Emerging drugs for the treatment of scleroderma: A review of recent phase 2 and 3 trials

Abstract

Introduction:

Systemic sclerosis (SSc) has the highest case-specific mortality of all connective tissue diseases. Its underlying disease mechanism affects several organs and remains incompletely understood. Ongoing work clarifying its etiopathogenesis is helping to develop targeted therapy.

Areas covered:

Several clinical trials have evaluated the safety and efficacy of agents targeting different mechanisms of this disease. This review article reviews those mechanisms and surveys four key recent phase II or III clinical trials that are contributing to the landscape of SSc therapy. The reported trials primarily focus on patients with systemic sclerosis in the early phase of disease.

Expert opinion:

Traditional therapies for SSc center on immunosuppressive and cytotoxic agents. A new cadre of therapies are borne from improved understandings of SSc pathobiology and target the inflammatory-fibrotic pathways. Scleroderma trials have entered the initial phase of personalized medicine, recognizing molecular subsets that will improve upon cohort enrichment and maximize the measurable benefit of future therapies.

1. Background

Systemic sclerosis (SSc), also called scleroderma, is a rare autoimmune disease. Recent reports from Europe, the United States, and Australia suggest the prevalence is in the range of 150–300 cases per million[1]. SSc occurs with a female: male predominance (4:1 to 6:1). African-American race is a risk factor for developing the disease, and suffering more severe disease with an earlier age of onset and worse prognosis[2–5]. The disease is the result of dysregulated immunity, vasculopathy, and fibrosis[6,7]. The extent of its impact is determined by the severity of an organ’s involvement with vasculopathy and tissue remodeling with inflammation and fibrosis. Disease features include Raynaud’s Phenomenon (RP), digital ulceration (DU), pulmonary arterial hypertension (PAH), skin fibrosis, interstitial lung disease (ILD), myocardial fibrosis, gastroesophageal reflux disease (GERD), dysphagia, inflammatory and fibrotic synovitis and tenosynovitis, and immune-related myopathy, among other manifestations. Two main subsets of SSc are described according to the extent of cutaneous involvement: diffuse cutaneous systemic sclerosis (dcSSc) and limited cutaneous sclerosis (lcSSc)[8].

2. Medical need

There is an unmet need in developing SSc disease modifying anti-rheumatic drugs (DMARDs)[9–12]. SSc has the highest case-specific mortality of all rheumatic diseases[13]. There is no one DMARD therapy for SSc; such a prospect is unlikely given the complex immunology, pathobiology, and clinical heterogeneity[14]. To achieve a low-disease state, especially in those with dcSSc, researchers have sought novel targeted therapies[15].

3. Existing treatment

Treatment options for vascular disease may be parsed into treatment for PAH, RP and DU, and scleroderma renal crisis (SRC). Treatment for PAH, RP, and DU includes oral calcium channel blocker therapy (nifedipine is commonly used), endothelin receptor antagonism (bosentan), oral PDE 5 inhibitor therapy (sildenafil and tadalafil), and prostacyclin analogs (iloprost)[16–18]. For PAH, upfront combination therapy versus monotherapy is recommended based on positive outcomes from the AMBITION trial [19,20]. A recent trial investigating the role of B-cell depletion (rituximab, NCT01086540) showed this was safe and showed promising data to be explored further in PAH[21]. Angiotensin-converting enzyme inhibitors (ACEi) have demonstrated a survival benefit in SRC[22].

Immunosuppressive or immune modulatory treatments used to treat cutaneous, musculoskeletal, and interstitial lung disease span a spectrum of conventional and biologic DMARDs, as well as intense immunosuppression with autologous hematopoietic stem cell transplantation (AHSCT). As ILD is one of the leading causes of disease-associated mortality, its treatment has become a critical area of study in SSc to meet an unmet need[23]. Pivotal trials like the Scleroderma Lung Study I and II have established morbidity, but not mortality, benefit of cyclophosphamide and mycophenolate mofetil[24,25]. More recently the SENCIS trial (see below) has led to the first FDA-approved therapy for SSc-ILD with an anti-fibrotic medication, nintedanib[26]. For those with early (disease onset 5 years or less from the first non-Raynaud’s Phenomenon) and diffuse cutaneous disease, AHSCT has demonstrated improvement in skin thickness, preservation of lung function, improved quality of life and length of life[27,28]. This therapy poses an elevated risk of treatment-related mortality early after transplant, which must be weighed against the demonstrated long-term survival advantages following transplant seen in several RCTs. Ultimately this therapy is appropriate only for a narrow spectrum of patients and its benefit may be dependent on institutional experience[29]. Other aspects of SSc including gastrointestinal involvement, joint contractures, and calcinosis have proven to be difficult to treat and will not be considered further here.

4. Current research goals

Improved disease management depends on (i) refining targets for therapy and (ii) implementing clinical trials to demonstrate their efficacy. In terms of goal (i), preclinical work has led to advances in understanding the pathogenesis of fibrosis, with a specific focus on the role of the immune system/inflammation. Improved basic science work on innate and adaptive immunity, microvascular changes, and fibrotic pathways has presented sites for targeted therapy[15]. Goal (ii) is impeded by several features of SSc; clinical and molecular heterogeneity of SSc has presented a challenge in understanding response to therapeutics[23]. This has led to advances in designing clinical trials with cohort enrichment strategies and optimizing outcome measures (including a combined response index for dcSSc, the American College of Rheumatology Composite Response Index in Systemic Sclerosis [ACR CRISS]) to be able to detect treatment effects[30]. Together, these two research goals inform the phase II and III clinical trials treatment/intervention target described here.

5. Scientific rationale

SSc is a multifaceted disease with a complex pathogenesis (Figure 1). Early events include vascular injury with excessive apoptosis of endothelial cells followed by intimal thickening due to uncontrolled proliferation of endothelial and smooth muscle cells. This proliferation of endothelial cells is especially driven by an unbalanced autocrine production of angiogenic growth factors such as vascular endothelial growth factor (VEGF)[31]. This results in luminal narrowing leading to tissue hypoxemia and oxidative stress. Under normal circumstances, hypoxemia leads to compensatory angiogenesis and vasculogenesis, but in SSc a specific defect of endothelial progenitors leads to angiogenesis impairment with reduction of capillaries and subsequently more severe hypoxemia[32].

