A prospective observational study of associated anomalies in Hirschsprung’s disease

Hirschsprung’s disease (HSCR) is a congenital multifactorial developmental disorder of the enteric nervous system characterized by the absence of ganglion cells in the hindgut with variable distal bowel involvement. This is a rare disease, which occurs with an incidence of 1 into 5000 live births as a consequence of premature arrest of cranio-caudal migration of neural crest derived neuroblasts (NCN) in the hindgut, and is therefore regarded as a neurocristopathy [1‐3]. The neural crest is one of the earliest organs to form within the developing embryo, during formation of the neural tube. Cells of the neural crest are pluripotential and follow migratory pathways depending on their axial level of origin. NCN differentiate into crucial cell types ancestors and participate in the development of various organs such as adrenal medulla, melanocytes, craniofacial cartilage and bone, smooth muscle, neuroendocrine cells and sympathetic and parasympathetic nervous systems, including the enteric nervous system [3, 4]. Alterations in any of the genes involved in the enteric nervous system development may interfere with the colonization process of NCN and represents a primary aetiology for HSCR [5, 6]. The major gene responsible for HSCR is RET, whose mutations are detected in up to 50% of familial and in 7-35% of sporadic cases [7]. RET is crucial for the embryologic development of the enteric nervous system but also of other organs, including kidneys and urinary tract [8]. So far, a number of other minor HSCR susceptibility genes have been identified in less than 5% of patients [7]. The involvement of heterogeneous genetic pathways and pluripotential cell lineage explains why HSCR disease can be associated to malformations basically involving all organs and system.

Associated anomalies have been detected in between 20% and 30% of HSCR patients according to literature [5, 9, 10]. In 2006, Moore SW performed a systematic review of 18 series with 4328 patients reported in the last forty years and suggested an average percentage of associated anomalies of 21.1%, ranging between 5% and 35% [10]. Similarly, Amiel J and co-workers reported a percentage of associated anomalies of around 30% [9]. Nonetheless, both Authors suggested a measure of underestimation [9, 10]. In accordance to these considerations, we recently demonstrated that Congenital Anomalies of the Kidney and Urinary Tract (CAKUT) occur 4-folds more frequently than previously reported [8]. As a consequence, it comes clear that the exact clinical phenotype of HSCR patients is yet unknown and deserves definition to provide our patients with reliable prognostic expectations. A better knowledge of all associated anomalies in HSCR will also reveal undisclosed genetic implications that will improve the comprehension of HSCR pathophysiology. We therefore implemented a prospective observational cross-sectional study on 106 consecutive HSCR patients aimed at defining the percentage of associated anomalies and at subsequently implementing an up-to-date diagnostic algorithm.

All consecutive HSCR patients who were admitted to Giannina Gaslini Research Institute between January 2010 and December 2012 were eligible for this prospective observational cross-sectional study. Inclusion criterion was the presence of a reliable diagnosis of HSCR achieved with adequate pathological assessment [11]. Exclusion criteria were uncertain diagnosis and/or refusal to participate.

This study received Institution Review Board approval on the 15^th of October 2009. After enrolment, patients and parents were interviewed to collect demographic data, personal and family history and any information regarding known associated anomalies. All families were asked to sign a specific Informed Consent form and in case of acceptance each patient underwent an advanced non-invasive diagnostic algorithm, including ultrasound scan (US) of the kidney and urinary tract, cardiologic assessment with cardiac US, audiometry and ENT assessment, ophthalmologic assessment, cerebral US (< 1 year-old), and further specialist investigations based on clinical features. All these non-invasive investigations were performed according to widely accepted guidelines. The following data were collected according to the Personal Data Protection Act:

This is the first prospective observational study aimed at defining the spectrum of possible phenotypes in a large cohort of HSCR patients. A reason why none previously addressed this issue probably relies on the rarity of the disease and on the difficulty to organize and perform such complex multidisciplinary prospective clinical assessments. In fact, literature reports regarding associated anomalies in HSCR are merely retrospective assessments of surgical series or systematic literature reviews [5, 9, 10].

