microRNA fingerprinting of CLL patients with chromosome 17p deletion identify a miR-21 score that stratifies early survival

Abstract

Aberrant expression of microRNAs (miRNAs) has been associated with clinical outcome in patients with chronic lymphocytic leukemia (CLL). To identify a powerful and easily assessable miRNA bio-marker of prognosis and survival, we performed quantitative reverse-transcription polymerase chain reaction (qRT-PCR) profiling in 104 CLL patients with a well-defined chromosome 17p status, and we validated our findings with miRNA microarray data from an independent cohort of 80 patients. We found that miR-15a, miR-21, miR-34a, miR-155, and miR-181b were differentially expressed between CLLs with chromosome 17p deletion and CLLs with normal 17p and normal karyotype, and that miR-181b was down-regulated in therapy-refractory cases. miR-21 expression levels were significantly higher in patients with poor prognosis and predicted overall survival (OS), and miR-181b expression levels significantly predicted treatment-free survival. We developed a 21FK score (miR-21 qRT-PCR, fluorescence in situ hybridization, Karyotype) to stratify patients according to OS and found that patients with a low score had a significantly longer OS time. When we evaluated the relative power of the 21FK score with the most used prognostic factors, the score was the most significant in both CLL cohorts. We conclude that the 21FK score represents a useful tool for distinguishing between good-prognosis and poor-prognosis CLL patients.

Figure 1

microRNA signatures associated with 17pDEL in patients with CLL. (A) The definitions of patient categories. (B) The ratios and P values were obtained by applying the t test in the 7 studied miRNAs among the patients with CLL, classified according to the patients' properties on FISH and KARYO. (C) Clusters of differentially expressed miRNAs. Unsupervised clustering was performed on 104 patients with qRT-PCR expression of miR-15a, miR-21, miR-34a, miR-155, miR-181b, and miR-497. Single linkage and the Euclidean distance metric were used.

Figure 2

Kaplan-Meier survival curves for 104 patients with CLL, including 68 CLLs with low Rai stage. (A) OS for patients with CLL was classified into 3 levels according to the 21FK score (0/2, 1/2, and 2/2). (B) OS for patients with CLL was classified into 3 levels according to the 181bFK score (0/2, 1/2, and 2/2). (C) Among the 104 patients, the 21FK score was validated in 68 patients with low Rai stage CLL. The 21FK score predicted patients' survival, even if they had a Rai disease stage of 0, 1, or 2. The survival data were compared, and the reported P values were calculated according to the log-rank test. In the main text, the P values were reported according to Cox regression analyses associated with the hazard ratio computation.

Figure 3

Validation by microarray data evaluation: miR-21 and 21FK score stratify OS risk in an independent validation cohort. Kaplan-Meier survival curves have been reported. miR-21 expression levels were measured by microarray, and cutoffs were based on the median expression levels. (A) Patients with high miR-21 expression had a significantly poorer prognosis in terms of OS. (B) Patients with a high 21FK score (2/2 and 1/2) showed a significantly poorer prognosis in terms of OS than for patients with a low 21FK score (0/2). Statistical differences between curves were calculated with the use of the log-rank test. In the main text the P values were reported according to Cox regression analyses associated to hazard ratio computation.