Gary W. Wood 1. 2

Robert R. Lukin 1

Thomas A. Tomsick 1

A. Alan Chambers 1

This article appears in the March / April 1983 issue of AJNR and the May 1983 issue of AJR.

Received April 14, 1982; accepted after rev i­sion September 24, 1982.

Presented at the XII Symposium Neurorad iolo­gicum, Washington, DC, October 1982.

I Department of Radiology, Un iversity of Cin­c innati Medical Center , 234 Goodman St., Cinc in­nati, OH 45267. Address reprint requests to R. R. Lukin .

2 Present address: Department of Radiology, Albany Medical Center, Albany, NY 12208.

AJNR 4:125-129, March / April 1983 0 195-6108 / 83 / 0402-0125 $00.00 © American Roentgen Ray Society

Digital Subtraction Angiography with Intravenous Injection: Assessment of 1,000 Carotid Bifurcations

125

Digital subtraction angiography (DSA) with intravenous contrast injection was per­formed on 500 consecutive adult patients and evaluated for image quality of the carotid artery bifurcations. Diagnostic quality examinations were obtained in 974 common, 925 internal, and 904 external carotid artery segments. Sixty-two patients had standard carotid arteriography around the same time as DSA. Agreement of standard arterio­grams with diagnostic quality DSA examinations was noted in 97 of 98 common, 94 of 95 internal , and 79 of 91 external carotid artery segments. All cases of complete carotid occlusion (14 of 14) were correctly interpreted by DSA. To identify a population with clinically significant stenosis, a 60% or greater reduction in diameter of the internal carotid was defined as a positive examination . Applying this criterion , the sensitivity, specificity, and accuracy of DSA as compared with standard arteriography was about 94%.

Several clinical tri als of digital subtraction angiography (DSA) with intravenous contrast injection have been published setting preliminary indications for its use [1-5]. Clinical acceptance of this technique has been enthusiastic and many antic ipate results of DSA to rival those of se lective contrast arteriog raphy . Thi s study was designed to determine the foll owing: (1) the frequency with which a diagnostic quality examination of the c.arotid bifurcat ion for vascular occlusive disease can be obtained with DSA; (2) the accuracy of DSA in determining the degree of carotid stenosis using selective arterial angiography as the standard ; and (3) the utility of DSA in defining a population with substantial carotid occlusive disease .

Subjects and Methods

The study popu lation comprised the fi rst 500 patients referred to the University of Cinc innati Medical Center for carot id DSA during a 19 month period. Of these, 62 also had standard caroti d arteriog raphy near the time of the digi tal examinat ion. This yielded 98 carotid bifurcations examined by both methods; no pat ient had both examinations for correlative purposes alone. Typically patients were 50-80 years old and about two-thirds of all cases were outpatients. Indicat ions for DSA inc luded asymptomatic carotid bruits; c linically evident ischemic symptoms; nonspecific, possibly ischemic, symptoms; postsur­gical follow-up, and contrai ndications fo r direct arteriography.

A commercially available digital vascular imaging system (Philips) was used for all DSA examinations, and a prototype system has undergone c linical tes ting at the University of Wisconsin, Madison , and operat ional characteristics have been described [6). Serial imaging at one frame/ sec with four summations / frame was used for neck examinations using radiographic technique of 55-75 kVp and 200-300 mAo The system employs a dual mode 5 or 9 inch (1 2.7 or 22.9 cm) cesiu m-iodide image intensif ier with a Plumbicon camera. Digital conversion and processing occu r instantly with real-time display and postprocessing abi lity on a 256 x 5 12 matrix wi th resolution of 8 bits. Performance spec ifications inc lude 1 % con trast resolution and a 650:1 system signal:noise ratio .

