ASCO GU 2023: Best of the Journals: Prostate Cancer - Pathology

(UroToday.com) The first day of the American Society for Clinical Oncology (ASCO) Genitourinary Cancer Symposium 2023 began with four talks highlighting key data that, from the perspective of medical oncologists, urologic oncologists, radiation oncologists, and pathologists, advanced the needle in prostate cancer care in 2023.


Dr. Nancy Greenland provided the fourth and final of these talks, emphasizing key developments in prostate cancer care from the perspective of the pathologist.

The first study she presented was from Yu et al., assessing the effect of cribriform pattern 4 and intraductal prostatic carcinoma on NCCN and CAPRA risk stratification. To highlight the importance of these data, she presented the case of a 64 year old man diagnosed with prostate adenocarcinoma with a Gleason score of 3+4=7 and 40% of cores positive for cancer. The pathology report of the biopsy further notes the presence of cribriform pattern. Based on available clinicopathologic data, his CAPRA score is 5. However, it is unclear to what extent the presence of cribriform pattern will affect his risk stratification.

She noted that the International Society of Uro-Pathologists (ISUP) defines cribriform pattern as a “confluent sheet of contiguous malignant epithelial cells with multiple glandular lumina easily visible at low power” with “no intervening stroma or mucin separating individual or fused glandular structures”.

ASCOGU Nancy Greenland-0.jpg

Well predating the publication of this definition, Dr. Greenland highlighted that there have been numerous publications demonstrating that patients with cribriform morphology have aggressive biology.

Nancy Greenland-1.jpg

However, the effect of both cribriform pattern and intraductal carcinoma on nomogram predictions of tumor behaviour are relatively underexplored. In the paper by Yu and colleagues that she highlighted, Dr. Greenland noted that the authors analysed NCCN risk scores and CAPRA scores for patients with and without these features. They subsequently assessed the effect on oncologic outcomes including biochemical recurrence, metastasis and death from prostate cancer (in aggregate, event-free survival). This analysis included 612 patients who were followed for a mean of 5.3 years. As highlighted in the figure below, Dr. Greenland noted that the CAPRA score discriminated three distinct groups for biochemical recurrence which only the high-risk subset stratified from the others for event-free survival. Considering the NCCN risk stratification, there were two groups evident (very low/low/favourable intermediate vs unfavourable intermediate/high/very high) for biochemical recurrence and three evident for event-free survival .

Nancy Greenland-2.jpg

With the addition of the information regarding cribriform pattern and intraductal carcinoma, authors found important prognostic value, particularly among patients with intermediate risk CAPRA scores (3-5 for biochemical recurrence; 3-10 for event-free survival).

Nancy Greenland-3.jpg

Additionally, patients with relatively high risk disease (unfavourable intermediate, high, and very-high risk disease) according to NCCN risk groups were again further sub-stratified by the presence or absence of cribriform pattern and intraductal carcinoma. Further, the addition of this information improved model performance with higher concordance indices for both prognostic models and both endpoints.

ASCOGU Nancy Greenland-4.jpg

Dr. Greenland then moved to the second clinical case of a 68-year old man who underwent radical prostatectomy and was found to have Gleason score 3+4=7 on final pathology. The report further notes that “large cribriform” is present. She first noted that there have been differing definitions of what this means, though overall the literature suggests and association with aggressive disease behaviour. One of the first definitions, dating from 2011, noted large cribriform disease to be that where there was rounded acinar spaces with greater than 12 lumens and no solid growth. Subsequently, in 2019, Hollemans and colleagues defined large cribriform patterns as having a diameter of at least twice the size of adjacent pre-existent normal glands. The goal of the work by Chan and colleagues highlighted here by Dr. Greenland is to provide an outcome-based quantitative cut-off to distinguish large from small cribriform pattern, using tissue micro-array based sampling from the CANARY cohort of patients undergoing radical prostatectomy. Among 1287 patients in the cohort, cribriform pattern was identified in 207 (24%). The figures highlighted below demonstrate the author’s derivation of the measurements to distinguish small and large cribriform patterns.

