Leor Weinberger: Can we create vaccines that mutate and spread?

Recorded atMarch 03, 2020
EventTEDMED 2020
Duration (min:sec)20:45
Video TypeTED Stage Talk
Words per minute145.34 very slow
Readability (FK)58.27 easy
SpeakerLeor Weinberger

Official TED page for this talk


Viruses mutate and spread from person to person, a dynamic process that often leaves us playing catch-up when there's a new disease outbreak. What if vaccines worked the same way? Virologist Leor Weinberger shares a scientific breakthrough: "hijacker therapy," a type of medical treatment that could attack, modify and spread alongside a virus, potentially treating afflicted individuals and slowing the spread of infections like HIV.

Text Highlight (experimental)
100:00 I want to tell you about a student of mine.
200:03 A few weeks ago, she was on vacation
300:06 when at 3am she got a phone call.
400:10 It was her husband telling her
500:11 that the city that she was in would be quarantined by 10am.
600:17 She was in Wuhan, China.
700:19 The epicenter of the coronavirus outbreak.
800:23 By 8am she was on the road in a friend's car
900:26 driving 500 miles to Shanghai Airport.
1000:30 The entire time, she was terrified the authorities would arrest her.
1100:34 She made it out on one of the last flights.
1200:38 So we're all relieved that she's back home,
1300:41 safe in the US.
1400:43 But what if I told you she was in this room?
1500:48 What if I told you she was sitting next to you?
1600:52 There's no vaccine.
1700:54 There won't be a vaccine for at least 12 months.
1800:58 And this virus is mutating,
1901:01 so there's no guarantee that the vaccine that we produce 12 months from now
2001:05 will match the mutated virus.
2101:10 We're always playing catch-up.
2201:12 And this is the scenario we're in every time there's an outbreak.
2301:17 Our quarantines are porous,
2401:19 our medical responses are flat-footed.
2501:24 The fundamental problem that we face in controlling these outbreaks
2601:28 is that viruses and other infections do two things really well:
2701:33 they mutate
2801:36 and they transmit.
2901:40 Our medical tools do neither of these two things.
3001:43 Our medical tools don't mutate or transmit.
3101:46 We have a fundamental mismatch between our tools,
3201:50 which are static,
3301:52 and the infections, which are dynamic.
3401:56 This mismatch is why we're always flat-footed.
3501:59 This mismatch is why we're playing catch-up.
3602:02 And this mismatch is universal.
3702:06 It's the reason that we have antibiotic-resistant infections
3802:09 that killed 40,000 Americans last year,
3902:13 and it's the reason that the flu vaccine couldn't prevent the flu
4002:17 from killing 60,000 Americans last year.
4102:23 So how do we resolve this fundamental mismatch?
4202:28 I've devoted my career to studying and solving this problem.
4302:34 For 100 years,
4402:36 we've used chemicals as drugs to treat infections.
4502:40 Chemicals will never mutate or transmit.
4602:44 Similarly, our vaccines are not designed to mutate or transmit.
4702:51 20 years ago,
4802:53 I had a radical idea
4902:55 to use the viruses themselves as therapies --
5003:00 as building blocks for therapies.
5103:02 To build therapies that could mutate and transmit.
5203:08 Let me share with you how we've had a major breakthrough,
5303:11 and we're already testing these new therapies.
5403:16 HIV is the most devastating pandemic of our era.
5503:22 75 million infected;
5603:24 33 million dead.
5703:27 Most of us think that HIV is a solved problem.
5803:31 We have truly amazing antiviral drugs:
5903:33 they're safe and effective,
6003:35 and we've spent 15 years
6103:36 and billions of dollars deploying these drugs
6203:39 around the world.
6303:41 So let's look at the numbers.
6403:44 In 2019,
6503:46 1.7 million people acquired HIV.
6603:51 This number only hit home for me last year
6703:53 when I visited this rural village outside of Durban, South Africa.
6803:59 I was driving along this dirt road
6904:02 with my 10-year-old daughter in the back seat,
7004:04 and walking next to us on the side of the road were girls,
7104:09 the same age as my daughter,
7204:11 barefoot.
7304:12 My daughter asked about why these girls are barefoot
7404:15 and I had to explain disparity to her,
7504:18 which was hard.
7604:21 But what really shocked me was when my colleagues,
7704:25 the local African scientists explained to me
7804:29 that these young girls the same age as my daughter --
7904:32 (Breathes)
8004:33 sorry --
8104:34 had an 80 percent chance of acquiring HIV in their lifetime.
