Primrose syndrome: Characterization of the phenotype in 42 patients.

Primrose syndrome (PS; MIM# 259050) is characterized by intellectual disability (ID), macrocephaly, unusual facial features (frontal bossing, deeply set eyes, down-slanting palpebral fissures), calcified external ears, sparse body hair and distal muscle wasting. The syndrome is caused by de novo heterozygous missense variants in ZBTB20. Most of the 29 published patients are adults as characteristics appear more recognizable with age. We present 13 hitherto unpublished individuals and summarize the clinical and molecular findings in all 42 patients. Several signs and symptoms of PS develop during childhood, but the cardinal features, such as calcification of the external ears, cystic bone lesions, muscle wasting, and contractures typically develop between 10 and 16 years of age. Biochemically, anemia and increased alpha-fetoprotein levels are often present. Two adult males with PS developed a testicular tumor. Although PS should be regarded as a progressive entity, there are no indications that cognition becomes more impaired with age. No obvious genotype-phenotype correlation is present. A subgroup of patients with ZBTB20 variants may be associated with mild, nonspecific ID. Metabolic investigations suggest a disturbed mitochondrial fatty acid oxidation. We suggest a regular surveillance in all adult males with PS until it is clear whether or not there is a truly elevated risk of testicular cancer.

INTRODUCTION

Primrose syndrome (PS; MIM# 259050) is an infrequently described condition characterized by increased postnatal growth in height and head circumference, unusual facial features (frontal bossing, deeply set eyes, down‐slanting palpebral fissures), cognitive deficit associated with autism spectrum disorder, and ectopic calcifications. With age, distal muscle atrophy, hearing loss, cataract, sparse body hair, and a disturbed glucose metabolism can become clear.2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18 Until recently most reported affected individuals have been adults as the phenotype may become more easily recognizable over time.

PS is mostly caused by de novo heterozygous missense variants in the N‐terminal portion of the DNA binding domain of ZBTB20 (MIM* 606025), a transcriptional repressor. Two patients carrying truncating variants or small deletions have also been reported.6, 14 The protein is a member of the broad complex tramtrack bric‐a‐brac (BTB) zinc‐finger (ZnF) family and is characterized by an N‐terminal BTB domain that is involved in protein‐protein interaction, and five C2H2 zinc fingers at the C‐terminus mediating protein binding to regulatory sites within promoters of target genes.19, 20, 21, 22 ZBTB20 acts as a regulator of neurogenesis, fetal liver development, somatic growth, detoxification and glucose metabolism.23, 24, 25 Thus far, all ZBTB20 variants causing PS have been missense variants that affect amino acid residues in the first and second ZnF motifs.10, 11

Here we summarized the collective data from 42 patients with PS, 13 of whom have not been reported before, present the clinical, biochemical and molecular characteristics, and emphasize their evolution over time.

SUBJECTS AND METHODS

Subjects

The present series were gathered by contacting authors who have previously published on PS or because collaborators contacted one of us (RCH) because of his experience with PS. Data were collected through a table specifically designed for the study (Supplemental data Table S1). Clinical pictures, results of formal testing of cognitive development, and results of biochemical tests were also gathered. No biochemical or genetic studies were performed specifically for the present study. We gathered data from 29 patients reported previously1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18 and 13 hitherto unpublished patients. One stillbirth was also included. Intellectual disability (ID) was classified as mild/moderate‐severe based on neuropsychological consultations; IQ scores were included if available. The study was approved by the Medical Ethics Committee of the Amsterdam UMC (NL45451.018.13).

Molecular analyses

Molecular studies were performed either by whole‐exome sequencing (WES) using a panel aimed at detecting variants in genes known to cause ID if mutated, or by Sanger sequencing. In 32 patients, a ZBTB20 variant was detected using panel sequencing for ID, after which the clinical diagnosis was established. In four patients, the diagnosis was clinically based and the ZBTB20 variant was subsequently detected by Sanger sequencing. In one patient, the diagnosis was established based on SNP array. In five patients reported in literature, no information on methods of molecular analysis was available (all these patients showed normal karyotype).

