Peter Attia MD346 - Scaling biotech and improving global health: lessons from an extraordinary career in medicine
CHAPTERS
- 0:00 – 7:50
Residency at UCSF and the Early AIDS Epidemic
Susan Desmond-Hellmann describes growing up in Reno, her path to UCSF residency in 1982, and what the first AIDS patients looked like. She recalls Pneumocystis pneumonia, Kaposi sarcoma, and working gowned and gloved like in an operating room.
- •Catholic school and University of Nevada through medical school, then UCSF internal medicine
- •In 1982 AIDS was linked to the 'three H's' and residents took part in a study drawing their own blood
- •Pneumocystis pneumonia and Kaposi sarcoma were the hallmark presentations
- •Her memories center on sadness and her patients, not fear of contagion
- 7:50 – 12:55
Choosing Oncology and the State of the Field in the 1980s
She explains why she chose oncology, citing an inspiring VA attending, and Peter Attia shares a similar story about Steve Rosenberg. She describes a stifled field with decades-old breast cancer chemotherapy and her turn to epidemiology and a public health master's.
- •Oncology combines the compassion gene and the nerdy science gene
- •Attia relays Rosenberg's view that cancer either draws families together or blows them apart
- •Breast cancer therapy relied on decades-old drugs like Cytoxan, methotrexate and 5-FU
- •Epidemiology and biostatistics training later carried over into drug development
- 12:55 – 29:55
Uganda: Frontline Care in the HIV/AIDS Crisis
Funded by a Rockefeller Foundation grant to study heterosexual HIV transmission, she and her husband moved to Uganda in 1989. She describes the scale of the epidemic, triaging scarce chemotherapy, and how they coped.
- •Only the two of them went, and Uganda's infrastructure had been damaged under Idi Amin
- •About a third of patients seeking care had Kaposi sarcoma; half of 16-year-old girls at the STD clinic were HIV positive
- •Untreated STDs such as herpes and chancroid drove transmission; the government ran its 'Zero Grazing' campaign
- •She triaged patients, treating only the sickest with vincristine and bleomycin
- •She coped by leaning in and refusing to be a martyr, with trips and shared stories at home
- 29:55 – 41:21
From Kentucky Practice to Taxol at Bristol Myers Squibb
Back in the US, she joined a private oncology practice in Kentucky, then moved to Connecticut when her husband was recruited by Bristol Myers Squibb. She started in drug safety and went on to lead the Taxol project team.
- •She and her husband lacked academic standing at UCSF, so they went into private practice in Lexington
- •Bristol Myers Squibb hired her as a consultant after her husband refused to come without a job for her
- •Teaching herself SAS in Uganda helped her earn credibility with statisticians
- •Taxol is a microtubule poison that blocks cell division; she explains its neutropenia and allergy issues
- •Taxol was approved for breast cancer and became Bristol Myers Squibb's top drug
- 41:21 – 52:33
Genentech's Origins and Recombinant DNA Medicines
Desmond-Hellmann recounts Genentech's founding in 1976 and how it used genetic engineering to make protein drugs like growth hormone. She also covers its early products Activase and Pulmozyme and the company's struggles when she arrived in 1995.
- •Genentech was founded on proving genetic technology could make large protein medicines
- •Cells in tanks are taught to secrete high amounts of the target protein, which is purified into vials
- •Growth hormone replaced cadaver-derived hormone, backed by a registry following every treated child
- •Activase (tPA) was a large simple-trial bet in cardiovascular medicine, later eroded by stents
- •Genentech was struggling in 1995, and she left for the oncology opportunity under Art Levinson
- 52:33 – 1:00:56
Thrombopoietin and the Herceptin Breakthrough
Her first Genentech project, thrombopoietin, proved very difficult. She then describes Herceptin, an antibody targeting HER2, how internal skepticism about antibodies was overcome, and the contrast with the fraudulent bone marrow transplant data at ASCO.
- •Thrombopoietin was meant to complete a stool with EPO and Neupogen but its kinetics made it a poor drug
- •Herceptin targets HER2, overexpressed in about one in four breast cancers, which carries a worse prognosis
- •Many at Genentech thought antibodies were hype that couldn't treat solid tumors
- •At the same ASCO, Herceptin's promise appeared as bone marrow transplant for breast cancer was debunked
- 1:00:56 – 1:06:39
Rituxan and How Antibodies Differ from Chemotherapy
She explains how Genentech's deal with the cash-strapped IDEC brought in Rituxan, which changed her view of antibodies. The discussion covers chimeric antibodies, durable remission, retreatment, and early pricing.
- •IDEC's CD20 antibody was close to running out of money before the Genentech deal
- •A case of tumor lysis syndrome from a naked antibody convinced her of its power
- •Concerns about human anti-chimeric antibodies did not materialize clinically
- •Unlike chemotherapy, antibodies like Herceptin and Rituxan can be given again after relapse
- •Rituxan's sales rose very fast, and these drugs carried early big price tags
- 1:06:39 – 1:12:10
Avastin and the Anti-VEGF Hypothesis
Judah Folkman's idea that tumors need new blood vessels led Genentech to build an antibody against VEGF. She describes first-in-human fears, choosing cancers to test, and the breast cancer failure followed by colon cancer success.
