VAX-CAN: An Independent Study of COVID-19 Vaccination, Circulating Tumour DNA and Cancer Outcomes

By Editorial Team - October 19, 2026

WHITE PAPER | VERSION 1.2
Prepared: 19 October 2026
Prepared by: OneDayMD Research Team
Proposed programme: VAX-CAN Independent Research Programme
Status: Concept paper for scientific review, ethics consultation and funding discussions

A research proposal to investigate whether COVID-19 vaccination is associated with persistent cancer-related molecular signals, incident cancer diagnoses or objectively documented cancer progression.

Proposed approach: A prospective, independently funded research programme combining a randomized biomarker study where ethically justified, longitudinal cancer surveillance, validated molecular assays and long-term clinical outcome assessment.

Important statement of purpose

This proposal does not assume that COVID-19 vaccines cause cancer, accelerate cancer growth or produce a phenomenon commonly called “turbo cancer.” It proposes a transparent study to investigate these questions using prespecified methods, appropriate comparison groups and clinically meaningful outcomes.

The programme is a research concept. It has not been approved by a research ethics committee, funded, registered or initiated. Its design, sample size and procedures require review by qualified investigators and relevant authorities before implementation.

Abstract

Public controversy continues over whether COVID-19 vaccination might be associated with cancer development, recurrence or unusually rapid tumour progression. Case reports, retrospective database analyses and proposed biological mechanisms have attracted attention, but they do not by themselves establish a causal relationship. Meanwhile, the absence of conclusive evidence of causation does not eliminate the value of well-designed studies capable of testing specific safety hypotheses.

This white paper proposes VAX-CAN, an independent research programme designed to examine these questions through three complementary components: a prospective biomarker study, a longer-term cancer-outcomes cohort and a separate cancer-surveillance cohort. Randomization of vaccination timing may be considered for a limited biomarker component only when clinical equipoise exists and an ethics committee determines that the design does not impose unacceptable risk or withhold indicated care.

The proposed molecular assessment includes circulating tumour DNA (ctDNA), selected cell-free DNA measurements and, where analytically and ethically appropriate, multi-cancer early detection (MCED) assays. Biomarker results would be interpreted alongside clinical records, diagnostic investigations and independently adjudicated cancer outcomes. The study would account for pre-existing disease, SARS-CoV-2 infection, healthcare utilisation, cancer screening, vaccine characteristics and other important confounders.

The programme would be supported by non-pharmaceutical sponsors under a governance agreement that protects investigators' independence, prespecifies analyses, discloses conflicts of interest and guarantees publication regardless of results. Its central objective is to generate reproducible evidence that can distinguish transient molecular changes from persistent tumour-associated signals and clinically confirmed cancer outcomes.

Keywords: COVID-19 vaccine safety, cancer incidence, cancer progression, circulating tumour DNA, ctDNA, multi-cancer early detection, MCED, randomized trial, observational cohort, independent medical research.

Executive Summary

VAX-CAN is a proposed independent investigation into a disputed scientific question: whether COVID-19 vaccination is associated with measurable cancer-related molecular changes or subsequent cancer outcomes.

The proposal recognises two obligations. First, cancer-related claims should be evaluated rather than dismissed solely because they are controversial. Second, a study must not begin with the conclusion that a vaccine causes cancer or interpret any positive blood test as proof that it does.

Research question

Does vaccination produce a reproducible, persistent change in prespecified cancer-related biomarkers, and are any observed changes associated with confirmed clinical cancer outcomes?

Core technology

Validated ctDNA methods, appropriate cell-free DNA analysis, exploratory MCED testing and clinical cancer outcome verification.

Study structure

A randomized biomarker component only when ethical, a large prospective cohort and a separately analysed cancer-surveillance cohort.

Funding principle

Non-pharmaceutical sponsorship with independent governance, transparent financial disclosures and no sponsor veto over publication.

Five principles of the proposal

Measure first, interpret second. Define the endpoints and assay procedures before looking at the results.
Use real comparison groups. A before-and-after blood test alone cannot establish that vaccination caused a change.
Distinguish biomarkers from disease. A positive ctDNA or MCED test is not automatically a cancer diagnosis.
Protect participants. Do not withhold clinically indicated vaccination or cancer treatment simply to create a comparison group.
Publish every result. Positive, negative, inconclusive and unexpected findings should be reported transparently.

1. Background and Current Evidence

1.1 What is established, and what remains uncertain?

The US National Cancer Institute states that there is no evidence that COVID-19 vaccines cause cancer, cancer recurrence or disease progression. It also notes that vaccination can cause temporary enlargement of lymph nodes, especially in the armpit near the injection site, which may affect the interpretation of some imaging findings. See the NCI's COVID-19 vaccines and people with cancer guidance.^[1]

This is the appropriate starting point for a research proposal. The existing evidence does not establish that vaccination causes cancer, but a properly constructed research programme can examine defined questions about biomarkers, incidence and progression.

Source: COVID-19 mRNA-Induced "Turbo Cancers" - Dr Paul Marik and Dr Justus Hope

The crucial distinction is between a hypothesis that merits testing and a conclusion that has already been demonstrated. A temporal sequence—vaccination followed by a cancer diagnosis—does not alone show that vaccination caused the disease. The cancer may have existed before vaccination, may have been diagnosed because of additional healthcare contact, or may have developed independently.

