Purpose Undifferentiated connective tissue disease (UCTD) is characterised by symptoms and immunology suggestive of a systemic autoimmune diseases that are not sufficient to diagnose a defined connective tissue disease (CTD). Approximately one third of patients with UCTD will develop a defined CTD, most commonly systemic lupus erythematosus (SLE). The identification of profiles predictive of progression has clinical, therapeutic and prognostic implications. The aim of this systematic review and meta-analysis was to identify whether demographics, clinical and immunological parameters, and novel biomarkers can predict progression from UCTD to SLE.

Methods MEDLINE, EMBASE and the Cochrane Central Register of Randomized Controlled Trials were systematically searched from inception until February 2021. Abstracts and full-text manuscripts were screened by two reviewers. Publications were included if they included at least 20 UCTD patients, a minimum of six months of follow up, and provided data on at least one risk factor for developing a defined CTD. QUIPS tool was used to assess risk of bias and GRADE approach for grading the quality of the evidence. For predictors reported in at least two studies, meta-analysis was carried out using random-effects models to pool effect sizes. Heterogeneity was assessed using the standard chi-squared test and I2 statistic. Influence analysis was carried out to identify outlier studies with extreme effect sizes. Publication bias was assessed using visual inspection of funnel plots and Egger’s test. The study is registered with PROSPERO (ID: CRD42021237725)

Results A total of 3871 articles were initially identified via the literature search; 2559 abstracts were screened and 196 full-texts were reviewed for eligibility. Forty-five studies were included in the systematic review, and thirty-three in the meta-analysis. Key results are summarised in Table 1. The predictors for progression to SLE with the highest quality of evidence included those with younger age, serositis or the presence of anti-dsDNA antibodies. Other clinical predictors included renal involvement, mucocutaneous involvement (malar rash, alopecia, photosensitivity), thrombocytopenia and a positive Coombs’ test. Immunological parameters associated with progression included a homogenous pattern of ANA, hypocomplementaemia, positive anti-Smith, anti-cardiolipin and/or anti-SSA antibodies. No novel biomarkers were included in the meta-analysis. HLA antigens, T-regulatory cell shift, and complement activation products were reported as potential predictors in single studies. All studies were rated as high or moderate risk of bias. Significant publication bias was not observed.

Conclusions Demographic, clinical and immunological parameters may predict which patients with UCTD progress to SLE. The baseline predictors with the highest quality of evidence included those with younger age, serositis or presence of anti-dsDNA antibodies. Further work is required to investigate the role of novel biomarkers in predicting progression from UCTD to SLE. High study heterogeneity, risk of bias and low quality of evidence limits the extrapolation of these results.