By playing, you agree to uListen's Terms and Privacy Policy and YouTube's Terms of Service.
Peter Attia MD267 ‒ The latest in cancer therapeutics, diagnostics, and early detection | Keith Flaherty, M.D.
Summaries, chapters, transcripts, and FAQs on this page are made by uListen, not YouTube.
CHAPTERS
- 0:00 – 4:56
Keith Flaherty's background in translational oncology
Peter Attia welcomes Keith Flaherty back four years after their first conversation and asks about his background. Flaherty describes 23 years as a medical oncologist, from the first genetic-insight therapy (Gleevec) through melanoma and kidney cancer work to co-founding biotech companies.
- •Flaherty has been a medical oncologist for 23 years, starting when Gleevec first went into patients
- •He began with melanoma and kidney cancer because molecular insights were ripe for translation
- •He built a clinical program and a bedside-to-bench translational group at Mass General
- •He co-founded biotech companies including Loxo Oncology and Scorpion Therapeutics
- •His focus is translating molecular understanding into therapies and diagnostics
- 4:56 – 10:33
How cancer mortality shifts across decades of life
Attia presents data on the share of deaths from cancer by age group and cancer's rank among causes of death. Flaherty adds which cancer types lead in younger versus older people and the role of carcinogens like UV light and smoking.
- •Cancer deaths peak in the middle decades, around 30–31% of deaths at ages 55–74
- •Cancer always ranks first to third as a cause of death at every age bracket
- •Brain tumors, leukemia and melanoma lead in younger populations
- •Smoking-related cancers rise in later decades and include head and neck and bladder cancers
- •Breast and prostate cancers are hormone-linked and age-related
- 10:33 – 18:42
Hormone receptors, growth factor signaling, and the haves and have-nots of cancer
Attia notes that five cancers account for just over half of cancer deaths and that hormone levels are low when breast and prostate cancers appear. Flaherty explains how cancer cells become independent of growth factors by mutating receptors and downstream signaling, and contrasts treatable hormone-driven cancers with pancreatic cancer.
- •Lung, breast, prostate, colorectal and pancreatic cancers account for over half of cancer deaths
- •Cancer cells become autonomous of circulating hormones while still using the receptors
- •Lower testosterone at prostate cancer diagnosis is linked to worse outcomes, and ER/PR-negative breast cancers are worse
- •Hormone receptor–positive breast cancer has gained new approved drug combinations
- •Pancreatic cancer remains a have-not with only pre-2000 classical chemotherapy
- 18:42 – 23:41
Obesity as a carcinogen: insulin signaling and inflammation
Attia raises obesity as the second most prevalent environmental trigger of cancer after smoking. Flaherty links it to insulin and IGF signaling feeding the same growth factor pathways, along with chronic inflammation.
- •Obesity is tied to roughly 25–27 cancers according to Attia
- •Attia suspects inflammation and growth factors like insulin and IGF-1 rather than adiposity itself
- •Flaherty calls obesity the third highest ranking carcinogen
- •Insulin signaling connects to the PI3 kinase pathway, a known cancer driver
- •Chronic inflammation can directly cause cancers at a tissue site
- 23:41 – 35:19
Progress in metastatic cancer survival and the dominance of PD-1
Attia recalls Steve Rosenberg's statement about no progress in metastatic epithelial cancers and asks where things stand now. Flaherty estimates 15–20% ten-year survival, explains what 'metastatic' means, and says half of the gain comes from PD-1 immunotherapy with the rest from targeted drugs.
- •Ten-year survival for overt metastatic disease is around 15%, possibly heading toward 20%
- •Microscopic metastatic disease explains why 30–50% of surgically treated patients later relapse
- •PD-1 antibody therapy accounts for half of the gain
- •52 FDA approvals reflect only 19 unique mechanisms, about 10 truly unique targets
- •CTLA-4 adds benefit for only a small percentage of patients
- 35:19 – 48:43
How the immune system sees cancer: antigens, MHC, and checkpoints
Attia asks for a refresher on immunotherapy, and Flaherty explains how cells present protein fragments to the immune system and how cancer evades it. Using Attia's house-and-curb analogy, he covers immune surveillance, the PD-1/PD-L1 brake, and suppressive cells in the tumor microenvironment.
- •Cancer cells carry differences such as developmental markers, lineage antigens, and mutated proteins presented on MHC
- •Natural killer cells attack cells that stop presenting antigens
- •Being diagnosed with cancer means immune surveillance failed
- •Cancers express PD-L1 to press the brake on T cells, and PD-1 drugs block it
- •Cancers also recruit suppressive immune cells to build a protective microenvironment
- 48:43 – 59:32
Neoantigens in 80% of epithelial cancers and why immunotherapy works
Attia asks why the finding that about 80% of epithelial tumors have neoantigens is so significant. Flaherty explains how mutation load explains PD-1 responses in UV- and smoking-related cancers, and Attia ties this to spontaneous remissions and interleukin-2.
- •Cancers accumulate dozens to thousands of mutations, only some presented as neoantigens
- •Highly mutated UV- and smoking-related cancers respond best to PD-1 drugs
- •T cells can recognize a single amino acid change in a presented peptide
- •Cancers that survive tend to carry poorly immunogenic mutations
- •Attia links spontaneous remissions and IL-2 responses to the same neoantigen logic
- 59:32 – 1:10:15
Engineered T cell therapy and the specificity problem
Flaherty traces cell therapy from Rosenberg's expansion of tumor-infiltrating lymphocytes to engineered cells and explains that current approaches target lineage markers like CD19. Attia and Flaherty discuss why this doesn't extend to solid organs without cancer-specific targets, and Flaherty describes AND-gate designs and cost hurdles.
