Prognostic Value of CA125 Serum Levels in Female Patients with Acute Coronary Syndrome

Background: carbohydrate antigen 125 (CA125) is an increasingly promising biomarker of heart failure (HF), but its prognostic value in female patients with acute coronary syndrome (ACS) is unclear. We aimed to determine the short-term and mid-term prognostic value of CA125 serum levels in female ACS patients. Methods: A total of 131 consecutive female patients with ACS were retrospective enrolled. Their CA125 levels, B-type natriuretic peptide (BNP) levels and biochemical parameters were measured, and echocardiography was performed at admission. All-cause mortality during hospitalization and two-year follow-up was investigated for the prognosis. Results: The median value of CA125 serum level in the entire ACS patients was 13.85 U/mL. Patients in Killip (cid:0) had the highest values of CA125 level, followed by Killip (cid:0) and then Killip (cid:0) (p < 0.05). However, no statical difference was observed between Killip (cid:0) and (cid:0) - (cid:0) groups respectively (P > 0.05). The CA125 serum levels showed weak positive correlation with left ventricular end-diastolic diameter (LVEDD) (r = 0.3, P < 0.01) and a weak negative correlation with left ventricular ejection fraction (LVEF) (r = –0.23, p < 0.01). A receive operating characteristic (ROC) curve analysis showed that the AUC of CA125 in predicting acute heart failure (AHF) in ACS patients during hospitalization was 0.912, exhibiting higher sensitivity and specicity than BNP (0.846). The optimal cut-off value for CA125 in predicting AHF was 16.4 U/mL with a sensitivity of 0.916 and specicity of 0.893. The Kaplan-Meier survival analysis demonstrated that patients with high values of CA125 level had a poor overall survival than those with low values of CA125 level (log-rank, p < 0.001), whether during hospitalization or mid-term follow-up.

levels showed weak positive correlation with left ventricular end-diastolic diameter (LVEDD) (r = 0.3, P < 0.01) and a weak negative correlation with left ventricular ejection fraction (LVEF) (r = -0.23, p < 0.01). A receive operating characteristic (ROC) curve analysis showed that the AUC of CA125 in predicting acute heart failure (AHF) in ACS patients during hospitalization was 0.912, exhibiting higher sensitivity and speci city than BNP (0.846). The optimal cut-off value for CA125 in predicting AHF was 16.4 U/mL with a sensitivity of 0.916 and speci city of 0.893. The Kaplan-Meier survival analysis demonstrated that patients with high values of CA125 level had a poor overall survival than those with low values of CA125 level (log-rank, p < 0.001), whether during hospitalization or mid-term follow-up.
Conclusion: Elevated CA125 level can be used to predict AHF in female ACS patients. Patients with elevated CA125 levels had higher mortality in short-term and mid-term than those with low CA125 levels.

Background
Currently, acute coronary syndrome (ACS) is a major cause of morbidity and mortality worldwide despite advances in therapeutic techniques. Heart failure (HF) is the most common complication and leading cause across all types of ACS [1]. Women with ACS have higher adjusted risk than men of dying and developing HF during hospitalization and subsequent ve years [2,3].
Carbohydrate antigen 125 (CA125) is a well-known tumor biomarker associated with ovarian malignancy [4,5]; moreover, CA125 level is also elevated in congestive HF (CHF) [6][7][8]. Several studies demonstrated that the CA125 levels correlate with clinical severity, parallel with NYHA class in CHF patients [9][10][11]. In the past two decades, the prognostic value of CA125 in HF has been extensively investigated, and increased CA125 levels have been identi ed as an independent predictor of poor short-and mid-term outcomes, including mortality and rehospitalization rates, in acute and chronic HF of any etiology [12][13][14]. De Gennaro L et al. evaluated the relationship between CA125 concentration and AHF in ACS patients, indicating that the CA125 levels were signi cantly proportional to Killip class [15].
However, the precise relationship between elevated CA125 concentrations and short-term and mid-term prognosis in ACS patients has not yet been su ciently studied. The aim of this study was to determine the short-term and mid-term prognostic value of CA125 serum levels in female ACS patients.

Collection of case data
This was a retrospective study including 131 consecutive female ACS patients in our hospital between January and December 2018. Thirty-one patients were diagnosed with unstable angina (UA); 52 patients were diagnosed with non-ST-elevation myocardial infarction (NSTEMI); and 48 patients were diagnosed Continuous variables were expressed as mean value ± standard deviation (SD), where categorical variables were expressed as absolute numbers (percentages). Pearson's test was conducted to examine the correlations between two parameters. Receive operating characteristic (ROC) curves were used to illustrate the diagnostic ability of CA125 and BNP. The area under the ROC curve (AUC), true positive rate (also known as sensitivity or recall), and false positive rate (speci city) are shown in a graphical plot [19]. Kaplan-Meier survival analysis of differences between and within groups was performed by conducting a log-rank test. A p value of less than 0.05 was considered statically signi cant. All the analyses conducted in this study were carried out using R statistical software (version 3.6.1).

