Evidence-Based Imaging - part 7 potx

angiogra-䊏 A strong recommendation can be made for performing a noninvasiveimaging examination [e.g., single photon emission computed tomo-graphy SPECT or stress echo] prior to coronary

cumulative radiation dose to the breast was 10.8 cGy (range, 0–170).Seventy-seven breast cancer deaths were observed compared with 45.6expected deaths on the basis of United States mortality rates Women withscoliosis had a 1.7-fold risk of dying of breast cancer (95% CI, 1.3–2.1) whencompared with the general population The data suggested that radiationwas the causative factor, with risk increasing significantly with the number

of radiographic exposures and the cumulative radiation dose (moderateevidence) Potential confounding was noted because the severity of diseasewas related to radiation exposure and reproductive history; patients withmore severe disease were less likely to become pregnant and had a greaterrisk of breast cancer

In a large retrospective cohort study of 2039 patients, Levy and leagues (75) found an excess lifetime cancer risk of 1% to 2% (12 to 25 casesper 1000 population) among women (moderate evidence) The same groupsuggested that supplanting the anteroposterior (AP) view with the pos-teroanterior (PA) view would result in a three- to sevenfold reduction incumulative doses to the thyroid gland and the female breast, three- to four-fold reductions in the lifetime risk of breast cancer, and a halving of thelifetime risk of thyroid cancer (76) The same cohort of women was evalu-ated for adverse reproductive outcomes (77) Of the initial group of 1793young women evaluated for scoliosis between 1960 and 1979, 1292 womenreturned questionnaires in 1990 This cohort was compared with a refer-ence group of 1134 women selected randomly from the general population.The adolescent idiopathic scoliosis cohort had a higher risk of spontaneousabortions [odds ratio (OR), 1.35; 95% CI, 1.06–1.73] (moderate evidence).The odds of unsuccessful attempts at pregnancy (OR, 1.33; 95% CI,0.84–2.13) and of congenital malformations (OR, 1.2; 95% CI, 0.78–1.84)were also higher but not statistically significant (moderate evidence).Digital radiography seems to reduce radiation exposure The results arevaried (78–80), and the technology is evolving (limited evidence) Studiesreport an 18-fold reduction with some systems (73) versus an almosttwofold increase with others (81)

col-VII What Is the Use of Magnetic Resonance Imaging (MRI) for Severe Idiopathic Scoliosis?

Summary of Evidence: There is increasing concern about the association

of idiopathic scoliosis with structural abnormalities of the neural axis.Minimal tonsillar ectopia (<5mm) is significantly prevalent in scoliosis andcorrelates with abnormalities in somatosensory-evoked potentials andwith the severity of scoliosis (4) (moderate evidence) Otherwise, a paucity

of significant findings on MRI of patients evaluated for idiopathic sis is noted, even in severe cases (4)

scolio-Supporting Evidence: Cheng and colleagues (82) studied 36 healthy control

subjects, 135 patients with moderately severe adolescent idiopathic sis (Cobb angle less than 45 degrees), and 29 similar patients with Cobbangles greater than 45 degrees All of the patients were evaluated pros-pectively with MRI looking specifically for tonsillar ectopia and withsomatosensory-evoked potentials Tonsillar herniation was found in none

scolio-of the controls versus four scolio-of 135 (3%) and eight scolio-of 29 (27.6%) scolio-of the two

scoliotic groups (p< 001) (moderate evidence) Similarly, the percentages

344 L.S Medina et al.

of patients with abnormal somatosensory-evoked potentials were 0%,

11.9%, and 27.6%, respectively There was a significant association between

tonsillar ectopia and abnormal somatosensory function (p< 001;

correla-tion coefficient, 0.672) (moderate evidence) Tonsillar ectopia was defined

as any inferior displacement of the tonsils, and none of the patients had a

displacement greater than 5 mm, which is considered the usual threshold

for the diagnosis (83–85)

Several studies have addressed the prevalence of MR abnormalities in

patients with severe idiopathic scoliosis who are otherwise asymptomatic

Do and colleagues (86) studied a consecutive series of 327 patients with

idiopathic scoliosis requiring surgical intervention (average preoperative

curve of 57 degrees) but without neurologic findings The patients, aged

10 to 19 years, were evaluated from the base of the skull to the sacrum

Seven patients had abnormal MRI, including two with syrinx, four with

Chiari malformation type I, and one with a fatty vertebral body None of

them required specific treatment for these findings (moderate evidence)

In four other cases, equivocal MRI findings necessitated additional

workup In a similar prospective double-blinded study of 140 patients

eval-uated preoperatively, Winter et al (87) found four patients with

abnor-malities, three with Chiari I malformations, and one with a small syrinx,

none of whom required treatment In another study of MRI examinations

performed preoperatively, Maiocco et al (88) found two of 45 patients with

syrinx, one requiring decompression (moderate evidence)

To study whether the severity of the curve increased the risk of

associ-ated abnormalities, O’Brien et al (89) performed MR evaluation on 33

con-secutive patients with adolescent idiopathic scoliosis and Cobb angles

greater than 70 degrees No neural axis abnormalities were found (limited

evidence)

VIII What Is the Use of MRI for High-Risk

Subgroups of Scoliosis?

Summary of Evidence: Unlike adolescent idiopathic scoliosis, juvenile and

infantile idiopathic scoliosis and congenital scoliosis have a high incidence

of neural axis abnormalities (limited evidence) Increased incidence of

neural axis abnormalities have been seen with atypical idiopathic scoliosis

and left (levoconvex) thoracic scoliosis (Figs 18.3 and 18.4) (4) (limited

evidence)

Supporting Evidence: Several studies have shown that, with scoliosis types

that are different from the typical adolescent idiopathic form, there is a

high prevalence of neural abnormalities (4) Of 30 consecutive children

with congenital scoliosis studied by Prahinski and colleagues (90), nine had

syringomyelia Of these children, one required release of the tethered cord

and one correction of a diastematomyelia (limited evidence) Two studies

of prepubertal children suggest a high incidence of neural abnormalities

in juvenile and infantile scoliosis In a study of 26 consecutive children

aged less than 11 years, Lewonowski and colleagues (91) found five (19.2%)

with abnormalities of the cord Three required surgical intervention, two

with hydromyelia and one with a mass (91) (limited evidence) Gupta and

colleagues (92) found that six of 34 patients under 10 years of age studied

prospectively had neural axis abnormalities, including two patients with

syrinx requiring syringopleural shunting (one with a Chiari I tion) Other abnormalities included dural ectasia, tethered cord, and abrainstem astrocytoma (limited evidence).

malforma-In a retrospective review of 95 patients with idiopathic scoliosis who hadbeen studied for various indications, Schwend and colleagues (93) foundthat 12 had a syrinx, one a cord astrocytoma, and one dural ectasia (limitedevidence) Left thoracic scoliosis was the most important predictor ofabnormality (10 abnormalities in 43 patients) Mejia et al (94) then per-formed a prospective study (level II) of 29 consecutive patients with idio-pathic left thoracic scoliosis, finding only two with syrinx and no otherabnormalities (limited evidence) Barnes and colleagues (37) retrospec-tively analyzed 30 patients with atypical idiopathic scoliosis and found 17abnormalities in 11 patients, including seven cases of syringohydromyeliaand five Chiari I malformations (limited evidence)

346 L.S Medina et al.

Figure 18.5. Suggested decision tree for use in newborns with suspected occult spinal dysraphism For those patients in the high-risk group MRI is recommended For patients in the intermediate-risk group ultrasound (US) or MRI is the strategy

of choice, while for the low-risk group ultrasound or no imaging is recommended For patients with negative imaging studies close clinical follow-up with periodic

reassessment is recommended [Source: Medina et al (30), with permission.]

these two strategies per risk group may be based on individual and

insti-tutional diagnostic performance and cost per test In newborns with

sus-pected occult dysraphism, appropriate selection of patients for imaging

based on these risk groups may maximize health outcomes for patients and

improve health care resource allocation

How Should Scoliosis Be Evaluated?

Figure 18.6 summarizes the decision tree for patients with suspected

scoliosis

Imaging Case Study of Spinal Dysraphism

This imaging case study illustrates a child with skin stigmata (Fig 18.1)

who has an occult dysraphic lesion of the intradural lipoma type (Fig 18.2)

Imaging Case Study of Scoliosis

This imaging case study illustrates a child with atypical levoconvex

tho-racic scoliosis (Fig 18.3) who has neurofibromatosis type 1 with

underly-ing plexiform neurofibromas (Fig 18.4)

Figure 18.6. Suggested decision tree for use in patients with suspected scoliosis.

