Tutorial: ICH Guidelines Integration with CTWise API

Difficulty: Intermediate Time to Complete: 20 minutes Prerequisites: CTWise API key, familiarity with ICH guidelines

---

Goal

Learn how to integrate ICH (International Council for Harmonisation) guidelines into your clinical trial workflows using the CTWise API. By the end of this tutorial, you'll be able to:

1. Query ICH guidelines by topic area (E, S, M, Q series)
2. Cross-reference ICH with FDA regulations
3. Build a compliance matrix using both sources
4. Track guideline updates and revisions

---

Prerequisites

Before you begin, ensure you have:

• A CTWise API key (Starter tier or higher recommended)
• Understanding of ICH guideline structure (E, S, M, Q series)
• Python 3.8+ or Node.js 18+ for code examples

---

ICH Guideline Series Overview

Series	Topic	Examples
E	Efficacy	E6 (GCP), E8 (Clinical Studies), E9 (Statistics)
S	Safety	S1 (Carcinogenicity), S7 (Pharmacology), S9 (Anticancer)
M	Multidisciplinary	M4 (CTD), M7 (Mutagenic Impurities), M10 (Bioanalytical)
Q	Quality	Q1-Q14 (Manufacturing, Stability, etc.)

---

Step 1: List Available ICH Guidelines

First, retrieve all ICH guidelines available in CTWise:

export CTWISE_API_KEY="your-api-key-here"
export CTWISE_API_BASE="https://kj2xss9yg9.execute-api.us-east-1.amazonaws.com/dev0/v1"

curl -s "${CTWISE_API_BASE}/rules?source=ich&limit=50" \
  -H "x-api-key: ${CTWISE_API_KEY}" | jq '.rules[] | {id, title, category}'

Expected Output:

{"id": "ich-e6-r2", "title": "ICH E6(R2) Good Clinical Practice", "category": "efficacy"}
{"id": "ich-e8-r1", "title": "ICH E8(R1) General Considerations for Clinical Studies", "category": "efficacy"}
{"id": "ich-e9-r1", "title": "ICH E9(R1) Statistical Principles for Clinical Trials", "category": "efficacy"}
{"id": "ich-s7a", "title": "ICH S7A Safety Pharmacology Studies", "category": "safety"}
...

---

Step 2: Query Efficacy (E-Series) Guidelines

For clinical trial design, focus on E-series guidelines:

curl -s "${CTWISE_API_BASE}/rules?source=ich&category=efficacy&limit=20" \
  -H "x-api-key: ${CTWISE_API_KEY}" | jq .

Key E-Series Guidelines:

Guideline	Topic	Key Requirements
E6(R2)	GCP	Sponsor, investigator, IRB responsibilities
E8(R1)	Clinical Studies	Study design principles
E9(R1)	Statistics	Statistical methodology, estimands
E10	Choice of Control	Placebo, active control selection
E14	QT/QTc	Cardiac safety evaluation
E17	Multi-Regional	Multi-regional clinical trials

---

Step 3: Cross-Reference ICH with FDA

A powerful CTWise feature is cross-referencing ICH guidelines with FDA regulations:

# Get ICH E6(R2) with related FDA regulations
curl -s "${CTWISE_API_BASE}/rules/ich-e6-r2" \
  -H "x-api-key: ${CTWISE_API_KEY}" | jq '.references'

Expected Output:

{
  "references": [
    {"id": "fda-21cfr312", "relationship": "implements", "description": "FDA IND regulations"},
    {"id": "fda-21cfr50", "relationship": "harmonized_with", "description": "FDA informed consent"},
    {"id": "fda-21cfr56", "relationship": "harmonized_with", "description": "FDA IRB requirements"},
    {"id": "fda-21cfr11", "relationship": "referenced_by", "description": "Electronic records"}
  ]
}

---

Step 4: Build a Compliance Matrix

Create a compliance matrix that maps ICH guidelines to FDA regulations:

import requests
import json

API_KEY = "your-api-key-here"
API_BASE = "https://kj2xss9yg9.execute-api.us-east-1.amazonaws.com/dev0/v1"

headers = {
    "x-api-key": API_KEY,
    "Content-Type": "application/json"
}

def build_compliance_matrix(ich_guidelines: list) -> dict:
    """
    Build a matrix mapping ICH guidelines to FDA regulations.

