pituitary stalk interruption syndrome in 83 patients: novel hesx1

Pituitary Stalk Interruption Syndrome in 83 patients: novel HESX1 mutation and severe 1

hormonal prognosis in malformative forms. 2

Reynaud R.1, 5

(MD, PhD)*, Albarel F.2, 5

(MD)*, Saveanu A.2, 3, 5

(MD, PhD), Kaffel N.6

(MD), 3

Castinetti F.2, 5

(MD), Lecomte P.7

(MD), Brauner R.8

(MD, PhD), Simonin G.1

(MD), Gaudart J.4,9 4

(MD, PhD), Carmona E. 3, Enjalbert A.

3, 5 (PhD), Barlier A.

2, 3, 5 (MD, PhD), Brue T.

2, 5 (MD, PhD). 5

*both authors contributed equally to this work 6

1Department of Pediatrics , Pediatric Endocrinology Unit, and

2Department of Endocrinology, Hôpital 7

de la Timone, Centre de Référence des Maladies Rares d’Origine Hypophysaire ; 3Laboratory of 8

Biochemistry and Molecular Biology, Hôpital de la Conception, 4Service de Santé Publique et 9

Information Médicale, Hôpital de la Timone, Assistance Publique-Hôpitaux de Marseille, 13385 10

Marseille, 5

Centre de Recherche en Neurobiologie et Neurophysiologie de Marseille (CRN2M), 11

Centre National de la Recherche Scientifique, Unité Mixte de Recherche 6231, Institut Fédératif Jean-12

Roche, Faculté de Médecine de Marseille, Université de la Méditerranée, Marseille, France. 13

6Department of Endocrinology and Diabetology, Centre Hospitalier Universitaire H. Chaker, Sfax, 14

Tunisia, 7

Departement of Endocrinology, Centre Hospitalier Régional Universitaire Bretonneau, 15

Tours, France, 8

Pediatric Endocrinology Unit, Université Paris-Descartes and Assistance Publique –16

Hôpitaux de Paris, Hôpital Bicêtre, Le Kremlin Bicêtre, 94270, France. 9

Laboratoire d'Enseignement 17

et de Recherche sur le Traitement de l'Information Médicale, EA 3283, Faculté de Médecine Unité 18

Fonctionnelle de Biostatistique et Méthodologie de la Recherche Clinique, Faculté de Médecine de 19

Marseille, Université de la Méditerranée, , Marseille, France 20

Corresponding author: Dr Rachel Reynaud Hôpital de la Timone Enfant, Unité d'Endocrinologie 21

pédiatrique, 264 rue St Pierre, 13385 Marseille Cedex 05, France; [email protected]; Phone: + 22

33 4 91 38 71 13, Fax: + 33 4 91 38 43 11 23

24

Key words: Hypopituitarism; LHX4; OTX2; SOX3; HESX1; transcription factor; growth hormone 25

deficiency; pituitary development; Ectopic Posterior Pituitary; CPHD; MPHD 26

27

28

of 24 Accepted Preprint first posted on 26 January 2011 as Manuscript EJE-10-0892

Copyright © 2011 European Society of Endocrinology.

mailto:[email protected]

2

Short running title: Pituitary Stalk Interruption Syndrome 29

30

Word counts: Abstract : 197 ; 31

Full text (abstract, legends, acknowledgments, list of abbreviations included): 4076 32

of 24

3

ABSTRACT: Pituitary stalk interruption syndrome (PSIS) is a particular entity in the population of 33

patients with hypopituitarism. Only rare cases have a known genetic cause. 34

Objectives: (1) To compare subgroups with or without extra-pituitary malformations in a cohort of 35

PSIS patients to identify predictive factors of evolution, (2) to determine the incidence of mutations of 36

the known pituitary transcription factor genes in PSIS. 37

Study design: We analyzed features of 83 PSIS patients from 80 pedigrees and screened HESX1, 38

LHX4, OTX2, and SOX3 genes. 39

Results: PSIS had a male predominance and was rarely familial (5%). Pituitary hypoplasia was 40

observed only in the group with extra-pituitary malformations. Multiple hormone deficits were 41

observed significantly more often with vs. without extra-pituitary malformations (87.5% vs. 69.5%, 42

respectively)). Posterior pituitary location along the stalk was a significant protective factor regarding 43

severity of hormonal phenotype. A novel HESX1 causative mutation was found in a consanguineous 44

family and two LHX4 mutations were present in familial PSIS. 45

Conclusion: PSIS patients with extra-pituitary malformations had a more severe hormonal disorder 46

and pituitary imaging status, suggesting an antenatal origin. HESX1 or LHX4 mutations accounted for 47

less than 5% of cases and were found in consanguineous or familial cases. 48

49

50

51

52

53

54

55

56

57

58

59

60

of 24

4

Pituitary stalk interruption syndrome (PSIS) is characterized by the presence of a thin or absent 61

pituitary stalk. It is commonly associated with a hypoplastic or aplastic anterior pituitary or an ectopic 62

posterior pituitary. PSIS was identified by magnetic resonance imaging (MRI), which provides precise 63

visualization of abnormalities in the hypothalamic and pituitary regions (1). This anatomical 64

abnormality can be associated with other midline abnormalities and variable endocrine disorders, 65

ranging from isolated GH deficiency (IGHD) to combined pituitary hormone deficiency (CPHD). The 66

endocrine outcome seems to be a progressive onset of hormone deficiencies leading to 67

panhypopituitarism (2); posterior pituitary function is usually maintained (3). 68

The causes of PSIS are still unknown. Former theories implicated traumatic birth, considering the high 69

rate of perinatal events in this population (4). More recent studies favor organogenesis defects that 70

may be of genetic origin or due to environmental factors during pregnancy (2, 5, 6). Over the last 71

decade, mutations of transcriptional factor genes involved in pituitary development have been reported 72

in human congenital hypopituitarism (7). Genetic screening for GH-N, GHRH receptor, POU1F1, 73

LHX3, or PROP1 in several cohorts of patients with PSIS failed to reveal any mutation (3, 8-12).Rare 74

mutations of HESX1, LHX4, and more recently OTX2 or SOX3 genes have been reported in PSIS (see 75

review (7)). HESX1 mutations are responsible for a wide spectrum of clinical features, from IGHD 76

with no midline forebrain or optic defect to midline brain defects as part of septo-optic dysplasia 77

