I nfants with cyanotic congenital heart disease(CHD) often require a surgical systemic-to-pulmonary artery shunt (e.g., Blalock-Taussig
shunt [BTS]). BTS stenosis may result in progressivecyanosis whereas thrombotic occlusion can result inacute, severe cyanosis and, in patients who had theBTS is the sole source of pulmonary blood flow,cardiac arrest or death (1,2).
Several studies have demonstrated that trans-catheter interventions are effective treatment for BTSocclusion and restoring flow through the shunt (3–5).Transcatheter stenting of an occluded or stenotic BTShas been shown to be an effective rescue strategy torestore patency in the case of thrombotic obstructionof a modified BTS (6,7). In cases of severe cyanosisor an unstable infant dependent on the BTS, a rapid
m the Children’s Healthcare of Atlanta, Department of Pediatrics, Emory
thors have reported that they have no relationships relevant to the conte
nuscript received February 2, 2017; revised manuscript received May 3, 2
re-establishment of BTS patency is therefore critical.Although the standard approach uses the femoralartery (FA) for these interventions, the carotid artery(CA) may be a preferable, alternative access site inmany cases. It is recognized that access via the CAprovides a more direct route for interventionsdirected at some BTS—particularly those BTS thatarise from the common CA itself—and published datasuggests favorable outcomes for procedures per-formed via percutaneous puncture of the CA withoutsurgical cutdown (8,9).
We adopted a percutaneous CA approach initiallyin cases where the origin of the BTS was the commonCA, but since have used the CA for most cases whereintervention on the BTS is anticipated. The goal of thecurrent study is to compare outcomes following the
University School of Medicine, Atlanta, Georgia. The
J A C C : C A R D I O V A S C U L A R I N T E R V E N T I O N S V O L . 1 0 , N O . 1 7 , 2 0 1 7 Ligon et al.S E P T E M B E R 1 1 , 2 0 1 7 : 1 7 3 8 – 4 4 Carotid Access Is Superior for BT Shunt Interventions
1739
CA approach with outcomes following the FAapproach for BTS interventions in infants with CHD.We hypothesized that there would be notable im-provements in efficiency as well as overall successwhen using the CA for access.
FIGURE 1 Breakdown of Patients Undergoing Interventional
Cardiac Catheterization for BTS Interventions
42 BTS Interven�ons
34 pa�ents
32 FA 10 CA
5 cases converted to CA with subsequent
success3 unsuccessful* all 10
successful
*1 patient was taken straight to the operating room after
intervention was unsuccessful from the femoral artery (FA);
another required device plug placement by vascular surgery due
to a FA tear and the case was abandoned; another had FA
spasm while removing a stent that had been attempted.
BTS ¼ Blalock-Taussig shunt; CA ¼ carotid artery.
SEE PAGE 1745
CHD = congenital heart disease
femoral artery
METHODS
A retrospective chart review was conducted of allpatients who had undergone cardiac catheterizationfor intervention of a systemic-pulmonary arteryshunt obstruction performed via FA or CA access.All cases of catheter-based intervention for BTSobstruction between May 2012 (marking when ourinstitution first began using carotid access) untilJanuary 2017 were reviewed. This study wasapproved by our institutional review board.
Cases of BTS intervention from the CA approachwere compared with those undergoing interventionfrom the FA approach. The primary endpoint of ourstudy was procedural success, measured as successin performing planned intervention (either stentimplantation or balloon angioplasty). Secondaryendpoints included time from access to successfulintervention, time from intention to intervene tothe successful intervention (i.e., time after angio-graphic determination of BTS narrowing tosuccessful BTS intervention), radiation exposure(fluoroscopy time and dose area product), and pro-cedural complications. Vessel patency (FA or CA)following BTS intervention was also noted. All CAcases had a follow-up carotid ultrasound with colorDoppler within 24 h of the procedure to rule outvascular injury, stenosis, or thrombus, whereas FAcases were followed clinically with imaging of theFA in cases where there was clinical concern. Com-parison between the FA and CA cohorts was per-formed with an as-treated analysis, given that5 cases were transitioned from femoral to carotidaccess. The data that were collected during thesespecific cases was calculated starting from the firstattempt to obtain CA access, and any proceduretime contributed during FA access was subtractedfrom the total as appropriate. Independent-samplesWilcoxon rank sum test was used as appropriatefor comparison between the FA and CA cohorts. Chi-square analysis for categorical outcomes suchas procedural success and vessel patency wasperformed. Statistical analysis was performedusing SPSS version 23 (IBM Corporation, Armonk,New York).