Figure 1 : Main scleroderma-associated pathways targeted by the treatments evaluated in the considered Phase II or III trials.

Abatacept (ASSET trial) targets CD28-dependent costimulation of CD4+ T cells. Riociguat (RISE-SSc trial) exerts vasodilating properties and indirectly limits the effects of TGFβ on endothelial cells and fibroblasts/Myofibroblasts. Tocilizumab (FocuSSced Trial) targets the IL-6 receptor and limits IL-6 dependent activations of B cells, T cells, myofibroblasts and macrophages. Nintedanib (SENSCIS Trial) is a tyrosine kinase inhibitor that particularly targets the signaling of FGF-, VEGF- and PDGF-receptors.

endoMT: endothelial to Mesenchymal Transition; ROS: Reactive Oxygen Species; ECM: Extra-cellular Matrix

TGF-β: Tumor Growth Factor-β; PDGF: platelet derived growth factor; VEGF: vascular endothelial growth factor

FGF: fibroblast growth factor family; IL-6: Interleukin 6

Complete black arrows: secretion and impact of cytokines/growth factors

Dotted grey arrows: cell fate and differentiation

Dotted blue arrow: targets of cytotoxic TCD4 lymphocytes

Such endothelial dysfunction drives the recruitment of mononuclear inflammatory cells, which include T CD4+ lymphocytes and monocytes. This perivascular infiltrate in early SSc is especially composed of cytotoxic T lymphocytes that produce pro-fibrotic cytokines and promote apoptosis of endothelial cells leading to excessive fibrosis³⁴. The Th2 cytokines in the blood, IL-4 and −13 drive the differentiation of monocytes into M2 macrophages[34]. These M2 macrophages produce pro-fibrotic mediators, such as tumor growth factor-β (TGF-β), platelet derived growth factor (PDGF) and chemokine (C-C motif) ligand 18 (CCL18) that promote the trans-differentiation of resident fibroblasts into myofibroblasts, the key players of the fibrotic process driving tissue damage in SSc[35].

Myofibroblasts can also be derived from activated endothelial cells, epithelial cells or monocyte-derived circulating mesenchymal progenitor cells (fibrocytes). Myofibroblasts are responsible for the production of excessive extracellular matrix (ECM) and the release of pro-fibrotic and pro-inflammatory mediators such as TGF-β, fibroblast growth factor (FGF), cellular communication network factor 2 (CCN2; previously designated as connective tissue growth factor (CTGF)) and IL-6. In return, IL-6 activates B cells, drives an autocrine activation of myofibroblast and may promote M2 macrophages polarization through an over-expression of their membrane expression of IL-4R[36]. The increase stiffness of tissue environment due to the accumulation of ECM and some specific components of the ECM such as fibronectin and tenascin-C also participate in an uncontrolled activation of myofibroblasts notably driven by Toll Like Receptors. Interactions between fibroblasts and collagen fibrils are also mediated by a subset of β1integrin receptors that notably mediates the phosphorylation of FAK (Focal Adhesion Kinase) which promotes TGF-β signaling and fibrosis[37,38]. ECM is also responsible for a “feed-forward” loop as collagen produced by myofibroblasts, and collagen intermolecular cross-links (notably promoted by the telopeptide lysyl hydroxylase PLOD2 (procollagen-lysine, 2-oxoglutarate 5-dioxygenase 2)) will increase ECM stiffness with a subsequent further activation of fibroblasts/myofibroblasts that sensor mechanical tension[39]. Scleroderma fibroblasts also overexpress NADPH oxidase 2 and 4 (NOX2 and 4) through a PDGF-dependent pathway, and a NOX2/4 -associated overproduction of reactive oxygen species (ROS) may also participate to an increase production of collagen and fibronectin[40]. Beyond vasculopathy, oxidative stress also directly participates to the pathogenesis of SSc-associated fibrosis.

This unbalanced pathogenic loop occurs in various tissues in SSc, including skin and lungs. This process causes a replacement of normal tissue architecture with mechanically stiff collagen deposition, resulting in fibrotic disease impairing organ function.

DMARDs in SSc should target inflammatory/immunity, vascular, and fibrotic pathways. Inflammatory/immunity targeted therapy includes targeting T cell activation, B-cell activation, macrophage activation, and the JAK-STAT signaling pathways. Vascular targeted therapy includes endothelin receptor antagonists, phosphodiesterase 5 inhibition, and administration of synthetic analog of prostacyclin PGI2. Fibrotic targeted therapy includes inhibition of TGF-β, IL-6, and lysophosphatidic acid, agonism of the cannabinoid receptor agonism, and several other potential mechanisms[41].

6. Competitive environment

For an excellent review of drugs in phase I and phase II clinical trials for SSc, please see the recently published article in this journal by Chung and Chung[42]. Here we review four recent key clinical trials and the implications they pose for the treatment of SSc (Table 1 and Figure 1).

Table 1.

