|
Back to 2014 Annual Meeting Abstracts
Using microRNA expression profiles to differentiate between benign and malignant renal masses
Drew A. Palmer, MD, Casey G. Kowalik, MD, Kari H. Bailey, MD, Shiv B. Patel, MD, Travis B. Sullivan, MS, John A. Libertino, MD, Kimberly M. Rieger-Christ, PhD.
Lahey Hospital and Medical Center, Burlington, MA, USA.
BACKGROUND:
The use of percutaneous renal mass biopsy (RMB) in diagnostic algorithms for small renal masses is on the rise. The distinction between benign and malignant pathology on biopsy has significant implications for management and can be challenging for pathologists. While diagnostic accuracy of RMB approaches 90%, molecular markers have been suggested as an adjunct to aid in diagnosis. MicroRNAs (miRNA) are involved in the regulation of mRNA and have been implicated in the tumorigenesis of clear cell renal cell carcinoma (ccRCC). Our goal was to develop a molecular expression profile using miRNA to distinguish between benign renal masses and ccRCC.
METHODS:
Total RNA was extracted from formalin fixed paraffin embedded renal tissue specimens obtained via partial or radical nephrectomy from a single tertiary care hospital. MiRNA expression profiles were identified for ccRCC using microarray analysis on pooled samples. Validation of the differential expression of these miRNA was assessed in 62 benign (oncocytoma, angiomyolipoma, and normal parenchyma) and 95 ccRCC specimens using qRT-PCR. The expression profiles were then analyzed using a binary logistic regression model. Statistical analysis was performed using SPSS v21.
RESULTS:
Microarray analysis of the pooled samples showed differential expression of 257 miRNA between benign and malignant tissue. Validation of miRNA expression profiles was assessed in the benign and ccRCC specimens with significant differences in normalized fold change in miRNA expression demonstrated for 6 miRNAs at p < 0.001. Using two of the miRNA expression profiles, a binary logistic regression model was created for differentiation between benign renal masses and ccRCC with an area under the ROC curve of 0.98. The sensitivity and specificity of this model were 0.96 and 0.87, respectively.
CONCLUSIONS:
We have identified a panel of 6 miRNA whose expression profiles are statistically different between benign renal tissue and ccRCC. Using two of these miRNA expression profiles, an algorithm has been created that can distinguish benign and malignant renal masses. As the use of percutaneous RMB continues to rise, this model has the potential to further increase its diagnostic accuracy.
Back to 2014 Annual Meeting Abstracts
|
|
