A three-long noncoding RNA signature as a diagnostic biomarker for differentiating between triple-negative and non-triple-negative breast cancers

Abstract

Background: Triple-negative breast cancer (TNBC) is an aggressive cancer with unfavorable outcome and it is useful to explore noninvasive biomarkers for its early diagnosis. Here, we identified differentially expressed long noncoding RNAs (lncRNAs) in blood samples of patients with TNBC to assess their diagnostic value.

Methods: Differential expression of lncRNAs in plasma of patients with TNBC (n = 25) and non-TNBC (NTNBC; n = 35) and in healthy controls was compared by microarray analysis and validated by real-time PCR. lncRNA expression between plasma and BC tissues was compared using Pearson correlation test. Logit model was used to obtain a new lncRNA-based score. Receiver operating characteristic analysis was performed to assess the diagnostic value of the selected lncRNAs.

Results: Microarray data showed that 41 lncRNAs were aberrantly expressed. Among these, antisense noncoding RNA in the INK4 locus (ANRIL), hypoxia inducible factor 1alpha antisense RNA-2 (HIF1A-AS2), and urothelial carcinoma-associated 1 (UCA1) were markedly upregulated in plasma of patients with TNBC compared with patients with NTNBC (P < 0.01). HIF1A-AS2 expression was positively associated with its tissue levels (r = 0.670, P < 0.01). AUC (95% CI) of ANRIL, HIF1A-AS2, and UCA1 was 0.785 (0.660-0.881), 0.739 (0.610-0.844), and 0.817 (0.696-0.905), respectively. TNBCSigLnc-3, a new score obtained using the logit model, showed excellent diagnostic performance, with AUC of 0.934 (0.839-0.982), sensitivity of 76.0%, and specificity of 97.1%.

Conclusion: ANRIL, HIF1A-AS2, and UCA1 expression was significantly increased in plasma of patients with TNBC, suggesting their use as TNBC-specific diagnostic biomarkers.

Figure 1

Representative immunostaining patterns in prechemotherapy biopsy samples of patients with triple-negative breast cancer, according to pathological response.

Figure 2

Microarray profiling of lncRNAs in the plasma samples of patients with breast cancer and healthy individuals. A heat map representation of differentially expressed lncRNAs in the 3 study groups; results represent a cutoff P value of 0.05 and a fold change of >1.5. Green and red bars indicate downregulated and upregulated lncRNAs, respectively. lncRNAs = long noncoding RNAs.

Figure 3

Relative expression levels of ANRIL, HIF1A-AS2, and UCA1 in the plasma samples of healthy individuals, patients with triple-negative breast cancer, and patients with non-triple-negative breast cancer. ANRIL = antisense noncoding RNA in the INK4 locus, HIF1A-AS2 = hypoxia inducible factor 1alpha antisense RNA-2, UCA1 = urothelial carcinoma-associated 1.

Figure 4

Correlation of lncRNA levels between the plasma and breast cancer tissues of patients with triple-negative breast cancer. lncRNAs = long noncoding RNAs.

Figure 5

Diagnostic performance of ANRIL, HIF1A-AS2, and UCA1 in the plasma samples of healthy individuals and patients with triple-negative breast cancer. Receiver operating characteristic curve analysis. ANRIL = antisense noncoding RNA in the INK4 locus, HIF1A-AS2 = hypoxia inducible factor 1alpha antisense RNA-2, UCA1 = urothelial carcinoma-associated 1.

Figure 6

Diagnostic performance of ANRIL, HIF1A-AS2, and UCA1 in the plasma samples of patients with triple-negative breast cancer and patients with non-triple-negative breast cancer patients. Receiver operating characteristic curve analysis. ANRIL = antisense noncoding RNA in the INK4 locus, HIF1A-AS2 = hypoxia inducible factor 1alpha antisense RNA-2, UCA1 = urothelial carcinoma-associated 1.

Figure 7

Diagnostic performance of TNBCSigLnc-3 in the plasma samples of patients with breast cancer. A, Orderly distribution of TNBCSigLnc-3 values between patients with triple-negative breast cancer and patients with nontriple-negative breast cancer. B, Pairwise comparison of receiver operating characteristic curves of 4 subjects. TNBC = triple-negative breast cancer.