Prostate cancer is the most common cancer in men in the Netherlands. The incidence is increasing and more and more men are diagnosed in an early stage. Consequently, the number of treatments with curative intent, like radical prostatectomy, is also rising(19).
Figure 1: Kaplan-Meier curves of the PSA progression free interval of patients with a tertiary pattern (gl3+; ȱȱǼȱȱȱ ȱȱ¢ȱȱǻřƺDzȱ ȱȱǼǯ
Although radical prostatectomy is a routine operation with good functional results, there is still a substantial percentage of patients that experience a PSA relapse within 5 years. These men have not been cured in spite of the previous operation20. Biochemical recurrence is observed in approximately 40% of patients who undergo radical prostatectomy, with most of the relapses (95%) in the first 5 years5.Prediction of failure is extremely important, but remains difficult. The best predictors of PSA failure after radical prostatectomy are Gleason score, pathological staging, status of surgical margins and tumor volume21. The Gleason score is the most widely used grading system for prostate cancer. The reproducibility of this scoring system in several studies is in an acceptable range of about 70%(22;23) with a better intraobserver and interobserver reproducibility than the WHO system(24). A potential new predictive factor is the presence or absence of a tertiary Gleason pattern. The Gleason
concerning the prevalence of a tertiary Gleason pattern in prostate cancer. Aihara et al. found an average of 2.7 Gleason patterns per case, and more than 50% of the cases containing at least three different grades (14). Gleason reported in 1992 that more than 50% of tumors contained two or more patterns(13). In an earlier study we found that 16% of all the cancer lesions consisted of 3 or more different histological grades(11). In this study 48% of our patients had a tertiary grade at histopathological investigation.
So far only two studies were published dealing with the prognostic significance of the tertiary Gleason pattern. In the first study Pan et al. analyzed 114 radical prostatectomies with small tertiary patterns. The volumes of the tumor with tertiary components ranged from 1% to 15% of the total tumor volume (mean 2%)(15). These were compared to a prostatectomy database of 2,276 cases without a tertiary pattern. They defined typical Gleason score 5 or 6 tumors as tumors with Gleason scores 2 + 3 = 5, 3 + 2 = 5, 3 + 3 = 6. Both typical Gleason score 5 to 6 and Gleason score 7 tumors with worse tertiary patterns revealed significantly higher progression rates than typical Gleason score 5 to 6 tumors and Gleason score 7 tumors without tertiary patterns. In case of a higher grade tertiary pattern behavior was more like typical Gleason score 7 and 8 tumors respectively. Moreover, high-grade tertiary patterns were associated with a more advanced pathological stage. They also found that tertiary patterns with volumes over 1% of the total tumor volume had a shorter disease free survival as compared to cases where the tertiary tumor was less than 1% of the volume. In all, Pan et al. concluded that tertiary high-grade patterns adversely influence the prognosis, and proposed to modify the Gleason grading system by taking these small volumes of patterns 4 and 5 into account. Modifications of the Gleason scoring system based on a tertiary pattern were also discussed by Deshmukh and Foster in 1998(25). They described how to approach a case with a tertiary pattern. However, there is still no consensus in this respect. The second study from Mosse et al. compared the pathological stage of 277 radical prostatectomies with tertiary pattern Gleason grade 5 to the pathological stage of 604 radical prostatectomies lacking a tertiary pattern(16). The relative effects of a tertiary pattern of 5 were largest when the primary and the secondary Gleason grades were low. Even the score of the primary pattern was of importance. For example, the impact of a Gleason 5 pattern was
much more outspoken in Gleason score 3 + 4 tumors as compared to Gleason score 4 + 3 tumors. The significance of a tertiary Gleason 5 pattern is obscured in case of higher primary and secondary Gleason grades by the already aggressive nature of these advanced primary and secondary patterns. They concluded that, except for very high-grade tumors, the tertiary pattern at radical prostatectomy should be reported as it has prognostic significance.
In this study we enrolled only patients with an organ-confined tumor who underwent RRP, since that is the group of patients in which the application of an additional prognostic marker would be of use, in other words, in whom surgery could have been prevented. We found a tertiary Gleason grade in, on average, 7% of the total volume. Our study also indicates the importance of a tertiary Gleason pattern. Patients with a tertiary pattern had a 5-year risk of PSA progression of 37% versus 13% among patients without a tertiary pattern (p = 0.0002). There was no difference between patients with a higher tertiary pattern as compared to those with a lower tertiary pattern (p = 0.12). So, as we expected, we found a worse outcome in the high-grade tertiary pattern cohort, but we also found a worse outcome in case of a better tertiary pattern. In comparison to Pan et al. and Mosse et al., we found that merely the existence of a tertiary grade, whether low or high grade influences the prognosis. The reason for this difference could be that in the studies of Pan et al. and Mosse et al. only the largest (index) tumor in the prostate was recorded and small multifocal lower grade tumors were not recorded, not taking tumor multifocality and heterogeneity into account. Ruijter et al. found in their study that one third of the penetrating tumors in multifocal disease were not index-tumors. Furthermore in 9% of the prostate specimens the satellite tumors were morphologically more malignant, according to Gleason score, than the index-tumors in the same specimens. This indicates that the largest (index) tumor lesion may not be indicative for the prognosis(11). In all, we suggest to take into account all tertiary patterns for which submission of the entire prostatectomy specimen is needed.