Hantavirus, in two chapters

The headline picture, in two chapters

Hantaviruses cause two distinct severe diseases — pulmonary syndrome (HPS) in the Americas and renal syndrome (HFRS) in Eurasia — with case-fatality rates from under one percent (Puumala) to roughly thirty-eight percent (Sin Nombre, Andes). The 2025–2026 story breaks neatly in two:

Chapter I · Quiet US baseline drift (2024 → April 2026)

The Hackman household death in Santa Fe (February 2025) and the three-fatality Mammoth Lakes cluster (February–April 2025) were unusual but biologically unsurprising — Sin Nombre virus spilling over from local Peromyscus maniculatus, no human-to-human transmission, no detectable change in viral biology. Annual US case counts (CDC + PAHO surveillance) are drifting upward but inside historical envelopes.

Chapter II · The MV Hondius Andes-virus cluster (April–May 2026)

The most consequential hantavirus event of the decade. Andes virus is the only orthohantavirus with documented person-to-person transmission, and the Hondius incident appears to be the largest ANDV cluster outside the historical Argentina/Chile core ever recorded. WHO DON-600 (2026-05-08): 8 cases, 3 deaths, CFR 38%, multi-country, with onward transmission to a ship's doctor and at least one passenger contact.

Chapter II is why hantavirus suddenly entered the public, the rationalist, and the prediction-market conversation in early May 2026 — and why, across these pages, we treat the Hondius response (good and bad) as a case study in how the post-COVID public-health system actually behaves when a low-prior, high-consequence pathogen story breaks.

Key numbers

A note on reconciliation: the recent-outbreaks file flags that the PAHO 2025 US count (7 cases through epi-week 47) appears to undercount state-DOH reporting (NM alone reports 7 in 2025 per file 07). Treat headline annual numbers as lower bounds; the canonical CDC year-by-year breakdown is not published on the public dashboard.

Three recent dossiers

Mini-timeline · MV Hondius

Virology & epidemiology, in compressed form

Hantaviruses are tripartite negative-sense RNA viruses (genome segments S, M, L) in family Hantaviridae, order Bunyavirales. Of 53 species across 8 genera, only the rodent- and insectivore-borne Orthohantavirus contains confirmed human pathogens. Reservoir infection is typically persistent and asymptomatic; spillover to humans is by aerosolisation of urine/feces/saliva, occasionally by bite. Andes virus is the lone exception with documented person-to-person transmission.

Reservoir-host pairings

Case-fatality rates by syndrome and species

Treatment & vaccines, briefly

• HPS: supportive care, ECMO for severe cases (~80% survival when started early). Ribavirin not effective (two trials, meta-analysed). Monoclonal candidates: SNV-42 (Vanderbilt/USAMRIID), JL16/MIB22 for ANDV (hamster models), SAB-163 quadrivalent polyclonal (IND-enabled 2024). Favipiravir 100% survival in ANDV hamster model; no human trials.
• HFRS: ribavirin does work (Huggins 1991, n=242, 7× mortality reduction). Hantavax (Korea, inactivated, 1990); Chinese bivalent inactivated HFRS vaccines (~2M doses/year; HFRS cases fell 37,814 → 11,248 in 2000–2007).
• Pipeline: USAMRIID DNA vaccines (HTNV, PUUV) Phase 2a; ANDV DNA vaccine completed Phase 1 (98% reactogenicity rate); rVSV-HTNV-GP animal data 2024; Korea University / Moderna mRNA collaboration (2023; preclinical 2025). No FDA/EMA/WHO-approved hantavirus vaccine exists.

Prediction markets · ~$4.7M+ volume

No hantavirus prediction markets existed before 2026-05-04. The MV Hondius WHO notification triggered the first burst of market creation in early May 2026.