RESULTS
During the study period, 34 patients under-went 42 BTS interventions in our cardiaccatheterization laboratory (Figure 1). In 32catheterization cases, intervention began us-ing the FA approach, whereas in 10 cases the
CA was accessed at the outset of the procedure. Allcatheterizations were performed specifically to assessand potentially intervene on the BTS, as suspicionfor a stenotic shunt was due to either patient cyanosis(2 of 42, 64%) or shunt occlusion by echocardiography(15 of 42, 36%). The median patient age at catheteri-zation was 107 (interquartile range: 8 to 1,634) daysand median weight was 4.55 (interquartile range: 3.0to 14.1) kg. There was no statistically significantdifference between the cohorts in regard to the age(p ¼ 0.56), weight of the patients (p ¼ 0.59), numberof stents placed (p ¼ 0.36), fluoroscopy time(p ¼ 0.17), or dose (p ¼ 0.14) (Table 1). There were 2patients from each cohort (4 cases total) that were onextracorporeal membrane oxygenation during thecatheterization and thus fully heparinized at thetime of the procedure. A 4-F hydrophilic shortsheath was used in all 10 CA cases, and a 4-F longsheath used in 26 FA cases, with a short sheath in 6FA cases. A single stent was used in 18 cases, 2 stents
FA =
TABLE 1 Comparing the Femoral Arterial Approach Versus Carotid Arterial Approach
Femoral Access(n ¼ 27)
Carotid Access(n ¼ 15) p Value
Age, days 103 (59–140) 143 (33–224) 0.56
Weight, kg 4.6 (4.0–6.4) 4.75 (3.5–8.4) 0.59
Diagnosis
Single ventricle 8 (29.6) 5 (33.3)
Tetralogy of Fallot/DORV 11 (40.8) 5 (33.3) N/A
Pulmonary atresia/MAPCAs 8 (29.6) 5 (33.3)
ECMO at time of Intervention 2 (7) 2 (13) 0.54
Carotid artery origin of BTS 9 (33) 6 (43) 0.34
Other interventions performed 8 (29.6) 7 (50) 0.31
Procedure time, min 104 (75–172) 62 (40–106) 0.01
Time to arterial access, min 9.3 (6–12) 4.0 (3–7) <0.01
Time from sheath placement to guidewirethrough shunt, min
13 (10–19) 6.5 (4–9) <0.01
Time from sheath placement to stentdeployment, min
20 (16–29) 9 (7–17) <0.01
Anesthesia time, min 151 (115–222) 119 (88–148) 0.01
Fluoroscopy time, min 32 (20–50.8) 16 (10–59.3) 0.17
Radiation dose, mGy total 100 (67–214) 53.5 (36–330.25) 0.14
Total contrast volume, ml 27 (22–43) 24 (17–32) 0.49
Number of stents 1 (0–2) 1 (1–2) 0.36
Complications 3* 0 N/A
Unsuccessful intervention 8 (30) 0 (0) 0.035
Values are median (25th to 75th percentile) or n (%). *One patient in the femoral cohort required device plugplacement by vascular surgery due to a femoral artery tear; another had a vascular obstruction due to athrombus; another had femoral arterial spasm while removing a stent that had been attempted.
BTS ¼ Blalock-Taussig shunt; DORV ¼ double outlet right ventricle; ECMO ¼ extracorporeal membraneoxygenation; MAPCA ¼ major aortopulmonary collateral artery; N/A ¼ inadequate numbers for statisticalcomparison.
