Sociodemographic characteristics show that the mean age at diagnosis of cases was high, at 67.13 ± 8.17 years. This result is not different from that of Kaboré et al. [39] who report an average age of 71.5 years in Burkina Faso. These results indicate that the age at diagnosis in Burkina Faso is high as observed elsewhere in West Africa [40]. But our results are contrary to those obtained in various studies reporting when black men have an age at early diagnosis [41–43]. PSA levels at diagnosis were very high in our study with a mean of 627.85 ± 1153.42 ng / ml. Our results are in agreement with those obtained by Kaboré et al. in Burkina Faso with an average PSA of 537 ng / ml [35]. Our results corroborate those of Niang et al. in Senegal and Ofoha and Magnus in Nigeria [37, 44]. Tengue et al. in Togo also found PSA levels at diagnosis greater than 100 ng / ml [36]. Among the cases with their Gleason score, 82.14% have a score less than or equal to 7. This shows that the majority of these cases presented a moderately differentiated tumor at diagnosis. These different results show, on the one hand, that the diagnosis of prostate cancer is made at advanced stages of the disease and, on the other hand, the absence of prostate cancer screening programs. Regarding the family history of prostate cancer, of the 32 cases with a family history, 43.8% have a family history while 56.2% did not. These results could show that the majority of prostate cancer cases in our study population are not familial. But this trend could be due to the fact that the information was collected on the basis of verbal testimonies and not on the basis of medical records. Indeed, patients could confuse other prostate conditions (benign hypertrophy...) and prostate cancer.
Regarding alcohol consumption, our results are similar to those obtained by Dennis et al. [45] who found only a strong association between alcohol consumption and prostate cancer mortality. Our results do not support those obtained by Middleton et al. in 2009 [46] and Rota et al. in 2012 [47] in their meta-analyzes. As for cigarette consumption, our results are different from those obtained by Jones et al. in England; Cerhan et al. in the United States and Giovannucci et al. also in the United States [48–50]. All these different studies have only shown a slight increase in the risk of developing prostate cancer while a strong association was found with mortality. No association was found between physical activity and prostate cancer in our study. Our results do not corroborate those of Guéritat in France. This study demonstrated that physical exercise prevents the progression of prostate cancer either by regulating redox status and redox-dependent signaling pathways, or via the modulation of cholesterolemia or even of the expression profile of miRNAs [51]. Considering the consumption of fatty meat, our results corroborate those of Park et al. in their study of a population of Hawaii and Los Angeles and those of Dennis et al. in their meta-analysis of 4 cohort studies [45, 52].
Linkage analyzes of families at high risk for prostate cancer have provided convincing evidence that the HPC1 locus is likely to harbor a prostate cancer susceptibility gene [53]. The RNASEL gene has been proposed as a putative tumor suppressor gene located in this region by the positional cloning technique and by the candidate gene approach [54]. Association analyzes of the R462Q and D541E variants within the RNASEL gene with the Prostate cancer have achieved controversial results. Analysis of the different genotypes of the R462Q variant in our study population showed no association of this variant with prostate cancer. Our results support the conclusions of Wei et al.; Noonan et al.; and Alvarez et al. [21, 29, 30] as well as those of Fredrik et al. [34]. These studies found no association between the R462Q variant and prostate cancer. However, Casey et al. and Xiang et al. [30, 31] show that the AA genotype of the R462Q variant is significantly associated with prostate cancer. Regarding the D541E variant, our study found no association with prostate cancer. This goes hand in hand with the studies of Wei et al.; Ignacio et al.; Shook et al. [38, 55, 56] as well as several other authors [29, 30, 57, 58]. Contrary to our results, Noonan-Wheeler et al. and Wiklund et al. [33, 34] in a Swedish population observed an association between the GG genotype and an increased risk of prostate cancer.
Our results showed an association between the R462Q mutation and the degree of tumor differentiation (p ˃ 0.001). Indeed, carriers of heterozygous AG genotype (33.3%) and normal GG genotype (10.0%) presented undifferentiated tumors (Gleason ˃7) unlike carriers of mutated genotype. Our results are identical to those obtained by Alvarez-Cubero et al. in Spain [59]. On the other hand, San Francisco et al. found no association between the R462Q mutation and Gleason score [56]. For the D541E mutation, we also found an association with the degree of tumor differentiation. By observing closely, it can be seen that 21.1% of the undifferentiated tumors were carriers of the mutated GG genotype against 7.7% and 0% for carriers of the heterozygous and homozygous TT genotype respectively (p ˃ 0.001). The same result was obtained by San Francisco et al. in Chile [56]. In contrast, Alvarez et al. found no association between this mutation and the Gleason score [59]. We found no association between R462Q and D541E mutations with PSA levels at diagnosis. This shows that these two mutations in the RNASEL gene are not associated with the level of risk of the tumor (PSA level at diagnosis). Indeed, the PSA level at diagnosis makes it possible to measure the level of risk of tumor progression. For PSA values at diagnosis greater than 20 ng / ml, the tumor is considered to be at high risk [60].
The differences between our results and other studies may, on the one hand, be justified by the difference in sample sizes; the method of selection of controls and, on the other hand, by the ethno-geographic differences of the study populations. Indeed, the small size of the samples lacks the statistical power to detect associations. Also, the genetic predisposition to prostate cancer is heterogeneous (contribution of environmental and genetic factors) in its hereditary form [61] and involves the predisposition genes in a variable way depending on ethno-geographic origins.