Phenotypic convergence and shared signalling networks

Macrophages are key participants in innate immunity and in the maintenance of pulmonary homeostasis, exhibiting high plasticity across different pathological contexts (Ref. Reference Aegerter, Lambrecht and Jakubzick125). During the progression of IPF, macrophages act as the primary regulatory hub within the immune–cell interaction network by communicating intensely with specific populations, including T cells, natural killer T (NKT) cells, neutrophils and DCs (Ref. Reference Lv, Gao, Ma, Liu, Tian, Yang, Li, Zhao, Li, Zhang, Zhu, Zhang and Wu126). Furthermore, they play pivotal roles in the establishment of the tumour-associated microenvironment and in tumour immune evasion, invasion and metastasis (Ref. Reference De Zuani, Xue, Park, Dentro, Seferbekova, Tessier, Curras-Alonso, Hadjipanayis, Athanasiadis, Gerstung, Bayraktar and Cvejic127). In both IPF and LC, M2 polarization emerges not merely as a phenotypic shift but as a convergent molecular response driven by shared signalling networks under chronic tissue injury.

Under disease-associated signals, macrophages predominantly polarize into the M2 phenotype. Studies have shown that in IPF-LC mouse models, M2-type macrophages with high expression of Arg1 and programmed death-ligand 1 (PD-L1) are markedly increased, accompanied by a reduction and functional suppression of CD8⁺ T cells. A prevailing consensus in the literature indicates that these alterations collectively create a profoundly immunosuppressive microenvironment that accelerates tumour progression (Ref. Reference Barravecchia, Lee, Manassa, Magnuson, Ferris, Cavanaugh, Steele, Espinoza, Galban, Ramnath, Frankel, Pasca di Magliano and Galban21). However, it remains unclear whether this M2 polarization reflects identical epigenetic alterations in both diseases or independent processes that merely overlap in their downstream fibrotic and immunosuppressive phenotypes.

Individually, M2 macrophages in IPF drive fibrogenesis by secreting profibrotic mediators such as TGF-β and PDGF, which activate fibroblasts via shared gene regulatory networks, including TGF-β1/SMAD3 and Wnt/β–catenin pathways (Refs Reference Hou, Shi, Chen, Lv, Chen, Cao, Xiang and Han13–Reference Wu, Zhang, Jiang, He, Zhang, Chen, Tong, Nie and Pang15). Similarly, in LC, M2-like tumour-associated macrophages (TAMs) predominate, recruited by tumour-derived factors such as prostaglandin E2 (PGE2), monocyte chemoattractant protein-1 (MCP-1) and gamma-aminobutyric acid (GABA) (Refs Reference Dong, Wang, Nie, Xu, Bai, Lu, Jian, Wang and Zheng16–Reference Kawaguchi, Ohshio, Watanabe, Shiratori, Okamoto, Ueda, Kataoka, Suzuki and Hanaoka18). These TAMs remodel the tumour microenvironment (TME) by promoting EMT via the ERK1/2/Fra-1/Slug axis, stimulating angiogenesis and suppressing immune surveillance through IL-10 secretion (Refs Reference Guo, Song, Hao, Zhao, du, Li, Kuang, Yang, Wang, Deng and Wang19, Reference Zhao, Zhang, Liu, Xue, Chen, Zou and Jiang20). Despite varying initial triggers, IPF and LC converge on similar macrophage-driven cascades to maintain a fibrotic and tumorigenic microenvironment.

Metabolic reprogramming and immune suppression

In the coexisting IPF-LC microenvironment, macrophages not only undergo phenotypic changes but also exhibit features of functional suppression and altered metabolic states. During IPF-to-LC progression, single-cell RNA sequencing (scRNA-seq) and genetic analyses reveal that macrophages lose their pro-inflammatory capacity and IFN-γ responsiveness, accompanied by widespread T- and B-cell suppression. Moreover, fibroblasts can communicate with macrophages via the highly active macrophage migration inhibitory factor (MIF) pathway, ultimately orchestrating an immune-evasive microenvironment (Ref. Reference Gu, Xu, Jiao, Cai, Gu, Wu, Cheng, Chen, Chen and Liu22). Notably, in microenvironments where IPF and LC coexist, macrophages exhibit pronounced metabolic plasticity. Genetically, this metabolic reprogramming is tightly regulated by shared transcriptional networks, such as the JUN/SOD2-glycolysis axis, while simultaneously polarizing towards a profibrotic phenotype driven by elevated lipid metabolism and oxidative phosphorylation. Despite the differences in energy metabolism, the outcome converges on the polarization of macrophages towards M2/TAM-like phenotypes. These reprogrammed macrophages secrete multiple immunoregulatory and profibrotic factors, including TGF-β, IL-10 and CCL18, thereby promoting EMT and the formation of an immunosuppressive microenvironment, ultimately driving the comorbidity of IPF and LC (Refs Reference Gu, Xu, Jiao, Cai, Gu, Wu, Cheng, Chen, Chen and Liu22, Reference Ren, Zi, Zhang, Liao and Chen23).

