1. Home
  2. Medical News
  3. Oncology
advertisement

Six-RNA Lung Adenocarcinoma Panel: Sensitivity Falls in Validation

Illustrated lungs with a small localized early stage lung adenocarcinoma focus in one lung
11/20/2026

Key Takeaways

  • In pathologically confirmed stage I–II lung adenocarcinoma versus healthy controls, six-RNA panel sensitivity was lower in the separate validation sample set than in discovery.
  • Panel specificity was similar across the discovery and validation sets.
  • Eleven candidate small RNAs showed significantly different expression in cases versus controls.
  • In a small subset, the authors found no larger fold changes for the assessed RNAs in exosomes and concluded that exosomal analysis offered no superior diagnostic utility.
Early-stage lung adenocarcinoma is difficult to identify with conventional serum markers. Circulating small RNAs offer a possible biological signal, but their ability to distinguish cancer consistently across distinct serum samples remains uncertain.

Investigators conducted an observational case–control serum-sequencing comparison in a study of a six-small-RNA serum panel. Cases had pathologically diagnosed stage I–II primary lung adenocarcinoma and had not received neoadjuvant chemotherapy; controls were healthy rather than people with pulmonary disease. Two independently prepared and sequenced discovery sets included 43 cases overall; the separate within-study validation set included 29 cases and 27 healthy controls. Individual serum samples underwent small-RNA sequencing. Investigators selected consistently differentially expressed microRNAs (miRNAs) and transfer RNA–derived fragments (tRFs), constructed stepwise multiple-regression panels, and selected model D based on discrimination and collinearity. Its analytes were miR-193a-5p, miR-423-5p(21nt), miR-20a-5p, tRF-Gly(CCC)(29nt), tRF-iMet(CAT)(29nt), and tRF-Val(AAC). Ancillary work compared RNAs extracted directly from serum with exosomal RNAs; a separate machine-learning reanalysis pooled the sample sets and randomly divided them for training and testing.

In discovery, the six-RNA model D had sensitivity 0.954, specificity 0.925, and area under the curve (AUC) 0.974. Candidate small RNAs included markers more abundant in cases and others less abundant in cases. Those individual signals informed panel selection but did not constitute a separately validated assay.

In the separate within-study validation set, model D had sensitivity 0.690, specificity 0.926, and AUC 0.924. Model values did not differ significantly between pathological stages IA and IB/II in that set. In the pooled reanalysis, extreme gradient boosting (XGBoost) performed better than regression model D on accuracy, recall, and F-measure; the comparison did not involve an independent external cohort.

Distinguishing known lung adenocarcinoma cases from healthy controls does not establish performance in a lung cancer screening population or among people with other pulmonary conditions. Although validation samples were separate within the study, the investigators did not externally validate the panel in an independent cohort. The sample was relatively small, particularly for the exosomal comparison, and longitudinal measurements were unavailable. Control data were unavailable for comparisons with carcinoembryonic antigen (CEA) or cytokeratin 19 fragment (CYFRA 21-1). Limits of blank and detection and technical reproducibility were not assessed. The biological mechanism behind the signature remains unclear.

The authors characterized serum small RNAs as a potentially promising noninvasive approach to early detection of lung adenocarcinoma. Their signature distinguished early-stage cases from healthy controls, but sensitivity was lower in the separate within-study validation set.

Clinician Questions

Did the six-RNA serum panel reflect progression of stage I–II lung adenocarcinoma?

Not established. Panel values did not differ significantly between the small pathological-stage groups, but longitudinal progression was not evaluated.

Why was a hemolysis-sensitive microRNA excluded from the lung adenocarcinoma biomarker candidates?

The investigators excluded miR-451a-5p*(24nt), although it was frequently more abundant in cases, because its expression was affected by hemolysis.

This content is intended for healthcare professionals for educational purposes. The article and its image were developed with AI assistance and reviewed by members of the ReachMD team prior to publication. Please refer to the original source linked for the complete and most up-to-date information.

Register

We’re glad to see you’re enjoying ReachMD…
but how about a more personalized experience?

Register for free