Decision tree emphasizes importance of clinical history, physical exam, and

radiographs in determining the need for MRI.

Suggested Imaging Protocols for Spinal Dysraphism

Spinal Ultrasound

Spinal ultrasound should be performed in patients before the age of 3months to avoid the limited acoustic window from mineralization of theposterior elements An experienced operator should perform the studyusing a high-frequency, 5- to 15-MHz linear array transducer (52)

Entire Spine MRI

A retrospective case-control study including 101 patients (moderate dence) suspected of having occult lumbosacral dysraphism demonstratedthat conventional three-plane T1-weighted lumbosacral MRI in childrenand young adults provided better diagnostic information than a fastscreening two-plane T1-weighted MRI because of its higher specificity andinterobserver agreement (20) T2-weighted images in the axial and sagittalplane are often added to the protocol to assess intrinsic cord abnormali-ties Intravenous paramagnetic contrast is not routinely used, unless thepatient has a communicating dorsal dermal sinus tract or clinical concerns

Entire Spine MRI

Patients with scoliosis may represent an imaging challenge In patientswith scoliosis being evaluated with MRI, the entire spine should becovered Three plain T1- and T2-weighted images should be obtained withdifferent obliquities to optimize imaging information Another approach is

to obtain three-dimensional fast spin echo (FSE) volumetric imaging Weinberger and colleagues (95) recommend using a TR of 500 ms, TEeffof

21 ms, echo train length (ETL) of 8, 20- to 38-cm field of view, 256 ¥ 256 inplane matrix, 1-mm sagittal partition thickness, one excitation, and 16 kHz

of receive bandwidth Intravenous paramagnetic contrast is important inthe evaluation of intramedullary and extramedullary neoplasm

Future Research

• Formal cost-effectiveness analysis of imaging in children with scoliosis

• Further development of low or no radiation imaging techniques forpatients with scoliosis

References

1 Pacheco-Jacome E, Ballesteros MC, Jayakar P, Morrison G, Ragheb J, Medina

LS Neuroimag Clin North Am 2003;13:327–334.

348 L.S Medina et al.

2 Soonawala N, Overweg-Plandsoen WCG, Brouwer OF Clin Neurol Neurosurg

6 Shahcheraghi GH, Hobbi MH J Pediatr Orthop 1999;19(6):766–775.

7 DiPietro MA Radiology 1993;188:149–153.

8 Rowland Hill CA, Gibson PJ AJNR 1995;16:469–472.

9 Wilson DA, Prince JR AJNR 1989;10:259–262.

10 Beek FJ, de Vries LS, Gerards LJ, Mali WP Neuroradiology 1996;

38(suppl):S174–177.

11 Barson AJ J Anat 1970;106:489–497.

12 Warder DE, Oakes WJ Neurosurgery 1993;33:374–378.

13 Haworth JC, Zachary RB Lancet 1955;2:10–14.

14 Milhorat TH, Miller JI Neurosurgery In: Avery GB, Fletcher MA, Mhairi GM,

eds Neonatology (4th ed) Philadelphia: JB Lippincott, 1994:1155–1163.

15 Kalter H, Warkany J Medical progress N Engl J Med 1983;308:424–431.

16 Bird TD, Hall JG Neurology 1977;27:1057–1060.

17 Vintzileos AM, Ananth CV, Fisher AJ, et al Am J Obstet Gynecol 1999;

180(5):1227–1233.

18 Knight GJ, Palomaki GF Maternal serum alpha-fetoprotein and the detection

of open neural tube defects In: Elias S, Simpson JL, eds Maternal Serum

Screening for Fetal Genetic Disorders New York: Churchill Livingstone, 1992:

41–58.

19 Egelhoff JC, Prenger EC, Coley BD The spine In: Ball W Jr, ed Pediatric

Neuroradiology Philadelphia: Lippincott-Raven, 1997:717–778.

20 Medina LS, Al-Orfali M, Zurakowski D, Poussaint TY, DiCanzio J, Barnes PD.

Radiology 1999;211:767–771.

21 Appignani BA, Jaramillo D, Barnes PD, Poussaint TY AJR 1994;163:1199–1203.

22 Raghavan N, Barkovich AJ, Edwards M, Norman D AJR 1989;152:843–852.

23 Brophy JD, Sutton LN, Zimmerman RA, Bury E, Schut L Neurosurgery

1989;25:336–340.

24 Moufarrij NE, Palmer JM, Hahn JF, Weinstein MA Neurosurgery 1989;25:

341–346.

25 Milhorat TH, Miller JI Neurosurgery In: Avery GB, Fletcher MA, Mhairi GM,

eds Neonatology (4th ed) Philadelphia: JB Lippincott, 1994:1155–1163.

26 Volpe JJ Neurology of the Newborn, 4th ed Philadelphia: WB Saunders, 2001.

27 Hoffman HJ, Hendrick EB, Humphreys RP Childs Brain 1976;2:145–155.

28 Hoffman HJ, Taecholarn C, Hendrick EB, Humphreys RP J Neurosurg

1985;62:1–8.

29 Kriss VM, Desai NS AJR 1998;171:1687–1692.

30 Medina LS, Crone K, Kuntz KM Pediatrics 2001;108:E101.

31 Stirling AJ, Howel D, Millner PA, Sadiq S, Sharples D, Dickson RA J Bone Joint

Surg Am 1996;78(9):1330–1336.

32 Newton PO, Wenger DR Idiopathic and congenital scoliosis In: Morrissy RT,

Weinstein SL, eds Lovell & Winter’s Pediatric Orthopaedics, 5th ed

Philadel-phia: Lippincott Williams & Wilkins, 2000:677–740.

33 Rogala EJ, Drummond DS, Gurr J J Bone Joint Surg Am 1978;60(2):173–176.

34 Al-Arjani AM, Al-Sebai MW, Al-Khawashki HM, Saadeddin MF Saudi Med J

2000;21(6):554–557.

35 Weinstein SL Spine 1999;24(24):2592–2600.

36 Weinstein SL, Ponseti IV J Bone Joint Surg Am 1983;65(4):447–455.

37 Barnes PD, Brody JD, Jaramillo D, Akbar JU, Emans JB Radiology 1993;

186(1):247–253.

38 Horton D, Barnes P, Pendleton BD, Polly M J Okla State Med Assoc 1989; 82:15–19.

39 Rohrschneider WK, Forsting M, Darge K, Tröger J Radiology 1996;200:383– 388.

40 Powell KR, Cherry JD, Hougen TJB, Blinderman EE, Dunn MC J Pediatr 1975;87:744–750.

41 Byrd SE, Darling CF, McLone DG Radiol Clin North Am 1991;29:711–752.

42 Estin MD, Cohen AR Neurosurg Clin North Am 1995;6:377–391.

43 Herman TE, Oser RF, Shackelford GD Clin Pediatr 1993;32:627–628.

44 Kaplan JO, Quencer RM Radiology 1980;137:387–391.

45 McLone DG, Naidich TP The tethered spinal cord In: McLaurin RL, Schof L, Venes JL, Epstein F, eds Surgery of the Developing Nervous System Philadel- phia: WB Saunders, 1989:71–96.

46 Yamada S, Iacono RP, Andrade T, Mandybur G, Yamada BS Neurosurg Clin North Am 1995;6:311–323.

47 Davis PC, Hoffman JC, Ball TI, et al Radiology 1988;166:679–685.

48 Scatliff JH, Kendall BE, Kingsley DPE, Britton J, Grant DN, Hayward RD AJNR 1989;10:269–277.

49 Pang D, Wilberger JE J Neurosurg 1982;57:32–47.

50 Fone PD, Vapnek JM, Litwiller SE, et al J Urol 1997;157.

51 Satar N, Bauer SB, Shefner J, Kelly MD, Darbey MM J Urol 1995;154:754– 758.

52 Coley BD, Siegel MJ Spinal ultrasonography In: Siegel MJ, ed Pediatric graphy Philadelphia: Lippincott Williams & Wilkins, 2002:671–698.

Sono-53 Rubin JM, Di Pietro MA, Chandler WF, et al Radiol Clin North Am 1988; 26:1–27.

54 Page LK Occult spinal dysraphism and related disorders In: Wilkins RH, Rengachary SS, eds Neurosurgery New York: McGraw-Hill, 1992:2053–2058.

