Abstract
Introduction: Antigen carbohydrate 125 (CA125) has emerged as a proxy of fluid overload and inflammation in acute heart failure (AHF). We aimed to evaluate the influence of dapagliflozin on CA125 levels within the first weeks after discharge and whether CA125 changes were related to 6-month adverse clinical outcomes. Methods: In this retrospective observational study, data from 956 AHF patients discharged from a tertiary hospital were analyzed. CA125 levels were assessed during the index admission (visit 1) and at a median of 26 (15–39) days after discharge (visit 2). The primary endpoint was changes in CA125 and its correlation with the risk of 6-month death and recurrent readmissions (any or AHF-related). Multivariable mixed regression and a two-equation count model regression were used for the analyses. Results: The mean age of the cohort was 73.1 ± 11.1 years, 54.8% were males, 43.5% showed left ventricular ejection fraction ≥50%, and 18.7% of patients received dapagliflozin at discharge. Dapagliflozin treatment was associated with a greater reduction in CA125 levels at follow-up (−24 U/mL) compared to non-dapagliflozin patients (−14 U/mL, p = 0.034). The magnitude of CA125 reduction (per decrease in 10 U/mL) was significantly associated with a lower risk of 6-month death (incidence rate ratio [IRR] = 0.98, 95% CI = 0.96–0.99; p = 0.049), all-cause readmissions (IRR = 0.99, 95% CI = 0.98–0.99; p = 0.003), and HF readmissions (IRR = 0.98, 95% CI = 0.97–0.99; p < 0.001). Conclusion: Dapagliflozin treatment at discharge following an episode of AHF was associated with a greater reduction in CA125 during the first weeks after discharge. The greater CA125 reduction identified patients with a lower risk of 6-month adverse clinical outcomes.
Introduction
Acute heart failure (AHF) is a clinical condition associated with high morbidity and mortality rates, [1]. Decompensations often lead to repeated hospital admissions, significantly burdening healthcare resources. Recent developments, such as the introduction of sodium-glucose cotransporter 2 inhibitors (SGLT2i), have shown promising results in improving clinical outcomes in AHF setting [1‒4].
Fluid overload is a central component in AHF patients, directly influencing patient symptoms and morbimortality risk [5, 6]. While natriuretic peptides have traditionally been used to assess congestion, their effectiveness can be limited in some frequent scenarios such as elderly, kidney dysfunction, obesity, and patients with predominant systemic interstitial fluid overload [7, 8]. Circulating levels of antigen carbohydrate 125 (CA125) have emerged as a surrogate marker of fluid overload and inflammation in AHF [9‒11]. Elevated CA125 levels are associated with greater fluid overload, especially extravascular, and poorer outcomes after AHF [9‒11]. Likewise, changes over time are closely related to clinical status and risk of adverse outcomes [9]. Additionally, two randomized clinical trials showed their utility for guiding diuretic therapy [12, 13].
Conversely to natriuretic peptides, the utility of CA125 seems not significantly influenced by age, renal impairment, and left ventricular ejection fraction (LVEF) status [7, 9]. Prior studies suggested a significant CA125 decrease following SGLT2i initiation in stable ambulatory heart failure (HF) patients [14, 15]. However, the impact of SGLT2i treatment on CA125 levels following hospitalization for AHF and whether these changes are related to the risk of adverse clinical outcomes remains unknown. This study aimed to investigate the impact of dapagliflozin, an SGLT2i, on short-term CA125 changes following AHF and to determine whether changes in CA125 levels are associated with adverse clinical outcomes.
Methods
Study Design and Participants
This retrospective observational study utilized a cohort of 1,088 consecutive patients discharged alive following hospitalization for AHF in a third-level center in Spain (Cardiology Department of Hospital Clínico Universitario de Valencia-Spain). Only patients with a primary diagnosis of AHF were considered. AHF was defined as a rapid or gradual onset of symptoms and/or signs of HF, severe enough for the patient to seek urgent medical attention, leading to an unplanned hospital admission or emergency department visit. All patients included in the study were diagnosed by cardiologists including measurement of natriuretic peptides and echocardiography assessment revealing cardiac structural or functional abnormalities. Patients with AHF not requiring hospital admission, those with cardiogenic shock or requiring mechanical circulatory support (n = 10), and those with known contraindications to the use of dapagliflozin or other SGLT2 inhibitors, such as advanced renal insufficiency (glomerular filtration rate <20 mL/min/1.73 m2) (n = 32) or history of diabetic ketoacidosis (n = 6) were excluded. Additionally, patients receiving other SGLT2 inhibitors besides dapagliflozin (n = 45), as well as those discharged to receive ambulatory follow-up outside our health center (n = 87), were excluded from the study.
Dapagliflozin was generally prescribed during admission after clinical stabilization by the attending cardiologist, based on the patient’s stability and eligibility for outpatient HF management. After discharge, patients were responsible for obtaining dapagliflozin from a pharmacy using a medical prescription. An echocardiographic assessment was performed in all patients during the index hospitalization. The study period spanned from October 14, 2019, to December 31, 2022. Following discharge, structured follow-up was conducted in a specialized HF unit. This study was conducted in compliance with the principles of the Declaration of Helsinki and was approved by an Institutional Review Committee.
