Greater susceptibility of patients with idiopathic pulmonary fibrosis to basal cell carcinoma: a combined genomics and Mendelian randomization analysis

Introduction

Idiopathic pulmonary fibrosis (IPF) is a chronic progressive disease characterized by the destruction of alveolar epithelium, which leads to a reduction in gas exchange and lung function. The median survival time of patients with IPF after diagnosis is 3–5 years (1), and the 5-year survival rate is lower than 40% (2). The development of IPF is associated with a variety of factors, including age, exposure to tobacco smoke, and pulmonary and extrapulmonary complications. These factors contribute through various mechanisms, including lung epithelial damage and enhanced resistance to apoptosis in myofibroblasts, which ultimately leading to lung fibrosis. IPF occurs predominantly in the older adult population and is associated with the shortening of telomeres, representing a classic premature aging syndrome closely associated with lung fibrosis. Although IPF is considered a disease limited to the lungs, its risk factors are similar to those of many comorbidities [e.g., cardiovascular and degenerative diseases (3,4)], which may play a prominent role in the pathophysiological process of IPF.

Among the patients with IPF, only approximately 10% have no comorbidities. In contrast, up to 60% of patients exhibit one to three associated comorbidities, while about 30% have more than three. This phenomenon may reflect a genetic predisposition linking IPF with these diseases or the existence of common exogenous risk factors (5). The co-occurrence of IPF with other disorders significantly impacts patients’ quality of life and overall survival. Previous studies have indicated that individuals with IPF may be at increased risk of developing various cancers, particularly lung cancer (4-6). A significant association has also been established between IPF and an elevated incidence of skin cancer (6). These findings have contributed to a paradigm shift: previously, IPF was viewed solely as a pulmonary disorder but has more recently been recognized as part of a systemic disease process. This broader perspective underscores the need for comprehensive patient management and tailored cancer prevention strategies.

In this study, we used publicly available data from open gene-wide association studies (GWASs) within a Mendelian randomization (MR) framework to determine whether IPF promotes the formation of basal cell carcinoma (BCC). One of the aims of this research was to generate evidence for etiological studies of BCC, which is the most common type of cancer in the world. The etiology and pathogenesis of BCC are complex, involving the complicated interplay between environmental, phenotypic, and genetic factors. Moreover, as the early onset of the disease is slow and asymptomatic, BCC can be easily overlooked by patients and challenging for clinicians to diagnose accurately. This is further complicated by a disease development and invasion that produces a variety of clinical manifestations and outcomes (7). BCC can initially manifest as a subtle, pearly papule with slow growth kinetics and is often overlooked as a benign acne lesion or mole. The disease follows a pattern of local invasiveness, characteristically progressing through an orchestrated sequence from in situ hyperplasia to invasive expansion. Although BCC rarely metastasizes, a failure to intervene results in substantial local tissue damage (8). In contrast, the pathology of IPF is characterized by a dynamic, step-wise, fibrotic cascade. This process is initiated by alveolar epithelial cell damage, which triggers the recruitment of inflammatory and immune cells and subsequently activates the proliferation of fibroblasts. Ultimately, the process culminates in the effective “sealing” of normal alveolar spaces by scar tissue—a distinct, orderly pathophysiological trajectory (9). Although BCC and IPF affect anatomically and functionally distinct organs, they exhibit a shared pathogenic paradigm at the macroscopic level, including a multifactorial etiology, clinically silent early phases, and a progression governed by a complex but predictable biological sequence (5). Therefore, we undertook an in-depth investigation into the pathogenesis of BCC. Our results indicated that patients with IPF have a higher risk of developing BCC due to common pathogenic mechanisms and immune cell effects, which may provide early clinical signs of tumor emergence and therapeutic options related to pathway mechanisms. We present this article in accordance with the STROBE-MR reporting checklist (available at https://tcr.amegroups.com/article/view/10.21037/tcr-2025-1-2853/rc).

Methods

Microarray data analysis

Microarray data (GSE10667, GSE24206, and GSE53845) were downloaded from the Gene Expression Omnibus (GEO) database (https://www.ncbi.nlm.nih.gov/geo/) (10), including 31 patients with IPF and 15 healthy individuals from GSE10667, 17 patients with IPF and 6 healthy individuals from GSE24026, and 40 patients with IPF and 8 healthy individuals from GSE53845. Clinical information, such as age and sex, was also collected. Data normalization, integration, and principal component analysis (PCA) were performed via R language version 4.3.0 (The R Foundation of Statistical Computing, Vienna, Austria) with the limma (11), sva (12), ggplot2 (13), and ggpubr (14) packages.

Determination of differentially expressed genes (DEGs)

Data standardization and DEG analysis were performed via the limma package in R, and DEGs were screened according to the thresholds of |log₂ fold change |>1 and P<0.05. The pheatmap (15) and ggplot2 (13) R packages were used to generate DEG expression heatmaps and volcano plots.

Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genome analyses

After obtaining the DEGs, GO and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analyses were performed using the clusterProfiler R package (16). KEGG serves as a database resource for systematic functional analysis (17), whereas GO analysis examines gene functions across three categories: biological processes (BP), cellular components (CC), and molecular functions (MF). This study aimed to elucidate the molecular processes and key pathways involved in IPF and BCC, and to employ KEGG enrichment analysis to substantiate the close relationship between their respective pathways. A P value <0.05 was considered statistically significant.

MR

MR analysis was conducted based on data from publicly available GWAS sources (18) in order to determine the causal relationship between BCC and IPF. We selected key pathways of BCC (GWAS disease cohorts: ukb-d-C3_SKIN) and IPF (GWAS disease cohorts: finn-b-IPF) from the IEU OpenGWAS project (mrcieu.ac.uk) in the MR analysis. To evaluate the relationship between BCC and IPF, we used the R package GitHub (Input program instructions: MRCIEU/TwoSampleMR) to input the exposure factor and outcome factor, apply the MR method, and complete inverse-variance weighting (IVW). Additional sensitivity analysis was conducted with the MR-Egger method (18-20). P value <0.05 was considered statistically significant.

Identification of core genes in pathways and construction of gene-drug and competing endogenous RNA (ceRNA) network diagrams

KEGG analysis was implemented via the R (17) packages BiocManager (21), ggplot2 (13), and ggpubr (14), among others, in order to obtain the core genes of the BCC pathway and to draw a gene volcano plot. The “DGIdb” tool (22) predicted and visualized the interaction network between core genes and drugs. The miRanda, miRDB, miRWalk, and TargetScan tool were used to validate the target genes of microRNA (miRNA), and the spongeScan database was used to predict the long noncoding RNA (lncRNA) targeting miRNA (23). Finally, based on the interaction of lncRNA with the related genes, a ceRNA network (lncRNA-miRNA-mRNA) was constructed, which was visualized via Cytoscape software v. 3.9.1 (24).

