Fifty-seven sporadic BCCs of 55 patients (27 males and 28 females; median age at diagnosis: 70 years, range 32-93 years) were included in the study. Demographic characteristics and clinico-pathologic features of patients and BCC lesions are illustrated in Table 1. Thirty-one patients had one BCC, 11 patients two BCCs and 13 patients > 2 BCCs. Anatomical sites of BCC lesions included the head/neck region (15/57, 26.3%), trunk (38/57, 66.7%) and extremities (4/57, 7.0%). Histopathologic examination showed that 26/57 (45.6%) were superficial BCCs and 31/57 (54.4%) nodular, with 47/57 BCCs (82.5%) having no ulceration.
We analysed the mutational profile of 13 cancer genes (CSMD1, CSMD2, PTCH1, SMO, GLI1, NOTCH1, NOTCH2, TP53, ITIH2, DPP10, STEAP4, DPH3 promoter and TERT promoter) (Supplementary Table S1) in a total of 57 BCCs, including superficial and nodular subtypes. Somatic variants were detected across 7 genes (PTCH1, CSMD1, TP53, NOTCH1, TERT promoter, DPH3 promoter, DPP10), with the 98.2% (56/57) of BCCs showing at least one alteration (Figure 1). Overall, 81.0% of single-nucleotide variants (SNVs) were C>T changes, consistent with UV-induced mutagenesis.
PTCH1 mutations were found in 41/57 (71.9%) BCCs, with 16/57 (28.1%) lesions carrying more than one PTCH1 variants. Overall, 63 variants in the PTCH1 gene were identified, distributing within the entire coding sequence with no specific hotspot region (Figure 2a). The most frequent nucleotide change was the C > T transition identified in 18/57(31.6%) tumours, followed by the CC > TT tandem variants in 8/57 (14.0%) and C>A mutations in 3/57 (5.3%) tumours. The majority (27/63, 42.9%) were truncating mutations, including 10 frameshift deletions and/or insertions and 17 nonsense mutations. All the identified truncating variants were included as likely oncogenic for clinical significance and likely loss-of-function for biological effect in the OncoKB database.21,22 In addition, missense mutations accounted for 20.6% (13/63) of PTCH1 variants. None of the identified PTCH1 missense mutations were known for oncogenic significance. However, 6/13 (46.1%) missense variants, were predicted to be deleterious based on the dbNSFP v2.0 applied prediction tool. Finally, 8 splice sites mutations (8/63, 12.7%) were identified (Supplementary Table S2).
TP53 mutations were found in 26/57 (45.6%) BCCs, with hot spot positions including p.H179, p.S241, p.G245 and p.R280 (Supplementary Table S2).24 Single nucleotide missense variants clustering in the DNA binding domain (95-288 ammino-acid residues) were the most prevalent (15/30, 50.0%), and truncating mutations (all nonsense except for one frameshift deletion) represented 33.3% (10/30) of alterations (Figure 2b).
In addition, mutations in CSMD1 gene were identified in 36 of 57 (63.2%) BCCs, with missense mutations being the most prevalent (68/85, 80.0%). More than one CSMD1 alterations was found in 22/57 (38.6%) tumours, and 5/57 BCCs (8.8%) harboured more than six CSMD1 somatic mutations. All the identified CSMD1 point mutations were unknown for oncogenic significance. However, the dbNSFP v2.0 tool predicted that 41.2% (35/85) of CSMD1 alterations were deleterious for protein function (Supplementary Table S2), including the most recurrent p.R671C (c.2011C>T) amino-acid change, which was detected in 3/57 (5.3%) BCC lesions (Figure 2c).
NOTCH1 mutations were found in 25/57 (43.8%) BCCs. In total, we detected 35 NOTCH1 mutations, including truncating (14/35, 40.0%) and missense variants (13/35, 37.1%). Of the 35 identified mutations, 16 (16/35, 45.7%) are known oncogenic or predicted oncogenic according to the OncoKB database (Figure 2d, Supplementary Table S2).
Mutations in DPP10 were detected in 35.1% (20/57) of BCCs, with missense variants being the most frequent mutational event (20/30, 66.7%) (Figure 2e).
Eight TERT promoter mutations were found in 33/57 (57.9%) BCC lesions. The most frequent somatic change was the -146 C>T, which was detected in 19/57 BCCs (33.3%), followed by the -124 C>T transition and the -138/139 CC>TT tandem variation accounting for 8.8% (5/57) and 5.3% (3/57) of BCC lesions, respectively. Supplementary Table S3 illustrates all TERT mutations.
Overall, 9 DPH3 promoter mutations were identified in 49.1% (28/57) of BCCs. The most prevalent variant was the -121C>T transition (19/57, 33.3%), followed by the -122 C>T (4/57, 7.0%), -125 C>T (2/57, 3.5%) and -150 C>T (2/57, 3.5%) (Supplementary Table S3).
Statistical analysis of the mutational profile
We performed pairwise exclusivity and co-occurrence analysis for the 7 identified mutated loci (DPP10, CSMD1, PTCH1, NOTCH1, TP53, TERT promoter, DPH3 promoter) and found significant concurrent variants of TP53 gene with CSMD1 (p=0.002), PTCH1 (p=0.011) and DPH3 promoter (p=0.006) (Figure 1). However, after Benjamini-Hochberg false discovery rate (FDR) correction, only the association of CSMD1 variants with the TP53 mutation rate remained significant (q=0.045).
By evaluating the distribution of somatic alterations within 7 different mutated genes across 57 BCC lesions, we identified 43 different combinations. Among them, the two most frequent (4/57; 7.0%) were: 1) concurrent PTCH1 and TERT promoter mutations in a setting of wild type DPP10, CSMD1, NOTCH1, TP53 and DPH3 promoter genes, BCCs, and 2) the coexistence of DPP10, CSMD1, PTCH1, NOTCH1, TP53, TERT promoter and DPH3 promoter mutations. None of these two most frequent combinations were significantly associated with superficial or nodular subtypes (p= 0.117).
Focusing on the analysis of single mutated gene according to the specific BCC subtypes, PTCH1 and NOTCH1 mutations were found significantly associated with superficial BCCs (p=0.018 and p=0.020, respectively). In details, PTCH1 variants were 1.6 times (OR= 5.537, 95% CI=1.367-22.43) and NOTCH1 mutations 2.0 times more frequent (OR= 4.457, 95% CI=1.304-15.24) in superficial than in nodular BCCs.
The Principal Component Analysis (PCA) multivariate approach confirmed a significant association between PTCH1 mutations and the superficial BCC subtypes that were recognized as genetically similar group for PTCH1 mutations in a separate cluster of the PCA diagram (Figure 3).
The analysis of the mutational status according to patients and tumour characteristics revealed that PTCH1 mutations were significantly associated with intermittent sun exposure (p=0.046), and with the occurrence of single BCC lesions (p=0.021), and NOTCH1 mutations were more frequent in BCCs arising on the trunk compared to the head/neck and extremities (p=0.001).