Quantification of the individual risk of each Gleason pattern, including tertiary Gleason pattern 5, after radical prostatectomy: development of the modified Gleason grade grouping (mGGG) model

Background While the new Gleason grade grouping (GGG), which started in 2016, has been widely validated in prostate cancer, it does not incorporate the concept of tertiary Gleason pattern 5. Furthermore, no study has “quantified” the individual risk of each Gleason pattern, including tertiary Gleason pattern 5, after radical prostatectomy. Methods We reviewed 1022 men with adjuvant-treatment-naïve prostate cancer who underwent radical prostatectomy between 2005 and 2017. The primary endpoint was biochemical recurrence-free survival, defined as two consecutive prostate-specific antigen measurements ≥0.2 ng/ml after surgery. The individual quantitative risk score (IQRS) of each amount (primary/secondary/tertiary) of each Gleason pattern (3/4/5) was calculated using the Cox regression model. On the basis of the IQRS, the modified Gleason grade grouping (mGGG) model was developed. As a robustness analysis of the mGGG model, salvage treatment-free survival was also assessed. Results During a median follow-up of 45 months, 229 of 1022 (22.4%) patients developed biochemical recurrence. The IQRS of each Gleason pattern was as follows: primary 5, 1.81 points (hazard ratio [HR] 6.13); secondary 5, 1.37 points (HR 3.92); tertiary 5, 0.87 points (HR 2.39); primary 4, 1.07 points (HR 2.91); secondary 4, 0.79 points (HR 2.21); and any Gleason pattern 3, 0 points (HR 1). Based on the IQRS, the mGGG model was developed, which classified patients into the following five groups: I (3 + 3 or less); II (3 + 4); III (4 + 3); IV (3 + 4 + t5, 4 + 3 + t5, 3 + 5, 5 + 3, and 4 + 4); V (4 + 4 + t5, 4 + 5, 5 + 4, and 5 + 5). The c-index for biochemical recurrence-free survival was significantly improved from 0.655 of the original GGG model to 0.672 of the mGGG model (P < 0.05). In the robustness analysis, the c-index for salvage treatment-free survival was also significantly improved from 0.619 of the original GGG model to 0.638 of the mGGG model (P < 0.05). Conclusions The quantitative risk of tertiary (< 5%) Gleason pattern 5 is slightly higher than that of secondary (5–50%) Gleason pattern 4. Our newly developed mGGG model more accurately predicts outcomes after radical prostatectomy than the original GGG model.


Background
Since the Gleason scoring system for pathological diagnosis of prostate cancer (PC) was published in 1966 [1], it has been widely used in clinical practice and has evolved over time. The concept of tertiary (< 5%) Gleason pattern 5 [2] has been implemented since the 2005 International Society of Urological Pathology (ISUP) Consensus Conference on Gleason Grading of Prostatic Carcinoma [3]. The prognostic significance of tertiary Gleason pattern 5 has been widely validated in various settings of PC [4][5][6][7][8][9][10][11][12][13][14][15][16]. In 2016, the new Gleason grade grouping (GGG) [17] started to be used in clinical practice, according to the 2014 ISUP Consensus Conference on Gleason Grading of Prostatic Carcinoma [18]. Although the prognostic value of GGG has been confirmed in several settings [19][20][21], it does not incorporate the concept of tertiary Gleason pattern 5. A recent study reported that integrating tertiary Gleason pattern 5 into GGG improved the accuracy of predicting the patient's outcome after radical prostatectomy [16]. However, no study has assessed the actual effect of tertiary Gleason pattern 5 on outcomes of patients with PC. Therefore, the present study aimed to quantify the risk of each Gleason pattern, including tertiary Gleason pattern 5, after radical prostatectomy and to develop the modified Gleason grade grouping (mGGG) model.

Methods
The internal institutional review board of the Graduate School of Medicine and Faculty of Medicine, The University of Tokyo approved this retrospective study (approval number: 3124). We reviewed 1167 patients with PC who underwent radical prostatectomy at The University of Tokyo Hospital between 2005 and 2017. We excluded 32 patients who had received neoadjuvant androgen deprivation therapy (ADT) to guarantee an accurate pathological diagnosis. We also excluded 113 patients who received adjuvant treatment after surgery (radiotherapy, n = 14; ADT, n = 64; and both, n = 35) to assess the pure oncological outcome of surgery alone. Eventually, 1022 adjuvanttreatment-naïve patients were available for analysis ( Fig. 1). All prostatectomy specimens were pathologically reviewed according to the 2005 ISUP Consensus Conference on Gleason Grading of Prostatic Carcinoma [3].
The primary endpoint was biochemical recurrencefree survival (BRFS), defined as two consecutive prostate-specific antigen (PSA) measurements ≥0.2 ng/ ml after surgery [22]. Salvage radiotherapy and/or ADT was implemented when biochemical and/or radiological recurrence was observed [22][23][24][25]. We used Kaplan-Meier analysis with the log-rank test to determine survival differences among categories. Univariate and multivariate Cox proportional hazards regression models were also used to predict outcomes. Follow-up information was obtained as of March 2018.
The individual quantitative risk score (IQRS, points) of each amount (primary/secondary/tertiary) of each Gleason pattern (3/4/5) was calculated using the Cox regression model, and reported with the corresponding hazard ratio (HR). The HR was calculated from the IQRS by the following formula: HR = e IQRS . The IQRS of any Gleason pattern 3 was set as 0 points (HR 1), because a Gleason score (GS) of 3 + 3 rarely correlates with PC-specific death and metastasis, despite being pathologically malignant [26,27]. On the basis of the IQRS, the mGGG model was finally developed. Concordance (c) indices were calculated both for the mGGG model and the original GGG model. As a robustness analysis of the newly developed mGGG model, we also assessed another endpoint of salvage treatment-free survival.
All statistical analyses were performed using JMP Pro version 14.0.0 and SAS version 9.4 (SAS Institute, Cary, NC, USA). P < 0.05 was considered to indicate a significant difference.