Characterization of shared cellular subpopulations

Macrophages play a key role in both fibrosis and tumour progression, and based on this, researchers have gradually identified specific macrophage subpopulations in the IPF-LC microenvironment. For instance, Barravecchia et al. found a distinct subpopulation of TAMs (CD45⁺, CD206^low, CD11c⁺) significantly increased in the IPF-LC mouse model, suggesting the presence of polarized macrophage populations in the fibrosis–cancer coexisting microenvironment (Ref. Reference Barravecchia, Lee, Manassa, Magnuson, Ferris, Cavanaugh, Steele, Espinoza, Galban, Ramnath, Frankel, Pasca di Magliano and Galban21). Among the various macrophage subpopulations, secreted phosphoprotein 1 (SPP1)⁺ macrophages have garnered significant attention from researchers. Analysis based on the Human Lung Cell Atlas (HLCA) revealed the coexistence of SPP1⁺ profibrotic monocyte-derived macrophages in both pulmonary fibrosis and LC (Ref. Reference Sikkema, Ramírez-Suástegui, Strobl, Gillett, Zappia, Madissoon, Markov, Zaragosi, Ji, Ansari, Arguel, Apperloo, Banchero, Bécavin, Berg, Chichelnitskiy, Chung, Collin, Gay, Gote-Schniering, Hooshiar Kashani, Inecik, Jain, Kapellos, Kole, Leroy, Mayr, Oliver, von Papen, Peter, Taylor, Walzthoeni, Xu, Bui, de Donno, Dony, Faiz, Guo, Gutierrez, Heumos, Huang, Ibarra, Jackson, Kadur Lakshminarasimha Murthy, Lotfollahi, Tabib, Talavera-López, Travaglini, Wilbrey-Clark, Worlock, Yoshida, Chen, Hagood, Agami, Horvath, Lundeberg, Marquette, Pryhuber, Samakovlis, Sun, Ware, Zhang, van den Berge, Bossé, Desai, Eickelberg, Kaminski, Krasnow, Lafyatis, Nikolic, Powell, Rajagopal, Rojas, Rozenblatt-Rosen, Seibold, Sheppard, Shepherd, Sin, Timens, Tsankov, Whitsett, Xu, Banovich, Barbry, Duong, Falk, Meyer, Kropski, Pe’er, Schiller, Tata, Schultze, Teichmann, Misharin, Nawijn, Luecken and Theis24). Further, Chen et al. utilized scRNA-seq to identify that SPP1⁺ macrophages, located at the terminal end of the developmental trajectory, may play a key role in driving the progression from IPF to LC. The underlying mechanism may involve not only the induction of fibroblast transformation into cancer-associated fibroblasts (CAFs), but also interactions with these inflammation-associated CAFs and stimulation of EMT in alveolar epithelial cells (Ref. Reference Chen, Guo, Ai, Wang, Wang, Chen, Liu, Zhuang, Liu, Zhao, Liang and Gu25). These findings establish SPP1⁺ macrophages as a mechanistic link bridging fibrogenesis and tumorigenesis, rather than representing a phenotypic overlap.

Clinical implications and targeted therapies

Consequently, targeting these shared populations or their regulatory pathways, such as AXL, PI3K/Akt or triggering receptor expressed on myeloid cells-2 (TREM2), represents a promising therapeutic strategy (Refs Reference Kim, Kim, Im, Baek, Choi, Lee, Kim, Lee, Jeong, Ban, Kim, Ji, Lee, Kim, Lee, Yang, Yun, Choi and Rho128–Reference Li, Yuan, Li, He, Shen, Luo, Tai, Li, Deng and Luo131). Given the critical role of macrophages in the IPF and LC microenvironments, expanding efforts are directed towards macrophage-targeted therapies. Liu et al. developed a novel delivery system for treating LC combined with pulmonary fibrosis: macrophage membrane-coated rapamycin micelles (MM@TPGS/F127@RAPA) administered via intratracheal instillation. The design leverages the high similarity in macrophage biological behaviours between IPF and LC, including the recruitment of macrophages mediated by similar chemokines and the immune microenvironment characterized by M2 polarization. The encapsulated rapamycin targets shared signalling pathways, such as TGF-β1/SMAD3 and PI3K/AKT/mTOR, thereby enabling precise drug delivery and synergistic therapy (Ref. Reference Liu, Cai, Sun, Zhao, xu, Gu, Peng, di, Zhou, Gu, Ai and Yang132). Notably, with the continuous identification of specialized macrophage subsets, such as SPP1⁺ macrophages, targeting these cells that exhibit both profibrotic and pro-tumour characteristics holds promise as a new treatment approach to intervene in fibrosis and reduce the risk of LC.