ASCOGU Nancy Greenland-5.jpg

Using Kaplan Meier estimates of recurrence-free survival, the authors identified 0.25mm as the optimal cut-off to identify patients with more aggressive disease biology. Stratifying in this manner, as highlighted in figures C and D below, clearly distinguishes subgroups of patients with a higher likelihood of disease recurrence.

ASCOGU Nancy Greenland-6.jpg

The authors further validated their findings in an independent set of 419 patients in whom they stratified into three groups: those without cribriform glands >0.25mm or Gleason pattern 5, those with cribriform glands >0.25mm, and those with any Gleason pattern 5. As highlighted below, whether assessing recurrence-free survival, distant metastasis-free survival, or prostate cancer specific survival, these three groups clearly differed and, thus, this measure of large cribriform pattern provided important prognostic information.

ASCOGU Nancy Greenland-7.jpg

Based on these results, the authors advocate reporting cribriform gland size and the use of a 0.25mm cut-off to identify large cribriform glands where are associated with increased potential for metastatic disease, independent of Gleason pattern 5 disease.

Finally, Dr. Greenland presented a third patient case, this of a 73 year old man initially diagnosed with prostate cancer seven years ago, now on androgen deprivation therapy. Recently, a liver biopsy identified metastatic prostate adenocarcinoma “with neuroendocrine features”. Thus, she sought to address the question of whether we should change his therapeutic regime to one tailored to treatment of small cell disease.

In this context, she noted that prostate adenocarcinoma is predominantly androgen driven. Neuroendocrine prostate cancer (NEPC) is characterized by decreased androgen receptor signaling and an insensitivity to androgen-receptor targeting therapies. While it may occur de novo, NEPC is often treatment emergent, under the selective pressure of treatment. However, it is important to identify as the treatment paradigms are quite different. Unfortunately, metastatic biopsies are often difficult to obtain, particularly bone biopsies. Thus, there is an unmet need for an alternative method to identify NEPC.

In the work of Zhao et al. that she highlighted, Dr. Greenland noted that circulating tumor cells (CTCs) were collected from a prospective cohort of 17 patients (10 with adenocarcinoma and 7 with NEPC) as well as 139 patients enrolled in three phase II trials of androgen-receptor signaling inhibitors.

The authors performed analytic validation using 22RV1, a castration-resistant model. All gene expression targets were consistently identified from quantities of DNA corresponding to approximately 0.4 cells.

ASCOGU Nancy Greenland-8.jpg

Using the resulting liquid biomarker assay, the authors found a sensitivity of 51%, specificity of 91%, and overall accuracy of 78% to identify NEPC on a per sample basis among 116 longitudinal samples from 17 patients prospectively collected of whom 7 had NEPC.

ASCOGU Nancy Greenland-9.jpg

They further evaluated specificity by examining baseline samples from patients in two phase II trials which had required adenocarcinoma as an inclusion criterion. The liquid biopsy assay did not identify any patients as having NEPC, thus providing a specificity of 100%. In a similar trial, one baseline sample among 91 patients was classified as NEPC (specificity 99%), though that patient failed androgen-receptor targeted therapies quickly, in keeping with an aggressive disease phenotype.

Dr. Greenland further noted that the authors found that longitudinal sampling over multiple time points improved the diagnostic accuracy of the assay. When all serial blood samples were used to classify each patient, the AUC was 1. If a cut-off of 33% of serial samples read as positive was used to indicate NEPC, the resulting accuracy was 100%.

Utilizing data from the three leverage phase II trials, the authors found that those patients with expression of neuroendocrine markers (even in the presence of preserved AR target expression) had significantly worse clinical outcomes, including overall survival and time to treatment failure. The authors, therefore, concluded that this CTC-based assay could detect NEPC with high accuracy when using serial samples. Thus, using this liquid biopsy for early detection of the transition from adenocarcinoma to a neuroendocrine differentiation may allow for earlier clinical intervention.

Presented by: Nancy Greenland, MD, PhD, University of California, San Francisco, San Francisco, CA