Contents lists available at ScienceDirect

Journal of Radiology Nursing

journal homepage: www.radiologynursing.org

Journal of Radiology Nursing 35 (2016) 261e274

Carbon Dioxide Digital Subtraction Angiography: Everything You Needto Know and More

Lorena Garza, MD a, *, Christian Fauria, MD, MSW, MPH b, James G. Caridi, MD, FSIR c

a Department of Radiology, Tulane University Medical Center, New Orleans, LAb Tulane University Medical Center, New Orleans, LAc Tulane University Medical Center, New Orleans, LA

Keywords:Carbon dioxideContrast allergyNephrotoxicity

* Corresponding author: Lorena Garza, Departmentsity Medical Center, 430 South Pierce Street, New Orl

E-mail address: lgarzaga@tulane.edu (L. Garza).

This activity has been submitted to AlNurses Association is accredited as anCenter’s Commission on Accreditatio

http://dx.doi.org/10.1016/j.jradnu.2016.09.0041546-0843/$36.00/Copyright © 2016 by the Associati

a b s t r a c t

In 1971, during a routine celiac axis injection, 70 cc of room air was inadvertently injected into a patientinstead of iodinated contrast. Fortunately, there were no ill effects, and despite the use of cut film at thetime, Dr. Hawkins visualized the celiac axis and its branches as a negative image. Because of this incident,in combination with his previous knowledge of carbon dioxide (CO2) in venous imaging, he began tostudy the intra-arterial use of CO2 in animals. After the safe and successful use in animals, he applied thesame principles to humans. As technology continued to improve, CO2 evolved into a viable vascularimaging agent. Although used initially for renal failure and iodinated contrast allergy, the many uniqueproperties of CO2 yielded multiple advantages, which are now used in a multitude of scenarios alone orin combination with traditional contrast. Despite somewhat awkward delivery devices in the past, it hasnow been used with great success in both adults and children for more than 3 decades with only limitedreportable complications. Its safe use in children has been described, and when performed in this agegroup, the same principles apply as for adults. This article describes the history and technique of CO2

angiography for vascular procedures.Copyright © 2016 by the Association for Radiologic & Imaging Nursing.

History

It was not long after the discovery of X-rays by Conrad Roentgenin 1895 that gas was first used as an imaging agent. In 1914, roomair was used with radiographs in an attempt to visualize intra-abdominal contents (Rotenberg, 1914). Less than a decade later,room air, oxygen, and carbon dioxide (CO2) were insufflated in theretroperitoneum to evaluate for masses (Carelli & Sorddelli, 1921;Rosenstein, 1921). Because of the problem of air emboli, room airand oxygen, less soluble than CO2, were eventually replaced withCO2.

In the 1950s and 1960s, CO2 was used as a venous contrast agentto evaluate for pericardial effusion (Bendib, Tourni, & Boudjella,

of Radiology, Tulane Univer-eans, LA 70119.

abama State Nurses Assoapprover of continuing

n.

on for Radiologic & Imaging Nursin

1977; Paul, Durant, Oppenheimer, & Stauffer, 1957; Phillips,Burch, & Hellinger, 1961; Scatliff, Kummer, & Janzen, 1959).Bendib et al. (1977) performed 1,600 cases without complications,using a peripheral injection of 100 to 200 cc of CO2 to evaluate forpericardial effusion (Figure 1). In addition, in 1969, Hipona, Ferris,and Pick (1969) reported the safe use of CO2 for the evaluation ofthe inferior vena cava (IVC).

Subsequently, the biggest advancement in the use of CO2 as acontrast agent was the fortuitous epiphany by Hawkins (Figure 2).After numerous safety studies in animals, he applied these princi-ples to adults and children (Caridi et al., 2003; Hawkins & Caridi,1998; Kriss, Cottrill, & Gurley, 1997). As technology continued toimprove, CO2 evolved into a viable vascular imaging agent.Although used initially for renal failure and iodinated contrast al-lergy, the many unique properties of CO2 yielded multiple advan-tages, which are now used in a multitude of scenarios alone or incombination with traditional contrast.

ciation for approval to award contact hours. Alabama Statenursing education by the American Nurses Credentialing

g.

Figure 1. Left lateral decubitus chest X-ray after the administration of intravenouscarbon dioxide for pericardial evaluation.

Table 1Physical properties of CO2 and physiologically related gases

Properties CO2 O2 N2

Molecular weight 42 32 28Solubility 0.87 0.03 0.016

CO2 ¼ Carbon dioxide; O2 ¼ oxygen; N2 ¼ nitrogen.

L. Garza et al. / Journal of Radiology Nursing 35 (2016) 261e274262

Properties

To use CO2 appropriately, a basic knowledge of its properties isessential. CO2 is a nontoxic, nonflammable, buoyant, andcompressible gas that has low viscosity and is produced endoge-nously at approximately 200 to 250 cc per min. It is a naturalbyproduct, and there are approximately 120 L of CO2 stored in thesoft tissues at one time (Roussos & Koutsoukou, 2003).

Because CO2 is present endogenously, there is no concern forallergy or renal toxicity (Hawkins,1982; Hawkins& Caridi,1998). Itsviscosity is 1/400 that of iodinated contrast, and therefore, it is lessocclusive than other gases (Table 1). When administered intravas-cularly, it tends to dissolvewithin a vessel in 30 s. If CO2 persists in avessel for more than this, it is either trapped or there is room aircontamination.

As opposed to traditional liquid agents, CO2 does not mix withblood. To render a representative image, it must displace the bloodin the vessel. As a result, the vessel is less dense, and a negativeimage is obtained with digital subtraction angiography (DSA). Thequality and accuracy of the image will depend on the amount of

Figure 2. The arrow points to negative image from room air in the celiac circulation.

blood displaced by the CO2 (Figure 3). Typically, smaller vessels,especially those 10 mm or smaller, demonstrate a better correlationwith iodinated contrast (Figure 4) (Ehrman, Taber, Gaylord, Brown,& Hage, 1994; McLennan et al., 2001; Moresco et al., 2000).

Buoyancy must also be considered when using CO2. Typically, agas will rise to the nondependent surface of the vessel. If all theblood is not displaced from the vessel, it will readily demonstratethe anterior structures but potentially generate a spuriously smallerimage of the larger feeding vessel. It is imperative to displace asmuch blood as possible to generate a comparable image.

In addition to its buoyancy, when CO2 is administered into thevessel via a catheter, it has the potential to fragment into randombubbles depending on how it is delivered. In an attempt to avoid this,the catheter should be purged before definitive delivery, and acontinuous and controlled delivery of the volume of choice should begiven. Dr. Cho studied the best catheter to administer a uniform andorganized bolus of gas to minimalize the bubbling effect. He foundthat an end-hole catheter yielded the best results (Figure 5) (Cho,2007).

Againwith respect to buoyancy, there is one instance inwhich itcan be a significant detriment. When a blood gas interface is notpresent in an anterior structure such as an abdominal aorticaneurysm, the CO2 may sit without dissolving and trap in that po-sition (Figure 6). Trapping could potentially lead to ischemiamanifested by pain. The reported incidence of this occurrence isnegligible.