CD8⁺ T cells

As classical cytotoxic effector cells, CD8⁺ T cells primarily function by eliminating infected or malignant cells. However, in IPF, their function becomes markedly dysregulated. Although CD8⁺ T cells are abundantly infiltrated in the lung tissue of patients with IPF, they do not exhibit effective cytotoxic activity. Instead, subsets of these cells display features of exhaustion, undergoing metabolic reprogramming and activation of signalling pathways such as JAK–STAT, NF–κB and MAPK, thereby promoting fibroblast activation and driving the progression of fibrosis (Ref. Reference Wei, Jin, Li, Wang, Zheng, Feng, shi, Kong, Ma and Wang26). In both diseases, exhausted CD8⁺ T cells highly express immune checkpoint molecules such as programmed cell death protein 1 (PD-1) and cytotoxic T lymphocyte–associated protein 4 (CTLA4), which impair immune surveillance and the clearance of senescent or malignant cells (Refs Reference Yadav, Anbalagan, Khatun, Prabhakaran, Matsunaga, Manges, McLachlan, Lasky, Kolls and Thannickal27, Reference Guo, Zhang, Zheng, Zheng, Song, Zhang, Kang, Liu, Jin, Xing, Gao, Zhang, Dong, Hu, Ren, Kirchhoff, Roider, Yan and Zhang38). This progressive loss of anti-fibrotic and anti-tumour capacity represents a crucial mechanism for chronic disease progression and a primary target for immunotherapy.

CD4⁺ T cells

CD4⁺ T cells possess a variety of immune regulatory functions, primarily categorized into T helper (Th) cells and regulatory T cells (Tregs). A consistent consensus in the literature indicates that both IPF and LC are characterized by a systemic shift towards Th2 dominance, representing a shared immune–metabolic adaptation. In bleomycin (BLM)-induced pulmonary fibrosis, this is evidenced by a significant reduction in Th1 cytokine IFN-γ levels and a notable increase in Th2 cytokine IL-4 levels, leading to a decreased IFN-γ/IL-4 ratio (Refs Reference Xing, Qiang, Bao, Yang, Hou, Tao, Meng, Zhang, Zhang and Sun28, Reference Mojiri-Forushani, Hemmati, Khodadadi and Rashno29). Similarly, in the TME of NSCLC, a distinct Th2 skewing is observed, with Th2 and Treg cells predominating in both the tumour epithelium and stroma, while Th1 cells are significantly reduced (Ref. Reference Frafjord, Buer, Hammarström, Aamodt, Woldbæk, Brustugun, Helland, Øynebråten and Corthay39).

Other CD4⁺ subsets, including Th9 and Th17 cells, act as pathogenic drivers by activating STAT3 and SMAD2/3 signalling. Specifically, IL-9 and IL-17A secreted by these cells induce collagen production in fibroblasts, promote EMT in epithelial cells and enhance tumour cell migration and chemoresistance (Refs Reference Deng, Yang, Luo, She, Yu and Tang30, Reference Celada, Kropski, Herazo-Maya, Luo, Creecy, Abad, Chioma, Lee, Hassell, Shaginurova, Wang, Johnson, Kerrigan, Mason, Baughman, Ayers, Bernard, Culver, Montgomery, Maher, Molyneaux, Noth, Mutsaers, Prele, Stokes Peebles, Newcomb, Kaminski, Blackwell, van Kaer and Drake31, Reference Wang, Yang, Zhang, Wang, Zhang, He, Chen, Zhang, Qin, Wang and Zhang40–Reference Heim, Yang, Tausche, Hohenberger, Chiriac, Koelle, Geppert, Kachler, Miksch, Graser, Friedrich, Kharwadkar, Rieker, Trufa, Sirbu, Neurath, Kaplan and Finotto42).

