Measured Levels & Durations
Human circulating concentrations (pg/mL), tissue persistence durations, and experimental in vitro/in vivo concentrations — with dose-relevance assessment from the Spike Protein Dose scoping review.
Concentration database
Compiled from Yonker LM, et al. (2023)Opens in new tab, Selem E, Raszek M, Varon J, Halma MTJ (2026)Opens in new tab, and the primary experimental studies cited in each row below.
Dose translation summary
Most in vitro pathological effects occur at concentrations far above circulating human pg/mL levels, but some mechanisms (complement, endothelial activation, autoimmunity, BBB) may show plausible overlap at lower effective doses.
- Human circulating
- Low pg/mL (positive samples) to undetectable
- In vitro typical
- ng/mL to µg/mL for cytotoxicity; lower for some non-cytotoxic effects
- Magnitude gap
- ~10³–10⁶ between human circulating and in vitro cytotoxicity concentrations
Observed in humans
| Concentration | Endpoint / finding | Source | |||
|---|---|---|---|---|---|
| Human — circulating | Plasma (adolescents, mRNA myocarditis) | ~34 pg/mL | Free spike protein | Detected in myocarditis cases; not in asymptomatic controls | Yonker LM, et al. (2023)Opens in new tab |
| Human — circulating | Plasma (post-vaccine sequelae cohort) | Low pg/mL (subset positive) pg/mL | Circulating spike | Persistent detection months post-vaccination in symptomatic subset | Swank DL, et al. (2023)Opens in new tab |
| Human — circulating | Plasma (research assays) | Sub-pg/mL to low pg/mL pg/mL | Spike protein | Assay limit of detection | Selem E, Raszek M, Varon J, Halma MTJ (2026)Opens in new tab |
| Human — circulating | Blood/plasma | Trace–low pg/mL pg/mL | Spike protein | Persistence up to ~709 days in outlier cohorts | Selem E, Raszek M, Varon J, Halma MTJ (2026)Opens in new tab |
| Human — circulating | Serum (post-COVID ME/CFS cohort) | Variable; often below assay LOD pg/mL | Spike protein | Spike persistence not associated with ME/CFS diagnosis in this cohort | Various (2025)Opens in new tab |
| Human — tissue | Blood, heart, liver, spleen, brain-meninges (autopsy/biopsy) | Qualitative IHC/PCR positive qualitative | Vaccine-derived spike + mRNA | Up to ~6 months; brain borders up to ~4 years (limited N) | Fehrer C, et al. (2024)Opens in new tab |
| Human — tissue | Lymph node germinal centers | Localized antigen (not circulating) qualitative | Vaccine antigen | Persistence 8+ months | Röltgen K, et al. (2022)Opens in new tab |
Experimental & animal models
| Concentration | Endpoint / finding | Source | |||
|---|---|---|---|---|---|
| In vitro (experimental) | Mixed neuronal–glial cell cultures | 50 µg/mL | Exogenous S1 subunit | Induced neuronal cell death; stimulated microglial proliferation (first 3 days) | Various (2026)Opens in new tab |
| In vitro (experimental) | Microvascular endothelial cells | ~10 (~50 nM S1) µg/mL | S1 subunit | ~50% reduction in cell viability (approximate LD50) | Perico L, et al. (2022)Opens in new tab |
| In vitro (experimental) | Endothelial cells + complement | µg/mL range (S1) µg/mL | S1 subunit | Complement activation, platelet aggregation | Perico L, et al. (2022)Opens in new tab |
| In vitro (experimental) | Cell-free recombinant spike | 2.5–25 µg/mL | Recombinant spike (degradation by nattokinase) | Spike degradation demonstrated | Tanikawa T, et al. (2022)Opens in new tab |
| In vitro (experimental) | Cell-free fibrinogen | Low µg/mL range (S1) µg/mL | S1 subunit | Amyloidogenic fibrin resistant to fibrinolysis | Grobbelaar LM, Venter C, Pretorius E (2021)Opens in new tab |
