Genetics of diverse phenotypes in Hirschsprung disease : extension of aganglionosis, heredity and medullary thyroid carcinoma

(1)

Section of Pediatric Surgery Children’s Hospital

Helsinki University Central Hospital Faculty of Medicine

University of Helsinki Helsinki, Finland

Doctoral Programme in Clinical Research

GENETICS OF DIVERSE PHENOTYPES IN HIRSCHSPRUNG DISEASE – EXTENSION OF AGANGLIONOSIS, HEREDITY AND MEDULLARY

THYROID CARCINOMA

Valtter Virtanen

ACADEMIC DISSERTATION To be presented with the permission of The Faculty of Medicine of University of Helsinki,

for public examination in the Niilo Hallman Auditorium, Children’s Hospital, On 1^st of February 2019, at 12 noon.

Helsinki 2019

(2)

Supervisors

Professor Mikko Pakarinen Professor Markus Perola Section of Pediatric Surgery National Institute for Children´s Hospital Health and Welfare

Helsinki University Central Hospital Helsinki, Finland Reviewers

Professor Jacob Langer Docent Marja Hietala The Hospital for Sick Children Clinical genetics

University of Toronto Turku University Hospital

Toronto, Canada University of Turku

Opponent

Allesio Pini Prato, MD

Consultant Pediatric Surgeon

Director in Chief of Pediatric Surgery

Umberto Bosio Center for Digestive Diseases Colorectal Center, The Children's Hospital AON SS Antonio e Biagio e Cesare Arrigo

Alessandria, Italy

ISBN 978-951-51-4831-5 (paperback) ISBN 978-951-51-4832-2 (pdf)

http://ethesis.helsinki.fi Unigrafia Oy

Helsinki 2019

The Faculty of Medicine uses the Urkund system (plagiarism recognition) to examine all doctoral dissertations.

(3)

3 To Hirschsprung disease patients

(4)

4

1 LIST OF ORIGINAL PUBLICATIONS

This thesis is based on the following publications. The publications are referred to in the text by their Roman numerals and are reprinted with the permission of the publisher. In addition, some unpublished data are presented.

I. Virtanen VB, Pukkala E, Kivisaari R, Salo PP, Koivusalo A, Arola J, Miettinen PJ, Rintala RJ, Perola M, Pakarinen MP. Thyroid cancer and co-occurring RET mutations in Hirschsprung disease. Endocr Relat Cancer 2013;20:595-602.

II. Virtanen VB, Salo PP, Cao J, Löf-Granström A, Milani L, Metspalu A, Rintala RJ, Saarenpää-Heikkilä O, Paunio T, Wester T, Nordenskjöld A, Perola M, Pakarinen MP. Noncoding RET variants explain the strong association with Hirschsprung disease in patients without rare coding sequence variant. Eur J Med Genet 2018 in Press.

III. Virtanen VB, Salo PP, Perola M, Pakarinen MP. Rare coding sequence variants explain minority of Hirschsprung disease cases.

2018 Submitted.

(8)

8

2 ABBREVIATIONS

CCHS Congenital central hypoventilation syndrome ENS Enteric nervous system

FMTC Familiar medullary thyroid carcinoma FNB Fine needle biopsy

GRR Genotype relative risk

GWAS Genome wide association study HD Hirschsprung disease

LS Long segment

MAF Minor allele frequency MEN Multiple endocrine neoplasia MEN2A Multiple endocrine neoplasia 2A MLSA Multi-locus sequence analysis MTC Medullary thyroid carcinoma

PAF Population attributable fractions PTC Papillary thyroid carcinoma

RAIR Rectoanal inhibitory reflex RS Rectosigmoid

RV Rare variant

SIR Standardized incidence ratio SNP Single nucleotide polymorphism TCA Total colon aganglionosis US Ultrasound

WS Waardenburg syndrome

(9)

9

3 ABSTRACT

The aim of this doctoral thesis was to identify the genetic variations of Hirschsprung disease (HD) among Finnish HD patients. How the different genotypes affect different HD phenotypes, such as the extension of aganglionosis, heredity, and HD-associated medullary thyroid cancer (MTC) was also explored.

The entire patient cohort consisted of 112 HD patients. Of them, 91 adult HD patients participated the screening study for thyroid cancer. The participants went through a cross-sectional study where sequencing of the coding region of the RET gene was combined with blood tests and

ultrasound of the thyroid gland. If necessary, a fine needle biopsy was also performed. An additional 21 familial HD patients participated in the second stage of the study where diverse genetic analyses were performed.

In the second stage of the study, the material was analyzed

comprehensively using four complementary methods. Of these, capillary sequencing of RET exons and targeted sequencing of all genes known to be associated with HD was targeted to those gene loci that have shown an association with HD in previous studies. Additionally, a genome wide association study (GWAS) and whole-exome sequencing were performed to examine the whole genome.

In the screening study of thyroid cancer, two cases of MTC and one case of papillary thyroid cancer were observed. Both patients with MTC had RET variants, [p.Cys611Arg] and [p.Cys620Arg], which are known to be associated with MTC. In addition, MTC-associated RET variants were observed in four patients with no clinical signs of thyroid cancer; these

(10)

10 patients were selected for follow-up. Based on this study, the risk of MTC among HD patients is over 300 times higher than in the normal population.

The sequencing of all RET exons revealed 10 rare variants that affected gene function in 16 (14%) patients. An EDNRB frameshift variant was identified in two patients from the same family using whole-exome sequencing.

GWAS confirmed the strong association of the RET gene with HD. About half of the cases in our entire sample may be statistically attributed to the common non-coding RET variants.

Sequencing of all genes known to be associated with HD revealed 10 variants affecting gene function in nine patients. Overall, performing

extensive gene analysis revealed a coding sequence variant affecting gene function in 31 (28%) patients from whole study cohort.

Screening and gene testing are important to significantly improve the diagnostics, treatment planning, and patient counseling of HD. In particular, patients with genetic defects associated with an increased risk of thyroid cancer can be identified and treated in a timely manner.

(11)

11

4 ABSTRAKTI (ABSTRACT IN FINNISH)

Tämän väitöskirjatyön tavoitteena oli tunnistaa Hirschsprungin tautiin liittyvät geneettiset variaatiot suomalaisten Hirschsprungin tautia

sairastavien potilaiden joukossa. Lisäksi tavoitteena oli selittää, kuinka eri genotyypit vaikuttavat Hirschsprungin taudin eri fenotyyppeihin, kuten agangliottisen suolen osan pituuteen, periytyvyyteen ja Hirschsprungin tautiin liittyvään medulaariseen kilpirauhasen syöpään.