Data Collection
Patient demographics, vital signs, physical examination, medical history, laboratory test, 12-lead electrocardiogram, echocardiogram results, and medications were collected during the index admission. Treatment with diuretics, SGLT2i, angiotensin-converting enzyme inhibitors (ACEI), angiotensin receptor blockers (ARB), angiotensin receptor-neprilysin inhibitors, β-blockers, aldosterone antagonists (MRA), diuretics, anticoagulants, and other therapeutic strategies were individualized according to established guidelines and clinical judgment [1, 2]. During the transitional phase (first weeks after discharge-visit 2), laboratory and clinical assessments were registered.
CA125 was measured at visit 1 (72 ± 24 h after admission) and visit 2 (26 days after discharge [p25% to p75%: 15–39]) using a commercially available immunoassay kit (Elecsys CA125 II assay-Roche Diagnostics). The normal range of CA125 established for the assays was 35 U/mL. Serum creatinine, blood urea nitrogen, hemoglobin, electrolytes (serum sodium and potassium), and amino-terminal pro-brain natriuretic peptide (NT-proBNP) were simultaneously measured using commercially available immunoassays at the same visits.
Post-Discharge Follow-Up
Patients’ follow-up continued until one of the following events occurred during the 6 months following visit 2. The outcomes were assessed by verifying the patients’ survival status or occurrence of hospital admissions by reviewing electronic medical records of the public healthcare system in the Valencian Community. This assessment utilized data from the SIA-GAIA and Orion Clinics electronic databases, which comprehensively record all medical interactions occurring in the public healthcare system of the Valencian Community.
Endpoints
The main endpoint was between-treatment (dapagliflozin vs. non-dapagliflozin at discharge), CA125 changes (between visits 1 and 2). As secondary objectives, we analyzed whether CA125 changes were related to 6-month adverse clinical outcomes, including all-cause readmissions, HF-related readmissions, and all-cause mortality. HF readmission was defined as an unscheduled hospital stay lasting more than 24 h during which the patient required parenteral diuretics or vasoactive drugs. Researchers in charge of endpoint ascertainment were blinded to the medical treatment at discharge and CA125 status.
Statistical Methods
Continuous variables were summarized as mean ± standard deviation for normally distributed data or median (interquartile range) for skewed data. Categorical variables were expressed as percentages. Group comparisons – between patients receiving or not receiving dapagliflozin and between patients with normal vs. elevated CA125 levels – were conducted using independent t tests or Wilcoxon rank-sum tests for continuous variables and chi-square tests for categorical variables.
Differences in CA125 were calculated by subtracting the value measured from visit 1 to visit 2. A quantile regression model [16] was applied to account for the skewed distribution of CA125 differences, with the CA125 values at visits 1 and 2 serving as the dependent variables. The model included an interaction term for dapagliflozin treatment (binary coded as 0/1) and the time (in days) between visits 1 and 2. It was adjusted for baseline characteristics, including the following covariates: age, sex, first hospitalization status, comorbid conditions, New York Heart Association (NYHA) class, systolic blood pressure, heart rate, hemoglobin, eGFR, NT-proBNP, LVEF, and medications at discharge (furosemide-equivalent dose, β-blockers, ACEI/ARB/sacubitril-valsartan, and MRA). Due to the longitudinal data, we clustered repeated CA125 measures by patient ID to account for within-patient correlation. The model outputs the median differences in CA125 between the dapagliflozin and non-dapagliflozin arms.
The study also utilized a two-equation count model (“bivcnto”), appropriate for analyzing correlated count outcomes and adjusting for informative censoring due to death [17]. This model evaluated the rates of all-cause mortality and recurrent all-cause and HF hospitalizations and incorporated dapagliflozin and CA125 differences as primary exposures and adjusted for the previously mentioned covariates, including the CA125 level at index admission. The following baseline covariates adjusted the final model: CA125 value at visit 1, age sex, NYHA class, first hospitalization status, ischemic heart disease, diabetes, atrial fibrillation, systolic blood pressure, heart rate, peripheral edema, pleural effusion, hemoglobin, eGFR, NT-proBNP, LVEF, furosemide equivalent dose at discharge, and treatment at discharge with β-blockers, ACEI/ARB/sacubitril-valsartan, and MRA. Results were reported as incidence rate ratios (IRRs) with 95% confidence intervals (CIs). All analyses were conducted at a significant level of p < 0.05. Data management and analysis were performed using STATA 17.0, with the “bivcnto” package for count regression analysis [18].
