002

Acquired 7 (3.3) 2 (1.8) 3 (3.3) 2 (4.7) 2 (14.3) 0 (0) 16 (3) 0.132Arteritis 9 (4) 4 (4) 1 (1) 14 (33) 5 (36) 0 (0) 33 (6) <0.0001

ACVD indicates acquired cardiovascular disease; AV, aortic valve; CABG, coronary artery bypass graft; CAD, coronary artery disease; CHD, congenital heart disease;CHF, congestive heart failure; CTD, connective tissue disease; F, female; CMD-GCA, cystic medial degeneration and giant cell aortitis; M, male; MI, myocardial infarct.



temporal arteritis in 21, Takayasu arteritis in 9, poly-arteritis nodosa in 1, and immune complex arteritis in 1.Women with arteritis were younger on average than men(60±23 and 70±7, respectively), although this differ-ence did not achieve statistical significance (P=0.382).

A single patient had a history of syphilis andhistologically had acute AD and CMD, without aortitis.

General Pathology FeaturesThe mean±SD for the total number of aortic

sections (pieces) examined per case was 6±3 (range 1 to24), with total mean±SD length of 12±7.0 cm (range0.4 to 59.8 cm). The mean±SD for the percent of tissueuninvolved by a medial process, excluding patients with

histologically normal aorta, was 38±41%. For indivi-dual patients, some sections showed no medial abnorm-ality, whereas others were completely abnormal. Thedominant medial process was CMD in 209 patients(41%), followed by AD in 109 (21%) (Table 2).

There were significant differences in age amongindividuals with different primary medial processes(P<0.0001, Table 2). Patients with normal aortic tissuewere the youngest (mean±SD=54±17 y) and thosewith CMD-GCA were the oldest (mean±SD=71±8y). There was also a significant difference in thegender distributions for the various microscopic patterns(P<0.0001, Table 2). Overall, more men than womenunderwent surgical resection of ascending aortic tissue(59% of the 513 subjects). However, the majority of thosewith CMD-GCA (79% of 14 subjects) and aortitis (72%of 43 subjects) were women.

CMDComorbid factors commonly associated with CMD

(systemic hypertension, BAV, and ICTD) were foundin 208 (99%) of the 209 patients, and were multiple in76 (36%). The mean extent of tissue involved by CMDwas 8.6 cm with mean involvement of 68%. On the basisof the extent of transmural involvement, primary CMDwas severe in 129 patients (61%), moderate in 41, andmild in 40.

Because 81 patients had tissue involved by CMD asa secondary medial process, additional analyses wereperformed on a total of 290 patients. Table 3 containssimple comparisons between several clinical features andCMD severity. In the univariate analysis, factors with thestrongest associations with CMD and its severity includedcongenital heart disease (P=0.0020), in particular BAV

FIGURE 5. Mean ages, with 1 SD intervals, for men andwomen who have comorbid factors commonly associatedwith disease of the ascending aorta (ACVD indicates acquiredcardiovascular disease).

TABLE 3. Clinical Features of Patients With CMD

No CMD Mild CMD Moderate CMD Severe CMD

N=223 N=69 N=61 N=160 P

Age 0.0977Mean (SD) 57.5 (17.1) 59.4 (17.1) 59.6 (17.6) 61.3 (17.1)Median 60.0 63.0 64.0 67.5Range 2.0 to 82.0 6.0 to 87.0 9.0 to 83.0 14.0 to 89.0

Sex 0.1722F (%) 94 (42.2) 20 (29.0) 25 (41.0) 71 (44.4)M (%) 129 (57.8) 41 (71.0) 36 (59.0) 89 (55.6)

Concomitant surgery 184 (82.5) 58 (84.1) 46 (75.4) 129 (80.6) 0.5726AV surgery 146 (65.5) 53 (76.8) 41 (67.2) 120 (75.0) 0.1194CABG 54 (24.2) 9 (13.0) 12 (19.7) 24 (15) 0.0688

