ASCO 2021: Have We Optimized the Detection and Risk Stratification in Men With Localized Prostate Cancer?

( Following three presentations regarding both radiologic and genomic approaches for prostate cancer diagnosis and risk stratification in the Prostate, Testicular, Penile Poster Discussion session at the 2021 American Society of Clinical Oncology (ASCO) Annual Meeting, Dr. Taneja provided a discussant overview of these data and helped to contextualize how we may apply them to our practices.

He began by highlighting the numerous historical tools for prostate cancer detection and risk stratification including:

  • Clinical history and examination: digital rectal examination, family history, and ethnicity
  • Serum biomarkers: PSA (including density, velocity, and free/total ratio), as well as the prostate health index, 4k score, PCA3, and Select MDx
  • Imaging: magnetic resonance imaging, ultrasound (including micro-ultrasound and multi-parametric ultrasound), and novel molecular imaging approaches
  • Genetic testing: germline assessment, somatic mutations, epigenetic evaluation, and genomic classifiers/panels.

In this talk, he focused on the role of MRI and genomic classifiers.

He first discussed Abstract 5007, summarizing the results of the CADMUS trial a paired comparison of multi-parametric ultrasound targeted biopsy with multi-parametric MRI targeted biopsies in men with elevated PSA. Notably, ultrasonography utilized here was multi-parametric, which is not the standard approach used by most clinicians. He emphasized that the two imaging approaches had generally good concordance in terms of lesion identification and in terms of targeted identification of cancer. However, some cancers were identified uniquely by each technique. As highlighted below, Dr. Taneja highlighted cancers missed by each approach. On the right, he presented data from a comparison of MRI-targeted biopsy and standard systematic biopsy. In each case, there are tumors missed by each approach. Together, he suggested that the findings of the CADMUS trial may not reflect the capabilities of multi-parametric ultrasound but rather the multifocality of prostate cancer.


Similarly, comparing to the MRI-FIRST trial, he highlighted that targeted and systematic sampling each demonstrated most cancers while missing a few. The combination of approaches therefore resulted in the highest detection of cancer. Thus, regardless of the definition of significant cancer, some tumors are detected uniquely by systematic biopsy. He concluded that the strengths of the CADMUS trial come from a strong comparative design and the use of an accessible, low-cost ultrasound approach. However, the generalizability of multi-parametric ultrasonography is unclear and it may have low sensitivity so it may not substantially reduce unnecessary biopsy. Moving forward, he suggested that an independent evaluation of multi-parametric ultrasound in warranted as well as consideration of micro-ultrasound, integration of molecular imaging with PSMA-based ligands, and a better understanding of the true significance of cancers which are not apparent on MRI.

Dr. Taneja then moved to discussion of abstract 5009, a comparison of MRI targeted biopsy with transperineal mapping biopsy in patients with recurrent disease following prostate radiotherapy, leading to targeted ablation. The authors concluded that prostate MRI and targeted biopsies can accurately detect and localized recurrent prostate cancer. Further, salvage focal ablation provides good cancer control with preservation of continence. Dr. Taneja highlighted that this trial demonstrated that increasing MRI Likert score was associated with an increasing likelihood of cancer based on biopsy, in keeping with what is seen in primary detection. Further, targeted biopsy missed some cancers that were detected based on transperineal mapping biopsy, though this was only 4 of 76 cases (a 95% sensitivity). However, he emphasized that it is not clear what is meant by missed cancer. He emphasized that, when considering focal therapy, it is not just cancer detection, but accurate localization, that is important. He then discussed a number of prior studies showing that, in the primary setting, while MRI misses some cancers outright, much more commonly it misses many more in estimating focality. In particular, when assessed compared to a whole mount radical prostatectomy specimen, MRI misses non-index lesions at a much higher rate (approximately 50% of high-grade tumors).

Dr. Taneja further highlighted that it is difficult to assess the relevance of oncologic outcomes with two years of follow-up following treatment for radiorecurrent disease. It is unclear whether the favourable outcomes observed represent the effect of treatment or the natural history of the disease. He concluded that this study addresses an important question in salvage therapy; however, methods are poorly described as is the primary outcome. Additionally, follow-up of therapeutic outcomes is short and local recurrence was not measured. Moving forward, focal salvage therapy remains of great interest though there is a pressing need to define selection criteria, optimal methods of mapping, optimal treatment planning, long-term follow-up, and effects on mortality.

Finally, he discussed abstract 5010 providing a validation of the Decipher Genomic Classifier in a randomized trial of dose-escalation in salvage radiotherapy among men who did not receive androgen-deprivation therapy. The authors demonstrated that patients with high GC scores were more than twice as likely to experience biochemical and clinical progression, and the use of salvage hormonal therapy, independent of standard clinicopathologic characteristics and radiotherapy dose.

Importantly, he emphasized that the classifier strongly predicted the development of metastasis. Dr. Taneja highlighted that this paper is strengthened by a pre-defined analysis with a biologic hypothesis and a strong correlation across all endpoints including clinically meaningful ones such as metastasis. However, it’s unclear whether there is an interaction between GC with treatment intervention, which would suggest a potential for a predictive marker. Thus, moving forward, assessment of concomitant androgen deprivation therapy in men with high GC scores is warranted.

In concluding, Dr. Taneja emphasized that the detection and risk stratification of prostate cancer has rapidly and significantly evolved as a result of advances in both imaging and genetic/genomic evaluation. Further advances in imaging may improve detection but need to consider differences between radiographic visible and invisible tumors. Thus, combining imaging and genomic approaches offers the chance to personalize the care of patients with prostate cancer.

Presented by: Samir Taneja, MD, Urologic Oncologost, NYU Langone

Written by: Christopher J.D. Wallis, Urologic Oncology Fellow, Vanderbilt University Medical Center, Contact: @WallisCJD on Twitter at the 2021 American Society of Clinical Oncology (ASCO) Annual Meeting, Virtual Annual Meeting #ASCO21, June, 4-8, 2021

email news signup