Results
Overall Baseline Characteristics
The mean age of the sample was 73.1 ± 11.1 years; 524 (54.8%) patients were men, and 179 (18.7%) received dapagliflozin at discharge. The index admission for AHF corresponded to the first admission in 74.6% of the patients, and the proportion of patients with LVEF ≤40%, 41–49%, and ≥50% were 42.4%, 14.1%, and 43.5%, respectively. The median (p25% to p75%) values of NT-proBNP and CA125 during admission were 4,629 pg/mL (2,614–9,054), and 75 U/mL (34–156), respectively. The median length of stay was 7 days (p25% to p75%: 5–11) versus 6 days (5–9), p = 0.001, for the non-SGLT2i and the dapagliflozin group. The median values of CA125 were 76.9 U/mL (35–157) versus 70.0 U/mL (32–152), p = 0.336, for the non-SGLT2i and the dapagliflozin group. The proportion of patients with CA125 >35 U/mL at baseline was similar across treatment arms (dapagliflozin: 72.6% vs. non-SGLT2i: 74.9%, p = 0.523).
Baseline Characteristics Stratified by Baseline CA125
Baseline characteristics across CA125 (≤35 U/mL vs. >35 U/mL) are presented in Table 1. Patients with CA125 >35 U/mL at baseline were, on average, younger and more likely to be men and showed a greater comorbidity burden. These patients also showed greater facts of congestion and both left and right systolic dysfunction. At discharge, they received higher diuretic doses.
Baseline characteristics across CA125 value at baseline
Variable . | CA125 ≤35 U/mL (n = 243) . | CA125 >35 U/mL (n = 713) . | p value . |
---|---|---|---|
Demographics and medical history | |||
Age, years | 76±10 | 72±11 | <0.001 |
Sex (male), n (%) | 115 (47.3) | 409 (57.4) | 0.007 |
Hypertension, n (%) | 211 (86.8) | 533 (74.8) | <0.001 |
Dislipidemia, n (%) | 140 (57.6) | 402 (56.4) | 0.738 |
Diabetes, n (%) | 113 (46.5) | 328 (46.0) | 0.893 |
Smoking history, n (%) | 56 (23.0) | 203 (28.5) | 0.100 |
First admission for AHF, n (%) | 171 (70.4) | 542 (76.0) | 0.081 |
CHD etiology, n (%) | 78 (32.1) | 227 (31.8) | 0.940 |
Charlson index, pointsa | 5.0 (4.0–6.0) | 5.0 (3.0–6.0) | 0.017 |
Last NYHA class under stable conditions prior to admission, n (%) | 0.968 | ||
I | 86 (35.4) | 246 (34.5) | |
II | 126 (51.9) | 374 (52.5) | |
III | 31 (12.8) | 93 (13.0) | |
Pleural effusion, n (%) | 81 (33.3) | 432 (60.6) | <0.001 |
Lower limbs edema, n (%) | 129 (53.1) | 519 (72.8) | <0.001 |
ECG | |||
BBB, n (%) | 92 (37.9) | 236 (33.1) | 0.177 |
Atrial fibrillation, n (%) | 108 (44.4) | 338 (47.4) | 0.424 |
Vital signs on admission | |||
Heart rate, bpm | 99±28 | 100±28 | 0.751 |
SBP, mm Hg | 150±34 | 143±29 | 0.001 |
DBP, mm Hg | 84±22 | 82±19 | 0.250 |
Echocardiography | |||
LVEF, % | 50±15 | 45±16 | <0.001 |
LAD, mm | 43±6 | 44±7 | 0.001 |
TAPSE, mma | 19.3 (17.1–21.1) | 18.0 (16.0–20.0) | <0.001 |
Laboratory data | |||
Hemoglobin, g/dL | 12.7±2.1 | 12.5±2.1 | 0.120 |
eGFR (CKD-EPI), mL/min/1.73 m2 | 59.1±23.7 | 60.4±26.3 | 0.486 |
Serum sodium, mEq/L | 139±4 | 139±4 | 0.582 |
Serum potassium, mEq/L | 4.1±0.4 | 4.1±0.5 | 0.717 |
NT-proBNP, pg/mLa | 4,060 (2,386–8,125) | 4,800 (2,713–9,595) | 0.007 |
CA125, U/mLa | 19 (14–28) | 105 (65–186) | <0.001 |
Treatment at discharge | |||
FED, mg/day | 71±38 | 84±46 | <0.001 |
Beta-blockers, n (%) | 173 (71.2) | 541 (75.9) | 0.147 |
ACEI/ARB/ARNI, n (%) | 194 (79.8) | 540 (75.7) | 0.327 |
MRA, n (%) | 76 (31.3) | 283 (39.7) | 0.150 |