Taudin geenitutkimuksiin osallistui 112 Hirschsprungin tautia sairastavaa potilasta. Näistä kilpirauhasen syövän seulontatutkimuksessa oli mukana 91 aikuista potilasta. Tutkimukseen osallistuneet kävivät läpi polikliinisen poikkileikkaustutkimuksen, jossa heidän kilpirauhasensa kuvattiin

ultraäänellä ja seerumin kalsitoniini- pitoisuus mitattiin. Kilpirauhasen pesäkemuutoksista otettiin ohutneulanäyte. Seulontatutkimuksessa tutkittiin lisäksi RET geenin koodaavan alueen variaatioita. Seulontatutkimuksen lisäksi geenitutkimuksiin osallistui 21 familiaalista Hirschsprungin tautia sairastavaa potilasta ja heidän 18 tervettä sukulaista.

Aineisto analysointiin kattavasti neljällä toisiaan täydentävällä menetelmällä. Näistä RET geenin eksonien kapillaarisekvensointi ja

kaikkien aiemmin Hirschsprungin tautiin yhdistettyjen geenien sekvensointi rinnakkaisella sekvensoinnilla kohdistettiin niihin perimän alueisiin, joiden aiempien tutkimusten perusteella oletettiin vaikuttavan taudin syntyyn.

Lisäksi perimänlaajuisen assosiaatioanalyysin ja kokoeksomisekvensoinnin avulla saimme koko perimän kattavaa geenitietoa.

Kilpirauhasen seulontatutkimuksessa havaittiin kaksi medullaarista ja yksi papillaarinen kilpirauhassyöpä. Molemmilta medullaarisen kilpirauhasen syövän sairastaneelta potilaalta löydettiin tunnetut syövälle altistavat

(12)

12 geenivariantit [p.Cys611Arg] ja [p.Cys620Arg] RET geenistä. Lisäksi

kilpirauhasen syöpään assosioituvia variantteja todettiin neljällä potilaalla, joilla syöpää tai sen esiastetta ei kliinisissä tutkimuksissa todettu. Heidän osalta aloitettiin seuranta endokrinologian poliklinikalla. Tutkimuksen perusteella medullaarisen kilpirauhasen syövän riski Hirschsprungin tautia sairastavilla potilailla on monisatakertainen muuhun populaatioon

verrattuna.

RET geenin kaikkien eksonien sekvensoinnissa löytyi kaikkiaan 10 harvinaista tautia aiheuttavaa varianttia 16 (14%) potilaalta.

Eksomisekvensoinnin avulla identifioitiin yhden perheen molemmalta Hirschsprungin tautia sairastavalta potilaalta variantti EDNRB geenistä.

Genominlaajuinen assosiaatiotutkimus vahvisti RET geenin vahvan yhteyden Hirschsprungin tautiin. Tutkimuksissa löytyneen tilastollisesti vahvimman SNP:n (Single nucleotide polymorphism) rs2505988 perusteella pystyttiin osoittamaan RET geenin yhteys Hirschsprungin tautiin lähes puolelta tutkimuspopulaation potilaista, joilta ei aiemmissa tutkimuksissa ollut löytynyt selvää tautia aiheuttavaa varianttia.

Sekvensoimalla kaikki aiemmin Hirschsprungin tautin yhdistetyt geenit pystyttiin identifioimaan 10 tautia aiheuttavaa varianttia yhdeksältä Hirschsprungin tautia sairastavalta potilaalta. Kaiken kaikkiaan laajojen geenitutkimusten perusteella pystyttiin osoittamaan tautia aiheuttava variantti 31 (28%) potilaalla.

Seulonta ja geenitutkimus ovat merkittäviä, jotta taudin diagnostiikka ja hoidon suunnittelu sekä perinnöllisyysneuvonta parantuisivat merkittävästi nykyisestä. Erityisesti potilaat, joiden geenivirheisiin liittyy lisääntynyt kilpirauhasen syövän riski, pystytään tunnistamaan ja hoitamaan riittävän ajoissa.

(13)

13

5 INTRODUCTION

Hirschsprung disease (HD) is the most common congenital intestinal motility disorder and occurs in approximately 1/4000 European newborns (Suita 2005, Best 2014). Disordered migration and differentiation of neural crest cells during embryonic development is thought to be the main cause of the absence of intramural ganglion cells in the myenteric and submucosal plexuses in the distal bowel (Martuciello G 2000, Heanue TA 2007).

Therefore, HD is regarded as a neurocristopathy. In addition, the neural crest gives rise to other neuronal, endocrine, and paraendocrine tissues that explain the increased incidence of other neurocristopathies in HD, like familial medullary thyroid carcinoma (FMTC), multiple endocrine neoplasia (MEN) syndromes, and neuroblastoma (Le Douarin N 1999). In most cases (80%), ganglion cells are absent at the rectosigmoid level. Sometimes, the disorder extends more to the proximal colon (15-20%) or even to the small intestine (5%), resulting in a more severe form of HD (Neuvonen 2017).

HD is treated by surgical removal of the aganglionic part of the intestine (Neuvonen 2017). In some rare cases of near-total intestinal aganglionosis, an intestinal transplantation might be necessary to avoid permanent

parenteral nutrition (Pakarinen MP 2013). Despite good surgical treatment of HD, anorectal and bowel function disorders may appear after the operation (Neuvonen 2017). If untreated, aganglionosis leads to intestinal obstruction, enterocolitis, malnutrition, and death.

The pathogenesis of HD is complex. The RET proto-oncogene is the major gene associated with HD (Emison ES 2005, Emison ES 2010). In addition to HD, RET variants play a key role in the development of MEN2

syndromes. Loss-of-function variants of RET are responsible for

(14)

14 development of HD, whereas gain-of-function variants cause MEN2

syndromes and MTC (Arighi E 2004).

Nevertheless, only a small portion of the HD cases can be explained by rare coding sequence variants of the RET gene. This has led to the discovery of additional disease genes. Currently, over 20 genes are defined as disease or candidate genes for HD, including EDNRB, SOX10, SEMA3C, SEMA3D, and other genes involved in the NTF-3/TRKC, prokineticin, NRGs, and SEMA signaling pathways (Ruiz-Ferrer M 2008, Fernández RM 2009, Garcia-Barcelo MM 2009, Ruiz-Ferrer M 2011, Wang LL 2011, Jiang Q 2012, Yang J 2013).

The complex inheritance pattern of HD, the incomplete penetrance of RET variants, and the small number of identified causative coding sequence variants in other genes underlines the role of factors other than those that affect RET coding sequence in the development of HD. The aim of this doctoral thesis is to identify HD-associated variants in Finnish HD patients and to clarify how genotypes affect diverse HD phenotypes (such as

extension of aganglionosis), heredity, and HD-associated medullary thyroid carcinoma (MTC).

