The Operating System for Human Health

• Fitbit becomes Google Health (May 19 rollout; in stores May 26). The Fitbit app auto-updates to the new Google Health app — with a Gemini-powered “Health Coach”, medical-record integration, and data feeds from Apple Health, Peloton, and MyFitnessPal. Fitbit Premium → Google Health Premium. Same launch window: screenless $99 Fitbit Air (Steph Curry edition $129.99).

  ↳ Continues Fitbit Air $99 launch (Issue #5). The category is shedding “fitness tracker” identity — the wearable is now the front end of a health platform, AI as the interface, medical records inside the same app.

  👉 Why it matters: A step counter that now hosts your labs, your meds, and a 24/7 AI coach. Google is positioning to own the line between consumer wellness and clinical care.  Google Health | Droid-Life →

• AstraZeneca licenses Owkin’s K Pro “AI Scientist” for 3 years (May 13) — another top pharma commits to agentic AI that generates hypotheses, interrogates literature, surfaces drug targets, prioritizes indications, designs experiments, and maps the competitive landscape of clinical trials.

  ↳ Continues a pattern seen across Issues #4 and #5: pharma partners with AI vendors.

  👉 Why it matters: Pharma licenses the AI engine rather than building it. Lilly gets molecules, Insilico keeps the platform. AZ gets agents, Owkin keeps K Pro. The defensibility sits on the AI side.  Press release | MobiHealthNews →

• Omada Health joins Lilly Employer Connect — Omada (Nasdaq: OMDA) becomes an independent program administrator in Lilly’s Employer Connect, offering its GLP-1 Care Track (clinical evaluation, prescribing, and ongoing medication management). The integration provides transparent net cost for Zepbound and lets employers set defined contribution levels. Same day, Omada reported Q1 2026 revenue of $78M (+42% YoY), 1.02M Total Members (+51% YoY), and a narrowed operating loss. Stock rose ~9% on the week. This is bigger than a distribution deal. It signals pharma’s transition into longitudinal digital care infrastructure — drug companies no longer just selling therapeutics, but building behavioral operating systems around them.

  ↳ Continues the “Two Front Doors” thesis (Issue #5) — now three channels, three unit economics, one drug. Omada’s coordinated lifestyle support is the adherence wrap that makes the employer channel work.

  👉 Why it matters: Pharma is shifting into longitudinal digital care infrastructure. The GLP-1 wars are evolving from molecule competition to adherence, coaching, and outcomes. The moat is continuous patient engagement around the drug.  Omada PR →

• CMS ACCESS is ready for the July 5 launch (150+ first-cohort participants have enrolled by May 15 deadline).

  How ACCESS works

  Voluntary 10-year payment model (CMS Innovation Center) for Medicare Part B providers. Replaces fee-for-service with Outcome-Aligned Payments (OAPs): monthly prospective per-beneficiary payment + annual reconciliation tied to BP, HbA1c, depression scores, MSK pain.

  4 clinical tracks · FHIR-based outcome reporting · quarterly cohorts through 2033, model runs through June 2036.

  ↳ Continues the ACCESS thread (Issue #3 → Issue #4). Insurers representing ~165M Americans across commercial, Medicare Advantage, and Medicaid have aligned their reimbursement with the ACCESS approach.

  👉 Why it matters: July 5 stress-tests the “wearables → reimbursable medical signal” bridge. ACCESS pays for outcomes, not visits. Q4 2026 / Q1 2027 data will separate the digital-therapeutic companies that deliver from those that market.  CMS ACCESS →

• 🔬 INDUSTRY-FIRST APPROVAL — FDA approves atezolizumab (Tecentriq) + Signatera CDx for ctDNA-positive MIBC (May 15). Two precedents: (1) Tecentriq as adjuvant for MIBC with molecular residual disease (MRD) post-cystectomy; (2) Natera’s Signatera — the first FDA-approved blood-based MRD companion diagnostic. Based on IMvigor011 Phase 3 (250 patients, 2:1 randomized; Genentech / Roche).

  ↳ Resolves the SERENA-6 question from Issue #5. SERENA-6 had immature OS; IMvigor011 brought mature OS (32.8 vs 21.1 mo, HR 0.59) with the CDx co-developed from trial design. FDA signal: ctDNA-guided treatment works when OS is mature and the CDx is built in.