The incidence of gas trapping most commonly arises if a typicallarge tank (usually 3 million cc) of CO2 under pressure is mis-connected to the delivery catheter allowing unfettered flow of gasinto the vessel. It is best to wait at least 30 to 60 s between in-jections to allow for CO2 to be dissolved.

As with any interventional procedure, patients undergo routineprocedural monitoring. As a safety measure, Cho (2007) recom-mends monitoring blood pressure 1, 2, and 3 min after the first CO2injection.

Asides buoyancy, another important divergent property of CO2when compared with liquid agents is its compressibility. BecauseCO2 is a gas it is extremely compressible under pressure. A 20 ccsyringe can hold a volume of 200 cc if compressed sufficiently.Therefore, the delivery system should be purged to atmosphericpressure to eliminate the compression and avoid administering alarger dose than the amount indicated on the syringe.

In addition, we have found that when a compressed dose isdelivered, it can be explosive causing pain and degrade acquiredimages. Compression can also cause CO2 to reflux into inappro-priate vessels, potentially resulting in complication. This is espe-cially true of the cerebral vessels in which CO2 should be avoided.Finally, to completely avoid explosive delivery, the delivery cathetershould be purged of saline or blood before the definitive dose andnot be under pressure.

Whatever system is being used should be purged to the atmo-sphere for equilibrium. After this, the delivery syringe plungershould be gently advanced to purge the diagnostic catheter of salineor blood. Subsequently, the CO2 can be delivered in a controlled andnonexplosive manner. Note that unlike a liquid contrast syringe, inwhich a smaller syringe generates more pressure, a larger syringe(usually 20 to 35 cc) should be used with CO2. If not, the gas may

Figure 3. Comparison of carbon dioxide and contrast, respectively. Less than 10 mm vessels yield comparable images. This is a good quality CO2 image given adequate amounts ofblood were displaced.

L. Garza et al. / Journal of Radiology Nursing 35 (2016) 261e274 263

simply compress in the syringe without delivery, especially if thecatheter is not purged.

Invisibility is the most significant property of CO2 and the onethat causes the most concern and discourages competent operatorsfrom using it. Because it is invisible, contamination with moreocclusive room air is a major concern of many operators. Knowl-edge of the sources of contamination and how to avoid them isimperative but relatively simple. It is essential, therefore,that disposable sources of at least medical grade CO2 be used(Table 2).

Contaminationmay also result when a syringe of CO2 is left opento room air. This can be avoided with the use of a closed deliverysystem that is not open to room air. Delivery systems, especiallythose that use a bag reservoir, should be purged three times to ridthe system of residual room air. Finally, delivery system connectionsshould be secured, either with glue or luer lock. Loose connectionscan lead to the aspiration of room air. Some individuals also hy-pothesize that stopcocks are not impervious to gas introductionwith aspiration. Therefore, stopcocks and aspiration should be keptto a minimum.

Contraindications, caveats, and disadvantages

When using gas as an imaging agent, the biggest concern is thepossibility of embolic occlusion and secondary ischemia. The rapid

Figure 4. Anterior vessels are well demonstrated by carbon dioxide digital subtraction angtransplant renal arteries.

solubility of CO2 allows for intravascular use, but the coronary, thoracicaortic, and cerebral vessels are less forgiving, and delivery into thesevessels should be avoided. In addition, because of the tendency of CO2to reflux, it is prudent to avoid intra-arterial injections above the dia-phragm. To reduce the possibility of central cerebral reflux, the patientcan be placed in the Trendelenburg position.

Other clinical scenarios that predispose a patient to untowardembolization include right-to-left shunts and the combination ofpulmonary arterial hypertension and a patent foramen ovale. Anadditional rare contraindication is the use of nitrous oxide generalanesthesia when using intravenous CO2.

A recurrent concern for novice operators is the use of CO2 inpatients with chronic obstructive pulmonary disorder (COPD). As aprecautionary measure in COPD patients, it is suggested to allowmore time between injections. Instead of the recommended 30 to60 s between injections in most routine patients, those with COPDshould be increased to 2 min to allow for definitive dissolution.

It is important to remember the potential increase in radiationexposure to the operator and the patient when using CO2 DSA. It isrecommended that the frame rate for acquiring CO2 imagesapproximate 6/s or more.

In addition to the aforementioned contraindications, there are afew minor disadvantages that exist when changing from a fluid-based vascular contrast to CO2. The primary disadvantage islearning how to use a gas-based delivery system as opposed to

iography. (A) Superior mesenteric artery celiac, (B) reimplanted renal arteries, and (C)

Figure 5. End hole vs multiside hole catheters. CO2 delivery via end hole catheters (column A) demonstrates a more uniform column of CO2 than those catheters with multipleside holes (columns B and C).

L. Garza et al. / Journal of Radiology Nursing 35 (2016) 261e274264

iodinated contrast. CO2 is invisible, colorless, odorless, and cannot beseen or felt; therefore, the comfort level for use is much less thaniodinated contrast. The operator must learn and feel confident in thetype of delivery system. In addition, CO2 vascular imaging is typicallynot as dense, and patient motion can seriously affect the finalproduct.

Advantages

Nonallergic Quality of CO2

Because CO2 is disseminated throughout the body's soft tissues,it is nonallergenic. This advantage is extremely beneficial when anoperator is confrontedwith an emergent intravascular procedure ina patient with a severe allergy to iodinated contrast. Currently, mostoperators use low osmolar nonionic contrast, which has an inci-dence of allergic reaction of 0.7% to 3% and severe anaphylaxis of

Figure 6. (A) Cross table lateral showing carbon dioxide (CO2) pooling anteriorly in abdominan AAA.

0.02% to 0.04% (Caro, Trindade, & McGregor, 1991; Lieberman &Seigle, 1999; Trcka et al., 2008).

Non-nephrotoxic Quality of CO2

Undoubtedly, the best advantage of CO2 DSA is its lack ofnephrotoxicity (Beese, Bees, & Belli, 2000; Caridi, Stavropolous,Hawkins, 1999; Hawkins & Caridi, 1998; Hawkins, Cho, &Caridi, 2009; Hawkins et al., 1992; Hawkins, Wilcox, Kerns, &Sabatelli, 1994; Thomson et al., 1996). Contrast-induced ne-phropathy (CIN) is the third leading cause of hospital-acquiredrenal failure behind decreased renal perfusion and nephrotoxicmedications (Nash, Hafeez, & Hou, 2002). The incidence ofhospital-acquired CIN approximates 7% (Bartholomew, Harjai, &Dukkipati, 2004).

Early animal studies by Hawkins (1982) and others showedthat CO2 as an intravascular contrast agent did not affect therenal function. Hawkins later went on to demonstrate this in

al aortic aneurysm (AAA). (B) Delayed image demonstrating persistent CO2 anteriorly in

Table 2Grades of CO2

Type Purity (%)

Research grade 99.999SFC grade 99.998Instrument grade 99.99Coleman grade 99.99Anerobic grade 99.9Medical grade 99.5Standard grade 99.5

CO2 ¼ Carbon dioxide; SFC ¼ Supercritical Fluid Chromatography.

