Introduction

Carbon dioxide (CO2) has been used as an intra-arterial contrast agent during angiography for more than four decades [1]. Renal insufficiency is an important indication for the use of CO2 angiography, given the absence of nephrotoxicity with this agent [2]. Takayasu arteritis (TA), a chronic idiopathic granulomatous large vessel vasculitis affecting the aorta and its main branches [3], can produce renal insufficiency [4], mainly by causing renal artery stenosis [5]. At our center, prevalence of renal dysfunction in 118 patients with TA presenting during the period 1963–1977 was 13.6% [6]. A recent study by Li et al. [4] showed an overall prevalence of renal dysfunction of 11.4% in 411 patients with TA. There are no reports of the use of CO2 angiography-guided interventions in TA apart from a single case report from our center in 2003 [7]; the present study describes our 16-year single-center experience with CO2 angiography in guiding renal-related percutaneous interventions in patients with TA.

Methods

Patients

Data on percutaneous interventions performed on patients with TA at our center (a large tertiary care referral hospital in South India) that were collected prospectively over 23 years and archived digitally were scrutinized. Cases where CO2 angiography was utilized to guide percutaneous interventions were selected, and the relevant case records and angiographic images were analyzed. Follow-up was obtained in all patients at 6- to 12-month intervals if stable, but more frequently when indicated. Clinical status, imaging information and additional procedures done at follow-up were studied. All procedures were performed after obtaining written informed consent. The institutional review board approved this retrospective study. For this type of study formal consent is not required.

Carbon dioxide Angiography System

The CO2 angiography system used (Fig. 1A) was developed in-house and has been operational for the last 17 years. It has no moving parts, and CO2 injection is pressure-driven. Pressurized medical grade CO2 obtained from a storage cylinder is passed through a Millipore filter into a disposable 50 ml plastic collection syringe that is held firmly between fixed restraints on a metal platform. The piston of the collection syringe is variably restrained by an assembly of swivelling blocks; the volume of the syringe is determined by the number of blocks in use in the piston stopper assembly. A two-way stopcock attached to the collection syringe nozzle ensures that at no stage can the flow of CO2 bypass the syringe and go directly from cylinder to patient. Gas pressure in the collection syringe is kept at 4 kg/cm2 (nearly 4 atm) by appropriately adjusting the knob in the dual pressure gauge assembly which separately indicates cylinder (inflow) and tubing (outflow) pressures. At this setting, the volume of CO2 delivered into the vessel is approximately three times the selected volume in the collection syringe; this can be verified by trial injection of CO2 into a collapsed plastic bag attached to the three-way stopcock along the tubing. Prior to angiography, the tubing and angiographic catheter are flushed with CO2 to remove air and avoid explosive delivery of CO2. For CO2 injection, the two-way stopcock is turned 90° to allow CO2 to flow from the collection syringe to the patient. Standard protocols were used for patient preparation and digital subtraction angiography. Typically, the collection syringe volume was set at 20 ml for flush aortograms (to deliver 60 ml of CO2) and 10 ml for selective angiograms (to deliver 30 ml of CO2). The quality of images obtained using this CO2 angiography system compared favorably with that obtained with 50% iodixanol, a diluted alternative contrast agent, that is sometimes used in renal insufficiency (Fig. 1B, C).

Fig. 1
figure 1

A Diagrammatic representation of the apparatus used for carbon dioxide angiography. Every alternate block of the piston stopper assembly has been left un-shaded to reveal the perpendicular rod around which the blocks can swivel 180°; all the blocks are identical. B, C. Consecutive abdominal aortograms performed in antero-posterior projection in a patient with renal failure using carbon dioxide and 50% iodixanol, respectively, with identical settings of digital subtraction angiography

Full size image

Results

Of 692 patients with TA who underwent percutaneous interventions to treat 1834 diseased arteries (including 425 renal arteries) over a 23-year period, seven patients (1.01%) underwent CO2 angiography-guided percutaneous interventions because of renal insufficiency (Table 1). All seven patients (1 male/6 female, mean age 38 years) met both the American College of Rheumatology [8] and the modified Ishikawa (clinical) criteria [9] for the diagnosis of TA; all were hypertensive on multiple medications. All patients were symptomatic at presentation, often with recent worsening.

Table 1 Case and procedure details
Full size table

Renal Interventions

Five patients with significant renal artery stenosis underwent CO2 angiography-guided renal angioplasty or stenting using percutaneous femoral arterial access (Table 1, cases 1–5; Fig. 2). In cases 1, 2 and 4, the renal artery was engaged with a 7F renal-guiding catheter and the stenotic lesion was crossed with a 0.018-inch nitinol guidewire. In cases 3 and 5, the renal arteries were engaged with a 6F Judkins Right diagnostic catheter and a 0.035-inch angled hydrophilic guidewire was used to cross the renal artery lesion; this was replaced with a 0.035-inch stiff guidewire with 1-cm soft tip over which a long 7F femoral sheath was advanced up to the renal artery ostium. Balloon-expandable bare metal stents were deployed in all the lesions except in case 5 where a stent had been deployed earlier. Balloon-dilatation pressures of 8 to 14 atm were used. A satisfactory result was obtained in all vessels without complications.

