Two independent teams have sequenced small-cell lung cancer exomes and transcriptomes, identifying several potential therapeutic leads.
Both teams — one from Genentech and Johns Hopkins University and the other comprising a large international consortium led by the University of Cologne in Germany — published their findings this week in Nature Genetics.
As a result of the findings from the University of Cologne team, researchers have now started a clinical trial for small-cell lung cancer patients with FGFR1 inhibitors, Roman Thomas, senior author of the study, told Clinical Sequencing News.
Thomas, who chairs the Department of Translational Genomics at the University of Cologne, said that the clinical trial is being set up as a "direct result of the study," which found that the FGFR1 tyrosine kinase gene was amplified in about 6 percent of patients.
The trial will assess the safety and tolerability of FGFR1 inhibitors, which are already being tested in squamous-cell lung cancer patients.
Meantime, among the findings from the Genentech and Johns Hopkins team were alterations within the PI3K pathway, for which Genentech has compounds in trial, Somasekar Seshagiri, a principal scientist at Genentech, told Clinical Sequencing News. However, the findings need to be validated and functional studies in cell lines will have to be done to assess the compounds' effectiveness, he said.
The studies, among the first to genomically characterize small-cell lung cancer, "really give us a picture of the landscape of this very deadly tumor type," said Charles Rudin, director of the FAMRI Center of Excellence at Johns Hopkins University.
"The next phase will be to turn that into better treatment," he said, "and that still remains a major challenge."
Small-cell lung cancer comprises around 15 percent of all lung cancers and typically occurs in smokers. Tumor genomes tend to be highly mutated and, as researchers from the Wellcome Trust Sanger Institute reported in 2009, often bear mutational signatures left by tobacco carcinogens (IS 12/22/2009).
The cancer is aggressive, often metastasizes, and leads to early death. There is currently no effective targeted therapy.
To further study this cancer, the German team sequenced the exomes of 27 tumor samples and two cell lines, two tumor whole genomes, and 15 transcriptomes from tumor samples.
In all cases, they found that known tumor suppressors TP53 and RB1 were inactivated. FGFR1 amplifications were identified in 6 percent of cases. Additionally, the histone modification pathway was significantly mutated, with recurrent mutations to the CREBBP, EP300, and MLL genes.
The tumors were also highly mutated, with an average of 7.4 protein-changing mutations per million base pairs. Moving forward, the team plans over the next year to sequence at least 100 more small-cell lung cancer whole genomes and transcriptomes on the Illumina HiSeq 2000, Thomas said.
To identify relevant pathways and mutated genes from the thousands of mutations identified, the team applied a series of filters that enabled them to home in on the likely drivers and significant pathways.
"The major technical advance in our paper is to create statistical filters that sort out the genes that are unlikely to be of biological relevance," Thomas said.
In this regard, the RNA-seq data was critical, because it allowed the team to filter out mutations to genes that were not expressed. Of 22 significantly mutated genes, only six were expressed.
Transcriptome sequencing also allowed the team to identify gene fusions, amplifications, and rearrangements.
Next, they looked at the geographical location of mutations, identifying significant mutation clusters. "The reason for that," Thomas explained, is that "the prototypic oncogene and tumor suppressor[s], if mutated, typically bear areas of recurrent mutations. So we looked for genes where mutations were not randomly distributed across the gene, but really clustered in certain areas."
The identification of CREBBP is a good example of a gene that passed these filters, Thomas said. Mutations were clustered in functional areas of the gene, and the team also identified a gene fusion involving CREBBP, he said. EP300 also contained mutations clustered in functional areas.
Both of those genes are involved in the histone modification pathway, which was the most significantly mutated pathway, Thomas said.
The next step will be to further study these genes and the functional consequences of the mutations, "hopefully to identify cellular functions that can be targeted by inhibitors," Thomas said.
The finding that all tumors contained mutations to TP53 and RB1, leading to the loss of function of those genes, may be important for explaining the biology of the cancer, Thomas said. However, it is unclear whether these genes can be targeted by drugs. So far, attempts at targeting TP53, which is a well-known tumor suppressor involved in many types of cancer, have been unsuccessful. However, there are some strategies that are currently being tested at the preclinical stage.
Thomas said that aside from sequencing many more genomes, the team is pursuing several other avenues of research, including trying to link the genomic findings to clinical outcomes.
For instance, some patients are initially very sensitive to chemotherapy, but then relapse quickly, and "that's something we don't understand on a mechanistic level."
Additionally, the researchers will further study the role of histone modification in small-cell lung cancer as well as begin doing functional studies in cell lines and mice.
Similar to the University of Cologne team's findings, the Johns Hopkins and Genentech researchers identified mutations to TP53 and RB1, although not in all cases. They also identified the histone modification and chromatin remodeling pathway to be important, with mutations to EP300 and MLL2, among other genes in that pathway.
However, they reported their most significant finding to be mutations to the SOX2 gene, which were found in 27 percent of all cases and appeared to be correlated with clinical outcome.
The Genentech team sequenced the exomes and transcriptomes of 36 primary tumors and matched normal samples, 17 cell lines and matched normals, as well as four primary tumors and 23 cell lines without matched normal data. The team also sequenced one whole genome and matched normal sample.
The SOX2 gene is particularly interesting because it is "known to be a key regulator of stem cell differentiation," said Rudin. In the study, the researchers found SOX2 to be amplified in 27 percent of cases, and when they looked at an independent set of 110 tumors, "it correlated with more advanced-stage disease, although it needs to be confirmed" he added. "That tells us it may be an important driver."
Additionally, the team was able to demonstrate in cell lines that suppressing the gene inhibited proliferation, suggesting a potential avenue to pursue therapeutically.
However, SOX2 is a transcription factor, which are notoriously difficult to target, Rudin said.
The team also found a recurrent fusion between RLF and MYCL1 that appeared to be important. When MYCL1 was silenced in cell lines with the fusion, cell proliferation decreased. The MYC gene family has been implicated in cancer for a long time, said Rudin, but, like SOX2, it is a transcription factor and difficult to target, he said.
Seshagiri added that there is currently work underway in figuring out how to "block protein-DNA interactions, but these are all very early stages," he said, and don't have "the rigor that's needed for drug development."
In total, the team identified 22 significantly mutated genes, some of which overlapped with the German group's findings.
Transcriptome sequencing was critical not only for figuring out which mutations were actually being expressed, but also for identifying fusions and confirming amplifications, Seshagiri said.
Seshagiri said that Genentech researchers are now going through the list of significantly mutated genes and fusions to validate the findings and also to search for either targets of compounds the company is already working on, or targets for which drugs could be developed.
In particular, Seshagiri noted that there were alterations in the PI3K pathway that Genentech already has compounds in trials for, but "we'll need to delve into this deeper and do a more comprehensive analysis," he said.
According to the company's development pipeline, it has four PI3K inhibitors in clinical trials: GDC-0032, GDC-0084, GDC-0941, and GDC-0980. None are currently in trials for small-cell lung cancer.