L. Garza et al. / Journal of Radiology Nursing 35 (2016) 261e274 265

humans as well. Comparing iodinated contrast, gadolinium, andCO2 in renal-insufficient patients, CO2 was the only agent notdemonstrating an elevation in creatinine (Spinosa et al., 1998,2000). It should be the first-line imaging agent in patientswith renal insufficiency requiring vascular evaluation or inter-vention. Even if there are limitations to the CO2 imaging, it canbe used in conjunction with limited diluted doses of iodinatedcontrast.

Low Viscosity

The viscosity of CO2 is 1/400 that of iodinated contrast,permitting its delivery through smaller and less invasive cathetersand needles. This is advantageous when using microcatheters inwhich sufficient volumes of thicker contrast may be difficult todeliver. CO2 can easily be administered in significant doses and hasthe advantage of central reflux resulting in opacification of theentire vascular structure (Figure 7).

Likewise, CO2 can be injected through (ultrafine) needles assmall as 27 gauge. These needles are much less invasive and havebeen used successfully with CO2 in the liver and spleen as well asperipheral venography (Figure 8).

Because of its low viscosity, CO2 can be injected through acatheter with the wire in place using a Y-adapter (Figure 9). This isadvantageous when performing invasive procedures and at timeswhen it is preferable to maintain wire access.

Other

Finally, one of the biggest advantages of CO2 is its cost. Thetypical cost for CO2 is 3 cents per 100 cc, which is exponentiallycheaper than iodinated contrast.

Figure 7. (A and B). The low viscosity of carbon dioxide permits delivery between wire andrequire removal of the guidewire.

Indications: alone or as an adjunct to iodinated contrastallergy

Iodinated contrast allergy is explained previously. Emergentpatients or those who for a variety of reasons did not receiveprednisone preparation can use CO2 DSA if necessary.

Contrast-induced Nephropathy

There are many clinical scenarios that predispose a patient toCIN. These include myeloma, diabetes, acute cardiac abnormalities,hypotension, nephrotoxic drugs, and underlying renal disease. Inthese patients, the incidence of CIN is increased but can be lessenedwith the use of CO2 DSA. Most commonly, the incidence of CIN isrelated to the volume of iodinated contrast, the route of adminis-tration (intra-arterial carries higher risk than intravenous), and pre-existing renal insufficiency (Kooiman et al., 2012; Moresco et al.,1998; Murakami et al., 2012; Seeger, Self, Harward, Flynn, &Hawkins, 1993).

When performing procedures requiring significant volumes ofcontrast, CO2 can be used alone or as an adjunct to decrease thispossibility. It should be the preferred contrast in evaluation andintervention of the renal artery in renal transplants and renalarterial reimplantation cases (See Figure 11).

Hemorrhage

CO2 is extremely beneficial in the demonstration of acute arte-rial and venous hemorrhage (Figures 10 and 12) and is 2.5 timesmore sensitive than the thicker contrast (Hashimoto, Hashimoto, &Soto, 1997; Hawkins, Caridi, LeVeen, Klioze, & Mladinich, 2000;Hawkins, Caridi, Wiechman, & Kerns, 1997; Krajina et al., 2004;Sandhu, Buckenham, & Belli, 1999). Regardless of the etiology,whether iatrogenic, traumatic, or gastrointestinal bleed, delin-eating the origin of bleeding and treating it precipitously can lead tosignificantly less morbidity and mortality. Hawkins reported thatthe use of CO2 DSA has approximately 2.5 times the sensitivity fordefining the acute hemorrhage when compared with iodinatedcontrast (Hawkins et al., 1997).

Peripheral Arterial Occlusive Disease

The incidence of peripheral arterial occlusive disease (PAD),chronic limb ischemia, and endovascular repair is increasing. CO2

balloon to allow assessment of stenosis and position. (C) Follow-up imaging does not

Figure 8. (A) Large patient with poor opacification of central structures using iodinated contrast. (B) Carbon dioxide that can be delivered via 25-gauge peripheral needle, does notmix with blood, and readily demonstrates central structures.

L. Garza et al. / Journal of Radiology Nursing 35 (2016) 261e274266

DSA is vital and extremely useful in this clinical scenario, but it iscurrently underused. Successful correlation of CO2 and iodinatedcontrast in PAD was demonstrated by Seeger's group in the mid1990s (Seeger et al., 1993). They demonstrated a 92% correlationwith an increase to 100%when a small amount of iodinated contrastwas administered (Figure 13).

A large number of patients with PAD/chronic limb ischemiacommonly have concomitant renal artery disease and other pre-disposition to renal insufficiency. Considering these facts and thesusceptibility of this group to CIN, it would seem intuitive to useCO2 as a contrast agent whenever possible.

Endovascular Abdominal Aneurysm Repair

One current use of CO2 DSA has been in the placement andevaluation of abdominal aortic endografts. This use has beenproliferating among numerous operators for a variety of reasons,including decrease of iodinated contrast load, more sensitive eval-uation of endoleaks, safety, and cost (Beese et al., 2000; Caro et al.,1991; Hawkins et al., 1994, 2009).

With thematurationof endovascular abdominal aneurysmrepair(EVAR), it has become clear that there is a propensity for developing

Figure 9. (A) Arteriovenous fistula (AVF) from previous common femoral artery approach.more selective injection. (C) After stent placement demonstrating resolution of the AVF, all

renal failure that is permanent and cumulative (Chao et al., 2007;Criado, Kabbani, & Cho, 2008; Gahlen et al., 2001; Walsh, Tang, &Boyle, 2008). The typical EVAR patient is usually older than70 years. This group of individuals has approximately a 30% inci-dence of abnormally low glomerular filtration rate that may not bereflected by serum creatinine levels alone. They also tend to havecomorbid conditions that predispose them to renal insult. The inci-dence of renal insufficiency in patients undergoing EVAR approxi-mates 7% to 25% with acute renal failure occurring in 2% to 16%,resulting in an associatedmortality of 30% to 50%. CO2 can be used asthe exclusive contrast agent or in addition to smaller volumes ofiodinated contrast, allowing accurate and complete endovascularrepair without inducing renal compromise. In addition, because ofits low viscosity, it has also been noted by several of the aforemen-tioned authors that CO2 is more sensitive for detecting endoleaks.

Interventional oncology

Interventional oncology is a recent catchall phrase for a specialtythat includes a multitude of minimally invasive procedures in thetreatment of a variety of tumors. One aspect of this specialty is

Nonselective injection demonstrates an early draining vein. (B) Magnified view with awith carbon dioxide.

Figure 10. (A) Posthepatic trauma, no evidence of acute hemorrhage with liquid contrast. (B) Gross carbon dioxide extravasation demonstrating site of bleeding for embolization.

L. Garza et al. / Journal of Radiology Nursing 35 (2016) 261e274 267

catheter directed and requires angiography/venography andembolization. The contrast load for these patients is often high.

