Magnetic resonance imaging (MRI) is a standard diagnostic procedure for evaluation of prostate cancer (PCa) and good image quality is essential. This could potentially be improved by adding three-dimensional T2 weighted imaging (3D-T2WI) to the standard protocol as it enables reconstruction of images in any given plane, potentially allowing us to better demonstrate the tumour/capsular interface. Independent of scanner parameters, patient factors such as bowel peristalsis and rectal distension by gas/stool can also affect image quality. The Prostate Imaging-Reporting and Data System (PI-RADS) guidelines acknowledge that there is currently lack of evidence in both areas and therefore no direct recommendation on patient preparation is given.
To compare 3D-T2WI with conventional 2D-T2WI in assessment of local staging of PCa. We also examined the effect of anti-peristaltic agent hyoscine butylbromide (HBB) and rectal distension on the quality of anatomical and functional MR imaging of prostate.
Patients and methods
All studies were performed at 3T MRI scanner using PI-RADS recommended protocol including T2 weighted imaging (T2W), T1 weighted imaging (T1WI), diffusion weighted imaging (DWI) with maximum b-value at 2000s/mm2, apparent diffusion coefficient (ADC) maps and dynamic contrast enhanced (DCE) images.
The first study included 75 patients undergoing MRI before radical prostatectomy. PI-RADS ≥4 lesions were assessed for extracapsular extension (ECE) on 2D-T2W images using a 5-point Likert scale and the length of tumour prostatic capsular contact. A second read using additional 3D-T2W images and reformats evaluated ECE and the maximal 3D capsular contact length and surface.
The second study included 87 patients receiving 20 mg HBB prior to scanning (HBB group) and 86 patients who did not (non-HBB group). Two radiologists evaluated image quality of T2WI, DWI and ADC maps, using a 5-point Likert scale. DWI was further assessed for distortion and artefact, and T2WI for the presence of motion artefact or blurring. DCE image quality was assessed by recording the number of corrupt data points from the curve.
In the third study, images of 173 patients were evaluated by two radiologists. Rectal volumes were calculated and a subjective assessment of rectal distension was made using a 5-point Likert scale (1 = no stool/gas, 5 = large amount of stool/gas). DWI was scored for image quality, distortion and artefact. T2W images were evaluated for image sharpness and the presence of motion artefact. The stability of the DCE was assessed as in the second study.
In the first study a total of 106 lesions were analysed for the presence of ECE. ECE was present in 54% of lesions with PI-RADS score of ≥ 4 and in 64% of all patients. ECE was reported as focal in 28% and established in 72%. In terms of subjective reads, sensitivity and specificity for 3D-T2 was 75.4% versus 64.9% for conventional 2D-T2 (p = 0.06), respectively and 83.7% versus 85.7% (p = 0.71) for conventional 2D-T2, respectively. 3D-T2 reads showed a significantly higher mean Likert score of 3.7 ± 1.4 versus 3.3 ± 1.4 (p < 0.01) in ECE positive lesions and non-significantly lower Likert score for ECE negative lesions with average score of 1.5 ± 1 versus 1.6 ± 0.9 (p = 0.27) as compared to 2D-T2. 3D-contact assessment had significantly higher sensitivity (73.7%) than 2D-contact (59.6%, p = 0.03), whilst maintaining the same specificity of 87.8% (p = 1). 3D-surface area evaluation increased sensitivity from 59.6% to 82.5% (p < 0.01) but this was associated with a loss of specificity from 87.8% to 71.4% (p < 0.01). High grade group (HGG) tumours showed significantly higher ECE prevalence (88%) than low grade group (LGG) at 44%, p < 0.01. In addition, in HGG a PPV of 90.9% was achieved at ≥ 5 mm, conversely for LGG, a PPV of 90.4% was achieved at a cut-off at ≥ 12.5 mm.
The second study showed significantly higher T2W image quality in the HBB group than in the non-HBB group (3.6 ± 1.1 versus 2.8 ± 0.9); p = < 0.01. The HBB group also showed significantly less T2W motion and T2W blur than the non-HBB group (23.0% and 51.7% versus 53.5% and 83.7%, respectively; p = < 0.01). However, there was no significant improvement in DWI or ADC image quality, or DWI degree of distortion or artefact. There was a trend towards a lower number of corrupted data points from the contrast curve (2.5 ± 2.4 versus 3.7 ± 2.6), p = 0.05.
The third study showed a significant correlation between increased rectal distension and both reduced DW image quality (r = -0.628, p < 0.01), and increased DW image distortion (r=0.814, p < 0.01). There was a significant trend for rectal distension to increase artefact at DWI (r = 0.154, p = 0.04) and to increase motion artefact on T2 (p < 0.01). There was no relationship between rectal distension and T2-sharpness or DCE image quality. 63 patients underwent lesion-targeted biopsy post MRI, there was a trend to higher positive predictive values in patients with minor rectal distension compared to those with moderate/marked distension.
The use of 3D-T2 showed significantly increased sensitivity and confidence in calling ECE over conventional 2D-T2. In addition, capsular contact length threshold was significantly longer for LGG than HGG lesions. Rectal distension had a significant negative effect on the quality of both T2W and DW images whereas the administration of HBB was significantly associated with improved image quality of T2W images.
Consideration should therefore be given to 3D-T2W imaging for improved detection of ECE as well as to bowel preparation and the routine use of HBB to optimise image quality of mpMRI of prostate.