New Imaging Modalities for Identifying Higher Grade Prostate Cancer in Men on Active Surveillance

Abstract

In this review, we discuss the potential utility of newer imaging modalities including microultrasound and PSMA PET for the detection of clinically significant prostate cancer. These technologies may become helpful adjuncts to mpMRI in the active surveillance setting.

Introduction

In recent decades, the routine use of prostate-specific antigen (PSA) testing has led to an increase in the number of men undergoing prostate biopsy. By some estimates, more than 1 million prostate biopsies are performed annually in the United States alone.[1] Given the large number of these procedures performed throughout the world, significant effort has been spent on determining the most accurate method for performing this procedure with the best rates of cancer detection and lowest cost while maintaining accessibility for patients.

At the present time, prostate biopsy is most commonly performed via an ultrasound-guided, multicore approach, either transrectally (TR) or transperineally (TP). Recently, important clinical guidelines have recommended the use of biopsies targeted to magnetic resonance imaging (MRI) detected lesions, as they have been shown to improve the detection of clinically significant prostate cancer while reducing the detection of non-clinically significant prostate cancer.[2]–[4] As a result, multiple biopsy platforms have been developed that allow for the fusion of MRI and live transrectal ultrasound (TRUS) imaging to augment the ability to detect prostate cancer.[5], [6] While mpMRI has been well-studied and has been shown to improve the likelihood of identifying prostate cancer compared to conventional ultrasound, it has still been found to miss up to 23% of Grade Group 2 or higher lesions.[7] Despite certain benefits, the diagnostic pathway that relies on MRI creates unique challenges for access to such care and potentially excludes men with poor kidney function, claustrophobia, pelvic hardware such as hip replacements, and/or cardiac implants. Furthermore, one must consider the increased cost of care and variable expertise amongst radiologists in interpreting such studies, with one study reporting only 59% agreement for peripheral zone lesions even amongst prostate-specialized readers.[8]

mpMRI has entered prostate cancer active surveillance (AS) programs worldwide as an important adjunct to monitoring patients. While guidelines are only cautiously endorsing its use in AS, clinical implementation is widespread in 2021.[9] Given the above, there is intense research into developing new imaging modalities and optimizing image-fusion strategies for targeting prostate biopsies to the most suspicious regions of interest. New developments in imaging beyond mpMRI and likely in conjunction with mpMRI are on the horizon, and may well improve the detection of higher grade prostate cancer. Such modalities may therefore be increasingly applicable to men considering AS, as well as to men already on surveillance who may harbor more aggressive undersampled elements and we discuss the most promising here.

Micro-ultrasound

In recent years there has been growing interest in the use of high-resolution micro-ultrasound (microUS), a novel imaging modality that operates at 29 MHz compared to the 9-12 MHz frequency of conventional transrectal urological ultrasound. Like conventional TRUS, microUS provides a transrectal modality for targeting regions of interest in real time. Within the context of active surveillance for the prostate cancer population, men can potentially remain in such a program for many years and therefore, undergo many prostate biopsies during that time period. As illustrated in Figure 1, a modality like microUS that improves spatial resolution compared to conventional ultrasound could be used for standard systematic samples as well as for real-time targeting of suspicious lesions during a single biopsy procedure, and thus may be of utility in an AS population.[10]

Figure 1. MicroUS view demonstrating an obvious hypoechoic ROI that proved to be a Gleason 4+4=8 prostate.

The sagittal view shows the increase in resolution noted with microUS, a real-time modality that can be used in lieu of conventional US to guide prostate biopsy needles to suspicious areas based on a 5-point scoring system (PRI-MUS).

The largest scale assessment of microUS for prostate cancer detection was a recently published randomized clinical trial that compared microUS to conventional TRUS for the detection of clinically significant prostate cancer (csPCa) on prostate biopsy.[11] This trial enrolled over 1,600 patients over 5 sites in the US and Canada, and included the very first patients in whom microUS was used clinically in the world. MicroUS is a relatively new platform using a novel angled side-fire transducer at high frequencies that improves resolution by 300% at the cost of limiting depth of penetration to 5cm (90^th percentile of prostate height). The first outcome from this large trial was the development of a standardized reporting protocol termed PRI-MUS (prostate risk identification using micro-ultrasound) designed to help practitioners understand what csPCa looks like on real-time microUS.[12] This 5-point grading system, much like PI-RADS for mpMRI, is not without its limitations, as the PRI-MUS score focuses exclusively on peripheral zone lesions; there is currently no well-established reference for anterior, transition zone, or central zone lesions. Nevertheless, instruction in PRI-MUS scoring, which was implemented mid-trial after PRI-MUS was developed, dramatically improved sensitivity in the microUS arm, especially compared to conventional US.