(15)

15

6 REVIEW OF THE LITERATURE

6.1 Hirschsprung disease

6.1.1 Definition of HD

HD is the most common congenital intestinal motility disorder and is characterized by the absence of intramural ganglion cells in the myenteric and submucosal plexuses of the distal bowel (Martuciello G 2000, Heanue TA 2007). Disordered migration and differentiation of neural crest cells during embryonic development is thought to be the main cause of this neurocristopathy.

6.1.2 History of HD

HD was described for the first time already over 300 years ago by Frederick Ruysch (Cass D 1986). There were several subsequent reports of

conditions similar to HD, but it took nearly 200 years before the Danish pediatrician Dr. Harald Hirschsprung (1830-1916) made the official definition of the disease. It was the year 1886 when he presented the findings of autopsies of two infants who died from complications of bowel obstruction. The main findings were a dilated and hypertrophied colon and a rectum of almost normal appearance. Therefore, he named the disease congenital megacolon (Hirschsprung H 1888).

During the next 50 years, more cases of HD were reported and theories for the pathogenesis were developed. Eventually it was possible to define HD by demonstrating the absence of ganglion cells in the disordered segment.

As there are also other conditions that cause a dilated colon, demonstrating the absence of ganglion cells was key for the diagnosis of HD. Drs.

(16)

16 Swenson, Neuhauser, and Pickett started to use the barium enema as a diagnostic tool for HD in the mid-20^th century (Swenson O 1999). Dr.

Swenson and Bill also performed also a life-saving proximal colostomy in six patients. Poor results of this procedure led to the development of the procedure. The Swenson procedure was first performed in 1948 and is the first successful operative procedure for HD. The procedures involves partial resection of the colon and creation of a coloanal anastomosis (Swenson O 2004). Bernard Duhamel later described an alternative procedure for the treatment for the HD, which involves retrorectal dissection and longitudinal stapled colorectal anastomosis (Duhamel B 1956). A few years later, the Soave procedure was developed (Soave F 1964). These three procedures are still the most commonly performed operative treatments for HD.

As treatment for HD has not changed significantly in recent decades, the development of diagnostic tools and especially genetic tests took much longer. As late as the early 1990s, the RET gene was discovered as a disease gene (Pasini B 1995). At the same time, some studies confirmed the association of HD with other neurocristopathies. In this way other susceptibility genes, such as PHOX2B and SOX10, were also identified (Maris JM 1997, Pingault V 1998). At the beginning of 21^st century methods for gene analysis were rapidly developing, thus enabling the discovery of new disease and candidate genes.

6.1.3 Epidemiology

HD is the most common congenital intestinal motility disorder. The incidence varies between 1/2500 and 1/10000; the incidence is

approximately 1/4000 in European newborns (Suita S 2005, Best KE 2014).

In Asia the incidence is relatively high (approximately 1/2500) (Puri P 2008). The reasons for this difference re currently unclear. Males are at

(17)

17 increased risk for HD (sex ratio 4:1) and the sex ratio decreases as the aganglionosis segment becomes longer. Siblings have also an increased risk for HD (4%) compared to the general population incidence (0.02%).

The recurrence risk for siblings increases when the aganglionosis becomes more extensive (Badner 1990).

HD is associated with additional congenital anomalies in approximately 20% to 30% of HD cases. Chromosomal anomalies are involved in over 10% of HD cases, with trisomy 21 being the most common anomaly.

Despite the clear association between HD and trisomy 21, there is limited evidence on the genetic factors behind this association.

6.1.4 Etiology and pathophysiology 6.1.4.1 Genetic factors

Genetic disorders are currently identified in less than half of HD cases.

Modern analytical methods have led to the identification of additional disease genes, which have made it possible to confirm genetic disorders in an increasing number of HD patients. A genetic disorder causes a mainly isolated form of HD but some genes are associated with a syndromic form of HD. Genetic factors tend to play larger role in familial than in sporadic HD cases.

The number of identified genes involved in the pathogenesis of HD is increasing (Amiel J 2008, Borrego S 2013). The RET proto-oncogene is still the major gene associated with HD (Emison ES 2005, Emison ES 2010).

RET encodes a receptor tyrosine kinase. RET variants are more common in male than female HD patients. The penetrance of the variants is also incomplete (varying between 50% and 70%) and is higher in male patients (Attie T 1995). Causative coding sequence variants of RET explain

(18)

18 approximately 40% of familial and 20% of the sporadic cases (Emison ES 2005, Emison ES 2010). Common non-coding variants with low penetrance have also shown a clear association with HD. The most studied single nucleotide polymorphism (SNP) is rs2435357, which is located within a conserved enhancer element in the first intron of RET. The association has been confirmed in several studies (Gunadi 2014, Jiang Q 2015). The SNP rs2435357 seems to influence the expression of RET in the gastrointestinal tract (Gunadi 2014, Jiang Q 2015).

Despite greater possibilities in analytical methods, a genetic disorder in genes other than RET has only been found in a small proportion of HD cases. The RET signaling pathway involves interactions with many other genes, including ARTN, GDNF, GFRA1, NRTN, and PSPN (Salomon 1996 R,Sánchez-Mejías A 2010a, Ruiz-Ferrer M 2011) (Figure 1). RET is a receptor for GDNF ligands. For activation of RET, both GDNF and the coreceptor GFRA1 are required. In addition to GDNF, three other growth factor ligands of RET are involved in the RET signaling pathway, namely NRTN, PSPN, and ARTN. Damage to RET, GDNF, or GFRA1 cause colonization failure of the enteric neural crest-derived cells to the distal parts of the intestine. This disturbance in enteric nervous system (ENS) is thought to be a mixture of migration failure and cell apoptosis (Taraviras S 1999). GDNF variants are rare and only a few HD cases with GDNF variants have been reported in the literature (Salomon 1996 R). GFRA1 and NRTN variants have been more recently reported in HD patients (Sánchez-Mejías A 2010b). Most of these variants have been discovered together with additional RET variants.

(19)

19 Figure 1. Illustration of the genes involved in the RET signaling pathway (reproduced from Takhashi M 2001.Cytokine Growth Factor Rev. 2001;12:361-373 with permission from Elsevier)

The genes EDNRB, EDN3, and ECE1 interact in the EDNRB signaling transduction pathway and also play an important role in the pathogenesis of HD (Hofstra RM 1999, Sánchez-Mejías A 2010c). Variants that affect this pathway have been identified both in isolated and syndromic HD. The most commonly associated syndromes include congenital central hypoventilation syndrome (CCHS) and Waardenburg syndrome (WS). The interaction of all these genes is required for the successful transduction pathway. EDNRB transfers signals through EDN1, which is processed by ECE1 (Figure 2).