Ligon et al. J A C C : C A R D I O V A S C U L A R I N T E R V E N T I O N S V O L . 1 0 , N O . 1 7 , 2 0 1 7
Carotid Access Is Superior for BT Shunt Interventions S E P T E M B E R 1 1 , 2 0 1 7 : 1 7 3 8 – 4 4
1740
were used in 13 cases, and 3 stents were implantedin 2 cases; 9 cases underwent balloon angioplastyalone for intervention.
Overall, 39 of 42 (93%) interventions on the BTSwere ultimately successful. In all 10 cases that beganwith CA access, BTS intervention was successful.Among the FA cohort that began with FA access(32 cases), however, BTS intervention was successfulin only 24 of 32 (75%) of cases, with inability to stentthe BTS from the FA approach in 8 cases. The mostcommon reason for lack of success was difficulty withstent tracking over guidewire (n ¼ 3) or completeocclusion of the BTS with thrombus (n ¼ 5) withinability to traverse thrombus from the FA approach.In 5 of these 8 unsuccessful BTS cases from the FA thedecision was made to intervene on the BTS using theCA after multiple attempts were made from the FAapproach. In these 5 cases, access was obtained fromthe CA approach during the same catheterization(Figure 2). Subsequent successful BTS interventionwas achieved in all 5 catheterizations. In theremaining 3 unsuccessful FA cases, 1 patient wastaken directly to the operating room for surgical BTSrevision. Another unsuccessful intervention occurred
in a patient who had complete thrombosis of his BTSand subsequently experienced a FA tear occurredduring long sheath advancement, resulting in cessa-tion of the procedure and device plug placement byvascular surgery. The third FA failure had an existingBTS stent; attempted stent placement in the distalportion of BTS was unsuccessful due to the curvaturethrough the innominate artery. Angiography andDoppler demonstrated an occluded right FA statuspost-stent removal. This patient underwent success-ful BTS stenting from the CA 1 week later.
Among the 10 cases that began with CA access, allhad successful BTS stent (n ¼ 9) or angioplasty (n ¼ 1).The angioplasty of the BTS was the intention in apatient with partially occlusive thrombus of the distalBTS in the immediate post-operative period. In 3cases from the CA, complete occlusion of the BTS wasnoted. In 2 of these cases, the patients were emer-gently transferred to the catheterization laboratoryfor BTS stenting due to sudden profound cyanosis. Inthese cases, rapid carotid access (2 and 4 min,respectively) was achieved with facile recanalizationof a completely thrombosed BTS performed (Figure 3).In both of these cases, stenting was then performedafter aggressive angioplasty and maceration of BTSthrombus. In the third case of BTS occlusion, aninfant with pulmonary atresia and major aorto-pulmonary collaterals that had undergone a saphe-nous vein graft BTS was referred for interventionafter thrombus was noted at the region of a suspectedvenous valve within the graft. Only with the use ofthe tip of the short dilator with a guidewire was thevenous valve ultimately crossed and the BTS angio-plastied and thereafter stented (Figure 4).
All CA patients underwent surveillance carotid ul-trasound within 24 h of the procedure, demonstratingno thrombus and complete CA patency in all cases.When we included the 5 patients who crossed overfrom FA to CA approach, procedure time was lower inthe CA cohort (62 min vs. 104 min FA; p ¼ 0.01), aswas anesthesia time (119 min vs. 151 min FA; p ¼ 0.01).Notably, time to arterial access was shorter (4.0 minvs. 9.3 min FA; p < 0.01) and so was time to placementof the guidewire through the BTS (6.5 min vs. 13 minFA; p < 0.01). Additionally, time from the finalsheath to BTS stent implantation was lower (9 min vs.20 min; p < 0.01) (Table 1).
We also analyzed the data on an intention-to-treatbasis in which the 5 patients who were transitionedfrom FA to CA approach were excluded from the CAcohort. The procedure time remained lower in the CAcohort (49 min vs. 105 min FA; p < 0.01) as wasanesthesia time (92.5 min vs. 146 min FA; p ¼ 0.01).Further, the CA cohort had a shorter time to arterial
FIGURE 2 Femoral Arterial Access Transitioned to Carotid Arterial Access
(A) Initially, the right femoral artery was used for angiography and intended intervention, demonstrating the complete shunt obstruction proximally (arrow). (B) Due to
difficulty with recanalization, access was transitioned to carotid arterial access with a much-improved angle of intervention. (C) After stenting, there is much-improved
caliber of the shunt.
