临床试验设计可行性研究｜个人思维图谱小站

正文

获取原文件

SKILL.md

中文导读

从终点、人群、比较方案、效应量、周期及监管范围整理试验设计的证据与可行性资料。

输入材料

研究问题与拟定试验条件
公开试验、流行病学及药物资料

交付结果

试验可行性报告
设计依据与尚待验证事项

阅读与使用

本页保存指定版本的原始 Skill 和配套文件，可查看中文导读、原文与参考资料。获取完整包后，按原文准备所需工具与材料。执行需要 ToolUniverse SDK 或 MCP 服务，部分数据库需要独立访问权限。医疗资料供研究、资料整理和专业人员复核，试验资格、用药与诊疗结论需要核实当前来源。

来源仓库为 mims-harvard/ToolUniverse，采用 Apache-2.0 许可。Stars 表示仓库关注度，核验日期为 2026-10-15。分类表示适用的工作环节，实际执行条件以原文为准。

原始 Skill

Systematically assess clinical trial feasibility by analyzing 6 research dimensions. Produces comprehensive feasibility reports with quantitative enrollment projections, endpoint recommendations, and regulatory pathway analysis.

IMPORTANT: Always use English terms in tool calls (drug names, disease names, biomarker names), even if the user writes in another language. Only try original-language terms as a fallback if English returns no results. Respond in the user's language.

Reasoning Before Searching

Trial design starts with the question, not the methods. Answer these four questions before running any tools — they determine everything else:

What is the primary endpoint? Is it overall survival (gold standard but slow), PFS (faster but surrogate), ORR (single-arm friendly but not always accepted), or a biomarker (needs validation as surrogate first)? The endpoint determines FDA pathway, statistical design, and duration.
Who is the population? Broad unselected vs. biomarker-enriched. Enriched populations have higher response rates, allowing smaller trials — but require a validated companion diagnostic and reduce the eligible patient pool.
What is the comparator? Placebo (only if no standard of care exists), active control (requires non-inferiority or superiority framing), or single-arm with historical control (acceptable for rare diseases or breakthrough designations, but FDA scrutiny is high).
Is the effect size realistic given the mechanism? A 20% improvement in ORR over SOC requires ~100 patients per arm. A 50% improvement requires ~30. If the mechanism only justifies a 10% improvement, the trial may be underpowered regardless of design. Check precedent effect sizes in similar trials before committing to an endpoint.

These four answers determine sample size, duration, and trial design. Look them up from precedent trials and FDA guidance — do not derive them from first principles.

LOOK UP DON'T GUESS: Never assume what the standard of care is for an indication — look it up with DrugBank and FDA tools. Never assume an endpoint is FDA-accepted — verify with search_clinical_trials precedents and OpenFDA_get_approval_history. Never estimate prevalence from memory — use OpenTargets, gnomAD, or COSMIC.

Core Principles

1. Report-First Approach (MANDATORY)

DO NOT show tool outputs to user. Instead:

Create [INDICATION]_trial_feasibility_report.md FIRST
Initialize with all section headers
Progressively update as data arrives
Present only the final report

2. Evidence Grading System

Grade	Symbol	Criteria	Examples
A	3-star	Regulatory acceptance, multiple precedents	FDA-approved endpoint in same indication
B	2-star	Clinical validation, single precedent	Phase 3 trial in related indication
C	1-star	Preclinical or exploratory	Phase 1 use, biomarker validation ongoing
D	0-star	Proposed, no validation	Novel endpoint, no precedent

3. Feasibility Score (0-100)

Weighted composite score:

Patient Availability (30%): Population size x biomarker prevalence x geography
Endpoint Precedent (25%): Historical use, regulatory acceptance
Regulatory Clarity (20%): Pathway defined, precedents exist
Comparator Feasibility (15%): Standard of care availability
Safety Monitoring (10%): Known risks, monitoring established

Interpretation: >=75 HIGH (proceed), 50-74 MODERATE (additional validation), <50 LOW (de-risking required)

When to Use This Skill

Apply when users:

Plan early-phase trials (Phase 1/2 emphasis)
Need enrollment feasibility assessment
Design biomarker-selected trials
Evaluate endpoint strategies
Assess regulatory pathways
Compare trial design options
Need safety monitoring plans

Trigger phrases: "clinical trial design", "trial feasibility", "enrollment projections", "endpoint selection", "trial planning", "Phase 1/2 design", "basket trial", "biomarker trial"

Core Strategy: 6 Research Paths

Execute 6 parallel research dimensions. See STUDY_DESIGN_PROCEDURES.md for detailed steps per path.