Fig. 2
figure 2

Carbon dioxide-guided renal artery interventions in patients with Takayasu arteritis and renal insufficiency. All images are carbon dioxide angiograms unless stated otherwise. A to C. Case 1. Baseline angiogram (A) shows right renal artery stenosis (arrow). After stent positioning (B) and deployment, the final angiogram (C) showed a good outcome. D to F. Case 2. Baseline angiogram (D) shows ostial left renal artery stenosis (arrow) and long infra-renal aorta narrowing. Renal function normalized after left renal stenting (E). Conventional angiogram (F) obtained 11 years later (5 years after interval infra-renal aortic stenting) shows good long-term outcome. G to I. Case 3. Baseline angiogram (G) shows right renal artery occlusion (black arrow) and left renal artery stenosis (white arrow). Bilateral renal artery stenting was performed (H) leading to normalization of renal function. Conventional angiogram done 4 months later (I) shows good short-term outcome

Full size image

Aortic Interventions

Two patients underwent CO2 angiography-guided aortic and ancillary interventions with successful outcomes and without complications (Table 1, cases 6, 7; Fig. 3).

Fig. 3
figure 3

Carbon dioxide-guided aortic and ancillary interventions in patients with Takayasu arteritis and renal insufficiency. A to G. Case 6. Magnetic resonance angiograms in transverse (A) and coronal (B) planes and carbon dioxide angiogram early (C) and late (D) frames in antero-posterior (AP) projection show a single (left) renal artery, focal abdominal aortic dissection, narrowing of the infra-renal aorta and both common iliac arteries and a large false lumen aneurysm on the left lateral aspect of infra-renal aorta (black asterisk). Carbon dioxide AP (E) and lateral (F) angiograms after infra-renal aortic and bilateral iliac artery stenting show relief of stenosis; the superior mesenteric and right renal arteries are not visualized. Carbon dioxide AP angiogram (G) after deployment of a tapered endograft in the abdominal aorta and chimney grafts in the left renal and superior mesenteric arteries shows patency of these arteries; the false lumen aneurysm was no longer visualized in the late frames. H to K. Case 7. Carbon dioxide AP angiogram (H) obtained 7 months after stenting of the lower thoracic-upper abdominal and infra-renal aorta (renal function had deteriorated since then) shows residual stenosis and non-visualization of the renal arteries. The stents were further expanded by balloon dilatation (I), and the right kidney was auto-transplanted (J) resulting is normalization of renal function. Conventional AP aortogram (K) obtained 11 years later shows widely patent aortic stents

Full size image

In case 6, extensive endovascular repair was performed for spontaneous focal abdominal aortic dissection with false lumen aneurysm formation and aorto-iliac true lumen narrowing with 103 mmHg systolic pressure gradient. The celiac and left renal arteries were the only patent visceral aortic branches. In the first sitting, aorto-iliac stenting was performed with a covered stent in the infra-renal aorta and bare metal stents in both common iliac arteries. In a second sitting 7 days later, a tapered aorto-uni-iliac endograft was deployed in the abdominal aorta starting above the upper limit of the aortic dissection (at the celiac artery ostium level) and overlapping the earlier deployed stent below. Flow into the left renal artery was preserved by constructing a chimney graft. Flow into the celiac artery was preserved using a bare metal stent deployed in chimney fashion. A satisfactory result was obtained with exclusion of the false lumen aneurysm, abolition of the aorto-iliac pressure gradient and preserved flow into the left renal and celiac arteries.

Case 7 had presented 7 months earlier with occlusions of both renal arteries and of the aorta at two levels; the aortic occlusions were recanalized and stented with balloon-expandable stents, but the lesions were resistant to dilatation, and only partial opening was achieved with a 9-mm-diameter non-compliant balloon. Renal function had subsequently deteriorated, and in the present sitting CO2 angiography-guided balloon dilatation of both stents was performed using a 10-mm-diameter non-compliant balloon which resulted in further stent expansion and reduction in the aortic pressure gradient. This provided a sufficient pressure head for the right kidney to be auto-transplanted 2 weeks later to the right iliac fossa with the right renal artery being anastomosed end-to-end to the right internal iliac artery.

