Funding sources This work was supported by the National Institute for Health Research Translational Research Collaboration for Rare Diseases (NIHR-TRC) and by the NIHR Biomedical Research Centre based at Guy's and St Thomas’ NHS Foundation Trust and King's College London.
Recessive forms of congenital ichthyosis encompass a group of rare inherited disorders of keratinization leading to dry, scaly skin. So far, 13 genes have been implicated, but there is a paucity of data on genotype–phenotype correlation in some populations.
We compiled an English cohort of 146 individuals with recessive ichthyosis and assessed genotype–phenotype correlation.
Deep phenotyping was undertaken by history-taking and clinical examination. DNA was screened for mutations using a next-generation sequencing ichthyosis gene panel and Sanger sequencing.
Cases were recruited from 13 National Health Service sites in England, with 65% of patients aged < 16 years at enrolment. Pathogenic biallelic mutations were found in 83% of cases, with the candidate gene spread as follows: TGM1 29%, NIPAL4 12%, ABCA12 12%, ALOX12B 9%, ALOXE3 7%, SLC27A4 5%, CERS3 3%, CYP4F22 3%, PNPLA1 2%, SDR9C7 1%. Clinically, a new sign, an anteriorly overfolded ear at birth, was noted in 43% of patients with ALOX12B mutations. The need for intensive care stay (P = 0·004), and hand deformities (P < 0·001), were associated with ABCA12 mutations. Self-improving collodion ichthyosis occurred in 8% of the cases (mostly TGM1 and ALOX12B mutations) but could not be predicted precisely from neonatal phenotype or genotype.
These data refine genotype–phenotype correlation for recessive forms of ichthyosis in England, demonstrating the spectrum of disease features and comorbidities, as well as the gene pathologies therein. Collectively, the data from these patients provide a valuable resource for further clinical assessment, improving clinical care and the possibility of future stratified management.
What's already known about this topic?
Recessive forms of ichthyosis are rare but often difficult to diagnose.
Mutations in 13 genes are known to cause recessive forms of ichthyosis: ABCA12, ALOX12B, ALOXE3, CERS3, CYP4F22, LIPN, NIPAL4, PNPLA1, SDR9C7, SLC27A4, SULT2B1, ST14 and TGM1.
Some phenotypic features may associate with certain gene mutations, but paradigms for genotype–phenotype correlation need refining.
What does this study add?
The genotypic spectrum of recessive ichthyosis in England (based on 146 cases) comprises TGM1 (29%), NIPAL4 (12%), ABCA12 (12%), ALOX12B (9%), ALOXE3 (7%), SLC27A4 (5%), CERS3 (3%), CYP4F22 (3%), PNPLA1 (2%) and SDR9C7 (1%).
New or particular phenotypic clues were defined for mutations in ALOX12B, ABCA12, CYP4F22, NIPAL4, SDR9C7 and TGM1, either in neonates or in later life, which allow for greater diagnostic precision.
In around 17% of cases, the molecular basis of recessive ichthyosis remains unknown.
