INDEX An alternative scanning protocol to eliminate perivenous streak artifacts in thoracic spiral computed tomography: The usefulness of ankle vein contrast injection

Masaki Hara, M.D.1
Hiroyuki Ogino, M.D. 1
Masato Itoh, M.D. 1
Satoru Ohba, M.D. 1
Keiichi Andoh, M.D. 1
Masanori Kitase, M.D. 1
Philip C. Goodman, M.D. 2

1 Department of Radiology, Nagoya City University Medical School, 1 Kawasumi, Mizuho-cho, Mizuho-ku, Nagoya, 467-8601 Japan
2 Department of Radiology, Duke University Medical Center, Durham, North Carolina 27710.

Running title
Ankle vein contrast injection for spiral chest CT

Key Words.
Thoracic Computed Tomography (CT), Contrast Enhanced Computed Tomography (CT), Artifact, Ankle Vein Injection.

Abbreviations used: CT, computed tomography; 3D, three dimension; SD, standard deviation; and HU, Hounsfield unit.

Please address all correspondence and reprint requests to: Masaki Hara
Department of Radiology
Nagoya City University Medical School
1 Kawasumi, Mizuho-cho, Mizuho-ku, Nagoya, 467-8601 Japan
Phone: 81-52-853-8276
Facsimile: 81-52- 852-5244

Purpose. To eliminate perivenous streak artifact from contrast-enhanced thoracic spiral CT using superficial ankle vein injection (ankle vein method).
Methods. Forty-four consecutive inpatients had thoracic spiral CT following ankle vein method and were compared to 30 patients who had conventional antecubital vein injection. Non-ionic 300 mg/mL contrast material were injected at 1-1.5 mL/sec with a power injector. Twenty mL of normal saline was injected immediately thereafter as a flushing bolus. Qualitative scores of perivenous artifact (1 to 5 = extensive) and vascular enhancement (1 to 5 = excellent) were recorded. Complications were investigated clinically.
Results. The mean score of perivenous artifact was 1 with the ankle vein method and 4.4 with the conventional method (P<.0001). Scores of pulmonary artery opacification were almost equal (4.2 and 4.5 respectively, P =.11). No complications were observed.
Conclusions. The ankle vein method is an effective method to prevent perivenous artifact during thoracic spiral CT.

The popularity of contrast enhanced thoracic spiral CT has led to a reevaluation of the occasional image-degrading artifacts caused by antecubital vein injection of iodinated contrast medium. Rubin and colleagues (1) recently demonstrated that contrast concentrations of 150 mg/mL iodine injected at a rate of 2.5 mL/sec reduced perivenous artifact while maintaining adequate vessel enhancement. Yet they reported Hounsfield numbers of approximately 400, 800, and 1150 in the superior vena cava, brachiocephalic vein and subclavian vein respectively. At these levels some streak artifacts will occur and there may still be difficulty evaluating intravascular lesions (Fig.1.) with a risk of the extravasation of contrast media using relatively high flow rate. Prokop and colleagues (2) have shown that while concentrations of over 150 mg/mL iodine produced streak artifacts, even levels below 150 mg/mL iodine can cause minor artifacts which can degrade the quality of 3D images. More diluted 60 mg/mL iodine concentrations with high flow rates of 7 mL/sec have been recommended in another report (3), but this suggestion has not been widely accepted.
Administration of contrast medium through a lower extremity injection site may be a way to eliminate intrathoracic perivenous artifacts, but has been avoided due to the large potential space for pooling of iodinated material and also because of the risk of lower extremity thrombosis.
With spiral CT however data acquisition time can be shortened to a single breath hold and thus the contrast material need not be in contact with lower extremity veins over an extended time thus reducing the possibility of complications (4). In this study we evaluated the value of ankle vein injection for contrast-enhanced thoracic spiral CT.

