Pathogenesis of SSc-ILD

Abstract

Interstitial lung disease (ILD) is the most common internal organ manifestation of systemic sclerosis (SSc) and remains the leading cause of disease-related mortality. The disease course is variable among patients, for although ILD develops in the majority of patients, only a subset progress to severe fibrotic lung disease, reflecting considerable biological heterogeneity. SSc-ILD evolves through distinct pathogenic phases, with early immune-mediated epithelial and endothelial injury giving way to autonomous fibroblast activation, extracellular matrix remodelling and self-sustaining fibrotic pathways. Insights from genetic and epigenetic studies, single-cell technologies, experimental models and, more recently, biosamples collected during clinical trials have substantially advanced our understanding of disease mechanisms and highlighted differences between early inflammatory and established fibrotic disease. This review aims to summarise the current understanding of the cellular and molecular mechanisms underlying SSc-ILD and discuss how these advances are shaping precision medicine approaches.

Introduction

Interstitial lung disease (ILD) is a major internal organ complication of systemic sclerosis (SSc), and a leading cause of disease-related mortality and morbidity.^1,2 Subclinical interstitial abnormalities are common (up to 80%); however, the trajectory of SSc-ILD is very heterogeneous.³ There are differences in prevalence based on disease subgroup, with 25% of limited cutaneous SSc (lcSSc) developing ILD and 50% of diffuse cutaneous SSc (dcSSc). The factors determining if the disease remains stable, or patients have a more progressive form, with substantial loss of lung function, are incompletely understood, but reflect interactions between genetic susceptibility, immune dysregulation, tissue injury and aberrant repair.²

SSc-ILD evolves through distinct pathogenic phases, where inflammation and fibrosis have differing impacts at different stages of the disease. Early disease is dominated by immune-mediated epithelial and endothelial injury which promotes fibroblast activation. More established disease is driven by autonomous fibroblast activation, altered tissue mechanics and feed-forward profibrotic signalling. This review examines the genetics, experimental models, and therapeutic trials which have helped refine our understanding of the heterogeneity of SSc-ILD.

Clinical risk factors and epidemiology

It is appreciated that the course of SSc-ILD is not linear, and varies over time and not all patients have an equal risk. Some patients develop ILD which remains limited with minimal or no progression over time, other patients experience a more aggressive progression throughout their disease course. However, recent large cohort studies have shown that ILD progression is not linear, and can go through phases of stability and late-stage progression.⁴

Clinical heterogeneity in SSc-ILD is apparent both in susceptibility to lung involvement and in subsequent disease trajectory. Anti-topoisomerase (ATA) autoantibody positivity remains the most consistently identified risk factor for developing SSc-ILD particularly within the first 3 years of disease onset, and this is regardless of cutaneous disease subtypes.⁵ However, later stage disease has a higher prevalence in lcSSc compared with dcSSc.

One of the largest observational studies has come from EUSTAR (European Scleroderma Trials and Research), which included patients within the international EUSTAR database. High risk factors for new-onset ILD included high inflammatory markers, New York Heart Association stage ≥2, muscle weakness, high inflammatory markers and certain SSc specific autoantibodies, notably ATA and PM-Scl. Those with anti-centromere antibody (ACA) positivity and arthritis had a lower risk of developing SSc-ILD compared with the mean. Interestingly, they did not find disease duration to be associated with risk of ILD.⁶

Gender differences have also been noted. Male patients with ILD tend to be younger, with shorter disease duration and frequently dcSSc,⁷ with a higher mortality compared with female SSc-ILD patients.

Genetic studies

Genetic factors have been shown to contribute to the susceptibility to developing SSc-ILD, with some associations appearing distinct from those that confer susceptibility to developing SSc itself. Association studies including genome-wide association study meta-analyses have identified several single-nucleotide polymorphisms (SNPs) linked with the development of SSc-ILD.

