Genetic Study Reveals Clues to Pancreatic Neuroendocrine Tumors
Researchers at the Johns Hopkins Kimmel Cancer Center have sequenced the genes of patients with an uncommon form of pancreatic cancer, producing the first overview of the genetic basis of the disease, called neuroendocrine tumors. The investigators identified several genes that may provide prognostic information, as well as mutations in genes in a signaling pathway for which targeted therapies currently exist. Their findings appeared online in Science Express on January 20.
Patients whose tumors had mutations in three commonly mutated genes lived longer than patients without these mutations, the researchers found. In addition, some patients had mutations in genes that encode proteins in the mTOR pathway. These individuals could be candidates for drugs such as rapamycin (everolimus), which block growth-promoting signals through the mTOR pathway.
“It’s really exciting that some potential clinical applications have emerged from our first report on this disease,” said lead investigator Dr. Nickolas Papadopoulos, who directs translational genetics at the Johns Hopkins Ludwig Center for Cancer Genetics & Therapeutics. But more follow-up work is needed, he noted, such as a clinical trial to evaluate the use of mTOR inhibitors in selected patients.
“This study is a very important step forward,” said Dr. Philip A. Philip, who studies cancers of the pancreas at the Barbara Ann Karmanos Cancer Institute in Detroit, MI, and was not involved in the research. “The results underscore the fact that the only way we can make any headway in our search for more effective therapies is to understand disease biology at a level that is much higher than we know so far—and that applies to all cancers.”
Pancreatic neuroendocrine tumors, also called islet cell tumors, make up less than 5 percent of all pancreatic cancers. These patients have a better prognosis than patients who develop what is most commonly called pancreatic cancer, ductal adenocarcinoma of the pancreas. But some neuroendocrine tumors grow silently and are discovered only after they have spread to other organs, so more effective treatments beyond surgery are needed.
To gain insights into the genetic basis of the tumor, the researchers sequenced nearly all of the protein-coding genes in tumor and normal tissue samples from 10 patients with pancreatic neuroendocrine tumors. The patients did not have an inherited form of the disease.
The three most commonly mutated genes in the 10 tumors—MEN-1, DAXX, and ATRX—all play a role in the packaging of DNA. Known as chromatin remodeling, this is an epigenetic process that controls the activity of genes without causing changes in DNA sequence. But mutations in genes involved in chromatin remodeling may affect the regulation of genes elsewhere in the genome.
“We are seeing a connection between genetics and epigenetics, and also how the genetics may actually confer the epigenetic changes we’ve known about for so long,” said Dr. Papadopoulos. His group recently published genome studies of ovarian cancer and medulloblastoma, and in both cases mutations were found in genes involved in epigenetic processes. (See “Seeking Better Treatments for Brain Tumors in Children” in this issue.)
The stability of the chromatin is essential for the normal expression of genes, noted Dr. Laufey Amundadottir of NCI’s Division of Cancer Epidemiology and Genetics (DCEG), who co-led a genome-wide association study of pancreatic cancer. “When chromatin is not maintained in its original form,” she explained, “gene expression can go awry.”
After identifying gene mutations in the 10 samples, the Hopkins team sequenced these genes in samples from 58 additional patients with pancreatic neuroendocrine tumors. Among the 68 tumors, 44 percent had somatic (non-inherited) mutations that inactivated MEN-1, and 43 percent had mutations in either DAXX or ATRX, which produce proteins that are components of the same molecular complex.
Patients with mutations in MEN1 or DAXX/ATRX tended to live longer than patients without the mutations. The difference was particularly striking among patients who had metastatic disease and whose tumors had mutations in both MEN1 and DAXX/ATRX. These patients all survived at least 10 years, whereas more than 60 percent of the patients without the mutations died within 5 years of diagnosis.
Patients with mutations in MEN1 and DAXX/ATRX may have a biologically distinct subtype of pancreatic cancer, the researchers suggested, and this could explain the difference in survival.
Another category of tumors—14 percent of the total—were those with mutations in mTOR pathway genes, including TSC2, PTEN, and PIK3CA. “This finding obviously needs to be validated in large clinical trials, but it is an exciting example of ‘personalized medicine,’” said co-author Dr. Ralph Hruban, who directs the Sol Goldman Pancreatic Cancer Research Center at Johns Hopkins.
Additional evidence that mTOR is a relevant target for treating some patients with pancreatic neuroendocrine tumors comes from several recent clinical trials, noted Dr. Jack Welch of NCI’s Division of Cancer Treatment and Diagnosis (DCTD). Preliminary results reported last year indicated that everolimus improved progression-free survival in a subset of patients, while another study suggested that everolimus can treat patients whose disease progresses despite chemotherapy.
Based on these results, the NCI Clinical Trials Cooperative Group Program has launched a national phase II clinical trial (CALGB-80701) of everolimus with or without bevacizumab in patients with locally advanced or metastatic pancreatic neuroendocrine tumors.
Additional studies are under development, noted Dr. Welch. “These trials will include correlative laboratory studies to expand the understanding of the underlying biology of this disease in patients like those in the Johns Hopkins study,” he said.
In 2008, the Johns Hopkins team surveyed tumors from patients with pancreatic ductal adenocarcinoma. With the new results, it is now clear that there are some obvious genetic differences between this disease and pancreatic neuroendocrine tumors. For example, fewer genes were mutated in pancreatic endocrine tumors than in pancreatic ductal adenocarcinomas, and the most commonly mutated genes in each type were different.
One possible explanation, the researchers suggest, is that mutations in each of these cancers may arise through different mechanisms, perhaps due to exposure to different environmental carcinogens or through the action of different pathways for repairing damaged DNA.“This study really demonstrates the power of new DNA sequencing technologies,” said Dr. James V. Tricoli of NCI’s Cancer Diagnosis Program, who was not involved in the research. “As more groups start to use these approaches, we are going to identify additional mutations that can not only subcategorize tumors but also reveal to us better ways to treat these patients—and that’s the most important outcome.”