Immune cell analysis

The gene expression data of patients were uploaded to the CIBERSORTx database (https://cibersortx.stanford.edu) (25), and the infiltration levels of 22 immune cells were determined. The R limma package was used to analyze the correlation between hub genes and immune cells. This analysis aimed to elucidate how the immune microenvironment interacts with these key disease-associated genes in both BCC and IPF.

Data validation

For the validation dataset GSE38958 (https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE38958) (26), which included 70 IPF patients and 45 healthy controls, data normalization and correction were performed using the R packages limma and ggpubr. Following this, we experimentally validated the expression of the 12 candidate genes in both the IPF and control groups. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments.

Results

Microarray data analysis

The IPF datasets (GSE10667, GSE24026, and GSE53845) were obtained from the GEO database. After data normalization, PCA, and merging, the three data sets were found to have no overlap, indicating that they could be further investigated (Figure 1A) and were merged for further analyses after sample normalization (Figure 1B). Differential analysis of gene microarrays yielded a total of 1,333 DEGs, including 1,088 upregulated genes and 245 downregulated genes between the patients with IPF and the control group (available online: https://cdn.amegroups.cn/static/public/tcr-2025-1-2853-1.XLSX) (P<0.05). Figure 1C is a volcano plot visualizing the distribution of DEGs, with upregulated genes indicated in red and downregulated genes in blue. Using the pheatmap and ggplot2 R packages, the expression patterns of the top 50 upregulated and downregulated genes in the control and IPF groups are visualized in the heatmap shown in Figure 1D.

Figure 1 Visualization of IPF datasets (GSE10667, GSE24026, and GSE53845) from the GEO database. (A) PCA revealed no overlap between the three groups of data. (B) The samples were standardized and merged for further analysis. (C) Volcano plots of DEGs in the standard group and the IPF group, with upregulated genes in red and downregulated genes in blue. (D) Heatmaps of the top 50 most up- and downregulated genes in the different groups. DEGs, differentially expressed genes; FC, fold change; GEO, Gene Expression Omnibus; IPF, idiopathic pulmonary fibrosis; PC, principal component; PCA, principal component analysis.

GO and KEGG analyses

We performed GO and KEGG analyses to further examine the function of the identified DEGs. The arrows in the directed acyclic graph in Figure 2A indicate the hierarchical relationship between GO terms. The darker the color is, the more significant the enrichment; information within the box includes the GO term number, functional description, and the number of differential genes enriched in that term.

Figure 2 GO and KEGG enrichment analyses. (A) Directed acyclic graph of GO analysis results. (B) Bar plot of GO functional enrichment in BP, MF, and CC, with longer bars indicating a greater number genes and more significant enrichment. (C) KEGG pathway enrichment analysis of DEGs in different functional categories. BP, biological processes; CC, cellular components; DEGs, differentially expressed genes; ECM, extracellular matrix; GO, Gene Ontology; KEGG, Kyoto Encyclopedia of Genes and Genomes; MF, molecular functions; PPAR, peroxisome proliferator-activated receptor.

Figure 2B shows a bar graph of the top 20 most significantly enriched GO terms, with longer bars indicating a greater number of genes and more significant enrichment. The y-axis displays the enriched entries, with different colors representing three ontologies enriched in GO, including BP, MF, and CC. The key genes were found to be mainly involved in the closely related functions of extracellular matrix (ECM), external encapsulated structural organization, hormone level regulation, intercellular junctions, and signaling receptor activation activities, suggesting their critical function in ECM formation, promotion of fibroblast differentiation, induction of collagen production and migration, cell adhesion, and signaling receptor activation (Figure 2B) (P<0.05). Additionally, KEGG pathway enrichment analysis of DEGs indicated enrichment in cytokine-cytokine receptor interactions, ECM-receptor interactions, BCC, and chemokine signaling pathways (Figure 2C). We further found that BCC may be closely associated with IPF.

Increased risk of BCC in patients with IPF

KEGG pathway enrichment analysis of DEGs in available online: https://cdn.amegroups.cn/static/public/tcr-2025-1-2853-2.XLS suggested a potential association between IPF and BCC. To investigate the causal effect of IPF on BCC risk, a two-sample MR analysis was performed.

We selected independent genetic variants significantly associated with IPF at the genome-wide level (P<5×10⁻⁸) from the FinnGen consortium as instrumental variables. The F-statistics of these variants ranged from 30.99 to 90.29, with a mean of 54.23, all well above the conventional threshold of 10, indicating strong instrument strength. Summary-level data for skin cancer were obtained from the UK Biobank (361,194 participants). Although the C44-coded skin malignancies represent a composite endpoint, BCC constitutes approximately 80% of these cases (27,28), making this dataset a valid proxy for BCC genetics.

The primary MR analysis using the IVW method showed a borderline significant association between genetically predicted IPF and increased skin cancer risk (P=0.042). The MR-Egger method, which is more robust to potential pleiotropy, yielded a similar but slightly stronger association (P=0.013). All supplementary MR methods consistently supported a positive association, including the weighted median (P=0.024), simple mode (P=0.0048), and weighted mode (P=0.0025) approaches, suggesting a positive correlation between BCC and IPF (Figure 3A). A forest plot displaying effect estimates and confidence intervals further confirmed this positive relationship—all MR-Egger and IVW estimates were greater than zero (Figure 3B). These results indicate that IPF may be associated with an elevated risk of BCC.

Figure 3 MR analysis results. (A) Scatterplot of genetic randomization visualizing the potential relationship between IPF and the risk of developing BCC. (B) Forest plot demonstrating the causal relationship between BCC and IPF, with all MR-Egger and all inverse-variance weighted values being greater than 0 and IPF being positively correlated with BCC. (C) Symmetrically distributed funnel plot indicating the lack of bias and low heterogeneity of the study results. (D) Leave-one-out plot showing the effect of excluding each SNP on BCC, indicating a lack of bias. BCC, basal cell carcinoma; IPF, idiopathic pulmonary fibrosis; MR, Mendelian randomization; SNP, single-nucleotide polymorphism.