Some of the strongest genetic associations are linked to specific antinuclear autoantibody (ANA) subgroups, which reflects the close relationship between ANA and class II major histocompatibility complex haplotypes. Particularly, ATA, which has the highest associated risk of ILD, is strongly associated with HLA-DRB1, HLA-DRB5, HLA-DQB1 and HLA-DPB1 haplotypes. In contrast, differing HLA haplotypes are associated with a lower risk of ILD, and alternative antibody profiles, such as ACA with HLA-DQA1.^2,8,9

There are a number of additional candidate genes that have been identified as associated with SSc-ILD, although the majority are derived from small studies. Most of the genes are associated with immune responses. STAT4 polymorphisms have been shown to be protective against the development of SSc-ILD,¹⁰ whereas associations between IRF5 and ILD, including its interaction with STAT4 have been less consistent.^11,12 Polymorphisms in NLRP1, an important component required for the assembly of the inflammasome, are associated with ATA positivity and ILD, while other implicated immune-related genes include IL1A and IRAK1.^13–15

Genetic associations outside of immune regulatory pathways have also been described and impact the development of SSc-ILD. A well-phenotyped single-centre study identified a functional polymorphism in the CTGF promoter gene, which was strongly associated with SSc-ILD.⁸ However, this link has been inconsistently replicated,^9,16,17 therefore suggesting multiple interacting genetic and non-genetic factors are also essential for SSc-ILD development.

Epigenetic mechanisms provide an additional link between genetic susceptibility and SSc-ILD. Altered DNA methylation, histone modifications and dysregulated non-coding RNAs, particularly microRNAs (miRNAs) (eg, miR-21, miR-29 and miR-155), alter the expression of genes involved in immune regulation and fibrosis. These influence pathways including transforming growth factor-β (TGFβ) signalling and therefore myofibroblast differentiation and progressive ILD.¹⁸ Given these are also potentially reversible, epigenetic changes are also being explored as biomarkers of disease progression and potential therapeutic targets in SSc-ILD.¹⁹

Taken together, these findings suggest that susceptibility to SSc-ILD is unlikely to be determined by a single genetic programme. Rather, genetic and epigenetic factors likely contribute to distinct immunological and fibrotic endotypes. The particularly strong relationship between HLA genotype, ATA positivity and ILD are important links between inherited susceptibility and subsequent immune phenotype.

Cellular pathogenesis

Cellular pathogenesis reflects overlapping processes in genetically susceptible individuals, where injury to the epithelial and endothelial cells initiates an innate immune response, which evolves into the activation of the adaptive immune system and fibroblast activation. Once established, extracellular matrix (ECM) accumulation and altered tissue mechanics generate self-sustaining feed-forward loops that perpetuate fibrosis independently of the initiating inflammatory insult. This can be viewed clinically on high-resolution CT as a progression from early inflammatory/fibrotic changes, to more severe architectural distortion with honeycombing (figure 1).

Figure 1. CT scan images of progressive changes in lung fibrosis over time. Early inflammatory changes characterised with ground-glass changes, whereas the later stages have more fibrotic architectural distortion with honeycombing and traction bronchiectasis.

Initial alveolar epithelial cell/endothelial injury

Damage to the alveolar epithelial cells (AECs) and the endothelial cells is increasingly appreciated as one of the earliest pathogenic events in SSc-ILD, preceding inflammatory cell infiltration. Disruption to the alveolar-capillary barrier releases damage-associated molecular patterns (DAMPS), recruits inflammatory cells and activates tissue repair signalling.²⁰ In health, it is type II alveolar cells which restore epithelial integrity. However, in SSc, this repair process is disrupted. The failure of epithelial repair, creates a local profibrotic environment resulting in fibroblast recruitment and proliferation, myofibroblast differentiation and matrix deposition tissue.²¹ This is further driven by type II pneumocytes undergoing senescence, losing their regenerative capability and secreting profibrotic mediators.²²

Innate immune response: monocytes and macrophages

Inflammation and immune activation are also a key early pathogenic process occurring in SSc ILD. There is notable infiltration of monocytes and macrophages into the lung parenchyma and the alveolar space,²³ particularly the macrophages which are abnormal polarised towards a pro-fibrotic phenotype²⁴ (alternatively activated macrophages or M2) more than towards proinflammatory macrophage phenotype. However, in the presence of CXCL4 (a platelet-activating chemokine), which is elevated in SSc bronchoalveolar lavage (BAL), the macrophages exhibit impaired efferocytosis (the defective removal of apoptotic cells).²⁵ Increased MAP kinase (MAPK) signalling pathway molecules²⁶ also contribute to the inflammatory state and maladaptive repair. However, it has also been shown that pulmonary macrophages increase the expression of their Fcγ receptors, which enhances their phagocytic and inflammatory responses, contributing to tissue damage.