Dapagliflozin, n (%) | 49 (20.2) | 130 (18.2) | 0.505 |
Variable . | CA125 ≤35 U/mL (n = 243) . | CA125 >35 U/mL (n = 713) . | p value . |
---|---|---|---|
Demographics and medical history | |||
Age, years | 76±10 | 72±11 | <0.001 |
Sex (male), n (%) | 115 (47.3) | 409 (57.4) | 0.007 |
Hypertension, n (%) | 211 (86.8) | 533 (74.8) | <0.001 |
Dislipidemia, n (%) | 140 (57.6) | 402 (56.4) | 0.738 |
Diabetes, n (%) | 113 (46.5) | 328 (46.0) | 0.893 |
Smoking history, n (%) | 56 (23.0) | 203 (28.5) | 0.100 |
First admission for AHF, n (%) | 171 (70.4) | 542 (76.0) | 0.081 |
CHD etiology, n (%) | 78 (32.1) | 227 (31.8) | 0.940 |
Charlson index, pointsa | 5.0 (4.0–6.0) | 5.0 (3.0–6.0) | 0.017 |
Last NYHA class under stable conditions prior to admission, n (%) | 0.968 | ||
I | 86 (35.4) | 246 (34.5) | |
II | 126 (51.9) | 374 (52.5) | |
III | 31 (12.8) | 93 (13.0) | |
Pleural effusion, n (%) | 81 (33.3) | 432 (60.6) | <0.001 |
Lower limbs edema, n (%) | 129 (53.1) | 519 (72.8) | <0.001 |
ECG | |||
BBB, n (%) | 92 (37.9) | 236 (33.1) | 0.177 |
Atrial fibrillation, n (%) | 108 (44.4) | 338 (47.4) | 0.424 |
Vital signs on admission | |||
Heart rate, bpm | 99±28 | 100±28 | 0.751 |
SBP, mm Hg | 150±34 | 143±29 | 0.001 |
DBP, mm Hg | 84±22 | 82±19 | 0.250 |
Echocardiography | |||
LVEF, % | 50±15 | 45±16 | <0.001 |
LAD, mm | 43±6 | 44±7 | 0.001 |
TAPSE, mma | 19.3 (17.1–21.1) | 18.0 (16.0–20.0) | <0.001 |
Laboratory data | |||
Hemoglobin, g/dL | 12.7±2.1 | 12.5±2.1 | 0.120 |
eGFR (CKD-EPI), mL/min/1.73 m2 | 59.1±23.7 | 60.4±26.3 | 0.486 |
Serum sodium, mEq/L | 139±4 | 139±4 | 0.582 |
Serum potassium, mEq/L | 4.1±0.4 | 4.1±0.5 | 0.717 |
NT-proBNP, pg/mLa | 4,060 (2,386–8,125) | 4,800 (2,713–9,595) | 0.007 |
CA125, U/mLa | 19 (14–28) | 105 (65–186) | <0.001 |
Treatment at discharge | |||
FED, mg/day | 71±38 | 84±46 | <0.001 |
Beta-blockers, n (%) | 173 (71.2) | 541 (75.9) | 0.147 |
ACEI/ARB/ARNI, n (%) | 194 (79.8) | 540 (75.7) | 0.327 |
MRA, n (%) | 76 (31.3) | 283 (39.7) | 0.150 |
Dapagliflozin, n (%) | 49 (20.2) | 130 (18.2) | 0.505 |
Values for continuous variables are expressed as mean ± standard deviation.
ACEI, angiotensin-converting enzyme inhibitors; AHF, acute heart failure; AMI, acute myocardial infarction; ARB, angiotensin receptor blockers; BBB, bundle branch block; BUN, blood urea nitrogen; CA125, carbohydrate antigen 125; CHD, coronary heart disease; eGFR, estimated glomerular filtration rate; FED, furosemide-equivalent dose; LAD, left atrial diameter; MDRD, Modification of Diet in Renal Disease; MRA, mineralcorticoid receptor antagonist; NT-proBNP, amino-terminal pro-brain natriuretic peptide; NYHA, New York Heart Association; PASP, pulmonary artery systolic pressure; TAPSE, tricuspid annular plane systolic excursion; VHD, valve heart disease.
aValues expressed as mean (interquartile range).
Baseline Characteristics Stratified by Dapagliflozin Treatment
Patients receiving dapagliflozin were more likely to be men and showed higher rates of dyslipidemia, diabetes mellitus, and higher comorbidities but lower rates of ischemic etiology. They exhibited greater parameters of congestion and lower left and right systolic function. Patients treated with dapagliflozin were more likely to be discharged on neurohormonal inhibition agents. No differences in the diuretic regimen were found. Significant variations in eGFR and NT-proBNP were also found (Table 2).
