Editorial: The prognostic value of prostate biopsy grade – Forever a product of sampling
The ability to project clinical outcomes based on limited data is crucial to the practice of medicine. This principle is particularly germane to the management of prostate cancer, where clinical outcomes vary widely. In the current issue of BJUI, Danneman et al. [1] assess pathological grade concordance between diagnostic needle biopsy and subsequent radical prostatectomy specimens from 2000 to 2012. The authors observed increased concordance of biopsy and prostatectomy Gleason scores over the time period (from 55% in 2000 to 68% in 2012) with the majority of improvement occurring before 2005. Interestingly, concordance decreased over time (from 68 to 57%) with use of the newly revised grading system. These and other findings led to the proposal that increased concordance was attributable more to the elimination of Gleason scores 2–5 than the systematic change in grading itself.
We commend the authors for exploring this important topic. Our ability to derive meaningful information on disease biology and behaviour from biopsy specimens is essential to counselling patients on the many available management options. At the same time, biopsy grading is inherently limited in its ability to predict overall prostate pathology because it is not only dependent on architecture and morphology, but also on the, admittedly minimal, sample of tissue obtained. As such, we should be cautious in using terms such as ‘undergrading’ in describing biopsy specimens, which may have been properly graded, but simply lacked the higher grade tumour observed at prostatectomy. In reality, such a phenomenon represents undersampling rather than undergrading, and there is hope that such undersampling will decrease with improved methods of detection, such as multiparametric MRI/TRUS fusion-guided biopsy.
Notably, the authors refer to the updated grading system, which was first described by Dr Epstein and validated in a multi-institutional study [2] before the 2014 International Society of Urological Pathology (ISUP) consensus conference, as ISUP grades 1–5. For clarity, it should be noted that the initial report and validation of the new system [2], the 2014 ISUP consensus conference proceedings [3] and the WHO 2016 edition of Pathology and Genetics: Tumours of the Urinary System and Male Genital Organs [4], have all described the new system based on grade groups 1–5. Consistent use of the adopted terminology will be helpful moving forward.
Nonetheless, there are several potential explanations for the patterns observed in the present study. As the authors note, lower concordance based on the grade group system can be largely explained by the more precise classification of Gleason score 7 cancers. Based on evidence of disparate outcomes in Gleason score 3+4 = 7 and 4+3 = 7 disease [5], the ISUP system distinctly classifies these cancers as prognostic grades 2 and 3, respectively. Certainly, when compared with a system in which Gleason score 7 represents a single classification, one would expect poorer concordance in the more widely distributed group. We believe the clinical utility of separating these classifications far outweighs a modest decrease in concordance, which may be explained by other factors in any case. Previous studies have shown the importance of subdividing the Gleason score 7 population when comparing grading systems [6]. Furthermore, details are not provided as to whether a global grade was assigned to biopsy, a common practice in Sweden, which is not the currently recommended practice. That 5–7% of specimens received a Gleason score < 6 calls into question whether contemporary recommendations were fully adopted during the study period.
Regardless, Danneman et al. elegantly highlight the frequency with which biopsy and prostatectomy grades are discordant, and the fact that, to date, pathological grading remains a subjective practice. As noted, there are widespread efforts to address both of these issues, including the use of targeted biopsies and tissue-based genomic markers. Until these practices are well-validated and widely implemented, there are several reasons to believe the most recent grade group system will improve contemporary practice, despite limited concordance. For one, use of a more intuitive scale ranging from 1 to 5 should prove easier for patients to understand, a significant consideration in light of the information overload patients absorb with a new diagnosis of cancer. Furthermore, available data to this point demonstrate excellent prognostic value. In one study from Johns Hopkins, the revised Grade Group system showed improved accuracy for predicting 5-year metastasis (C-index 0.80 vs 0.70) and 10-year prostate cancer-specific mortality (C-index 0.77 vs 0.64) as compared with the original Gleason score [7].
Until truly objective methods of pathological assessment emerge, additional validation of the new grade group system is likely to further support its use moving forward. As Danneman et al. point out, however, we must keep in mind that biopsy, although perhaps our most useful tool, captures only a small fraction of the overall picture.
Re: Prognostic value of prostate biopsy grade: forever
a product of sampling
Lars Egevad1, Daniela Danneman1, Hemamali Samaratunga2, Brett Delahunt3
1Department of Oncology-Pathology, Karolinska Institutet, Stockholm, Sweden
2Aquesta Pathology and University of Queensland, Brisbane, Queensland, Australia
3Department of Pathology and Molecular Medicine, Wellington School of Medicine and Health Sciences, University of Otago, Wellington, New Zealand
We thank Tosoian and Epstein for their interest in our article Accuracy of prostate biopsies for predicting Gleason score in radical prostatectomy specimens: nationwide trends 2000-2012.
We agree that the underestimation of prostate cancer grade by core biopsies is a serious limitation in pre-operative patient counseling. The authors are correct in suggesting that the term under-grading may be misinterpreted as a diagnostic error in the reading of biopsies, while a substantial part of the observed discrepancy between biopsy and prostatectomy grade is more likely due to sampling error. However, sampling problems are not the only component in biopsy under-grading. Multiple studies have demonstrated a considerable interobserver variation in grading of biopsy specimens and this may also lead to underestimation of grade. [1] Replacing the term under-grading with under-sampling in this context would indicate that the only problem is tumor tissue sampling, while pattern misinterpretation in a small specimen may play a role as well. The term under-sampling furthermore suggests that too little cancer has been sampled, although the biopsy material may be fully adequate.