We then evaluated the validity of the causal inference. The funnel plot was approximately symmetric, indicating no substantial directional bias (Figure 3C). Sensitivity analyses supported the validity of the MR assumptions. MR-Egger intercept test did not provide evidence of directional pleiotropy (intercept =−0.00402, P=0.135). Heterogeneity among the genetic instruments was low, as indicated by Cochran’s Q statistics for MR-Egger (Q=3.10, P=0.376, I²=3.2%) and IVW (Q=7.36, P=0.118, I²=45.7%). These findings suggest that patients with IPF may have a higher risk of developing BCC.

To assess the robustness of the results, a leave one out sensitivity analysis was performed. This method iteratively excludes individual single nucleotide polymorphisms (SNPs) and recalculates the overall association. As shown in Figure 3D, all iterations yielded robust estimates consistently above zero, confirming that the observed causal effect was not driven by any single outlier SNP. This further validates the reliability of the MR results, supporting a positive association between IPF and BCC and indicating that individuals with IPF may be at an increased risk of BCC compared with the general population.

Construction of a core gene-drug network diagram

In this study, we used R language for the KEGG pathway analysis of DEGs. We successfully identified the core genes of all pathways and found 12 core genes in the BCC pathway. As shown in Figure 4A, we used a volcano plot to clearly distinguish the upregulated and downregulated genes, with upregulated genes (SMO, GLI1, GLI2, WNT10A, BMP4, and FZD3) show in red color, and the down-regulated genes (FZD5, HHIP, WNT7A, FZD8, FZD4, and BMP2) show in green. These genes have been reported in several studies to play a large role in the progression of both IPF and BCC via epithelial-mesenchymal transition (EMT), specifically through the interactions of Hedgehog (HH), Wnt, and transforming growth factor-beta (TGF-β) pathways (29,30). In addition, the interactions between these core genes and drugs were identified using the Drug Gene Interaction Database (DGIdb). To translate our network pharmacology findings into therapeutic insights, we first constructed a drug-gene interaction network using Cytoscape. From this analysis, we identified and examined several drugs targeting key pathways, such as Hedgehog signaling (Figure 4B). These included SMO-targeted inhibitors (e.g., vismodegib, sonidegib, glasdegib) with clinical validation in cancers like BCC and small cell lung cancer (31,32), and GLI inhibitors being tested for SMO inhibitor-resistant BCC (33) (available online: https://cdn.amegroups.cn/static/public/tcr-2025-1-2853-3.XLS). Furthermore, we evaluated pirfenidone, an approved IPF agent. Its action of selectively destabilizing the HH-mediated activators GLI1 and GLI2 aligns directly with our network’s core pathway, indicating its potential applicability in lung cancer or IPF-lung cancer co-morbidity (34). Collectively, this work positions these compounds as candidates for validating the dysregulated pathways identified in our study. Whether drugs targeting the identified related genes can be used for the cotreatment of IPF and BCC warrants further exploration.

Figure 4 Core gene-drug network diagram. (A) A network diagram of genes with significant up- and downregulation in BCC and (B) the interactions between core genes such as SMO, GLI1, and FZD and drugs. BCC, basal cell carcinoma; FC, fold change.

Immunological interplay between IPF and BCC: gene expression, immune cell infiltration, and pathway activation

Correlation analyses were performed between hub genes in BCC and immune cells (Figure 5, https://cdn.amegroups.cn/static/public/tcr-2025-1-2853-4.XLS) to clarify the relationship between the immune microenvironment and these BCC-related genes. Among the downregulated genes in the BCC pathway, FZD5, FZD8, HHIP, and BMP2 showed a positive correlation with M2 macrophage infiltration. Upregulated genes such as SMO, GLI1, and GLI2 were positively correlated with monocyte infiltration, while upregulated genes such as GL1 and GL2 were negatively correlated with M0 macrophage infiltration. In IPF, TGF-β is considered to be a significant profibrotic cytokine and to play a critical role in sustained fibroblast activation and myofibroblast differentiation in fibrotic diseases (35). TGF-β is the most well-characterized promoter of ECM production and is considered the most potent chemotactic factor for immune cells (e.g., monocytes and macrophages). Inflammation also plays a vital role in IPF, with macrophages producing cytokines that induce an inflammatory response and participate in the transition to the healing environment through the recruitment of fibroblasts, epithelial cells, and endothelial cells. If an injury persists, neutrophils and monocytes are recruited, and reactive oxygen species production exacerbates epithelial damage (36). Finally, monocytes and macrophages produce platelet-derived growth factor (PDGF), C-C motif chemokine ligand 2 (CCL2), macrophage colony-stimulating factor, and colony-stimulating factor 1. These proteins may also have direct profibrotic effects (37). In vivo, dynamic changes in macrophage activation can be observed, with M1 predominantly secreting pro-inflammatory factors, M2 expressing inhibitory inflammatory factors, and M2 functioning mainly in the late stages of inflammation and being capable of suppressing inflammatory factors and inhibiting the inflammatory response (38). Our study found that as IPF emerges and progresses, the upregulated genes of the BCC pathway become active, with the increase of macrophages and monocytes, promoting pathway activation and further facilitating BCC formation. Our findings may provide clues to the immunological correlation between IPF and BCC.

Figure 5 Analysis of the correlation between hub genes and immune cells. NK, natural killer.

Validation of pathway genes and structure of the ceRNA system

Based on normalized GSE38958 data, we further validated the expression of the 12 genes in the IPF and control groups, confirming the high expression of genes that were upregulated in the IPF patient group (e.g., SMO, GLI1, GLI2, WNT10A, BMP4, and FZD3) (Figure 6A). The overexpression of these genes may be closely related to the pathogenesis of IPF, providing important insights for further investigation of the specific roles of these genes in the development of IPF. We used the spongeScan database to validate the target genes of miRNAs and predict lncRNA against these miRNAs. This step allowed us to better understand the role of miRNAs in regulating gene expression and to identify the potential regulatory mechanisms. Based on the complex interactions between lncRNA and related genes with miRNAs (available online: https://cdn.amegroups.cn/static/public/tcr-2025-1-2853-5.XLS), we constructed a ceRNA map containing lncRNA, miRNAs, and mRNAs (Figure 6B). The construction of this ceRNA network can clarify the interregulatory relationship between lncRNA, miRNAs, and mRNAs and support the theoretical basis for further research on therapeutic interventions against IPF combined with BCC.