1.2 Case reports and biological hypotheses

A 2026 scoping review by Kuperwasser and El-Deiry examined published reports of malignancies or malignant progression occurring in temporal association with COVID-19 vaccination or SARS-CoV-2 infection. The review identified 69 publications describing 333 individual patients, with many publications consisting of single-patient case reports or small case series.^[2]

These reports may help investigators identify hypotheses, possible patient characteristics or clinical patterns that warrant further examination. However, they cannot reliably estimate the incidence of cancer in exposed versus unexposed populations, and they cannot establish causation merely from the order in which events occurred.

Mechanistic questions involving inflammation, immune function, tumour surveillance and pre-existing malignant disease could be included as exploratory research questions. They should not be described as demonstrated pathways from vaccination to cancer unless supported by appropriate evidence.

1.3 The South Korean population study

A retrospective population-based study published in Biomarker Research on 26 September 2025 examined Korean National Health Insurance data involving approximately 8.4 million individuals. The authors reported statistical associations between vaccination and six cancer types during a one-year follow-up period: thyroid, gastric, colorectal, lung, breast and prostate cancer.^[3]

The paper is relevant to the rationale for further investigation, but its results should be interpreted carefully. It was observational, not randomized, and its reported associations do not by themselves establish that vaccination caused the cancers.

Important publication-status qualification

On 22 October 2025, the journal's publisher posted an editorial notice stating that concerns about this article had been raised with the editors and that editorial action would follow investigation. The notice is displayed on the original article page.^[3]

This notice is a reason to check the paper's publication status whenever citing it. It is not, by itself, a finding that the study is invalid or that its conclusions are correct. VAX-CAN should independently assess the quality, limitations and reproducibility of relevant observational findings rather than adopt their conclusions in advance.

A rigorous follow-up investigation would need to evaluate baseline differences between vaccinated and comparator populations, the timing and intensity of cancer screening, prior infections, healthcare utilisation, vaccine exposure classification, outcome verification and the possibility of residual confounding.

1.4 Defining “turbo cancer” scientifically

“Turbo cancer” is not a standardized clinical diagnosis or a validated research endpoint. Its use can combine several distinct events: a new diagnosis shortly after vaccination, a previously undetected malignancy, cancer recurrence, treatment-resistant disease and objectively rapid progression.

These events cannot be treated as interchangeable. VAX-CAN would replace the colloquial term with explicit, clinically meaningful definitions.

Claim or concern	Scientific endpoint	Evidence required
New cancer appearing after vaccination	Incident, confirmed malignancy	Diagnostic records, pathology where appropriate, dates and prior history
Return of a previous cancer	Confirmed recurrence	Original diagnosis, disease-free status, follow-up records and appropriate diagnostic evidence
Unusually rapid progression	Prespecified, cancer-specific progression endpoint	Objective imaging, pathology, validated molecular measures or established clinical progression criteria
Unexpected molecular signal	Persistent, assay-defined cancer-associated biomarker signal	Validated laboratory procedures, repeat confirmation and clinical adjudication where indicated

The study should not create a new disease category simply to match a popular term. It should investigate whether there is a reproducible signal in well-defined outcomes.

2. Research Gap and Rationale

Several important questions remain distinct and require different methods of investigation.

Short-term molecular changes: Does vaccination produce measurable changes in cell-free DNA or other selected biomarkers, and do they resolve?
Persistent cancer-associated signals: Does a newly detected signal persist on repeat testing, and can it be attributed to tumour-derived material?
Cancer incidence: Is vaccination associated with a difference in the rate of confirmed new malignancies after accounting for relevant confounders?
Cancer progression: Among participants with established cancer, is vaccination associated with measurable differences in recurrence or progression?
Clinical significance: Do biomarker findings predict outcomes that matter to patients, rather than only changes in a laboratory measurement?

A paired blood-draw study can address the first question, but not necessarily the others. The third and fourth questions require appropriately sized populations, validated clinical endpoints and sufficient follow-up.

VAX-CAN therefore proposes a research programme instead of a single test interpreted as a comprehensive answer to cancer risk.

3. Study Objectives and Hypotheses

3.1 Primary scientific objective

To estimate whether the defined vaccination exposure is associated with a difference in the prespecified primary cancer-related biomarker endpoint at a specified follow-up time, compared with an ethically appropriate contemporaneous control group.

3.2 Secondary objectives

Characterize the time course of selected molecular changes from the pre-vaccination baseline through early and later follow-up.
Determine whether new assay-positive results persist on repeat testing and whether they can be attributed to a plausible biological source.
Measure the frequency of confirmed new cancers during longer-term follow-up.
Assess cancer recurrence and progression in a separately defined population already undergoing appropriate cancer surveillance.
Evaluate whether findings differ by relevant prespecified factors, such as age, sex, vaccine platform, previous infection and prior cancer history.
Assess false-positive findings, false-negative findings, incidental diagnoses and the downstream consequences of additional diagnostic investigations.

3.3 Competing hypotheses

The protocol should explicitly allow for multiple possible results.

No detectable effect: No meaningful difference in the prespecified biomarker outcome is observed within the precision of the study.
Transient molecular change: A difference appears soon after vaccination but resolves and does not translate into a confirmed cancer outcome.
Persistent biomarker difference: A reproducible difference remains detectable, requiring further investigation of its source and clinical significance.
Clinical outcome difference: A difference in confirmed cancer incidence or progression is observed and requires assessment for causal interpretation.
Inconclusive findings: Assay limitations, insufficient statistical precision, attrition or conflicting outcomes prevent a reliable conclusion.

