Improvement of long-term clinical outcomes by successful PCI in the very elderly women with ACS

Background Whether very elderly women with acute coronary syndromes (ACS) should receive aggressive percutaneous coronary intervention (PCI) is still controversial. We assessed the effectiveness and long-term clinical outcomes of successful PCI in this population and identified prognostic factors which might contribute to the incidence of major adverse cardiovascular and cerebrovascular events (MACCE) in the very elderly female PCI cohort. Methods Female ACS patients aged ≥ 80 years were consecutively enrolled (n = 729) into the study. All the patients were divided into female PCI group (n = 232) and medical group (n = 497). MACCE was followed up, including non-fatal myocardial infarction (MI), stroke, heart failure requiring hospitalization (HFRH), cardiovascular (CV) death, and the composite of them. After propensity score matching (1:1), the incidences of MACCE were compared between the two groups. Clinical and coronary artery lesion characteristics were compared between the female PCI patients with (n = 56) and without MACCE (n = 176). Multivariate Cox regression analysis was performed to identify risk factors which independently associated with MACCE in the female PCI patients. MACCE of male PCI patients, who aged ≥ 80 years and hospitalized in the same period (n = 264), was also compared with that of the female PCI patients. Results A total of 32% very elderly female ACS patients received PCI in the present study. (1) Compared to female medical group, PCI procedure significantly alleviated the risks of MACCE: non-fatal MI (6.2% vs. 20.2%, P < 0.001), HFRH (10.9% vs. 22.5%, P = 0.012), CV death (12.4% vs. 28.7%, P < 0.001) and the composite MACCE (24.0% vs. 44.2%, P < 0.001) during the median follow-up period of 36 months. (2) Between very elderly female and male PCI patients, there were no significant differences in occurrence of MACCE (P = 0.232) and CV death (P = 0.951). (3) Multivariate Cox analysis revealed that ST-segment elevation myocardial infarction (STEMI) (HR 1.944, 95% CI 1.11–3.403, P = 0.02) and elevated log- N-Terminal pro-brain natriuretic peptide (NT-proBNP) (HR 1.689, 95% CI 1.029–2.773, P = 0.038) were independently associated with the incidence of MACCE in the female PCI patients. Conclusions PCI procedure significantly attenuated the risk of MACCE and improved the long-term clinical outcomes in very elderly female ACS patients. Aggressive PCI strategy may be reasonable in this population.


Background
Acute coronary syndrome (ACS) has been the leading cause of death worldwide [1,2]. The extensive application of percutaneous coronary intervention (PCI) has significantly improved the clinical outcomes of ACS patients [3,4]. However, there are still controversies and concerns about PCI treatment in very elderly female patients with ACS [5,6]. Very elderly women are a special cohort with their own clinical characteristics, which may affect the clinical outcomes of PCI in this population. For example, very elderly women with ACS were detected to have more risk factors or comorbidities, such as hypertension, diabetes and cerebrovascular disease [7], higher bleeding complications when treated with dual antiplatelet therapy (DAPT) [8,9], and more severe coronary artery lesions [10,11].
In clinical practice, cardiologists are more likely to avoid PCI procedure for very elderly women because of the potential higher risk of complications. Limited studies had also revealed that very elderly female ACS patients had received significantly less invasive angiography and timely revascularization [12,13] for higher mortality rates [12,14]. Furthermore, very elderly women treated with PCI for ACS were often excluded from large international clinical trials [15,16]. As a result, there are less clinical studies and clinical experiences of PCI procedure in this population. The aims of the present study were to investigate the effectiveness and clinical outcomes of PCI procedure in the very elderly female ACS patients in the era of drug-eluting stents, and to identify the possible risk factors which contributed to the incidence of MACCE in the very elderly female PCI cohort.