RESULTS

The study included 22 males and 20 females, varying in age between 9 months and 49 years. The mean and the median age at diagnosis were 17.3 ± 15.4 years and 11.0 ± 15.4 years. The main clinical characteristics of the study participants are summarized in Tables 1 and 2 and illustrated in Figures 1 and 2 and Figure S1. The data in the tables are shown separately for children (0‐16 years) and adults (>16 years). Detailed information for each patient is available in Table S1, see Supplement. In the text, only data for which information is not reported in the tables are discussed. Single patient number is indicated between brackets if specific findings are mentioned.

TABLE 1

Growth, development, and behavior in the 42 individuals with Primrose syndrome

	Children	Adults	All
	n = 29	n = 13	n = 42
Growth at birth
Length (cm)			49.7 ± 3.60
Length > 2SD			1/22
Weight (kg)			3.19 ± 0.64
Weight > 2SD			3/29
Head circumference (cm)			35.91 ± 2.25
Head circumference > 2SD			9/22
Postnatal growth
Mean age at last clinical evaluation (y)	7.74 ± 4.22	37.38 ± 10.34	17.80 ± 15.6
Height (cm)	125.83 ± 29.33	177.50 ± 10.71
Height > 2SD	3/25 (12%)	0/9 (0%)	3/34 (9%)
Weight (kg)	31.91 ± 22.45	72.80 ± 16.09
Weight > 2SD	6/25 (24%)	0/6 (0%)	6/31 (19%)
Head circumference (cm)	54.71 ± 3.69	58.75 ± 2.46
Head circumference > 2SD	21/26 (81%)	8/12 (67%)	29/38 (76%)
Development
Intellectual disability mild	5/27 (19%)	2/13 (15%)	7/39 (18%)
Intellectual disability moderate‐severe	22/27 (81%)	11/13 (85%)	33/39 (85%)
Behavior
Autism	16/24 (67%)	4/9 (44%)	20/33 (61%)
Self‐injurious behavior	7/19 (37%)	4/7 (57%)	11/26 (42%)
Sleep disturbances	8/19 (42%)	2/7 (29%)	10/26 (38%)

Note: Only data of at term born newborns (38‐42 wk) have been used.

TABLE 2

Clinical features of the 42 individuals with Primrose syndrome

	HPO ID	Children	Adults	All
Clinical sign		N = 29	N = 13	N = 42
Morphology
Brachycephaly	0000248	8/18 (44%)	5/9 (56%)	13/27 (48%)
Frontal bossing	0002007	15/20 (75%)	7/9 (78%)	22/29 (76%)
Ptosis	0000508	10/18 (56%)	10/10 (100%)	20/28 (71%)
Downslanted palpebral fissures	0000494	11/22 (50%)	7/12 (58%)	18/34 (53%)
Deeply set eyes	0000490	16/21 (76%)	10/11 (91%)	26/32 (81%)
Highly arched palate	0002705	7/17 (41%)	2/6 (33%)	9/23 (39%)
Torus palatinus	189 700	1/16 (6%)	6/11 (55%)	7/27 (26%)
Large jaw	0040309	8/17 (47%)	8/11 (73%)	16/28 (57%)
Large ears	0000400	14/25 (56%)	10/11 (91%)	24/36 (67%)
Calcification of ears	0005103	2/17 (12%)	12/12 (100%)	14/28 (50%)
Neuromuscular findings
Seizures	0001250	2/20 (10%)	4/9 (44%)	6/29 (21%)
Ataxia	0001251	6/18 (33%)	2/5 (40%)	8/23 (35%)
Hypotonia	0001252	21/25 (84%)	5/9 (56%)	26/34 (76%)
Distal muscle wasting	0003693	1/22 (5%)	11/11100%)	12/33 (36%)
Flexion contractures	0001371	5/24 (21%)	8/8 (100%)	13/31 (42%)
Delayed myelination	0012448	1/23 (4%)	2/11 (18%)	3/34 (9%)
Brain calcification	0002514	3/23 (13%)	1/11 (9%)	4/34 (12%)
Corpus callosum anomaly	0001273	11/23 (48%)	4/11 (36%)	15/34 (44%)
System involvement
Cataract	0000518	0/20 (0%)	6/10 (60%)	6/30 (20%)
Strabismus	0000486	10/21 (48%)	0/10 (0%)	10/31 (32%)
Hearing loss	0000365	21/27 (78%)	12/13 (92%)	33/40 (83%)
Scoliosis	0002650	9/23 (39%)	6/10 (60%)	15/33 (45%)
Cystic bone lesions	0012062	0/9 (0%)	5/9 (56%)	5/18 (28%)
Decreased BMD	0004349	3/8 (38%)	6/9 (67%)	9/17 (53%)
Hip dysplasia	0001385	1/17 (6%)	4/8 (50%)	5/25 (20%)
Thin nail	0001816	6/20 (30%)	4/7 (57%)	10/27 (37%)
Sparse body hair	0002231		11/12 (92%)	11/12 (92%)
Delayed puberty	0000823		3/11 (27%)	3/11 (27%)
Cryptorchidism	0000028	5/10 (50%)	2/6 (33%)	7/16 (44%)
Tumors	0002664	0/15 (0%)	2/9 (22%)	2/24 (8%)
Diabetes mellitus	0000819	2/16 (13%)	6/9 (67%)	8/25 (32%)
Anemia	0001903	4/16 (25%)	1/5 (20%)	5/21 (24%)
Elevated serum AFP levels	0006254	4/11 (36%)	5/7 (71%)	9/18 (50%)