Transarterial Chemoembolization, Drug-eluting Beads, and Yttrium-90

There are procedures that use angiographically direc-ted therapy predominantly for liver tumors. Transarterial che-moembolization, drug-eluting beads, and yttrium-90 therapy areused with liver tumor patients, who tend to be older and usuallyhave other comorbidities, such as renal insufficiency, diabetes, andhypertension. Patients with primary hepatic tumors also have

Figure 11. Renal transplant carbon dioxide (CO2) digital subtraction angiography demonstrtransplant kidney.

significant underlying liver disease and are at risk for hepatorenalcompromise. Approximately 75% of patients with cirrhosis willhave renal insufficiency at some time during the course of theirdisease.

Commonly, a large volume of contrast is required for interro-gation of the vasculature, treatment, and follow-up examination.These risks can be exacerbated by postembolization syndromefollowed by decreased oral intake or very rarely nontarget embo-lization and tumor lysis syndrome. All these factors, especially thehigh volume of contrast, place the patient at risk for CIN. Probably,the most significant scenario is that some of these patients may be

ating ostial stenosis treated with a stent. All performed with CO2 without insulting the

Figure 12. Renal angiogram after robotic surgery for renal cell carcinoma and potential postoperative bleeding. (A) Traditional contrast not only does make the diagnosis but alsojeopardizes a compromised kidney. (B and C) Arrows indicate source of bleeding using carbon dioxide digital subtraction angiography.

L. Garza et al. / Journal of Radiology Nursing 35 (2016) 261e274268

denied life-prolonging therapy because of their elevated creatinineand risk for CIN (Figure 14).

Another procedure, uterine fibroid embolization, is usuallyperformed in young and healthy individuals. Less invasive

Figure 13. Advanced arterial disease can be dem

transarterial therapy can be precluded if there is an elevation of theserum creatinine. In fact, renal insufficiency is listed as one of therelative contraindications (Stokes et al., 2010) and has beendescribed (Rastogi, Wu, Shlansky-Goldberg, & Stavropoulos, 2004).

onstrated using refined imaging techniques.

Figure 14. (A) Patient had previous transarterial chemoembolization and needs repeat delivery in a right hepatic artery branch. Iodinated contrast injected goes peripherally, andtumor is not seen. (B) Carbon dioxide digital subtraction angiography demonstrates central reflux and filling of the culprit vessel so inadvertent embolization is avoided and thecatheter can be redirected for appropriate treatment.

L. Garza et al. / Journal of Radiology Nursing 35 (2016) 261e274 269

As with transarterial chemoembolization procedures, theymay alsobe subjected to high volumes of iodinated contrast and developpostembolization syndrome (Figure 15).

Similarly, some patients with renal cell carcinoma requireembolization of the primary tumor or the hypervascularmetastasis. Obviously, many of these patients are susceptibleto CIN.

In each of the clinical scenarios delineated previously, CO2can usually be used as the predominant contrast agent with theaddition of a small or limited amount of dilute iodinatedcontrast.

Figure 15. Carbon dioxide digital subtraction angiography of pelvis demonstratinghypertrophied uterine arteries and tumor vascularity in the uterine fibroid.

Venous evaluation and treatment

As noted, CO2 was used safely in the venous system early in the1960s. Since the discovery and refinement of DSA, the myriad usesfor CO2 as a venous vascular contrast agent have proliferated. Thegaseous properties noted previously make it an ideal venouscontrast agent.

In the extremities, the low viscosity of CO2 permits delivery ofsufficient volumes through smaller 25-gauge needles. This is lessinvasive and less painful to the patient. Because CO2 does not mixwith blood, it is not diluted, and a peripheral hand injection willyield good opacification centrally. This is unlike contrast.

Figure 16. Right upper extremity venogram demonstrating central venous occlusionand patency of collaterals and right internal jugular vein. This can be used to planaccess in patients with limited veins.

Figure 17. Fistulagram. (A) Central veins. (B and C) Peripheral vein with stenosis. (D) Balloon waist from stenosis.

L. Garza et al. / Journal of Radiology Nursing 35 (2016) 261e274270

In the presence of venous occlusion, the properties of low vis-cosity and reflux will often demonstrate collateral cervical andthoracic veins bilaterally (Figure 16). This is critical in patients withrenal insufficiency and chronic venous occlusive disease. Because ofCIN and nephrogenic systemic fibrosis, respectively, venouscomputed tomography and magnetic resonance cannot be used fora road map before venous access procedures. This can be circum-vented nicely with CO2 extremity venography. It can help deter-mine venous patency and potential access sites. CO2 venographycan also be used in patients with iodinated contrast allergy.

CO2 DSA can be extremely useful in the examination of dialysisfistulas or interposition grafts (Figure 17). It is excellent atdemonstrating the veins and central circulation.

Another good use is in IVC evaluation, especially for filterplacement. Many times, these are emergent and patients may haveallergy, have elevated creatinine, or do not need the additionalvolume of iodinated contrast boluses. CO2 has been shown to be avery effective imaging agent for the evaluation of the IVC (Figure 18)(Boyd-Kranis, Sullivan, Eschelman, Bonn, & Gardiner, 1999;Holtzman et al., 2003; Pessanha de Rezende et al., 2011; Sing,Stackhouse, Jacobs, & Heniford, 2001).

CO2 renal venography can also be performed in other procedures(adrenal vein sampling and balloon-occluded retrograde trans-venous obliteration of varices) where evaluation of veins isessential.

Figure 18. Intravenous cavagram from jugular approach.

A common venous use of CO2 involves a variety of clinical sce-narios in the liver and spleen. It has been shown by Culp andHawkins that, as opposed to iodinated contrast, CO2 does not haveany negative effect on the splenocytes or hepatocytes when injec-ted directly into the respective parenchyma (Culp, Mladinich, &Hawkins, 1999). Hawkins showed that an injection of as little as12 cc/s had a negative effect as compared with CO2 at 200 cc/s.

One of the more unusual but effective uses of CO2 is in patientswith abnormalities of the splenoportal system (Figure 19) (Burke,Weeks, Mauro, & Jaques, 2004; Caridi et al., 2003; Cho & Cho,2003; Teng et al., 2006). The low viscosity of CO2 will cause opa-cification the small caliber splenoportal system, identifying theanatomy without significant bleeding (Burke et al., 2004; Caridiet al., 2003; Cho & Cho, 2003; Teng et al., 2006).

More commonly, it has become routine for many operators touse CO2 in transjugular intrahepatic portosystemic shunt (TIPS)procedures (Figure 20). The incidence of renal compromise afterTIPS approximates 2% to 3%. The entire procedure, including hepaticand portal venography as well as tract measurement, portal local-ization, and post-TIPS placement, can be performed with CO2,reducing the possibility of CIN (Hawkins& Caridi, 1998). Where CO2is most helpful in TIPS is in the localization of the portal vein. Thelow viscosity permits excellent visualization of the portal vein ingreater than 80% of cases and visualizes more of the system thaniodinated contrast without most of the consequences. However,because of a few complications of capsular rupture with catheter-

Figure 19. Splenoportography using a 25-gauge spinal needle and carbon dioxide.

