Original Research

Factors associated with biochemical recurrence of prostate cancer post radical prostatectomy

Ezekiel E. Radebe, Frederik M. Claassen

Received: 07 Nov. 2023; Accepted: 11 Apr. 2024; Published: 31 May 2024

Copyright: © 2024. The Author(s). Licensee: AOSIS.
This is an Open Access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Abstract

Background: Prostate cancer (PCa) is a common solid organ male malignancy with high global mortality. In South Africa, one in 28 men develop PCa. Determining biochemical recurrence (BCR) factors after radical prostatectomy will enable close surveillance and early management.

Aim: The study aims to determine pathological parameters associated with BCR post-radical prostatectomy at a tertiary hospital in the Free State, South Africa.

Setting: All patients (N = 200) who underwent radical prostatectomy in 2013–2018 at the Department of Urology, Universitas Academic Hospital, Bloemfontein were included.

Methods: This was a retrospective descriptive study. Relevant data were extracted from the hospital’s electronic database.

Results: The patients’ median age was 66 years (range: 46–80). A total of 70 patients (35.0%) had BCR, of whom 31 (44.3%) had positive surgical margins (PSM) alone and recurrence occurred at various intervals in their follow-up period, from 9 to 48 months. Patients who had PSMs, extracapsular expansion (ECE), and seminal vesicle invasion (SVI) made up 10.0% (n = 7) of these and had a relapse within 18 months of follow-up. Patients with negative post-operative specimens experienced BCR, ranging from 15 to 48 months, made up 25.7% (n = 18). Extracapsular extension and PSM as combined parameters were 10.0% (n = 7). One patient had SVI as an independent parameter.

Conclusion: These pathological parameters can be linked to BCR after radical prostatectomy. Positive surgical margins prove a strong predictor of BCR after radical prostatectomy.

Contribution: Knowing the most sensitive predictive pathological parameters – either in isolation or combination – is essential to optimise patient management and tighten follow-up schedules.

Keywords: biochemical recurrence; prostate-specific antigen; prostate cancer; radical prostatectomy; surgical margins; extracapsular extension.

Introduction

According to the Global Cancer Observatory (GLOBOCAN) 2020 report, prostate cancer (PCa) was the second most frequent solid organ malignancy, after lung cancer, worldwide in men, with an estimated incidence of 7.3% and mortality rate of 3.8% among its patients globally. The average age of patients diagnosed with PCa is 66 years. There is a notable variation across geographical regions and ethnicity. The incidence of PCa is three times higher in transitioned countries than in transitioning countries, with less variable death counts.¹

In patients treated for PCa with potentially curative therapies, measuring prostate-specific antigen (PSA) levels is the most important part of the follow-up strategy. An increase in serum PSA levels may be indicative of the first sign of relapse, thereby making clinicians aware of the possibility of a diagnosis of biochemical recurrence (BCR).² Biochemical recurrence is defined as a PSA level > 0.2 ng/mL with a second confirmed rise.^3,4 Not all quantifiable PSA levels will result in disease visible in the clinical setting.³ Clinical failure is the pathologic or radiographic evidence of metastases, a climbing PSA level while getting androgen deprivation therapy or resultant death secondary to PCa.^3,4

Radical prostatectomy is the removal of an entire prostate gland, seminal vesicle, and distal part of vas deferens performed for organ-confined, low-risk prostate malignancy with curative intent.² Salvage treatment is an option for patients with local recurrence following radical prostatectomy in radiation or focal therapy.⁵

Model treatment method for localised PCa remains controversial but radical prostatectomy offers a low BCR rate estimated to be from 15% to 35% compared to brachytherapy and external beam radiation therapy (30% to 50%) within 10 years.² In patients with BRC, positive surgical margin (PSM) is the most significant of the parameters for the development of BCR and the subsequent decision regarding adjuvant radiotherapy.²