55 Westcott MA, Dynes MC, Remer EM, Donaldson JS, Dias LS Radiographics 1992;12:1155–1173.

56 Atala A, Bauer SB, Dyro FM, et al J Urol 1992;148:592–594.

57 Reigel DH, Tchernoukha K, Bazmi B, Kortyna R, Rotenstein D Pediatr surg 1994;20:30–42.

Neuro-58 Law DA, Aronoff SC Pediatr Infect Dis J 1992;11:968–971.

59 Rogg JM, Benzil DL, Haas RL, Knuckey NW AJNR 1993;14:1393–1395.

60 DiTullio MV Jr Surg Neurol 1977;7:351–354.

61 Feigen RD, Cherry JD Textbook of Pediatric Infectious Diseases, 2nd ed Philadelphia: WB Saunders, 1987.

62 Givner LB, Baker CJ Pediatr Infec Dis J 1983;2:385–387.

63 Capitanucci ML, Iacobelli BD, Silveri M, Mosiello G, De Gennaro M Eur J Pediatr Surg 1996;6(suppl 1):25–26.

64 Mills JL Teratology 1982;25:385–394.

65 Rusnak SL, Discoll SG Pediatrics 1965;35:989–995.

66 Becerra JE, Khoury MJ, Cordero JF, Ericson JD Pediatrics 1990;85:1–9.

67 Long FR, Hunter JV, Mahboubi S, Kalmus A, Templeton JM Tethered cord and associated vertebral anomalies in children and infants with imperforate anus: evaluation with MR imaging and plain radiography Radiology 1996;200:3 77–382.

68 Morrissy RT, Goldsmith GS, Hall EC, Kehl D, Cowie GH J Bone Joint Surg Am 1990;72(3):320–327.

69 Carman DL, Browne RH, Birch JG J Bone Joint Surg Am 1990;72(3):328–333.

70 Shea KG, Stevens PM, Nelson M, Smith JT, Masters KS, Yandow S Spine 1998; 23(5):551–555.

71 Loder RT, Urquhart A, Steen H, et al J Bone Joint Surg Br 1995;77(5):768–770.

72 Crockett HC, Wright JM, Burke S, Boachie-Adjei O Spine 1999;24(19): 2007–2009; discussion 2010.

73 Kalifa G, Charpak Y, Maccia C, et al Pediatr Radiol 1998;28(7):557–561.

350 L.S Medina et al.

74 Morin Doody M, Lonstein JE, Stovall M, Hacker DG, Luckyanov N, Land CE.

78 Kalmar JA, Jones JP, Merritt CR Spine 1994;19(7):818–823.

79 Kling TF Jr, Cohen MJ, Lindseth RE, De Rosa GP Spine 1990;15(9):880–885.

80 Stringer DA, Cairns RA, Poskitt KJ, Bray H, Milner R, Kennedy B Pediatr Radiol

1994;24(1):1–5.

81 Geijer H, Beckman K, Jonsson B, Andersson T, Persliden J Radiology

2001;218(2):402–410.

82 Cheng JC, Guo X, Sher AH, Chan YL, Metreweli C Spine 1999;24(16):1679–1684.

83 Barkovich AJ, Wippold FJ, Sherman JL, Citrin CM AJNR 1986;7(5):795–799.

84 Elster AD, Chen MY Radiology 1992;183(2):347–353.

85 Mikulis DJ, Diaz O, Egglin TK, Sanchez R Radiology 1992;183(3):725–728.

86 Do T, Fras C, Burke S, Widmann RF, Rawlins B, Boachie-Adjei O J Bone Joint

Surg Am 2001;83–A(4):577–579.

87 Winter RB, Lonstein JE, Heithoff KB, Kirkham JA Spine 1997;22(8):855–858.

88 Maiocco B, Deeney VF, Coulon R, Parks PF Jr Spine 1997;22(21):2537–2541.

89 O’Brien MF, Lenke LG, Bridwell KH, Blanke K, Baldus C Spine 1994;

19(14):1606–1610.

90 Prahinski JR, Polly DW Jr, McHale KA, Ellenbogen RG J Pediatr Orthop 2000;

20(1):59–63.

91 Lewonowski K, King JD, Nelson MD Spine 1992;17(6 suppl):S109–116.

92 Gupta P, Lenke LG, Bridwell KH Spine 1998;23(2):206–210.

93 Schwend RM, Hennrikus W, Hall JE, Emans JB J Bone Joint Surg Am 1995;

Cardiac Evaluation: The Current

Status of Outcomes-Based Imaging

Andrew J Bierhals and Pamela K Woodard

I Does coronary artery calcification scoring predict outcome?

II Special case: high-risk patientsIII Which patients should undergo coronary angiography?

IV Which patients should undergo noninvasive imaging of the heart?

V What is the appropriate use of coronary artery computed phy and magnetic resonance?

tomogra-352

Issues

䊏 A strong recommendation can be made for initial coronary phy among high-risk patients and those who are post–myocardialinfarction (MI) that was transmural or with ischemic symptoms(strong evidence)

angiogra-䊏 A strong recommendation can be made for performing a noninvasiveimaging examination [e.g., single photon emission computed tomo-graphy (SPECT) or stress echo] prior to coronary angiography in low-risk patients and those who have had a non–Q-wave MI (strongevidence)

䊏 Aside from coronary angiography, the appropriate usage of cardiacimaging studies remains unclear, and more research is required toevaluate the outcomes, as well as the cost-effectiveness of the afore-mentioned modalities (insufficient evidence)

䊏 Coronary artery calcium scoring has been shown in asymptomaticpatients to be predictive of coronary artery disease; however, there havebeen no data to support the position of added predictive value overand above the clinical Framingham model (insufficient evidence)

Key Points

Definition and Pathophysiology

The etiology of coronary artery disease (CAD) is multifactorial involvingboth interaction of lifestyle and genetic predispositions While some factorsare not modifiable, those risks that may be altered are often neglected untilthere evidence of disease As a result, a multitude of tests and clinical

assessment tools have been developed to risk stratify patients in order to

direct short- and long-term treatments The modifiable risk factors (e.g.,

hypertension, hyperlipidemia, and diabetes) have been on the rise over the

past decade (1,2); therefore, a greater urgency has arisen to identify patients

with CAD

Coronary artery disease begins as fatty streaks in the coronary arteries

that may begin as early as 3 years of age The fatty streaks are composed of

large cells with intracellular lipids (foam cells) that are located in the

suben-dothelial region As patients age, the fatty streaks develop into fibrous

plaques that narrow the vessel lumen, reducing blood flow The fibrous

plaques over time may calcify, reducing vessel compliance and increasing

fragility This further reduces blood flow and increases the chance of the

plaque rupturing, resulting in an acute coronary artery occlusion

Epidemiology

Coronary artery disease is a nationwide epidemic involving 6.4% of the

entire population (3,4) and is the largest cause of mortality, accounting for

one in every five deaths (4) This translates into a death rate of 177.8 per

100,000 (based on 2001 estimates) (4) In the United States, over 1.5 million

people will have a myocardial infarction, and the majority of the patients

will initially present with symptoms in their 50s and 60s

A large volume of literature has been generated investigating these

modalities, but little has focused on the impact the modalities have on the

patient outcomes even though there has been a steady increase in the use

of costly diagnostic testing and treatment (5) This chapter reviews the

literature on the outcomes research of cardiac imaging, and makes

recommendations concerning the utilization of the techniques in patient

management

Overall Cost to Society

In the United States, the estimated 2004 cost of heart disease to society is

$238 billion, with over half secondary to CAD ($133 billion) (4,6) The cost

of heart disease is substantial in comparison to other disease processes,

such as cancers ($189 billion) and AIDS ($29 billion) (4,6) The costs of CAD

include direct health care of $66 billion, and $67 billion in indirect costs

(e.g., loss of productivity secondary to morbidity and mortality) (4,6)

The expenditures for health care are consistently increasing, because of

new technologies and the current medicolegal environment An

ever-declining budget results in a need for clinicians to incorporate

cost-effective strategies in patient evaluations However, cost-cost-effective does not

mean withholding evaluations or always ordering the seemingly least

expensive test, but rather understanding what is most efficient with respect

to a specific clinical situation, based on current research The purpose of

this approach is to direct a finite amount of resources and limit costs to

society without affecting the quality of health care This chapter reviews

the cost-effectiveness and outcomes of various imaging modalities of heart

disease, and makes recommendations concerning these techniques in

patient care Specifically, coronary artery calcification scoring, myocardial

SPECT, angiography, stress echocardiography, and cardiac magnetic

reso-nance (MR) and computed tomography (CT) will be evaluated in theirpotential roles in the evaluation of heart disease.