Trial Design Query
|
+-- PATH 1: Patient Population Sizing
|   Disease prevalence, biomarker prevalence, geographic distribution,
|   eligibility criteria impact, enrollment projections
|
+-- PATH 2: Biomarker Prevalence & Testing
|   Mutation frequency, testing availability, turnaround time,
|   cost/reimbursement, alternative biomarkers
|
+-- PATH 3: Comparator Selection
|   Standard of care, approved comparators, historical controls,
|   placebo appropriateness, combination therapy
|
+-- PATH 4: Endpoint Selection
|   Primary endpoint precedents, FDA acceptance history,
|   measurement feasibility, surrogate vs clinical endpoints
|
+-- PATH 5: Safety Endpoints & Monitoring
|   Mechanism-based toxicity, class effects, organ-specific monitoring,
|   DLT history, safety monitoring plan
|
+-- PATH 6: Regulatory Pathway
    Regulatory precedents (505(b)(1), 505(b)(2)), breakthrough therapy,
    orphan drug, fast track, FDA guidance

Report Structure (14 Sections)

Create [INDICATION]_trial_feasibility_report.md with all 14 sections. See REPORT_TEMPLATE.md for full templates with fillable fields.

Executive Summary - Feasibility score, key findings, go/no-go recommendation
Disease Background - Prevalence, incidence, SOC, unmet need
Patient Population Analysis - Base population, biomarker selection, eligibility funnel, enrollment projections
Biomarker Strategy - Primary biomarker, alternatives, testing logistics
Endpoint Selection & Justification - Primary/secondary/exploratory endpoints, statistical considerations
Comparator Analysis - SOC, trial design options (single-arm vs randomized vs non-inferiority), drug sourcing
Safety Endpoints & Monitoring Plan - DLT definition, mechanism-based toxicities, organ monitoring, SMC
Study Design Recommendations - Phase, design type, schema, eligibility, treatment plan, assessment schedule
Enrollment & Site Strategy - Site selection, enrollment projections, recruitment strategies
Regulatory Pathway - FDA pathway, precedents, pre-IND meeting, IND timeline
Budget & Resource Considerations - Cost drivers, timeline, FTE requirements
Risk Assessment - Feasibility risks, scientific risks, mitigation strategies
Success Criteria & Go/No-Go Decision - Phase 1/2 criteria, interim analysis, feasibility scorecard
Recommendations & Next Steps - Final recommendation, critical path to IND, alternative designs

Tool Reference by Research Path

PATH 1: Patient Population Sizing

OpenTargets_get_disease_id_description_by_name - Disease lookup
OpenTargets_get_diseases_phenotypes_by_target_ensembl - Prevalence data
ClinVar_search_variants - Biomarker mutation frequency
gnomad_search_variants - Population allele frequencies
PubMed_search_articles - Epidemiology literature
search_clinical_trials - Enrollment feasibility from past trials

PATH 2: Biomarker Prevalence & Testing

ClinVar_get_variant_details - Variant pathogenicity
COSMIC_search_mutations - Cancer-specific mutation frequencies
gnomad_get_variant - Population genetics
PubMed_search_articles - CDx test performance, guidelines

PATH 3: Comparator Selection

drugbank_get_drug_basic_info_by_drug_name_or_id - Drug info
drugbank_get_indications_by_drug_name_or_drugbank_id - Approved indications
drugbank_get_pharmacology_by_drug_name_or_drugbank_id - Mechanism
FDA_OrangeBook_search_drug - Generic availability
OpenFDA_get_approval_history - Approval details
search_clinical_trials - Historical control data

PATH 4: Endpoint Selection

search_clinical_trials - Precedent trials, endpoints used
PubMed_search_articles - FDA acceptance history, endpoint validation
OpenFDA_get_approval_history - Approved endpoints by indication

PATH 5: Safety Endpoints & Monitoring

drugbank_get_pharmacology_by_drug_name_or_drugbank_id - Mechanism toxicity
FDA_get_warnings_and_cautions_by_drug_name - FDA black box warnings
FAERS_search_reports_by_drug_and_reaction - Real-world adverse events
FAERS_count_reactions_by_drug_event - AE frequency
FAERS_count_death_related_by_drug - Serious outcomes
PubMed_search_articles - DLT definitions, monitoring strategies