    Args:
        ich_guidelines: List of ICH guideline IDs (e.g., ["ich-e6-r2", "ich-e8-r1"])

    Returns:
        Dictionary with ICH → FDA mappings
    """
    matrix = {}

    for guideline_id in ich_guidelines:
        # Get guideline details with references
        response = requests.get(
            f"{API_BASE}/rules/{guideline_id}",
            headers=headers
        )
        response.raise_for_status()
        data = response.json()

        # Extract FDA references
        fda_refs = [
            ref for ref in data.get("references", [])
            if ref["id"].startswith("fda-")
        ]

        matrix[guideline_id] = {
            "title": data["title"],
            "category": data["category"],
            "fda_references": fda_refs,
            "harmonization_status": "Full" if len(fda_refs) > 0 else "Partial"
        }

    return matrix

# Build matrix for key E-series guidelines
guidelines = ["ich-e6-r2", "ich-e8-r1", "ich-e9-r1", "ich-e10", "ich-e14"]
matrix = build_compliance_matrix(guidelines)

# Print matrix
print("ICH-FDA Compliance Matrix")
print("=" * 60)
for ich_id, data in matrix.items():
    print(f"\n{ich_id}: {data['title']}")
    print(f"  Category: {data['category']}")
    print(f"  Harmonization: {data['harmonization_status']}")
    print(f"  FDA References:")
    for ref in data["fda_references"]:
        print(f"    - {ref['id']}: {ref['description']}")

Sample Output:

ICH-FDA Compliance Matrix
============================================================

ich-e6-r2: ICH E6(R2) Good Clinical Practice
  Category: efficacy
  Harmonization: Full
  FDA References:
    - fda-21cfr312: FDA IND regulations
    - fda-21cfr50: FDA informed consent
    - fda-21cfr56: FDA IRB requirements

ich-e9-r1: ICH E9(R1) Statistical Principles for Clinical Trials
  Category: efficacy
  Harmonization: Full
  FDA References:
    - fda-guidance-statistical: FDA statistical guidance

---

Step 5: Semantic Search Across ICH and FDA

Search both sources simultaneously for a specific topic using either POST or GET methods:

POST Method (Recommended)

curl -s -X POST "${CTWISE_API_BASE}/semantic-search" \
  -H "x-api-key: ${CTWISE_API_KEY}" \
  -H "Content-Type: application/json" \
  -d '{
    "query": "requirements for informed consent in pediatric clinical trials",
    "sources": ["ich", "fda"],
    "limit": 10
  }' | jq '.results[] | {id: .rule_id, title, source, score: .similarity_score}'

GET Method (Alternative)

curl -s "${CTWISE_API_BASE}/rules/search?q=informed+consent+pediatric+trials&sources=ich,fda&limit=10" \
  -H "x-api-key: ${CTWISE_API_KEY}" | jq '.results[] | {id: .rule_id, title, source, score: .similarity_score}'

Expected Output:

{"id": "ich-e11", "title": "ICH E11 Clinical Investigation of Medicinal Products in the Pediatric Population", "source": "ich", "relevance_score": 0.94}
{"id": "fda-21cfr50-subpart-d", "title": "21 CFR 50 Subpart D - Additional Safeguards for Children", "source": "fda", "relevance_score": 0.91}
{"id": "ich-e6-r2", "title": "ICH E6(R2) Good Clinical Practice", "source": "ich", "relevance_score": 0.78}
{"id": "fda-21cfr50", "title": "21 CFR 50 - Protection of Human Subjects", "source": "fda", "relevance_score": 0.75}

