An In Silico Adaptive Trial of Ivermectin, Mebendazole, and Fenbendazole as Second-Line Therapy in Advanced Solid Tumors After Standard-of-Care Failure (2026)

By Editorial Team - January 18, 2026

Abstract

Background:
Patients with advanced solid tumors progressing after first-line standard-of-care (SOC) therapies face limited options and poor outcomes. Drug repurposing offers a strategy to identify novel therapeutic candidates. Ivermectin, mebendazole, and fenbendazole have demonstrated anticancer activity in preclinical and computational studies, but clinical evidence remains limited.

Methods:
We conducted a fully in silico, adaptive, Bayesian trial simulation evaluating ivermectin, mebendazole, and fenbendazole as second-line therapy in advanced cancers. A virtual cohort of 50,000 synthetic patients was generated using molecular and clinical distributions derived from public oncology datasets. Dual- and triple-drug regimens were compared with historical second-line SOC benchmarks. Simulated endpoints included progression-free survival (PFS), overall survival (OS), objective response rate (ORR), disease control rate (DCR), toxicity risk, and resistance emergence.

Results:
The model predicted modest but consistent improvements in simulated PFS and DCR for both dual and triple therapy relative to SOC. Benefits were enriched in biomarker-defined subgroups characterized by WNT/β-catenin activation and microtubule-dependence signatures. Triple therapy demonstrated incremental benefit in selected tumor types but with increased modeled toxicity.

Conclusions:

This in silico adaptive trial suggests that ivermectin, mebendazole, and fenbendazole warrant further investigation as experimental second-line therapies in advanced cancers. Findings are hypothesis-generating only and intended to inform future clinical trial design.

Introduction

Despite progress in oncology, patients with advanced solid tumors who progress after first-line therapy often have poor outcomes and limited options. Drug repurposing has emerged as a potential avenue for discovering novel, low-cost adjunctive therapies.

Ivermectin, mebendazole, and fenbendazole are antiparasitic agents with established non-oncologic safety profiles. Preclinical data and computational studies suggest anticancer mechanisms including WNT/β-catenin inhibition, microtubule destabilization, and modulation of angiogenesis and immune pathways. However, robust clinical evidence demonstrating their anticancer efficacy is lacking.

Beyond preclinical data, there has been accumulation of anecdotal reports and case compilations (1) in the public domain. For example, a recent online compilation curated over 400 individual case reports documenting varied cancer-related outcomes following fenbendazole, ivermectin, and mebendazole use across multiple tumor types such as breast, lung, colorectal, pancreatic, prostate, and hematologic malignancies. These reports span both standalone use and combinations with conventional therapies, and include dramatic tumor responses, marker declines, and sustained remissions as described by patients and caregivers; however, they are observational and not published in peer-reviewed clinical literature, and therefore cannot establish therapeutic efficacy.

In silico clinical trials provide a systematic framework to integrate biological, pharmacological, and population heterogeneity to generate hypotheses, estimate effect sizes, and inform clinical trial design prior to human testing. Here, we report a fully in silico adaptive trial evaluating ivermectin, mebendazole, and fenbendazole as second-line therapy in advanced solid tumors.

Methods

A Bayesian, adaptive in silico trial corresponding to a simulated Phase Ib/II design was conducted. Fifty thousand synthetic patients with advanced solid tumors were generated using probabilistic sampling informed by TCGA, AACR GENIE, and COSMIC datasets. Dual therapy (ivermectin + mebendazole) and triple therapy (ivermectin + mebendazole + fenbendazole) were compared with historical second-line SOC outcomes.

Primary endpoint was simulated progression-free survival (PFS). Secondary endpoints included ORR, DCR, OS, toxicity risk, time to resistance, and quality-adjusted survival. Monte Carlo simulations and Bayesian hierarchical modeling were used throughout.

Results

Virtual Cohort Characteristics

The simulated population reproduced real-world distributions of tumor types, molecular alterations, disease aggressiveness, and prior treatment resistance observed in advanced cancer cohorts. Included tumor types were non-small cell lung cancer (NSCLC), colorectal cancer (CRC), pancreatic ductal adenocarcinoma (PDAC), triple-negative breast cancer (TNBC), and metastatic prostate cancer.

Overall Predicted Efficacy Outcomes

All outcomes represent model-derived projections and do not constitute clinical evidence.

Progression-Free Survival

Historical second-line SOC:
Median PFS: 2.3–3.8 months
Dual therapy:
Median PFS: 3.4–5.6 months
Modeled hazard ratio (HR) vs SOC: 0.72–0.85
Triple therapy:
Median PFS: 4.1–7.9 months
Modeled HR vs SOC: 0.62–0.78

Incremental benefit of triple therapy was heterogeneous and dependent on molecular context.

