Follicular disorganization of the spleen in sickle cell disease identified by a convolutional neural network model

Abstract The spleen is the largest immune organ and is among the first to be functionally impaired in sickle cell disease (SCD). Histopathologic studies have consistently documented marginal zone depletion in the lymphoid follicles of SCD spleens. Our group and others have reported alterations in B, T, and dendritic cell populations in the peripheral blood of persons with SCD, paralleling changes observed in splenic tissue. We aim to directly compare immune cell phenotypes in peripheral blood with spatial patterns in spleen tissue from patients with SCD. Here, we describe our novel analysis of human spleen tissue. Fresh spleen tissue was collected from 12 patients undergoing total splenectomy (10 SCD, 2 hereditary spherocytosis [HS] controls). Formalin-fixed, paraffin-embedded sections were stained with hematoxylin & eosin (H&E) and immunohistochemical markers: IgD (unswitched memory B cells), CD21 (naïve follicular B cells and dendritic cells), CD3 (T cells), and Pax5 (B cell nuclei). Whole-slide images were acquired using a high-resolution scanner and analyzed using a convolutional neural network (cNN) model adapted from Cellpose (Stringer et al., 2021) to identify and segment follicles from the H&E images. Pathologist-in-the-loop training of the model improved follicle identification precision from 75% to 90%. Follicle metrics were derived, and distances from follicle centers to immunostained cells were calculated. Group comparisons were performed using one-way ANOVA with Tukey's post hoc test. Across all samples, a median of 33 follicles per subject were analyzed (IQR: 8–57). Compared with HS controls, SCD spleens showed similar mean follicle count (33.0 vs. 30.6, n.s.) and follicle density (1.48 vs. 1.76 follicles/1M pixels, n.s.). For all stains, mean intensity per follicle and the percent of positive cells were significantly higher (27-57% and 86-145%, p < 0.001) in subjects with SCD compared with controls. Spatial intensity profiles of immunohistochemical stains in controls displayed the expected pattern of highest signal at the follicle center with progressive decline outward, whereas SCD samples demonstrated a shifted distribution of staining, indicating follicular disorganization of CD3+ cells and potential loss of IgD+ and Pax5+ cells in adjacent tissue. These findings reveal structural disorganization and altered immune cell localization within SCD spleens beyond what is seen by H&E alone, consistent with functional impairment. While interpretation is limited by the small control group, ongoing recruitment and integration with peripheral blood phenotyping aim to elucidate mechanisms driving altered B cell differentiation and immune dysregulation in SCD.

Authors

Institutions

Publication Details

Journal
Blood
Published
2025-11-03
DOI
https://doi.org/10.1182/blood-2025-1159
Primary Topic
Hemoglobinopathies and Related Disorders
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
article

Follicular disorganization of the spleen in sickle cell disease identified by a convolutional neural network model

Norma Estrada, Venée N. Tubman, Choladda V. Curry, Adam T. Szafran et al.
Blood
Hemoglobinopathies and Related Disorders
article

Follicular disorganization of the spleen in sickle cell disease identified by a convolutional neural network model

Norma Estrada, Venée N. Tubman, Choladda V. Curry, Adam T. Szafran, Jonathan M. Flanagan
article en

Abstract

Abstract The spleen is the largest immune organ and is among the first to be functionally impaired in sickle cell disease (SCD). Histopathologic studies have consistently documented marginal zone depletion in the lymphoid follicles of SCD spleens. Our group and others have reported alterations in B, T, and dendritic cell populations in the peripheral blood of persons with SCD, paralleling changes observed in splenic tissue. We aim to directly compare immune cell phenotypes in peripheral blood with spatial patterns in spleen tissue from patients with SCD. Here, we describe our novel analysis of human spleen tissue. Fresh spleen tissue was collected from 12 patients undergoing total splenectomy (10 SCD, 2 hereditary spherocytosis [HS] controls). Formalin-fixed, paraffin-embedded sections were stained with hematoxylin & eosin (H&E) and immunohistochemical markers: IgD (unswitched memory B cells), CD21 (naïve follicular B cells and dendritic cells), CD3 (T cells), and Pax5 (B cell nuclei). Whole-slide images were acquired using a high-resolution scanner and analyzed using a convolutional neural network (cNN) model adapted from Cellpose (Stringer et al., 2021) to identify and segment follicles from the H&E images. Pathologist-in-the-loop training of the model improved follicle identification precision from 75% to 90%. Follicle metrics were derived, and distances from follicle centers to immunostained cells were calculated. Group comparisons were performed using one-way ANOVA with Tukey's post hoc test. Across all samples, a median of 33 follicles per subject were analyzed (IQR: 8–57). Compared with HS controls, SCD spleens showed similar mean follicle count (33.0 vs. 30.6, n.s.) and follicle density (1.48 vs. 1.76 follicles/1M pixels, n.s.). For all stains, mean intensity per follicle and the percent of positive cells were significantly higher (27-57% and 86-145%, p < 0.001) in subjects with SCD compared with controls. Spatial intensity profiles of immunohistochemical stains in controls displayed the expected pattern of highest signal at the follicle center with progressive decline outward, whereas SCD samples demonstrated a shifted distribution of staining, indicating follicular disorganization of CD3+ cells and potential loss of IgD+ and Pax5+ cells in adjacent tissue. These findings reveal structural disorganization and altered immune cell localization within SCD spleens beyond what is seen by H&E alone, consistent with functional impairment. While interpretation is limited by the small control group, ongoing recruitment and integration with peripheral blood phenotyping aim to elucidate mechanisms driving altered B cell differentiation and immune dysregulation in SCD.

BloodVol. 146(Supplement 1)
Baylor College of Medicine (US), Texas Children's Hospital (US)
Openalex Percentile: Top 42%
Hemoglobinopathies and Related Disorders
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.