The results of our study are remarkable and confirmed a measure of underestimation already suggested by some [5, 8‐10, 26, 27]. In particular, we demonstrated that the percentage of associated anomalies is higher than expected with involvement of basically all organs and systems.

Delayed diagnosis or mild symptoms can explain the reason for such a discrepancy. In fact, refractive anomalies or visual impairment (VI), mild CAKUT or unilateral hearing impairment (HI) can be diagnosed with significant delay in the absence of specific screening programmes [8, 19, 28].

With regard to phenotype considerations, our study confirmed the high percentage of CAKUT in HSCR patients, as previously published [8]. In fact, despite some overlap of patients included in both studies (17% of cases), the percentage of associated CAKUT remained constantly high as well as the preponderance of hypoplasia, vesicoureteric reflux and hydronephrosis amongst detected malformations. Based on the results of our studies, we thus confirm that renal US should be included as a routine diagnostic investigation in all HSCR patients.

We could not detect any major ocular anomaly in our series of HSCR patients but nearly 10% of VI and more than 40% of refraction abnormalities. The latter are frequent in HSCR patients as well as in the general population. In 2009, the Baltimore Pediatric Eye Disease Study indicated in 0,7% the prevalence of myopia and nearly 9% that of hyperopia, in 1030 white Americans younger than 6 years of age [22]. Similarly, in 2011 the Sydney Pediatric Eye Disease Study examined 2461 children younger than 6 years of age and suggested an incidence of 6,4% of VI mostly due to refractive errors, amblyopia, and strabismus [23]. In 2008 Jobke and co-workers performed a population-based study on 516 German children, adolescent and young adults aged between 2 and 35 years and ended up with a prevalence of over 75% of emmetropia, nearly 20% myopia and 5% hyperopia [24]. Due to the heterogeneity of literature data regarding the general population, we can hardly determine the relative risk of HSCR patients of having refraction abnormalities. Nonetheless, based on embryologic considerations regarding the shared neural crest origin of enteric nervous system and specific ocular structures and on the relatively high percentage of patients with refractive errors and VI observed in our study, we strongly recommend to perform routine ophthalmologic assessments (as those performed in the general population) in order to early detect and prevent VI, in accordance to WHO suggestions [28, 29].

The prevalence of major or moderate CHD has been reported to be around 7,2 per 1000 live birth in the general population [30], whereas the proportion of CHD in HSCR patients has been reported to be around 5%, according to literature [5, 9, 10]. We could basically confirm these data (4,7% in our series). In particular, CHD detected during phenotype screening mostly occurred in patients with chromosomal abnormalities and were essentially represented by septal defects. This is in accordance with the embryologic role of NCN in the development of both enteric nervous system and cardiac outflow septation [31]. On the other hand, none of the patients presented conotruncal heart defects, whose pathogenesis is similarly related to the abnormal neural crest cell proliferation and migration [31]. The percentage of minor CHD or trivial lesions without clinical significance in our series accounted for another 6,6% of patients, well within literature ranges [30]. Based on these results we do suggest performing cardiac US only in the subpopulation of known or suspected syndromic HSCR patients.

In the general population, the incidence of HI in subjects without audiological risk factor is around 1,5‰, whereas that of subjects with audiological risk factors is 4,5% [32, 33]. With specific regard to HSCR patients, syndromic HI can be found in patients with Waardenburg-Shah type 4, due to mutations of SOX10 gene [5, 9, 10]. Of note in 1973 Skinner described 4 patients with congenital deafness and HSCR and suggested that this association could not be fortuitous but based on shared embryological background as only one case had known risk factors for HI (streptomycin administration) [34]. Based on these considerations and on the results of our study (4,7% of HI), we recommend that HSCR itself should be included amongst risk factors for HI. Subsequently HSCR patients should undergo specific audiologic screening programmes as those suggested by the Joint Committee of Infant Hearing in 2007 [19].