The purpose is to discriminate between these possibilities, not to presuppose which will occur.

4. Proposed Study Architecture

The programme should consist of a feasibility stage and three analytically distinct components. The randomized component should proceed only if an acceptable intervention and comparator can be justified ethically.

Stage 0: Feasibility and validation

Confirm assay performance, laboratory reproducibility, sample handling, recruitment feasibility, baseline signal rates and the practical response to unexpected findings.

Output: Final protocol parameters and estimates for power calculations.

Component A: Randomized biomarker study

Compare biomarker measurements under a randomized vaccination-timing design only where genuine clinical equipoise exists and delaying vaccination does not impose unacceptable risk.

Output: Comparative short-term biomarker evidence.

Component B: Prospective population cohort

Follow eligible participants with verified vaccination and infection histories to evaluate confirmed incident cancer outcomes over several years.

Output: Longitudinal estimates of cancer incidence and relevant confounding.

Component C: Cancer-surveillance cohort

Study recurrence and progression among cancer survivors or patients already receiving clinically indicated molecular surveillance.

Output: Separate estimates of molecular and clinical progression outcomes.

4.1 Why a hybrid design is preferable

Randomization can strengthen causal inference for outcomes measured during the randomized period. However, a short biomarker trial is unlikely to be adequately powered to assess rare cancers or long-term cancer mortality.

Conversely, a large observational cohort can follow participants for many years, but is vulnerable to differences between people who receive vaccination and those who do not.

Combining these complementary designs allows the research team to investigate early molecular effects while separately examining long-term clinical outcomes.

4.2 Ethical limits on randomization

The randomized component should not require participants to forgo an indicated vaccination merely to create an unvaccinated comparison group. Researchers must follow current local vaccination guidance, individual clinical circumstances and the judgement of the appropriate research ethics committee.

Where there is genuine clinical equipoise, a randomized comparison of vaccination timing may be considered. Both timing strategies must be clinically acceptable, and delayed participants must have access to vaccination when scheduled or earlier if medically indicated.

If such a comparison is not ethical or feasible, the programme should use a prospective observational design rather than force an RCT. The World Medical Association's Declaration of Helsinki provides the relevant international ethical framework for medical research involving human participants.^[4]

5. Study Populations

5.1 Cohort 1: Adults without known active cancer

The primary population would consist of eligible adults without a known active malignancy at enrolment. This cohort is intended to assess selected cancer-associated biomarkers and confirmed incident cancer outcomes.

Proposed inclusion criteria:

Adults aged 18 years or older who can provide informed consent.
A documented baseline assessment and ability to complete the planned sampling schedule.
A verifiable vaccination history and collection of relevant prior-infection information.
Access to suitable clinical follow-up, subject to participant consent and applicable privacy requirements.

Potential exclusions or separate strata:

Known active cancer at enrolment.
Recent cancer treatment or a condition likely to substantially affect biomarker interpretation.
Inability to obtain the required clinical history or follow-up.

These criteria require refinement. Prior cancers, hereditary cancer risk and recent diagnostic investigations should not simply be ignored or excluded automatically. The protocol must specify whether such individuals are ineligible, eligible for a separate stratum or included with an appropriate adjustment.

5.2 Cohort 2: Cancer survivors and patients under surveillance

This cohort must be analysed separately because the probability of detecting ctDNA and the likelihood of recurrence depend strongly on the cancer type, stage, treatment and time since treatment.

Participants should have a documented malignancy and an appropriate clinical follow-up pathway. Tumour-informed ctDNA testing should be used where the assay is validated for the specific disease and relevant clinical question. Existing treatment and surveillance decisions must remain under the treating clinical team.

The study must never use an exploratory biomarker result as the sole basis for stopping effective cancer treatment, delaying treatment or initiating an unproven intervention.

5.3 Baseline variables

At enrolment, investigators should record variables likely to influence cancer detection, disease risk or biomarker interpretation.

Variable group	Required information	Reason for collection
Demographics	Age, sex and relevant demographic variables	Adjustment for baseline cancer risk and subgroup analyses
Vaccination	Product, platform, dose number and administration dates	Define and verify exposure
Infection	Documented SARS-CoV-2 infection dates, testing and relevant severity measures	Separate infection-related associations from vaccination-related associations
Cancer risk	Smoking, relevant medical history, family history and established risk factors	Control confounding and describe generalizability
Screening	Prior tests, imaging, abnormal findings and dates of diagnostic work-up	Account for differences in cancer detection
Medication and treatment	Relevant immunomodulatory medication, cancer treatment and major changes in therapy	Interpret immune measures, cell-free DNA and cancer outcomes

6. Biomarker Strategy: ctDNA and MCED

6.1 Circulating tumour DNA

Circulating tumour DNA refers to a fraction of cell-free DNA released by tumour cells into the bloodstream. Depending on the assay and clinical setting, ctDNA can provide information about tumour-associated alterations, molecular residual disease or changes in a known malignancy.

Its value depends on the cancer type, disease burden, biological characteristics of the tumour, assay design and timing of blood collection. Small or early-stage cancers may shed too little tumour-derived DNA to be reliably detected.

Accordingly, a negative ctDNA result in an otherwise healthy participant cannot establish that cancer is absent. A detectable signal also requires careful interpretation: not every DNA alteration in plasma originates from a solid tumour.

