Aortic root translocation (Nikaidoh) procedure for complex transposition of the great arteries with left ventricular outflow tract obstruction

Background Several surgical techniques have been developed for the management of complex transposition of the great arteries with ventricular septal defect and left ventricular outflow tract obstruction (TGA/VSD/LVOTO). Aortic root translocation, or the Nikaidoh operation, offers the most anatomic biventricular repair in these patients. However, the Nikaidoh operation commonly has been limited to patients with “typical” anatomy, including a conoventricular VSD and usual coronary anatomy. We sought to describe a single surgeon’s experience with aortic root translocation for complex TGA/VSD/LVOTO. Methods We present a series of 12 patients with complex anatomy who underwent the Nikaidoh operation over the last 13 years. Results We report good mid- to long-term results, excellent performance of the reconstructed left ventricular outflow tract, aortic valve competence, and no coronary insufficiency. Conclusions Our experience suggests that the Nikaidoh operation is a valid option even for patients with complex TGA/VSD/LVOTO.

with its long and angulated subaortic baffle, is now reserved for highly selected patients.Surgical management of complex or atypical forms of TGA/VSD/LVOTO remains a challenge owing to wide anatomic variability in the arrangement of the great arteries; size and location of the VSD, the atrioventricular valve (AVV), and its subvalvular apparatus; and coronary anatomy.With increasing experience with the Nikaidoh operation, some surgeons have extended its use to complex forms of TGA/VSD/LVOTO, [2][3][4][5][6][7][8] which may involve the following anatomic variants: congenitally corrected transposition of the great arteries (cc-TGA), a remote inlet or complete atrioventricular canal-type VSD with no outlet extension, unusual coronary patterns, or straddling AVVs.Here we present a single surgeon's experience (E.A.B.) with aortic root translocation for complex TGA/VSD/LVOTO over the last 13 years.

METHODS Patient Cohort
Between January 2010 and February 2023, a total of 16 patients with a diagnosis of TGA or double-outlet right ventricle (RV) with VSD and LVOTO underwent an aortic root translocation by a single surgeon (E.A.B.) at our institution.Of those, 12 had complex anatomy and were included in the present series.The decision to perform the Nikaidoh operation instead of a Rastelli-type repair was at the discretion of the surgeon and was based mainly on the conal septal anatomy, coronary pattern, and mechanism of LVOTO.
This study was approved by our hospital's Institutional Review Board (approval AAAU5073, January 12, 2023), with a waiver of individual consent.

Operative Technique
A median sternotomy is performed, and cardiopulmonary bypass is established with bicaval cannulation and moderate hypothermia.Any prior palliative shunts and the ligamentum arteriosum are ligated and divided.The great vessels are dissected, and the branch pulmonary arteries are extensively mobilized to the hilum.The coronary arteries are mobilized circumferentially as far distally as possible to allow for en bloc translocation of the aortic root with the coronary arteries attached.We do not reimplant the coronary buttons, as we strongly believe that nonanatomic reimplantation results in a higher incidence of aortic regurgitation.
Following aortic cross-clamping and cold cardioplegic cardiac arrest, the right atrium is opened, and the intracardiac anatomy is inspected.The ascending aorta is transected at the level of the pulmonary artery (PA) bifurcation, and the main PA is transected just above the sinotubular junction.We assess the location of the VSD and conal septum, the size of the pulmonary annulus, and AVV morphology to verify the appropriate anatomy for a Nikaidoh procedure.
The aortic root is then harvested from the RV (with the coronaries left in situ), starting a few millimeters below the aortic annulus on the right ventricular outflow tract (RVOT) and continuing circumferentially below the valve, leaving a muscular rim of 3 to 5 mm.The conal septum is then transected down into the VSD, thereby relieving the LVOTO.Any chordal attachments to the septum are carefully noted and kept on their correct side; chordae that are straddling the septum are detached and later reattached to the VSD patch.
The aortic root is translocated posteriorly ensuring no tension or kinking of the coronary arteries and sutured to the neo-LVOT using interrupted, mattressed polypropylene sutures.The posterior two-thirds of the aortic root are sutured to the rim of pulmonary annulus and the cut edge of the conal septum.Anteriorly, the LVOT is reconstructed by closure of the VSD with a patch extending from the crest to the aortic neo-annulus.The patch should protrude beyond the free edge of the aortic root, as it later can serve as the floor of the RVOT if a nonconduit RVOT reconstruction is planned.Any transected chordal attachments are reattached to the patch.The remaining one-third of the aortic root is then anchored to the free edge of the VSD patch to complete the LVOT reconstruction.A Lecompte maneuver is performed, and the ascending aorta is reanastomosed-often after shortening it, to avoid a posterior bulge and compression of the PA-and the RVOT is reconstructed with an orthotopic conduit or as a transannular patch with the cross-clamp applied.For these long cases, we prefer to have a competent pulmonary valve present as opposed to free pulmonary regurgitation and hence lean toward using a conduit.In many cases, the native RVOT opening is very wide and crescent-shaped; these corners can be troublesome bleeding spots and thus are closed primarily to create a more circular orifice.Video 1 demonstrates intraoperative footage of the Nikaidoh operation for complex TGA/VSD/LVOTO.

