Take Home Message
Higher tumor volume was associated with worse genomic features in intermediate- and high-risk disease, but not in low-risk prostate cancer. These results encourage utilization of active surveillance in the majority of low-risk patients with a high tumor burden.
Keywords: Active surveillance, Biomarker, Decipher, Genomic risk classifier, Prostate cancer, Risk stratification, Tumor volume
Abstract
Background
Despite the historic association of higher prostate cancer volume with worse oncologic outcomes, little is known about the impact of tumor volume on cancer biology.
Objective
To characterize the relationship between tumor volume (measured by percent positive cores [PPC]) and cancer biology (measured by Decipher genomic risk classifier [GC]) in men who underwent prostate biopsy.
Design, setting, and participants
Prostate biopsies from 52 272 men profiled with Decipher captured in a population-based prospective tumor registry were collected from 2016 to 2021.
Outcome measurements and statistical analysis
The degree of distribution and correlation of PPC with a GC score across grade group (GG) strata were examined using the Mann-Whitney U test, Pearson correlation coefficient, and multivariable linear regression controlled for clinicopathologic characteristics.
Results and limitations
A total of 38 921 (74%) biopsies passed quality control (14 331 GG1, 16 159 GG2, 5661 GG3, 1775 GG4, and 995 GG5). Median PPC and GC increased with sequentially higher GG. There was an increasingly positive correlation (r) between PPC and GC in GG2–5 prostate cancer (r [95% confidence interval {CI}]: 0.07 [0.5, 0.8] in GG2, 0.15 [0.12, 0.17] in GG3, 0.20 [0.15, 0.24] in GG4, and 0.25 [0.19, 0.31] in GG5), with no correlation in GG1 disease (r = 0.01, 95% CI [–0.001, 0.03]). In multivariable linear regression, GC was significantly associated with higher PPC for GG2–5 (all p < 0.05) but was not significantly associated with PPC for GG1. Limitations include retrospective design and a lack of final pathology from radical prostatectomy specimens.
Conclusions
Higher tumor volume was associated with worse GC for GG2–5 prostate cancer, whereas tumor volume was not associated with worse GC for GG1 disease. The finding that tumor volume is not associated with worse cancer biology in GG1 disease encourages active surveillance for most patients irrespective of tumor volume.
Patient summary
We studied the relationship between prostate cancer tumor volume and cancer biology, as measured by the Decipher genomic risk score, in men who underwent prostate biopsy. We found that tumor volume was not associated with worse cancer biology for low-grade prostate cancer. Our findings reassuringly support recent guidelines to recommend active surveillance for grade group 1 prostate cancer in most patients, irrespective of tumor volume.
1. Introduction
Tumor volume has widely been incorporated in staging classification systems for solid malignancies as it reflects cancer progression, strongly correlates with prognosis, and is a critical component in assessing the objective treatment response [1], [2], [3], [4]. Tumor volume is thought to contribute to the biologic aggressiveness of prostate cancer with early landmark studies ushering in tumor volume as a commonly accepted predictor of worse oncologic outcomes, which continues to be prominently featured in contemporary guidelines [5], [6], [7], [8].
However, subsequent studies have not reached consensus on whether tumor volume, measured either as percent positive biopsy cores (PPC) or tumor volume on radical prostatectomy specimen, represents a true independent predictor of worse outcomes for any Gleason grade group (GG). While some studies demonstrated that higher tumor volume on radical prostatectomy specimen is associated with biochemical recurrence [7], [9], regional metastasis [10], and adverse pathologic features [11], other series demonstrated tumor volume to offer minimal independent prognostic information, with some arguing that tumor volume should not be measured routinely [12], [13], [14], [15]. The controversial relationship between tumor volume and cancer aggressiveness is further complicated by evidence demonstrating that high PPC is prognostic of worse outcomes in low-, intermediate-, and high-risk prostate cancer [16], [17], [18]. Importantly, the majority of these studies included the original Gleason system in which many pattern 3 morphologies would today be considered pattern 4 [19] and predate contemporary, routine use of genomic risk classifiers (GCs). Despite the conflicting evidence, the recently updated National Comprehensive Center Network (NCCN) guidelines on prostate cancer include tumor volume and genomic risk as factors associated with worse outcomes, such as biopsy reclassification on active surveillance (AS) and increased probability of adverse pathologic features on radical prostatectomy pathology [6].
While a high genomic risk [20], [21], [22] mediates worse outcomes via more aggressive disease, higher tumor volume has not been associated with more aggressive cancer biology or adverse genomic features. Consequently, uncertainty regarding the management of men with a high tumor burden may be allayed with a better understanding of the relationship between tumor volume and cancer biology. Decipher, a 22-gene GC, is validated to improve patient selection for AS [23], [24], predicts early metastasis after prostatectomy [25] and use of androgen deprivation therapy with radiation [26], correlates with worse tumor biology [20], [27], and has recently been endorsed by the NCCN to improve risk stratification after prostate biopsy [6]. Therefore, we aimed to characterize the relationship between tumor volume (measured by PPC) and cancer biology (measured by GC). Our hypothesis is that higher PPC is associated with worse GC, and thereby worse cancer biology.
