Epidemiological data
We identified 102 cases of prostate cancer in a single year. Tengue et al. in 2014 found an annual frequency of 77.3 cases [9]. This demonstrates a clear increase in the incidence of prostate cancer in Togo. The average age of our patients was 67.79 ± 8.5 years with age ranges of 50 and 86 years. The age range [65; 70[ years was the most represented with a percentage of 24.5. Similar age averages were reported by Tengue et al. (68.5 years), as well as Mbey et al. (68.7 years) [9, 10]. Indeed, prostate cancer remains the prerogative of the elderly according to the literature [1, 8, 11]. A family history of prostate cancer was noted in 17.7% of patients. Darré et al. found a family history of prostate cancer in 41.38% [12].
Data related to late diagnosis of prostate cancer
Functional urinary signs were the most frequently recorded revealing signs with 95 cases of pollakiuria and 6 4 cases of dysuria. These results are similar to those of Tengue et al. who described symptoms of the lower urinary tract in 89.2% [9]. Darré et al. had also found functional urinary signs in 37.93% [12]. This high frequency of functional urinary disorders could be explained by the fact that they constitute a nuisance in daily life forcing the patient to consult. Bone pain and deterioration of general condition were found in 41 cases and 47 cases respectively. Tengue et al. also found an alteration of the general condition respectively in 82.3% [9]. Overall, the existence of functional manifestations during prostate cancer disease would reflect an advanced stage of this pathology [13]. The TNM classification was specified in all our respondents, which allowed us to establish the classification into stages. The entire series consisted only of advanced cases, classified stage PT3NxMx (31.37%) and pT4NxMx (68.63%). These results are similar to those reported by Tengue et al. which found 75.9% of metastasized tumors [9]. Likewise Sine et al. found 96.5% of advanced cancers [14]. The current trend shows that approximately 15% of patients with prostate cancer are diagnosed at an advanced stage [15]. Histologically, these were adenocarcinomas in 99% and squamous cell carcinoma in 1%. These were only adenocarcinomas in the series by Darré et al. [12]. A predominance of prostate adenocarcinomas was also reported by Ndoye et al. in 97.7% of cases [16]. According to the literature, prostate adenocarcinoma represents more than 90% of cases of prostate cancer; squamous cell carcinoma and neuroendocrine carcinoma being very rare [2, 3]. The Gleason score was specified in all patients. In our study, ISUP grade 5 corresponding to Gleason score 9 was the most represented with 30.39%. Bouanika et al. as well as boulfhrad et al. also reported a predominance of score 9 [17, 18]. The average PSA level at the time of diagnosis was 521.39 ng/ml with extremes of 11 and 6140.5 ng/ml). Omisanjo et al. found a similar mean PSA of 563.2ng/ml [19]. Niang et al. as well as Sine et al. found a mean PSA value before any treatment of 1447.57 ng/ml and 1367 ng/ml respectively [14, 20]. Screening for prostate cancer in the presence of elevated serum PSA levels has been used in men aged over 50 years since at least 1992 [21]. Indeed, screening based on prostate specific antigen (PSA) is one of the most common methods of diagnosing prostate cancer. However PSA expression has also been reported in other tissues such as normal prostate epithelium, small intestine, kidney cells and salivary organs, but the expression level in these parts is 100 to 1000 times lower than that of prostate cancer [22]. In 2018, the U.S. Preventive Services Task Force suggested that for all men aged 70 and older, prostate cancer screening based on a PSA blood test be suspended; for men aged 55 to 69, individual screening based on an informed discussion with the doctor and no screening for men under 55 [23, 24]. In routine practice, guidelines are often not followed and one of the main reasons for this is the lack of knowledge of the recommendations among primary care general practitioners [25, 26]. These generally high PSA levels in our context could be explained by an often late diagnosis.
In our series, only 36.3% of patients had to travel a distance of more than 20 km compared to 67.7% who had traveled distances between 0 and 20 km as part of a consultation in a health structure. These results show us that patients living far away consulted less. Also Amborggi et al. reported that the burden of travel between a patient's residence and health centers was an important problem that could influence the early diagnosis and choice of cancer treatment [27]. Thomas et al. demonstrated that travel burden (measured as travel distance or time) can lead to diagnostic delays and can influence treatment choice for a variety of common cancers [28]. In the series by Kaboré et al. patients with poor geographic access or living in rural areas (18.3%) were twice as likely to have advanced stages of cancer than urban residents (9.6%) [29]. Only 3.92% of our respondents were aware of screening and the means of screening for prostate cancer. This result is much lower than that reported by Kaboré et al. which was 29.7% [29]. In the series by Darré et al. 51.43% of participants had no knowledge of the PSA blood test, with a statistically significant correlation between this knowledge of the PSA blood test with the level of study and the effectiveness of an internship in a health service. urology [30]. In our study we had difficulty determining the precise consultation period because patients cannot remember exactly the duration before their first consultations. Nevertheless, we used the medical records of these patients. The average consultation time in our series was 20.73 ± 10.17 months. The majority of patients (53.9%) consulted within less than 21 months compared to 46.1% who consulted beyond 21 months. This could be explained on the one hand by the fact that in our society we consider that urinary problems are normal with age; on the other hand, by the lack of knowledge of the signs of the pathology, illiteracy, the preeminence of traditional therapy and by the poverty of the population, the vast majority of which is rural or unemployed in our country. Factors such as poverty, religion, culture, communication barriers, language and costs are thought to affect men's participation in prostate cancer screening [31].
Factors associated with late diagnosis of Prostate Cancer with statistical analyzes
After univariate analysis, five factors were associated with the long diagnosis delay, namely the absence of health insurance, lack of financial means, type of first consultation, diagnostic error and taking a second opinion. Multivariate analysis of all these parameters reveals that two factors were associated with late diagnosis of prostate cancer. The first factor was diagnostic error (with a p-value = 0.032; ORa = 1.30; 95% CI [1.10–4.9]). Thus, for a patient whose initial diagnosis was other than prostate cancer, there would be a 30% higher risk of late diagnosis. This may be explained by insufficient knowledge of the signs and symptoms of prostate cancer by healthcare providers, especially at the primary level. The second factor was the taking of a second opinion (with p-value = 0.04; OR = 3.26; 95% CI [1.06–10.1]; which means that a patient whose initial diagnosis required a second opinion, would have a 3.26 higher risk of late diagnosis. This second opinion may be linked on the one hand to a lack of information and awareness and on the other hand to the strong cultural tendency of population towards traditional therapy. The existence of health insurance would promote early diagnosis [31–33]. Kabore et al. reported the lack of screening program and the insufficiency of the technical platform as associated with the late diagnosis of prostate cancer in our African context [29].
Limitations
The study was carried out among patients followed for prostate cancer in the urology department of the Sylvanus Olympio University Hospital Center (CHU) in Lomé; which limits its representativeness to other patients outside this center or in different care contexts. The results cannot be generalized to all patients with prostate cancer in Togo. The convenience sampling method used in the study may introduce selection bias because participants were chosen based on their availability and access to health care. The data was collected using a questionnaire developed on the basis of a literature review and translated into different languages for a better understanding of the respondents, which may introduce information bias. Participants' responses may be influenced by comprehension errors linked to approximate translations of the questions or by a desire to provide socially acceptable answers.