Lupus Science & Medicine
demonstrated by our group 12 13 and others 14 32–40 to reli- ably detect active LN and to correlate with histological disease activity in the KB. Importantly, UBs, evaluated as single and multimarker urinary panels, have been shown to have prognostic utility in evaluating the response to therapy in paediatric 41 42 and adult LN cohorts. 13 38 43 Parodis et al , found that patients with a higher urinary sVCAM/creatinine ratio measured at the time of the LN flare were at an increased risk of a 25% decline in eGFR. 44 In our results, sVCAM was associated with a clin- ically significant decline in eGFR in the whole cohort but not in the cohort of patients who achieved PERR. However, Parodis et al 44 measured sVCAM at the time of the flare, which could explain the discrepancy in results. Additionally, our PERR subcohort had a lower number of patients compared with the study from Parodis et al 44 (47 vs 111), which may have reduced our power to detect this association. The originality of our study lies in the fact that we measured UBs 2 years after an LN flare and evaluated their association with adverse long-term renal outcomes in patients who had achieved PERR, exploring the potential ability of these UBs to detect subclinical kidney inflamma- tion. This approach is particularly relevant given recent studies showing a disconnect between proteinuria and histological activity, 45 as well as evidence suggesting that ongoing histological activity is associated with a higher risk of future flares. 8 9 Our results indicate that due to the high sensitivities, NPVs and −LRs associated with having elevated MCP-1, patients with MCP-1 levels <28.1 (pg/mmol) are at a low risk of experiencing a subsequent LN flare. Addition- ally, all combinations, except for CD163 + MCP-1 eleva- tions, exhibited excellent specificities, good PPVs and +LR above 10. Therefore, patients with these biomarker combinations would be at high risk for an LN flare within the next 3 years. In terms of eGFR decline, adiponectin demonstrated the highest predictive accuracy. In addition, all of the combinations of biomarkers had excellent specificities and NPVs, with most achieving excellent +LRs, except for the combinations of CD163 + PF4, MCP-1 + PF4and PF4 + anti-dsDNA abs, which had +LRs below 10. These find- ings suggest that CD163, MCP-1 and adiponectin are the most informative UBs for identifying patients at risk of future LN flares and clinically meaningful eGFR decline. Incorporating these markers alongside conventional biomarkers such as anti-dsDNA abs, C3 and proteinuria could enhance the ability to accurately identify individ- uals at risk for poorer long-term renal outcomes, avoiding the need for repeat KBs. While our primary analysis focused on the association between UBs and renal outcomes in patients who achieved PERR, we also analysed the whole cohort. Since only patients who achieved a good clinical response (PERR group) are at risk of true relapse, flare analysis is more meaningful in this group. Nevertheless, both the PERR and whole cohorts provide valuable insights into renal
disease progression. Although persistent proteinuria is a well-established predictor of poor outcomes, 46 47 our find- ings demonstrate that UBs, CD163, MCP-1, adiponectin, sVCAM and PF4 are independently associated with a ≥ 30% decline in eGFR in the whole cohort and CD163, MCP-1, adiponectin and PF4 in the cohort who achieved PERR. Importantly, these associations remain signifi- cant even after adjusting for proteinuria in the whole cohort, suggesting that these biomarkers offer additional, complementary prognostic value. Our findings further support prior research linking monocyte/macrophage infiltration with the severity of LN. 48 49 Infiltration of CD163 + macrophages is linked to histopathological inflammation in LN. 37 Wei Chen et al also found CD163 + dendritic cells enriched in LN kidneys, correlating with disease severity and promoting T helper cell 1/T helper cell 17 polarisation. 50 MCP-1, produced by renal cells and leukocytes in response to type I inter- ferons and proinflammatory cytokines, regulates mono- cyte/macrophage recruitment and contributes to renal injury. 51 52 sVCAM and PF4 are key mediators involved in monocyte/macrophage recruitment. Higher levels of serum adiponectin have been found in ALN. 53 54 The pathogenic role of adiponectin in LN is still unclear, with several studies supporting an anti-inflammatory 55 and even reno-protective 56 action, whereas other studies have highlighted the proinflammatory properties of the low molecular weight isoform of adiponectin. 57–60 These find- ings raise the possibility that under inflammatory condi- tions, a shift in the predominant isoform of adiponectin could occur, converting it from anti-inflammatory to proinflammatory. Our study has several limitations. The lack of paired KB at the 2-year mark limits our ability to determine whether elevated UBs reflect ongoing histological activity. Also, UBs were measured at a single time point, preventing eval- uation of their longitudinal dynamics. As most patients in this cohort had proliferative or mixed LN, our findings primarily reflect biomarker performance in this group. The small number of non-proliferative cases limited subgroup analysis and prevented inclusion of histolog- ical class in multivariable models. As a single-centre study with a small sample size and few events, our findings may have limited generalisability and should be validated in independent cohorts. We did not include social determi- nants of health in our analysis, which may limit the inter- pretation of our findings given their known impact on LN outcomes. Finally, as treatments have shifted toward upfront multitarget therapies, the generalisability of our real-world findings may be limited. 18 61 Nonetheless, the link between persistent kidney inflammation, indicated by elevated UBs after treatment and poorer prognosis, is likely to remain applicable despite evolving treatment approaches. This study has several strengths, including a racially diverse cohort with prospective follow-up and regular monitoring, enabling comprehensive assessment of renal outcomes. The long-term follow-up also allowed
10
Baker R, et al . Lupus Science & Medicine 2026; 13 :e001724. doi:10.1136/lupus-2025-001724
Powered by FlippingBook