In the literature, several factors have been explored to predict the risk of BCR with precision, but the conclusions have been cumbersome because of different studies and definitions used by various authors. Cooperberg et al.⁶ developed the Cancer of the Prostate Risk Assessment (CAPRA) score and created a new postsurgical score (CAPRA-S). Looking at the clinical stage as a parameter, they concluded that clinical staging was not a predictive factor for BCR in their study. However, the digital rectal examination (DRE) was still regarded as an integral part of patient assessment.⁶

Numerous nomograms⁴ have been developed for the prediction of BCR following radical prostatectomy, and prognostic variables have been specified in these nomograms. Kattan’s nomogram predicted BCR based on PSA levels, Gleason score, and pathologic stage. The Standard Cox model is a proportional hazards model based on clinical failure as the resultant event. The covariates selected for the Standard Cox model are pre-specified and chosen based on their reliability in predicting cancer-related consequences after radical prostatectomy. Predictors included PSA, Gleason score, and stage, where the stage is presented as a categoric variable: 1 (patient with any of the following: PSMs, seminal vesicle invasion (SCI), extracapsular extension (ECE), or lymph node invasion (LNI)) and 0 (if otherwise).⁴ The nomograms’ three most often used variables are pathological T stage, pathological Gleason score, and preoperative PSA level.^7,8,9

Different studies reported on preoperative PSA and PSA density (PSAD) levels. Radwan et al.¹⁰ shared that PSAD (prostate volume measured with either abdominal or transrectal ultrasound/PSA value) is more predictive for BCR than PSA. Freedland et al.¹¹ recently found that preoperative PSAD, compared to PSA, provided slight improvement for BCR prediction following radical prostatectomy. Brasell et al.¹² compared PSAD and PSA results using u/s volume and the actual prostate volume. As a predictor of BCR, PSA was reported to be better than PSAD. Thus, it was extrapolated that PSAD added no value over PSA alone as a predictive factor.²

The primary aim of this research was to determine the pathological parameters associated with BCR in patients with low- and intermediate risk PCa post-radical prostatectomy.

Methods

Study design and setting

This was a retrospective, descriptive study. This single-centre study was performed at the Department of Urology at the Universitas Academic Hospital (UAH), Bloemfontein, in the Free State province of South Africa. Universitas Academic Hospital is a referral and tertiary academic hospital that caters to patients in the Free State, Northern Cape and Eastern Cape provinces and selected patients referred from Lesotho, a small neighbouring country. It is estimated that UAH serves approximately seven million of the South African population per year. The hospital has 636 beds, of which 15 are allocated to the Department of Urology.

Study population and sampling strategy

All patients who underwent a radical prostatectomy at UAH between 2013 and 2018 were included in the study. The researcher identified patients with BCR as those with a rise in PSA levels ≥ 0.2 ng/mL, 6–8 weeks post-operative with a second confirmatory rise at 12 weeks after the initial reading.³

The inclusion criteria were:

• low-risk PCa (Gleason score ≥ 6; PSA ≥ 10 ng/mL; clinically T1C–T2A)^13,14
• intermediate-risk PCa (Gleason score 7; PSA 10–20 ng/mL; clinically T2B).^13,14

The exclusion criteria were:

• high-risk PCa (Gleason score ≥ 8; PSA ≥ 20 ng/mL; clinically T2C and above)^13,14
• patients who received adjuvant treatment
• patients who were lost to follow-up after surgery in less than two follow-up visits
• patients with PSA persistence.

Data collection

A list kept by the Department of Urology was used to identify all patients treated with radical prostatectomy between 2013 and 2018. The patient file numbers were used to access the patients’ files on the hospital’s electronic database (MediTech®).

Relevant information was entered into an Excel spreadsheet and included age, dates of initial and repeat PSA tests, clinical stage, Gleason score, number and/or frequency of positive cores, and date of radical prostatectomy. The post-radical prostatectomy specimen pathological parameters determined in this study included: ECE, LNI, PSM, and SVI. The PSA results for follow-up visits were also captured.

