Multi-parametric MRI without artificial erection for preoperative assessment of primary penile carcinoma: A pilot study on the correlation between imaging and histopathological findings

Abstract

Purpose: We aimed to evaluate the diagnostic potential of non-erectile multi-parametric magnetic resonance imaging (mpMRI) for preoperative assessment of primary penile squamous cell carcinoma (SCC).

Method: Twenty-five patients who underwent surgery for penile SCC were included. Preoperative mpMRI without artificial erection was performed in all patients. The preoperative MRI protocol consisted of high-resolution morphological and functional sequences (diffusion-weighted imaging and dynamic contrast-enhanced MRI perfusion) covering the penis and lower pelvis. T and N staging, according to the 8th edition of the Union for International Cancer Control TNM classification, as well as the largest diameter and thickness/infiltration depth of the primary lesions were determined in all patients. Imaging data were retrospectively collected and compared with the final histopathology reports.

Results: Very good agreement was observed between MRI and histopathology for the involvement of corpus spongiosum (p = 0.002) and good agreement was observed for the involvement of penile urethra and tunica albuginea/corpus cavernosum (p < 0.001 and p = 0.007, respectively). Good agreement was observed between MRI and histopathology for overall T staging and weaker, but still good agreement was observed for N staging (p < 0.001 and p = 0.002, respectively). A strong and significant correlation was observed between MRI and histopathology for the largest diameter and thickness/infiltration depth of the primary lesions (p < 0.001).

Conclusions: Good concordance was observed between MRI and histopathological findings. Our initial findings indicate that non-erectile mpMRI is useful in preoperative assessment of primary penile SCC.

Keywords: Histopathology; Multi-parametric MRI; Primary penile carcinoma; Surgery.

Fig. 1

Principles of the multi-parametric magnetic resonance imaging protocol for preoperative evaluation of primary penile carcinoma. The interpretation is based on the integrated assessment of tumor morphology (upper row), cellularity (middle row), and vascularization (lower row).

Fig. 2

Preoperative multi-parametric magnetic resonance imaging (mpMRI) and histopathological findings in a patient with primary penile carcinoma. Sagittal T2-weighted sequence (A), diffusion-weighted imaging (B), and apparent diffusion coefficient (ADC) map (C) shows a large T2-tumor invading glans (dashed line). Tumor has homogenous, low diffusion (B, C) (ADC = 600 μmm²/s) and a rapid, initial permeability with wash-out on dynamic contrast-enhanced MRI perfusion (D–G; D, acquisition image; E, K^trans; F, k_ep; G, perfusion curve) (arrows), and delayed, contrast-enhanced T1-weighted Dixon sequence (H, arrow). Photomicrograph of whole-mount hematoxylin and eosin-stained section with outlined tumor invading glans/corpus spongiosum (I, dashed line). There was no involvement of tunica albuginea or corpus cavernosum. Photomicrographs of histologic specimen confirm infiltration of moderately to poorly differentiated keratinizing squamous cell carcinoma, usual type (J, magnification × 10; K, magnification × 20).

Fig. 3

Scatterplot demonstrates comparison of largest tumor diameter (A) and infiltration depth (B) as determined by multi-parametric magnetic resonance imaging (mpMRI) and histopathology. Correlation coefficient (ρ) and p-value from Spearman rank-order correlation are shown. An example on correlation between mpMRI findings (C-F) and histopathology is also shown (G–I). Tumor is difficult to outline on T2-weighted sequence (C). However, diffusion-weighted sequence (D) and apparent diffusion coefficient map (E) clearly illustrate tumor extension (arrows). Tumor extension is larger on K^trans perfusion map (F), probably due to peritumoral edema and inflammation. Photomicrograph of whole-mount hematoxylin and eosin-stained section with 6 × 18 mm T2-tumor invading glans (G, dashed lines). Photomicrographs of histologic specimen confirm infiltration of poorly differentiated keratinizing squamous cell carcinoma, usual type (H, magnification × 10; I, magnification × 20).