From January 2015 to January 2023, a total of 2661 consecutive STEMI patients admitted, among which 226 patients had incomplete data or did not receive coronary angiography, the remaining 2435 patients were included in this study. The mean age of this cohort was 63 years (SD, 13) years and 1947 (80.0%) were male. During hospitalization, a total of 90 (3.70%) patients died and 110 (4.52%) MACEs occurred.
Compared with survivors, non-survivors had a higher level of BUN (9.0 ± 4.9 vs. 6.2 ± 2.7 mmol/L, p < 0.001), lower LVEF (45 ± 11% vs. 54 ± 8%, p < 0.001) and higher BLR (12.46 (17.03, 22.57) vs. 10.68 (8.49, 13.61), p < 0.001). The ROC of BUN, LVEF, creatinine, NT-proBNP, TnI, GRACE score, TIMI score, and BLR for predicting in-hospital mortality and MACE were presented in Fig. 1 and Fig. 2. The AUC of BUN, LVEF, creatinine, NT-proBNP, TnI, GRACE score, TIMI score and BLR for in-hospital mortality were 0.714, 0.741, 0.682, 0.712, 0.630, 0.784, 0.770, and 0.789, respectively; for MACE AUCs were 0.705, 0.698, 0.668, 0.704, 0.630, 0.771, 0.763, and 0.762, respectively, indicating BLR had highest prognostic power for in-hospital mortality and MACE. DeLong test showed that the predictive value of BLR was significantly higher than BUN, LVEF, creatinine, NT-proBNP, and TnI (all p < 0.05), but was comparable to GRACE score and TIMI scores (all p > 0.05). The optimal cut-off value of BLR for predicting in-hospital all-cause mortality was 12.54 determined by Yonden index with a sensitivity of 75.6% and a specificity of 67.6%. Then patients were divided into two groups, BLR < 12.54 and BLR ≥ 12.54 for further analysis.
Table 1 showed the baseline characteristics and treatment of the 2 groups divided according to BLR cut-off value. Patients with higher BLR tended to be older (67 ± 12 vs. 61 ± 12 yeas, p < 0.001), had lower body mass index (23.74 ± 3.43 vs. 24.54 ± 3.43 kg/m2, p < 0.001), and had a higher proportion of previous MI (6.1% vs. 4.0%, p = 0.027), PCI (5.4% vs. 3.1%, p = 0.005), hypertension (59.2% vs. 49.3%, p < 0.001), diabetes (33.5% vs. 23.3%, p < 0.001), renal dysfunction (32.6% vs. 6.2%, p < 0.001), and previous stroke (6.8% vs. 4.4%, p = 0.010). At admission, more patients in higher BLR group presented with higher Killip class, lower blood pressure, higher heart rate, and cardiac shock (all p < 0.05).
Table 1
Comparison of baseline characteristics divided by BLR
| Total (n = 2435) | BLR ≥ 12.54 (n = 826) | BLR < 12.54 (n = 1609) | P value |
Demographics |
Age (years) | 63 ± 13 | 67 ± 12 | 61 ± 12 | < 0.001 |
Male (n, %) | 1947 (80.0%) | 650 (78.7%) | 1297 (80.6%) | 0.263 |
BMI (Kg/m2) | 24.27 ± 3.45 | 23.74 ± 3.43 | 24.54 ± 3.43 | < 0.001 |
Co-morbidities (n, %) |
Previous MI | 115 (4.7%) | 50 (6.1%) | 65 (4.0%) | 0.027 |
Previous PCI | 95 (3.9%) | 45 (5.4%) | 50 (3.1%) | 0.005 |
Hypertension | 1282 (52.6%) | 489 (59.2%) | 793 (49.3%) | < 0.001 |
Diabetes mellitus | 652 (26.8%) | 277 (33.5%) | 375 (23.3%) | < 0.001 |
Dislipidemia | 370 (15.2%) | 116 (14.0%) | 254 (15.8%) | 0.257 |
Renal dysfunctiona | 368 (15.1%) | 269 (32.6%) | 99 (6.2%) | < 0.001 |
Previous stroke | 126 (5.2%) | 56 (6.8%) | 70 (4.4%) | 0.010 |
Current smoker | 1613 (66.2%) | 529 (64.0%) | 1084 (67.4%) | 0.100 |
Clinical presentation (n, %) |
Cardiac arrest before admission | 108 (4.4%) | 38 (4.6%) | 70 (4.4%) | 0.777 |