Tregs play a central role in maintaining immune system homeostasis, but their function is disrupted in both IPF and LC. However, the literature presents contradictory evidence regarding their precise role in fibrosis. While some studies suggest Tregs exert anti-fibrotic effects, others indicate they undergo phenotypic instability and drive a pro-fibrotic shift from Th1 to Th2 cells (Refs Reference Kamio, Azuma, Matsuda, Usuki, Inomata, Morinaga, Kashiwada, Nishijima, Itakura, Kokuho, Atsumi, Hayashi, Yamaguchi, Fujita, Saito, Abe, Kubota and Gemma32–Reference Birjandi, Palchevskiy, Xue, Nunez, Kern, Weigt, Lynch, Chatila and Belperio35). Furthermore, in the LC microenvironment, activated Tregs directly inhibit CD8⁺ T-cell proliferation and IFN-γ production via PD-1 and PD-L1 interactions, thereby promoting immune evasion (Refs Reference Guo, Zhang, Zheng, Zheng, Song, Zhang, Kang, Liu, Jin, Xing, Gao, Zhang, Dong, Hu, Ren, Kirchhoff, Roider, Yan and Zhang38, Reference Kim, Park, Son, Lee, Chung, Cho, Shim, Park, Kim, in Yoon, Kim, Jung, Cho, Park, Rha and Ha43).

γδ T cells

Additionally, non-conventional T cells, such as γδ T cells, are activated in both diseases and participate in early immune responses by secreting cytokines such as IL-17 and IFN-γ, which influence chronic inflammation, fibrosis and TME remodelling. However, their functional roles differ between the two diseases. In the BLM-induced pulmonary fibrosis model, γδ T cells are considered key regulatory cells, playing a protective role in lung fibrosis by producing IFN-γ and suppressing Th17 immune responses (Refs Reference Segawa, Goto, Iizuka, Kaneko, Yokosawa, Kondo, Matsumoto and Sumida36, Reference Russo, Savino, Mirolo, Buracchi, Germano, Anselmo, Zammataro, Pasqualini, Mantovani, Locati and Teixeira37). In contrast, the role of γδ T cells in NSCLC are dualistic: γδ T cells producing IFN-γ exhibit anti-tumour activity and are associated with favourable long-term overall survival, while γδ T17 cells producing IL-17 may promote tumour progression by mediating pro-tumour inflammatory responses (Refs Reference Xu, Ke, Shang, Liu, Fu, Wang and Jin44, Reference Song, Ma, Chen, Miao, Tao and Liu45). It remains unclear whether these contrasting roles reflect disease-specific mechanisms or dynamic phenotypic shifts dictated by the evolving tissue microenvironment.

IPF and LC exhibit several key similarities in T cell-mediated immunity, including CD8⁺ exhaustion, Th2 skewing, activation of the Th9/Th17 axis and dysfunction of Tregs. Notably, high expression of PD-1 across multiple T-cell subsets acts as a central driver of immune tolerance in both microenvironments. However, there is currently a lack of reports on the exact spatial distribution and epigenetic state of T cells in the coexisting IPF-LC microenvironment. Further multi-omics investigation is needed to elucidate their immune characteristics and provide a foundation for the development of targeted therapeutic strategies.

Cellular origins and phenotypic plasticity

Firstly, from the perspective of cellular origins, fibroblasts in both IPF and LC exhibit pronounced heterogeneity and derive from multiple sources. While resident pulmonary fibroblasts serve as the primary reservoir for expansion, the fibrotic and tumour stroma are significantly augmented by alternative lineages through shared epigenetic reprogramming and transdifferentiation programmes (Refs Reference Cao, Li, Zheng, Zhang, Qian, Ren, Wang, Wang, Wu, Zhang and Xu81, Reference Walsh, Worrell, Fabre, Hinz, Kane and Keane82). Specifically, both microenvironments recruit bone-marrow–derived fibrocytes or mesenchymal stem cells and harness the plasticity of non-mesenchymal cells via EMT and endothelial–mesenchymal transition (EndMT) (Refs Reference Walsh, Worrell, Fabre, Hinz, Kane and Keane82–Reference Wong, Cheung, Chen, Chan, Yu, Lo, Kang and To85). This cellular repertoire is further diversified by the transdifferentiation of pericytes, adipocytes and pleural mesothelial cells undergoing mesothelial–mesenchymal transition (MMT) (Refs Reference Wong, Cheung, Chen, Chan, Yu, Lo, Kang and To85–Reference Liu, Yu, Lu, Cheng, Liang, Wang, Chen, Huang, Song, He, Xiong, Xin, Ma and Ye88). Notably, recent studies have identified SMAD3-driven macrophage-to-myofibroblast transition as a novel mechanism contributing to CAF generation, revealing immune cells as an unexpected progenitor pool (Ref. Reference Tang, Chung, Xue, Xiao, Meng, Huang, Zhou, Chan, Tsang, Cheng, Lee, Leung, Lam, To, Tang and Lan89). These diverse origins reflect a profound phenotypic plasticity, where distinct cellular lineages converge into a unified pathogenic stroma under chronic injury.