Study population and protocol
A total of 1449 ACS patients aged 80 years or older, who were admitted to Cardiovascular center, Beijing Friendship Hospital between January 2013, and December 2018, were enrolled into the study. The medical retrospective data were recruited and recorded in CBD-BANK (Cardiovascular Center Beijing Friendship Hospital Database Bank) supported by DHC SOFTWARE system (DongHua company SOFTWARE CO., LTD). ACS, including ST-elevation myocardial infarction (STEMI), non-ST-elevation MI (NSTEMI), and unstable angina pectoris (UAP), were diagnosed, and confirmed based on symptoms, electrocardiogram (ECG) changes, cardiac biomarkers. The treatment options of PCI or medication only depended on the patients' clinical situation, and decisions were made by two cardiologists simultaneously according to the international standards and guideline [17].
(3) Within the female group, patients were categorized into the PCI group (n = 232) and the medical group (n = 497). (4) Between the female PCI group and medical group, clinical characteristics, and follow-up MACCE were recorded and compared before and after Propensity score (PS) matching (1:1). (5) Within female PCI group, clinical and coronary artery characteristics were analyzed and compared in female PCI patients with (n = 56) and without MACCE (n = 176). Independent prognostic factors for the incidence of MACCE were detected in the female PCI patients. (6) MACCE of male PCI patients, aged ≥ 80 years old and hospitalized in the same period (n = 264), was also compared with the female PCI patients. The study protocol was approved by Institutional Ethics Committee of Beijing Friendship Hospital, and the study was also in accordance with the 1964 Declaration of Helsinki and its later amendments or comparable ethical standards.

Data collection
Baseline data were collected and analyzed, including demographic information, initial clinical presentation (heart rate, systolic/diastolic blood pressure at admission), and past medical history (hypertension, diabetes, dyslipidemia, prior MI/stroke and smoking). Laboratory examination results involved hemoglobin A1c, lipid spectrum, creatinine, alanine aminotransferase (ALT), N-Terminal pro-brain natriuretic peptide (NT-proBNP), and creatine kinase-MB (CK-MB), measured during hospitalization. M-mode and two-dimensional echocardiography (ECHO) were performed (Philips IE33) for routine parameters, such as left ventricular ejection fraction (LVEF). Characteristics of coronary artery lesion and stent implantation information were evaluated by angiographic and PCI procedure, presenting in the medical documents. Severe coronary artery stenosis was defined as epicardial main coronary artery stenosis ≥ 75% [18]. Chronic total occlusion (CTO) referred to 100% coronary artery occlusion with TIMI grade 0 flow and more than three months [19]. Successful PCI was identified as the attainment of a residual vessel diameter stenosis ≤ 20% and normal epicardial coronary flow (TIMI-3 flow) [20]. Hemorrhage was confirmed as "major" if there was a reduction of hemoglobin ≥ 5 g/dL or any intracranial bleeding [21]. Medication after discharge was determined from the medical records or regular telephone follow-up.

Primary and secondary endpoints and follow-up
Composite MACCE was defined as the primary endpoint, which was the combination of cardiovascular (CV) death, non-fatal MI, stroke, and heart failure requiring hospitalization (HFRH). The secondary endpoints included all-cause death and each of the adverse events mentioned above. CV death was referred to death from MI, HF, or documented cardiac sudden death. Stroke was new ischemic stroke that occurred during follow-up, confirmed by symptoms, results of computed tomography scan. Non-fatal MI and HFRH were determined by the information from medical records, including symptom, physical sign, results of ECG and ECHO, and the value of biomarkers. All MACCE were confirmed by two separate cardiologists simultaneously. Regular follow-up was conducted by clinic visits or phone interviews every 1-3 months until Dec 2019.

Statistical analysis
Continuous variables were expressed as median with interquartile range and were compared by Mann-Whitney U test. Categorical data were expressed as frequencies or percentages and were compared by Chi-square or Fisher's exact statistics. Propensity score matching (1:1) was performed between female PCI group and medical group to reduce the effect of treatment-selection bias and potential confounding factors in the present study.
The primary and secondary endpoints were compared between female PCI and medical groups, as well as female PCI group and male PCI group (≥ 80 years) hospitalized at the same period by Chi-square test. Survival curves were conducted by Kaplan-Meier method and compared with the log-rank test. Univariable and multivariable Cox proportional hazards analysis were used to estimate the hazard ratios (HR) and 95% confidence intervals (CI) for composite MACCE in the very elderly female PCI cohort. Based on possible confounding variables identified from the univariate analysis, multivariable Cox regression models in an all-enter way were constructed to determine the independent predictors for the incidence of composite MACCE in the female PCI group.
A two-sided P value < 0.05 was statistically significant. All the statistical analysis was conducted by the SPSS software version 23.0 (IBM, Armonk, NY, USA).
Between the very elderly female PCI and medical groups, there were no significant differences in age, BMI, heart rate, systolic/diastolic blood pressure at admission, and also in the medical history of hypertension, diabetes mellitus, dyslipidemia, prior MI and stroke. When compared to medical group, the proportions of STEMI (28.4% vs. 10.1%, P < 0.001) and smoking (9.9% vs. 5.8%, P = 0.046) were much higher in female PCI group, and the rate of UAP was much lower (47.8% vs. 69%, P < 0.001). In laboratory findings, higher levels of LDL-C (P = 0.012), ALT (P = 0.024) and CK-MB (P = 0.002) and lower level of creatinine (P = 0.016) were detected in female PCI group. Notably, there were no significant differences in length of stay and log [NT-proBNP] between the two groups. After PS matching, the differences in baseline data were disappeared, except for log [NT-proBNP] (P < 0.001), which was significantly increased in the medical group.

