Digital pathology in pediatric nodular lymphocyte predominant Hodgkin lymphoma: correlation with treatment response.
Sergej SeredaAnanth G ShankarLuise WeberAlan D RamsayGeorgina W HallJanis HaywardWilliam Hamish Beith WallaceJudith Landman-ParkerAndreas BraeuningerDirk HasencleverAstrid SchneiderChristine Mauz-KoerholzDieter KörholzStefan GattenloehnerPublished in: Blood advances (2023)
Early-stage pediatric nodular lymphocyte predominant Hodgkin lymphoma (NLPHL) can be treated effectively with low-intensity chemotherapy, most frequently cyclophosphamide in combination with vinblastine and prednisone (CVP). Descriptive histological risk factors based on the disease-defining lymphocyte predominant cells (LP cells) as used within the Fan classification, are less predictive in early-stage patients. We used deep learning-based cell detection and spatial analysis on digitized biopsy slides from 53 early-stage pediatric NLPHL patients to quantitatively assess LP cell histomorphometry. We found that poor responding patients had significantly fewer LP cells per cluster and lower LP cell density than good responding patients. In our exploratory analysis, we found no correlation between Fan classes or B cell pattern variables and therapy response. We hypothesize that the relationship between poor treatment response and decreased LP cell density may be explained by differences in LP cell proliferation.
Keyphrases
- early stage
- end stage renal disease
- ejection fraction
- hodgkin lymphoma
- newly diagnosed
- cell proliferation
- deep learning
- chronic kidney disease
- risk factors
- prognostic factors
- induced apoptosis
- single cell
- cell therapy
- peritoneal dialysis
- machine learning
- squamous cell carcinoma
- stem cells
- signaling pathway
- radiation therapy
- patient reported outcomes
- cell death
- mesenchymal stem cells
- cross sectional
- quantum dots
- smoking cessation
- pi k akt
- locally advanced
- replacement therapy
- loop mediated isothermal amplification