Recessive forms of ichthyosis include both syndromic and nonsyndromic disorders, with the term autosomal recessive congenital ichthyosis (ARCI) describing the nonsyndromic variants.1 The prevalence of ARCI has been estimated at 1–1·6 per 100 000 people in European populations.2, 3 ARCI is characterized by widespread scaly skin with or without erythema, and approximately 70% of individuals are born with a collodion membrane.1 ARCI was traditionally classified based on phenotype, into several subgroups: lamellar ichthyosis, congenital ichthyosiform erythroderma (CIE), harlequin ichthyosis (HI), self-healing collodion baby, acral self-healing collodion baby and bathing suit ichthyosis.1 Over time, the ‘self-healing’ phenotypic term was refined to the alternative nosology, self-improving collodion ichthyosis (SICI).4 Historically, ichthyosis prematurity syndrome (IPS) has also been included within the spectrum of ARCI, although it is now classified as a syndromic type of ichthyosis.1
However, in addition to clinical subtyping, recent advances in gene identification and genetic testing have generated the possibility for a molecular pathology-based classification of ARCI.1 Currently, there are 13 known genes implicated: ABCA12,5ALOX12B,6ALOXE3,6CERS3,7CYP4F22,8LIPN,9NIPAL4,10PNPLA1,11SDR9C7,12SLC27A4,13SULT2B1,14ST1415 and TGM1.16 Molecular characterization has shown that these genes mainly encode proteins in diverse pathways involved in lipid metabolism and transport within the stratum corneum, or formation of the cornified envelopes.17 Previous studies have also shown that TGM1 mutations account for 30–50% of cases of ARCI, with ALOXE3 and ALOX12B together accounting for a further 20–30%, NIPAL4 and CYP4F22 10% each, and ABCA12 5%, with the other genes all being much rarer causes of ARCI.3, 18, 19 However, in at least 15% of cases of ARCI no mutation is identifiable.3, 19
Previous attempts to establish robust genotype–phenotype correlation have been restricted by assessment of small numbers of cases and the genetic heterogeneity of ARCI.1 Based on current knowledge, TGM1 mutations are often associated with dark plate-like scale, although some TGM1 mutations lead to marked erythema.20, 21ALOX12B and ALOXE3 mutations typically result in erythema with fine white scale and often give a mild phenotype.22NIPAL4 mutations may be associated with yellow palmoplantar keratoderma (PPK) and types of scale that may vary from dark to white.23ABCA12 mutations may be associated with a severe disease course with high rates of infant mortality, severe erythema and scaling, ectropion, eclabium and absent retroauricular folds and alopecia.24, 25 However, ABCA12 mutations may also have a phenotype with dark scale or a milder CIE appearance.25 Recent reports in rarer forms of ARCI with CERS3, PNPLA1 and ST14 mutations have shown great variability in phenotype.26-29
A better understanding of the phenotypic and genotypic spectrum of ARCI, as well the syndromic forms of recessive ichthyosis, is fundamental to improving clinical care through pre-emptive clinical monitoring, optimal prescribing and a platform for stratified or personalized clinical care. Therefore, the aim of this study was to expand genotype–phenotype correlation in recessive forms of congenital ichthyosis. Our study focuses mainly on individuals with ARCI but also includes individuals with IPS, a syndromic form of recessive ichthyosis.1
Materials and methods
The study was approved by the West Midlands Solihull Ethics Committee, with further Research and Development permission granted through the NIHR UK Rare Genetic Disease Research Consortium Musketeers' Memorandum, which consolidates a single approval for multiple sites in a national study. Data were collected from 13 National Health Service (NHS) trusts within England. Sites were chosen following a national consultation with consultant dermatologists to identify NHS trusts with appropriate cases of ichthyosis. Inclusion criteria consisted of a phenotype of suspected autosomal recessive ichthyosis or genetic diagnosis based on mutations in the following genes: ABCA12, ALOX12B, ALOXE3, CERS3, CYP4F22, LIPN, NIPAL4, PNPLA1, SLC27A4, ST14, TGM1 and chromosome location 12q11 (now identified as SDR9C7).12SULT2B1 was identified after study initiation and subsequently added as an inclusion criterion.
All age groups were included. Patients, and parents or guardians of children under the age of 17 years, provided written informed consent, with patient assent where appropriate. Deceased patients were enrolled with the permission of first-degree relatives. Patients with autosomal dominant and X-linked inherited forms of ichthyosis, and ichthyosis thought to be secondary to other underlying conditions were excluded. Those with onset of scaling at > 6 months of age were also excluded.
Clinical details were recorded in an electronic database (MedSciNet). Phenotypic data collected included demographics; birth history; past medical history; features of skin, hair, nails and oral mucosae; evolution of disease; current symptoms; response to treatments; and genotype if known. The aim was to generate information about autosomal recessive ichthyosis in the neonatal period as well as at later time points.