Adaptive immune cells: T cells and B cells

Within the adaptive immune system, the changes in the innate immune system result in a notable increase in natural killer (NK) cells and a shift from resting to effector tissue resident CD8+T cells, a cell type also implicated in other autoimmune diseases. There is an infiltration of CD4+T cells, which differentiate into Th2 phenotypes within the lung tissue. They produce key cytokines including IL4, IL-13, IL-17, endothelin 1, PDGF (platelet-derived growth factor), oncostatin M, as well as activating TGFβ.^24,27

Although sampling BAL has proven highly insightful in understanding sequential pathology, the results of cellular composition have not been shown to be useful prognostically, and therefore the test is not performed routinely.

B cells

B cells also play a key role in SSc ILD, particularly in patients with ATA positivity, where increased CD19 levels are noted. The imbalance seen with between elevated IL-6 producing effector B cells and the IL10 producing regulatory B cells is strongly associated with extensive ILD and severe skin disease.²⁸ The ATA producing CD27+B cells with a high affinity for topoisomerase I have been shown to produce increased pro-inflammatory cytokines including IL6 and IL23. However, the B cells with a low affinity to ATA tend to have a higher frequency of anti-inflammatory cytokine production.²⁹ There are also abnormalities, particularly in the transitional B cells which show dysregulation in SSc, overproduction of IL6,^30,31 which is more apparent in the ATA positive patients. This supports the clinical finding that patients with ATA positivity have a much higher risk of developing SSc-ILD and implicate the B cells in driving this link.

There is also a link with topoisomerase-1 specific CD4 T cells restricted to ATA positive patients, which has a Th17 proinflammatory phenotype and is associated with SSc-ILD progression, even if found in small numbers.

Fibroblast activation

Fibroblast activation is the pivotal event that links inflammation to irreversible tissue fibrosis in SSc-ILD. In response to persistent epithelial injury and the release of profibrotic mediators, fibroblasts initially acquire an activated phenotype characterised by increased proliferation, migration and ECM production, and subsequently they differentiate into myofibroblasts, with high levels of expression of αsmooth muscle actin (ACTA2/αSMA). The excessive production of collagen distorts the lung architecture further.^32–35

Activated fibroblasts can arise from multiple sources, including primarily the expansion of the interstitial fibroblast population and the resident lung progenitor mesenchymal cells,³² as well as originating from transdifferentiation of epithelial cells, endothelial cells, pericytes and recruited from circulating cells such as fibrocytes and monocytes.³⁶

Single-cell sequencing studies have demonstrated heterogeneity within the fibroblast subpopulations in SSc-ILD. Fibroblasts, with unique gene expression compared with healthy control (HC) fibroblasts, have been identified, including those with high levels of SPINT2, MFAP5 and WIF1.^32,33 Compared with healthy lung fibroblasts, the SSc lung fibroblasts show increased expression of ECM-associated genes including COL1A1 and COL3A1, as well as activation of profibrotic pathway genes such as TGFβ2, IGF2, IGFBP3, IGFBP5 and WNT5A, while expression of antifibrotic and matrix-remodelling genes such as MMP1, MMP9, CTSL, SFRP1 and IL33 is reduced, especially after TGFβ stimulation.^37,38 Therefore this transcriptional shift favours matrix accumulation and impaired tissue remodelling.

TGFβ is the dominant driver of fibroblast activation and promotes the differentiation into myofibroblasts, via the canonical SMAD2/3 signalling and interactions with Wnt/β-catenin pathway.³⁸ As fibrosis progresses, fibroblast activation becomes self-perpetuating. The activation of mechanotransduction pathways including yes-associated protein (YAP) and myocardin-related transcription factor-A (MRTF-A), by increased matrix stiffness further stimulates TGFβ signalling, collagen synthesis and myofibroblast persistence. This therefore creates a feed forward cycle where tissue stiffness and fibroblast activation reinforce each other, even without ongoing exposure to the initial trigger.³⁹

Molecular pathways

The process of pulmonary fibrosis in SSc involves several interacting signalling pathways and molecular mediators, driven by inflammatory cells, endothelial cells and fibroblasts. They can broadly be grouped into master profibrotic pathways, immune amplifiers, chemokine networks and innate danger-sensing pathways. The signalling cascades are closely interconnected, but nearly all converge on fibroblast activation.