Tosoian and Epstein note that they suspect that global Gleason score (GS) has been employed in the biopsy reports of our study and claim that the grading is therefore not in line with current recommendations. They are indeed correct in their assumption that, in addition to assigning a GS to each core, virtually all Swedish pathologists also provide a global GS in the summary diagnosis, i.e. a GS based on all cancer seen in the cores rather than the highest GS of an individual core. However, we disagree with the statement that this is contrary to recommended practice. There are no international recommendations that a global GS should not be reported. The International Society of Urological Pathology (ISUP) 2005 consensus proceedings state that a GS should be given for each core (or container), but there was no consensus as to whether grading should be core or case based. [2] At the 2014 ISUP consensus conference, this issue was not discussed. [3] The International Collaboration for Cancer Reporting (ICCR) guidelines accept both options by stating that depending on local practice, the different elements of grade data may be reported at either core or specimen level or as a composite (global) grade, based upon all cancer present in the biopsy cores or a combination of both. In a survey among 266 European uropathologists, it was found that 77% would calculate a global GS, while only 17% would use the highest score present within a single core. [4] There are sound arguments for using either highest or global GS and we will not revisit the literature on this. It is noteworthy; however, that on a population basis, the highest GS may better correlate with prostatectomy GS, which is not surprising as biopsies tend to under-grade (under-sample) the cancer present. However, in the individual patient, the reporting of highest GS may be misleading as a small biopsy fragment containing e.g. GS 8 cancer may overestimate the GS of a tumor that is actually GS 7. There is a risk that such a practice would further fuel Gleason inflation. Moreover, in one study, global GS was shown to be an independent prognostic factor in endocrine-treated prostate cancer, while highest GS was not. [5] In reality there can be no certainty whether two cores with conflicting GS have sampled the same tumor focus or separate tumors. The tradition in Sweden and major parts of Europe is to err on the lower side.
It is also claimed that since 5-7% of biopsy specimens were reported with GS <6, contemporary recommendations were not fully adopted during the study period. However, as is shown in Fig. 1, the reporting of such scores was prevalent during the years 2000 – 2005, i.e. before the publication of the ISUP 2005 recommendations, while by 2009 this had been almost completely abandoned. This is in line with what can be expected when new guidelines are introduced and in reality shows a very rapid adjustment to the, then newly adopted, ISUP grading recommendations.
As the authors mention the terminology for prostate cancer grading has been a subject of intense debate since the ISUP 2014 consensus conference. From the outset the ISUP had determined that the correct terminology for the grading system would be “ISUP Grade” and this was unanimously ratified by the Council of the Society at its annual meeting in March 2015. This term is now utilized internationally and is the formal designation adopted by the International Collaboration for Cancer Reporting (ICCR) for use in all of its prostate cancer protocols. It has come to our attention that the 2016 guidelines on screening, diagnosis and local treatment with curative intent of prostate carcinoma, produced by the European Association of Urology (EAU), the European Society for Radiotherapy (ESTRO) and the International Society of Geriatric Oncology (SIOG) also refer to contemporary prostate cancer grading as ISUP Grade. Because of conflicting opinions on the matter the WHO editorial meeting in 2015 decided that no formal recommendation on nomenclature should be given in the Blue Book. [6] The descriptive term grade groups was used in the text with the explicit purpose of allowing the pathology community to develop their preferred terminology. Some authors have misunderstood this as a WHO recommendation. Grade group is a confusing term as ISUP grades do not group Gleason grades but rather Gleason scores. Furthermore, grade group is often abbreviated GG which used to be the abbreviation for Gleason grade (or pattern). Unfortunately the ISUP 2014 consensus meeting proceedings may give the impression that the grade group terminology was supported by the consensus conference, which was not the case. [3] This was a last minute addition to the proof made by one of the authors of the proceedings, without the knowledge of at least some of the co-authors. It is therefore our standpoint that the term ISUP grade is preferred in order to acknowledge the central role of the ISUP in its development and dissemination.
1 Engers R. Reproducibility and reliability of tumor grading in urological neoplasms. World J Urol 2007; 25: 595-605
2 Epstein JI, Allsbrook WC, Jr., Amin MB, Egevad L. The 2005 International Society of Urological Pathology (ISUP) Consensus Conference on Gleason Grading of Prostatic Carcinoma. Am J Surg Pathol 2005; 29: 1228-42
3 Epstein JI, Egevad L, Amin MB, Delahunt B, Srigley JR, Humphrey PA. The 2014 International Society of Urological Pathology (ISUP) Consensus Conference on Gleason Grading of Prostatic Carcinoma: Definition of Grading Patterns and Proposal for a New Grading System. Am J Surg Pathol 2016; 40: 244-52
4 Berney DM, Algaba F, Camparo P, et al. Variation in reporting of cancer extent and benign histology in prostate biopsies among European pathologists. Virchows Arch 2014; 464: 583-7
5 Tolonen TT, Kujala PM, Tammela TL, Tuominen VJ, Isola JJ, Visakorpi T. Overall and worst gleason scores are equally good predictors of prostate cancer progression. BMC Urol 2011; 11: 21
6 Moch H, Humphrey PA, Ulbright TM, Reuter VR. WHO Classification of Tumours of the Urinary System and Male Genital Organs. 4th edn, Lyon: WHO Press, 2016
Conflicts of interest: None