Goals

The goals of imaging related to CAD are based on the a priori risk to the

patient In a low-risk population, the goals of imaging are to identify thosewith early disease Subsequently, interventions directed toward risk factorsand lifestyle may be initiated in order to reverse disease or halt progres-sion before any long-term effects result However, risk stratificationbecomes the goal of cardiac imaging among those patients who are con-sidered high risk The imaging in the aforementioned population is todetermine if any coronary artery intervention (i.e., endovascular or bypassgraft) is required over and above medical management

A similar literature review was performed for coronary angiography

using Medline The keyword search from 1999 to 2004 included coronary angiography and outcomes and coronary angiography and cost-effectiveness.

Over 5000 reports were identified, with approximately 100 addressingpatient outcomes and 10 evaluating cost-effectiveness

Lastly, a literature review was performed on Medline from 1999 to 2004for noninvasive techniques including SPECT, positron emission tomogra-phy (PET), echocardiogram, and coronary CT and MR using the samemethod, as described above The review yielded over 100 articles address-ing patient outcomes and five evaluating cost-effectiveness; however, therewere no reports that evaluated either topic for MR or CT angiography

I Does Coronary Artery Calcification Scoring Predict Outcome?

Summary of Evidence: Coronary artery calcium scoring has been shown in

asymptomatic patients to be predictive of CAD; however, there have been

no data to support the position of added predictive value over and abovethe clinical Framingham model Therefore, coronary artery calcificationscoring cannot be recommended as a screening tool at this time The lack

of cost-effectiveness data necessitates further investigations before a finalposition can be determined on the utility of calcium scoring (insufficientevidence)

Supporting Evidence: Coronary artery calcium scoring performed by

com-puted tomography (CT) has been utilized in asymptomatic patients toassess their risk of an acute coronary event (7) However, the literature hasdebated the utility of calcium scoring Some researchers support its use

354 A.J Bierhals and P.K Woodard

(8,9), while others are less enthusiastic concerning the utilization in patient

care (10)

Computed tomography calcium scoring, despite conflicting reports, has

been shown to be associated with a fourfold increased risk in myocardial

infarction and coronary death in a meta-analysis by O’Malley et al (11) in

2000 The study included nine reports that had a diverse asymptomatic

population that was evaluated for coronary artery calcification by electron

beam CT The authors also reported a ninefold increased risk of coronary

events (i.e., nonfatal MI, sudden death, or revascularization) among those

with a coronary artery calcium score above the median There is moderate

evidence to suggest that coronary artery calcification score is predictive of

coronary events

More recent reports have echoed these results regarding the predictive

value of CT calcium scoring A 2003 study by Shaw et al (8) developed a

multivariate model on a sample of greater than 10,000 asymptomatic

indi-viduals incorporating calcium score with typical clinical risk factors (i.e.,

hypertension, hypercholesterolemia, diabetes, age, and sex) to predict

all-cause mortality The results of the study indicated that calcium score

pre-dicted all-cause mortality (p< 001) over and above the effects of other risk

factors The study also found that there was a trend with the coronary

artery calcium score such that as the calcium burden increased there was

a greater risk of all-cause mortality The relative risk in patients with

ele-vated calcium scores ranged from 1.6 to 4.0 above individuals with the

lowest calcium burden; as the calcium burden increased, the risk increased

(Fig 19.1) Based on the results, the authors concluded that calcium scoring

of the coronary arteries provides additional information in the prediction

of all-cause mortality (8); however, morbidity and mortality secondary to

Figure 19.1. Graph shows risk stratification for each category of Framingham risk

(from low to high) according to baseline calcium score Event rate is predicted

mor-tality at 5 years (8) Low risk <0.30 (no risk factors), intermediate risk <0.70 (one to

two risk factors), and high risk <0.9 (three or more risk factors) probability of cardiac

disease [Source: Shaw et al (8), with permission from the Radiological Society of

North America.]

CAD was not specifically addressed In addition, the authors did not tigate if the added explanation would have any clinical impact and thusprovide information that would have proved clinically important.

inves-Other authors have found similar results in the prediction of mortalityfrom calcium scoring For example, Arad et al (12) demonstrated that mod-erate calcium scores were associated with a 10 times increase in cardiacdeath or MI In addition, a small study of 676 subjects demonstrated thatcoronary artery calcification scores incrementally predicted cardiac events(13) These studies, as with the aforementioned larger sample, were able

to show that coronary artery calcification on CT predicted health comes (e.g., MI and mortality) But of all the studies that have been eval-uated, none has shown any extra value in risk stratification and patientmanagement

out-Aside from the earlier described reports, there has been a multitude ofsimilar studies with varying patient population that have reached the sameconclusion concerning the ability of coronary artery calcium scoring topredict heart disease and mortality (14–19) Other investigators utilizedcalcium scoring in conjunction with laboratory tests, such as C-reactiveprotein to model the mortality of heart disease (20), but no interactiveeffects were noted, although each independently predicted coronary eventsand mortality However, a review of the literature to date has failed to iden-tify any direct data suggesting that calcium scoring has any clinical benefitover the current Framingham risk model (21)

Currently, coronary artery calcium scoring on CT is utilized as a riskstratification tool for CAD The major proportion of the data to date hasshown that calcium scoring can predict CAD as well as mortality related

to heart disease among asymptomatic patients A literature review did notuncover any data that show that calcium scoring adds any additional infor-mation over current clinical predictive models in the asymptomatic patient

In addition, there have been no studies specifically evaluating the effectiveness of coronary calcium scoring as a screening tool As a result,calcium scoring, while predictive of CAD and mortality, has yet to beshown to add any additional information over and above current clinicalmodels Therefore, at this time there is insufficient data to recommendcalcium scoring as a screening or risk stratification tool in the asympto-matic population However, the dearth of cost-effectiveness data precludesstating that calcium scoring should not be preformed as a screening test.Subsequently, additional cost-effectiveness studies should be instituted toevaluate the role of calcium scoring in the screening for CAD

cost-II Special Case: High-Risk Patients

Summary of Evidence: Among high-risk symptomatic populations

coro-nary artery calcium scoring on CT has failed to show any predictive valuefor a coronary event or mortality Thus, among high-risk populationscalcium scoring cannot be recommended for screening or risk stratification(Insufficient Evidence)

Supporting Evidence: The data in the asymptomatic populations

consis-tently indicated that coronary artery calcium scoring can predict cardiacevents and may be helpful in risk stratifying patients However, the results

356 A.J Bierhals and P.K Woodard

in populations with a known risk are not as straightforward Qu et al (22)

evaluated calcium scoring in a diabetic population The data showed that

when adjusting for other risk factors in a diabetic sample, calcium scores

did not predict coronary events, but calcium scoring was predictive among

nondiabetics (Fig 19.2) Although the results have not been as clear among

an elderly population that coronary artery calcification is associated with

the degree of CAD, some researchers have found that the calcium score

has variability among an elderly population, and thus may have the

poten-tial to discriminate risk within this group (23) However, other authors

have concluded that there is limited utility of using calcium scoring among

elderly patients (24,25) because of comorbidities limiting the effect of

inter-ventions Lastly, Detrano et al (10) concluded that neither clinical risk

assessment nor calcium scoring is an accurate predictor of cardiac events

in a high-risk population, based on the Framingham model Currently,

there is insufficient evidence to recommend coronary artery calcium

scoring in a high-risk population as a means of risk predicting coronary

events (insufficient evidence)

III Which Patients Should Undergo

Coronary Angiography?