Figure 20. Transjugular intrahepatic portosystemic shunt procedure using only carbon dioxide.

L. Garza et al. / Journal of Radiology Nursing 35 (2016) 261e274 271

wedged CO2 injections, balloon occlusion of the hepatic vein withvenous CO2 injections became the preferred method of choice(Figure 21) (Rees, Niblettt, Lee, Diamond, & Crippin, 1994; Semba,Saperstein, Nyman, & Dake, 1996; Taylor, Smith, Watkins, Kohne,& Suh, 1999; Theuerkauf et al., 2001). Complications are fewerwith the advantage of better portal visualization.

Visualizing the portal vein has also become more importantwith the rise in interventional oncology. Transhepatic portal veinaccess is necessary for certain procedures. More specifically, it iscritical in performing selective portal vein embolization to generatehepatic parenchymal hypertrophy of the remaining future liverremnant when extended hepatectomy is necessary.

Finally, there have been additional reports of limited numbers ofpatients using CO2 for atypical procedures. One of these is in painmanagement. CO2 has been used instead of iodinated contrastbefore neurolysis when the patient is highly allergic (Hirata, Higa,Shono, Hirota, & Shinokuma, 2003). Another newer area of userequiring additional investigation is for intraosseous venography inpercutaneous vertebroplasty (Tanigawa, Komemushi, Kariya,Kojima, & Sawada, 2005). CO2 can also be used in angioscopy(Silverman, Mladinich, Hawkins, Abela, & Seeger, 1989). It persistslonger than saline, giving a clearer view of the vessel. Similarly, it

Figure 21. Capsular rupture with wedged hepatic venography using carbon dioxide.

can be used percutaneously to separate organs before and duringablative procedures.

CO2 delivery

Since the advent of intravascular CO2 delivery, there has beena number of innovative methods of delivery developed(Alexander, 2011; Cherian et al., 2009; Cronin, Patel, Kessel,Robertson, & McPherson, 2005; Hawkins, Caridi, & Kerns,1995; Mendes, Wolosker, & Krutman, 2013). CO2 delivery be-gins with a source. As stated previously, a medical-grade CO2should be used, and because of potential impurities over time,the source should also be disposable. Next, there must be amechanism to deliver the CO2 from the source. One must beaware that many of the commercial canisters contain 3 millioncc of pressurized gas.

To avoid the inadvertent overload of pressurized gas and thecumbersome presence of a large canister, we introduced the use of aflaccid reservoir bag with a series of one-way valves (Figures 22).This method used a converted fluid management system byAngiodynamics called the Angioflush III system (Angiodynamics,Inc., Queensbury, NY). A similar system by Merit Medical is alsoused successfully (Figure 23). The theory behind these systems wasthat they would be a nonpressurized flaccid reservoir of CO2 thatwould avoid explosive delivery and excessive volumes.

The one-way glued valves are intended to eliminate stopcocks,prevent room air contamination, and eliminate the necessity toremove the delivery syringe. The inherent problem in each appa-ratus is that the systems required assembly. Regardless of thetraining and simplicity, incorrect assembly can result in airembolus. In addition, the bag must be filled and purged three timesto remove residual room air. This step is somewhat cumbersomeand time consuming, especially when the decision to use CO2 oc-curs spontaneously during the procedure. Other similar systems doexist.

For years, we have reported that the patient should never beconnected directly to the cylinder. This has been modified recentlybecause of the development of a K valve stopcock that precludesthe possibility of CO2 passing directly from the canister to thepatient (Figure 24). The next generation of delivery systems uses acompact regulator that uses a small 10,000 cc canister ofpharmaceutical-grade CO2 (Figure 25). The smaller system can beplaced in a sterile sleeve or left beneath the sterile drape. It isconnected to a series of two tubes with one-way valves, as well as

Figure 22. AngioFill Bag Collection System and Angioflush (Angiodynamics, Inc., Queensbury, NY). 1,500 cc nondistended plastic bag is attached to a three-port fitting with two one-way check valves. The three-port fitting is connected to a 100 cm tubing, then to a second fitting with two inline check valves and a three-way stopcock. This system is no longeravailable.

L. Garza et al. / Journal of Radiology Nursing 35 (2016) 261e274272

a K valve and a reservoir and delivery syringe. The K valve pre-vents direct communication with the patient. CO2 is introducedinto the reservoir syringe. From the reservoir, the gas should bepushed, not aspirated, to the delivery syringe to avoid the unlikelypossibility of air contamination through the K valve. Equilibriumwith the atmosphere can be achieved with a three-way stopcock

Figure 23. Merit Medical Custom Waste Bag and Contrast Delivery Set (Merit MedicalSystems, Inc., South Jordan, UT). This system is not approved by the Food and DrugAdministration. Similar to the AngioFill, it uses a flaccid plastic bag, one-way valves,and tubing attached to a three-way stopcock for delivery.

on the delivery catheter. The system does not require assemblyand is extremely user friendly. Setup for use takes approximately1 min.

Another recently developed type of delivery is the AngiodroidCO2 injector (Angiodroid, Angiodynamics, Inc., Queensbury, NY),which uses digital versus hand injection.

Finally, some operators filter the CO2 before it enters the vascu-lature. CO2 can be delivered and has been delivered without filtra-tion for arteriographyandvenography, as the gas is not injected as anarterial contrast agent above the diaphragm. However, the use of afilter (0.2 mm pore size) can effectively remove particulate contam-ination and bacteria (0.5 to 5.0 mm).

Conclusion

CO2 is not the quintessential imaging agent, but it offers uniqueproperties when used alone or in combination with iodinatedcontrast that can expand diagnostic and therapeutic options in avariety of clinical scenarios. Used appropriately, it is safe and cannot only prevent CIN but also offer life-extending procedures topatients who would otherwise have been precluded because oftheir underlying renal status. It can also be used less invasively orwhen traditional contrast fails to make diagnoses and avoids moresignificant intervention. Current technology permits simple andsafe administration with images comparable to liquid contrast. It isan inexpensive and versatile tool that should be added to everyinterventionist's toolbox.

Figure 24. Proprietary K valve. Product can only be transferred along the lines of the valve stem. Product cannot go from the original source to patient directly. It must traverse boththe reservoir and delivery syringe.

L. Garza et al. / Journal of Radiology Nursing 35 (2016) 261e274 273

References

Alexander, J.Q. (2011). CO2 angiography in lower extremity arterial disease. Endo-vascular Today, pp. 27-34. Retrieved from http://evtoday.com/2011/09/cosub2sub-angiography-in-lower-extremity-arterial-disease.

Bartholomew, B.A., Harjai, K.J., & Dukkipati, S. (2004). Impact of nephropathy afterpercutaneous coronary intervention and a method for risk stratification.American Journal of Cardiology, 93(12), 1515-1519.

Beese, R.C., Bees, N.R., & Belli, A.M. (2000). Renal angiography using carbon dioxide.British Journal of Radiology, 73, 3-6.

Bendib, M., Tourni, M., & Boudjellab, A. (1977). CO2 angiography and enlarged CO2 angi-ography in cardiology (author's transl) Annales de Radiologie (Paris), 20, 673-686.