Preoperative Characteristics
The median age and weight at surgery was 2.5 years (range, 0.3-13.0years) and 14.5 kg (range, 5.8-45.0kg), respectively.The great vessel relationship was d-TGA in 9 patients (75.0%), of whom 4 (44.4%) had DORV and 3 (25.0%)had cc-TGA.One patient (8.3%) had situs inversus.All patients had severe LVOTO at the time of surgery, which ranged from valvar pulmonary stenosis to tunnel-like    Seven patients (58.3%) had at least 1 abnormal AVV, which included a straddling right AVV in 2 patients and moderate right AVV hypoplasia and stenosis, a right AVV with important chordal attachments to the conal septum, a left AVV with important chordal attachments to the VSD crest, a common AVV with a large cleft, and fibrous continuity of both AVVs with the aortic valve in 1 patient each.Nine patients (75.0%) had undergone a prior palliative procedure, including a bidirectional Glenn in 6, Blalock-Taussig-Thomas shunt in 5, balloon atrial septostomy in 2, and PA band in 2. Patients' anatomic and clinical data are presented in Table 1.

Early Surgical Outcomes
There were no operative mortalities.The median cardiopulmonary bypass time was 193 minutes (range, 166-377 minutes), and the median cross-clamp time was 125 minutes (range, 101-184 minutes).Two patients underwent early reoperations.Patient 8, the smallest patient in the series at 5.8 kg, underwent reoperation on postoperative day 1 for iatrogenic aortic insufficiency (AI) owing to a malpositioned annular suture causing distortion of the posterior aortic valve cusp; removing the suture restored aortic valve competency.Patient 9 had a complex postoperative course complicated by VSD patch dehiscence, recurrent subpulmonic outflow obstruction, and a recalcitrant chylothorax necessitating VSD patch revision, RVOT plasty, and takedown of the superior cavopulmonary connection 2 months later.He recovered well from the procedure, but was returned to the operating room 3 weeks later for repair of an RVOT pseudoaneurysm.Patient 11 had low cardiac output early due to massive aortopulmonary collaterals and required coiling on postoperative day 4.
No patient required an unplanned permanent pacemaker.Patient 8 developed de novo high-grade atrioventricular block preoperatively and underwent concomitant epicardial pacemaker insertion.Patients 2 and 12 were discharged with intermittent first-degree atrioventricular block and junctional rhythm (also present preoperatively), respectively.The remaining patients were in normal sinus rhythm at discharge.
The median postoperative length of stay was 9.5 days (range, 6.0-83.0days).Except for 1 patient with moderate left ventricular dysfunction (patient 9), all had normal biventricular function, no hemodynamically significant outflow tract obstruction, and no more than mild AI at discharge.

Late Follow-up
At a median follow-up of 4.3 years (range, 0.3-12.3years), there were 2 late deaths (Table 2).Patient 5 had heterotaxy syndrome with asplenia and died 12 months postoperatively from infectious causes.Patient 9 died 6 months after his last surgery following a 2-week history of viral illness and emesis that resulted in severe metabolic derangements and ultimately in cardiac arrest at home.In both patients, the latest echocardiogram showed an excellent repair with no more than mild systolic dysfunction and no AI or residual outflow tract obstruction.
Two patients underwent late surgical reintervention.Patient 1 required balloon angioplasty of the distal conduit 4.0 years later and conduit replacement 9.3 years later.Patient 2 developed conduit failure and underwent transcatheter pulmonary valve implantation 9.7 years later, which was complicated by transection of a tricuspid papillary muscle and severe tricuspid regurgitation.The patient subsequently underwent open pulmonary valve replacement and tricuspid valve repair the next day.Patient 7 underwent a conduit balloon angioplasty and bilateral branch PA stenting 1.9 years later, followed by stent placement within the distal conduit and proximal right PA 4.7 years after that.At the time of this report, no patients had more than mild LVOTO or mild AI at the latest follow-up.Patient 2 remained in intermittent first-degree atrioventricular block, and the remaining patients were in normal sinus rhythm.All survivors are currently New York Heart Association class I.

DISCUSSION
The Nikaidoh procedure was initially developed for standard forms of d-TGA with a conoventricular VSD and LVOTO.Despite offering the advantages of a more anatomic repair, the Nikaidoh procedure involves a higher degree of technical complexity owing to manipulation of the aortic root.Consequently, the risks of iatrogenic aortic valve incompetence and coronary ischemia are significantly higher than with a Rastelli operation.In complex forms of TGA/VSD/LVOTO, such as those involving unusual coronaries or a remote VSD, surgeons may revert to using a Rastelli-type operation or even a single ventricle approach.Although the Nikaidoh is particularly challenging in this setting, our series demonstrates that aortic translocation can be safely extended to this group of patients with excellent durability and long-term outcomes.This subset of patients presents a unique set of technical considerations, as discussed in detail below.

