Adverse Pathological Findings in Needle Biopsy Gleason Score 6 Prostate Cancers with Low and Intermediate Preoperative PSA Levels Following Radical Prostatectomy

Discussion

GS is the most powerful predictor of PCa prognosis and in determining therapy (1, 5). The discrepancy between the GS recorded for the needle biopsy and that for surgical specimen often results in improper assessment of the disease and its treatment, consequently influencing prognosis and outcome of patients (8, 16, 17). Thus correct staging and grading of PCa is of importance in making the correct therapeutic decision.

In the present study we investigated the GS after RP in 788 patients diagnosed with GS 6 after needle biopsy. We found that disease in 36.2% of patients remained GS 6 after RP, however 52.4% were undergraded and 11.4% were overgraded in needle biopsy. Studies investigating GS correlation between biopsy and prostatectomy specimens have shown considerable discrepancy. Oliveira et al., for example, found agreement with biopsy GS 6 in 77.9% of 408 surgical specimens. They found undergrading and overgrading in the biopsy in only 18.5% and 2.6% of cases, respectively (11). Berg et al. also found complete agreement of primary and final GS in 76.9% in a cohort of 350 patients (10).

Another large study including 6,922 patients found coherence between biopsy GS and RP GS in 68.8%. The authors found undergrading in the biopsy to be associated with poorer biochemical recurrence-free survival compared to patients who retained GS after RP (6).

Montironi et al. reviewed several studies and found undergrading and overgrading occurring in 42% and 15% of all cases, respectively (18). Another group reported biopsy undergrading and overgrading in 54% and 15% of patients investigated, respectively (19).

To our knowledge, our present study is the largest, comparing the agreement of GS 6 PCa before and after RP. Thus, one reason explaining our increased number of cases of undergraded GS6 PCa in comparison to other studies, might be the high patient number included in our study. Fine et al. ascribe the discrepancy between needle biopsy and RP specimen in their own study as differences in pathologists (20). In a large study, they investigated the impact of the pathologist itself on GS. They found that at their own Central Department of Pathology at Memorial Sloan-Kettering Cancer Center, the discrepancy between GS before and after RP was significantly lower than at outside institutions (20). Furthermore, borderline cases, as well as intra-observer and inter-observer variability might be responsible for certain variations. Other factors which have been described as having influenced the discrepancy between needle biopsy and RP GS are age at-diagnosis, biopsy Gleason sum, PSA, prostate weight, biopsy positive-to-total core ratio and maximal percent of tumors in cores (21).

The undergrading rate of more than 50% in our patient collective underscores the risk and consequence of incorrect grade evaluation at needle biopsy. Watchful waiting, active surveillance or low-dose brachytherapy, for example, are typically reserved for patients with GS 6 PCa (1). Especially for those patients, a proper recognition of histological grade of PCa is imperative. Thus, all patients undergoing these therapeutic options should be informed about the risk of underdiagnosis.

Additionally, we investigated the risk associated with undergrading GS 6 PCa at biopsy, stratified according to preoperative PSA levels. We found that undergrading of GS 6 in needle biopsy occurs more often in the intermediate-PSA group with an incidence of 57.5% vs. 45.7% in the low-PSA group. In accordance with our results, other groups also found preoperative PSA to be predictive for PCa undergrading (22, 23).

The tumor, lymph node and metastasis (TNM) staging system for PCa defines pT3a as extension of tumor into periprostatic soft tissue (1). In general, extraprostatic extension and positive surgical margins (R1 resection) of PCa are known to correlate with a worse prognosis and thus, with shortened survival time (15).

In the present study, we show that undergrading resulted in a significantly higher incidence of ≥pT3a tumors than of organ-confined PCa. Additionally, we found that 67.7% of all ≥pT3a tumors were associated with intermediate perioperative PSA levels (PSA 4-10 ng/ml).

Investigating the impact of R1 resections our data clearly show that 75.1% of all R1 resections were underdiagnosed in needle biopsy. Stratifying all R1 resections according to PSA levels, 75.9% of patients had intermediate preoperative PSA levels (PSA 4-10 ng/ml).

In accordance with the extra-prostatic extension, R1 resections were observed more often in underdiagnosed PCa (22.1%) than in the group which remained GS6 in the RP specimen (10.6%). Stratifying the underdiagnosed cases according to PSA levels, R1 resections occured significantly more often in the intermediate PSA group. In contrast to our findings, Bulbul et al. compared the incidence of laterality of PCa or surgical margins in patients and found that increased PSA levels were not associated with a higher incidence of positive surgical margins (24).

There are several limitations in our study. The retrospective nature of our study and the lack of a multicenter character are inherent limitations. Randomized controlled multicenter studies need to confirm the present findings, moreover, not all specimens were assessed by the same pathologist, resulting in an inter-observer variability. As previously described by our group, the estimates of underdiagnosis and overdiagnosis are highly dependent on the definition used, and are subject to debate (12). Another possible limitation is that there were two different biopsy techniques, in the periods 1996-2000 and 2000-2012. In a study by our group, investigating on different prostate biopsy techniques, no differences in the biological behavior of the PCa were detected, nor for the pathological stage; there was only a difference in core detection rates (12).

As the GS represents the most important tool for therapeutic decisions in PCa, the goal should be to find a way to eliminate errors in grading needle biopsy specimens. Recently, additional tools for predicting high-grade PCa have been described. Diffusion-weighted magnetic resonance imaging, for example, is known to predict the presence of high-grade tumors in patients with GS 6 PCa (25). Moreover, measurement of pro-PSA isoforms results, were a significant independent predictor of the GS and non-organ-confined PCa in RP specimens [reviewed in (26)].

In summary, our study provides evidence that more than 50% of GS6 PCa diagnosed with needle biopsy exhibit an intermediate tumor grade in the RP specimen, especially in patients with intermediate preoperative PSA levels (4-10 ng/ml). Moreover we show that the incidence of adverse pathological findings, including extraprostatic extension and positive surgical margins is significantly higher in patients with undergraded biopsy GS6. Low preoperative PSA levels improved the correlation between primary and final GS and led to reduction of unfavorable pathological findings.