Pilot study

The first patient of each year of the study period was included in the pilot study. Data were entered into the pre-designed Excel spreadsheet. No changes were made to the spreadsheet and the five patients were included in the data analysis.

Data analysis

Data were analysed by the Department of Biostatistics, Faculty of Health Sciences at the University of the Free State, South Africa, using SAS^® 9.4. It was decided to offer descriptive statistics to provide a summary of the frequency, percentage, cumulative frequency, and cumulative percentage. In connection with the aim of the study, the frequencies were compared, and conclusions were drawn from those comparisons.

All the required data of the patients who met the inclusion criteria were entered into the Excel spreadsheet. The researcher reviewed the raw data from each patient. Based on the PSA value after post-radical prostatectomy, patients were identified with BCR.

Ethical considerations

An application for full ethical approval was made to the Health Sciences Research Ethics Committee of the University of the Free State and ethics consent was received on 6 January 2022. The ethics approval number is HSD2021/1463/2501. Pseudo-anonymisation was performed to ensure the protection of personal information. The patient’s personal information was protected from unauthorised access as the data files were kept in the Department of Urology, where only the researcher had access to the files. This is in accordance with the Protection of Personal Information Act (POPI Act 4 of 2013). Because of the retrospective nature of this study, individual patient consent was not required by the Health Sciences Research Ethics Committee.

Results

Patients

In total, 201 patient files were considered eligible for inclusion in the study. During the data review, one patient was excluded as it was found that the patient met one of the exclusion criteria (Gleason score ≥ 8 and received adjuvant treatment). Therefore, the information of 200 patients was analysed. The median age of the patient population was 66 years (range: 46–80 years; interquartile range [IQR]: 62–70 years).

Factors associated with biochemical recurrence prostate cancer post radical prostatectomy

Of the 200 patients, 70 (35.0%) had BCR post-radical prostatectomy. Table 1 summarises the occurrence of the histological parameters in BCR patients. As shown in Figure 1, 33 patients (47.1%) had only one of the histological parameters. The remaining 37 patients (52.9%) had either no histological parameter or a combination of parameters.

FIGURE 1: Flow diagram of biochemical recurrence and relationship to histological parameters (N = 70).

Of the patients with BCR, 18 (25.7%) had no histological features. For these patients, BCR occurred much later in their follow-up (range: 15–48 months).

The highest percentage of patients (n = 31; 44.3%) with BCR presented with only PSM. Biochemical recurrence occurred at varying times in the follow-up term ranging from 9 to 48 months. Five patients (7.1%) with SVI and PSM had BCR between 9 and 42 months post-radical prostatectomy. Seven (10.0%) patients presented with a combination of ECE, SVI, and PSM. These patients had BCR within 18 months of follow-up.

Patients with the following features all had BRC between 15 and 48 months post-radical prostatectomy: only SVI (n = 1; 1.4%), only ECE (n = 1; 1.4%), or ECE with PSM (n = 7; 10.0%). None of the patients had LNI.

Histological features of total study population

Table 2 describes the histological parameters found in the post-radical prostatectomy specimens of all 200 patients, regardless of their BCR status.

Of the total study population, 105 (52.5%) had PSM, 35 (17.5%) ECE, and 29 (10.5%) SVI. None of the patients had LNI. The PSM, ECE, and SVI combinations were not investigated for the study population.

Discussion

The study’s findings presented results from patients’ data who underwent open radical prostatectomy between the years 2013 and 2018 at UAH, Bloemfontein. The post-surgical radical prostatectomy specimen was assessed for the following parameters: ECE, SVI, LNI, and PSM. In our setting, limited pelvic lymphadenectomy (bilateral obturator lymph nodes) is performed with every radical prostatectomy as a staging and not as a curative measure; thus, the lymph nodes are also being assessed for cancer invasion. Results showed a strong relationship between PSM and BCR after radical prostatectomy. In addition, PSM was linked with unfavourable pathological outcomes (quicker disease progression) post radical prostatectomy.

