Quantification of Gleason pattern 4 metrics identifies pathologic progression in patients with Grade Group 2 prostate cancer on active surveillance

Abstract

Background:

During active surveillance (AS) for Grade Group (GG) 2 prostate cancer, pathologic progression to GG3 on surveillance biopsy is a trigger for intervention. However, this ratio of GP3:GP4, may be obscured by increases of relatively indolent disease. We aimed to explore changes in GP4 quantity during AS and propose alternative definitions for progression based on GP4 changes.

Design, Setting, and Participants:

We assessed patients enrolled on AS between November 2014 and March 2020 with GG2 disease on diagnostic biopsy and subsequent surveillance biopsy approximately 1 year later. Outcome measures included change in overall %GP4 and total length GP4 (mm).

Results and Limitations:

61 patients met the inclusion criteria, the median change in total length of GP4 and %GP4 was −0.12 mm (IQR −0.31, 0.09) and −2.5% (IQR −8.6, 0.0), respectively. Excluding the 35 patients with no evidence of GP4 on surveillance biopsy, median change in total GP4 length and %GP4 was 0.19 mm (IQR −0.04, 0.67) and 1.2% (IQR −1.6, 6.6), respectively. Three patients progressed to GG3 disease on surveillance biopsy, one of whom had only a small increase in %GP4. Conversely, an additional two patients who did not meet the criterion for GG3 had a large increase (> 1 mm) in total GP4 length.

Conclusions:

Presence of GG3 disease on surveillance biopsy as a trigger for treatment in men on AS is of questionable use alone; we suggest including other measures that do not depend on a ratio, such as an increase in total GP4 length.

MICROABSTRACT

Pathological progression to Grade Group 3 (GG3) disease in active surveillance prompts intervention, but the optimal method for Gleason pattern 4 (GP4) quantification remains unclear. Significant increases in GP4 length do not always meet GG3 criteria, while upgrading to GG3 can happen with minimal changes in GP4 length. We present an alternative method for grading GP4 is presented, highlighting limitations in current assessment practices.

Introduction

Active surveillance (AS) is increasingly recognized as a management approach for favorable Grade Group 2 (GG2) prostate cancer¹. Nonetheless, such patients do have a risk of metastatic progression. In the Toronto AS series reported by Klotz et al, for instance, patients with GG2 disease had a 3-fold increased risk of metastasis compared to men with GG1². However, when performed appropriately with careful patient selection, AS in GG2 disease can provide acceptable oncologic outcomes even if definitive treatment is eventually required^3,4.

While our experience of AS in patients with GG2 disease matures, there is a need to understand the prognostic implications of pathologic findings. Within GG2, Gleason pattern 4 (GP4) disease is considered more aggressive than the GP3 component. Therefore, the role of more precise quantification of GP4 is increasingly being recognized by several institutions, including the most recent World Health Organization genitourinary tumor classification, which recommends reporting the percentage of GP4 disease (%GP4)⁵. Various methods of GP4 quantification have been proposed, including overall %GP4, maximal core %GP4, maximal GP4 core length (mm), and total GP4 length (mm)^6,7. Of these metrics, recent publications by our group^6,8 demonstrated that, in men with GG2 disease, total length of GP4 was most predictive of adverse pathology and oncologic outcomes after RP. For men with 2 mm of GP4 on biopsy, the risk of adverse pathology was 45% and 3-year biochemical recurrence (BCR) rate was 13%; for each 1-mm increase in GP4, the risk of adverse pathology increased by 6–8% and 3-year BCR rates increased by 1–2%^6,8.

A key consideration for AS is the criteria for intervention. Presently, many groups advocate that intervention should be triggered by detection of GG3 disease. This is problematic given that the difference between GG2 and GG3 depends on the ratio of GP3 to GP4. A patient with a dramatic increase in GP4 volume may not be classified as GG3 disease if a proportionate increase in GP3 occurs; similarly, a patient who has a surveillance biopsy showing a reduced amount of GP4 might be reclassified as GG3 if that biopsy shows a concomitantly larger decrease in GP3.