Biomechanical and metabolic convergence

Excessive fibroblast activation constitutes a central pathological engine that drives the aberrant ECM remodelling shared by IPF and LC. In both diseases, this activation represents a convergent cellular response to continuous mechanical stress and biochemical stimuli. In IPF, hyperactivated fibroblasts secrete excessive collagens, leading to increased ECM stiffness, which acts as an abnormal mechanotransduction platform that sustains a vicious cycle of fibroblast activation (Ref. Reference Yan, Wang, Su, Zou, Zhou, Shou and Huo68). Similarly, in the TME, myofibroblast-derived collagens such as COL5A1 elevate intratumoral mechanical stress, mechanically triggering EMT and enhancing therapeutic resistance (Ref. Reference Zhu, Wang, Wang, Huang, Li, Chen, Chen, Zhang, Li, Liu and Chen71).

Genetically and metabolically, this structural remodelling is underpinned by shared transcriptional networks. Fibroblasts in both diseases display enhanced glycolysis and lactate secretion. In IPF, the glycolysis driven by 6-phosphofructo-2-kinase/fructose-2,6-biphosphatase 3 (PFKFB3) stabilizes HIF-1α via succinate, promoting fibroblast-to-myofibroblast differentiation (Ref. Reference Xie, Tan, Banerjee, Cui, Ge, Liu, Bernard, Thannickal and Liu69). Additionally, epigenetic modifications play a crucial role: glycolysis-derived lactate functions as a paracrine signal, to induce histone lactylation in macrophages and promote the expression of pro-fibrotic genes (Ref. Reference Cui, Xie, Banerjee, Ge, Jiang, Dey, Matthews, Liu and Liu70). In LC, this metabolic shift manifests as the ‘Warburg’ and ‘reverse Warburg’ effects, where CAFs secrete lactate and energy substrates to support tumour growth, while cancer cell-derived lactate promotes M2 macrophage polarization via CAFs (Refs Reference Cruz-Bermúdez, Laza-Briviesca, Vicente-Blanco, García-Grande, Coronado, Laine-Menéndez, Alfaro, Sanchez, Franco, Calvo, Romero, Martin-Acosta, Salas, Garcia and Provencio72–Reference Duda, Janczara, McCubrey, Gizak and Rakus74). Beyond direct metabolite secretion, CAFs can also modulate tumour metabolism through extracellular vesicle communication. Specifically, CAF-derived exosomes are enriched in long noncoding RNA (lncRNA) ANRIL, which acts as a competing endogenous RNA to sponge miR-186-5p. This relieves the inhibition of hypoxia-inducible factor 1α (HIF-1α), thereby enhancing glycolytic activity and proliferation of LC cells (Ref. Reference Zhang, Huang, Tang and Hu75). This metabolic and structural convergence is underpinned by a shared molecular program. Bioinformatics analyses consistently confirm this molecular convergence, identifying shared hub genes, such as COL1A1, COL3A1, COL4A1, POSTN, TIMP3 and MMP1/3, which are enriched in both IPF fibroblast foci and LC stroma (Refs Reference Kreus, Lehtonen, Skarp and Kaarteenaho90, Reference Yao, Li and Gao91). Furthermore, research has found that a HER2-driven regulatory programme promotes cell invasion and matrix remodelling across both fibrotic foci and lung adenocarcinoma stroma (Ref. Reference Liu, Geng, Liang, Coelho, Yao, Deng, Wang, Dai, Huang, Xie, Liu, Rowan, Taghavifar, Kulur, Liu, Stripp, Hogaboam, Jiang and Noble92).