PATH 6: Regulatory Pathway

OpenFDA_get_approval_history - Precedent approvals
PubMed_search_articles - Breakthrough designations, FDA guidance
search_clinical_trials - Regulatory precedents (accelerated approval)

Quick Start Example

from tooluniverse import ToolUniverse

tu = ToolUniverse()
tu.load_tools()

# Example: EGFR+ NSCLC trial feasibility
# Step 1: Disease prevalence
disease_info = tu.tools.OpenTargets_get_disease_id_description_by_name(
    diseaseName="non-small cell lung cancer"
)
# Returns HPO phenotypes only, not prevalence figures; get prevalence from literature
prevalence = tu.tools.OpenTargets_get_associated_phenotypes_by_disease_efoId(
    efoId=disease_info['data']['search']['hits'][0]['id']
)

# Step 2: Biomarker prevalence
variants = tu.tools.ClinVar_search_variants(gene="EGFR", significance="pathogenic")

# Step 3: Precedent trials
trials = tu.tools.ClinicalTrials_search_studies(
    condition="EGFR positive non-small cell lung cancer",
    filter_status="COMPLETED", filter_phase="PHASE2"
)

# Step 4: Standard of care comparator
soc = tu.tools.FDA_OrangeBook_search_drug(generic_name="osimertinib")

# Compile into feasibility report...

See WORKFLOW_DETAILS.md for the complete 6-path Python workflow and use case examples.

Next Steps

Once this feasibility assessment is complete, use tooluniverse-clinical-trial-protocol to turn the 6-dimension findings (endpoint, population, comparator, effect size, duration, regulatory pathway) into an actual protocol document — full ICH E6(R2)/FDA-style sections plus a real sample-size calculation. That skill consumes this report's output rather than re-deriving feasibility.

Integration with Other Skills

tooluniverse-drug-research: Investigate mechanism, preclinical data
tooluniverse-disease-research: Deep dive on disease biology
tooluniverse-target-research: Validate drug target, essentiality
tooluniverse-pharmacovigilance: Post-market safety for comparator drugs
tooluniverse-precision-oncology: Biomarker biology, resistance mechanisms

Programmatic Access (Beyond Tools)

When ToolUniverse tools return limited trial metadata, use the ClinicalTrials.gov v2 API directly:

import requests, pandas as pd

# Search with pagination (all lung cancer immunotherapy trials with results)
all_studies = []
token = None
while True:
    params = {"query.cond": "lung cancer", "query.intr": "immunotherapy",
              "filter.overallStatus": "COMPLETED", "filter.results": "WITH_RESULTS", "pageSize": 100}
    if token: params["pageToken"] = token
    resp = requests.get("https://clinicaltrials.gov/api/v2/studies", params=params).json()
    all_studies.extend(resp.get("studies", []))
    token = resp.get("nextPageToken")
    if not token: break

# Extract structured data
rows = []
for s in all_studies:
    proto = s.get("protocolSection", {})
    rows.append({
        "nctId": proto.get("identificationModule", {}).get("nctId"),
        "title": proto.get("identificationModule", {}).get("briefTitle"),
        "enrollment": proto.get("designModule", {}).get("enrollmentInfo", {}).get("count"),
        "phase": proto.get("designModule", {}).get("phases", [None])[0] if proto.get("designModule", {}).get("phases") else None,
    })
df = pd.DataFrame(rows)

# FDA drug approval history
drug = "pembrolizumab"
fda = requests.get(f"https://api.fda.gov/drug/drugsfda.json?search=openfda.brand_name:{drug}&limit=10").json()

See tooluniverse-data-wrangling skill for pagination, error handling, and bulk download patterns.

Reference Files

File	Content
`REPORT_TEMPLATE.md`	Full 14-section report template with fillable fields
`STUDY_DESIGN_PROCEDURES.md`	Detailed steps for each of the 6 research paths
`WORKFLOW_DETAILS.md`	Complete Python example workflow and 5 use case summaries
`BEST_PRACTICES.md`	Best practices, common pitfalls, output format requirements
`EXAMPLES.md`	Additional examples
`QUICK_START.md`	Quick start guide

Version Information

Version: 1.0.0
Last Updated: February 2026
Compatible with: ToolUniverse 0.5+
Focus: Phase 1/2 early clinical development

临床试验设计可行性研究

正文