2. Patients and methods
2.1. Data source
We retrieved prostate biopsy characteristics and GC results from the Decipher Genomic Resource for Intelligent Discovery (GRID) database, a prospectively maintained genome-wide expression registry (NCT02609269), as described previously [28]. Deidentified transcriptome profiles, and basic demographic and clinicopathologic data—age at biopsy, self-reported race, prebiopsy prostate-specific antigen (PSA), clinical stage, continuous GC scores (0–1), and International Society of Urologic Pathology Gleason GG—were collected from clinical use of the Decipher prostate genomic classifier (Veracyte, Inc., San Diego, CA, USA). A waiver of informed consent was obtained (WIRB #20172337).
2.2. Study design
We performed a retrospective study of all men who underwent prostate biopsy captured by the GRID registry from February 2016 to November 2021. Of the 52 272 men with a Decipher biopsy genomic risk score, 38 921 met our selection criteria (Supplementary Fig. 1). Men with incomplete clinicopathologic data or with results not reported due to assay quality control failure were excluded (n = 13 351). Our primary outcome was PPC on prostate biopsy, defined as the total number of cores that contain cancer divided by the total number of cores obtained. To facilitate discussion, we defined low-, intermediate-, and high-volume prostate cancer as fewer than three, between three and six, and six or more positive biopsy cores, respectively. Continuous GC scores (generated on a scale of 0–1) were calculated based on a previously described commercially available GC model [29].
2.3. Statistical analysis
Summary statistics were reported as medians (interquartile ranges [IQRs]) and counts (percentages) for continuous and categorical clinicopathologic variables, respectively. The statistical significance of differences in continuous and categorical variables across PPC strata was assessed via the two-sided Kruskal-Wallis and chi-squared tests, respectively. Pairwise comparisons of continuous variables across GG strata were performed via the Mann-Whitney U test. Pearson correlation coefficients with 95% confidence intervals (CIs) between PPC and GC were reported by GG. Univariable and multivariable linear regression models were used to investigate the relationships between PPC and GC and clinicopathologic variables, which included continuous GC scores (0–1), age at biopsy, race, prebiopsy PSA, clinical stage, and GG; in a post hoc manner, the multivariable linear model was fit to include an interaction term between GC and GG to account for the increasingly positive correlation between these variables.
As a secondary analysis, we evaluated differences in GC between the highest-volume PPC strata (>83%) of one GG with the lowest-volume PPC strata (<17%) of the subsequent GG with the Mann-Whitney U test. We considered p values of <0.05 to indicate statistical significance. All analyses were performed in R version 3.6.3 (R Foundation for Statistical Computing, Vienna, Austria).
3. Results
The median age was 68 yr (IQR 62–73). The median prebiopsy PSA was 6.2 (IQR 4.7–8.6). Most men were not Black (97.3%) and had cT1 disease (84.3%), <50% PPC (73.0%), and GG1 (36.8%) or GG2 (41.5%) disease (Table 1). For men with GG1 prostate cancer, less than half (41.0%) had intermediate- or high-volume disease; there was no statistically significant difference in the distribution or median GC across PPC strata (p = 0.172; Supplementary Table 1).
Table 1.
Clinical characteristics of the analytic cohort stratified by percentage positive cores
| Percent positive cores (PPC) |
Overall | p value | |||||||
|---|---|---|---|---|---|---|---|---|---|
| <17% | 18–25% | 26–3% | 34–50% | 51–66% | 67–83% | >84% | |||
| N | 11 955 (30.7) | 5743 (14.8) | 2004 (5.1) | 12 085 (31.1) | 2320 (6.0) | 2482 (6.4) | 2332 (6.0) | 38 921 (100.0) | |
| Age at biopsy | |||||||||
| Median (Q1, Q3) | 67 (62, 72) | 68 (62, 73) | 68 (63, 73) | 68 (62, 73) | 68 (62, 73) | 68 (62, 73) | 68 (62, 74) | 68 (62, 73) | <0.001a |
| Race | |||||||||
| Black | 271 (2.3) | 139 (2.4) | 38 (1.9) | 324 (2.7) | 69 (3.0) | 75 (3.0) | 79 (3.4) | 995 (2.6) | 0.004b |
