Practice Patterns in Reporting Tertiary Grades at Radical Prostatectomy: Survey of a Large Group of Experienced Urologic Pathologists Open Access

In the spring of 2018, a survey was designed and circulated to 105 experienced international urologic pathologists via an online survey program, SurveyMonkey Inc (San Mateo, California). Pathologists invited to participate were identified solely by the authors as academically engaged genitourinary pathologists from a range of institutions. The 10 multiple choice-style questions circulated are summarized in Table 1. The survey informed potential respondents that it would evaluate their perspectives/practice patterns related to tertiary grading in RP specimens. The first question assessed how participants define a “tertiary” pattern, which was followed by 8 case-based grading scenarios. The final question asked if reporting tertiary patterns would affect assignment of Grade Group for an individual RP case. All replies were tabulated and analyzed with descriptive statistics.

Ninety-five pathologists responded to the online survey and are listed in Supplemental Table 1 (see supplemental digital content at www.archivesofpathology.org in the March 2020 table of contents). Responses were tabulated on a per question basis, as well as an interaction of responses (ie, depending on how a respondent defined “tertiary,” were his/her subsequent responses consistent with that definition), as delineated below.

Fifty-five of 95 pathologists (58%) responded that TP is a minor pattern/less than 5% of the tumor, while 40 (42%) defined TP as the third most common pattern.

In a tumor with overwhelming pattern 3 and less than 5% pattern 4, of the 95 participants, 56 (59%) would assign 3 + 4 = 7, with 40 also including a comment regarding percentage of pattern 4, while 35 (37%) would assign 3 + 3 = 6 with a minor/tertiary component pattern 4. Four respondents would not document pattern 4 in this scenario, assigning 3 + 3 = 6.

In a tumor with overwhelming pattern 4 and less than 5% pattern 5, of the 95 participants, 51 (54%) would assign 4 + 4 = 8 with a minor/tertiary component of pattern 5, while 43 (45%) would assign 4 + 5 = 9, with 23 including a comment regarding percentage of pattern 5. One respondent would not document pattern 5 in this scenario, assigning 4 + 4 = 8.

All respondents would address the presence of pattern 5. The responses for Group 1 are summarized in Table 2.

In a scenario with less than 5% pattern 5, of the 95 respondents, 93 (98%) would assign 3 + 4 = 7 with TP5. In scenarios with either 15% or 25% pattern 5, of the 95 respondents, 66 (69%) and 75 (79%) would assign 3 + 5 = 8, with half including a comment, respectively. Conversely, 29 of 95 (31%) and 20 of 95 (21%) would assign 3 + 4 = 7 with TP5.

All respondents would address the presence of pattern 5. The responses for Group 2 are summarized in Table 3.

In a scenario with less than 5% pattern 5, of the 95 respondents, 88 (93%) would assign 4 + 3 = 7 with TP5, while the remaining 7 would assign 4 + 5 = 9. In scenarios with either 15% or 25% pattern 5, of the 95 respondents, 67 (71%) and 76 (80%) would assign 4 + 5 = 9, respectively, with one-third including a comment. Conversely, 28 of 95 (29%) and 19 of 95 (20%) would assign 4 + 3 = 7 with TP5.

For group 1 (order of most common patterns = 3, 4, and 5), among 55 respondents who defined TP as a minor pattern/less than 5% of the tumor, 98%, 9%, and 3% would assign 3 + 4 = 7 with TP5 in scenarios with percentage pattern 5 of less than 5%, 15%, and 25%, respectively. Among the 40 respondents who defined TP as the third most common pattern, 98%, 60%, and 45% would assign 3 + 4 = 7 with TP5 in the same scenarios.

For Group 2 (order of most common patterns = 4, 3, and 5), among 55 respondents who defined TP as a minor pattern/less than 5% of the tumor, 95%, 11%, and 4%, would assign 4 + 3 = 7 with TP5 in scenarios with percentage pattern 5 of less than 5%, 15%, and 25%, respectively. Among the 40 respondents who defined TP as the third most common pattern, 90%, 55%, and 43% would assign 4 + 3 = 7 with TP5 in the same scenarios.

Eighty-five of 95 (89%) responded that a TP would not impact their assignment of Grade Group to an individual case.