A 16 gauge, 30.5 cm, straight , end-hole cath eter (In tracath, Deseret) , p laced in the

126 WOOD ET AL. AJNR:4, Mar. / Apr. 1983

A 8

basilic or cephalic vein , was used for about 75% of the injec tions. Fluoroscopic contrast testing of ca theter placement was performed to avoid injec tion into inadequate or banch veins. In three or four patients with small or tortuous veins, contrast ex travasation in the soft tissues of the arm resulted in local pain without lasting compli­cation. Early in our experience, many examinations were performed th rough an end- and multi-side-holed 5 French straight Teflon catheter positioned in the superior vena cava, but this method was abandoned after eight instances of mediastinal contrast extravasa­tion. Occasional use of pigtail catheters in the superior vena cava and catheter placement in the inferior vena cava via the femoral route was made without observed complication . Typical injection rates of 8 -1 2 ml / sec of MD-76 (M allinckrodt) (diatrizoate meglu­mine and diatrizoate sodium, 370 mg iodine/ ml) were used for a total volume of 35-50 ml of contrast material followed by a 20-30 ml 5% dextrose in water bolus flush. All examinations were physi­c ian-controlled. Typically three, but from one to five , angiographic sequences in various oblique projections were used per patient.

Each DSA was subsequently interpreted by one of three experi­enced neurorad iologists (R . R. L. , T. A. T. , or A. A. C.). In question­able instances, one or both of the other staff neuroradiologists were consulted for a consensus opinion. Subjective image quality was assessed for each common, intern al, and external carotid artery segment at the bifurcation. Excellent , diagnostic or unsatisfactory ratings were assigned on the basis of artery projection , contrast and spatial resolution, artifac ts, and image correlation among var­ious oblique views (fig . 1). Subsequently, percentage stenosis was measured relative to the normal or projected normal artery lumen. An unsatisfactory rating was given when the arterial segment was overlapped or obscured on all images. Excellent and diagnostic examinations were interpreted fo r deg ree of arterial stenosis. Un­satisfactory examinations were recorded as to patency, but no assessment of stenosis was attempted.

Consistency with other reports was desired at the study outset, and the methods of Chilcote et al. [3] were chosen for grading

c

Fig. 1 .-DSA quality criteria. A, Ex­ce llent visualizati on of carotid bifurca­tion. B, Diagnostic visualization of ca­rotid bi furcation. Misregistrati on artifact from hyoid bone motion (arrow) . C, In­adequate visualization due to patient mo­tion.

TABLE 1: Quality of Demonstration by DSA of Carotid Bifurcation Artery Segment

Segment of Carotid No. Arteries

Artery Bi furca tion Excellent Diagnostic Unsatisfactory

Common 455 51 9 26 External 474 430 96 Internal 482 443 75

arterial stenosis and determining angiographic correlation . Arterial stenosis of each bifurcation element was classified into the following categories: 0 = normal; 1 = minimal stenosis (1 %-20%); 2 = mild stenosis (20%-40%); 3 = moderate stenosis (40%-60%); 4 = moderate to marked stenosis (60-80%); 5 = marked stenosis (80%-99%); and 6 = occluded (100%). DSA was considered correct if the category assigned agreed with standard arteriography.

To identify a population of patients with clinically significant carotid stenosis, DSA was considered positive if the internal carotid artery segment stenosis was equal to or greater than 60% (category 4) [7-9]. Sensitivity , specific ity, and accuracy were determined relative to standard arteriography .

Results

DSA examinations of at least diagnostic quality were obtained in 974 (97%) common , 925 (93%) internal, and 904 (90%) external carotid artery segments at the bifurca­tion (table 1). In no case was the internal carotid artery satisfactorily imaged when the common carotid was not.

The estimation of degree of arterial stenosis with excellent or diagnostic DSA examinations agreed with standard arte­riography in 97 of 98 common , 94 of 95 internal , and 79 of

AJNR :4 , Mar. l Apr. 1983 DSA OF CAROTID BIFURCATIONS 127

A 8

Fig . 2.-Evaluation for carotid bruit. A, DSA shows high-grade (80%) stenosis of proximal internal carotid artery . B , Se lec tive common caroti d arteriogram of same patient.