| Not Black | 11 684 (97.7) | 5604 (97.6) | 1966 (98.1) | 11 761 (97.3) | 2251 (97.0) | 2407 (97.0) | 2253 (96.6) | 37 926 (97.4) | |
| PSA | |||||||||
| Median (Q1, Q3) | 5.94 (4.5, 8.19) | 6 (4.66, 8.16) | 6.11 (4.6, 8.5) | 6.2 (4.8, 8.46) | 6.5 (4.92, 9) | 6.71 (5.07, 10) | 7.7 (5.4, 13.3) | 6.2 (4.7, 8.6) | <0.001a |
| PSA <4 | 1869 (15.6) | 776 (13.5) | 279 (13.9) | 1362 (11.3) | 206 (8.9) | 231 (9.3) | 209 (9.0) | 4932 (12.7) | <0.001 b |
| 4 ≤ PSA < 10 | 8225 (68.8) | 4124 (71.8) | 1390 (69.4) | 8738 (72.3) | 1652 (71.2) | 1625 (65.5) | 1300 (55.7) | 27 054 (69.5) | |
| 10 ≤ PSA < 20 | 1624 (13.6) | 736 (12.8) | 297 (14.8) | 1681 (13.9) | 373 (16.1) | 456 (18.4) | 467 (20.0) | 5634 (14.5) | |
| PSA ≥20 | 237 (2.0) | 107 (1.9) | 38 (1.9) | 304 (2.5) | 89 (3.8) | 170 (6.8) | 356 (15.3) | 1301 (3.3) | |
| Clinical stage | |||||||||
| T1 | 10 769 (90.1) | 5016 (87.3) | 1691 (84.4) | 10 010 (82.8) | 1785 (76.9) | 1885 (75.9) | 1661 (71.2) | 32 817 (84.3) | <0.001 b |
| T2a | 812 (6.8) | 450 (7.8) | 206 (10.3) | 1114 (9.2) | 282 (12.2) | 268 (10.8) | 223 (9.6) | 3355 (8.6) | |
| T2b/c | 351 (2.9) | 255 (4.4) | 102 (5.1) | 883 (7.3) | 232 (10.0) | 293 (11.8) | 346 (14.8) | 2462 (6.3) | |
| T3/4 | 23 (0.2) | 22 (0.4) | 5 (0.2) | 78 (0.6) | 21 (0.9) | 36 (1.5) | 102 (4.4) | 287 (0.7) | |
| ISUP grade group | |||||||||
| 1 | 6159 (51.5) | 2303 (40.1) | 633 (31.6) | 3782 (31.3) | 504 (21.7) | 530 (21.4) | 420 (18.0) | 14 331 (36.8) | <0.001 b |
| 2 | 4281 (35.8) | 2395 (41.7) | 947 (47.3) | 5488 (45.4) | 1108 (47.8) | 1077 (43.4) | 863 (37.0) | 16 159 (41.5) | |
| 3 | 1160 (9.7) | 763 (13.3) | 300 (15.0) | 2006 (16.6) | 447 (19.3) | 510 (20.5) | 475 (20.4) | 5661 (14.5) | |
| 4 | 293 (2.5) | 220 (3.8) | 98 (4.9) | 568 (4.7) | 152 (6.6) | 202 (8.1) | 242 (10.4) | 1775 (4.6) | |
| 5 | 62 (0.5) | 62 (1.1) | 26 (1.3) | 241 (2.0) | 109 (4.7) | 163 (6.6) | 332 (14.2) | 995 (2.6) | |
| GC score | |||||||||
| Median (Q1, Q3) | 0.39 (0.25, 0.58) | 0.41 (0.25, 0.62) | 0.43 (0.27, 0.66) | 0.45 (0.28, 0.68) | 0.52 (0.31, 0.75) | 0.52 (0.302, 0.77) | 0.61 (0.35, 0.88) | 0.43 (0.27, 0.67) | <0.001 a |
GC = genomic risk classifier; ISUP = International Society of Urological Pathology; PPC = percent positive cores; PSA = prostate-specific antigen; Q1 = quartile 1; Q3 = quartile 3.
Kruskal-Wallis test.
Chi-squared test.
In our study population, the median PPC was 25.0% for GG1 (IQR 14.3–37.5), 33.3% for GG2 (IQR 16.7–50.0), 36.8% for GG3 (IQR 23.1–53.8), 41.7% for GG4 (IQR 25.0–65.7), and 64.3% for GG5 (IQR 41.7–91.7). Similarly, median GC was 0.32 for GG1 (IQR 0.21–0.48), 0.46 for GG2 (IQR 0.29–0.66), 0.61 for GG3 (IQR 0.41–0.82), 0.76 for GG4 (IQR 0.53–0.92), and 0.88 for GG5 (IQR 0.69–0.97; Table 2 and Fig. 1). There was a statistically significant difference in both median PPC and GC between GG strata sequentially (all p < 0.001).
Table 2.
Median (with interquartile ranges) percent positive cores and genomic risk classifier by grade group, with Pearson correlations and slope estimates (with 95% confidence intervals) derived from linear regressions between percent positive cores and genomic risk classifier by grade group
| ISUP grade group |
|||||
|---|---|---|---|---|---|
| 1 | 2 | 3 | 4 | 5 | |
| N (%) | 14 331 (36.8) | 16 159 (41.5) | 5661 (14.5) | 1775 (4.6) | 995 (2.6) |
| Median PPC (Q1, Q3) | 25 (14.3, 37.5) | 33.3 (16.7, 50) | 36.8 (23.1, 53.8) | 41.7 (25, 65.7) | 64.3 (41.7, 91.7) |
| Median GC (Q1, Q3) | 0.32 (0.21, 0.48) | 0.46 (0.29, 0.66) | 0.61 (0.41, 0.82) | 0.76 (0.53, 0.92) | 0.88 (0.69, 0.97) |
| Pearson correlation (95% CI) | 0.01 (–0.00, 0.03) | 0.07 (0.05, 0.08) | 0.15 (0.12, 0.17) | 0.20 (0.15, 0.24) | 0.25 (0.19, 0.31) |
| Slope (95% CI) | 0.12 (–0.05, 0.30) | 0.61 (0.47, 0.76) | 1.34 (1.11, 1.57) | 1.74 (1.32, 2.16) | 2.88 (2.25, 3.52) |
CI = confidence interval; GC = genomic risk classifier; ISUP = International Society of Urological Pathology; PPC = percent positive cores; Q1 = quartile 1; Q3 = quartile 3.
Fig. 1.
Boxplots of percent positive cores and genomic risk classifier (GC) by International Society of Urological Pathology (ISUP) grade group. Sample sizes with medians (first quartile, third quartile) are superimposed above each boxplot.