Although Moore reported a relatively high proportion of brain abnormalities in HSCR patients, we could only confirm a case of corpus callosum agenesis in a patient of ours who was already diagnosed with prenatal US [10, 26]. The child had no further syndromic features and a near-normal neurologic development (mild dyslexia). In our series of patients we could also detect some ventricle asymmetry and subarachnoid spaces dilatations that were found in less than 20% of cases, mostly on the left hand side. Of note, asymmetry between the sizes of the ventricles can be observed in up to 40% of healthy infants, being the left ventricle often larger than the right [25]. Based on these results and in accordance with literature data, we do not recommend performing routine cerebral US in all HSCR patients but only in those with clear syndromic features who have higher likelihood of carrying cerebral malformations (i.e. Mowat-Wilson Syndrome) [10, 26, 35].

Amongst other associated anomalies (Table 6) tumors were exclusively represented by FMTC, well known to be associated with specific genotypes at the RET gene locus in HSCR patients [7]. As expected, after testing for mutations of the RET gene (data not shown), we could verify that the two patients of our panel who are affected with HSCR + FMTC syndromic association are indeed carriers of the p.C609Y and p.Y791F mutations, respectively. These mutations have already been reported in association with this combined phenotype and also described as “Janus mutations” [36].

The percentage of anomalies that could have been tracked down retrospectively by notes reviews or questionnaire administration resembles that of literature reports [5, 9, 10]. In fact, 26,4% of patients was already aware of their association when they entered our screening programme. This aspect further confirms the reliability of our results and that a deeper clinical screening could have detected the missed associated anomalies (“the more you look for, the more you find”).

Besides providing a novel view on clinical associations and variable symptoms of HSCR disease in children, the present data prompt us to deepen into follow-up data from adults who were treated for HSCR, which are lacking at the moment. Indeed, we cannot exclude that additional organs or systems can be involved and become impaired later in adult life. Indeed, typical adulthood associated diseases have not been described yet and are unknown at the moment. For instance, association with specific cancers, nervous degeneration, intestinal chronic inflammation and/or other late-onset disorders cannot be ruled out and further investigation in an adult set of HSCR patients would be advisable.

Our Institution is known to be a referral centre for HSCR with an intrinsic subsequent risk of inclusion BIAS. This concern was confirmed by the higher than expected percentage of ultralong forms of the disease in our series (17,9%). Nonetheless, given the fact that the percentage and type of associated anomalies did not correlate to gender or length of aganglionosis, the results of our study maintained their strength. Furthermore, the percentage of Down Syndrome as well as the strong male preponderance observed in our series of patients are coherent with literature data and confirm that the results of our study are representative of the whole HSCR population [5, 9, 10].

The relatively small population of this observational study represents another potential limitation. As a consequence, the estimates we could provide have the intrinsic limitation of a wide variability and should be taken with care (see Tables 1 and 7 for details). Nonetheless, HSCR is a rare disease and the number of patients enrolled in such a relatively small time-span implies strong commitment and a multidisciplinary approach that deserve consideration.

It is evident that the implementation of a prospective multicentre research project is warranted. In fact, a larger series of patients could increase the strength of the results and possibly confirm the cost-effectiveness of this proposed diagnostic algorithm for a significant change in clinical practice for HSCR management (Figure 1).

Based on the results of our study, we suggest performing US of the kidney and urinary tract as well as audiologic investigation in all cases, whereas heart US, cardiologic assessment, cerebral US and ophthalmologic assessment should be performed basing on clinical features and according to the standards of care adopted for the general population. On the other hand, all investigations should be considered for patients with known or suspected syndromes or chromosomal abnormalities in order to promptly apply specific prevention strategies, as we already suggested (i.e. preoperative prophylactic stoma and postoperative rectal irrigations in case of associated CHD) [37]. Figure 1 shows a suggested diagnostic algorithm. It includes early clinical genetic assessment in order to detect symptoms and signs suggestive of syndromes that would lead to further investigations.

Early diagnosis of associated anomalies provides more reliable prognostic expectations, prompt establishment of prevention strategies and implementation of adequate rehabilitation treatments. In accordance to our previous publication and to present results, we hypothesize that intestinal aganglionosis in HSCR patients represents the intestinal phenotype of a more complex syndrome driven by the interaction of neural crest maldevelopment and predisposing genetic background [5, 7‐10, 29, 31]. The investigation of genetic background of individuals presenting with associated anomalies might be the next step to explore this intriguing multifactorial congenital disease.