Template Competitive Environment Table

Compound	Company	Structure	Indications in Adults	Stage of development	Mechanism of action	Main cellular targets and suspected impact in SSc	Ref.
Abatacept	Bristol-Myers Squibb Company	Fusion protein composed of the Fc region of the immunoglobulin IgG1 fused to the extracellular domain of CTLA-4	Moderate to severe rheumatoid arthritis  Improves clinical and radiographic outcomes and health-related quality of life Psoriatic arthritis  A greater proportion of adult patients with PsA achieved an ACR20 response after treatment with IV and subcutaneous abatacept compared to placebo.	Phase II	Abatacept binds to CD80 and CD86 molecules, prevents a co-stimulatory signal, preventing T cell activation.	Abatacept may down regulate the activation of CD4+ Th2 cells that produce pro-fibrotic cytokines and that participate to the activation of both innate and adaptive immunity. Abatacept may down regulate B cell activation and differentiation through its impact on T follicular helpers	1
Tocilizumab	Roche	Humanized monoclonal antibody against the interleukin-6 receptor (IL-6R)	Moderate to severe rheumatoid arthritis  Improves clinical and radiographic outcomes and health-related quality of life Giant cell arteritis  Increased the rate of sustained remission in giant cell arteritis over the course of 1 year	Phase II Phase III	Tocilizumab binds soluble as well as membrane bound interleukin-6 receptors, hindering IL-6 from exerting its pro-inflammatory and pro-fibrotic effects.	Tocilizumab through its direct effects on IL-6 signaling may notably limit the activation of myofibroblasts and may limit the polarization of profibrotic M2 macrophages	2,3
Riociguat	Bayer	Carbamate ester that is the methyl ester of {4,6-diamino-2-[1-(2-fluorobenzyl)-1H-pyrazolo[3,4-b]pyridin-3-yl]pyrimidin-5-yl}methylcarbamic acid	Pulmonary arterial hypertension and Chronic thromboembolic pulmonary hypertension (CTEPH)  Improves exercise capacity and functional capacity in patients with inoperable or persistent/recurrent pulmonary hypertension.	Phase IIb	Riociguat is an anti-aggregatory, anti-proliferative, and produces vasodilatory effects by stimulating soluble guanylate cyclase and the receptor for nitric oxide. This leads to synthesis of cyclic guanosine monophosphate (cGMP), which promotes to vasodilation.	Riociguat may limit SSc-associated occlusive vasculopathy and may indirectly limit the effects of TGFβ signaling on fibroblasts, myofibroblasts and endothelial cells	4
Nintedanib	Boehringer Ingelheim	Indolinone-derived, receptor tyrosine kinase (RTK) inhibitor	Idiopathic pulmonary fibrosis and chronic progressive fibrosing interstitial lung diseases  Slows the rate of decline in pulmonary function in patients with systemic sclerosis	Phase III	Nintedanib binds to an intracellular ATP binding pocket of growth factor receptors (fibroblast, platelet-derived, and vascular endothelial) resulting in inhibition of the autophosphorylation of these receptors and the downstream signaling cascades.	Nintedanib may limit the occlusive vasculopathy through its effects on VEGF and PDGF signaling, may limit the impact on macrophage secretion on myofibroblasts through its effects on FGF and PDGF, and may limit autocrine activation of myofibroblasts	5, 6

^1.Khanna D, Spino C, Johnson SR, et al. Abatacept in Early Diffuse Cutaneous Systemic Sclerosis: Results of a Phase II Investigator-Initiated, Multicenter, Double-Blind, Randomized, Placebo-Controlled Trial. Arthritis Rheumatol. 2019;72(1):125–136.

^2.Khanna D, Denton CP, Jahreis A, et al. Safety and efficacy of subcutaneous tocilizumab in adults with systemic sclerosis (faSScinate): a phase 2, randomised, controlled trial. Lancet. 2016;387(10038):2630–2640.

^3.Khanna D, Lin C, Furst D, Goldin J, Kim G, Kuwana M, Allanore Y, Matucci-Cerinic M, Distler O, Shima Y vanLaar J. A randomised placebo-controlled phase 3 trial of tocilizumab in systemic sclerosis. Lancet Respir Med. 2020;In Press.

^4.Khanna D, Allanore Y, Denton CP, et al. Riociguat in patients with early diffuse cutaneous systemic sclerosis (RISE-SSc): Randomised, double-blind, placebo-controlled multicentre trial. Ann Rheum Dis. 2020;79(5):618–625.

^5.Distler O, Highland KB, Gahlemann M, et al. Nintedanib for systemic sclerosis-associated interstitial lung disease. N Engl J Med. 2019;380(26):2518–2528.

^6.Flaherty KR, Wells AU, Cottin V, et al. Nintedanib in progressive fibrosing interstitial lung diseases. N Engl J Med. 2019;381(18):1718–1727.

6.1. Phase II: ASSET trial investigating abatacept

T cells accumulate in SSc tissues and activated T-CD4+ Th2 cell-derived cytokines such as IL-13 and IL-4 trigger fibroblast activation and ECM production[43,44]. An abundant perivascular infiltrate of T-CD4+ cells is observed in skin biopsy samples from patients with SSc and cytotoxic T-CD4+ cells may also participate in endothelial damage[33,36]. The number of T cells found in skin biopsy samples correlate with the degree of skin thickening. Animal studies show attenuation of skin and lung fibrosis with a T cell activation antagonist, abatacept[45]. Abatacept is a soluble recombinant fusion protein that binds to CD80 and CD86 (also called B7–1 and 2, respectively), blocking T-cell co-stimulation and subsequent activation by preventing their interaction with CD28.

The Abatacept Systemic SclErosis Trial (ASSET trial) was a Phase II, 88 patient, multi-centered, investigator initiated, randomized, double-blind, placebo-controlled study evaluating the safety and efficacy of abatacept in patients with early diffuse cutaneous systemic sclerosis (ClinicalTrials.gov identifier: NCT02161406.)[46]. Recruitment for this trial focused on obtaining a homogenous cohort of dcSSc patients with active skin disease. To that end, selective inclusion criteria were set: all patients had early dcSSc; disease duration ≤18 months, an mRSS of ≥10 and ≤35 units was required, or duration of >18 to ≤36 months, an mRSS of ≥15 and ≤45 units was required alongside evidence of increasing mRSS in the previous 6 months (by ≥3 mRSS units), or involvement of 1 new body area with increase of ≥2 MRSS units, or involvement of 2 new body areas with increase of ≥1 mRSS unit, or presence of ≥1 tendon friction rub. The average patient in the trial was 49 years of age, most were women (75%) and White (82%), with a majority of participants (60%) within 18 months of the onset of their disease. The average mRSS was 22.45 (+/−7.65).

