1. Introduction
Single-ventricle physiology is often taught using a complete-mixing model: systemic venous and pulmonary venous blood enter a common ventricular chamber, mix uniformly, and are then distributed between the pulmonary and systemic circulations. That model is useful, but it is not always anatomically or physiologically correct. In many functionally univentricular hearts, the incoming blood streams retain directional momentum and remain partially separated as they cross the ventricle. The result is streaming—preferential routing of one venous stream toward one arterial outlet.
Unfavorable streaming occurs when relatively desaturated systemic venous blood is preferentially directed toward the aorta while relatively oxygenated pulmonary venous blood is preferentially directed toward the pulmonary artery. Systemic arterial saturation may therefore be substantially lower than predicted from pulmonary blood flow alone. This is particularly relevant in double-inlet left ventricle (DILV) and other single-ventricle morphologies with malposed great arteries, where atrioventricular inflow direction, ventricular looping, the position of the outlet chamber, and the ventricular septal communication together determine the path followed by blood through the heart. [1–3]
The central principle is simple: a common ventricle does not necessarily behave as a well-stirred reservoir. Cyanosis in single-ventricle physiology reflects not only how much blood reaches the lungs, but also which blood preferentially reaches the aorta.
2. Anatomical Substrate for Unfavorable Streaming
2.1 Double-inlet left ventricle as a model
In DILV, both atrioventricular connections enter a dominant ventricle of left ventricular morphology. A smaller morphologically right ventricular outlet chamber is commonly present and communicates with the dominant ventricle through a ventricular septal defect, traditionally termed the bulboventricular foramen. Ventriculoarterial connections are frequently discordant, with the aorta arising from the rudimentary outlet chamber and the pulmonary artery arising from the dominant ventricle. [1,2]
This architecture creates the substrate for preferential flow. Blood entering through the right-sided atrioventricular valve does not immediately mix homogeneously with blood entering through the left-sided valve. Each inflow jet has a specific three-dimensional trajectory. If systemic venous inflow from the right atrium is aligned with the bulboventricular foramen and the pathway to the aorta, deoxygenated blood may be preferentially ejected systemically. At the same time, pulmonary venous blood entering from the left atrium may be directed more efficiently toward the pulmonary outlet.
The DILV, D-loop, d-transposition configuration is a useful teaching example of this unfavorable relationship. The important concept, however, is alignment rather than simple proximity. An aorta that appears “close” to the right atrial inflow on a schematic diagram does not by itself determine streaming. The relevant variables are the orientation of the atrioventricular valve orifices, direction of the inflow jets, ventricular geometry, location and size of the bulboventricular foramen, position of the outlet chamber, and spatial relationship of both semilunar valves.
2.2 Three-dimensional inflow–outflow relationships
Morphological work supports the importance of these three-dimensional relationships. Meyer and colleagues examined 39 postmortem DILV specimens without cavopulmonary connection. Sixteen hearts had a relatively parallel relationship between inflow and outflow tracts, a configuration considered capable of maintaining separation between blood streams, whereas 10 had a more perpendicular relationship that would be expected to promote mixing. The study also demonstrated substantial variation in great-artery relationships and in the geometry of the ventricular septal communication. [3]
These findings reinforce a practical point: ventricular looping or the label d-TGA versus l-TGA cannot independently predict streaming. Historical catheterization studies similarly showed that the position of the outlet chamber and the detailed ventricular geometry were important determinants of systemic arterial oxygenation. [4–6]
3. Hemodynamic Consequences
3.1 Complete mixing, favorable streaming, and unfavorable streaming
Three conceptual patterns can occur:
- Complete mixing: pulmonary and systemic venous blood become nearly homogeneous before reaching either arterial outlet.
- Favorable streaming: pulmonary venous blood is preferentially directed toward the aorta, improving systemic arterial saturation.
- Unfavorable streaming: systemic venous blood is preferentially directed toward the aorta, worsening systemic arterial desaturation.