(SOD) (13-16). LHX4 mutations have been reported to cause CPHD with variable neuroradiological 78

abnormalities: PSIS, sella turcica or pituitary hypoplasia, ectopic posterior lobe, defect of the corpus 79

callosum and hindbrain (17-19). OTX2 gene defects, first observed in association with eye 80

malformations and later with variable pituitary disorders, were recently reported in three PSIS patients 81

with no ocular abnormalities(20, 21). Both over- or under-dosage of SOX3 may be associated with X-82

linked hypopituitarism, PSIS, and mental retardation, but two cases of SOX3 abnormality have been 83

reported without mental delay (22, 23). Genetic mutations therefore seemed to be associated with extra 84

pituitary malformations (EPM). In this study, we analyzed clinical, hormonal, and neuroradiological 85

features and conducted molecular analyses of HESX1, LHX4, OTX2, and SOX3 in a cohort of PSIS 86

patients to define prognostic factors of the endocrine outcome, and to determine the frequency of 87

genetic origin of this syndrome. We found that patients with extra-pituitary malformations (EPM+) 88

of 24

5

more frequently harbored a severe hormonal and imaging status than those without EPM (EPM-) and 89

we identified LHX4 and HESX1 mutations only in familial or consanguineous cases. 90

91

Subjects and Methods 92

Patients 93

Eighty-three patients were recruited from both national and international centers between 1996 and 94

2007: 69 patients through the GENHYPOPIT network, a multicentric study in pediatric and adult 95

endocrinology centers for the screening of pituitary genetic determinants of CPHD, and 14 additional 96

patients for whom DNA was not available. All patients presented with PSIS, defined on the basis 97

of MRI findings showing a thin or absent pituitary stalk and no normal posterior lobe 98

hypersignal in the sella turcica, and all had at least one anterior pituitary hormone deficiency, 99

including GHD as defined below. The clinical features were collected by use of a specific clinical 100

report form. Detailed phenotypic characterization of these 83 patients (58 males) is shown in Table 1. 101

Genomic DNA was screened from 69 patients for LHX4, HESX1, and OTX2 genes, and in males for 102

SOX3. Written informed consent from all patients or parents of minors was obtained for this study 103

approved by our institutional ethics committee. 104

Two groups were determined according to the presence (EPM+ group: 24 patients) or absence (EPM- 105

group: 59 patients) of extra-pituitary malformations, whether central nervous system (CNS), ocular, 106

dental, craniofacial, and/or cardiac. Cases with a family history of IGHD or CPHD with at least one 107

PSIS case among first-degree relatives were considered as familial cases. 108

Clinical studies 109

Neonatal features and description of physical malformations were provided by the referring medical 110

centers and interpreted according to gestational age, delivery conditions, and presence or absence of 111

neonatal distress. Birth weight and length were expressed as SDS for gestational age and sex, and 112

patients were defined as small for gestation age (SGA) when birth weight and/or length was below -2 113

standard deviation score (SDS) (8). Breech presentation and neonatal hypoxemia (defined by an Apgar 114

score <7 at 5 minutes after birth or need for neonatal resuscitation) were recorded. Height at diagnosis 115

of 24

6

was reported by each referring center and expressed in standard deviation score (SDS) according to 116

Sempé et al. (24). 117

Hormonal studies were performed in all index patients, in each referring medical center. Normal 118

values for each center were taken into account. The plasma GH response was studied with at least two 119

provocative tests in each patient: insulin intolerance test (0.05 U/kg), GHRH infusion test (80 µg, 120

Somatoreline; Choay/Sanofi, Gentilly, France) or propanolol-glucagon test (0.25 mg/kg propanolol 121

orally and 1 mg glucagon, im). A diagnosis of TSH deficiency (TSHD) was made if serum T4 122

concentration was subnormal (free T4 < 12.0 pmol/liter or total T4 < 65 nmol/liter) with an 123

inappropriately low serum TSH concentration (<5 µU/ml). Basal plasma ACTH and cortisol were 124

measured at 0800 h (normal range for cortisol: 210–560 nmol/liter). The plasma ACTH and cortisol 125

response was also determined during an insulin tolerance test. Gonadotroph axis was investigated only 126

in patients of postpubertal age,i.e. over 15 yr for female and 17 yr in male subjects. FSH-LH 127

deficiency was diagnosed on the basis of delayed or absent pubertal development with low serum 128

testosterone or estradiol levels and blunted LH/FSH response to a GnRH stimulation test. Prolactin 129

deficiency and hyperprolactinemia were defined as basal plasma PRL levels lower than 5 ng/mL and 130

above 25 ng/mL, respectively. CPHD was defined as GHD associated with at least one other anterior 131

pituitary hormone deficiency. 132

Pituitary MRI with gadolinium injection was performed in all patients, in each referring medical 133

center according to standardized procedures, using precontrast sagittal and coronal spin echo T1-134

weighted images followed by post-gadolinium T1-weighted imaging (25). Pituitary height was 135

measured and compared with normal value for age (25). The pituitary stalk was considered “thin” 136

when it had a continuous but extremely thin appearance with below normal size (1), or “interrupted” 137

when discontinuous, or “invisible” when not observed on any section. The location of the ectopic 138

posterior pituitary lobe was described either as localized at the median eminence, along the pituitary 139

stalk, or not visible. Associated abnormalities of the brain, cerebellum, optic nerves, optic chiasm, 140

sella turcica, or midline structures were also systematically sought. MRI films were not centrally read. 141

Molecular analysis of transcription factor genes 142

of 24

7

Genomic analysis of HESX1, LHX4, OTX2 and SOX3 was performed by direct sequencing. Genomic 143

DNA was extracted from peripheral blood lymphocytes. The coding regions and intron-exon 144

boundaries of HESX1, LHX4, and OTX2 as well as the single exon of SOX3 were amplified by 145

polymerase chain reaction (PCR) (Kit HotStarTaq DNA polymerase, Qiagen, Paris, France) using 146

exon-flanking primers (published as supplemental data on web site at http://eje.org). The same primers 147

were used for sequencing, with analysis on an AB3130xl (Applied Biosystem, Foster City, California, 148