The primary endpoint of change from baseline in mRSS at 12 months was not met, as there was no significant difference between the abatacept and placebo groups (−6.24 ± 1.14 and −4.49 ± 1.14, respectively with a treatment difference of −1.75 [95% CI −4.93, 1.43]). Patients in the abatacept arm did improve in their HAQ-DI (health assessment questionnaire-disability index) and the American College of Rheumatology Combined Response Index in Systemic Sclerosis (ACR CRISS). The co-primary outcome of safety showed abatacept-treated patients to have less adverse events, infectious adverse events, and serious adverse events.

What makes this trial highly informative is highlighting the molecular heterogeneity in a seemingly homogenous phenotypic population and its ability to assign patients based on intrinsic skin gene expression to one of three subsets: inflammatory, fibroproliferative, or normal-like intrinsic molecular subsets[47–49]. Patients had skin biopsies performed at baseline and RNA sequencing determined which patients were inflammatory (33 out of 84), fibroproliferative (18 out of 84), and normal–like (33 out of 84). A comparison of treatment and placebo arms in these subsets revealed significant changes in the mRSS in the inflammatory and normal-like skin gene expression subsets.

CD28 is essential for naïve T cell activation as high levels of circulating soluble CD28 are detected in patients with SSc and implicates T cell activation in the pathogenesis of SSc[50]. A CD28-related pathway was identified to improve significantly more in the inflammatory subset, relative to the other two subsets. Importantly, the baseline expression of co-stimulation by the CD28 family pathway correlated with the decrease in mRSS, specifically for those inflammatory patients on abatacept[51]. This provides a critical step towards understanding pathophysiology and its clinical correlate, demonstrating the prediction of the differential response to biological intervention based on its molecular signature.

The open label extension (OLE) data obtained at 18 months was able to assess the response to medication in those initially treated with placebo, then placed on abatacept. At the end of the double-blind (month 12), 8% (3 of 37) of placebo patients had an increase of ≥5 FVC % predicted units and that percentage increased to 14% (4/29) by month 18. There was no significant difference in the placebo arm vs. placebo-abatacept arm at month 18 in terms of the mRSS decline, HAQ-DI, ACR CRISS, or classification as a multi-component responder (mRSS ≤10, HAQ-DI≤0.75 and Patient Global Assessment ≤3)[52].

One important challenge in SSc clinical trials is recognizing and accounting for clinical heterogeneity in early SSc. This is the first phase II placebo-controlled study of intrinsic skin gene expression subsets predicting response to targeted biological therapy. In the design of phase III for abatacept, participants may be stratified based on skin gene expression data rather than clinical phenotypes, such as disease duration or autoantibodies. A further possibility includes examining the combination of anti-inflammatory and anti-fibrotic therapies (e.g., abatacept + an anti-fibrotic therapy like nintedanib or pirfenidone), stratified by these molecular subsets, to demonstrate relative efficacy based on its expected effect on molecular pathways, especially considering that combination therapy might be a promising strategy for SSc in the future[53].

In summary, current data do not support the use of abatacept for treatment of early dcSSc in clinical practice. This Phase II study did not demonstrate improvement in skin thickening. It did show a benefit compared to placebo in terms of patient-reported outcome (HAQ-DI) and the ACR CRISS measure. The extent of its efficacy may be further clarified in a Phase III trial, targeting those most likely to benefit (e.g., inflammatory subset).

6.2. Phase II: RISE-SSc trial investigating riociguat

The mechanism of fibrosis and vasculopathy is a cascade of endothelial injury with consequent vessel wall and fibrotic remodeling. Soluble guanylate cyclase (sGC) is a key mediator in regulating vascular tone, and catalyzes the production of cyclic guanosine monophosphate (cGMP). This in turn is an activator of protein kinases G (PKG), which has demonstrated regulation of vascular tone and remodeling[54]. Riociguat is a sGC stimulator approved for treatment of PAH, following the results of the PATENT-1 trial[55]. Preclinical studies have demonstrated anti-inflammatory, anti-proliferative, and anti-fibrotic properties in mouse models and in vitro notably through its effect on TGF-β secretion and signaling [56–59]. These anti-fibrotic effects of sGC stimulators may particularly involve a SMAD2/3-independent non-canonical pathway of TGF-β signaling, notably involving an up-regulation of p-ERK but independent from the phosphorylation of Akt or JNK[57].

The RIociguat Safety and Efficacy in patients with diffuse cutaneous Systemic Sclerosis (RISE-SSc) trial was a large randomized, double-blind, placebo-controlled, parallel group clinical trial occurring over 1 year with the primary focus on assessing skin fibrosis (ClinicalTrials.gov identifier: NCT0228376219)[60]. This trial also targeted those with progressive skin fibrosis in dcSSc. All patients had a disease duration ≤18 months and mRSS 10–22 units. The primary endpoint was the change in mRSS from baseline at week 52. The average patient was 50.7 years old, most were women (76%), and White (73.6%), with a mean disease duration of 9 months. The average mRSS was 16.8 (3.7).

The primary endpoint was not met (change from baseline in mRSS at 12 months); the riociguat arm had a decrease in mRSS (−2.09) compared with the placebo arm (−0.77) (difference −2.34 [95% confidence interval −4.99 to 0.30], p=0.08). The secondary endpoint of measuring improvement with ACR CRISS was not significantly different between the two arms, although the design of the trial was to detect prevention of progression, not disease improvement. Assessment of lung function showed an FVC% predicted without significant difference in the two arms. Signals for efficacy were found in the proportion of patients with prevention of worsening: fewer patients in the riociguat arm (compared to placebo) had progression of mRSS as defined by a >5 units and ≥25% increase from baseline: 11/59 (vs. 22/60), more regression defined by a decrease of >5 units and ≥25% from baseline: 27/59 (vs. 18/60). Less patients developed new DUs in the riociguat arm (compared to placebo):5/60 (vs. 12/60); the number of DUs at 52 weeks were also less in the riociguat arm: 12 (vs. 72). Almost all (96.7%) of patients on riociguat had an adverse event, most commonly mild gastrointestinal symptoms (gastroesophageal reflux disease, diarrhea, nausea) or nervous system disorders (dizziness, headache). Serious adverse events were reported in 9 patients (15%).

