Medbot
Archive

2026-08-19 · cycle 0003 · UNREVIEWED

skin

Queued by a human. Must resolve against Open Targets before it is evidence.

Wipeout. Every candidate for skin died. That is a real result.

Check this yourself

  • Disease id opens Open Targets, not our copy of it.
  • Every trial is an NCT on ClinicalTrials.gov. Every paper is a PMID on PubMed.
  • Challenger and critic saw the same evidence pack. Models are named below.
  • Stamp is UNREVIEWED. No clinician signed this.

How this was run

Disease id
QUERY_skin on Open Targets
Review
UNREVIEWED. No oncologist signed this.
Challenger
gpt-5-6-luna
Critic
gpt-5-6-luna
Evidence fetched
2026-08-19 16:51:02 UTC

Evidence pack both models saw

Same pack for challenger and critic. Links go to the public records, not to us.

Targets from Open Targets

  • None returned

Known drugs

  • None returned

Trials on ClinicalTrials.gov

  • None returned

Papers on PubMed

  • PMID 42127392 2026 · N Engl J Med
    Inflammatory Myopathies.
  • PMID 40562027 2025 · Cell
    An iPSC-derived CD19/BCMA CAR-NK therapy in a patient with systemic sclerosis.
  • PMID 39447204 2025 · Dermatol Surg
    Exosomes for Treating Hair Loss: A Review of Clinical Studies.
  • PMID 37053311 2023 · Science
    Engineered skin bacteria induce antitumor T cell responses against melanoma.
  • PMID 38193084 2023 · Front Immunol
    A review of skin immune processes in acne.
  • PMID 38338862 2024 · Int J Mol Sci
    PRAME Updated: Diagnostic, Prognostic, and Therapeutic Role in Skin Cancer.
Rounds
Entered01020304
  • Pembrolizumab
    killed · falsification
  • Nivolumab plus relatlimab
    killed · falsification
  • Dabrafenib plus trametinib
    killed · falsification
  • Cemiplimab
    killed · falsification
  • Vismodegib
    killed · falsification
  • Lifileucel
    killed · falsification
  • Tebentafusp
    killed · triage
  • Ipilimumab
    killed · mechanism

0 standing · 8 crossed out

Standing · 0

Wipeout. Nothing survived.

Killed · 8

Pembrolizumab

killed · falsification

anti-PD-1 monoclonal antibody

KNOWNINFERRED
Thesis
[KNOWN] Use established PD-1 blockade as the conservative benchmark for advanced melanoma and selected advanced cutaneous squamous cell carcinoma. [INFERRED] The next testable increment is identifying biomarker-defined nonresponders rather than treating broad skin cancer as one disease.
Mechanism
[KNOWN] Pembrolizumab blocks PD-1 signaling and can restore antitumor T-cell activity. [INFERRED] Response heterogeneity may reflect tumor antigen presentation, T-cell exclusion, or an immunosuppressive microenvironment.
Kill experiment
In a prospectively defined melanoma or cutaneous squamous cell carcinoma cohort, demonstrate that PD-L1 status, antigen-presentation markers, T-cell exclusion, and immune-suppressive features do not predict response or resistance beyond clinical covariates, or show no activity in the specified treatment-naive indication.
Why it died
This is an established PD-1 therapy for selected melanoma and cutaneous squamous cell carcinoma, not an unreviewed hypothesis. The pack contains no Open Targets result, known-drug record, trial record, or evidence supporting extrapolation to undifferentiated "skin" disease. Its proposed biomarker increment is generic inference, not a defined testable target.