Liu et al.¹⁵ identified six predictive parameters for BCR following their report. For recurrence-free survival following radical prostatectomy, they asserted that SVI, PSM, ECE, LNI, and peri-neural invasion (PNI) were all statistically significant predictors (p = 0.001).

Extracapsular extension (pT3a) affects the BCR and is shown to be an independent predictive factor in PCa-specific mortality.² Also, ECE has been shown to increase the risk of BCR up to 1.5 times over organ-confined PCa and this is in accordance with literature.²

In the BCR group, one patient (1.4%) had only ECE while 14 patients (20.0%) had either ECE with PSM or with PSM and SVI. In the total study population, 17.5% patients had ECE. The effect of the ECE degree was studied following different methods, such as the Wheeler method and Epstein’s criteria.³ The ECE was shown as an independent BCR predictor in PCa even without PSM; however, with the combination of both parameters, the patients achieved BCR earlier in their follow-up. The results and literature show substantial evidence that ECE is a predictor of BCR.³

Similar to the ECE results, one BCR patient (1.4%) had only SVI while 12 BCR patients (17.1%) had either SVI with PSM or with ECE and PSM. Only 10.5% of the total study population had SVI. Thus, SVI is a predictor of BCR. The invasion of seminal vesicles on the radical prostatectomy specimen is linked with poor prognosis.^2,3 Seminal vesicle invasion increases the risk of BCR 2.3 times and the mortality rate specific to PCa by 22%. It is suggested that there is a poorer prognosis when comparing patients with pT3b and those with pT3a on the radical prostatectomy specimen. The prognosis is even poorer when associated with other parameters.³

Even though when men have high-grade PCa at diagnosis and poorer outcomes, researchers have found a subset of patients with SVI who have favourable prognoses. Isolated SVI with lower PSA levels and lower Gleason Score/ISUP (International Society for Urologic Pathology) grading, negative lymph nodes, and negative surgical margins represent a small percentage of men with pT3 PCa, carrying a better prognosis.⁵ The literature supports our finding that SVI is a significant predictor of BCR, with the risk increasing when there is an association with other pathological parameters.

None of the patients who underwent the radical prostatectomy with limited pelvic lymph-node dissection and were found to have positive lymph nodes on their post-surgical specimen fulfilled the inclusion criteria. The only patient with nodal disease never reached PSA levels > 0.2 ng/mL and thus did not fit the selected definition of BCR. In literature, people diagnosed with positive nodal disease have 1.7-fold higher recurrence rates than those with negative nodal disease. Metastasis lymph node disease further increases cancer-specific mortality, with more than 40% in patients followed-up for longer durations.²

The progression time can also be correlated significantly to the total number of diseased nodes. The clinical and biochemical time of relapse can range from 24 to 46 months and 7 to 28 months if considering patients with one positive node or more than two positive nodes, respectively. Patients with minimal positive lymph nodes undergoing radical prostatectomy with meticulous dissection of pelvic lymph nodes might be free of BCR for more than 10 years.² In the literature, however, BCR occurred whereby 65% of participants had one positive node, and 13.57% had five positive nodes, which shows strong evidence of BCR.⁵ However, this parameter tested poorly in our study, as there was no reproducible evidence.