Polymerase chain reaction and Sanger sequencing
Following informed consent, DNA was obtained from peripheral blood in ethylenediaminetetraacetic acid tubes or from saliva samples taken with Oragene kits (GenoTek, Kanata, ON, Canada). DNA was extracted using precipitation with ethanol. DNA samples were initially screened for TGM1 mutations by means of Sanger sequencing, given that this is the most frequently mutated gene in ARCI. Briefly, polymerase chain reaction (PCR) amplification was performed using Amplitaq gold DNA polymerase (Thermo Fischer Scientific Inc., Waltham, MA, U.S.A.) and 14 primer pairs (Sigma-Aldrich, St Louis, MO, U.S.A.) spanning all coding exons and intronic flanking regions (Table S1; see Supporting Information). PCR products were purified with ExoSAP_IT (GE Healthcare, Chicago, IL, U.S.A.) and were bidirectionally sequenced using BigDye Terminator v3·1 chemistry on a 3730xl DNA analyser (both Thermo Fischer Scientific Inc.). Mutation-position reporting was in reference to TGM1 mRNA accession number NM_000359·2. Sequencing chromatograms were visualized using Sequencher 5·0 software tool (Gene Codes Corporation, Ann Arbor, MI, U.S.A.).
Next-generation sequencing and data analysis
DNA was extracted from saliva or peripheral blood, as described above. DNA concentration was quantified using a Qubit fluorometer (Thermo Fischer Scientific Inc.). Target enrichment was performed using the TruSeq Custom Amplicon kit (Illumina, San Diego, CA, U.S.A.). DesignStudio (Illumina) was used for library design. The designed library targeted all coding exons, at least 20 bp of the intron at each intron–exon boundary, and up to 50 bp of the 3′-untranslated regions of the following 38 genes: ABCA12, ABHD5, AGPS, ALDH3A2, ALOX12B, ALOXE3, AP1S1, ARSE, CERS3, CLDN1, CYP4F22, EBP, ELOVL4, GJB2, GJB3, GJB4, GJB6, KRT1, KRT10, KRT2, KRT9, LIPN, LOR, NIPAL4, PEX7, PHYH, PNPLA1, PNPLA2, POMP, SLC27A4, SNAP29, SPINK5, ST14, STS, TGM1, TGM5, VPS33B and ZMPSTE24. Variant calling was performed by platform GATK HaplotypeCaller (Broad Institute, Cambridge, MA, U.S.A.). Identification of pathogenic variants was performed using in silico pathogenicity prediction tools: CADD (https://cadd.gs.washington.edu/info), SIFT (https://sift.bii.a-star.edu.sg) and PolyPhen-2 (http://genetics.bwh.harvard.edu/pph2). Variants were visualized using integrated genome viewer (IGV) and variants with suboptimal coverage (< 10 reads) were confirmed by means of Sanger sequencing.
Sequencing of SDR9C7 and SULT2B1
Cases in which no mutations were identified via next-generation sequencing were also screened by Sanger sequencing for possible mutations in SDR9C7 and SULT2B1, which were not included in the gene panel. Primers (Sigma-Aldrich) were designed using oligoevaluator.com (Table S1; see Supporting Information).
Fisher's exact tests were used to identify significant phenotypic associations between cases of ARCI with different genotypes. Ordinal logistic regression was used to model the odds of increasing scale size and presence of TGM1 mutation. R (version 3·3·1; R Foundation, Vienna, Austria) was used for analyses, and an alpha level of 0·05 was considered significant.
