Single-stage arterial switch operation conversion with dextro-transposition of the great arteries and pulmonary arterial hypertension due to baffle leak after a Senning procedure

Abstract

Patients who have undergone an atrial switch operation for dextro-transposition of the great arteries (dTGA) sometimes suffer from right ventricular dysfunction, tricuspid regurgitation, arrhythmias, or baffle leaks. We report the first case of single-stage arterial switch operation conversion in an adult patient with dTGA and pulmonary arterial hypertension (PAH) due to a baffle leak after a Senning procedure. Perioperative mechanical circulatory support was required for 3 weeks. Although chronic kidney disease persisted, her left ventricular function and pulmonary hypertension improved over the first postoperative year. This is a rare case, and we believe that the patient's PAH might have helped the left ventricle withstand systemic pumping for over two decades.

Learning objective

Patients with dextro-transposition of the great arteries may face problems after an atrial switch operation because their right ventricle is responsible for systemic pumping. Staged arterial switch operation conversion is a strategy that can combat these problems but increases the risk of perioperative mortality. Pulmonary arterial hypertension allows for single-stage arterial switch operation, but the operative risk may be higher.

Introduction

Atrial switch operations were performed to treat patients with dextro-transposition of the great arteries (dTGA) until the 1980s, when the arterial switch operation (ASO) replaced it as the standard treatment for dTGA. More recently, atrial switch surgery has been associated with several complications. A recent systematic review showed that late mortality in patients who had undergone an atrial switch was 11 % [1]. The major cardiac causes included arrhythmias (45 %) and congestive heart failure as a result of right ventricular (RV) dysfunction (21 %), which gradually worsened over time.

Patients who have undergone atrial switch operations sometimes require tricuspid valve surgery [2]. A few reports have suggested staged ASO conversion. The first stage consists of pulmonary artery banding to train the left ventricular (LV) myocardium to act as the systemic pump. The following stage consists of ASO, removal of the atrial switch connection, and reconstruction of the interatrial septum [3]. Although ASO conversion seems to be a better strategy than tricuspid valve repair, the mortality of this staged procedure is as high as 20–30 % since the LV is not adapted to systemic pumping [4]. The only other alternative is heart transplantation, which is rarely performed in Japan.

We report a case of single-stage ASO conversion after a Senning procedure. Our case had three notable differences in comparison with cases in previous reports. First, the patient's RV function was preserved. Second, she had severe pulmonary arterial hypertension (PAH) since she was 4 years old. Third, single-stage ASO conversion has been rarely reported. Only 11 known successful cases have been reported in the literature to date, all of which had native LV outflow tract stenosis or atrial baffle obstruction [2], [5], [6], [7].

Case report

A 32-year-old woman with congenital dTGA with intact ventricular septum had undergone a Senning procedure at 5 months of age. Postoperatively, her mean pulmonary arterial pressure gradually increased to 20, 45, 45, 48, and 80 mmHg at 1, 4, 15, 20, and 23 years of age, respectively. Her mean pulmonary capillary wedge pressure had never been >15 mmHg; therefore, the cause of her PAH was considered to be a result of atrial shunt of a Senning baffle leak. Transcatheter baffle leak closure was attempted but failed due to the shape and size of the defect. Hence, bosentan and beraprost (PAH-specific medications) were given, and subsequent surgical repair was planned.

Atrial tachycardia occurred at the age of 30 years. Catheter ablation treatment was performed, and the arrhythmia did not recur. During this period, the patient received several second opinions on her future course of treatment.

At the age of 32 years, during her first visit to our hospital, the patient experienced fatigue and dyspnea due to pulmonary congestion. Her New York Heart Association functional classification was II. Cardiac catheterization and magnetic resonance imaging were performed (Table 1). Her PAH was mildly reduced with tadalafil, macitentan, and selexipag, and oxygen inhalation significantly improved acute vasoreactivity. There was no tricuspid regurgitation (TR) observed during transthoracic echocardiography.

Table 1.

Cardiac catheter and MRI data before initiation of PAH-specific medications, preoperatively, and one year postoperatively.

	Before PAH medications	Preoperatively	One year postoperatively
Age (years)	23	32	33
mPAP (mmHg)	80 (55a)	51 (47a)	23
LVP (mmHg)	115/edp10	77/edp8	104/edp9
RVP (mmHg)	120/edp8	105/edp10	51/edp9
Pulmonary/systemic pressure ratio	0.96	0.73	0.49
Qp/Qs	2.2	1.8 (3.0a)	1.0
PVRI (unit·m2)	14.2 (13.0a)	8.3 (4.4a)	3.6
C.I. (l/min/m2)	2.4	3.1	3.3
LVEDVI (ml/m2)	–	202 (MRI)	128 (Cath)
LVEF	–	0.54 (MRI)	0.54 (Cath)
RVEF		0.59 (MRI)	0.56 (Cath)
LV mass (g/m2)		78 (MRI)

PAH-specific medications improved the patient's PAH, and acute vasoreactivity testing by oxygen inhalation showed a significant response. Cardiac MRI was performed instead of preoperative angiography to minimize the risk of PAH deterioration. Arterial switch operation conversion led to a significant improvement in PAH and preserved systemic circulation.

mPAP, mean pulmonary arterial pressure; LVP, left ventricular pressure; edp, end-diastolic pressure; RVP, right ventricular pressure; Qp/Qs, pulmonary to systemic flow ratio; PVRI, pulmonary vascular resistance index, C.I., cardiac output index; LVEDVI, left ventricular end-diastolic volume index; LVEF, left ventricular ejection fraction; RVEF, right ventricular ejection fraction; PAH, pulmonary arterial hypertension; MRI, magnetic resonance imaging; Cath, cardiac catheter.