Materials and Methods
Between November 1995 and March 1997, 44 consecutive pre-surgical inpatients were examined by spiral CT using an ankle vein injection method after obtaining informed consent from the each patient. As control subjects 30 patients studied by conventional antecubital vein injection in the same period were enrolled.
Two patients with a past history of lower extremity thrombophlebitis or varices were excluded from this study. There were 10 women and 34 men (age range, 17-77 years; average age, 59.3 years) in the study group. There were 8 women and 22 men (age range, 34-81 years; average age, 59.3 years) in the control group.
CT scans were performed with a Somatom Plus S scanner (Siemens Medical Instrumentation, Erlangen, Germany). Spiral scans were obtained at 120 kVp and 165 mAs with 2-10 mm collimation and pitch of 1.0 within a single breath hold. All images were reconstructed at intervals of 10 mm using a algorithm which was suitable for both mediastinum and lung, and were viewed at lung (window width, 1550 HU; window level, - 550 HU) and mediastinal (window width, 280 - 400 HU; window level, 40 - 100 HU) settings. The CT scans were interpreted by two thoracic radiologists (M.H., K.A.), with decisions made by consensus. Each patient received 300 mg/mL nonionic iodinated contrast medium. Patients examined by ankle vein injection were divided into two groups. Because our initial preliminary group (n=7) received injection amount of 100 mL contrast medium at 1.5 mL/sec caused relatively high contrast enhancement within the aorta (approximately 200 HU on the average) (Fig. 2a), we reduced the amount of contrast medium. The other following group (n=37) received 70 mL contrast medium at 1 mL/sec. Ankle vein access was achieved with a 21 G scalp vein needle inserted into a great saphenous or medial marginal vein. A power injector was used. After contrast administration 20 mL of normal saline was injected over 10 seconds to clear lower extremity veins of contrast material. A scan delay of 60 seconds was selected for both groups because of relatively large venous capacity of the inferior vena cava and lower extremity .
Patients examined by antecubital vein injection received 100 mL contrast medium at 1.5 mL/sec to reduce the risk of extravasation. Scanning was initiated after at least 60 mL of contrast medium had been administered, following a scan delay of approximately 40 seconds without injection of normal saline. Venous access was achieved with a 21 G or 18 G scalp vein needle.

The CT numbers of bilateral brachiocephalic veins, superior vena cava, pulmonary artery, aorta, and great vessels were measured as continuously as possible in all 7 preliminary cases with 100 mL injection, in 13 of 37 cases with 70 mL ankle vein injection and in 16 of 30 cases with conventional antecubital vein injection. Because we had limited access to the CT scanner, complete measurements of enhancement for 74 patients could not be made. Mean time-density curves of the studied patients were calculated. Opacification of the pulmonary artery and superior vena cava was classified as follows: 1= poor, insufficient for diagnosis; 2= fair, bare minimum for diagnosis; 3= good, diagnostic quality; 4= very good, between 3 and 5; 5= excellent, high degree of vascular opacification. Perivenous artifacts were classified as follows: 1= none; 2= mild, between 1 and 3; 3= moderate, partially obscuring adjacent structures; 4= between 3 and 5; 5= extensive, completely obscuring adjacent structures.
Patients were examined for localized swelling, redness and clinical signs of thrombosis at the injection site immediately after the CT scan and 24 hours later.
A Mann-Whitney test, and P values were calculated for each comparison.

With the 100 mL ankle vein method, on average the thoracic aorta enhanced more than 200 HU by the start of scanning (Fig. 2a). The superior vena cava enhanced from 40 - 50 HU to about 150 HU after contrast administration without any artifacts (Fig. 2b).
On average using the 70 mL ankle vein method the thoracic aorta enhanced more than 150 HU by the start of scanning (Fig. 3a). Peak opacification in the pulmonary artery was observed between 12 and 23 seconds (Fig. 3b). Bilateral brachiocephalic veins and the superior vena cava enhanced from 40 - 50 HU to 100 HU after contrast administration (Fig. 3c).
With the 100 mL ankle vein method, the aorta and superior vena cava achieved an additional 50 HU enhancement compared with the 70 mL method.

Perivenous artifacts (Table 1)
The mean scores of perivenous artifacts with 70 mL ankle vein, 100 mL ankle vein, and conventional antecubital injections were 1, 1, and 4.4 respectively (Fig. 3c). A statistically significant difference in observed artifacts was noted when comparing ankle vein with antecubital injections (P <.0001).

Pulmonary artery opacification (Table 1)
The mean scores for pulmonary artery opacification with 70 mL ankle vein, 100 mL ankle vein, and antecubital vein injection were 4.1, 5.0, and 4.5 respectively (Fig. 3b). The combined average score for 70 mL and 100 mL ankle vein injection was 4.2, so that the antecubital method was slightly superior but not statistically significant (P =.11).