Individuals with a suspected diagnosis of autosomal recessive ichthyosis (n = 291) were identified and invited to join the study (Fig. S1; see Supporting Information). Recruitment occurred between April 2014 and November 2016 (n = 163). Subsequently, 17 cases were withdrawn. Reasons included missing or insufficient DNA samples or data indicating an alternative likely diagnosis; one case was withdrawn following the participant's request. The final cohort for evaluation comprised 146 patients from 124 families, who had both gene screening and phenotypic data. Two deaths of enrolled participants occurred during the course of the study: one infant with an ABCA12 mutation and another from whom no DNA sample was available for genotyping; the latter case was subsequently withdrawn. Individual details of phenotypes and genotypes are provided in Table S2 (see Supporting Information). Clinical illustration of key phenotypic features is shown in Figure 1.
Key phenotypic features from study participants. (a) Patchy erythema associated with NIPAL4 mutations. (b) Hand abnormalities associated with ABCA12 mutations. (c) Overfolded ear sign associated with 43% of ALOX12B mutations. (d) Onycholysis associated with homozygous mutation in SDR9C7, c.355G>A, p.Glu119Lys.
In our cohort, 50% of the cases were white, 34% Asian and 7% black. Overall, 65% were aged < 16 years at the time of enrolment, 54% were female and 37% came from related-parent families (Table 1). The most common phenotype was nonbullous congenital ichthyosiform erythroderma (NBCIE, 38%), followed by lamellar ichthyosis (28%), other unspecified ichthyosis (9%), HI (9%), SICI (7%), IPS (5%) and bathing suit ichthyosis (4%).
Table 1. Demographics and phenotypic classification of an English cohort of patients with autosomal recessive ichthyosis (n = 146)
Overall, 31% of the cases were born preterm (< 37 weeks). At birth, 71% were born with a collodion membrane. The majority were erythematous at birth, with only 13% of cases reporting absence of erythema. Forty-nine per cent of the cases were born with ectropion and 25% had eclabium; however, this resolved in 51% and 86% of the cases, respectively. Symptoms including pruritus, anhidrosis and heat intolerance were common across multiple genotypes and were not confined to phenotypes with thick scale. Other medical conditions associated most frequently in this cohort were hearing loss (35%) (predominantly conductive), recurrent skin infections (24%), vitamin D deficiency (14%), arthritis or arthralgia (8%), developmental delay (being less mentally or physically developed than children of a similar age) (8%), anaemia (6%), cardiac manifestations (5%) (including patients with valvular anomalies, rhythm disturbances, one case of ventricular septal defect and heart failure in the neonatal period and one myocardial infarction, aged 44 years) and conjunctivitis (5%). Thirty-eight per cent of cases had been treated with retinoids and 6% reported stopping retinoids due to side-effects or lack of benefit.
Mutations were detected in 121 of the 146 cases (83%), with the following genes implicated: TGM1 (29%), NIPAL4 (12%), ABCA12 (12%), ALOX12B (9%), ALOXE3 (7%), SLC27A4 (5%), CERS3 (3%), CYP4F22 (3%), PNPLA1 (2%) and SDR9C7 (1%) (Fig. 2). One case had biallelic pathogenic compound heterozygous mutations in TGM1 (IVS7–2A>G and exon 3: c.401A>G, p.Tyr134Cys) and two additional heterozygous missense mutations in ABCA12 (exon 3: c.259T>C, p.Tyr87His and c.252C>A, p.Asp84Glu) potentially contributing to the HI phenotype.
Pie chart showing the genotypic spectrum of the 146 cases studied.
A summary of the genotype–phenotype correlation is provided in Table S3 (see Supporting Information).
There were 42 cases with TGM1 mutations, of which 90% were born with a collodion membrane. Those with TGM1 mutations were more likely to present with collodion at birth (P = 0·018) than cases without TGM1 mutation. While erythema at birth, following shedding of the collodion, was common in TGM1 cases, this subsided in the majority over the first year of life. TGM1 mutations were significantly associated with darker scale that was adherent and plate like. Scale colour depended on skin type, being light brown in the patients with white skin type and darker in Asian and African skin types. The odds of presenting with large plates of scale on the lower legs were greater for cases with TGM1 mutation than for cases without (P < 0·001, ordinal logistic regression). This feature was seen from as young as 3 months of age. PPK was seen in 50% of cases with TGM1 mutations and was often thick and fissured.
