ABSTRACT
BACKGROUND: Prostate cancer (PCa) is the most common malignancy in Mexican men. Serum prostate-specific antigen (PSA) is the usual noninvasive biomarker used for its detection. Its low specificity can increase the number of unnecessary prostate biopsies and the incidence of unpleasant complications for patients. The androgen-receptor gene (AR-CAG) repeat length and the percentage of promoter methylation (PPM) of genes glutathione-S-transferase P1 (GSTP1) and Ras association domain family 1 isoform A (RASSF1A) improve PCa detection. As an option for noninvasive assessment, we evaluated a combined analysis of all these biomarkers. PATIENTS AND METHODS: A total of 186 patients scheduled for biopsy were included in the present study. PSA and AR-CAG repeats were analyzed in blood samples. The PPM of GSTP1 and RASSF1A genes was estimated in prostate tissue and urinary sediment cells (USCs) and plasma DNA using quantitative methylation-specific polymerase chain reaction. The predictive values for PCa and benign prostatic hyperplasia (BPH), logistic regression analysis, receiver operating characteristic curve, and decision curve analysis were used to assess the differential diagnosis. RESULTS: Statistically significant differences between PCa and BPH patients were observed for all biomarkers, with higher positive and negative predictive values when all biomarkers were included in the analysis, attaining USC values of 89.2% and 78.0%, respectively. The differential diagnosis accuracy of PSA (area under the curve, 0.59) increased to 0.70 and 0.68, respectively, when the combined analysis of PPM of RASSF1Aplasma or GSTP1AUSC and AR-CAG repeats was performed. Decision curve analysis showed the utility of the combined analysis to decrease the number of unnecessary biopsies. CONCLUSION: The results showed that combined analysis of the proposed biomarkers in plasma and USCs significantly increased the confidence for the differential diagnosis for PCa and BPH. This noninvasive practice might help in the early detection of PCa and patient follow-up, avoiding to some extent unnecessary prostate biopsies.
Subject(s)
DNA Methylation , Glutathione S-Transferase pi/genetics , Prostatic Hyperplasia/diagnosis , Prostatic Neoplasms/diagnosis , Receptors, Androgen/genetics , Trinucleotide Repeats , Tumor Suppressor Proteins/genetics , Adult , Aged , Aged, 80 and over , Diagnosis, Differential , Epigenesis, Genetic , Humans , Logistic Models , Male , Mexico , Middle Aged , Promoter Regions, Genetic , Prostate-Specific Antigen/blood , Prostatic Hyperplasia/genetics , Prostatic Hyperplasia/metabolism , Prostatic Neoplasms/genetics , Prostatic Neoplasms/metabolism , ROC Curve , Sensitivity and SpecificityABSTRACT
BACKGROUND: Diabetes mellitus affects 3 to 10% of pregnant women. The reported frequency of congenital malformations (CM) in diabetic mothers is 5.5 to 10%, contributing these defects to approximately 40% of the neonatal mortality in children of diabetic mothers (CDM). OBJECTIVE: To investigate the frequency and type of congenital malformations in a sample of livebirths of diabetic mothers from the Mexican population. MATERIAL AND METHODS: The analyzed information was obtained from the RYVEMCE (Registro y Vigilancia Epidemiológica de Malformaciones Congénitas). We analysed the type and frequency of the different CM diagnosed. These frequencies were compared with the whole amount of those CM included in the database of our registry (20,381). As part of the analysis, other maternal and neonatal variables were also compared between CDM and the control group. RESULTS: A total of 314 CDM (0.77%), 234 malformed and 80 non malformed were identified. The frequencies of cleft palate (CP), limb reduction defect (LRD) and microcephaly (MIC) were significantly higher in CDM than in the rest of malformed newborns of not diabetic mothers of the RYVEMCE (OR: 9.9, 3.8 and 10.0, respectively). A higher frequency of macrosomia was observed in CDM (18.0%) than in controls (6.1%), OR: 3.4, p < 0.001, in the frequency of preterm birth (28.5% than controls 13.0%), OR: 3.02, p < 0.0001 and in caesarean delivery (71.5% than controls 44.4%) OR: 3.15, p < 0.00001. CONCLUSIONS: Our results confirm the higher frequency of CM in CDM and in particular a higher risk for CP, LRD and MIC. Pregnancy and delivery complications such as macrosomia and preterm and caesarean delivery were also more frequent in CDM that controls. Certain associations of CM not described previously, were observed in the studied sample. Our results confirm the need of pregnancy planning including a pre-gestational and gestational control of maternal glycaemia, particularly in a population with such a high prevalence of diabetes mellitus as the observed in the Mexican one.
