How non-committed is non-committed?

In the current issue of this journal, Kari et al. [1] from the Freiburg University Heart Center present a series of 47 patients with the transposition of great arteries with left ventricular outflow tract obstruction (LVOTO) and non-committed VSD treated with different surgical techniques. The author’s main question was if alternative techniques to the Rastelli procedure will help reduce the need for reoperations in the LVOT in the setting of ‘complex baffles’ for non-committed VSD. Results show that patients with Rastelli repairs underwent more reoperations compared to those without intraventricular baffle (14% vs 50%) and that survival was the same (98% vs 90% at 15 years P = 0.48). They concluded that Rastelli operation yield robust results in the setting of non-committed VSD and that late results in neonatal ASO are outstanding.

In this group of such a complex anatomy, the authors and the surgeons performing these procedures should be congratulated for the results. On the other hand, many other studies have shown that the Rastelli procedure (in TGA VSD LVOTO including patients with committed VSD) presents higher incidence of reoperations and mortality, as the Boston children’s [2] study, with a late survival of 52% at 20 years and freedom from RVOTO of 21% at 20 years. In the same way, the Mayo Clinic study [3], the largest published series encompassing >30 years of experience and 231 patients, reported an early mortality of 10%, but a disappointing late survival of only 48% at 20 years and freedom from reintervention of only 18% at 20 years.

Arterial switch operation (ASO), as the authors concluded, should be the foremost strategy of choice. The evaluation of the pulmonary valve and annuls by echocardiography needs to be as precise as possible to determine the strategy. A pulmonary annuls up to Z-2 DS or bicuspid valve are indeed suitable for ASO. An aggressive and meticulous resection of the infundibular septum is crucial. Obviously, extreme care must be taken to avoid the damage of the pulmonary valve. Although these valves may eventually fail in the long term, is clearly better to have an aortic valve replacement in the future than multiple RV-PA replacement and RV failure with others surgical strategies.

The Nikaidoh aortic root translocation is best indicated in the setting of a moderate small pulmonary annuls (Z-2 to Z-5), restrictive or non-committed VSD or/and straddling tricuspid valve. The main advantage in this tecnique is a perfectly aligned LVOT and RVOT, but significant early morbidity is more common [4] and requires RV–PA conduit (therefore failure and replacements). Late aortic valve regurgitation has been reported [5].

Nevertheless, in a recent publication from the STS database, the Rastelli procedure is still the most common procedure in the management of LTGA VSD and LVOTO with 56.3% of the cases, and the Nikaidoih was 40.6%. Operative mortality was similar, with 3.1% for Rastelli and 4.4% for Nikaidoh, but prevalence of cardiac arrest and permanent pacemaker placement were higher in the Nikaidoh group [4].

It would be interesting for the article by Kari et al., to see the correlation of preoperative imaging and the surgeon’s findings of the z-score values of the pulmonary valves and position of the VSD to define the surgical strategy. So, the question is how non-committed is non-committed? Defining non-committed only by operative notes may be misleading. Nowadays, the role of three-dimensional heart-printed model is mandatory in the management of these complex patients and helpful in establishing the location and size of the VSD and the relationship with the great arteries. Also, with new and better software programmes, the possibility to navigate inside the heart and extract different layers of it is already here. The decision-making in what kind of procedure suits better for each one of these patients is not simple and achieving all the available information will help to get the best results. In our experience with TGA VSD and LVOTO, and other lesions like DORV, the 3D-printed model had helped us tremendously in finding the optimal size and shape of the baffle from the VSD to the targeted great vessel.

Still, the success of a procedure in congenital heart surgery should not be based on survival or freedom from reoperation, and especially when we compared different surgical techniques for the same anomaly. Functional status must be evaluated so we can conclude if it is a successful result. In many cases, we ‘forced’ for example a biventricular repair with very complex baffles and the patient functional performance is worse than with a univentricular (Fontan) repair. Therefore, it might look pretty, it might sound pretty, but it may well perform poorly.

In conclusion, TGA with non-committed VSD and LVOT demand full attention to every anatomic detail is mandatory to select the correct surgical approach. Innovation in new technology, like 3-D computed tomography image and printed models, is the excellent tool to help define how non-committed is the VSD to the great vessels and the size of the patch. ASO should be performed if possible, avoiding the long and complex baffles and RV–PA conduits replacement of Rastelli procedure.

REFERENCES