Master profibrotic pathways

Transforming growth factor beta

Cytokines and growth factors play a vital role in SSc-ILD. TGFβ is one of the major regulators of connective tissue growth and repair, and a key driver of fibroblast activation and differentiation into myofibroblasts. It is a potent inducer of the ECM, promoting the synthesis, secretion, processing and cross-linking of fibrillar collagens (particularly I, III and V).⁴⁰ TGFβ is sequestered within the ECM and activated through several mechanisms including angiotensin II, Toll-like receptor signalling and proinflammatory cytokines such as IL1β, TNFα and IL6. It promotes the secretion of key matrix components such as fibronectin and thrombospondin (THBS) and induces fibroblast migration, proliferation and differentiation.

SSc lung fibroblasts express higher levels of TGFβ1 and TGFβ2 compared with HC.⁴¹ TGFβ is also expressed at high levels by the alternatively activated macrophages in SSc lung tissue, together with MBD2 (methyl-CpG-binding domain 2) protein, which suppresses the expression of an upstream inhibitor of TGFβ signalling.⁴²

TGFβ further promotes ECM accumulation by suppressing matrix degradation. This occurs partly through induction of protease inhibitors such as plasminogen activator inhibitor-1 (PAI-1) and tissue metalloprotease inhibitors (TIMPs), therefore favouring the persistence of the fibrotic matrix.

Platelet-derived growth factor

PDGF is one of the important factors required to maintain the profibrotic environment and promote ECM deposition. It is produced by the alveolar macrophages, endothelial cells and fibroblasts. Elevated levels of PDGF receptor subunits, a chemoattractant for mesenchymal cells, are found in patients with SSc in the skin and BAL,⁴³ and correlate with lung disease.⁴⁴

There is unique crosstalk between PDGF and TGFβ in SSc lung fibroblasts not seen in HC fibroblasts.⁴⁵ TGFβ increases PDGFRA expression in SSc fibroblasts, increasing their sensitivity to PDGF and enhancing the fibrogenic response. Key pathways activated include the PI3K pathway and the Ras-MAPK and, via the melanin-concentrating hormone receptor 1 (MCHR1), induces TGFB1 and CTGF expression in fibroblasts.

Nintedanib, an anti-fibrotic approved for use in SSc-ILD, works through blocking the ATP-binding pocket of PDGFR, thus blocking the PDGF-induced differentiation of lung fibroblasts into myofibroblasts, reduces proliferation and migration, and suppresses the production of collagen and fibronectin.

Immune amplifiers

Interleukin-6

IL6 has important roles in both SSc skin fibrosis and ILD. Elevated IL6 concentrations are predictive of forced vital capacity (FVC) and DLCO (Diffusing Capacity of the lungs for Carbon Monoxide) decline within the first year of disease onset and of death within the first 30 months since diagnosis.⁴⁶

IL6 acts through three main mechanisms. IL6 has a direct role in fibrosis by promoting the activation of fibroblasts via the STAT3/JAK2 (Janus kinase) signalling pathway, resulting in increased production of collagen I and other components of the ECM, as well as promoting the interaction of fibroblast and macrophages.⁴⁷ In conjunction with TGFβ, IL6 promotes the differentiation of CD4+T cells to Th17 cells, increasing the production of IL17, which has a potent proinflammatory effect. Finally IL6 stimulates the maturation and activation of B cells, increasing the production of autoantibodies.

Tumour necrosis factor α

Serum levels of TNFα correlate with the presence of pulmonary fibrosis and reduced FVC in patients with SSc-ILD.⁴⁸ It forms an interactive cytokine network which amplifies inflammation and regulates fibroblast activation along with IL17, IL6 and TGFβ. It also modulates IL17 production and signalling, resulting in potentiation of the inflammatory response.⁴⁹

IL1β promotes the expression of metalloproteinases which are involved in the remodelling of the ECM. It also promotes the production of IL6, IL17 and TNFα. However, conversely, IL1β has direct antifibrotic effects on the fibroblasts themselves, highlighting the complex interaction of these molecules.^50,51

IL17, plays an important proinflammatory and pro-fibrotic role in SSc-ILD. It is produced by Th17 cells, known to be elevated in SSc, and correlates with extent of ILD.⁵² IL-17 promotes the activation and proliferation of lung fibroblasts, leading to increased collagen and ECM production, as well as stimulating the release of profibrotic factors such as TGFβ and CTGF.⁵³

IL-11 has been shown to drive activation of lung fibroblasts both in mouse models and SSc derived lung fibroblasts. It is significantly upregulated in SSc lung fibroblasts. IL31 is also involved in SSc ILD, with overexpression of the receptor IL31RA in SSc derived lung fibroblasts, and increase in serum IL31 in patients with SSc compared with HC.