ACVD 210 (94.2) 64 (92.8) 55 (90.2) 131 (81.9) 0.0011

HTN 113 (50.7) 40 (58.0) 42 (68.9) 84 (52.5) 0.0728CAD 74 (33.2) 12 (17.4) 15 (24.6) 28 (17.5) 0.0020

CHF 26 (11.7) 8 (11.6) 11 (18) 20 (12.5) 0.6003MI 23 (10.3) 5 (7.2) 7 (11.5) 11 (6.9) 0.5572

Congenital HD 88 (39.5) 31 (44.9) 14 (23) 22 (13.8) <0.0001

BAV 73 (32.7) 25 (36.2) 12 (19.7) 17 (10.6) <0.0001

Other 23 (10.3) 10 (14.5) 2 (3.3) 9 (5.6) 0.0481

ICTD 10 (4.5) 5 (7.2) 13 (21.3) 39 (24.4) <0.0001

ACTD 4 (1.8) 3 (4.3) 4 (6.6) 5 (3.1) 0.2585

ACVD indicates acquired cardiovascular disease; AV, aortic valve; CABG, coronary artery bypass graft; CAD, coronary arterydisease; CHF, congestive heart failure; CTD, connective tissue disease; F, female; HTN, hypertension; M, male; MI, myocardial infarct.



(P<0.0001), in which CMD was usually either absent ormild. In contrast, patients with ICTD had more severeCMD (P<0.0001). There was an insignificant trendtoward a positive association between CMD severity andage (P=0.0977).

Following the multivariate analysis that jointlyexamined multiple correlates of CMD severity, 3 vari-ables were statistically significant. These included ICTD,BAV, and age. Of these, ICTD was the most stronglyassociated with CMD severity. Subjects with ICTD hadan odds of an increased CMD severity that was nearly 8times that of subjects without ICTD [P<0.0001; OR,95% confidence interval (CI)=7.61, 4.11-14.12]. Whilecontrolling for all other variables in the model, the oddsof higher CMD severity was less for those with BAV(P=0.0102; OR=0.58, 95% CI=0.39-0.88) than forthose without BAV. The odds for increased CMDseverity was higher for those of higher age; each 10-yearincrease in age was associated with a 31% increase in theodds for higher CMD severity (P<0.0001; OR, 95%CI=1.31, 1.30-1.33).

ADAmong 109 patients, 13 (12%) had a past history of

AD, 8 of whom had prior surgical repair. Another 13 hadpreviously undergone coronary artery bypass surgery.

Comorbid risk factors were identified in 97 patients,and hypertension was the most common, affecting 70patients (Tables 2 and 4). In the univariate analysis,hypertension was more commonly seen in patients with

AD than without (P=0.0202). In contrast, AD was lesscommonly seen in patients with BAV.

Five variables were retained in the multivariateanalysis that jointly examined multiple correlates ofAD. These variables included BAV, ICTD, hypertension,age, and sex. The strongest association was between BAVand AD, and patients with BAV had a third of the oddsof having AD (P=0.0022; OR, 95% CI=0.35, 0.18-0.68) compared with patients without BAV. Whilecontrolling for all other variables, the odds of ADfor patients with ICTD were 2.26 (P=0.0339, 95%CI=1.06-4.81) times that of patients without ICTD.Similarly, the odds of AD for patients with hypertensionwere 1.56 (P=0.0709, 95% CI=0.96-2.52) times thatof patients without hypertension, but this did not reachstatistical significance. Men had a 69% (P=0.0263; OR,95% CI=1.69, 1.06-2.69) increased odds of AD, andeach 10-year increase in age was associated with a 24%increase in the odds of AD (P=0.0199; OR, 95%CI=1.24, 1.03-1.48).

ADs were acute in 57 patients (52%), healed in 45,and both acute and healed in 8. CMD was present in 55AD patients (50%), and was severe in 17.

Normal MediaFor the 90 patients with histologically normal

media, 66% were men and the mean±SD age was54±17 years (Table 2). Among these, 81% underwentconcomitant aortic valve surgery, for pure regurgitationin 34 of the 73. BAV was the most common comorbidfactor in these subjects, and was significantly morefrequent than in those without normal media(P<0.0001, 57% and 18%, respectively). Hypertensionwas significantly less common than with other medialprocesses (P=0.0004, 37% and 58%, respectively).ICTD was suspected in only 3 patients, and ACTD inanother 3. A single patient had a history of Takayasuarteritis.

The mean maximum ascending aortic diameter was4.9 cm (range 2.9 to 6.6 cm). Atherosclerosis was identi-fied in only 11 specimens (12%) and was diffuse andcalcific in only 6.