Figure 6 Validation of core genes and the ceRNA network. (A) Differences in the expression of the 12 genes in the IPF group and the control group. The upregulated expressed bases in BCC were significantly expressed in the IPF group (*, P<0.05; **, P<0.01; ***, P<0.001). (B) ceRNA network diagram demonstrating the regulatory relationship between lncRNAs, miRNAs, and mRNAs. BCC, basal cell carcinoma; ceRNA, competing endogenous RNA; IPF, idiopathic pulmonary fibrosis; lncRNA, long noncoding RNA; mRNA, messenger RNA; miRNA, microRNA.

Discussion

IPF is a progressive lung disease characterized by scarring of the lungs and a high mortality rate (39). The pathogenesis of IPF involves complex interactions between different cell types and signaling pathways. The consensus on IPF is that recurrent alveolar epithelial cell injury and an aberrant wound healing response lead to dysregulated intercellular interactions, resulting in extensive deposition of dense fibrous tissue, reduced cystatin C levels, and impaired gas exchange (40). Although IPF is considered a disease confined to the lungs, its risk factors are similar to those of many comorbidities (e.g., cardiovascular and degenerative diseases) and play a critical role in the course of IPF (4,5).

Research has revealed that senescent lung fibroblasts from patients with IPF can significantly enhance the proliferation, invasion, and migration capabilities of non-small cell lung cancer cells (41). In the pathological progression of IPF, the aging and aberrant activation of fibroblasts are considered key factors. These senescent fibroblasts promote the formation of the tumor microenvironment through the secretion of exosomes and matrix metalloproteinases (such as MMP1), thereby accelerating the progression of non-small cell lung cancer (42). Studies have indicated that activated fibroblasts produce abundant ECM components (e.g., collagen), forming a physical barrier that impedes the infiltration of cytotoxic immune cells such as T cells into the tumor core. On a pan-cancer scale, collagen triple helix repeat containing 1-positive (CTHRC1+) fibroblasts and secretory leukocyte peptidase inhibitor-positive (SLPI+) macrophages have been found to establish a profibrotic niche, shaping an immunosuppressive microenvironment associated with poor prognosis, which further underscores the relationship between pulmonary fibrosis and tumor development (43). Additionally, correlations have been observed between BCC and IPF. COL10A1 has been identified as a novel marker for high-risk BCC—particularly the sclerosing/infiltrative subtype—is expressed by specific cancer-associated fibroblasts (CAFs), and is associated with ECM remodeling and invasiveness (44). Telomere shortening has been recognized as a key factor inducing cellular senescence in patients with IPF, with telomere dysfunction serving as a central mechanism driving senescence in this context (45,46). Telomere dysfunction synergizes with ultraviolet (UV) radiation to promote the development of skin cancers, including BCC. Telomere shortening can lead to genomic instability, rendering cells more susceptible to the accumulation of carcinogenic mutations following UV-induced damage (47).

Epithelial cells in patients with IPF, including those in the skin, may exhibit a generalized “vulnerable” phenotype. This vulnerability may stem from telomere dysfunction, persistent endoplasmic reticulum stress, mitochondrial dysfunction, and other factors. Such alterations make the skin epithelium more prone to the accumulation of DNA damage and less capable of repair upon UV exposure, ultimately leading to carcinogenesis (48). Several studies have reported that patients with IPF carrying the MUC5B risk allele have a higher incidence of skin cancer, including of BCC. Although the precise mechanism underlying this relationship remains unclear, it is hypothesized that MUC5B may facilitate a microenvironment conducive to cancer cell growth by modulating EMT or interacting with the HH signaling pathway—a key pathway in BCC pathogenesis (49). Furthermore, a distinct population of basal cells has been identified in the lungs of patients with IPF, characterized by high expression of keratin 17 (KRT17). KRT17 is typically associated with dedifferentiation, wound healing, and cancer but is rarely detected in the airways of healthy lungs (50). Moreover, WNT7A derived from basal cells has been shown to contribute the emergence of a fibrotic environment, further suggesting that basal cells are a key link between IPF and skin cancer (51). Although BCC and IPF affect distinct organs, mainly the skin and lungs, respectively, and differ in their clinical manifestations, they share common pathological themes at a macroscopic level. These include multifactorial etiology, early asymptomatic progression, and the involvement of complex but ordered BP. Recognizing these shared characteristics can provide a broader perspective for understanding the pathogenesis of complex diseases.

In this study, we identified 1,333 DEGs between patients with IPF and healthy individuals, and enrichment analysis showed that genes in the BCC pathway were enriched in DEGs. MR studies further confirmed that patients with IPF are more likely to develop BCC. We further identified 12 core genes mediating the BCC pathway, among which SMO, GLI1, GLI2, WNT10A, BMP4, and FZD3 were upregulated in BCC; moreover, these genes were more highly expressed in patients with IPF than in controls. The co-occurrence of IPF and BCC is associated with more rapid disease progression. Furthermore, the pathogenic mechanisms underlying the elevated susceptibility of IPF patients to BCC, relative to other populations, are likely complex and multifactorial. TGF-β1 is considered to be the major profibrotic cytokine in the development of IPF and plays a crucial role in sustained fibroblast activation and myofibroblast differentiation in fibrotic disease (35). TGF-β1 signaling occurs only transiently during wound healing but sustains its activity in fibrotic disease and fibroblasts isolated from the lungs of patients with IPF. TGF-β signaling increases the expression of wingless/integrated-1 proteins (WNT) ligands and Frizzled receptors (FZD) receptors in human lung fibroblasts, and the WNT signaling pathway figures prominently in tissue homeostasis and remodeling in many organ systems, including the lungs. WNT ligands bind to the FZD receptor, thereby controlling cellular differentiation, growth, and polarity in various cellular systems (29,30). The FZD receptor activates either the canonical pathway, which signals through β-catenin, or one of the noncanonical pathways, which primarily signal through calcium (the WNT/Ca2+ pathway) or Ras homolog family member A (RhoA)/c-Jun N-terminal kinase (JNK) (RhoA/JNK) (the WNT/planar cell polarity pathway) (52). FZD3, a member of this receptor family, promotes fibrogenesis and epithelial tumor proliferation via WNT pathway activation (53,54). In contrast, BMP4 is a member of the TGF-β family, which is highly expressed in many tissues and plays a crucial role in embryogenesis, embryonic development, and organ homeostasis (55). In various organs, including the lungs, BMP4 can cause injury and promote inflammation and is associated with fibrotic diseases (56). BMP4 also exerts a critical regulatory effect on MMPs, which contribute substantially to tumor invasion and metastasis; another mechanism by which MMPs promote metastasis is via the regulation of EMT progression through interactions with calcineurin in a dynamically reversible process by which epithelial cancer cells pass through EMT to acquire viability and invasiveness to perform multiple steps, including the invasion-metastasis cascade (57). This further supports the critical role of BMP4 in BCC. There is a growing body of evidence suggesting that the HH pathway is a central regulator of many processes in embryonic development and is reactivated after adult lung injury (29). In typical HH signaling, binding HH ligands to their receptor proteins patched homolog 1 (PTCH-1) and 2 (PTCH-2) leads to the suppression of SMO, resulting in the activation of the GLI1, GLI2, and GLI3 transcription factors. In atypical HH signaling, GLI activity is regulated independently of SMO by other signaling pathways, including TGF-β (29). In the lungs of patients with IPF, the expression of HH ligand sonic HH (SHH) and the transcription factor GLI2 is elevated in epithelial cells. In contrast, the SHH receptors PTCH-1, SMO, and GLI1 are upregulated in fibroblasts/myofibroblasts (29). Studies have shown, therapeutic intervention into the upstream components of the HH cascade, such as SHH or SMO, exerted little effect, whereas blockade of downstream components of the GLI family inhibited experimental pulmonary fibrosis (58,59). These findings suggest that atypical HH pathway regulation and GLI function contribute to IPF. Dysregulated HH signaling has been shown to drive tumorigenesis through loss of homeostatic control and to sustain aggressive oncogenic phenotypes—including tumor initiation, progression, metastasis, and therapy resistance—in human cancers. Aberrant activation of this pathway has been associated with several cancer types, particularly BCC (58). Activated SMO delivers HH-activated signals to the cytoplasm, releasing GLI transcription factors from cytoplasmic segregation via the suppressor of fusion (SUFU). This results in the translocation of GLI to the nucleus, where it binds to transcriptional targets to regulate cellular gene expression and promote tumor development (60).

