Logistic regression analysis of low grade spindle cell lesions. A cytologic study

Area of Science:

• Cytopathology
• Surgical Pathology
• Oncology

Background:

• Low grade spindle cell lesions encompass a diverse group of benign and malignant neoplasms.
• Accurate cytologic diagnosis is crucial for appropriate patient management and treatment.
• Distinguishing between benign and malignant spindle cell lesions can be challenging based on traditional criteria.

Purpose of the Study:

• To identify key diagnostic cytologic criteria for various low grade spindle cell lesions.
• To statistically validate features predictive of malignancy, specifically synovial sarcomas, benign neural tumors, and reparative lesions.

Main Methods:

• Review of cytologic specimens from 20 synovial sarcomas, 18 benign neural tumors, 10 reparative lesions, 24 other benign, and 27 malignant low grade spindle cell lesions.
• Systematic coding of specimens for 30 cytologic variables.
• Logistic regression analysis to identify variables predictive of specific lesion types.

Main Results:

• High cellularity, short spindle cells, small nucleoli, and absence of tissue culture appearance were significant criteria for malignant neoplasms.
• Synovial sarcomas were characterized by tissue fragments, high cellularity, absence of long filamentous cells, and myxoid background.
• Benign neural tumors showed fibrillar ground substance and absence of ovoid/round nuclei; reparative lesions exhibited tissue culture appearance.

Conclusions:

• Statistical analysis significantly narrowed down the cytologic features for evaluating low grade spindle cell lesions.
• A few key features are highly significant for differentiating between malignant and benign spindle cell lesions.
• These statistically validated criteria can improve diagnostic accuracy in cytopathology.