The most common TGM1 mutation was c.877–2A>G (IVS5–2A>G), a splice-site variant found in nine patients (21% of all TGM1 mutant alleles). However, six patients with this same mutation presented with more erythema and less scale leading to clinical classification as NBCIE, while three others were classified as having lamellar ichthyosis (dark plate-like scale).
There were 18 patients with NIPAL4 mutations, 28% of whom were born with a collodion membrane. Of note, 17 cases from various ethnic backgrounds (nine white British, four white European, one Turkish, two Asian, one black African) were either compound heterozygous or homozygous for the p.Ala176Asp variant. This amino acid substitution has a reported allele frequency of 6·65 × 10−4 in the Genome Aggregation Database (gnomAD) and it has previously been reported as pathogenic.30 Moreover, p.Ala176Asp is predicted to be ‘deleterious’ by the CADD pathogenicity software tool (CADD score 29·4), further supporting its pathogenicity. NIPAL4 mutations were usually associated with a phenotype of widespread coarse grey-white scale. Patchy erythema, which was almost plaque like with reasonably well-demarcated nonerythematous areas, was associated with NIPAL4 mutations, and four of 18 patients had a past medical history of psoriasis. In addition, PPK was present in one-third of patients, which was predominantly focal, localized to pressure areas and yellow in colour.
There were 17 patients with ABCA12 mutations, the majority of whom were described as harlequin babies (76%). Most cases with ABCA12 mutations were born with a thick opaque, armour-like membrane, which after shedding revealed moderate-to-severe erythema (P < 0·001) and large white scale (P = 0·025). High rates of ectropion (P = 0·001), eclabium (P = 0·014), level 3 care (P = 0·007) and hand deformities (P < 0·001) were noted, which were significantly more common than in cases without ABCA12 mutations. Fingers exhibited flexion deformities, loss of pulp and proximal webbing giving rise to a ‘long palm–short fingers’ appearance. ABCA12 mutations were significantly associated with alopecia (P < 0·001), usually at the hair margins, and absent retroauricular folds (P < 0·001). Arthritis or arthralgia was significantly associated with ABCA12 mutations (P < 0·001). Retinoid usage was frequent (P < 0·001), with 82% of cases having taken a retinoid at some stage. Four patients had a milder CIE phenotype; of these, one had a homozygous missense mutation and the others were compound heterozygous for a nonsense mutation and splice-site mutations. Two cases had prominent brown scale as a phenotypic feature: one case had a homozygous missense mutation in the first ATP binding domain (exon 28: c.4139A>G, p.Asn1380Ser) and the other had additional concurrent biallelic mutations in TGM1 as mentioned above.
There were 14 patients with ALOX12B mutations, 71% of whom were born with a collodion membrane. Cases were typically erythematous with fine, white spicules of nonadherent scale. Six (43%) showed marked improvement through childhood (P < 0·001). An anteriorly overfolded ear was present from birth in six cases (43%). This finding was not seen in cases with mutations in other genes and therefore it might serve as a useful clinical diagnostic marker. PPK was present in four cases and was mild and diffuse. Ridging and pseudoclubbing of nails were frequent.
There were 11 patients with ALOXE3 mutations, 45% of whom were born with a collodion membrane. The predominant phenotype was similar to that with the other lipoxygenase, ALOX12B, with mild-to-moderate erythema and fine white nonadherent scale. Although relatively mild clinically, none was described as truly ‘self-healing’ because of residual spicules of scale; however, several did show improvement with age. Mild PPK was present in three cases. Nail changes with ridging and pseudoclubbing were common.