Killip class | | | | < 0.001 |
I | 1841 (75.6%) | 534 (64.6%) | 1307 (81.2%) | |
II | 278 (11.4%) | 129 (15.6%) | 149 (9.3%) | |
III | 55 (2.3%) | 34 (4.1%) | 21 (1.3%) | |
IV | 261 (10.7%) | 129 (15.6%) | 132 (8.2%) | |
Cardiogenic shock | 196 (8.0%) | 144 (13.8%) | 82 (5.1%) | < 0.001 |
Admission vital signs |
Systolic blood pressure (mmHg) | 125 ± 25 | 122 ± 26 | 127 ± 25 | < 0.001 |
Diastolic blood pressure (mmHg) | 78 ± 17 | 76 ± 17 | 79 ± 16 | < 0.001 |
Heart rate (bpm) | 82 ± 18 | 83 ± 20 | 82 ± 17 | 0.031 |
Location of MI (n, %) |
Anterior MI | 1281 (52.6%) | 428 (51.3%) | 853 (53.0%) | 0.575 |
Lateral MI | 267 (11.0%) | 104 (12.6%) | 161 (10.1%) | 0.066 |
Inferior MI | 1200 (49.3%) | 419 (50.7%) | 781 (48.5%) | 0.307 |
Right ventricle MI | 262 (10.8%) | 95 (11.5%) | 167 (10.4%) | 0.398 |
Posterior MI | 324 (13.3%) | 119 (14.4%) | 205 (12.7%) | 0.252 |
Culprit vessel (n, %) |
LM | 12 (0.5%) | 9 (1.1%) | 3 (0.2%) | 0.003 |
LAD | 1265 (52.0%) | 417 (50.5%) | 848 (52.7%) | 0.292 |
LCX | 278 (11.4%) | 93 (11.3%) | 185 (11.5%) | 0.857 |
RCA | 940 (38.6%) | 329 (39.8%) | 611 (38.0%) | 0.379 |
Laboratory findings |
Troponin I (ng/mL) | 3.12 (0.30, 14.60) | 4.24 (0.50, 18.10) | 2.58 (0.25, 12.90) | < 0.001 |
NT-proBNP (pg/mL) | 137 (39, 505) | 270 (79, 1155) | 98 (31, 320) | < 0.001 |
D-dimer (ng/mL) | 285 (100, 659) | 485 (189, 1020) | 217 (100, 487) | < 0.001 |
White blood cell counts (× 109 /L) | 11.31 ± 3.87 | 11.68 ± 4.24 | 11.11 ± 3.65 | < 0.001 |
Hemoglobin (g/L) | 138 ± 19 | 134 ± 20 | 141 ± 19 | < 0.001 |
Blood urea nitrogen (mmol/L) | 6.3 ± 2.9 | 8.7 ± 3.8 | 5.1 ± 1.1 | < 0.001 |
Creatinine (µmol/L) | 75 (63, 90) | 87 (71, 115) | 71 (60, 82) | < 0.001 |
LDL (mmol/L) | 2.84 ± 0.93 | 2.72 ± 0.92 | 2.91 ± 0.93 | < 0.001 |
HDL (mmol/L) | 1.09 ± 0.31 | 1.12 ± 0.33 | 1.07 ± 0.29 | < 0.001 |
Echocardiography findings |
LVEF (%) | 54 ± 8 | 49 ± 9 | 56 ± 6 | < 0.001 |
LVEDD (mm) | 49 ± 5 | 50 ± 6 | 48 ± 4 | < 0.001 |
Regional wall motion abnormality (n, %) | 2092 (88.7%) | 728 (91.4%) | 1364 (87.4%) | 0.005 |
Ventricular aneurysm (n, %) | 105 (4.4%) | 63 (7.9%) | 42 (2.7%) | < 0.001 |
Medication use in hospital (n, %) |
Aspirin | 2358 (96.8%) | 794 (96.1%) | 1564 (97.2%) | 0.150 |
P2Y12 receptor inhibitors | 2404 (98.7%) | 812 (98.3%) | 1592 (98.9%) | 0.183 |
Statins | 2398 (98.5%) | 803 (97.2%) | 1595 (99.1%) | < 0.001 |
β blockers | 1323 (54.3%) | 380 (46.0%) | 943 (58.6%) | < 0.001 |
ACEI/ARB | 815 (33.5%) | 218 (26.4%) | 597 (37.1%) | < 0.001 |
Anticoagulant drug | 101 (4.1%) | 49 (5.9%) | 52 (3.2%) | 0.002 |
PPI | 2159 (88.7%) | 754 (91.3%) | 1405 (87.3%) | 0.004 |
Percutaneous coronary intervention | 2292 (94.1%) | 773 (93.6%) | 1519 (94.4%) | 0.414 |
TIMI score | 4.84 ± 2.52 | 5.73 ± 2.51 | 4.38 ± 2.40 | < 0.001 |
GEACE score | 128 (106, 155) | 145 (120, 171) | 120 (101, 145) | < 0.001 |
BLR: Blood urea nitrogen to left ventricular ejection fraction ratio, BMI: body mass index, MI: myocardial infarction, PCI: percutaneous coronary intervention, LM: left main artery, LAD: left anterior descending artery, LCX: left circumfex artery, RCA: right coronary artery, NT-proBNP: N-terminal-pro-B-type-natriuretic-peptide, LDL: low-density lipoprotein, HDL: high-density lipoprotein, LVEF: left ventricular ejection fraction, LVEDD: left ventricular end-diastolic dimension, ACEI: angiotensin-converting enzyme inhibitor, ARB: angiotensin-converting receptor blocker, PPI: proton pump inhibitor). |
