Abortive transcript (AT) is a 2-19 nt long non-coding RNA that is produced in the abortive initiation stage. Abortive initiation was found to be closely related to RNA polymerase through in vitro experiments. Therefore, the distribution of AT length and the scale of abortive initiation are correlated to the promoter, discriminator, and transcription initiation sequence, and can be affected by transcription elongation factors. AT plays an important role in the occurrence and development of various diseases. Here we summarize the discovery of AT, the factors responsible for AT formation, the detection methods and biological functions of AT, to provide new clues for finding potential targets in the early diagnosis and treatment of cancers.

Lung cancer (LC) is the leading cause of cancer-related deaths worldwide. Tumor marker (TM) detection can indicate the existence and growth of a tumor and has therefore been used extensively for diagnosing LC. Here, we conducted a bibliometric analysis to examine TM-related publications for LC diagnosis to illustrate the current state and future trends of this field, as well as to identify additional promising TMs with high sensitivity.

Publications regarding TMs in LC diagnosis were downloaded from the Web of Science Core Collection. CiteSpace was applied to perform a bibliometric analysis of journals, cocitation authors, keywords, and references related to this field. VOSviewer was used to generate concise diagrams about countries, institutions, authors, and keywords. Changes in the TM research frontier were analyzed through citation burst detection.

A total of 990 studies were analyzed in this work. The collaboration network analysis revealed that the People's Republic of China, Yonsei University, and Molina R were the most productive country, institution, and scholar, respectively. Additionally, Molina R was the author with the most citations. The National Natural Science Foundation of China was the largest funding source. "Carcinoembryonic antigen (CEA) as tumor marker in lung cancer" was the top reference with the most citations, Lung Cancer was the core journal, and "serum tumor marker" experienced a citation burst over the past 5 years.

This bibliometric analysis of TMs in LC diagnosis presents the current trends and frontiers in this field. We summarized the research status of this field and the methods to improve the diagnostic efficacy of traditional serum TMs, as well as provided new directions and ideas for improving the LC clinical detection rate. Priority should be given to the transformation of computer-assisted diagnostic technology for clinical applications. In addition, circulating tumor cells, exosomes, and microRNAs were the current most cutting-edge TMs.

To evaluate the diagnostic value of the serum tumor markers carcinoembryonic antigen (CEA), cytokeratin-19-fragment (CYFRA21-1), squamous cell carcinoma associated antigen (SCCAg), neuron-specificenolase (NSE) and pro-gastrin-releasing peptide (ProGRP) for lung cancers of different pathological types.

This study was conducted among patients with established diagnoses of lung adenocarcinoma (LADC, n=137), lung squamous cell carcinoma (LSCC, n=82), small cell lung carcinoma (SCLC, n=59), and benign chest disease (BCD, n=102). The serum tumor markers were detected for all the patients for comparison of the positivity rates and their serum levels. ROC curve was used for analysis of the diagnostic efficacy of these tumor markers either alone or in different combinations.

In patients with LADC, the positivity rate and serum level of CEA were significantly higher than those in the other groups (P < 0.05); the patients with LSCC had the highest positivity rate and serum level of SCCAg among the 4 groups (P < 0.05). The positivity rates and serum levels of ProGRP and NSE were significantly higher in SCLC group than in the other groups (P < 0.05). CYFRA21-1 showed the highest positivity rate and serum level in LADC group and LSCC group. With the patients with BCD as control, CEA showed a diagnostic sensitivity of 62.8% and a specificity of 93.1% for LADC, and the sensitivity and specificity of SCCAg for diagnosing LSCC were 64.6% and 91.2%, respectively. CYFRA21-1 had the highest diagnostic sensitivity for LADC and LSCC. The diagnostic sensitivity and specificity of ProGRP for SCLC were 83.1% and 98.0%, respectively. When combined, CYFRA21-1 and CEA showed a high sensitivity (78.8%) and specificity (86.3%) for diagnosing LADC with an AUC of 0.891; CYFRA21-1 and SCCAg had a high sensitivity (84.1%) and specificity (87.3%) for diagnosing LSCC with an AUC of 0.912. NSE combined with ProGRP was highly sensitive (88.1%) and specific (98.0%) for diagnosis of SCLC, with an AUC of 0.952. For lung cancers of different pathological types, the combination of all the 5 tumor markers showed no significant differences in the diagnostic power from a combined detection with any two of the markers (P>0.05).

CEA, CYFRA21-1, SCCAg, NSE and ProGRP are all related to the pathological type of lung cancers and can be used in different combinations as useful diagnostic indicators for lung cancers.