Summary of Evidence: Coronary angiography has been studied with a

greater degree of rigor than the other modalities, with several studies

investigating the cost-effectiveness Based on the large amount of extant

3.4 p<0.0001

p<0.06

1.7 1.7

1.4 1

1

Figure 19.2. A: Relative risks (RRs), stratified by diabetes status, of nonfatal

myocardial infarction (MI), or coronary death associated with calcium score risk

groups (low, <2.8; medium, 2.8–117.8; high, >117.8) B: RRs, stratified by diabetes

status, of nonfatal MI, coronary death, percutaneous transluminal coronary

angio-plasty (PTCA), coronary artery bypass graft (CABG), or stroke associated with

calcium score risk groups (low, <2.8; medium, 2.8–117.8; high, >117.8) [Source: From

Qu et al (22).]

data, a strong recommendation can be made for initial coronary phy among high-risk patients and those who are post-MI that was trans-mural or with ischemic symptoms Also, a strong recommendation can bemade for performing a noninvasive imaging examination (i.e., SPECT orstress echo) prior to coronary angiography in low-risk patients and thosewho have had a non–Q-wave MI (Fig 19.3) (strong evidence).

angiogra-Supporting Evidence: Over the past 20 years, coronary angiography has

been the mainstay in the diagnosis of acute occlusion of the coronary ies as well as in the quantification of CAD to direct management, whethersurgical, medical, or endovascular Throughout this period, angiographyhas become the gold standard for the diagnosis of CAD, but unlike otherimaging studies of the heart there is greater risk associated with the pro-cedure Subsequently, the risk and technical factors preclude all patientsfrom undergoing an angiogram

arter-Several cost-effectiveness models have been proposed to evaluate therole of coronary angiography in the diagnosis of coronary artery disease(26–28) Patterson et al (27) utilized decision analysis to evaluate angiog-raphy versus other noninvasive modalities [i.e., SPECT, PET, exercise elec-trocardiogram (ECG)] This model incorporated both direct and indirectcosts as well as quality-adjusted life years (QALYs) to evaluate the differ-ent diagnostic modalities The diagnostic evaluations included non-invasive testing followed by angiography (among those with an initialabnormal test) or angiography alone The results of the study indicate thatcost-effectiveness of the diagnostic modality is based on the initial pretestlikelihood of disease The authors found angiography was the most

cost-effective modality in those with a high pretest probability (p> 70)

However, populations with low risk (p < 70) noninvasive testing was the most cost-effective with PET > SPECT > exercise ECG In addition, theauthors found that there was little impact on the cost-effectiveness

358 A.J Bierhals and P.K Woodard

Figure 19.3. The recommended decision tree for the evaluation of CAD based on the patients’ initial clinical status *Noninvasive study can represent SPECT or stress echo depending on the institutional performance characteristics of the imaging study.

from the differing treatment modalities (i.e., surgical, medical, or

endovas-cular) Similar results have been described by Garber and Solomon (28)

Their decision analysis demonstrated that while stress echocardiography

was the least costly per QALY saved, immediate angiography was an

acceptable cost-effective alternative to SPECT and stress echocardiography

among patients who are at high risk of cardiac disease In their model, the

relative cost-effectiveness for the modalities remained the same regardless

of the patient’s age or gender (Fig 19.4) There is strong evidence to

rec-ommend that among low-risk populations a noninvasive cardiac imaging

study should be performed prior to coronary angiography (strong

evidence)

Coronary angiography seemingly has a specific role in the diagnosis and

risk stratification of patients with heart disease and has been shown to be

cost-effective in given populations (26–28); however, the data in post-MI

populations is not as clear In a decision analysis by Kuntz et al (29), the

decision analytic model incorporated clinical history and symptoms in the

post-MI patient to evaluate the cost-effectiveness of angiography versus

medial care While the authors incorporated clinical elements into the

analyses, there was a failure to account for type of MI to address the issue

of noninvasive evaluation of cardiac perfusion (e.g., SPECT or stress echo)

Based on the model outcomes, angiography was found to be cost-effective

in almost all patients in the post-MI setting, and among those at highest

risk the cost-effectiveness ratios were less than $50,000 for each QALY

saved Only in those women at low to moderate risk for coronary disease

was angiography found not to be cost-effective Similar results on the

patient survival and outcomes have been found in other studies that have

included all post-MI patients (30,31), and the largest effects were among

the patients with transmural infarctions There is strong evidence to

support the use of angiography in the transmural infarction while those

with a nontransmural infarction should undergo a noninvasive study prior

to angiography (strong evidence)

Several authors have evaluated low to moderate risk (probability of

CAD < 7) subpopulations in the post-MI state to determine the

cost-effectiveness and outcomes among those treated with noninvasive image

guidance versus immediate angiography Barnett et al (32) utilized a

randomized controlled trial to evaluate the cost-effectiveness of

angiogra-phy versus selective angiograangiogra-phy (i.e., performing angiograangiogra-phy in patients

with an abnormal finding on a noninvasive study) in a population with a

non–Q-wave MI The results indicate a conservative management program

is more cost-effective than immediate angiography in patients with a

non–Q-wave MI In the acute setting, image-directed angiography resulted

in a cost of $14,700 versus $19,200 for immediate angiography and

per-sisted after 2 years of follow-up, at which time there was an approximate

$2100 difference in cost In addition, the conservative group had a better

survival (1.86 years) over a 2-year follow-up relative to immediate

angiography (1.76 years) Thus, conservative management (i.e.,

noninva-sive image-directed angiography) is the dominant strategy over

angiogra-phy in the non–Q-wave post-MI patient with resulting lower cost and

improved outcome There is strong evidence to show that

noninva-sive testing prior to angiography is more cost-effective than angiography

alone in patients who have had a nontransmural infarction (strong

evidence)

Figure 19.4. A: Cost-effectiveness of tests for coronary artery disease, in thousands

of 1996 U.S dollars per quality-adjusted life year (QALY), for men at 50% pretest risk for disease B: Cost-effectiveness of tests for coronary artery disease, in thou- sands of 1996 U.S dollars per QALY, for women at 50% pretest risk for disease ECHO, stress echocardiography; ETT, exercise electrocardiography; PET, positron emission tomography; SPECT, single photon emission computed tomography.

[Source: Garber and Solomon (28).]

Figure 19.4. Continued

362 A.J Bierhals and P.K Woodard

An earlier report by Boden et al (33) came to a supporting conclusionregarding patient outcome in the post-MI setting They evaluated theimpact of post-MI angiography in a population with non–Q-wave MIs.Through 2 years of follow-up among the aforementioned patient popula-tion (Fig 19.5), a noninvasive image-directed approach to patient man-agement was found to have a significantly lower mortality and reinfarctionrates than those patients who had undergone an initial angiogram in theacute MI state The findings have been supported by the recommendations

of other groups and researchers (34,35)

Coronary angiography has a specific role in the evaluation of heartdisease that is based on the patient’s clinical history and symptoms Thedata support the position that in an asymptomatic population with a lowclinical suspicion of heart disease, noninvasive testing should be per-formed prior to angiography (26–28,36), whereas in situations were there

is a high clinical suspicion of CAD, angiography should be the initial test

of choice A similar picture develops in post-MI patients For instance,individuals who have had a transmural MI or who have clinical signs ofischemia should undergo a coronary angiogram, but those with a non–Q-wave MI without clinical ischemia would best be evaluated by noninva-sive imaging (29–35) Therefore, the utilization of coronary angiography isbased on the clinical situation and the initial use may not always be themost prudent or cost-effective method to manage patients who are sus-pected of CAD or recently in the post-MI state

Figure 19.5. Kaplan-Meier analysis of the probability of survival according to egy group during 12 to 44 months of follow-up Death from any cause was included

strat-in this analysis The Cox proportional-hazards ratio for the conservative as pared with the invasive strategy was 0.72 (95% confidence interval, 0.51–1.01).

com-[Source: Boden et al (33).]

IV Which Patients Should Undergo Noninvasive

Imaging of the Heart?

Summary of Evidence: There is a moderate amount of support to suggest

that stress echo should be recommended prior to coronary angiography in

the low-risk patients However, several authors have suggested that stress

echo is highly operator dependent and at times SPECT may be a viable

alternative Both modalities have an acceptable cost-effectiveness profile;

as a result, there is insufficient evidence to recommend SPECT over stress

echo More comprehensive cost-effectiveness reports are needed to

com-pletely evaluate these modalities (insufficient evidence)

Supporting Evidence: A few cost-effectiveness evaluations have been

performed incorporating the aforementioned noninvasive studies that

have had some conflicting results A decision analysis was performed by

Kuntz et al (36) that modeled immediate angiography versus a stepwise

approach to angiography In this situation angiography would be

per-formed only if the initial noninvasive test were positive The analysis

incor-porated SPECT, stress echocardiography, and stress electrocardiography

The results indicated that stress echocardiography was more cost-effective

than SPECT in the low-risk population with an incremental cost

effective-ness ratio of $26,800/QALY versus $27,600/QALY, respectively Although

the model does assume an idealized performance of echocardiography,

slight changes in sensitivity of either SPECT or echo affect the results of

the model Thus, decisions concerning the performance of a specific

test should be based on the test characteristics at a given institution (36)

