Skip to main content
The Texas Heart Institute Journal logoLink to The Texas Heart Institute Journal
. 2003;30(3):208–213.

Coronary Arteries in Crisscross Heart

Paolo Angelini 1, Alberto Lopez 1, Roberto Lufschanowski 1, Margit A Nemeth 1, Scott D Flamm 1
PMCID: PMC197319  PMID: 12959204

Abstract

Crisscross heart, or superoinferior ventricles, is a complex and often confusing congenital anomaly. We report a heretofore unreported presentation of “isolated” crisscross heart, in situs inversus levocardia, which allows us to more clearly define the typical features of crisscross ventricles. The case of this 41-year-old woman, who had a peculiar coronary anatomy, underscores the concept that coronary artery anatomy is strictly related to the myocardial mass served. In complex congenital heart defects, development of an anterior descending artery is possible (as a primary artery, along with the circumflex and right coronary arteries) only if the ventricular septum develops properly and is aligned with the semilunar valves. We use the present case of crisscross heart to illustrate the spectrum of anomalies that can occur during formation of the cardiac apex; this spectrum ranges from a normal apex, to a diverted apex (as in dextroversion in situs solitus), to a crisscross anomaly. (Tex Heart Inst J 2003;30:208–13)

Key words: Coronary angiography; coronary vessel anomalies; crisscross heart; heart defects, congenital/classification/diagnosis; heart/embryology; magnetic resonance imaging; superoinferior ventricles

Crisscross heart (CCH), also called superoinferior ventricles or upstairs-downstairs heart, is a complex, hard-to-decipher congenital anomaly that has been characterized as the “Rosetta stone” of congenital heart disease. 1 Although many investigators have attempted to define CCH, these definitions have been vague. For instance, the best recognized students of such pathology, Anderson and colleagues, stated that “the atrioventricular connection may appear to cross.” 2 Elsewhere, these same authors wrote that “the essence is a rotational abnormality in the ventricular mass, caused by the postseptational rotation of the ventricular mass around its long axis.” 3 Later, they asserted that “the term crisscross itself describes a particular relationship of the ventricular inlets.” 4 These and similar statements have been challenged by other authors, who have concluded that the “crisscrossed arrangement is an illusion.” 5 Clearly, the essence of CCH is not some kind of defective atrioventricular or ventriculoarterial connection.

The nature of any defect can best be understood if that defect occurs in isolation. Cases of CCH commonly feature multiple additional defects, 6 the presence of which tends to preclude a biventricular repair. Crisscross heart almost never occurs as an isolated presentation: 1 in fact, a review of the world medical literature does not reveal any such cases. We describe an exceptional case of levocardia with CCH in an otherwise “normal” heart, and we use magnetic resonance images (MRI) to illustrate the basic anatomic defect. We also discuss the implications of the peculiar coronary angiographic findings in this patient.

Case Report

A 41-year-old woman, the mother of 1 child, presented at our institution because of the recent onset of recurrent dizzy spells and rapid cardiac palpitations, which culminated in an episode of syncope from which she recovered spontaneously within a few minutes. The patient had a prior history of good health and no cardiovascular symptoms. On her admission to our hospital, a chest radiograph and cardiac MRI (Fig. 1) showed situs inversus of the liver and lungs with levocardia—the apex of the heart was situated in the left side of the chest. On the electrocardiogram, the unusual P-wave axis was consistent with situs inversus: an abnormal septal vector (deep Q waves in the inferior leads and in lead V6) suggested ventricular septal malorientation consistent with a superiorly located right ventricle (Fig. 2). Electrocardiographically, the left ventricle and the apex were left-sided, as deduced from the presence of the highest precordial potentials in lead V6. A Holter monitor revealed intermittent supraventricular tachycardia (133 beats/min).

graphic file with name 9FF1.jpg

Fig. 1 Magnetic resonance image in a frontal cross-section shows the main features of our case: liver on the left, spleen on the right, right aortic arch, cardiac apex on the left. In this tomographic section, the left ventricle (LV) is aligned with the frontal plane, and the ventricular septum is located superiorly to the left ventricle (LV).

graphic file with name 9FF2.jpg

Fig. 2 Electrocardiogram (see text).

