NEW YORK (GenomeWeb News) – By folding genomic information about patients' lung cancer tumors in with other clinical data to inform treatment choice, patients generally exhibited improved survival, researchers from the Clinical Lung Cancer Genome Project and the Network Genomic Medicine reported in Science Translational Medicine today.
Lung cancers may be divvied up into a number of histomorphological and immunohistochemical categories: non-small cell lung cancer, small cell lung cancer, and carcinoid cancer. And then NSCLCs can be further classified into adenocarcinomas, squamous cell carcinomas, and large cell carcinomas. Adenocarcinomas can then even be placed into subtypes based mainly on growth patterns.
Overlaying that, though, are genetic mutations. Researchers led by Jürgen Wolf at the University of Cologne in Germany, sought to determine which cancer genomic changes are linked to specific histomorphological and immunohistochemical tumor subtypes. By characterizing 1,255 clinically annotated lung tumors from a range of histological subtypes, they found that subgroups tend to follow certain genomic alteration patterns. The majority — 55 percent — of cases they examined contained a mutation for which there is a personalized therapy under study.
"[W]e provide a blueprint for genomic diagnosis of lung tumors," Wolf and his colleagues wrote. "Determination of immunohistochemical, genomic, and clinical features may thus be combined to yield classes of tumors that are biologically relevant, afford genomically tailored stratification of patients into clinical trials, and improve overall survival of patients with lung cancer."
To systematically examine the genomic subtypes of lung cancer, Wolf and his colleagues collected and analyzed 1,255 surgically resected, fresh-frozen specimens along with their clinical annotations. Somatic copy-number alteration, SNP, and histology data indicated that there were subtype-specific genomic alterations. For example, small cell lung cancers and some large cell lung cancers had a number of changes at the chromosome-arm level while adenocarcinomas and squamous cell carcinomas contained more focal point changes.
A number of those amplified or deleted regions were in areas known to include proto-oncogenes or tumor suppressor genes. Adenocarcinoma, for instance, tended to have amplifications near EGFR, MYC, CCND1, MDM2, NKX2-1, and ERBB2.
By consensus clustering, 98 percent of adenocarcinoma cases, 84 percent of small cell lung cancer cases, and 77 percent of squamous cell lung cancer cases clustered into their own groups.
Large cell lung cancer, though, did not fall into its own cluster of mutations. It, the researchers noted, contained genomic alterations found across the spectrum of the other lung cancer subtypes.
"LC exhibits a general diagnostic plasticity when considering data on chromosomal copy number, gene mutations, gene expression, and immunohistochemistry," Wolf and his colleagues noted. "Combined immunohistochemical and genomic analysis is therefore ideal to classify this heterogeneous group" as adenocarcinomas, squamous cell carcinomas, or neuroendocrine carcinomas.
Overall, though, as there appeared to be genomic alterations specific to most cancer subtypes, Wolf and his colleagues developed a model to determine whether genomic alterations could be used to predict subtype.
They combined mutation and somatic copy-number alteration data to determine which genomic alterations marked each subtype: ALK, BRAF, and EGFR mutations, among others, mainly occurred in adenocarcinomas while MYCN mutations were contained largely within small-cell lung cancers.
Using those findings, they devised a statistical model to classify tumors based on their genomic changes. For adenocarcinoma, small-cell lung cancer, and squamous cell carcinoma, they reported that the diagnosis their model predicted tracked closely with the one determined from pathological review.
"Most initial diagnoses of AD, SCLC, and SQ were confirmed by both our model and pathology review," the investigators said.
A number of large cell lung cancers were reassigned to other subtypes, a move that was also in concordance with pathological review.
Wolf and his colleagues also evaluated how their combination of genomics and immunohistochemistry could be used in the clinic. They turned to a cohort of 5,145 lung cancer patients for whom they added immunohistochemical and genomic characterization to their diagnostic workup.
For this analysis, they searched for certain key mutations — ALK, BRAF, DDR2, EGFR, ERBB2, FGFR1, KRAS, and PIK3CA — in paraffin-embedded tumor samples.
Based on the results, the researchers made recommendations regarding possible treatments, suggesting either approved targeted therapies or ones being evaluated in clinical trials. For example, based on these findings, some 64 of the 84 advanced-stage patients with an EGFR mutation received either erlotinib or gefitinib, and half of the advanced-stage patients with ALK translocations received crizotinib.
By comparing patients who had been genotyped to those who were not — due to, for example, a lack of tissue — Wolf and his colleagues found that genotyping affects survival, independent of stage and histology.
"Although this observation most likely results from the favorable outcome in patients treated with kinase inhibitors, it demonstrates that genotyping is mandatory for patients to benefit from targeted therapeutic intervention," they added.
Broken down by mutation type, if there was a targeted therapy available, the researchers generally found that those patients with the mutation who received the therapy had an increased survival time.
"In our outreach study, genotyping alone — as a prerequisite for personalized treatment — was associated with improved patient survival," Wolf and his colleagues said. "These epidemiological results emphasize the need for broad availability of systematic and comprehensive genomic lung cancer diagnoses."