Intercellular communication in the coexisting microenvironment

Within the coexisting IPF-LC microenvironment, activated IPF fibroblasts function as a pivotal communication nexus bridging the pathological gap between fibrogenesis and malignancy (Figure 2). This crosstalk is primarily driven by extracellular vesicles (EVs), exosomes and secreted soluble factors. For instance, senescent IPF fibroblasts release EVs enriched with miR-19a to activate c-Myc signalling, while exosomal MMP1 triggers the PI3K–AKT–mTOR axis in NSCLC cells. Collectively, these dual pathways accelerate tumour proliferation, invasion and EMT (Refs Reference Lei, Zhong, Zhang, Zheng, Xu, Li, Huang and Ren76, Reference Fujita, Fujimoto, Miyamoto, Kaneko, Kadota, Watanabe, Kizawa, Kawamoto, Watanabe, Utsumi, Wakui, Minagawa, Araya, Ohtsuka, Ochiya and Kuwano77). Beyond vesicular transport, IPF fibroblasts secrete Periostin, which engages integrin β3 receptors to stimulate the Erk/Akt pathway and fuel tumour progression (Ref. Reference Yamato, Kimura, Fukui, Kanou, Ose, Funaki, Minami and Shintani78). Notably, molecules such as SPP1 and prostaglandin E synthase (PTGES) serve as dual-pathological drivers. While SPP1 activates the PI3K/Akt/mTOR pathway to promote both fibroblast expansion and cancer cell proliferation, PTGES facilitates myofibroblast differentiation by modulating AKT and the HIF-1α pathway (Refs Reference Lin, Lee, Liao, Chou and Yang79, Reference Yue, Xiong, Chen, Yang, Zhao, Shao, Wei, Gao, Huang and Chen80). However, it is still unclear whether IPF-specific myofibroblasts can directly transdifferentiate into tumour-promoting CAFs within this shared niche, or if they represent independent parallel lineages. Elucidating the spatiotemporal transcriptomic trajectories of these specific stromal cells is urgently needed.

Figure 2.

Intercellular communication of fibroblasts in the IPF-LC microenvironment (created with BioRender.com). Notes: EMT: epithelial–mesenchymal transition; HIF-1α: Hypoxia-inducible factor 1α; IPF: idiopathic pulmonary fibrosis; IPF-LC: idiopathic pulmonary fibrosis associated with lung cancer; MMP1: matrix metalloproteinase 1; NSCLC: non-small–cell lung cancer; PTGES: prostaglandin E synthase; SPP1: secreted phosphoprotein 1.

Figure 2. Long description

The diagram is titled Intercellular Communication in the I P F-L C Microenvironment. At the top, a header box labeled Activated I P F Fibroblasts states they act as a communication nexus to accelerate tumor proliferation, invasion, and E M T.

Below this, a blue star-shaped I P F fibroblast is the primary source. It has three main output paths:

* A horizontal black arrow points directly to a cluster of pink N S C L C cells.

* A downward-curving blue arrow represents secreted factors leading to a Periostin node, which then points to integrin beta 3 receptors to stimulate the E r k forward slash A k t pathway, targeting a central N S C L C cell.

* Two vertical black arrows lead to signaling cascades: one through m i R-19a to c-Myc signaling, and another through exosomal M M P 1 to the P I 3 K forward slash A K T forward slash m T O R axis. Both cascades point toward the central N S C L C cell.

At the bottom, under the heading Dual-pathological drivers, are two oval nodes:

* S P P 1 on the left points upward to P I 3 K forward slash A K T forward slash m T O R leading to fibroblast expansion and cancer cell proliferation, which then points to the central N S C L C cell.

* P T G E S on the right points upward to A K T and the H I F-1 alpha pathway leading to myofibroblast differentiation, which also points to the central N S C L C cell.