Abbreviations: AFP, alpha‐fetp protein; BMD, bone mineral density; HPO ID, human phenotype ontology identifier.

FIGURE 1

Features from selected individuals with Primrose syndrome. (A) Faces from youngest to oldest at age 1.5 years (A), 2.5 years (B), 3 years (C), 4 years (D), 4 years (E), 5 years (F), 6 years (G), 8 years (H), 9 years (I), 11 years (J), 12 years (K), 13 years (L), 18 years (M), 31 years (N), 33 years (O), and 53 years (P). The patient identification number is indicated underneath the panels. (B) Other clinical features include alobar calcified ear (1), calcified ear on X‐ray (2), incomplete extension of fingers and small nails (3), joint hypermobility (4), distal muscle wasting in an adult (5), markedly small and thin nails (6), and malformed callosal body (7) [Colour figure can be viewed at wileyonlinelibrary.com]

FIGURE 2

Changes with age of cognition, muscle wasting, and serum alpha‐fetoprotein (AFP) in individuals with Primrose syndrome. A, Cognition. No evident correlation. B, Muscle wasting; data are presented based on age of first appearance. Increase with age evident. C, AFP serum levels. Each symbol represents a single individual; course over time in single patients is depicted if available. Elevated levels in almost every individual; no clear change with age in a single individual

Growth

The mean duration of pregnancy was 38.8 ± 2.0 weeks. Three pregnancies (P5, P21, and P28) were complicated by oligohydramnios, one pregnancy resulted in intrauterine demise (P28). Postnatal growth in height and weight is usually between the 50th and 90th centile but in males sometimes is >98th centile (Supplemental Figure [Link], [Link]).

Development and behavior

IQ score was available in seven patients (P7, P9, P21, P33‐36; six children, IQ 25‐77, one adult, IQ 25). Infrequent findings included attention deficit hyperactivity disorder (ADHD) (P9, P36, P37) and delayed speech (P7, P8). One child showed hyperphagia (P32), one adult patient also showed schizophrenia (P42). Patients' intellectual function seems not to change over time (Figure 2A). Insufficient data were available to evaluate reliably whether behavioral problems were progressive with time or not.

Morphological signs

No morphological sign is present in all affected individuals (Table 2), and the phenotype is variable indeed (Figure 1). Infrequently reported findings included cleft palate (P37, P38) and short philtrum (P21, P22, P31, P34, and P35). Four children also showed hypertrichosis (P2, P17, P28, and P29).

Neuromuscular findings

Muscle wasting was first noticed at age 11 years and shows a clear increase with age (Table 2; Figure 2B). A muscle biopsy was available in patient 23 only, which demonstrated neurogenic atrophy. Contractures were first noticed at age 10 years and became more prominent with age as well. Hypertonia probably due to spasticity, was present in two patients (P25, P42) and was recognized first in adulthood. Infrequent findings included joint hypermobility of the upper limbs (P17, P26, P30), and Chiari malformation (P30).