While there was no correlation between PPC and GC for GG1 prostate cancer (r = 0.01, 95% CI [–0.001, 0.03]), there was an increasingly positive correlation between PPC and Decipher risk score in GG2 (r = 0.07, 95% CI [0.05–0.08]), GG3 (r = 0.15, 95% CI [0.12–0.17]), GG4 (r = 0.20, 95% CI [0.15–0.24]), and GG5 disease (r = 0.25, 95% CI [0.19–0.31]; Table 2 and Fig. 2).
Fig. 2.
Smoothed generalized additive model fits for percent positive cores by genomic risk classifier (GC) stratified by International Society of Urological Pathology (ISUP) grade group.
In univariable analyses, age, Black race, clinical stage, PSA, GC, and GG were associated with higher PPC (all p < 0.05). In multivariable analyses including an interaction term, all variables that were statistically significant in univariable analyses remained so, except race; furthermore, there was no association between GC and PPC for men with GG1 and a statistically significant increase in the degree of association between GC and GG from GG1 to GG2, GG2 to GG3, and GG4 to GG5, with similar parameter estimates between GC and GG between GG3 and GG4 (Table 3).
Table 3.
Univariable and multivariable linear regression (with and without interaction term between genomic classifier and grade group) for percent positive cores
| Regression term | Univariable |
Multivariable |
Multivariable + interaction |
|||
|---|---|---|---|---|---|---|
| Estimate (95% CI) | p value | Estimate (95% CI) | p value | Estimate (95% CI) | p value | |
| (Intercept) | – | – | 31.9 (29.9, 33.8) | <0.001 | 33.9 (31.9, 35.9) | <0.001 |
| GC (per 0.1) | 1.8 (1.7, 1.9) | <0.001 | 0.7 (0.6, 0.8) | <0.001 | 0.1 (–0.1, 0.3) | 0.177 |
| Age at biopsy (per 5 yr) | 0.4 (0.3, 0.5) | <0.001 | –0.8 (–0.9, –0.6) | <0.001 | –0.8 (–0.9, –0.6) | <0.001 |
| Race: Black | 2.7 (1.3, 4.2) | <0.001 | 1.2 (–0.1, 2.6) | 0.079 | 1.3 (0.0, 2.7) | 0.056 |
| 4 ≤ PSA < 10 vs >4 | 3.6 (2.9, 4.3) | <0.001 | 3.6 (2.9, 4.2) | <0.001 | 3.6 (2.9, 4.2) | <0.001 |
| 10 ≤ PSA < 20 vs >4 | 7.2 (6.3, 8.1) | <0.001 | 5.3 (4.4, 6.1) | <0.001 | 5.2 (4.4, 6.1) | <0.001 |
| PSA ≥20 vs >4 | 23.6 (22.2, 25.0) | <0.001 | 15.0 (13.7, 16.4) | <0.001 | 14.6 (13.2, 15.9) | <0.001 |
| ISUP grade group 2 vs 1 | 7.9 (7.4, 8.4) | <0.001 | 6.9 (6.4, 7.4) | <0.001 | 5.2 (4.1, 6.2) | <0.001 |
| ISUP grade group 3 vs 1 | 12.7 (12.0, 13.3) | <0.001 | 10.0 (9.3, 10.7) | <0.001 | 4.1 (2.4, 5.7) | <0.001 |
| ISUP grade group 4 vs 1 | 17.7 (16.6, 18.8) | <0.001 | 12.8 (11.7, 14.0) | <0.001 | 3.5 (0.3, 6.7) | 0.030 |
| ISUP grade group 5 vs 1 | 34.8 (33.4, 36.2) | <0.001 | 27.4 (25.9, 28.8) | <0.001 | 7.9 (2.5, 13.2) | 0.004 |
| Clinical stage T2b/c vs T1-T2a | 13.6 (12.7, 14.5) | <0.001 | 9.8 (8.9, 10.7) | <0.001 | 9.8 (8.9, 10.6) | <0.001 |
| Clinical stage T3/4 vs T1-T2a | 27.7 (25.1, 30.3) | <0.001 | 15.0 (12.5, 17.6) | <0.001 | 14.6 (12.0, 17.1) | <0.001 |
| GC (per 0.1): ISUP grade 2 vs 1 | – | – | – | – | 0.5 (0.3, 0.7) | <0.001 |
| GC (per 0.1): ISUP grade 3 vs 1 | – | – | – | – | 1.2 (0.9, 1.5) | <0.001 |
| GC (per 0.1): ISUP grade 4 vs 1 | – | – | – | – | 1.6 (1.2, 2.1) | <0.001 |
| GC (per 0.1): ISUP grade 5 vs 1 | – | – | – | – | 2.8 (2.1, 3.4) | <0.001 |
CI = confidence interval; GC = genomic risk classifier; ISUP = International Society of Urological Pathology; PSA = prostate-specific antigen.
In a secondary analysis, the mean GC for high-volume GG1 was statistically significantly lower than that for low-volume GG2 (0.37 vs 0.47, p < 0.001), and the mean GC for high-volume GG2 was statistically significantly lower than that for low-volume GG3 (0.52 vs 0.57, p < 0.001). Conversely, the mean GC for high-volume GG3 was not statistically significantly different from that for low-volume GG4 (0.69 vs 0.67, p = 0.232), and the mean GC for high-volume GG4 was statistically significantly higher than that for low-volume GG5 (0.79 vs 0.71, p = 0.02; Fig. 3). Using a cutoff proposed previously as a very–low-risk threshold for metastatic progression, 26% of high-volume GG1 and 12% of low-volume GG2 had GC <0.2 [30].