USA). 149

Statistical analyses 150

Results are expressed as mean ± SD for age and bone age or in percentages. Univariate comparisons 151

between groups were performed by one-tailed Student’s unpaired t-test (or Wilcoxon test if necessary) 152

or when appropriate by one-tailed Pearson’s chi2 test (or Fischer exact test if necessary). Multivariate 153

studies were also performed as logistic regression (with descendant stepwise analysis based on 154

likelihood ratio). P values were considered significant when <0.05. 155

156

RESULTS 157

General features of PSIS patients (Table 1): 158

Eighty-three patients were investigated from clinical centers based in 7 countries (France, n= 67; 159

Tunisia, n=8; Turkey, n=3; Algeria, n=2; Argentina, n=1; Egypt, n=1; and Lebanon, n=1). Sex ratio 160

was 2.3 (58 males/25 females). Nine index cases (11.25%) were born from consanguineous parents, 161

and 4 index cases (5%) had a familial history of CPHD. 162

Mean age at diagnosis was 9.6 ± 8.8 years (1 day-39 years), with a height retardation of -3.5 ± 1 SDS. 163

Eighteen patients were diagnosed before or at 2 years of age and 5 patients were diagnosed at adult 164

age (>18 years). The mean bone age at diagnosis was 5.2 ± 3.6 years, with a bone age retardation of -165

3.2 ± 2.3 years. Patients were referred for growth retardation (86.8%), hypoglycemia (15.8%), 166

jaundice (5.3%), and/or salt wasting (1.3%). Mean age at the study analysis was 21.8 ± 11.6 years. All 167

of the PSIS patients had GHD, 79.5% had deficiency in TSH, 67.5% in ACTH, 65.1% in LH/FSH, 168

14.5% in PRL, and 16.9% had hyperprolactinemia. IGHD was found in only 6% of patients. Among 169

of 24

8

the 58 males, 23 presented micropenis (39.7%) and 6 had cryptorchidism (10.3%), presumably related 170

to GH and/or LH/FSH deficiency. 171

Comparison of phenotype features in patients with (EPM+) or without (EPM-) extra-pituitary 172

malformation (Tables 1-2). 173

Among these 83 patients, 24 (28.9%) referred to as EPM+ presented extra pituitary malformations 174

(central nervous system, ocular, dental, craniofacial, sella turcica, or cardiac malformations) (table 2); 175

59 (71.1%) had no EPM (EPM-). Mean age at diagnosis was similar for the two groups (EPM+: 9.4 ± 176

11.6 years; EPM-: 9.6 ± 7.2 years; p=0.46). The incidence of neonatal distress and breech delivery was 177

not significantly different (18.8% EPM+ vs. 21.3% EPM-, and 13.3% EPM+ vs. 19.6% EPM-, 178

respectively) as shown on table 1. 179

Concerning hormonal status, the proportion of patients with IGHD was very low (6%) and similar in 180

both groups (p=0.29). Univariate studies showed significantly more patients bearing GHD in 181

association with at least two other hormonal abnormalities in the EPM+ than the EPM- group (87.5% 182

EPM+ vs. 69.5% EPM-, p=0.04) (Figure 1). Multivariate studies confirmed these data, showing that 183

EPM was an independent risk factor in terms of severity of hormonal status (OR=6.4 CI95% [1.13; 184

36.4]; p=0.04), when adjusted on imaging covariate. 185

On MRI, severe pituitary hypoplasia (lack of detectable residual pituitary tissue) was observed only in 186

the EPM+ group (8.7%, p<0.05). Multivariate statistical studies showed that having a posterior 187

pituitary along the pituitary stalk was a protective factor regarding severity of hormonal phenotype 188

(OR=0.05 CI95% [0.04; 0.602]; p=0.02), when adjusted on presence or absence of EPM. 189

Molecular analyses. 190

Gene screening was performed in 69 PSIS patients from 65 pedigrees; among them, 7 had a familial 191

history of PSIS (4 pedigrees) and 9 were born from consanguineous parents (7 pedigrees). Among 192

these 69 PSIS patients, we found one new HESX1 mutation. This HESX1 gene defect consists of a 193

homozygous nonsense mutation leading to a stop codon in the second exon (c.325C>T, p.Arg109X) 194

predicting a severely truncated protein (loss of 74 aminoacids including the homeodomain). The 195

propositus was a boy, born post-term with caesarean delivery from consanguineous parents of Turkish 196

origin. At one year of age, he was referred to a pediatric department for recurrent hypoglycemia and 197

of 24

9

growth retardation. He also had micropenis and cryptorchidism. He was treated for GH and ACTH 198

deficiencies. At 14 years of age, endocrine deficits also included TSH, LH/FSH, and PRL deficiencies. 199

Pituitary MRI was performed and showed a hypoplastic anterior pituitary gland, an ectopic posterior 200

pituitary lobe located at median eminence, and an interrupted pituitary stalk (Figure 2). No septo-optic 201

dysplasia, midline brain defects, or other malformations were observed. In the family of the 202

propositus, we found the same mutation at heterozygous state in his father, his mother, one of his two 203

sisters, and one cousin, who were all asymptomatic. Furthermore, in one patient (EPM+, patient 21, 204

table 2) with no familial or consanguinity history, a novel heterozygous HESX1 allelic variation was 205

found: p. Ser67Thr (c.200G>C). Analysis of other members of the family showed that both the father 206

and a sister, phenotypically normal, also had the same allelic variation, which was not carried by the 207

mother and a brother. The new c.200G>C heterozygous HESX1 allelic variation affects a poorly 208

conserved nucleotide. The Serine to Threonine substition was suggested to have minor physico-209

chemical effects by in silico analysis with Alamut software. 210

One familial case (group EPM+, patient 17, table 2) had an already reported (18) LHX4 gene defect: a 211

heterozygous intronic point mutation involving the splice-acceptor site preceding exon5 (c.607-1G>C 212

heterozygote). A second LHX4 heterozygous mutation (c.293_294insC, p.Thr99AsnfsX53) was 213

recently reported by our group and confirmed as a functionally defective variant on the basis of in 214

vitro studies (17): two brothers born from non consanguineous parents presented somatolactotroph and 215

thyrotroph deficiencies. MRI showed poorly developed sella turcica and pituitary hypoplasia for both, 216

