Ed Weber, M.D., Editor
January- February, 2015
A technically well performed and expertly interpreted multiparametric MRI combining standard T2 and diffusion weighted, and contrast enhanced imaging is the most accurate method of assessing the true size and location of cancer within the prostate gland short of a pathologist’s assessment of a whole-mount prostate specimen. A biopsy targeted into a region of high suspicion on a mpMRI confers a greater likelihood of establishing the true Gleason score of the cancer as compared to a random 12-core TRUS guided biopsy. Therefore, it is not surprising that mpMRI is assuming a primary role in the selection of men for active surveillance, and being evaluated as the preferred tool for monitoring during the course of active surveillance (AS).
Maxwell Meng, MD, (UCSF) at the meeting of the Society of Urologic Oncology (SUO), Bethesda, December 2013, weighed in on this topic with his presentation: “Role of MRI: Can MRI replace the biopsy or not?” He answered his own question with data supporting the substitution. The thrust of his remarks was summarized for UroToday: “The potential benefits of MRI in AS include improvement in preoperative risk stratification, increased acceptance of AS by low risk patients, identification of candidates for early intervention, and improved monitoring during follow-up.”
MRI AND PATIENT SELECTION FOR ACTIVE SURVEILLANCE:
An abundance of studies has established that patients initially considered as good candidates for AS based on a TRUS-guided biopsy undergo a 20-30% reclassification on a confirmatory biopsy or on a prostatectomy specimen. Dr. Meng (UCLA) cited his study of 500 men thought to have low-risk disease on initial work-up. At 6 – 12 months thereafter these men underwent an MRI and biopsy. Meng confirmed what others have found: a baseline MRI score indicating low suspicion for cancer (≤ 2 on a scale of 1, definitely no tumor, to 5, definitely tumor) was associated with an excellent negative predictive value (0.98-1.0) for predicting subsequent upgrading and the presence of significant cancer. Of the men classified as low-risk at the initial assessment, 38% showed no region of suspicion on subsequent MRI. Rebiopsy in this group led to a reclassification rate of only 3.5%.
The study supported the added value of a MRI guided biopsy as compared to the standard 12-core sampling. In 582 patient (320 with a negative first 12-core biopsy) the MRI targeted biopsy upgraded 32% overall; 22% in men with an initial Gleason score of ≤3+4.
The superiority of a targeted MRI guided/fusion biopsy over the standard 12-core was again supported in a presentation at the 2014 SUO meeting in a report by Dr. Salami (The North Shore Institute for Urology, Hofstra North Shore, N.Y.). The study focused on 140 men in whom the initial TRUS biopsy(s) was negative. The second biopsy session employed an MRI fusion guided biopsy and a random 12-core TRUS guided biopsy. The cancer detection rate was 65% for those men with one prior negative biopsy. Although the detection of cancer of any grade was similar with both techniques, i.e., ~50%, the fusion biopsy detected significant cancer in 48% vs 31%. The standard technique missed 19 of 91 clinically significant cancers whereas the fusion biopsy missed only 3.5%.
As seen in previous studies, the cancers missed by the TRUS technique were predominantly located in the anterior and central portions of the gland, regions reached with difficulty in the standard transrectal biopsy but well within the range of a transperineal approach.
Considering the studies cited above, and supported by an abundance of other studies with the same conclusion, patient selection for AS based on an mpMRI targeted biopsy would appear to be the preferred technique.
Pathologic diagnosis of the specimen obtained at prostatectomy is unquestionably the reference standard for comprehensive information about the location, size, and Gleason score of prostate cancer. But an initial prostatectomy can not be performed and then later choose AS for only favorable candidates. However, the MRI-guided method has the better record than TRUS-guided for determining both significant cancer (i.e. those greater than 1 cm diameter) and cancer prompting reclassification (and possible intervention).
A comparison of the index tumor biopsied using image-guided MRI/fusion with the step sectioned prostatectomy specimen in 135 men was reported by Ukimura, Eur Urol Sept 2014. The index tumor (IT) was “defined as lesion with the highest Gleason score or largest volume or extraprostatic extension. Their conclusion: The concordance between IT location on biopsy and RP specimens was 95% (128-125). … [and] “The concordance of primary Gleason pattern between targeted biopsy and RP specimen was 90%.”
Mullins, Partin, Epstein et al., at Johns Hopkins, (BJU Int. 2013 June) reported that mpMRI “demonstrated excellent specificity (0.974) and a negative predictive value (0.897) for the detection of pathological index lesions.”
