As a malignant tumor with extremely high mortality, lung cancer has gaining great attention and extensive researches during recent decades. With the development of surgical techniques, concurrent radiotherapy and chemotherapy, and imaging examination technology have greatly improved the prognosis of lung cancer patients. Nevertheless, the most effective way to improve the survival of lung cancer patients lies in early diagnosis and targeted treatment. Therefore, a large amount of researchers are committed to finding suitable non-invasive biomarkers to predict the diagnosis or prognosis of lung cancer, and provide directions for clinical treatment of lung cancer .
As vital regulators of various biological processes in cancer, miRNAs were regarded as perfect non-invasive biomarkers for human cancers [35, 36]. MiR-155 was widely studied to participate in the occurrence and progression of diverse cancers, including lung cancer . Several researches suggested that up-regulated miR-155 is positively correlated with the pathogenesis of lung cancer, indicating that miR-155 acts as an oncogene in lung cancer [37, 38]. Zang et al revealed that miR-155 was involved in the drug resistance of lung cancer. In their study, miR-155 was shown to modulate celluar poptosis and DNA damage via Apaf-1 regulated pathways to decrease the sensitivity of lung cancer cells to cisplatin . Moreover, another research conducted by Katrien et al found that miR-155 increases resistance to chemotherapy in lung cancer cells by forming a feedback loop with TP53 . In particular, they also found that over-expression of miR-155 is significantly linked to poor OS of lung cancer patients. These results indicated that miR-155 has the potential to be an ideal biomarker for lung cancer. In addition to the above studies focused on the molecular mechanism of miR-155 regulation in lung cancer cells, accumulating cohort studies have reported the coorelationship between miR-155 levels in different individuals with lung cancer diagnosis or prognosis to determine whether miR-155 acts as an ideal biomarker [40, 41]. However, these results have not been corroborated and even contradictory. Thus, this meta-analysis appears to be necessary to figure out the diagnostic and prognostic value of miR-155 for lung cancer.
In the diagnositic meta-analysis, the total DOR with 95% CI of miR-155 was 15.99 (95% CI: 8.11-31.52). In addition, AUC and corresponding 95% CI were 0.85 (95% CI: 0.82-0.88), indicating that miR-155 could act as a moderate marker in the lung cancer diagnosis compared to healthy individuals. Subgroup analysis of Assay type revealed that studies based on SYBR had a higher DOR of 22.69 (95% CI: 13.90-37.04) and the higher AUC of 0.89 (95% CI: 0.86-0.91), which might be the possible sources of heterogeneity. Nowadays, several tumor biomarkers have been applied for detecting early lung cancer clinically, such as CA-125, CEA, CYFRA21-1, NSE and so on. However, limited sensitivity and specificity of these existing biomakers restricted their diagnostic accuracy. Based on the 6 included articles, miR-155 can be stably detected in the plasma of lung cancer patients with marked differences when compared with control samples, suggesting it can serve as a serum-based biomarker for lung cancer detection individually. As Currently, miR-155 hasn’t been applied as a clinical diagnostic tool in patients that had not previously been diagnosed with lung cancer. And clinical detection of lung cancer usually involves not only a single miRNA, but a combination of miRNAs.What’s more, miR-155 could be combined with traditional biomarkers for the diagnosis of lung cancer, so as to improve the diagnosis accuracy in the future.
By the way, as polymorphisms in genes encoding miRNAs may alter the expression of the corresponding miRNA and thus confer susceptibility to multiple diseases such as cancers, it might be meaningful to investigate the association between polymorphisms in genes encoding miR-155 and lung cancer susceptibility. For example, previous study published by Xie et al identified that rs767649 (A>T) in regulatory regions of miR-155 was associated with the increased risk and poor prognosis of lung cancer . What’s more, they found four target genes of miR-155 including HBP1, TJP1, SMAD5 and PRKAR1A involved in the oxidative stress process of lung cancer. Given that miR-155 is a typical oncogene in lung cancer, more well-designed studies in the future could confirm its diagnostic value, and more importantly, further researches colud focus on gene polymorphisms encoding miR-155, which can manually regulate miRNA levels, leading to changes in cancer-associated downstream protein signaling pathways.
On the other hand, the prognostic meta-analysis suggested that up-regulated miR-155 might not be associated with poor clinical outcomes of lung cancer patients, which was 1.26-fold higher risk for poor OS and 1.28-fold higher risk for poor DFS/PFS. These results might caused by different genetic backgrounds, environmental exposures and detection methods. Recently, accumulating studies worldwide have shown that expression levels of miRNAs in different individuals have significant predictive value in cancers. Currently, the detection of miRNAs in tissue samples has been applied to current tumor prognosis studies, but the detection of serum/plasma samples and other human body fluids appears to be more portable, non-invasive, and can effectively assess survival prognosis at any time before or after treatment. It can even play a role in the patient's life-long disease surveillance and is of great help to clinical thearapy. This meta-analysis found that miR-155 has no obvious prognostic effect on lung cancer, which is inconsistent with results of some previous prognostic studies, while the result is consistent with the prognostic value of miR-155 of NSCLC reported in a meta-analysis published by Lamichhane SR et al in 2018 . However, the sample size included in our meta-analysis is larger than previous mata-analysis, more researches with sufficient data will be needed to verify this result.
Ultimately, several limitations still existed in this meta-analysis as follows: (1) Racial factors were not comprehensive enough, and the population is too monotonous. For example, the diagnostic meta-analysis is mainly for Asians and Africans while the prognostic meta-analysis only focused on Caucasians and Asians. Therefore, more researchers should pay attention to the impact of racial factors in the subsequent studies. (2) Unpublished studies may contain negative results, but we are not available include them, which potentially lead to lack of credibility in the data. (3) We only included articles published in English and Chinese, but did not cover articles in other languages. (4) The sample size was still relatively small, including only 19 studies, which may undermine the reliability of our findings. Therefore, more well-designed studies based on larger samples and sufficient data are required to verify the diagnostic and prognostic value of circulating miR-155 in lung cancer. (5) Adjusted estimates could not be performed in our meta analysis without enough data for the adjustment by other covariates such as TNM stage, histological type, mean of age, gender and so on.. Therefore, further high-quality researches in the risk of lung cancer might be performed to draw more accuracy results in subsequent years.