Discussion

Coronary artery development is a poorly understood phenomenon with tremendous clinical implications. In the third/fourth week of gestation, as the myocardium thickens, a vascular plexus without connection to the aorta forms and develops into an immature coronary artery.^{Reference Angelini15} The exact mechanism of establishing connection of the immature coronary artery to the aorta is unknown, but thought to be related to “in-growth” of the immature coronary artery with signalling from neural crest cells, vascular endothelial growth factor, and chemokine CXCL12.^{Reference de Oliveira Silva-Junior, da Silva Miranda and Mandarim-de-Lacerda16,Reference Ramai, Lai, Monzidelis and Reddy17} Disruptions in those pathways, whether genetic or environmental, likely contribute to the formation of coronary artery anomalies and anomalous origin of a coronary artery from the pulmonary artery.^{Reference Pérez-Pomares, de la Pompa and Franco18} Review of associated lesions seen in TCAPA may shed some light into the pathways associated with its development. Although our sample size was low, bicuspid aortic valve was seen in 8.8% of patients in this cohort, which differs from the 0.5–2% incidence found in the general population.^{Reference Tutar, Ekici, Atalay and Nacar19} Associations between bicuspid aortic valve and anomalous origin of a coronary artery from the pulmonary artery have been described in animal studies as well as in ACxAPA/ALADAPA and again may hint towards a common embryologic pathway.^{Reference Fernández, Durán and Real20} One interesting case identified in this review was a patient diagnosed with TCAPA at autopsy who had a sibling diagnosed with ALCAPA.^{Reference Goldblatt, Adams, Ross, Savage and Morris21} Though it is possible that the association of these two rare diagnoses was a mere coincidence, this observation supports, but does not demonstrate, a genetic or environmental basis for anomalous origin of a coronary artery from the pulmonary artery.

An understanding of the direction of blood flow in anomalous origin of a coronary artery from the pulmonary artery is necessary in understanding the associated pathophysiology. In 1886, John Brooks hypothesised the direction of blood flow in the anomalous vessel to be retrograde based on pathologic review of a patient with ARCAPA.^{Reference Brooks22} This theory remained unsupported until 1960 when Drs. David Sabiston Jr, Catherine Neill, and Helen Taussig published their findings regarding a 2.5-month-old boy with ALCAPA.^{Reference Sabiston, Neill and Taussig23} They sampled blood from the left coronary artery and pulmonary artery and found the oxygen saturation in the left coronary artery to be 100% compared to 76% in the pulmonary artery, thereby confirming retrograde flow. It is now known, based on detailed angiographic/echocardiographic findings in ALCAPA/ARCAPA, that antegrade flow of deoxygenated blood occurs in the anomalous vessel within the first 2 weeks of life when pulmonary resistance is high.^{Reference Singh, Di Carli, Sullivan, Leonen and Morrow24} Once pulmonary resistance decreases and collateralisation with the aortic-derived coronary artery is developed, retrograde flow of blood ensues and becomes reliant on the higher pressure from the aortic-derived coronary artery. However, in TCAPA, an aortic-derived coronary artery is not present and antegrade flow of deoxygenated blood continues after pulmonary resistance decreases.^{Reference Abdelmohsen, Alkhushi, Bahaidarah, Shihata and Jamjoom25} In this scenario, coronary perfusion pressure is directly related to pulmonary artery pressure. Cardiac lesions that increase pulmonary artery pressure such as ventricular septal defect, aorto-pulmonary window, and truncus arteriosus would therefore augment coronary perfusion pressure. Alternatively, lesions associated with left to right shunting (such as a ventricular septal defect or patent ductus arteriosus) would additionally increase the mixed venous oxygen saturation and would theoretically augment coronary artery oxygen saturation and in this cohort were associated with later ages of presentation in patients with concomitant TCAPA.

