Prostate cancer is the most common malignancy in men, and various treatment methods have been developed in recent decades. Surgery remains one of the most popular treatment options with robotic-assisted radical prostatectomy (RARP) becoming increasingly popular in many centers (1-2). Prostate-specific antigen (PSA) is a sensitive marker for prostate cancer, typically decreased to undetectable levels after radical prostatectomy. However, in some cases, PSA recurs after initially becoming undetectable. A Japanese study that analyzed the outcomes of cT1-2N0M0 prostate cancer patients who underwent radical prostatectomy without concomitant hormone therapy between 1993 and 2000 reported that PSA recurrence was observed in 25.3% of patients during a 45.6-month follow-up period (3).

Salvage radiotherapy is often performed in cases of PSA recurrence in patients following total prostatectomy (4). However, in a study on a Japanese population, the PSA recurrence rate for postoperative salvage radiotherapy was 41.0%, suggesting that further studies are needed to improve the outcomes (5). Therefore, identifying factors that predict PSA recurrence after irradiation is crucial.

Predictors of prostate cancer recurrence include the rate of positive scores on prostate biopsy, PSA levels before prostatectomy, Gleason score, and T stage after surgery, all of which are associated with PSA recurrence. These postoperative risk factors contribute to predicting PSA recurrence (6). However, predictive factors specific to prostate cancer patients treated with salvage radiotherapy have not been fully evaluated. This study focused on histopathological factors as potential predictors of PSA recurrence in patients who underwent salvage radiotherapy. Specifically, we evaluated the cribriform pattern, Gleason score pattern 5 ratio, and perineural invasion, which are histopathological factors in prostate cancer specimens detected by our pathology department to investigate their association with postoperative PSA recurrence after irradiation.

By performing a pathological analysis using surgical specimens from patients who underwent total prostatectomy at our Institution, we aimed to determine whether there is a correlation between these pathological findings and PSA recurrence after postoperative salvage radiotherapy. This study aimed to clarify the relationship between pathological findings and PSA recurrence, potentially improving patient outcomes by identifying significant prognostic factors.

Patients. All 50 patients who underwent radiotherapy after total prostatectomy at our Center between 2010 and 2020 after total prostatectomy and pathological analysis were retrospectively evaluated. The prostate floor was delineated for the clinical target volume. The planning target volume included the clinical target volume with a 5-10 mm margin for set-up error. All patients provided written informed consent prior to radiotherapy. Gleason score pattern 5 ratio, cribriform pattern, and perineural invasion were evaluated as none, mild, moderate, or severe using surgical pathology specimens at our Hospital’s pathology Department. Immunohistochemical analysis was performed to evaluate ETV1 and ERG protein expressions, together with PTEN loss, which were known as major chromosomal abnormalities in prostate cancer. Statistical analysis was performed using t-test and log-rank test. After treatment, PSA was measured every 3 months, and PSA >0.2 was considered PSA recurrence. p<0.05 was considered statistically significant. Analyses were performed using the SPSS software. This study was reviewed and approved by our Institutional Review Board (S20-055) and was performed in accordance with the Declaration of Helsinki.

Ethics approval and consent to participate. This study was reviewed and approved by our Institutional Review Board (S20-055) and was performed in accordance with the Declaration of Helsinki.

Patient characteristics and treatment details. The median age was 67 (54-81) years. Among patients who underwent radiotherapy after total prostatectomy, the median time from prostatectomy to PSA recurrence was 12.0 months (range=1.94-60.35 months). The median time from prostatectomy to the start of irradiation was 13.6 months (2.6-61.8 months). Open radical prostatectomy was performed on 31 patients and RARP was performed on 19 patients. Postoperative radiotherapy was administered to 2 patients and salvage radiotherapy was administered to 48 patients. Three-dimensional conformal radiotherapy based on Computed Tomography (CT) simulation was performed on 38 patients, and the radiation dose was 64 Gy in 32 fractions. Intensity modulated radiation therapy based on CT simulation was performed on 12 patients, and the radiation dose was 66 Gy in 33 fractions. Hormonal therapy was not administered in all cases prior to undergoing total prostatectomy. Overall, 21 (42.0%) patients received neo-adjuvant or concurrent hormonal therapy (Androgen deprivation therapy: 14 patients, antiandrogen therapy: 7 patients).

