Quantcast
Channel: Endpoints News
Viewing all articles
Browse latest Browse all 2609

The Endpoints Slack interview: Siddhartha Mukherjee on the doctor-writer worldview, AI, and the future of cancer

$
0
0

Siddhartha Mukherjee is something of a biotech Renaissance man.

First and foremost, he sees himself as an oncologist. But that core interest has led him in many directions, including running a research lab at Columbia University, co-founding a growing list of biotech startups, and writing books on science and medicine, including the Pulitzer Prize-winning “The Emperor of All Maladies,” a history of cancer.

Mukherjee joined the Endpoints News Slack channel to talk about his latest work across academia, medicine, writing and business. The chat took place on May 10 and has been minimally edited.

Andrew Dunn
good morning Sid! thanks for joining our Slack room

where do you find yourself chatting from today?

Siddhartha Mukherjee
I am in New York City
Andrew Dunn
I am as well — although you have me beat on residency length by just a bit having moved here a couple months ago.
Siddhartha Mukherjee
Well, I’ve been here for more than 10 years
Andrew Dunn
for starters, what do you make of AlphaFold 3? how big of a deal is that for your own research, and for cancer research more broadly?
Siddhartha Mukherjee
I’m pretty impressed. The field seems to be moving rapidly. I have not tried the new version, which predicts interactions as well, but I am hoping this will really open up new arenas.

One area that this could really open up is determining protein-protein interactions. It’s very important for cancer biology to understand the topology of these interactions. Take antibody and antigen interactions. If we could predict these and model these virtually, it would be a huge benefit to the field

Andrew Dunn
on the broader topic of AI, where do you find yourself on the enthusiasm and skepticism scale … both for AI overall, but also in applying AI to the world of biology and medicine

we are clearly in a moment full of hype, but also real progress

Siddhartha Mukherjee
Complicated question, but I hope to shed some light. AI interfaces with medicine in multiple ways, so we need to think through these systematically. First, there’s AI in diagnostics. That’s been around for a while. Think of ways AI could help interpret a mammogram or a path slide. Clearly a powerful use case, and important.

Then, there’s patient-facing AI. I am thinking of co-pilots and chatbots and non-synchronous chronic care management. That’s also important, but there’s a lot of hype around that, and I think the space is way too crowded.

Then, there’s AI in the so-called “back office”. Management of revenue cycles, improvements in billing and so forth. That’s a powerful use case, and something that I am working on.

Finally, there’s AI in discovery, such as drug discovery. You’ll see a couple of announcements from my lab and group in this area soon, but it’s the one that I am the most excited about. This is the true use of GENERATIVE AI in health care — and it’s the area that is the least understood and, I suspect, the most valuable.

Andrew Dunn
it seems to me there’s also a way to split uses between productivity/efficiency and creativity

i.e. the back office use (efficient!) versus, say, de novo antibody generation (creative!)

Siddhartha Mukherjee
Yes, I think people are using the word AI too loosely in medicine. We need to be accurate about use cases, and also determine the difference between productivity and creativity.
The use cases that I described pretty much follow the productivity-versus-creativity aspect.  Of course, there’s a mix: you can be really creative in generating asynchronous care models. But the real creative aspects of AI lie in making new kinds of matter, or learning to design or read clinical trials more effectively.

And speaking of trials, I thought we should also catch up on our metabolic program with Faeth.

Andrew Dunn
We can jump to Faeth, as I did want to ask you about your biotech work broadly too. Can you catch us up with very quick updates on the biotechs you’re involved with today?
Siddhartha Mukherjee
Gosh. Yes, of course. There’s Vor biopharma, which is directed at Acute Leukemia, marrow transplant and related diseases. I’m not sure Vor has really advertised this, but we were among the first to transplant human bone marrow that had been gene edited. We are in mid trial now, so it’s a very exciting time.
Then, Immuneel. We’ve just obtained the commercial license to manufacture CAR-T in India, and are about to dose our first patient. That’s been an incredible success story.
Myeloid therapeutics is in mid-trial too, with a novel myeloid based cell therapy. That’s going through dose escalation.

And, aside from Faeth, you’ll soon hear about two companies in stealth mode that are really bananas. Perhaps you’ll have to have me back to discuss those. We’ll start disclosing those in about a month or two.

Andrew Dunn
consider my interest piqued. i’ll bother you real soon on that lead.

and for Faeth, I just caught up with the CEO, but for our audience what’s the latest on their progress and the idea of metabolics in cancer?

Siddhartha Mukherjee
Well, it’s my job to keep you interested. But back to Faeth. Once again, we just completed a pivotal study using our “PIKTOR” program, which is a combination of inhibition of PI3Kinase and mTOR, along with chemo, in patients that had been heavily pretreated with chemo alone. We saw a strong, pretty unprecedented signal in endometrial cancer patients, and that is what is being followed up.
Faeth isn’t a “diet” company, but our mission is to tackle cancer metabolism as a whole. And PIKTOR, which attacks two crucial nodes of cancer metabolism is the best example. I don’t think anyone has seen this sort of a response in endometrial cancer — multiple CRs in heavily pre-treated patients. So it’s a nail-biting time for us.

I should say Faeth isn’t JUST a “diet” company. It’s a company that focuses on all aspects of metabolism. Sometimes by changing the metabolic state of a patient (eg. by depletion of a nutrient, or changing metabolic signaling). But the PIKTOR program uses the same principles to attack the cancer cell’s internal metabolic program.