There were seven patients with SLC27A4 mutations, although in three cases (now adults) birth history details were not recallable. For the other four cases, collodion membrane at birth was seen in two patients, but all four presented with capital hyperkeratosis. These four cases were born prematurely (< 37 weeks) (P = 0·02) and had respiratory impairment necessitating respiratory support (P < 0·001), and in three cases level 3 care (neonatal intensive care) was needed. Erythema was not reported beyond the neonatal period and a follicular ‘goosebump’-like scale was common. The phenotype improved spontaneously and was generally mild after the neonatal period. No individuals had PPK. The key clinical point about IPS is the risk of neonatal asphyxia. Moreover, the very specific course of IPS should discriminate it from a nonsyndromic disease; it is better classified as a syndromic type of ichthyosis.1
There were four patients with CERS3 mutations, all of whom were born with a collodion membrane. Generalized monomorphic grey–light brown, flat, diamond-shaped scale with mild erythema was observed. All four were of Pakistani ancestry and all described anhidrosis and heat intolerance. No hair abnormalities were observed.
There were five patients with CYP4F22 mutations, all of whom were born with a collodion membrane. In most of the cases, the scale was dark grey–light brown and nonadherent, with minimal erythema. None had ectropion or eclabium. Scale was most severe in the flexural areas including the neck, groin, axillae and across the trunk. In contrast to cases with TGM1 mutations, the scale was more monomorphic and did not increase in size towards the lower extremities. Two siblings showed improvement in scaling through childhood.
There were three patients with PNPLA1 mutations, with various types of scale: one with darker scale and two cases with white scale. The phenotype was generally mild with some unaffected areas. There was no erythema and palmoplantar skin was normal.
We identified a single case with a new homozygous mutation in SDR9C7 (c.355G>A; p.Glu119Lys). This patient presented with a fine semiadherent greyish-white scale and notably had onycholysis affecting several fingernails, present since the age of 1 year but improving during childhood. Mycological assessment was not performed; clinically, the appearances resembled a distal onychomycosis or possibly postviral onychomadesis. Hair examination showed normal growth with no evident alopecia; palmoplantar skin appeared normal.
Self-improving collodion ichthyosis
SICI was diagnosed when following shedding of the collodion membrane the majority of the skin was normal or near normal without showing residual signs of ichthyosis. Skin xerosis or focal scaling was occasionally seen.4 SICI occurred in 11 cases, comprising five cases with TGM1 mutations and five with ALOX12B mutations, as well as one case in which no mutation was found. Of note, SICI usually could not be predicted from the mutations identified, although one compound heterozygous ALOX12B case of SICI showed variants previously reported with a SICI phenotype, namely the missense mutations p.Tyr521Cys and p.Ala597Glu.4 A further 13% of cases showed a more gradual improvement in scale and/or erythema throughout childhood, with a range of mutations but no consistent genotype correlation.
We gathered deep phenotypic and genotypic data on a large cohort of 146 English individuals with autosomal recessive ichthyosis from 124 families. Using a targeted gene panel and individual gene sequencing we identified biallelic mutations in 83% of cases. Our mutation detection rate is similar to findings in other international cohorts over the last decade.18, 19, 29 However, of note, even with the advent of next-generation sequencing, around 15–20% of patients continue to have no defined molecular pathology.29 This persistent shortfall in mutation detection raises the possibility of inherited abnormalities in other regulatory or noncoding parts of the genome, although supportive evidence for this type of alternative genomic pathology is currently lacking.
One of the key objectives of deep phenotyping autosomal recessive ichthyosis is to try to improve clinical management in recognizing clinical associations or complications and in devising better clinical care plans. Our study was able to identify several clinical characteristics that were associated with particular genotypes (Table 2). For example, anteriorly overfolded ears were noted with ALOX12B mutations, babies with ABCA12 mutations typically demonstrated hand deformities and were more likely to require neonatal intensive care, and onycholysis may be a predictor of SDR9C7 mutations.