Our overarching goal was to further explore how best to use GP4 to define disease reclassification or progression for triggering treatment in patients on AS. We aimed to assess the role of quantification methods of GP4 including not only Gleason grade group, but also overall %GP4, change in GP4 and total length GP4. Although MRI is widely used in practice, including at our own institution, there is no widely accepted method of Gleason pattern quantification for cores from MRI-targeted biopsies. For instance, if three cores are targeted to a given lesion, and length of pattern 4 is 0, 0.5 and 1.3mm for each, it is unclear whether the amount of pattern 4 should be based on the total (1.8), the average (0.6) or maximum (1.3). Combining data from patients who received varying numbers of targeted cores and the interpretation of Gleason pattern grading presents its own challenges. Therefore, we limited our analysis to cores obtained from systematic biopsies only. As such, our results are intended to be proof-of-principle that would require further evaluation, most specifically in MRI-targeted biopsy before implementation in clinical practice.

Patients and Methods

We identified patients with GG2 disease on diagnostic biopsy who enrolled on AS at Memorial Sloan Kettering Cancer Center (MSK) from November 2014 to March 2020. This period was selected due to the nature of pathologic reporting allowing for quantification of GP4 disease, where all prostate biopsies are reported by dedicated genitourinary pathologists. Next, we identified the subset of patients who had an initial diagnostic biopsy followed by a repeat (surveillance) biopsy after approximately one year, allowing a window of 6–24 mo. We included patients with GG2 disease enrolled on AS who progressed to GG3 disease or demonstrated a substantial increase in GP4 volume, as defined as an increase in total mm GP4 of ≥1 mm, resulting in at least 2 mm GP4. This GP4 quantification parameter was based on the aforementioned findings from research previously published⁸.

We excluded any outside biopsies lacking individual core data for review. While some patients may have undergone MRI-targeted biopsy, we included only those biopsies with 12–14 systematic cores and only a single systematic core taken per region. MRI-targeted cores, whether image-fusion, cognitive-fusion, or in-bore biopsy, were excluded from the analysis. We reviewed individual core data for the systematic biopsy components for the diagnostic biopsies for all patients, as well as and their surveillance biopsy occurring approximately one year after their diagnostic biopsy. Ultimately, we identified 61 patients for whom both the diagnostic and 1-yr surveillance biopsy had 12–14 systematic cores and only one systematic core per region.

From 2015 onwards, % cancer per core, mm cancer per core, % pattern 4 (in 3+4=7 bearing cores) were routinely reported in the cohort as a standard component of pathology reporting. Pathologic reporting was performed by seven dedicated urologic pathologists. Length of GP4 for each core was calculated by multiplying the %GP4 by the length of cancer. We performed a descriptive analysis comparing the diagnostic biopsy findings with the surveillance biopsy findings in terms of the GP4 quantification metrics as well as the change in these metrics from diagnostic to 1-year surveillance biopsy. All analyses were conducted using R 4.1.0⁹.

Results

Between November 2014 and March 2020, we identified 1757 patients with enrolled in AS, of whom 185 had GP4 disease on the diagnostic biopsy. Of these 185 patients, 61 had diagnostic and 1-year surveillance biopsy pairs. Demographic characteristics of this cohort are displayed in Table 1. On the 1-year surveillance biopsy, 11 (18%) did not have any cancer detected, 23 (38%) did not have any GP4 (ie. had GG1 disease), 22 (36%) remained at GG2, and three patients (5%) had grade group progression, 2 to GG3 and 1 to GG4. Thus, a majority of patients (34) had zero for total length of GP4 and overall %GP4 on 12-month surveillance biopsy, and therefore the overall median for both GP4 measures at surveillance biopsy was zero. Distributions of total length of GP4 and overall %GP4 by biopsy status are displayed in Figures 1a and 1b, respectively.

Table 1: Patient demographics^a.

Characteristic	N = 61
Age at diagnostic biopsy (yr)	65 (60, 69)
Raceb
Asian	3 (5.1%)
Black	4 (6.8%)
Other	1 (1.7%)
White	51 (86%)
Unknown	2
Clinical stageb
T1	57 (97%)
T2	2 (3.4%)
Unknown	2
MRI-targeted biopsy	22 (36%)
Total number of systematic cores per biopsy
12	26 (43%)
13	5 (8.2%)
14	30 (49%)
Total PSA at diagnosis (ng/mL)b (n=58)	5.1 (4.1, 6.6)

^aData are expressed as median (IQR) or n (%).

^bPercentages or median (IQR) calculated disregarding the 2 patients with unknown data.

MRI = magnetic resonance imaging, PSA = prostate-specific antigen.