(20)

20 Damaged EDN3, ECE1, or EDNRB disturb the development of the enteric nervous system, thus causing the absence of enteric ganglia in the distal part of the intestine. This is caused by disturbed or failed migration, reproduction, survival, or differentiation of neural crest-derived precursors that form the neurons and glial cells of the enteric nervous system

(Bondurand N 2018). After RET variants, variants in EDNRB have been observed most frequently in HD patients and cover 5% of isolated HD cases (Sánchez-Mejías A 2010, Borrego S 2013). EDN3 and ECE1 variants have been rarely identified in HD patients (Hofstra RM 1999, Sánchez-Mejías A 2010b).

Figure 2. Illustration of the genes involved in the EDNRB signaling pathway (reproduced fromBondurand N 2018. Dev Biol. 2018 in press with permission from Elsevier)

(21)

21 Additionally, other signaling pathways influence the migration and

differentiation of neural crest cells. The genes NTF-3 and NTRK3

participate in the NTF-3/TRKC signaling pathway and are considered as candidate genes for HD (Fernández RM 2009, Ruiz-Ferrer M 2008).

PROK1, PROKR1, and PROKR2 are a part of the prokineticins signaling pathway and have a weak connection to HD (Ruiz-Ferrer M 2011). Other pathways involving in the development of the enteric nervous system include the NRG and SEMA signaling pathways. Variants in the genes NRG1, NRG3, SEMA3A, and SEMA3D also have a weak association with HD (Garcia-Barcelo 2009 MM, Wang LL 2011, Jiang Q 2012, Yang J 2013, Gunadi 2014). The remaining disease or candidate genes include

PHOX2B, SOX10, ZFHX1B, KIAA1279, L1CAM, DSCAM, BDNF, AXIN2, and TCF4 (Weese-Mayer DE 2002, Zweier C 2007, Gao H 2008,

Nakakimura S 2008, Sánchez-Mejías A 2010c, Jiang Q 2011, Fernández RM 2013, Jannot AS 2013). Currently 27 genes in the literature are reported to have an established or a suspected association with the development of HD (Table 1).

(22)

Table 1. Genes reported in the literature with an established or a suspected pathogenic role in the development of HD (Study III) GSS, Gerstmann-Sträussler-Scheinker syndrome; GWAS, Genome-wide association study GeneChromosome Evidence levelSequenced locusReferences RET10q11.2Knownwhole geneEmisson ES 2005 & 2010 ART1p34.1Possibleall exonsRuiz-Ferrer M 2011 GDNF5p12-p13.1Knownall exonsSalomon R 1996 GFRA110q25-q26Knownall exonsSánchez-Mejías A 2010a NRTN 19p13.3possible together with RETall exonsSánchez-Mejías A 2010a PSPN19p13.3Possibleall exonsRuiz-Ferrer M 2011 EDNRB13q22Knownall exonsSánchez-Mejías A 2010b EDN320q13.32Knownall exonsSánchez-Mejías A 2010b ECE11p36.1knownall exonsHofstra RM 1999 NTF3 12p13possibleall exonsRuiz-Ferrer M 2008 NTRK315q24-25possible with RET-splicing variants all exonsFernández RM 2009 PROKR12p14possible with RETand GDNF variants all exonsRuiz-Ferrer M 2011 PROK11p21possible with RETand GDNF variantsall exonsRuiz-Ferrer M 2011 PROKR220p12.3possible with RETand GDNF variantsall exonsRuiz-Ferrer M 2011 NRG18p12possible (GWAS) all exonsGarcia-Barcelo MM 2009 NRG310q23.1possible (exome sequencing)all exonsYang J 2013 SEMA3A7p12.1possible (GWAS) all exonsWang LL 2011 SEMA3D7q21.11possible (GWAS) all exonsJiang Q 2015 PHOX2B4p13knownall exonsFernández RM 2013 L1CAMXq28possible with X-linked hydrocephalusall exonsNakakimura S 2008 SOX1022q13.1knownall exonsSánchez-Mejías A 2010c KIAA127910q22.1possible with GSSall exonsJiang Q 2011 ZFHX1B2q22.3possible with Mowath Wilson syndrome all exonsJiang Q 2011 TCF4 18q21.1weakall exonsZweier C 2007 AXIN217q24.1weakall exonsGao H 2008 BDNF11p14.1weakall exonsWeese-Mayer DE 2002 DSCAM 21q22.2-q22.3possible (GWAS) all exonsJannot AS 2013

(23)

23 6.1.4.2 Non-genetic factors

Due to the complex inheritance pattern of HD, the incomplete penetrance of RET variants, and the small number of identified causative coding

sequence variants in other genes, it is unclear in how many HD patients a genetic disorder is the actual cause of the disease. This has led to

increasing interest in studies on non-genetic and epigenetic factors in the pathogenesis of HD (Heuckeroth RO 2016, Torroglosa A 2016).

The control of gene expression is complex and many different mechanisms are involved. These include mechanisms such as DNA modification

(epigenetic mechanism) and regulation of transcription (controlled by enhancers, transcription factors, repressors, and silencers). In addition, post-transcriptional regulation and regulation of translation play a key role (Torroglosa A 2016). For example, alternations in DNA methylation

regulates RET expression, which might also be a mechanism implicated in the onset of HD (Torroglosa A 2016).

In addition, some maternal risk factors have been identified. Obesity is the most significant risk factor for developing HD (Löf Granström A 2016).

6.1.4.3 Pathophysiology

The basic pathophysiological explanation for HD is an obstruction in the distal bowel. An aganglionic part of colon is narrowed, which prevents the transmission of peristaltic waves. This causes a dilated colon just before the aganglionic segment. The complete pathophysiology of HD is still not understood (Figure 3).

(24)

24 Figure 3. Basic pathophysiology and clinical symptoms of HD (reproduced from Heuckeroth RO 2018. Nat Rev Gastroenterol Hepatol. 2018;15:152-167 with permission from Springer Nature)

(25)

25 The ganglion cells are special nerve cells in the intestine and are

responsible for the peristaltic movement towards the rectum. Most of the intestinal neural cells are located in the myenteric plexus. A disturbed migration of the ganglion cells causes the absence of ganglion cells in the myenteric and submucosal plexuses. The myenteric plexus is responsible for motility, while the submucosal plexus controls the motility, blood flow, and transport of ions across the intestinal epithelium. The absence of ganglion cells in the myenteric plexus results in an elevated synaptic activity of the enzyme acetylcholinesterase. The increased activity of the enzyme lowers the activity of acetylcholine, which causes a decrease of nitric oxide. The decreased level of nitric oxide disturbs the relaxation of the smooth muscle in the endothelium of the distal colon. The disturbed

relaxation of the smooth muscles, mechanical obstruction, and disordered peristalsis are caused by the lack of ganglion cells and derangements of the enteric nervous system. This leads to the accumulation of stool and dilatation of the proximal colon (Butler Tjaden NE 2013).