Epithelial plasticity and the EMT signalling axis

Phenotypic evidence of this epithelial plasticity has been well-validated in the IPF-LC mouse model, where the significant loss of CK19⁺ and CK7⁺ epithelial cells confirms an active EMT programme that promotes tumour invasion and metastasis (Ref. Reference Barravecchia, Lee, Manassa, Magnuson, Ferris, Cavanaugh, Steele, Espinoza, Galban, Ramnath, Frankel, Pasca di Magliano and Galban21). This process is orchestrated by several distinct but converging signalling axes. At the core of this transition is the TGF-β/SMAD axis, acting as the master regulator in both diseases. It downregulates E-cadherin while upregulating mesenchymal markers such as vimentin and α-SMA, thereby conferring migratory capacities to alveolar epithelial cells in IPF and metastatic potential to cancer cells in NSCLC (Refs Reference Zhang, Van Der Zon, Ma, Mei, Cabukusta, Agaser, Madunić, Wuhrer, Zhang and Ten Dijke105–Reference Ding, Shen, Yu, Zheng, Liu, Wang, Liu, Zhao, Nian and Liu107). Working in tandem with TGF-β, the WNT/β–catenin pathway represents a critical convergent node. Specifically, differentiated embryo-chondrocyte expressed gene 1 (DEC1) facilitates fibrotic progression in IPF by modulating the integrated PI3K/AKT/GSK-3β/β-catenin cascade (Ref. Reference Hu, Zou, Ni, Zhang, Zheng, Liu and Cheng108). Furthermore, specific genetic mutations within epithelial cells link fibrotic remodelling to malignant transformation. For instance, the ATII marker surfactant protein C gene (SFTPC) serves as a dual driver: its mutation triggers epithelial dysfunction and inflammation underlying IPF, while its downregulation in NSCLC relieves SOX7-mediated inhibition of WNT signalling to fuel EMT (Refs Reference Zhang, An, Liu, Zhu, Pan, Gu, Zhao, Pu and Zhu109, Reference Nureki, Tomer, Venosa, Katzen, Russo, Jamil, Barrett, Nguyen, Kopp, Mulugeta and Beers138). Concurrently, the reactivation of developmental and mechanosensitive pathways further entrenches this plasticity. Aberrant Notch1 signalling impairs surfactant processing in IPF while driving tumour invasion via downstream effectors like SIX1 in LC (Refs Reference Huang, Lin, Wang, Gao, Yuan, Zhang, Chen and Chu110, Reference Wasnick, Korfei, Piskulak, Henneke, Wilhelm, Mahavadi, Dartsch, von der Beck, Koch, Shalashova, Weiss, Klymenko, Askevold, Fink, Witt, Hackstein, el Agha, Bellusci, Klepetko, Königshoff, Eickelberg, Schermuly, Braun, Seeger, Ruppert and Guenther111). Parallel to this, the PI3K/AKT/mTOR and MAPK/ERK1/2 pathways synergistically drive EMT, cytoskeletal remodelling, migration and invasion in NSCLC (Ref. Reference Zhou, Luo, Gan, Zhi, Zhou, Lv, Huang and Liang112). Furthermore, YAP/TAZ signalling acts as a mechanotransductive bridge, exacerbating ECM stiffening in IPF while maintaining the metastatic phenotype in LC (Refs Reference Ishikawa, Okubo, Matsushita, Tashiro, Warita and Naito113, Reference Zeyada, Eraky and El-Shishtawy114). These converging pathways illustrate that epithelial plasticity is not merely a phenotypic shift, but a shared pathological vulnerability driven by continuous microenvironmental stress.

Cellular senescence and the SASP microenvironment

ATII cell senescence, primarily triggered by p53 pathway activation, constitutes a cornerstone of IPF pathogenesis. Consequently, the resulting senescence-associated secretory phenotype (SASP) establishes a persistent profibrotic microenvironment (Ref. Reference Yao, Guan, Carraro, Parimon, Liu, Huang, Mulay, Soukiasian, David, Weigt, Belperio, Chen, Jiang, Noble and Stripp93). This senescent landscape is mechanistically orchestrated by three convergent clusters: mitochondrial dysfunction driven by the loss of regulators such as PTEN-induced putative kinase 1 (PINK1), adenine nucleotide translocase 1 (ANT1) and yes-associated protein 1 (YAP1) (Refs Reference Su, Guo, Wang, Ma, Zhang, Zhang, Geng, Jin, Guo, Yang, Niu, Ren, Wang, Ning, Li, He, Sun, Li, Li, Shan and Liang94–Reference Sui, Boatz, Shi, Hu, Li, Zhang, Königshoff and Kliment96); DNA damage-induced cell-cycle arrest, in which deficiencies in insulin-like growth factor binding protein 2 (IGFBP2), nuclear receptor subfamily 2 group F member 2 (NR2F2) and leucine-rich repeat kinase 2 (LRRK2) activate the p21 pathway by impairing autophagy (Refs Reference Chin, Ravichandran, Sanborn, Fleming, Wheatcroft, Kearney, Tokman, Walia, Smith, Flint, Mohanakumar, Bremner and Sureshbabu97–Reference Tian, Lv, Su, Wu, Li, Hu, Zhang and Wu99); and amplified inflammatory signalling, particularly the activation of the NF-κB axis following phosphatase and tensin homolog (PTEN) loss (Ref. Reference Tian, Li, Qiu, Dai, Zhang, Chen and Cai100).

However, during LC progression, cellular senescence and SASP play a dual, context-dependent role. In the early stages, epithelial cell senescence acts as a crucial barrier that suppresses tumour initiation (Ref. Reference Liu, Li, Wang, Liu, Feng, Wang, Huang, Huang, Wu, Cai, Cai, Xu, Cai and Li101). Yet, SASP derived from senescent bronchial epithelial cells paradoxically creates a microenvironment that enhances tumour cell adhesion, angiogenesis and migration, collectively supporting malignant progression (Ref. Reference Kuźnar-Kamińska, Mikuła-Pietrasik, Witucka, Romaniuk, Konieczna, Rubiś, Książek, Tykarski and Batura-Gabryel102). Thus, senescent epithelial cells bridge the gap between fibrogenesis and carcinogenesis, providing a unified molecular framework for the elevated LC risk observed in IPF lungs. This paradox reveals that while early senescence protects against tumorigenesis, chronic SASP eventually establishes a tumour-promoting microenvironment. Future studies must define the precise mechanisms driving this pro-oncogenic shift in fibrotic lungs.