Figure 1: Distribution of GP4 tissue observed on diagnostic biopsies (blue shading) and surveillance biopsies (orange shading). A) Total length GP4 (mm) excluding 2 surveillance biopsies with 10.5 and 12.6 mm. B) Overall percentage (%GP4) excluding 2 surveillance biopsies with 55% and 70%.

The GP4 quantification metrics and changes in those metrics are summarized in Table 2. Table 3 shows changes in GP4 metrics excluding the patients who met definitions of progression. Table 4 shows individual patient data for the five patients who either progressed to GG3 or who had a >1 mm increase in GP4 to a total GP4 length ≥2 mm. It is notable that there is considerable discordance between the two definitions: one patient who progressed to GG3 had only a very small increase in GP4; conversely, two patients had extremely large increases in length of GP4, 10 and 12 mm, but both remained GG2 on systematic surveillance biopsy.

Table 2: Gleason pattern 4 quantification metrics and change between diagnostic and surveillance biopsies^a.

	All patients	Excluding patients without GP4 on surveillance biopsy
Characteristic	N = 61	N = 26
Gleason score at surveillance biopsy
6	23 (38%)
3+4	24 (39%)	23 (88%)
4+3	2 (3.3%)	2 (7.7%)
9	1 (1.6%)	1 (3.8%)
No cancer	11 (18%)
Total length of GP4 (mm)
Diagnostic biopsy	0.25 (0.12, 0.44)	0.25 (0.12, 0.44)
Surveillance biopsy	0.00 (0.00, 0.28)	0.46 (0.21, 1.0)
Change	−0.12 (−0.31, 0.09)	0.19 (−0.04, 0.67)
Overall percent of GP4
Diagnostic biopsy	5.0 (2.2, 10)	5.0 (2.2, 10)
Surveillance biopsy	0.00 (0.00, 4.9)	5.0 (4.1, 13)
Change	−2.5 (−8.6, 0.00)	1.2 (−1.6, 6.6)

^aData are expressed as median (IQR) or n (%).

GP4 = Gleason pattern 4.

Table 3: Gleason Pattern 4 metrics and changes excluding patients who met definitions of progression.

Characteristic	Median (IQR)
Summary across all biopsy cores excluding 3 surveillance biopsies with GG3 disease (n=119)
Total length of GP4 on biopsy (mm)	0.15 (0.00, 0.36)
Overall percent GP4	3.2 (0.00, 8.0)
Change in GP4, excluding 3 men with GG3 on surveillance biopsy (n=58)
Change in total GP4 length (mm)	−0.13 (−0.32, 0.02)
Change in overall percent GP4	−2.6 (−8.9, −0.24)
Change in GP4, excluding 5 men with GG3 OR increase of >1 mm with at least ≥2 mm total on surveillance biopsy (n=56)
Change in total GP4 length (mm)	−0.15 (−0.32, −0.04)
Change in overall percent GP4	−2.8 (−9.2, −0.42)

GG3: Grade Group 3, GP4: Gleason pattern 4.

Table 4: Gleason Pattern 4 features among 5 patients who, on surveillance biopsy, had GG3 disease or an increase of >1 mm in total pattern 4 that resulted in ≥2 mm pattern 4 across all cores. Each row represents a unique patient who satisfied either progression criterion.

Patient	Progression Criterion	Total length of GP4 on diagnostic biopsy (mm)	Increase in total length of GP4 (mm)	Total length of GP4 on surveillance biopsy (mm)	Overall % GP4 on diagnostic biopsy	Increase in overall % GP4	Overall % GP4 on surveillance biopsy
1	GG3	0.36	0.69	1.0	3.9	66	70
2	GG3	0.81	1.4	2.2	25	30	55
3	GG4	0.15	13	13	1.2	34	36
4	Increase in GP4	0.44	10	10	9.6	28	37
5	Increase in GP4	0.08	2.1	2.2	5.0	8.3	13

The relationship between the GP4 metrics and progression is displayed in Figure 2. One patient with GG3 on surveillance biopsy only had 1.0 mm of GP4. Those who have GG3 disease have the highest values of overall %GP4 (Figure 2). Between the diagnostic and the 1-year surveillance biopsy, this patient’s overall %GP4 increased from 9.6% to 37% and their total length of GP4 increased from 0.44 to 10.4 mm. Although this patient did not satisfy the standard definition of progression on AS (ie, GG3) they did satisfy our proposed definition which includes, in addition to those with GG3 disease, those that have an increase of >1 mm of length of GP4 which results in over 2 mm of GP4 total (Figure 2). The other patient who satisfied our updated proposed definition had an increase of 2.1 mm and an increase of 8.3% resulting in total length of 2.2 mm pattern 4 and 13% overall pattern 4 (Figure 2). Figure 2 also demonstrates larger increases in overall %GP4 among 12 patients who did not satisfy our proposed criteria. This shows, as expected, a strong positive relationship between length GP4 and %GP4. However, clearly there are some patients with a high %GP4 but limited total GP4 length and vice versa.