Outcome and Follow-Up

CO2 angiography provided sufficiently clear visualization of the vascular anatomy and enabled achievement of successful outcomes in all the patients. Some patients felt mild transient abdominal discomfort immediately after CO2 injection, but none had nausea, vomiting, hypotension, narcosis or air contamination-related problems. There were no late complications related to CO2 angiography. All patients were followed-up after the CO2 angiography-guided intervention (median duration 5 years, range 2 months to 16 years; Table 1). None of the patients had deterioration in renal function or required dialysis; rather, renal function improved in 6 patients and stabilized in one. All patients experienced resolution of their symptoms. Blood pressure control improved, as also left ventricular systolic function that was initially depressed in some patients. Restenotic renal artery lesions seen at follow-up in cases 3 and 4 were treated by balloon dilatation; in case 3 this produced distal edge dissection in the right renal artery and required deployment of a stent. In case 7, progressive dilatation of the aortic stents was performed during subsequent follow-up visits.

Comparison with Iodinated Contrast

During the same period of time, 46 TA patients with elevated serum creatinine levels (> 1.4 mg/dl) underwent interventions using iodinated contrast after intravenous hydration; of these, none with serum creatinine levels below 3.0 mg/dl developed sustained worsening of renal function or required dialysis, but one of two patients with serum creatinine ≥ 3.0 mg/dl required long-term dialysis. On the other hand, none of the TA patients who underwent CO2 angiography-guided interventions, including four with serum creatinine ≥ 3.0 mg/dl, had worsening of renal function or required dialysis.

Discussion

Contrast-induced nephropathy is one of the most common causes of hospital-acquired renal insufficiency and is associated with a mortality of 14%; pre-existing renal insufficiency poses the greatest risk for developing this condition [10]. Absence of nephrotoxicity is perhaps the biggest advantage CO2 provides as a contrast agent, though it has several other benefits including being non-allergic and inexpensive, having low viscosity, allowing use of unlimited total volumes and not being diluted by blood [11]. With modern technology such as high-resolution digital subtraction angiography, stacking software, tilting tables and reliable delivery systems, the quality of images obtained with CO2 angiography has improved considerably [2]. In many cases, such as the ones presented in this report, the entire percutaneous intervention can be performed without use of iodinated contrast. In case 6, a single angiogram using 4 ml of iodinated contrast diluted with saline was used to confirm adequate collateral flow from the celiac artery to the superior and inferior mesenteric artery territories; this was done before deploying aortic endografts across the ostia of these already occluded vessels because recanalization of these vessels would no longer be possible. An important prerequisite for successful CO2 angiography is gentle, controlled, non-explosive delivery of the gas using a closed and ideally non-pressurized system; this can be achieved by use a series of one-way valves with a flaccid reservoir or blood bag [11, 12]. The CO2 angiography system used in this study was a pressurized system, but with adequate attention to outflow pressure and syringe volume and by purging the system with CO2 gas before use, no significant problems have been encountered; the system has worked well in more than 100 patients over 17 years, is simple, inexpensive and easy to use and has guided aorto-iliac, lower extremity, venous and complex endovascular procedures.

Renal insufficiency is not uncommon in TA despite these patients being young and having few co-morbidities [4, 6]. Renal insufficiency in TA is usually attributed to renal ischemia caused by vascular obstruction and renal parenchymal damage induced by systemic hypertension; glomerular disease is considered exceptional [13]. Recent clinical studies [5, 14] support this contention. Hong et al. [5] found that 9.7% of TA patients with renal artery involvement developed chronic renal insufficiency over a 90-month follow-up period. Similarly, Obiagwu et al. [14] showed that renal artery revascularization procedures were effective in salvaging renal function in children with TA-induced renal artery stenosis. However, an autopsy study [15] on 25 TA patients showed that whereas non-specific, ischemic and/or hypertensive glomerular changes were present in 44% of kidney specimens, 56% showed specific glomerular pathologies, most commonly diffuse mesangial proliferative glomerulonephritis; extent of large arterial inflammatory infiltrates assessed by morphometric analysis was most in the latter condition suggesting a relationship between the two phenomena.

This report highlights some characteristic features of percutaneous interventions in TA. Firstly, the high rate of restenosis seen after bare metal renal artery stenting in TA. Secondly, repeat intervention, mostly balloon angioplasty alone, is usually successful in dealing with the problem of restenosis. Thirdly, stenotic lesions in TA may be very resistant to dilatation, but can be progressively opened up with serial balloon dilatations over multiple sittings. Lastly, spontaneous dissections and aneurysms seen in TA can be effectively treated by endovascular techniques with minimal morbidity.

Conclusion

Renal-related percutaneous interventions can be effectively and safely performed guided by CO2 angiography in patients with TA and renal insufficiency. Such procedures result in relief of symptoms, better blood pressure control, improvement in left ventricular systolic function and recovery or stabilization of renal function. CO2 angiography is a useful adjunct to the interventional armamentarium available to treat patients with TA and renal insufficiency.