Fig. 3.
Smoothed genomic risk classifier (GC) score densities by grade group and percent positive cores (PPC) strata. The value and solid vertical line within each density represent the median GC score.
4. Discussion
In this novel, population-based study of Decipher biopsy genomic risk assay, we found that higher tumor volume was associated with a higher genomic risk for men with GG2–5 prostate cancer, with stronger interactions associated with progressively higher GGs. However, higher tumor volume was not associated with higher GC for GG1 prostate cancer. We demonstrate that higher-volume GG2–5 prostate cancer is more biologically aggressive than low-volume GG2–5 tumor, whereas GG1 prostate cancer has a similar genomic risk regardless of tumor volume. In other words, in contrast to GG1 prostate cancer, tumor volume is associated with worse tumor biology for GG2–5 disease. Additionally, our secondary analysis suggests that high-volume GG3 and GG4 may have similar biologic aggressiveness to low-volume GG4 and GG5, respectively.
The natural history of prostate cancer is largely indolent, and better prediction of tumor biology may obviate treatment in older men. This is the impetus for improving prognostic models to predict a long-term oncologic outcome after diagnosis with localized prostate cancer. This led to the development of criteria (eg, those by Epstein et al [8]), risk grouping systems (eg, NCCN and D’Amico), risk assessment scoring systems (eg, Cancer of the Prostate Risk Assessment [CAPRA]), and nomograms that have been refined to improve prognostic performance by including variables such as PPC and primary Gleason pattern [16]. Despite this, existing models do not currently reflect recent advancements in our understanding of prostate cancer as a complex, tremendously heterogeneous disease with a unique genomic landscape [20] that may explain the relatively poor performance of the above classification systems in predicting metastasis and cancer-specific mortality [31]. For instance, that a small, but statistically significant, proportion of low-risk prostate cancer is associated with a high genomic risk may provide an explanation for why some men may harbor occult intermediate- or high-risk disease [18], [32] and experience worse outcomes [20].
Our findings may offer a potential explanation for the controversy regarding the relationship between tumor volume and cancer biology by demonstrating that higher PPC is associated with worse GC in GG2–5 disease, while strongly implying that GG1 prostate cancer, which involves the population of most interest regarding the future use of genomic classifiers to confirm candidacy for AS, may represent a distinct biologic entity in which high tumor volume is not a hallmark of worse tumor biology. While contemporary studies have shown that men with intermediate- and high-volume GG1 tumors have a higher risk of biopsy reclassification [23] or harbor occult clinically significant disease [18], [32], these cohorts did not include GCs, which are currently recommended by professional guidelines [6]. Given recently updated NCCN guidelines recommending that men with low-risk prostate cancer undergo GCs after diagnostic biopsy, our findings have important implications to improve the nuance and accuracy of risk stratification and determining AS candidacy [6]. Indeed, results from early efforts to incorporate novel prognostic commercial tissue-based molecular assays, such as the Decipher genomic classifier, into traditional risk classification systems to generate integrated clinical-genomic risk groups have led to significant prognostic reclassification as well as improved independent prediction over existing prognostic models, which currently include variables such as tumor volume that do not directly measure biologic aggressiveness [20], [33].
For example, in our study, 5236 men have intermediate- or high-volume GG1 prostate cancer that may not strictly meet the NCCN “very low” or “low” risk criteria for which AS is a “preferred” option. Based on previously established cutoffs of a low or intermediate (<0.60) Decipher risk score [34], 12 271 men in our study who would potentially be recommended to pursue earlier definitive treatment may have been suitable candidates for surveillance; this potentially includes 2090 men with high-volume GG1 disease—“unfavorable intermediate” risk—who currently would not be deemed suitable AS candidates. Although more study is needed, use of genomic classifiers may support the continued use of AS in some men with high-volume GG1 prostate cancer [35], [36].
Separately from tumor volume, GCs may contain other potential avenues to better the management of prostate cancer. Analogous to GG1 disease, genomic classifiers may help identify men with GG2 cancer as candidates for AS and promote the concept of biologic aggressiveness when counseling men with GG3–5 cancer. Its use may help refine the distinction between “favorable intermediate-” and “unfavorable intermediate”-risk disease, which may affect the utilization of androgen deprivation therapy after radiation, and the need for cross-sectional imaging and metastatic evaluation prior to definitive treatment. Genomic classifiers may also improve our current understanding of the prognostic significance of the largest (index) cancer focus as the lesion that determines disease progression, with particular relevance to the emerging field of focal therapy. Furthermore, along with an increasing number of positive biopsy cores that often triggers definitive treatment [37], [38], [39], use of repeat genomic analysis on subsequent surveillance prostate biopsies can iteratively inform management for men with intermediate- or high-volume prostate cancer. Future research will integrate prostate biopsy GC with final radical prostatectomy specimen pathology to further validate the use of biomarkers in prostate cancer; this goal has been cited as an important goal by the National Cancer Institute Early Detection Research Network.