Therefore, several components of the HH pathway are being investigated in the context of targeted cancer therapy, especially GLI1 and SMO. Aberrant HH pathway activation is a significant driver of BCC. The discovery that PTCH1 mutations contribute to Gorlin syndrome (also known as BCC syndrome), which is characterized by susceptibility to BCC and other cancers, established the causal role of aberrant HH signaling in this malignancy (60,61). We speculate that IPF’s susceptibility to BCC may be mediated through the HH pathway acting via various mechanisms and gene regulations: both classical HH signaling through genes such as SMO, GLI1, and GLI2; and atypical HH signaling involving factors such as TGF-β, BMP4, WNT ligands, and FZD receptors. Together, these processes may promote a microenvironment conducive to the development of BCC. Our study analyzed the interregulatory effects of BCC pathway genes and drugs. Other large-scale genomic studies have shed light on the genetic basis of sporadic BCC in the general population. For instance, mutations in HH pathway genes (e.g., PTCH1, SMO, and SUFU) were found to occur in approximately 85% of sporadic BCC cases. Somatic PTCH1 gene mutations have been detected in 70–75% of cases, while 10–20% of patients exhibit activating mutations in SMO, an oncogene repressed by PTCH1 (62). Although surgical resection is the treatment of choice for those with BCC, it can lead to disfigurement or functional impairment, and for advanced and rare cases, targeted drug therapy is the only option. Inhibitors of the HH pathway have been established as the mainstay of treatment for locally advanced or metastatic BCC (63). The HH inhibitors, vismodegib and sonidegib (SMO inhibitors), are orally administered. Vismodegib has been approved in Europe for the symptomatic treatment of metastatic BCC, while sonidegib is used for select patients with BCC who may be difficult to treat due unsuitability for surgery or radiotherapy or for those with postoperative recurrence (63). These facts, when considered in light of our results, suggest that targeted gene therapy may be effective in patients IPF-related BCC.

Pirfenidone is an antifibrotic drug with anti-inflammatory, antioxidant, and antifibrotic properties. Pirfenidone selectively destabilizes GLI proteins, which are significant activators of HH-mediated gene transcription. As a result, pirfenidone reduces the overall activity of the HH pathway in lung fibroblasts among patients with IPF, decreasing cell migration and proliferation. Pirfenidone blocks many TGF-β-triggered cellular events, such as fibroblast proliferation and migration, ECM protein production, and EMT (64). Blockade of the HH pathway through use of the clinically approved SMO antagonist vismodegib does not affect fibrosis development, representing an efficient approach (65). Given that signaling pathways regulated by GLI proteins such as TGF-β and HH are involved in the pathological processes of many cases, the GLI-inhibition function of pirfenidone provides a theoretical rationale for the application of this drug to other human lung diseases (e.g., lung cancer) (38,61). Fibrosis can increase the risk of lung cancer by 7% to 20%. There are a variety of standard genetic, molecular, and cellular processes that link pulmonary fibrosis to lung cancer and which may predispose patients to IPF and lung cancer, including myofibroblast/mesenchymal transition, myofibroblast activation and uncontrolled proliferation, endoplasmic reticulum stress, altered growth factor expression, oxidative stress, and significant genetic and epigenetic variants. The Wnt/β-catenin pathway is overexpressed in lung tissues of patients with IPF and lung cancer (29,41,53), and overexpression of the SHH pathway increases the susceptibility to epithelial apoptosis and the resistance to fibroblast apoptosis. In the early stages of tumorigenesis, SHH is reactivated by cancer stem cells and leads to paracrine effects on other tumor cells, resulting in tumor growth, tumor spread, and EMT (58,59). Therefore, the use of pirfenidone in patients with lung cancer or in patients with both IPF and lung cancer is justified, and Zykadia has been approved as a second-line therapeutic agent for patients non-small cell lung cancer (34,38). The importance of HH signaling in a wide range of human cancers has stimulated a substantial amount of interest in targeting this pathway for cancer therapy. HH pathway inhibitors, including a variety of SMO antagonists such as vismodegib, sonidegib, and glasdegib, have been examined in clinical trials for the treatment of BCC and small cell lung cancer (31,32,34), among other cancers, and the GLI inhibitor arsenic trioxide has been approved as a second-line treatment for non-small cell lung cancer (6,32). Targeting such a central regulatory unit may explain the clinical efficacy of this drug in the treatment of pulmonary fibrosis. Its efficacy may support its use in other glial cell-driven diseases such as cancer. Gastrointestinal side effects commonly complicate treatment with the standard IPF therapies, Their association with treatment persistence and clinical outcomes, including disease progression and survival, remains a significant concern in IPF care (66-68). In our study, we further constructed core gene-miRNA-lncRNA network, which may inform the development of novel disease treatments.

