Table 5.
Peripheral blood biomarker prognostication of RA‐ILD
| Biomarker (citations) | Outcome |
|---|---|
| Autoantibodies | |
| Anti‐CCP (33, 40, 98) | Three studies, all negative. One retrospective cohort study found high titer ACPA was associated with acute exacerbations and/or all‐cause mortality in univariate analysis (OR 3.949, 95% CI: 1.119‐13.932, P = 0.033) but did not persist after multivariable adjustment (OR 0.722, 95% CI: 0.262‐1.989, P = 0.528). Two other case‐control studies with unadjusted analyses found no difference in anti‐CCP positivity and/or titer between patients with RA‐ILD who died vs. survived. |
| RF (40, 42, 98) | Three case‐control studies performed, with two negative and one positive. One found that over a 10‐year period, patients with RA‐ILD who died had a greater RF titer than those who survived (774.5 vs. 345.2 IU/mL, P = 0.001) but no difference in proportion RF‐positive (82.1% vs. 70.8%, P = 0.161). Multivariable logistic regression found no association between high RF titer (>3 times ULN) and mortality (OR 1.90, 95% CI: 0.56‐6.43). Another found no association between RF titer and mortality (HR 1.00, 95% CI: 0.999‐1.001, P = 0.31). Another of patients with RA‐ILD found that nonsurvivors had greater mean RF titer than survivors (349 vs. 86.1 IU/ml), P = 0.016), with no difference between groups in RF positivity (27 [nonsurvivors] vs. 39 [survivors], P = 0.673). In univariate analysis, RF >88 IU/ml was associated with mortality (HR 2.246, 95% CI: 1.066‐4.732, P = 0.033). |
| Lung epithelial–related proteins | |
| KL‐6 (38‐42) | Five studies, four positive, one negative. The only retrospective cohort design of the group was a study of RA‐UIP in which unadjusted analysis found an association with disease progression at 1 year (OR 1.001, 95% CI: 1.000‐1.002, P = 0.008) that was no longer significant after adjustment (OR 1.001, 95% CI: 1.000‐1.003, P = 0.077). KL‐6 levels were significantly greater during acute exacerbations compared with baseline disease (2147 vs. 794 U/ml, P < 0.001). Multivariate analysis revealed that high levels of KL‐6 (≥933 U/ml) were an independent predictor of mortality (HR 3.035, 95% CI: 1.168‐7.885, P = 0.023). The remaining four were case‐control designs. In one study of patients with RA‐ILD, KL‐6 was found to be a significant predictor of death, with an optimal cutoff of 685 U/ml (C‐index = 0.687, P = 0.004). High levels of KL‐6 (>685U/ml) were associated with mortality in multivariable analysis (HR 2.984, 95% CI: 1.227‐7.257, P = 0.016). When stratified by ILD pattern, the findings remained for the RA‐UIP group, but no association with mortality was found for the non‐UIP group. Another adjusted study found no association between KL‐6 and mortality (HR 1.003, 95% CI: 0.999‐1.006, P = 0.068). Other unadjusted studies found that baseline KL‐6 levels were significantly higher in those who experienced disease progression (1987 vs. 799 U/ml), P = 0.027) or did not survive a 1‐year follow‐up period (OR 1.016, 95% CI: 1.01‐1.02). |
| SP‐A (41) | Single retrospective cohort study of RA‐UIP. Unadjusted analysis found no association with SP‐A levels and disease progression at 1 year (OR 1.004, P = 0.418). Levels were similar during acute exacerbations compared with baseline disease (P = 0.265). |
| SP‐D (38, 42) | Two case‐control studies, both negative. In an adjusted analysis, there was no association with SP‐D and mortality (HR 1.001, P = 0.203). Another unadjusted analysis found no association between SP‐D concentrations and disease progression over a mean follow‐up of 3 years. |
| Genetic polymorphisms | |
| MUC5B mutation (33) | Single retrospective cohort study found that MUC5B mutations (rare [MAF <0.01] and deleterious variants not including rs35705950) were associated with acute exacerbations and/or all‐cause mortality in a univariate LR (OR 2.308 P = 0.043), though this was not statistically significant after multivariable adjustment (OR 2.312, 95% CI: 0.951‐5.620, P = 0.065). |
| Cytokines and chemokines | |