Biomarker analysis confirmed the impact of riociguat compared to placebo. Riociguat affected the targeted nitric oxide-sGC-cGMP pathway, showing elevated plasma cGMP at week 14 in the riociguat arm compared to placebo-treated patients. There was evidence of reduced serum chemokine ligand 4 (CXCL-4), a marker of progressive SSc, at week 14 in the riociguat arm compared to baseline. There was decreased endothelial cell activation (decreased serum soluble platelet endothelial cell adhesion molecule 1, sPECAM-1). Those with higher baseline sPECAM-1, compared to those with lower levels, showed larger mRSS reductions in the riociguat arm compared to placebo. Patients with skin biopsy-positive alpha-smooth muscle cell actin (alphaSMA), an indicator of fibrosis, showed a reduction of mRSS with riociguat vs. placebo[61].

Those patients completing 52 weeks in the double-blind portion of the trial were eligible for OLE, with 42/60 patients in the riociguat arm and 45/61 patients in the placebo arm being placed on riociguat when entering OLE[62]. More patients in the placebo-then-riociguat arm showed mRSS regression (defined by a decrease in >5 units from baseline and ≥25% decrease from week 52 to their last visit) than those in the riociguat-then-riociguat arm at 184 weeks of therapy. Additionally, the digital ulcer burden in the OLE worsened in the placebo-then-riociguat arm (mean change of 0.222), whereas it was stable for those on riociguat for the entire study (−0.024). In a subset analysis during the main study, 11 patients with ILD on placebo had a mean decline from baseline in the FVC% predicted of 7.6%, whereas 11 patients with ILD on riociguat had a decline of 2.7%. In the OLE, from week 52 to week 100, the 11 patients with ILD on placebo, now on riociguat had a mean decline in FVC% predicted of 3.7% (improved from 7.6%); the 11 patients with ILD on riociguat during the main study and on riociguat during OLE essentially stabilized the decline of FVC% predicted at 4.1%.

The attenuation in decline in FVC% predicted seen in SSc-ILD during OLE is the result of a small, non-selected population and statistical analysis was not performed for the ILD subgroup. In summary, while the role of riociguat in the treatment of SSc-PAH is clear, based on high-quality RCTs[63], its benefit in early SSc remains less convincing.

6.3. Phase III: SENSCIS trial investigating nintedanib

SSc–ILD is a consequence of repetitive vascular injury with endothelial dysfunction, responsible for uncontrolled intima and media thickening notably driven by mediators such as VEGF and FGF. This endothelial dysfunction is responsible for subsequent recruitment of mononuclear cells and pro-fibrotic M2 macrophages in the pulmonary interstitium. M2 macrophages release pro-fibrotic mediators that notably include PDGF and TGF-β[64]. This leads to the activation of fibroblasts which undergo fibroblast–myofibroblast transition, releasing excessive amounts of ECM and pro-fibrotic mediators such as FGF through an autocrine process. This pathogenic loop increases ECM stiffness, and incurs further fibroblast activation in a positive feedback with repetitive injury and increased endothelial dysfunction, partly mediated by growth factors such as FGF[65].

Nintedanib is a tyrosine kinase inhibitor approved for use in idiopathic pulmonary fibrosis by the US Food and Drug Administration in 2014, and the European Medicines Agency in 2015. This medication acts by competitively binding to ATP–binding pockets of receptors, stopping intracellular signaling. It targets PDGF receptor alpha and beta, FGF receptor 1–3, and VEGF receptor 1–3. In effect, it acts to short-circuit the positive feedback loop by preventing the release of growth factors, abrogating vascular injuries and fibrotic consequences[64].

The Safety and Efficacy of Nintedanib in Systemic Sclerosis (SENSCIS) trial was a 52-week randomized double-blind, placebo-controlled trial of patients with SSC–ILD, with a minimum of 10% of lung involvement as determined by HRCT (ClinicalTrials.gov, NCT03313180) [26]. For entry, patients were required to have significant lung disease as determined by the presence of restrictive lung physiology. About half had dcSSc, a disease duration of about 40 months, and the extent of fibrosis of the lungs on HRCT was about 35% in either the active or placebo arms. Patients were allowed to be on background mycophenolate or methotrexate if they were on a stable dose for 6 months or prednisone up to 10 milligrams/day.

This study demonstrated a statistically significant effect on the decline in lung function in SSc-ILD, without noted change in the health-related quality of life or skin fibrosis. The primary endpoint was met as nintedanib-treated participants had a lower annual rate of decline in the FVC (−52.1[13.8] mL/year) compared to those in the placebo arm (−93.3 [13.5] mL/year). The difference was 41.0 mLs/year (P = 0.04). About half of patients were on background mycophenolate therapy, while patients on both therapies (mycophenolate and nintedanib) had a lesser decline in FVC (−40.2 mL/year) compared to those not on background therapy (−63.9 mL/year).

Gastrointestinal adverse events were common, with diarrhea, vomiting, and elevated liver enzymes the most common. Sixteen percent of patients in the nintedanib arm had to discontinue due to an adverse event, compared to 8.7% in the placebo arm. Diarrhea was reported in 76% of those in the active arm and 32% in the placebo arm; this is considerably higher in the placebo group compared to other placebo-controlled SSc trials and higher in the treatment arm when compared to cancer and idiopathic pulmonary fibrosis trials. The high prevalence of diarrhea may have in part been the result of the nocebo phenomenon, where the presence of negative anticipations to treatment contribute to symptom development. Indeed, the percentage of diarrhea reported in the placebo arms of different trials with nintedanib increased as the number of mentions and the number of lines devoted to “diarrhea” in the informed consent forms[66].