| In vitro (experimental) | Human cardiac pericytes | Low ng–µg/mL range ng/mL | Full spike | Pericyte dysfunction via CD147 (non-infective) | Avolio E, et al. (2021)Opens in new tab |
| In vitro (experimental) | Primary human trophoblasts + BeWo cell line | 100 ng/mL | S1 subunit | Inflammatory response, cytotoxic/antiproliferative effect, G2-M arrest, disrupted glucose/folate/amino-acid transporters | Various (2026)Opens in new tab |
| In vitro (experimental) | Co-cultured human alveolar epithelial cells + macrophages | 0.1–10 µg/mL | Recombinant S1 protein | Dose-dependent LDH release, IL-6 and TNF-α production; 10 µg/mL matched LPS-induced injury by day 3 | Various (2025)Opens in new tab |
| In vitro (experimental) | BEAS-2B lung epithelial cells + MRC-5 fibroblasts | 1,000 ng/mL | Spike protein | Increased IL-6, TNF-α, CXCL1, CXCL3; epithelial-mesenchymal transition via GADD45A upregulation | Various (2023)Opens in new tab |
| In vitro (experimental) | Citrated whole blood (platelets) | 2–20 ng/mL | Recombinant spike (Ancestral, Alpha, Delta, Omicron variants) | Decreased platelet count, increased mean platelet volume, platelet clumping/activation — most pronounced with Delta/Alpha at 20 ng/mL | Various (2024)Opens in new tab |
| In vitro (experimental) | Ex vivo whole blood (platelets) | 5 µg/mL | Recombinant spike (Alpha, Beta, Gamma, Delta variants) | No significant change in platelet aggregability, P-selectin, PAC-1 binding, or thromboelastography parameters | Various (2023)Opens in new tab |
| In vitro (experimental) | A549 lung epithelial cells | 100 ng/mL | S1 subunit | NLRP3 inflammasome activation, IL-6/IL-1β/IL-18 release | Various (2024)Opens in new tab |
| In vitro (experimental) | BEAS-2B and A549 epithelial cells | 1 µg/mL | S1 or S2 subunit | S1 increased IL-6 in BEAS-2B; miR-149-5p largely unaffected (contrasts with poly(I:C) response) | Various (2024)Opens in new tab |
| In vitro (experimental) | Mouse lung epithelial cells | 500 ng/mL | S2 subunit | Proinflammatory cytokine/chemokine induction (Tnf-α, Il1β, Il6, Ccr2, Ccr5), attenuated by IGFBP2 overexpression | Various (2025)Opens in new tab |
| In vitro (experimental) | A549 lung epithelial cells | 100 ng/mL | S1 subunit (Wuhan and Omicron variants) | NLRP3/MAPK-mediated IL-6, IL-1β, IL-18 secretion, dose-dependently reduced by luteolin/H. perforata extract | Various (2025)Opens in new tab |
| In vitro (experimental) | A549 lung epithelial cells | 100 ng/mL | S1 subunit | NLRP3 inflammasome activation via JAK1/STAT3, reduced by Perilla frutescens luteolin fraction | Various (2022)Opens in new tab |
| In vitro (experimental) | A549 lung epithelial cells | 100 ng/mL | S1 subunit | NLRP3 inflammasome activation via Akt/MAPK/AP-1, reduced by hesperetin | Various (2022)Opens in new tab |
| In vitro (experimental) | Xenopus oocytes expressing human nAChR subunits | 1 µg/mL | Spike receptor-binding domain (RBD) | Marked reduction in current amplitude at α4β2 and α4α6β2 nAChR subtypes; equivocal at α3α5β4; absent at α3β4/α7 | Various (2023)Opens in new tab |
| In vivo — animal models | Mouse (mRNA vaccination model) | Systemic exposure (vaccine dose) varies | Vaccine-encoded spike mRNA | Myocarditis recapitulated; ameliorated by CXCL10/IFN-γ inhibition | Cao X, et al. (2025)Opens in new tab |
| In vivo — animal models | hACE2 mice (S1 protein administration) | Systemic S1 exposure varies | Spike S1 subunit | Brain fog–like cognitive phenotype (reduced learning, memory, nesting); ameliorated by NRICM101 | Chang CC, et al. (2024)Opens in new tab |
Assays
- Simoa (Single Molecule Array) — ultra-sensitive digital ELISA capable of sub-pg/mL detection; used in Yonker et al. and most positive-detection circulating studies.
