In cases of multiple lung cancers, individual tumors may represent either a primary lung cancer or both primary and metastatic lung cancers. Treatment selection varies depending on such features, and this discrimination is critically important in predicting prognosis. The present study was undertaken to determine the efficacy and validity of mutation analysis as a means of determining whether multiple lung cancers are primary or metastatic in nature. The study involved 12 patients who underwent surgery in our department for multiple lung cancers between July 2014 and March 2016. Tumor cells were collected from formalin-fixed paraffin-embedded tissues of the primary lesions by using laser capture microdissection, and targeted sequencing of 53 lung cancer-related genes was performed. In surgically treated patients with multiple lung cancers, the driver mutation profile differed among the individual tumors. Meanwhile, in a case of a solitary lung tumor that appeared after surgery for double primary lung cancers, gene mutation analysis using a bronchoscopic biopsy sample revealed a gene mutation profile consistent with the surgically resected specimen, thus demonstrating that the tumor in this case was metastatic. In cases of multiple lung cancers, the comparison of driver mutation profiles clarifies the clonal origin of the tumors and enables discrimination between primary and metastatic tumors.

ObjectivesPulmonary sarcomatoid carcinomas are rare and generally aggressive tumors composed of carcinomatous and sarcomatous components; however, the evolution of sarcomatoid cancer has not been elucidated. Here, we aimed to evaluate the mutational profiles and phylogeny of sarcomatoid carcinomas using next generation sequencing and in-silico analysis to facilitate the development of novel therapies.MethodsFour patients who underwent surgery for sarcomatoid cancer were enrolled. Cancer cells were collected from carcinomatous and sarcomatous components in each tumor by laser capture microdissection. Next-generation sequencing was performed in each component, and the mutation profiles were compared. For further inference of phylogenies, phylogenetic and PyClone analyses were performed. Mismatch repair disturbance and programmed death ligand-1 (PD-L1) expression were also evaluated.ResultsComparative genetic analysis of different histological areas revealed that the separate components shared several common mutations, which showed relatively high cellular prevalence in the PyClone statistical inference. Phylogenetic analysis showed that the sarcomatous component had ramified from the carcinomatous component in the early phase of the evolution process and accumulated a number of mutations that were different from those of the carcinomatous component. Moreover, microsatellite instability was detected in a case of sarcomatoid cancer and PD-L1 was strongly positive (≥ 50%) in all sarcomatoid cancers.ConclusionsOur data suggest that sarcomatoid carcinoma evolves from a common ancestral clone, and its phylogenetic features may reflect high-grade malignancy in pulmonary sarcomatoid carcinoma. High tumor mutation burden and strong PD-L1 staining may provide a rationale for the use of targeted immunotherapies in pulmonary sarcomatoid carcinomas.

In cases of multiple lung cancers, individual tumors may represent either a primary lung cancer or both primary and metastatic lung cancers. In this study, we investigated the differences between clinical/histopathological and genomic diagnoses to determine whether they are primary or metastatic. 37 patients with multiple lung cancers were enrolled in this study. Tumor cells were selected from tissue samples using laser capture microdissection. DNA was extracted from those cells and subjected to targeted deep sequencing. In multicentric primary lung cancers, the driver mutation profile was mutually exclusive among the individual tumors, while it was consistent between metastasized tumors and the primary lesion. In 11 patients (29.7%), discrepancies were observed between genomic and clinical/histopathological diagnoses. For the lymph node metastatic lesions, the mutation profile was consistent with only one of the two primary lesions. In three of five cases with lymph node metastases, the lymph node metastatic route detected by genomic diagnosis differed from the clinical and/or pathological diagnoses. In conclusion, in patients with multiple primary lung cancers, cancer-specific mutations can serve as clonal markers, affording a more accurate understanding of the pathology of multiple lung cancers and their lymphatic metastases and thus improving both the treatment selection and outcome.

Circulating tumor DNA (ctDNA), extracted from plasma, is a non-invasive surrogate biomarker. However, the distribution of ctDNA in the body still remains to be elucidated. In this study, resected lung tumors, with simultaneous blood and bone marrow samples, were analyzed to elucidate the distribution of ctDNA. Rib bone marrow, pulmonary venous blood (Pul.V) and peripheral blood (Peri.B) were obtained from 30 patients. The liquid samples were divided into cell pellets and supernatant by centrifugation; a total of 212 DNA samples were subjected to massively parallel sequencing. ctDNA was detected in 5 patients. Given that the frequency of mutations in the primary tumor was considered to be 100%, those in the other specimens were as follows; Pul.V plasma 20%, Peri.B plasma 11%, and the other samples 0%. Furthermore, ctDNA reflected the predominant mutations in the primary lesion. Clinically, the presence of ctDNA was associated with significantly poorer survival. These results suggest ctDNA “spill over” into an immediate outflow tract (Pul.V), and from there is disseminated to the entire body. Thus, it can be inferred that ctDNA reflects the cancer progression and could function as a prognostic marker.

Pulmonary invasive mucinous adenocarcinoma (IMA) is considered a variant of lung adenocarcinomas based on the current World Health Organization classification of lung tumors. However, the molecular mechanism driving IMA development and progression is not well understood. Thus, we surveyed the genomic characteristics of IMA in association with immune-checkpoint expression to investigate new potential therapeutic strategies. Tumor cells were collected from surgical specimens of primary IMA, and sequenced to survey 53 genes associated with lung cancer. The mutational profiles thus obtained were compared in silico to conventional adenocarcinomas and other histologic carcinomas, thereby establishing the genomic clustering of lung cancers. Immunostaining was also performed to compare expression of programmed death ligand 1 (PD-L1) and B7-H3 in IMA and conventional adenocarcinomas. Mutations in Kirsten rat sarcoma viral oncogene homolog (KRAS) were detected in 75% of IMAs, but in only 11.6% of conventional adenocarcinomas. On the other hand, the frequency of mutations in epidermal growth factor receptor (EGFR) and tumor protein p53 (TP53) genes was 5% and 10%, respectively, in the former, but 48.8% and 34.9%, respectively, in the latter. Clustering of all 78 lung cancers indicated that IMA is distinct from conventional adenocarcinoma or squamous cell carcinoma. Strikingly, expression of PD-L1 in ≥1% of cells was observed in only 6.1% of IMAs, but in 59.7% of conventional adenocarcinomas. Finally, 42.4% and 19.4% of IMAs and conventional adenocarcinomas, respectively, tested positive for B7-H3. Although currently classified as a variant of lung adenocarcinoma, it is also reasonable to consider IMA as fundamentally distinct, based on mutation profiles and genetic clustering as well as immune-checkpoint status. The immunohistochemistry data suggest that B7-H3 may be a new and promising therapeutic target for immune checkpoint therapy.