A 8

9 1 extern al carotid artery seg ments at the bifurcati on. DSA overestimated arteri al stenosis by two grades in each of two cases of common and intern al carotid artery stenosis. Exact correlation of DSA with standard arteriography was ob­served in all g rade 5 (80%-99% stenos is) and grade 6 (complete occlusion) les ions (fi gs. 2 and 3).

Table 2 presents the compari son of intern al carot id DSA with standard arteri ograms in the same fashion as Chil cote et al. [3]. No stenosis was greater in the common than the corresponding internal carotid artery segment.

Standard arteri ography was used to defin e the positi ve and negative pati ent populati ons for carotid occlusive d is­ease at the 60% or greater level of stenosis. DSA identified

TABLE 2: Stenosis of Internal Carotid Arteries Judged Excellent or Diagnostic by DSA Correlated with Standard Arteriography

No. Arteries with Stenosis by OSA ( n = 95) Stenosis by Standard Arteriography - --------------

0 2 3 5 6

0 , Normal 11 4 1, 1%- 20% 10 8 4 2,20%-40% 11 7 2 3, 40%-60% 2 2 3 1 ' 4,60%- 80% 2 4 1-r 5, 80%-99% 9 6, Occluded 14

. OSA overestimation by two grades of arlerial stenosis. t DSA overestimation by one grade of internal carotid stenosis. DSA overestimation of

a lesser grade common carotid stenosis by two grades occurred in this patient.

c Fig. 3. -A, DSA shows right intern al carotid occ lusion (arro w) and sign ificant stenosis o f left internal caroti d artery . Selective right (B) and left (C) common

carotid arteriograms.

128 WOOD ET AL. AJNR:4, Mar. l Apr. 1983

4 5A

28 of 30 true-positives (~60% stenosis) and 61 of 65 true­negatives «60% stenosis) for a sensitivity of 93%, specific­ity of 94%, and accuracy of 94%. These percentages would increase slightly if a clini ca lly significant stenosis were di­fin ed as 40% or greater.

We have now obtained 2, 000 digital angiograms at Cin­cinnati General Hospital. There have been no known related mortalities. Those cases of contrast extravasation have re­sulted in local pain of minutes to hours without lasting sequelae. We know of one case each of the following related to contrast administration: exacerbation of preexisting renal insuffic iency, congestive heart failure, and myocardial in­farcti on 6 hr after the procedure in a patient with a history of infarct. We have encountered a somewhat disturbing incidence of anginal attacks c lose ly related to contrast injecti on (perhaps 5%-10% of patients with known preex­isting angina). Urticari a, nausea, and vomiting occur at about the same rate as in intravenous urography. At present , patients with renal insuffic iency, diabetes, or known ather­oscleroti c heart disease are carefully evaluated as to their indications for the procedure. Most patients tolerate the procedure well (two-thirds are outpatients) and many pa­ti ents have stated that having also experienced standard arteri og raphy, they greatly prefer the digital intravenous examinati on.

Discussion

Results of this study compare favorably with previous reports regarding the accuracy of DSA in defin ing carotid artery stenosis [1 , 2]. High sensitivity, spec ificitY,and accu­racy characterize DSA in identifying pati ents with signifi cant carotid occ lusive disease.

58

Fig. 4.-Elderly woman with nonspe­cific symptoms, normal carotid bifu rca­tion. Left carotid DSA using 7 ml l sec contrast injection via Intracath placed in periph eral arm vein.

Fig. 5. -Middle-aged man wi th facial numbness. A, Carotid DSA shows 80% internal ca rotid stenosis with ulceration. B, Selective common carotid arterio­gram.

Fig. 6. -0blique projection of intracranial c irculation.