Nivolumab plus relatlimab

killed · falsification

anti-PD-1 plus anti-LAG-3 monoclonal antibodies

KNOWNINFERRED
Thesis
[KNOWN] This is an established immunotherapy option in advanced melanoma. [INFERRED] It is a rational comparator for tumors with inadequate response or rapid progression on PD-1 pathway therapy, but the benefit-risk tradeoff should be tested by subtype and prior exposure.
Mechanism
[KNOWN] Nivolumab inhibits PD-1, while relatlimab inhibits LAG-3. [INFERRED] Dual checkpoint blockade may improve T-cell function where both inhibitory pathways are active.
Kill experiment
In a biomarker-selected melanoma study enriched for concurrent PD-1 and LAG-3 activity, show that adding relatlimab produces no clinically meaningful improvement over nivolumab alone in response durability or survival, with acceptable assay quality and exposure.
Why it died
This is an established combination for advanced melanoma, not an unreviewed hypothesis for skin disease. The pack provides no trial or target evidence for use outside that histology. The claim that dual PD-1/LAG-3 blockade benefits PD-1-refractory tumors is an inference without a defined population, endpoint, or causal evidence, and the candidate offers no basis for treating generic skin disease.

Dabrafenib plus trametinib

killed · falsification

BRAF inhibitor plus MEK inhibitor

KNOWNINFERRED
Thesis
[KNOWN] This combination is a standard targeted option for advanced melanoma harboring an activating BRAF V600 alteration. [INFERRED] It is a strong molecularly selected branch, not a general treatment for unselected skin cancer.
Mechanism
[KNOWN] Dabrafenib inhibits mutant BRAF and trametinib inhibits MEK1/2 downstream in the MAPK pathway. [INFERRED] Combined pathway suppression can reduce adaptive MAPK reactivation compared with single-agent BRAF inhibition.
Kill experiment
In independently confirmed BRAF V600-positive advanced melanoma, demonstrate in a controlled comparison that the combination does not improve progression control or duration of response over an appropriate targeted comparator, or test a non-BRAF-V600 population and show no genotype-specific activity.
Why it died
This is a standard targeted regimen for BRAF V600-mutant advanced melanoma, not an unreviewed hypothesis. Its relevance is limited to a molecularly defined melanoma subgroup and is unsupported for undifferentiated skin disease; the pack has no target or trial evidence. The adaptive-MAPK-resistance rationale is stated as inference and does not establish a new candidate.

Cemiplimab

killed · falsification

anti-PD-1 monoclonal antibody

KNOWNINFERRED
Thesis
[KNOWN] Cemiplimab is an established systemic treatment for advanced cutaneous squamous cell carcinoma when local therapy is unsuitable or insufficient. [INFERRED] It provides a histology-specific benchmark distinct from melanoma immunotherapy datasets.
Mechanism
[KNOWN] Cemiplimab blocks PD-1 signaling to enhance antitumor T-cell responses. [INFERRED] Lack of response may be investigated through tumor antigen presentation and immune-cell exclusion measurements.
Kill experiment
In advanced cutaneous squamous cell carcinoma meeting the stated systemic-treatment criteria, show prospectively that cemiplimab has no reproducible objective response or durable disease control, or that proposed antigen-presentation and immune-exclusion features fail to distinguish responders from nonresponders.
Why it died
Cemiplimab is an established systemic treatment for advanced cutaneous squamous cell carcinoma, not an unreviewed hypothesis. It cannot represent a general skin-disease solution, and the pack contains no supporting Open Targets, drug, trial, or disease-specific evidence. The proposed antigen-presentation and immune-exclusion explanation for nonresponse is generic inference rather than a validated hypothesis.

Vismodegib

killed · falsification

Hedgehog-pathway inhibitor

KNOWNINFERRED
Thesis
[KNOWN] Vismodegib is an established option for advanced or locally advanced basal cell carcinoma in appropriate settings. [INFERRED] It is a plausible conservative candidate only for Hedgehog-dependent basal cell carcinoma, not for melanoma or cutaneous squamous cell carcinoma.
Mechanism
[KNOWN] Vismodegib inhibits Smoothened, suppressing aberrant Hedgehog pathway signaling driven commonly by PTCH1 or SMO alterations in basal cell carcinoma. [INFERRED] Resistance could arise through downstream pathway reactivation.
Kill experiment
In molecularly confirmed Hedgehog-pathway-dependent advanced basal cell carcinoma, show no meaningful response or progression control at clinically tolerated exposure, or show that PTCH1/SMO pathway activation does not enrich for benefit compared with pathway-negative tumors.
Why it died
Vismodegib is an established Hedgehog inhibitor for appropriately selected advanced or locally advanced basal cell carcinoma, not an unreviewed candidate. Its mechanism and disease restriction are already known, while the pack supplies no Open Targets, drug, trial, or new evidence. It is irrelevant to melanoma, cutaneous squamous cell carcinoma, and undifferentiated "skin" disease; resistance by downstream reactivation is only an inference.

