Jimmy Lin: A simple new blood test that can catch cancer early

Recorded atApril 24, 2017
Duration (min:sec)11:56
Video TypeTED Stage Talk
Words per minute179.02 medium
Readability (FK)48.9 difficult
SpeakerJimmy Lin

Official TED page for this talk


Jimmy Lin is developing technologies to catch cancer months to years before current methods. He shares a breakthrough technique that looks for small signals of cancer's presence via a simple blood test, detecting the recurrence of some forms of the disease 100 days earlier than traditional methods. It could be a ray of hope in a fight where early detection makes all the difference.

Text Highlight (experimental)
100:13 Cancer.
200:15 Many of us have lost family, friends or loved ones to this horrible disease.
300:19 I know there are some of you in the audience who are cancer survivors, or who are fighting cancer at this moment.
400:25 My heart goes out to you.
500:27 While this word often conjures up emotions of sadness and anger and fear, I bring you good news from the front lines of cancer research.
600:36 The fact is, we are starting to win the war on cancer.
700:41 In fact, we lie at the intersection of the three of the most exciting developments within cancer research.
800:47 The first is cancer genomics.
900:49 The genome is a composition of all the genetic information encoded by DNA in an organism.
1000:55 In cancers, changes in the DNA called mutations are what drive these cancers to go out of control.
1101:01 Around 10 years ago, I was part of the team at Johns Hopkins that first mapped the mutations of cancers.
1201:07 We did this first for colorectal, breast, pancreatic and brain cancers.
1301:12 And since then, there have been over 90 projects in 70 countries all over the world, working to understand the genetic basis of these diseases.
1401:21 Today, tens of thousands of cancers are understood down to exquisite molecular detail.
1501:28 The second revolution is precision medicine, also known as "personalized medicine."
1601:32 Instead of one-size-fits-all methods to be able to treat cancers, there is a whole new class of drugs that are able to target cancers based on their unique genetic profile.
1701:42 Today, there are a host of these tailor-made drugs, called targeted therapies, available to physicians even today to be able to personalize their therapy for their patients, and many others are in development.
1801:55 The third exciting revolution is immunotherapy, and this is really exciting.
1902:00 Scientists have been able to leverage the immune system in the fight against cancer.
2002:05 For example, there have been ways where we find the off switches of cancer, and new drugs have been able to turn the immune system back on, to be able to fight cancer.
2102:14 In addition, there are ways where you can take away immune cells from the body, train them, engineer them and put them back into the body to fight cancer.
2202:24 Almost sounds like science fiction, doesn't it?
2302:27 While I was a researcher at the National Cancer Institute, I had the privilege of working with some of the pioneers of this field and watched the development firsthand.
2402:36 It's been pretty amazing.
2502:38 Today, over 600 clinical trials are open, actively recruiting patients to explore all aspects in immunotherapy.
2602:46 While these three exciting revolutions are ongoing, unfortunately, this is only the beginning, and there are still many, many challenges.
2702:54 Let me illustrate with a patient.
2802:58 Here is a patient with a skin cancer called melanoma.
2903:00 It's horrible; the cancer has gone everywhere.
3003:05 However, scientists were able to map the mutations of this cancer and give a specific treatment that targets one of the mutations.
3103:13 And the result is almost miraculous.
3203:16 Tumors almost seem to melt away.
3303:19 Unfortunately, this is not the end of the story.
3403:22 A few months later, this picture is taken.
3503:25 The tumor has come back.
3603:28 The question is: Why?
3703:30 The answer is tumor heterogeneity.
3803:33 Let me explain.
3903:36 Even a cancer as small as one centimeter in diameter harbors over a hundred million different cells.
4003:43 While genetically similar, there are small differences in these different cancers that make them differently prone to different drugs.
4103:51 So even if you have a drug that's highly effective, that kills almost all the cells, there is a chance that there's a small population that's resistant to the drug.
4204:00 This ultimately is the population that comes back, and takes over the patient.
4304:05 So then the question is: What do we do with this information?
4404:08 Well, the key, then, is to apply all these exciting advancements in cancer therapy earlier, as soon as we can, before these resistance clones emerge.
4504:19 The key to cancer and curing cancer is early detection.
4604:24 And we intuitively know this.
4704:25 Finding cancer early results in better outcomes, and the numbers show this as well.
4804:31 For example, in ovarian cancer, if you detect cancer in stage four, only 17 percent of the women survive at five years.
4904:39 However, if you are able to detect this cancer as early as stage one, over 92 percent of women will survive.
5004:46 But the sad fact is, only 15 percent of women are detected at stage one, whereas the vast majority, 70 percent, are detected in stages three and four.
5104:57 We desperately need better detection mechanisms for cancers.
5205:01 The current best ways to screen cancer fall into one of three categories.
5305:06 First is medical procedures, which is like colonoscopy for colon cancer.
5405:11 Second is protein biomarkers, like PSA for prostate cancer.
5505:15 Or third, imaging techniques, such as mammography for breast cancer.
5605:22 Medical procedures are the gold standard; however, they are highly invasive and require a large infrastructure to implement.
5705:30 Protein markers, while effective in some populations, are not very specific in some circumstances, resulting in high numbers of false positives, which then results in unnecessary work-ups and unnecessary procedures.