Outcomes. The tumor stage was distributed as follows: pT2 (n=15; 30.0%), pT3a (n=18; 36.0%), and pT3b (n=17; 34.0%). A positive surgical margin was present in 27 cases (54.0%). The PSA levels before radiotherapy was distributed as follows: <0.2 (n=12; 24.0%), 0.2-0.5 (n=18; 36.0%), 0.5-1.0 (n=10; 20.0%), and >1.0 (n=9; 18.0%). The cribriform pattern was distributed as follows: none, <1% (n=1; 2.0%); small, 1<20% (n=16; 32.0%); moderate, 20<50% (n=17; 34.0%); and severe, >50% (n=16; 32.0%). Perineural infiltration was distributed as follows: none (n=6; 12.0%), small (n=22; 44.0%), moderate (n=15; 30.0%), and severe (n=7; 14.0%). Gleason pattern 5 ratio was distributed as follows: 0% (n=15; 30.0%), 0-10% (n=15; 30.0%), 10-30% (n=9; 18.0%), and >30% (n=11, 22.0%). No deaths related to prostate cancer were observed during follow-up; however, one patient died 48 months after radiotherapy owing to atrial fibrillation. Eighteen patients (36.0%) had PSA recurrence.

Statistical analysis. The correlation between clinical factors and PSA recurrence rate is shown in Table I, and the correlation between pathological factors and PSA recurrence rate is shown in Table II. The PSA recurrence rates of patients treated with and without hormonal therapy were 19.0% and 48.3%, respectively. The PSA recurrence rate based on the cribriform pattern was 17.6% in the none to moderate group (34 patients) and 75.0% in the severe group (16 patients). The PSA recurrence rate was significantly higher in cases with a severe cribriform pattern (p=0.001; Figure 1). The PSA recurrence rate for perineural invasion within the capsule was 34.9% for the none to moderate group (43 cases) and 42.9% for the severe group (7 cases), with no significant difference (p=0.679). The PSA recurrence rate for the group with a Gleason pattern 5 ratio of >30% was 27.3% in 11 cases, which was not significantly different (p=0.333).

ERG, and ETV1 chromosomal rearrangement were evaluated by immunostainings. ERG positive case was observed only 1, and ETV1 was positive in 19 cases, respectively. In addition, substantial areas with loss of PTEN expression was observed in 17 patients. However, these findings were not associated with cribriform pattern (Table II).

This study investigated whether cribriform pattern was associated with biochemical recurrence-free survival in the 50 patients who received radiotherapy after radical prostatectomy. The PSA recurrence rates when the two groups were divided by the cribriform pattern of the prostate of surgical pathology specimens were 17.6% for the none to moderate (34 cases) group and 75.0% for the severe (16 cases) group; there was a significant difference (p=0.001). Other histopathological factors, such as the perineural invasion inside and outside the capsule and Gleeson pattern 5 ratio, did not show significance for PSA recurrence.

Gleason grade 4 prostate cancer comprises four main architectural growth patterns, specifically poorly formed, fused, and cribriform glands (Figure 2) (6-7). It is now well-established that, among these growth patterns, cribriform architecture in biopsies or radical prostatectomy specimens is strongly associated with a less favorable outcome (8-9).

Several reports have shown postoperative and radical radiotherapy outcomes for cribriform patterns using biopsy and resected pathology specimens. Ramotar et al. studied the improvement of treatment methods based on the results of radiation therapy and treatment after prostatectomy by observing the intraductal carcinoma and cribriform architecture and by using a 22-gene decipher genomic classifier. The study revealed that intraductal carcinoma, cribriform architecture, and the 22-gene decipher genomic classifier predict biochemical relapse and metastasis beyond conventional clinicopathologic indexes in a post-prostatectomy radiotherapy setting (10).

Hollemans et al. found that both invasive cribriform and intraductal carcinoma of the prostate have an impact on prognosis in radical total prostatectomy (11). The study reviewed 420 prostatectomy specimens from patients with The International Society of Urological Pathology (ISUP) grade 2 prostate cancer, assessed the percentages of Gleason grade 4 and tertiary 5, and performed immunohistochemistry for basal cells to differentiate intraductal carcinoma from invasive cribriform growth. Severe cribriform architecture was associated with older age, higher percentage of Gleason grade 4, extraprostatic expansion, and more frequent lymph node metastases. In multivariable Cox regression analysis, pT-stage, positive surgical margins, and severe cribriform growth were independent predictors for biochemical recurrence-free survival, whereas intraductal carcinoma, small cribriform growth, and percentage of Gleason grade 4 were not.