Results
The mean age of enrolled patients was 69.3±11.6 years; 94 patients (71.8%) had hypertension; 47 patients (35.9%) had diabetes mellitus, and 17 patients (13.0%) had atrial brillation. The clinical characteristics of the patient cohort are shown in Table 1. The median value of CA125 serum level in the entire ACS patients was 13.85 U/mL. CA125 levels in the STEMI and NSTEMI groups were signi cantly higher than those in the UA group (p < 0.05 and p < 0.05, respectively), but no signi cant difference was observed between the two groups (p > 0.05, Figure 1-A).
Additionally, the serum levels of CA125 were higher in patients with pleural and pericardial effusion than those without pleural and pericardial effusion (p < 0.05, Figure 2-B). Patients who had developed AHF during hospitalization had higher levels of CA125 than those without AHF (p < 0.05, Figure 1-C). Patients in Killip demonstrated having the highest levels of CA125, followed by Killip , and then Killip (p < 0.05, Figure 1-D). However, no statistical difference was observed between Killip and -groups (P > 0.05, Figure 1-D).
The CA125 serum levels were evaluated to determine their correlation to echocardiographic parameters and plasma BNP. The CA125 levels showed a weak positive correlation with LVEDD (r = 0.3, P < 0.01, Figure 2-A), and weak negative correlation with LVEF (r = -0.23, p < 0.01, Figure 2-B). No signi cant correlation was observed between the CA125 concentrations and plasma BNP levels (r = 0.15, p > 0.05).
A total of 26 patients developed AHF during hospitalization: nine patients died during hospitalization and ve patients died during follow-up. The diagnosis of AHF requires the presence of crackles (Killip ≥ ) or the use of intravenous diuretic agents or intravenous inotropes. The ROC curve analysis showed that the AUC of CA125 in predicting AHF in female ACS patients during hospitalization was 0.912, and that of BNP was 0.846 ( Figure 3). Compared with BNP, CA125 showed higher sensitivity and speci city in predicting HF during hospitalization in ACS patients. The optimal cut-off value for CA125 in predicting AHF was 16.4 U/mL with a sensitivity of 0.916 and speci city of 0.893.
After a mean of 21.5±7.0 months of follow-up duration, a total of 21 deaths were identi ed, 11 deaths during hospitalization and 10 deaths at follow-up. The patients were divided into high and low groups according to the optimal cut-off value of CA125 that predicted AHF during hospitalization. The Kaplan-Meier survival analysis demonstrated that patients with high levels of CA125 had a poor overall survival than those with low levels (log-rank, p < 0.001), whether during hospitalization or mid-term follow-up ( Figure 4).