8204:40 It blew me away.
8304:42 How can these girls have an 80 percent risk
8404:44 if we have safe and effective drugs?
8504:49 The reason is the fundamental mismatch.
8604:52 It creates barriers to controlling infectious disease,
8704:55 particularly in resource-limited settings.
8805:00 The first barrier is mutation:
8905:02 viruses mutate, our drugs don't.
9005:05 The second barrier: adherence.
9105:08 It's really hard to get these drugs to those who need the most.
9205:13 I can't even adhere to a week-long antibiotic regimen in this country.
9305:20 We're asking those in resource-limited settings
9405:23 who face enormous adversity
9505:24 to adhere to lifelong regimens.
9605:27 And the third barrier is deployment,
9705:31 or access.
9805:32 It's super hard to get these drugs to those who need them most.
9905:38 Not these girls but those who engage in high-risk activities,
10005:42 such as commercial sex work and injection drug use.
10105:48 In the epidemiological parlance,
10205:49 these groups are called "super-spreaders."
10305:52 And in the 1900s,
10405:54 a small subset of super-spreaders
10505:57 drove HIV along the Trans-Africa Highway
10606:00 and spread the virus through the continent like this.
10706:04 These groups are exceptionally hard to identify,
10806:07 they're small,
10906:08 they face enormous social stigma so they don't self-identify
11006:12 and they're the ones we need to get to the most.
11106:15 All of these barriers combined created the situation we have today,
11206:21 where greater than 15 percent of people in Southern Africa live with HIV.
11306:29 Now, conventional wisdom is:
11406:31 what we need to do is get more drugs to more people more effectively.
11506:37 I'd argue we need to solve the fundamental mismatch;
11606:40 we need to eliminate these barriers.
11706:44 If we can build therapies that mutate and transmit,
11806:48 we can overcome drug resistance,
11906:50 overcome adherence barriers,
12006:54 and if we do it properly,
12106:56 we will convert the super-spreaders from the greatest barrier
12207:00 to the most powerful deployment strategy that we can imagine.
12307:06 This is a radical concept.
12407:09 It has huge potential payoff,
12507:12 but there's a catch.
12607:15 And it's serious.
12707:19 Before we deploy a therapy that may transmit,
12807:24 even if it's only in a limited population of already infected individuals,
12907:28 before we do this,
13007:30 we need to be exceptionally careful,
13107:32 and we need to test safety in the most rigorous ways possible.
13207:37 The reason I'm here today is because for the first time 20 years,
13307:41 we got it to work,
13407:43 and this is the first time I'm sharing it publicly.
13507:47 (Applause and cheers)
13607:56 Last two times I did this I cried, so --
13707:58 (Laughter)
13808:01 So in order to help you understand this breakthrough,
13908:05 let me take you back 20 years to 1999.
14008:10 I was a graduate student in Berkeley, California,
14108:13 studying the biophysics of HIV.
14208:15 For such a complex epidemic,
14308:17 the simplicity of this virus fascinated me.
14408:22 HIV, like all viruses, is just an instruction set --
14508:26 malware.
14608:27 It hijacks a cell and converts that cell into a factory to do one thing:
14708:32 single-mindedly churn out more virus.
14808:35 All the normal functions of the cell get wiped away.
14908:38 HIV infects the white blood cell that keeps us healthy.
15008:41 This cell has already been hijacked and converted into a factory.
15108:46 And if we magnify,
15208:48 we can see the anatomy of the virus.
15308:50 Those squiggly red lines in the middle?
15408:53 That's HIV's instruction set.
15508:55 Its genetic material.
15608:58 This genetic material directs the hijacking process,
15709:02 converting that cell into a factory first
15809:05 to make more copies of the instruction set,
15909:09 and then all the other components of the virus
16009:11 and assemble them into a particle.
16109:15 This is how the virus replicates.
16209:19 Each of these particles can go in and hijack a new cell.
16309:24 This is how the virus transmits.
16409:26 And every time a cell is hijacked,
16509:29 little mistakes are made in copying the genetic material.
16609:33 That's how the virus mutates.
16709:36 This ability to replicate, transmit and mutate
16809:39 is something that our current drugs cannot do.
16909:43 So, being young and naïve and a little bit ignorant,
17009:47 I thought:
17109:49 why can't we create therapies that mutate and transmit and replicate?