Figure 2: Scatterplot of total length of GP4 versus overall percentage of GP44. The black circles represent those without GG3 disease, white circles those ≥ GG3, and white squares those who had an increase of >1 mm resulting in a total length of ≥2 mm pattern.

Discussion

The findings of the current study provide an insight into changes in GP4 in men on AS for GG2 prostate cancer. Additionally, our findings highlight the limitations of using identification of GG3 disease on surveillance as a sole trigger for intervention. We found select patients with large increases in GP4 length who did not qualify to be upgrading to GG3 and, conversely, a patient upgraded to GG3 despite a small increase in GP4 length. Although not of immediate clinical relevance for patients on AS - due to our exclusion of MRI-targeted biopsy cores – our findings raise questions about current approaches to grading and triggers for AS progression.

Quantification of GP4 disease is increasingly being recognized as a meaningful pathologic marker to further sub-stratify GG2. Traditionally, the use of the ratio of GP3:GP4 dictates treatment, with a ratio of GP4 ≥50% characterizing GG3 disease. Better defined strata of GP3:GP4 ratios on biopsy have demonstrated utility in predicting adverse pathology on prostatectomy and biochemical failure^10,11. Sauter et al reported that with increasing deciles of GP3:GP4%, the odds of adverse pathology increased progressively¹¹, a finding which has been corroborated by comparable series¹⁰.

There are biologic reasons to suggest that quantification of GP4 (mm) is preferable to a ratio of GP3:GP4 disease. For example, the use of such ratios suggests oncologic aggressiveness dependent on the amount of comparatively indolent GP3 disease, rather than the absolute quantity of GP4 disease. As such, our group has compared the utility of various methods of quantification of GP4, including total length of GP4 (mm) and total % of GP4^6,12. Increasing length of GP4 on initial biopsy is associated with increased risk of post-prostatectomy adverse pathology⁶ and poorer oncologic outcomes⁸. Specifically, total length of GP4 was more predictive of adverse pathology than total %GP4⁶. These findings for total length GP4 remained true for risk of BCR (HR=2.48, 95% CI 1.36–4.52) compared to overall %GP4 (HR 1.61, 95% CI 1.21–2.15)⁸. Recently, Delahunt et al compared total length of GP4 to total %GP4 using data from the TROG 03.04 trial⁷. The resulting Harrell’s concordance index for relative prognostic significance for %GP4 at total length GP4 was comparable with respect to risk of prostate cancer– specific mortality (0.736, 95% CI 0.647–0.773 versus 0.765, 95% CI 0.685–0.796). In the TROG 03.04 trial, the precise methodology for calculating total GP4 length was not clarified and thus direct comparison with the current study results is not possible. Nevertheless, our experience highlights the utility of precise quantification of the volume of GP4 in place of a ratio of GP3:GP4 disease.

Assessment of GP4 metrics and quantification of GP4 on surveillance biopsy highlights the limitations of relying on a GP3:GP4 ratio cutpoint as a sole trigger for intervention in AS patients. In this study, we proposed an additional criteria of substantial change in GP4, defined as an increase in GP4 length of >1 mm resulting in at least 2 mm GP4. Using this additional measure of pathologic progression, we identified two patients who satisfied this criterion but not the criteria for GG3. Of these, one patient had an increase in GP4 from 0.44 mm to 10 mm on surveillance biopsy, with a corresponding change in %GP4 from 9% to 37%. Conversely, one patient had an increase in GP4 length from 0.36 mm to 0.69 mm, with a corresponding change in %GP4 from 4% to 70%; this patient met the criteria for GG3 but not the GP4 length criterion. There is no doubt that the disease identified in the first scenario represents a more significant disease progression than the second case. Indeed, based on the aforementioned work by Dean et al, an increase of 10 mm of GP4 increases the risk of adverse pathology by over 70%⁶. While these were select patients, our findings challenge the notion of using progression to GG3 alone as trigger for intervention. It should be noted that while the limitations of our study preclude formal recommendations based on these findings, our results provide a ‘proof-of-principle’ for a composite measure for determining pathologic progression in patients with GG2 disease on AS that includes quantification of GP4 length.