In a true complete-mixing state, oxygen saturation in the aorta and pulmonary artery should be similar before any additional downstream shunt or source of flow. With unfavorable streaming, aortic saturation may be lower than pulmonary arterial saturation, because the pulmonary artery receives a disproportionate amount of oxygenated pulmonary venous return.
Classic invasive studies established that complete mixing is far from universal. Rahimtoola and colleagues demonstrated preferential streaming in patients with common or single ventricles. In the later series by Macartney and colleagues, among patients with detailed hemodynamic assessment and without atrioventricular valve atresia, complete mixing was observed in 36%, favorable streaming in 52%, and unfavorable streaming in 12%. These percentages reflect selected historical cohorts and should not be interpreted as contemporary prevalence estimates, but they provide strong physiological evidence that arterial oxygenation cannot be understood by a complete-mixing model alone. [4,5]
Mocellin and Sauer similarly found that systemic arterial saturation in univentricular hearts was influenced by outlet-chamber position and streaming characteristics, particularly in patients with transposed great arteries. [6]
3.2 Why cyanosis may appear disproportionate to pulmonary blood flow
The most clinically important consequence of unfavorable streaming is cyanosis that is more severe than expected from the amount of pulmonary blood flow. A patient may have adequate—or even excessive—Qp yet remain markedly desaturated because oxygenated pulmonary venous blood is preferentially recirculated to the pulmonary artery rather than delivered to the systemic circulation.
This distinction is critical because the therapeutic implications are opposite. If cyanosis is caused by inadequate pulmonary blood flow, increasing Qp may be appropriate. If pulmonary blood flow is already adequate and unfavorable streaming is the principal problem, simply increasing Qp may fail to improve systemic saturation and may worsen pulmonary overcirculation, ventricular volume load, and systemic perfusion.
4. Interaction With Qp/Qs and Oxygen Delivery
For complete-mixing physiology, pulmonary-to-systemic flow ratio can be estimated by the Fick relationship:
Qp/Qs = (SaO₂ − SvO₂) / (SpvO₂ − SpaO₂)
where SaO₂ is systemic arterial saturation, SvO₂ is mixed systemic venous saturation, SpvO₂ is pulmonary venous saturation, and SpaO₂ is pulmonary arterial saturation.
In a completely mixed single-ventricle circulation, SaO₂ and SpaO₂ may be similar enough that simplified bedside assumptions are sometimes used. Those assumptions become unreliable when significant streaming is present. The aorta and pulmonary artery are sampling different mixtures, so systemic arterial saturation cannot be used as a surrogate for the oxygen content entering the pulmonary artery.
The same caution applies to systemic oxygen delivery. Oxygen delivery is determined by:
DO₂ = Qs × CaO₂
A higher arterial saturation does not necessarily mean better oxygen delivery if systemic blood flow falls. Conversely, severe arterial desaturation caused by unfavorable streaming may reduce arterial oxygen content despite preserved Qs. Modeling studies of parallel single-ventricle circulation have repeatedly shown that saturation, Qp/Qs, systemic blood flow, and oxygen delivery are related but not interchangeable physiological variables. [7,8]
Therefore, the clinical target should not be the highest possible pulse-oximetry value. The goal is adequate systemic oxygen delivery, assessed through systemic perfusion, venous saturation when available, lactate, urine output, ventricular function, and the balance between Qp and Qs.
5. Clinical Recognition
Unfavorable streaming should be considered when systemic arterial saturation is unexpectedly low despite apparently adequate pulmonary blood flow. Several situations may make the phenomenon more apparent:
- after pulmonary artery banding, when the velocity and direction of ventricular outflow change;
- after ductal closure or ligation, when an additional source of mixing or pulmonary blood flow is removed;
- after changes in pulmonary or systemic vascular resistance;
- when an atrial communication becomes restrictive;
- when ventricular loading conditions or contractility alter the geometry and momentum of inflow jets.
A poor response of systemic saturation to supplemental oxygen does not prove unfavorable streaming, but when lung disease, pulmonary venous desaturation, and critically reduced pulmonary blood flow have been excluded, it should raise concern that the oxygenated pulmonary venous stream is not reaching the aorta effectively.
