Autoimmune Diseases Unit, Department of Internal Medicine, Bellvitge University Hospital, Bellvitge Biomedical Research Institute-IDIBELL, University of Barcelona, Barcelona, Spain.
Autoimmune Diseases Unit, Department of Internal Medicine, Bellvitge University Hospital, Bellvitge Biomedical Research Institute-IDIBELL, University of Barcelona, Barcelona, Spain; Faculty of Medicine and Health Sciences, Universitat Internacional de Catalunya, Barcelona, Spain; Evaluation of Health Determinants and Health Policies Group, Hestia Chair in Integrated Health and Social Care, Barcelona, Spain.
Eur J Intern Med. 2020 Jul;77:59-65. doi: 10.1016/j.ejim.2020.04.024. Epub 2020 Apr 21.
Sarcoidosis is a heterogeneous disease with high variability in natural history and clinical spectrum. The study aimed to reveal different clinical phenotypes of patients with similar characteristics and prognosis.
Cluster analysis including 26 phenotypic variables was performed in a large cohort of 694 sarcoidosis patients, collected and followed-up from 1976 to 2018 at Bellvitge University Hospital, Barcelona, Spain.
Six homogeneous groups were identified after cluster analysis: C1 (n=47; 6.8%), C2 (n=85; 12.2%), C3 (n=153; 22%), C4 (n=29; 4.2%), C5 (n=168; 24.2%), and C6 (n=212; 30.5%). Presence of bilateral hilar lymphadenopathy (BHL) ranged from 65.5% (C4) to 97.9% (C1). Patients with Löfgren syndrome (LS) were distributed across 3 phenotypes (C1, C2, and C3). In contrast, phenotypes with pulmonary (PS) and/or extrapulmonary sarcoidosis (EPS) were represented by groups C4 (PS 100% with no EPS), C5 (PS 88.7% plus EPS), and C6 (EPS). EPS was concentrated in groups C5 (skin lesions, peripheral and abdominal lymph nodes, and hepatosplenic involvement) and C6 (skin lesions, peripheral lymph nodes, and neurological and ocular involvement). Unlike patients from LS groups, most patients with PS and/or EPS were treated with immunosuppressive therapy, and evolved to chronicity in higher proportion. Finally, the cluster model worked moderately well as a predictive model of chronicity (AUC=0.705).
Cluster analysis identified 6 different clinical patterns with similar phenotypic variables and predicted chronicity in our large cohort of patients with sarcoidosis. Classification of sarcoidosis into phenotypes with prognostic value may help physicians to improve the efficacy of clinical decisions.
结节病是一种异质性疾病，其自然病史和临床表现谱差异很大。本研究旨在揭示具有相似特征和预后的患者的不同临床表型。
对西班牙巴塞罗那贝尔维奇大学医院 1976 年至 2018 年间收集和随访的 694 例结节病患者的 26 个表型变量进行聚类分析。
聚类分析后确定了 6 个同质组：C1（n=47；6.8%）、C2（n=85；12.2%）、C3（n=153；22%）、C4（n=29；4.2%）、C5（n=168；24.2%）和 C6（n=212；30.5%）。双侧肺门淋巴结肿大（BHL）的存在率从 C4（65.5%）到 C1（97.9%）不等。存在 Löfgren 综合征（LS）的患者分布在 3 种表型（C1、C2 和 C3）中。相比之下，有肺部（PS）和/或肺外结节病（EPS）的表型由 C4（PS 为 100%，无 EPS）、C5（PS 为 88.7%，加 EPS）和 C6（EPS）代表。EPS 集中在 C5（皮肤病变、外周和腹部淋巴结以及肝脾受累）和 C6（皮肤病变、外周淋巴结以及神经和眼部受累）。与 LS 组的患者不同，大多数 PS 和/或 EPS 患者接受了免疫抑制治疗，并且更有可能发展为慢性。最后，该聚类模型作为我们大型结节病患者慢性的预测模型效果中等（AUC=0.705）。
聚类分析确定了 6 种具有相似表型变量的不同临床模式，并预测了我们大型结节病患者的慢性。将结节病分类为具有预后价值的表型可能有助于医生改善临床决策的效果。
