Prostate Cancer in a Sub-Saharan Tertiary Level Setting: Profile of Contemporary Gleason Grading Over 13 years

Introduction

Prostate cancer (PCa) is the second most frequently diagnosed cancer among men (after lung cancer) and the fifth leading cause of death worldwide.¹ Despite the lack of large population-based studies, available evidence suggests that low-income countries have the highest prevalence, incidence, and mortality rates of PCa.² This imbalance is mainly due to a generalized lack of access to health care, limited prostate-specific antigen (PSA) screening services, and poor disease awareness. Furthermore, the scarcity of published data on PCa, underdeveloped cancer registries, and inconsistent data collection lead to inaccurate estimates of PCa-related morbidity and mortality rates.¹ While multiple studies have reported a high prevalence of PCa among Black American men compared to White men,³ little is known about the prevalence of PCa among Black men living in Africa, especially SSA. The Gleason score—developed between 1966 and 1974 by Donald Gleason and the Veterans Administration Cooperative Urologic Research Group⁴—remains the most accurate tool for predicting the aggressiveness and prognosis of PCa. The International Society of Urological Pathology (ISUP) revised the Gleason scoring system in 2005⁵ and again in 2014,⁶ resulting in a standardized five-tiered grade group (GG) system.⁷

To contribute to the growing data on PCa in SSA, we conducted a retrospective review of PCa cases at King Faisal Hospital (KFH), a tertiary referral hospital in Rwanda, from 2010 to 2022. These cases were classified according to the 2014 ISUP grading system, and we aimed to characterize the pathological profile of each PCa patient.

Methods

Study design and setting

This retrospective, cross-sectional, descriptive study included patients diagnosed with PCa from 2010 to 2022 at KFH, a top national referral and teaching hospital in Rwanda. KFH is a 160-bed facility currently undergoing major infrastructure expansion and subspecialty training.

Data collection

The variables of interest included nationality, age at diagnosis, the origin and type of prostatic specimen, presenting symptoms, PSA levels, pathological GG, and treatment modalities. All PCa cases were identified based on pathology logs, operative logs, as well as paper and electronic medical records. Pathology logs were used as a reliable primary source of information for case identification, Gleason scores, and related GGs. Operative theatre logs provided clinical information related to surgical treatment modalities, especially for patients who underwent radical prostatectomy or tunnel transurethral resection of the prostate (TURP) with or without orchiectomy for advanced disease. The remaining clinical information was retrieved from both paper and electronic-based medical records. To ensure that all cases were pertinent to the contemporary ISUP grading system, all slides that corresponded to pathological PCa and that were analyzed before 2015 (ie, the year the updated Gleason grading system was adopted at KFH) were re-reviewed by 2 independent pathologists using hematoxylin and eosin staining.

Data analysis

Data entry and analysis were conducted using SPSS version 21. Descriptive data are reported as frequencies and percentages.

Ethical considerations

This study was reviewed and approved by the Ethics and Research Committee at KFH.

Results

Over the study period, a total of 245 cases of PCa were reported, with 190 cases occurring between 2015 and 2022. Seventy-nine cases were diagnosed between 2010 and 2014, and histopathological re-evaluation revealed that 24 (30.7%) of these cases were benign prostatic hyperplasia, as indicated by a Gleason score <6. The remaining 55 PCa cases were combined with the other 190 cases for analysis.

As shown in Table 1, 197 (80.4%) PCa cases were diagnosed and managed at KFH, while 48 (19.6%) cases were referred from other facilities. Thirty-six patients (14.7%) were non-Rwandan nationals, primarily from neighboring countries. Over half of the patients (132, 53.9%) were aged 70 years or younger at diagnosis, with a mean age at diagnosis of 68 years. The most common primary complaint at presentation was lower urinary tract symptoms (LUTS), reported by 93 (38%) patients. Only 48 (19.6%) patients had PSA levels ≤20 ng/mL at diagnosis.

Most cancer specimens (147, 60.0%) were prostate core biopsies, followed by prostatic chips obtained from tunnel TURP (43, 17.6%). Histological confirmation accounted for 82.1% of diagnoses, including all types of prostate specimens. In comparison, 1.2% of cases were diagnosed based solely on clinical findings such as a hard prostate on digital rectal examination, markedly elevated PSA levels, and/or radiologic evidence of metastases. The remaining 16.7% of cases lacked information on the diagnostic modality. The most frequent GGs were GG 4 (77, 31.4%) and GG 5 (59, 24.1%), together accounting for over half of all cases. Despite some missing data on management options (99, 40.4%), ADT was the most prevalent initial treatment modality (65, 26.5%). All of this information is summarized in Table 1.

Discussion

On average, 19 patients were diagnosed with PCa every year at KFH, of whom 81.2% were diagnosed via histological confirmation. Slightly more than half of patients were aged 70 or younger at the time of diagnosis, and all of the patients were at least 50 years old. Among patients with complete data, the primary complaint at presentation was LUTS and/or a referral to KFH due to a high PSA level. Based on the contemporary ISUP grading system, over half of all PCa patients in this cohort had high-grade disease. The most common initial treatment was ADT, accounting for 26.5% of all treatment modalities after excluding missing data.