Superior vena cava opacification (Table 1)
The mean scores of superior vena cava opacification with 70 mL ankle vein, 100 mL ankle vein, and antecubital vein injection were 2.5, 3.7, and 4.9 respectively. Even though the 100 mL ankle vein method did not opacify the superior vena cava as well as the antecubital technique it was most diagnostic considering the absence of artifacts with the ankle vein technique.

No complications at the injection site and no episodes of lower extremity thrombosis were encountered immediately after the CT scan and 24 hours later.

Contrast enhancement for thoracic CT may be valuable for some indications, such as when improved discrimination between vessels and surrounding structures is necessary (e.g. hilar abnormalities). Furthermore at equilibrium contrast enhancement of suspected lesions may occur.
For conventional CT 100 mL or more of contrast medium is the standard amount injected. A recent study revealed that a reduced amount, 60 mL, of contrast would suffice for spiral CT (5). However other investigators still recommend that the usual greater amount and concentration of contrast medium be used to evaluate vessels and/or the internal structure of solid lesions (6)(7).
On contrast enhanced thoracic spiral CT examinations performed with standard amounts of undiluted contrast, perivenous artifacts emanating from the intrathoracic veins can seriously degrade the images. Several techniques using diluted contrast medium have already been investigated in an attempt to improve image quality in this setting (1)(2)(3)(8)(9). Kodaira and colleagues (8) described the use of 60 mL of 240 mg/mL iodinated contrast at an injection rate of 2 mL/sec. They achieved CT numbers in the superior vena cava of 279 HU +- SD 111 but were not able to eliminate perivascular artifacts well enough to permit evaluation of intravascular abnormalities. Rubin and colleagues (1) used 100 mL of 150 mg/mL iodinated contrast at an injection rate of 2.5 mL/sec producing superior vena cava enhancement up to 400 HU However perivascular artifacts still occurred. Lesser concentrations of iodine or smaller injection rates were examined by Storto and colleagues (9) who emphasized that a reduction in iodine concentration below 200 mg/mL or an infusion rate less than 2 mL/sec through an antecubital vein significantly worsened image quality.
For thoracic applications, a lower extremity approach can reliably and totally eliminate perivenous streak artifacts. Although the inferior vena cava has been regarded as a large potential reservoir of contrast medium, in our study we obtained excellent contrast images except in one patient with hypothyroidism. A flushing bolus injection of 20 mL of normal saline was effective in decreasing pooling of contrast medium in the inferior vena cava. A scan delay of 60 seconds resulted in successful contrast enhanced imaging but the timing may have to be adjusted in patients with slower circulations such as our patient with hypothyroidism and perhaps in patients with cardiovascular disease. In addition to eliminating perivenous streak artifact there was another advantage to the ankle injection method. Unlike the conventional antecubital injection technique where contralateral veins remain devoid of contrast, lower extremity injections and larger scan delays result in bilateral brachiocephalic vein and superior vena cava enhancement to approximately 100 and 150 HU respectively. Thus the ankle vein method might be useful for cases of lung apex and thoracic inlet diseases. Femoral venous injections have been proposed (3) but this method is more invasive and time-consuming.
The most anticipated problem about the ankle vein method has been the possibility of post-injection thrombophlebitis and thrombosis. The incidence (> 50%) of this complication has already been established for phlebography using ionic high osmolar contrast medium (4)(10)(11)(12). However this risk diminishes to 0 - 7 % using nonionic low osmolar solutions (10)(11)(12). In this group of patients, most had surgery shortly after their CT scans. In order to distinguish complications of the ankle vein injection from those of surgery we had to limit our follow-up period to 24 hours after CT. Although there were no complications in our investigation, further trials are essential to establish the safety of this method as we excluded high risk patients and the number of patients was limited. We believe the technique will prove safe and that the lack of complications in this procedure as compared to phlebography was due to the diminished time of contact between the vessel walls and contrast medium. With upright ascending phlebography this time amounts to several minutes (4,10) whereas with the ankle vein method contrast is out of the lower extremity vessels within 70 seconds.
In conclusion we suggest that ankle vein injection technique might be useful for contrast enhanced CT scanning in situations (e.g. intravascular lesions, mediastinal abnormalities near the great veins) where streak artifacts are likely to produce significant image degradation.

