The Ire of IRE1α: Overexpression of IRE1α at Myeloma Diagnosis Is Associated with Decreased Survival While Downregulation of IRE1α Expression Is Predictive of Therapy Resistance

Introduction: Inositol-requiring enzyme 1α (IRE1α) is one of three principal transmembrane sensors in the endoplasmic reticulum (ER) involved in regulating the unfolded protein response (UPR). It splices X-box protein 1 (XBP1) mRNA yielding a transcription factor intimately involved in the plasma cell differentiation and proliferation. The IRE1α-XBP1 axis exhibits positive feedback functions that promote the pro-survival UPR and allows plasma cells to survive the high biosynthetic burden of immunoglobulin production, although IRE1α also has functions that promote cell death depending on the chronicity and severity of ER stress. Functional IRE1α has been identified as critical for in vitro cytotoxicity of bortezomib by Leung-Hagesteijn et al. (Cancer Cell 2013), with loss of IRE1α leading to in vitro resistance to proteasome inhibition. However, Mimura et al (Blood 2012) identified in vitro synergy with IRE1α-inhibition and bortezomib, and Papandeou et al. (Blood 2011) identified in vitro and in vivo anti-myeloma activity of a different IRE1α endonuclease inhibitor. The clinical implications of IRE1α are still poorly understood. In this study, we sought to identify associations between IRE1α mRNA expression (the level at which it is principally regulated) with treatment resistance and patient outcomes.

Methods: We performed a secondary analysis of data from the MMRF CoMMpass study (IA14). In the CoMMpass study, RNAseq on CD138-enriched bone marrow cells was performed using Illumina TruSeq RNA library kits. Cox Regression analysis was used to compare the association of IRE1α expression with progression-free and overall survival. To determine if IRE1α expression is modulated following disease progression, we also identified patients who had pre- and post-treatment RNAseq data. We analyzed expression of IRE1α from the pre- and post-treatment specimens using paired T-tests comparing patients who had relapsed disease (relapse following completion of therapy) and refractory disease (progression on or within 60 days of completing therapy).

Results: We included 768 patients who had IRE1α expression data at myeloma diagnosis. The median age was 63 (range 27-93) and 50% were male. Thirty percent of patients (n = 231) were considered to have high IRE1a expression at myeloma diagnosis. High IRE1α expression was associated with worse clinical outcome. Those with high expression had a 37% (aHR 1.37; 95% CI 1.10-1.72; p = 0.006) increase hazard for progression and a 55% (aHR 1.55; 95% CI 1.13-2.15; p = 0.008) increase hazard for death.

We then identified 46 patients who had pre- and post-treatment RNAseq data. Twenty-two had relapsed disease while off of treatment and 24 had developed refractory disease while on treatment. Patients who had refractory disease were noted to have a significant decrease in the expression of IRE1α (p = 0.008). This was similar when we performed subset analysis of patients who were refractory to PI-based treatments (n=8; p = 0.044) and IMID-based treatments (n = 13; p = 0.054). In contrast, there was not a significant difference between pre-treatment and post-treatment IRE1α in relapsed disease (p = 0.312). Expression of other UPR elements, specifically PERK and XBP1, was not significantly downregulated in relapsed or refractory disease.

Conclusion: Our study demonstrates that high expression of IRE1α at diagnosis is associated with worse outcomes, but that resistance to therapy is associated with a significant decrease in IRE1α expression. While this seems counterintuitive, it highlights the complex nature of the UPR and the roles of IRE1α. More aggressive/proliferative disease relies heavily on an adapted pro-survival UPR, highlighted by increased expression of IRE1α de novo. However, IRE1α has previously been demonstrated to be required for anti-myeloma activity of novel therapies; potentially either by clonal selection, dynamic alterations in UPR signaling, or epigenetic modifications, treatment-refractory myeloma cells downregulate the expression of IRE1α, a phenomenon not seen in those who completed treatment but subsequently relapsed. IRE1α inhibition could initially potentiate the clinical effects of novel therapies in a higher-risk population, but may also promote later resistance to them. Importantly, IRE1α could serve as a prognostic biomarker for risk-stratification at diagnosis as well as one predictive of impending resistance to therapy.