In his report in BJU Int. 2014, “Mutiparametric magnetic resonance imaging (MRI) and active surveillance for prostate cancer: future directions,” Mullins stated: “It is currently the practice at our institution [Johns Hopkins] to refer eligible men for multiparametric MRI before enrollment in AS.”
MRI IN MONITORING PATIENTS DURING ACTIVE SURVEILLANCE:
The application of monitoring the course of AS with multiparametric MRI alone (with biopsy only as indicated) is a work in progress requiring evaluation and validation. However, substituting mpMRI for sequential TRUS-guided biopsies has great potential and is well worth pursuing. Routine monitoring biopsies are recommended at various intervals in different regimens ranging from yearly to as long as every four years. Considering that many well-selected men for AS may be followed without intervention for 7 to 10 years, this substitution would avoid repeated trauma, risk of infection, biopsy related scarring complicating subsequent surgery, and would address patient inconvenience and distaste for biopsies.
In his SUO 2013 presentation Meng offered his opinion on this subject: “The best way to incorporate MRI may be by using an initial MRI targeted biopsy to find clinically significant tumors. This will allow more accurate risk assessment prior to initiation of AS. MRI provides a high NPV [negative predictive value], the ability to avoid the need for confirmatory biopsy, and the potential to identify patients who may have more favorable outcomes based on the absence of MRI-detected lesions.”
MRI studies with low suspicion scores, i.e. scores of 1 -2 , have a high negative predictive value (NPV) for ruling out significant cancers. In various studies the NPVs range from 90%, 96%, 98%, to 100%.
With a baseline MRI as comparison, subsequent MRIs have the interpretative advantage of evaluating patient-specific changes in the size of the index lesion and changes in its density (i.e. the ‘apparent diffusion coefficient’ on which the grade of the lesion is determined). This information could be augmented by the rate of change of the PSA (PSA velocity) and change in the calculated PSA density.
A review article on the role of MRI in AS by Schoots, Klotz et al., Eur Urol. Nov 2014, states, ”There are few data to assess the use of MRI as a monitoring tool during surveillance, so there is a need to define significant disease on MRI and significant changes over time.”
In personal communication, Dr. Klotz, Professor, Department of Surgery, University of Toronto, kindly indicated studies that are in progress attempting to define and validate the issues which are necessary to support MRI at various points in a schema for AS relying on MRI:
- The Canadian ’ASIST’ trial, NCT01354171, currently 80% accrued, designed to evaluate the rate of upgrading of patients with an AS entry Gleason score of 6 to a score higher than 7 by comparing systematic TRUS-guided biopsies with MRI-targeted biopsies at one year;
- The English ‘PROMIS’ trial, NCT01292291, designed to evaluate whether additional TRUS-guided systematic biopsies from regions beyond the MRI index lesion are necessary for diagnostic accuracy compared to MRI-targeted biopsies alone. A second objective is to determine the “proportion of men who could safely avoid biopsy as determined by specificity and negative predictive values;”
- And a third planned large randomized Canadian trial of men with elevated PSA values comparing systematic TRUS biopsies with MRI biopsies only into regions of MRI suspicion.
The use of mpMRI pre-biopsy and its substitution for a TRUS-guided biopsy in monitoring AS are issues currently fraught with controversy. At the recent SUO14 meeting a panel was conducted and moderated by Dr. Klotz in which arguments pro and con were presented: “Should MR be the standard of care before biopsy?” The pro, argued by Dr. Taneja, NYU School of Medicine, closely followed the material presented above.
Dr. Stephenson, Cleveland Clinic, assumed the role of the ‘devils advocate’ saying: not all studies report greater accuracy of MRI over standard systematic biopsy; an MRI targeted biopsy misses 14% of significant cancers and may need to be combined with systematic biopsies; the data regarding upgrading and missing cancer has wide variation when comparing the MRI and TRUS systems; the negative predictive value for MRI in finding significant disease can be as low as 19-48% in comparison to much higher values reported to be in the 90+ range — and at the Cleveland Clinic adding an MRI to guide an initial biopsy MRI raises the cost by 25%.
Clearly future studies will be required to resolve these different positions.
The use of a targeted mpMRI biopsy before patient selection for AS has substantial evidential support. A role for MRI in monitoring patients during AS is not ready as yet to become a recommendation in the NCCN guidelines. But the future for development and refinement in this area is promising. The goal of employing mpMRI to reduce unnecessary biopsies and achieve greater biopsy accuracy in diagnosis is worthy. It would be surprising if future developments did not proceed in that direction.