Purpose: To identify biomarkers for diagnosis and classification of interstitial cystitis/bladder pain syndrome (IC/BPS) by urinary lipidomics coupled with machine learning. Methods: Urine samples from 138 patients with IC/BPS, including 116 with Hunner lesion (HL) and 22 with no HL, and 71 controls were assessed by lipid chromatography-tandem mass spectrometry. Single and paired lipid analyses of differentially expressed lipids in each group were conducted to assess their diagnostic ability. Machine learning models were constructed based on the identified urinary lipids and patient demographic data, and a five-fold cross-validation method was applied for internal validation. Levels of urinary lipids were adjusted to account for urinary creatinine levels. Results: A total of 218 urinary lipids were identified. Single lipid analysis revealed that urinary levels of C24 ceramide and LPC (14:0) distinguished HL and no HL, with an area under the receiver operating characteristics curve of 0.792 and 0.656, respectively. Paired lipid analysis revealed that summed urinary levels of C24 ceramide and LPI (18:3), and subtraction of PG (36:5) from PC (38:2) distinguished HL and no HL even more accurately, with an area under the curve of 0.805 and 0.752, respectively. A machine learning model distinguished HL and no HL, with the highest area under the curve being 0.873 and 0.750, respectively. Limitations include the opaque black box nature of machine learning techniques. Conclusions: Urinary levels of C24 ceramide, along with those of C24 ceramide plus LPI (18:3), could be potential biomarkers for HL. Machine learning-coupled urinary lipidomics may play an important role in the next-generation AI- driven diagnostic systems for IC/BPS.
Background/Objectives: Lysophosphatidic acid (LPA) is a bioactive lipid mediator implicated in neuropathic pain (NP). Depressive symptoms frequently accompany NP and adversely affect outcomes, yet objective biomarkers remain limited. This study aimed to identify cerebrospinal fluid (CSF) LPA molecular species associated with depressive symptoms in patients with NP. Methods: CSF samples were obtained from 48 patients, and LPA species, including positional isomers (sn-1 and sn-2), were quantified using liquid chromatography–tandem mass spectrometry. Depressive symptoms were assessed using the depression subscale of the Hospital Anxiety and Depression Scale (HADS-D). Statistical analyses were performed to evaluate associations among depressive symptoms, clinical variables, and LPA species. Results: CSF LPA18:0 was significantly associated with HADS-D scores (r = 0.380, p = 0.010), with the strongest association observed for the sn-1 isoform (r = 0.385, p = 0.009). In multivariable analysis, LPA18:0 sn-1 remained independently associated with depressive symptoms, alongside pain intensity and pain catastrophizing (R2 = 0.386). Structural equation modeling supported an association between LPA18:0 sn-1 and depressive symptoms independent of pain-related factors. Conclusions: These findings suggest that CSF LPA18:0 sn-1 may be associated with a biological dimension of depressive distress and may represent a candidate biomarker of depressive distress in NP. However, these findings should be interpreted cautiously, as the analyses were exploratory, and further validation in independent and longitudinal cohorts is warranted.