The overall trial failed to provide a true picture of the utility of microUS in larger part due to the delay in implementation of PRI-MUS, the first-generation nature of the microUS system used, and issues with apical under-sampling in the microUS arm due to prototype probe configuration. Overall, detection of csPCa was not found to be different regardless of modality (microUS or conventional US) at approximately 36% in each arm. However, these issues have since been addressed by the manufacturer and more recent large-scale trials on the commercial microUS instruments have demonstrated its considerable value. In particular, a registry trial of over 1,000 men at 11 sites in Europe, Canada, and the United States concluded that microUS had similar sensitivity and specificity to mpMRI for diagnosing csPCa.[13] A separate prospective, blinded comparison study of 320 men found csPCa detected in the same proportion of men with a microUS-based biopsy strategy as with an mpMRI fusion-based strategy.[14] Refinements in PRI-MUS to include anterior, transition and central zone lesion identification and ongoing work to further improve depth of imaging should only improve on these early results.

There have been some attempts to assess microUS in AS populations, most notably by Eure et al.[15] and Wiemer et al.[16] These authors have assessed patients with known prostate cancer using MRI and microUS in combination. There have been improvements in csPCa detection without increasing the amounts of non-clinically significant cancer found in these studies, the first a small cohort of nine men and the latter a larger study of 159 men, 29% of whom were on AS at the time. microUS indeed appeared additive with MRI in terms of finding csPCa, despite some overlap in ROIs as expected. The results were not however broken down into AS and non-AS populations. The additive nature of microUS and mpMRI is particularly promising given the ability to sample mpMRI lesions with microUS guidance, which has demonstrated improved accuracy in early studies.[17], [18] While more data is required in AS eligible men, this technology has clear potential in both diagnostic and surveillance applications, with or without the addition of microUS-guided MRI-fusion biopsies.

Molecular Imaging

Prostate cancer is a highly heterogeneous disease and appropriate risk stratification at the time of initial diagnosis can be difficult to achieve. While the serum PSA is a tissue-specific and sensitive tumor marker, it does not provide disease localization within the prostate or the location of metastatic spread.[19] PSA-based imaging has been of limited utility up until now because it is a serum protein released into systemic circulation and therefore a poor marker for localizing disease,[20] but the addition of mpMRI and targeted fusion biopsy has increased the ability of biopsies to find the highest-grade cancers, though some higher grade areas are still missed resulting in sampling error.[21] Imaging plays a very important role in the management of prostate cancer patients for diagnosis and staging, during surveillance, and in the settings of biochemical recurrence and metastatic disease and there is certainly a lot to be improved upon as discussed below.

Metabolism-based PET imaging agents, most notably 2-deoxy-2-[¹⁸F]fluoro-D-glucose (¹⁸F-FDG), have been available for many years, but their high cost in relation to poor to moderate sensitivity have limited their use for prostate cancer. There are a number of metabolic radiotracers currently in use that can characterize subsets of prostate cancer including ¹⁸F-sodium fluoride (¹⁸F-NaF), ¹⁸F-FDG, ¹⁸F-fluoroacetate or ¹¹C-acetate, and ¹⁸F-fluorocholine or ¹¹C-choline. Each of these radiotracers has specific applications. For instance, ¹⁸F-NaF targets hydroxyapatite, making it particularly useful to measure active bone remodeling and to detect bony metastases. ¹⁸F-FDG PET is a glucose analog to measure glucose metabolism and is most useful for monitoring response to therapy in metastatic castration-resistant disease, a state in which prostate cancer cells are often highly glycolytically active. ¹⁸F-fluoroacetate or ¹¹C-acetate or choline derivatives target fatty acid synthase and choline kinase, respectively. These PET radiotracers help measure fatty acid metabolism and are useful in detecting biochemical recurrence, but unfortunately are less helpful in the evaluation of primary prostate cancer due to their inability to discriminate malignant from benign prostate tissue.

