Relentless pursuit of improved patient outcomes

The Castle story

Our story is simple. Our science is complex. From the beginning, Castle Biosciences was driven by a desire to help patients receive the best care possible. Our founder believed that the traditional methods to determine cancer patients' prognoses could be improved by harnessing the biology of their tumors. Through this belief, Castle was born.

Who we are

We are an innovator in proprietary laboratory-developed tests, which provide personalized, clinically actionable information that can help healthcare providers and patients make more informed disease management decisions.

Our mission

Improving health through innovative tests that guide patient care

Our vision

To transform disease management by keeping people first: patients, clinicians, employees and investors

Our history

2023
Opened new laboratory in Pittsburgh, doubling footprint of prior facility

Validation of the neural network algorithm incorporating clinicopathologic factors with the 31-GEP class score for personalized SLNB positivity risk assessment.

  • Of all factors, the 31-GEP was the most predictive of a positive SLN
  • The i31-SLNB algorithm reclassified 63% of patients with T1 tumors, originally classified with a 5-10% likelihood of SLN positivity.
  • Increase from 8.5% to 27.7% in patients with T1-T4 tumors predicted to have <5% SLN positive likelihood, with 98% NPV
2021
Acquisition of Cernostics and the TissueCypher® Barrett’s Esophagus test

Validation of the neural network algorithm incorporating clinicopathologic factors with the 31-GEP class score for personalized SLNB positivity risk assessment.

  • Of all factors, the 31-GEP was the most predictive of a positive SLN
  • The i31-SLNB algorithm reclassified 63% of patients with T1 tumors, originally classified with a 5-10% likelihood of SLN positivity.
  • Increase from 8.5% to 27.7% in patients with T1-T4 tumors predicted to have <5% SLN positive likelihood, with 98% NPV
2021
Acquisition of the MyPath® Melanoma laboratory and MyPath Melanoma test

Validation of the neural network algorithm incorporating clinicopathologic factors with the 31-GEP class score for personalized SLNB positivity risk assessment.

  • Of all factors, the 31-GEP was the most predictive of a positive SLN
  • The i31-SLNB algorithm reclassified 63% of patients with T1 tumors, originally classified with a 5-10% likelihood of SLN positivity.
  • Increase from 8.5% to 27.7% in patients with T1-T4 tumors predicted to have <5% SLN positive likelihood, with 98% NPV
2021
New corporate headquarters in Friendswood, TX

Validation of the neural network algorithm incorporating clinicopathologic factors with the 31-GEP class score for personalized SLNB positivity risk assessment.

  • Of all factors, the 31-GEP was the most predictive of a positive SLN
  • The i31-SLNB algorithm reclassified 63% of patients with T1 tumors, originally classified with a 5-10% likelihood of SLN positivity.
  • Increase from 8.5% to 27.7% in patients with T1-T4 tumors predicted to have <5% SLN positive likelihood, with 98% NPV
2021
Introduction of DiffDx®-Melanoma

Validation of the neural network algorithm incorporating clinicopathologic factors with the 31-GEP class score for personalized SLNB positivity risk assessment.

  • Of all factors, the 31-GEP was the most predictive of a positive SLN
  • The i31-SLNB algorithm reclassified 63% of patients with T1 tumors, originally classified with a 5-10% likelihood of SLN positivity.
  • Increase from 8.5% to 27.7% in patients with T1-T4 tumors predicted to have <5% SLN positive likelihood, with 98% NPV
2021
Introduction of DecisionDx®-SCC

Validation of the neural network algorithm incorporating clinicopathologic factors with the 31-GEP class score for personalized SLNB positivity risk assessment.

  • Of all factors, the 31-GEP was the most predictive of a positive SLN
  • The i31-SLNB algorithm reclassified 63% of patients with T1 tumors, originally classified with a 5-10% likelihood of SLN positivity.
  • Increase from 8.5% to 27.7% in patients with T1-T4 tumors predicted to have <5% SLN positive likelihood, with 98% NPV
2020
Doubled footprint of Phoenix laboratory

Validation of the neural network algorithm incorporating clinicopathologic factors with the 31-GEP class score for personalized SLNB positivity risk assessment.

  • Of all factors, the 31-GEP was the most predictive of a positive SLN
  • The i31-SLNB algorithm reclassified 63% of patients with T1 tumors, originally classified with a 5-10% likelihood of SLN positivity.
  • Increase from 8.5% to 27.7% in patients with T1-T4 tumors predicted to have <5% SLN positive likelihood, with 98% NPV
2019
Initial public offering completed

Validation of the neural network algorithm incorporating clinicopathologic factors with the 31-GEP class score for personalized SLNB positivity risk assessment.

