Objectives Awareness of paediatric-specific predictors of damage in childhood lupus is needed to inform mitigation measures. The objective of this study was to ascertain how clinical and demographic variables correlate with damage accrual and identify predictors of damage. Methods This analysis included UK JSLE Cohort Study participants. Univariable and multivariable Prentice-Williams-Peterson models investigated how demographic and clinical factors influenced the hazards of new damage. Analyses were performed across the entire cohort, in patients with minimal disease activity marked by a time-adjusted average SLEDAI-2K score (AMS) of ≤2, in patients with low activity (AMS of ≤4), patients with moderate-to-high activity (AMS of >4) and patients with no CS use. Results Within the entire cohort (n = 430), factors associated with damage included: any methylprednisolone [hazard ratio, HR 2.20 (CI 1.33–3.62)], time-adjusted mean Physician’s Global Assessment (PGA) [HR 2.87 (CI 1.48–5.56)] and AMS score [HR 1.13 (CI 1.03–1.24), all P < 0.05]. Within the low activity subgroup, any methylprednisolone [HR 2.61 (CI 1.04–6.53)] and time-adjusted mean PGA [HR 3.41 (CI 1.52–7.76)] were associated with damage (both P < 0.05). Within the moderate-to-high activity subgroup, any methylprednisolone [HR 2.29 (CI 1.31–4.00)], time-adjusted mean PGA [HR 2.66, (CI 1.20–5.87)] and AMS score [HR 1.15 (CI 1.03–1.29)] were predictive of damage (all P < 0.05). Baseline organ damage was predictive of subsequent damage accrual in the minimal disease activity subgroup [HR 1.33 (CI 1.78–8.08)] and the no CSs subgroup [HR 3.64 (CI 1.83–7.24), both P < 0.005]. Conclusion Disease activity levels (AMS/PGA) and proxy indicators (methylprednisolone exposure, baseline damage) were found to be key predictors of damage accrual. This highlights the importance of practical strategies, such as treat-to-target, for reducing disease activity and long-term treatment toxicity.

中文翻译：

---

儿童狼疮器官损伤的促成因素：皮质类固醇的使用和疾病活动度


目标 需要了解儿童狼疮损伤的儿科特异性预测因子，以便为缓解措施提供信息。本研究的目的是确定临床和人口统计学变量如何与损伤累积相关，并确定损伤的预测因子。方法 该分析包括英国 JSLE 队列研究参与者。单变量和多变量 Prentice-Williams-Peterson 模型研究了人口统计学和临床因素如何影响新损伤的危害。对整个队列进行了分析，在时间调整后平均 SLEDAI-2K 评分 （AMS） 为 ≤2 的轻微疾病活动患者中，在低活动度患者 （AMS 为 ≤4） 、中度至高度活动患者 （AMS 为 >4） 和未使用 CS 的患者。结果在整个队列 （n = 430） 中，与损伤相关的因素包括：任何甲泼尼龙 [风险比，HR 2.20 （CI 1.33-3.62）]、时间调整平均医师整体评估 （PGA） [HR 2.87 （CI 1.48-5.56）] 和 AMS 评分 [HR 1.13 （CI 1.03-1.24），均 P < 0.05]。在低活性亚组中，任何甲泼尼龙 [HR 2.61 （CI 1.04–6.53）] 和时间调整平均 PGA [HR 3.41 （CI 1.52–7.76）] 都与损伤相关 （均 P < 0.05）。在中度至高活性亚组中，任何甲泼尼龙 [HR 2.29 （CI 1.31-4.00）]、时间调整平均 PGA [HR 2.66，（CI 1.20-5.87）] 和 AMS 评分 [HR 1.15 （CI 1.03-1.29）] 均可预测损伤 （均 P < 0.05）。基线器官损伤可预测微小病活动亚组 [HR 1.33 （CI 1.78–8.08）] 和无 CSs 亚组 [HR 3.64 （CI 1.83–7.24），均 P < 0.005]。 结论 疾病活动水平 （AMS/PGA） 和代理指标 （甲泼尼龙暴露、基线损伤） 是损伤累积的关键预测因子。这凸显了实用策略（例如按目标治疗）对于减少疾病活动和长期治疗毒性的重要性。