Chemokines

The epithelial cells produce key cytokines including fibroblast growth factor, vascular endothelial growth factor and chemokines such as CCL (chemokine ligand)-2, CCL7 and CCL18.^52,54–57 Although these are not disease specific, they likely represent mediators of tissue repair and responses to cellular injury and have been explored as potential future therapeutic targets.

CCL2 has consistently shown a pivotal role in both SSc skin fibrosis and severity of ILD.^52,54 CCL2 is produced by fibroblasts, endothelial cells and macrophages, and is involved in leucocyte trafficking and activation of monocytes and T cells, as well as fibroblast differentiation to myofibroblasts.⁵⁸ Higher serum concentrations of CCL2 are predictive of a faster decline in FVC in early SSc. Fibroblasts expressing CCL2 receptor have a strong profibrotic phenotype, and produce increased amounts of αSMA, CCL2 and CTGF, however selective inhibition of this receptor only attenuates the overexpression of αSMA, and not CTGF in skin.⁵⁹

CXCL4

CXCL4 (also called platelet factor 4) is another chemokine known to play a vital role in wound repair. It stimulates the expression of IL4 and IL13, and acts as a chemotactic factor for neutrophils, fibroblasts and monocytes.⁵³ At high levels, it impairs the macrophage’s ability to remove apoptotic cells. CXCL4 is associated with ILD, and predictive of disease progression,⁵³ with elevated levels being associated with a more rapid decline in DLCO, and low levels predicting improvement in lung function tests.⁵³

CXCL11 is an interferon induced chemokine, where both serum and BAL levels show strong association with ILD, specifically BAL predicts subsequent decline in SSc patients, even without established ILD.^60,61

Danger sensing

Pattern recognition receptors play a vital role in SSc pathogenesis, through stimulating interferon (IFN)−1 and other cytokines, and activating fibroblasts. TLR4 (toll like receptor) is overexpressed on SSc fibroblasts, and its inhibition reduces tissue fibrosis. Other TLRs have also been implicated, and they interact with circulating nucleic acids, and complexes including CXCL4, to enhance IFN-1 production. Mitochondrial DNA also activates immune pathways through TLR9 and cytosolic GAS (cGAS)/STING pathways in SSc, impacting disease progression.⁶²

DAMPs play a crucial role in initiating immune response after cellular injury. They activate the innate immune cells, as well as epithelial cells and fibroblasts, through an interaction with TLR4.⁶³ Elevated serum S100A4 levels are associated with SSc ILD.⁶⁴

Biomarkers of disease activity

Circulating pneumoproteins provide evidence of ongoing epithelial injury in SSc-ILD. SP-D (surfactant protein), KL-6 (Krebs von den Lungen-6) are predominantly derived from AECs, and correlate with the presence and extent of ILD,⁶⁵ whereas macrophage-derived CCL18 links innate immune activation with fibroblast collagen production and is a predictive marker of disease progression.^43,66–68

KL6 is one of the more reliable markers for SSc-ILD,^66,69 and correlates inversely with percentage predicted DLCO, DLCO decline, and positively with disease extent on HRCT.⁷⁰ It has been proposed as a prognostic marker predicting end-stage disease or death at higher concentrations.⁷¹ Through meta-analysis, KL-6 is superior to SP-D in differentiating ILD from non-ILD,⁷² however, SP-D is more specific,⁷³ with decline in SP-D levels correlating with treatment response.⁷⁴ Elhai et al showed that SP-D in combination with ATA status improved accuracy of ILD detection, but did not correlate with progression.⁶⁶

Higher levels of both KL-6 and CCL18 are both predictive of a decline in FVC% over 12 months, independent of autoantibody positivity and immunosuppression.^67,71,75

Preclinical models

Preclinical models provide an opportunity to interrogate individual components of SSc-ILD pathogenesis, although no mouse model recapitulates the full complexity of SSc. Instead, different models offer complementary insight into initiation, susceptibility and persistence of pulmonary fibrosis.