Results from our study suggest that IPF increases the risk of BCC, though several limitations should be noted. Although statistically significant, the observed effect size [odds ratio (OR) <1.10] may not be biologically or clinically meaningful. This effect is considerably smaller than those of established environmental risk factors, such as UV light exposure (OR ≈2–3), indicating that even if a true effect exists, its magnitude is likely modest. Second, while our sensitivity analyses showed no evidence of directional pleiotropy or significant heterogeneity, we cannot completely rule out the possibility of residual horizontal pleiotropy. Third, due to limited availability of large-scale IPF genetic data and challenges in obtaining certain instrumental variables, the number of genetic instruments used in this study was relatively small, which may have limited the statistical power and precision of our estimates. Fourth, our analyses were conducted in populations of European ancestry (FinnGen and UK Biobank), which restricts the generalizability of our findings to other ethnic groups. Differences in population-specific genetic architecture and environmental exposures may influence the transferability of these results. Fifth, we analyzed skin cancer as a composite outcome, which included histological subtypes with potentially distinct etiologies (e.g., BCC, squamous cell carcinoma, and melanoma). The observed association may vary across subtypes, but our data did not allow for subtype-specific analyses. Finally, our MR approach estimated the lifelong effect of genetic predisposition to IPF, which may differ from the effects of clinically diagnosed IPF or its treatment. Chronic inflammatory states and immunosuppressive therapies used in IPF management could influence BCC risk in ways that differ from the effects of genetic susceptibility alone.

Nevertheless, because potential confounders may not have been fully adjusted for, more robust and validated approaches are needed to conclusively evaluate the causal relationship between these two conditions. Although randomized controlled trials are considered the gold standard for establishing causality, they face major limitations in this context in terms of practical feasibility, cost, and ethical considerations (9,69). With the growing integration of factors such as genomics into studies of underlying pathophysiological mechanisms and human disease treatment strategies, MR analysis has emerged as a powerful epidemiological tool (70). This method leverages genetic variation as an unbiased instrumental variable to investigate causal relationships between exposures and clinical disease outcomes. Using publicly available GWAS summary statistics, we examined the genetically predicted causal relationship between BCC and IPF within an MR framework. This study contributes to a more comprehensive understanding and assessment of the potential association between IPF and BCC. We also analyzed interactions between pathway genes and immune function to further elucidate the mechanism underlying the IPF-BCC link and to generate novel insights into the pathogenesis and treatment of these diseases. Given the limitations noted above, future studies with more comprehensive designs are warranted to reliably evaluate this association.

Conclusions

This study establishes that patients with IPF have an increased risk of developing BCC, driven by shared genetic and molecular pathways. Our findings provide a mechanistic basis for this association and highlight the need for skin cancer surveillance in IPF patients. The identified common pathological pathways offer novel opportunities for developing targeted therapies applicable to both conditions. Future research should focus on multi-ethnic validation and translational exploration of shared therapeutic targets.

Acknowledgments

None.

Footnote

Reporting Checklist: The authors have completed the STROBE-MR reporting checklist. Available at https://tcr.amegroups.com/article/view/10.21037/tcr-2025-1-2853/rc

Peer Review File: Available at https://tcr.amegroups.com/article/view/10.21037/tcr-2025-1-2853/prf

Funding: This work was supported by the National Major Science and Technology Project for Prevention and Treatment of Cancer, Cardiovascular and Cerebrovascular Diseases, Respiratory and Metabolic Diseases (No. 2024ZD0529503), administered by the Center for Medical Science and Technology Development, National Health Commission of the People’s Republic of China, and undertaken by the Chinese PLA General Hospital (Principal Investigator: W.Z.).