  • Of all factors, the 31-GEP was the most predictive of a positive SLN
  • The i31-SLNB algorithm reclassified 63% of patients with T1 tumors, originally classified with a 5-10% likelihood of SLN positivity.
  • Increase from 8.5% to 27.7% in patients with T1-T4 tumors predicted to have <5% SLN positive likelihood, with 98% NPV
2018
Formal Medicare Local Coverage Determination (LCD) for DecisionDx®-Melanoma

Validation of the neural network algorithm incorporating clinicopathologic factors with the 31-GEP class score for personalized SLNB positivity risk assessment.

  • Of all factors, the 31-GEP was the most predictive of a positive SLN
  • The i31-SLNB algorithm reclassified 63% of patients with T1 tumors, originally classified with a 5-10% likelihood of SLN positivity.
  • Increase from 8.5% to 27.7% in patients with T1-T4 tumors predicted to have <5% SLN positive likelihood, with 98% NPV
2018
DecisionDx®-UM included in National Comprehensive Cancer Network (NCCN) guidelines

Validation of the neural network algorithm incorporating clinicopathologic factors with the 31-GEP class score for personalized SLNB positivity risk assessment.

  • Of all factors, the 31-GEP was the most predictive of a positive SLN
  • The i31-SLNB algorithm reclassified 63% of patients with T1 tumors, originally classified with a 5-10% likelihood of SLN positivity.
  • Increase from 8.5% to 27.7% in patients with T1-T4 tumors predicted to have <5% SLN positive likelihood, with 98% NPV
2017
Formal Medicare Local Coverage Determination (LCD) for DecisionDx®-UM

Validation of the neural network algorithm incorporating clinicopathologic factors with the 31-GEP class score for personalized SLNB positivity risk assessment.

  • Of all factors, the 31-GEP was the most predictive of a positive SLN
  • The i31-SLNB algorithm reclassified 63% of patients with T1 tumors, originally classified with a 5-10% likelihood of SLN positivity.
  • Increase from 8.5% to 27.7% in patients with T1-T4 tumors predicted to have <5% SLN positive likelihood, with 98% NPV
2013
Introduction of DecisionDx®-Melanoma

Validation of the neural network algorithm incorporating clinicopathologic factors with the 31-GEP class score for personalized SLNB positivity risk assessment.

  • Of all factors, the 31-GEP was the most predictive of a positive SLN
  • The i31-SLNB algorithm reclassified 63% of patients with T1 tumors, originally classified with a 5-10% likelihood of SLN positivity.
  • Increase from 8.5% to 27.7% in patients with T1-T4 tumors predicted to have <5% SLN positive likelihood, with 98% NPV
2010
DecisionDx®-UM included in AJCC guidelines

Validation of the neural network algorithm incorporating clinicopathologic factors with the 31-GEP class score for personalized SLNB positivity risk assessment.

  • Of all factors, the 31-GEP was the most predictive of a positive SLN
  • The i31-SLNB algorithm reclassified 63% of patients with T1 tumors, originally classified with a 5-10% likelihood of SLN positivity.
  • Increase from 8.5% to 27.7% in patients with T1-T4 tumors predicted to have <5% SLN positive likelihood, with 98% NPV
2009
Introduction of DecisionDx®-UM

Validation of the neural network algorithm incorporating clinicopathologic factors with the 31-GEP class score for personalized SLNB positivity risk assessment.

  • Of all factors, the 31-GEP was the most predictive of a positive SLN
  • The i31-SLNB algorithm reclassified 63% of patients with T1 tumors, originally classified with a 5-10% likelihood of SLN positivity.
  • Increase from 8.5% to 27.7% in patients with T1-T4 tumors predicted to have <5% SLN positive likelihood, with 98% NPV
2008
Began operations

Validation of the neural network algorithm incorporating clinicopathologic factors with the 31-GEP class score for personalized SLNB positivity risk assessment.

  • Of all factors, the 31-GEP was the most predictive of a positive SLN
  • The i31-SLNB algorithm reclassified 63% of patients with T1 tumors, originally classified with a 5-10% likelihood of SLN positivity.
  • Increase from 8.5% to 27.7% in patients with T1-T4 tumors predicted to have <5% SLN positive likelihood, with 98% NPV