Baseline characteristics across treatment with dapagliflozin at discharge
Variable . | No dapagliflozin (n = 777) . | Dapagliflozin (n = 179) . | p value . |
---|---|---|---|
Demographics and medical history | |||
Age, years | 73±11 | 73±11 | 0.403 |
Sex (male), n (%) | 409 (52.6) | 115 (64.2) | 0.005 |
Hypertension, n (%) | 604 (77.7) | 140 (78.2) | 0.890 |
Dislipidemia, n (%) | 422 (54.3) | 120 (67.0) | 0.002 |
Diabetes, n (%) | 338 (43.5) | 103 (57.5) | 0.001 |
Smoking history, n (%) | 208 (26.8) | 51 (28.5) | 0.640 |
First admission for AHF, n (%) | 570 (73.4) | 143 (79.9) | 0.070 |
CHD etiology, n (%) | 270 (34.7) | 35 (19.6) | <0.001 |
Charlson indexa | 5.0 (3.0–6.0) | 5.0 (4.0–7.0) | 0.024 |
Last NYHA class under stable conditions prior to admission, n (%) | 0.001 | ||
I | 266 (34.2) | 66 (36.9) | |
II | 395 (50.8) | 105 (58.7) | |
III | 116 (14.9) | 8 (4.5) | |
Pleural effusion, n (%) | 400 (51.5) | 113 (63.1) | 0.005 |
Lower limbs edema, n (%) | 514 (66.2) | 134 (74.9) | 0.025 |
ECG | |||
BBB, n (%) | 270 (34.7) | 58 (32.4) | 0.551 |
Atrial fibrillation, n (%) | 365 (47.0) | 81 (45.3) | 0.677 |
Vital signs on admission | |||
Heart rate, bpm | 99±28 | 102±28 | 0.231 |
SBP, mm Hg | 145±32 | 141±26 | 0.113 |
DBP, mm Hg | 82±19 | 85±21 | 0.070 |
Echocardiography | |||
LVEF, % | 47±16 | 42±15 | <0.001 |
LAD, mm | 44±7 | 44±5 | 0.898 |
TAPSE, mma | 18 (16–20) | 17 (15–20) | 0.007 |
Laboratory data | |||
Hemoglobin, g/dL | 12.5±2.0 | 12.8±2.3 | 0.056 |
Creatinine at admission, mg/dL | 1.35±0.64 | 1.24±0.55 | 0.031 |
eGFR (CKD-EPI), mL/min/1.73 m2 | 58.8±25.3 | 65.6±26.6 | 0.001 |
Serum sodium, mEq/L | 139±4 | 139±4 | 0.844 |
Serum potassium, mEq/L | 4.1±0.4 | 4.1±0.5 | 0.091 |
NT-proBNP, pg/mLa | 4,774 (2,788–9,315) | 3,719 (1,826–7,019) | 0.001 |
CA125, U/mLa | 77 (35–157) | 70 (32–152) | 0.322 |
CA125 >35 U/mL at baseline, n (%) | 583 (75.0) | 130 (72.6) | 0.505 |
Treatment at discharge | |||
Diuretics, n (%) | 764 (98.3) | 174 (97.2) | 0.320 |
FED, mg/day | 81±44 | 81±47 | 0.966 |
Beta-blockers, n (%) | 560 (72.1) | 154 (86.0) | <0.001 |
ACEI/ARB/ARNI, N (%) | 521 | 135 | 0.252 |
MRA, n (%) | 255 (32.8) | 104 (58.1) | 0.010 |
Variable . | No dapagliflozin (n = 777) . | Dapagliflozin (n = 179) . | p value . |
---|---|---|---|
Demographics and medical history | |||
Age, years | 73±11 | 73±11 | 0.403 |
Sex (male), n (%) | 409 (52.6) | 115 (64.2) | 0.005 |
Hypertension, n (%) | 604 (77.7) | 140 (78.2) | 0.890 |
Dislipidemia, n (%) | 422 (54.3) | 120 (67.0) | 0.002 |
Diabetes, n (%) | 338 (43.5) | 103 (57.5) | 0.001 |
Smoking history, n (%) | 208 (26.8) | 51 (28.5) | 0.640 |
First admission for AHF, n (%) | 570 (73.4) | 143 (79.9) | 0.070 |
CHD etiology, n (%) | 270 (34.7) | 35 (19.6) | <0.001 |
Charlson indexa | 5.0 (3.0–6.0) | 5.0 (4.0–7.0) | 0.024 |
Last NYHA class under stable conditions prior to admission, n (%) | 0.001 | ||
I | 266 (34.2) | 66 (36.9) | |
II | 395 (50.8) | 105 (58.7) | |
III | 116 (14.9) | 8 (4.5) | |
Pleural effusion, n (%) | 400 (51.5) | 113 (63.1) | 0.005 |
Lower limbs edema, n (%) | 514 (66.2) | 134 (74.9) | 0.025 |
ECG | |||
BBB, n (%) | 270 (34.7) | 58 (32.4) | 0.551 |
Atrial fibrillation, n (%) | 365 (47.0) | 81 (45.3) | 0.677 |
Vital signs on admission | |||
Heart rate, bpm | 99±28 | 102±28 | 0.231 |
SBP, mm Hg | 145±32 | 141±26 | 0.113 |
DBP, mm Hg | 82±19 | 85±21 | 0.070 |
Echocardiography | |||
LVEF, % | 47±16 | 42±15 | <0.001 |
LAD, mm | 44±7 | 44±5 | 0.898 |
TAPSE, mma | 18 (16–20) | 17 (15–20) | 0.007 |
Laboratory data | |||
Hemoglobin, g/dL | 12.5±2.0 | 12.8±2.3 | 0.056 |
Creatinine at admission, mg/dL | 1.35±0.64 | 1.24±0.55 | 0.031 |
eGFR (CKD-EPI), mL/min/1.73 m2 | 58.8±25.3 | 65.6±26.6 | 0.001 |
Serum sodium, mEq/L | 139±4 | 139±4 | 0.844 |
Serum potassium, mEq/L | 4.1±0.4 | 4.1±0.5 | 0.091 |
NT-proBNP, pg/mLa | 4,774 (2,788–9,315) | 3,719 (1,826–7,019) | 0.001 |
CA125, U/mLa | 77 (35–157) | 70 (32–152) | 0.322 |
CA125 >35 U/mL at baseline, n (%) | 583 (75.0) | 130 (72.6) | 0.505 |
Treatment at discharge | |||
Diuretics, n (%) | 764 (98.3) | 174 (97.2) | 0.320 |
FED, mg/day | 81±44 | 81±47 | 0.966 |
Beta-blockers, n (%) | 560 (72.1) | 154 (86.0) | <0.001 |
ACEI/ARB/ARNI, N (%) | 521 | 135 | 0.252 |
MRA, n (%) | 255 (32.8) | 104 (58.1) | 0.010 |
Values for continuous variables are expressed as mean ± standard deviation.
