Search

Background/Objectives Circulating endothelial cells (CECs) are indicative of vascular injury and correlate with severity of vascular diseases. A pilot study showed that the ratio of CEC to platelet count (CEC/PC) was effective in predicting cirrhosis. Therefore, we evaluated CEC/PC in a larger cohort of patients, correlated it with cirrhosis, and compared its operating characteristics with previously described biomarker for cirrhosis, the AST/platelet ratio index (APRI). Methods Fifty-three patients with cirrhosis, 20 matched healthy controls, and 9 patients with noncirrhotic liver disease were recruited. Peripheral blood sample was collected and analyzed to enumerate nucleated CEC CD146+, CD105+, CD45– using a commercial assay. Results Median CEC counts were significantly higher in patients with cirrhosis (62 cells/4 mL, interquartile range [IQR]: 43.5–121) as compared with controls (31 cells/4 mL, IQR: 22.2–40). The CEC/PC was also significantly elevated in cirrhotics (0.69, IQR: 0.39–1.48) compared with controls (0.12, IQR: 0.09–0.20) and noncirrhotics (0.21, IQR: 0.08–0.43). Receiver operator characteristic (ROC) analysis revealed that CEC cutoff value of ≥37 cells/4 mL showed sensitivity of 81% and specificity of 75% for differentiating cirrhosis from controls (area under the curve [AUC]: 0.80; 95% confidence interval [CI] 0.67–0.91). The CEC/PC ratio cutoff value of ≥0.23 showed sensitivity of 91% and specificity of 82% (AUC: 0.92; 95% CI 0.83–0.99). The APRI cutoff value of ≥0.4 showed sensitivity of 94% and specificity of 85% for differentiating cirrhosis from control patients (AUC: 0.96; 95% CI 0.90–1.0). A product of CEC and APRI, termed CAPRI (CEC-APRI), effectively distinguished patients with cirrhosis from controls; with cutoff value of ≥12.7, showing higher sensitivity of 98% and specificity of 85% (AUC: 0.98; 95% CI 0.96–1.0). Conclusion The CEC/PC ratio is significantly elevated in patients with cirrhosis and demonstrates comparable operating characteristics to previously described APRI. Furthermore, CAPRI, compiled as product of CEC to APRI showed outstanding ability to distinguish patients with cirrhosis from controls, although larger studies are necessary for validation.

Background/Aims:Muscle cramps are common in patients with cirrhosis, particularly when diuretics are used for ascites. Its pathophysiologic mechanism remains uncertain. In this work, we conducted field testing of a questionnaire to measure the extent and severity of muscle cramps in patients with cirrhosis and explored plasma metabolomic biomarkers for muscle cramps. Methods: Patients with an established diagnosis of cirrhosis were prospectively contacted prior to their follow-up appointment in our liver clinic. Patients were asked to fill out a muscle cramps questionnaire (mMCQ) which asks about occurrence, frequency, location and impact (sleep and daily living) of muscle cramps. Blood samples were drawn after a minimum of 12 hours of fasting, which were promptly cold-centrifuged and plasma was separated for metabolomics assays. Results: In this on-going study, 109 patients have been contacted to date, of whom 46 patients have responded to the questionnaire (response rate=42% to date). The respondents were 57.8 ± 11.1 years of age and 59% were men. The mean MELD score was 12.8 ± 5.5. Approximately half (46%) reported history of hepatic encephalopathy and 6.5% refractory ascites requiring therapeutic paracentesis. In this patient cohort, the prevalence of muscle cramps was 76%. Of those who reported muscle cramps, 50% had moderate to severe symptoms (occuring at least daily or causing at least moderate disturbance of activities of daily living). To date, we have assayed 25 amino acid in the plasma samples (n=43), including taurine, threonine, serine, asparagine, glutamic acid, argininosuccinic acid, glutamine, proline, glycine, alanine, citruline, α-amino-N-butyric acid, valine, cysteine, methionine, isoleucine, leucine, tyrosine, phenyl alanine, β-alanine, ornithine, lysine, histidine, arginine, and allo-isoleucine. Of these, plasma cysteine levels were significantly different according to the severity of muscle cramps. In patients without muscle cramps the median plasma level was 59 nmol/mL with an interquartile range (IQR) of 52-75 nmol/mL, compared to those with cramps (median=82 nmol/mL, IQR=61-94). In the figure, patients with severe cramps had the highest cysteine levels (median 90 nmol/ mL, IQR=78-106) compared to those with mild cramps (median:73 nmol/mL, IQR=51-86). Conclusion: Muscle cramps are extremely common in patients with cirrhosis receiving ongoing care at a liver clinic. Plasma cysteine levels may be a potential biomarker for muscle cramps, as they correlate with their occurrence and severity.