The patient underwent heart catheterization, during which all the cardiac chambers were entered, and situs inversus with a crisscross deformity of the apex in levoversion was revealed. The catheter passed from the left-sided, morphologic right atrium (m-RA), to the right-sided, morphologic left atrium (m-LA) via a patent foramen ovale. All the pressures and saturations were normal, and no shunts were present. Selective coronary angiography showed 2 nearly symmetrical coronary arteries (Fig. 3) that ran along the right and left borders of the heart, respectively; each artery provided 2 vertically oriented branches that crossed both ventricles. The anterior bifurcation of each coronary artery furnished a horizontal branch, which provided multiple penetrating branches to a horizontally oriented septum, as seen during the myocardial blush phase of angiography. The venous coronary system featured 1 large vein that drained into the m-LA and 2 other large veins that drained into the m-RA. No coronary sinus was seen.

graphic file with name 9FF3.jpg

Fig. 3 Selective coronary angiograms (anteroposterior projection) of the right-sided (A) and left-sided (B) coronary arteries, both of which provide smaller transverse branches that lead to perforating septal vessels.

LAD = left anterior descending coronary artery

Electrophysiologic testing confirmed the presence of 2 atrioventricular nodal reentry circuits. Cardiac MRI clearly showed the unusual and complex misalignment of the ventricles, which had the typical features of crisscross inlets and superoinferior ventricles (Figs. 4 and 5).

graphic file with name 9FF4.jpg

Fig. 4 A) Near-horizontal magnetic resonance image (MRI) cross-section shows the right atrium (RA), right ventricle (RV), and main pulmonary artery (MPA) encircling the aortic valve. B) Oblique sagittal MRI cross-section shows the left ventricular (LV) mass (mid-section), encircled by the right atrium (RA) and the right ventricle (RV). The RV apex is seen as a thin apical recess that nearly reaches the diaphragmatic surface of the heart.

Ant = anterior; L = left; Post = posterior; R = right

graphic file with name 9FF5.jpg

Fig. 5 Cardiac magnetic resonance images, 3-dimensional reconstruction. The right ventricle and pulmonary artery are white, and the left atrium, left ventricle, and aorta are gray. A) Frontal, anterior view, and B) left lateral view.

A = anterior; I = inferior, L = lateral; LV = left ventricular; P = posterior; R = right; RV = right ventricular; S = superior

Using a segmental approach, we made the final diagnosis of situs inversus of the atria, atrioventricular concordant connections, crisscross deformity of the ventricles in L-loop-like morphology, and normally crossed great vessels in L-loop-like morphology of the aortopulmonary trunks (I,L,L), in levocardia.

The patient underwent successful ablation of the atrioventricular reentry circuits.

Discussion

Our case offers a unique opportunity to reexamine the essential features of crisscross (superoinferior) ventricles, although multiple variations can, of course, occur. This case, especially with the use of cardiac MRI, clearly illustrates the typical, defining features of crisscross ventricles:

  • The inlets of the 2 ventricles are elongated and cross each other, with the anatomic right ventricular inlet situated superiorly. In our patient, the tricuspid valve was situated to the left, and the right ventricular inlet crossed toward the right.

  • The apical portions of the ventricles are situated opposite their respective inlet valves, that is, opposite the expected locations. In our patient, CCH was complicated by the coexistence of levocardia (an apical malformation that is independent of the crisscross defect).

  • The ventricular outlets have normal intrinsic anatomic features and connections; the m-RV outlet connects with an anteriorly located outflow tract, which crosses the m-LV outlet superiorly. In our patient, the left ventricular outflow tract seemed elongated but maintained normal fibrous continuity with the mitral valve. In this “uncomplicated” case of CCH, the right ventricular cavity was normal in size but greatly displaced in location. With respect to a normal situs inversus heart in levocardia, this patient's heart had undergone a clockwise rotation along the ventricular long axis, as seen from the apex (Fig. 6). This rotation occurred at an unknown stage of embryologic development and by means of an unknown morphogenetic mechanism. However, the distortion must have occurred at the same time that the heart, having developed in situs inversus, underwent an abnormal leftward apical rotation (rather than the expected rightward rotation), thus evolving into situs inversus levocardia.

graphic file with name 9FF6.jpg

Fig. 6 Schematic drawing of “all possible” apical morphologies: situs solitus, D-loop, and normally crossed great vessels. A) Normal; B) normal with crisscross-heart ventricles (CCH); C) normal with dextroversion (apex to the right); and D) normal with dextroversion and CCH.

Detailed comments about the coronary anatomy in CCH have previously been made only by Van Praagh and associates. 7 They reviewed the coronary arteries in 10 cases of superoinferior ventricles, two of which involved CCH but none of which entailed levocardia and situs inversus. Those authors concluded that the coronary patterns were “always compatible with the type of ventricular loop present.” Additionally, they mentioned that “the coronary artery that emerges between or close to the semilunar valve is the anterior descending artery.”

