Hirschsprung's disease (HSCR) is one of the most common congenital digestive tract disorders in neonates with an incidence of approximately 1 in 5,000 live births (1). In HSCR patients, the absence of ganglion cells in a variable length of the distal gut, beginning at the internal sphincter and extending proximally, results in absent peristalsis in the affected bowel. The final diagnosis depends on rectal suction biopsy by identifying the absence of ganglion cells (2).

HSCR is caused by deficient cranio-caudal migration, proliferation, differentiation and colonization of neural crest cells in the hindgut (3). Neural crest cells are pluripotential and then gives rise to crucial cell types ancestors, based on their migratory pathway, terminal location, and differential abilities (4). Cardiac neural crest cells, a smaller specified subgroup of neural crest cells, significantly contribute to proper cardiovascular formation, including development of the smooth muscle, septation of the cardiac outflow tract, and patterning of the septa, valves and arterial truncus (5). The developmental deficiencies in cardiac neural crest cells are considered to cause a variety of cardiac malformations, i.e., congenital heart defect (CHD) (6).

CHD is the most common congenital structural disorder in newborns affecting approximately 1% of live births (7). It is also the leading cause of mortality from birth defects (8). There is a view that both HSCR and CHD could be regarded as neurocristopathies. Several studies have reported the high incidence of CHD, being about 1.4%–17% in HSCR patients (9–14). Most of the cases were mild or moderate forms, including atrial or ventricular septal defects, mild aortic or tricuspid valve abnormalities and coarctation of the aorta. The sample size of those studies was small (ranging from 53 to 207 patients). One report described the association of hypoplastic left heart syndrome with HSCR (13). Other than this, there has been no report, to the best of our knowledge, about the occurrence of severe types of CHD in HSCR patients.

Our institution has established the Provincial Key Laboratory for Structural Birth Defects with a large number of HSCR patients. We therefore implemented a 20-year retrospective study on the patients with HSCR aiming to assess the incidence and phenotypes of associated CHDs based upon a fairly large cohort.

This study was approved by the Institutional Review Board of Guangzhou Women and Children's Medical Center (No. 2018052406). Informed consent was waived owing to the nature of this retrospective medical record review study. All HSCR patients admitted for further evaluation and/or surgical treatment to the Guangzhou Women and Children's Medical Center (Guangzhou, Guangdong province, China) between January 2003 and December 2020 were retrospectively reviewed. Inclusion criteria were patients with pathologically confirmed diagnosis of HSCR. Exclusion criteria were unavailable medical records for review.

Demographic and clinical data were collected from the patient's records. During the hospitalization, all patients were assessed by clinical observation, physical examination and further specialist investigations based on clinical features. Echocardiographic examination was performed when a cardiovascular abnormality was suspected based on symptoms, depressed pulse oximetry, heart murmur, family history of congenital heart disease or abnormal prenatal ultrasonic findings. Detailed echo-scan was performed through a transthoracic approach with pulsed, continuous and color-Doppler using echocardiographic systems.

In our cohort, 4 types of HSCR were identified according to the length of aganglionosis: short segment HSCR (aganglionosis extending up to the left descending colon), long segment HSCR (aganglionosis extending up to the right transverse or ascending colon), total colonic aganglionosis (aganglionosis extending to the whole colon) and total intestinal aganglionosis (aganglionosis extending over the jejunum with less than 20 cm of normo-ganglionic bowel).

CHD was defined as a gross structural abnormality of the heart or intrathoracic great vessels that is actually or potentially of functional significance (15), diagnosed with echocardiography by the pediatric cardiologists. In according with many other epidemiologic studies, the definition of CHD excluded functionless abnormalities of the great veins (such as persistent left superior cava), bicuspid aortic valves, mitral valve prolapse, Marfan syndrome, cardiomyopathies, and congenital arrhythmias without an associated structural heart lesion (such as long Q-T syndrome). Atrial septal defects within oval fossa, patent foramen ovale and patent ductus arteriosus throughout the first 14 days of life were also excluded because they are considered normal findings. The CHD patients were grouped by disease severity. The category of severe CHD included those who were severely ill in early infancy (16).

HSCR was first described in 1,886 by a Danish physician, Harald Hirschsprung, and the chief culprit was revealed as aganglionosis in all or distal part of the colon. Colonic lesions fail to relax, causing functional colonic obstruction over time (17). HSCR occurs as an isolated phenotype in most cases, while a number of HSCR-associated anomalies such as gastrointestinal, neurologic, and genitourinary anomalies have been reported (10). In particular, concomitant CHDs have been drawn attention on account of its influence on the long-term prognosis.

