A microRNA Signature Predicts Response to Anti-CD19 Therapy (MEDI-551) in B-Cell Malignancies

Background:

microRNAs (miRNA, miR) are a class of highly conserved short noncoding, 17–25 nucleotide long RNA products (Ambros; Nature, 2004) that regulate gene expression at the posttranscriptional level (Bartel; Cell, 2004). Specific miRNA expression signatures can distinguish cancer from benign tissues and may provide the basis for developing new diagnostic and therapeutic strategies (Nelson et al.; Mol Can Ther, 2008). Despite numerous studies focused on understanding the regulation of miRNA in cancer development and progression, the relationship between miRNA expression and response to B-cell therapy has not been fully explored; therefore, we sought to identify microRNAs that would identify patients with B cell malignancies who may derive improved benefit from MEDI-551, an afucosylated, affinity-optimized anti--CD19 antibody with enhanced antibody dependent cell cytotoxicity (ADCC).

Methods:

A miRNA signature was found to be highly differentially expressed between cell lines of high sensitivity (n = 4) versus low sensitivity (n = 3) to in vitro ADCC with MEDI-551. miRNA expression patterns were confirmed across four broad profiling platforms to ensure their reproducibility. This miRNA signature was pre-specified for testing in a Phase 1 and Phase 2 trials of MEDI-551 in B-cell malignancies to assess its clinical utility in predicting patient response to MEDI-551 treatment. For this analysis, quantitative PCR (TaqMan) was utilized on pre-treatment (baseline) whole blood or PBMC samples.

Results:

In patient samples, we found that the miRNA signature displayed consistent results between in vitro and in vivo experiments. The in vitro data showed a 12-fold (p < 0.001) lower median miRNA signature in cell lines highly sensitive to in vitro ADCC with MEDI-551 compared to those with lower sensitivity. In a Phase I trial evaluating multiple doses of MEDI-551, diffuse large B cell lymphoma (DLBCL) patients who responded to MEDI-551 (CR/PR, n = 5) had significantly lower (7-fold; p < 0.0001) median miRNA signature expression compared to patients who did not respond (PD, n = 10) (Figure 1). Similar results were observed in patients (n=34) with follicular lymphoma (FL) included in the Phase I trial and a Phase II trial for chronic lymphocytic leukemia (CLL) patients (n=160). Importantly, the microRNA signature did not appear to predict response to rituximab. Correlation of the miRNA signature with prognostic factors is covered in an abstract titled “MEDI-551 MicroRNA Signature is Not Correlated With Prognostic Markers in Patients With Relapsed or Refractory Chronic Lymphocytic Leukemia From a Phase 2 Study of MEDI-551 and Bendamustine vs Rituximab and Bendamustine.”

To ensure that performance of the molecular test was sufficient for testing in a late stage clinical development, the original research assay was altered to make it more amenable to clinical testing and a validation plan was designed and executed to create a sensitive and specific miRNA signature test that demonstrated equivalence with the original assay utilized to evaluate clinical trial samples.

Figure 1: miRNA signature expression is lower in DLBCL patient samples who respond to MEDI-551 compared to non-responders. Expression of the miRNA signature was evaluated in pre-treatment whole blood samples from DLBCL patients (n=26) enrolled in a Phase I trial by TaqMan qPCR and compared with expression in blood samples from healthy donors (n=13) to generate fold change values.

Conclusions:

Potential use of miRNA signatures as molecular biomarkers in human cancers is supported by a large body of literature, but their clinical implementation has been hindered by technical hurdles. Using cell lines with differing sensitivities to ADCC mediated by MEDI-551, we discovered a microRNA signature that correlates with response to MEDI-551.in Phase I and II clinical trials in patients with B-cell malignancies. This work, combined with the creation of an analytically validated miRNA signature test show promise as a potential predictive marker for response to MEDI-551. Additional larger trials will be necessary to confirm the utility of the microRNA signature as a potential companion diagnostic for MEDI-551 in the treatment of B cell malignancies. In addition, ongoing work is focused on gaining an understanding of the biological significance of this miRNA signature in B cell malignancies and its relationship to the mechanism of action for MEDI-551.