In our series, an accurate assessment of internal carotid stenosis at the bifurcation was obtained with DSA in more than 90% of individual artery evaluations. Examinations were performed in a representative clinical practice situation with an occasional uncooperative patient accounting for most of our unsatisfactory examinations. These were mainly due to patient motion and swallowing artifacts. Only two patients were deferred because of their inability to cooper­ate, and in only one were we unable to place a venous catheter properly. This latter patient was a bilateral, above­the-knee amputee with poor arm veins and multiple femoral scars from previous vascular graft placements . Problems such as venous reflux and vessel overlap were usually solved by varying oblique projections .

Contrast resolution was only rarely insufficient for diag-

AJNR:4, Mar./ Apr. 1983 DSA OF CAROTID BIFURCATIONS 129

nostic carotid bifurcation imaging . In two patients this oc­curred with very sluggish venoarterial circulat ion, despite a superior cava catheter and 12 ml / sec injection rate . Insuf­ficient contrast at the carotid bifurcation has not been a problem with peripheral injection sites. We have performed diagnostic DSA with as low as 7 ml / sec injection rates (fig . 4). From the point of view of patient convenience and laboratory throughput, a peripheral injection site is preferred to the centrally placed catheter. Some technical facility is necessary, however, particularly in patients with inadequate veins . In other applications a more compact central bolus contrast injection may be beneficial.

Important to maximizing information is viewing of the entire serial angiographic sequence and the ability to post­process and remask images during the procedure . We frequently return to the video disk or tape for reprocessing and reviewing during final examination interpretation.

We often noted satisfactory DSA imaging of carotid ulcer­ation and intracran ial artery segments during this study, although our experience in these areas has been inconsist­ent to date (figs. 5 and 6). In anticipation of this, we excluded analysis of angiographic features other than carotid bifur­cation stenosis for the purposes of this study.

REFERENCES

1. Strother CM , Sackett JF, Crummy AB , et al. Clinical applica­

tions of computerized fluoroscopy. Th e extracranial carotid

arteries. Radiology 1980;1 36: 781-783 2. Christenson PC , Ovitt TW, Fisher HD III , Frost MM , Nudelman

S, Roehrig H. Intravenous angiog raphy using digital video subtraction: intravenous cervico-ce rebrovascular angiog ra­phy . AJNR 1980;1 :379- 386

3. Ch ilcote WA , Modic MT, Pavliceh WA, et al. Digital subtraction angiography of the carotid arteries: a comparative stud y in 100 patients. Radiology 1981 ; 139: 287 -295

4 . Little JR , Furlan AJ , Modic MT, Bryerton B, Weinstein MA. Intravenous digital subtraction angiog raphy: application to cer­ebrovascular surgery. Neurosurgery 1981 ;9: 1 29-136

5. Seeger JF, Weinstein PR, Carmody RF, Ovitt TW, Fisher HD III , Capp MP. Digital video subtraction angiog raphy of the cerv ical and cerebra l vascu lature. J Neurosurg 1982;56: 1 73-179

6. Ludwig JW, Engels PHC . Early c linical experience with the Philips digital vascu lar imag ing device. In : Mistretta CA, Crummy AB, Strother CM , Sackett JF, eds. Digita l subtraction arteriography: an applica tion of computerized fluoroscopy. Chicago : Year Book Medical, 1982 : 125-1 32

7 . Blaisdell WF, Cl auss RH , Galbraith JG , Imparato AM , Wylie EJ . Joint study of ex tracran ial arterial occlusion. IV . A review of surgica l considerat ions. JAMA 1969;209 : 1 889-1895

8. Sol is OJ, Roberson GR , Taveras JM , Mohr J , Pessin M. Cer­ebral angiography in acu te general infarc tion. Rev In teram Radio/1977 ;2: 19-25

9. Thiele BL, You ng JV, Chikos PM , Hirsch JH , Strand ness DE. Correlation of arter iographic findings and symptoms in cere­

brovascular disease. Neurology (NY) 1980;30: 1 041 -1 046