There are also similar conditions that resemble HD with some degree of gastrointestinal tract malfunction. In most of the cases, ganglion cells are present but the amount and the distribution of the ganglion cells is

abnormal. While these conditions represent various conditions of the ENS, they could also affect similar developmental disorders which makes the diagnosis and treatment difficult (Moore SW 2017).

(26)

26 6.1.5 Diagnostics

6.1.5.1 Primary diagnostics

The diagnosis of HD is normally clear and the disease can be identified in most cases during the neonatal period. The most common symptom is the absence of meconium within the first 48 hours of life. Other symptoms include abdominal distension and vomiting due to intestinal obstruction and enterocolitis (Figure 3). In some milder forms of HD, the diagnosis is made later during childhood or even in adulthood based on symptoms such as severe constipation, chronic abdominal distension, vomiting, and failure to thrive (Parc R 1984).

If the physician has a clinical suspicion of HD, series of examinations can be performed to confirm the diagnosis. An abdominal X-ray is normally the first step. Contrast enema is used to assess the level of the transition zone between the ganglionic and the aganglionic colon. The aganglionic part can be identified in the image as a narrowed distal bowel segment with proximal dilatation. In most of the cases, the rectum appears small and the

contractions are uncoordinated. In the repetition of the imaging, evacuation of the contrast enema is observed (Proctor ML 2004).

Anorectal manometry is also a useful diagnostic tool. The normal rectoanal inhibitory reflex (RAIR) is absent in HD. RAIR can be identified in children with HD. A normal finding excludes the HD diagnosis, but the absence can be false positive and should be followed by a rectal biopsy. Anorectal manometry cannot be performed during first 2 weeks of life due to the absent normal rectoenteric reflex (Emir H 1999). Confirmation of the diagnosis must be performed by taking a rectal biopsy (Friedmacher F 2015). The area between healthy and aganglionic bowel must also be identified for the planning of the surgery. A series of biopsies that cover the

(27)

27 full thickness should be taken. This way the both the myenteric and the submucosal plexuses can be examined.

6.1.5.2 Differential diagnosis

HD is the most common reason for the failure to pass meconium within the first 48 hours after the birth. Other diseases have similar clinical findings, including meconium ileus caused by cystic fibrosis and functional intestinal obstruction caused by infection or intoxication of the mother. Some other ENS anomalies and intestinal malformations can also resemble HD.

6.1.5.3 Genetic testing

Genetic testing is currently not used for primary diagnostics. It is more useful as a tool for medical genetics and prevention. The early identification of high-risk familial HD patients would ensure more careful and effective treatment of HD after the birth. In the future, gene technical treatments or other treatments may be available already during pregnancy.

6.1.6 Treatment of HD

The only curative treatment for HD is surgery. HD is treated by surgical removal of the aganglionic part of the intestine (Neuvonen 2017).

Depending on the time of the diagnosis and the length of the aganglionic segment, there are a few surgical possibilities available. A one-stage procedure can be performed for young infants with early diagnosis. If the diagnosis is late and there is already colonic dilation, a primary colostomy is required. In the Swenson procedure, a part of the colon is resected and a coloanal anastomosis is performed (Swenson O 2004). The Duhamel procedure is an alternative approach and is executed by performing a retrorectal dissection and longitudinal stapled colorectal anastomosis. The

(28)

28 Soave procedure consists of an endorectal pull-through with mucosectomy (Soave F 1964). Recently, minimal access techniques, including

laparoscopic and transanal approaches, have also been developed.

Transanal pull-through procedures can be performed in most cases as a one-stage procedure. The need for intra-abdominal dissection can be avoided in this case. These three procedures are the most commonly performed operative treatments for HD. Most TCA cases are treated by modifications of the Duhamel and endorectal pull-through procedures (Yeh Y 2015). In some rare cases of near-total intestinal aganglionosis, an intestinal transplantation might be needed to avoid permanent parenteral nutrition (Pakarinen MP 2013).

Despite good surgical treatment of HD, disorders of anorectal and bowel function may appear after the operation (Neuvonen 2017). Short-term complications include leak or stenosis of the anastomosis and enterocolitis.

Enterocolitis can also be a long-term complication. The other most common long-term complications include chronic constipation and soiling. The most common cause of death in treated HD is unrecognized enterocolitis. If untreated, aganglionosis leads to intestinal obstruction, enterocolitis, malnutrition, and death. The mortality increases among HD patients with long-segment disease; mortality is highest among patients with TCA (Tsuji H 1999). The most recent estimates for the mortality rate of HD varies between 1% and 10% (Pini Prato A 2013).

(29)

29 6.1.7 Associated syndromes and anomalies

The neural crest gives rise to other neuronal, endocrine, paraendocrine, craniofacial, conotruncal heart, and pigmentary tissues; this explains the increased incidence of other neurocristopathies in HD. HD-associated syndromes and anomalies include familial medullary thyroid carcinoma (FMTC), multiple endocrine neoplasia (MEN) syndromes, neuroblastoma, congenital central hypoventilation syndrome (CCHS), WS, Mowat-Wilson syndrome, cartilage hair hypoplasia, and conotruncal heart defects (Le Douarin N 1999).

The MEN 2 syndromes consist of three types of cancer predisposition:

FMTC, MEN 2A, and MEN 2B. Patients with MEN 2A typically have MTC (70% by the age of 70 years), phaeochromocytoma (half of the cases), and hyperplasia of the parathyroidglands (15-35% of the cases). MEN 2B patients additionally have oral neuromas, marfanoid habitus, and hyperganglionosis of the hindgut (Amiel J 2001).

Neural crest-derived tumors (such as neuroblastoma) are rare conditions.

CCHS is also a rare syndrome and in some cases presents together with neuroblastoma and other neural crest-derived tumors. The main symptom of CCHS is disturbed response to hypoxia and hypercapnia. Similar to HD, most of the identified variants have been discovered in genes belonging to the RET and EDNRB signaling pathways. These genes include RET, GDNF, and EDN3 (Amiel J 1996).

Similar to HD, WS is characterized by diverse genotypes and phenotypes.

In some rare cases they have been observed together (Attie T 1995, Pingault V 1998). The identified variants have been observed in SOX10 and genes involving the EDNRB pathway (Attie T 1995, Pingault V 1998).