Tumour-informed and tumour-naive methods

Approach	Possible role	Principal limitation
Tumour-informed ctDNA	Track known tumour-associated alterations in a participant with an established cancer diagnosis	Requires suitable tumour material or a validated patient-specific assay; sensitivity varies by disease and setting
Tumour-naive cfDNA analysis	Explore molecular signals without a known tumour profile	Signals may be nonspecific or originate from non-tumour tissues; a DNA alteration is not itself a cancer diagnosis
Serial quantitative analysis	Assess change over time using a validated method	Technical variability, biological variability and assay detection limits must be accounted for

6.2 Multi-cancer early detection

MCED assays are designed to detect signals potentially associated with multiple cancer types through blood-based analysis. Some methods assess DNA methylation patterns or other molecular characteristics to estimate whether a cancer-associated signal may be present and, in certain assays, its likely tissue of origin.

MCED testing could be considered an exploratory component of the cancer-free cohort. However, it should not be treated as an established universal screening solution.

A systematic review published through the UK National Institute for Health and Care Research concluded that the available evidence did not support strong conclusions about the effectiveness of MCED testing in general-population screening. It identified the need for adequately designed randomized trials with sufficiently long follow-up and outcomes that matter directly to patients.^[5]

The National Cancer Institute also explains that MCED tests have not established a reduction in cancer mortality in randomized trials and that the possible harms of screening, including false positives and downstream investigations, remain important considerations.^[6]

These limitations make independent evaluation valuable, but they also mean the study must not equate an MCED signal with a cancer diagnosis or assume that repeated MCED testing is harmless.

6.3 Clonal haematopoiesis and false-positive interpretation

A critical laboratory issue is clonal haematopoiesis. As people age, some blood-cell populations acquire somatic mutations and expand. Mutations from these cells can appear in plasma cell-free DNA and may be confused with tumour-derived DNA if an assay does not adequately distinguish their origins.

Research on clonal haematopoiesis in liquid biopsy describes this as an important interpretive challenge for molecular cancer testing.^[7]

The protocol should therefore consider matched white-blood-cell DNA sequencing, tumour-tissue concordance where available and prespecified filtering procedures appropriate to the assay. These controls are particularly important in older participants and cancer survivors previously exposed to cytotoxic therapies.

6.4 Proposed laboratory quality standards

All participating laboratories should follow standardized, documented procedures for:

Blood collection, tube type, processing time, storage and transport.
Sample labelling and blinded identification of study allocation.
DNA extraction, assay performance, quality-control thresholds and batch effects.
Repeat testing and confirmation of unexpected findings.
Documentation of assay detection limits, reproducibility and any changes to the testing platform.
Data retention, chain of custody, controlled access and auditability.

The central laboratory should report the result generated by the assay and its validated limitations. It should not label an unconfirmed signal as vaccine-induced cancer.

7. Proposed Sampling Schedule

The schedule below is a planning framework. The final protocol should select timepoints according to the assay's validated characteristics, pilot data, scientific rationale, cost and participant burden.

Timepoint	Core activities	Interpretive purpose
Baseline, before vaccination	Clinical assessment, exposure history and baseline blood collection	Establish the reference measurement before the defined exposure
Day 7–14	Prespecified early biomarker assessment	Explore acute molecular or inflammatory changes
Day 28–30	Repeat core blood measurements	Evaluate persistence or resolution of early changes
Day 90	Primary biomarker endpoint assessment	Measure the main prespecified short-term biomarker outcome
Month 6	Follow-up sampling where justified	Explore longer persistence and detect emerging changes
Month 12	Clinical follow-up and selected repeat assays	Relate molecular findings to one-year clinical outcomes
Years 2–5 and potentially longer	Medical-record follow-up and cancer-registry linkage	Investigate cancer incidence, recurrence and progression

MCED testing need not be performed at every blood draw. A limited, prespecified schedule may be more appropriate because repeated testing increases cost and the cumulative probability of obtaining false-positive results.

The observational and cancer-surveillance cohorts may need different sampling schedules. For patients with established cancer, assay timing should be compatible with the relevant disease, treatment and validated surveillance pathway.

8. Randomized Trial Design

8.1 The primary randomized question

If an ethically acceptable randomized vaccination-timing study can be conducted, its purpose should be narrowly defined: to test whether the selected vaccination exposure produces a difference in a prespecified biomarker endpoint over a defined short-term period.

It would not be designed to establish that vaccination does or does not cause rare cancers over many years. That question requires a substantially larger and longer investigation.

8.2 Proposed randomized allocation

Arm	Intervention	Assessment
Arm A	Vaccination at the usual planned time	Baseline and prespecified post-vaccination measurements
Arm B	Vaccination after a prespecified, clinically acceptable delay, only if ethically approved	The same measurements during the comparison period before the delayed dose

Both groups should receive the same sampling instructions, laboratory procedures, clinical follow-up and endpoint assessment. Laboratory personnel and outcome adjudicators should remain blinded to allocation wherever practicable.

The protocol must specify rules for unplanned vaccination, infections, treatment changes, missed samples and withdrawals. Participants should retain access to appropriate clinical care throughout.

8.3 If randomization is not permissible

The scientific value of the programme does not depend on forcing a randomized design. If withholding or delaying vaccination would create unacceptable risk, recruitment should proceed through an observational design with documented exposure histories, contemporaneous comparators and prespecified adjustment methods.