ACEI, angiotensin-converting enzyme inhibitors; AHF, acute heart failure; AMI, acute myocardial infarction; ARB, angiotensin receptor blockers; BBB, bundle branch block; BUN, blood urea nitrogen; CA125, carbohydrate antigen 125; CHD, coronary heart disease; eGFR, estimated glomerular filtration rate; FED, furosemide-equivalent dose; LAD, left atrial diameter; MDRD, modification of diet in Renal Disease; MRA, mineralcorticoid receptor antagonist; NT-proBNP, amino-terminal pro-brain natriuretic peptide; NYHA, New York Heart Association; PASP, pulmonary artery systolic pressure; TAPSE, tricuspid annular plane systolic excursion; VHD, valve heart disease; ARNI, angiotensin receptor-neprilysin inhibitors.
aValues expressed as mean (interquartile range).
CA125 Changes at the First Ambulatory Visit
The first ambulatory visit after discharge occurred at a median of 26 days after discharge (p25% to p75%: 15–39) with no differences across treatment (non-dapagliflozin: 27 days [16–36] vs. dapagliflozin: 23 days [21–26], p = 0.120). In the overall cohort, the median CA125 at visit 2 was 41.3 U/mL (19.0–96.0), with an estimated difference between visits 1 and 2 of −15.3 U/mL (−60 to −1.0) (p = 0.035). At visits 1 and 2, the frequency of patients with CA125 >35 U/mL was 713 (74.5%) and 524 (54.8%), respectively.
CA125 Changes across Dapagliflozin Treatment
Patients receiving dapagliflozin experienced a higher reduction in CA125 values at visit 2. The medians of CA125 difference between the treatment groups were −14 U/mL (−60 to 3) and −24 U/mL (−61 to −2) for those not receiving and receiving dapagliflozin, respectively (p = 0.034). Similar significant reductions were found regarding relative reductions from baseline (−26.5% vs. −41.2%, p = 0.007) and CA125 logarithm (−0.31 vs. −0.52, p = 0.008) for those not receiving and receiving dapagliflozin, respectively. Likewise, the proportion of patients with CA125 ≤35 U/mL at visit 2 was higher in those on dapagliflozin versus non-dapagliflozin (103 [57.5%] vs. 330 [42.4%], p = 0.001) (shown in Fig. 1).
Proportion of patients with CA125 values >35 U/mL at visits 1 and 2 in those receiving and not receiving SGLT2i. CA125, carbohydrate antigen 125; SGLT2i, sodium-glucose cotransporter 2 inhibitors.
Proportion of patients with CA125 values >35 U/mL at visits 1 and 2 in those receiving and not receiving SGLT2i. CA125, carbohydrate antigen 125; SGLT2i, sodium-glucose cotransporter 2 inhibitors.
Multivariable quantile regression analyses showed that dapagliflozin treatment was associated with a greater reduction in CA125 (shown in Fig. 2). Indeed, the predicted difference at visit 2 in the median of CA125 for the control and dapagliflozin groups were 52.9 U/mL and 24.6 U/mL, respectively (p = 0.004).
Predicted difference in the median of CA125 from visit 1 to visit 2 in those receiving and not receiving dapagliflozin. CA125, carbohydrate antigen 125.
Predicted difference in the median of CA125 from visit 1 to visit 2 in those receiving and not receiving dapagliflozin. CA125, carbohydrate antigen 125.
CA125 Changes and Adverse Clinical Endpoints
At 6-month follow-up after visit 2, a total of 90 patients (9.4%) died, resulting in a crude mortality incidence rate of 0.246 person-years (P-Y). We also registered 292 and 152 all-cause and HF-related readmission events, respectively. The estimated crude incidence rates of all-cause and HF readmissions were 1.09 P-Y (95% CI: 0.98–1.20) and 0.51 P-Y (95% CI: 0.44–0.59), respectively. The distribution of HF readmissions among patients was 128 (13.4%) had 1 admission, 18 (1.9%) had 2 admissions, 5 (0.5%) had 3 admissions, and 1 (0.1%) had 4 readmissions.