Systemic findings

Someother findings show a difference in occurrence with age as well (Table 2), although some can occur at an early age as well. Dysplastic hip joint changes, cystic bone lesions, and cataract were found only in adults. Infrequent reported findings included reduced tear production (P2, P13), microphthalmia (P25, P38), unilateral blindness due to glaucoma (P2), kyphosis (P20, P32, P39), hyperlordosis (P32), pectus abnormalities (P10, P14), pulmonary artery stenosis in an adult patient (P16), small penis (P37), hypothyroidism (P3, P5, P18), and GH deficiency (P10, P37). Baseline adrenal cortex hormones were also checked in four patients, with normal results. Alpha‐fetoprotein (AFP) levels showed that levels were typically elevated but not in all affected individuals, and did not show a marked change over time (Figure 2C). One patient showed selective IgG2 deficiency (P31). One patient developed a testis carcinoma at 27 years and a (fatal) seminoma in the other testis at 40 years of age Another male developed a germ cell tumor at 28 years and also a seminoma at that age.

Metabolic investigations

Plasma amino acids were investigated in nine patients and tested normal. Plasma acylcarnitines were available in four patients showing increased C2, C4OH, C5OH, C6OH, C14, and C14:2 levels in two of them. Mild ketonuria was found in four patients, and two of these four (P14, P21) also showed mild dicarboxylic aciduria, together with increased ethylmalonic acid and glutaric acid excretion.

Molecular testing

ZBTB20 variants for all reported individuals are tabulated in Table 3, and depicted in Figure 3. None was present in the public database gnomAD (Table S2). All variants were either missense changes or insertion/deletions, acting as a frameshift, and have been classified as class 4 and class 5 according to the criteria of the American College of Medical Genetics. No variants were detected in the BTB site or in the distal part of the ZnF_C2H2 site in individuals with a phenotype‐fitting PS. No indications for mosaicism were detected in any patient. In all patients in whom one or both parents were available (n = 26), the variant was found to be de novo. No familial occurrence has been reported. Mean paternal age at birth was 33.9 ± 7.5 years; mean maternal age at birth was 30.3 ± 4.9.

TABLE 3

Molecular characteristics of the 42 individuals with Primrose syndrome, compared to the major clinical manifestations