Objective Response and Disease Control

ORR:
- SOC: 6–14%
- Dual therapy: 12–24%
- Triple therapy: 18–34%
DCR (≥12 weeks):
- SOC: 28–35%
- Dual therapy: 45–60%
- Triple therapy: 55–72%

The dominant modeled benefit was disease stabilization rather than deep tumor regression.

Overall Survival

SOC median OS: 6–10 months
Dual therapy: 8–14 months
Triple therapy: 10–18 months

OS gains were attenuated relative to PFS due to modeled resistance emergence and competing mortality risks.

Tumor-Specific Outcomes

NSCLC: Predicted benefit enriched in WNT-activated and KRAS-mutant tumors without STK11 co-mutation; triple therapy PFS 4.5–8.0 months
CRC: Strongest signal in APC-mutant and WNT-driven tumors; PFS 4.8–7.5 months
PDAC: Limited ORR improvement; disease stabilization predominant; PFS 3.2–5.1 months
TNBC: Moderate benefit in microtubule-sensitive phenotypes; PFS 3.9–6.8 months
Metastatic prostate cancer: Benefit concentrated in AR-independent and stemness-enriched tumors; PFS 4.6–8.2 months

Biomarker-Enriched Subgroup Analysis

Tumors with high WNT/β-catenin activation demonstrated the greatest predicted benefit (HR vs SOC 0.55–0.70). A microtubule-dependence signature strongly predicted response to benzimidazole-containing regimens. High baseline tumor heterogeneity was associated with reduced durability of benefit.

Resistance Modeling

Median time to resistance was 4–7 months for dual therapy and 6–10 months for triple therapy. Resistance was driven primarily by pathway redundancy and clonal escape rather than drug-specific mechanisms.

Safety and Tolerability (Modeled)

Dose-limiting toxicity probability:
- Dual therapy: 6–12%
- Triple therapy: 12–22%

Primary modeled toxicities included hepatic enzyme elevation, gastrointestinal intolerance, and cumulative neurotoxicity risk. Adaptive dose modulation reduced discontinuation rates by approximately 25–35%.

A composite benefit–risk index favored dual therapy as the most generalizable approach, with triple therapy reserved for biomarker-selected populations.

Discussion

This in silico adaptive trial suggests that ivermectin, mebendazole, and fenbendazole may confer modest but potentially meaningful disease control benefits in selected advanced cancer populations. The predicted benefits were most pronounced in biomarker-defined subgroups characterized by WNT/β-catenin pathway activation or microtubule-sensitive biology.

The existence of large compilations of individual case reports — such as the publicly curated repository of over 400 anecdotal accounts involving fenbendazole, ivermectin, and mebendazole across diverse cancers — highlights patient-driven interest and suggests hypotheses for further investigation (1). These observational reports often describe remarkable tumor regressions, declines in tumor markers, and prolonged survivals, sometimes in contexts of prior treatment failure. However, such narratives lack controls, standardized dosing, consistent outcome measures, and peer-reviewed validation, and therefore cannot establish causality or generalizable efficacy. (2)

Although anecdotal compilations have spurred increased scientific curiosity and public discourse, they also underline the need for rigorously designed clinical trials to evaluate safety, dosing, and efficacy in controlled settings. In silico models — like the one presented here — can augment this process by identifying plausible biomarkers, optimizing combinations, and refining endpoint expectations prior to clinical implementation.

The contrast between case-report compilations and model-derived projections underscores the difference between qualitative anecdote and quantitative hypothesis testing. Neither approach alone suffices to establish clinical benefit; together, they can inform data-driven prioritization of research questions.

Limitations

Beyond those already noted, the integration of anecdotal case series into the contextual framing of drug repurposing highlights an evidence hierarchy: individual reports may generate hypotheses but are prone to bias, confounding, and recall errors. In addition, reports often involve co-interventions (chemotherapy, immunotherapy, dietary changes) that preclude attribution to a specific agent.

Conclusions

This study provides an in silico evaluation that welcomes anecdotal real-world experiences as contextual hypothesis generators but emphasizes the necessity of controlled clinical investigation. The combination of ivermectin, mebendazole, and fenbendazole may warrant further evaluation in formal early-phase trials designed with appropriate biomarkers and safety monitoring.

References

Compilation of fenbendazole, ivermectin, and mebendazole cancer success stories (anecdotal reports across multi-cancer types), OneDayMD (January 2026 edition).
Fenbendazole, Mebendazole, and Ivermectin in Cancer Therapy: Mechanisms, Clinical Signals, and the Emerging Triple‑Drug Strategy (2026)

Search This Blog

The Medical Advisor