Lifileucel

killed · falsification

autologous tumor-infiltrating lymphocyte therapy

KNOWNINFERRED
Thesis
[KNOWN] Lifileucel is an established cellular therapy option for selected advanced melanoma after prior systemic treatment. [INFERRED] A testable increment is determining whether pretreatment features of T-cell infiltration or tumor antigenicity predict durable benefit.
Mechanism
[KNOWN] Lifileucel expands and reinfuses autologous tumor-infiltrating lymphocytes. [INFERRED] Its activity depends on recovering tumor-reactive T-cell populations capable of recognizing patient-specific tumor antigens.
Kill experiment
In consecutively treated eligible melanoma patients, prospectively measure baseline T-cell infiltration and tumor antigenicity and show that these features do not predict product generation, objective response, or durable benefit, or demonstrate that manufacturing failure or post-infusion persistence makes the proposed biomarker-selected strategy clinically infeasible.
Why it died
Lifileucel is an established autologous TIL therapy for selected previously treated advanced melanoma, not an unreviewed hypothesis. The pack contains no trial or target evidence and does not justify extending it to generic skin disease. The proposed predictors, T-cell infiltration and tumor antigenicity, are broad inferred correlates without a specified assay or validated causal link.

Tebentafusp

killed · triage

gp100-directed ImmTAC T-cell engager

KNOWNINFERRED
Thesis
[KNOWN] Tebentafusp has clinical activity in HLA-A*02:01-positive uveal melanoma. [INFERRED] It is a wrong-but-plausible decoy for a broad skin-cancer query because melanoma biology is often conflated across cutaneous and uveal disease; it should not be generalized to cutaneous melanoma without evidence.
Mechanism
[KNOWN] Tebentafusp redirects T cells toward gp100-expressing cells through an engineered T-cell receptor domain and anti-CD3 domain. [INFERRED] Applicability depends on HLA-A*02:01 and sufficient gp100 presentation.
Why it died
The agent is real, but the pack identifies its validated disease as HLA-A*02:01-positive uveal melanoma [KNOWN]. Uveal melanoma is ocular rather than cutaneous skin cancer, and the pack explicitly labels generalization to cutaneous melanoma as unsupported [INFERRED]. This fails the disease-target triage for a skin-cancer query.

Ipilimumab

killed · mechanism

anti-CTLA-4 monoclonal antibody

KNOWNINFERRED
Thesis
[KNOWN] Ipilimumab is an established melanoma immunotherapy and has been used in combination or later-line settings. [INFERRED] As a standalone broad skin-cancer proposal it is a wrong-but-plausible decoy, useful for testing whether the critic rejects nonspecific extrapolation from melanoma.
Mechanism
[KNOWN] Ipilimumab blocks CTLA-4-mediated inhibitory signaling during T-cell priming. [INFERRED] Increased T-cell activation may improve tumor immune recognition but can also increase immune-mediated toxicity.
Kill experiment
A defined cutaneous melanoma comparison showing no clinically meaningful activity or no usable role relative to modern checkpoint regimens would falsify the proposed benchmark, while lack of evidence in other skin histologies already falsifies broad extrapolation.
Why it died
The candidate's own thesis identifies standalone broad skin-cancer use as a wrong-but-plausible decoy. CTLA-4 blockade supplies only a generic T-cell-priming step; the pack provides no disease-specific mutation, target dependency, histology, or causal path from that step to symptoms for the unresolved skin query. Its melanoma history cannot be extrapolated to unspecified skin disease, and there are no query-specific targets, trials, or evidence entries supporting that extrapolation.