J A C C : C A R D I O V A S C U L A R I N T E R V E N T I O N S V O L . 1 0 , N O . 1 7 , 2 0 1 7 Ligon et al.S E P T E M B E R 1 1 , 2 0 1 7 : 1 7 3 8 – 4 4 Carotid Access Is Superior for BT Shunt Interventions
1741
access (3.5 min vs. 9 min FA; p < 0.01) and shortertime to placement of the guidewire through the BTS(5.5 min vs. 12 min FA; p < 0.01), and time from thefinal sheath to BTS stent implantation was lower (8.5min vs. 19.5 min; p < 0.01).
DISCUSSION
In reviewing our experience intervening on surgicalsystemic-pulmonary artery shunts via the FA and CAapproaches, we found a higher success rate inrelieving shunt obstruction when using CA access.Further, there were notable improvements in effi-ciency in regard to shorter time to access and inter-vention. No complications or adverse events wereassociated with CA access and there were no concernsfor carotid patency on follow-up imaging viaultrasound.
Catheter-based intervention to stent or angioplastystenotic or occluded BTS can be technically chal-lenging, and often occurs in an urgent or emergentclinical setting. Particularly in the patient withpulmonary blood flow solely arising from the BTS, astenotic or occluded shunt causes severe, life-threatening cyanosis. In several of our patients BTSocclusion was complete and associated with clinicaldeterioration. Particularly in the cases of completeocclusion—and in the inverted saphenous vein graftwith venous valve occlusion—the carotid routeallowed for quick recanalization not only because of
the proximity of the BTS to the origin of the sheath,but also because the angle of approach was so straightthat direct, linear force could be applied to theperforating wire permitting recanalization. Our datasuggest that timely intervention is more achievableby choosing the optimal access route and that the CAshould be considered when BTS intervention isanticipated.
The percutaneous entry to the CA has been largelyabjured among the pediatric interventional cardiol-ogy community because of concerns regarding injuryto the artery itself or potential risk for embolic stroke.We have been very methodical in our percutaneouscarotid cases and have always obtained carotid ul-trasound with Doppler within 24 h of the procedure.We found that CA access has a low rate of thrombosis,and risks of vascular complications were in fact lowerthan with the use of the FA. These findings areconfirmed by a recent publication by Justino andPetit (8), which demonstrated excellent technicalsuccess with interventions using the carotidapproach. Otherwise there has been little publisheddata on the topic of percutaneous carotid access forcatheter-based interventions, with most reportsdemonstrating success using the surgical cutdowntechnique (10,11), most commonly for aortic valvu-loplasty (9,12,13). However, recent published data hasbegun to shift attention toward the encouragingsafety profile of percutaneous puncture in the CA fortranscatheter interventions (8). Our study highlights
FIGURE 3 Emergency Recanalization of a BTS
Recanalization of a left systemic-pulmonary artery shunt was performed in a patient with life-threatening cyanosis (saturations 11% to 29%
on 100% fraction inspired oxygen) via the left carotid artery. (A) Angiography demonstrates the point of obstruction within the shunt
(arrow). (B) The entire procedure from access attempt until recanalization with 2 balloon angioplasties of the shunt required only 20 min. (C)
In a different patient, the right carotid artery was used for angiography, demonstrating the region of narrowing at the systemic arterial end of
the shunt (arrow). (D) Two stents were implanted through the same short sheath with resolution of shunt stenosis. The procedure was 15
min from first attempt to gain access until deployment of second stent—with a total sheath time of 27 min. BTS ¼ Blalock-Taussig shunt.
Ligon et al. J A C C : C A R D I O V A S C U L A R I N T E R V E N T I O N S V O L . 1 0 , N O . 1 7 , 2 0 1 7
Carotid Access Is Superior for BT Shunt Interventions S E P T E M B E R 1 1 , 2 0 1 7 : 1 7 3 8 – 4 4
1742
this comparison between the traditional FA route tothe CA approach with regard to procedural successand efficiency.
