Background In our previous work, we developed an RT-qPCR based platform to identify B cell lineage biomarkers (BM) of clinical response to RTX in rheumatoid arthritis (RA). Fc receptor-like protein 5 (FCRL5), preferentially expressed on naïve and memory B cells, was put forward as candidate BM.
Objectives The goal of this study was to determine, whether FCRL5 mRNA expression at baseline could serve as a predictive BM for achieving clinical response to RTX in patients with GPA or MPA (RAVE trial).
Methods FCRL5 mRNA expression analysis was carried out on whole blood PAX gene samples from the RAVE trial (randomized, double-blind noninferiority trial comparing RTX with cyclophosphamide (CYC) followed by azathioprine for remission induction in GPA or MPA). Clinical and efficacy data was available for all patients. Primary outcome variable was complete remission (CR) defined as no disease activity (BVAS/WG=0) off glucocorticosteroids at 6 months.
Results FCRL5 mRNA gene expression data was available for 188 patients (97 RTX and 91 CYC). In the RTX arm, baseline FCRL5 expression was significantly higher in patients who achieved CR at 6 months compared to those who did not (median 0.005 mRNA expression unit, IQR 0.003–0.012 vs. median 0.004, IQR 0.002-0.006; p=0.026) No difference was observed in the CYC group. For the purpose of binary analysis; patients with baseline FCRL5 expression threshold of ≥0.01 were termed FCRL5high, all others FCRL5low. At 6 months, 84% of FCRL5high patients in the RTX arm achieved CR as compared to 57% of FCRL5low patients (p=0.016). There was no difference in CR in FCRL5 subgroups in the CYC arm (56 vs 54%, p=0.6). In FCRL5high patients, there was a 30 percent difference in CR between RTX and CYC arms, as compared to 11 percent difference in all comers.
At 12 months, 68% of FCRL5high patients in the RTX arm achieved CR (versus 40% of FCRL5low; p=0.02) compared to 29% of FCRL5high patients in the CYC arm (vs 44% of FCRL5low, P=0.23), equating to a 39 percent difference in the RTX arm. That trend was maintained at 18months (RTX arm: CR 68% in FCRL5high vs 29% in FCRL5low, p=0.0009; CYC arm: CR 24%in FCRL5high vs 38% in FCRL5low, p=0.3).
A multivariate logistic regression analysis including pre-defined variables known to be associated with clinical response to RTX, such as relapsed disease (vs. newly diagnosed), PR3 serology (vs MPO) and type of GPA vasculitis (vs MPA) was carried out, which confirmed FCRL5 as an independent biomarker for response for RTX but not CYC arm (OR for CR in RTX arm 5.67, 95%CI 2.04-17.4, p adjusted=0.0007; OR for CR in CYC arm 0.4, 95%CI 0.1-1.3, p=0.13).
Conclusions In the RAVE trial patients who achieved CR at 6 months had significantly higher baseline FCRL5 expression in the RTX but not CYC arms. In this exploratory biomarker analysis, higher FCRL5 expression has been shown to predict CR at 6, 12 and 18 months in patients receiving RTX but not CYC, implying a predictive rather than prognostic role, independent of other clinical and serological variables.
Disclosure of Interest None declared