Tuberous sclerosis complex (TSC) is a genetic syndrome due to mutations in either TSC1 or TSC2, leading to the development of hamartomatous tumours at multiple body sites, including facial skin (facial angiofibroma (FAF)), brain (cortical tubers) and kidney (angiomyolipoma (AML)). In this report, we describe an individual with minimal TSC clinical features, who had ‘no mutation identified’ (NMI) by prior genetic testing in a clinical laboratory. Our massively parallel sequencing (MPS) analysis of multiple samples from different body sites and tumours (including blood, saliva, normal skin, AML and FAF) revealed an extraordinary situation in which FAF and AML had completely independent inactivating biallelic variants in TSC2, not present in other matched samples. This suggests that the two different lesions (AML and FAF) are not due to the same underlying germline or mosaic mutation, rather both are likely sporadic events. This case demonstrates the relevance of thorough clinical examination, high-coverage MPS of multiple tumours and matched normal tissues, and appropriate genetic counselling for individuals with marginal TSC features and possible TSC1 or TSC2 mosaicism.
- genetic counseling
Statistics from Altmetric.com
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.
Tuberous sclerosis complex (TSC) is an autosomal dominant syndrome due to inactivating mutations in either TSC1 or TSC2, leading to activation of mammalian target of rapamycin complex 1 (mTORC1) and development of hamartomatous tumours. TSC-related tumours affect multiple tissues and organs, including skin (facial angiofibromas (FAFs) and ungual fibromas), kidney (angiomyolipomas (AMLs)) and brain (cortical tubers).1 TSC tumours develop through the Knudson two-hit mechanism, so that inactivation of one allele of either TSC1 or TSC2 (germline heterozygous or mosaic ‘first hit’) is followed by a somatic ‘second hit’ which inactivates the other allele of the same gene in each tumour.2 Each tumour in a TSC subject develops independently and is triggered by a distinct second hit, which most often occurs through copy-neutral loss of heterozygosity (LOH) or second point mutations.3–7 Massively parallel sequencing (MPS) analyses have shown that about 10%–15% of TSC cases occur due to generalised mosaicism for an inactivating TSC1/TSC2 variant. The range of variant allele frequencies (VAFs) for such mosaic variants is broad, from <1% to 20%–30%, median 1.7% in blood in our experience.6 8 It is often present at different levels in different cell types and tissues, and is enriched in TSC tumours. In consequence, genetic analysis of TSC-related tumours, including AML and FAF, facilitates identification of mosaic TSC-causative TSC1/TSC2 variants.6 We have shown that most patients with TSC who had ‘no mutation identified’ (NMI) by conventional testing were mosaic for a TSC1/TSC2 variant.6 9
Here, we report an individual with minimal TSC clinical features, who had NMI by prior genetic testing in a clinical laboratory. Our MPS analysis of multiple samples from different body sites and tumours revealed an extraordinary situation in which FAF and AML had completely independent biallelic variants in TSC2.
Materials and methods
Patient recruitment, clinical evaluation and sample collection
Demographic and clinical data were collected based on patient self-report, and clinical and imaging evaluation by EAT, as well as further examination by DJK and JMG. A solitary FAF was subject to a 2 mm punch biopsy reaching the dermal fibroblast layer, and a 5 mm normal skin biopsy was taken from the upper arm by JMG. Skin biopsies were subject to immediate DNA extraction. An AML needle biopsy was performed under CT guidance and also used for DNA extraction. Blood and saliva samples were also collected.
DNA extraction and quantification
Genomic DNA extraction from peripheral blood lymphocytes, saliva, skin and AML biopsies was performed following the manufacturer’s protocol (QIAamp Qiagen Mini Kit, Germany). DNA quantification was performed by both QUBIT dsDNA HS and PicoGreen assays.
