Imagerie par résonance magnétique et cancer de la prostate

Magnetic resonance microscopy may enable distinction between normal histomorphological features and prostate cancer in the resected prostate gland

OBJECTIVES

To determine imaging protocol parameters for characterization of prostate tissue at histological length scales.

MATERIAL AND METHODS

Rapid acquisition with relaxation enhancement, spin echo and gradient echo fast low angle shot data were acquired using ex vivo 3-Tesla or 7-Tesla magnetic field strengths from fresh prostatectomy specimens (n = 15) obtained from either organ donor or patients with prostate cancer (PCa). To achieve the closest correspondence between histopathological components and magnetic resonance imaging (MRI) results, in terms of resolution and sectioning planes, multiple high-resolution imaging protocols (ranging from a few minutes to overnight) were tested. Ductograms were generated as part of image post-processing. Specimens were subsequently submitted for histopathological evaluation.

RESULTS

A total of seven imaging protocols were tested. Ex vivo 7-Tesla MRI identified normal components of prostate glands, including ducts, blood vessels, concretions and stroma at a spatial resolution of 60 × 60 × 60 μm³ to 107 × 107 × 500 μm³ . Malignant glands and nests of tumour cells identified at 60 × 60 × 90 μm³ were highly similar to low-magnification (×2) histopathology. Ductograms enhanced the differentiation between benign and malignant glands. The results of the present study were encouraging, and further work is warranted with a larger sample size.

CONCLUSION

We showed that critical histopathological features of the prostate gland can be identified with high-resolution ex vivo MRI examination and this offers promise that MRI microscopy of PCa will ultimately be possible in vivo.

[Value of MRI in detection of extracapsular extension in prostate cancer: a prospective study comparing imaging and histology]

OBJECTIVE

To establish MRI's performances for the detection of extracapsular progression of prostate carcinoma, in a single center, analyzing the correlation between MRI imaging and histological analysis of prostate specimen.

METHODS

From February 2008 to June 2012, all the patients selected for prostatectomy had a pre-operative MRI. Diffusion, T2 and dynamic T1 with gadolinium injection sequences were realized on a 1.5T-MRI with external antenna. All imaging data was analyzed by a specialized radiologist. Prostate specimens were histologically analyzed throughout large blades for utmost topographic comparison. The histological TNM was compared to the MRI data. MRI's capacity in determining the existence and the size of extracapsular progression was studied.

RESULTS

One hundred and fifty-eight patients (median age 62 years old, mean PSA 8.6 ng/mL) were included, among which 45% of d'Amico low risk and 55% of intermediate and high risk. Histological results were 63% of pT2 and 37% of pT3. MRI's sensibility and specificity for detecting extracapsular progression were 0.30 and 0.85 respectively (PPV 0.54; NPV 0.67), with a 65% accuracy. In the low risk group, sensibility equaled to 0.16.

CONCLUSION

In our experience, MRI results were not reliable to influence the choice of treatment. It should be executed by expert radiologists, who are still very few.

A new sign for detection of extracapsular extension of prostatic adenocarcinoma: sum of positive sextants per lobe combining interpretation of magnetic resonance imaging and prostate biopsy results

BACKGROUND

The objective of this study was to assess the possibility of improving the reliability of preoperative detection of extracapsular extension (ECE) in each prostate lobe by using a new sign called sum of positive sextants per lobe (SPS-L), combining interpretation of MRI and prostate biopsy results.

PATIENTS AND METHODS

We reviewed the charts of 590 patients undergoing radical prostatectomy between 2002 and 2007. All patients were assessed by preoperative 1.5 Tesla MRI using an integrated endorectal and pelvic phased array coil. A sextant was defined as "positive" when tumor infiltration was observed on a biopsy sample or a pathologic image was observed on MRI (0 = absent, 1 = present). A score, called sum of positive sextants per lobe (SPS-L), was defined as the sum of positive sextants on biopsy samples and positive sextants on MRI (MRI-L) for each lobe. A second score taking into account the presence or absence of ECE visualized on MRI (T3 MRI-L + SPS-L) was also tested for the detection of ECE per lobe.

RESULTS

On multivariate analysis, the SPS-L and T3 MRI-L + SPS-L scores were significantly higher in the presence of ECE and extensive ECE (P < .0001). The areas under receiver operating characteristic (ROC) curves were significantly greater for the T3 MRI-L + SPS-L score than for the positive biopsy result per lobe (PB(+)/L) rate (P < .0001).

CONCLUSION

The use of indirect signs (SPS-L) associated with direct signs (T3 MRI) allows the preoperative detection of ECE per lobe by endorectal 1.5 Tesla MRI with high sensitivity.

[Functional MRI by general radiologists in prediction of side-specific extracapsular extension after radical prostatectomy: what value in daily practice?]

OBJECTIVE

To evaluate the performance of functional MRI (FMRI) performed by general radiologists (GR) in detection of side-specific extracapsular extension (SSECE) prostate cancer (PCa).