Tom et al. evaluated the effect of cribriform pattern and/or intraductal carcinoma on radiotherapy for prostate cancer patients with a Gleeson score of 7 (12). The study revealed that cribriform pattern with intraductal carcinoma was associated with adverse outcomes in men with Gleason 7 prostate cancer treated with radiotherapy, whereas cribriform pattern without intraductal carcinoma was not associated with adverse outcomes.

Radiation therapy for prostate cancer is effective in combination with radical external radiation and hormone therapy, and as a control for recurrent cases after total prostatectomy, external irradiation to the prostate bed is effective as salvage therapy. However, the control rate of salvage radiotherapy was lower than that of radical radiotherapy. Therefore, we believe that some countermeasures are necessary in cases with severe cribriform disease. For this purpose, it is necessary to increase the prescribed irradiation dose. In our study, salvage radiotherapy was performed at 64-66 Gy; however, it is necessary to use a dose of 66 Gy or higher to avoid adverse events.

Pisansky et al. determined whether a dose-response relationship exists for salvage radiotherapy of biochemical failure after prostatectomy for prostate cancer. The study revealed that dose escalation to levels >66 Gy are associated with a reduced likelihood of biochemical failure after salvage radiotherapy in patients with post-radical prostatectomy (RP) detectable PSA levels (4).

King et al. analyzed through a systematic review and a meta-analysis, from 1996 to 2015, identified the pathologic, clinical, and treatment factors associated with relapse-free survival after salvage radiotherapy. The study reviewed level 2a evidence for dose-escalated salvage radiotherapy dose of 70 Gy or higher. Salvage radiotherapy dose–response for microscopic disease supports the hypothesis that prostate cancer is inherently radioresistant (13).

We also considered the significance of hormonal therapy in salvage treatment for the control of recurrent cases after total prostatectomy. In this study, we found that PSA recurrence after irradiation was suppressed in patients who received concomitant hormone therapy. Hormonal therapy in prostate cancer increases the control rate when combined with radiation therapy. The principle of hormone therapy is to reduce or block the male hormones (androgens) that contribute to prostate cancer development, thereby reducing the effect of testosterone on prostate cancer and temporarily suppressing its growth (14-16). Radiation therapy for localized prostate cancer is generally categorized according to risk, presence or absence of adjuvant hormone therapy, and duration of adjuvant hormone therapy (12-14). In the current study, there was a significant difference in PSA recurrence rates between patients with and without hormonal therapy for the first PSA after prostate cancer surgery. This suggests that hormonal therapy may be beneficial in patients with a large cribriform pattern (17,18). Recently, a brand-new drug, oral recombinant methioninase, has been reported to suppress PSA level in advanced prostate cancer patients (19). Further studies are warranted to establish the strategy for a combination of drugs and radiotherapy.

Clinical and pathological factors that predict cribriform pattern are important. Mikoshi et al. reported that MRI-detectability of prostate cancer was associated with cribriform pattern (20). ERG, PTEN, and ETV1 rearrangements were reported as potential molecular markers for cribriform lesion (21,22); However, this study did not show the associations. Future studies are needed to identify radiological or molecular markers that can assist in diagnosing the cribriform pattern.

Recently, the efficacy of prostate-specific membrane antigen positron emission tomography/computed tomography (PSMA PET/CT)-guided salvage radiotherapy has been reported (23). A complete clinical response of approximately 70% at one year after irradiation especially in patients with low PSA levels, indicating that early detection of PSA recurrence and appropriate salvage radiotherapy contribute to improved prognosis. Juracek et al. reported that a logistic regression model combining the miR-335:miR-501 ratio, PSA, and prostate volume data was significantly more discriminating in predicting prostate biopsy results than standalone parameters (24). Considering the results of our study on Cribriform pattern and prognosis prediction, it is expected that the relationship between Cribriform pattern and PSMA PET/CT or miR-335:miR-501 can lead to the development of more sophisticated prognostic predictions (23,24).

The present study had several limitations. It was a retrospective single-center study with a small sample size. Further prospective studies are warranted to evaluate cribriform pattern as a predictive factor in prostate cancer.

Cribriform pattern in surgical pathology specimens was found to be a useful predictor of PSA recurrence after postoperative radiotherapy.

The Authors declare that there are no conflicts of interest.

Kawahara M, Tanaka A and Shirai K established the study design and directed the analysis. Tanaka A, Hiruta M and Oshiro H contributed to validation of pathological data. Kawahara M and Shirai K analyzed the data and contributed to the analysis of the results and performed the statistical analysis. Tanaka A, Miyagawa T and Shirai K supervised the project. All Authors reviewed the manuscript.

We would like to thank Editage (www.editage.com) for English language editing.