Results
Patients' characteristics are shown in Table 1. Findings of extraprostatic extension and seminal vesicle invasion were included in pathological T stage as pT3a and pT3b, respectively. During a median follow-up of 45 (interquartile range: 21-77) months, 229 of 1022 (22.4%) patients developed biochemical recurrence, and of them, three died of PC. Eight of 1022 (0.8%) patients died from other causes without recurrence, resulting in 237 events for BRFS (Fig. 1).
In the robustness analysis assessing salvage treatmentfree survival, significant differences were observed in the survival profiles of the five groups of the mGGG model (P < 0.0001) (Fig. 3a), whereas the discrimination between group III and group IV of the original GGG model seemed insufficient (Fig. 3b). Also for salvage treatment-free survival, the mGGG model achieved a significantly higher predictive accuracy (c-index: 0.638) than the original GGG model (c-index: 0.619) (P = 0.029).

Discussion
To the best of our knowledge, the present study is the first quantification of the individual risk of each Gleason pattern, including tertiary Gleason pattern 5, after radical prostatectomy. Notably, the quantitative risk of tertiary (< 5%) Gleason pattern 5 was slightly higher than that of secondary (5-50%) Gleason pattern 4. Furthermore, our newly developed mGGG model more accurately predicted outcomes after radical prostatectomy than the original GGG model. The pathological diagnostic system of PC is unique and complex in that it considers the worst and second worst lesions, as well as their percentages of areas [1,3,18], whereas those of other cancers usually simply use the worst lesion. As described above, the Gleason scoring system, which was published in 1966 [1], has evolved over time. The concept of tertiary (< 5%) Gleason pattern 5 [2] has been implemented since 2005 [3] and the GGG [17] has been proposed since 2014 [18], although it does not incorporate the concept of tertiary Gleason pattern 5. Recently, a Japanese research group reported that integrating tertiary Gleason pattern 5 into GGG improved the prediction accuracy of biochemical recurrence after radical prostatectomy using a retrospective cohort of 1000 patients [16]. Similarly, Lucca et al. previously investigated the significance of tertiary Gleason pattern 5 in a large cohort (n = 4146) of patients with localized GS 7 PC treated by radical prostatectomy [13]. They identified the presence of tertiary Gleason pattern 5 as an independent predictor of biochemical recurrence and developed a prognostic model for 416 patients with GS 7 (3 + 4 or 4 + 3) and tertiary Gleason pattern 5. This study judiciously highlighted the importance of tertiary Gleason pattern 5 among patients with GS 7, however, it did not evaluate the relative importance of tertiary Gleason pattern 5 in patients with all GS categories. Actually, no study so far has "quantified" the actual effect of tertiary Gleason pattern 5 on outcomes after radical prostatectomy. In this context, the present study clearly identified the IQRS of each Gleason pattern (3/4/5) and its echelon (primary/secondary/tertiary) ( Table 2). Notably, the IQRS of tertiary (< 5%) Gleason pattern 5 was slightly higher than that of secondary (5-50%) Gleason pattern 4. This finding is in accordance with previous reports, which emphasized the importance of (tertiary) Gleason pattern 5 in various settings of PC [4][5][6][7][8][9][10][11][12][13][14][15][16]. The present study confirmed the paramount importance of Gleason pattern 5, regardless of its amount, in predicting outcomes after surgery.
On the basis of the IQRS of each Gleason pattern, the present study developed the mGGG model, in which the original GGG model was used as a backbone with GS 3 + 4 + t5 (originally group II) and GS 4 + 3 + t5 (originally group II) shifted into group IV and with GS 4 + 4 + t5 (originally group IV) shifted into group V ( Table 3). Given that the mGGG model enables accurate  HR is calculated from the IQRS by the following formula: HR = e IQRS prediction of BRFS after surgery with a higher cindex than the original GGG model, we deem that the mGGG model is a better scoring system than the original GGG. Additionally, the robustness of the mGGG model was confirmed by assessing salvage treatment-free survival, a treatment-oriented endpoint, other than BRFS, a pure oncological endpoint. Recently, the importance of treatment-oriented endpoints has been recognized in the field of oncology and some endpoints such as ADT-free survival have already been assessed in studies of PC [28]. Both for BRFS and salvage treatment-free survival, the predictive accuracy of the mGGG model was higher than that of the original GGG model. This might mean the usefulness of the mGGG model in the real-world clinical practice. This study had several limitations. Firstly, it was a retrospective analysis of a single institution with a relatively short follow-up. Secondly, the pathological review of prostatectomy specimens was based on the 2005, but not 2014, ISUP Consensus Conference on Gleason Grading of Prostatic Carcinoma, although their criteria were generally identical except for some minor revisions. Thirdly, we could not assess GS 3 + 3 with tertiary Gleason pattern 5 and GS 5 + 5 because of lack of applicable cases. Given that the IQRS for GS 3 + 3 + t5 and GS 5 + 5 are theoretically 0.87 points (HR 2.39) and 3.18 points (HR 24.06), respectively, GS 3 + 3 + t5 and GS 5 + 5    would be assigned to the mGGG category II and V, respectively. Further studies should be conducted to externally validate the mGGG model and to finally establish the optimal pathological diagnostic system of PC.

Conclusions
The quantitative risk of tertiary (< 5%) Gleason pattern 5 is slightly higher than that of secondary (5-50%) Gleason pattern 4. Our newly developed mGGG model more accurately predicts outcomes after radical prostatectomy than the original GGG model.