Unfortunately, well known to each provider per-forming pediatric cardiac interventions are both theconcerning acute and long-term complications of FAocclusion. Studies have suggested that low bodyweight is a risk factor for access complications duringFA catheterizations (14,15) as well for procedure fail-ure in BTS interventions (6). Interestingly, our studyreinforces the previously published findings thateven at small weights (<3 kg) percutaneous CA accessis safe, with no cases of longstanding arterial occlu-sion (8,16–18).
Our study provides a comparison of FA access toCA in a single institution using the same protocols forcatheterization as well as methods to standardize
data collection. Further, these patients were treatedover the same timeframe, by the same intervention-alists, controlling for era effect. Given similar patientsand the same interventionalists, our data supportgreater procedural success with the use of CAapproach. The presence of extracorporeal membraneoxygenation and being fully heparinized did not fac-tor into our approach method and no complicationsfrom the 4 patients (2 FA and 2 CA) were directlyattributed to this.
STUDY LIMITATIONS. Limitations of our findingsinclude that this is a small cohort from a retrospec-tive single-center study. The results may not benecessarily generalizable to every pediatric heartcenter. Further, percutaneous carotid access requiresa uniquely patient and methodical approach as well
FIGURE 4 Complete BTS Occlusion
(A) Suspected venous valve within the saphenous vein graft (arrow). (B) Guidewire easily passed given advantageous angle from access.
(C) Stent deployment in the Blalock-Taussig shunt (BTS). (D) Status post-intervention with resolution of BTS occlusion following stent
implantation (arrow).
J A C C : C A R D I O V A S C U L A R I N T E R V E N T I O N S V O L . 1 0 , N O . 1 7 , 2 0 1 7 Ligon et al.S E P T E M B E R 1 1 , 2 0 1 7 : 1 7 3 8 – 4 4 Carotid Access Is Superior for BT Shunt Interventions
1743
as specific equipment described in a prior publica-tion (8). As with any new practice, there was alearning period where these new methods requiredmore time to be performed. It was noted that thetime to arterial access was shorter in the carotidaccess patients that were performed in the latter halfof the dataset. Finally, although ultrasound guid-ance was used for all CA access, it was rarely usedfor FA access and this may account for some of thedifference in time to arterial access. However,ultrasound guidance would not be expected to play
any role in the rate of success in BTS intervention ortime to BTS intervention.
Although most centers use the FA for interventionson surgical systemic-pulmonary artery shunts, ourapproach has included the usage of the CA as analternative access site. At our institution, the CAapproach has been associated with greater proceduralsuccess, shorter time to gain arterial vascular access,quicker recanalization of the shunt with placement ofa guidewire, and shorter time from final sheath up-grade to stent placement. Operators should consider
PERSPECTIVES
WHAT IS KNOWN? Studies have shown that
transcatheter BTS interventions are effective.
WHAT IS NEW? Our study shows that access from
the CA for BTS intervention is an effective and effi-
cient method when compared to traditional FA access.
WHAT IS NEXT? The next step is to evaluate the
medium- and long-term safety of CA access.
Ligon et al. J A C C : C A R D I O V A S C U L A R I N T E R V E N T I O N S V O L . 1 0 , N O . 1 7 , 2 0 1 7
Carotid Access Is Superior for BT Shunt Interventions S E P T E M B E R 1 1 , 2 0 1 7 : 1 7 3 8 – 4 4
1744
the route of access to the surgical systemic-pulmonary arterial shunt deliberately, particularlywhen urgent intervention is needed.
CONCLUSIONS
Our data indicate that percutaneous interventions onBT shunts are often performed more rapidly and withgreater success when the carotid approach isemployed. While percutaneous carotid access requirescareful approach, fastidious technique, and post-procedural surveillance, we have found use of carotidapproach to be safe and highly efficacious, particularlyfor this high-risk population of infants with occludedor stenotic shunts. We believe that considerationshould be given for use of the carotid approach whenintervention on the BT shunt is expected.
ADDRESS FOR CORRESPONDENCE: Dr. ChristopherJ. Petit, 2835 Brandywine Road, Suite 300, Atlanta,Georgia 30341. E-mail: [email protected].