The bleomycin mouse model has provided important insight into the initiation and resolution of injury induced pulmonary fibrosis. Initially acute inflammation develops in the first 5–7 days after bleomycin exposure, followed by fibrosis from days 10 to 14.⁷⁶ In wild-type mice, the lung fibrosis returns to normal lung architecture after 60 days. However in genetically modified mice that develop mild ILD, the response to bleomycin administration is often more severe, highlighting the importance of host factors in determining the response to pulmonary injury.

Genetically altered mouse models further demonstrated that susceptibility to fibrosis is strongly dependent on the underlying host phenotype. Type 1 and type 2 tight skin (Tsk) mouse strains are characterised by germline mutations in fibrillin-1 and collagen 3a1 respectively, develop prominent skin fibrosis, but do not develop ILD. This divergence between cutaneous and pulmonary fibrosis highlights the organ-specific nature of fibrotic response.

The Fra-2 (Fos-related antigen 2) transgenic strain develops both fibrotic and vascular manifestations of SSc, including pulmonary fibrosis. It emphasises the importance of fibroblast activation by TGFβ signalling in driving fibrosis.^77,78 Conversely, mice with postnatal deletion of type 2 TGFβ receptor do not develop lung fibrosis after exposure to bleomycin, thus supporting the importance of resident lung fibroblasts in coordinating the fibrotic response.⁷⁹

The TβRII∆k-fib model, where TGFβ signalling is altered specifically in fibroblasts, has been informative in understanding the impact of failure of fibrosis to resolve. These mice express a non-signalling type 2 TGFβ receptor on fibroblasts, resulting in an imbalance of TGFβ receptors and ligand-dependent TGFβ pathway activation. Following minor epithelial injury with acidic unbuffered saline, about 25% of mice develop ILD at 6 weeks.^79,80 This is proposed as an experimental model to chronic gastro-oesophageal reflux, and therefore represents potential mechanisms for initiation or amplification of fibrosis in SSc patients who are susceptible. Importantly this mouse model also fails to resolve fibrosis following bleomycin exposure, thus reflecting reduction in epithelial cell regeneration and proliferation after injury, and the persistence of myofibroblasts in the lung resulting in chronic fibrosis.

Taken together, these models suggest that the critical abnormality in SSc-ILD may not be pulmonary injury itself, but the response and resolution to injury. In a susceptible host, relatively minor epithelial injury may initiate a fibroblast activation loop which fails to terminate, providing evidence for the feed-forward model of persistent fibrosis seen in humans.

Insights from experimental medicine studies and clinical trials

Intervention studies offer an opportunity to interrogate the relative importance of specific pathways at different stages of SSc-ILD. Our knowledge, however, is limited by the recruitment criteria, which tends to be early stage progressive SSc-ILD.

The FaSScinate and FocuSSced clinical trials were phase 2 and phase 3, respectively, randomised placebo-controlled studies, comparing tocilizumab, an anti-IL6 against placebo. It highlighted the importance of IL-6 in the development of SSc-ILD. Quantitative CT analysis confirmed improvement in all parameters of ILD on tocilizumab, whereas there was a decline in the placebo group.⁴⁶ Mild cases recruited to the trial showed near-complete attenuation of their lung disease. Of note, analysis of isolated skin fibroblasts, confirmed almost complete reversal of the profibrotic phenotype after tocilizumab; however, the skin improvement did not reach significance, suggesting this subgroup of fibroblasts may better reflect those driving the early stages of SSc-ILD, but do not seem as relevant in the later more established lung disease cohort.⁴⁷ Post hoc analysis revealed that it was the ATA+ patients who really drove the group effect in stabilisation and improvement in SSc-ILD, whereas the other autoantibody subgroups showed a less marked effect from tocilizumab,⁸¹ suggesting that IL6 is a driver in early SSc in ATA+patients more than other subgroups.

