The concordance between the volume hotspot and the grade hotspot: a 3-D reconstructive model using the pathology outputs from the PROMIS trial.

Résumé 0

The rationale for directing targeted biopsy towards the centre of lesions has been questioned in light of prostate cancer grade heterogeneity. In this study, we assess the assumption that the maximum cancer Gleason grade (Gleason grade hotspot) lies within the maximum dimension (volume hotspot) of a prostate cancer lesion. 3-D histopathological models were reconstructed using the outputs of the 5-mm transperineal mapping (TPM) biopsies used as the reference test in the pilot phase of Prostate Mri Imaging Study (PROMIS), a paired validating cohort study investigating the performance of multi-parametric magnetic resonance imaging (MRI) against transrectal ultrasound (TRUS) biopsies. The prostate was fully sampled with 5 mm intervals; each core was separately labelled, inked and orientated in space to register 3-D cancer lesions location. The data from the histopathology results were used to create a 3-D interpolated reconstruction of each lesion and identify the spatial coordinates of the largest dimension (volume hot spot) and highest Gleason grade (Gleason grade hotspot) and assess their concordance. Ninety-four men, with median age 62 years (interquartile range, IQR= 58-68) and median PSA 6.5 ng ml(-1) (4.6-8.8), had a median of 80 (I69-89) cores each with a median of 4.5 positive cores (0-12). In the primary analysis, the prevalence of homogeneous lesions was 148 (76%; 95% confidence interval (CI) ±6.0%). In all, 184 (94±3.2%) lesions showed concordant hotspots and 11/47 (23±12.1%) of heterogeneous lesions showed discordant hotspots. The median 3-D distance between discordant hotspots was 12.8 mm (9.9-15.5). These figures remained stable on secondary analyses using alternative reconstructive assumptions. Limitations include a certain degree of error within reconstructed models. Guiding one biopsy needle to the maximum cancer diameter would lead to correct Gleason grade attribution in 94% of all lesions and 79% of heterogeneous ones if a true hit was obtained. Further correlation of histological lesions, their MRI appearance and the detectability of these hotspots on MRI will be undertaken once PROMIS results are released.