OBJECTIVES: Leftward displacement of the septum primum is usually described as associated with hypoplastic left heart syndrome or visceral heterotaxy. This rare malformation results in partially or totally anomalous pulmonary venous drainage with a normal connection of the pulmonary veins to the left atrium, depending on the degree of septal shift. We report the 3D echocardiographic and anatomic findings as well as the surgical repair in a series of isolated severe leftward displacement of the septum primum, responsible for totally anomalous pulmonary venous drainage.
METHODS: Three patients presenting with situs solitus and extreme leftward displacement of the septum primum were included. All of the pulmonary veins drained anomalously into the anatomical right atrium, and the distance between the mitral valve and the abnormal septum primum was greatly reduced, compromising the size of the left atrial chamber, but with normal left ventricle diameters. Preoperative 3D echocardiographic findings are reported. We achieved a biventricular surgical repair in all cases. The atrial septation was accomplished using an autologous pericardial patch after removing the abnormal septal membrane.
RESULTS: The postoperative course was free from any cardiovascular complications. Echocardiographic scans showed a harmonious reconstruction without pulmonary venous obstructions or stenosis.
CONCLUSIONS: This article reports the severe leftward displacement of the septum primum presented as an isolated cardiac malformation; 3D transthoracic echocardiography allowed an accurate diagnosis of this malformation and helped in choosing the best surgical strategy.
Leftward displacement of the septum primum (LDSP) is a rare malformation strongly associated with different degrees of anomalous pulmonary venous drainage .
A normally developed atrial septum defines the limit between the left atrium (LA) and the right atrium (RA). The septum primum (SP) grows to the left of the opening of the sinus venosus, which leaves the evagination of the common pulmonary vein opening into the LA .
The inferior part of the interatrial septum is derived from the musculature of the vestibular spine and the mesenchymal cap of the SP, thus producing a buttress for the anteroinferior oval fossa and forming the basis of the ﬁbrous septal structures of the heart .
The cephalic border of the SP normally attaches to the superior limbic band of the septum secundum, on its left side, which results from the infolding of the atrial roof.
This process of infolding, when complete, provides the buttress against which the SP ﬂap valve can close in postnatal life .
Concerning the pulmonary venous return, the venous primordium acquires its lumen only after its leftward shift, concomitant with the division of the common atrial chamber into right and left components by the growth of the primary atrial septum .
If the septum secundum fails to develop, the cephalic border of the SP loses its connection to the superior limbic band. Consequently, the SP could be shifted leftwards (in situs solitus) by the blood stream that moves from the RA towards the LA during foetal circulation .
In patients with situs inversus of the viscera and atria, SP can be displaced to the right into the right-sided LA.
Depending on the degree of displacement towards the LA, the scenario varies from partially to totally anomalous pulmonary venous drainage (TAPVD), presenting externally with a normal connection of the pulmonary veins with the morphological LA.
LDSP with no other cardiac malformations is an unusual finding. It was originally described as being strongly associated with other syndromes such as heterodaxy or hypoplastic left heart syndrome (HLHS) .
We describe the management of 3 patients affected by an isolated displacement of the SP. In all cases, the extreme leftward shift of the SP resulted in TAPVD with a normal pulmonary venous connection to the morphological LA.
MATERIAL AND METHODS
Between September 2014 and November 2015, 3 patients were diagnosed by 2D transthoracic echocardiography (2D-TTE) as having an isolated LDSP causing TAPVD.
All 3 patients were treated by our surgical team. Two patients were operated on in Toulouse, France and 1 child was operated on during a humanitarian mission in Port-au-Prince, Haiti. All patients presented with symptoms of cardiac failure.
Case 1: A 5.5-year-old child who presented with severe dyspnoea was scheduled for the surgical repair of an isolated atrial septal defect (ASD) during our humanitarian mission in Haiti. Weight and height were 19.0 kg and 110 cm, respectively. During the preoperative screening process, 2D-TTE indicated a different diagnosis. The SP was severely shifted towards the LA, and all 4 pulmonary veins, normally connected to the heart, drained into the RA (Fig. 1).
Case 2: A 6.5-year-old child with a history of intrauterine growth retardation presented with dyspnoea and was scheduled for an LDSP associated with a TAPVD surgical repair in Toulouse. Weight and height were 14.5 kg and 110 cm, respectively (Figs 2 and 3)
Case 3: A 3-month-old child who presented with signs of cardiac failure and sinus bradycardia was referred for an LDSP associated with a TAPVD surgical repair in Toulouse. Weight and height were 3.7 kg and 50 cm, respectively (Fig. 7).
The diagnosis was made in all 3 cases by 2D-TTE (iE 33, Philips Medical Systems, Andover, MA, USA), not 3D.
