Background Human herpes virus-8 (HHV8) negative/idiopathic multicentric Castleman disease (iMCD) is a rare hematological disorder with one FDA-approved treatment and no early indicators of patient response after treatment is commenced. Clinical onset can be sudden and severe, involving polyclonal lymphoproliferation with characteristic dysmorphic germinal centers, constitutional symptoms, systemic inflammation and life-threatening cytokine storm-driven multi-organ failure. CXCL13, a key regulator of lymph node germinal center development, was recently found to be the most elevated cytokine in iMCD flare, but the clinical significance of this finding is not yet clear. Interleukin-6 (IL-6) is the known driver of pathogenesis in a portion of patients and the target of the only FDA-approved treatment, siltuximab. In the Phase II study of siltuximab (NCT01024036), one-third of patients met response criteria, which were assessed after a minimum of 48 weeks. Early indicators of response to siltuximab are urgently needed to inform clinicians about the likelihood of patient response to therapy, adjust treatments if needed, and identify novel therapeutic targets for siltuximab non-responders. Methods Clinical data and serum samples were collected as part of NCT01024036. We measured serum protein analytes in the 52 subjects who were treated with anti-IL6 therapy, as well as the 26 patients in the control arm, at day 1, day 8, and day 64 of therapy (infusions administered every 21 days). Serum samples from 44 healthy donors were also analyzed. Of the 1,305 analytes measured using SomaLogic SOMAscan, 1,178 passed QC and were included in analyses. Each analyte was log2 transformed and capped at the 2.5th and 97.5th percentiles. Response to anti-IL6 therapy was determined by independent review in NCT01024036. Data processing was performed using the Medidata Rave Omics machine learning platform and R v3.4.4. Linear mixed effects models were used to detect whether kinetic changes in protein expression were associated with anti-IL6 response. Upon running the full model with the selected covariates, the p-values of the interaction between time point and response were used to test for differences between responders and non-responders. A separate model was fitted using each protein, and False Discovery Rates (FDR) were estimated by the Benjamini-Hochberg method with alpha < 0.05. Results Seven days after siltuximab was first administered (day 8), 9 proteins were significantly different between responders and non-responders: IgA, BCMA, NPS-PLA2, ART, IL-18 BPa, CD5L, b2-Microglobulin, CXCL13, and NRP1. All 9 of these proteins were significantly decreased in responders compared to non-responders. At day 64, the number of significantly different proteins increased to 121, including 8 of the 9 proteins from day 8; NPS-PLA2 did not achieve significance at day 64. This result indicates that there may be early indicators of response in serum as early as day 8. Given that CXCL13 was recently discovered as a key cytokine in iMCD, the early and significant decline of CXCL13 in responders versus non-responders was highly notable (Day 8: FDR = 0.02, Day 64: FDR = 0.005). Prior to treatment, CXCL13 was significantly higher in this cohort of iMCD patients than in a group of age-matched healthy donors (p = 8.19e-09). By day 64, CXCL13 levels in siltuximab responders decreased to levels approaching the healthy donor range but remained elevated in non-responders and placebo patients. Conclusions This analysis represents the first use of high-quality serum proteomics data to study early indicators of response to treatment in a rare hyperinflammatory, lymphoproliferative disorder. The decline in CXCL13 levels in responders and continued elevation in non-responders suggests that CXCL13 is downstream of IL-6 in responders and independent of IL-6 signaling in non-responders. CXCL13, along with several other proteins that demonstrated significant decline by day 8 including IgA and beta2-microglobulin, can be routinely measured and could serve as a panel that indicates the likelihood of response soon after commencing therapy, if validated in a separate cohort. These proteins could also provide a more continuous scale of response than traditional outcome measures. Given that iMCD may have a sudden and severe onset, early indicators of response to anti-IL6 therapy are critical for timely treatment administration. Figure Disclosures Katz: Medidata Solutions: Employment. Nabel:Thermo Fisher: Patents & Royalties: royalty income. Diamond:Brigham and Women's Hospital and Harvard Medical School's Department of Medicine, Division of Genetics, Genomes2People Translational Genomics Research Program: Consultancy. Karvir:Medidata Solutions: Employment. Zhang:Medidata Solutions: Employment. Reddy:Janssen Pharmaceuticals: Employment, Equity Ownership. Guilfoyle:Janssen Pharmaceuticals: Employment, Equity Ownership. Tendler:Janssen Pharmaceuticals: Employment, Equity Ownership. van Rhee:Takeda: Consultancy; Sanofi Genzyme: Consultancy; Castleman Disease Collaborative Network: Consultancy; EUSA: Consultancy; Adicet Bio: Consultancy; Kite Pharma: Consultancy; Karyopharm Therapeutics: Consultancy. Beineke:Medidata Solutions: Employment. Oromendia:Medidata Solutions: Employment, Equity Ownership. Fajgenbaum:Janssen Pharmaceuticals: Research Funding.
Read full abstract