Anatomic Biventricular Outflow Tract Reconstruction and Management of Inlet VSDs
At several centers, the Nikaidoh is now the operation of choice for patients with d-TGA/VSD/LVOTO.Benefits over the Rastelli operation include a more natural, direct Moreover, because the aortic root is moved posteriorly, RV-PA continuity may be established at the level of the explanted aortic root (ie, the natural RVOT orifice), away from the sternum, thereby avoiding compression and facilitating future reentry.Early concerns regarding AI have been largely resolved by avoiding coronary button takedown and reimplantation, as well as with a very precise geometric root transfer. 7Therefore, we generally prefer aortic root translocation over double-root translocation where the coronaries must be taken down and reimplanted by necessity.Aortic translocation is particularly advantageous in the setting of an inlet or remote VSD, where the Rastelli either carries a great risk of LVOTO due to the curvaceous tunnel or is simply impossible.By dividing and resecting the conal septum, a straight LVOT can be achieved.In some cases, it is possible to avoid using a VSD patch completely and reconstruct the outlet septum with the muscular cuff, as was done in 2 cases in this series.Inlet VSDs can be managed similarly by approaching them from above, connecting the aortic annular opening (after enucleation of the aortic root) with the VSD, and leaving all tricuspid valve attachments on the right side.This generally has been sufficient to create a sufficient LVOT without formal enlargement of the VSD, because the conal division is then closed by the VSD patch.

Management of Complex Coronary Artery Patterns
Unusual coronary anatomy (ie, atypical coronaries for d-TGA) was previously considered a relative contraindication to the Nikaidoh procedure. 8With more experience, we now routinely harvest the aortic root en bloc and leave the coronary arteries in situ, as we have found that the harvest and reimplantation of coronary buttons tends to distort aortic root geometry and is likely the cause of neo-AI.Our single instance of early postoperative AI was due to a technical error.An annular suture had distorted the aortic valve, and removal of the offending suture resolved the issue.
When translocating the root en bloc, the key is to mobilize the coronary arteries circumferentially far enough out on the myocardium.In this series, 2 patients had inverted coronary arteries (1LR-2Cx), and the anterior and posterior loops were managed with extensive circumferential dissection to avoid kinking or stretching of the vessels.In both cases, the right coronary artery crossed over the RVOT.Because RV-PA continuity is constructed in the native RVOT, the crossing coronary remains anterior to the RVOT and not underneath a heterotopic conduit as in a Rastelli.Therefore, coronary compression is not an issue with the Nikaidoh procedure as it is with the Rastelli operation.In fact, the initial plan in 1 of these patients (patient 3) was to perform a Rastelli operation; however, we pivoted intraoperatively to an aortic translocation procedure on visualization of the right coronary artery crossing the RVOT, as this technique avoids division of the RVOT conus.

Management of cc-TGA
Three patients in our cohort had cc-TGA, including 2 with {S,L,L} segmental anatomy and 1 with situs inversus {I,D,D} segmental anatomy.All had usual coronaries for their anatomy.In these patients, the operation does not vary much from aortic translocation for d-TGA/ VSD/LVOTO, except that the plane of the translocation is not anterior/posterior but rather more oblique, in a right/left transverse orientation, as the great arteries are often side-by-side or obliquely related.Given that the atrial switch is rarely performed today, we prefer to wait until these patients weigh at least 8 to 10 kg.In {S,L,L} patients in particular, the conduction system is at risk and often runs anterior and superior to the pulmonary valve before descending along the anterior margin of the VSD.Therefore, division of the conus should occur as leftward as safely possible, and care should be taken when suturing the VSD patch to avoid injury to conduction tissue.No patients in our series developed postoperative complete atrioventricular block necessitating a permanent pacemaker.

Limitations
Limitations of this series include its retrospective nature and potential lack of external validity given the small cohort size and single-surgeon experience.Additionally, given the rarity and anatomic heterogeneity of this lesion, a multicenter collaboration with long-term follow-up is needed to evaluate the durability and efficacy of the Nikaidoh operation for complex TGA/VSD/LVOTO.

CONCLUSIONS
Overall, the present series adds to the few isolated reports on aortic translocation and biventricular outflow tract reconstruction for complex or atypical forms of TGA/ VSD/LVOTO.In experienced centers, the Nikaidoh operation is a particularly valuable option for patients with unfavorable anatomy for a Rastelli procedure.We demonstrate that the benefits of the Nikaidoh operation-the most anatomic biventricular repair possible-may be extended to these patients with variable and challenging anatomy, with excellent results. of interest.The editors and reviewers of this article have no conflicts of interest.

TABLE 1 .
Summary of patient anatomic and operative data JTCVS Techniques c Volume 22, Number C 245 Bacha et al Congenital: Transposition of the Great Arteries