All patients had an extreme LDSP responsible for TAPVD to the RA. However, the pulmonary venous connection was normal, i.e. to the left side of the atrial body. The superior limbic band was absent in all 3 cases. The orientation of the SP was clearly displayed from the apical and subcostal 4-chamber views (Fig. 1). In all cases, there was an atrial shunt via an ASD located between the SP and the posterior atrial wall.
ASD form and location were seen more clearly by 3D transthoracic echocardiography (3D-TTE) (Fig. 7C) as was the failure of the formation of the superior limbic band between the superior vena cava orifice and the right upper pulmonary vein (Figs 3B and 4C) using X5-1 matrix probes (Philips Medical Systems, Andover, MA, USA). The LA was small, whereas the RA and ventricle were severely enlarged. The pulmonary arterial pressure was normal. In all cases, there was situs solitus with ventricular D-loop, atrioventricular concordance and normally related great arteries. The systemic venous return was normal.
Anatomic findings and surgical repair
After a full median sternotomy, the pericardium was opened longitudinally. The extracorporeal circulation was started through an aortobicaval cannulation. After aortic cross-clamping, the heart was arrested by cool-blood cardioplegia, the RA was opened and the surgical inspection was started.
At first glance, the atrial cavity looked like a common atrium; the limits between the right and left atrial sides were totally undeveloped (Fig. 4). The superior limbic band of the septum secundum was absent, and the anteroposterior rims of the ASD were not represented (Fig. 4). All 4 pulmonary veins drained into this ‘common atrium’. At the inferior part of this large ASD, a membrane extended from the tendon of Todaro, at the level of the atrioventricular junction (AVJ), to the free wall of the LA, between the inferior pulmonary vein orifices and the mitral valve. The membrane presented a large ‘perforation’ at its posterosuperior aspect (Fig. 4). This membrane divided the left atrial cavity like in a ‘cor triatriatum’, and the perforation assumed the role of a common orifice for pulmonary venous drainage. From the outside, the Sondergaard groove was not represented, but the pulmonary venous connections looked normal. The vestibular component of the RA and the AVJ were normal (Fig. 4).
For this reason, in all our cases, the membrane was identified as the SP, shifted leftwards. The lack of the septum secundum superior limbic band, associated with the posterosuperior perforation, demonstrated the embryological origin of this membrane: It represents the SP just after the confluence of septal perforations to create the ostium secundum. Analysed through ancestral ostium secundum, the mitral valve appeared to be normal (Fig. 5).
After this meticulous anatomical inspection, the SP was carefully excised (Fig. 5). An anatomical atrial septation was obtained in all cases, using a large autologous pericardial patch sewed with a 5/0 polypropylene running suture without mitral valve damage, or pulmonary venous orifices distortion or stenosis. The RA was closed in a standard fashion, and the patients were easily weaned off cardiopulmonary bypass without inotropic support.
The aortic clamping times were 30, 46 and 40 min, and the duration of the extracorporeal circulation was 48, 66 and 64 min, respectively. No inotropic support was necessary, and all patients were extubated on Day 1 following the operation.
The postoperative course was free from any cardiovascular complications. 2D-TTE confirmed the good result with the absence of a residual shunt and showed a harmonious reconstruction with no pulmonary vein obstructions or stenosis. All patients were discharged from the hospital 4 days after surgical repair. Further regular postoperative reviews demonstrated a good clinical condition without any underlying residual cardiac lesions in all patients.
LDSP was clinically described by Stella Van Praagh in 1995 as ‘septum primum malposition’. Van Praagh and her colleagues described an interatrial communication associated with the absence of the superior limbic band of the septum secundum and totally or partially anomalous pulmonary venous drainage due to malalignment of the SP. Furthermore, in this series, the septal abnormality was strongly associated with visceral heterotaxy .
The malposition of the primary atrial septum was also demonstrated in the ‘double outlet right atrium’, a different entity that depends on the deviation of the primary atrial septum, but in the setting of a common AVJ [5, 6].
In 2013, Park et al.  described LDSP in patients presenting with HLHS. The authors elucidated the relationship between atrial septation and left heart haemodynamic development, affirming that LDSP was ‘unique in patients with HLHS and has not been reported in other types of congenital heart disease’.
Silvestri et al.  commented on the article by Park et al.  comparing LDSP to the different anatomical patterns of the superior limbic band that are present in children with HLHS and absent in those with polysplenia syndrome. The investigators suggested that LDSP may be pathogenetically related to a left-sided obstruction in children with HLHS (usually not associated with anomalous pulmonary venous drainage) or systemic obstructions in those with polysplenia.
As suggested by Van Praagh and colleagues, Silvestri confirmed that the absence of the superior limbic band seems to be embryogenetically related to the malposition of the SP associated with TAPVD.
In 2005, Tomar et al.  described LDSP as a rare variant of anomalous pulmonary venous drainage and provided 2D echocardiographic scans: In their paper, the subxiphoid coronal, apical 4-chamber and parasternal long-axis views show the deviation of the SP most clearly. Tomar et al. described LDSP as an isolated malformation in 1/3 of his patients but he makes no mention of the anatomy of the superior limbic band .