The findings of the current study highlight the progression in disease course in these patients and provide an understanding of the underlying disease process. While the risk of pathologic upgrading on surveillance biopsy and associated risk of adverse oncologic outcomes has been well established^2,13,14, assessment of pathologic GP4 changes in this setting has been poorly characterized. Our findings highlight that on surveillance biopsy, most patients had no demonstrable GP4 disease. In patients with demonstrable GP4 disease on the surveillance biopsy, the mean change is likely of minimal clinical consequence (increase of 0.19 mm). Comparable literature assessing GP4 change over time is lacking, particularly in the setting of active surveillance. Indeed, in this setting, assessment of tumor volume progression has been best studied radiologically rather than pathologically. Serial MRI suggests tumor volume progression of approximately 40% per year^15,16. However, it is clear that these radiologic measures cannot be used as a surrogate for pathologic outcomes. Accordingly, we provide an insight into the changes in GP4 disease during AS, which may be critical in future identification of patients at risk of adverse oncologic outcomes based on unfavorable changes in GP4 over time. Discordant grading based off either the ratio or length (mm) method has a variety of clinical implications in scenarios that rely on accurate quantification of pattern 4. Differences in quantification of GP4 results in decisions to either continue active surveillance or recommend treatment, whether to add or omit systemic therapy in intermediate risk prostate cancer as well the decision to perform a pelvic lymph node dissection in radical prostatectomy.

There are several limitations to the current study, but most pertinently, this pertains to MRI-targeted biopsies. In the current analysis, we opted to exclude biopsies that were collected with MRI targeting. MRI-targeted cores were heterogenous across the cohort with respect to method and number of cores, largely based on surgeon and institutional preference. There is no widely accepted method for quantifying GP4 for multiple cores targeted at the same MRI lesion. In particular, simple summing GP4 from multiple targeted cores would likely overrepresent the quantity of GP4. Exclusion of MRI-targeted cores allows interrogation of the underlying disease biology of GP4 disease in men on AS without introducing the unknown element of the method of GP4 quantification. Additionally, it has been recently questioned whether cancers found by MRI targeting are, grade-for-grade, equivalent to those found on systematic biopsy, particularly as many high-grade cancers are found by MRI targeting in patients with negative systematic biopsies, a group known to have extremely low long-term prostate cancer mortality^17,18. Considering the exclusion of MRI-targeted cores, meaningful clinical recommendations cannot be drawn.

Indeed, the purpose of this study was to provide insights into GP4 biology rather than recommendations to guide clinical practice. We demonstrate an alternate method to quantify GP4 as a proof of concept. Future research must carefully consider how to approach the incorporation of MRI-targeted cores to avoid overrepresentation of GP4 disease.

Conclusion

Our study provides insight into GP4 changes during the course of AS in men with GG2 prostate cancer. We demonstrated that some patients had large increases in GP4 length (mm) but did not meet criteria for diagnosis of GG3 disease based on the ratio of GP3 to GP4 disease. Conversely, one patient met criteria for upgrading to GG3 despite a minimal change in length of GP4. Our results suggest that length of GP4, in addition to reclassification to GG3 disease, should be considered in assessing disease progression or reclassification in men with GG2 disease on AS. Further studies are required to confirm our findings, and to discern an optimal approach to inclusion of MRI-targeted cores.

Highlights.

• During active surveillance (AS) for Grade Group (GG) 2 prostate cancer, pathologic progression to GG3 on surveillance biopsy is a trigger for intervention.

• The ratio of GP3:GP4, may be obscured by increases of relatively indolent disease and therefore quantification of GP4 remains ill defined

• We found large increases in GP4 length (mm) do not always meet criteria for diagnosis of GG3 disease whilst upgrading to GG3 can occur despite a minimal change in length of GP4 (mm).

• Our results suggest that length of GP4, in addition to reclassification to GG3 disease, should be considered in assessing disease progression or reclassification in men with GG2 disease on AS.