Omentum Pathsampling Analysis

Evidence-Based Protocol for Detecting Occult Metastases

Author

Serdar Balcı

Published

December 1, 2025

Overview

This project presents a comprehensive pathsampling analysis of 1,096 omentum specimens from gynecological malignancy cases to establish evidence-based tissue sampling protocols for detecting microscopic-only metastases in grossly normal omentum.

ImportantKey Clinical Recommendation

Sample 4 cassettes from grossly normal omentum to achieve 95% detection confidence

  • Validated across multiple analytical approaches
  • Based on 46 microscopic-only cases with systematic detection tracking
  • Sensitivity: 95.3% (binomial) to 97.6% (empirical)

Main Documents

📊 Main Pathsampling Analysis

Comprehensive reproducible analysis with:

  • Complete dataset analysis (n=1,096 cases)
  • Microscopic-only subset focus (n=46 tracked cases)
  • Bootstrap validation (10,000 iterations)
  • Risk-stratified recommendations
  • Clinical protocol implementation guide

Key Findings: - Detection probability (q): 0.5349 (95% CI: 0.451-0.639) - Mean first detection: 1.87 cassettes - Recommended minimum: 4 cassettes for 95% confidence

📋 Executive Summary Report

Quick reference guide including:

  • Validated recommendations across all methods
  • Comparison: All cases vs microscopic-only
  • Perfect concordance with jamovi analysis
  • Copy-ready text for publications

📈 Detailed Results Report

In-depth analysis featuring:

  • Complete statistical methodology
  • Bootstrap distribution analysis
  • Subgroup analyses by location and grade
  • Quality assessment metrics

✅ Jamovi Validation Analysis

Independent validation using jamovi GUI:

  • Analysis type: All-cases approach (61 positive out of 1,096 total)
  • Includes: All tumor types (microscopic-only + abundant + other)
  • Detection probability: 56.0% (all positive cases)
  • Recommendation: 4 cassettes for 95% confidence
  • Bootstrap sensitivity: 96.7% at 4 cassettes
  • Binomial sensitivity: 96.2% at 4 cassettes
  • Validation: Perfect match with R implementation

Study Design

Dataset Characteristics

Characteristic Value
Total specimens 1,096
Microscopic-only cases 60 (5.5%)
Cases with tracking 46 (76.7% of microscopic-only)
Primary locations Ovary, Endometrium, Synchronous
Data source Single academic center
Study design Retrospective consecutive series

Analysis Approach

  1. Geometric probability model for sequential detection
  2. Maximum likelihood estimation for detection probability (q)
  3. Bootstrap confidence intervals (10,000 iterations)
  4. Multiple validation methods (jamovi, R, manual calculations)

Clinical Context

Microscopic-Only Metastases

Definition: Grossly normal omentum with occult metastases found only on microscopic examination

Clinical Significance:

  • Occurs in 5.5% of cases
  • Not visible at gross examination or surgery
  • Upstages disease and impacts treatment decisions
  • Requires systematic sampling protocol

Current Practice Gap

Many institutions sample 1-2 sections of normal-appearing omentum:

  • Detection rate with 2 cassettes: only 78.4%
  • Missing 21.6% of occult metastases
  • Based on expert opinion rather than systematic data

Evidence-Based Protocol (4 cassettes)

  • Detection rate: 95.3%
  • Missing only 4.7% of occult metastases
  • Captures an additional 17% of cases compared to 2-cassette protocol

Risk-Stratified Recommendations

Risk Level Clinical Scenario Cassettes Detection Rate
Low Early stage, low-grade 3 93.8%
Standard Stage III, high-grade 4 95.3%
High Stage IV, post-chemotherapy 5-7 98-99%

Data Files

Primary Data

  • omentum_new.xlsx - Complete dataset (1,096 cases)
    • Original data with all variables
    • Includes tracking information for sequential detection

Analysis Outputs

Supporting Documentation

Validation Summary

Three Independent Analyses

Analysis Dataset Positive Cases Detection (q) Cassettes Sensitivity
Jamovi (All) 1,096 total 61 with tracking 56.0% 4 96.7%
R (All) 1,096 total 61 with tracking 56.0% 4 96.7%
R (Micro-only) 60 micro-only 46 with tracking 53.5% 4 95.3%

Key Points:

  • 61 cases = All positive cases with first detection tracking:
    • 46 Microscopic-Only (macro=Absent, micro=Present)
    • 15 Abundant/Obvious (macro=Present, micro=Present)
  • 60 cases = Total microscopic-only cases (46 tracked + 14 untracked)
  • 46 cases = Microscopic-only subset used for primary QMD analysis
  • Different q values (56.0% vs 53.5%) due to different populations
  • Same recommendation = 4 cassettes for all approaches

Note: The 15 abundant cases have BOTH macroscopic and microscopic tumor. They are included in the jamovi “all cases” analysis but excluded from the microscopic-only focused analysis.

Result: ✅ Perfect concordance across all methods

Reproducibility

Render Main Analysis

cd /Users/serdarbalci/Desktop/omentum
quarto render omentum_pathsampling_analysis.qmd

Render Complete Website

quarto render

View Results in R

# Load microscopic-only results
result <- readRDS("results_microscopic.rds")
print(result$keyResults)
print(result$recommendTable$asDF)

# View data
library(readxl)
data <- read_excel("omentum_new.xlsx")

Citation

When using this analysis, please cite:

Balcı S. Evidence-Based Pathsampling Protocol for Omentum in Gynecological Malignancies: A Systematic Analysis of 1,096 Consecutive Cases. [Journal pending submission]. 2025.

Statistical Software

  • R version: 4.x
  • Quarto: Latest version
  • ClinicoPath package: Custom jamovi module
  • Key packages: tidyverse, readxl, ggplot2, patchwork

Contact

Principal Investigator: Serdar Balcı, MD Institution: [Academic Medical Center] Analysis Date: November 28-December 1, 2025

Acknowledgments

This analysis uses:

  • ClinicoPath jamovi module for pathsampling function
  • Quarto for reproducible research documents
  • Bootstrap methods for robust confidence intervals
  • Geometric probability model for sequential detection

Last updated: December 1, 2025