- Top of page
- Conflict of interest
The TNM classification system by the UICC and the AJCC is used worldwide for the stage classification of malignancies, including PC. This TNM classification correlates the prognoses for PC and is used to select the treatment modalities. However, serum levels of PSA and a histological scoring system known as GS correlate strongly with the prognosis, as well as with TNM staging. Thus, several risk classification systems have been developed for localized or locally advanced PC in combination with multifactorial prognostic factors.[1-3] D'Amico's classification was a system of risk stratification originally developed to guide in the selection of radiation therapy for localized or locally advanced PC. D'Amico's classification has also been validated for surgical treatment in several studies. Based on this historical background, the UICC and AJCC developed PG for PC in the TNM seventh edition by incorporating PSA and GS into TNM SG in 2011.[6, 7] The PG classifies localized PC (SG-I and II) into three groups (PG-I, IIA and IIB) using PSA cut-off values of 10 and 20 ng/mL, and GS cut-off values of 6 or less, 7 and 8 or more. This classification system is almost identical to the risk classification by D'Amico, where low, intermediate and high risk correspond to PG-I, PG-IIA and PG-IIB/III, respectively.
As PC is an androgen-sensitive malignant tumor, PADT is used worldwide for patients with metastatic disease or locally advanced PC. Surgical treatment or radiation therapy is the gold standard for treating localized PC, and PADT is an option to provide sufficient outcomes. To improve the prognosis for these patients, ADT in combination with such definitive therapies is recommended as combination therapy.[8-10] PADT is also a reasonable alternative option for local or locally advanced disease in various patient conditions, such as in elderly patients, those with poor performance status or those with particular hopes (rejection for operation and radiation therapy).[11, 12] In fact, PADT including complete androgen blockade, castration, anti-androgen agent monotherapy and intermittent administration is used in 14.1–49.8% of patients with localized PC in Japan and Western countries.[13-16] Thus, ADT plays crucial roles in the treatment of various stages of PC; however, mostly in metastatic disease and not in all stages, and it is important to elucidate the clinical outcomes of PADT.
To understand the reality and treatment outcomes of ADT, we have started a nationwide longitudinal cohort study called the J-CaP study. In the J-CaP study, more than 26 000 PC patients treated by ADT were enrolled between 2001 and 2003, and have been followed up until now. Several crucial findings have been reported from this study.[17-19] In particular, we provided a new risk-scoring system known as J-CAPRA for patients treated with PADT in combination with the Cancer of the Prostate Strategic Urologic Research Endeavor database in the USA.
In the present study, we validated the PG of the TNM seventh edition using this J-CaP database, to elucidate whether the PG could stratify the prognosis of PC patients. Furthermore, we provided mPG for PADT patients.
- Top of page
- Conflict of interest
In the present study, the use of PG in the TNM seventh edition was validated using the data of PC patients who were treated with PADT from the J-CaP database. Our data showed that PG could stratify the prognosis of patients, including OS, CSS and PFS for the patients who were treated with PADT. However, the OS of PG-I to III were relatively close. In contrast, the OS of PG-IV was far poorer than those of PG-I to III. One of the features of the present study was that the prognosis of subgroups stratified by TNM, PSA and GS could be analyzed, because our database included 15 259 patients treated with PADT, and each subgroup consisted of a median of 206 patients (range 86–1618). Through these analyses, it became clear that the prognosis of the subgroups in each PG varied widely. Interestingly, our data clearly showed that there were two groups with distinct prognosis in PG-IIB and IV, which were subclassified as PG-IIB1/IIB2 and PG-IV1/IV2. The OS of PG-IIB1 was almost equivalent to those of PG-I and IIA, whereas the OS of PG-IIB2 was almost equivalent to that of PG-III. Similarly, the patients of T4N0M0 and TXN1M0 showed almost equivalent survival to those of PG-III. Even if they had metastatis, the patients of GS 7 or less and PSA 100 ng/mL or less showed a relatively better prognosis, which was similar to that of PG-III. Furthermore, we developed a modified PG for the patients treated with PADT, based on our findings. The c-index of mPG was higher than that of SG and PG in the analysis of all kinds of end-points, including OS, CSS and PFS.
The survival of patients with localized PC who were treated by radical prostatectomy or radiation therapy was stratified by the D'Amico classification.[20-24] Generally, the prognosis of patients with high-risk PC was clearly worse than those of patients with low- or intermediate-risk PC, when they were treated by operation or radiation monotherapy. To improve the clinical outcomes of these patients, optimization of treatment protocols depending on the risk group was carried out, such as dose escalation of radiotherapy, introduction of new radiation techniques or combination use of ADT.[25, 26] Long-term adjuvant therapy using ADT was believed to be especially important for high-risk PC, which had possibly concomitant micrometastasis. Our data showed that the 5-year OS and CSS of PG-I, PG-IIA and PG-IIB/III who were treated by PADT were slightly worse, but almost comparable with those of the patients with the same risk of PC treated by radical prostatectomy or radiation therapy.[20-24] Our data also showed that PADT was more effective for patients with low GS PC than for those with high GS. These data provide useful information for predicting the effect of ADT before or after radical therapy.
Generally, PADT is the gold standard of the treatment for PG-IV PC. As shown in the present study, however, the patients of PG-IV have various clinical backgrounds, and the treatment results of PADT also differ greatly. To improve the clinical outcomes, recent new approaches in addition to PADT or other than PADT have been developed for this group. There are some reports of the combination use of ADT with radiotherapy for N1 cases and of prostatectomy after ADT or chemotherapy for T4 or N1 cases to improve clinical outcomes.[27-29] In contrast, PADT is also effective for patients with metastasis at first, but the problem was to become castration-resistant prostate cancer at shorter duration. The PG-IV is a very heterogenous population and probably needs more subdivision than just PG-IV1 and IV2. The present results indicated that among metastatic disease, PC such as PG-IV2 presented a very poor prognosis, and we should improve the clinical outcomes of this subpopulation if possible. For this population, the introduction of new agents, including docetaxel, cabazitaxel and molecular target drugs, in the early phase of treatment has been reported. The patients of PG-IV2 (M1, PSA >100 ng/mL or GS ≥ 8) might be chosen first in the future for the use of those drugs. Thus, it might be important to stratify the risk of PG-IV to develop new treatments or compare the outcomes of these treatments. Although there are few reports about classifying PG-IV so far, the results of the present study are very useful and applicable for choosing the treatment and evaluating the prognosis when considering the medical treatment that ranks second to PADT.
When interpreting the results of this research, there are some points that should be noted. First, in the present study, we validated the use of PG only for the cohort of patients treated by PADT; therefore, further studies are required to validate the study results for the cohort treated by definitive treatments. The second point was that patients were registered to the database by the TNM fifth edition in the present study, and we had to substitute to the seventh edition from the fifth edition for patients in the T2 stage. Although we tried to classify the T2a and T2b of the fifth edition as T2b of the seventh edition, the patients of T1, T2a and T2b had actually almost the same prognosis. The third point was that the present study involved large-scale research with many institutions, and as diagnosis is not based on the work of a central pathologist, pathological assessment might not be unified.
The present study is the first report that verified the validity of the PG of the TNM seventh edition using a large-scale database. Although this classification could appropriately stratify the prognosis, each group still consisted of subgroups with a wide variety in prognosis of patients treated with PADT. The mPG might provide improved accuracy in predicting prognosis in PC treated by PADT and should be validated using a large database of PC patients treated by other modalities.