Can We Predict Remission in Children with Juvenile Idiopathic Arthritis?

Abstract

Background: Long-term consequences of active disease in juvenile idiopathic arthritis (JIA) include persistent pain, disability and potential joint replacement surgery. The aims of treatment are therefore clinically inactive disease (CID) and, ideally, sustained disease remission. This thesis set out to understand CID and remission in JIA: how they are defined, how commonly they are achieved, long-term outcomes following their achievement and whether they are associated with factors early in disease. Methods: The setting for this thesis was the UK Childhood Arthritis Prospective Study (CAPS), the largest multicentre, prospective inception cohort of JIA globally. Children and young people (CYP) were selected for each paper based on their dates of recruitment and categories of disease. At one year following initial presentation to paediatric rheumatology, CYP were classified as to whether they had fulfilled published criteria for CID. Initial analyses explored whether different definitions for CID identified the same groups of CYP. Outcomes to five years were then compared between those achieving all, some or none of the criteria for CID using multivariable, multilevel logistic (absence of limited joints), linear (quality of life) and zero-inflated negative binomial (functional ability) regression models. Finally, risk factors for remission (CID maintained over two annual follow-ups) were explored using multivariable logistic regression models. Throughout, multiple imputation under various assumptions was implemented for missing data. Results: The majority of CYP in CAPS were female (65%) and had oligoarthritis (50%). Median age at initial presentation was eight years (IQR 4, 12). At one year, fewer than 50% of CYP had achieved CID according to published definitions. There was poor overlap (44%) between groups of CYP identified by two validated definitions, whose main difference was the inclusion or exclusion of patient-reported wellbeing. The odds of no limited joints in the long-term did not differ between CYP fulfilling either definition of CID. However, CYP who achieved CID on scores which included wellbeing had superior long-term function (OR for no disability: 2.5, 95% CI 1.8, 3.5) and quality of life (β: 3.9, 95% CI 1.6, 6.2) to those who either did not achieve CID or only achieved it using inflammatory criteria (i.e. had persistent symptoms despite the absence of inflammation). Finally, there were few factors at initial presentation which were associated with remission. However, greater improvements in both physician and patient-reported variables over the first year following initial presentation were associated with higher odds of remission in the first three years. Conclusions: The disease burden in JIA remains high with over 50% of CYP not achieving CID within the first year of disease. Current definitions of CID available for use in clinical practice as potential treatment targets do not classify the same groups of CYP and are associated with different long-term outcomes. Further study is required to define the best outcome measures for CYP with JIA.

Date of Award	31 Dec 2018
Original language	English
Awarding Institution	The University of Manchester
Supervisor	Wendy Thomson (Supervisor), Kimme Hyrich (Supervisor) & Suzanne Verstappen (Supervisor)