AortitisAmong 43 patients with arteritis, coexistent sys-

temic arteritis included Takayasu arteritis in 8 patientsand temporal arteritis in 6 (Table 5). Two patients hadan ACTD (rheumatoid arthritis and Reiter syndrome).Of the 6 patients with infectious aortitis, blood ortissue cultures were positive in each (4 streptococcus,1 staphylococcus, and 1 salmonella). Four of the 6patients also underwent concomitant replacement of aninfected and severely regurgitant aortic valve. Althoughsubjects with aortitis were most frequently women(72% of 43 cases), the reverse was true for infectiousaortitis, which primarily affected men (83% of 6 cases)(P=0.0124). Among 6 patients with healed aortitis,4 were women, and the mean±SD age was 57±15years. Two patients had a history of Takayasu arteritis.

TABLE 4. Clinical Features of Patients With AD

No AD AD

N=404 N=109 P

Age (y) 0.0169

Mean (SD) 58.1 (17.8) 63.1 (14.1)Median 63.5 67.0Range 2.0 to 89.0 22.0 to 87.0

Sex 0.3105F (%) 170 (42.1) 40 (36.7)M (%) 234 (57.9) 69 (63.3)

Concomitant surgery 356 (88.1) 61 (56) <0.0001

AV surgery 310 (76.7) 50 (45.9) <0.0001

CABG 86 (21.3) 13 (11.9) 0.0280

ACVD 366 (90.6) 94 (86.2) 0.1849HTN 209 (51.7) 70 (64.2) 0.0202

CAD 96 (23.8) 33 (30.3) 0.1643CHF 54 (13.4) 11 (10.1) 0.3617MI 29 (7.2) 17 (15.6) 0.0063

Congenital HD 141 (34.9) 14 (12.8) <0.0001

BAV 115 (28.5) 12 (11) 0.0002

Other 40 (9.9) 4 (3.7) 0.0392

ICTD 49 (12.1) 18 (16.5) 0.2279ACTD 14 (3.5) 2 (1.8) 0.3848Arteritis 29 (7.2) 4 (3.7) 0.1852

ACVD indicates acquired cardiovascular disease; AV, aortic valve; CABG,coronary artery bypass graft; CAD, coronary artery disease; CHF, congestiveheart failure; CTD, connective tissue disease; F, female; HTN, hypertension; M,male; MI, myocardial infarct.



The mean maximum ascending aortic diameter was5.4 cm (range 2.8 to 10.0 cm). The most commonhistologic type of aortitis was giant cell (58%)(Table 5). The 25 cases of giant cell aortitis included5 initially diagnosed as lymphoplasmacytic aortitis, inwhich medial giant cells were detected in deeper sections.Medial inflammation was severe in 88% and involved92% of the tissue specimens. Lymphoplasmacytic aortitisaffected 6 patients and was severe in 83% and involved97% of the pieces examined.

Noninflammatory features were seen in 17 patients(13 with LMN, 3 with CMD, and 1 with both processes).Medial degeneration was mild to moderate and involved18% of the sections (range 8% to 29%). Intimalatherosclerosis was focal in 11 patients and diffuse in 7.

CMD and AortitisAscending aortic tissue from 14 patients showed a

similar extent of medial degeneration and giant cellaortitis. Gender distribution was similar to patients withaortitis only, with women being predominantly affected(Table 2). Five had a history of temporal (giant cell)arteritis, and 2 had an ACTD.

Microscopically, CMD was severe in 12 patientsand moderate in 2, and giant cell aortitis was severe in11 and moderate in 3. All 14 had focal LMN. Intimalatherosclerosis was present in 5 patients and was focal in3 and diffuse in 2.

Other Aortic DiseaseLMN was the sole abnormality in 2 patients and

coexisted with other lesions in 98 (36 with CMD, 23 AD,14 CMD-GCA, 13 aortitis, and 12 with other abnorm-alities).

Other processes represented the principal micro-scopic abnormality in 46 cases. Findings were consistentwith previous surgery in 32 patients and includedfibrocalcific tissue and graft or suture material in 26 andtissue from a false aneurysm in 6. Diffuse calcific

atherosclerosis was the main microscopic finding in6 patients, 2 of whom had penetrating atheroscleroticulcers.