The study should not recruit participants by encouraging them to refuse medical care. Nor should it imply that a randomized biomarker result would settle the long-term cancer question.

9. Endpoints and Outcome Definitions

9.1 Primary endpoint for the randomized biomarker component

The proposed primary endpoint is:

Between-group difference in the proportion of participants with a new, assay-defined cancer-associated molecular signal at Day 90, among participants meeting the prespecified baseline-negative criteria, with confirmation according to a predefined laboratory procedure.

The precise assay, positivity definition, confirmation method and handling of missing or indeterminate results must be finalized before recruitment. The endpoint is a molecular research measure, not a diagnosis of cancer.

If pilot data show that this endpoint is too rare, insufficiently reproducible or poorly suited to a randomized comparison, the steering committee should select an alternative scientifically justified endpoint before the definitive trial starts. Such a change must be documented transparently.

9.2 Secondary biomarker endpoints

Change in selected quantitative molecular measurements from baseline.
Time to resolution or persistence of a newly detected signal.
Agreement between the initial assay and the repeat or confirmatory assay.
Proportion of positive results that lead to a confirmed diagnosis, remain unresolved or prove non-confirmatory.
Frequency and type of downstream diagnostic investigations initiated after a research finding.

9.3 Clinical endpoints in the prospective population cohort

The following should be evaluated separately from the randomized biomarker endpoint:

Incident, appropriately confirmed invasive cancer.
Cancer-specific incidence, using prespecified diagnostic categories.
Stage and other relevant characteristics at diagnosis.
Cancer-specific mortality and all-cause mortality, where the cohort and follow-up are sufficient.

9.4 Clinical endpoints in the cancer-surveillance cohort

Confirmed cancer recurrence.
Progression based on established cancer-specific clinical criteria.
Change in a validated, tumour-informed ctDNA measurement where appropriate.
Time to clinically confirmed progression, analyzed in relation to the original disease and treatment history.

No single progression definition should be applied indiscriminately to every cancer. A change in ctDNA is not necessarily equivalent to radiological progression, symptomatic deterioration or reduced survival.

10. Statistical Analysis Plan

10.1 Prespecification

A complete statistical analysis plan should be signed off before treatment allocation is revealed or the definitive analysis begins. It should specify the primary endpoint, analysis population, model, covariates, handling of missing data, subgroup analyses and decision criteria.

10.2 Randomized analysis

The primary randomized analysis should follow the intention-to-treat principle, comparing participants according to their assigned group. A baseline-adjusted analysis should be selected in advance, with the model suited to the endpoint and randomization procedure.

Per-protocol and exposure-adjusted analyses may provide supplementary information, but they must not replace the primary analysis simply because they produce a more favourable result.

10.3 Observational analysis

The population cohort should use a clearly documented target-trial framework where feasible, defining eligibility, exposure strategies, time zero, follow-up and outcome assessment before analysis. This reduces avoidable design problems but cannot eliminate all unmeasured confounding.

Depending on the available data, methods may include propensity-score weighting or matching, time-varying exposure models, multivariable regression and sensitivity analyses for unmeasured confounding. The method should be chosen for the data structure and causal question rather than selected after reviewing the outcome.

10.4 Important sources of bias

Potential bias	Why it matters	Proposed mitigation
Detection bias	One group may undergo more screening or diagnostic work-up	Record screening intensity, healthcare use and diagnostic pathways
Confounding by indication	Health status and clinical risk may influence vaccination decisions	Define eligibility and exposure strategies; adjust for measured differences
Prior SARS-CoV-2 infection	Infection may be associated with inflammatory changes and healthcare contact	Collect infection histories and conduct prespecified sensitivity analyses
Immortal-time or time-zero bias	Incorrect timing definitions can distort comparisons	Align eligibility, exposure classification and follow-up initiation
Multiple comparisons	Testing many cancer types, assays and timepoints raises the risk of chance findings	Prespecify primary endpoints and control multiplicity
Informative attrition	Participants who leave the study may differ in relevant ways	Track losses to follow-up and use justified missing-data methods

10.5 Subgroup analyses

Potential subgroup variables include age, sex, vaccine platform, dose number, prior SARS-CoV-2 infection, baseline cancer risk and established cancer history. These should be prioritized according to scientific plausibility and data availability.

Subgroup analyses should be interpreted cautiously, especially when sample sizes are small. A statistically significant result in one subgroup and a non-significant result in another does not, by itself, demonstrate that the groups differ. Appropriate interaction tests and multiplicity controls are required.

11. Sample Size and Statistical Power

A defensible sample size cannot be specified from the concept paper alone. It must be calculated after the feasibility stage establishes the expected event rate, assay performance, baseline prevalence of positive results, loss to follow-up and the smallest effect worth detecting.

The planning process should define:

The expected frequency of the primary endpoint in each group.
The minimum clinically or scientifically meaningful difference.
The required statistical power and significance threshold.
The planned multiplicity adjustment, if applicable.
The anticipated rate of missing or unusable samples.
The additional follow-up needed to accrue sufficient confirmed clinical events.

The feasibility study should be designed to estimate these parameters, not marketed as proof for or against a cancer-risk hypothesis.

A definitive incidence study may require a much larger population and several years of follow-up because cancer is heterogeneous and many individual malignancies are uncommon. A biomarker study and a clinical cancer-outcome study must therefore have separate power calculations.

