Considering that chronic epithelial injury and dysfunctional wound healing collectively trigger the fibrotic cascade, fibroblasts-one of the principal non-immune cell populations implicated in tissue repair-occupy a pivotal position in fILD pathogenesis (29). Based on single-cell RNA sequencing (scRNA-seq) analysis and DNA methylation profiling of fibroblast phenotypes, activated fibroblasts are increasingly recognized as heterogeneous cellular populations endowed with profibrotic properties (30, 31). The unrelenting activation of fibroblasts in fILD encompasses multiple pathological processes, including fibroblast-to-myofibroblast transformation (FMT), enhanced migratory capacity, resistance to apoptotic clearance mechanisms, and excessive deposition of ECM proteins that distort normal lung architecture (32, 33).

Numerous metabolites and soluble paracrine factors produced by macrophages are conventionally regarded as essential mediators orchestrating the biological transition between macrophage polarization states and fibroblast activation-collectively termed macrophage-fibroblast crosstalk-during the pathological progression of fILD (34). S100a4, also designated fibroblast-specific protein-1 (FSP-1), was initially considered a protein exclusively expressed by fibroblasts; however, contemporary experimental evidence indicates that it can potently induce fibrogenic phenotypes in mesenchymal progenitor cells in IPF (35–37). Recent findings suggest that S100a4 released from macrophages contributes substantially to fILD pathogenesis by promoting fibroblast differentiation and activation (35). Furthermore, the abundant expression of CX3CR1 in macrophage populations has prompted researchers to systematically evaluate the functional significance of the CX3CL1–CX3CR1 chemokine axis in fILD (38, 39). Although CX3CR1 deficiency may not quantitatively impact macrophage populations following bleomycin (BLM) administration, it nevertheless profoundly influences macrophage polarization and subsequent fibrotic processes, including myofibroblast activation and ECM remodeling (40).

Under certain pathological conditions, neither pro-inflammatory cytokine deficiency nor aggressive steroid and immunosuppressive therapeutic regimens can effectively limit fILD progression (41, 42). Based on accumulating research evidence, it has become imperative to consider that the establishment of an immunoregulatory microenvironment-predominantly comprising regulatory lymphocytes and myeloid-derived suppressor cells-may be fundamentally associated with pathological pulmonary fibrosis (43, 44). Within the type 2 immune response framework, interleukin-10 (IL-10) functions as an inducer of fibrogenic cellular phenotypes, particularly promoting M2-like and Th2-like pro-fibrotic cell differentiation (45, 46). Consistent with the actions of transforming growth factor-β1 (TGF-β1), the predominant profibrogenic cytokine, investigations have revealed that the anti-inflammatory cytokine IL-10 is abundantly produced by alveolar macrophages (46, 47). Additional studies have further demonstrated that IL-10 paradoxically suppresses lung fibrosis through TGF-β1-dependent mechanisms, wherein IL-10 overexpression simultaneously regulates macrophages polarization via the CCL2/CCR2 chemokine axis (48). In this context, accumulating evidence suggests that immunoregulatory cytokines-paradigmatically TGF-β1 and IL-10-serve as principal orchestrators of the pro-fibrotic microenvironment, triggering cascades of fibroproliferative wound healing processes intimately linked to inappropriate communication between M2-polarized macrophages and activated fibroblasts within pulmonary tissues (Figure 2) (49, 50). Phenotypic distinctions between macrophage subpopulations in normal versus fibrotic lungs substantiate the concept of specific innate immune factors playing causal roles in fILD pathogenesis (51, 52). To emphasize this critical point, co-localization analyses and causal modeling approaches have been systematically employed to identify macrophages highly expressing secreted phosphoprotein 1 (SPP1) and MER proto-oncogene tyrosine kinase (MERTK)—designated SPP1hi macrophages—and these identified macrophage subsets prove crucial in accelerating fibroblast activation within lung tissues (51, 53).

