Nintedanib reduces pulmonary fibrosis in a model of rheumatoid arthritis-associated interstitial lung disease

Abstract

Rheumatoid arthritis (RA)-associated interstitial lung disease (RA-ILD) develops in ~20% of patients with RA. SKG mice, which are genetically prone to development of autoimmune arthritis, develop a pulmonary interstitial pneumonia that resembles human cellular and fibrotic nonspecific interstitial pneumonia. Nintedanib, a tyrosine kinase inhibitor approved for treatment of idiopathic pulmonary fibrosis, has been shown to reduce the decline in lung function. Therefore, we investigated the effect of nintedanib on development of pulmonary fibrosis and joint disease in female SKG mice with arthritis induced by intraperitoneal injection of zymosan (5 mg). Nintedanib (60 mg·kg^-1·day^-1 via oral gavage) was started 5 or 10 wk after injection of zymosan. Arthritis and lung fibrosis outcome measures were assessed after 6 wk of treatment with nintedanib. A significant reduction in lung collagen levels, determined by measuring hydroxyproline levels and staining for collagen, was observed after 6 wk in nintedanib-treated mice with established arthritis and lung disease. Early intervention with nintedanib significantly reduced development of arthritis based on joint assessment and high-resolution μ-CT. This study impacts the RA and ILD fields by facilitating identification of a therapeutic treatment that may improve both diseases. As this model replicates the characteristics of RA-ILD, the results may be translatable to the human disease.

Keywords: SKG mice; inflammation; interstitial lung disease; rheumatoid arthritis.

Fig. 1.

Late treatment with nintedanib did not alter established arthritis in SKG mice. A: daily oral gavage of mice with nintedanib or saline beginning 10 wk after zymosan injection. B and C: there was no significant change in weight or survival in nintedanib- compared with saline-treated mice. D: visual arthritis score was not significantly different in nintedanib- compared with saline-treated mice until week 16. Values are means ± SE; n = 50 mice per treatment group. ***P < 0.001.

Fig. 2.

Late treatment with nintedanib did not alter lung compliance but significantly reduced collagen levels in the lungs. A: quasi-static compliance was significantly reduced in zymosan-injected, saline-treated mice compared with naïve controls. Treatment with nintedanib did not improve C_st. B: hydroxyproline levels were significantly reduced in nintedanib- compared with saline-treated mice and were not different from naïve controls. Data were obtained at week 16 and are shown as median ± interquartile range (n = 50 mice per treatment group and 10 naïve controls). **P < 0.01; ***P < 0.001.

Fig. 3.

Collagen was not increased in lungs harvested from nintedanib-treated mice at week 16. Hematoxylin-eosin (H&E)-stained lung sections from 3 representative animals demonstrate similar patchy areas of inflammation and disease. Picrosirius red (PSR)-stained serial sections show decreased collagen (red) in nintedanib- compared with saline-treated animals. Final magnification ×200, except insets (×300).

Fig. 4.

Late nintedanib treatment significantly reduced the presence of α-smooth muscle actin (SMA)-positive (myo)fibroblasts, as shown by immunostaining (final magnification ×400) and semiquantitative analysis of α-SMA-positive cells. Data were obtained at week 16 and are shown as median ± interquartile range (n = 10 mice per group). *P < 0.01.

Fig. 5.

Late treatment with nintedanib altered bronchoalveolar lavage (BAL) cell and fluid composition. A: differential counts of BAL cells showed no difference in the number of macrophages (Mac) but a significant increase in the number of lymphocytes (Lymph) and neutrophils (Neuts). B–D: BAL fluid levels of TNF-α, IL-4, and KC were increased in nintedanib- compared with saline-treated mice. E: levels of VCAM-1 were significantly reduced after nintedanib treatment. F: IFN-γ was not significantly changed after nintedanib treatment. Data were obtained at week 16 and are shown as median ± interquartile range (n = 20 mice per treatment group and 10 controls). *P < 0.05; **P < 0.01.

Fig. 6.