(30)

30 In addition, many chromosomal anomalies have been observed in HD patients. Down syndrome is the most frequently encountered chromosomal anomaly and is associated with 2% to 10% of ascertained HD cases

(Garver KL 1985). Despite extensive genetic screening of chromosome 21, only a few variants have been identified, mainly in the DSCAM gene

(Jannot AS 2013). Other chromosomal anomalies known to be associated with HD include DiGeorge syndrome, tetrasomy 9p, XXY chromosomal constitution, partial duplication of chromosome 2q, mosaic trisomy 8, chromosome 13 microdeletion, deletion of the long arm of the chromosome 10 and 20p deletion (Bottani A 1991, Lurie IW 1994, Martucciello G 1998, Mowat DR 1998, Amiel 2001). Apart from Down syndrome, the association of these anomalies with HD is extremely rare. In these extremely rare cases it is also possible that the co-occurrence may just be a coincidence without any reliable evidence.

(31)

31

6.2 HD-associated medullary thyroid carcinoma (MTC)

6.2.1 Epidemiology of MTC in HD

There is minimal information on the incidence of MTC in HD. The most recent studies have estimated that the incidence of HD-associated MTC varies between 2.5% to 5% (Decker RA 1998, Sijmons RH 1998, Pakarinen MP 2005). MTC occurs normally in middle-aged patients. In many cases the association between MTC and HD might be missed and not reported;

estimation of the incidence is thus difficult.

6.2.2 Etiology of HD-associated MTC

The RET proto-oncogene has a significant association with both HD and MEN 2 syndromes (Mulligan LM 1993, Donis-Keller H 1993, Hofstra RM 1994, Carlson KM 1994, Attié T 1995). It is interesting how a dysfunction of one gene can cause such different diseases. Loss-of-function variants of the RET gene are responsible for HD, whereas gain-of-function variants lead to MEN 2 syndromes (Arighi E 2004).

Only a few variants are known to be responsible for HD-associated MEN 2A. The focus has been in RET exon 10, of which most of the variants are in one of four codons (609, 611, 618, and 620) (Mulligan LM 1994a, Nishikawa M 2003, Romeo G 1998, Sijmons RH 1998). MEN 2A patients without HD have an additional locus for high-risk variants in exon 11,

particularly in codons 630 and 634 (Donis-Keller H 1993, Mulligan LM 1993, Mulligan LM 1994b, Hansford JR 2000). Rare variants in codon 791 of the RET gene have also shown a weak association with MEN 2A and FMTC (Frank-Raue K 2008, Erlic Z 2010).

(32)

32 6.2.3 Diagnostics of MTC in HD

6.2.3.1 Genetic testing

HD-associated MTC is an extremely rare condition and it is challenging to identify high-risk patients for MTC among HD patients. Genetic screening for MTC-associated variants has not been standardized and there are currently no clear recommendations if genetic analysis should be performed for all HD patients or only for selected patients. Most studies have

examined RET variants among selected HD or MTC patients (or both) or their families (Nishikawa M 2003, De Groot JW 2005, Bütter A 2007, Fialkowski EA 2008). Provided that any phenotype subgroup cannot be excluded, screening all HD patients would be the only way to identify all MTC-associated variants. Several studies have confirmed that almost all MTC-associated variants in HD patients are in RET exon 10 and 13, which makes genetic testing reliable and inexpensive.

6.2.3.2 Blood tests

The only reliable marker in the blood for MTC is serum calcitonin. The test must be controlled in susceptible cases with elevated serum calcitonin.

Pentagastrin-stimulated measurement of serum calcitonin is also another possibility among patients with suspected MTC without highly elevated serum calcitonin. In the very early stage of the disease, the most sensitive method to detect MTC is the pentagastrin-stimulated calcitonin

measurement (Roque M 1999). This method is also more sensitive than thyroid ultrasound (US).

6.2.3.3 Thyroid ultrasound and fine needle biopsy (FNB)

Thyroid US and FNB are the gold standards as a primary diagnostic tool for any thyroidal carcinoma and management that requires a pathologist and a radiologist. Wide availability, non-ionizing radiation, and the combination of

(33)

33 FNB make thyroid US a widely used and reliable diagnostic tool for

evaluating thyroid nodules (Wong KT 2005).

6.2.4 Prevention

6.2.4.1 Screening among HD patients

In previous studies, screening for MTC or MTC-associated variants among HD patients was focused only on genetic testing. Therefore, the selection of patients for screening was similar as mentioned above (Nishikawa M 2003, De Groot JW 2005, Bütter A 2007, Fialkowski EA 2008). Considering the complexity and diversity of the genetics of HD, a selection of patients for screening may exclude some high-risk patients.

6.2.4.2 Follow-up

There is no standard follow-up protocol for MTC among HD patients with an identified MTC-associated variant. In most cases, follow-up is individually planned according to several factors, including patient age at identification of the MTC-associated variant and family history of MTC. In some cases, a prophylactic thyroidectomy is recommended, whereas in some cases routine US controls and life-long calcitonin screening are preferred (Frank- Raue K 2008, Frank-Raue K 2011).

(34)

34

7 AIMS OF THE STUDY

The aim of this doctoral thesis was to identify HD-associated variants in Finnish HD patients and how genotypes affect diverse HD phenotypes, including extension of aganglionosis, heredity, and HD-associated MTC.

(35)

35

8 PATIENTS AND METHODS

8.1 Patients and study design

This doctoral thesis consists of two main parts. In the first part, a thyroid cancer screening study was performed on volunteer adult HD patients. In the second stage of the study, additional HD patients (including children) participated in the study and diverse gene analyses were performed on all existing HD patients to obtain further information on the genetic etiology of HD.

The study protocol was approved by the Ethics committee of the Hospital for Children and Adolescents (Statement of ethics committee 228_E7_03).

In addition, the study protocol of study II was approved by the Ethics committee of the Pirkanmaa Hospital District and by the Ethics Committee at Karolinska Institute, Stockholm.

8.1.1 Study I

The first study was a population-based screening study of thyroid cancer among adult HD patients. Participation in the screening study was offered to all eligible HD patients who had been operated on for HD at the Children’s Hospital, Helsinki University Central Hospital from 1950 to 1986, which covers most of the HD cases in Finland during that period. To determine eligibility, hospital records were searched and all HD patients alive after 1 January 1967 were enrolled. Since 1967, the Population Register Center of Finland provides all Finnish residents with a personal identity (ID) code. Each ID code from all potential participants was validated by comparing the individuals with those listed under the Population Register Center of

(36)

36 Finland. The Cause-of-death register of Statistics Finland was used

for verifying the vital status of the individuals. Possible cancer cases of the individuals were validated until 31 December 2010 by reviewing the Finnish Cancer Registry, which maintains complete records of all cancer cases in Finland since 1953.

A total of 156 HD patients were verified. Five had moved abroad and 11 died from causes other than thyroid cancer, leaving 140 eligible HD patients. The database of the Population Register Centre was used to track the patients and the patients were then contacted by mail during the years 2007 to 2009. Of 140 patients, 91 were willing to participate in the study and a cross-sectional study was performed for the

participants during their outpatient visit. No exact reason was noted for non-participation.