Primary and secondary endpoints in female PCI and medical groups
Before PS matching, the risks of CV death (10.8% vs. 18.1%, P = 0.011), all-cause death (16.4% vs. 24.3%, P = 0.015) and HFRH (9.1% vs. 17.5%, P = 0.003) were Table 1 Clinical characteristics for PCI and medical group in very elderly female ACS patients *P2Y12 receptor antagonist within 12 months after PCI. P, level of statistical significance BMI body mass index, BP blood pressure, MI myocardial infarction, PAD peripheral artery disease, UAP unstable angina pectoris, NSTEMI non-ST-segment elevation myocardial infarction, STEMI ST-elevation myocardial infarction, HbA1c Hemoglobin A1c, TC total cholesterol, LDL-C low-density lipoprotein cholesterol, TG triglyceride, ALT alanine aminotransferase, CK-MB creatine kinase-MB, NT-proBNP N-Terminal pro-brain natriuretic peptide, LVEF left ventricular ejection fraction, ACEI angiotensinconverting enzyme inhibitors, ARB angiotensin receptor blockers  (Table 2), compared to the female medical group. Kaplan-Meier curves illustrated the incidences of primary and secondary endpoints of the two groups in details ( Fig. 2a-f ).
As shown in Fig. 3a, b, the incidences of MACCE were also compared between very elderly female and male PCI group (ACS, ≥ 80 years, n = 264) hospitalized at the same period. The occurrences of composite MACCE (24.1% vs. 19.7%, P = 0.232), non-fatal MI (8.2% vs. 4.5%, P = 0.094) and HFRH (9.1% vs. 5.7%, P = 0.149) tended to be higher in female PCI group, but there were no significant differences between the two groups. Also, there were no remarkable differences in CV death (10.8% vs. 10.6%, P = 0.951) and all-cause death (17.7% vs. 18.9%, P = 0.456) between the very elderly female and male PCI cohorts.

Characteristics of the very elderly female PCI patients
Among the very elderly female PCI patients, 24% (n = 56) had MACCE during the follow-up period. In the female PCI patients with MACCE, the proportions of STEMI (44.6% vs. 23.3%, P = 0.002) and primary PCI (41.1% vs. 16.5%, P < 0.001) were much higher than that in PCI patients without MACCE. The median time of door-to-balloon (D-to-B) was 111 min (72.25, 147.5 min), and there was no significant difference in this time between the female PCI patients with and without MACCE. In addition, compared to female PCI patients without MACCE, the levels of creatinine (81.3 vs. 74.2 umol/L, P = 0.029) and log NT-proBNP (3.23 vs. 2.83, P < 0.001) were higher and the rate of LVEF ≥ 50% (72.2% vs. 91.9%, P < 0.001) was lower in female PCI patients with MACCE. Also, the risk of TIMI major bleeding was much higher in female PCI patients with MACCE (12.5% vs. 4.5%, P = 0.035). As for medication during follow-up, female PCI patients with MACCE had much lower proportion of uses in Aspirin (71.4% vs. 84.7%, P = 0.026), P2Y12 receptor antagonist (80.4% vs. 96.6%, P < 0.001) and β-blocker (42.9% vs. 65.9%, P = 0.002) than that in PCI patients without MACCE (Table 3).
All female PCI patients (100%, n = 232) received drug-eluting stent (DES) implantation, and the procedural success rate of PCI was 97% in the very elderly female cohort, as well as 98% in the very elderly male ACS patients in the present study (P = 0.605). There were no significant differences in stent number > 2 (32.1% vs. 36.4%, P = 0.565) and post-PCI TIMI 3 flow (94.6% vs. 98.3%, P = 0.134) between female PCI patients with and without MACCE. Also, there were no significant differences in D-to-B time of primary PCI