DISCUSSIONThe data presented in the current investigation

represents the only comprehensive surgical pathologystudy of the ascending aorta published in the past 20years. In comparison to previous surgical pathology seriespublished by Pomerance et al in 197728and by Klimaet al20 in 1983, the present series has less malepreponderance and somewhat older patients20 (Table 6).Both trends can be attributed primarily to the relativelylarge number of older women with aortitis in the currentseries. As in previous series, CMD was the most commonhistologic finding, followed by AD; however, in contrastto previous series, histologically normal media was asignificant finding in our study.

CMDMen are affected by CMD at least twice as often as

women. One autopsy and 3 surgical series showed asimilar male predilection of 68% to 84%.20,25,28,34 In thecurrent study, men accounted for 62% of all patients withmedial degeneration and for 87% of those who also had aBAV.

Reported risk factors for CMD have includedMarfan syndrome, increasing age, aortic valve disease,and systemic hypertension. However, some authors havedescribed severe CMD in the ascending aortas of patientswith Marfan syndrome,20 whereas others have not.31,34 Inour study, patients with ICTD were at greatest risk ofdeveloping severe CMD, independent of other factors.Klima et al20 emphasized an inverse relationship betweenthe severity of CMD and age, and Savunen et al34 did notshow significant difference in the severity of CMD inolder patients. In contrast, our study and that ofSchlatman and Becker35demonstrated increased severityof CMD in older patients.

TABLE 5. Clinical Features Versus Histologic Type of Aortitis (43 Patients)

Clinical Feature Giant Cell Lympho-plasmacytic Infectious Healed Total

PatientsNumber 25 6 6 6 43Percentage 58 14 14 14 100

Age (y)Mean 62 63 65 57 62Range 19 to 82 16 to 80 37 to 85 38 to 76 16 to 85

SexF (%) 21 (84) 5 (83) 1 (17) 4 (67) 31 (72)M (%) 4 (16) 1 (17) 5 (83) 2 (33) 12 (28)F:M ratio 5.2:1 5.0:1 1:5.0 2.0:1 2.6:1

Systemic arteritisGiant cell 6 (24) 0 (0) 0 (0) 0 (0) 6 (14)Takayasu 5 (20) 1 (17) 0 (0) 2 (33) 8 (19)None 14 (56) 5 (83) 6 (100) 4 (67) 29 (67)

ACTD 1 (5) 0 (0) 0 (0) 1 (17) 2 (5)

F indicates female; M, male.



Aortic valve disease is commonly described inpatients with aneurysms of the ascending aorta andCMD,20 and as in our study, BAV is the most commonvalvular abnormality.33 A developmental defect not onlyof the aortic valve but also of the entire ascending aorta isthought to explain this association.2,6,40 In our study,BAV was not associated with an increased risk ofdeveloping severe CMD and was commonly associatedwith histologically normal media. In fact, the degree ofCMD in patients with regurgitant tricuspid aortic valvewas not significantly different, as also reported in anotherstudy.2 It is therefore probable that mechanisms otherthan morphologic abnormalities underlie the associationof BAV and ascending aortic aneurysms.

Finally, hypertension has been correlated with thesize of aortic aneurysms and thus plays a role in CMD.However, several studies have shown a lower or similardegree of CMD in hypertensive patients as in normoten-sive patients.20,35 In our study, hypertension did not seemto predict for the presence of CMD.

In the current investigation, the extent of medialdegeneration varied considerably from site to site withinindividual specimens. Frequently, even among patientswith severe disease, some sections showed no disease oronly mild abnormality. This emphasized the possibility ofsampling error. Accordingly, it is recommended thatmultiple pieces be taken from different areas of thespecimen. On the basis of our study, we recommend anaverage of 6 sections of ascending aortic tissue, eachabout 2 cm in length.

ADAD was the second most common histologic finding

in our series, similar to previous series.20,28 However, theincidence of 21% was lower than previously reported.Indeed, dissection accounted for 43% of the resectedascending aortic aneurysms described by Frist and

Miller12 and affected 39% of those reported by Klimaet al.20 However, a study of 93 patients from theCleveland Clinic showed dissection in 30%.25 And itrepresented only 29% of the cases of Pomerance et al28

and 21% of the present cases. These latter figures andours are attributable primarily to the higher percentageof patients with aortitis (22% in Pomerance et al28 and11% in the present study vs. 1% in Klima et al20 and notenumerated by Frist and Miller12).

Dissection most commonly involves theascending aorta and affects men twice as often aswomen14,23,27,31,39and in our study, male gender provedto be an independent risk factor for dissection.

