Prostate tumor size often underestimated by MRI

January 8, 2021
Jason M. Broderick

The investigators drew specific attention to the impact of their findings on prostate ablation therapy.

Magnetic resonance imaging (MRI) frequently underestimates tumor size in patients with prostate cancer, according to a study published in the Journal of Urology.1,2

In the study, the underestimation of tumor size by MRI occurred most frequently when measuring smaller tumors and those with a low PI-RADS, a score used to measure the likelihood that a suspicious area is a cancer.

MRI is a key tool in the prostate cancer armamentarium. With this study, the investigators, led by researchers at the UCLA Jonsson Comprehensive Cancer Center, drew specific attention to the impact of their findings on prostate ablation therapy.

Ablation techniques, such as cryotherapy, HIFU, and laser ablation, are highly dependent on the accuracy of tumor size estimates. Thus, underestimation of tumor size creates the high potential for undertreatment.

“Multiparametric magnetic resonance imaging frequently underestimates pathological tumor size and the degree of underestimation increases with smaller radiologic tumor size and lower PI-RADSv2 scores. Therefore, a larger ablation margin may be required for smaller tumors and lesions with lower PI-RADSv2 scores. These variables must be considered when estimating treatment margins in focal therapy,” the authors wrote in their conclusion.

The analysis included 461 lesions in 441 patients with concordant 3T mpMRI and pathological region of interest. The median patient age was 62 years (range, 41-80), the median PSA was 6.3 ng/ml (range, 0.6-139.2), and the median PSA density was 0.174 (range, 0.02-3.395).

Based on ISUP grade distribution, the highest Gleason Grade Group (GG) distribution was GG1 in 18% of patients, GG2 in 38.18%, GG3 in 14.53%, GG4 in 19.31%, and GG5 in 9.98%.Overall, 17.14% of patients had a PI-RADSv2 score of 3 (cancer equivocal), 44.25% had a score of 4 (cancer likely), and 38.61% had a score of 5 (cancer highly likely).

The mean radiologic tumor size was 1.572 cm and the mean pathologic tumor size was 2.373 cm.

“Radiologic tumor size consistently underestimated pathological tumor size regardless of the preoperative covariates, and the degree of underestimation increased with smaller radiologic tumor size and lower PI-RADSv2 scores,” the authors wrote.

Compared to GG1, PTS was significantly larger for biopsy GG5 compared to GG1(mean change, 0.37 cm; P = .014),for PI-RADSv2 5 lesions compared to PI-RADSv2 4 lesions (mean change, 0.26, P = .006), and higher prostate specific antigen density.

According to biopsy GG and radiologic covariates, the correlations between RTS versus PTS were low overall, with correlation coefficients that ranged between 0.1 and 0.65.

Summarizing the significance of the findings in an accompanying editorial, Joseph L. Chin, MD, FRCSC, Western University, London Ontario, Canada, wrote, “With this type of magnetic resonance versus surgical pathology size comparison study, due to challenges in co-registration and potential pitfalls in spatial resolution and sectioning geometrics as discussed by the authors, urologists performing focal therapy cannot assume they can fully compensate for any radiologic underestimation of tumor size with literal margin expansion of the magnetic resonance target (by the amounts reported in this paper).

“Simple margin expansion would only be completely valid if the centroid of the lesion on MRI aligned exactly with the centroid of the lesion on histology, and if the shape of the lesion on MRI matched exactly with that histology.

“Nevertheless, the tendency of magnetic resonance to underestimate the actual tumor size (in the majority of cases) and the tendency for smaller tumors and lower Gleason Grade cancers to have greater degrees of underestimation are important messages.”


1. Pooli A, Johnson DC, Shirk J, et al. Predicting pathological tumor size in prostate cancer based on multiparametric prostate magnetic resonance imaging and preoperative findings. J Urol. 2021;205(2):444-451. doi: 10.1097/JU.0000000000001389

2. Heady D. UCLA Research Brief. MRI frequently underestimates tumor size in prostate cancer. Published online January 7, 2021. Accessed January 8, 2021.

3. Chin JL. Editorial Comment. J Urol. 2021;205(2):451. doi: 10.1097/JU.0000000000001389.01

Related Content:

Prostate Cancer