8.1.2 Study II-III

In the second part of the study, an additional 21 HD patients (including children) volunteered to participate in the genetics part of the study, where diverse gene analyses were performed on all existing HD patients,

including 91 adult HD patients from the screening study. Parental assent was provided for minor patients. Genetic analyses were performed between 2010 to 2014.

8.2 Controls (I-III)

The first study was a screening study and therefore a control group was not necessary. During the genetics part of the study, several different control groups were used. In whole exome sequencing, the results were compared to seven healthy controls and to public sequence databases (1000

Genomes Project, NCBI dbSNP). For sequencing all exons of RET, five

(37)

37 healthy first-degree relatives from familial HD patients were used as

controls. In GWAS, 386 healthy representatives of the general population from the Child Sleep Study were chosen for a control group. In the same study, the replication of previously identified variants in and near SEMA and NRG1 and the replication of our identified genome-wide significantly

associated variants was performed as a co-operation with Karolinska Institute. The replication population included 154 HD patients and 177 healthy Caucasian controls. For the third study, one healthy control sample was chosen for targeted resequencing.

8.3 Methods

In the first stage of the study, the screening for thyroid cancer among adult HD patients included several different examinations. Later, in the genetic part of the doctoral thesis, the material was analyzed comprehensively using four different complementary methods. Capillary sequencing of RET exons and targeted sequencing of all known HD genes were focused on the location of the genome according to previous studies associated with the pathogenesis of HD. In addition, GWAS and whole exome sequencing were performed to obtain genome-wide information on the genetics of HD.

8.3.1 Study I

In the screening study, all participants (91/140, 65%) completed a cross- sectional study where genetic testing was combined with a serum calcitonin measurement and thyroid gland US. FNB was also performed if necessary.

In addition, cancer history was validated from the Finnish Cancer Registry.

The Finnish Cancer Registry is a national organization where all new cancer cases with accompanying patient ID from every medical facility in Finland is registered. As Finland has a small population and well-organized

(38)

38 health care, this means practically complete coverage of the population with high accuracy (Teppo L 1994, Pukkala E 2011). The cancer history of all 156 HD patients was validated automatically using ID codes. The start of the follow up was set at the date of the patient’s birth or 1 January 1967, whichever was later. The end of the follow up was set on the patient’s death or 31 December 2010, whichever occurred first. As some patients were born before 1 January 1967, the cancer history from these older patients was validated by reviewing medical records from 1950 to 1967. In this way, all MTC cases were captured considering the possible early onset of the disease.

All US examinations were performed by a radiologist. The examination included US of the whole neck area with a focus on thyroid gland

morphology and the surrounding lymph nodes. The radiologist performed a FNB from any solid or cystic solid hypoechoic or isoechoic solitary lesions in the thyroid gland or from extra-glandular focal lesions with a low

threshold. Findings such as simple cysts or typical multiple goiter nodules with a sonolucent halo sign were not an indication for FNB. If the first US examination and the cytology were not confirmed as a benign lesion, repeat US examination with or without FNB was indicated.

Standard hospital laboratory methods were used for measurement of serum calcitonin (https://huslab.fi/ohjekirja/2008.html). These included an

overnight fast and upper normal limit of 1.7 pmol/l for women and 3.4 pmol/l for men. If the serum calcitonin level was elevated or if there were technical difficulties with the analysis, the measurement was repeated using the same standard methods. As an alternative method for repetition of standard serum calcitonin measurement, a pentagastrin-stimulated measurement of serum calcitonin was used for patients with slightly elevated serum

calcitonin or with clinical suspicion of MTC. Standard hospital laboratory

(39)

39 methods were used for the pentagastrin-stimulated measurement of serum calcitonin (https://huslab.fi/ohjekirja/3719.html).

For sequencing data, 30-ml whole blood samples collected in EDTA tubes were drawn from each HD patient and DNA was extracted from the

samples. The RET cDNA from selected loci was PCR amplified and nucleotide sequences were resolved with the ABI3730xl DNA Analyzer capillary electrophoresis instrument (Applied Biosystems) using standard methods. Considering previous knowledge on the location of MTC- associated variants, exons 10, 11, 13, and 16 were sequenced from all participants as a standard screening. After completing and analyzing all screening examinations, 43 patients were chosen for sequencing all remaining exons of the RET gene. Criteria for the additional sequencing included an identified thyroid cancer, an identified RET variant in the first sequencing, suspicious clinical findings, familial or long-segment disease, and a group of randomly selected patients.

8.3.2 Study II-III

After thyroid cancer screening was performed, there were an additional 21 HD patients who participated in the study. In the screening study, RET exon sequencing was performed partially incomplete, meaning that all exons of RET had not been sequenced from all HD patients. Therefore, as an outset of the gene studies, all exons of RET were sequenced from all existing HD patients. All exons of the RET gene were PCR amplified and the nucleotide sequences were resolved with ABI3730xl DNA Analyzer capillary

electrophoresis instrument.

In the next stage, eight HD cases were chosen for whole exome

sequencing. Six were familial HD patients from four different families and

(40)

40 two sporadic HD patients with long-segment disease. DNA samples were fragmented and enriched using the NimbleGen SeqCap EZ target kit and sequenced with Illumina Genome Analyzer IIx sequencer. The results of the whole exome sequencing were compared to seven healthy controls and to public sequence databases (1000 Genomes Project, NCBI dbSNP). In this way, a group of new candidate genes were identified where the identified rare coding sequence variants have been observed only in HD patients.

To support the sequencing-based methods, a GWAS was performed. The samples of all 112 HD patients, their healthy relatives participating in the study, and 386 controls from the Child Sleep Study were genotyped using the Illumina HumanOmniExpress BeadChip using standard protocols at the Core Facility of the Estonian Genome Center, University of Tartu, Estonia.

Human reference GRCh37/hg19 was used as the reference genome. SNPs and individuals with genotyping rate <95% and SNPs with P < 5 × 10^-3 were removed for an exact test of Hardy-Weinberg equilibrium in controls, SNPs with P < 0.05 for a test of differential missingness between cases and controls (PLINK --test-missing), and SNPs violating Mendelian inheritance in any of the genotyped pedigrees. The data was then re-phased and imputated with shapeIT v2.r644 and Impute v2.3.0 using the 1000 Genomes Project (1000G) integrated phase1 v3 reference haplotypes.

Eventually, SNPs with minor allele frequency (MAF) < 0.01 or Impute “info”

metric < 0.9 were removed.

