Our Methodology: How We Grade Cancer Prevention Evidence

How We Grade Evidence

Every compound, intervention, and test on this site receives an evidence grade based on the quality of human clinical evidence. Not animal studies. Not cell cultures. Not anecdotes. Human data.

The Grading System

Grade A: Strong Evidence

Phase II+ randomized controlled trial (RCT) with a positive signal in a clinically meaningful endpoint (survival, tumor response, validated biomarkers). The intervention has been tested in humans under controlled conditions and showed benefit.

Examples: High-Dose IV Vitamin C (Phase II RCT, doubled survival in pancreatic cancer), Exercise (massive epidemiological evidence with consistent dose-response).

Grade B: Promising

Phase I/II human data with a positive signal, OR strong, replicated epidemiological evidence with a plausible and confirmed mechanism. The intervention has been tested in humans and showed enough promise to warrant further investigation.

Examples: Mebendazole (Phase I glioblastoma + Phase 2a GI cancer + RCT colorectal), High-Dose Melatonin (meta-analysis of 8 RCTs), Vitamin D3 (VITAL trial mortality signal).

Grade C: Early / Limited

In vitro (cell) and/or in vivo (animal) data only, OR very early Phase I human data without efficacy signals. The mechanism is interesting but unproven in humans. "Works in a dish" is here.

Examples: Fenbendazole (preclinical only, zero human trials), Ivermectin (impressive preclinical, first Phase I/II underway), DCA (Phase I with mechanism confirmation but no efficacy RCT).

Grade D: Weak / Mixed

RCT data exists but showed null or negative results, OR previously promising observational data was contradicted by a rigorous trial. The hypothesis has been tested and did not hold up.

Examples: Metformin for Cancer (MA.32 trial: HR 1.01, no benefit).

Debunked

Retracted research, scientific fraud, actively harmful mechanisms, or clear evidence of danger. These are interventions people should actively avoid.

Examples: GcMAF (retracted papers, criminal fraud conviction), BPC-157 (pro-angiogenic mechanism promotes tumor growth).

Our Research Process

Identify the compound/intervention through literature review, clinical trial registries, or reader request
Search primary literature: PubMed, PMC, ClinicalTrials.gov, Cochrane Library, Google Scholar
Prioritize evidence hierarchy: Systematic reviews > RCTs > Phase I/II trials > epidemiological studies > case series > preclinical > in vitro
Assess the mechanism: Is it well-characterized? Plausible? Confirmed in human tissue?
Check for replication: Has more than one group found the same result?
Evaluate safety: What are the risks? Are they characterized in humans?
Assign evidence grade based on the criteria above
Write the review with citations to primary sources
Update when new evidence emerges (article dates reflect last review)

What We Don't Do

We don't extrapolate animal results to human recommendations
We don't treat anecdotes or case reports as evidence of efficacy
We don't let commercial relationships influence evidence grades
We don't ignore negative results or failed trials
We don't present preclinical data as though it proves a treatment works in humans

Why This Matters

Cancer patients are bombarded with information of wildly varying quality. Supplement companies claim their products fight cancer based on cell studies. Social media influencers promote veterinary drugs based on one person's story. Well-meaning friends share articles about compounds that were debunked years ago. Our grading system exists to cut through the noise. When we say Grade A, we mean it. When we say Debunked, we mean that too.