UICC and AJCC 8th edition tumor-nodes-metastasis (TNM) classifications for patients treated with radical prostatectomy: reliable but not infallible prognostic tools

We read with interest the article by Herden and colleagues (1) regarding the validation of the UICC and AJCC 8^th edition tumor-nodes-metastasis (TNM) system on a German community-based dataset. In this study, the authors analyzed data derived from the HAROW study (n=2,957), a prospective, non-interventional, health service research study examining various treatment options for patients with localized prostate cancer (≤ cT2c). Within their study, Herden and colleagues selected all patients that underwent radical prostatectomy (RP) (n=1,738, 58.8%) with available pathologic information and follow-up data (n=515). For validation purpose, the authors stratified the patients in either cT1 or cT2, S I (≤ T2a) or S II (T2b/c), and prognostic stage group (PSG) I, PSG IIA, PSG IIB or PSG ≥ IIC according to the TNM classification and World Health Organization (WHO) grading. All groups were compared regarding proportion of organ confined disease (≤ pT2) and extraprostatic extension (≥ pT3) after RP, as well as favorable (≤2) and unfavorable (≥3) WHO grade group, or favorable (≤7a) and unfavorable (≥7b) Gleason Score after RP, respectively. Finally, biochemical recurrence free-survival (BFS) was examined and compared between groups in each classification systems.

The results showed that cT1/cT2 and S I/II subgrouping are feasible to predict a different pT-category and a favorable WHO grade after RP, but failed to predict a different BFS rates. Indeed, no BFS rate differences were found between cT1 and cT2 or SI and SII, after a median follow-up of 31.7 months. Conversely, the authors found statistically significant differences within PSG groups when pathological or BFS end-points were considered.

The main take home message of this study is that the combination rather than singular tumor characteristics may increase the prediction rate of a staging system. In other words, the combined use of TNM and Gleason grade groups performed better than TNM system alone. As the authors mentioned in the study, several models are available in clinical practice for prediction of histological finding (2,3) or survival outcomes (4) after RP. All of them have showed better accuracy when compared with TNM system or singular tumor factors. Indeed, the inclusion of more variables and the use of logistic or Cox-based models provide better discrimination power for predicting models. Nevertheless, we should always keep in mind that a model, even if it is highly accurate, should be also easy to use in the daily practice (5). Multiple input models with logarithmic variables and nomogram-based calculation might discourage the end-user from the routine use of such tool in clinical practice. Consequently, the AJCC 8^th with integrated PSG grouping might represent a good compromise between accuracy and feasibility in every-day practice.

The strength of the study is the large sample size and the homogeneity of the HAROW population. Inclusion criteria and study design were also appropriate. On the other hand, the study is affected by a high drop-out rate and a short-follow-up, which might have introduced a selection bias, as pointed out by the authors. Moreover, hard clinical end-points, such as clinical recurrence or cancer-specific mortality should be preferred over pathological finding or BFS to better discriminate the clinical utility of prognostic system, especially for patients with localized prostate cancer. Additionally, we do believe that the absence of ethnic variability (only Caucasian patients were enrolled) might have affected the results of this study. It is well known that African-American patients are not accurately classified by prostate cancer scores for Caucasian patients (6,7). Last but not least, missing information on surgical margins, seminal vesicle invasion and lymph node involvement are a major limitation, when BFS is observed. Patients with these risk factors in the RP specimen have worse oncologic outcomes (8-11). Moreover, these patients might have received further treatments after RP, resulting in various BFS.

Taken together, the study of Herden et al. provides interesting insights in the field of prostate cancer staging. Moreover, it underlies the need of a better staging system, which includes more tumor biological variables, but remains down-to-earth for end-users.

Acknowledgements

None.

Footnote

Conflicts of Interest: The authors have no conflicts of interest to declare.

References

1. Herden J, Heidenreich A, Wittekind C, et al. Predictive value of the UICC and AJCC 8th edition tumor-nodes-metastasis (TNM) classification for patients treated with radical prostatectomy. Cancer Epidemiol 2018;56:126-32. [Crossref] [PubMed]
2. Gandaglia G, Fossati N, Zaffuto E, et al. Development and Internal Validation of a Novel Model to Identify the Candidates for Extended Pelvic Lymph Node Dissection in Prostate Cancer. Eur Urol 2017;72:632-40. [Crossref] [PubMed]
3. Bandini M, Marchioni M, Pompe RS, et al. First North American validation and head-to-head comparison of four preoperative nomograms for prediction of lymph node invasion before radical prostatectomy. BJU Int 2018;121:592-9. [Crossref] [PubMed]
4. Pompe RS, Bandini M, Preisser F, et al. Contemporary approach to predict early biochemical recurrence after radical prostatectomy: update of the Walz nomogram. Prostate Cancer Prostatic Dis 2018;21:386-93. [Crossref] [PubMed]
5. Bandini M, Fossati N, Briganti A. Nomograms in urologic oncology, advantages and disadvantages. Curr Opin Urol 2019;29:42-51. [Crossref] [PubMed]
6. Bandini M, Marchioni M, Preisser F, et al. A Head-to-head Comparison of Four Prognostic Models for Prediction of Lymph Node Invasion in African American and Caucasian Individuals. Eur Urol Focus 2017. [Epub ahead of print]. [Crossref] [PubMed]
7. Bandini M, Preisser F, Nazzani S, et al. The Effect of Other-cause Mortality Adjustment on Access to Alternative Treatment Modalities for Localized Prostate Cancer Among African American Patients. Eur Urol Oncol 2018;1:215-22. [Crossref]
8. Preisser F, Mazzone E, Knipper S, et al. Rates of Positive Surgical Margins and Their Effect on Cancer-specific Mortality at Radical Prostatectomy for Patients With Clinically Localized Prostate Cancer. Clin Genitourin Cancer 2019;17:e130-9. [Crossref] [PubMed]
9. Preisser F, Marchioni M, Nazzani S, et al. The Impact of Lymph Node Metastases Burden at Radical Prostatectomy. Eur Urol Focus 2018. [Epub ahead of print]. [Crossref] [PubMed]
10. Preisser F, Bandini M, Marchioni M, et al. Extent of lymph node dissection improves survival in prostate cancer patients treated with radical prostatectomy without lymph node invasion. Prostate 2018;78:469-75. [Crossref] [PubMed]
11. Preisser F, Bandini M, Graefen M, et al. Reply to the letter to the editor: RE: Preisser F, et al. Extent of lymph node dissection improves survival in prostate cancer patients treated with radical prostatectomy without lymph node invasion. The Prostate. 2018;1-7. Prostate 2018;78:692.