An increasing number of studies show that miRNAs from blood can be used as biomarkers for various human diseases. The spectrum of diseases analyzed includes cancer as well as non-cancer diseases. While the majority of these studies analyzed serum or plasma, there are also a considerable number of investigations on miRNA expression in PBMCs or whole blood. Independent of the question whether serum, plasma, PBMCs or whole blood was analyzed, these studies provide strong evidence that blood based miRNA signatures can be used to differentiate between controls and patients. In a comparative analysis of blood-borne miRNA profiles in 14 diseases, we showed recently that miRNA signatures could even differentiate between different diseases of the same organ as for example between lung cancer and COPD with accuracy over 90%. It was also shown that miRNA expression profiles could not only indicate the probability of a diseases but also the probability of the absence of a disease .
Tissue based miRNA expression analyses even revealed specific miRNA expression pattern in different histological cancer subtypes, including lung cancer. For example, miRNA expression analysis allowed to correctly differentiating transthoracic needle aspiration biopsy specimens from NSCLC patients in adenocarcinoma and squamous cell lung carcinoma . Sets of two miRNAs accurately discriminated between NSCLC and SCLC and between adenocarcinoma and squamous cell lung carcinoma in bronchial brushing specimens . MiRNA analysis also distinguished NSCLC from SCLC cell lines . A small subset of eight miRNAs classified small pre-operative biopsies into squamous cell carcinoma, non-squamous non–small cell lung cancer, carcinoid, and small cell carcinoma . Interestingly, we found a high overlap of deregulated miRNAs identified in the above mentioned studies on tissue samples from different histological lung cancer subgroups and the miRNAs detected in the blood samples analyzed in the present study. However, there is considerable less evidence that blood-borne miRNA signatures can identify specific subcategories within a disease as for example a specific tumor grading or histological cancer subtypes. As a first attempt to address this question we analyzed three histological subtypes of lung cancer. Lung cancer was chosen since the best possible lung cancer treatment requires accurate subclassification. The clinically most important differentiation is between NSCLC and SCLC as they differ not only histologically but also in their behavior. SCLCs are very fast growing tumors making them the most aggressive form of lung cancer with chemotherapy or radiotherapy or a combination of both as common treatment. NSCLC treatment depends on cancer stage; surgery is the treatment of choice for early stage NSCLC (stage I and II) [3, 4]. However, as about 70% of NSCLCs are diagnosed at advanced stages, systemic therapy (chemo- or targeted therapy) and/or radiation is recommended .
Our results show that relatively few blood-borne miRNAs were differentially expressed between patients with adenocarcinoma and with squamous cell lung carcinoma, and even less between NSCLC and SCLC. The similarity between the miRNA signature did not allow a separation between the different histotypes, as it was previously reported for miRNAs that were derived from tumor tissues [12–15, 30–34]. Interestingly, the miRNA hsa-miR-205 that was shown in several publications to distinguish between tissue samples from adenocarcinoma and squamous cell lung cancer patients was not detected in any of the analyzed blood samples [12–15, 31–34]. The results that we obtained for the blood-borne miRNA signatures are, however, consistent with the results recently obtained for miRNA isolated from serum that were also independent from histology [35, 36]. Serum derived miRNA signatures showed a comparable risk of cancer death between patients with adenocarcinoma or squamous cell carcinoma. Likewise, serum derived miRNA signatures showed a comparable risk of cancer death between patients with either stage I, II, or IIIa carcinomas . Lin and colleagues  identified two serum miRNAs, namely hsa-miR-126 and hsa-miR-183 that may serve as potential serum biomarkers for metastatic non-small-cell lung cancer. This prompted us to investigate if the miRNA expression pattern of whole blood might also be influenced by metastases. Using a set of three to four miRNAs we were able to differentiate between metastatic and non-metastatic adenocarcinoma and SCLC samples, but this was not possible for squamous cell lung cancer samples. The two miRNAs hsa-miR-126 and hsa-miR-183 identified by Lin et al., were indeed detected in all of our analyzed samples, however, they were not included in our set of miRNAs. A classification with these two miRNAs and our samples was not possible. The three miRNAs that classified between metastatic and non-metastatic adenocarcinoma were expressed in all analyzed blood samples, but were all up-regulated in metastatic adenocarcinoma (between 1.4 and 1.7 fold). These miRNAs have never been related to cancer metastasis in literature. However, hsa-miR-361-5p was identified to be a regulator of VEGFA and thus associated with skin cancer . The four miRNAs that classified between metastatic and non-metastatic SCLC were also expressed in all analyzed blood samples. Two miRNAs were up-regulated and two miRNAs were down-regulated between 1.2 and 1.5 fold in metastatic SCLC. Hsa-miR-328 that was 1.4 fold up-regulated in metastatic SCLC has previously been shown to mediate NSCLC migration and to be associated with NSCLC brain metastasis . In colorectal cancer tissue hsa-miR-103 might promote metastasis by targeting the known metastasis suppressors death-associated protein kinase (DAPK) and Krüppel-like factor 4 (KLF4) . However, in blood of metastatic SCLC patients, this miRNA was slightly down-regulated (1.2 fold). The two other miRNAs hsa-miR-15b and hsa-miR-199a-5p have not been associated with metastasis or lung cancer.