A New Method For Cancer Diagnosis: A Window Into How To Interpret Medical Articles

I want to tell you about a new technology for cancer detection and monitoring. This is called a liquid biopsy. Liquid biopsy is a blood test to detect cancer. Even though a solid tumor has a specific location in the body, sometimes cancer cells and other cellular debris break off from the tumor and circulate in your blood. This test detects these free-floating cells and/or cellular debris. What makes a liquid biopsy so exciting, is it has the ability to test for 50+ cancer types with one blood test.   

However, in order to tell you more about this blood test, we first have to learn how to gauge the accuracy of a diagnostic test in general.

You may think if you get a positive result on a test, that means for sure you have the disease the test was testing for. However, the truth is that a test can give you either a false positive (a test telling you that you have a disease when you don’t) or a false negative (a test telling you that you don’t have a disease when you do). The accuracy of a diagnostic test is measured primarily with 2 values: sensitivity and specificity.

Sensitivity: is the odds that a test can detect a disease among a disease carrier. In other words, if you had 100 people that you know beforehand have a disease, the sensitivity of the test would be how many people out of a 100 this test could detect.

Specificity: is the odds that a test can correctly identify a non-diseased person. In other words, if you had 100 people that you knew did not have a disease, the specificity of the test would be how many people out of 100 this test could correctly report as not having the disease.

Why do you need 2 numbers to measure the accuracy of a test? Sensitivity seems valuable, but why do we need specificity?

Let's say I told you I invented a test. I told you I could detect 100% of all cancers without even so much as a needle prick. Would you be interested? Chances are you would think such a test would be revolutionary. However, what if I told you my test was to measure someone with a ruler, and everyone over 1 foot tall had cancer. Chances are you would be less impressed, to say the least. However, if you think about it, this test would detect 100% of all cancers: as you would be unlikely to find someone under 1 foot tall, the test would give a positive result for all people being tested, and hence it wouldn’t miss a single case of people that actually did have cancer. However, you might reply, “but wait, this test would give a positive result for everyone, even people without cancer, that test would be useless”. And you would be right. This is where specificity comes in. Because while sensitivity measures the ability of a test to correctly identify a diseased patient, specificity measures a test’s ability to identify a healthy patient. So in the example I gave you, although my ruler test would have a high sensitivity, it would have a very low specificity (and hence, once again, would be useless).

In general, when sensitivity goes up, specificity goes down. This is because as you increase sensitivity, you are lowering the threshold for saying a patient has a disease, and so you will get more false positives, and hence specificity will decrease. This means tests with high sensitivity make good initial screening tests. However, as doctors, if a test is testing for a serious medical condition, we generally don’t just assume that the test is accurate. In many cases, we will order another follow-up test with a high specificity to confirm the diagnosis.

So why not just start with the test with a high specificity? Well because the opposite is also true: as specificity goes up, often sensitivity goes down. So if we start with a test with higher specificity, the sensitivity will suffer, and we run the risk of missing the disease diagnosis altogether.

So what is a good sensitivity and specificity? The very unsatisfying answer is that it depends. The best way to get a lay of the land is to compare whatever test you are looking at against what is called the gold standard. The gold standard is the test that is commonly accepted by the medical community as the go-to test to detect the disease you are interested in. So to do this, let's go back to the liquid biopsy technology.

For some cancers, a liquid biopsy has the potential to be game-changing for detection. Take, for example, pancreatic cancer. The challenge with pancreatic cancer is that most of the time, by the time it has been diagnosed it is too late. Meanwhile, the sensitivity and specificity of a liquid biopsy are 93% and 92% respectively [1]. For comparison, CT, one of the mainstays of diagnosing pancreatic tumors, only has a sensitivity and specificity in the low 80%s.  For some cancers, the liquid biopsy fails to perform as well as the gold standard for detection. The gold standard for diagnosing colon cancer is a colonoscopy. According to Uptodate, a leading resource for the synthesis of well-established medical knowledge (like a version of google that doctors use), the sensitivity of a colonoscopy in detecting colon cancer is 95% and the specificity of a colonoscopy is 86% [2]. The sensitivity of a liquid biopsy for colon cancer is 79%-86% depending on the method [3]. So for colon cancer, a colonoscopy is still the preferred screening method of colon cancer screening. BUT: many people find the colonoscopy process cumbersome. It is both an invasive procedure and the prep is unpleasant. Colon cancer is the 3rd leading cause of all cancer deaths in the US. So less invasive testing methods lower the barriers to colon cancer detection and hence are a benefit.

Hopefully the above ramblings have been helpful and the technical details are not too dry. Please let me know if I hit the mark with the amount of technical detail I included in this article or if I need to include more/less. Thank you!

Joshua Engle, MD

References:

1.)  Zhu Y, Zhang H, Chen N, Hao J, Jin H, Ma X. Diagnostic value of various liquid biopsy methods for pancreatic cancer: A systematic review and meta-analysis. Medicine (Baltimore). 2020;99(3):e18581. doi:10.1097/MD.0000000000018581

2.)  URL: https://www.uptodate.com/contents/image/print?imageKey=PC%2F116366. Google search: “uptodate, colonscopy, sensitivity and specificity”. Uptodate original  data sources:

a.)  Zauber A, Knudsen A, Rutter CM, et al. Evaluating the Benefits and Harms of Colorectal Cancer Screening Strategies: A Collaborative Modeling Approach. AHRQ Publication No. 14-05203-EF-2. Rockville, MD: Agency for Healthcare Research and Quality; October 2015.

b.)  Knudsen AB, Zauber AG, Rutter CM, et al. Estimation of Benefits, Burden, and Harms of Colorectal Cancer Screening Strategies: Modeling Study for the US Preventive Services Task Force. JAMA 2016; 315:2595.

3.)  Zhu Y, Yang T, Wu Q, et al. Diagnostic performance of various liquid biopsy methods in detecting colorectal cancer: A meta-analysis. Cancer Med. 2020;9(16):5699-5707. doi:10.1002/cam4.3276