Inflammatory cell subsets in lungs of SKG mice after late treatment with nintedanib. Animals were treated with nintedanib or saline for 6 wk, and lung digests were analyzed by flow cytometry analysis at week 16. A: total cells; B: CD4⁺ T cells; C: CD8⁺ T cells; D: B220⁺ B cells; E: tissue/interstitial macrophages; F: inflammatory macrophages; G: alveolar macrophages; H: monocytes; I: CD11b⁺ dendritic cells; J: CD103c⁺ dendritic cells; K: neutrophils. Values are means ± SE; n = 50 mice per treatment group. *P < 0.05.

Fig. 7.

Early treatment with nintedanib reduced development of arthritis in SKG mice. A: daily oral gavage of mice with nintedanib or saline beginning 5 wk after zymosan injection. B and C: there was no significant change in weight or survival in nintedanib- compared with saline-treated mice. D: visual arthritis score declined significantly beginning 1 wk after nintedanib treatment compared with saline-treated animals. Values are means ± SE; n = 50 mice per treatment group. ***P < 0.001.

Fig. 8.

Early treatment with nintedanib reduced bone loss and remodeling. A and B: volume of hind talus and calcaneus bones was significantly greater after early treatment with nintedanib than saline. Data were obtained at week 10. Values are means ± SE; n = 10 mice per treatment group. *P < 0.05. C and D: bone surface as visualized by 3D reconstruction of talus and calcaneus bones from saline- and nintedanib-treated mice. Bones of saline-treated mice were irregular and deformed, suggesting extensive erosion, while bones of nintedanib-treated mice were almost normal. Representative saline- and nintedanib-treated mice had arthritis scores equal to the reported means of 4.5 and 1.7, respectively, at week 10 (n = 10; Fig. 4D). E: 3D reconstruction of talus and calcaneus bones from a naïve SKG mouse from a similar μ-CT image volume that was acquired at somewhat lower (18-μm) resolution.

Fig. 9.

Early treatment with nintedanib during disease development did not improve static lung compliance and did not change collagen levels in the lungs. A: quasi-static compliance was not altered in zymosan-injected, saline- compared with nintedanib-treated mice. B: hydroxyproline levels (a measure of collagen) were not significantly altered among all cohorts of mice. Data were obtained at week 10 and are shown as median ± interquartile range; n = 50 mice per treatment group and 10 naïve controls. *P < 0.05.

Fig. 10.

Early treatment with nintedanib did not alter histological appearance of the lungs. Hematoxylin-eosin (H&E)-stained lung sections from 3 representative animals demonstrated similar patchy areas of inflammation and disease. Picrosirius red (PSR)-stained serial sections showed similar amounts of collagen (red) in nintedanib- and saline-treated mice. Final magnification ×200, except insets (×300).

Fig. 11.

Early treatment with nintedanib did not significantly reduce the presence of α-smooth muscle actin (SMA)-positive (myo)fibroblasts, as shown by immunostaining and semiquantitative analysis of α-SMA-positive cells. Data were obtained at week 10 and are shown as median ± interquartile range (n = 10 mice per group). Final magnification ×400.

Fig. 12.

Early treatment with nintedanib did not alter bronchoalveloar lavage (BAL) cell counts or fluid composition. A: differential counts of BAL cells showed no difference in the number of macrophages (Mac), lymphocytes (Lymph), or neutrophils (Neuts). B, C, and E: there was no difference in levels of TNF-α, IL-4, or VCAM-1. D and F: KC and IL-1β were significantly reduced in nintedanib- compared with saline-treated mice. Data were obtained at week 10 and are shown as median ± interquartile range (n = 20 mice per treatment group and 10 naïve controls). *P < 0.05.

Fig. 13.

Inflammatory cell subsets in lungs of SKG mice after early treatment with nintedanib. Animals were treated with nintedanib or saline for 6 wk, and lung digests were analyzed by flow cytometry at week 10. A: total cells; B: CD4⁺ T cells; C: CD8⁺ T cells; D: B220⁺ B cells; E: tissue/interstitial macrophages; F: inflammatory macrophages; G: alveolar macrophages; H: monocytes; I: CD11b⁺ dendritic cells; J: CD103c⁺ dendritic cells; K: neutrophils. Values are means ± SE; n = 50 mice per treatment group. *P < 0.05; ***P < 0.001.