As implied by the observation of widely variable coronary anatomy in normal (as well as structurally defective) hearts, coronary arteries have the essential function of nourishing the ventricular myocardium, and their anatomy varies in relation to the structure of the dependent ventricular masses. 8 Specifically, the location and extent of development of the ventricular septum influences the location and size of the anterior and posterior descending arteries. Accordingly, in the extreme case of agenesis of the ventricular septum (common ventricle), the left anterior descending artery (LAD) is absent; in tricuspid atresia, that artery is grossly displaced to the right. 8 A major reason for our interest in the coronary morphology of CCH is that the ventricular septum is oriented quite abnormally in this anomaly—in a horizontal plane rather than a near-vertical one, as in a normal heart. In our patient, coronary angiography suggested that the LAD did not develop as the usual anterior bifurcation of the left main trunk; rather, the ventricular septum received perforating branches from 2 horizontally oriented secondary branches that were situated on the anterior surface of the heart and that originated from each of the 2 main, opposite coronary arteries. The latter arteries were grossly of the same size and branching morphology; they did not present any intrinsic feature of the right or circumflex coronary arteries proper, since the right and left coronary arteries ran on the surface of both superoinferiorly oriented ventricles and did not encircle the atrioventricular grooves.

The LAD has been defined as “an artery that runs over the anterior interventricular groove, while providing septal perforating branches.” 9 It is not essential for the LAD either to originate from a specific sinus of Valsalva or to have a common trunk with the circumflex artery. In our patient, the 2 branches that ran horizontally from each margin of the heart, along the interventricular groove, were considered to constitute the LAD.

The principle that the coronary arteries should be named on the basis of their dependent myocardial territory 8 is confirmed by the arrangement of the coronary arteries in our case of CCH. If other groups confirm these findings, it might be possible to use coronary angiography to diagnose CCH by the presence of epicardial arteries that cross each ventricular mass and provide horizontally oriented branches to the ventricular septum. We found it particularly helpful to visualize the horizontally displaced ventricular septum during the myocardial blush phase of angiography. We were not able to identify a posterior descending artery.

Normal embryologic formation of the definitive cardiac apex is not necessarily determined by the orientation of the bulboventricular loop. Indeed, in a normal embryo, the bulboventricular loop is turned to the right, but the cardiac apex eventually develops to the left. 10 During early embryonic development, it is only when the left ventricular mass grows selectively, quickly, and predominantly, when ventricular septation is completed, and when compactization of the ventricular masses occurs, that the apex becomes defined (normally toward the left). Interestingly, the coronary pattern is established at the time of ventricular myocardial compactization, 11 not later than at that stage. In our particular case of CCH, organization of the ventricular masses must have been completed by the time the coronary arterial pattern became defined; this is confirmed by the unusual coronary tree, in which the anterior descending artery system is directed at right angles with respect to the main coronary arteries. This finding contradicts the frequently cited but unproven claim that CCH is the product of a postlooping defect. 1,3 Most likely, an abnormal bulboventricular loop causes the essential defect in this complex anomaly. In view of the features of our case, the hypothesis of Van Praagh and co-authors, 5 that superoinferior ventricles are “caused by” underdevelopment of the right ventricular inlet, appears less sustainable. In our patient, the tricuspid valve annulus had a normal diameter (35 mm) compared with the mitral annulus, and the right ventricular inlet was elongated but not stenotic (Figs. 4A and 4B). This uncomplicated case suggests that the essential morphogenetic mechanism of CCH consists of abnormal formation of the apex. Indeed, deformed bulboventricular loops (featuring an extra bend that could lead to CCH) have been observed in experimental embryologic studies, especially those studies using genetic mutations in experimental animals. 12

As de la Cruz illustrated so well in describing the morphogenesis of the living heart, 13 first the situs of the atria is established (cardiogenic areas, prelooping stage), then the laterality of the atrioventricular valves and connections is irreversibly established by bulbo-ventricular loop formation, and finally the definitive formation of the apex occurs. At this stage, the apex could be situated “normally” (for situs solitus, to the left; for situs inversus, to the right) or “discordant” (dextrocardia in situs solitus; levocardia in situs inversus). Additionally, the apex could be “normally twisted” (opposite the direction of the bulboventricular loop) or “abnormally twisted,” as in CCH or supero-inferior ventricles, due to an abnormal rotation of the apical portions of the ventricles along the longitudinal cardiac axis (Figs. 6 and 7). Normal formation of the apex from the C-loop stage (when the prospective ventricular apical portions are situated initially in a superoinferior pattern and then eventually side by side) involves a longitudinal twisting, which causes the right ventricular apex to be anterior and to the left of the left ventricular apex during S-loop formation. 13 Normally, the apical portions of the 2 ventricles are situated on the same side of the inlets. In CCH, the apical portions of the 2 ventricles become separated and crossed with respect to the inlets (Figs. 6 and 7).

graphic file with name 9FF7.jpg

Fig. 7 Schematic drawing of “all possible” apical morphologies: situs inversus, L-loop, and normally crossed great vessels. A) Mirror-image dextrocardia (normal, in situs inversus); B) mirror-image dextrocardia with isolated crisscross-heart ventricles (CCH); C) situs inversus levocardia with L-loop; and D) situs inversus levocardia, with L-loop and CCH (as in our patient).

