Introduction: Multiparametric magnetic resonance imaging (mpMRI) has improved the accuracy of prostate cancer detection. Prior data suggests that in-bore biopsy may have improved accuracy over other fusion modalities, such as MRI-US fusion using software registration. We sought to compare the relative impact of selected MRI spatial parameters on prostate cancer detection between MRI-US fusion and MRI in-bore biopsy.
Methods: Patients with at least one PIRADS 4 or higher lesion on mpMRI that underwent MRI-US or MRI in-bore biopsy at one institution between 2013 and 2020 were included. The majority of patients had 3T MRI with endorectal coil. Patients on active surveillance were excluded. The primary endpoint was the detection of Gleason Grade Group 2 or higher (GG2+) on targeted biopsy on a per lesion basis. Propensity score matching was performed, and the overall GG2+ detection rate was compared using the Chi-square test. Generalized estimated equations were used to test interaction terms between biopsy modality and pre-specified variables: maximum lesion dimension, lesion volume, prostate volume, and lesion location (transitional zone vs. peripheral zone). Inclusion of number of biopsy cores to each model was also pre-specified. Analysis was performed using SAS, version 9.4.
Results: 53 MRI-US fusion patients and 133 MRI in-bore patients were evaluated for matching. 68 patients and 84 lesions were then matched for clinical and MRI factors, and all prebiopsy covariates were balanced. Overall GG2+ detection rate was higher amongst lesions sampled with in-bore biopsy compared to MRI-US fusion biopsy (59.5% v 45.2%, p = 0.190). Lesion maximum dimension, lesion volume, prostate volume, and lesion location did not affect GG2+ detection for in-bore biopsy technique. However, the interaction between biopsy modality and lesion volume (OR 1.2, CI 1.03-1.43, p-value 0.024) and between modality and prostate volume were significant for the detection of GG2+ with MRI-US fusion biopsy (OR 0.97, CI 0.96-1.00, p-value 0.030).
Conclusions: In a propensity score matched cohort of patients without prostate cancer undergoing MRI-US fusion biopsy or in-bore biopsy for PIRADS 4 or 5 lesions, overall GG2+ detection rate was higher for patients undergoing in-bore biopsy, which was independent of lesion volume or prostate size. GG2+ detection rate was lower for MRI-US fusion biopsy patients with smaller lesions or larger prostates, which may reflect inaccurate coregistration or increased tissue deformation with this approach.
.jpg)
.jpg)