Patient		Variant type	Nucleotide change	Protein change	Macrocephaly	Moderate/severe ID	Autism	Self‐injurious behavior	Distal muscle wasting	Cystic bone lesions	Cataract	Reference
Number	Age
1	0.9 y	Missense	c.626A>G	p.Gly209Arg	−	−	+	+	−	−	−	Current study
2	32 y	Missense	c.1739G>A	p.Cys580Tyr	−	+	+	+	+	n.a.	+	Current study
3	3 y	Missense	c.1749C>G	p.Cys583Trp	−	+	−	−	−	n.a.	n.a.	Cleaver et al 16
4	4.7 y	Missense	c.1760 T>C	p.Phe587Ser	+	+	−	+	−	−	−	Current study
5	3 y	Missense	c.1766C>A	p.Ala589Asp	n.a.	+	+	−	−	−	−	Current study
6	35 y	Missense	c.1768A>C	p.Lys590Gln	+	−	−	−	n.a.	−	−	Posmyjk et al 2011 8
7	31 y	Missense	c.1771C>G	p.Gln591Glu	+	+	+	+	+	+	−	Mathijssen et al 5
8	9 y	Missense	c.1787A>Gc.2002G>A	p.His596Argp.Gly668Arg	+	+	+	−	+	n.a.	−	Casertano et al 12
9	9 y	Missense	c.1794C>G	p.Phe598Leu	+	−	−	−	−	−	−	Current study
10	15.2 y	Missense	c.1800C>G	p.His600Gln	+	+	n.a.	−	−	−	−	Grimsdottir et al 2018 15
11	49 y	Missense	c.1802C>T	p.Thr601Ile	−	−	−	+	+	+	−	Cordeddu et al 10
12	45 y	Missense	c.1805G>C	p.Gly602Ala	+	+	−	−	+	−	−	Cordeddu et al 10
13	2.2 y	Missense	c.1811A>C	p.Lys604Thr	+	n.a.	+	−	−	−	−	Cordeddu et al 10
14	5.3 y	Missense	c.1813C>T	p.Pro605Ser	−	+	n.a.	+	−	n.a.	−	Current study
15	2.6 y	Missense	c.1822C>T	p.Cys608Arg	+	+	−	−	−	n.a.	n.a.	Ferreira et al 17
16	16 y	Missense	c.1832G>A	p.Cys611Tyr	+	+	+	−	−	n.a.	−	Alby et al 13
17	11 y	Missense	c.1837C>T	p.Arg613Cys	+	−	+	−	−	n.a.	−	Alby et al 13
18	5.3 y	Missense	c.1847C>Tc.2221G>A	p.Ser616Phep.Gly741Arg	+	+	+	+	−	n.a.	−	Mattioli et al 11
19		Missense	c.1850 T>C	p.Leu617Ser	+	+	−	+	−	n.a.	n.a.	Cleaver et al 16
20	30 y	Missense	c.1861C>T	p.Leu621Phe	+	n.a.	−	−	+	−	−	Carvalho et al 7
21	3.1 y	Missense	c.1869G>C	p.Lys623Asn	+	+	+	+	−	−	−	Casertano et al 12
22	1.1 y	Missense	c.1871A>C	p.His624Pro	+	n.a.	+	−	−	−	−	Current study
23	2.5 y	Missense	c.1873A>G	p.Met625Val	+	−	n.a.	−	−	n.a.	−	Current study
24	27 y	Missense	c.1873A>G	p.Met625Val	−	+	+	−	+	n.a.	n.a.	Ferreira et al 17
25	49 y	Missense	c.1876G>A	p.Val626Met	n.a.	+	n.a.	n.a.	+	+	n.a.	Battisti et al 4
26	8 y	Missense	c.1879A>G	p.Thr627Ala	−	n.a.	−	−	−	n.a.	−	Cleaver et al 16
27	9.3 y	Missense	c.1898C>T	p.Ala633Val	+	+	+	n.a.	−	n.a.	−	Current study
28	IUD	Missense	c.1906 T>C	p.Cys636Arg	−	n.a.	n.a.	n.a.	n.a.	n.a.	n.a.	Alby et al 13
29	3.4 y	Missense	c.1931C>T	p.Thr644Ile	+	n.a.	n.a.	n.a.	n.a.	n.a.	n.a.	Stellacci et al 14
30	11.3 y	Missense	c.1943C>T	p.Ser648Phe	+	+	−	−	−	n.a.	n.a.	Cleaver et al 16
31	6 y	Missense	c.1945C>T	p.Leu649Phe	−	+	+	n.a.	−	n.a.	−	Yamamoto‐Shimojima et al 25
32	13.4 y	Missense	c.1967A>G	p.His656Arg	+	+	−	+	−	n.a.	n.a.	Cleaver et al 16
33	5.7 y	Insertion/deletion	c.1203del	p.Asp401fsGlufs*26	+	+	+	−	−	n.a.	−	Current study
34	12.4 y	Insertion/deletion	c.1844_1846del	p.615_616del	+	−	+	−	−	−	n.a.	Current study
35	8.5 y	Insertion/deletion	c.1024delC	p.Gln342Serfs*42	+	−	n.a.	−	n.a.	n.a.	n.a.	Stellacci et al 14
36	11 y	Insertion/deletion	c.1568delC	p.Pro523fs	−	−	−	−	−	n.a.	n.a.	Current study
37	13 y	Insertion/deletion	c.1568delC	p.Pro523fs	+	−	+	−	−	n.a.	n.a.	Current study
38	53 y	Insertion/deletion	Del11rs12275693– rs1442927		−	+	−	+	+	+	+	Dalal et al 6
39	31 y	n.a.	n.a.	n.a.	+	+	n.a.	n.a.	+	n.a.	+	Liebrecht et al 9
40	33 y	n.a.	n.a.	n.a.	+	+	n.a.	n.a.	+	+	+	Primrose 1
41	39 y	n.a.	n.a.	n.a.	+	+	n.a.	n.a.	+	+	+	Collacott et al 2
42	43 y	n.a.	n.a.	n.a.	−	n.a.	n.a.	n.a.	+	−	+	Lindor et al 3

Note: + present; − absent; n.a. not available.