The model also supported the results of other angiographic studies in

which immediate angiography is more cost-effective in the high-risk

patient

Another decision analysis performed by Garber and Solomon (28)

included PET in their analyses along with angiography, stress echo, planar

thallium, exercise electrocardiography, and SPECT The results indicated

that the initial use of stress echo was the most cost-effective followed by

SPECT and angiography (Fig 19.4) Positron emission tomography was

not cost-effective in the diagnosis, resulting in higher cost without

improved outcomes The study also brings to the forefront the idea that

there is variability in cost and performance of SPECT and stress echo;

sub-sequently, SPECT may be the initial modality of choice in some regions

(28,38)

However, a single study evaluating the cost-effectiveness of SPECT

versus exercise electrocardiography was performed to evaluate any

addi-tional prognostic value of SPECT (37) The authors found that SPECT

pro-vided additional information, which translated into $5500 per level of risk

reclassification

Other researchers have also included PET in decision analysis along with

SPECT and angiography (26) The findings of this study contradicted the

prior model, such that PET was found to be the most cost-effective

modal-ity in diagnosing CAD among low-risk patients (28) Aside from the two

prior studies, no other reports were found in the literature review to

eval-uate the cost-effectiveness of PET in the diagnosis of CAD Subsequently,

there is insufficient evidence to recommend PET in the evaluation of CAD

(insufficient evidence)

Similarly, only the previously described studies could be found to uate the cost-effectiveness of stress echocardiography (28,36,38) However,several other studies evaluating the cost-effectiveness of SPECT were iden-

eval-tified in the literature review In a small patient sample (n= 29), SPECTwas found to increase the diagnostic ability in cardiologist who were treat-ing emergency room patients with acute chest pain (39) The study alsofound a decrease in hospitalizations and a savings of $800 per patient (39),although this study had a small sample size and did not rigorously eval-uate cost and outcomes Lastly, Udelson et al (40) assessed the effect ofSPECT in the evaluation of acute chest pain in the emergency department.There was a lower hospitalization rate among patients without coronaryischemia who had undergone a SPECT in the emergency department (42%)versus usual care (52%) The results suggest that SPECT may have an effect

on decision making and possibly lower the costs by reducing tion; however, to date there is insufficient evidence to recommend SPECT

hospitaliza-in the emergency setthospitaliza-ing

In conclusion, multiple decision analyses and randomized studies agreethat in a low-risk patient a noninvasive study should be preformed prior

to an angiogram Also, the models seem to support stress phy as the most cost-effective, but also have suggested that SPECT may be

echocardiogra-as cost-effective depending on the institutional performance Subsequently,there is little definitive data to use one of these studies over the other The use of SPECT or echo should be based on the institutional efficacy.Although there is an early suggestion that SPECT may be useful in theemergent chest pain setting for patient triage, there is not enough data atthis time to support this position Lastly, there is conflicting evidence con-cerning the cost-effectiveness of PET in the diagnosis of CAD and ischemia;more studies are needed to determine the role of PET in the cardiac eval-uation (insufficient evidence)

In symptomatic post-MI patients or those at high risk for CAD, coronaryangiography is the most cost-effective method to evaluate, diagnose, andplan treatments However, among those without symptoms, noninvasivemodalities (i.e., PET, SPECT, and stress echocardiography) are the morecost-effective means to evaluate heart disease But the research to date issomewhat unclear as to the utilization of the aforementioned modalities.The current literature is somewhat limited in the cost-effective evaluations

of noninvasive studies

V What Is the Appropriate Use of Coronary Artery Computed Tomography and Magnetic Resonance?

Summary of Evidence: The newer noninvasive modalities of cardiac MR

and CT have a paucity of cost-effectiveness research and outcomes dataavailable at this time and cannot be recommended for the evaluation ofischemic cardiac disease (insufficient evidence)

Supporting Evidence: In the past decade there have been advances in CT

and MR in the evaluation of many aspects of the heart and heart disease.The current literature has limited data on the performance of MR and CTwith respect to evaluation of the coronary arteries or for assessment of ath-erosclerosis aside from calcium scoring However, our literature reviewfound no reports evaluating the cost-effectiveness of either modality

364 A.J Bierhals and P.K Woodard

Huniak et al (41) performed a decision analysis and developed a model

incorporating current initial diagnostic modalities (i.e., SPECT and stress

echo) prior to coronary angiography In addition, coronary MR and CT

were also included to determine those cost and performance

characteris-tics necessary for the new modalities to possess in order to be

cost-effective For a new diagnostic study to be more cost-effective than stress

echo, a cost of less than $1000 and a sensitivity and specificity greater than

89% and 88%, respectively, should be obtained The results were similar

for replacing SPECT, such that the new imaging study must have a

sensi-tivity and specificity greater than 85% and 80%, respectively Lastly, as

would be expected, a new testing modality required a sensitivity and

speci-ficity of 99% to replace angiography (41) While the prior study is a good

start in the evaluation of the cost-effectiveness of coronary MR and CT,

dedicated studies are required to fully evaluate these aspects of the

imaging modalities, in order to have a complete understanding of their role

in patient care

In addition, as opposed to the traditional modalities, cardiac MRI can

assess simultaneously a multitude of aspects of the heart and cardiac

func-tion Thus, a modality with such versatility may have higher costs that

are offset by evaluating several cardiac dimensions at once, resulting in

a greater cost-effective modality Therefore, studies need to be designed to

address cardiac MR’s role in a complete heart evaluation encompassing

ejection fraction, wall motion, coronary arteries, perfusion, and valvular

disease All of these aspects of cardiac MR have been addressed, but no

single study has encompassed all aspects to evaluate cost-effectiveness

Studies have shown that cardiac perfusion abnormalities can be detected

with similar sensitivity and specificity with MR, SPECT, and PET (42–44)

Cardiac MRI has been found to comparable to stress echo in the

evalua-tion of wall moevalua-tion (44,45) In addievalua-tion, it is better than SPECT in the

assess-ment of myocardial viability as it is of higher resolution and able to

differentiate between subendocardial and transmural infarct Cardiac MR

has also been utilized to evaluate the coronary arteries for aberrant vessel

course and bypass graft complications, all with a relatively high degree of

sensitivity of about 90% (44) Cardiac MR has been found to correlate with

Doppler ultrasound findings in the estimation of valvular area size (46,47)

Aside from the potential utilization for heart disease, cardiac MR has been

shown to have applications for patients with congenital heart disease (48)

that assist with surgical planning and medical management The current

cardiac MR data are extremely promising but remain limited and require

further investigation regarding a future role in patient care

Cardiac CT also suffers from a paucity of data evaluating the

cost-effectiveness in patient management; as a result, its role in patient care

remains unclear Cardiac CT has made great strides over the past 5 years

with the introduction of multidetector scanners, which has improved

res-olution and speed, allowing for improved performance of multiphase and

arterial phase studies These characteristics do provide some advantage

over MR in terms of speed and in the evaluation of stents and patients with

pacemakers But due to the novelty of the modality, the literature remains

more limited than that for cardiac MR Therefore, even before

cost-effectiveness studies can be performed, data must be generated on the

per-formance of cardiac CT Preliminary studies have shown that cardiac CT

can evaluate coronary artery stents (49), and others have used cardiac CT

to evaluate congenital heart disease (50) Also, preliminary data have been

generated in the use of cardiac CT for coronary angiography (51); however,the sample sizes are not substantial enough to generate any accurateassessment of performance.

Recommended Imaging Protocols Based on the Evidence

Cardiac Catheterization

• Selective injection of left coronary artery with at least the projectionsanteroposterior (AP), left anterior oblique (LAO) cranial, and right ante-rior oblique (RAO) caudal is the minimum needed to cover the course

of the left main anterior descending and circumflex arteries

• Selective injection of the right coronary artery with at least the tions lateral, RAO, LAO, and LAO cranial are required to evaluate theright coronary artery

projec-Stress Echo

In a nonpharmacologic stress echocardiogram, the target for an adequatestudy is similar to that of SPECT or a treadmill test Failure to meet thestress limits the sensitivity of the examination The heart rate should reach

at least 85% of predicted However, the study should be terminated ifcardiac symptoms arise or there are ECG changes

• Adenosine is infused intravenously at 140mg/kg/minute over 4 to 6minutes The thallium 201 is injected 3 minutes after infusion Adenosine

is contraindicated in individuals with heart block and bronchospasm

Future Research

• In the future, cost-effectiveness research should focus on incorporatingcalcium scoring and clinical risk stratification in the screening for earlyheart disease Coronary artery calcification scoring has been shown inthe asymptomatic patient to predict a coronary event, but cost-effectiveness has not been adequately evaluated By evaluating calciumscoring in this manner, a determination can be made concerning themodalities’ additional benefits as well as the cost that may be incurred

• Future research should focus on the potential utilization and outcomes

of novel coronary artery imaging modalities, such as CT and MRI Thesemodalities are promising for the evaluation of coronary arteries in mul-tiple clinical circumstances (52) Prior to any cost-effectiveness studies,

an understanding of modality performance characteristics (e.g., tivity and specificity) is needed, along with evaluation of the impact onpatient management and outcome

sensi-366 A.J Bierhals and P.K Woodard

1 Hurst W The Heart, Arteries, and Veins, 10th ed New York: McGraw-Hill, 2002.

2 CDC/NCHS National Health and Nutrition Examination Survey III (NHANES

III), 1988–1994.

