Predicting Immunotherapy Response via Spatial Cell Proximity

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current methods lack effectiveness in predicting which cancer patients will respond positively to immunotherapy, as they do not adequately assess the spatial proximity of PD-1 and PD-L1 cells and biomarker positivity in tumor tissue, leading to variable treatment outcomes.

Innovation Solution

The method involves scoring tumor tissue samples based on the spatial proximity between PD-1 and PD-L1 cells and deriving a % biomarker positivity value for HLA-DR+ cells expressing IDO-1, comparing these scores to threshold values to predict patient response to immunotherapy.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If current prediction methods are used, then treatment outcomes vary, but the methods do not adequately assess spatial proximity of PD-1 and PD-L1 cells

Engineering Contradiction:
Improveprediction accuracyVSAvoidmethod complexity
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The method segments the assessment into two distinct scoring components: (1) an interaction score that quantifies spatial proximity between PD-1 and PD-L1 cells, and (2) a biomarker positivity percentage for HLA-DR+ cells. This segmentation allows each component to be evaluated independently and then integrated, improving prediction reliability while maintaining manageable complexity through modular assessment

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The invention adds a spatial dimension to the assessment by measuring the physical proximity between PD-1 and PD-L1 cells in tissue samples. Rather than only assessing biomarker presence, the method incorporates spatial coordinates and distance measurements, creating a two-dimensional evaluation framework (presence + location) that significantly improves prediction accuracy

Inventive Principle:
Principle #17Another dimension (Dimensionality change)

2Measurement precision

If spatial proximity assessment is implemented, then prediction accuracy improves, but measurement complexity increases

Engineering Contradiction:
Improvespatial proximity measurementVSAvoidmeasurement difficulty
Core Design Contradiction:
Measurement precisionVSDifficulty of detecting and measuring

Solution Approach 1:

The method uses image analysis software as an intermediary to automatically calculate spatial proximity scores from tissue sample images. The software processes the complex spatial relationships between cells, applying algorithms that measure distances and interactions without requiring manual measurement. This intermediary tool handles the measurement complexity while providing precise, reproducible spatial proximity assessments

Inventive Principle:
Principle #24Intermediary (Mediator)

Solution Approach 2:

The invention transforms complex spatial relationship data into a simplified numerical interaction score parameter. By converting multi-dimensional spatial coordinates into a single quantitative score that represents the degree of proximity between PD-1 and PD-L1 cells, the method maintains measurement precision while facilitating easier interpretation and comparison across samples

Inventive Principle:
Principle #35Parameter changes

Data Source

PatentEP3576757B1Method of predicting response to immunotherapy
Publication Date: 2021.06.09 NOVARTIS AG
  • EP3576757B1 patent drawingFigure 1
  • EP3576757B1 patent drawingFigure 2a~2b
  • EP3576757B1 patent drawingFigure 3a~3b

AI summary

The invention relates, in part, to methods of predicting a likelihood that a cancer patient will respond positively to immunotherapy. The methods include scoring a sample containing tumor tissue from a cancer patient, wherein the score is representative of a spatial proximity between at least one pair of cells, a first member of the at least one pair of cells expressing a first biomarker and a second member of the at least one pair of cells expressing a second biomarker that is different from the first biomarker, and deriving a value for % biomarker positivity (PBP) for all cells or optionally, one or more subsets thereof, present in a field of view of a tissue sample from the cancer patient.