Calibration signals

• Hantavirus pandemic 2026 (8%) < "any new pandemic 2026" (12%) — markets are pricing hantavirus as a below-average pandemic risk among possible escalations.
• Kalshi PHEIC ~23% >> Polymarket "pandemic" 8% — structural, not arbitrage. PHEIC has a much lower bar than WHO's word-of-art "pandemic" characterisation.
• Lab-leak market at 3% — consistent with no credible reporting of a laboratory origin. The market exists at all because of a residual COVID-19 reflex; volume ($104K) is modest.
• FDA vaccine in 2026 at 11% — the page text explicitly notes no candidate is in Phase 3, with full efficacy studies typically requiring 4–7 years. The 11% reflects optionality on emergency-use pathways, not realistic full approval.

Null finds: Metaculus, Good Judgment Open, INFER, PredictIt, Smarkets, Insight Prediction, Augur, Hedgehog — all empty on hantavirus as of 2026-05-10.

The rationalist take: sparse, late, occasionally sharp

Until the Hondius story broke, the rat-sphere essentially ignored hantavirus — including the Hackman case, despite its mainstream reach. The cluster generated a roughly two-week burst of LessWrong / EA Forum / Substack discussion in early May 2026. The substantive material is below.

Notable absences (the more interesting finding)

• No ACX / Slate Star Codex post on hantavirus, the Hackman case, or the Hondius cluster — despite the H5N1 deep-dive precedent ("H5N1: Much More Than You Wanted to Know").
• No SecureBio / NAO post; no 80,000 Hours, Open Philanthropy, Asterisk, or Works in Progress piece.
• No Yudkowsky / Hanson / Aaronson tweet surfaced; no Saloni Dattani OWID piece; no Samotsvety / Sentinel forecast write-up.
• No LessWrong post on Andes-virus H2H GCBR fit despite the obvious topical relevance.
• The Feb 2025 Hackman case, despite mainstream reach, generated essentially zero rat-sphere engagement. Discourse switched on only for the H2H story.

Ecology & climate drivers

Two mechanisms compete for the 2024–26 US picture, and crucially — both make the same prediction for Mono County 2025.

Mechanism A · Trophic cascade (the textbook)

The classic Mills/Yates framework: El Niño → enhanced winter precipitation → seed/nut boom → 10× Peromyscus maniculatus population (e.g. spring 1993 Four Corners) → spillover to humans. Three documented cascade chains: 1991-92, 1997-98, 2004-05 precipitation → outbreaks the following year. Lag ≈ 6–9 months. The 2024–26 winters were La Niña (drought-promoting) — opposite of the canonical mechanism. The cascade explanation must therefore reach back to California's 2022–23 atmospheric-river season (283% of normal Sierra snowpack) with rodent peaks materialising 18–24 months later. Plausible but inferred (C2/C3); no Eastern Sierra trap-census data was located.

Mechanism B · Arid behavioural incursion

A 2016 mechanistic re-analysis found a strong negative precipitation–HPS relationship at the national scale (r² = 0.78), with cases rising exponentially below ~1.5 mm/day. Mechanism: arid conditions push mice into peridomestic structures. La Niña 2024–26 fits this directly. Both mechanisms can be true simultaneously and they converge on the Mono County 2025 prediction.

Argentina 2025–26 · the colilargo story

Cases up ~77% (101 vs 57) and a geographic shift from Patagonia to Buenos Aires province. Expert attribution: milder winters → higher overwinter Oligoryzomys longicaudatus survival, plus periurban expansion. This is independent corroboration that something has shifted in South American reservoir dynamics.

Climate-projection literature, in one paragraph

A PNAS 2025 meta-analysis ranks hantaviruses as the second-most-studied pathogen group in climate-zoonosis research; 69% of temperature effects and 64% of precipitation effects in modelled studies were statistically significant. Projections under RCP4.5/8.5 for 2040–2050 estimate 31–34% more people at risk globally. The European Puumala analog confirms the mast-year mechanism but also illustrates that warming has opposite effects in different systems (West-Central Europe: more outbreaks; Scandinavia: fewer due to snow-cover loss). Generic "climate change increases zoonotic risk" framings flatten this — see file 06 for the texture.