1. Rubin GD, Lane MJ, Bloch DA, Leung AN, Stark P. Optimization of Thoracic Spiral CT: Effects of iodinated contrast medium concentration. Radiology 1996; 201: 785-791.
2. Prokop M, Schaefer CM, Leppert AGA, Galanski M. Spiral CT angiography of thoracic aorta: femoral or antecubital injection site for intravenous administration of contrast material? (abst.) Radiology 1993; 189(P): 111.
3. Remy-Jardin M, Remy J, Wattinne L, Giraud F. Central Pulmonary Thromboembolism: Diagnosis with Spiral Volumetric CT with the Single-Breath-Hold Technique-Comparison with Pulmonary Angiography. Radiology 1992; 185: 381-387.
4. Coel MN, Dodge W. Complication rate with Pupine Phlebography. AJR 131:821-822, 1978.
5. Costello PC, Duppy DE, Ecker CP, Tello R. Spiral CT of the Thorax with Reduced Volume of Contrast Material: A Comparative Study. Radiology 1992; 183: 663-666.
6. Swensen SJ, Brown LR, Colby TV, Weaver AL, Midthun DE. Lung Nodule Enhancement at CT: Prospective Findings. Radiology 1996; 201: 447-455.
7. Zhang M, Kono M. Solitary Pulmonary Nodules: Evaluation of Blood Flow Patterns with Dynamic CT. Radiology 1997; 205: 471-478.
8. Kodaira T, Shimamoto K, Hirota H, Itoh S, Itoh K, Kamata N, et al. Optimal Concentration of Contrast Medium in Helical CT of the Thorax. Nippon Igaku Hoshasen Gakkai Zasshi - Nippon Acta Radiologica 1995; 55:138-44.
9. Storto ML, Ciccotosto C, Patea RL, Spinazzi A, Bonomo L. Spiral CT of the Mediastinum: Optimization of Contrast Medium Use. Eur. J. Radiol. 1994; 18(suppl): S83-S87.
10. Albrechtsson U, Olsson CG. Thrombosis after Phlebography: A Comparison of Two Contrast Media. Cardiovasc. Radiol 1979; 2: 9-18.
11. Walters HL, Clemenson J, Browse NL, Thomas ML. 125I-fibrinogen Uptake Following Phlebography of the Leg: Comparison of Ionic and Nonionic Contrast Media. Radiology 1980; 135: 619-621.
12. Bettmann MA, Robbins A, Braun SD, Watzner S, Dunnick NR, Finkelstein J. Contrast Venography of the Leg: Diagnostic Efficacy, Tolerance, and Complication Rates with Ionic and Nonionic Contrast Media. Radiology 1987; 165: 113-116.
Fig. 1. Helical CT scan in patient with lung cancer in the left lower lobe shows a conspicuous artifact on the superior vena cava caused by high- concentration contrast material.

Fig. 2a, b. With the 100 mL ankle vein method, the aorta (a) aorta enhanced more than 200 HU on average by the start of scanning. The superior vena cava enhanced from 40 - 50 HU to 150 HU after contrast administration.

Fig. 3a. The thoracic aorta initially enhanced to approximately 150 HU with the 70 mL ankle vein method.

Fig. 3b. The peak of pulmonary artery opacification was observed between 12 and 23 seconds after beginning the scan.

Fig. 3c. Bilateral brachiocephalic veins and superior vena cava enhanced from 40 - 50 HU to 100 HU after contrast administration.
Table 1. Results of grading of perivenous artifacts and opacification of aorta, pulmonary artery and superior vena cava

Figs. 4. A 54-year-old man with squamous cell carcinoma of the right lower lobe was examined using the 70 mL ankle vein method (a, b, c, d). The score of superior vena cava artifacts and pulmonary artery, and superior vena cava opacification were 1, 5, 2 respectively. Image quality was considered adequate for diagnosis. The maximum CT number of the pulmonary artery was 238 HU. The time-density-curve of this case is shown (e).

Figs. 5. A 77-year-old woman with transverse colon carcinoma was examined using the 70 mL ankle vein method to evaluate lung metastases (a, b, c). The score of superior vena cava artifacts and pulmonary artery and superior vena cava opacification was 1, 5, 3 respectively. An incidental thrombosis of the right pulmonary artery was observed (arrow).

Nagoya City University
Medical School