ABSTRACT Chronic lymphocytic leukemia (CLL) is a highly heterogeneous B‐cell malignancy, spanning a spectrum from indolent conventional forms (C‐CLL) to Richter transformation. Accelerated CLL (A‐CLL) represents an intermediate subtype, histologically and clinically more aggressive than C‐CLL, yet its diagnosis relies on subjective histological criteria. We hypothesized that high‐dimensional single‐cell morphometrics from routine hematoxylin and eosin‐stained lymph node biopsies could provide complementary metrics for positioning cases along the disease continuum. We implemented a novel, open‐source methodology using CellProfiler and Seurat to segment cells and measure hundreds of parameters from two indolent and two aggressive cases. Datasets were integrated to mitigate technical batch effects. Our analysis combined an unsupervised approach with trajectory analysis and a pathologist‐guided supervised machine learning model. Both unsupervised and supervised approaches successfully and independently distinguished aggressive from indolent cases. Aggressive cases were enriched in morphometric clusters corresponding to paraimmunoblasts, while indolent cases were enriched in those corresponding to small lymphocytes. Pseudo‐bulk analysis using a subset of key parameters also successfully classified patients. In this highly exploratory, proof‐of‐concept study utilizing a small, well‐characterized cohort ( n = 4 CLL patients plus 2 RT reference cases), we present a robust and traceable single‐cell morphometric pipeline. By providing objective metrics of cellular and subcellular morphology, our method reveals strong segregation patterns reflecting the CLL continuum. Our findings warrant validation in larger cohorts, but offer novel, quantifiable insights that could potentially complement diagnostic criteria and aid in patient stratification.
Lorenzo et al. (Tue,) studied this question.