For several years, researchers have pursued a test that offered cancer detection before symptoms ever appeared. If this could be developed for mass use, then it could save billions in healthcare expenditures in addition to lives. An ounce of prevention is better than a pound of cure. But developing a blood test for cancer that is both sensitive and specific has proven tough. Wvery year, scientists get a little close to making this a reality. And machine learning techniques are helping.
Recently, researchers completed an extensive longitudinal study spanning nearly a decade. In the study, asymptomatic people were followed over time and monitored for cancer development. By combining machine learning with unique blood tests for cancer, they were able to accurately predict cancer occurrence. In fact, their cancer detection ability was 91 percent with it being specific to the cancer type in 95 percent. Naturally, this has everyone who is involved in cancer prevention testing quite hopeful.
“We can’t say for sure that the patients didn’t have any symptoms, but we detected the cancer years before they ever walked into the hospital.” – Kun Zhang, PhD, Professor of Bioengineering, University of California San Diego
The Cancer Detection Research Study
The research described was performed in China and involved over 123,000 healthy subjects. Between 2007 and 2014, a blood test for cancer was administered to the subjects periodically. The results of these tests were then matched to patient histories over time using cancer registry data. In patients who eventually presented with cancer, the blood test for cancer identified 91 percent. The time between cancer detection on the test and their presentation to a clinic was between 1 and 4 years. In other words, the test predicted cancer long before any symptoms appeared.
The most interesting aspect of the study involved how the blood test for cancer, called PanSeer, was developed. The researchers used machine learning to help develop the test by providing data regarding healthy people and those with cancer. The protocol only involved 5 types of cancer, which included stomach, esophageal, bowel, lung, and liver cancers. Based on the data, a testing model was designed. Specifically, the cancer detection involves looking for a specific DNA methylation signature in the blood. And based on the results of the study, it looks to be highly accurate.
“For any test, it is always more prudent to begin with a high-risk population. You want to see some benefit with the high-risk population first, and then it can slowly be extended to others at lower risk.” – Kun Zhang, PhD
Other Screening Blood Tests for Cancer
The research was conducted at the University of California in San Diego. However, researchers worked in conjunction with Singlera Genomics, based in La Jolla, to develop PanSeer. Their study is unique since it’s the first to examine individuals over a prolonged period of time before cancer appeared. Other researchers have studied cancer detection tests using samples for patients already diagnosed. And while their results have also been promising, they are not as robust as Singlera’s data.
Several other companies have reported progress in cancer detection as well. For example, CancerSEEK offers a blood test for cancer that screens for 8 different types of cancer. Like PanSeer, it also using DNA methylation and protein markers to identify possible cancer cells. Grail is another company whose blood test for cancer checks for 12 cancer types. However, its cancer detection sensitivity is reported to be only 67 percent. Finally, CellMax Life FirstSight blood test for cancer is specific for colorectal cancer and adenomas. Though promising, it similarly has a lower detection rate as well. So far, the best results have come with this most recent study.
“If there is a positive finding [on PanSeer], then which cancer do you look for? It has an issue in that regard, and that’s the problem with this type of [cancer detection] test, as it is easier if there is one site of origin.” – Benjamin Weinberg, MD, Assistant Professor of Medicine, Division of Hematology and Oncology, Georgetown University
Challenges with Cancer Detection Tests
In an ideal world, it would be nice to identify cancer as soon as the first cell appeared. In this way, therapies could be administered when they might be most effective. But this isn’t a practical approach for a couple of reasons. For one, cancer detection of such a small volume of cells would be challenging to say the least. But also, the body’s immune system often eliminates cancer cells on its own. A blood test for cancer that is too sensitive might therefore subject people to unnecessary treatments. This violates the first rule of medicine, which is to first do no harm.
In addition to these challenges, the current tests available screen for several cancer types. Should a blood test for cancer return a positive result, the difficulty then is knowing where to look. Even for PanSeer, which checks for 5 different cancer types, it might require extensive testing to locate a potential cancer. That is why it’s important for cancer detection methods to be both sensitive and specific. Before widespread blood testing for cancer can be considered, these issues must be further resolved.
“This is something that’s used to assess for minimal residual disease in patients who have undergone surgery and appear to be ‘cured’ of the disease. The test is looking to see if there is any circulating tumor DNA being shed from whatever tumor is left behind.” – Benjamin Weinberg, MD
Reasons to Be Optimistic
Despite the challenges described, the results of these latest research findings are exciting for 2 important reasons. First, the ability to leverage machine learning to advance cancer research suggests tremendous innovations may be around the corner. And secondly, tests like PanSeer may offer immediate benefit to people who have already undergone treatment and need close monitoring. This type of blood test for cancer could provide early detection of recurrence that could guide better care. All of this provides reasons why optimism and hope surround these latest cancer research findings.
To read more about one person’s heroic battle against cancer, check out Ed Kopko’s PROJECT BOLD LIFE: The Proven Formula to Take on Challenges and Achieve Happiness and Success, which delves into Pam Sloate’s personal fight.