Multiply Logo

Integration

The module integrates multiple features within a multidimensional framework, making it possible to assess whether patients can be meaningfully clustered using only the available clinical variables.

Features

The following clinical variables were included in the analysis:

Ki-67

Proliferation index.

MIPI

Mantle Cell Lymphoma International Prognostic Index.

LDH

Lactate dehydrogenase.

ECOG

Performance status score (0-4) describing the patient's daily functional ability.

Age

Patient age at diagnosis.

Sex

Patient sex.

P53 status

TP53 mutation/expression status.

Bulky

Presence of bulky disease.

Stage

Ann Arbor stage.

Dimensionality Reduction Analysis

A dimensionality reduction approach based on Principal Component Analysis (PCA) was applied to summarize the multidimensional clinical information and visualize patient distribution in a two-dimensional space.

Scatterplot of PCA for overall survival
Scatterplot of PCA for progression-free survival

The resulting PCA scatter plot represents each patient according to the first two principal components, which capture the largest proportion of variability in the original dataset.

Loadings plot of PCA

Contribution of each clinical variable to the principal components

The plot enables the visual assessment of potential patient clusters, group separation, overlapping clinical profiles, and possible outliers. The percentage of variance explained by each principal component is reported on the corresponding axis and should be considered when interpreting the observed distribution.

Hierarchical Clustering and Clinical Profiles

Based on the principal components obtained from the dimensionality reduction analysis, hierarchical clustering was performed, identifying four distinct patient groups.

Cluster plot of PCA

Patient clusters projected onto the first two principal components

The distribution of the clinical features across the four clusters is shown below, allowing the characterization of the specific clinical profile associated with each group.

Distribution of clinical features across clusters

Distribution of clinical variables by cluster

Survival Analysis

Kaplan-Meier curves for progression-free survival (PFS) and overall survival (OS), according to the four-cluster patient stratification, are also reported below.

Kaplan-Meier curves for PFS

Kaplan-Meier curves for Progression-Free Survival (PFS)

Kaplan-Meier curves for OS

Kaplan-Meier curves for Overall Survival (OS)