Hybrid capture MPS for 50 mTORC1 pathway genes, including the entire genomic extent of TSC1 and TSC2 (as described previously),6 was performed for FAF and AML samples. Mean target coverage for FAF and AML was 779× and 796×, respectively; >99% target bases were at 100× sequencing depth for both samples. Single nucleotide variants (SNVs), indels and large mutations (multiexonic deletions/duplications) were called using our custom computational pipeline (Unix/Python/Matlab).6 The identified SNVs/indels were validated using targeted amplicon MPS (Illumina platform, San Diego, California, USA), as described,6 10 for FAF, AML and other samples available from the patient along with two negative controls for each identified mutation (read depth range 5324–33,415×, median 19 201×). All findings were reviewed manually using IGV.11
Fingerprinting analysis was performed using 45 polymorphic loci, for FAF and AML samples. Picard Tools GenotypeConcordance was used to calculate the concordance in all pairwise combinations of samples in the MPS sample batch.
A woman was referred for consideration of TSC when she had several renal AML identified on an abdominal ultrasound performed for other reasons. She had no history indicative or suggestive of TSC in any way, with normal development and no seizure history. She had no manifestations of TSC on physical examination apart from a single angiofibroma in the nasal fold. Retinal examination showed no TSC lesions. MRI abdomen showed bilateral fat containing solid renal masses consistent with AML with size of 3.8×3 cm and 2×2.7 cm in the right kidney, and 0.9×0.7 cm in the left kidney; a 1.5×1 cm T2 hypointense, T1 intermediate lesion in the liver without associated enhancement or internal fat (figure 1A). MRI of the brain showed a non-enhancing T2/fluid attenuated inversion recovery (FLAIR) hyperintense mass along the right postcentral gyrus without associated radial-glial band or mineralisation, consistent with a cortical hamartoma (tuber) or a low-grade glioma (figure 1B); and prominent FLAIR hyperintensity along the right periatrial white matter consistent with a subcortical tuber, though non-specific. Genetic testing for TSC gene variants was performed at a clinically certified laboratory with negative findings. In summary, a total of one definite (bilateral AMLs) and two possible (solitary FAF and two presumed cortical/subcortical tubers) TSC- related features were observed in this subject.1
She was referred for in-depth molecular genetic testing for mosaic variants in TSC1/TSC2, and provided informed consent. Hybrid capture MPS analysis of an AML biopsy revealed two different TSC2 mutations, a frameshift c.4351dup (p.Arg1451Profs*73) and missense c.4712A>G (p.Tyr1571Cys) mutation, at similar VAFs of 15.8% and 21.1%, respectively. c.4351dup (p.Arg1451Profs*73) has been reported nine times in the LOVD database (https://databases.lovd.nl/shared/genes/TSC2),12 and classified as a definite pathogenic mutation. c.4712A>G (p.Tyr1571Cys) occurs within the TSC2 GAP domain and has been reported three times in LOVD. Functional studies have shown that another missense variant affecting the same TSC2 aa (c.4711T>A, p.Tyr1571Asn) affects TSC2 stability and is inactivating13; therefore, p.Tyr1571Cys is also very likely inactivating (figure 1C, table 1).
Hybrid-capture MPS analysis of a biopsy of the solitary FAF lesion revealed two different nonsense TSC2 mutations: c.1248_1249delinsTT (p.Gln417*) and c.3319G>T(p.Glu1107*), at VAFs 2.2% and 2.4%, respectively. The c.1248_1249delinsTT (p.Gln417*) variant is a CC>TT substitution, well-known to be due to UV-induced DNA damage. Such mutations are commonly observed as second hits in TSC FAF.6–8 Both mutations are definitely pathogenic, although not reported previously (figure 1C, table 1).
Amplicon MPS analysis of all mutations was performed on all patient samples (table 1). The two FAF mutations were seen only in the FAF samples, confirming them as somatic and FAF-specific. The two AML mutations were seen only in the AML DNA (table 1). (Note that two of these variants, both single nucleotide variants, were seen in a single other sample in a single read, VAF <0.01%, below our VAF threshold for significance (table 1).) SNP fingerprinting analysis of 45 polymorphic loci (outside the targeted genes of interest), as well as SNPs in TSC1 and TSC2 regions (n=23 and 20, respectively), confirmed that the FAF and AML were derived from the same individual (figure 1D). No evidence for LOH was seen for SNP VAFs for either AML or FAF.