- •Rosenberg's approach expanded immune cells from a patient's tumor and cured a significant minority
- •TIL therapy alone is on the verge of FDA approval for melanoma
- •Current engineered cell therapy targets B-cell markers like CD19, accepting loss of normal B cells
- •Attia shares a friend's complete response to anti-PD-1 that also destroyed all his pancreatic cells and left him needing insulin
- •Finding truly cancer-specific surface targets is the field's main struggle, and AND-switch designs and cost are open issues
- 1:10:15 – 1:16:46
Why TIL therapy fails and the case for multimodal treatment
Attia asks why TIL therapy doesn't work every time. Flaherty describes layered defenses including altered interferon pathways, suppressive myeloid cells, and a harsh metabolic microenvironment, and argues that curing cancer will need a combination of therapies from different categories.
- •Some cancers alter intracellular interferon pathways so immune cells cannot kill them
- •Suppressive myeloid cells and trafficking barriers block cell therapy
- •Low oxygen and nutrient competition make the microenvironment hostile to immune cells
- •Fibroblasts in pancreatic cancer appear to contribute to the force field
- •Flaherty argues for multimodal regimens drawing from different therapeutic categories, not one cocktail for all
- 1:16:46 – 1:31:50
The next five years: metabolism, epigenetics, and inducing mutations
Attia asks what it would take to double durable response rates within five years. Flaherty says efforts to add more T-cell gas pedals and brakes have largely failed, then describes metabolism- and epigenetics-targeted drugs, his group's work on mitochondrial biogenesis regulators, and the idea of pushing cancer cells with more mutations to boost immune recognition.
- •Drugging additional T-cell regulators on top of PD-1 has systematically failed without patient selection
- •Flaherty's group identified a mitochondrial biogenesis regulator that distinguishes cancers, with a preclinical proto-drug
- •Oxidative phosphorylation switching is a common theme in resistance to targeted therapy and immunotherapy
- •Adding mutations may boost immune recognition or push cancer cells toward cell death
- •A colleague's mouse data suggest alkylating chemotherapy over a targeted therapy backbone could enhance PD-1 therapy, and human trials are planned
- 1:31:50 – 1:38:36
Why microscopic disease is curable and macroscopic disease is not
Attia contrasts five-year survival in stage 3 and stage 4 colon cancer treated with the same chemotherapy and asks why. Flaherty offers two hypotheses, clonal heterogeneity and immune mop-up, and Attia adds that small tumor burden limits a hostile environment, setting up the case for early detection.
- •Stage 3 colon cancer patients have around 60–70% five-year survival, while stage 4 patients with liver metastases have almost none
- •Larger tumors have subclones, and a resistance mutation may already exist in a small subclone
- •Curative therapies likely rely on the immune system to eradicate remaining cells after drugs kill most of them
- •Targeted drugs can cure a substantial fraction in the adjuvant setting but a trivial fraction with overt metastases
- •This underlies enthusiasm that early detection will make existing drugs far more effective
- 1:38:36 – 1:49:22
Current screening limits and the promise of circulating tumor DNA
Flaherty reviews how modest current screening is for cervical, breast, colon and prostate cancer, and Attia argues that stacking tests improves them. Flaherty then explains how sequencing, fragment sizes and methylation in circulating tumor DNA can detect cancer earlier and suggest the tissue of origin, including use after surgery.
- •Mammography cuts breast cancer death by about a third, and colon screening by about 25–30%
- •Attia argues that stacking tests with different sensitivities and specificities makes screening far better than any single test
- •Circulating tumor DNA can be sequenced and its methylation pattern can point to the tissue of origin
- •Serial sampling and much more sensitive research methods are coming
- •Detecting minimal disease post-surgery is easier, but earlier detection in unselected people is the harder and more important problem
- 1:49:22 – 1:57:23
Commercial cancer detection tests and what differentiates them
Attia asks about tests people can get today. Flaherty names several commercially available blood tests, cautions that positive results without clinical expertise can cause anxiety, and explains how mutations, fragment length, and methylation differentiate the companies.
- •Flaherty says it is reasonable for early adopters to get a test but it isn't yet imperative
- •He names tests from Thrive (acquired by Exact), Delphi, and Grail (which Attia uses with patients)
- •The medical community has not yet worked out how to manage positive results with no visible tumor
- •Fragmentomics and methylation are key differentiators beyond mutation detection
- •Investment in diagnostics is rising, with about ten more companies coming
- 1:57:23 – 2:05:38
Closing the gap in cancer care access and the outlook for progress
Attia asks what patients can do to get expert guidance given disparities in cancer care. Flaherty proposes pooling expert opinion and using AI-based decision support for typical cases while experts handle edge cases, and mentions N of 1 and xCures. They close with optimism about the convergence of diagnostics and therapeutics.
- •Care disparity means it matters who you know and which expert directs you
- •Flaherty wants to pool insights across institutions and use technology to build a database of expert opinions
- •Typical cases could get decision support from past similar cases while experts focus on edge cases
- •N of 1 and xCures are mentioned as navigation efforts, but a full insight machine does not yet exist
- •Flaherty says the convergence of diagnostics and therapeutics is the take-home message and should raise 10-year survival