Our findings must be interpreted in the context of the study design. First, we did not have a centralized pathology review of biopsy specimens, true of any population-based study, but GRID is representative of both community and academic settings, thus increasing generalizability. Second, because our study did not include characteristics of the radical prostatectomy specimen or long-term follow-up, we were unable to assess the rates of pathologic upgrading, upstaging, metastasis, or mortality. However, given that Decipher is validated for predicting progression, rates of metastasis, and mortality, its use allows for significant prognostic associations between tumor volume and intermediate- and long-term outcomes. Third, the GRID dataset lacks sufficient granularity to enable further analysis based on common racial categories and does not capture information regarding prostate magnetic resonance imaging. Finally, the GRID registry does not capture the maximal cancer core length or percent of cancer within a positive core. However, tumor core length and PPC are closely correlated, with one study stating that the two measures together did not add prognostic information [40].
5. Conclusions
Higher tumor volume was associated with a worse genomic risk score for GG2–5 disease, whereas tumor volume was not associated with a worse genomic risk score for GG1 disease. Therefore, tumor volume has a greater, independent role as a prognostic factor with increasing GG, mediated by more aggressive tumor biology. Our findings support the emerging role of genomic classifiers in the risk stratification of prostate cancer and encourage AS in the majority of men with higher-volume GG1 disease.
Author contributions: Ashwin Ramaswamy had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: Ross, Schaeffer, Davicioni, Hu.
Acquisition of data: Proudfoot, Davicioni.
Analysis and interpretation of data: Proudfoot.
Drafting of the manuscript: Ramaswamy, Proudfoot.
Critical revision of the manuscript for important intellectual content: Ramaswamy, Proudfoot, Ross, Schaeffer, Davicioni, Hu.
Statistical analysis: Proudfoot.
Obtaining funding: None.
Administrative, technical, or material support: Ramaswamy, Davicioni, Hu.
Supervision: Davicioni, Hu.
Other: None.
Financial disclosures: Ashwin Ramaswamy certifies that all conflicts of interest, including specific financial interests and relationships and affiliations relevant to the subject matter or materials discussed in the manuscript (eg, employment/affiliation, grants or funding, consultancies, honoraria, stock ownership or options, expert testimony, royalties, or patents filed, received, or pending), are the following: Ashwin Ramaswamy reports no conflicts of interest. James A. Proudfoot is an employee of Veracyte, Inc., which has rights to the Decipher genomic classifier assay and maintains the Decipher GRID database. Ashley E. Ross is a consultant for Veracyte, Inc.; Astellas, Inc.; Bayer, Inc.; Blue Earth, Inc.; Janssen, Inc.; and Tempus, Inc. Elai Davicioni is an employee of Veracyte, Inc., which has rights to the Decipher genomic classifier assay and maintains the Decipher GRID database. Edward M. Schaeffer is a consultant for Pfizer, Inc.; Astellas, Inc.; and Lantheus, Inc.; as well as receiving salary support from NIH R01CA241758 and NIH R01CA259173. Jim C. Hu receives research support from the Frederick J. and Theresa Dow Wallace Fund of the New York Community Trust, as well as receiving salary support from NIH R01CA241758, NIH R01CA259173, Prostate Cancer Foundation, PCORI CER-2019C1-15682, and CER-2019C2-17372.
Funding/Support and role of the sponsor: This work was supported by Veracyte Inc.
Data sharing: Data for this study were generated from the Decipher GRID database, a prospective, genome-wide expression registry (NCT02609269) evaluating clinical use of the Decipher prostate genomic classifier (Veracyte, Inc.). The data underlying this article will be shared on reasonable request to the corresponding author.
Associate Editor: Guillaume Ploussard
Footnotes
Supplementary data to this article can be found online at https://doi.org/10.1016/j.euros.2022.12.002.
Appendix A. Supplementary data
The following are the Supplementary data to this article:
References
- 1.Therasse P., Arbuck S.G., Eisenhauer E.A., et al. New guidelines to evaluate the response to treatment in solid tumors. European Organization for Research and Treatment of Cancer, National Cancer Institute of the United States, National Cancer Institute of Canada. J Natl Cancer Inst. 2000;92:205–216. doi: 10.1093/jnci/92.3.205. [DOI] [PubMed] [Google Scholar]
- 2.Takenaka T., Yamazaki K., Miura N., Mori R., Takeo S. The prognostic impact of tumor volume in patients with clinical stage IA non-small cell lung cancer. J Thorac Oncol. 2016;11:1074–1080. doi: 10.1016/j.jtho.2016.02.005. [DOI] [PubMed] [Google Scholar]
- 3.Sze W.M., Lee A.W.M., Yau T.K., et al. Primary tumor volume of nasopharyngeal carcinoma: prognostic significance for local control. Int J Radiat Oncol Biol Phys. 2004;59:21–27. doi: 10.1016/j.ijrobp.2003.10.027. [DOI] [PubMed] [Google Scholar]
- 4.Giuliani L, Giberti C, Martorana G, Rovida S. Radical extensive surgery for renal cell carcinoma: long-term results and prognostic factors. J Urol 1990;143:468–73; discussion 473–4. [DOI] [PubMed]
- 5.McNeal J.E. Cancer volume and site of origin of adenocarcinoma in the prostate: relationship to local and distant spread. Hum Pathol. 1992;23:258–266. doi: 10.1016/0046-8177(92)90106-d. [DOI] [PubMed] [Google Scholar]
- 6.NCCN. Guidelines detail. https://www.nccn.org/guidelines/guidelines-detail?category=1&id=1459.