associated with a hypoplastic corpus callosum, thin pituitary stalk and invisible posterior pituitary for 217

the younger one (EPM+, patient 24, table 2), but with eutopic posterior pituitary and normal corpus 218

callosum for the older one. Their father carried the same heterozygous mutation. At age 28, he 219

harbored a partial gonadotroph deficiency and a hyperplasic pituitary without pituitary stalk or 220

posterior pituitary abnormalities on MRI (17). 221

We also identified two LHX4 variants (c.1108G>A heterozygote, p.Gly370Ser) and (c.296 C>T 222

heterozygote, p.Thr90Met), with normal DNA binding and transactivation properties in 223

transfection studies as we reported (17). Moreover, 56.3% of patients presented the c.983G>A 224

of 24

10

(p.Asn328Ser) LHX4 polymorphism, in a heterozygous (40%) or homozygous (16.3%) way. This 225

polymorphism represented 36.4% of all alleles studied. 226

No OTX2 mutation was identified, even in patients bearing eye defects. The OTX2 polymorphisms 227

c.98-46C>A (rs2277499) and 3'UTR+10G>A (rs171978) represented 7.2% and 11.6% of all alleles 228

studied, respectively. No abnormality detectable by sequencing was found in SOX3 gene in the males 229

from our series. 230

231

DISCUSSION 232

In this study we analyzed neonatal, hormonal, neuroimaging data and genetic features from patients 233

with PSIS to better characterize this syndrome of unknown pathogenesis. Several clinical points of 234

interest appeared when comparing two groups, one without EPM and one with EPM (CNS, ocular, 235

dental, craniofacial, sella turcica, or cardiac malformations). First, the EPM+ group was associated 236

with a more severe hormonal status, as found in other studies (3, 6).This calls for repeat, careful long-237

term follow-up in these patients.Second, the presence of EPM was unfortunately not associated with 238

an earlier diagnosis in our series, although earlier diagnosis has been reported in previous series (about 239

9.5 years vs. less than 6 years) (2, 3). Indeed, diagnosis in the overall population should have been 240

made earlier, considering the proportions of neonatal events (20.6%), SGA (9.2%), hypoglycemia 241

(16%), jaundice (5.3%), extra pituitary malformations (29.3%), micropenis (40.4%), and 242

cryptorchidism (10.5%). Such neonatal events should lead to earlier evaluation of pituitary function 243

and stalk morphology by MRI to preserve height and mental development (2). Third, the number of 244

perinatal events did not differ between both groups, even though the EPM- group had a slightly higher 245

proportion of neonatal distress and breech presentation. Fourth, the EPM+ group included a 246

significantly greater proportion of patients with no visible anterior pituitary. No association was found 247

between location of posterior pituitary and presence of EPM, as described in other cohorts of PSIS 248

patients (5, 6). 249

MRI pituitary aspect appears important for analysis and has been related to severity of hormonal status 250

(3 , 9 , 26). Indeed, regardless of the presence or absence of EPM, a correlation has previously been 251

reported between anterior pituitary function and anatomical characteristics of the hypothalamic-252

of 24

11

pituitary region, particularly a better functional prognostic value of the visibility of the pituitary stalk 253

(5, 26 , 27). In our population, posterior pituitary along the stalk represented a protective factor 254

regarding severity of hormonal status. Note that, in the literature, location of ectopic posterior lobe 255

was also associated with functional prognosis, with a greater number of hormone deficiencies when 256

posterior lobe is localized at the median eminence or hypothalamic region (26) 257

258

In the last decade, PSIS has been described in patients with gene defects affecting pituitary 259

transcription factors: HESX1, LHX4, OTX2 or SOX3 (7). The present study reports systematic genetic 260

analysis in patients with PSIS on the basis of current knowledge on genes involved in this condition. 261

First, we found a novel HESX1 homozygous nonsense mutation generating a truncated protein, 262

resulting in total loss of homeodomain and co repressor binding. This HESX1 p.Arg109X mutation 263

leads to a severely truncated protein including the homeodomain that is involved in DNA binding and 264

required to recruit components of N-CoR associated corepressors as reported by Dasen et al. (28). 265

HESX1 homozygous mutations have most often been related to severe hormonal, imaging, and clinical 266

presentation features, whereas heterozygous mutations are associated with a milder phenotype with 267

presence of different midline brain defects or optic nerve hypoplasia (29). Surprisingly, despite this 268

severe homozygous genotype, the phenotype of our patient was relatively light, with no midline brain 269

defect or septo-optic dysplasia (SOD). To date in the literature only one patient, also born from a 270

consanguineous union, has been reported bearing HESX1 homozygous mutation with pituitary stalk 271

abnormality but no midline brain defect or SOD (29). So HESX1 homozygous mutation may be 272

associated with a less severe phenotype than previously considered. We also identified a novel 273

heterozyguous variation of HESX1 in a patient and in his unaffected father and sister. In silico analysis 274

did not favor a functional impact of this allelic variation. We thus considered this monoallelic 275

variation as unlikely to have a pathogenic significance. However, as this variation was not found 276

among 100 allelic controls (although not population matched), we cannot rule out a variant of 277

functional significance but incomplete penetrance. 278

We also found two LHX4 allelic variations of unknown significance (p.Gly370Ser and 279

p.Thr90Met), and one common polymorphism of LHX4. Concerning the 2 variants of unknown 280

of 24

12

significance, whereas Gly to Ser aminoacid substitution was not predicted to have significant 281

physico-chemical effects, Thr90Met aminoacid substitution affects the LIM domain and is 282

likely associated with important physico-chemical changes. In our in vitro studies, these 283

variants did not modify the DNA-binding and transactivation properties of the protein (17). 284

Considering the limitations of functional studies however, it cannot be ruled out that these 285

variants may somehow be functionally relevant. . The overall incidence of LHX4 defects was very 286

low (n=2, 2.9%), but it was high in a subset of patients with familial history of PSIS (50%). Including 287

our results, 60% of cases with LHX4 mutation currently reported in the literature had familial history 288

of CPHD (17-19). Note that the prevalence of familial CPHD with PSIS was similar in our cohort and 289

in the cohort of Pinto et al (6.2% and 5.9%, respectively) (3) but lower than in the CPHD cohort 290

without PSIS previously reported by our group (18 familial cases among 120 unrelated CPHD patients 291