Considering that intercellular communication mediated through metabolite exchange has been extensively characterized in cancer cells and carcinoma-associated fibroblasts, it becomes necessary to provide mechanistic insights into the role of immunometabolic modulators in fILD as intercellular messengers governing macrophage-fibroblast interactions (51, 54). Glycolytic upregulation in lung fibroblasts, smooth muscle cells, and endothelial cells has been causally implicated in the pathological progression of pulmonary fibrosis (55). Subsequent metabolomic analyses have demonstrated significant elevation of glycolysis-derived lactate in conditioned media from TGF-β1-induced lung myofibroblasts and in bronchoalveolar lavage fluid (BALF) from murine fibrosis models (56–58). Macrophages cultured with fibroblast-conditioned media undergo phenotypic switching toward pro-fibrotic states characterized by elevated expression of select pro-fibrotic mediators, thereby establishing a positive feedback loop wherein myofibroblast glycolysis accelerates the pro-fibrotic activity of macrophages in fILD (59, 60). Fatty acid (FA) metabolism in cells regulates diverse biological activities and energy production through FA oxidation pathways (61). Altered FA metabolite concentrations and compositional profiles have been consistently identified in patients with fILD and in experimental animal models, suggesting the critical role of lipid metabolism in promoting pro-fibrotic phenotypes in both macrophages and fibroblasts/myofibroblasts (62, 63). For example, macrophage polarization toward the M2 phenotype depends not only upon metabolic energy supplied through FA oxidation but is also transcriptionally activated by peroxisome proliferator-activated receptor gamma (PPAR-γ), a transcription factor initially identified in adipose tissue for its regulatory role in FA storage and lipid homeostasis (64, 65). Additionally, genetic cell lineage tracing studies targeting lung fibroblasts in mice have revealed that both abundant pro-fibrotic cytokines secreted by polarized macrophages and PPAR-γ activation can induce phenotypic switching between lipid-laden fibroblasts and contractile myofibroblasts during both progression and resolution phases of lung fibrosis (66, 67).

Previous investigations have conclusively demonstrated that FAP is aberrantly expressed in lesional tissues of fILD, with particularly prominent expression within fibrotic interstitium and concentrated in fibroblast foci characteristic of IPF patients (68). FAP belongs to the prolyl peptidase enzyme family, sharing structural and functional homology with dipeptidyl peptidase IV (DPPIV), with which it exhibits approximately 70% amino acid sequence identity (69). Regarding its endopeptidase enzymatic activity, FAP paradoxically possesses antifibrotic proteolytic capacity, as it functionally complements matrix metalloproteinase 1 (MMP1) to proteolytically cleave fibrillar collagens type I and III. Moreover, a soluble circulating form of FAP has been identified and designated anti-plasmin cleaving enzyme (APCE) (70). APCE functions analogously to membrane-bound FAP and synergistically complements MMP1 in collagen degradation. Additionally, APCE proteolytically cleaves the N-terminal region of α2-antiplasmin, generating a modified substrate that enhances clot stability by reducing blood clot degradation and delaying fibrinolytic processes. Notably, human and murine FAP share approximately 89% amino acid sequence homology and exhibit comparable enzymatic activities, facilitating translational research between species (71).

The chronic bleomycin-induced fibrosis model possesses the distinct advantage of inducing progressive, irreversible lung fibrosis with histopathological lesions that more closely recapitulate those observed in IPF patients compared to acute injury models (72). Investigations employing both FAP-overexpressing and FAP knockout (FAPKO) transgenic mouse lines revealed exacerbated fILD severity, with FAPKO mice exhibiting enhanced pulmonary infiltration of immune cell populations (73). However, these studies did not observe statistically significant differences in fibrosis severity between transgenic and wild-type (WT) mice when fILD was induced through constitutive expression of active TGF-β (74). This suggests that bleomycin preferentially activates fibroblasts into FAP-positive cells exhibiting a proteolytic phenotypic profile, whereas TGF-β signaling predominantly mediates fibroblast-to-myofibroblast differentiation and subsequent tissue contractility (74). Conversely, Egger and colleagues observed pro-fibrotic functional properties attributable to FAP (12). They induced fibrosis in WT mice using the chronic bleomycin administration model and therapeutically treated animals with PT100, a pharmacological FAP inhibitor. PT100-treated mice demonstrated significantly improved survival and exhibited quantitatively reduced pulmonary fibrotic areas compared to vehicle-treated controls (12). It merits noting that PT100 functions as a pan-inhibitor targeting the entire dipeptidyl peptidase (DPP) enzyme family rather than exclusively inhibiting FAP. The contradictory experimental outcomes may therefore be mechanistically explained by this lack of pharmacological specificity, which contrasts with the genetic precision afforded by mouse models employing FAPKO mice or selective depletion strategies targeting FAP-positive cells (75).