| CCL18 (PARC) (38) | Single case‐control study of RA‐ILD. Unadjusted analyses found no relationship between CCL18 concentrations and disease progression at mean follow‐up of 3 years. Data not reported. |
| IL‐32 (41) | Single retrospective cohort study of RA‐UIP. Unadjusted analysis found no association with IL‐32 levels and disease progression at 1 year (OR 0.999, P = 0.218). Levels were no different during acute exacerbations compared with baseline disease (P = 0.461). |
| IL‐6 (41) | Single retrospective cohort study of RA‐UIP. Adjusted analysis revealed an association with IL‐6 levels and disease progression at 1 year (OR 1.040, P = 0.039). IL‐6 levels tended to be higher during acute exacerbations, though not statistically significant (P = 0.068). |
| Extracellular matrix proteins | |
| MMP‐7 (41) | Single retrospective cohort study of RA‐UIP. Unadjusted analysis found no association between MMP‐7 levels and disease progression at 1 year (OR 1.099, P = 0.394). Levels were no different during acute exacerbations compared with baseline disease (P = 0.580). |
| Others | |
| CRP (40, 41, 42, 97, 98) | Five studies, with two positive and three negative. In a case‐control study of RA‐ILD, baseline CRP levels were higher in those who had died vs. those who had survived over a 10‐year period (37.2 vs. 25.1 mg/l, P = 0.009), though multivariable regression found no association with mortality (OR 1.34, 95% CI: 0.95‐1.88). In a retrospective cohort study, multivariable analysis found higher CRP concentrations were associated with fatal outcomes (OR 1.072, 95% CI: 1.000‐1.150, P = 0.049). An adjusted case‐control study found no association between CRP and mortality (HR 1.027, 95% CI: 0.954‐1.094, P = 0.458). In a retrospective cohort study of RA‐UIP, unadjusted analysis found higher CRP levels in patients whose disease progressed at 1 year compared with those whose disease did not progress (5.7 vs. 1.3 [units not reported], P = 0.013). Lastly, in a case‐control study of patients with RA‐ILD, there was no difference in mean CRP titers between survivors and nonsurvivors (3.9 [non‐survivors] vs. 2.7 mg/dl, P = 0.303), and univariate analysis found no association with mortality (HR 1.030, 95% CI: 0.972‐1.092, P = 0.322). |
| DKK‐1 (97) | Single retrospective cohort study performed multivariable analysis, finding that DKK‐1 was associated with fatal outcomes (OR 15.764, 95% CI: 1.086‐228.843, P = 0.043). Median survival was longer for DKK‐1‐negative patients (5.1 vs. 2.7 years, P = 0.041). |
| ESR (41, 98) | Two studies. In one case‐control study of RA‐ILD, baseline ESR was higher in those who had died vs. those who had survived over a 10‐year period (58 vs. 42.2 mm/h, P = 0.008), though multivariable analysis of mortality was null (OR 1.00, 95% CI: 0.97‐1.02). In another retrospective cohort study of RA‐UIP, unadjusted analysis found higher ESR in patients whose disease progressed at 1 year compared with those whose disease did not progress (74 vs. 41 mm/hour, P = 0.001). |
| Leukocyte indices (42) | Single adjusted case‐control study found that higher monocyte count (HR 1.020, 95% CI: 1.004‐1.035, P = 0.018) and neutrophil count (HR 1.001, 95% CI: 1.001‐1.117, P = 0.026) were associated with mortality in RA‐ILD. No association was observed with lymphocyte count (HR 1.002, P = 0.385). Patients with high monocyte and neutrophil counts had worse survival than those with no (P < 0.001) or just one high lineage (P = 0.001). |
Abbreviations: ACPA, anti–citrullinated protein antibody; CCL, chemokine ligand; CCP, cyclic citrullinated peptide; CI, confidence interval; CRP, C‐reactive protein; DKK, dickkopf; ESR, erythrocyte sedimentation rate; HR, hazard ratio; IL, interleukin; ILD, interstitial lung disease; IU, international unit; KL, Krebs von den Lungen; LR, logistic regression; MAF, minor allele frequency; ml, milliliter; mm, millimeter; MMP, matrix metallopeptidase; MUC5B, mucin5B; OR, odds ratio; PARC, pulmonary and activation‐regulated chemokine; RA, rheumatoid arthritis; RF, rheumatoid factor; SP‐A, surfactant protein A; SP‐D, lung epithelial–derived surfactant protein D; U, unit; UIP, usual interstitial pneumonia; ULN, upper limit of normal.