Andrew Dunn
i wanted to ask on the diet element because that was a really fascinating provocation from the start of Faeth, the notion that cancer patients could benefit by having a specific diet in addition to cancer treatments
Siddhartha Mukherjee
I agree that diet has always been a relatively neglected part of cancer therapeutics. You go to your oncologist and ask what you should eat, and he or she might shrug their shoulders and say “eat what makes you feel good”. So Faeth BEGAN with diets in combination with targeted therapy. But then the thinking evolved. We began to wonder whether the internal nodes of cancer metabolism — eg MTOR or PI3K signaling — might be druggable even without changing diet, and that’s what the endometrial cancer data shows. Incredible responses by metabolic modulation. We haven’t lost sight of diet, and will add it to the programs as they advance, but right now, just metabolic modulation with the PIKTOR program seems to show some potent responses
Andrew Dunn
a decade from now, do you think nutrition will be a standard element of cancer care? or is that too optimistic, too short of a timeline for this to play out in your mind?

by standard, i really mean what you outlined above — a patient going to their typical cancer doctor

Siddhartha Mukherjee
I would hope that we get a handle on cancer nutrition. Again, that’s where AI could potentially help. Deconvolute the complex nutrition that patients have and try to get a handle on what’s best. But I suspect that observational studies, even with large data sets, aren’t going to be enough. We will need interventional, or experimental, studies.
As far as timeline goes, observational studies will take a lot of time. I think we need experimental data, and we are ready to really dive into that now.

But I guess that one lesson of Faeth is that you can go from diet to metabolic modulation. In some senses, you are attacking the same axis. But this time, with drugs rather than diet or diet alone. Diet turned out to be a portal for Faeth to enter the cancer metabolism space, and that has been very very neglected

Andrew Dunn
gotcha, tons of interesting work to watch across all those companies. i’d be remiss if i didn’t switch gears to ask about another part of your career in your writing
Siddhartha Mukherjee
Of course
Andrew Dunn
did you always know you wanted to write? and not just articles, but big, Pulitzer-winning books?
Siddhartha Mukherjee
Yes, I did, although I never imagined myself as a “writer”. I sort of wandered into it, because the questions were so enticing. I never saw myself writing 600-page books — and now three of them.
And once again, the writing turns out to be a portal into a world of ideas. I wrote an article about AI for the New Yorker long before AI was fashionable. But that got me interested in AI in general, and so here I am, about five years later, digging into generative AI for drug discovery. For me, everything works like a discovery portal. You see a window and you go through it, just to look at what’s on the other side

Or take CAR T. Again, I started with writing about CAR T for a piece for the New Yorker. And then it became a portal: why can’t we make CAR T in India ? And that’s how Immuneel got started

Andrew Dunn
has writing for the public made you a better researcher or doctor? if so, how so?
Siddhartha Mukherjee
And yes, writing has made me both a better researcher and doctor. When you write, you REALLY need to think about a subject. It’s like doing a fresh sabbatical every time, or learning about a patient’s intimate experience. And that makes me a better doctor and researcher I hope
Andrew Dunn
that’s really fascinating. it sounds like that writing and researching process is like an antidote to maintain a curiosity in what’s happening, where you don’t get siloed in to a narrow speciality like oncology (even as huge as cancer research is in itself)
Siddhartha Mukherjee
Your use of the word “antidote” is right. There are a lot of cancer researchers that I deeply respect that work on one and only one aspect of cancer biology. My practice is fundamentally different. Like I said, every time I write something, I take a sabbatical and learn a new world. And I see where I can make a difference in that world. CAR-T is an example. Talking to a patient and discovering that I had no recommendation about her diet is another.
Andrew Dunn
we have just a few minutes, so i wanted to touch on another world you inhabit in academia
you’re a Columbia professor and you’re set to speak at University of Pennsylvania’s graduation in just a few days.

To put it lightly, this is a turbulent time on college campuses. What’s your advice to students graduating and entering the professional world out of this current college culture?

Siddhartha Mukherjee
It’s really turbulent, and I understand the huge implications that are swirling around. I don’t see myself as an advice giver, so I try to refrain from preaching. I’m a cancer physician and I will try to talk to students about empathy and forgiveness. It’s a message that I think will resonate beyond just being a doctor.
Andrew Dunn
Okay, and to close out with a rapid round if that sounds good quick, short answers!
Siddhartha Mukherjee
OK
Andrew Dunn
Would you rather run a biotech startup or a government health agency?
Siddhartha Mukherjee
Biotech startup
Andrew Dunn
The first time you made life-changing money was …
Siddhartha Mukherjee
When I started as a young professor
Andrew Dunn
What’s the best book you’ve read recently?
Siddhartha Mukherjee
Just finished “Age of Reason” by Holmes
Andrew Dunn
FINAL Q — and feel free to take a beat on this one —

How do you personally define success for your career and life?

Siddhartha Mukherjee
Curing, treating and preventing cancer

Didn’t miss that beat

Andrew Dunn
Beautiful. Thanks so much for taking the time Sid, it’s always a pleasure

That’s a true rapid round!

Siddhartha Mukherjee
My pleasure
Andrew Dunn
Oh, who should we have on next?
Siddhartha Mukherjee
Gosh. Try Reid Hoffman or Kevin Scott on AI
Andrew Dunn
Reid would be fascinating — well, thanks again Sid! This was a ton of fun.
Siddhartha Mukherjee
Thanks. Signing off

Viewing all articles
Browse latest Browse all 2609

Trending Articles