5805:45 Imaging methods, while useful in some populations, expose patients to harmful radiation.
5905:51 In addition, it is not applicable to all patients.
6005:54 For example, mammography has problems in women with dense breasts.
6105:58 So what we need is a method that is noninvasive, that is light in infrastructure, that is highly specific, that also does not have false positives, does not use any radiation and is applicable to large populations.
6206:14 Even more importantly, we need a method to be able to detect cancers before they're 100 million cells in size.
6306:20 Does such a technology exist?
6406:22 Well, I wouldn't be up here giving a talk if it didn't.
6506:26 I'm excited to tell you about this latest technology we've developed.
6606:31 Central to our technology is a simple blood test.
6706:34 The blood circulatory system, while seemingly mundane, is essential for you to survive, providing oxygen and nutrients to your cells, and removing waste and carbon dioxide.
6806:45 Here's a key biological insight: Cancer cells grow and die faster than normal cells, and when they die, DNA is shed into the blood system.
6906:55 Since we know the signatures of these cancer cells from all the different cancer genome sequencing projects, we can look for those signals in the blood to be able to detect these cancers early.
7007:06 So instead of waiting for cancers to be large enough to cause symptoms, or for them to be dense enough to show up on imaging, or for them to be prominent enough for you to be able to visualize on medical procedures, we can start looking for cancers while they are relatively pretty small, by looking for these small amounts of DNA in the blood.
7107:27 So let me tell you how we do this.
7207:29 First, like I said, we start off with a simple blood test -- no radiation, no complicated equipment -- a simple blood test.
7307:36 Then the blood is shipped to us, and what we do is extract the DNA out of it.
7407:40 While your body is mostly healthy cells, most of the DNA that's detected will be from healthy cells.
7507:47 However, there will be a small amount, less than one percent, that comes from the cancer cells.
7607:53 Then we use molecular biology methods to be able to enrich this DNA for areas of the genome which are known to be associated with cancer, based on the information from the cancer genomics projects.
7708:05 We're able to then put this DNA into DNA-sequencing machines and are able to digitize the DNA into A's, C's, T's and G's and have this final readout.
7808:16 Ultimately, we have information of billions of letters that output from this run.
7908:26 We then apply statistical and computational methods to be able to find the small signal that's present, indicative of the small amount of cancer DNA in the blood.
8008:37 So does this actually work in patients?
8108:39 Well, because there's no way of really predicting right now which patients will get cancer, we use the next best population: cancers in remission; specifically, lung cancer.
8208:52 The sad fact is, even with the best drugs that we have today, most lung cancers come back.
8308:57 The key, then, is to see whether we're able to detect these recurrences of cancers earlier than with standard methods.
8409:05 We just finished a major trial with Professor Charles Swanton at University College London, examining this.
8509:12 Let me walk you through an example of one patient.
8609:16 Here's an example of one patient who undergoes surgery at time point zero, and then undergoes chemotherapy.
8709:23 Then the patient is under remission.
8809:26 He is monitored using clinical exams and imaging methods.
8909:30 Around day 450, unfortunately, the cancer comes back.
9009:37 The question is: Are we able to catch this earlier?
9109:39 During this whole time, we've been collecting blood serially to be able to measure the amount of ctDNA in the blood.
9209:47 So at the initial time point, as expected, there's a high level of cancer DNA in the blood.
9309:54 However, this goes away to zero in subsequent time points and remains negligible after subsequent points.
9410:02 However, around day 340, we see the rise of cancer DNA in the blood, and eventually, it goes up higher for days 400 and 450.
9510:13 Here's the key, if you've missed it: At day 340, we see the rise in the cancer DNA in the blood.
9610:20 That means we are catching this cancer over a hundred days earlier than traditional methods.
9710:26 This is a hundred days earlier where we can give therapies, a hundred days earlier where we can do surgical interventions, or even a hundred days less for the cancer to grow or a hundred days less for resistance to occur.
9810:40 For some patients, this hundred days means the matter of life and death.
9910:45 We're really excited about this information.
10010:48 Because of this assignment, we've done additional studies now in other cancers, including breast cancer, lung cancer and ovarian cancer, and I can't wait to see how much earlier we can find these cancers.
10111:04 Ultimately, I have a dream, a dream of two vials of blood, and that, in the future, as part of all of our standard physical exams, we'll have two vials of blood drawn.
10211:15 And from these two vials of blood we will be able to compare the DNA from all known signatures of cancer, and hopefully then detect cancers months to even years earlier.
10311:27 Even with the therapies we have currently, this could mean that millions of lives could be saved.
10411:31 And if you add on to that recent advancements in immunotherapy and targeted therapies, the end of cancer is in sight.
10511:40 The next time you hear the word "cancer," I want you to add to the emotions: hope.
10611:46 Hold on.
10711:48 Cancer researchers all around the world are working feverishly to beat this disease, and tremendous progress is being made.
10811:55 This is the beginning of the end.
10911:58 We will win the war on cancer.
11012:00 And to me, this is amazing news.
11112:03 Thank you.
11212:04 (Applause)