3 American Heart Association Annual Report 1996.

4 American Heart Association Heart Disease and Stroke Statistics—2004 update.

5 Health Resources Utilization Branch, CDC/NCHS National Hospital

Discharge Survey.

6 National Health Expenditures, Amounts, and Average Annual Percent Change,

by Type of Expenditure: Selected Calendar Years 1980–2012 cms.hhs.gov.

7 O’Rourke RA, Brundage BH, Froelicher VF, et al., American College of

10 Detrano RC, Wong ND, Doherty TM, et al Circulation 1999;99:2633–2638.

11 O’Malley PG, Taylor AJ, Jackson JL, et al Am J Cardiol 2000;85:945–948.

12 Arad Y, Spadaro LA, Goodman K, Newstein D, Guerci AD J Am Coll Cardiol

2000;36:1253–1260.

13 Raggi P, Cooil B, Callister TQ Am Heart J 2001;141:375–382.

14 Detrano RC, Doherty TM, Davies MJ, Stary HC Curr Probl Cardiol 2000;25:

374–402.

15 Raggi P Herz 2001;26:252–259.

16 Raggi P, Callister TQ, Cooil B, et al Circulation 2000;101:850–855.

17 Shaw LJ, O’Rourke RA J Am Coll Cardiol 2000;36:1261–1264.

18 Detrano R, Hsiai T, Wang S, et al J Am Coll Cardiol 1996;27:285–290.

19 Arad Y, Spadaro LA, Goodman K, et al Circulation 1996;93:1951–1953.

20 Park R, Detrano R, Xiang M, et al Circulation 2002;106:2073–2077.

21 Lipid Research Clinics Program JAMA 1984;251:365–374.

22 Qu W, Le TT, Azen ST, et al Diabetes Care 2003;26:905–910.

23 Newman AB, Naydeck BL, Sutton-Tyrrell K, Feldman A, Edmundowicz D,

Kuller L Circulation 2001;104:2679–2684.

24 Sangiorgi G, Rumberger JA, Severson A, et al J Am Coll Cardiol 1998;31:

126–133.

25 Janowitz WR, Agatston AS, Kaplan G, et al Am J Cardiol 1993;72:247–254.

26 Patterson RE, Eisner RL, Horowitz SF Circulation 1995;91:54–65.

27 Patterson RE, Eng C, Horowitz SF, Gorlin R, Goldstein SR J Am Coll Cardiol

1984;4:278–289.

28 Garber AM, Solomon NA Ann Intern Med 1999;130:719–728.

29 Kuntz K, Tsevat J, Goldman L, Weinstein MC Circulation 1996;94:957–965.

30 FRagmin and Fast Revascularisation during InStability in Coronary artery

disease (FRISC II) Investigators Lancet 1999;354:708–715.

31 Cannon CP, Weintraub WS, Demopoulos LA, et al N Engl J Med 2001;344:

1879–1887.

32 Barnett PG, Chen S, Boden WE, et al Circulation 2002;105:680–684.

33 Boden W, O’rourke R, Crawford M, et al N Engl J Med 1998;338:1785–1792.

34 Pepine CJ, Allen HD, Bashore TM, et al J Am Coll Cardiol 1991;18:1149–1182.

35 Ryan TJ, Anderson JL, Antman EM, et al J Am Coll Cardiol 1996;28:1328–1428.

36 Kuntz K, Fleischmann KE, Hunick MGM, Douglas PS Ann Intern Med 1999;

130:709–718.

37 Hachamovich R, Berman DS, Kiat H, et al Circulation 105:823–829.

38 Fleischmann KE, Hunink MG, Kuntz KM, Douglas PS JAMA 1998;280:913–920.

39 Weissman IA, Dickinson CZ, Dworkin HJ, Oneil WW, Juni JE Radiology 199:

353–357.

40 Udelson JE, Beshansky JR, Ballin DS, et al JAMA 288:2693–2700.

41 Hunink MGM, Kuntz K, Fleischmann KE, Brady TJ Ann Intern Med 1999;131: 673–680.

42 Hartnell G, Cerel A, Kamalesh M, et al AJR 1994;163:1061–1067.

43 Muzik O, Duvernoy C, Beanlands RS, et al J Am Coll Cardiol 1998;31:534–540.

44 Wagner A, Mahrholdt H, Sechtem U, Kim R, Judd R Magn Reson Imaging Clin North Am 2003;11:49–66.

45 Baer FM, Voth E, Larosee K, et al Am J Cardiol 1996;78:415–419.

46 Nishimura T, Yamada N, Itoh A, et al AJR 1989;153:721–724.

47 Aurigemma G, Reichek N, Schiebler M, et al Am J Cardiol 1990;66:621–625.

48 Boxt LM, Rozenshtein A Mag Reson Imaging Clin North Am 2003;11:27–48.

49 Fidler JL, Cheatham JP, Fletcher L, et al AJR 2000;174:355–359.

50 Choi BW, Park YH, Choi JY, et al AJR 2001;177:1045–1049.

51 Herzog C, Dogan S, Diebold T, et al Radiology 2003;229:200–208.

52 Schoepf UJ, Becker CR, Ohnesorge BM, Yucel EK Radiology 2004;232:18–37.

368 A.J Bierhals and P.K Woodard

Aorta and Peripheral

Vascular Disease

Max P Rosen

I Aorta: what are the appropriate imaging studies for suspected acute

aortic dissection or traumatic rupture?

II Aorta: what is the impact and cost-effectiveness of screening for

abdominal aortic aneurysms on mortality from abdominal aortic

aneurysms rupture?

III Aorta: endovascular vs surgical treatment of abdominal aortic

aneurysms: which is the best choice?

IV Peripheral vascular disease: what are the appropriate noninvasive

imaging studies for patients with suspected peripheral vascular

disease?

A Magnetic resonance angiography

B Computed tomography angiography

V Special case: evaluation of abdominal aortic aneurysms graft

endoleak

VI Special case: evaluation of the renal donor

VII Special case: evaluation of renal artery stenosis

369

䊏 Due to the need for rapid diagnosis of patients with suspected acute

aortic rupture or dissection, computed tomographic angiography

(CTA) is preferable to magnetic resonance angiography (MRA)

(limited evidence)

䊏 Screening with ultrasound for abdominal aortic aneurysm (AAA)

among men between the ages of 60 and 74 has been shown to be

cost-effective with a mean cost-cost-effectiveness ratio of £28,400 per life year

gained (strong evidence)

䊏 Endovascular repair of AAA has been shown to significantly reduce

30-day mortality from repair of AAA rupture However, the

proce-dural cost of endovascular repair is greater than that for open

surgi-cal repair (strong evidence)

䊏 Computed tomographic angiography is preferred to catheter

angiog-raphy for detection of aortic stent-graft endoleak (moderate evidence)

Issues

Key Points

Definition, Pathophysiology, and Epidemiology

Imaging of the aorta and peripheral vascular disease poses a unique set ofchallenges and benefits in medical imaging For almost all clinical settings,the gold standard is catheter-based angiography While advances incatheter design and imaging equipment over the past decade have greatlyenhanced the field of diagnostic angiography, the basic tenets of the fieldhave changed little in the past 20 years Thus, there is an extensive body

of literature based on catheter-based imaging With the advent of detector CT scans and concurrent advances in MRA, CTA and MRA havebecome viable alternatives to catheter-based diagnostic angiography.However, unlike any other diagnostic modality, a catheter-based diagnos-tic study may rapidly be converted to an interventional procedure Thus,any new modality for imaging the aorta or peripheral vascular diseasemust be compared to the gold standard of angiography, both for its diag-nostic accuracy and for its cost-effectiveness in the context of immediatelyconverting a catheter-based diagnostic study to a therapeutic intervention.Aortic rupture is usually cased by blunt or penetrating trauma Aorticdissection can be precipitated by traumatic or nontraumatic causes such ashypertension and aortitis; the latter may be infectious or inflammatory innature Aortic aneurysms are caused by a weakening in the aortic wallresulting in either saccular or fusiform dilatation

multi-While most AAAs are the result of atherosclerosis, they may also havetraumatic, infectious, and inflammatory etiologies In men over the age of