CA125 Reduction and 6-Month Adverse Clinical Outcomes across Treatment Allocation
Rates of 6-month mortality were 83 (10.7%) and 7 (3.9%) (p = 0.005) for those not receiving and receiving dapagliflozin, respectively. Likewise, we found a significant reduction in the rates of 6-month all-cause and HF readmissions in patients discharged with dapagliflozin (all-cause readmission: no dapagliflozin 1.20 P-Y [95% CI: 1.08–1.34] and dapagliflozin 0.60 P-Y [95% CI: 0.44–0.82], p < 0.001, and 2) HF readmissions: no dapagliflozin 0.59 P-Y [95% CI: 0.51–0.69] and dapagliflozin 0.18 P-Y [95% CI: 0.10–0.31], p < 0.001).
In the whole sample, CA125 reduction at visit 2 identified those with a better risk profile (reduction across endpoints) as depicted in Figure 3a (mortality), Figure 3b (any readmissions), and Figure 3c (HF readmissions). The effect of 6-month driven CA125 difference (per decrease in 10 U/mL) on clinical endpoints was borderline associated with reduction in all-cause mortality: IRR = 0.98 P-Y, 95% CI = 0.96–1.00; p = 0.050, and a statistically significant reduction in all-cause readmission (IRR = 0.99, 95% CI = 0.98–0.99; p = 0.003) and HF readmission (IRR = 0.98, 95% CI = 0.97–0.99; p < 0.001). We did not find statistical evidence of heterogeneity across treatment arms for any of the endpoints (death [p value for interaction = 0.920], any readmissions [p value for interaction = 0.366], and HF readmissions [p value for interaction = 0.325]).
Association of CA125 changes with 6-month clinical events. a All-cause mortality. b All-cause rehospitalization. c HF-rehospitalization. CA125, carbohydrate antigen 125; HF, heart failure.
Association of CA125 changes with 6-month clinical events. a All-cause mortality. b All-cause rehospitalization. c HF-rehospitalization. CA125, carbohydrate antigen 125; HF, heart failure.
Discussion
In this all-corner cohort of patients with AHF, treatment with dapagliflozin was associated with a significant decrease in CA125 values the first weeks after discharge, suggesting this agent provides an additional decongestive and anti-inflammatory effect. Likewise, the greater magnitude of CA125 reduction during the first weeks after discharge identified patients at lower risk of 6-month death and all-cause and HF readmissions. These findings confirm prior observations reported in ambulatory stable HF patients.
Mechanisms behind the Benefits of SGLT2i in HF
The evolving landscape of HF treatment has been significantly reshaped by the introduction of SGLT2i, including dapagliflozin [1]. These agents, initially developed for type 2 diabetes management, have demonstrated remarkable efficacy in improving clinical outcomes in patients with chronic HF, regardless of diabetes status and LVEF [2‒4]. The DAPA-HF and DELIVER trials showed that dapagliflozin significantly reduced the risk of cardiovascular death and HF hospitalizations [2, 4]. In AHF, the evidence is less consistent but also indicates that SGLT2i led to a greater decongestion, less requirement of loop diuretics, and improved surrogates of decongestion [19‒21].
The mechanisms behind the beneficial effect of SGLT2i in HF seem multifactorial but not fully understood. In AHF, the diuretic effect of SGLT2 inhibitors is primarily osmotic diuresis, resulting from the inhibition of sodium-glucose co-transporters in the renal tubules, leading to increased excretion of glucose and sodium [22], has been proposed as a crucial mechanism explaining the benefits of these drugs in decompensated HF [23, 24]. Conversely to traditional diuretics, some investigations suggest these drugs lead to hemoconcentration [25, 26] and increased osmolarity, favoring plasma refill, fluid redistribution and tissue decongestion rather than a reduction in total body water [23, 24, 26].
SGLT2i has also shown anti-inflammatory properties that may mediate the beneficial effects in situations with a greater inflammatory milieu, such as AHF [27]. Indeed, a systematic review and meta-analysis of randomized controlled trials have provided substantial evidence that SGLT2i, including dapagliflozin, significantly reduces markers of inflammation such as C-reactive protein [28].
Interpreting the Impact of Dapagliflozin on CA125 Levels and Clinical Outcomes
In the AHF setting, current evidence endorsed the role of CA125 as a surrogate marker of the severity of fluid overload (especially extravascular data of congestion) and immunoinflammatory activity in AHF [9, 10]. This relationship is bidirectional, as increased inflammatory activity enhances vascular permeability, thus facilitating interstitial fluid overload [5, 30]. Our study builds upon this understanding by showing that dapagliflozin significantly reduces CA125 levels, suggesting its potential role in ameliorating inflammation and fluid overload in the first weeks following an episode of AHF patients. Consistent with these observations, we hypothesize that changes in CA125 over weeks could serve as a valuable indicator for assessing the decongestive and anti-inflammatory effects of SGLT2i. It is important to note that CA125 has a relatively long half-life (7–12 days), and then its changes are sustained but require more time than other surrogates of decongestion, such as natriuretic peptides.