(15%)) (30). 292

The male preponderance observed in our study (sex ratio: 2.3/1) is in keeping with other studies (2, 3, 293

5, 9, 26) suggesting an X-linked inheritance. Neither HESX1, OTX2, nor LHX4 gene is located on X 294

chromosome. Recently, X-linked inherited SOX3 abnormalities were reported to be responsible for 295

hypopituitarism in patients with PSIS (22, 23). One genetic screening of SOX3 has been been 296

performed by Alatzoglou et al, in 42 CPHD and 21 IGHD with ectopic posterior pituitary 297

pedigrees without any sequence or gene copy number abnormalities (8). In our study, molecular 298

genetics analysis through DNA sequencing revealed neither in frame-duplication nor deletion in the 299

first polyalanine tract of the SOX3 gene, the only abnormalities responsible for molecular defects 300

reported to date. Note that a large duplication cannot be excluded as no cytogenetic study could be 301

performed on patients’ extracted genomic DNA collected through Genhypopit. 302

Finally, no OTX2 mutation was found in our series. Heterozygous OTX2 mutations have been 303

primarily reported in patients with highly variable ocular malformations and more recently 304

hypopituitarism with or without ocular anomalies (20, 21). Dateki et al. recently emphasized a 305

low prevalence of OTX2 mutation in 61 Japanese patients with pituitary dysfunction without 306

ocular anomalies, in line with our results (20). 307

of 24

13

308

Overall, the present study suggests that at least two clinical subsets of PSIS patients might be 309

distinguished. The first one, with extra-pituitary malformations, has more severe hormonal (CPHD) 310

and pituitary imaging features, including pituitary hypoplasia. The second one, without extra-pituitary 311

malformation, has a higher proportion of neonatal distress but a lighter hormonal and pituitary 312

imaging profile. This wide phenotypic spectrum might depend on genetic and/or environmental 313

factors. Among this cohort of PSIS patients, we found only three unequivocally causative genetic 314

defects. Noteworthy, mutation of HESX1 was found in a patient without SOD, born from one of the 315

seven consanguineous unions reported in our study. Second, both mutations of LHX4 were found 316

among the four index cases with a familial history of PSIS. Considering the rarity of familial or 317

consanguineous cases of PSIS, HESX1 and LHX4 screening should thus be recommended in such a 318

context. Our study shows that HESX1 or LHX4 gene defects can be ruled out as common causes of 319

sporadic PSIS, a finding that does not currently support cost effectiveness of systematic genetic 320

screening of these factors in this condition. We did not find any OTX2 mutation in this cohort, but only 321

6 patients harbored ocular malformations. OTX2 genetic screening should be performed when ocular 322

malformation with or without hypopituitarism is present. However, pituitary phenotype in OTX2 323

mutations without ocular anomaly remains to be further studied. 324

Phenotype variability among PSIS patients and complexity of pituitary embryogenesis suggest that 325

this syndrome involves several mechanisms, such as a multigenic pattern or, as in septo-optic 326

dysplasia, environmental influences (29). To date the underlying mechanisms involved in most cases 327

of PSIS thus remain to be identified. For PSIS, we thus recommend a careful and long-term follow-up 328

with repeat hormonal evaluation, especially in patients with associated extra-pituitary malformations, 329

given the severity of hormonal prognosis in this group of patients. Molecular diagnostic analyses 330

should be discussed according to familial or consanguinity status, and pattern of syndromic features. 331

332

Acknowledgments 333

We thank the clinicians who sent blood samples of their PSIS patients for genetic screening in the 334

GENHYPOPIT network: Dr M. Abid (Sfax, Tunisia), Dr A. Bennet (Toulouse, France), Dr P. Berlier 335

of 24

14

(Lyon, France), Prof. F. Borzon-Chazot (Lyon, France), Prof. P. Bouchard (Paris, France), Prof. P. 336

Bougneres (Paris, France), Prof. O. Bruno (Buenos Aires, Argentina), Prof. P. Caron (Toulouse, 337

France), Prof. O. Chabre (Grenoble, France), Prof. P. Chatelain (Lyon, France), Prof. F. Chentli 338

(Algiers, Algeria), Prof. S. Christin-Maitre (Paris, France), Prof. B. Conte-Devolx (Marseille, France), 339

Prof. M. David (Lyon, France), Dr G. Diene (Toulouse, France), Prof. M. El Kholy (Cairo, Egypt), Dr 340

O. Evliyaoglu (Ankara, Turkey), Dr C. Fedou (Montpellier, France), Dr A.M. Guedj (Nimes, France), 341

Prof. G. Halaby (Beirut Lebanon), Prof. M.L. Kottler (Caen, France), Prof. B. Leheup (Nancy, 342

France), Dr M. Manavela (Buenos Aires, Argentina), Dr J.C. Mas (Nice, France), Prof. M. Pugeat 343

(Lyon, France), Prof. V. Rohmer (Angers, France), Dr F. Soumeya (Algiers, Algeria), Dr C. Stuckens 344

(Lille, France), Prof. A. Tabarin (Bordeaux, France), Dr C. Teinturier (Paris, France), Dr Z. Turki 345

(Tunis, Tunisia), Dr M.C. Vantyghem (Lille, France). We also thank Anne Carle, Anne-Laure 346

Germanetti, Morgane Pertuit, Nicole Peyrol and Nadine Pluchino (Molecular Biology Laboratory) for 347

the genetic analysis of transcription factors. 348

349

List of abbreviations 350

CNS: central nervous system; CPHD: combined pituitary hormone deficiency; EPM: extra pituitary 351

malformations; GH: growth hormone; IGHD: isolated growth hormone deficient; MRI: magnetic 352

resonance imaging; OD: odds ratio; PSIS: pituitary stalk interruption syndrome; SOD: septo-optic 353

dysplasia; SGA: small for gestational age; SDS: standard deviation. 354

355

Disclosure statement: there is no conflict of interest that could be perceived as prejudicing the 356

impartiality of the research reported. 357

Grant support: Association pour le Développement des Recherches Biologiques et Médicales au 358