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://tcr.amegroups.com/article/view/10.21037/tcr-2025-1-2853/coif). The authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. King TE Jr, Albera C, Bradford WZ, et al. All-cause mortality rate in patients with idiopathic pulmonary fibrosis. Implications for the design and execution of clinical trials. Am J Respir Crit Care Med 2014;189:825-31. [Crossref] [PubMed]
2. Cai M, Zhu M, Ban C, et al. Clinical features and outcomes of 210 patients with idiopathic pulmonary fibrosis. Chin Med J (Engl) 2014;127:1868-73.
3. Raghu G, Rochwerg B, Zhang Y, et al. An Official ATS/ERS/JRS/ALAT Clinical Practice Guideline: Treatment of Idiopathic Pulmonary Fibrosis. An Update of the 2011 Clinical Practice Guideline. Am J Respir Crit Care Med 2015;192:e3-19. [Crossref] [PubMed]
4. Yang Y, Sheng YH, Carreira P, et al. Genome-wide assessment of shared genetic landscape of idiopathic pulmonary fibrosis and its comorbidities. Hum Genet 2024;143:1223-39. [Crossref] [PubMed]
5. Kreuter M, Ehlers-Tenenbaum S, Palmowski K, et al. Impact of Comorbidities on Mortality in Patients with Idiopathic Pulmonary Fibrosis. PLoS One 2016;11:e0151425. [Crossref] [PubMed]
6. Lee HY, Lee J, Lee CH, et al. Risk of cancer incidence in patients with idiopathic pulmonary fibrosis: A nationwide cohort study. Respirology 2021;26:180-7. [Crossref] [PubMed]
7. Peris K, Licitra L, Ascierto PA, et al. Identifying locally advanced basal cell carcinoma eligible for treatment with vismodegib: an expert panel consensus. Future Oncol 2015;11:703-12. [Crossref] [PubMed]
8. Dika E, Scarfì F, Ferracin M, et al. Basal Cell Carcinoma: A Comprehensive Review. Int J Mol Sci 2020;21:5572. [Crossref] [PubMed]
9. Travis WD, Costabel U, Hansell DM, et al. An official American Thoracic Society/European Respiratory Society statement: Update of the international multidisciplinary classification of the idiopathic interstitial pneumonias. Am J Respir Crit Care Med 2013;188:733-48. [Crossref] [PubMed]
10. Barrett T, Troup DB, Wilhite SE, et al. NCBI GEO: mining tens of millions of expression profiles--database and tools update. Nucleic Acids Res 2007;35:D760-5. [Crossref] [PubMed]
11. Ritchie ME, Phipson B, Wu D, et al. limma powers differential expression analyses for RNA-sequencing and microarray studies. Nucleic Acids Res 2015;43:e47. [Crossref] [PubMed]
12. Leek JT, Johnson WE, Parker HS, et al. sva: Surrogate Variable Analysis. doi: 10.18129/B9.bioc.sva. R package version 3.58.0. 2025. Available online: https://bioconductor.org/packages/sva
13. Wickham H. ggplot2: Elegant Graphics for Data Analysis. Springer-Verlag New York. 2016. Available online: https://ggplot2.tidyverse.org
14. Kassambara A. ggpubr: ‘ggplot2’ Based Publication Ready Plots. R package version 0.6.0. 2023. Available online: https://CRAN.R-project.org/package=ggpubr
15. Kolde R. Pheatmap: Pretty Heatmaps. R package version 1.0.12. 2019. Available online: https://CRAN.R-project.org/package=pheatmap
16. Wu T, Hu E, Xu S, et al. clusterProfiler 4.0: A universal enrichment tool for interpreting omics data. Innovation (Camb) 2021;2:100141. [Crossref] [PubMed]
17. Kanehisa M, Furumichi M, Sato Y, et al. KEGG for taxonomy-based analysis of pathways and genomes. Nucleic Acids Res 2023;51:D587-92. [Crossref] [PubMed]
18. Hemani G, Zheng J, Elsworth B, et al. The MR-Base platform supports systematic causal inference across the human phenome. Elife 2018;7:e34408. [Crossref] [PubMed]
19. Emdin CA, Khera AV, Kathiresan S. Mendelian Randomization. JAMA 2017;318:1925-6. [Crossref] [PubMed]
20. Dudbridge F. Polygenic Mendelian Randomization. Cold Spring Harb Perspect Med 2021;11:a039586. [Crossref] [PubMed]
21. Morgan M, Ramos M. BiocManager: Access the Bioconductor Project Package Repository. R package version 1.30.22. 2023. Available online: https://CRAN.R-project.org/package=BiocManager
22. Cannon M, Stevenson J, Stahl K, et al. DGIdb 5.0: rebuilding the drug-gene interaction database for precision medicine and drug discovery platforms. Nucleic Acids Res 2024;52:D1227-35. [Crossref] [PubMed]
23. Huang HY, Lin YC, Cui S, et al. miRTarBase update 2022: an informative resource for experimentally validated miRNA-target interactions. Nucleic Acids Res 2022;50:D222-30. [Crossref] [PubMed]
24. Shannon P, Markiel A, Ozier O, et al. Cytoscape: a software environment for integrated models of biomolecular interaction networks. Genome Res 2003;13:2498-504. [Crossref] [PubMed]
25. Newman AM, Steen CB, Liu CL, et al. Determining cell type abundance and expression from bulk tissues with digital cytometry. Nat Biotechnol 2019;37:773-82. [Crossref] [PubMed]
26. National Center for Biotechnology Information. “Gene Expression Omnibus. GSE38958.” Bethesda, MD: National Library of Medicine. 2013. Available online: https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE38958
27. Rogers HW, Weinstock MA, Feldman SR, et al. Incidence Estimate of Nonmelanoma Skin Cancer (Keratinocyte Carcinomas) in the U.S. Population, 2012. JAMA Dermatol 2015;151:1081-6. [Crossref] [PubMed]
28. Lomas A, Leonardi-Bee J, Bath-Hextall F. A systematic review of worldwide incidence of nonmelanoma skin cancer. Br J Dermatol 2012;166:1069-80. [Crossref] [PubMed]
29. Bolaños AL, Milla CM, Lira JC, et al. Role of Sonic Hedgehog in idiopathic pulmonary fibrosis. Am J Physiol Lung Cell Mol Physiol 2012;303:L978-90. [Crossref] [PubMed]
30. Javelaud D, Pierrat MJ, Mauviel A. Crosstalk between TGF-β and hedgehog signaling in cancer. FEBS Lett 2012;586:2016-25. [Crossref] [PubMed]
31. Jiang J. Hedgehog signaling mechanism and role in cancer. Semin Cancer Biol 2022;85:107-22. [Crossref] [PubMed]
32. Pietanza MC, Litvak AM, Varghese AM, et al. A phase I trial of the Hedgehog inhibitor, sonidegib (LDE225), in combination with etoposide and cisplatin for the initial treatment of extensive stage small cell lung cancer. Lung Cancer 2016;99:23-30. [Crossref] [PubMed]
33. Xie H, Paradise BD, Ma WW, et al. Recent Advances in the Clinical Targeting of Hedgehog/GLI Signaling in Cancer. Cells 2019;8:394. [Crossref] [PubMed]
34. Didiasova M, Singh R, Wilhelm J, et al. Pirfenidone exerts antifibrotic effects through inhibition of GLI transcription factors. FASEB J 2017;31:1916-28. [Crossref] [PubMed]
35. Leask A, Abraham DJ. TGF-beta signaling and the fibrotic response. FASEB J 2004;18:816-27. [Crossref] [PubMed]