17209:57 Here was the idea.
17309:59 If we can take the virus
17410:02 and engineer it to amputate the genetic material in blue here,
17510:06 this amputated instruction set can no longer hijack a cell.
17610:11 But this amputated instruction set can do something very special.
17710:16 In an already infected cell,
17810:18 this amputated instruction set can hijack the hijacker.
17910:23 It can commandeer HIV's machinery to make more copies of itself,
18010:29 the amputated instruction set,
18110:31 and then each of these copies can steal all of HIV's other components
18210:36 so that the cell gets converted from a factory that produces virus
18310:41 to a factory that produces therapy.
18410:44 Hijackers.
18510:46 These carry no disease.
18610:48 This dramatically lowers HIV levels
18710:51 and keeps the cell healthier.
18810:55 This idea consumed me for months.
18910:58 It was the most intense intellectual experience of my life.
19011:03 On every bike ride to lab,
19111:06 on every walk to the coffee shop,
19211:09 on every run in the hills above campus,
19311:11 the ideas, the arguments,
19411:13 the counterarguments,
19511:15 they all came so rapidly in my head, in my inner monologue,
19611:18 that I was physically out of breath.
19711:20 I thought if we can create a therapy that replicates,
19811:24 it would only need to be taken once.
19911:27 It could mutate along with the virus
20011:30 and possibly it could transmit between infected individuals
20111:34 to treat them.
20211:36 It was a therapy that could do all of the things that the virus could do.
20311:41 It solved the fundamental mismatch.
20411:46 The most radical part of this concept
20511:49 was that the super-spreaders would also be converted
20611:52 from transmitting the virus to now transmitting the therapy.
20711:58 It was a therapy that would go viral along with the virus.
20812:04 This scared some people.
20912:06 But there's already a precedent:
21012:09 we already inadvertently use therapies that transmit.
21112:14 The vaccine that eradicated polio in the US,
21212:17 the oral polio vaccine,
21312:20 transmits between people.
21412:23 It's not well-known,
21512:24 but the fact that this vaccine transmits
21612:27 is part of the reason that it was chosen for the worldwide eradication effort
21712:32 despite its safety issues.
21812:36 So the bigger problem was that these hijacker therapies didn't exist.
21912:42 My Berkeley advisors said to me,
22012:45 "Lovely idea, so sad it will never work,"
22112:49 or, "Regulators will never allow it,
22212:51 drop it."
22312:54 But the idea wouldn't drop me.
22412:59 If it ever worked,
22513:01 it would solve the fundamental mismatch.
22613:04 So we tried for years to build it.
22713:07 We tried every trick in the book
22813:10 and failed over and over again.
22913:13 We --
23013:15 every time we thought we had a good idea,
23113:17 we'd spend months, sometimes years working on it
23213:21 only to come up empty.
23313:24 We once spent five years
23413:28 building 150,000 versions of a hijacker therapy.
23513:33 Every single one failed.
23613:37 I once asked a really bright student what he hoped to learn from me
23713:40 during his PhD --
23813:42 (Laughter)
23913:44 And he replied,
24013:46 "How to keep going,
24113:48 how to continue moving forward
24213:50 despite zero evidence that there's anything there."
24313:53 (Laughter)
24413:56 I wonder if he was trying to tell me something.
24513:58 (Laughter)
24614:01 By 2018, things looked bad.
24714:06 There was no evidence
24814:08 that a hijacker therapy could be engineered.
24914:10 In fact, we had evidence that it might be impossible.
25014:15 It was time to face the hard truth.
25114:18 This solution that I'd wanted so badly,
25214:20 this hijacker therapy just didn't exist.
25314:26 For 20 years,
25414:27 I had been chasing a ghost.
25514:32 Then one day,
25614:35 Elizabeth, a postdoc in my lab,
25714:37 came to me with this picture.
25814:40 It doesn't look like much.
25914:42 My wife thinks it looks like a pregnancy test.
26014:45 (Laughter)
26114:47 But this little band down here --
26214:51 that was the amputated genetic material
26314:54 that we had been looking for for 20 years.
26414:56 The entire time that we had been trying to build it and failing,
26515:01 it had evolved by itself in a flask in the back of the lab.
26615:04 (Audience) What?
26715:05 (Laughter)
26815:07 We finally had a foothold.
26915:08 And we used this to build the first hijacker.
27015:13 But we had no evidence that what we built was a therapy.