While currently FDA-approved only for the evaluation of suspected prostate cancer post-treatment recurrence, ¹⁸F-fluciclovine PET/CT (Axumin, 2020 Blue Earth Diagnostics) is an artificial-amino-acid-based radiotracer that has also been shown to accurately identify cancer within the prostate in men with a histopathological diagnosis of prostate cancer, with higher uptake of the radiotracer in intraprostatic tumor foci when compared to normal prostate tissue.[22] Turkbey et al., however, demonstrated that while the mean SUV_max of the tumor was significantly higher than normal prostate tissue, the mean tumor SUV_max did not differ significantly from benign prostatic hyperplasia nodules, making it non-specific for prostate cancer.[23] While Schuster et al. demonstrated a correlation of SUV_max with Gleason score and differences in SUV_max between malignant and benign sextants in a small study of 10 patients, there remained a lack of specificity, highlighting the limitations of ¹⁸F-fluciclovine-based PET/CT. However, the addition of mpMRI to fluciclovine PET increased the PPV from 50% for ¹⁸F-fluciclovine alone and 76% for mpMRI alone to 82% for a combination of all methods, suggesting that the combination could improve diagnostic performance for characterization of a primary prostatic tumor and for biopsy planning.[24] To this point, Elschot et al. aimed to identify the optimal integrated ¹⁸F-fluciclovine PET/MRI protocol for the identification and characterization of primary prostate cancer. In a study of 28 patients diagnosed with high-risk prostate cancer, the authors found that that late-window PET imaging (33–38 min post-injection) was best for differentiating between prostate tumors and benign tissue, as well as between high-grade and low- or intermediate-grade tumors.[25] While this integrated PET/MRI protocol shows promise in the initial characterization of primary prostate tumors, additional study to evaluate its performance across all risk groups is warranted.

PET agents targeting prostate-specific membrane antigen (PSMA) are exciting new molecules that may revolutionize multiple indications for prostate cancer imaging. Furthest in development of the PSMA imaging agents are ⁶⁸Ga-PSMA-11 and ¹⁸F-DCFPyL PET. These PSMA PET agents may aid in the diagnosis and treatment of prostate cancer due to PSMA’s unique properties as a prostate cancer biomarker. PSMA is a type II transmembrane glycoprotein that is primarily expressed in prostatic tissue and over-expressed in primary prostatic tumors and metastatic and hormone-resistant disease. In addition to the diagnostic utility of PSMA, there is ongoing development of PSMA ligands labeled with therapeutic radionucleotides that have shown promising results in the treatment of advanced prostate cancer.[19], [26]

The initial assessments of ⁶⁸Ga-PSMA-11 (sometimes referred to in the literature as ⁶⁸Ga-PSMA-HBED-CC or simply ⁶⁸Ga-PSMA) and ¹⁸F-DCFPyL PET have been in the areas of metastatic disease and BCR. ⁶⁸Ga-PSMA-11 has received the most attention, primarily in Europe and Australia, and has gained acceptance (and United States FDA approval in 2020) as a highly sensitive and specific prostate cancer imaging agent. ⁶⁸Ga-PSMA-11 has been utilized for evaluating the extent of disease, with results comparable to choline PET/CT. ⁶⁸Ga-PSMA-11 PET/CT utilizes a low-molecular weight ligand for human PSMA radiolabeled with ⁶⁸Ga that binds to the extracellular part of the PSMA receptor and are then internalized into the prostate cancer cell. PSMA is not appreciably released into the systemic circulation, and after internalization, the PSMA ligand undergoes endosomal recycling leading to enhanced uptake in tumor cells. These characteristics make for high levels of radiotracer uptake locally, and result in high quality images for diagnostic procedures.[19], [27]

Rahbar et al. investigated the utility of local prostate imaging using PSMA-targeted PET. In patients with high-risk prostate cancer who had undergone ⁶⁸Ga-PSMA-11 PET/CT prior to radical prostatectomy, the Gleason score of each segment of the prostate after surgery was compared to the ⁶⁸Ga-PSMA-11 PET/CT images mapped according to the histologic axis. Based on this analysis, the sensitivity and specificity of ⁶⁸Ga-PSMA PET/CT were each found to be 92%, with a negative predictive value (NPV) of 85% for detection of any prostate cancer, which provided early evidence that the localization of cancer within the prostate can be estimated with high accuracy using PSMA-targeted imaging.[28] Although this was a very different group of patients than the AS population, the high NPV of PSMA PET suggests that it could have a potential role in the evaluation of patients on AS.