The study should publish confidence intervals and absolute event rates alongside relative measures. A non-significant result is not proof of zero risk, particularly if the study is underpowered or follow-up is too short.

12. Independent Sponsorship and Funding Model

VAX-CAN is intended to be funded independently of pharmaceutical manufacturers. The objective is to build a scientifically credible programme whose methods, analysis and reporting remain independent of the commercial or ideological interests surrounding the controversy.

However, non-pharmaceutical sponsorship alone does not guarantee independence. A diagnostics company may benefit from greater demand for MCED testing. An insurer, technology company or organization involved in vaccine litigation may have other financial or reputational interests.

All sponsors must therefore disclose their interests and accept consistent scientific-governance requirements.

12.1 Potential sponsor categories

Independent foundations

Philanthropic foundations and public-interest research funds willing to support pre-specified methods and publication regardless of findings.

AI Companies

AI companies willing to contribute and train to development of original studies instead of citing already established third party studies as the output to public related questions.

Non-pharmaceutical companies

Technology, data infrastructure, laboratory service, insurance or other non-pharma organizations willing to contribute under an independent governance agreement.

Academic and research institutions

Universities, cancer research centres and epidemiology groups providing scientific leadership, clinical infrastructure and methodological oversight.

Public-interest research consortium

A pooled funding mechanism that diversifies support and reduces dependence on a single sponsor.

12.2 Proposed allocation of responsibilities

Function	Responsible party	Independence safeguard
Scientific leadership	Academic principal investigators and independent steering committee	Sponsors cannot dictate conclusions
Clinical recruitment and follow-up	Participating hospitals and research sites	Approved consent and documented procedures
Laboratory analysis	Qualified central laboratories	Blinded samples and documented quality control
Statistical analysis	Independent statistical core	Prespecified analysis plan and auditable code
Participant protection	Research ethics committees and independent safety oversight	Authority to request modifications or stop unsafe procedures
Publication	Investigators under a public publication policy	No sponsor veto over negative or unfavourable results

12.3 Required sponsor agreement

Before accepting funding, the study leadership should publish or make available a governance agreement covering:

Full disclosure of funding sources, financial relationships and relevant non-financial interests.
Investigator control over the protocol, endpoints, analysis and interpretation.
Protection from sponsor interference with the results or publication.
Registration and publication of prespecified methods before recruitment or primary analysis.
Secure access to the necessary data for independent verification.
Documentation of any sponsor-provided assay, software, equipment or laboratory service.
A publication policy covering positive, negative, inconclusive and unexpected findings.

Where a sponsor supplies an assay, the independent team should document the assay's limitations and the sponsor's role. An independent replication cohort should ideally use a different laboratory or independently controlled testing pathway when technically feasible.

13. Ethics, Participant Safety and Incidental Findings

The ethical management of unexpected results is central to the proposal. A molecular study may uncover a suspected cancer signal without having established how the participant should be investigated or what the result means clinically.

13.1 Informed consent

Participants should be told that the study is designed to answer an uncertain research question. They should understand that a blood test may produce false-positive or false-negative results, that a positive result may not represent cancer and that additional investigations may be required.

Consent should explain the study's purpose, blood collection procedures, potential risks, data use, privacy protections, withdrawal options, handling of incidental findings and availability of follow-up. Participation must be voluntary and should not require agreement with a particular view about vaccines.

13.2 Management of positive results

Before recruitment, the programme must establish a funded pathway for clinically significant incidental findings.

Laboratory review: Assess sample quality and apply prespecified repeat or confirmation procedures.
Clinical review: Have an appropriately qualified clinician review potentially important findings.
Diagnostic referral: Where indicated, arrange appropriate clinical evaluation using established diagnostic pathways.
Follow-up: Record whether a diagnosis was confirmed, excluded, remained unresolved or required continued monitoring.
Safety reporting: Report relevant adverse events and unexpected problems through the approved research-governance system.

The protocol must not create an unmanageable stream of incidental results that participants cannot afford to investigate. Ethical and financial arrangements for follow-up should be approved before the first participant is enrolled.

13.3 Avoid inappropriate clinical action

MCED and experimental molecular results should not independently trigger unvalidated cancer treatments, unnecessary surgery or discontinuation of standard care. Decisions should be based on appropriate clinical evaluation and established evidence.

13.4 Data protection and biological samples

The study should use coded identifiers, restricted access and secure data systems. Sample storage, genetic data processing, future research use and cross-border transfers must comply with applicable privacy laws and consent requirements.

For international studies, participating sites must obtain the necessary institutional and regulatory approvals. Malaysian sites, if included, would need review under the applicable Malaysian research-ethics and personal-data requirements in addition to the relevant institutional procedures.

14. Transparency and Open Science

Independent funding will only be meaningful if the research process is independently verifiable.

Before recruitment, the study should:

Register the applicable trial or study in a recognized public registry.
Publish the protocol and statistical analysis plan.
Define the primary endpoints, assays and confirmation rules.
Publish the sponsor and investigator conflict-of-interest disclosures.
Establish an independent data-monitoring and safety procedure where appropriate.

The SPIRIT–CONSORT guidance provides standards for reporting randomized trial protocols and results, while the STROBE Statement supports transparent reporting of observational studies.^[8]^[9]

The programme should release a clear summary of its results even if publication in a journal is delayed. Where participant privacy and consent permit, de-identified data, analysis code and relevant methodological documentation should be made available under a documented access policy.