27115:18 The first hurdle that any therapy has to clear is testing in a mouse.
27215:22 It can be risky.
27315:24 In our case, if our mice died,
27415:27 so did our funding,
27515:29 and with it, any hopes of this becoming a therapy,
27615:32 let alone a transformative one.
27715:36 After so many failures,
27815:37 we were all pretty skeptical but we didn't really have an alternative.
27915:42 We had to give it a shot; we had to try.
28015:46 Amazingly, the hijacker therapy worked in a mouse,
28115:52 and it worked exactly as we'd predicted 20 years before.
28215:56 It protected the cells in a mouse from HIV.
28315:59 Here are the cells under a microscope.
28416:01 First, HIV in red infects those cells,
28516:05 and then the hijacker, in blue, can be activated,
28616:09 protects those cells and transmits to other cells.
28716:14 We'd finally built the hijacker after 20 years.
28816:18 Everyone in the lab was elated.
28916:21 For me, this was proof of concept.
29016:25 If we could do it for one virus we could do it for others.
29116:30 To understand how this hijacker might impact HIV levels worldwide,
29216:35 we ran computer simulations.
29316:37 Epidemiological models.
29416:39 And the results were pretty amazing.
29516:43 If we do nothing in the hardest-hit parts of Africa,
29616:46 HIV prevalence will stay between 25 and 30 percent.
29716:52 If we manage to introduce drugs to three-quarters of the population
29816:56 or if we ever get the long-sought-after vaccine,
29917:00 we would reduce those numbers down to 20 percent.
30017:04 But those are best-case scenarios.
30117:07 If HIV evolves resistance
30217:10 or if people change their behaviors,
30317:12 these numbers could go right back up to 30 percent
30417:15 or even beyond.
30517:17 The blue is the hijacker therapy.
30617:22 And we've not found a way,
30717:24 either theoretically or experimentally,
30817:27 that HIV can evolve resistance to the hijacker.
30917:32 The reason this hijacker works so well is the super-spreaders.
31017:37 If the hijacker is introduced in one place over here,
31117:42 the super-spreaders can pick it up and transmit it through the population.
31217:51 Imagine if 10 years from now, HIV is no longer a pandemic.
31317:57 To get there,
31417:58 we have to start large-scale clinical trials in five years,
31518:02 which means initial human tests next year.
31618:06 The FDA has cleared us to start testing in HIV-positive patients
31718:11 who have a terminal cancer
31818:14 and have less than a year left to live.
31918:20 Volunteering for this trial
32018:21 is their last incredibly generous gift to the world.
32118:25 They're called the "Last Gift cohort."
32218:27 And to test in these altruistic patients next year,
32318:32 we have to finish our preclinical tests this year ...
32418:36 and I think we will.
32518:38 I still meet colleagues who push back
32618:40 and are stridently opposed to letting us move forward with testing.
32718:44 They say, "What if something goes wrong?
32818:47 You can't unrelease it."
32918:50 They say, "There are ethical issues;
33018:52 people can't consent."
33118:55 Well, oral polio vaccine faced similar ethical and safety concerns.
33218:59 In fact, oral polio vaccine faced such an effective misinformation campaign
33319:04 that most people still don't know that it transmits,
33419:07 that it successfully eradicated polio in many countries
33519:10 or that this vaccine is the basis for new vaccines.
33619:15 When we presented this hijacker therapy idea in Africa last year,
33719:20 the African scientists had a different response.
33819:25 They said, "How can you not test this?"
33919:28 They said it's unethical to not test it.
34019:32 So even though we might fail,
34119:35 I think the stakes are too high not to try.
34219:38 If we do nothing,
34319:40 those girls outside of Durban will acquire HIV,
34419:43 and the next time that there's a new virus that emerges,
34519:48 we'll be in the same vulnerable position that we are today
34619:53 with quarantines that are porous
34719:57 and vaccines that we need to wait months for that may not match.
34820:04 I think it's time for a new approach
34920:06 that's different than the static approaches of the last century.
35020:11 It's time for a new technology that is less reactive
35120:14 and more proactive.
35220:16 I think it's time for treatments
35320:18 that don't just benefit the most affluent among us
35420:22 but also those who face the greatest adversity.
35520:30 I think it's time for a new type of weapon
35620:33 that matches,
35720:35 or that solves the fundamental mismatch.
35820:38 I think it's time for therapies that can go viral.
35920:40 Thank you very much.
36020:42 (Applause)