Pandemic-prep & the GCBR question

Official biothreat-list status (current)

• US Federal Select Agent Program — not listed. The December 2024 biennial review explicitly considered SNV, ANDV, HTNV, DOBV and declined all four. Stated reason: limited H2H, hard to culture, low weaponisation feasibility. Effective list date: 2025-01-16. (selectagents.gov)
• WHO Pathogens Prioritization 2024 — included. Hantaviridae family is named; O. sinnombrense (SNV) is the designated prototype pathogen; O. hantanense (HTNV) the second priority. This is an upgrade from the 2018 R&D Blueprint where hantavirus ranked P13 and was off the primary list.
• NIAID — Category A (highest research-priority tier; grouped under Viral Hemorrhagic Fevers / Bunyaviruses). CDC's operational classification is Category C. The discrepancy reflects different agency mandates (research funding vs operational response).
• CEPI — not on the formal priority list of 8. Adjacent activity: Korea University / Moderna mRNA collaboration via Moderna's mRNA Access program (active since 2023).

Pandemic-potential consensus (and what would change it)

Across the CDC, ECDC, WHO, and the published literature, hantaviruses are rated low pandemic-potential. The reasons are stable: no presymptomatic transmission documented; rodent-borne maintenance route; environmentally fragile virions; difficult lab culture (BSL-3, low yields). Andes virus is the documented exception, but its R₀ in the 2018-19 Epuyén cluster was 2.12 pre-intervention → 0.96 post-isolation — once isolation works, the chain crashes.

Three things would change the verdict:

1. Sustained ≥3rd-generation H2H in the Hondius cluster. Not observed yet — the ship's-doctor and the British evac case are both first-generation contacts of the index couple. As of 2026-05-10 the chain looks short.
2. Reassortment between SNV and ANDV producing an SNV-like host range with ANDV-like H2H. Demonstrated experimentally between SNV and BCCV (Black Creek Canal) in lab settings. Theoretical, not observed in nature.
3. A laboratory escape — currently priced at 3% on Polymarket. No credible reporting suggests one.

Under the Johns Hopkins working definition of a GCBR (a biological event with civilization-scale, systemic impact), hantavirus does not currently qualify.

Public-health response · graded on the curve

What worked

• Mono County Public Health was widely praised for timely, specific, transparent messaging during the Mammoth Lakes cluster — three press releases naming the PHO and detailing exposure circumstances.
• WHO issued DON-599 within 48 hours of formal notification and DON-600 four days later. CDC issued HAN 00528 on 2026-05-08 with explicit clinical guidance and 42-day contact-monitoring recommendations.
• ECDC deployed a field expert and published a Threat Assessment Brief within 4 days of WHO notification.
• Diagnostics: a September 2024 ScienceDirect study reported LFIA hantavirus-IgM/IgG prototypes at 100% sensitivity / 97.5–99.3% specificity — point-of-care progress, though no FDA-cleared rapid test is yet on market.

What didn't

• The Saint Helena gap. ~30 Hondius passengers disembarked on 2026-04-24 without contact tracing — 13 days after the first death. The widow's KL592 boarding compounded the gap. This is the single defining operational failure of the response.
• The 3-week WHO-notification lag. The first death was 2026-04-11; WHO notification came 2026-05-02. The chain of decision-making across the operator, the ship's medical staff, and Saint Helena port-health is not yet on the public record.
• NMDOH messaging on the Hackman case was reactive around the celebrity dimension; the substantive prevention guidance circulated, but framing was driven by the press cycle.
• CDC HAN tier confusion. Some outlets framed the response as "Level 3" (the agency's lowest activation tier) and called the HAN itself "Level 3"; the HAN page is in fact a Health Advisory (mid-severity HAN tier — Alert > Advisory > Update). Worth sorting out for future signal-reading.

Vaccine pipeline regulatory status (today)

• USAMRIID DNA vaccines — HTNV, PUUV in Phase 2a; ANDV completed Phase 1.
• Moderna · Korea University mRNA — preclinical mouse efficacy data published Feb 2025; Phase 1 in humans not yet confirmed.
• Approved products globally: South Korea's Hantavax (HTNV, since 1990) and Chinese bivalent inactivated HFRS vaccines. No FDA / EMA / WHO-approved hantavirus vaccine.