^aMeasured during oxygen inhalation to test acute vasoreactivity.

These data indicated that the baffle leak was significant and may have led to a poor prognosis, and her LV systolic function was expected to tolerate her systemic pressure. Hence, ASO conversion was performed.

The operation was performed without unexpected complications. However, the patient's LV function had deteriorated significantly, and she could not be weaned from cardiopulmonary bypass. In addition, hemostasis was difficult to achieve. Thus, a percutaneous LV assist device, Impella 2.5 (Abiomed, Tokyo, Japan), was inserted. Despite circulatory assistance, central extracorporeal membrane oxygenation was required in her second postoperative week. Extracorporeal membrane oxygenation was weaned with intra-aortic balloon pumping, and all mechanical circulatory supports were removed on postoperative day (POD) 16. Respiratory supports were stopped on POD 43. Renal replacement therapy was necessary from POD 2 to POD 58. Brain magnetic resonance imaging showed no signs of abnormality. She was discharged on POD 80 and was provided with home oxygen therapy.

One year after the surgery, cardiac catheterization was performed (Table 1). PAH had improved, and the LV was functioning efficiently. Her serum creatinine level had increased from 0.57 mg/dL to 0.84 mg/dL, and her estimated glomerular filtration rate had reduced from 98 mg/min/1.73 m² to 64 mg/min/1.73 m², indicating that chronic kidney disease occurred since the operation. Her New York Heart Association functional classification remained at class II.

Discussion

We discuss the first reported case of single-stage ASO conversion for an adult patient with dTGA and PAH due to a baffle leak after a Senning procedure. We considered three issues regarding this case.

First, we had to determine whether the PAH of this patient was a contraindication for surgery. Baffle leaks are a well-known adverse outcome of atrial switch operations. A recent review reported that 8 % of patients in a cohort needed reintervention following an atrial switch operation, of which 56 % required baffle repair [1].

In Japan, surgical or transcatheter repair is generally not recommended in adult patients diagnosed with PAH having non-restrictive left-to-right shunt when the pulmonary vascular resistance index is >8 Wood units･m² [8]. Although this criterion is similar to the guidelines from developed countries, there is no prospective evidence regarding long-term outcomes after a treat-and-repair approach in patients with congenital heart disease and PAH [9]. In our case, the RV served as the pulmonic pump after ASO conversion; it had previously worked as the systemic pump, and its functionality was preserved. Thus, it was presumed to prevent postoperative pulmonary circulatory collapse. Hence, we concluded that the operation was not contraindicated despite its high mortality rate.

Second, we had to determine which procedure was preferable: baffle leak closure only or baffle leak closure with ASO. Our usual strategy for baffle leaks is either transcatheter or surgical repair. There are a few reports on ASO conversion after a failed atrial switch in dTGA [3], [5], [6], [7], [10]. These cases had clear indications for the conversion procedure, such as RV dysfunction or significant TR. However, our patient had no TR, and her RV function was preserved. Hence, we were concerned that the risk of the invasive procedure may have outweighed its potential benefit. Furthermore, once the ASO is performed, it is impossible to return to the previous form, and it poses a risk for systemic circulatory failure. Studies have assessed the LV/RV systolic pressure ratio (>0.65–0.9 is the criterion for LV preparedness), LV ejection fraction (>0.55), LV mass (>65 g/m²), and other factors to estimate the potential LV function of dTGA patients after atrial switch [3], [6], [10]. However, most of these studies were performed in younger populations, and the eligibility criteria for ASO conversion remained unclear. Conversely, it is important to countermeasure the increasing prevalence of morphological RV failure when considering the life expectancy of older patients who have undergone atrial switch procedures. Nonetheless, because our patient had met at least one of the criteria above, we decided to proceed with the ASO conversion.

Third, we had to determine whether LV retraining was necessary. A small study showed the efficacy of LV retraining in patients with dTGA after an atrial switch procedure [3], [6], [10]. In our case, PAH had unintentionally trained the patient's LV for over 20 years, leading to an LV/RV systolic pressure ratio of >0.7. Nevertheless, two-stage ASO conversion may confirm LV preparedness for systemic pumping and may have a lower risk of perioperative mortality than the single-stage procedure. However, it also has some demerits, such as complications due to the additional surgical invasion, the financial and temporal burden of multiple inpatient treatments, and transient cyanosis from the right-to-left shunt of the baffle leak (rare occurrence). After considering the advantages and disadvantages of the procedures, we decided to implement a single-stage ASO conversion. However, two-stage ASO might have avoided the need for mechanical circulatory assistance.

We had great difficulty managing intensive care during the perioperative period. However, the patient's LV had adapted to the systemic circulation, and her PAH improved within a year. Furthermore, the procedure changed her systemic pump from the vulnerable RV to the tough LV, thus avoiding systemic ventricular dysfunction or atrioventricular valve insufficiency. These benefits are expected to improve her long-term prognosis.

This is the first reported case of an adult with dTGA and PAH who had undergone single-stage ASO conversion. There are several strategies for circumventing the residual symptoms or complications associated with atrial switch procedures in adults with dTGA. The risks, benefits, and prognosis of each case should be considered. Although the postoperative course has been relatively uncomplicated for our patient, our strategy may not be reproducible for unpredictable LV function, severe PAH, and cases where long-term follow-up is required.

Declaration of competing interest

The authors declare that there is no conflict of interest.