Table 2. Summary of key clinical features providing diagnostic clues to the genotype in an English cohort of patients with autosomal recessive ichthyosis
Need for level 3 care
Absent retroauricular fold
Anteriorly overfolded ear
Eclabium at birth
Ectropion at birth
Older age groups
Moderate-to-severe erythema on body
Larger plates of scale on lower body
Focal yellow keratoderma
Arthritis or joint movement limitations
TGM1, ABCA12, NIPAL4
Monomorphic brown scale
TGM1, ALOX12B, SLC27A4
Nevertheless, we also demonstrated that identical mutations in a specific gene can have diverse phenotypic consequences. Notably, some patients with the TGM1 acceptor splice-site mutation c.877–2A>G were classified as having NBCIE or LI because of variable severity of the erythema. Other studies have also shown considerable phenotypic heterogeneity within a single genotype.22 While a satisfactory explanation for genotype–phenotype disparity in such cases is lacking, we did find evidence for potential genetic modifiers. Notably, we observed HI with brown scale on a genetic background of pathogenic variants in two genes, TGM1 and ABCA12. Indeed, some candidate genes are already known to exert modifying effects on the expression of others.30
One potential common genetic modifier that we did not address specifically in our study is the profilaggrin gene, FLG, mutations in which underlie the semidominant condition ichthyosis vulgaris and are further implicated in a spectrum of atopic and allergic skin manifestations.31FLG was not incorporated into our gene panel for technical reasons but, clinically, ichthyosis vulgaris was excluded both by the assessment of the referring dermatology consultants and again by the recruiting study team. We also excluded patients with onset of ichthyosis at > 6 months of age. All parents of participating children were asked if they had any affected skin (to exclude autosomal dominant or semidominant inheritance) and recruited adult patients were asked whether their parents were affected; none was included in our cohort. Nevertheless, as with all other dermatoses, inherited or acquired, the potential impact of intragenic copy number variation or loss-of-function mutations in FLG, with superimposed xerosis, ichthyosis vulgaris or atopic dermatitis, should always be remembered in assessing individual patients.
It is important for clinicians managing patients with autosomal recessive ichthyosis to be aware of associated pathologies or comorbidities. We confirm previous observations of an association of arthritis with ABCA12 mutations.25 Of new potential concern is the finding that around 5% of our cohort had cardiac anomalies, although no specific mutant gene association was identified and the clinical pathology was highly variable (ischaemic events, electrical disturbances and structural defects). Based on these findings, no specific monitoring recommendation can be made for now, other than to raise clinical awareness. Conversely, our data endorse current guidelines for possible complications affecting the ears: assessments should be carried out every 6 months in those under 6 years old, with specialist referral in cases of pruritus or pain in the ear, ear discharge, a feeling of clogged ears or hearing loss.32 Likewise, routine monitoring should include assessment of vitamin D levels and possible ophthalmic complications.32 In our cohort, we found rates of anaemia similar to those in the general population.33
Currently, management of autosomal recessive ichthyosis is largely supportive. However, new treatments specific to the inherent skin pathology are in clinical development, such as topical transglutaminase 1 protein replacement therapy,34 which emphasize the likely future importance of stratification by genotype in delivering tailored treatments. Our study findings in this English cohort add new phenotypic data that will sharpen clinical skills, as well as creating a database for future translational research. This can evolve into better clinical understanding and care of patients with autosomal recessive ichthyosis, as well as new personalized therapies.
We would like to thank all of the NHS Trusts that took part across England. We would also like to acknowledge all of the patients and families who have kindly contributed samples, and acknowledge the Ichthyosis Support Group for help with recruitment. We were grateful for the support of the NIHR CRN in recruiting to the study and the National NIHR CRN Genetics Specialty for support through the NIHR U.K. Rare Genetic Disease Research Consortium Agreement (Musketeers’ Memorandum).
Table S3 Genotype–phenotype correlation in 146 cases of autosomal recessive ichthyosis.
Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.
1Oji V, Tadini G, Akiyama M et al. Revised nomenclature and classification of inherited ichthyoses: results of the First Ichthyosis Consensus Conference in Soreze 2009. J Am Acad Dermatol 2010; 63: 607– 41.