Contradictions, gaps, and bias-watch

Cross-file tensions, surfaced not papered over

C5 unknowns flagged across files

• Detailed Eastern Sierra small-mammal trap-census data 2024–25
• Specific MMWR articles on US hantavirus 2024–2025 (search returned none; may exist behind direct mmwr.cdc.gov queries)
• WHO PHEIC committee proceedings on the Hondius cluster (likely confidential)
• Genome-sequence comparison of Hondius ANDV strain vs Argentine reference (presumably in progress)
• Any private rat-sphere Discord / Twitter DM discourse beyond what surfaced in public search

Bias-watch (per RESEARCH-BEST-PRACTICES.md §3.2)

• Optimistic framing. Public-health "low risk to the public" framing for Hondius is plausibly correct, but crew/passenger risk is rated moderate by WHO. We've surfaced both.
• Single-source reliance. Some Hackman-property details rest on a single CNN or Fox piece — flagged in file 01.
• Recency bias. The whole project is shaped by Hondius being live news; we have explicitly preserved Mono / Hackman / Argentina material so the picture isn't reduced to one event.
• Speculative-as-base-case. We have not treated mRNA vaccine availability or pan-hantavirus mAbs as base-case assumptions. Pipeline only.
• Sycophantic framing of the rat-sphere avoided: file 03 explicitly notes the silence on the Hackman case and on H2H GCBR fit, rather than only quoting existing posts approvingly.

Sources & methodology

Seven topic files, ~2,500 lines of structured claims, ~80 distinct WebSearch queries and ~50 WebFetch requests against authoritative sources. Every claim carries a confidence tier (C1–C5) and at least one URL. Below is the canonical primary-source register.

Confidence legend

C1 verified   C2 well-sourced   C3 inferred   C4 anecdotal   C5 unknown

Primary sources

• WHO DON-599 (2026-05-04) — link
• WHO DON-600 (2026-05-08) — link
• WHO statement, Tedros (2026-05-07) — link
• WHO hantavirus factsheet — link
• CDC HAN 00528 — link
• CDC hantavirus cases page — link
• CDC MMWR Yosemite 2012 (mm6146a5) — link
• EID Yosemite 2014 (Núñez et al.) — link
• NMDOH press release (2025-03-07) — link
• Mono County PH 1st death — link
• Mono County PH 2nd death — link
• Mono County PH 3rd death — link
• PAHO Dec-2025 epidemiological alert — link
• NEJM Epuyén 2020 — Martínez VP et al. — PubMed search
• FSAP final rule (Dec 2024) — link
• FSAP list (effective 2025-01-16) — link
• WHO 2024 Pathogens Prioritization — PMC11627490
• NIAID biodefense pathogens — link
• ICTV Hantaviridae profile 2024 — PMC11094369
• Mills/Yates 2002 BioScience — link
• JID 2001 El Niño / Four Corners — link
• HPS 25th Anniversary review (PMC6199996) — link
• Mechanistic Understanding (PMC4776718) — link
• Linch · LessWrong (2026-05-09) — link
• Kelsey Piper · The Argument (2026-05-08) — link
• Polymarket: Hantavirus pandemic 2026 — link
• Kalshi: WHO PHEIC 2026 — link
• Manifold mirror — link
• NEON small-mammal surveillance reports — see file 06
• Korea CDC HFRS surveillance — see file 02 / 07

Methodology, in one paragraph

This monograph was assembled on 2026-05-10 by an AI research workflow described in /workspace/overview/RESEARCH-BEST-PRACTICES.md and /workspace/overview/research-agent-instructions.md. Seven topic-specialist sub-agents searched, fetched, and extracted claims into /workspace/health/hantavirus/research/ (files 01–07), each enforcing C1–C5 confidence tagging and "URL or it didn't happen" sourcing. A coordinator agent then produced this synthesis and the webapp. Content has been AI-compiled and AI fact-checked, not personally verified. URLs in this document are the URLs the search tools actually returned; if a link 404s, the canonical archive (Wayback) should be consulted before treating the absence as a contradiction.