Centre Hospitalier Régional de Marseille (ADEREM) 359

360

361

362

363

of 24

15

REFERENCES 364

1. Simmons GE, Suchnicki JE, Rak KM, Damiano TR MR imaging of the pituitary stalk: 365

size, shape, and enhancement pattern. AJR Am J Roentgenol 1992 159:375-377 366

2. Tauber M, Chevrel J, Diene G, Moulin P, Jouret B, Oliver I, Pienkowski C, Sevely A 367

Long-term evolution of endocrine disorders and effect of GH therapy in 35 patients 368

with pituitary stalk interruption syndrome. Horm Res 2005 64:266-273 369

3. Pinto G, Netchine I, Sobrier ML, Brunelle F, Souberbielle JC, Brauner R Pituitary 370

stalk interruption syndrome: a clinical-biological-genetic assessment of its 371

pathogenesis. J Clin Endocrinol Metab 1997 82:3450-3454 372

4. Maghnie M, Larizza D, Triulzi F, Sampaolo P, Scotti G, Severi F Hypopituitarism and 373

stalk agenesis: a congenital syndrome worsened by breech delivery? Horm Res 1991 374

35:104-108 375

5. Melo ME, Marui S, Carvalho LR, Arnhold IJ, Leite CC, Mendonca BB, 376

Knoepfelmacher M Hormonal, pituitary magnetic resonance, LHX4 and HESX1 377

evaluation in patients with hypopituitarism and ectopic posterior pituitary lobe. Clin 378

Endocrinol (Oxf) 2007 66:95-102 379

6. Simon D, Hadjiathanasiou C, Garel C, Czernichow P, Leger J Phenotypic variability 380

in children with growth hormone deficiency associated with posterior pituitary 381

ectopia. Clin Endocrinol (Oxf) 2006 64:416-422 382

7. Davis SW, Castinetti F, Carvalho LR, Ellsworth BS, Potok MA, Lyons RH, 383

Brinkmeier ML, Raetzman LT, Carninci P, Mortensen AH, Hayashizaki Y, Arnhold 384

IJ, Mendonca BB, Brue T, Camper SA Molecular mechanisms of pituitary 385

organogenesis: In search of novel regulatory genes. Mol Cell Endocrinol 2009 323:4-386

19 387

of 24

16

8. Alatzoglou KS, Turton JP, Kelberman D, Clayton PE, Mehta A, Buchanan C, Aylwin 388

S, Crowne EC, Christesen HT, Hertel NT, Trainer PJ, Savage MO, Raza J, Banerjee 389

K, Sinha SK, Ten S, Mushtaq T, Brauner R, Cheetham TD, Hindmarsh PC, Mullis PE, 390

Dattani MT Expanding the spectrum of mutations in GH1 and GHRHR: genetic 391

screening in a large cohort of patients with congenital isolated growth hormone 392

deficiency. J Clin Endocrinol Metab 2009 94:3191-3199 393

9. Arrigo T, Wasniewska M, De Luca F, Valenzise M, Lombardo F, Vivenza D, Vaccaro 394

T, Coradi E, Biason-Lauber A Congenital adenohypophysis aplasia: clinical features 395

and analysis of the transcriptional factors for embryonic pituitary development. J 396

Endocrinol Invest 2006 29:208-213 397

10. Kim SS, Kim Y, Shin YL, Kim GH, Kim TU, Yoo HW Clinical characteristics and 398

molecular analysis of PIT1, PROP1,LHX3, and HESX1 in combined pituitary 399

hormone deficiency patients with abnormal pituitary MR imaging. Horm Res 2003 400

60:277-283 401

11. Sloop KW, Walvoord EC, Showalter AD, Pescovitz OH, Rhodes SJ Molecular 402

analysis of LHX3 and PROP-1 in pituitary hormone deficiency patients with posterior 403

pituitary ectopia. J Clin Endocrinol Metab 2000 85:2701-2708 404

12. Turton JP, Reynaud R, Mehta A, Torpiano J, Saveanu A, Woods KS, Tiulpakov A, 405

Zdravkovic V, Hamilton J, Attard-Montalto S, Parascandalo R, Vella C, Clayton PE, 406

Shalet S, Barton J, Brue T, Dattani MT Novel mutations within the POU1F1 gene 407

associated with variable combined pituitary hormone deficiency. J Clin Endocrinol 408

Metab 2005 90:4762-4770 409

13. McNay DE, Turton JP, Kelberman D, Woods KS, Brauner R, Papadimitriou A, Keller 410

E, Keller A, Haufs N, Krude H, Shalet SM, Dattani MT HESX1 mutations are an 411

of 24

17

uncommon cause of septooptic dysplasia and hypopituitarism. J Clin Endocrinol 412

Metab 2007 92:691-697 413

14. Sobrier ML, Maghnie M, Vie-Luton MP, Secco A, di Iorgi N, Lorini R, Amselem S 414

Novel HESX1 mutations associated with a life-threatening neonatal phenotype, 415

pituitary aplasia, but normally located posterior pituitary and no optic nerve 416

abnormalities. J Clin Endocrinol Metab 2006 91:4528-4536 417

15. Sobrier ML, Netchine I, Heinrichs C, Thibaud N, Vie-Luton MP, Van Vliet G, 418

Amselem S Alu-element insertion in the homeodomain of HESX1 and aplasia of the 419

anterior pituitary. Hum Mutat 2005 25:503 420

16. Thomas PQ, Dattani MT, Brickman JM, McNay D, Warne G, Zacharin M, Cameron 421

F, Hurst J, Woods K, Dunger D, Stanhope R, Forrest S, Robinson IC, Beddington RS 422