In conclusion, CCH is a complex congenital abnormality of the cardiac anatomy that features a unique morphology. This heretofore unreported presentation of CCH in situs inversus levocardia, imaged with cardiac MRI and coronary angiography, provides valuable insight into the complex nature of this anomaly. It is apparent that CCH essentially consists of twisting of the apical portions of the ventricles along the long axis of the heart, while the great vessels and the atrioventricular valves remain substantially in the positions expected for the situs and the bulboventricular loop.

Footnotes

Address for reprints: Paolo Angelini, MD, 6624 Fannin Street, Suite 2780, Houston, TX 77030

E-mail: pangelinimd@houston.rr.com

References

  • 1.Freedom RM. Supero-inferior ventricles, criss-cross atrio-ventricular connections and the straddling atrioventricular valve. In: Freedom RM, Benson LN, Smallhorn JF, editors. Neonatal heart disease. London: Springer-Verlag; 1992. p. 667–78.
  • 2.Anderson RH, Shinebourne EA, Gerlis LM. Criss-cross atrioventricular relationships producing paradoxical atrioventricular concordance or discordance. Their significance to nomenclature of congenital heart disease. Circulation 1974;50:176–80. [DOI] [PubMed]
  • 3.Anderson RH. Cross-cross hearts revisited. Pediatr Cardiol 1982;3:305–13. [DOI] [PubMed]
  • 4.Anderson RH. Abnormal positions and relationships of the heart. In: Anderson RH, Baker EJ, Macartney FJ, Rigby ML, Shinebourne EA, Tynan M, editors. Pediatric cardiology. 2nd ed. Vol 2. London: Churchill Livingstone; 2002. p. 1577–93.
  • 5.Van Praagh R, Weinberg PM, Matsuoka R, Van Praagh S. Malpositions of the heart. In: Adams FH, Emmanouilides GC, editors. Moss' heart disease in infants, children, and adolescents. Baltimore: Williams & Wilkins; 1983. p. 422–58.
  • 6.Marino B, Guccione P, Carotti A. Malposition of the heart. In: Moller JH, Hoffmann JI, editors. Pediatric cardiovascular medicine. New York: Churchill Livingstone; 2000. p. 621–7.
  • 7.Van Praagh S, LaCorte M, Fellows KE, Bossina K, Bush HJ, Keck EW, et al. Superioinferior ventricles: anatomic and angiographic findings in ten postmortem cases. In: Van Praagh R, Takao A, editors. Etiology and morphogenesis of congenital heart disease. Mount Kisco (NY): Futura Publishing Co.; 1980. p. 317–78.
  • 8.Angelini P, de la Cruz MV. The relationship between the coronary arteries and congenital heart disease. In: Angelini P, editor. Coronary artery anomalies: a comprehensive approach. Philadelphia: Lippincott Williams & Wilkins; 1999. p. 173–89.
  • 9.Angelini P. Coronary artery anomalies—current clinical issues: definitions, classification, incidence, clinical relevance, and treatment guidelines. Tex Heart Inst J 2002;29:271–8. [PMC free article] [PubMed]
  • 10.Angelini P, Lopez-Velarde P, Leachman RD. Ventriculoarterial discordance revisited. Tex Heart Inst J 1988;15:174–82. [PMC free article] [PubMed]
  • 11.de la Cruz MV, Moreno-Rodriguez R, Angelini P. Ontogeny of the coronary vessels. In: Angelini P, editor. Coronary artery anomalies: a comprehensive approach. Philadelphia: Lippincott Williams & Wilkins; 1999. p. 11–6.
  • 12.Layton WM, Manasek FJ. Cardiac looping in early IV/IV mouse embryos. In: Van Praagh R, Takao A, editors. Etiology and morphogenesis of congenital heart disease. Mount Kisco (NY): Futura Publishing Co.; 1980. p. 109–26.
  • 13.De la Cruz MV, Sanchez-Gomez C. Torsion and looping of the cardiac tube and primitive cardiac segments. Anatomical manifestations. In: De la Cruz MV, Markwald RR, editors. Living morphogenesis of the heart. Boston: Birkhauser; 1998. p. 99–120.

Articles from Texas Heart Institute Journal are provided here courtesy of Texas Heart Institute

RESOURCES