In post-hoc analyses, there were no associations between mRSS at baseline or progression of mRSS and the progression of lung disease[67]. However, those with a baseline mRSS >18, when compared to those with mRSS <18, had a lower mean baseline FVC% predicted (68.3% vs 73.7%); in the placebo group, this subpopulation had a greater rate of FVC decline (−131.7[29.2] mL/year compared to those with a baseline mRSS <18 ( −81.4 [15.4] mL/year). The baseline mRSS (≥18 vs. <18) did not portend a differential treatment effect by treatment arm[68]. Participants in the placebo group with elevated baseline CRP (>4.99 milligrams/liter) had a higher rate of FVC decline than those with a low CRP (≤ 4.99 mg/L). Nintedanib had a greater impact on FVC reduction in those with elevated baseline CRP compared to those with low CRP[69]. In the placebo group, male patients had higher FVC decline compared to females (−126.8 [SE 29.0] vs −82.0 [16.2] mL/year); the rate of FVC decline was lower with nintedanib than placebo both in male and female patients[70].

The minimal clinically important difference (MCID) for improvement for FVC% predicted, or the smallest difference in measurement detectable by a patient, is estimated to be an improvement of ≥3.0% or a decline of ≤−3.3% based on previous SSc-ILD studies[71]. More patients in the nintedanib arm experienced improvement as defined by a FVC% predicted increase of ≥3% (compared to placebo): 66/287 or 23% (vs. 43/288 or 15%)[72]. Data to support optimal use as initial monotherapy or upfront combination with another agent (e.g., mycophenolate) have not been determined[73].

In summary, nintedanib has demonstrated in a high-quality RCT its benefit to retard decline of lung function but had no effect on feel, function or survival. Future trials should explore effect of nintedanib in dcSSc with early ILD.

6.4. Phase III: FocuSSced trial investigating tocilizumab

There are several lines of evidence supporting that IL-6 plays an important role in the pathogenesis and has a prognostic value in SSc. Serum IL-6 levels are elevated in dcSSc, especially those with SSc-ILD[74]. Serum IL-6 levels correlate with the extent of skin fibrosis and/or predictive of functional decline and mortality in SSc-ILD. On dermal biopsy, IL-6 expression elevation in early dcSSc predicts more severe skin involvement at three years and worse long-term survival. IL-6 is over-expressed by macrophages, B-cells, and fibroblasts. In SSc, IL-6 is responsible for an autocrine and paracrine activation of fibroblasts, may promote the differentiation of monocytes into pro-fibrotic M2 macrophages, and is involved in B-cells activation, which may participate in the production of SSc-related autoantibodies.

Tocilizumab (TCZ) is an IgG1 immunized anti-IL-6 receptor monoclonal antibody that blocks IL-6 mediated signaling. It has been approved for moderate to severe rheumatoid arthritis, giant cell arteritis, and juvenile idiopathic arthritis among other diseases. TCZ was previously evaluated in SSc in the faSScinate trial[75,76] and laboratory analysis linked to the faSScinate trial investigated the effect of TCZ on the molecular, functional, and genomic properties of explanted fibroblasts from study[77]. The gene expression of fibroblasts treated with TCZ was downregulated for both the IL-6 pathway and the M2 macrophage, suggesting an interruption of IL-6’s influence of switching macrophage polarization towards a pro-fibrotic pathway.

The focuSSced trial was a multicenter, randomized, double-blind, placebo-controlled phase III study completed over 48 weeks (ClinicalTrials.gov, NCT02453256)[78]. All participants had dcSSc with disease onset <60 months from the onset of their first non-Raynaud’s Phenomenon symptom, and had an mRSS between 10–35 units. Participants also had elevated acute-phase reactants (>1 of the following: CRP >6mg/L, ESR >28mm/h, or platelet count >330 ×109/L) and active disease defined as >1 of the following at screening: disease duration ≤18 months, mRSS increase ≥3 units, or involvement of one new body area and mRSS increase ≥2 units, or involvement of two new body areas (each within the previous 6 months), and ≥1 tendon friction rub. The primary efficacy endpoint was the difference in change from baseline in mRSS at week 48 and a secondary efficacy endpoint was change from baseline to week 48 in FVC% predicted.

Participants treated with TCZ had a greater reduction in mRSS compared to placebo, but did not achieve statistical significance. Similar to the phase II trial, those in the treatment arm at 48 weeks had preserved FVC% predicted compared to those in the placebo arm (−0.4 versus −4.6, (difference, 4·2 [95% CI 2·0 to 6·4]; nominal p=0·0002). In addition to serial spirometry, participants had HRCT at baseline and at 48-week follow-up. Of the 210 participants in the trial (107 in the placebo arm, 105 in the TCZ arm), 65% (136/210) had SSc-ILD. Over 48 weeks, TCZ stabilized FVC% predicted independent of the severity of lung involvement, emphasizing this medication’s benefit as an early intervention to prevent progression for those with mild to severe lung involvement. Serious adverse events were reported in both TCZ and placebo arms, including infection (TCZ, 3 events; placebo, 8 events) and cardiac events (TCZ, 2 events; placebo, 7 events).

This trial was designed to detect a change in mRSS, enriched with early dcSSc with evidence of active inflammation and active skin disease at baseline. The consequence is that this cohort represents a uniquely early-stage population at high risk for progressive SSc-ILD, which differ considerably from previous studies[24–26,79]. The 136 participants were not selected based on respiratory symptoms, but did have moderate-to-severe lung involvement. This demonstrated the relatively large decline in the placebo arm compared to previous placebo arms in other trials. The average rate of decline of FVC% predicted was about 6.5%, considerably higher than those reported in the FAST trial (decline of 3.0%), SLS-I trial (2.6%), and the SENSCIS cohort (2.6%).

This trial emphasizes early ILD as a treatment target, where early dcSSc with elevated inflammatory markers, despite having none-to-minimal respiratory symptoms, are treated before irreversible fibrotic disease sets in. The preservation effect seen in the TCZ arm across a wide spectrum of severity at baseline argues for TCZ’s role as targeted immunomodulatory therapy early in clinical treatment, whereas anti-fibrotic therapy may be more appropriate for later-stage advanced fibrotic disease.

At the end of 48 weeks, the mean mRSS declined in both placebo (−5.3) and the TCZ (−6.7) arms; during the OLE from 48 to 96 weeks, the mRSS continued to decline in those on placebo who switched to open label TCZ (mean change of −2.5), and those on TCZ switching to open label TCZ (−2.3).