Boyd-Kranis, R., Sullivan, K.L., Eschelman, D.J., Bonn, J., & Gardiner, G.A. (1999).Accuracy and safety of carbon dioxide inferior vena cavography. Journal ofVascular and Interventional Radiology, 10(9), 1183-1189.

Burke, C.T., Weeks, S.M., Mauro, M.A., & Jaques, P.F. (2004). CO2 splenoportographyfor evaluating the splenic and portal veins before or after liver transplantation.Journal of Vascular and Interventional Radiology, 15(10), 1161-1165.

Carelli, H.H., & Sorddelli, E. (1921). A new procedure for examining the kidney.Revista de la Asociaci�on M�edica Argentina, 34, 18-24.

Caridi, J.G., Hawkins, I.F., Jr., Cho, K., et al. (2003). CO2 splenoportography: Pre-liminary results. AJR. American Journal of Roentgenology, 180(5), 1375-1378.

Caridi, J.G., Stavropolous, W., & Hawkins, I.F., Jr. (1999). CO2 digital subtractionangiography for renal artery angioplasty in high-risk patients. AJR. AmericanJournal of Roentgenology, 173(6), 1551-1556.

Caro, J.J., Trindade, E., & McGregor, M. (1991). The risks of death and of severenonfatal reactions with high- vs low-osmolality contrast media: A meta-anal-ysis. AJR. American Journal of Roentgenology, 156(4), 825-832.

Chao, A., Major, K., Kumar, S.R., et al. (2007). Carbon dioxide digital subtractionangiography-assisted endovascular aortic aneurysm repair in the azotemic pa-tient. Journal of Vascular Surgery, 45(3), 451-460.

Cherian, M.P., Mehta, P., Gupta, P., Kalyanpur, T.M., Jayesh, S.R., & Rupa, R. (2009).Technical note: A simple and effective CO2 delivery system for angiographyusing a blood bag. Indian Journal of Radiology and Imaging, 19(3), 203-205.

Cho, K.J. (2007). CO2 as a venous contrast agent: Safety and tolerance. In K.J. Cho, &I.F. Hawkins (Eds.), Carbon dioxide angiography: Principles, techniques and prac-tices (pp. 37-44). New York: Informa Healthcare.

Cho, K.J., & Cho, D.R. (2003). CO2 digital subtraction splenoportography with the“skinny” needle experimental study in a swine model. Cardiovascular andInterventional Radiology, 26(3), 273-276.

Criado, E., Kabbani, L., & Cho, K. (2008). Catheter-less angiography for endovascularaortic aneurysm repair: A new application of carbon dioxide as a contrast agent.Journal of Vascular Surgery, 48(3), 527-534.

Figure 25. Commander and AngiAssist delivery system. This uses a dedicated phar-maceutical grade carbon dioxide 10,000 cc cylinder containing a compact user-friendlycompressor. The AngiAssist is one piece preassembled containing one-way valves and apriority K valve, which prevents direct delivery from the source to the patient.

Cronin, P., Patel, J.V., Kessel, D.O., Robertson, I., & McPherson, S.J. (2005). Carbondioxide angiography: A simple and safe system of delivery. Clinical Radiology,60(1), 123-125.

Culp, W.C., Mladinich, C.R., & Hawkins, I.F., Jr. (1999). Comparison of hepatic damagefrom direct injections of iodinated contrast agents and carbon dioxide. Journal ofVascular and Interventional Radiology, 10(9), 1265-1270.

Ehrman, K.O., Taber, T.E., Gaylord, G.M., Brown, P.B., & Hage, J.P. (1994). Comparisonof diagnostic accuracy with carbon dioxide versus iodinated contrast material inthe imaging of hemodialysis access fistulas. Journal of Vascular and Interven-tional Radiology, 5(5), 771-775.

Gahlen, J., Hansmann, J., Schumacher, H., Seelos, R., Richter, G.M., & Allenberg, J.R.(2001). Carbon dioxide angiography for endovascular grafting in high risk pa-tients with infrarenal abdominal aortic aneurysms. Journal of Vascular Surgery,33(3), 646-649.

Hashimoto, S., Hashimoto, K., & Soto, M. (1997). CO2 as an intra-arterial digitalsubtraction angiography (IADSA) agent in the management of trauma. Seminarsin Interventional Radiology, 14, 163-173.

Hawkins, I.F. (1982). Carbon dioxide digital subtraction arteriography. AJR. AmericanJournal of Roentgenology, 139(1), 19-24.

Hawkins, I.F., & Caridi, J.G. (1998). Carbon dioxide (CO2) digital subtraction angi-ography: 26 year experience at the University of Florida. European Radiology,8(3), 391-402.

Hawkins, I.F., Jr., Caridi, J.G., & Kerns, S.R. (1995). Plastic bag delivery system for hand in-jection of carbon dioxide. AJR. American Journal of Roentgenology, 165(6), 1487-1489.

Hawkins, I.F., Jr., Caridi, J.G., LeVeen, R.F., Klioze, S.D., & Mladinich, C.R.J. (2000). Useof carbon dioxide for the detection of gastrointestinal bleeding. Techniques inVascular and Interventional Radiology, 3(3), 130-138.

Hawkins, I.F., Caridi, J.G., Wiechman, B.N., & Kerns, S.R. (1997). Carbon dioxide (CO2)digital subtraction angiography in trauma patients. Seminars in InterventionalRadiology, 14, 175-180.

Hawkins, I.F., Cho, K.J., & Caridi, J.G. (2009). Carbon dioxide in angiography to reducethe risk of contrast-induced nephropathy. Radiologic Clinics of North America,47(5), 813-825.

Hawkins, I.F., Mladinich, C.J., Drane, W.E., et al. (1992). Effects of CO2 angiography onrenal function. Journal of Vascular and Interventional Radiology, 3, 6.

Hawkins, I.F., Jr., Wilcox, C.S., Kerns, S.R., & Sabatelli, F.W. (1994). CO2 digital angi-ography: A safer contrast agent for renal vascular imaging? American Journal ofKidney Diseases, 24(4), 685-694.

Hipona, F.A., Ferris, E.J., & Pick, R. (1969). Capnocavography: A new technique forexamination of the inferior vena cava. Radiology, 92, 606-609.

Hirata, K., Higa, K., Shono, S., Hirota, K., & Shinokuma, T. (2003). Splanchnic neu-rolysis using carbon dioxide as the contrast agent. Regional Anesthesia and PainMedicine, 28(1), 68-69.

Holtzman, R., Lottenberg, L., Bass, T., Saridakis, A., Bennett, V.J., & Carrillo, E.H.(2003). Comparison of carbon dioxide and iodinated contrast for cavographyprior to inferior vena cava filter placement. American Journal of Surgery, 185(4),364-368.

Kooiman, J., Pasha, S.M., Zondag, W., et al. (2012). Meta-analysis: Serum creatininechanges following contrast-enhanced CT imaging. European Journal of Radiology,81(10), 2554-2561.