Importantly, unfavorable streaming must also be distinguished from systemic outflow tract obstruction. In DILV with transposed great arteries, systemic output may traverse the bulboventricular foramen before reaching the aorta. Progressive restriction can produce subaortic obstruction, increased ventricular pressure, low systemic output, and deterioration independent of the streaming pattern. Long-term series have shown that systemic outflow obstruction is an important determinant of outcome in DILV and related morphologies. [9]
6. Diagnostic Assessment
6.1 Echocardiography
Echocardiography remains the first-line modality. The objective is not merely to identify a “single ventricle,” but to reconstruct the complete flow pathway using sequential segmental analysis. The study should define:
- atrial situs and systemic and pulmonary venous connections;
- atrioventricular connections and valve morphology;
- ventricular morphology and looping;
- the size, position, and direction of the bulboventricular foramen or VSD;
- the location of the outlet chamber;
- ventriculoarterial connections and great-artery relationships;
- pulmonary and systemic outflow obstruction;
- aortic arch anatomy;
- atrial communication and its restrictiveness.
Color Doppler can demonstrate the direction of inflow and ejection jets, but it does not directly measure their oxygen content. When streaming is strongly suspected, contrast echocardiography from a systemic venous injection can sometimes provide direct evidence of preferential routing.
Miura and colleagues reported a neonate with a single left ventricle who developed severe desaturation after ductal ligation and pulmonary artery banding. Saturation fell to 68% despite 100% inspired oxygen. Contrast echocardiography demonstrated systemic venous blood streaming predominantly toward the aorta. Balloon atrial septostomy improved mixing, with saturation increasing to 85%. This case illustrates the diagnostic and therapeutic relevance of streaming, but it does not establish atrial septostomy as a routine treatment for all patients with unfavorable streaming. [10]
6.2 Catheterization and advanced imaging
When the mechanism remains uncertain, catheterization can be highly informative. Simultaneous or carefully timed samples from the pulmonary artery and aorta may demonstrate a meaningful saturation difference. Pressures should be measured across the bulboventricular foramen, systemic outflow, pulmonary outflow, and atrial level as appropriate. Interpretation must account for ductal flow, systemic-to-pulmonary collaterals, pulmonary venous desaturation, and sampling location.
Four-dimensional flow cardiac MRI offers a direct method to visualize complex intracardiac flow trajectories. Wong and colleagues used 4D flow MRI in a patient with TGA, a large VSD, and severe cyanosis to demonstrate adequate pulmonary blood flow but marked unfavorable streaming with poor intracardiac mixing. The imaging findings changed management away from an intervention designed simply to increase pulmonary flow and toward improving mixing. This remains specialized technology, but it illustrates how patient-specific flow visualization can separate flow quantity from flow distribution. [11]
7. Management Implications
Treatment should address the demonstrated physiological problem rather than the saturation number alone.
If pulmonary blood flow is inadequate, ductal support, catheter intervention, or a surgical source of pulmonary blood flow may be necessary. If pulmonary overcirculation is present, pulmonary artery banding or another strategy to limit Qp may be appropriate. However, either intervention can change intracardiac jet geometry and may unmask or worsen unfavorable streaming.
When unfavorable streaming is the dominant cause of cyanosis, improving mixing may be more effective than increasing pulmonary flow. Enlargement of a restrictive atrial communication by balloon septostomy or surgical septectomy has been used in selected patients, particularly when atrial-level separation contributes to persistent stream segregation. Evidence for this strategy consists largely of case reports and individualized physiological reasoning rather than comparative trials. [10,11]
In DILV with a systemic outlet dependent on the bulboventricular foramen, the entire systemic outflow pathway must also be assessed. A favorable or unfavorable streaming pattern does not eliminate the need to address clinically significant subaortic or arch obstruction. Depending on anatomy and stage of palliation, systemic outflow reconstruction may require a Damus–Kaye–Stansel connection or a Norwood-type pathway rather than isolated manipulation of pulmonary blood flow. [9]
8. Key Clinical Principles
- A common ventricular chamber does not guarantee complete mixing.
