Hsa_circ_0006010 and hsa_circ_0002903 in peripheral blood serve as novel diagnostic, surveillance and prognostic biomarkers for disease progression in chronic myeloid leukemia

Abstract

Background: In the era of tyrosine kinase inhibitor (TKI) treatment, the progression of chronic myeloid leukemia (CML) remains a significant clinical challenge, and genetic biomarkers for the early identification of CML patients at risk for progression are limited. This study explored whether essential circular RNAs (circRNAs) can be used as biomarkers for diagnosing and monitoring CML disease progression and assessing CML prognosis.

Methods: Peripheral blood (PB) samples were collected from 173 CML patients (138 patients with chronic phase CML [CML-CP] and 35 patients with accelerated phase/blast phase CML [CML-AP/BP]) and 63 healthy controls (HCs). High-throughput RNA sequencing (RNA-Seq) was used to screen dysregulated candidate circRNAs for a circRNA signature associated with CML disease progression. Quantitative real-time PCR (qRT-PCR) was used for preliminary verification and screening of candidate dysregulated genes, as well as subsequent exploration of clinical applications. Receiver operating characteristic (ROC) curve analysis, Spearman's rho correlation test, and the Kaplan-Meier method were used for statistical analysis.

Results: The aberrant expression of hsa_circ_0006010 and hsa_circ_0002903 during CML progression could serve as valuable biomarkers for differentiating CML-AP/BP patients from CMP-CP patients or HCs. In addition, the expression levels of hsa_circ_0006010 and hsa_circ_0002903 were significantly associated with the clinical features of CML patients but were not directly related to the four scoring systems. Furthermore, survival analysis revealed that high hsa_circ_0006010 expression and low hsa_circ_0002903 expression indicated poor progression-free survival (PFS) in CML patients. Finally, PB hsa_circ_0006010 and hsa_circ_0002903 expression at diagnosis may also serve as disease progression surveillance markers for CML patients but were not correlated with PB BCR-ABL1/ABL1^IS.

Conclusions: Our study demonstrated that PB levels of hsa_circ_0006010 and hsa_circ_0002903 may serve as novel diagnostic, surveillance, and prognostic biomarkers for CML disease progression and may contribute to assisting in the diagnosis of CML patients at risk for progression and accurate management of advanced CML patients.

Keywords: Chronic myeloid leukemia; Diagnosis; Disease progression; Hsa_circ_0002903; Hsa_circ_0006010; Prognostic biomarkers; Surveillance.

Fig. 1

Flow chart of the study design

Fig. 2

Profile of circRNA expression in CML. a Differentially expressed circRNA heatmap between CML-CP patients and HCs. b Volcano plot of differentially expressed circRNAs between CML-CP patients and HCs. c Differentially expressed circRNA heatmap between CML-CP and CML-AP/BP patients. d Volcano plot of differentially expressed circRNAs between CML-CP and CML-AP/BP patients. e Relevant heatmap results between CML-CP patients and HCs. f Relevant heatmap results between CML-CP and CML-AP/BP patients. Note: ^{a, c} On the left, circRNAs are clustered according to expression similarity. At the top of the figure, each sample is clustered according to the similarity of the expression spectrum, with the clustering intensity increasing from green to red. ^{b, d} The two vertical black lines represent twofold up- and downregulation, and the horizontal black lines symbolize a p value of 0.05. ^{e, f} The relative intensity increased from yellow to blue. N_1001, N_1002, and N_1003 represent 3 HCs; C_1001, C_1002, and C_1003 represent 3 CML-CP patients; and A_1001, A_1002, and A_1003 represent 3 CML-AP/BP patients

Fig. 3

Selection of candidate circRNAs related to CML disease progression. a Circos plot displaying the distribution of significantly dysregulated circRNAs between CML-AP/BP and CML-CP patients. The outermost layer is the human chromosome map. The inner 9 circles represent each sample analyzed by RNA-seq. The bar chart shows the expression levels of circRNAs. b Classification of dysregulated circRNAs according to genomic origin

Fig. 4

qRT-PCR validation of differentially expressed circRNAs in CML patients. a qRT-PCR validation and RNA-Seq analysis of dysregulated genes. b qRT-PCR validation of six circRNAs between CML-AP/BP and CML-CP patients. c qRT-PCR validation of six circRNAs between CML-AP/BP patients and HCs. d FC values of hsa_circ_0006010 and hsa_circ_0002903 expression when comparing CML-AP/BP/CML-CP and CML-AP/BP/HCs. *** p < 0.001

Fig. 5

Diagnostic value of PB hsa_circ_0006010 and hsa_circ_0002903 for CML disease progression. a ROC curve analysis of PB hsa_circ_0006010 and hsa_circ_0002903 for discriminating CML-AP/BP patients from CML-CP patients. b ROC curve analysis of PB hsa_circ_0006010 and hsa_circ_0002903 for discriminating CML-AP/BP patients from HCs. c ROC curve analysis of PB hsa_circ_0006010 and hsa_circ_0002903 for discriminating CML-CP patients from HCs

Fig. 6

Correlations between PB hsa_circ_0006010 and hsa_circ_0002903 expression and PFS in CML-CP patients. a K-M curve of PB hsa_circ_0006010. b K-M curve of PB hsa_circ_0002903

Fig. 7

PB hsa_circ_0006010 and hsa_circ_0002903 as potential disease progression surveillance markers for CML patients. a PB hsa_circ_0006010 expression was not correlated with PB BCR-ABL1/ABL1^IS in CML patients at diagnosis. b PB hsa_circ_0002903 expression was not correlated with PB BCR-ABL1/ABL1^IS in CML patients at diagnosis. c PB hsa_circ_0006010 expression in 6 pairs of samples from CML-AP/BP patients at diagnosis and after achieving CR through TKI treatment. d hsa_circ_0002903 expression in 6 pairs of samples from CML-AP/BP patients at diagnosis and after achieving CR through TKI treatment. e PB hsa_circ_0006010 expression in 5 pairs of PB samples from CML-CP patients who developed disease progression due to TKI resistance and who achieved CR after 2G-TKI treatment. f PB hsa_circ_0006010 expression in 5 pairs of PB samples from CML-CP patients who developed disease progression due to TKI resistance and after 2G-TKI treatment