FAP serves as a cell surface antigen marker denoting fibroblast activation state in cardiac muscle tissue and in pulmonary fibrotic lesions (75). It can be strategically exploited as a target molecule for engineering CAR T cells designed for IPF therapeutic intervention (Figure 3). A pioneering research team has successfully developed a CD5-targeted LNP-encapsulated FAP CAR mRNA delivery system (76). Given that CD5 is physiologically expressed on cell membranes of T lymphocytes and a minor subset of B lymphocytes, and considering it is dispensable for T lymphocyte effector functions, investigators conjugated CD5-specific antibody molecules to the LNP surface, thereby enabling the CD5-specific antibody-modified LNP-mRNA delivery platform to selectively target CD5-positive T lymphocytes in vivo (77). Further mechanistic investigations revealed spatial clustering of FAP CAR-positive T lymphocytes with FAP-positive activated fibroblasts within fibrotic lesions, accompanied by substantially reduced fibrosis severity in regions surrounding these cellular clusters, whereas fibrotic tissue architecture in spatially distant areas remained unaltered (78). This landmark study validated both the functional efficacy and T lymphocyte-targeting capability of the LNP-mRNA delivery system at the cellular and tissue levels, and for the first time experimentally demonstrated the feasibility of generating functional CAR T cells in vivo without ex vivo manipulation.

It warrants particular emphasis that CAR expression mediated by the CD5-targeted LNP-FAP CAR mRNA delivery system exhibits inherently transient kinetics, resulting in temporally limited CAR T cell persistence in vivo (77, 79). Should enhanced or sustained therapeutic efficacy be clinically required, repeated administrations can be safely performed to maintain therapeutic CAR T cell levels (77). For CAR T cell therapeutic applications in non-malignant diseases such as IPF, autoimmune disorders, and infectious diseases, transient CAR T cell expression profiles prove advantageous by substantially mitigating risks of long-term toxicities and enabling superior therapeutic controllability (80, 81). During clinical implementation, this transient expression strategy may afford greater dose titration flexibility and improved safety-efficacy balance (82). Consequently, transiently expressed CAR T cell platforms possess considerable developmental potential for future clinical translation in non-oncological indications.

Current preclinical animal models inadequately recapitulate the pathological complexity and pronounced heterogeneity characteristic of human fILD (88). The pulmonary microenvironment in fILD comprises extraordinarily diverse cellular populations-including phenotypically distinct fibroblast subsets, myoFbs, microvascular and lymphatic endothelial cells, and heterogeneous immune cell infiltrates-alongside biochemically varied ECM components exhibiting spatial and temporal heterogeneity (89–91). Substantial inter-patient and intra-patient heterogeneity in the pulmonary fibrotic microenvironment across fILD subtypes and progressive disease stages considerably complicates the rational design of broadly effective targeted immunotherapies (92). Unlike oncology applications, where abundant tumor-specific or tumor-associated antigens facilitate highly selective targeted treatments, fILD pathogenesis exhibits a relative paucity of validated disease-specific antigenic targets, thereby substantially restricting the therapeutic applicability of engineered immune cell platforms (including CAR T cells, CAR-equipped macrophages, or engineered natural killer cells) for clinical translation in fibrotic diseases (93–95). Additionally, immune cell-based therapeutic interventions carry inherent risks of precipitating immune-related adverse events, including cytokine release syndrome (CRS), immune effector cell-associated neurotoxicity syndrome (ICANS), and graft-versus-host disease (GVHD), with risks particularly elevated when employing allogeneic cell products (96). Consequently, stringent safety considerations assume paramount importance in fILD therapeutic applications to avoid inadvertently exacerbating pre-existing organ damage and to rigorously minimize off-target effects on healthy tissues.

Autologous CAR T cell manufacturing represents an emerging yet prohibitively expensive biotechnology platform (97). The substantial financial expenditures associated with cell product engineering, large-scale manufacturing under current Good Manufacturing Practice (cGMP) conditions, quality control testing, and clinical administration frequently exceed several hundred thousand US dollars per patient, presenting formidable economic barriers even for potentially curative therapeutic interventions (97). These extraordinarily high costs correlate directly with severely limited patient accessibility, effectively restricting CAR T cell therapy availability to a select number of specialized academic medical centers and comprehensive cancer centers possessing the requisite infrastructure and expertise (98). This personalized “living drug” manufacturing paradigm necessitates highly specialized cell processing facilities, sophisticated quality assurance systems, and multidisciplinary clinical expertise, thereby confining practical availability to a geographically limited number of major tertiary care centers (98). Consequently, substantial patient populations who might represent ideal candidates for CAR T cell therapy based on disease characteristics currently cannot access this potentially transformative treatment modality due to geographic, financial, or healthcare system-related barriers.