2Hernandez-Martin A, Garcia-Doval I, Aranegui B et al. Prevalence of autosomal recessive congenital ichthyosis: a population-based study using the capture–recpature method in Spain. J Am Acad Dermatol 2012; 67: 240– 4.
3Pigg MH, Bygum A, Ganemo A et al. Spectrum of autosomal recessive congenital ichthyosis in Scandinavia: clinical characteristics and novel and recurrent mutations in 132 patients. Acta Derm Venereol 2016; 96: 932– 7.
4Vahlquist A, Bygum A, Ganemo A et al. Genotypic and clinical spectrum of self-improving collodion ichthyosis: ALOX12B, ALOXE3, and TGM1 mutations in Scandinavian patients. J Invest Dermatol 2010; 130: 438– 43.
15Basel-Vanagaite L, Attia R, Ishida-Yamamoto A et al. Autosomal recessive ichthyosis with hypotrichosis caused by a mutation in ST14, encoding type II transmembrane serine protease matriptase. Am J Hum Genet 2007; 80: 467– 77.
20Farasat S, Wei MH, Herman M et al. Novel transglutaminase-1 mutations and genotype–phenotype investigations of 104 patients with autosomal recessive congenital ichthyosis in the USA. J Med Genet 2009; 46: 103– 11.
22Rodriguez-Pazos L, Ginarte M, Fachal L et al. Analysis of TGM1, ALOX12B, ALOXE3, NIPAL4 and CYP4F22 in autosomal recessive congenital ichthyosis from Galicia (NW Spain): evidence of founder effects. Br J Dermatol 2011; 165: 906– 11.
23Alavi A, Shahshahani MM, Klotzle B et al. Manifestation of diffuse yellowish keratoderma on the palms and soles in autosomal recessive congenital ichthyosis patients may be indicative of mutations in NIPAL4. J Dermatol 2012; 39: 375– 81.
26Vahidnezhad H, Youssefian L, Saeidian AH et al. Gene-targeted next generation sequencing identifies PNPLA1 mutations in patients with a phenotypic spectrum of autosomal recessive congenital ichthyosis: the impact of consanguinity. J Invest Dermatol 2017; 137: 678– 85.
28Dahlqvist J, Klar J, Hausser I et al. Congenital ichthyosis: mutations in ichthyin are associated with specific structural abnormalities in the granular layer of epidermis. J Med Genet 2007; 44: 615– 20.
29Youssefian L, Vahidnezhad H, Saedian AH et al. Autosomal recessive congenital ichthyosis: genomic landscape and phenotypic spectrum in a cohort of 125 consanguineous families. Hum Mutat 2019; 40: 288– 98.
30Li H, Lorie EP, Fischer J et al. The expression of epidermal lipoxygenases and transglutaminase-1 is perturbed by NIPAL4 mutations: indications of a common metabolic pathway essential for skin barrier homeostasis. J Invest Dermatol 2012; 132: 2368– 75.
33 Global Burden of Disease Pediatrics Collaboration, Kyu HH, Pinho C et al. Global and national burden of diseases and injuries among children and adolescents between 1990 and 2013: findings from the Global Burden of Disease 2013 Study. JAMA Pediatr 2016; 170: 267– 87.
Liangshan Li, Wenmiao Liu, Yinglei Xu, Miaomiao Li, Qian Tang, Bo Yu, Renmei Cai and Shiguo Liu, Targeted regions sequencing identified four novel PNPLA1 mutations in two Chinese families with autosomal recessive congenital ichthyosis, Molecular Genetics & Genomic Medicine, 8, 2, (2019).
Ruhong Cheng, Jianying Liang, Yue Li, Jia Zhang, Cheng Ni, Hong Yu, Xiangsheng Kong, Ming Li and Zhirong Yao, Next-generation sequencing through multi-gene panel testing for diagnosis of hereditary ichthyosis in Chinese, Clinical Genetics, , (2020).