The clinical presentation of patients with TCAPA was found to be quite variable. Pulmonary artery pressure decreases after birth as intrinsic pulmonary resistance decreases as the lung expands and the ductus arteriosus closes.^{Reference Teitel, Iwamoto and Rudolph26} This change is most dramatic in the first 2–3 days of life and adult range pulmonary artery pressures are achieved by 2 weeks of life.^{Reference Rudolph27} In TCAPA, this decrease in pulmonary artery pressure (therefore coronary perfusion pressure) is maladaptive and likely contributes to the cohort of patients presented within the first 2 weeks of life. Observations of extensive myocardial hypertrophy/remodelling, fibrosis, and other pathologic evidence of chronic ischemia have been seen at autopsy in some patients with TCAPA immediately after birth and are suggestive of prenatal insults.^{Reference Swann and Werthammer28} In normal fetal anatomy, a flap of tissue known as the Eustachian valve preferentially directs the flow of well-oxygenated blood from the inferior vena cava/placenta towards the foramen ovale and into the left atrium/ventricle, thereby preferentially delivering a relatively high concentration of oxygenated blood to the heart/brain. Alternatively, desaturated blood from the superior vena cava is preferentially directed towards the tricuspid valve and then away from the pulmonary circulation into the proximal descending aorta via the ductus arteriosus.^{Reference Murphy29} For this reason, relatively decreased coronary oxygen saturations in patients with TCAPA in utero may theoretically lead to cardiac dysfunction prior to delivery and thus explain the chronic appearing pathologic changes observed in some patients soon after birth.

Although only half of patients with TCAPA presented within the first 2 weeks of life, 91.2% of all patients became symptomatic within the first year of life. Reasons for the delay in presentation can most likely be attributed to the presence of associated cardiac lesions that maintained increased pulmonary pressure and mixed systemic venous oxygen saturation such as ventricular septal defect, patent ductus arteriosus, aorto-pulmonary window, and truncus arteriosus. The presence of associated cardiac lesions causing a right to left shunt is one explanation for the relatively high rate of cyanosis observed in patients with TCAPA. However, this explanation is not complete, as three patients with TCAPA and no other cardiac lesions presented with cyanosis. This might be explained by the myocardial ischemia associated with TCAPA leading to heart failure, as echoed by the observed decreased ejection fraction seen in this cohort of patients with TCAPA. The cohort that presented beyond 2 weeks of life without an associated cardiac lesion might be explained by the fact that as left ventricular function deteriorates due to ischemia, diastolic dysfunction ensues resulting in elevated left atrial pressure (leading to pulmonary hypertension) that then would augment flow to the anomalous coronary arteries.

Two-thirds of patients with TCAPA were associated with a single coronary artery arising from the pulmonary artery. Single coronary artery from the aorta is a well-described rare lesion with an estimated incidence between 0.024 and 0.066%.^{Reference Desmet, Vanhaecke and Vrolix30,Reference Lipton, Barry, Obrez, Silverman and Wexler31} Patients can present incidentally or with evidence of coronary ischemia. Depending on the course of the anomalous single coronary artery from the aorta, treatment is typically reserved for symptomatic patients.^{Reference Akcay, Tuncer and Batyraliev32} A classification system was described by Lipton et al. and then modified by Yamanaka et al. who detailed the origin and course of the single coronary artery from the aorta.^{Reference Lipton, Barry, Obrez, Silverman and Wexler31,Reference Yamanaka and Hobbs33} This classification system is not applicable to TCAPA, as components of this classification system describe the origin from the aortic sinus of Valsalva as well as branching from “normally” located coronary arteries. The characterisation in Figure 4 better represents not only the number of coronary arteries involved in TCAPA, but also the phenotypic distribution of their observed origin and course as it relates to the pulmonary artery.

