MCT1 As Molecularly Validated Predictive Marker for Lenalidomide-Maintenance Therapy in Multiple Myeloma

Introduction:

Lenalidomide-based maintenance therapy is the currently approved standard of care for multiple myeloma (MM) patients after high-dose melphalan and autologous stem cell transplantation (HD-Mel), which significantly prolongs progression-free (PFS) and overall survival (OS). For patients with del17p bortezomib based maintenance treatment is considered overcoming adverse prognosis of this aberration. Predictive markers of response to lenalidomide maintenance have remained elusive. We have previously shown that IMiDs exert their anti-MM activity via destabilization of MCT1 and CD147 and combined overexpression reduces response to lenalidomide-treatment in vitro and in an in vivo MM xenograft model (Eichner et al. Nature Medicine 2016).

Methods:

CD138-purified myeloma cell samples of 654 patients receiving high-dose melphalan therapy and autologous stem cell transplantation and either bortezomib (n=101), thalidomide (n=98) or lenalidomide (n=455) maintenance treatment were assessed by gene expression profiling (GEP) using U133 2.0 plus DNA microarrays, 316 by RNA-sequencing (RNA-seq). Expression of CD147 and MCT1 were assessed and correlated with PFS and OS data. Gene expression based risk scores, including UAMS70-gene, Rs-score and gene expression based proliferation index were assessed alongside routine iFISH-analysis. Survival curves and median time to progression were computed with nonparametric survival estimates for censored data using the Kaplan-Meier method. Difference between the curves were tested using the G-rho Log-rank test. Landmark analysis was performed by defining an alternative start point (landmark) at 12 months.

In vitro, CD147 and MCT1 were lentivirally overexpressed in MM1S cells, which were subjected to lenalidomide or bortezomib treatment and proliferation analysis. Xenografted MM-tumors were followed by ¹⁸FDG-PET and analyzed by immunohistochemistry.

Results:

Patients with high gene expression levels of MCT1 showed significantly reduced PFS (31.9 vs. 48.2months in MCT1^high vs. MCT1^low,P=.03) and OS (75.9 months vs. not reached (NR) months in MCT1^high vs. MCT1^low; P=.001) in case of lenalidomide maintenance. Likewise, patients with thalidomide maintenance showed reduced PFS (34.8 vs. 43.7 months in MCT1^high vs. MCT1^low, P=.23) and significantly shorter OS (83.6 months vs. not reached (NR) months in MCT1^high vs. MCT1^low;P=.03). For bortezomib based maintenance, MCT1 expression had no significant impact on PFS (39.8 months vs. 32.6 months in MCT1^high vs. MCT1^low) and OS (125.8 months vs. 129.8 months in MCT1^high vs. MCT1^low). No association with other prognostic factors was found.

As still differences between MCT1^high vs. MCT1^lowexpression myeloma cells might be attributed to undiscerned molecular factors and for functional validation, we lentivirally overexpressed CD147 and MCT1 in human myeloma cell lines. Overexpression of MCT1 significantly reduced cytotoxicity of lenalidomide, while no change was observed in MM cells treated with bortezomib. We subsequently validated our results in vivo. Functional investigations in the mechanism of MCT1 impact on cellular survival are ongoing.

Conclusion:

Taken together MCT1 expression as potential predictive marker for response to IMiD-based maintenance treatment. Both PFS and OS were significantly reduced in patients with high gene expression levels of MCT1. In vitro and in vivo (xenograft model), MCT1 overexpression reduced sensitivity to lenalidomide unlike bortezomib treatment.