  👉 Why it matters: Third drug–CDx co-approval in three weeks (after Labcorp+KEYTRUDA in Issue #4 and Veppanu+Guardant360 in Issue #5). Adjuvant immunotherapy runs ~$196K/patient-year; Signatera-negative patients (~half) showed 97% 2-year OS without treatment. MRD-guided care is the economic gatekeeper. Natera stock +5%.  FDA | Natera →

• FDA approves Enhertu (T-DXd) for two new HER2+ early breast cancer indications (May 15, Daiichi Sankyo / AstraZeneca). First ADC into curative-intent HER2+ early breast cancer: (1) neoadjuvant T-DXd → THP for Stage II/III (DESTINY-Breast11); (2) adjuvant for residual invasive disease after neoadjuvant trastuzumab+taxane (DESTINY-Breast05). DB05 cut invasive recurrence or death 53% (HR 0.47); 3-year iDFS 92.4%.

  Therapy timing — where this week’s approvals slot in

  Early stage (curative-intent) ————————→ Metastatic (palliative)

  Cancer / regimen	Neoadjuvant	Surgery	Residual-disease assessment	Adjuvant	Met 1L	Met 2L+
  HER2+ early breast cancer (T-DXd)	✨ T-DXd (DB11, NEW)	Mastectomy / lumpectomy	Pathologic eval of resected breast + nodes (histology)	✨ T-DXd (if pathologic RD; DB05, NEW)	T-DXd (2024)	T-DXd (2019–22)
  Muscle-invasive bladder cancer (atezolizumab + Signatera)	—	Radical cystectomy	✨ Signatera ctDNA-based MRD (NEW)	✨ Atezolizumab (if Signatera+; IMvigor011)	—	—

  ✨ = May 15, 2026 approval. Note the different residual-disease modalities: HER2+ breast cancer uses pathologic assessment of the surgical specimen (histology of resected tissue and lymph nodes — the pCR vs RD framework). MIBC uses molecular assessment via Signatera ctDNA — the first FDA-approved blood-based MRD CDx. Signatera is approved only in combination with atezolizumab for MIBC; it is not used in breast cancer.

  👉 Why it matters: Enhertu moves up the cancer-care timeline — from late-line metastatic into curative-intent neoadjuvant/adjuvant. Longer treatment durations, larger eligible population. Safety limits adoption: ILD/pneumonitis 4.4% (DB11) and 10% (DB05); boxed warning unchanged.  FDA | AstraZeneca →

👉 The shape of progress: 62% of these approvals are in oncology, 29% in rare disease, 9% “other.” The story isn’t cancer alone — it’s that the tooling built for oncology (biomarker-guided drug + CDx + AI) is now being ported into rare disease and chronic conditions.

Modality breakdown of the 16

Datroway and Emrelis are the two ADCs; Redemplo (apoC-III-directed) is the only siRNA — the most novel modality of the cohort. Zero bispecifics, zero cell therapies (those sit in the 5 gene/cell therapy tally below).

+1. Baby KJ’s bespoke CRISPR (custom-designed lipid-nanoparticle base-editing therapy, U Penn / CHOP team, IND-stage, not FDA-approved) — Severe CPS1 deficiency, n-of-1. A CRISPR therapy designed for one specific infant. FDA approved the expanded-access IND within one week of submission (Feb 14, 2025). By end of 2025, the infant was walking and talking. The first FDA-blessed playbook for n-of-1 gene editing — not a commercial product, but a regulatory precedent for ultra-rare individualized therapies.

👉 What it adds up to: 16 personalized NMEs + 5 gene/cell therapies + 16 CDx + 1 AI development tool + 1 bespoke CRISPR for a single patient — in one year. The signal isn’t any one approval. It’s that the regulatory plumbing for personalized medicine is now fully operational, end to end.

The epigenetic aging clocks — what they measure, what they need

Input in practice: all eight clocks require a single blood draw analyzed on a DNA methylation microarray (typically Illumina EPIC, ~850K CpG sites). Horvath2013 works across multiple tissues; the rest are blood-specific. The 2nd-generation clocks (PhenoAge, GrimAge) used additional clinical labs and plasma proteins in training, but at application require only the methylation profile.

*GrimAge was not included in the seven clocks measured by the UCL study. PhenoAge (the other 2nd-gen clock) and the two pace-of-aging clocks all showed slower aging; the four chronological-age clocks showed no effect. The biology of aging is moving from a single number to a panel.

The continuous loop — the new OS for human health

↳ and back to human behavior. 24/7. On the wrist and in the pocket.

👉Healthcare is being reorganized from a network of places into a single continuous loop of biological intelligence.