a:Renal dysfunction: estimate glomerular filtration rate< 60mL/(min*1.73^2) |
The location of MI on electrocardiogram was similar between the two groups (all p > 0.05). On coronary angiography, the distribution of culprit vessels were similar except that left main coronary artery was relatively more common in patients with higher BLR (1.1% vs. 0.2%, p = 0.003). The comparison of laboratory findings revealed that patients with higher BLR had a higher level of TnI, NT-proBNP, D-dimer, white blood cell counts, BUN, creatinine, and high-density lipoprotein, but had a lower level of hemoglobin and low-density lipoprotein (all p < 0.05). The echocardiography showed that patients in higher BLR group had lower LVEF (49 ± 9% vs. 56 ± 6%, p < 0.001), larger left ventricular end-diastolic dimension (50 ± 6 vs. 48 ± 4mm, p < 0.001), and more regional wall motion abnormality (91.4% vs. 87.4%, p = 0.005) ventricular aneurysm (7.9% vs. 2.7%, p < 0.001). During hospitalization, statins, β blockers, angiotensin-converting enzyme inhibitor/angiotensin-converting receptor blocker were more prescribed to patients with lower BLR, while proton pump inhibitor and anticoagulant drugs were more used in patients with higher BLR (all p < 0.05). The percentage of PCI was comparable between the two groups (p > 0.05). The mean TIMI score and GRACE score in patients with higher BLR was 5.73 and 145, respectively, and were significantly higher than in patients with lower BLR (all p < 0.001).
Figure 3 showed the in-hospital outcomes between patients with high and low BLR. The in-hospital mortality and MACE were significantly higher in BLR ≥ 12.54 group (8.23% vs. 1.37% for in-hospital mortality, 9.44% vs. 1.99% for MACE, all p < 0.001). The cardiovascular mortality was significantly higher in high BLR group (8.11% vs. 1.31%, p < 0.001), while the nonfatal stroke and nonfatal MI were similar between the two groups (1.21% vs. 0.81%, p = 0.331 for nonfatal stroke and 0.36% vs. 0.06%, p = 0.227 for nonfatal MI, respectively). Figure 4 displays the K-M curves of the two group patients and it revealed the cumulative survival and free of MACE in patients with higher BLR were significantly lower than that in patients with lower BLR (all Log rank p < 0.001).
Table 2 displayed the results from univirate and multivariate Cox regression for in-hospital mortality. As a continuous variable, BLR was positively associated increased risk of in-hospital all-cause mortality (HR = 1.061, 95%CI 1.047, 1.074, p < 0.001). As as a category variable and compared with BLR < 12.54, BLR ≥ 12.54 was associated with almost 5-fold increased risk of in-hospital all-cause mortality (HR = 4.773, 95%CI 2.935, 7.762, p < 0.001). After multivariate adjustment, BLR ≥ 12.54 was still an independent prognostic factor for in-hospital mortality (HR = 1.948, 95%CI 1.143, 3.318, p = 0.014). Other independent prognostic factors included heart rate (HR = 1.016, 95%CI 1.005, 1.026, p = 0.003), lactate (HR = 1.246, 95%CI 1.184, 1.312, p < 0.001), NT-proBNP > 265.5 pg/mL (HR = 2.139, 95%CI 1.284, 3.564, p = 0.004), and TNI > 4.50 ng/mL (HR = 1.733, 95%CI 1.079, 2.784, p = 0.023).