TSC occurs due to heterozygous or mosaic mutations in either TSC1 or TSC2, with estimated prevalence of 1 per 6000 live births.1 FAF, AML and cortical tubers, all seen in the subject described in this study, are among the most common TSC manifestations, each present in more than two-thirds of individuals with TSC.14
Mosaicism for variants in either TSC1 or TSC2 in TSC is increasingly recognised with MPS strategies designed for low VAF detection, since VAFs are typically <2%. It is present in 10%–15% of diagnosed TSC individuals and is associated with a milder clinical phenotype, as expected.6 8–10
In our patient, we identified four different TSC2 mutations, including two unique mutations in each of AML and FAF, a surprising finding. This suggests that the two different lesions are not due to the same underlying germ line or mosaic mutation. Rather both are likely sporadic events. Although based on our short read MPS sequencing we cannot exclude that these are mutations in cis, we suspect that they have occurred in trans leading to biallelic TSC2 inactivation in the respective lesions.
According to the consensus TSC diagnostic criteria,15 our patient presents one major TSC-related feature (‘≥2 AMLs’), and two other manifestations. She has an isolated single FAF, and two brain lesions that are possibly TSC-related. However, neither of these clinical features are major TSC criteria, since ≥3 FAFs and ‘multiple cortical tubers’ are required as major criteria. As a result, the patient can be diagnosed with possible but not definite TSC.15 Consequently, we propose that this patient does not have the TSC syndrome, but rather by chance has both kidney AML and a solitary FAF, as well as two brain lesions of uncertain origin and pathology.
Both AML and FAF occur in the general population, in individuals without TSC. AMLs are seen in ~0.1%–0.2% of the general adult population, and are more common in women than men.16 17 Solitary FAFs may occur as single isolated lesions (≤3 FAF papules) in as much as ~8% of general population.18 Consequently, we can estimate that patients with both solitary sporadic AML and <3 FAF lesions will occur in about 1 in 10 000 individuals. Bilateral AML are also seen in ~30% individuals who have AML, without other evidence of TSC,19 making the frequency of bilateral AML and <3 FAF about 1 in 30 000. Hence, this combination is not exceedingly rare. The co-occurrence of brain lesions consistent with TSC tubers raised significant clinical concern that the patient did have the TSC syndrome. As noted above, many patients with mosaic TSC have very few features of the disease, so that patients such as this one will not be rare, and may have either systemic mosaicism for a TSC gene mutation or by chance have an aggregation of symptoms that are similar to TSC.
In summary, our patient presents with an extraordinary situation which we have not seen previously in studying over a hundred patients with features consistent with mosaic TSC. This case highlights the importance of accurate clinical examination, comprehensive deep MPS of multiple tumours and matched normal tissues, and appropriate genetic counselling for individuals with marginal TSC features and possible TSC1 or TSC2 mosaicism. Concurrent sporadic clinical features of TSC may occur in an individual patient, without an underlying diagnosis of TSC.
Patient consent for publication
The patient provided written informed consent and the research protocol was approved by our Institutional Review Board, the Partners Human Research Committee (2013P002667).
We thank the patient for participation in this study.
Correction notice This article has been corrected since it was published Online First. The caption of figure 1 has been amended for clarity.
Contributors KK conceptualised the study, performed DNA extraction from all analysed samples, prepared amplicon MPS libraries, performed computational analysis of the MPS data and interpreted the results, wrote the manuscript and prepared the table and figure, acquired funding for the study. EAT performed clinical and imaging evaluation. JMG performed clinical evaluation and collected skin biopsies. ART participated in MPS data generation and preprocessing. DJK conceptualised and supervised the study, performed clinical evaluation, reviewed and participated in writing the manuscript, acquired funding for the study. All authors read and approved the manuscript.
Funding The study was funded by FY2020 TSC Alliance Postdoctoral Fellowship Award (KK) and Engles Family Fund for Research in TSC and LAM (DJK).
Competing interests None declared.
Provenance and peer review Not commissioned; externally peer reviewed.