Hypertension is the most commonly reported riskfactor for ascending AD and affects approximately 70%of patients.5,14,21,23,27,29,39 It was documented in 64% ofthe current patients and was significantly more commonin patients with dissection than without. Although therewas a trend for systemic hypertension to represent anindependent predictor of risk for dissection, this did notreach a statistical difference in our study.

Because of its rarity, Marfan syndrome generallyaccounts for only 3% to 5% of all patients withAD.14,23,39 The higher frequencies of 10% reported byNakashima et al,27 13% by Bachet et al,1 and 14% in thecurrent series may be related to referral bias. The risk ofdissection in patients with Marfan syndrome has beenattributed to concomitant hypertension.31 However, inour study, patients with ICTD represented the highestrisk population of developing dissection, independent ofother risk factors.

In addition, among patients with ascending AD,BAV has been reported in 8% to 14% of autopsy seriesand affected 11% in the present surgical series.23,30 In onestudy, the risk for dissection was reported to be 9 timeshigher in patients with BAV than those with tricuspidaortic valve23 and thus BAV was considered an important

TABLE 6. Comparison of Current Study With 2 Surgical Pathology Series

PublicationAuthors Homme et al Klima et al20 Pomerance et al28

Year published 1983 1977Years studied 1985 to 1999 1956 to 1977 1969 to 1976

PatientsNumber 514 339 63Age (y)Mean 59 50 63Range 2 to 89 19 to 87 21 to 74

SexMale (%) 303 (59) 232 (68) 51 (81)Female (%) 211 (41) 107 (32) 12 (19)

Microscopy (%)*CMD 279 (54) 238 (70) 45 (71)Dissection 109 (21) 133 (39) 18 (29)Normal 90 (18) 0 (0) 2 (3)Aortitis 57 (11) 3 (1) 14 (22)Atherosclerosis 6 (1) w 2 (3)Other 42 (8) 0 (0) 0 (0)

*Totals may exceed 100% because some patients had 2 major microscopic features.wAtherosclerosis affected 139 patients but was the major abnormality in an unknown number.



risk factor for dissection. In our study, in the multivariateanalysis, BAV was less likely to be associated withdissection than patients without, which seemed to supportstudies suggesting that BAV may not be a significantrisk factor.27,39 However, BAV still occurred 7 timesmore commonly among patients with AD, in comparisonto the 1.5% incidence in the general population. BAV istherefore likely to represent a risk factor for AD, andaneurysm of the ascending aorta.

CMD affected 51% of cases of AD in the presentstudy, and was usually focal and mild. However, inpatients with an ICTD, CMD was more often moderateor severe. These findings support the interpretation ofothers that CMD is primarily a risk factor for AD inpatients with ICTD.19,27,36,39

Histologically Normal AortaResected ascending aortic tissue without appreci-

able abnormality was reported in only 3% of the cases ofPomerance et al28 and in none of the cases of Klimaet al.20 In contrast, it represented the third most frequentcategory (18%) in the current study. Patients with thishistology were the youngest, and men accounted for 64%.Interestingly, 57% had a BAV, significantly more thanwith any other histologic finding. This would support atheory of a developmental defect involving both the aorticvalve and the entire ascending aorta, which might nothave obvious morphologic abnormalities by light micro-scopy.2,6,40 Indeed, an increased rate of apoptosis insmooth muscle cells2 and increased metalloproteinaseactivity3 have been suggested as causes of medial wallweakening and aneurysms of the ascending aorta inpatients with BAV, but these are not necessarily expressedin the development of CMD.

Active or Healed AortitisThe reported frequency of aortitis in surgical series

has varied greatly, from 1% by Klima et al20 to 22% byPomerance et al.28 Aortitis, either alone or associatedwith equally prominent CMD, was identified in 11% ofthe patients in the current investigation. Patients withboth processes more closely resembled patients withaortitis alone than isolated medial degeneration, suggest-ing aortitis to represent the most significant anddeterminant histologic finding. Thus, they have beencombined with aortitis patients in this discussion.