CRS constitutes the most clinically significant and frequently encountered toxicity syndrome associated with CAR T cell therapy (99). Before clinical application in novel disease contexts beyond oncology, rigorous preclinical and early-phase clinical evaluation of CRS-like symptomatology is essential (100, 101). Early-phase CAR T cell clinical trials conducted in systemic lupus erythematosus (SLE) patients reveal that comparatively lower target cell burdens characteristic of non-malignant autoimmune indications may potentially result in attenuated CRS severity compared to oncology applications; however, larger randomized controlled trials with extended follow-up are required for definitive confirmation of this hypothesis (102). Preclinical CAR T cell studies specifically targeting activated cardiac fibroblasts in heart failure animal models demonstrated the absence of serum cytokine elevations, no alterations in body weight parameters, and no impairment of physiological wound healing processes in treated murine subjects (103). Nevertheless, in human patients, therapeutic reduction of substantial established fibrotic tissue burdens could conceivably provoke clinically significant CRS-like inflammatory responses, warranting careful monitoring and potential intervention.

Therapeutic depletion of healthy cell subsets-exemplified by B cell ablation in SLE patients treated with anti-CD19 CAR T cells-may paradoxically confer clinical benefit through elimination of autoreactive B cell clones; however, long-term immunological consequences and potential infectious complications remain incompletely characterized (102). Preliminary clinical data emerging from ongoing trials suggest these cellular depletions may prove transient with spontaneous immune reconstitution; however, long-term follow-up studies spanning multiple years will be necessary to definitively clarify potential chronic complications, including hypogammaglobulinemia and opportunistic infections (81). Although pathological chronic fibrosis unequivocally drives organ dysfunction, it remains uncertain whether specific physiological contexts exist wherein fibroblast activation serves beneficial homeostatic functions that should be preserved (103). Senescent cells, while contributing to age-related pathologies, nevertheless perform essential physiological roles during embryonic development, acute wound healing responses, and tissue regeneration processes (104). Therefore, comprehensive evaluation of potential impacts of senolytic CAR T cell therapies on these beneficial physiological processes constitutes a critical research priority, emphasizing the importance of selectively targeting pathological senescent cell populations through identification and validation of disease-specific antigens (104). Transient CAR T cell expression platforms prove particularly attractive in non-oncological contexts, as inherently limited CAR T cell persistence substantially alleviates concerns regarding potential long-term safety risks associated with persistent CAR expression and chronic immune activation (105).

The rapidly expanding field of CAR T cell therapy beyond oncology encompasses numerous fundamental unanswered questions demanding systematic investigation in the forthcoming years (106). Determination of the optimal therapeutic index remains a pivotal research focus, given the criticality of appropriate CAR T cell infusion dosing strategies to achieve an optimal balance between therapeutic efficacy and acceptable safety profiles, particularly for mRNA-based transient CAR T cell platforms requiring repeated administrations (107). In both acute injury-driven and fILD, the clinical application paradigm for transient CAR T cell immunotherapy necessitates careful evaluation through well-designed dose-finding and pharmacodynamic studies. Optimal therapeutic timing relative to disease stage and progression kinetics remains uncertain and likely varies substantially according to underlying disease pathogenesis and individual patient factors (108). It proves imperative to definitively establish whether irreversible advanced disease stages exist beyond which therapeutic intervention becomes ineffective or potentially harmful due to loss of regenerative capacity (108). For tissues and organs retaining regenerative competence, clarifying whether selective clearance of pathological cell populations can successfully restore physiologically normal tissue function and architecture represents an essential translational question (109). Moreover, systematically understanding CAR T cell trafficking dynamics, tissue infiltration patterns, and persistence within non-lymphoid, non-tumor tissues warrants comprehensive investigation. Local or locoregional CAR T cell delivery strategies-including intratracheal, endobronchial, or image-guided direct tissue injection approaches-may substantially enhance therapeutic efficacy while simultaneously minimizing systemic exposure and associated toxicities (110). As this promising therapeutic field continues evolving, elucidating intrinsic cellular and extrinsic microenvironmental modulators of CAR T cell therapeutic activity will prove critical for rational optimization (Figure 4).