BOSTON – With a little chemical or genetic snooping, or both, clinicians may be able to pinpoint the source of nearly all metastatic neuroendocrine tumors of the small bowel or pancreas.
By looking at expression patterns of four genes, investigators were able to determine that a surgically obtained metastatic neuroendocrine tumor (NET) originated in the small bowel with more than 96% accuracy, and, with the use of an immunohistochemistry algorithm, they identified the pancreas as the primary source of metastases in 10 of 10 cases.
"All the NETs that were misclassified by one method were correctly identified by the other method," said Dr. Jessica Maxwell of the department of surgery at the University of Iowa Hospitals and Clinics in Iowa City.
In about 15%-20% of cases of metastatic NETs, the primary tumor site is unknown, but is most likely to be in the small bowel or pancreas. Failure to identify the primary tumor site, despite optimal work-up, could delay referral for surgery or complicate choice of systemic medical therapies, she said at the annual meeting of the American Association of Endocrine Surgeons.
The authors tested the mettle of immunohistochemistry and gene expression classification (GEC) methods on 136 metastatic NETs collected intraoperatively from 97 patients with small-bowel NETs (38 with metastases to liver and 59 with metastases to lymph nodes) and 39 with pancreatic NETs (17 liver and 22 lymph node metastases).
The GEC uses quantitative or "real-time" polymerase chain reaction (qPCR) to evaluate expression of four key genes, encoding for the secretin receptor (SCTR), oxytocin receptor (OXTR), bombesin-like receptor-3 (BRS3), and opioid receptor kappa-1 (OPRK1).
The differential patterns of gene expression mark the metastases as originating either in the pancreas or small bowel.
They also tested a two-tiered immunohistochemistry algorithm using the markers CDX2, PAX6, and ISLET1 for tier 1, and PrAP, PRm NESP55, and PDX1 in tier 2. They tested the algorithm on six primary tumors and validated their findings on 37 metastases.
The immunohistochemistry method can identify a primary tumor site with as few as three markers, but if the findings are indeterminate, the addition of the four tier 2 markers can help to nail down the tumor site, Dr. Maxwell said.
They found that the GEC accurately identified 94 of 97 small bowel NETs (96.9%), and 34 of 39 pancreatic NETS (87.2%).
In contrast, the immunohistochemistry algorithm correctly identified the primary site in 23 of 27 small bowel metastases (85.2%), and in 10 of 10 (100%) pancreatic metastases.
When the methods were compared head to head in 27 metastases, GEC had a 96.2% overall accuracy and immunohistochemistry an 85.2% accuracy.
As noted before, the methods were complementary, with all NETs misclassified by one method called accurately by the other.
The investigators suggest that because the methods are highly accurate and complementary, they may best be used sequentially, starting with immunohistochemistry which is both inexpensive and widely available, and if immunohistochemistry fails, moving on to GEC.
"Sequential use allows for identification of nearly all metastatic neuroendocrine tumors from small bowel or pancreatic sites," Dr. Maxwell said.
In the discussion, Dr. Eren Berber of the Center for Endocrine Surgery at the Cleveland Clinic, who was not involved in the study, questioned whether knowing the primary site had any practical implications for surgeons.
Dr. Maxwell noted that some pancreatic NETs are not detected by preoperative studies and that given the risks of pancreatectomy or pancreaticoduodenectomy, accurately identifying the source of an NET may be helpful for patient counseling and preoperative planning.
The study was supported by a grant from the National Institutes of Health. Dr. Maxwell and Dr. Berber reported having no financial disclosures.