- •Folkman's compelling hypothesis was that tumors can't outgrow a BB without new blood vessels
- •Napoleone Ferrara made the anti-VEGF antibody on the Herceptin backbone
- •Gwen Fife feared blood vessels might fall apart in the first Phase I patient
- •Renal cell carcinoma was biologically logical but hard to study
- •The breast cancer study failed, but a colon cancer phase III added roughly eight months of survival
- 1:12:10 – 1:24:40
How Drug Development Works: Preclinical to Phase III
Prompted by Peter Attia, she walks through the drug development process. Topics include small molecules versus antibodies, preclinical safety planning, companion diagnostics, the IND, and what each clinical phase is for.
- •Small molecules are chemistry and tend to have off-target surprises, while antibodies are biology and more precise
- •Preclinical work includes mini-fermenter production, animal models and a toxicology and safety plan
- •Genentech should have developed the HER2 diagnostic in parallel; FDA required an approved test before approving Herceptin
- •The IND is the request to ship an unapproved drug across state lines
- •Phase I tests safety, Phase II sets dose, schedule and patient selection, and Phase III confirms benefit
- 1:24:40 – 1:27:23
The Cost Controversy Around Avastin
Attia recalls the debate in medical school over Avastin's high price and limited survival gain, including the UK's refusal to fund it. Desmond-Hellmann says a Wall Street Journal headline about affordability shocked her and led to internal discussion about pricing.
- •The NHS declined to pay because of its quality-adjusted life year threshold
- •Avastin cost about $100,000 for roughly eight months of added median survival
- •Genentech's philosophy was that no patient should go without a drug because of inability to pay
- •Affordability concerns prompted more executive committee discussion about pricing
- 1:27:23 – 1:41:25
Chancellor of UCSF: Finances and Fundraising
After Roche's 2009 takeover of Genentech, she became UCSF chancellor. She describes the university's revenue streams, her lack of fundraising experience, and how philanthropy and grateful patients helped during a recession.
- •Genentech was bought by Roche in 2009, which she says was not what the company wanted
- •She pitched the search committee on business management, using assets well and finishing Mission Bay
- •UCSF revenue comes from NIH overhead, clinical income, grants and philanthropy, with almost nothing from the state
- •Fundraising advice from John Ford: share your hopes and dreams and see if there's a match
- •She spent about 40% of her time externally
- 1:41:25 – 1:51:27
CEO of the Gates Foundation: Strategy and Portfolio Management
She describes how Melinda and Bill Gates recruited her to lead the foundation. She explains the organization and its $8 billion budget, the portfolio management process she introduced, and the difficulty of managing the people side.
- •Both Bill and Melinda wanted her, and she saw merit in both their viewpoints
- •The foundation has a couple thousand staff and an annual budget of eight billion dollars
- •She brought a Genentech-style portfolio management process, despite Bill's view that it was all in his head
- •Her hardest challenge was holding managers to people standards when Bill valued technical experts
- 1:51:27 – 1:58:16
Leadership Philosophy and Building Culture
She shares what she learned about people and work, how she tried to reach beyond her direct reports, and her definition of culture. Examples from Genentech and the Gates Foundation illustrate it.
- •Lesson that work never makes up for a bad home life
- •She changed grant review so she could interact with people beyond her reports
- •She defines culture as an atmosphere that brings out the best in people, with ownership to tweak it
- •At Genentech, she and Art Levinson stopped a suggestion to profit from a looser HER2 test
- •At Gates, she coached presenters facing Bill's tough questioning
- 1:58:16 – 2:14:30
COVID, Lost Trust in Science, and Ivermectin
Attia and Desmond-Hellmann discuss the erosion of trust after COVID and the problem of combining scientific and advocacy roles. They also address ivermectin claims about cancer and why shaming people backfires.
- •Both say they are horrified at the current narrative and worry another pandemic would be handled worse
- •Attia argues one person can't be both science chief and public advocate
- •Admitting when science changes, as with bone marrow transplants, matters for trust
- •Desmond-Hellmann says silence cedes space, and facts should be offered rather than removing people from the dialogue
- •They explain why one drug can't cure all cancers, and Attia criticizes the 'horse dewormer' response
- •She co-chaired a PCAST report urging a broader public health workforce
- 2:14:30 – 2:28:17
AI in Medicine and Cancer Detection
Desmond-Hellmann, who serves on the OpenAI board, discusses how AI could speed clinical trials, reduce clinician burden and advance preclinical science. They then turn to biomarkers, liquid biopsies and the limits of early cancer detection.
- •Her hope is to cut the roughly six-to-seven-year IND-to-approval timeline by two years using AI on paperwork-heavy steps
- •AI could enable ongoing safety follow-up in every patient on a drug
- •The protein-folding Nobel is the most important AI promise for medicine so far, in her view
- •She wants better biomarkers like viral load, which enabled rapid HIV drug progress
- •She is pessimistic about liquid biopsies, noting tumors may not shed enough DNA, and would bet on protein for breast cancer
- •Attia adds that PSA, used with density and velocity, can be predictive