15. Proposed Implementation Roadmap

The following is a planning framework, not a prediction of the time required for approvals, recruitment or publication. Actual milestones depend on funding, clinical-site capacity, assay performance and ethics review.

Phase	Major activities	Decision gate
Phase 1: Protocol development	Assemble investigators, refine the research question, evaluate assays, consult patients and define governance	Independent scientific review of the concept
Phase 2: Feasibility and ethics	Evaluate assay performance, sampling logistics, recruitment, incidental-finding pathways and statistical assumptions	Ethics and scientific approval before enrolment
Phase 3: Pilot study	Test sample processing, repeatability, retention and laboratory procedures	Confirm feasibility and final sample-size assumptions
Phase 4: Definitive enrolment	Recruit approved cohorts and conduct blinded testing according to the protocol	Ongoing independent safety and quality monitoring
Phase 5: Primary analysis	Analyse the prespecified biomarker endpoint and publish the primary results	Independent statistical review
Phase 6: Long-term follow-up	Maintain clinical follow-up and registry linkage, subject to consent and approvals	Report cancer outcomes as sufficient data accrue
Phase 7: Replication	Replicate important findings in a separate cohort using independent analytical oversight	Assess reproducibility and generalizability

16. Expected Deliverables

The proposed programme should be funded around defined scientific outputs rather than a promised conclusion.

Published protocol: A complete document describing the design, endpoints, eligibility, assays and safeguards.
Feasibility report: Assay reproducibility, baseline signal rates, sample-handling performance and recruitment feasibility.
Primary biomarker report: The randomized or appropriately controlled comparison, including confidence intervals and all prespecified outcomes.
Clinical outcomes report: Confirmed incidence, recurrence and progression findings as adequate follow-up becomes available.
Methods and data documentation: Analysis code, relevant technical information and a data-sharing framework.
Public summary: A plain-language account of what the study found, what it did not find and the limitations that remain.

If findings are suggestive but inconclusive, the report should explain what additional sample size, follow-up or independent replication would be required. Unexpected findings should not be promoted as established causal effects before confirmation.

17. Limitations and Risks of the Proposed Programme

17.1 A biomarker is not a cancer diagnosis

Neither ctDNA nor MCED should be assumed to detect all cancers or reliably predict which molecular signals will become clinically consequential. A change in a blood-based result may reflect biological variation, technical variation, a non-tumour source or previously undetected disease.

17.2 Before-and-after comparisons are insufficient

A study that collects blood only before and after vaccination may detect a change but cannot reliably attribute it to vaccination. The change might also have occurred with time, infection, medication, another exposure or normal biological variability. A suitable control group and standardized sampling are essential.

17.3 Cancer has a long and heterogeneous natural history

Cancer may arise from accumulated molecular changes over many years. A negative biomarker result at Day 90 cannot rule out an effect on cancer incidence years later. Equally, a cancer diagnosed within weeks of vaccination may have been present but undetected beforehand.

17.4 Randomized studies may be ethically constrained

A randomized comparison of vaccination timing may not be permissible in many populations. The research programme must accept that limitation and select the strongest ethical alternative rather than withholding indicated care.

17.5 Observational studies remain vulnerable to confounding

Even large database studies can produce misleading associations when comparison groups differ in healthcare use, baseline risk, screening, vaccination eligibility, infections or other relevant factors. Statistical adjustment cannot guarantee that every source of bias has been removed.

17.6 Costs and downstream investigation

MCED testing, serial sequencing and prolonged follow-up may be expensive. False-positive findings can generate anxiety and additional testing. The study's budget must account for these consequences, not merely the laboratory price of the initial blood test.

17.7 No single study can settle every cancer-safety question

An adequately powered trial may address a limited biomarker question, while a longer-term cohort assesses broader clinical outcomes. A final assessment may require multiple studies, independent replication and different methods across cancer types.

18. Interpretation Framework

To limit exaggerated headlines, the study should publish its interpretation rules before the results become known.

Possible finding	Appropriate interpretation	Inappropriate conclusion
No meaningful biomarker difference detected	No difference was detected within the study's precision and limitations	All possible cancer risks have been ruled out
Temporary biomarker change	A time-limited molecular difference occurred; its biological significance requires interpretation	The vaccine caused new cancer
Persistent molecular signal	A reproducible signal warrants assessment of its source and clinical implications	The participant definitely has cancer or the vaccine caused it
Higher cancer incidence in an observational cohort	An association has been observed and requires causal assessment	Vaccination has been proved to cause cancer
Difference in the randomized biomarker endpoint	The randomized comparison supports a causal interpretation for the defined endpoint, within the trial's scope	The result proves increased long-term cancer incidence or mortality
Positive clinical outcome finding replicated independently	The total evidence warrants an updated causal assessment and further clinical interpretation	Every cancer type, vaccine platform and population has the same risk

The strength of each conclusion should depend on the endpoint, study design, statistical precision, consistency of evidence and quality of replication.

19. Funding Appeal

Invitation to independent research sponsors

OneDayMD proposes the development of VAX-CAN as an independent scientific programme supported by a consortium of non-pharmaceutical funders and academic partners.

The immediate funding priority should be the protocol-development and feasibility stage. This stage would determine whether the proposed assays can answer the defined research questions, establish realistic sample-size assumptions, identify participating laboratories and clinical centres, and develop an ethics-approved plan for unexpected results.