Dysregulated angiogenesis and vascular remodelling

Regarding vascular dynamics, both diseases exhibit aberrant angiogenesis and compromised barrier integrity. For example, Engelbrecht et al. demonstrated that expanded COL15A1⁺ bronchial endothelial cells in both IPF and lung adenocarcinoma display impaired vascular barrier function and a pro-angiogenic phenotype (Ref. Reference Engelbrecht, Kooistra, Burg, Hariri, Nguyen, Brazee, Hla, Gochuico and Knipe120). In IPF, endothelial angiogenic capacity exhibits a strict stage-dependent divergence. In early, softer tissue environments, pro-angiogenic signals like VEGF promote neovascularization; conversely, in stiff fibrotic foci, angiogenesis is suppressed and vascular density declines, though PDGF and FGF may partially compensate by promoting pericyte-mediated vessel maturation (Ref. Reference Leonard-Duke, Bruce, Peirce and Taite115). In parallel, LC exhibits a sustained angiogenic profile, where tumour cells secrete VEGF and bFGF to activate endothelial cells and stimulate angiogenesis to support tumour growth and metastasis (Ref. Reference Du, Shi, Chen, Zhou and Che117). Despite these divergent trajectories, both diseases converge on compromised vascular integrity. In the TME, high ECM stiffness disrupts endothelial junctions, leading to abnormal vessel architecture and substantial leakage, thereby facilitating invasion and hematogenous dissemination (Ref. Reference Cappelli, Kanugula, Adapala, Amin, Sharma, Midha, Paruchuri and Thodeti118). Mechanistically, the loss of endothelial stabilizers like sphingosine-1-phosphate receptor 1 (S1PR1) or the Ang/Tie2 axis further disrupts junctions in both diseases, leading to vascular leakage that fuels inflammation and fibrosis (Refs Reference Hou, Jia, Cao, Zhang, Zhang, Wei, Xie, Dong and Wang121–Reference Wu, Zhang, Wang, Xu and Zheng123).

Endothelial metabolic reprogramming

Underpinning this dysfunction is a metabolic switch to glycolysis, primarily driven by the stabilization of HIF-1α. In both diseases, the loss of specific regulatory proteins precipitates this metabolic reprogramming. Specifically, reduced TRIM34 expression in IPF enhances endothelial glycolysis via stabilization of HIF-1α, thereby promoting endothelial proliferation, migration and aberrant angiogenesis (Ref. Reference Yang, Liu, Chen, Jing, Wang, Chen, Song and Fu116). Similarly, in LC, the downregulation of Foxp1 promotes glycolysis-dependent angiogenesis by relieving repression of the HIF1α–Hk2 metabolic axis (Ref. Reference Pi, Liu, Chang, Chen, Pan, Zhang, Zhuang, Liu, Wang, Tomlinson, Chan, Cheng, Yu, Zhang, Zhao, Liu, Liu and Zhang119). This shared metabolic reprogramming drives aberrant vascular architecture and local hypoxia, which synergistically accelerate disease progression.

EndMT as a convergent cellular transformation

EndMT represents a commonly validated mechanism in both IPF and LC. In pulmonary fibrosis, EndMT serves as a crucial source of myofibroblasts, driving collagen deposition; whereas in LC, it promotes tumour fibrosis and remodels local cytokine profiles. Recent work indicates that manipulating genetic regulators, such as enhancing endothelial heat shock protein family B (small) member 1 (HSPB1) function or inhibiting the JAK–STAT pathway, can effectively suppress EndMT (Ref. Reference Choi, Nam, Kim, Jang, Jin, Lee, Park, Ji, Cho and Lee124).

In summary, endothelial cells drive the progression of IPF and LC through shared pathological mechanisms involving dysregulated angiogenesis, mechanical and metabolic perturbations and EndMT. Endothelial cells have thus emerged from their traditional role as structural vascular components to become key pathological drivers in both diseases, highlighting their potential as shared therapeutic targets. A critical gap in current knowledge is how endothelial cells behave precisely at the spatial interface where avascular fibrotic tissue meets highly vascularized tumour tissue. Understanding this dynamic requires advanced spatial transcriptomics.