MiR-19a as a prognostic indicator for cancer patients: a meta-analysis

Abstract

MiR-19a was aberrantly expressed in various types of cancers and was observed to be potentially associated with the prognosis of cancer patients. The present analysis aims to elucidate its precise predictive value in various human malignancies. Online electronic searches of PubMed, Web of Science (WOS), Embase in English and VIP, Wanfang, SinoMed, and the China National Knowledge Infrastructure (CNKI) in Chinese up to September 8, 2018 were conducted. As a result, in overall analysis, a significant association was identified between miR-19a levels and OS (HRs = 2.31, CI: 1.11-4.83). The relation of miR-19a expression to OS was further recognized by fixed model within the studies of sample size less than 150 (HRs = 1.68, CI: 1.35-2.08), NOS scores greater than or equal to 8 (HRs = 1.53, CI: 1.13-2.06) or less than 8 (HRs = 1.89, CI: 1.58-2.27), specimen derived from tumor (HRs = 1.73, CI: 1.42-2.12) or blood (HRs = 1.87, CI: 1.46-2.40) and the patients of osteosarcoma (HRs = 7.17, CI: 5.04-10.21). Sensitivity analyses revealed no significant results. The association between miR-19a expression level and DFS was also found to be significant (HRs = 2.03, CI: 1.13-3.66). Correlations between miR-19a levels and clinicopathological features were examined and revealed that lymph node metastasis was significantly associated with miR-19a expression levels (OR = 0.565, CI: 0.346-0.921). Summarily, the over expression of miR-19a was an underlying risk of poor prognosis in many human malignancies, especially in osteosarcoma. Moreover, elevated miR-19a expression was linked to the potential of lymph node metastasis.

Keywords: cancer; meta-analysis; miR-19a; prognosis.

Figure 1. The flow diagram of the meta-analysis

Figure 2. Prognostic effect of miR-19a expression levels on survival index

The association between miR-19a expression levels and (A) overall survivals, subgroup analyses of (B) sample sizes (≥150 and <150), (C) NOS scores (≥8 and <8), (D) specimen (tumor or blood), (E) tumor category (osteosarcoma, non-small cell lung cancer, and others), and (F) DFS (data from log rank tests).

Figure 3. Sensitivity analyses for pooling results from univarite or multivarite statistics

Sensitivity analyses for HRs of data for overall survivals extracted from (A) log tests and (B) cox multivariate regression.

Figure 4. Independent prognostic efficacy of miR-19a expression levels on survival index

The independent role of miR-19a as a prognostic detector for the overall survivals in patients of carcinoma, (A) overall survivals, subgroup analyses of (B) sample sizes (≥150 and <150), (C) NOS scores (≥8 and <8), (D) specimen (tumor and blood), (E) tumor category (osteosarcoma, non-small cell lung cancer, and others).