Despite limited data on PCa in SSA,^8,9 mortality rates for this disease are expected to double by 2040.¹⁰ Our findings are similar to those reported in existing literature. The mean age at diagnosis in our cohort (68 years) is in line with the 62.7-71.5 age range reported in other studies in SSA.¹¹ Among the 157 patients with recorded PSA levels, only 48 (30%) had PSA levels ≤ 20 ng/mL, indicating low or intermediate risk. Also, the fact that patients presented with LUTS or were referred to KFH because of elevated PSA levels rather than voluntarily presenting for PSA screenings suggests that patients tend to seek care only when symptoms appear. These observations are consistent with previous findings from studies in SSA, which have documented the symptomatic presentation of patients with late-stage PCa.¹² Because early PCa is typically asymptomatic, our findings suggest the need for initiatives that promote public awareness and early detection services among adult men in Rwanda and SSA. Additionally, population-wide research initiatives can help to elucidate the most accurate age at which to initiate PCa early detection, rather than relying solely on data from high-income countries.¹³

The primary objective of this study was to determine the pathological grade of PCa cases. The Gleason score is recognized as the best predictor of PCa prognosis.⁷ The updated 2014 ISUP Gleason score eliminated Gleason scores 2-5 from clinical practice; historically, 28% of PCa cases had Gleason scores of 2-5.¹⁴ In our study, 79 PCa slides—all of which were initially reported as PCa-positive between 2010 and 2014—were reexamined using the 2014 ISUP Gleason grading system. Twenty-four (30.7%) slides were reported as PCa-negative. This proportion is similar to the historical 28% rate of PCa cases with Gleason scores 2-5. This finding suggests that the 24 patients received some form of treatment for PCa even though they did not have the disease according to the contemporary Gleason scoring system. However, this overdiagnosis of PCa was a worldwide common practice until the contemporary grading system was implemented in 2014. These findings underscore the importance of staying up to date with emerging medical literature and updated pathology standards. In this study, the histological GGs associated with the poorest outcome (GG5 and GG4) were observed in 53.9% of patients, indicating that most patients had aggressive disease. Previous research has reported the existence of aggressive PCa cases in SSA, characterized by high Gleason scores and elevated serum PSA levels.¹⁵ Our findings further support this observation.

Because patients often present with an advanced stage of disease and due to the currently available treatment resources, most patients received ADT treatment. This is unfortunate, as most patients in this study were within the ideal age group (50-70) for curative treatment, especially surgery. In Rwanda, earlier diagnosis offers an opportunity to detect and treat localized PCa. Historically, patients with PCa who were candidates for surgery and radiotherapy were sent abroad for treatment. However, since 2017, KFH has hosted regular surgical workshops with visiting uro-oncologists to provide training on curative radical prostatectomy and on transferring surgical skills to local teams. Furthermore, since early 2019, the Rwanda Cancer Center has provided access to radiotherapy. The availability of diagnostic procedures such as PSA screening, magnetic resonance imaging, and ultrasound-guided prostate biopsy at KFH—alongside multidisciplinary tumor board discussions guided by the National Comprehensive Cancer Network Harmonized Guidelines for SSA—should help shift treatment trends from ADT to curative approaches.

However, this improvement in the standard of care should be accompanied by the enhancement of record-keeping and disease staging. In our study, tumor node metastasis staging was not reported because it was rarely documented. One way to improve PCa documentation would be to establish a clear collaboration between the Rwanda National Cancer Registry and hospital clinicians. This would promote the exchange of knowledge as well as the accuracy and consistency of data collection with respect to PCa.

In Rwanda, there are no consistent nationwide PCa awareness campaigns. Promoting cancer awareness and making screening services more accessible and affordable were among the top priorities of the Rwanda National Cancer Control Plan 2020-2024.¹⁶ However, PCa has yet to be included in the national screening campaigns. Given the high global prevalence of PCa, including in Rwanda, and its favorable prognosis when diagnosed at an early stage,¹⁷ more efforts should be invested in promoting public awareness and early detection services in Rwanda and across SSA. The strong positive correlation between prognostic groups and PSA levels (P value = 0.00), as shown in Table 2, further highlights this recommendation.

Limitations

This study has several limitations. As this is a retrospective review, there was a high rate of missing data due to inconsistent data documentation on the variables of interest. This was mainly due to technical difficulties associated with the transition from paper-based to an electronic medical record system during the initial stages of the study period. Therefore, clinical stage and survival outcomes could not be retrieved. A heavy workload among clinicians might have made documentation and patient follow-up difficult. Additionally, obtaining detailed clinical information for some patients was challenging, especially for the 17.9% of PCa specimens that were obtained from outside KFH.

Conclusion

This study reveals that a high proportion of PCa patients present at KFH with advanced, high-grade disease. These findings underscore the need for structured early detection programs and population-based research initiatives to determine the true burden of PCa in the region and to facilitate a transition to curative treatment options. Additionally, a multidisciplinary approach should be implemented to improve PCa data capture; this approach should involve clinicians, cancer registry officers, and central-level policymakers.

Ethical approval

This study was reviewed and approved by the Ethics and Research Committee at King Faisal hospital, Rwanda.

Informed consent

Informed consent was not required as the study used only secondary data.

Data availability

All data that support the findings of this study are available on request from the corresponding author (E.N.).

Conflict of interest

The authors have no conflict of interest to declare.