Abbreviation: IUD, intrauterine demise.

FIGURE 3

Schematic overview of the ZBTB gene and localization of mutations. It is noteworthy that patient carrying p.Gln209Arg mutation showed no macrocephaly and no ID. Autism and self‐injurious behavior were recorded [Colour figure can be viewed at wileyonlinelibrary.com]

Dalal et al

Genotype‐phenotype correlation

The genotype was available for 38 patients (Table 3). Obviously the phenotype in the four patients reported before the causative gene was found, was more severe due to ascertainment bias. No clear genotype‐ phenotype correlation was detected. Some individuals with a variant in exon 1 (P6: c.1768A>C; P9: c.1794C>G) and in exon 5 (P34: c.1844_1846del; P19: c.1861C>T) showed a less severe ID, and some also only a limited number of the other characteristics of PS. However, other patients carrying variants in nearby base pairs showed the classical phenotype. The difference in age of the affected individuals and the progressive nature of the findings further hamper to correlate phenotype and genotype reliably.

DISCUSSION

We present a series of hitherto unpublished individuals with PS and summarize the findings of these individuals and those that have been reported in literature. The present study confirms that PS can present as an overgrowth syndrome with respect to brain growth (71%), but increased growth in height and weight is less marked and present in a minority of the patients (21%). Indeed, some females grow below the third centile for height and weight. The growth pattern is already present at birth and the subsequent overgrowth is non‐progressive.

The cardinal findings of PS are the ID (mild 16%, moderate‐severe 84%), mildly increased growth (height and weight between 50th and 90th centile, macrocrania 78%), and as most characteristic signs the calcified external ears, sparse body hair, bone dysplasia, and distal muscle wasting. The calcification of the ears, cataract, torus palatinus, cystic bone lesions and muscle wasting with subsequently contracture formation are clearly age‐related and become often only apparent in puberty or thereafter, so percentages differ in the various age groups. Cognition does not seem to decline with age, although sufficiently detailed studies to conclude this with certainty are missing. Hearing loss is also common both in children and adults, mostly presenting as sensorineural hearing loss.

The progression in signs and symptoms with age may point to a metabolic disturbance. Biochemically, unexplained anemia, disturbed glucose metabolism, and increased AFP levels are cardinal features of PS. Further metabolic investigations demonstrated abnormal acylcarnitine and urine organic acid profiles in some PS individuals, including increased excretion of dicarboxylic acids, ethylmalonic and glutaric acids. In one individual (P8), this pattern became more abnormal with age. Over time, this patient showed progressive lipodystrophy and developed muscle wasting with limb atrophy by 11 years of age; at that time, Oral Glucose Tolerance test also showed impaired glucose tolerance. The findings suggest disruption of the mitochondrial fatty acid oxidation. One may speculate that this is linked to pleiotropic effects of ZBTB20 on lipid and glucose metabolism.19, 25 Mitochondrial dysfunction has been reported in Zbtb20 knock‐out mouse. Mitochondrial dysfunction has also been involved in the development of muscle atrophy and insulin‐resistance, type 2 diabetes, and cataract, but at the present, there is no proof that these signs can be explained in PS individuals due to mitochondrial malfunctioning. More detailed analyses of mitochondrial functioning are warranted.

Increased AFP levels constitute a remarkable sign in PS. It has been proposed that mutated ZBTB20 disrupts the AFP repression resulting in AFP increase and overgrowth. AFP levels appeared >2 SD higher than reference values by age during the first months of life and progressively decreased with age. Among the presently reported males, two adults developed a testis tumor. No female developed neoplasm. Despite reports of ZBTB20 expression being associated with tumorigenesis, including gastric cancer and hepatocellular carcinoma, it remains unclear whether an increased risk of malignancies is part of this syndrome.