The application of prostate cancer grading has undergone significant evolution during the past 2 decades, in part owing to systematic needle biopsy sampling as well as the centrality of tumor grading in predictive models.^1,21–25  The 2005 and 2014 ISUP grading conferences codified much of this evolution, including endorsement of a patient-centered prognostic Grade Group system.^19,20  Fundamentally, however, prostate cancer grading still relies on identifying the first and second most common histologic patterns to reach a Gleason score and corresponding Grade Group. For prostate glands harboring more than 2 grades, “tertiary” patterns, typically of higher grade, have a reported incidence of 4% to 50% in RP studies.^2,3,13,14,35  Several elements regarding TP remain controversial, including its definition (limited versus no limit), association with recurrence rates (intermediate between grades versus akin to next highest Grade Group), and more recently, its effect on Grade Group assignment. While the ISUP 2014 report suggested that a TP be defined exclusively as truly minor/limited (<5%) and should not alter the Grade Group, there is conflicting evidence in contemporary cohorts.^{9–12,15,16,18} 

Importantly, although reporting of TPs is recommended for RP specimens, there is discordance with clinical predictive models that do not include TP in a standard fashion.^23–25  This is likely due to the reality that almost all TP studies are single institution reports, may be underpowered to show clear prognostic significance in multivariable analysis, often have short follow-up, and have not yet shown utility in clinical decision-making. From the pathology perspective, there has been nearly no attention given to the variability of tertiary grading pattern definition and its application in specific grading scenarios. In that vein, the current survey of a large group of experienced genitourinary pathologists reveals a range of practice patterns for TP reporting.

The definition of tertiary pattern and its upper limit have been a source of ongoing controversy. In an early article by Pan et al³  and subsequent articles from Johns Hopkins Hospital, TP has exclusively referred to very limited amounts (<5%) of a third most common pattern, usually pattern 5.^4,11,12,18  While some investigators have adopted this definition, this has not been universal, with other studies having no specific cutoff.^6–8,10  In these latter reports, grading scenarios with up to ∼30% pattern 5 could still be considered a TP. In a survey on Gleason grading published in 2005 preceding the ISUP grading conference, 46% (27 of 59) of genitourinary pathologists responded that they would use a 5% limit for inclusion of a TP, whereas 54% had no set limit.³⁶  While there was consensus at the 2005 ISUP grading conference that the Gleason score for RP specimens is assigned on the basis of primary and secondary patterns with a comment as to a TP (if present), no limit for TP was specified.¹⁹ 

In a report addressing practice-related issues resulting from the proceedings of the 2014 ISUP grading conference, it was noted that although no specific cutoff was voted on, 83% of participants favored the term minor high-grade pattern over tertiary grade pattern,⁵  indicating that TP should not just be the third most common pattern but that it should also be minor or limited in extent. In the current survey, conducted in 2018, 58% responded that TP refers to a minor pattern/less than 5% of the tumor, while 42% would still define TP as simply the third most common pattern. Therefore, while an increasing number of experienced urologic pathologists would limit the definition of TP, many would not. As tertiary grades are seldom incorporated within nomograms and prognostic tables, there remains a significant concern that not having a limit for TP could lead to underestimation of risk in grading scenarios with relatively large amounts of pattern 5. Our findings indicate substantial variation in practice, with implications for comparing studies among institutions as well as prognostication post RP.

Among the issues with the “tertiary” terminology are RP grading scenarios with only 2 patterns, in which the second higher-grade pattern is of very limited extent. At the 2014 ISUP grading conference, 75% favored reporting a percentage of pattern 4 for Gleason score 7 cases (Grade Groups 2–3), in both the needle biopsy and RP specimen types.⁵  Theoretically, adoption of this reporting parameter would allow cases reported as Gleason score 3 + 3 = 6 with TP 4 to be reclassified as Gleason score 3 + 4 = 7 with a percentage of pattern 4 noted.³⁷  Similarly, although not specifically addressed at the 2014 conference, it has been posited that for cases formerly reported as Gleason score 4 + 4 = 8 with TP 5, these cases might now be recorded as Gleason score 4 + 5 = 9 with a minor component of pattern 5. Our survey of practice patterns, however, reveals that in these grading scenarios, 37% (for overwhelming pattern 3 and <5% pattern 4) and 54% (for overwhelming pattern 4 and <5% pattern 5) of survey respondents would still assign Gleason score on the basis of the predominant grade alone and note the minor higher-grade component as a tertiary pattern. These findings reveal that reporting a percentage of pattern 4 (and percentage of higher-grade patterns more globally) in RP specimens has not been uniformly embraced.