- 7.Stamey T.A., McNeal J.E., Yemoto C.M., Sigal B.M., Johnstone I.M. Biological determinants of cancer progression in men with prostate cancer. JAMA. 1999;281:1395–1400. doi: 10.1001/jama.281.15.1395. [DOI] [PubMed] [Google Scholar]
- 8.Epstein J.I., Walsh P.C., Carmichael M., Brendler C.B. Pathologic and clinical findings to predict tumor extent of nonpalpable (stage T1c) prostate cancer. JAMA. 1994;271:368–374. [PubMed] [Google Scholar]
- 9.Ramos C.G., Roehl K.A., Antenor J.A.V., Humphrey P.A., Catalona W.J. Percent carcinoma in prostatectomy specimen is associated with risk of recurrence after radical prostatectomy in patients with pathologically organ confined prostate cancer. J Urol. 2004;172:137–140. doi: 10.1097/01.ju.0000132139.40964.75. [DOI] [PubMed] [Google Scholar]
- 10.McNeal J.E., Villers A.A., Redwine E.A., Freiha F.S., Stamey T.A. Histologic differentiation, cancer volume, and pelvic lymph node metastasis in adenocarcinoma of the prostate. Cancer. 1990;66:1225–1233. doi: 10.1002/1097-0142(19900915)66:6<1225::aid-cncr2820660624>3.0.co;2-x. [DOI] [PubMed] [Google Scholar]
- 11.Chun F.K.H., Briganti A., Jeldres C., et al. Tumour volume and high grade tumour volume are the best predictors of pathologic stage and biochemical recurrence after radical prostatectomy. Eur J Cancer. 2007;43:536–543. doi: 10.1016/j.ejca.2006.10.018. [DOI] [PubMed] [Google Scholar]
- 12.Epstein J.I., Carmichael M., Partin A.W., Walsh P.C. Is tumor volume an independent predictor of progression following radical prostatectomy? A multivariate analysis of 185 clinical stage B adenocarcinomas of the prostate with 5 years of followup. J Urol. 1993;149:1478–1481. doi: 10.1016/s0022-5347(17)36421-2. [DOI] [PubMed] [Google Scholar]
- 13.Salomon L., Levrel O., Anastasiadis A.G., et al. Prognostic significance of tumor volume after radical prostatectomy: a multivariate analysis of pathological prognostic factors. Eur Urol. 2003;43:39–44. doi: 10.1016/s0302-2838(02)00493-1. [DOI] [PubMed] [Google Scholar]
- 14.Wolters T., Roobol M.J., van Leeuwen P.J., et al. Should pathologists routinely report prostate tumour volume? The prognostic value of tumour volume in prostate cancer. Eur Urol. 2010;57:821–829. doi: 10.1016/j.eururo.2009.07.027. [DOI] [PubMed] [Google Scholar]
- 15.Ito Y., Udo K., Vertosick E.A., et al. Clinical usefulness of prostate and tumor volume related parameters following radical prostatectomy for localized prostate cancer. J Urol. 2019;201:535–540. doi: 10.1016/j.juro.2018.09.060. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Zumsteg Z.S., Spratt D.E., Pei I., et al. A new risk classification system for therapeutic decision making with intermediate-risk prostate cancer patients undergoing dose-escalated external-beam radiation therapy. Eur Urol. 2013;64:895–902. doi: 10.1016/j.eururo.2013.03.033. [DOI] [PubMed] [Google Scholar]
- 17.D’Amico A.V., Renshaw A.A., Cote K., et al. Impact of the percentage of positive prostate cores on prostate cancer-specific mortality for patients with low or favorable intermediate-risk disease. J Clin Oncol. 2004;22:3726–3732. doi: 10.1200/JCO.2004.01.164. [DOI] [PubMed] [Google Scholar]
- 18.Dinh K.T., Muralidhar V., Mahal B.A., et al. Occult high-risk disease in clinically low-risk prostate cancer with ≥50% positive biopsy cores: should national guidelines stop calling them low risk? Urology. 2016;87:125–132. doi: 10.1016/j.urology.2015.08.026. [DOI] [PubMed] [Google Scholar]
- 19.Epstein J.I. Prostate cancer grading: a decade after the 2005 modified system. Mod Pathol. 2018;31:S47–S63. doi: 10.1038/modpathol.2017.133. [DOI] [PubMed] [Google Scholar]
- 20.Cooperberg M.R., Erho N., Chan J.M., et al. The diverse genomic landscape of clinically low-risk prostate cancer. Eur Urol. 2018;74:444–452. doi: 10.1016/j.eururo.2018.05.014. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Kornberg Z., Cooperberg M.R., Cowan J.E., et al. A 17-gene genomic prostate score as a predictor of adverse pathology in men on active surveillance. J Urol. 2019;202:702–709. doi: 10.1097/JU.0000000000000290. [DOI] [PubMed] [Google Scholar]
- 22.Lin D.W., Zheng Y., McKenney J.K., et al. 17-Gene genomic prostate score test results in the Canary Prostate Active Surveillance Study (PASS) cohort. J Clin Oncol. 2020;38:1549–1557. doi: 10.1200/JCO.19.02267. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Lonergan P.E., Washington S.L., 3rd, Cowan J.E., et al. Risk factors for biopsy reclassification over time in men on active surveillance for early stage prostate cancer. J Urol. 2020;204:1216–1221. doi: 10.1097/JU.0000000000001186. [DOI] [PubMed] [Google Scholar]