- High-sensitivity / standard ELISA — widely available but with a substantially higher limit of detection than Simoa; negative results do not rule out low-level circulating spike.
- IHC / ISH (immunohistochemistry / in situ hybridization) — used for tissue localization in autopsy and biopsy series; qualitative, not a mass concentration.
- RT-qPCR — detects vaccine mRNA rather than protein directly; used for blood, tissue, and breast milk persistence studies.
- Mass spectrometry — used in some systematic-review-cited studies for orthogonal confirmation of ELISA/Simoa findings.
Assay choice strongly affects reported detection rates: studies using Simoa consistently report higher positivity than those using standard ELISA, since much of the circulating spike protein literature is right at or below conventional assay limits of detection.
Chart — human vs. experimental concentrations
Human circulating Experimental (in vitro)
Values converted to a common pg/mL basis for log-scale comparison only; original units and study context differ substantially (see table above) and should not be treated as directly equivalent doses.
Table 2 — Bodily fluids
From Selem E, Raszek M, Varon J, Halma MTJ (2026)Opens in new tab — blood persistence up to ~709 days in outlier cohorts.
| Concentration | Source | ||||
|---|---|---|---|---|---|
| Blood (plasma/serum) | Spike protein | Up to ~709 days (~23 months) | 709 | Low pg/mL range; Yonker ~34 pg/mL in myocarditis cases | Selem E, Raszek M, Varon J, Halma MTJ (2026)Opens in new tab |
| Blood (whole blood, PBMCs) | Vaccine mRNA | Up to ~6 months reported | 180 | Trace levels; qualitative detection in subsets | Fehrer C, et al. (2024)Opens in new tab |
| Breast milk | Spike protein / vaccine mRNA | Up to ~45 days post-vaccination | 45 | Trace; transient | Selem E, Raszek M, Varon J, Halma MTJ (2026)Opens in new tab |
| Urine | Spike protein | Days to weeks | 30 | Low/trace | Selem E, Raszek M, Varon J, Halma MTJ (2026)Opens in new tab |
| Saliva | Spike protein | Days to weeks | 21 | Low/trace | Selem E, Raszek M, Varon J, Halma MTJ (2026)Opens in new tab |
Table 3 — Tissues (autopsy / biopsy)
| Method | Notes | ||||
|---|---|---|---|---|---|
| Lymph nodes (germinal centers) | Spike protein / vaccine mRNA | Months (up to ~8+ months in some reports) | 240 | IHC, ISH, flow cytometry | Germinal centers may act as immune reservoirs; duration varies by vaccine and individual. |
| Myocardium / cardiac tissue | Spike protein | Weeks to months | 120 | IHC, Simoa (circulating) | Association with myocarditis in some pediatric cases; causality not established. |
| Brain / meninges / skull border | Spike protein | Up to ~4 years reported in some tissue studies | 1460 | IHC, autopsy | Findings at brain-meningeal borders are provocative but based on limited autopsy series; replication needed. |
| Liver, spleen | Spike protein / vaccine mRNA | Up to ~6 months | 180 | IHC, RT-qPCR | Reticuloendothelial system may sequester vaccine components. |
| Adipose tissue | Vaccine mRNA / spike | Weeks to months | 90 | RT-qPCR, IHC | Lipid nanoparticle biodistribution studies inform but do not fully explain human persistence. |