As my good friend Bill Devine wrote several years ago, in a previous editorial published in Cardiology in the Young,¹ “What if Ivemark had suggested the term ‘Syndrome of Visceral Symmetry with Asplenia’ instead of ‘Asplenia, a Teratologic Syndrome of Visceral Heterotaxy’”? Bill was pointing out the fact that, in many instances, it was less than helpful to describe lesions within the heart on the basis of the state of the spleen, since presumed splenic structure was not always consonant with the observed cardiac malformations. At the time, the very concept of isomerism within the heart was controversial, with the Van Praagh's having argued forcefully that the concept of symmetry within the atrial chambers from patients with heterotaxy syndrome was “an erroneous concept”.² Since then, the evidence substantiating the concept of isomerism within the heart has continued to accrue. Nowadays, when based on the criterion of the extent of the pectinate muscles relative to the atrioventricular junctions,³ it is difficult to ignore the obvious presence of isomeric right or left atrial appendages in the two subsets of the syndrome that continues to be described as “visceral heterotaxy” (Figs 1 and 2). The evidence accruing from experiments using molecular biological techniques is now overwhelming, showing that, within the heart, it is the atrial appendages that are most consistently under the control of the genes producing morphological leftness as opposed to rightness. Thus, in hearts from mice in which the genes responsible for producing leftness, such as pitx2,⁴ or cited2,⁵ have been knocked out, there are appendages of morphologically right type on both sides of the common atrial chamber, along with marked symmetry in the formation of the venous tributaries and so-called venous valves. It is also now known that, during early development, other genes produce a midline barrier, preventing the genes responsible for leftness from moving into the right side of the developing embryonic disc. When these genes are knocked out, such as sonic hedgehog, then the embryos develop lungs on both sides of the heart that resemble the normal morphologically left lung,⁶ and there is evidence that both of the appendages are also of left morphology (Deborah Henderson – personal communication). When judged on the basis of the morphology of the appendages, therefore, the evidence supporting the existence of isomerism is incontrovertible. This is not to imply, of course, that the patient with isomerism of the eft atrial appendages will possess eight pulmonary veins! Such a suggestion reduces the debate to the absurd. It is a recognised fact, nonetheless, that patients with isomerism of the right atrial appendages never have normal pulmonary venoatrial connections, even when the pulmonary veins drain directly to the heart rather than to an extracardiac site.³ Furthermore, some degree of symmetry of the systemic venous tributaries is found in patients with both right and left isomerism. The precise types of venoatrial connection, however, are variable, and need to be described separately so as to provide a full account of the cardiac make-up of the patient with visceral heterotaxy. As Devine commented in 1992,¹ “by identifying the arrangement of the atrial appendages as the first step in sequential segmental analysis of hearts, even the most complex hearts, such as those found in conjoined twins, can easily be described and diagnosed”.

Figure 1. This heart, from a patient with visceral heterotaxy, has the pectinate muscles extending to the crux around both atrioventricular junctions. This is unequivocal evidence of presence of morphologically right atrial appendages on both sides of the common atrial chamber.

Figure 2. In this heart, from another patient with visceral heterotaxy, the vestibules of the atrioventricular junctions are smooth on both sides towards the crux, indicating the presence of atrial appendages bilaterally of left morphology.

All of this is pertinent to the excellent review made by Nagel et al.⁷ of a series of patients diagnosed in Birmingham Children's Hospital with visceral heterotaxy, and admitted for surgical treatment of their cardiac malformations. In this respect, it should be noted that “heterotaxy” remains a popular term for description of these children, and is likely to be continued to be used. When used with precision, nonetheless, the word should be applied to all patients in whom the arrangement is other than normal. If used in etymylogically correct fashion, therefore, “heterotaxy” should include those with mirror-imagery, or so-called “situs inversus”. Indeed, in the past it was frequent to find such patients described as having “partial situs inversus”. Nowadays, it has become conventional to use the term “heterotaxy” only to include the groups which many others still describe in terms of “asplenia” and “polysplenia”. As Nagel et al show in their review, while “heterotaxy” still remains a useful term for grouping together these patients, it is now much less than satisfactory to divide them on the presumed state of the spleen. This is because some patients with the findings anticipated for the syndrome of “asplenia” can possess a spleen, and solitary spleens can also be found in the patients who would be expected to have multiple spleens. Indeed, again as pointed out by Nagel et al in some instances the spleen can be absent when the cardiac features of the patients suggest that multiple spleens should be present. Even more significantly, Nagel et al.⁷ have shown that, if the presence of Howell–Jolly bodies is truly indicative of functional absence of splenic tissue, then such functional asplenia can develop in patients with visceral heterotaxy independent of the anatomical state of the spleen.

The key to rational analysis of all these potentially paradoxical findings is found in the crucially important philosophical principle introduced by Van Praagh et al. in 1980.⁸ When commenting on the illogical desire of some of my colleagues and myself, at that time, to define ventricles on the basis of their valve of entry, Van Praagh et al.⁸ argued that it was less than scientifically sound to define one variable feature on the basis of another variable. The logic underscoring this principle is indisputable. It means, for example, that atrial chambers should not be defined on the basis of venoatrial connections, since the venoatrial connections are themselves variable. In the setting of visceral heterotaxy, it means that cardiac structures should not be defined on the basis of splenic anatomy. This is not to downplay the importance of splenic anatomy, which needs to be described in its own right. But as Nagel et al. now show,⁷ just because there is splenic tissue present, we should not assume that it is working. In the setting of visceral heterotaxy, however, entirely different subsets of patients are created depending on whether they are stratified according to the state of the spleen or the structure of the atrial appendages.⁹ As Devine¹ concluded in 1992 “surely the cardiac defects are best characterised on the basis of the heart, using as the major criterion the morphology of the isomeric appendages”.