Based on our experience, we believe that an AVJ anomaly or a left sided or systemic obstruction should be researched in any case of LDSP associated with a TAPVD. On the other hand, we demonstrated in this article that this malformation can also occur as an isolated anomaly with no other abnormal cardiac or situs conditions, when the superior limbic band is absent and when the AVJ and ventricular chambers are normal.
Nevertheless, the diagnosis of LDSP could be confused with a ‘divided left atrium’ resulting from the association of ‘cor triatriatum sinister’ and a common atrium . Cor triatriatum was described in various unusual settings other than the ‘classical form’ . In the literature, no mention is made of the anatomy of the superior limbic band [9–11].
We believe that, if a divided LA is echocardiographically similar to the anomaly described in this article, LDSP is surgically, anatomically and embryogenetically different. The position of the ASD is crucial in support of the differential diagnosis. When a divided LA is associated with a common atrium, the ASD is ‘in low position’, near the AVJ , and it can communicate with the vestibular component of the RA (ASD is described as ostium primum) [9, 11]; furthermore this ‘membrane’ is free floating and not attached to the AVJ .
In LDSP, the ASD is higher than in the divided LA; it is in a superoposterior position, far from the AVJ and far from the vestibular atrial component. It represents the ancestral ostium secundum just after the confluence of the SP perforations. The crux cordis and the AVJ are normal, and the SP is well ‘anchored’ to the vestibular atrial component (Fig. 6).
In 2002, Robert H. Anderson reviewed the development of the atrial septum; he demonstrated that the infolding process of the septum secundum occurs concomitantly with the incorporation of the pulmonary vein into the body of the primary atrium .
Starting inferiorly at the 21st stage of the Carnegie horizon, this infolding process moves to the atrial roof, resulting in the superior limbic band adjacent to the mouth of the superior caval vein and forming the anterosuperior margin of the foramen ovale . An exhaustive description of the infolding process of the septum secundum in the literature fails to mention the infolding mechanism or ‘purpose’ . This topic would be an interesting point for future investigation.
By the 12th week of development, the superior right-sided pulmonary vein has become a separate tributary of the LA, and this process is concomitant with the infolding of the atrial roof (superior limbic band) .
Anderson clearly demonstrated that these 2 processes occur concomitantly. Therefore, we believe that they are definitely connected.
We supposed, as did Stella Van Praagh in 1995, that the failure of the development of the superior limbic band of the septum secundum derived from the failure of this ‘infolding process’ has to be considered the ‘primum mobile’.
This theory could be investigated preoperatively using echocardiographic scans and was confirmed by surgical inspection in our series. If 2D echocardiography is useful for diagnosis, 3D echocardiograms provide better images. Multiplanar reconstruction can add further information about ASD form and positioning  and demonstrates the absence of the superior limbic band (Figs 3B and 4C). Furthermore, 3D reconstruction gives real anatomical views of intracardiac structures, so it clearly guides operative strategy.
Hiramatsu et al.  proposed in 1998 a surgical septoplasty that avoided the use of any foreign-tissue patches. This attractive technique was not feasible for us because of the severe displacement of the SP and the lack of septal tissue available for interatrial wall reconstruction in our patients.
After resection of the SP, construction of a new, appropriately positioned atrial septum can be achieved with either pericardial or prosthetic material .
In the situation presented in this article, autologous pericardial patch reconstruction was the only reasonable option to guarantee a harmonious atrial septation without any pulmonary vein stenosis, obstruction or distortion or foreign tissue patches. In contrast to the situation with an ostium primum ASD or a common atrium, the conduction system lay in the usual position and was far away from the patch suture line. No AV blocks are described in our series.
Severe LDSP described as an isolated malformation is the topic of our article. Preoperative 3D-TTE is a valuable tool to easily view the intracardiac anatomy of LDSP and to ‘model’ the operative strategy. The feasibility of pericardial patch repair without any complications is demonstrated.
The further development of 3D-TTE technology  could guide some difficult preoperative decisions and facilitate the surgical management of more complex procedures.
This paper is dedicated to the memory of my dear friend Yves Durandy (1947–2016), world renowned paediatric anaesthetist and perfusionist. He spent his life working on technical improvements for paediatric extracorporeal circulation. He pioneered the use of warm-blood cardioplegia during cardiac operations in children. His devotion to humanitarian causes was strong until the end of his life.
Conflict of interest: none declared.
- pericardial sac
- hypoplastic left heart syndrome
- situs inversus
- left atrium
- right atrium
- congenital abnormality
- echocardiography, three-dimensional
- mitral valve
- left ventricle
- constriction, pathologic
- tissue membrane
- preoperative care
- pulmonary veins
- reconstructive surgical procedures
- surgical procedures, operative
- echocardiography, transthoracic
- cardiac complication
- blood outflow
- situs solitus
- septum primum