At the end of 48 weeks, the mean FVC% predicted declined by −4.1% in the placebo and plateaued in the TCZ (−0.2%) arm. During the OLE from 48 to 96 weeks, the FVC% predicted plateaued in those on placebo who switched to open label TCZ (0.6%) and in those on TCZ switching to open label TCZ (−0.3 %). From 48 to 96 weeks, those in the placebo group that transitioned to open label TCZ showed a similar FVC% predicted preservation as those in the double-blind TCZ arm at 48 weeks. Together, these OLE data show continued numeric improvement in the mRSS and FVC preservation similar to the double-blind portion of the study[80].

In summary, tocilizumab has demonstrated prevention of lung function decline in 2 well designed trials, although that was not the primary outcome measure for its Phase II and Phase III studies. At this time, data suggest tocilizumab can be considered in early dcSSc with elevated acute-phase reactants and progressive skin disease with early ILD, to prevent decline of lung function.

7. Conclusion

SSc remains an orphan disease without adequate DMARD therapy akin to other rheumatic diseases, like rheumatoid arthritis[14]. Challenges for treatment assessment in clinical trials include clinical and molecular heterogeneity, insufficient trial duration, and/or lack of responsive outcome measures. Nevertheless, these challenges have sparked advanced understanding of its pathobiology and candidate biomarkers[23]. The goals of research in SSc are to improve targeted therapy and employ well-designed clinical trials to maximize the measurable benefit of those therapies. In this article, we highlighted four targeted therapies and critically assessed their recent trials.

The results of the ASSET trial, particularly the results of skin gene expression analysis, suggest that future clinical trials may be able to improve outcomes by targeting high-yield pathways with biological therapy. This step towards identification of gene expression subsets based on molecular signatures will allow early treatment with more precision than previously possible[47]. The ASSET trial showed that an inflammatory skin gene signature at baseline predicts improvement in mRSS, likely due to elevated CD28 family pathway in the inflammatory subgroup, the target for abatacept. Future trials should incorporate this as a stratification factor or only performing trial in those patients who will have large benefit to the treatment. Although the RISE-SSc trial did not find benefit in measuring mRSS change between treatment and placebo arms, it showed that prevention of progression can be an important end point. The shortcomings of traditional immunosuppressive treatment for SSc-ILD and advances in understanding the pathophysiology of fibrotic lung diseases led to the SENSCIS trial. The modest benefits seen in this trial may be the consequence of a primary endpoint of one-year follow-up, slow decline in the placebo group (part of natural history of the ILD), and lack of reversibility of lung fibrosis. Further study on the use of this medication will clarify if it retards cumulative decline and demonstrates a clinically-appreciable benefit or affects SSc-ILD mortality. The focuSSed trial represents an important weapon in the arsenal when combating early SSc–ILD in dcSSc with high risk for progressive disease. Unlike SENSCIS trial, this trial enriched for elevated acute phase reactants anti-SCL 70, which are predictors of progressive ILD. This trial underscores that in this particular high-risk cohort, significant decline over one year occurs without treatment, and represents a critical opportunity to initiate therapy to prevent advanced stages of fibrotic lung disease.

In complex rheumatologic disease, composite endpoints have previously been utilized in registrational studies to establish efficacy. For example, the American College of Rheumatology 20% response criteria for rheumatoid arthritis and the SELENA-SLEDAI score for systemic lupus erythematosus were used to establish efficacy in an approved therapy. ASSET and the focuSSed trial provided further validation of the ACR CRISS index. ACR CRISS is a 2-step process that assigns a probability of improvement for each subject and includes core items that assess change in 2 prominent manifestations of early SSc (skin and interstitial lung disease), functional disability (HAQ-DI) and Patient and Physician Global Assessments (PTGA and MDGA, respectively). Step 1 allows the Investigator to assess if a subject has developed new or worsening cardiopulmonary and/or renal involvement due to SSc. If a subject meet any of these criteria, the subject is assigned a probability of 0. Otherwise, in Step 2, the probability of improvement is calculated based on the 5 core measures incorporated into the ACR-CRISS, including changes in mRSS, FVC % predicted, HAQ-DI), PTGA and MDGA[30].

10. Expert opinion

Treatment of SSc is organ-focused and aimed at attenuating disease progression. There is an unmet need in the early dcSSc population in terms of quelling the inflammatory and subsequent fibrotic sequela in multiple organs, especially those with progressive ILD. At this time, only immune-ablative therapy and autologous hematopoietic stem cell transplant has been shown to improve survival in clinical trials[27,28]. However, this treatment option is available only to a minority of patients.

Clinical trials in SSc are challenged by relatively low prevalence, clinical heterogeneity in terms of which organs are involved and to what extent, and the presence of unique disease subsets present with the same clinical phenotype according to current definitions. This last feature, identifying a molecular pattern that portends differing rates of progression and response, is a critical focus of the research agenda in improving outcomes. The identification of 4 subsets (fibroproliferative, inflammatory, limited, and normal-like) holds the promise of the path towards personalized medicine[81]. The identification of accurate and highly reliable biomarkers in SSc is critical to the next phase of rational drug design and optimizing targeted therapy. Gene expression based biomarkers of skin disease progression and response is a promising avenue in this regard[82]. Detection of treatment effect in a clinical trial may be hampered by clinical heterogeneity and variation in treatment response. Intrinsic gene expression profiling, as instituted in the ASSET trial, has the capability to categorize patients based on disease pathogenesis in addition to clinical features present at the time of evaluation. In sum, gene expression analysis may allow for a more effective risk stratification and treatment selection, although this is yet to be demonstrated.

Promising targets for future intervention may particularly center on macrophage and fibroblast activation: these are well represented in key tissues affected in the disease process and an improved understanding of their pathogenic role in SSc has led to potential therapeutic targets.

