In a concept-tested study, an international scientific team led by researchers at the Kimmel Johns Hopkins Cancer Center has shown that a laboratory test that uses artificial intelligence tools has the potential to more accurately classify people with cysts Pancreatics will develop the cancers pancreas.
The test, called CompCyst (for the integral analysis of the cyst), incorporates molecular and clinical markers in cyst fluids, and seems to be on the way to significantly improve clinical and conventional image testing, according to the research team.
Using information from more than 800 patients with pancreatic cysts who had liquids and cyst cysts at Johns Hopkins Hospital and 15 other medical centers around the world, researchers say that CompCyst is more frequent than the current standard methods identified correctly what patients needed and probably an opportunity to benefit from the surgery and who would not benefit from the surgery or just needed follow-up. Specifically, they discovered that the use of the test would not have saved the surgery more than half of the patients that were subjected to a removal of cysts that later were considered unnecessary, since the cysts are unlikely to have caused cancer .
A description of the work is published in the July 17 issue of Translational medicine of science.
"Our study shows the potential role of CompCyst as a complement to the clinical and image criteria existing at the time of evaluating pancreatic cysts," says Anne Marie Lennon, MBB Ch., PhD. Medicine and Director of the clinical multidisciplinary Johns Hopkins clinic. "It could provide a greater degree of confidence to doctors when they advise patients who do not require follow-up and can be given surveillance."
"Although we still have to validate this test prospectively, our results are exciting because they document a new and more objective way of managing patients with this disease," he adds. There are plans to start a prospective validation study the following year.
Pancreatic cysts are common. According to other published investigations, it is found in 4% of people aged 60 years and 8% of people over 70 years of age. This means that about 800,000 people with a pancreatic cyst are identified each year in the United States. On the contrary, only a small fraction of the cysts progresses towards cancer.
"The dilemma faced by patients and their doctors is the ability to distinguish precancerous cysts from cysts that will not progress toward cancer," says Lennon.
"Clinical and image tests currently available do not manage to distinguish the precancerous cysts of cysts that have little or no potential to become cancerous, which makes it difficult to determine which patients do not need to be tracked and which patients will need to follow up long-term or immediate surgical resection, "study researcher Christopher Wolfgang, MD, Ph.D., MS, John L. Cameron, professor of surgery, director of surgical oncology at the Johns Hopkins Kimmel Cancer Center and co-director of precision medicine Johns Hopkins Center of excellence for pancreatic cancer. "As a result, essentially all people diagnosed with a cyst are still long-term. Surgeons are confronted to make recommendations to patients based on the risks and benefits of surgery with limited information. We rarely have cancer , but it is at the expense of executing an operation that, in retrospect, has not been necessary. This study directly addresses these fundamental problems in the management of pancreatic cysts. "
The study confirmed the precise nature of the cysts examined using the histopathological analysis of resected surgical specimens. Then, the cysts were classified into three groups: those that did not have the potential to become cancerous, for which patients would not require periodic follow-up; Cysts that produce mucins that have a small risk of progressing towards cancer, for which patients can receive periodic follow-up of progression towards possible cancer; and cysts for which surgery is recommended because there is a high probability of progression towards cancer.
The CompCyst test, developed by researchers led by the Johns Hopkins Kimmel Cancer Center, was created with patient data, including clinical impressions and symptoms, tomography images and molecular characteristics such as DNA alterations in the cyst fluid.
In the study, the researchers evaluated the molecular profiles, including DNA mutations and changes in chromosomes, of a large number (862) of pancreatic cysts. They then fueled molecular information, together with clinical and radiological data, in a computer-based program that used artificial intelligence to classify patients in the three previous groups.
Based on the histopathological analysis of surgically resected cysts, the researchers found that surgery was not needed for 45% of patients who underwent surgery for their cysts. This unnecessary surgery was performed because clinicians could not determine if cysts were dangerous. In these patients, if you had used CompCyst, the researchers estimated that 60% to 74% of patients (depending on the type of cyst) could have been spared by these unnecessary surgeries.
The study had several limitations, according to the researchers, including that the pancreatic cyst fluid was obtained at the time of the surgery, and that the cysts evaluated are more atypical than those observed in the usual clinical practice.
"We believe that CompCyst has the ability to substantially reduce unnecessary surgeries of pancreatic cysts. During the next five years, we hope to use CompCyst in many more patients with cysts in an effort to guide surgical treatment – to determine when surgery is needed and when I do not need to – and evaluate the performance of the test, "explains Bert Vogelstein, MD, Clayton Oncology Professor, co-director of the Ludwig Center Cancer Center John Hopkins Kimmel and a researcher at Howard Hughes Medical Institute.