Heterozygous HESX1 mutations associated with isolated congenital pituitary 423

hypoplasia and septo-optic dysplasia. Hum Mol Genet 2001 10:39-45 424

17. Castinetti F, Saveanu A, Reynaud R, Quentien MH, Buffin A, Brauner R, Kaffel N, 425

Albarel F, Guedj AM, El Kholy M, Amin M, Enjalbert A, Barlier A, Brue T A novel 426

dysfunctional LHX4 mutation with high phenotypical variability in patients with 427

hypopituitarism. J Clin Endocrinol Metab 2008 93:2790-2799 428

18. Machinis K, Pantel J, Netchine I, Leger J, Camand OJ, Sobrier ML, Dastot-Le Moal 429

F, Duquesnoy P, Abitbol M, Czernichow P, Amselem S Syndromic short stature in 430

patients with a germline mutation in the LIM homeobox LHX4. Am J Hum Genet 431

2001 69:961-968 432

19. Pfaeffle RW, Hunter CS, Savage JJ, Duran-Prado M, Mullen RD, Neeb ZP, Eiholzer 433

U, Hesse V, Haddad NG, Stobbe HM, Blum WF, Weigel JF, Rhodes SJ Three novel 434

missense mutations within the LHX4 gene are associated with variable pituitary 435

hormone deficiencies. J Clin Endocrinol Metab 2008 93:1062-1071 436

of 24

18

20. Dateki S, Kosaka K, Hasegawa K, Tanaka H, Azuma N, Yokoya S, Muroya K, 437

Adachi M, Tajima T, Motomura K, Kinoshita E, Moriuchi H, Sato N, Fukami M, 438

Ogata T Heterozygous orthodenticle homeobox 2 mutations are associated with 439

variable pituitary phenotype. J Clin Endocrinol Metab 2010 95:756-764 440

21. Diaczok D, Romero C, Zunich J, Marshall I, Radovick S A novel dominant negative 441

mutation of OTX2 associated with combined pituitary hormone deficiency. J Clin 442

Endocrinol Metab 2008 93:4351-4359 443

22. Laumonnier F, Ronce N, Hamel BC, Thomas P, Lespinasse J, Raynaud M, Paringaux 444

C, Van Bokhoven H, Kalscheuer V, Fryns JP, Chelly J, Moraine C, Briault S 445

Transcription factor SOX3 is involved in X-linked mental retardation with growth 446

hormone deficiency. Am J Hum Genet 2002 71:1450-1455 447

23. Woods KS, Cundall M, Turton J, Rizotti K, Mehta A, Palmer R, Wong J, Chong WK, 448

Al-Zyoud M, El-Ali M, Otonkoski T, Martinez-Barbera JP, Thomas PQ, Robinson IC, 449

Lovell-Badge R, Woodward KJ, Dattani MT Over- and underdosage of SOX3 is 450

associated with infundibular hypoplasia and hypopituitarism. Am J Hum Genet 2005 451

76:833-849 452

24. Sempé M, Pedron G, Roy-Pernot M Auxologie, méthode et séquences.: Paris: 453

Théraplix 1979 454

25. Argyropoulou M, Perignon F, Brunelle F, Brauner R, Rappaport R Height of normal 455

pituitary gland as a function of age evaluated by magnetic resonance imaging in 456

children. Pediatr Radiol 1991 21:247-249 457

26. Chen S, Leger J, Garel C, Hassan M, Czernichow P Growth hormone deficiency with 458

ectopic neurohypophysis: anatomical variations and relationship between the visibility 459

of the pituitary stalk asserted by magnetic resonance imaging and anterior pituitary 460

function. J Clin Endocrinol Metab 1999 84:2408-2413 461

of 24

19

27. Arrigo T, De Luca F, Maghnie M, Blandino A, Lombardo F, Messina MF, 462

Wasniewska M, Ghizzoni L, Bozzola M Relationships between neuroradiological and 463

clinical features in apparently idiopathic hypopituitarism. Eur J Endocrinol 1998 464

139:84-88 465

28. Dasen JS, Barbera JP, Herman TS, Connell SO, Olson L, Ju B, Tollkuhn J, Baek SH, 466

Rose DW, Rosenfeld MG Temporal regulation of a paired-like homeodomain 467

repressor/TLE corepressor complex and a related activator is required for pituitary 468

organogenesis. Genes Dev 2001 15:3193-3207 469

29. Kelberman D, Dattani MT Septo-optic dysplasia - novel insights into the aetiology. 470

Horm Res 2008 69:257-265 471

30. Reynaud R, Gueydan M, Saveanu A, Vallette-Kasic S, Enjalbert A, Brue T, Barlier A 472

Genetic screening of combined pituitary hormone deficiency: experience in 195 473

patients. J Clin Endocrinol Metab 2006 91:3329-3336 474

475

Legends to Tables and Figures 476

Table 1: Clinical, perinatal, hormonal, and MRI features in 83 patients. Comparison between 477

the EPM+ group (with extra-pituitary malformations) and the EPM- group (with no extra-478

pituitary malformation). 479

First column, features in 83 PSIS patients. Second and third columns, characteristics in the EPM+ 480

group, with extra-pituitary malformation, and the EPM- group, without such malformation. The p 481

values for both groups by univariate comparisons are given in the last column. They were obtained by 482

one-tailed Student’s unpaired t-test (or Wilcoxon test if necessary) or when appropriate by one-tailed 483

Pearson’s chi2 test (or Fischer exact test if necessary). The values were considered significant (*) when 484

p<0.05). Numbers in parentheses represent documented cases for each feature (n=total, EPM+/EPM-) 485

in 83 patients. SGA: small for gestational age. IGHD= isolated GH deficiency. CPHD: combined 486

of 24

20

pituitary hormone deficiencies. ND= not documented. NS= not significant. °: PRL status known for 76 487

patients (10 PRL deficits, 52 normal, 14 hyperprolactinemia) 488

489

490

Table 2: Features of the EPM+ group, presenting extra-pituitary malformations. 491

ND= not documented. Sex: F=female; M=male. SGA: small for gestational age. IGHD=isolated GH 492

deficiency. Along PS=along the pituitary stalk. ME=median eminence. SOD: septo-optic dysplasia. 493

IVC: interventricular communication. IAC: interauricular communication. PRL: prolactin 494

* patients with LHX4 mutations; ** patient with HESX1 heterozygous polymorphism. 495

Figure 1: Hormone status in both groups of PSIS patients: EPM+, with extra-pituitary 496

malformation (EPM) and EPM-, without EPM. 497

The EPM+ group comprises 24 PSIS patients and the EPM- group 59 patients. GH deficiency (GHD) 498

+ 2 or more hormone abnormalities was significantly more common in the EPM+ than in the EPM- 499

group (87.5% vs. 69.5%, p=0.04). IGHD= isolated GHD.* : PRL status known for 76 patients 500