Several studies over the last few decades have reported CHD in fairly small populations of HSCR. A prospective observational study published in 2013 detected a prevalence rate of 4.7% out of 106 children with HSCR, with coarctation of the aorta in 1 and septal defects in 4 cases (9). In another prospective study, 133 consecutive HSCR patients underwent cardiac screening and 11 patients (8.3%) presented CHDs which were mostly septal defects or small patent ductus arteriosus (14). Two epidemiological studies reported the incidence of 4/207 (1.4%, 3 with ventricular septal defect and 1 with patent ductus arteriosus) and 2/126 (1.6%, no details given) respectively (10, 18). One report showed higher incidence of HSCR in patients with hypoplastic left heart syndrome (13). In our largest cohort of patients with HSCR ever reported, we found that 1.2% of children had concomitant CHD, which were mostly presented by atrial or ventricular septal defects. The incidence of severe CHD was 1.4‰ (N = 3), including complete atrioventricular canal, congenitally corrected transposition of the great arteries and double-outlet of right ventricle. None of them has ever been reported. In patients with syndromic disorders, the prevalence of associated CHD was 28.6%. Among 7 patients with Down's syndrome, 42.9% had associated CHD, in concordance with previous studies (9, 14, 19).

There are several reasons which could lead to a wide range in estimates of incidence. It may be attributed to the dissimilarities in definition of CHD and methods among the studies (16). High incidences were found in the prospective studies, as they detected mild or moderate CHDs, which are often asymptomatic and do not have significant murmurs. By contrast, estimates of prevalence of severe CHD are more reliable, since that few patients in the severe group will be misdiagnosed in the children's medical centers. Most epidemiological studies reported an estimated number of 1.0–1.7 cases per 1,000 live births for severe CHD (16, 20, 21). A similar incidence was found in our HSCR population.

More interestingly, 2 of the severe CHDs were conotruncal developmental anomalies, i.e., congenitally corrected transposition of the great arteries and double-outlet of right ventricle. This may be expected as the critical embryologic role of neural crest cells in the development of both enteric nervous system and cardiac outflow tract and septation. After undergoing epithelial-mesenchymal transition, multipotent cells of the neural crest migrate and subsequently differentiate into a wide variety of cell type ancestors that are responsible for the development of various organs (22, 23). Various animal models have repeatedly confirmed that vagal neural crest cells are the main source of the enteric nervous system constituents (3, 24). The crucial role of neural crest cells in conotruncal separation during proper heart formation have been demonstrated (5, 6, 25). Although progress has been made in revealing possible embryological connection between CHD and HSCR, the developmental processes of neural crest cells are extraordinary complexities. Future experimental work in this area is needed to determine the regulatory pathways and factors during the development of cardiovascular and enteric nervous system.

In our cohort, 14 patients (0.6%) had Down's syndrome or other syndromes. Several particular reasons probably account for the lower than reported percentage (4.3%–6.6%) of Down's syndrome in our live-born series (9, 10, 14, 19). In China, prenatal screening for Down's syndrome is offered as part of routine antenatal care, and termination of pregnancy for Down's syndrome is legal (26, 27). From the viewpoint of Down's syndrome, the most frequent associated anomalies are cardiac disorders, which has been reported to occur in roughly 44%–60% (28, 29). Besides, the clinical association between Down's syndrome and HSCR has been well-established (30, 31). However, the general incidence of CHD associated with HSCR-Down's syndrome has not substantially changed compared with all Down's syndrome patients – 3 out of 7 HSCR-Down's syndrome patients (42.9%) presented also an associated CHD and similar proportions in other studies (36.1%–62.5%) (9, 14, 19). The past three nationwide surveys in Japan found that cardiovascular anomalies were increased over time among HSCR-Down's syndrome patients (19). The incremental trends were consistent with the increasing diagnosed incidence in general population over the years, which primarily due to the rising frequency of echocardiographic ultrasound screening and the growing ability to detect mild lesions. Nonetheless, the associations between CHD, HSCR and genetic syndromes are still of significance in revealing the possible abnormal genetic background and pathogenesis and in evaluating the prognosis.

Our study has several limitations. The observational retrospective study had intrinsic risks of information bias and temporal bias, and the specific biological mechanisms are speculative. Besides, our echocardiographic evaluation was conducted in only 306 patients depending on referral, which tended to underestimate the number of mild lesions, particular atrial septal defect. But it is unlikely that major cardiac defects were missed. In addition, it represents a single-institutional experience for a referral center for HSCR and should extrapolate with care. A meta-analysis combining numerous studies may help to identify further associations. Furthermore, missing information about long-term follow-up resulted in inability to assess the effect of associated CHDs on the outcomes of patients with HSCR.

Our retrospective study during the last twenty years in China demonstrated CHDs in 1.2% of 2,174 HSCR patients. Echocardiography evaluation in 306 patients helped to identify 3 patients (1.4‰) with severe CHD including complete atrioventricular canal, congenitally corrected transposition of the great arteries and double-outlet of right ventricle, which have never been reported.