---

Step 6: Node.js Integration Example

const axios = require('axios');

const API_KEY = process.env.CTWISE_API_KEY;
const API_BASE = 'https://kj2xss9yg9.execute-api.us-east-1.amazonaws.com/dev0/v1';

const headers = {
  'x-api-key': API_KEY,
  'Content-Type': 'application/json'
};

async function getICHGuidelines(series) {
  /**
   * Get ICH guidelines by series (efficacy, safety, quality, multidisciplinary)
   */
  const response = await axios.get(`${API_BASE}/rules`, {
    headers,
    params: {
      source: 'ich',
      category: series,
      limit: 50
    }
  });
  return response.data;
}

async function crossReferenceWithFDA(ichGuidelineId) {
  /**
   * Get FDA regulations that implement or harmonize with an ICH guideline
   */
  const response = await axios.get(`${API_BASE}/rules/${ichGuidelineId}`, {
    headers
  });

  const ichRule = response.data;
  const fdaReferences = ichRule.references.filter(ref => ref.id.startsWith('fda-'));

  // Fetch details for each FDA reference
  const fdaDetails = await Promise.all(
    fdaReferences.map(async (ref) => {
      const fdaResponse = await axios.get(`${API_BASE}/rules/${ref.id}`, { headers });
      return {
        ...ref,
        details: fdaResponse.data
      };
    })
  );

  return {
    ich: ichRule,
    fda_references: fdaDetails
  };
}

// Example usage
(async () => {
  // Get all efficacy guidelines
  const efficacyGuidelines = await getICHGuidelines('efficacy');
  console.log(`Found ${efficacyGuidelines.total} efficacy guidelines`);

  // Cross-reference ICH E6(R2)
  const crossRef = await crossReferenceWithFDA('ich-e6-r2');
  console.log(`\nICH E6(R2) maps to ${crossRef.fda_references.length} FDA regulations`);

  crossRef.fda_references.forEach(ref => {
    console.log(`  - ${ref.id}: ${ref.description}`);
  });
})();

---

Use Case: Clinical Trial Compliance Checklist

Generate a compliance checklist for a Phase 3 clinical trial:

def generate_trial_compliance_checklist(trial_phase: str = "3") -> list:
    """
    Generate compliance checklist based on ICH and FDA requirements.
    """
    # Search for phase-specific requirements using semantic search
    response = requests.post(
        f"{API_BASE}/semantic-search",
        headers=headers,
        json={
            "query": f"Phase {trial_phase} clinical trial requirements",
            "sources": ["ich", "fda"],
            "limit": 20
        }
    )
    rules = response.json()["results"]

    checklist = []
    for rule in rules:
        checklist.append({
            "requirement": rule["title"],
            "source": rule["source"].upper(),
            "rule_id": rule["id"],
            "category": rule.get("category", "general"),
            "completed": False
        })

    return checklist

# Generate checklist
checklist = generate_trial_compliance_checklist("3")

print("Phase 3 Clinical Trial Compliance Checklist")
print("=" * 50)
for i, item in enumerate(checklist, 1):
    print(f"{i}. [{' '}] {item['requirement']}")
    print(f"      Source: {item['source']} | Rule: {item['rule_id']}")

---

Troubleshooting

Error	Cause	Solution
Empty ICH results	ICH not in your sources	Verify ICH source is available for your tier
Missing cross-references	Rule hasn't been mapped	Check rule's last_updated date
Slow response	Large result set	Use pagination, reduce limit
Rate limit errors	Too many requests	Implement exponential backoff

---

Next Steps

• Query FDA Rules Tutorial - Deep dive into FDA regulations
• Rule Updates Monitoring - Set up change alerts
• SDK Documentation - Use our Python and JavaScript SDKs

---

Additional Resources

• ICH Official Website
• FDA Guidance Documents
• CTWise API Reference

---

Feedback

Found an issue with this tutorial? Submit feedback or contact support@orchestraprime.ai.