METHODS

We retrospectively analyzed 79 patients who underwent FMRI with pelvic phased array coil before radical prostatectomy (RP) performed at University Hospital (UH) of Nîmes. Twelve GR (including three from UH) interpreted the images during their daily practice. FMRI results were dichotomized as positive or negative and confronted to pathological reports for SSECE and side-specific seminal vesicle invasion (SSSVI), with calculation of diagnostic values. The influence of interval between biopsy and FMRI, diffusion-weighted sequence (DWS) and intensity of FMRI, on the diagnostic performance were assessed by Fisher's exact test.

RESULTS

A SSECE and a SSSVI were observed at FMRI and pathology respectively on 14 (8.8%) and 38 (24.1%) prostate lobes, and on six (3.8%) and seven (4.4%) prostate lobes. The sensitivity, specificity, positive and negative predictive values of FMRI for SSECE were respectively 24%, 96%, 64% and 80%; and for SSSVI were 14%, 97%, 17% and 96% respectively. The time between biopsy and FMRI, intensity of FMRI and DWS, did not influence the sensitivity and specificity of fMRI at Fisher test.

CONCLUSION

This study found that preoperative prostate FMRI performed by GR has good specificity but poor sensitivity in predicting SSECE on pathological reports.

[Active surveillance for prostate cancer: usefulness of endorectal MR at 1.5 Tesla with pelvic phased array coil in detecting significant tumors]

PURPOSE

To describe and assess MRI signs of significant tumor in a series of patients who all underwent radical prostatectomy and also fulfilled criteria to choose active surveillance according to French "SurAcaP" protocol.

PATIENTS AND METHODS

The clinical reports of 681 consecutive patients operated on for prostate cancer between 2002 and 2007 were reviewed retrospectively. All patients had endorectal MR (1.5 Tesla) with pelvic phased array coil. (1.5 T erMR PPA). Sixty-one patients (8.9%) fulfilled "SurAcaP" protocol criteria. Preoperative data (MR+core biopsy) were assessed by comparison to whole-mount step section pathology.

RESULTS

85.3% of the 61 patients entering SurAcaP protocol had significant tumor at pathology. (Non Organ Confined Disease (Non OCD)=8.2%, Gleason sum score>6=39.2%). A new exclusion criterion has been assessed: T3MRI±NPS>1 as a predictor tool of significant tumor. ("T3MRI±NPS>1"=Non OCD at MR±number of positive sextants involved in tumor at MR and/or Core Biopsy > to 1). Sensitivity, specificity, PPV, NPV of the criterion "T3MRI±NPS>1" in predicting significant tumor were, respectively: 77%, 33%, 86%, 20%. Adding this criterion to other criteria of the "SurAcaP" protocol could allow the exclusion of all Non OCD, and a decrease in Gleason sum Score>6 rates (20%).

CONCLUSION

Endorectal MR at 1.5 Tesla with pelvic-phased array coil should be considered when selecting patients for active surveillance in the management of prostate cancer. A criterion based upon MR and core biopsy findings, called "T3MR±NSP>1" may represent an exclusion citeria due to its ability to predict significant tumor.

[Diffusion-weighted MR imaging of the prostate]

Functional imaging complements T2-weighted imaging in the evaluation of the prostate. The most frequently used techniques are diffusion-weighted imaging and perfusion imaging following the intravenous administration of contrast material. Perfusion imaging has high sensitivity and moderate specificity, even when using a quantitative technique, because contrast kinetics in some cases of peripheral zone prostatitis and benign transition zone hyperplasia may simulate cancer. Diffusion-weighted imaging is currently under evaluation but appears to be preferable to dynamic perfusion MR imaging because of its higher specificity and simpler acquisition. Functional imaging of the prostate is performed to detect cancers missed on biopsies or evaluate the volume of a newly diagnosed clinically localized cancer to assist in therapy selection. Future applications for image-guidance of targeted therapies to the tumor are currently investigational.

[Oncologic and functional outcomes after radical prostatectomy in T3 prostate cancer]

According to current literature, the gold standard treatment for T3 prostate cancer is the combination of radiotherapy and extended hormonotherapy. Clinical staging based on digital rectal examination seems useless nowadays, since 20% of T3 prostate cancer is overevaluated during physical examination. Prostatic MRI is extensively needed to evaluate extraprostatic extension during preoperative work-up. EAU guidelines recommend radical prostatecomy only in selected patients: less than or equal to cT3a, PSA less than 20 ng/ml and biopsy Gleason score less than or equal to 8. Carcinologic control obtained after radical prostatectomy is variable from one series to another, with biochemical free survival rate at 5, 10 and 15 years that range from 45 to 62%, 43 to 51%, and 38 to 49%. Specific survival rates at 5, 10 and 15 years are, respectively, of 84 to 98%, 85 to 91% and 76 to 84%. Surgical margins rate differ from 22 up to 61% corresponding to several operative techniques and surgeon's own experience. Regarding urinary continence, functional outcomes are in line with those of prostatectomy for localized prostate cancer. Considering erectile dysfunction, rates are linked with the type of surgery, which can be extensive or not. There is no impact on overall or specific survival of neoadjuvant treatments. One current question remains the efficacy of early adjuvant treatment after prostatectomy, especially adjuvant irradiation. Radical prostatectomy can be considered in selected cases as a viable alternative first-line treatment option. However, patients have to be warned that they may undergo complementary treatments during the postoperative course of the disease.