RE F E RENCE S
1. Hehir DA, Dominguez TE, Ballweg JA, et al. Riskfactors for interstage death after stage 1 recon-struction of hypoplastic left heart syndrome andvariants. J Thorac Cardiovasc Surg 2008;136:94–9,99 e1–3.
2. Fenton KN, Siewers RD, Rebovich B, Pigula FA.Interim mortality in infants with systemic-to-pulmonary artery shunts. Ann Thorac Surg 2003;76:152–6; discussion 156–7.
3. Petit CJ, Gillespie MJ, Kreutzer J, Rome JJ.Endovascular stents for relief of cyanosis insingle-ventricle patients with shunt or conduit-dependent pulmonary blood flow. CatheterCardiovasc Interv 2006;68:280–6.
4. Gillespie MJ, Rome JJ. Transcatheter treatmentfor systemic-to-pulmonary artery shunt obstruc-tion in infants and children. Catheter CardiovascInterv 2008;71:928–35.
5. Diab YA, Ramakrishnan K, Alfares FA, et al.Transcatheter treatment of thrombosis in thesingle ventricle pathway: an institutional experi-ence. Congenit Heart Dis 2016;11:39–44.
6. Bonnet M, Petit J, Lambert V, et al. Catheter-based interventions for modified Blalock-Taussigshunt obstruction: a 20-year experience. PediatrCardiol 2015;36:835–41.
7. Vaughn GR, Moore JW, Mallula KK, Lamberti JJ,El-Said HG. Transcatheter stenting of the
systemic-to-pulmonary artery shunt: a 7-yearexperience from a single tertiary center. CatheterCardiovasc Interv 2015;86:454–62.
8. Justino H, Petit CJ. Percutaneous common ca-rotid artery access for pediatric interventionalcardiac catheterization. Circ Cardiovasc Interv2016;9:e003003.
9. Davenport JJ, Lam L, Whalen-Glass R, et al. Thesuccessful use of alternative routes of vascularaccess for performing pediatric interventionalcardiac catheterization. Catheter Cardiovasc Interv2008;72:392–8.
10. Latson LA, Qureshi AM. Techniques fortranscatheter recanalization of completelyoccluded vessels and pathways in patients withcongenital heart disease. Ann Pediatr Cardiol2010;3:140–6.
11. Brzezinska-Rajszys G. Access from thecommon carotid artery. In: Sievert H,Qureshi S, Wilson N, Hijazi ZM, editors. Percu-taneous Interventions for Congenital HeartDisease. Hove, UK: Informa Healthcare, 2007:97–100.
12. Maeno Y, Akagi T, Hashino K, et al. Carotidartery approach to balloon aortic valvuloplasty ininfants with critical aortic valve stenosis. PediatrCardiol 1997;18:288–91.
13. Carminati M, Giusti S, Spadoni I, et al. Balloonaortic valvuloplasty in the first year of life. J IntervCardiol 1995;8:759–66.
14. Glatz AC, Shah SS, McCarthy AL, et al. Preva-lence of and risk factors for acute occlusive arterialinjury following pediatric cardiac catheterization: alarge single-center cohort study. Catheter Car-diovasc Interv 2013;82:454–62.
15. Brotschi B, Hug MI, Kretschmar O, Rizzi M,Albisetti M. Incidence and predictors of cardiaccatheterisation-related arterial thrombosis inchildren. Heart 2015;101:948–53.
16. Steinberg C, Weinstock DJ, Gold JP,Notterman DA. Measurements of central bloodvessels in infants and children: normal values.Cathet Cardiovasc Diagn 1992;27:197–201.
17. Gasparella M,Milanesi O, Biffanti R, et al. Carotidartery approach as an alternative to femoral accessfor balloon dilation of aortic valve stenosis in neo-nates and infants. J Vasc Access 2003;4:146–9.
18. Choudhry S, Balzer D, Murphy J, Nicolas R,Shahanavaz S. Percutaneous carotid artery accessin infants < 3 months of age. Catheter CardiovascInterv 2016;87:757–61.
KEY WORDS cardiac catheterization,carotid artery, pediatric cardiology