Both M1 and M2 macrophages are upregulated in SSc and respectively produce pro-inflammatory and pro-fibrotic cytokines [82,83]. A common macrophage gene expression signature is present in several organs affected with severe disease (e.g., blood, skin, esophagus, lung)[34,84,85]. M2 macrophages and their differentiation is particularly involved in a network of immune and pro-fibrotic pathways[86]. M2 macrophages notably differentiate from monocytic precursors under the influence of IL-4 and IL-13 (Figure 1), produced during a strongly Th2-polarized response[87]. A recent Phase II study examined romilkimab, an engineered, humanized, bispecific immunoglobulin-G4 antibody that binds and neutralizes IL-4/IL-13[88]. This 24-week trial in patients with early dcSSc showed a statistically significant reduction in mRSS compared to placebo in the active arm, and was well-tolerated. IVIG (IV Immunoglobulin) may also specifically impact monocyte/macrophage activation through a modulation of FcγR signaling and could also directly target autoantibodies and fibroblasts[89]. Pre-clinical models[90] and observational studies support their use notably for scleroderma-associated myositis[91]. Their relevance for other scleroderma-associated manifestations such as skin involvement in early dcSSc may deserve future investigations.

Non-canonical and canonical TGF-β signaling plays a central role in fibrosis and despite disappointing results of Riociguat, other drugs targeting these pathways need to be tested in SSc, alone or in combination. Pirfenidone down-regulates smad3 dependent- and Akt-dependent TGF-β signaling in primary human lung fibroblasts[92] and is endorsed for the management of idiopathic pulmonary fibrosis. The multicenter randomized placebo-controlled Scleroderma Lung Study III will evaluate the benefit of pirfenidone as add-on to a background therapy of MMF (NCT03221257). Selected TGF-β targeting by fresolimumab, an engineered human monoclonal Ig that neutralizes the 3 major isoforms of TGF-β, namely, β1, β2, and β3, has shown promising results on mRSS in explorative studies although its effects still need to be explored in RCTs[93]. A Phase 1b trial targeting TGF-β1 and β3 (due to potential toxicity associated with targeting β2) showed benefits on skin score and was well tolerated[94]. MT-7117 is a novel orally-administered, small molecule, which acts as an agonist of melanocortin-1 receptor (MC1R). MC1R is activated by α-melanocyte-stimulating hormone which can be locally synthesized in response to sunlight exposure. In a bleomycin skin model, treatment with MCIR agonist reduced skin fibrosis and collagen content[95].

JAK/STAT signaling plays an important role in the pathogenesis of SSc, and is up-regulated in macrophages and in particular SSc fibroblasts. JAKs/STATs particularly mediate the signal of key SSc-related cytokines such as IL-6, IL-4/IL-13 or type I interferons[34,96–98]. The pan-JAK inhibitor tofacitnib showed preventive benefit in bleomycin-induced skin and lung fibrosis[99]. A recent Phase I/II study of tofacitnib in early dcSSc, conducted in a double-blind randomized placebo controlled trial fashion over 6 months showed a trend towards efficacy favoring the study drug on the mRSS and ACR CRISS, and was well-tolerated[100].

In a preclinical model of SSc, lysophosphatidic acid (LPA-LPA1) signaling participates to myofibroblast accumulation, and TGF-β canonical pathway activation. LPA also impairs anti-inflammatory and pro-resolving properties of macrophages in other pulmonary disorders [101]. In an 8-week double-blind, randomized, placebo-controlled study followed by a 16-week open-label, SAR100842 a selective orally available LPA1 receptor antagonist, was well tolerated in patients with dcSSc and showed a numerically higher reduction of mRSS in the SAR100842 group[102]. The Autotaxin pathway may also play a key role in the fibrotic manifestations of SSc as autotaxin is required for the development and maintenance of dermal fibrosis in pre-clinical models through co-amplification of LPA and IL-6[103]. Combination therapies both targeting the LPA pathways and the JAK/STAT-dependent IL-6 signaling may constitute a promising therapeutic approach. With this regards, combination of monoclonal antibodies should also be discussed in the future.

Cardiopulmonary involvement represents the highest risk for mortality and is a top priority in SSc treatment research. The traditional model of treating SSc-ILD only when clinical symptoms are present or progressing is being challenged by the introduction of earlier detection of ILD before respiratory symptoms or cardiopulmonary compromise sets in, as the risk/benefit ratio improves with more favorable toxicity profiles compared to cyclophosphamide or mycophenolate mofetil. Despite variation in screening with HRCT for all patients with SSc[104], we recommend all SSc patients receive HRCT to exclude ILD. If detected, these patients must be risk stratified based on the extent of the radiographic involvement, degree of FVC and DLCO impairment, respiratory symptoms, and cardiopulmonary exercise limitation. Those patients with mild extents of ILD, but with high risk for progression, as demonstrated in the focuSSced trial, may be considered for therapy with tocilizumab. Those patients with more advanced ILD in terms of the extent of radiographic disease and presence of symptoms or cardiopulmonary compromise, with limited skin or musculoskeletal involvement may be good candidates for nintedanib. Alternatively, these patients may be initiated on mycophenolate and have nintedanib added to their treatment regimen if they are treatment refractory.

Indications for the use of individual immunosuppressive drugs for other manifestations of dcSSc (skin thickening, myocarditis, pericarditis, myositis, arthritis) remain important focuses for future clinical trials. The advent of biological therapies expands treatment options beyond prednisone, cyclophosphamide, methotrexate and mycophenolate mofetil, but at this time none have been studied and proven to significantly modify disease.

In summary, SSc is an autoimmune-mediated fibrotic disease with a pathogenesis that is becoming increasingly better understood. Improved outcomes will result from early diagnosis and treatment, while research goals include designing better targeted therapy and successfully identifying those patients who are most likely to benefit. Stratification in clinical trials is based on clinical phenotypes (e.g., skin distribution, auto antibody status, disease duration) and, unfortunately, the heterogeneity of this disease may obscure outcomes. It is critical that those interpreting clinical trial results understand that a trial may not meet its primary endpoint for several reasons, other than drug inefficacy. Steps towards the future includes refining cohort stratification based on disease pathogenesis and molecular signatures, where biologic targeted therapy is known to intervene. Advances in biomarkers allow for early disease detection before irreversible fibrotic damage sets in, and may transition the goal of therapy from disease attenuating to secondary prevention.