A novel algorithm-based risk classification for vascular damage in men with erectile dysfunction.

Abstract

Background: Penile dynamic color doppler duplex ultrasound (CDDU) is a relevant tool in assessing men with suspected vasculogenic erectile dysfunction (V-ED).

Aim: To investigate (1) factors potentially associated with V-ED to define risk classes useful in predicting V-ED; (2) the response to phosphodiesterase type 5 inhibitors (PDE5i); and (3) the onset of incident major cardiovascular (CV) events.

Methods: A cohort of men with ED and without known concomitant CVD was grouped into: patients undergoing CDDU (N. 301) and patients not undergoing CDDU but prospectively monitored for incident major CV events after initiating PDE5i (N. 127). Logistic regression and Chi-square Automatic Interaction Detectors (CHAID) methodology were employed to identify potential predictors and develop a novel risk classification system. Receiver operating characteristic (ROC) curves and decision curve analysis was performed to assess its accuracy.

Outcomes: Factors associated with V-ED useful to develop a novel risk classification system predicting incident major CV events and PDE5i response.

Results: The new classification defines patients as follows: Very Low Risk [age < 53, body mass index (BMI) < 25 Kg/m2], Low Risk (age < 53, BMI > 25 Kg/m2, non-smokers), Moderate Risk (age > 53, non-smokers), High Risk (age < 53, BMI > 25 Kg/m2, smokers), and Very High Risk (age > 53, smokers). Multivariable logistic regression analysis highlighted age, BMI, and smoking as significant predictors of V-ED. CHAID methodology yielded a risk classification system with an accuracy of 0.79. Notably, "Very High Risk" class was associated with a significantly increased risk of incident major CV events [odds ratio (OR) 4.00, 95% confidence interval (CI) 1.06-15.08, P < .05]. Moreover, patients belonging to "Very High Risk" and "High Risk" classes were also associated with diminished PDE5i response. At Kaplan-Meier analysis, men belonging to "Very High Risk" class depicted a notable risk of incident major CV events (P = .03).

Clinical implications: We propose a novel risk classification system which may have some clinical value in tailoring patients at significantly higher risk of V-ED. Although preliminary, current findings also suggest that the novel risk classification system could help tailoring men at potential increased risk of incident major CV events and those not responding to PDE5i.

Strengths and limitations: This study introduces a novel user-friendly risk stratification tool for V-ED, emphasizing the need for CV screening and alternative therapies for higher-risk groups. A limited number of events in the cohort with follow-up for major CV events and response to PDE5is constrains the interpretation of the results. Current findings need an external validation cohort.

Conclusion: Patients with ED categorized as either "Very High Risk" or "High Risk" should undergo a CDDU due to an increased risk of V-ED. Additionally, despite the clinical impact of these findings need further investigation, patients classified as "Very High Risk" could face a heightened risk of major CV events and a lower response to PDE5is.