The current series and numerous recent studies havedocumented the emerging role of aortitis in the pathogen-esis of ascending aortic aneurysms, particularly amongthe elderly.4,8–10,13,18,24,32 Giant cell aortitis represents themost common microscopic type and affects women 3 to 5times as often as men,8,24 as confirmed in the presentstudy. It occurs primarily in 2 groups of patients: (1)those <50 years old who have Takayasu aortitis, and (2)those >50 years old who have isolated giant cell aortitis,or less frequently giant cell/temporal arteritis.13,15,37

Lymphoplasmacytic aortitis is considerably lesscommon than giant cell aortitis and has been describedprimarily in patients with systemic lupus erythemato-

sus.10,18,38 Of 11 patients initially categorized with thispattern, deeper sections revealed giant cells in 5 and led totheir reinterpretation as giant cell aortitis. Like giant cellaortitis, lymphoplasmacytic aortitis was mainly as iso-lated finding, rarely associated with a systemic arteritis. Inour series, there was no specific association with ACTD.An in depth analysis of the surgical pathology ofnoninfectious aortitis is the subject of a companion studyfrom our institution.26

Although syphilitic aortitis and mycotic aneurysmsaccounted for most inflammatory disorders of the aortain autopsy series,7,12,16,17,22 they have been relativelyuncommon in surgical pathology series,20,28 includingours in which only 6 were infectious and none weresyphilitic. Mycotic aneurysms of the ascending aorta aregenerally related to infective endocarditis of native orprosthetic aortic valves,11 as was found in the currentstudy, and their decrease in incidence is most likelyrelated to early detection and effective treatment ofendocarditis.

REFERENCES1. Bachet JE, Termignon JL, Dreyfus G, et al. Aortic dissection:

prevalence, cause, and results of late reoperations. J ThoracCardiovasc Surg. 1994;108:199–205.

2. Bonderman D, Gharehbaghi-Schnell E, Wollenek G, et al.Mechanisms underlying aortic dilatation in congenital aortic valvemalformation. Circulation. 1999;99:2138–2143.

3. Boyum J, Fellinger EK, Schmoker JD, et al. Matrix metalloprotei-nase activity in thoracic aortic aneurysms associated with bicuspidand tricuspid aortic valves. J Thorac Cardiovasc Surg. 2004;127:686–691.

4. Brack A, Martinez-Taboada V, Stanson A, et al. Disease pattern incranial and large-vessel giant cell arteritis. Arthritis Rheum. 1999;42:311–317.

5. Coady MA, Rizzo JA, Goldstein LJ, et al. Natural history,pathogenesis, and etiology of thoracic aortic aneurysms anddissections. Cardiol Clin. 1999;17:615–635.

6. de Sa M, Moshkovitz Y, Butany J, et al. Histologic abnormalities ofthe ascending aorta and pulmonary trunk in patients with bicuspidaortic valve disease: clinical relevance to the Ross procedure.J Thorac Cardiovasc Surg. 1999;118:588–594.

7. Eliot RS, Woodburn RL, Edwards JE. Conditions of the ascendingaorta simulating aortic valvular incompetence. Am J Cardiol. 1964;14:679–694.

8. Evans JM, Bowles CA, Bjornsson J, et al. Thoracic aortic aneurysmand rupture in giant cell arteritis: a descriptive study of 41 cases.Arthritis Rheum. 1994;37:1539–1547.

9. Evans JM, O’Fallon WM, Hunder GG. Increased incidence ofaortic aneurysm and dissection in giant cell (temporal) arteritis:a population-based study. Ann Intern Med. 1995;122:502–507.

10. Faye-Petersen OM, Arnold MM, Grizzle WE, et al. Lymphoplas-macytic aortitis and acute aortic dissection: an uncommon associa-tion. Arch Pathol Lab Med. 1996;120:402–404.

11. Feigl D, Feigl A, Edwards JE. Mycotic aneurysms of the aortic root:a pathologic study of 20 cases. Chest. 1986;90:553–557.

12. Frist WH, Miller DC. Aneurysms of ascending thoracic aorta andtransverse aortic arch. Cardiovasc Clin. 1987;17:263–287.

13. Gravanis MB. Giant cell arteritis and Takayasu aortitis: morpho-logic, pathogenetic and etiologic factors. Int J Cardiol. 2000;75(Suppl 1):S21–S33.

14. Hagan PG, Nienaber CA, Isselbacher EM, et al. The InternationalRegistry of Acute Aortic Dissection (IRAD): new insights into anold disease. JAMA. 2000;283:897–903.

15. Hall S, Barr W, Lie JT, et al. Takayasu arteritis: a study of 32 NorthAmerican patients. Medicine. 1985;64:89–99.



16. Heggtveit HA. Syphilitic aortitis: a clinicopathologic autopsy studyof 100 cases (1950–1960). Circulation. 1964;29:346–355.