Table 2
The univariate and multivariate Cox regression analysis of in-hospital mortality
Predictors for in-hospital mortality | Univariate analysis HR (95%CI) P | Multivariate analysis HR (95%CI) P |
Age | 1.051 (1.031, 1.071) | < 0.001 | | |
Male | 1.830 (1.183, 2.831) | 0.007 | | |
Admission heart rate | 1.026 (1.017, 1.035) | < 0.001 | 1.016 (1.005, 1.026) | 0.003 |
Admission SBP | 0.985 (0.977, 0.992) | < 0.001 | | |
Lactate | 1.320 (1.266, 1.377) | < 0.001 | 1.246 (1.184, 1.312) | < 0.001 |
NT-proBNP > 265.5 | 3.361 (2.131, 5.303) | < 0.001 | 2.139 (1.284, 3.564) | 0.004 |
Troponin I > 4.50 | 2.112 (1.371, 3.254) | 0.001 | 1.733 (1.079, 2.784) | 0.023 |
Killip Class > I | 3.502 (2.286, 5.365) | < 0.001 | | |
GRACE score | 1.019 (1.015, 1.023) | < 0.001 | | |
TIMI score | 1.375 (1.271, 1.488) | < 0.001 | | |
BLR (≥ 12.54) | 4.773 (2.935, 7.762) | < 0.001 | 1.948 (1.143, 3.318) | 0.014 |
SBP: Systolic blood pressure; NT-proBNP: N-terminal-pro-B-type-natriuretic-peptide, BLR: blood urea nitrogen to left ventricular ejection fraction ratio. |
Table 3 showed the association of BLR with in-hospital MACE. Also, as a continuous variable, BLR was positively associated increased risk of in-hospital MACE (HR = 1.061, 95%CI 1.048, 1.073, p < 0.001). Compared with BLR < 12.54, BLR ≥ 12.54 was associated with almost 4-flold increased risk of in-hospital MACE (HR = 3.866, 95%CI 2.548, 5.865, p < 0.001). Similarly, BLR ≥ 12.54 was independently associated with increased risk of in-hospital MACE after multivariate adjustment (HR = 1.720, 95%CI 1.066, 2.774, p = 0.026). Other independent factors associated with in-hospital MACE included age (HR = 1.038, 95%CI 1.008, 1.068, p = 0.013), heart rate (HR = 1.019, 95%CI 1.009, 1.029, p < 0.001), lactate (HR = 1.225, 95%CI 1.166, 1.286, p < 0.001), NT-proBNP > 265.5 pg/mL (HR = 2.030, 95%CI 1.273, 3.239, p = 0.003), and TNI > 4.50 ng/mL (HR = 1.832, 95%CI 1.180, 2.844, p = 0.007).
Table 3
The univariate and multivariate Cox regression analysis of in-hospital MACE
Predictors for in-hospital mortality | Univariate analysis HR (95%CI) P | Multivariate analysis HR (95%CI) P |
Age | 1.054 (1.036, 1.072) | < 0.001 | 1.038 (1.008, 1.068) | 0.013 |
Male | 1.590 (1.061, 2.384) | 0.025 | | |
Admission heart rate | 1.026 (1.018, 1.035) | < 0.001 | 1.019 (1.009, 1.029) | < 0.001 |
Admission SBP | 0.987 (0.980, 0.995) | 0.001 | | |
Lactate | 1.304 (1.253, 1.358) | < 0.001 | 1.225 (1.166, 1.286) | < 0.001 |
NT-proBNP > 265.5 | 3.335 (2.217, 5.017) | < 0.001 | 2.030 (1.273, 3.239) | 0.003 |
Troponin I > 4.5 | 2.175 (1.471, 3.217) | < 0.001 | 1.832 (1.180, 2.844) | 0.007 |
Killip Class > I | 3.386 (2.307, 4.969) | < 0.001 | | |
GRACE score | 1.018 (1.014, 1.022) | < 0.001 | | |
TIMI score | 1.360 (1.267, 1.459) | < 0.001 | | |
BLR (≥ 12.54) | 3.866 (2.548, 5.865) | < 0.001 | 1.720 (1.066, 2.774) | 0.026 |
SBP: Systolic blood pressure; NT-proBNP: N-terminal-pro-B-type-natriuretic-peptide, BLR: blood urea nitrogen to left ventricular ejection fraction ratio. |
For a more detailed analysis of the association between BLR and in-hospital all-cause mortality and MACE in patients with different clinical profiling, subgroups were analyzed (Fig. 5). It revealed the effect of BLR on different subgroups were consistent except in anterior vs. non-anterior MI with more significant in patients with anterior MI compared with non-anterior MI (for in-hospital mortality, HR = 3.181, 95%CI 1.486, 6.807, p-interaction = 0.018; for MACE, HR = 2.507, 95%CI 1.300, 4.833, p-interaction = 0.027, respectively).