The SENSCIS clinical trial assessed the effect of nintedanib, an antifibrotic as add-on therapy with or without MMF for patients with progressive SSc-ILD. It demonstrated the additive benefit of combined treatment of immunosuppression with antifibrotic treatment.^82,83 It also confirmed the link between extent of ILD on CT and risk of progression.⁸² These findings are consistent with persistent fibroblast-directed pathways contributing to progression of ILD in established disease, alongside the effects of immunomodulation. Decoupling these drivers at differing stages and extents is important, as it is the late stage progressors who have the greatest mortality, either from the ILD itself or group 3 pulmonary hypertension.

Nerandomilast is a preferential phosphodiesterase 4B (PDE4B) inhibitor, just completed phase 3 clinical trials in progressive fibrosing ILD, where it significantly slowed progression of fibrotic ILD in a number of autoimmune diseases including SSc.⁸⁴ Preclinical studies demonstrated anti-inflammatory and anti-fibrotic properties in both in vitro and in vivo models. Inhibiting PDE4, results in increased cAMP levels, as well as reducing the infiltration of CD3+T lymphocytes and F4/80+macrophages in fibrotic lungs within the bleomycin mouse model of SSc-ILD. It has also been shown to suppress the gene expression of key cytokines including TNFα, IL6, IL13, TGFβ1, IL1β, IL17A and IL13 in these models.⁸⁵ Therefore, this suggests that nerandomilast works by modulating the macrophages and T lymphocytes and suppressing the gene expression of key inflammatory cytokines involved in the early stages of SSc-ILD pathogenesis.

Phases of pathogenesis of SSc-ILD

Early phase: susceptibility and triggering

Pathogenesis of SSc-ILD is likely to be multifactorial, involving a susceptible individual and an environmental trigger (figure 2). The susceptibility as discussed is likely genetic or epigenetic. While within the environment, these initiate a cascade in susceptible tissue to fibrotic scarring in response to exposure to, for example, chemical exposure, recurrent microaspirations or infection.

Figure 2. Schematic of pathogenesis of SSc-ILD. Three phases are highlighted with the early phase being immune dominant, intermediate phase with inflammatory cells activating the fibroblasts, and the third late phase involving a feed-forward loop of fibrosis. CXCL4, platelet factor 4; DAMPS, damage-associated molecular patterns; ECM, extracellular matrix; IL, interleukin; MAPK, MAP kinase; MRTF, myocardin-related transcription factor-A; NK, natural killer; PDGF, platelet derived growth factor; TAZ, transcriptional co-activator with PDZ-binding motif; TGFb, transforming growth factor; YAP, yes-associated protein.

It seems likely that similar processes in the early stages of lung disease are also occurring in the skin, with macrophage recruitment, endothelial activation, and T and B cell infiltrates. High resolution CTs confirm early inflammation in the lungs with amorphous ground-glass change, although this may represent fine fibrosis rather than pure inflammation in the early stages of the disease.

Cohort studies consistently highlight the early stages of disease, particularly within the diffuse cohort, are associated with a loss of lung function. The RISE-SSc clinical trial, where the average disease duration was less than 9 months, and the tocilizumab trials confirmed a meaningful decline in FVC% predicted over 52 weeks.⁸⁶

This early inflammation results in activated fibroblasts with a profibrotic phenotype and protein expression consistent with TGFβ activation.

The established phase: progression and failed resolution

It is at this point there is a change from the early inflammatory lesions to a more fibrotic phenotype, visualised on CT as the development of NSIP pattern that can be cellular or fibrotic.⁸⁷ The innate and adaptive immune system, interacting with fibroblasts, and resulting in increased ECM deposition are likely key drivers at this stage. It has been postulated the underlying disease process is linked to whether the lung disease remains mild and stable, or evolves into an extensive organ manifestation. Influencing factors include the intensity of the inflammatory process, recruitment of profibrotic cell populations, and failure of apoptosis of myofibroblasts.⁸⁸ As these likely reflect background susceptibility, it is here where ANA subtypes, as well as altered microbiomes and recurrent aspirations probably have the biggest influence.⁸⁹ Local activation of fibrotic pathways with feed-forward loops involving TGFβ dependent pathways are involved, including mechano-sensors such as MRTF-A and YAP-TAZ (transcriptional co-activator with PDZ-binding motif).

