# Risk factors of negative prostate biopsies: A CHAID decision tree analysis

## TL;DR

This study identifies risk factors for negative prostate biopsies using a decision tree analysis, focusing on prostate volume and MRI scores.

## Contribution

The study introduces a CHAID decision tree analysis to identify risk factors for negative prostate biopsies, particularly highlighting prostate volume and PIRADS score.

## Key findings

- 69.8% of patients had positive biopsies for prostate cancer, while 36.9% had clinically significant prostate cancer.
- PSA density and PIRADS score were independent predictive factors for clinically significant prostate cancer.
- A PIRADS score ≤4 combined with prostate volume >46 mL doubled the risk of negative biopsies.

## Abstract

This study aimed to perform a risk analysis of any prostate cancer (Pca) and of clinically significant prostate cancer (csPCa) in a contemporary cohort of prostatic biopsies.

We conducted a retrospective analysis of patients who underwent prostate biopsies in our centre between December 2020 and December 2022. We calculated Pca and csPCa rate (ISUP grade ≥2). Univariate and multivariate regression models were constructed to assess independent predictive factors for Pca and csPCa. We used χ
2 automatic interaction detection (CHAID) for decision tree analysis.

We included 255 patients in the analysis, of whom 69.8% had positive biopsies for Pca and 36.9% for csPCa. Multivariate analysis found PSA density (PSAd) (OR = 1.001) (1.000; 1.001), PIRADS score (OR = 1.393) (1.234; 1.571) as independent predictive factors of csPCa. For the detection of any PCa, CHAID analysis revealed that patients with PIRADS score ≤4 doubled the risk of negative biopsies (from 22.6% to 54.3%) when the prostate volume was >46 mL.

For patients with a PIRADS ≤4, a large prostate volume (>46 mL) was a predictor of negative biopsies, independently of PSAd. MRI interpretation and targeting in these patients should therefore be performed with particular caution.