Verification of new molecular markers for early diagnosis of lung tumors for follow-up examinations (PURE - Lung Verification)

Status:

completed 12/2020

Aims:

Numerous promising marker candidates for early detection of lung cancers were identified in a previous project (FP‐339). The current project was designed to verify these candidates in an independent patient collective recruited in the Bonn/Rhein‐Sieg area. Additionally, the previous marker spectrum of epigenetic markers and proteins should be expanded to include genomic and RNA‐based markers. The spectral histopathology already established in the previous project was to be methodically improved in order to achieve a more precise tissue classification, as well as to provide tissue samples by means of laser microdissection for further molecular characterisation. In addition to the recruitment of 400 patients with suspected lung tumours, a group of 100 patients with benign respiratory diseases ("COPD controls") was included subsequently as control group.

Activities/Methods:

A total of 423 patients, including 333 with suspected lung tumour and 90 COPD controls, were included in the study at two clinics of the Bonn/Rhein‐Sieg Lung Cancer Centre. Blood and saliva samples were collected from patients. Tissue samples were obtained during operations as well as bronchoscopies and histopathologically confirmed within the framework of the study. DNA mutations were identified by next‐generation sequencing (NGS) in tissue and plasma. Circulating DNA in plasma was additionally detected with a more sensitive method (digital droplet PCR ‐ ddPCR). Epigenetic analyses of DNA methylation, RNA, and copy number variations (CNV) were performed in tissue samples (surgical resections and EBUS/PE material from bronchoscopies), whole blood, plasma and saliva samples. Proteome analyses were performed in tissue and plasma samples. Identified proteins were then evaluated using various bioinformatics methods. Further, biospectroscopic tissue analyses were extended to include molecular detection of the mutation status. By means of laser microdissection, tumour areas and healthy areas were isolated and made available for proteome and mutation analyses.

Results:

Pathology: The analysis of ctDNA from plasma samples did not show sufficient detection of DNA mutations (<10% wild type) compared to tissue‐based analyses (<10% wild type), even after methodological modifications (blood collection system, more sensitive ddPCR). This may be due to an insufficient tumour burden at early tumour stages (low amounts of ctDNA shedding).

Epigenetics: Various markers at the DNA methylation level and evaluation of CNV enabled excellent differentiation between tumour and non‐tumour tissue. This was even possible in the smallest amounts of tissue from endobronchial ultrasound‐based (EBUS) biopsies. Differentiation of tumour patients and controls in body fluids (plasma, saliva) using newly developed methods for DNA methylation was also technically possible, however, no differentiation between tumour and COPD patients was achieved. This limitation also applied to RNA markers. However, marker performance was improved for small cell lung tumour subtypes as well as certain marker combinations (miRNA + protein) compared to single markers.

Biophysics: The biospectroscopic tissue classification from the preceding project (FP‐339) could be verified and the required measurement time was reduced to a few minutes by using quantum cascade lasers (QCL). Hereby, a labelfree molecular detection of driver mutations (TP53, KRAS, EGFR) has been established, which also allows the labelfree, automated pathological subclassification of driver mutations of adenocarcinoma of the lung with high sensitivity and specificity on the molecular level.

Proteomics: In the tissue‐based comparison of adenocarcinomas and controls with benign lung diseases, proteins could be identified that were also differentially detectable in the plasma samples. By combining proteomics analysis and bioinformatics approaches, a biomarker panel was identified in a study subsample that separated adenocarcinomas and COPD with a sensitivity of 97% and a specificity of 89%. Verification of this marker panel has not yet been carried out.

Conclusions: Overall, markers or marker combinations have not been verified in body fluids yet that are suitable for final validation in a longitudinal study for early cancer detection. Further analyses should in particular verify new marker combinations and protein markers in order to improve the insufficient differentiation between cancer patients and patients with chronic inflammatory lung diseases.

Last Update: