#06 – D.A. Wallach: music, medicine, longevity, and disruptive technologies

Attia's clinical position is that liquid biopsies, when they mature, should be used as confirmation — not as lead screening — because false positives from a blood test would cascade into invasive follow-up procedures (ERCP, biopsy) for findings that may resolve on their own. Wallach pushes back: if sensitivity and specificity are both near 100%, lead with the liquid biopsy precisely because it is cheap and non-invasive.

This is Wallach and Attia's shared vision of what personalized medicine actually means in practice. Today, sensitivity and specificity are discussed as properties of tests, not as properties of test-applied-to-this-specific-person. Bayesian medicine inverts this: the person's prior is primary, and the test value updates it. The implication is that recommending the same screening protocol to all women over 40 is a category error — the protocol should vary based on individualized disease probability.

Attia uses PSA as the canonical example of the threshold problem: set the PSA cutoff at 1 and you catch every prostate cancer but generate enormous false positive rates; set it at 20 and you miss cancers but confirm every positive. The field has aligned on this design priority — but Attia and Wallach clarify that the correct framing is 100% negative predictive value (never miss a true cancer), not 100% specificity, which are different statistical concepts.

Wallach's framework emerged from disillusionment with digital health: healthcare is largely a labor business (doctors, nurses) with legally constrained margins, not a content-distribution business where marginal cost goes to zero. Kaiser has 3-4% gross margins. The structural conclusion: if you want to generate outsized returns and outsized human benefit, bet on the molecules.

Sean Parker funded Nadeau's center because his own severe allergies led him to study the immune system, and he recognized a crossover between the allergy and cancer biology immune pathways. The allergy work is proof-of-concept for the broader immune modulation paradigm: the immune system can be reprogrammed in adults. Attia notes that allergy immunotherapy has always been called immunotherapy — long before cancer immunotherapy made the term fashionable.

Attia's cancer pessimism is specifically about curing detected cancer, not about preventing it. The liquid biopsy discussion is precisely about closing the cancer prevention gap — catching it before it metastasizes, in the window when immune-mediated intervention or surgery could be curative. The IL-1/methotrexate cardiovascular data was exciting precisely because it opened a new therapeutic lever beyond the lipid playbook.

The Oncoblot analysis was particularly instructive: the test was designed for metastatic cancer of unknown primary (where it worked well, distinguishing 27 tissue types) but was being marketed as a screening tool, where its poor specificity in healthy populations created unacceptable false-positive rates. Attia's interpretation: it may have been picking up cancers that the immune system normally winnows out — a deep dilemma about whether early detection of pre-cancer is beneficial or harmful. GRAIL (the Illumina spinout) takes a different approach: instead of a protein biomarker, deep sequencing of peripheral blood to find sub-0.1% concentrations of tumor DNA, relying on the same sequencing infrastructure that has revolutionized genomics.

GRAIL premises that by sequencing the blood with enough depth — sequencing it over and over and over again — you can detect just the DNA coming from the tumor and identify its tissue of origin. The challenge: at very early cancer stages, tumor DNA concentrations may be in the parts-per-million range. Attia frames this as the most promising near-term path to seeing cancers years before they produce symptoms.

Until we reach that singularity, we are stuck running one-variable-at-a-time controlled experiments on humans, mice, dogs, or monkeys. Clinical trials are both slow and brutal — many participants die from experimental medicines or endure enormous suffering without benefit. Wallach frames his investment portfolio as stepping stones toward that singularity, rather than individual bets on specific therapeutic categories.

Wallach applies the same framing to Bitcoin: whether Bitcoin replaces the dollar is almost irrelevant. What matters is that the smartest 19-year-old computer scientists shifted attention to the area, and talent concentration produces something significant regardless of the specific application.

The bottleneck is not science; it is the cascade of medical societies, payers, and government guidelines that slow proven interventions from becoming standard of care. Getting definitive science to become standard practice faster is itself a massive societal priority that medicine neglects.

Attia frames atherosclerosis as an immune system trade-off: the same inflammatory machinery that protects us from infection can, when dysregulated, progressively damage the endothelium. The allergy/immunotherapy analogy is apt: immune modulation already works for food allergies, and the same principles may apply to cardiovascular auto-inflammation.