Examining the Promise of Multicancer Early Detection Tests: Overview and Call for RCTs

In this first episode of OncChats: Examining the Promise of Multicancer Early Detection Tests, Toufic A. Kachaamy, MD, Madappa Kundranda, MD, PhD, and Niloy Jewel J. Samadder, MD, provide an overview on multicancer early detection tests and the need for randomized clinical trials (RCTs) to further explore their use.

Kachaamy: Hi, everyone. I’m Toufic Kachaamy, the chief of medicine at City of Hope in Phoenix, Arizona. We’re joined today by Dr Madappa Kundranda, the division chief of oncology at Banner MD Anderson and Dr Niloy Jewel Samadder, who is a professor of medicine at Mayo Clinic, College of Medicine. He is [also] the vice chair of medicine at Mayo Clinic, Arizona, and the enterprise co-director of individualized precision oncology at Mayo Clinic Comprehensive Cancer Center. His area of expertise is the use of genomics in cancer, and he does a lot of research on hereditary cancer syndromes and endoscopic oncology.

Today, we will be talking about multicancer early detection. We all know that cancer is expected to become the number one killer in many countries, and the United States is no exception. In 2020, for example, 10 million people died from cancer. To put things into perspective, 7 million [people] died from the COVID-19 pandemic and we put the world on hold for that. Every year, more people die from cancer than [those who] died from the COVID-19 pandemic in 3 years.

One of the major issues in fighting cancer is finding it late. For most cancers, the higher the stage, the less of a chance [there is for] long-term remission or cure. A promising tool in the fight against cancer is finding a test for early detection; [this] will be a big step in the right direction. In the United States, we currently screen for breast cancer, colorectal cancer [CRC], prostate cancer, and in some instances, in those who are high risk, for lung cancer. However, over 60% of cancer-related deaths come from cancers that we do not screen for. Dr Samadder, can you start by giving us an overview of where we are in multicancer early detection?

Samadder: Well, first, let me start by saying thank you, Dr Kachaamy, for inviting me onto this web video discussion today. This is a very hot area, but it is important for us to know the data, the science behind this technology, how it can be applied to our patients, and what is needed to move it from where it is today, at its infancy, to real-world use of it. I want to first reiterate the point you made that we only have screening available for 4 or 5 major cancers: colorectal, breast, cervical, some indications in prostate cancer, and lung cancer in high-risk groups, namely smokers. Seventy percent of cancers are not screened for, so the vast majority of cancers have no screening paradigm whatsoever that could identify an early-stage, potentially curable cancer.

So, what is multicancer early detection? Well, this is felt to be the holy grail of cancer medicine. Can we identify on a single blood sample that we would take from a patient once a year, or every 6 months, or at whatever frequency, and screen them for detection of 20, 30, 50 different types of cancers all caught at an early stage when they are highly curable without the use of systemic toxic therapies or life-altering invasive surgeries? The way this technology works is that it analyzes a combination of circulating free DNA that is being shed from any tumor that is growing in [the] body; methylation, which is a change in the various genes and proteins expressed by tumors; and protein biomarkers that are present in the blood. [This is] a machine algorithm or an [artificial intelligence] AI system that analyzes all the data from the circulating tumor DNA signature, methylation signature, and protein biomarkers and uses machine learning to identify what is a cancer signal vs just normal background in the human body.

Even within that cancer signal, can we predict site of origin of the cancer? Because it’s not enough just to say, “Red light, you have cancer.” It could be anywhere in [the] body. We need a test that is specific enough to say, “Well, we think that cancer is located in the gastrointestinal [GI] tract.” That then leads to further imaging, endoscopy, or other procedures that can identify whether that tumor or precancer is in the upper GI system like the stomach or esophagus, the small bowel, the colon, the pancreas, or some other related organ.

Kachaamy: One of the things that I was interested in is how the tests are developed and how AI is used. You kind of touched on that. The other question I have for you is, there are multiple tests that we hear about, and whether [they are] for the consumer or even for the prescribing physician, are all multicancer early detection tests good? Are they all good enough? Is there one that’s better than the other? Give us your insight on that.

Samadder: Thank you. [Those are] great questions. I’m going to start by saying that not all these tests are [for] multicancer detection. There are tests that are being defined and created for single cancers, or possibly a group of cancers all within a few organ systems like a GI cancer–specific test or a women’s cancer–specific test. As such, there are a group of tests that are going to screen for 20, 30, or more cancers, and a group of tests that will probably screen for between 1 and 5 cancers. There are tradeoffs to each of them.

Let’s step back and look at the data. There are about 3 major tests that are either on the market as a [laboratory-developed] test [LDT] that is available or is very close to being launched. I won’t mention any of those 3 tests, but people can Google and see which vendors they are [from] and the names of those tests. All 3 of these tests are multicancer early detection tests, so [they test] for 10-plus cancer types. Most of these tests have been developed by companies that either have a background in sequencing technology, so they are very good with the history of how to sequence DNA and have large repositories of DNA samples from patients with or without cancer, or from companies that have a lot of expertise in early cancer detection, for CRC, for example.

But to date, what [are] the data to support the use of these tests? Well, that is really the Achilles heel. Most of these companies have really done, in my opinion, limited science to date, using largely case-control studies, or short, prospective follow-up studies. What is the gold standard in medicine when we talk about taking a drug or any other treatment or surgical procedure from discovery, to validation, to patient bedside? It is a RCT. This involves taking a large number of patients, randomizing them so half of them will receive treatment with the new technology or experimental drug and half of them will get a placebo or a standard-of-care procedure or test. [Then, investigators need to] follow these patients for long enough to see which pathway ended up preventing disease, reducing the burden of disease, or reducing mortality or death as the ultimate indicator. [Those] RCTs [are] what is lacking in this field. Most of the studies to date have [been] very short term, [with] follow-up of 1, [or a] maximum 2 or 3 years, [where] patients have been offered testing with one of these multicancer early detection technologies without a control or placebo group. [These efforts] just offered them testing, and then saw how often the blood test, for example, picked up an earlier-stage cancer or one that would not have been detected otherwise. That really is not sufficient proof.

Sufficient proof would be an RCT, where half the patients in a large trial [would receive] the multicancer early detection test and half the patients [would] not receive the test but [instead would get] standard-of-care screening, including mammograms and colonoscopy or other stool-based colon cancer screening, lung and prostate screening as indicated, as well as cervical cancer screening. [This would follow] them out for at least 5 years, possibly 10 years, [to see] which strategy ultimately led to earlier diagnosis of cancer; shifting of the stage of cancer such that you’re detecting earlier-stage cancers which are potentially curable; and if the trial is run long enough, [which strategy improved] cancer-related mortality.

Check back on Wednesday for the next episode in this series.