65, ruptured AAAs are responsible for 2.1% of all deaths in England andWales (1) Approximately 50% of these deaths occur before the patientreaches the hospital Operative mortality for the 50% of patients with rup-tured AAAs who reach the hospital alive is between 30% and 70%.Peripheral vascular disease is most often caused by hypertension, dia-betes, hypercholesterolemia, or cigarette smoking and can be classified aseither acute or chronic Acute limb ischemia (ALI) is defined as a suddendecrease in limb perfusion that may result in threatened viability of theextremity Chronic manifestations of peripheral arterial disease (PAD) aredivided clinically into (1) intermittent claudication and (2) chronic criticallimb ischemia

Overall Cost to Society

Data on the societal cost of imaging for these indications is not available,except for the cost-effectiveness of screening for AAA with ultrasoundamong men 65 to 74 years of age (see I, below)

370 M.P Rosen

䊏 Computed tomographic angiography is comparable to MRA for uation of peripheral vascular disease and for the preoperative evalu-ation of renal artery stenosis (moderate evidence)

eval-䊏 The most cost-effective imaging strategy for the evaluation of the livingrenal donor varies and is dependent on the perspective of the analysis(renal donor or recipient), as well as the specificity of digital subtrac-tion angiography (DSA) (moderate evidence)

The goals and method of imaging of the aorta and peripheral vascular

branches depend on the clinical setting In the case of suspected traumatic

injury or aortic dissection, the goal of imaging is twofold The most

diate goal is to identify as quickly as possible the patients in need of

imme-diate surgical repair The secondary goal in this acute setting is to help the

surgeon identify the extent of vascular injury and plan the appropriate

repair

The goal of screening asymptomatic patients for AAA is to identify

patients with AAA and provide immediate intervention if the size of the

AAA at the time of screening warrants repair For those patients with AAA,

the size of which does not warrant immediate repair, the goal of screening

is to identify any change in the size of the AAA over time, and to initiate

therapy when the rate of expansion of the AAA reaches a threshold that

justifies repair

When vascular insufficiency or ischemia is suspected, the goal of

imaging is to identify the level and extent of the stenosis or occlusion The

optimal imaging strategy is somewhat dependent on the most likely

method for intervention If a catheter-based intervention is likely, then a

catheter-based imaging study is often warranted as the initial imaging

study On the other hand, if a surgical intervention is likely, then a less

invasive initial imaging study such as CTA or MRA may be optimal

Methodology

PubMed searches for the following index terms were performed from

January 2000 to August 2004: computed tomography (CT) angiography,

mag-netic resonance (MR), vascular studies, arteries, stenosis or occlusion,

angiogra-phy, comparative studies, aneurysms, aortic, cost-effectiveness, and abdominal

aortic aneurysms Relevant articles in English were obtained and read for

appropriateness The search was limited to articles published in January

2000 or later to ensure that only studies employing current noninvasive

technologies would be included Selected articles published before 2000

and after August 2004 (2) were also included at the time of manuscript

review by the book’s editors

I Aorta: What Are the Appropriate Imaging

Studies for Suspected Acute Aortic Dissection or

Traumatic Rupture?

Summary of Evidence: Due to the need for rapid diagnosis of patients with

suspected acute aortic rupture or dissection (Fig 20.1), CTA is preferable

to MRA Most modern emergency departments are equipped with helical

CT scanners, and unlike MRA, CTA of the entire aorta can be performed

in a less than 60 seconds

Supporting Evidence: Yoshida et al (3) assessed the sensitivity, specificity,

and accuracy of CTA among 57 patients with surgically proven type A

dis-section who underwent helical CT, and reported 100% sensitivity of helical

CT to detect aortic dissection in the thoracic aorta Sensitivity for detection

of arch branch vessel involvement was 95% and 83% for detection of cardial effusion (The authors explain that the lower sensitivity for detec-tion of pericardial effusion may be due to the delay between CTA andsurgery, with the pericardial effusion developing during the delay.) Due tothe lack of reported follow-up of the 64 patients in whom the CTA did notshow dissection, this study represents limited (level III) evidence Severalother studies support the use of CTA to exclude aortic injury (4,5) but arebased on older single detector technology Although not commonly avail-able in emergency situations, Pereles et al (6) reported excellent 100% sen-sitivity for diagnosis of thoracic aortic dissection using true fast imagingwith steady-state precision (FISP)

peri-Cost-Effectiveness Analysis: An older paper by Hunink and Bos (7)

pub-lished in 1995 evaluated the cost-effectiveness of CT compared with plainfilm chest radiography and immediate angiography in deciding whenangiography should be performed in hemodynamically stable patientswith suspected aortic injury after blunt chest trauma This study was per-formed before the widespread use of multidetector CT, and investigatedthe use of CT as a triage tool rather than as a definitive diagnostic study.The authors conclude that selecting patients for triage to angiographybased on the CT findings yielded higher effectiveness at a lower cost-effectiveness ratio than doing so based on chest radiographs, and that theincremental cost-effectiveness ratio was $242,000 per life saved for thestrategy of CT followed by angiography for positive cases

372 M.P Rosen

Figure 20.1. Coronal (A) and sagittal (B) computed tomographic angiography (CTA) demonstrating type B aortic dissection Both renal arteries are supplied from the true lumen (arrows).

II Aorta: What Is the Impact and Cost-Effectiveness of

Screening for Abdominal Aortic Aneurysms on Mortality

from Abdominal Aortic Aneurysms Rupture?

Summary of Evidence: The Multicenter Aneurysm Screening Study (MASS)

(1) investigated the impact of ultrasound screening for AAA in a

popula-tion of 67,800 men between the ages of 65 and 74 years The study was a

randomized controlled study conducted at four centers in the United

Kingdom and provides strong evidence that screening for AAA with

ultra-sound significantly reduced AAA related deaths

Supporting Evidence: The MASS group (1) investigated the effect of AAA

screening on mortality in men using a randomized controlled trial design

of 67,800 men aged 65 to 74 years Men in whom AAA (>3cm in diameter)

were detected were followed with repeat ultrasound for a mean of 4.1

years Surgery was considered if the diameter of the AAA was >5.5cm or

if the AAA expanded >1cm per year, or if symptoms related to the AAA

developed Health-related quality of life was measured using the

stan-dardized medical Outcomes Study short-form 36-item survey (SF-36) (8)

and the EuroQol EQ-5D (9) The primary outcome measure was mortality

related to AAA

There were 65 (0.19%) AAA-related deaths in the screened group, and

113 (0.33%) in the control group (p= 0002) with a 53% risk reduction [(95%

confidence interval (CI) 30–64%] among those who underwent screening

Thirty-day mortality following elective surgery was 6% vs 37% following

emergency surgery

Cost-Effectiveness Analysis: Data from the MASS study (1) were used to

esti-mate the cost-effectiveness of AAA screening using ultrasound over a

4-year period and they provide strong evidence Costs included in the

analysis were costs associated with the initial screening program: clinic

staff and study administration, office space, equipment, and costs

associ-ated with any follow-up scans Costs associassoci-ated with surgery were

calcu-lated from the actual costs incurred by the cohort of patients who

underwent surgery and any hospital admission during the 12 months after

surgery No costs related to patient death from aneurysm rupture were

included if the patient had not been admitted to the hospital for attempted

emergency surgery Cost-effectiveness was measured as survival free from

mortality related to AAA for each patient for up to 4 years and was

expressed as incremental cost per additional life year gained

Over 4 years, the mean estimated cost-effectiveness ratio for screening

was $51,000 per life year gained, equivalent to $64,600 per quality-adjusted

life year (QALY) gained

III Aorta: Endovascular vs Surgical Treatment of

Abdominal Aortic Aneurysms: Which Is the Best Choice?

Summary of Evidence: Endovascular treatment of AAA is associated with a

significant reduction in 30-day mortality and hospital length of stay,

com-pared to surgical repair However, the cost of endovascular repair is greater

than that of surgical repair, due to the cost of the endograft (strong evidence)