17. Jackman JDJ, Radolf JD. Cardiovascular syphilis. Am J Med. 1989;87:425–433.

18. Khan AS, Spiera H. Association of aortic aneurysm in patients withsystemic lupus erythematosus: a series of case reports and a reviewof the literature. J Rheumatol. 1998;25:2019–2021.

19. Kita Y, Nakamura K, Itoh H. Histologic and histometric study ofthe aortic media in dissecting aneurysm: comparison with trueaneurysm and age-matched controls. Acta Pathol Jpn. 1990;40:408–416.

20. Klima T, Spjut HJ, Coelho A, et al. The morphology of ascendingaortic aneurysms. Hum Pathol. 1983;14:810–817.

21. Kouchoukos NT, Dougenis D. Surgery of the thoracic aorta. N EnglJ Med. 1997;336:1876–1888.

22. Lande A, Berkmen YM. Aortitis: pathologic, clinical and arterio-graphic review. Radiol Clin North Am. 1976;14:219–240.

23. Larson EW, Edwards WD. Risk factors for aortic dissection: anecropsy study of 161 cases. Am J Cardiol. 1984;53:849–855.

24. Liu G, Shupak R, Chiu BK. Aortic dissection in giant-cell arteritis.Sem Arthritis Rheum. 1995;25:160–171.

25. Marsalese DL, Moodie DS, Lytle BW, et al. Cystic medial necrosisof the aorta in patients without Marfan’s syndrome: surgicaloutcome and long-term follow-up. J Am Coll Cardiol. 1990;16:68–73.

26. Miller DV, Isotalo PA, Weyand CM, et al. Surgical pathology ofnon-infectious aortitis, with emphasis on an under-recognizedisolated variant. Am J Surg Pathol. 2006.

27. Nakashima Y, Kurozumi T, Sueishi K, et al. Dissecting aneurysm: aclinicopathologic and histopathologic study of 111 autopsied cases.Hum Pathol. 1990;21:291–296.

28. Pomerance A, Yacoub MH, Gula G. The surgical pathology ofthoracic aortic aneurysms. Histopathology. 1977;1:257–276.

29. Pretre R, Von Segesser LK. Aortic dissection. Lancet. 1997;349:1461–1464.

30. Roberts CS, Roberts WC. Dissection of the aorta associated withcongenital malformation of the aortic valve. J Am Coll Cardiol.1991;17:712–716.

31. Roberts WC. Aortic dissection: anatomy, consequences, and causes.Am Heart J. 1981;101:195–214.

32. Rojo-Leyva F, Ratliff NB, Cosgrove DM, et al. Study of 52 patientswith idiopathic aortitis from a cohort of 1,204 surgical cases.Arthritis Rheum. 2000;43:901–907.

33. Sabet HY, Edwards WD, Tazelaar HD, et al. Congenitally bicuspidaortic valves: a surgical pathology study of 542 cases (1991 through1996) and a literature review of 2,715 additional cases. Mayo ClinProc. 1999;74:14–26.

34. Savunen T, Aho HJ. Annulo-aortic ectasia: light and electronmicroscopic changes in aortic media. Virchows Arch [A]. 1985;407:279–288.

35. Schlatmann TJ, Becker AE. Histologic changes in the normal agingaorta: implications for dissecting aortic aneurysm. Am J Cardiol.1977;39:13–20.

36. Schlatmann TJ, Becker AE. Pathogenesis of dissecting aneurysm ofaorta: comparative histopathologic study of significance of medialchanges. Am J Cardiol. 1977;39:21–26.

37. Virmani R, Lande A, McAllister HA. Pathological Aspects ofTakayasu’s Arteritis. New York: Raven Press; 1986:55–79.

38. Willett WF, Kahn MJ, Gerber MA. Lupus aortitis: a casereport and review of the literature. J LA State Med Soc. 1996;148:55–59.

39. Wilson SK, Hutchins GM. Aortic dissecting aneurysms: causa-tive factors in 204 subjects. Arch Pathol Lab Med. 1982;106:175–180.

40. Yun KL. Ascending aortic aneurysm and aortic root disease.Cor Art Dis. 2002;13:79–84.



surgical pathology of the ascending aorta: a clinicopathologic study of 513 cases

Documents