Interestingly, after SGLT2i initiation, NT-proBNP changes are not meaningful and do not seem to be an accurate parameter for evaluating clinical response [31]. In contrast, some studies suggest that the short-term increase in hemoglobin and hematocrit predicts clinical response to SGLT2i [26, 32]. Thus, we envision changes in hemoglobin/hematocrit and CA125 as complementary tools for predicting the short- and mid-term clinical response to SGLT2i. A decrease of CA125 following SGLT2i has also been reported in ambulatory HF patients, suggesting also the extrapolation of these postulates to chronic HF [14, 15].
Limitations and Future Directions
To our knowledge, this is the first study correlating treatment with dapagliflozin with a significant plasmatic CA125 reduction following hospitalization for AHF. These findings support the role of CA125 in monitoring dapagliflozin treatment response. However, important limitations need to be noted. First, our findings stem from a retrospective observational study, which inherently carries limitations regarding causality. Second, several unmeasured confounders may be playing a role. In this regard, compliance with dapagliflozin or diuretic doses during hospitalization or changes occurred after discharge were not assessed. Third, we evaluated patients discharged after admission for AHF; thus, we cannot extrapolate these findings to other forms of HF or clinical scenarios, such as patients with AHF not requiring admission, or patients with cardiogenic shock or requiring mechanical circulatory support. Additionally, these findings cannot be extrapolated to other SGLT2i other than dapagliflozin. Lastly, we did not routinely measure inflammatory markers, such as C-reactive protein, which would have allowed us to evaluate the concordance between CA125 changes and other proxies of inflammation. Prospective studies are needed to confirm these observations and elucidate the underlying pathophysiological mechanisms behind these findings.
Conclusions
The administration of dapagliflozin upon discharge after an AHF episode was linked to a significant decrease in the levels of CA125 during the initial weeks following discharge. This reduction in CA125 was a marker for identifying patients at a reduced risk of experiencing adverse clinical outcomes within the next 6 months. The implications of this finding suggest that monitoring CA125 levels could be a valuable strategy in assessing the effectiveness of dapagliflozin treatment in this scenario.
Statement of Ethics
This study protocol was reviewed and approved by the Ethics Committee of INCLIVA (Instituto de Investigación Sanitaria del Hospital Clínico Universitario de Valencia), Valencia, Spain, Approval No. 2022/281. As this is a retrospective observational study, it has been granted an exemption from requiring written informed consent (Ethical Committee of INCLIVA, Instituto de Investigación Sanitaria, Valencia, Spain).
Conflict of Interest Statement
Gema Miñana reports personal fees from Abbott Vascular and Prosmédica. Rafael de la Espriella reports personal fees from AstraZeneca, Novartis, Boehringer-Ingelheim, and NovoNordisk (outside the submitted work). Patricia Palau reports personal fees from Servier. Antoni Bayés-Genís has lectured and/or participated in Abbott, AstraZeneca, Bayer, Boehringer-Ingelheim, Novartis, Roche Diagnostics, and Vifor advisory boards (outside the submitted work). Juan Sanchis reports personal fees from Abbott Vascular and Prosmédica. Julio Nuñez reports personal fees or advisory boards from Alleviant, AstraZeneca, Boehringer Ingelheim, Bayer, Novartis, NovoNordisk, Rovi, and Vifor Pharma (outside the submitted work). The rest of the authors have no disclosures. There are no disclosures related to the content of this article.
Funding Sources
This work has been funded by a non-conditional grant from AstraZeneca (ESR-22-22099) and CIBER Cardiovascular (Grant No. 16/11/00403 and 16/11/00420). The founders had no role in the investigation nor in the elaboration of the manuscript.
Author Contributions
Gema Miñana and Rafael de la Espriella: conceptualization, data curation, investigation, methodology, project administration, validation, visualization, writing – original draft, and writing – review and editing. Miguel Lorenzo-Hernández, Enrique Rodriguez-Borja, Anna Mollar, Patricia Palau, Agustin Fernández-Cisnal, Ernesto Valero, Arturo Carratalá, Enrique Santas, Vicent Bodi, Juan Sanchis, Eduardo Nuñez, and Antoni Bayés-Genís: formal analysis, investigation, methodology, project administration, resources, software, supervision, validation, and writing – review and editing. Julio Nuñez: conceptualization, formal analysis, funding acquisition, investigation, methodology, project administration, resources, software, supervision, validation, visualization, writing – original draft, and writing – review and editing.
Additional Information
Gema Miñana and Rafael de la Espriella contributed equally to this work.
Data Availability Statement
All data generated or analyzed during this study are included in this article. Further inquiries can be directed to the corresponding author.