Krajina, A., Lojík, M., Rejchrt, S., et al. (2004). Carbon dioxide arteriography in thedetection of acute massive gastrointestinal bleeding. Folia GastroenterologyHepatology, 2(1), 8-12.

Kriss, V.M., Cottrill, C.M., & Gurley, J.C. (1997). Carbon dioxide (CO2) angiography inchildren. Pediatric Radiology, 27(10), 807-810.

Lieberman, P.L., & Seigle, R.L. (1999). Reactions to radiocontrast material: Anaphy-lactoid events in radiology. Clinical Reviews in Allergy & Immunology, 17(4), 469-496.

McLennan, G., Moresco, K.P., Patel, N.H., et al. (2001). Accuracy of CO2 angiographyin vessel diameter assessment: A comparative study of CO2 versus iodinatedcontrast material in a porcine model. Journal of Vascular and InterventionalRadiology, 12(8), 985-989.

Mendes, C.A., Wolosker, N., & Krutman, M. (2013). A simple home-made carbondioxide delivery system for endovascular procedures in the iliofemoral arteries.Circulation Journal, 77(3), 831.

Moresco, K.P., Patel, N., Johnson, M.S., Trobridge, D., Bergan, K.A., & Lalka, S.G.(2000). Accuracy of CO2 angiography in vessel diameter assessment: A

L. Garza et al. / Journal of Radiology Nursing 35 (2016) 261e274274

comparative study of CO2 versus iodinate contrast material in an aortoiliac flowmodel. Journal of Vascular and Interventional Radiology, 11(4), 437-444.

Moresco, K.P., Patel, N.H., Namyslowski, Y., Shah, H., Johnson, M.S., & Trerotola, S.O.(1998). Carbon dioxide angiography to the transplanted kidney: Technicalconsiderations and imaging findings. AJR. American Journal of Roentgenology,171(5), 1271-1276.

Murakami, R., Hayashi, H., Sugizaki, K., et al. (2012). Contrast- induced nephropathyin patients with renal insufficiency undergoing contrast-enhanced MDCT. Eu-ropean Radiology, 22(10), 2147-2152.

Nash, K., Hafeez, A., & Hou, S. (2002). Hospital-acquired renal insufficiency. Amer-ican Journal of Kidney Diseases, 39(5), 930-936.

Paul, R.E., Durant, T.M., Oppenheimer, M.J., & Stauffer, H.M. (1957). Intravenouscarbon dioxide intracardiac gas contrast in the roentgen diagnosis of pericardialeffusion and thickening. American Journal of Roentgenology, Radium Therapy, andNuclear Medicine, 78(2), 224-225.

Pessanha de Rezende, M., Massiere, B., Von Ristow, A., et al. (2011). Carbon dioxideuse as contrast for vena cava filter implantation: Case series. Journal of VascularBrasileira, 11(1), 18-21.

Phillips, J.H., Burch, G.E., & Hellinger, R. (1961). The use of intracardiac carbon di-oxide in the diagnosis of pericardial disease. American Heart Journal, 61, 748-755.

Rastogi, S., Wu, Y.H., Shlansky-Goldberg, R.D., & Stavropoulos, S.W. (2004). Acuterenal failure after uterine artery embolization. Cardiovascular and InterventionalRadiology, 27(5), 549-550.

Rees, C.R., Niblettt, R.L., Lee, S.P., Diamond, N.G., & Crippin, J.S. (1994). Use of carbondioxide as a contrast medium for transjugular intrahepatic portosystemic shuntprocedures. Journal of Vascular and Interventional Radiology, 5(2), 383-386.

Rosenstein, P. (1921). Pneumoradiology of kidney positioneA new technique for theradiological representation of the kidneys and neighboring organs (suprarenalgland, spleen, liver). Journal of Urology, 15, 447.

Rotenberg, E. (1914). Rontgenophotographie der leber, der milz, und des zwerch-fells. Deutsch Med Wschr, 40, 1205.

Roussos, C., & Koutsoukou, A. (2003). Respiratory failure. European RespiratoryJournal, 22(Suppl 47), 3s-14s.

Sandhu, C., Buckenham, T.M., & Belli, A.M. (1999). Using CO2-enhanced arteriog-raphy to investigate acute gastrointestinal hemorrhage. AJR. American Journal ofRoentgenology, 173(5), 1399-1401.

Scatliff, J.H., Kummer, A.J., & Janzen, A.H. (1959). The diagnosis of pericardial effu-sion with intracardiac carbon dioxide. Radiology, 73, 871-883.

Seeger, J.M., Self, S., Harward, T.R., Flynn, T.C., & Hawkins, I.F., Jr. (1993). Carbondioxide gas as an arterial contrast agent. Annals of Surgery, 217(6), 697-698.

Semba, C.P., Saperstein, L., Nyman, U., & Dake, M.D. (1996). Hepatic laceration fromwedged venography performed before transjugular intrahepatic portosystemicshunt placement. Journal of Vascular and Interventional Radiology, 7(1), 143-146.

Silverman, S.H., Mladinich, C.J., Hawkins, I.F., Jr., Abela, G.S., & Seeger, J.M. (1989).The use of carbon dioxide gas to displace flowing blood during angioscopy.Journal of Vascular Surgery, 10(3), 313-317.

Sing, R.F., Stackhouse, D.J., Jacobs, D.G., & Heniford, B.T. (2001). Safety and accuracy ofbedside carbon dioxide cavography for the insertion of inferior vena cava filters inthe intensive care unit. Journal of American College of Surgery, 192(2), 168-171.

Spinosa, D.J., Angle, J.F., Hagspiel, K.D., Kern, J.A., Hartwell, G.D., & Matsumoto, A.H.(2000). Lower extremity arteriography with use of iodinated contrast materialor gadodiamide to supplement CO2 angiography in patients with renal insuf-ficiency. Journal of Vascular and Interventional Radiology, 11(1), 35-43.

Spinosa, D.J., Matsumoto, A.H., Angle, J.F., et al. (1998). Gadolinium-based contrastand carbon dioxide angiography to evaluate renal transplants for vascularcauses of renal insufficiency and accelerated hypertension. Journal of Vascularand Interventional Radiology, 9(6), 909-916.

Stokes, L.S., Wallace, M.J., Godwin, R.B., et al. (2010). Quality improvement guide-lines for uterine artery embolization for symptomatic leiomyomas. Journal ofVascular and Interventional Radiology, 21(8), 1153-1163.

Tanigawa, N., Komemushi, A., Kariya, S., Kojima, H., & Sawada, S. (2005). Intra-osseous venography with carbon dioxide contrast agent in percutaneous ver-tebroplasty. AJR. American Journal of Roentgenology, 184(2), 567-570.

Taylor, F.C., Smith, D.C., Watkins, G.E., Kohne, R.E., & Suh, R.D. (1999). Balloon occlusionversus wedged hepatic venography using carbon dioxide for portal vein opacifica-tion during TIPS. Cardiovascular and Interventional Radiology, 22(2), 150-151.

Teng, G.J., Deng, G., Liu, Z.S., et al. (2006). Ultrafine needle CO2 splenoportography: Acomparative investigation with transarterial portography and MR portography.European Journal of Radiology, 59(3), 393-400.