Late phase: severe fibrosis in a subset of patients

Not all patients with ILD develop progressive fibrosis. Those most at risk are patients with a more extensive disease and are associated with a worse outcome, both from ILD and its complications including group 3 pulmonary hypertension.^90,91

Progressive disease is characterised by a failure of fibrosis resolution rather than continued escalation of inflammation. Progressive ECM deposition alters the mechanical properties of the lung, and the increased matrix stiffness activates mechanotransduction pathways important in driving myofibroblast differentiation.⁹² Persistent epithelial injury and cellular senescence may further impair restoration of normal alveolar architecture, while recurrent epithelial injury including from microaspirations from gastro-oesophageal reflux, provide additional stimuli for tissue injury.^93,94 These processes result in establishing a self-sustaining feed-forward fibrotic circuit where collagen deposition further increases tissue stiffness, which activates fibroblasts and results in ECM production.

While immunomodulatory treatment has the greatest impact in the early course of the disease, trials involving anti-fibrotics (eg, SenSCIS and INBUILD),⁸³ have shown that in the late stage of disease, there is more extensive architectural disruption and where the impact seems to be greatest. The process is unlikely to be linear or occur uniformly, and relative contributions from immune-mediated injury and fibrosis vary between patients and time points. This contributes to the heterogeneity of SSc-ILD.

Future directions

Understanding the pathogenesis of SSc, especially across different autoantibody subtypes, is important as disease management moves towards precision medicine. With increasing evidence that even in the VEDOSS patients (very early diagnosis of SSc), there is evidence of interstitial lung abnormalities and reduced pulmonary function tests, suggesting early lung disease.⁹⁵

Conclusions

SSc-ILD is not a linear disease, but rather a self-perpetuating system. It evolves through distinct pathogenic phases. Early immune-mediated injury gives way to self-sustaining fibroblast activation, maintained through mechanotransduction, ECM remodelling and failed tissue repair. This evolving understanding explains why immunomodulatory therapies are most effective early, where antifibrotic therapies have become increasingly used in more established fibrotic disease. Future precision medicine approaches will likely require treatment strategies tailored to both disease extent and the underlying pathogenic phase.

Table 1. Pathways associated with SSc-ILD, main cellular source, role in SSc-ILD, downstream effectors and our current methods of targeting them(table 139,40,41,45,47,52,53,62, 85).

Pathway	Main cellular source	Major biological role	Key downstream effectors	Therapeutic relevance
TGFβ	Fibroblasts	Master profibrotic pathway, myofibroblast activation	SMAD2/3, non-canonical MAPK, PI3K	Under investigation
PDGF	Fibroblasts	Fibroblast proliferation	PI3K, RAS/MAPK	Nintedanib- established progressive fibrosis
IL6	Macrophages/B cells	Immune amplification	JAK/STAT3	Tocilizumab- especially in early SSc-ILD
Chemokines (CXCL4, CCL2)	Platelets, fibroblasts, macrophages	Monocyte recruitment, macrophage activation and fibroblast stimulation	CCR2 signalling, CXCL3, NF-kB activation	Biomarker and potential future targets
Mechanotransduction	Fibroblasts,	Matrix stiffness, sustaining fibrosis	YAP/TAZ, MRTF-A	Emerging therapeutic targets
Innate danger sensing (TLRs)	Macrophages, dendritic cells, fibroblasts	Recognition of DAMPS and amplifies innate immune response	NF-kB	Experimental targets
PDE4B-cAMP	Macrophages, T lymphocytes, fibroblasts	Modulates inflammatory production and fibroblast activation	cAMP-PKA-CREB	Nerandomilast for progressive fibrosing ILD

cAMP, cyclic-AMP; CCL, C-C motif chemokine ligand; CCR, C-C motif chemokine receptor; CREB, cAMP response element binding protein; CXCL4, platelet factor 4; DAMPS, damage-associated molecular patterns; IL, interleukin; ILD, interstitial lung disease; JAK/STAT3, Janus kinase/signal transducer and activator of transcription; MAPK, mitogen-activated protein kinase; MRTF-A, myocardin-related transcription factor-A; NF-kB, nuclear factor kappa-light-chain-enhancer of activated B cells; PDE4B-cAMP, Phosphodiesterase 4B-cyclic AMP; PDGF, platelet derived growth factor; PKA, protein kinase A; RAS, rat sarcoma; SSc, systemic sclerosis; TGFβ, transforming growth factor; TLRs, Toll-like receptors; YAP/TAZ, yes-associated protein/ transcriptional co-activator with PDZ-binding motif.