162、PL-ED:基于代谢—炎症状态交互的勃起功能障碍治疗反应计算与可证伪预测(PL-ED: Computational Modeling and Falsifiable Prediction of Erectile Dysfunction Treatment Response Using Metabolic–Inflammatory State Interactions)

Phosphodiesterase type 5 inhibitors (PDE5is) remain the first-line pharmacological treatment for erectile dysfunction (ED), yet real-world unresponsiveness rates reach 48.1% in specific cohorts, and no pre-treatment tool reliably identifies the non-responsive phenotype. Current ED assessment—centered on IIEF-EF, EHS, morning testosterone, and selective specialized investigations—follows a “etiologic classification—severity grading—treatment selection” logic that does not explicitly organize systemic metabolic burden and inflammation–albumin-related burden into a testable predictive structure. Prim-Lex Medicine posits the Animal-Life Prim-Fire as its theoretical anchor, but its direct measurements (NAD(P)H FLIM, lactate/pyruvate ratio, mitochondrial function assays) are inaccessible in general hospitals, creating a methodological tension between theoretical clarity and clinical measurability. This paper develops the PL-ED Framework. Its core methodological shift is explicit: rather than “measuring the Prim-Fire” in general hospitals, the framework tests whether the Prim-Fire level has left reproducible treatment-response predictive traces—independent of conventional clinical variables—in routinely available hematological and clinical data. The detection layer comprises six groups of routine information: IIEF-EF/EHS, standard clinical data, fasting plasma glucose (FPG) and HbA1c, SIRI derived from complete blood count, hs-CRP, and morning total testosterone. The computation layer simplifies the Five-Element structure into two candidate modules: Fire (F) , an equal-weight aggregation of standardized FPG and HbA1c, positioned as a joint candidate proxy for host metabolic burden; and Wood (W) , an equal-weight aggregation of standardized log SIRI and log(1+hs-CRP), positioned as a joint candidate proxy for systemic inflammatory burden. The Fire–Wood interaction is the sole coupling structure entering formal testing. Three substantive methodological revisions are made to the prior framework. First, the primary outcome is frozen as the change in IIEF-EF from baseline to week 12 (with baseline-adjusted follow-up score as the primary analysis and change score as sensitivity analysis), rather than simultaneously suspending diagnostic, severity, and treatment-response outcomes. Second, a five-tier model comparison hierarchy is established: (clinical baseline), , , (baseline + four standardized raw variables), and ; the comparison of versus is the critical contrast for judging whether equal-weight aggregation confers parsimony advantages, rather than comparing with alone. Third, H1 is judged primarily by on the validation set; MAE, calibration (intercept/slope/plot), and predictive serve as supporting evidence. serves only as an exploratory target difference, not a universal clinical-importance threshold. DCA does not enter the H1 judgment for the continuous primary outcome; it is used only for exploratory analysis of the binary secondary outcome when a pre-specified, clinically defensible decision scenario exists. The framework explicitly states that LDH, FPG, SIRI, and hs-CRP are not direct measures of NADH/NAD⁺ or Prim-Fire coherence; that a statistically significant Fire–Wood interaction does not constitute proof of Five-Element generative or restrictive biological mechanisms; and that PL-ED scores must not replace PDE5i indication judgment, contraindication screening, or clinical decision-making. Its anticipated contribution is strictly limited: a minimum viable clinical validation version executable in secondary and tertiary hospitals, with actually collectable data and honestly falsifiable hypotheses.