In the last stage of the study, targeted sequencing was performed for those HD patients with no known causative variant in our previous studies. The literature of the genetics of HD was carefully searched and all genes (n=27) that are connected to the pathogenesis of either isolated or syndromic HD were chosen for sequencing. Even susceptible genes with weak evidence

(41)

41 were chosen. The main consideration was in coding sequence variants and therefore only exons with their three flanking base pairs flanking the exons and the promoter region (transcription start site ±250 bp) from all these genes were sequenced. In addition, the whole RET gene was sequenced, including the entire transcribed region and the flanking 1000 base pairs next to the transcription start and end sites (chr10:43 571 474-43 626 799) to cover possible major frameshift variants in the main gene of HD. The DNA from the targeted regions was enriched and sequenced as 100-bp paired reads using the Agilent HaloPlex platform and an Illumina GA IIx sequencer.

Variant classification

Throughout the whole study identified variants were classified into five categories based on their predicted effect on gene function using the Ensembl Variant Effect Predictor

(http://grch37.ensembl.org/info/docs/tools/vep/index.html): affects function, probably affects function, unknown, probably does not affect function, and does not affect function. In addition, the effects of amino acid substitutions were predicted using SIFT (http://provean.jcvi.org/index.php) (either tolerated or deleterious) and PolyPhen2

(http://genetics.bwh.harvard.edu/pph2/) (probably damaging, possibly damaging, benign, or unknown). The difference in the allele frequency between cases and controls was also used in the classification. For the variants identified by sequencing HD cases, the allele frequencies in the general (control) population were obtained from the Exome Aggregation Consortium database (ExAC; http://exac.broadinstitute.org) and the allele frequencies were compared to both Finnish and non-Finnish Europeans.

(42)

42 Variants were classified based on MAF in the general population as

common (MAF > 5%), low-frequency (MAF 1 to 5 %), or rare (MAF < 1 %).

(43)

43

8.4 Statistical analyses

8.4.1 Study I

In the first study, the standardized incidence ratio (SIR) was calculated for all types of thyroid cancer and MTC alone. For calculation of SIR, the observed number of cancer cases were divided with the expected number of cases. The amount of observed cancer cases among all HD patients was calculated by 5-year age groups, separately for men and women and four calendar periods (1967-1977, 1978-1988, 1989-1999, and 2000-2010), and for the number of person-years at risk. For the calculation of the expected number of cancer cases, the number of person-years in each stratum was multiplied by the corresponding cancer incidence in Finland. Exact 95% CIs were defined on the assumption that the data followed a Poisson

distribution.

8.4.2 Study II

The ABI Variant Reporter and Sequencher software packages were used for analyzing capillary sequencing traces. PLINK v1.90b3w was used for data management and association testing. The association between genetic variants and disease status was tested using all three genetic models (additive, recessive, and dominant) with a logistic regression model for variants with MAF > 0.10 and Fisher’s exact test for the rarer variants.

The P-value limits for statistical significance was set at 5 × 10^-8 for GWAS and at 0.05 for the replication. Population attributable fractions (PAF), or the proportion of the disease in the population that is attributable to exposure, were calculated using the R Epi package.

(44)

44 The genotype relative risk (GRR) was calculated by first calculating

penetrances by setting the population prevalence of the disease at 1/5000 and then using Bayes’ theorem for the calculated penetrances using the reported genotype frequencies. GRR was then calculated using the least penetrant genotype class as reference. The same principle has been used in previous studies (Rybicki B 2000).

8.4.3 Study III

Reads against the GRCh37 human reference genome were aligned using BWA v 0.6.2. Orphan reads were aligned to multiple genomic positions and the remaining sequence reads near possible insertions and deletions were realigned using GATK Lite v2.2.16. The resulting realigned sequence reads were used to detect variant sites and call genotypes in all sequenced samples simultaneously with the Platypus v 0.7.9 variant caller. Variants were annotated using the Ensembl Variant Effect Predictor with the Ensembl database v75.

(45)

45

9 RESULTS

9.1 Patient characteristics

The initial study population included 91 adult HD patients. In different stages of further genetic studies, a total of 21 additional HD patients (including children) participated in the study. This explains the small proportion of underage patients in the study. From the whole study population of 112 HD patients, there were 36 familial and 76 sporadic cases. Of the 36 familial patients, 31 had an affected first degree relative from 14 different families.

There was a statistically significant difference in the length of aganglionosis between familial and sporadic HD patients. Long-segment disease

(aganglionosis extending proximal to the sigmoid colon) and total colon aganglionosis were more common in familial patients (17% vs 4%, P=0.022 for long segment; 11% vs 0%, P=0.007 for total colon aganglionosis) (Table 2).

Common health issues and diseases such as asthma, diabetes mellitus, and arterial hypertension were evenly divided among HD patients and were present as often as in the healthy population. Congenital or genetic

disorders, or both, included Down syndrome in one sporadic female patient with short-segment disease, cartilage hair hypoplasia in one sporadic male patient with short-segment disease, paraplegia and deafness in one

sporadic female patient with short-segment disease, and Marfan syndrome with blindness and operation-treated aortic valve regurgitation in one

sporadic male patient with short-segment disease. One familial male patient with short-segment disease and his daughter with long-segment disease had central hypoventilation syndrome. During the study, the daughter was

(46)

46 also diagnosed with neuroblastoma and tumor in the suprarenal gland. In addition, one patient had been treated for medullary thyroid carcinoma and one for MEN2A.

Table 2. Patient characteristics

1 Range

2 Median

HD, Hirschsprung disease; TCA, total colon aganglionosis

Patient characteristics n=112

Mean age, years 38.5 (1-67.7)¹, (43.5)² Gender

Male 88 (78.6%)

Female 24 (21.4%)

Level of aganglionosis Rectosigmoid Long-segment

96 (85.7%)

9 (8.0%) TCA

Unclear

4 (3.6%)

3 (2.7%) Familial HD

Yes 36 (32.1%)

No 76 (67.9%)

Family history of thyroid cancers

Yes 3 (2.7%)

No 99 (88.4%)

Missing 10 (8.9%)

Associated anomalies and syndromes

Syndromic association 6 (5.4%)

Isolated anomaly 2 (1.8%)

(47)

47

9.2 HD-associated medullary thyroid carcinoma (I)

The screening study of MTC included the cancer survey through the Finnish Cancer Registry for the whole study cohort (n=156) and a cross- sectional study for all participants (91/140, 65%). The cross-sectional study included clinical examinations including measurement of serum calcitonin and thyroid gland US (combined with FNB if necessary). In addition, genetic analyses were performed to identify coding sequence variants in RET associated with MTC (Figure 4).

Figure 4. Screening study for MTC

*This PTC case was entered into the Finnish Cancer Registry after the discovery in the screening study

US, ultrasound; MTC, medullary thyroid carcinoma; PTC, papillary thyroid carcinoma