Related Experiment Videos
Individualized prognosis for melanoma patients
A J Cochran1, D Elashoff, D L Morton
1Department of Pathology, UCLA School of Medicine, Los Angeles, CA 90095-1732, USA.
Human Pathology
|April 4, 2000
Summary
This study developed a new melanoma prognostication tool using five key variables to predict individual survival probability. The model accurately assesses melanoma patient outcomes, improving upon existing risk category assignments.
Area of Science:
- Oncology
- Dermatology
- Biostatistics
Background:
- Malignant melanoma prognosis is highly variable.
- Current risk stratification methods lack individual patient precision.
- Accurate prognostication is crucial for effective melanoma management.
Purpose of the Study:
- To develop an individualized risk assessment model for melanoma patient survival.
- To identify key prognostic variables for predicting melanoma recurrence.
- To enhance the accuracy of melanoma prognostication beyond current methods.
Main Methods:
- Analysis of a melanoma histology database of 1,042 patients.
- Univariate Cox model and step-up multivariate analysis of 30 variables.
- Development of risk score formulas for survival and recurrence probability.
Main Results:
- Five variables (gender, site, age, Breslow thickness, ulceration) significantly predict survival.
- Individualized risk scores correlate with predicted 3-, 5-, and 10-year survival probabilities.
- Four variables (site, subtype, thickness, ulceration) predict melanoma recurrence likelihood.
Conclusions:
- A novel, simple, and accurate method for individual melanoma prognostication was established.
- The model integrates readily available demographic and histological data.
- This approach offers improved predictive accuracy compared to single-factor analyses.