Abstract
INTRODUCTION: Randomized controlled trials (RCTs) are considered the gold standard for assessing treatment efficacy. However, sampling bias can affect the generalization of results to routine clinical practice. Here we assessed whether patients with lupus nephritis (LN) seen in routine clinical practice would have satisfied entry criteria to the major published RCTs in LN. METHODS: A systematic literature search from January 1974 to May 2015 was carried out, identifying all RCTs investigating LN induction treatment. Patients diagnosed with proliferative or membranous LN between 1995 and 2013 were identified from the Barts Lupus Centre database; baseline characteristics were compared with each RCT's entry criteria to assess hypothetical inclusion or exclusion. RESULTS: Of 363 articles, 33 RCTs met inclusion criteria. Of 137 patients newly diagnosed with LN (111 with proliferative/mixed proliferative and 26 with pure membranous LN), 32% would have been excluded from RCT entry (range 8%-73%). The main reasons for exclusion would have been too severe disease, too mild disease, or prior immunosuppressant use, which were exclusion criteria in 26, 20, and 22 RCTs, respectively. A total of 27 patients with LN (20%) were re-biopsied due to flare; 68% of these would have been ineligible to enter RCTs. CONCLUSION: Published RCTs do not truly reflect the heterogeneity of patients with LN in routine practice at our lupus center. The external validity of RCTs could be improved by including more representative patient cohorts. RCTs should be used as a guide but consideration should be given to similarities between individual patients and the characteristics of the trial cohorts before treatment decisions being made.