To evaluate whether ZBTB20 variants are more common in men with testicular germ cell tumor (TGCT), we interrogated WES data from lymphocyte‐derived DNA from 919 TGCT cases of Western European ancestry (comprising 306 familial and 613 unselected TGCT cases) and 1609 healthy controls of Western European ancestry from the UK 1958 Birth Cohort, all analyzed via the same pipeline.33, 34 We compared between TGCT cases and healthy controls, the frequency of high quality, rare (minor allele frequency [MAF] < 0.01) non‐synonymous variants. In the TGCT series, three rare non‐synonymous ZBTB20 variants [p.(Thr514Ala), p.(Ala693Val), and p.(Gly712Val)] were identified in the constitutional DNA of men with familial TGCT and one in a man with non‐familial TGCT [p.(Gly712Val)]. These men developed their seminoma or teratoma at ages 28, 28, 32, and 33 years, respectively. No further data regarding serum biomarkers or clinical phenotype were available for these patients. No rare non‐synonymous ZBTB20 variants were detected in 1609 healthy controls. Paired tumor germline WES data were available for an additional 179 TGCT cases: no ZBTB20 variants were detected in the constitutional or tumor DNA. Thus, the frequency of germline ZBTB20 mutation in TGCT cases would appear elevated (4/1098 in cases, 0/1609 in controls, P exact < .05). Still, the absolute risk of TGCT is low (1 in 200 in Western European males, lower in other ethnicities) and TGCT typically has an excellent outcome. The two males with PS who developed testicular tumors have died because of their tumors. There is no recognized protocol for TGCT surveillance established as effective for subpopulations at significant elevation of risk (such as family history, prior contralateral disease, or cryptorchidism). In addition, self‐examination is not feasible in most men with PS. Accordingly, families and other caregivers of men with PS should be alerted to the possible modest elevation in relative risk of TGCT, reassured as to the low absolute risk, and advised regarding symptom awareness and testicular examination by caregivers.

There is no evident genotype‐phenotype correlation in the present series. However, numbers are small, and it may still be that if a larger series can be evaluated this will become clear.

A dominant‐negative effect of missense variants has been previously hypothesized. Cleaver et al provided very limited information on an individual with a de novo c.505G>C [p.(Glu169Gln)] variant in whom pathogenicity remained uncertain, presumably because the phenotype did not resemble PS. We follow a patient (not included in the present series) with ZBTB20 variant c.1775A>G [p.(Asn519Ser)] detected by WES because of unexplained mild ID. This adult woman, age 39 years, has macrocephaly but otherwise none of the characteristic signs or symptoms of PS is present. She did show short stature and an unusual face. No other potentially pathogenic variants have been detected by WES, and the ZBTB20 variant is absent in her parents. It remains uncertain whether the variant is pathogenic. If so, it indicates that ZBTB20 variants can lead to ID and brain anomalies without the other characteristics of PS. In this respect, it may be of interest that two individuals with nearby located variants (P6: c.1768A>C; P9: c.1794C>G) show a relatively mild phenotype with less severe ID as well. Data suggest that patients with frameshift variants may show a milder phenotype. However, the small number of patients and limited data hamper reliable conclusions on genotype‐phenotype correlations.

A major limitation of the present study is its retrospective nature. Early clinical data were sometimes lacking as the clinical suspicion for PS raised later in life. Additionally, several patients came to the attention of a physician only as adults, hampering a complete early clinical history.

We conclude that PS is an established clinical entity that is recognizable in adults but more difficult to recognize in infants and children. In a clinically suspicious child checking the AFP levels can be useful. The manifestations are progressive, and repeated evaluation for anemia, diabetes, and osteoporosis are indicated. At the present, there is no clear indication that cognition shows a decline with time as well. There may be an increased risk to develop testis tumors, and regular follow‐up for this from puberty onward seems indicated.

CONFLICTS OF INTEREST

The authors declare no potential conflict of interest.

Supplemental Table S1 Individual data of presently reported individuals with Primrose syndrome.

Click here for additional data file.

Supplemental Table S2 ZBTB20 new variants

Click here for additional data file.

Supplemental Figure 1 Growth pattern in length and weight of boys (A) and girls (B) and in head circumference in boys (C) and girls (D) of present series of individuals with Primrose syndrome.

Click here for additional data file.

Supplemental Figure 2 (A) Height SDS, (B) Weight SDS and (C) Head circumference SDS in children (black circle) and adults (gray square) with Primrose syndrome

Click here for additional data file.