Recent data supports the combined use of PET and MRI to increase the accuracy of prostate biopsy through lesion targeting.[29] mpMRI is already considered one of the most useful modalities for the identification of clinically significant primary prostate cancer and molecular imaging with PET has shown promising results to this effect as well. While ¹⁸F-FDG PET shows negative signal in low Gleason score cancers and choline PET performs poorly at differentiating malignant and benign prostatic tissue, ⁶⁸Ga-PSMA PET/CT has demonstrated high specificity for PSMA-expressing tumor cells suggesting potential utility in better identification of localized disease.[26] Investigators have started to consider this idea more seriously, and in a recent study of men with biopsy-proven prostate cancer, simultaneous ⁶⁸Ga-PSMA PET/CT and MRI improved diagnostic accuracy for prostate cancer localization compared with either imaging modality alone.[30] In fact, there have been instances to suggest that PSMA PET/CT may aid in the detection of tumor foci not seen on MRI and missed by prior negative biopsies.[31] As illustrated in Figure 2, mpMRI and PSMA PET can aid in the localization of intra-prostatic lesions to guide targeted biopsy based on axial T2 images or apparent diffusion coeffient maps on mpMRI. Similarly, PSMA-targeted PET demonstrates a strong signial within the prostate at the tumor focus that can be used to guide targeted biopsy. Early prospective investigations are currently being undertaken to evaluation the diagnostic accuracy of PSMA-targeted 18F-DCFPyl PET/CT for detecting clinically signiciant prostate cancer where both modalities are available (NCT03471650).[32] While this study is not being conducted in a selected AS population, it demonstrates the evolution of the use of PSMA-targeted PET/CT from patients with unfavorable high risk and worse disease to those who are in a screening population.

Figure 2. Targeted biopsy performed based on mpMRI and PSMA PET in a man found to have Gleason 4+5=9 prostate cancer.

(A) Axial T2 and (B) axial apparent diffusion coefficient (ADC) map from mpMRI. The target for biopsy in the right base peripheral zone demonstrates low T2 signal and restricted diffusion and was classified as a PI-RADS 5 lesions (red arrows). (C) Axial 18F-DCFPyL PET and (D) Axial 18F-DCFPyL PET/CT images demonstrate very intense uptake throughout the lesion (red arrows). In this case, either the mpMRI data or the PSMA PET data could be used for targeted biopsy.

Simopoulos et al. were the first to describe the use of ⁶⁸Ga-PSMA PET/CT in a patient with prior history of Grade Group 1 disease under AS with four prior negative MRIs and six negative prior biopsies with a rising PSA. ⁶⁸Ga-PSMA PET/CT revealed radiotracer uptake in the prostate near a prior TURP defect and in combination with MRI/TRUS fusion biopsy, the patient was found to have Grade Group 2 disease that was subsequently confirmed at radical prostatectomy.[31], [33] The reliability of ⁶⁸Ga-PSMA PET/CT to detect localized, clinically significant disease was further described by Rhee et al. who assessed the diagnostic performance of mpMRI and ⁶⁸Ga-PSMA-11 PET/CT in 20 men with localized prostate cancer who were imaged prior to radical prostatectomy. In these 20 men, 50 clinically significant lesions were identified from the whole mount histopathological analysis. The sensitivity, specificity, positive predictive value (PPV) and NPV of mpMRI were 44, 94, 81 and 76%, respectively. In contrast, the values with ⁶⁸GaPSMA-11 PET were 49, 95, 85 and 88%, respectively suggesting that ⁶⁸Ga-PSMA-11 PET may augment the ability of mpMRI to detect localized prostate cancer.[33], [34]

Regarding radioactive fluoride radiotracers, Rowe et al. have previously demonstrated the ability of N-[N-[(S)-1,3-dicarboxypropyl]carbamoyl]-4-¹⁸F-fluorobenzyl-L-cysteine (¹⁸F-DCFBC) to bind to prostate tumors with high PSMA expression.[35] In a prospective study, they evaluated the ability of ¹⁸F-DCFBC PET to detect and characterize primary prostate cancer in men undergoing definitive surgery with correlation to pelvic MR imaging and pathology post prostatectomy. In these men, correlation between histology and the imaging modalities was performed on a per-segment and per-lesion basis for cancer detection with both PET and MR imaging compared with histology at prostatectomy. The results of this prospective analysis showed that ¹⁸F-DCFBC PET is able to detect clinically significant, high-grade prostate cancer and can differentiate between benign lesions and BPH nodules.[35] However, ¹⁸F-DCFBC suffered from persistent high blood-pool activity levels and lacked the high tumoral uptake of newer generation PSMA-targeted radiotracers.