The programme should not solicit support on the basis that vaccine-related cancer has already been proved. It should solicit support because an important public controversy requires carefully designed, transparent and reproducible evidence.

Proposed funding commitments

Support independent scientific and statistical planning.
Fund feasibility testing, sample collection and laboratory validation.
Provide appropriate resources for participant follow-up and investigation of clinically important incidental findings.
Protect investigator independence and the publication of all results.
Enable longer-term clinical outcome analysis and independent replication.

The budget should be developed only after the sample-size calculations, laboratory quotations, site requirements, data-linkage costs and follow-up obligations have been defined. A single arbitrary funding figure would be premature at this stage.

20. Conclusion

COVID-19 vaccination and cancer is a scientifically sensitive subject where neither anecdotal reports nor a single observational study can settle the question of causation. A credible investigation needs to go beyond temporal associations and establish what was present before vaccination, what changed afterward, whether that change persists, and whether it corresponds to clinically confirmed cancer.

VAX-CAN proposes a staged, independently funded programme combining molecular measurement, ethically appropriate comparative methods and longitudinal clinical outcomes. It recognizes that short-term biomarker research and long-term cancer-risk assessment require separate endpoints, sample-size calculations and analytical strategies.

The proposed non-pharmaceutical sponsorship model would support independence only if accompanied by transparent funding disclosures, independent scientific governance, blinded laboratory analysis, prespecified methods, robust participant protections and a guarantee that results will be published regardless of direction.

The central proposition

The purpose of VAX-CAN is not to prove a predetermined claim. It is to design a study capable of producing credible evidence about a disputed question and to make that evidence available for independent scrutiny.

Prepared by: OneDayMD Research Team
Document type: Independent research concept white paper
Version: 1.2
Date: 19 October 2026

References and Further Reading

The following sources provide background on vaccine and cancer guidance, published reports, limitations of MCED screening, molecular assay interpretation, research ethics and reporting standards. They are not presented as proof that vaccination causes cancer.

[1] National Cancer Institute. COVID-19 Vaccines and People with Cancer. NCI guidance on COVID-19 vaccination, cancer treatment and questions about recurrence or progression.
https://www.cancer.gov/about-cancer/coronavirus/covid-19-vaccines-people-with-cancer

[2] Kuperwasser C, El-Deiry WS. COVID vaccination and post-infection cancer signals: Evaluating patterns and potential biological mechanisms. 2026. Scoping review of published reports involving malignancies or progression temporally associated with COVID-19 vaccination or infection. The reports do not, by themselves, establish causality or estimate population cancer risk.
https://pmc.ncbi.nlm.nih.gov/articles/PMC12893478/

[3] Kim HJ, Kim MH, Choi MG, et al. 1-year risks of cancers associated with COVID-19 vaccination: a large population-based cohort study in South Korea. Biomarker Research. 2025;13:114. doi:10.1186/s40364-025-00831-w. The journal's article page includes an editorial notice dated 22 October 2025 stating that concerns were raised with the editors and were under investigation.
https://doi.org/10.1186/s40364-025-00831-w

[4] World Medical Association. Declaration of Helsinki: Ethical Principles for Medical Research Involving Human Participants. Revised 2024. International ethical framework addressing research design, participant welfare, informed consent and the use of comparator interventions.
https://www.wma.net/policies-post/wma-declaration-of-helsinki/

[5] Wade R, Nevitt S, Liu Y, et al. Multi-cancer early detection tests for general population screening: a systematic literature review. Health Technology Assessment. 2025;29.02. National Institute for Health and Care Research. See the conclusions on evidence limitations and the need for adequately designed trials with patient-relevant outcomes.
https://www.ncbi.nlm.nih.gov/books/NBK611735/

[6] National Cancer Institute. Cancer Screening Overview (PDQ®)–Health Professional Version. NCI's overview of cancer screening, including the limitations, potential harms and evidence requirements for multi-cancer detection tests.
https://www.cancer.gov/about-cancer/screening/hp-screening-overview-pdq

[7] Clonal Hematopoiesis in Liquid Biopsy: From Biological Noise to Valuable Clinical Implications. Review of the implications of clonal haematopoiesis for circulating cell-free DNA analysis and molecular cancer testing.
https://pmc.ncbi.nlm.nih.gov/articles/PMC7463455/

[8] Chan AW, Boutron I, Hopewell S, et al. SPIRIT 2025 statement: updated guideline for protocols of randomised trials. BMJ. 2025;389:e081477. doi:10.1136/bmj-2024-081477.
https://www.consort-spirit.org/

[9] von Elm E, Altman DG, Egger M, et al. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) Statement: Guidelines for Reporting Observational Studies. Epidemiology. 2007;18(6):800–804. doi:10.1097/EDE.0b013e3181577654.
https://www.strobe-statement.org/

Research and Medical Disclaimer

This white paper describes a proposed research programme and is not a clinical trial protocol, clinical recommendation, medical diagnosis or proof of a causal relationship between vaccination and cancer. The programme is not approved, funded, registered or underway unless a future announcement explicitly establishes that status.

The scientific literature and clinical guidance may change. Any future study must use current evidence and applicable requirements at the time of approval and recruitment. COVID-19 vaccination decisions should be made in consultation with a qualified healthcare professional, taking individual circumstances and current local guidance into account.

Cancer screening and surveillance should follow appropriate clinical guidance. Research-only ctDNA or MCED findings should not replace established diagnostic procedures or lead to unvalidated treatment decisions.