36. Coward WR, Saini G, Jenkins G. The pathogenesis of idiopathic pulmonary fibrosis. Ther Adv Respir Dis 2010;4:367-88. [Crossref] [PubMed]
37. Zolak JS, de Andrade JA. Idiopathic pulmonary fibrosis. Immunol Allergy Clin North Am 2012;32:473-85. [Crossref] [PubMed]
38. Kinoshita T, Goto T. Molecular Mechanisms of Pulmonary Fibrogenesis and Its Progression to Lung Cancer: A Review. Int J Mol Sci 2019;20:1461. [Crossref] [PubMed]
39. Raghu G, Remy-Jardin M, Myers JL, et al. Diagnosis of Idiopathic Pulmonary Fibrosis. An Official ATS/ERS/JRS/ALAT Clinical Practice Guideline. Am J Respir Crit Care Med 2018;198:e44-68. [Crossref] [PubMed]
40. Lederer DJ, Martinez FJ. Idiopathic Pulmonary Fibrosis. N Engl J Med 2018;378:1811-23. [Crossref] [PubMed]
41. Lettieri S, Oggionni T, Lancia A, et al. Immune Stroma in Lung Cancer and Idiopathic Pulmonary Fibrosis: A Common Biologic Landscape? Int J Mol Sci 2021;22:2882. [Crossref] [PubMed]
42. Lei Y, Zhong C, Zhang J, et al. Senescent lung fibroblasts in idiopathic pulmonary fibrosis facilitate non-small cell lung cancer progression by secreting exosomal MMP1. Oncogene 2025;44:769-81. [Crossref] [PubMed]
43. Han Y, Zhang L, Sun D, et al. Spatiotemporal analyses of the pan-cancer single-cell landscape reveal widespread profibrotic ecotypes associated with tumor immunity. Nat Cancer 2025;6:1880-98. [Crossref] [PubMed]
44. Esposito M, Yerly L, Shukla P, et al. COL10A1 expression distinguishes a subset of cancer-associated fibroblasts present in the stroma of high-risk basal cell carcinoma. Br J Dermatol 2024;191:775-90. [Crossref] [PubMed]
45. Planas-Cerezales L, Arias-Salgado EG, Buendia-Roldán I, et al. Predictive factors and prognostic effect of telomere shortening in pulmonary fibrosis. Respirology 2019;24:146-53. [Crossref] [PubMed]
46. Schafer MJ, White TA, Iijima K, et al. Cellular senescence mediates fibrotic pulmonary disease. Nat Commun 2017;8:14532. [Crossref] [PubMed]
47. Srinivas N, Rachakonda S, Hielscher T, et al. Telomere length, arsenic exposure and risk of basal cell carcinoma of skin. Carcinogenesis 2019;40:715-23. [Crossref] [PubMed]
48. Naikawadi RP, Disayabutr S, Mallavia B, et al. Telomere dysfunction in alveolar epithelial cells causes lung remodeling and fibrosis. JCI Insight 2016;1:e86704. [Crossref] [PubMed]
49. Le Jeune I, Gribbin J, West J, et al. The incidence of cancer in patients with idiopathic pulmonary fibrosis and sarcoidosis in the UK. Respir Med 2007;101:2534-40. [Crossref] [PubMed]
50. Jaeger B, Schupp JC, Plappert L, et al. Airway basal cells show a dedifferentiated KRT17(high)Phenotype and promote fibrosis in idiopathic pulmonary fibrosis. Nat Commun 2022;13:5637. [Crossref] [PubMed]
51. Huang G, Liang J, Huang K, et al. Basal Cell-derived WNT7A Promotes Fibrogenesis at the Fibrotic Niche in Idiopathic Pulmonary Fibrosis. Am J Respir Cell Mol Biol 2023;68:302-13. [Crossref] [PubMed]
52. Hayat R, Manzoor M, Hussain A. Wnt signaling pathway: A comprehensive review. Cell Biol Int 2022;46:863-77. [Crossref] [PubMed]
53. Dees C, Schlottmann I, Funke R, et al. The Wnt antagonists DKK1 and SFRP1 are downregulated by promoter hypermethylation in systemic sclerosis. Ann Rheum Dis 2014;73:1232-9. [Crossref] [PubMed]
54. Liang J, Fang F, Gao X, et al. LncRNA NEAT1 promotes proliferation, migration, and invasion of laryngeal squamous cell carcinoma cells through miR-411-3p/FZD3-mediated Wnt signaling pathway. BMC Cancer 2024;24:904. [Crossref] [PubMed]
55. Wang RN, Green J, Wang Z, et al. Bone Morphogenetic Protein (BMP) signaling in development and human diseases. Genes Dis 2014;1:87-105. [Crossref] [PubMed]
56. Dong X, Mao Y, Gao P. The Role of Bone Morphogenetic Protein 4 in Lung Diseases. Curr Mol Med 2023;23:324-31. [Crossref] [PubMed]
57. Al-Bahrani R, Nagamori S, Leng R, et al. Differential Expression of Sonic Hedgehog Protein in Human Hepatocellular Carcinoma and Intrahepatic Cholangiocarcinoma. Pathol Oncol Res 2015;21:901-8. [Crossref] [PubMed]
58. Athar M, Tang X, Lee JL, et al. Hedgehog signalling in skin development and cancer. Exp Dermatol 2006;15:667-77. [Crossref] [PubMed]
59. Guo Y, Tong Y, Zhu H, et al. Quercetin suppresses pancreatic ductal adenocarcinoma progression via inhibition of SHH and TGF-β/Smad signaling pathways. Cell Biol Toxicol 2021;37:479-96. [Crossref] [PubMed]
60. Hahn H, Wicking C, Zaphiropoulous PG, et al. Mutations of the human homolog of Drosophila patched in the nevoid basal cell carcinoma syndrome. Cell 1996;85:841-51. [Crossref] [PubMed]
61. Bonilla X, Parmentier L, King B, et al. Genomic analysis identifies new drivers and progression pathways in skin basal cell carcinoma. Nat Genet 2016;48:398-406. [Crossref] [PubMed]
62. Basset-Seguin N, Hauschild A, Grob JJ, et al. Vismodegib in patients with advanced basal cell carcinoma (STEVIE): a pre-planned interim analysis of an international, open-label trial. Lancet Oncol 2015;16:729-36. [Crossref] [PubMed]
63. Gupta AK, Mays RR, Abramovits W, et al. Odomzo(®) (Sonidegib). Skinmed 2018;16:35-8.
64. Xiao H, Zhang GF, Liao XP, et al. Anti-fibrotic effects of pirfenidone by interference with the hedgehog signalling pathway in patients with systemic sclerosis-associated interstitial lung disease. Int J Rheum Dis 2018;21:477-86. [Crossref] [PubMed]
65. Glass DS, Grossfeld D, Renna HA, et al. Idiopathic pulmonary fibrosis: Current and future treatment. Clin Respir J 2022;16:84-96. [Crossref] [PubMed]
66. Jovanovic D, Mogulkoc N, Sterclova M, et al. Adverse side-effects of antifibrotic drugs and outcome of treatment in idiopathic pulmonary fibrosis. J Thorac Dis 2025;17:10908-23. [Crossref] [PubMed]
67. Kou M, Jiao Y, Li Z, et al. Real-world safety and effectiveness of pirfenidone and nintedanib in the treatment of idiopathic pulmonary fibrosis: a systematic review and meta-analysis. Eur J Clin Pharmacol 2024;80:1445-60. [Crossref] [PubMed]
68. Doubkova M, Kriegova E, Littnerova S, et al. DSP rs2076295 variants influence nintedanib and pirfenidone outcomes in idiopathic pulmonary fibrosis: a pilot study. Ther Adv Respir Dis 2021;15:17534666211042529. [Crossref] [PubMed]
69. Bothwell LE, Podolsky SH. The Emergence of the Randomized, Controlled Trial. N Engl J Med 2016;375:501-4. [Crossref] [PubMed]
70. Davies NM, Holmes MV, Davey Smith G. Reading Mendelian randomisation studies: a guide, glossary, and checklist for clinicians. BMJ 2018;362:k601. [Crossref] [PubMed]