- 24.Kim H.L., Li P., Huang H.C., et al. Validation of the Decipher test for predicting adverse pathology in candidates for prostate cancer active surveillance. Prostate Cancer Prostatic Dis. 2019;22:399–405. doi: 10.1038/s41391-018-0101-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Erho N., Crisan A., Vergara I.A., et al. Discovery and validation of a prostate cancer genomic classifier that predicts early metastasis following radical prostatectomy. PLoS One. 2013;8:e66855. doi: 10.1371/journal.pone.0066855. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Hamid A.A., Huang H.C., Wang V., et al. Transcriptional profiling of primary prostate tumor in metastatic hormone-sensitive prostate cancer and association with clinical outcomes: correlative analysis of the E3805 CHAARTED trial. Ann Oncol. 2021;32:1157–1166. doi: 10.1016/j.annonc.2021.06.003. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Greenland N.Y., Cooperberg M.R., Wong A.C., et al. Molecular risk classifier score and biochemical recurrence risk are associated with cribriform pattern type in Gleason 3 + 4 = 7 prostate cancer. Investig Clin Urol. 2022;63:27–33. doi: 10.4111/icu.20210262. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Hall W.A., Fishbane N., Liu Y., et al. Development and validation of a genomic tool to predict seminal vesicle invasion in adenocarcinoma of the prostate. JCO Precis Oncol. 2020;4:1228–1238. doi: 10.1200/PO.20.00013. [DOI] [PubMed] [Google Scholar]
- 29.Ross A.E., Johnson M.H., Yousefi K., et al. Tissue-based genomics augments post-prostatectomy risk stratification in a natural history cohort of intermediate- and high-risk men. Eur Urol. 2016;69:157–165. doi: 10.1016/j.eururo.2015.05.042. [DOI] [PubMed] [Google Scholar]
- 30.Nguyen P.L., Martin N.E., Choeurng V., et al. Utilization of biopsy-based genomic classifier to predict distant metastasis after definitive radiation and short-course ADT for intermediate and high-risk prostate cancer. Prostate Cancer Prostatic Dis. 2017;20:186–192. doi: 10.1038/pcan.2016.58. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Spratt D.E., Schaeffer E.M. Transcriptomic heterogeneity of favorable-risk prostate cancer: time to move past Cancer of the Prostate Risk Assessment (CAPRA) to clinical-genomic risk. Eur Urol. 2018;74:453–454. doi: 10.1016/j.eururo.2018.06.001. [DOI] [PubMed] [Google Scholar]
- 32.Dinh K.T., Mahal B.A., Ziehr D.R., et al. Incidence and predictors of upgrading and up staging among 10,000 contemporary patients with low risk prostate cancer. J Urol. 2015;194:343–349. doi: 10.1016/j.juro.2015.02.015. [DOI] [PubMed] [Google Scholar]
- 33.Spratt D.E., Zhang J., Santiago-Jiménez M., et al. Development and validation of a novel integrated clinical-genomic risk group classification for localized prostate cancer. J Clin Oncol. 2018;36:581–590. doi: 10.1200/JCO.2017.74.2940. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Jairath N.K., Dal Pra A., Vince R., Jr, et al. A systematic review of the evidence for the decipher genomic classifier in prostate cancer. Eur Urol. 2021;79:374–383. doi: 10.1016/j.eururo.2020.11.021. [DOI] [PubMed] [Google Scholar]
- 35.Cher M.L., Dhir A., Auffenberg G.B., et al. Appropriateness criteria for active surveillance of prostate cancer. J Urol. 2017;197:67–74. doi: 10.1016/j.juro.2016.07.005. [DOI] [PubMed] [Google Scholar]
- 36.Dall’Era M.A., Albertsen P.C., Bangma C., et al. Active surveillance for prostate cancer: a systematic review of the literature. Eur Urol. 2012;62:976–983. doi: 10.1016/j.eururo.2012.05.072. [DOI] [PubMed] [Google Scholar]
- 37.Bul M., Zhu X., Valdagni R., et al. Active surveillance for low-risk prostate cancer worldwide: the PRIAS study. Eur Urol. 2013;63:597–603. doi: 10.1016/j.eururo.2012.11.005. [DOI] [PubMed] [Google Scholar]
- 38.Soloway M.S., Soloway C.T., Eldefrawy A., Acosta K., Kava B., Manoharan M. Careful selection and close monitoring of low-risk prostate cancer patients on active surveillance minimizes the need for treatment. Eur Urol. 2010;58:831–835. doi: 10.1016/j.eururo.2010.08.027. [DOI] [PubMed] [Google Scholar]
- 39.Iremashvili V., Pelaez L., Manoharan M., Jorda M., Rosenberg D.L., Soloway M.S. Pathologic prostate cancer characteristics in patients eligible for active surveillance: a head-to-head comparison of contemporary protocols. Eur Urol. 2012;62:462–468. doi: 10.1016/j.eururo.2012.03.011. [DOI] [PubMed] [Google Scholar]
- 40.Vollmer R.T. Percentage of tumor and tumor length in prostate biopsy specimens: a study of American veterans. Am J Clin Pathol. 2008;130:940–943. doi: 10.1309/AJCP3VUXBYTEY3PU. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.