Figure 2: Pituitary MRI in a patient bearing an HESX1 gene mutation. 501

MRI was performed at 25 years of age. Non enhanced coronal (A & B) T1-weighted sequence 502

depicting an interrupted pituitary stalk (arrow, panel A), hypoplastic anterior pituitary gland 503

(above asterisk, panel A), and ectopic posterior pituitary (dotted arrow, panel B). (C): Coronal 504

MRI scan of the pituitary of a normal child the anterior pituitary and the pituitary stalk in the 505

normal sella turcica. 506

507

of 24

EPM- group (n=59)

5.1%25.4%

13.6%42.4%

13.5%

EPM+ group (n=24)

8.3%

4.2%

20.8%

45.9%

20.8%

IGHD

GHD+1defect

GHD+ 0 or 1

defect

GHD+ 2, 3, or 4* defects

12.5%

87.5%

GHD+ 0 or 1

defect30.5%

GHD+ 2, 3, or 4* defects

69.5%

GHD+2 defects

GHD+3 defects

GHD+4* defects

IGHD

GHD+1defect

GHD+2 defects

GHD+3 defects

GHD+4* defects

P = 0.04*

Figure 1.

of 24

C

Figure 2.

of 24

Table 1.

Features PSIS patients

(n=83)

EPM+

(n=24, 28.9%)

EPM-

(n=59, 71.1%)

P

EPM+ vs EPM-

Mean age at diagnosis (years) (n=70. 21/49) 9.6±8.8 9.4±11.6 9.51±7.3 NS

Perinatal events:

- IUGR (n=65. 17/48)

- Neonatal distress (n=63. 16/47)

- Breech presentation (n=61. 15/46)

9.2%

20.6%

18%

11.8%

18.8%

13.3%

8.3%

21.3%

19.6%

NS

NS

NS

Hormonal status (n=83. 24/59):

- IGHD

- CPHD: GHD+ 1 anterior pituitary hormone deficit

GHD+ 2 anterior pituitary hormone deficits

GHD+ 3 anterior pituitary hormone deficits

GHD+ 4 anterior pituitary hormone deficits°

6%

19.3%

15.6%

43.4%

15.7%

8.3%

4.2%

20.8%

45.9%

20.8%

5.1%

25.4%

13.6%

42.4%

13.5%

NS

0.01*

NS

NS

NS

MRI:

- Anterior pituitary: not visible

(n=79. 23/56) hypoplasia

normal

- Posterior pituitary lobe: median eminence

(n=81. 24/57) along pituitary stalk

invisible

ectopic unspecified

- Pituitary stalk: thin

(n=69. 21/48) interrupted

invisible

2.5%

77.2%

20.3%

45.7%

13.6%

7%

32.1%

11.6%

40.6%

47.8%

8.7%

69.6%

21.7%

45.8%

16.7%,

12.5%

25%

9.5%

47.6%

42.9%

0%

80.4%

19.6%

45.6%

12.3%

6%

35.1%

12.5%

37.5%

50%

0.01*

NS

NS

NS

NS

NS

NS

NS

NS

NS

of 24

Table 2.

n. Sex Perinatal

events

Hormonal

status

Anterior

pituitary

Posterior

pituitary

Pituitary

stalk

Associated malformations

1 F ND GHD+ 3 defects Hypoplasia Along PS Thin Absence of interventricular

septum

2 M ND GHD+ 3 defects Hypoplasia Ectopic ND Hydrocephalia. macrosomia.

blindness. SOD

3 M 0 GHD+ 1 defect

(PRL not done)

Normal Ectopic ND Chiari 1 malformation

4 M 0 GHD+ 3 defects Normal ME Interrupted Chiari 1 and chiasma

malformations

5 F Breech IGHD Hypoplasia ME Invisible Left optic nerve atrophy.

amblyopy. SOD

6 M 0 GHD+ 2 defects Hypoplasia Invisible Invisible Bilateral optic nerve atrophy.

SOD

7 F ND GHD+ 4 defects Hypoplasia ME Invisible Craniostenosis

8 F 0 GHD+ 3 defects Hypoplasia ME Interrupted Enlarged floor of the 3rd

ventricle

9 F 0 GHD+ 3 defects Hypoplasia Invisible Interrupted Myelinisation defect

10 M ND GHD+ 4 defects Hypoplasia Along PS Invisible Amblyopy. nystagmus. SOD

11 F 0 GHD+ 2 defects Hypoplasia ME Invisible Hypoplasia of sella turcica

12 M ND GHD+ 3 defects ND Ectopic ND Optic nerve hypoplasia. retinian

kyst. coloboma. microcephalia

13 M Breech

ND

GHD+ 3 defects Hypoplasia Ectopic Interrupted Hydrocephalia. frontal atrophia.

cervical rigidity. Chiari 1

malformation. Syringomyelia.

14 M 0 GHD+ 4 defects No pituitary

tissu visible

ME Invisible midline abnormalities

15 M 0 IGHD Hypoplasia Along PS Interrupted Chiari 1 malformation.

hypoplasia of sella turcica

16 M 0 GHD+3 defects Normal Ectopic Interrupted Cerebellar atrophy. hypotonia.

Little syndrome

17* M ND GHD+ 4 defects Hypoplasia ME Invisible Hypoplasia of sella turcica.

Chiari 1.

persistant craniopharyngeal canal

18 F 0 GHD+ 4 defects Hypoplasia ME invisible Left frontal porencephaly

19 M 0 GHD+ 3 defects Hypoplasia Ectopic Interrupted Cerebral anterior agenesy

20 F SGA GHD+ 2 defects No pituitary

tissu visible

ME Invisible Microcephalia. ventricle

dilatation. microphtalmia. SOD

corpus callosum, cerebellar and

sella turcica abnormalities.

21** M SGA.

breech

GHD+4 defects Normal ME Interrupted Cardiac malformations: IVC.

IAC

22 M ND GHD+2 defects

(PRL not done)

Hypoplasia Along PS Interrupted Incisor agenesy

23 M 0 GHD+3 defects Normal ME Interrupted Cardiac malformations: IVC.

Pulmonary stenosis

24* M ND GHD+2defects Hypoplasia Invisible Thin Hypoplasia of sella turcica and of

corpus callosum.

of 24

pituitary stalk interruption syndrome in 83 patients: novel hesx1

Documents