Surgical margin positivity is also associated with unfavourable pathological features, such as EPE, SVI, high pathological T stage, and elevated postoperative detectable PSA level. In addition, a high Gleason score on a radical prostatectomy specimen is also an important predictive factor in patients with PSM.⁶ Positive surgical margins is a predictive pathological measure that can affect BCR, recurrence-free survival, and related adjuvant therapy following radical prostatectomy, according to several studies.^{13,14,16,17,18}

More than half (52.5%) of the total study population had PSM. In line with this, 71.4% of the BCR patients had PSM; 44.3% as the only factor identified; and 27.1% in combination with ECE, with SVI, or with ECE and SVI. Positive surgical margin as an independent factor does not carry an increased risk of PCa-related death but does increase the risk of BCR up to 2–4 times and, thus, the need for secondary therapy.² Numerous studies tried to stratify the extent of PSM with no conclusive end-point, but the higher the number of positive margins, the higher the risk of BCR.²

The higher the burden of PSM, the higher the increased risk of BCR compared to the focal and solitary PSM; the clinical usefulness of this parameter as a predictive factor is very important. The findings from our study suggest that PSM is a strong predictor of BCR as it had the highest percentage as the solitary predictor and appeared in combination with all other predictors in most patients who developed BCR with an even poorer prognosis.

Positive surgical margins as predictive factors have been studied and conclusively linked to poorer recurrence-free survival. The position and burden of the surgical margin were assessed, and it was found that the apical position of the tumour and a PSM length of >3 mm, as well as PSM in more than three locations are associated with a higher risk of BCR.¹⁹

Patients with negative post-operative specimens who experienced BCR, ranging from 15 to 48 months post radical prostatectomy, made up 25.7% of the BCR patients. The other factor to consider in this group of patients is that their pre-operative ISUP score upgraded post-, and the median pre-operative PSA level was 13 ng/mL. In these cases, we accept that the pre-operative biopsy undermined or was not closely representative of the cancer burden.²⁰

Study’s limitations

This study has limitations, including the retrospective design, incomplete records, and small sample size. The study was limited to one centre in the Free State, and, as a result, the private sector is not represented, and findings cannot be generalised to the country as a whole. Although specific follow-up protocols exist in our department, adherence to this protocol is difficult to enforce because of the patients’ different circumstances and transport and referral systems being the main challenges.

Conclusion

In our study, BCR was diagnosed from 3 months after achieving nadir post radical prostatectomy in patients with PSM. Pathological parameters associated with BCR PCa post radical prostatectomy were determined. No positive LNI was found in our study although in literature, lymph nodes were found to be a predictor.^1,2 Both ECE and SVI appeared as significant predictors of BCR, which suggest that close PSA monitoring of patients with ECE is required. Positive surgical margin was the strongest and most common predictor of BCR; thus, patients with PSM should receive strict follow-up schedules for PSA monitoring for early BCR detection.

Early detection of BCR in PCa patients is crucial for treatment success and patient survival after radical prostatectomy. Knowing the most sensitive predictive pathological parameters – either in isolation or combination – is essential to optimise patient management and tighten follow-up schedules.

Acknowledgements

The authors would like to thank Dr W. Dahms, consultant at Pelonomi Provincial Hospital and leader of the Urology-Oncology firm at the Universitas Academic Hospital for his insights into structuring and planning the overall presentation of this research article. Special thanks to Mr C. van Rooyen, Department of Biostatistics, Faculty of Health Sciences, University of the Free State, for timeous assistance in the development statistical analysis and data report in this study, and Ms T. Mulder, medical editor, Faculty of Health Sciences, University of the Free State, for technical and editorial preparation of the manuscript.

The authors declare that they have no financial or personal relationship(s) that may have inappropriately influenced them in writing this article.

E.E.R. developed the research protocol, conducted the research, and wrote the report and article. F.M.C. acted as a study leader, was involved with concept development and proposal preparation, and assisted with the report and article. E.E.R. and F.M.C. read and approved the final manuscript.

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

The data supporting this study’s findings are available from the corresponding author, E.E.R., upon reasonable request.

The views and opinions expressed in this article are those of the authors and are the product of professional research. It does not necessarily reflect the official policy or position of any affiliated institution, funder, agency, or that of the publisher. The authors are responsible for this article’s results, findings, and content.

References