Gorin et al. prospectively studied the performance of a second-generation PSMA-targeted agent, ¹⁸F-DCFPyL, for PET/CT staging in men thought to have localized disease, but at high risk for harboring metastatic disease. Of the 25 men studied, this imaging protocol was found to have a 71.4% sensitivity and 88.9% specificity for identifying lymph node metastases. Nearly 30% of the men studied who were thought to have localized disease actually had nodal disease at the time of prostatectomy. Furthermore, this imaging test identified discrete foci of abnormal radiotracer uptake in the prostate gland of all imaged patients suggestive of its possible future use to aid in the mapping of intraprostatic tumor foci and to more accurately identify metastatic disease during initial staging.[36]

In addition to identifying localized disease, PSMA-targeted imaging may aid in the risk stratification process critical to men considering active surveillance. PSMA expression has been shown to correlate with pathologic grade group, with higher grade tumors showing increased PSMA staining. Therefore, imaging tests utilizing PSMA may determine the relative aggressiveness of a detected tumor.[33], [37] Given the rapidly evolving nature of the PSMA PET imaging technology and the potential for high background uptake or equivocal findings, Rowe et al. developed a PSMA-RADS (reporting and data system) similar to what currently exists for prostate MRI (PI-RADS) and for breast and liver imaging (BI-RADS and LI-RADS, respectively). PSMA-RADS version 1.0 is organized around a 5-point scale, with the lower number, PSMA-RADS-1 and PSMA-RADS-2 scans/lesions either certainly or almost certainly benign, respectively. At the other end of the scale, PSMA-RADS-4 indicates a high likelihood that PCa is present and PSMA-RADS-5 scans/lesions almost certainly represent PCa.[38] Of note, this system is intended for findings outside of the prostate and is not meant to replace the existing PI-RADS reporting system for prostate MRI.

Level 1 randomized evidence also exists supporting PET PSMA: specifically, the “proPSMA study”.[39] This multi-center study from Australia sought to determine the role of PSMA PET/CT in pre-treatment staging for prostate cancer, and was a two-arm randomized controlled trial among men with histologically confirmed high-risk disease who were being considered for curative intent by radical prostatectomy or radiotherapy. Patients were assigned in a 1:1 ratio to either conventional imaging using bone scan and CT, or to ⁶⁸Ga-PSMA PET/CT. In addition to the primary objective of determining the accuracy of first-line diagnostic imaging for the identification of either pelvic nodal or distant metastatic disease, a cost-effectiveness analysis based on data from this study found that on the basis of cost-per-accurate diagnosis, PSMA PET/CT cost less (in Australia) and had better accuracy in this setting.[40] Finally, Hope et al. through an international consortium conducted the pivotal trial that resulted in Ga-PSMA PET obtaining FDA approval for staging in the initial diagnostic setting for men at risk of metastases, but their recent abstract revealed little data regarding prostate gland uptake - clearly something we will hear more about over the coming years as this radiotracer sees increasing application.[41]

PET imaging has not yet been widely shown to be useful in considering men for AS. Most studies have been carried out in men with more aggressive primary tumors, partly because men who will undergo prostatectomy will contribute to a histopathologic gold standard that allows for a ground-truth determination of the performance of the imaging agent. Most PET imaging agents have also demonstrated trends towards higher uptake in higher grade tumors, suggesting there may be improved utility of PET imaging in men with higher-risk prostate cancer. These considerations have led to an emphasis in the literature on the sensitivity of PET imaging agents to detect the site of highest-grade disease in the gland in men with intermediate-risk and above prostate cancer. As PSMA PET becomes more widely available and a commonplace approach for imaging men with more advanced prostate cancer, it will be important to carry-out prospective studies in the AS population to determine the utility of PSMA PET for ruling out higher grade prostate tumors in these patients.