- Unfavorable streaming preferentially directs systemic venous blood toward the aorta, producing more severe cyanosis than expected from pulmonary blood flow alone.
- The relevant determinant is three-dimensional inflow–outflow alignment, not simply ventricular looping or the label d-TGA versus l-TGA.
- Aortic saturation may be lower than pulmonary arterial saturation when unfavorable streaming is substantial.
- SaO₂ alone cannot define Qp/Qs or systemic oxygen delivery. Streaming can invalidate simplified complete-mixing assumptions.
- Unexpected cyanosis requires a mechanism-based differential diagnosis: inadequate Qp, unfavorable streaming, pulmonary venous desaturation, atrial restriction, low cardiac output, and systemic outflow obstruction may coexist.
- Do not reflexively increase pulmonary blood flow solely to raise saturation. If Qp is already adequate, this may worsen pulmonary overcirculation without correcting systemic desaturation.
- Management must be individualized to anatomy and physiology. In selected patients, improving mixing rather than increasing Qp may be the appropriate intervention.
References
[1] Rao PS. Double-inlet left ventricle. Children (Basel). 2022;9:1274. DOI: 10.3390/children9091274. PMID: 36138583.
[2] Frescura C, Thiene G. The new concept of univentricular heart. Front Pediatr. 2014;2:62. DOI: 10.3389/fped.2014.00062. PMID: 25072035.
[3] Meyer SL, Jongbloed MR, Ho SY, Bartelings MM, McCarthy KP, Uemura H, Ebels T. Intracardiac anatomical relationships and potential for streaming in double inlet left ventricles. PLoS One. 2017;12:e0188048. DOI: 10.1371/journal.pone.0188048. PMID: 29190641.
[4] Rahimtoola SH, Ongley PA, Swan HJC. The hemodynamics of common (or single) ventricle. Circulation. 1966;34:14–23. DOI: 10.1161/01.CIR.34.1.14. PMID: 5940238.
[5] Macartney FJ, Partridge JB, Scott O, Deverall PB. Common or single ventricle: An angiocardiographic and hemodynamic study of 42 patients. Circulation. 1976;53:543–554. DOI: 10.1161/01.CIR.53.3.543. PMID: 1248087.
[6] Mocellin R, Sauer U. Haemodynamics studies in patients with univentricular hearts. Herz. 1979;4:242–247. PMID: 447189.
[7] Barnea O, Santamore WP, Rossi A, Salloum E, Chien S, Austin EH. Estimation of oxygen delivery in newborns with a univentricular circulation. Circulation. 1998;98:1407–1413. DOI: 10.1161/01.CIR.98.14.1407. PMID: 9760295.
[8] Francis DP, Willson K, Thorne SA, Davies LC, Coats AJ. Oxygenation in patients with a functionally univentricular circulation and complete mixing of blood: Are saturation and flow interchangeable? Circulation. 1999;100:2198–2203. DOI: 10.1161/01.CIR.100.21.2198. PMID: 10571980.
[9] Franken LC, Admiraal M, Verrall CE, et al. Improved long-term outcomes in double-inlet left ventricle and tricuspid atresia with transposed great arteries: systemic outflow tract obstruction present at birth defines long-term outcome. Eur J Cardiothorac Surg. 2017;51:1051–1057. DOI: 10.1093/ejcts/ezx022. PMID: 28329058.
[10] Miura S, Ookoshi Y, Momoki K, Hoshino K, Ueta I. Contrast echocardiogram for the evaluation of streaming in a single-ventricle patient. Pediatr Cardiol. 2019;40:216–218. DOI: 10.1007/s00246-018-1997-2. PMID: 30306199.
[11] Wong J, Pushparajah K, Hussain T, Giese D, Dedieu N, Mathur S, Greil GF, Razavi R, Bell A. Streaming in transposition of the great arteries by using cardiac magnetic resonance imaging. Circulation. 2014;129:1169–1170. DOI: 10.1161/CIRCULATIONAHA.113.002852. PMID: 24615966.