Discussion
In ACS patients, a number of novel biomarkers have been proposed in prognosis evaluation and risk strati cation. CA125 is a well-known tumor marker for diagnosis, monitoring, and risk strati cation in ovarian malignancy; it has become an increasingly promising biomarker in HF in the recent two decades [10,20,21].
In this study, the CA125 levels increased with the severity of HF in patients with Killip -, but the CA125 levels did not increased signi cantly in patients with Killip . This is probably because the Killip patients developed AHF several days after admission; however, the CA125 levels were measured only at admission. Hemodynamic changes after ACS resulted in an increase in congestion and hydrostatic pressure in the mesothelium, thereby provoking mesothelial cells to initiate CA125 synthesis [22]. The kinetics of CA125 release from mesothelial cells in ACS patients is not known; it may be similar to NT-proBNP. The values obtained after a few days after onset of symptoms may have superior prognostic value when compared with the measurements on admission [23].
In this study, 19.8% of total patients suffered from AHF, which was consistent with the incidence of postmyocardial infarction HF [24]. Patients with acute myocardial infarction are more likely to have higher CA125 levels compared with those with UA, whether STEMI or NSTEMI. The is probably because patients with myocardial infarction lose more functional cardiomyocytes due to myocardial stunning and necrosis compared to UA patients, increasing the risk of AHF. A retrospective cohort study of ACS patients also revealed that the incidence of HF among STEMI and NSTEMI was higher than UA patients during hospitalization and one-year follow-up [25].
So far, the relationship between echocardiographic parameters and CA125 is not conclusive due to the controversial results obtained from different studies. This study focused on the relationship between CA125 and LVEDD, and LVEF in female ACS patients. CA125 showed a weak positive correlation with LVEDD and negative correlation with LVEF. CA125 also demonstrated weak negative correlation with LVEF in CHF patients in another retrospective study [9]. However, in another study of ACS patients, CA125 was found to be related to LVEDD (r = 0.66, P < 0.001) and LVEF (r = -0.37, P < 0.01) [15]. CA125 levels were correlated with the deceleration time of early lling on transmitral Doppler (r = -0.63, p < 0.05), pulmonary artery pressure (r = 0.66, p < 0.05) and right atrial pressure (r = 0.69, p < 0.05). However, no signi cant correlation was observed between CA125 concentration and LVEF or LVEDD. In another study including 77 CHF patients, a weak correlation was observed between the CA125 levels and right ventricular systolic pressure, but no signi cant correlation was observed between CA125 and LVEDD, LVEF, or DT [9]. Hakki Yilmaz et al. analyzed the relationship between CA125 levels and left ventricular function in patients with end-stage renal disease on maintenance hemodialysis [26]. CA125 levels showed a signi cant positive correlation with LVEDD (r = 0.599, p < 0.001), left ventricular end-systolic diameter (LVESD, r = 0.750, p < 0.001), and left ventricular mass index (LVMI, r = 0.378, p < 0.05) and negative correlation with LVEF (r = -0.878, p < 0.001).
In contrast to previous studies, CA125, as shown in this study, is not related to BNP, but it is an excellent predictor of AHF in ACS patients during hospitalization, even better than BNP. Robust data showed that elevated BNP is highly predictive of AHF and all-cause mortality in ACS patients when compared with lower levels. Our results are also different from those of Azra Durak-Nalbantic1 et al., who found a signi cant although weak positive correlation of CA125 and BNP in CHF patients (r = 0.293, p < 0.05) [27]. Dursum Duman et al. studied the relationship between the CA125 and BNP concentrations in patients with advanced HF: LnCA125 was signi cantly correlated with LnBNP (r = 0.78, p < 0.001), and CA125 levels were independently associated with BNP (β = 0.58, p < 0.001) [11]. Luisa De Gennaro also revealed that elevated CA125 levels identi ed ACS patients with AHF with a higher speci city (97.1 vs. 31.4%), positive predictive value (83.3 vs. 33.3%), and accuracy (83.0 vs. 48.9%) when compared with BNP [15].
Many studies have already demonstrated that CA125 is a robust prognostic marker in both acute or chronic HF, and elevated CA125 level is associated with poor short-term and long-term outcomes, including mortality and rehospitalization rates. Julio Núñez et al. showed a positive trend between the quartile 1 to 4 of CA125 and 6 months' mortality in patients with AHF [28]; similar results also obtained in the latest research in patients with decompensated HF [21]. In a retrospective study of 55 CHF patients who underwent cardiac transplantation, the survival was signi cantly inferior in the group with elevated CA125 level compared with normal group at two-, ve-, and eight-years follow-up [29]. A recent systematic meta-analysis including 16 studies showed that elevated CA125 levels were associated with a 68% increase in all-cause mortality (eight studies, HRs: 1.68, 95% CI: 1.36 to 2.07; p < 0.001; I 2 : 74%) and 77% increase in HF-related readmissions ( ve studies, HRs: 1.77, 95% CI: 1.22 to 2.59; p < 0.01; I 2 : 73%) in AHF patients [30]. In this study, we also con rm that elevated CA125 levels are associated with higher mortality in female ACS patients both during hospitalization and mid-term follow-up. Similar results were obtained by Felipe Falcão et al.; patients with high CA125 values according to the optimal cutoff (11.48U/mL) had a higher rate of mortality at six months than those with low CA125 values in STEMI cohort [31].
Some limitations of this study should be highlighted. First, this is a relatively small-sized retrospective study of only female ACS patients conducted in a single center, which may limit the extrapolation of our results to all populations. The results of this study should be validated in larger trials and male crowd. Second, CA125 concentrations were determined only at hospital admission. The level might depend on the time elapsed from the onset of chest pain; serial measurement may provide more accurate prognostic value. Third, we cannot ignore the fact that factors that in uence mortality were not been considered in our analysis (e.g., treatment during hospitalization and follow-up).

Conclusions
Elevated CA125 levels can be used to predict developing AHF in female ACS patients. Patients with elevated CA125 levels had higher mortality in short-term and mid-term than those with low CA125 levels.