Theuerkauf, I., Strunk, H., Brensing, K.A., et al. (2001). Infarction and laceration ofliver parenchyma caused by wedged CO2 venography before TIPS insertion.Cardiovascular and Interventional Radiology, 24(1), 64-67.

Thomson, K.R., Tello, R., Sullivan, R., Dixon, R.G., Becker, G., & Mitchell, P.J. (1996).Carbon dioxide angiography. Asian Oceanian Journal of Radiology, 1, 20-23.

Trcka, J., Schmidt, C., Seitz, C.S., Br€ocker, E.-B., Gross, G.E., & Trautmann, A. (2008).Anaphylaxis to iodinated contrast material: Nonallergic hypersensitivity or IgE-mediated allergy? AJR. American Journal of Roentgenology, 190(3), 666-670.

Walsh, S.R., Tang, T.Y., & Boyle, J.R. (2008). Renal consequences of endovascularabdominal aortic aneurysm repair. Journal of Endovascular Therapy, 15(1), 73-82.

Suggested Reading

Castaneda-Zuniga, W.R., Jauregui, H., Rysavy, J.A., Formanek, A., & Amplatz, K.(1978). Complications of wedge hepatic venography. Radiology, 126(1), 53-56.

Choudhri, A.H., Cleland, J.G., Rowlands, P.C., Tran, T.L., McCarty, M., & al-Kutoubi, M.A. (1990). Unsuspected renal artery stenosis in peripheral vasculardisease. British Medical Journal, 301(6762), 1197-1198.

Coffey, R., Quisling, R.G., Mickle, J.P., Hawkins, I.F., Jr., & Ballinger, W.B. (1984). Thecerebrovascular effects of intraarterial CO2 in quantities required for diagnosticimaging. Radiology, 151(2), 405-410.

Dimakakos, P., Stefanopoulos, T., Doufas, A., et al. (1998). The cerebral effects ofcarbon dioxide digital subtraction angiography in the aortic arch and itsbranches in rabbits. AJNR. American Journal of Neuroradiology, 19(2), 261-266.

Hawkins, I.F. (2007). Aortogram and runoff. In K.J. Cho, & I.F. Hawkins (Eds.), Carbondioxide angiography: Principles, techniques, and practices (pp. 53-68). New York:Informa Healthcare.

Heye, S., Maleux, G., & Marchal, G.J. (2006). Upper-extremity venography: CO2versus iodinated contrast media. Radiology, 241(1), 291-297.

Huo, T.I., Wu, J.C., Lee, P.C., Chang, F.Y., & Lee, S.D. (2004). Incidence and risk factorsfor acute renal failure in patients with hepatocellular carcinoma undergoingtransarterial chemoembolization: A prospective study. Liver International, 24(3),210-215.

Jang, B.K., Lee, S.H., Chung, W.J., et al. (2008). Incidence and risk factors of acuterenal failure after transcatheter arterial chemoembolization for hepatocellularcarcinoma. Korean Journal of Hepatology, 14(2), 168-177.

Kappel, J., & Calissi, P. (2002). Nephrology: Safe drug prescribing for patients withrenal insufficiency. CMAJ, 166(4), 473-477.

Liss, P., Eklof, H., Hellberg, O., et al. (2005). Renal effects of CO2 and iodinatedcontrast media in patients undergoing renovascular intervention: A prospec-tive, randomized study. Journal of Vascular and Interventional Radiology, 16(1),57-65.

Marenzi, G., Lauri, G., Assanelli, E., et al. (2004). Contrast-induced nephropathy inpatients undergoing primary angioplasty for acute myocardial infarction. Jour-nal of the American College of Cardiology, 44(9), 1780-1785.

McCullough, P.A. (2008). Contrast-induced acute kidney injury. Journal of theAmerican College of Cardiology, 51(15), 1419-1428.

Missouris, C.G., Buckenham, T., Cappuccio, F.P., & MacGregor, G.A. (1994). Renalartery stenosis: A common and important problem in patients with peripheralvascular disease. American Journal of Medicine, 96(1), 10-14.

Moos, J.M., Ham, S.W., Han, S.M., et al. (2011). Safety of carbon dioxide digitalsubtraction angiography. Archives of Surgery, 146(12), 1428-1432.

Nikolsky, E., Mehran, R., Turcot, D.B., et al. (2004). Impact of chronic kidney diseaseon prognosis of patients with diabetes mellitus treated with percutaneouscoronary intervention. American Journal of Cardiology, 94(3), 300-305.

Penzkofer, T., Slebocki, K., Grommes, J., et al. (2013). Carbon dioxide-contrastedcomputed tomography angiography: High pitch protocols and adapted detec-tion parameters improve imaging quality. Rofo, 185(2), 128-135.

Rashid, S.T., Salman, M., Agarwal, S., & Hamilton, G. (2006). Occult renal impairmentis common in patients with peripheral vascular disease and normal serumcreatinine. European Journal of Vascular and Endovascular Surgery, 32(3), 294-299.

Rihal, C.S., Textor, S.C., Grill, D.E., et al. (2002). Incidence and prognostic importanceof acute renal failure after percutaneous coronary intervention. Circulation,105(19), 2259-2264.

Rolland, Y., Duvauferrier, R., Lucas, A., et al. (1998). Lower limb angiography: Aprospective study comparing carbon dioxide with iodinated contrast material in30 patients. AJR. American Journal of Roentgenology, 171(2), 333-337.

Shifrin, E., Plich, M., Verstandig, A.G., & Gomori, M. (1990). Cerebral angiographywith gaseous carbon dioxide CO2. Journal of Cardiovascular Surgery, 31(5), 603-606.

Sonoda, A., Nitta, N., Ushio, N., et al. 320-row multidetector CT angiography forhepatocellular carcinoma using CO2 gas instead of iodinated contrast agents:Experiment and preliminary clinical study. Poster presented at: EuropeanCongress of Radiology, March 3-7, 2011, Vienna, Austria. Poster C-1123.

Stram, E., & Molgaard, C.P. (1999). Use of a compression paddle to displace bowelgas for carbon dioxide digital subtraction angiography. Journal of Vascular andInterventional Radiology, 10(4), 405-408.

Sullivan, K., Bonn, J., Shapiro, M., & Gardiner, G.A. (1995). Venography with carbondioxide as a contrast agent. Cardiovascular and Interventional Radiology, 18(3),141-145.

Wilson, A.J., & Boxer, M.M. (2002). Neurotoxicity of angiographic carbon dioxide inthe cerebral vasculature. Investigative Radiology, 37(10), 542-551.

Wong, A.A., Charalel, R.A., Louie, J.D., & Sze, D.Y. (2013). Carbon dioxide contrastenhancement for c-arm CT utility for treatment planning during hepaticembolization procedures. Journal of Vascular and Interventional Radiology, 24(7),975-980.

Yamazaki, H., Oi, H., Matsushita, M., et al. (2001). Renal cortical retention on delayedCT and nephropathy following transcatheter arterial chemoembolisation. BritishJournal of Radiology, 74(884), 695-700.