Although the diagnosis of TCAPA was made at autopsy in 46% of patients, improvements in echocardiography, angiography, and cross-sectional imaging have led to an increased number of premortem diagnoses and the potential for operative intervention. Failure to obtain a pre-operative diagnosis was associated with death for two reasons: 1) failure to achieve adequate myocardial protection during administration of cardioplegia^{Reference Vitanova, Cleuziou, Deutsch, Ackermann and Schreiber34} and 2) correction of a cardiac lesion that then decreased pulmonary pressure, thereby unknowingly disrupting coronary perfusion.^{Reference Feldt, Ongley and Titus12} Angiography or CT angiography offers the advantage of identifying the origin/course of the anomalous vessel and should be considered the gold standard for diagnosis. The inability to identify coronary ostia or coronary flow on left-sided angiography prompted investigation via right-sided angiography and led to an angiographic diagnosis in 16 patients with TCAPA. Echocardiography has emerged as a non-invasive method to diagnose many coronary artery anomalies.^{Reference Angelini35} Assessments of ostial location and direction of flow through the anomalous vessel on echocardiogram can help differentiate variants of anomalous origin of a coronary artery from the pulmonary artery, although their sensitivity and specificity are not perfect. Numerous cases of TCAPA were incorrectly diagnosed as ALCAPA based on pre-operative echocardiogram, likely related to the inherent difficultly assessing the right coronary artery on echocardiography.^{Reference Zhang, Zhang and Gao36–Reference Donmez, Aykan and Yilmaz38} However, in ALCAPA, flow through the anomalous left coronary artery is more commonly retrograde, whereas in TCAPA flow is antegrade (from the pulmonary artery).^{Reference Abdelmohsen, Alkhushi, Bahaidarah, Shihata and Jamjoom25} The distinction between TCAPA and ALCAPA in the pre-operative setting is important because the protective effects of collateralisation from an aortic-derived coronary are not present in TCAPA, therefore increasing the urgency of operative intervention. Other echocardiographic findings suggestive of TCAPA, similarly seen in ALCAPA, include depressed left ventricular function and dilation, and often associated mitral regurgitation.^{Reference Thatte, Kirakosian, Kaza and Friedman39}

Treatment of TCAPA is aimed at creating an aortic-derived coronary artery system. Prior to surgical intervention, prostaglandin E1 should be considered in newborns to preserve the ductus arteriosus and augment coronary perfusion pressure and pulmonary artery oxygen saturation.^{Reference Karimi, Hulsebus and Lutin40} Although isolated percutaneous or surgical ligation of the anomalous vessel at its origin has been described in the treatment of ALCAPA/ARCAPA, this strategy has no role in TCAPA.^{Reference Collins, Colman, Benson, Hansen, Merchant and Horlick41,Reference Francis and Brown42} This is due to the dependency of the pulmonary artery for blood flow in TCAPA and the absence of coronary steal associated with the lesion. Likewise, ligation of the anomalous vessel at its origin with coronary artery bypass grafting has been a utilised treatment strategy in ALCAPA/ARCAPA. Though this method does establish an aortic-derived coronary artery system, it makes the entire coronary circulation reliant on bypass graft patency; a significant long-term concern given the age of presentation among patients with TCAPA. The two main treatment strategies that were used in TCAPA patients were aortic re-implantation or a Takeuchi repair using an intra-pulmonary baffle. An objective comparison between techniques is not possible in this review given a lack of standardisation, small sample size, and limitation in long-term follow-up/patency. Nevertheless, a greater than twofold increase in mortality was seen with the Takeuchi repair. This could be related to: 1) known complications related to the Takeuchi repair such as baffle leak, damage to the pulmonary valve, or supra-valvular pulmonary artery stenosis^{Reference Ginde, Earing, Bartz, Cava and Tweddell43} or 2) the median year of publication for patients undergoing the Takeuchi repair was 28 years earlier than the median year of publication for aortic re-implantation. Nevertheless, clinicians should proceed with caution if the Takeuchi repair is chosen over aortic re-implantation, given this observation in mortality.

There are numerous limitations to this review. First, there is an inherent bias in the cases that were included in our cohort as all came from published manuscripts. Although TCAPA is in itself rare, there is a push to publish rare presentations and associated lesions that may have altered some of our phenotypic observations. Next, some aspects assessed, such as presenting symptoms, physical exam findings, and outcomes are inherently subjective. Given the lack of standardisation, collective analysis becomes difficult and does introduce bias. Likewise, cases were included from 1931 to 2020. Major improvements in diagnostic modalities, cardiopulmonary bypass, and peri-operative care make collective analysis problematic. Furthermore, the data that were tabulated included only what was written by the author of each manuscript; data from all patients regarding cardiac catheterisation, echocardiography, and pathologic assessment of tissue were not present in each case. Although these could have been made inclusion criteria, they would have significantly decreased the overall sample size. Lastly, although nearly 20% of manuscripts had follow-up data, long-term follow-up is lacking which makes comparison of treatment strategies difficult. Nevertheless, analysis of case reports likely represents the only current avenue for gaining a better understanding of rare lesions such as TCAPA until clinical databases are robust enough to provide a sample size large enough for meaningful analysis.^{Reference Carey44}