In line with 2014 ISUP recommendations, we found that an overwhelming majority of respondents—98% for Group 1, 93% for Group 2—would report a TP5 in grading scenarios with 3 grades, in which there was less than 5% pattern 5 (ie, either 3 + 4 = 7 or 4 + 3 = 7 with TP5). Substantial variation, however, was found in grading scenarios with a larger percentage of pattern 5. Specifically, for Group 1 and 2 scenarios with 15% pattern 5, ∼30% of respondents would still assign TP5—as opposed to incorporating pattern 5 as a secondary grade—and in those scenarios with 25% pattern 5, 20% of respondents would still assign TP5, highlighting the degree to which the “minor/limited” definition of TP has not been universally adopted.

Unsurprisingly, when assessing responses to grading scenarios in Groups 1 and 2 through the lens of the “tertiary pattern” definition chosen, we found that for those defining TP as a minor pattern/less than 5% of the tumor, at least 95% would assign either 3 + 4 or 4 + 3 = 7 with TP5. Concordantly, the rate of assigning TP5 fell to approximately 10% for Group 1 and 2 scenarios with 15% pattern 5 and to less than 5% for scenarios with 25% pattern 5.

For those defining TP as the third most common pattern, in scenarios with less than 5% pattern 5, at least 90% would assign either 3 + 4 or 4 + 3 = 7 with TP5. While one might have expected similarly high rates of assigning TP5, owing to the definition chosen, we found that for scenarios with 15% and 25% pattern 5, approximately 40% (Group 1 scenarios) and 55% (Group 2 scenarios) chose to include pattern 5 as the secondary grade, even though it was the third most common pattern, indicating disparity between definition and utilization of TP. Stated differently, nearly half of respondents who defined TP as the third most common pattern adopted a “minor pattern” definition in application to specific grading scenarios.

A further recommendation from the 2014 ISUP grading conference was that minor high-grade patterns should not change Grade Group assignment.^5,20  The current survey results are in line with that recommendation, with nearly 90% responding that a tertiary pattern would not impact their assignment of Grade Group to an individual case. Literature following the 2014 conference—echoing work in the pre–Grade Group era—has produced divergent evidence regarding TP and its impact on Grade Group assignment, with some finding that a TP imparts rates of biochemical recurrence intermediate between grades (ie, Grade Group 3 with TP 5 compared with Grade Groups 3 and 4) and others showing biochemical recurrence rates akin to the next highest Grade Group (ie, Grade Group 3 with TP 5 similar to Grade Group 4).^{9–12,15,16,18}  The latter findings have led some to argue that unlike the 2014 ISUP recommendation, tertiary patterns should be incorporated into the Grade Group system rather than left as a comment.^10,15  Any such recommendation would be highly influenced by how the individual study defined TP and the uniformity in application of that definition in routine practice. The current practice pattern findings suggest great variability even among experienced urologic pathologists in this assessment.

Ultimately, data elements in standard pathology reports should have clinical import, for prognostication and/or management. To date, there has been only limited incorporation of TP in prognostic models and no studies incorporating a decision curve analysis³⁸  to demonstrate a net benefit in clinical decision-making, for example, for the provision of early adjuvant therapy post RP. The current study highlights differences in practice patterns, some governed definitionally, for reporting TP grading. Importantly, while just over half of pathologists surveyed defined TP as a minor pattern/less than 5% of the tumor, the greatest reproducibility among respondents was observed in grading scenarios in which the degree of pattern 5 was truly minor, that is, less than 5%, regardless of definition chosen. As a practice recommendation, these findings suggest that practicing pathologists should consider reporting a tertiary pattern only when a limited amount of a higher-grade, third most common pattern is present. At the same time, future practice might be informed by data from institutions in which percentage of all grading patterns are routinely recorded. Such studies can establish whether there is an optimal percentage above which tumors with 3 patterns no longer behave (eg, biochemical recurrence) akin to cancers with the same primary and secondary grades but lacking a TP. A final critical area for future study is how urologic oncologists use the presence of a TP in patient counseling and decision-making.