Results of Cox proportional hazards model analyses in female PCI cohort
All the baseline variables (except for variables with statistical collinearity) were included in univariate Cox proportional hazards model to identify the possible predictors for occurrence of MACCE in the female PCI cohort. As detailed in Table 5 Fig. 4).

Discussion
We investigated the efficacy and safety of PCI treatment in the very elderly women with ACS in the present study.
Our study demonstrated that PCI procedure significantly attenuated the risk of MACCE (24.0% vs. 44.2%, P < 0.001), and improved the long-term clinical outcomes in the very elderly female ACS population. In addition, STEMI (HR 1.944, 95% CI 1.11-3.403, P = 0.02) and elevated Log-NT-proBNP value (HR 1.689, 95% CI 1.029-2.773, P = 0.038) were independently associated with the incidence of MACCE after PCI in this cohort. Very elderly ACS patients had been historically underrepresented in clinical trials [16,22]. Limited studies had revealed a larger early hazard and less long-term benefit in the elderly female PCI population, including higher rates of MACCE, mortality and bleeding complication [23,24]. However, other study determined that coronary revascularization in elderly female ACS patients was associated with lower in-hospital MACCE and 1-year mortality, compared to no revascularized strategy [13]. We found that long-term clinical outcomes in the very elderly female PCI patients were distinctly superior to that in the patients with medical-only treatment during the median follow-up period of 36-month in the study. Compared to medical therapy alone, PCI procedure significantly alleviated the risks of MACCE in the very elderly female ACS cohort (P < 0.001), involving composite MACCE, non-fatal MI, HFRH (P = 0.012), CV death and all-cause death. Notably, PCI treatment did not extend the length of hospitalization. In earlier studies [15,25], there was always a lower DES use in elderly female ACS patients compared to the male patients, which might partly contribute to the poor clinical outcomes in this population. All the very elderly female patients received DES implantation in our study, and it was demonstrated that the use of DES in female ACS patients was more effective and safer than bare-metal stent (BMS) during long-term follow-up [26].
On the other hand, the effectiveness and benefits of PCI procedure were comparable between the very elderly female and male ACS patients (≥ 80 years), hospitalized in the same period in our study. Although the incidences of MACCE tended to be higher in female PCI group, there were no significant differences in primary and secondary endpoints between the very elderly female and male ACS patients, such as composite MACCE (P = 0.232), non-fatal MI (P = 0.094), HFRH (P = 0.149), CV death (P = 0.951) and all-cause death (P = 0.456). Based on the above results, we would like to suggest that very elderly female patients with ACS should be considered for more aggressive PCI strategy in the era of drugeluting stents. Fig. 3 Comparison of MACCE between very elderly female and male PCI patients. MACCE major adverse cardiovascular and cerebrovascular events, CV cardiovascular, MI myocardial infarction, HFRH heart failure requiring hospitalization, PCI percutaneous coronary intervention Very elderly patients always had more complex coronary artery lesions [27], including higher prevalence of multi-vessel disease, calcified lesions, tortuous lesions, and ostial lesions. We found that the coronary artery lesions were generally severe in this very elderly female cohort, but procedural success rate of PCI was relatively satisfactory, which was similar to previous studies of elderly patients (96-98%) [27,28]. Furthermore, the extent of coronary lesion in female PCI patients with MACCE was similar to that in female PCI patients without MACCE, and there were no significant differences in procedural success rate and number of stents implantation. But the study revealed that female PCI patients with MACCE had significantly lower proportion of pre-procedural TIMI 3 flow (P < 0.001), which meant the occurrence of AMI. In female patients with primary PCI, there was no significant difference in median time of door-to-balloon between patients with and without MACCE, which may be attributed to the fact that D-to-B time was relatively shorter Table 3 Clinical characteristics of patients with and without MACCE in female PCI group *P2Y12 receptor antagonist within 12 months after PCI. P, level of statistical significance BMI body mass index, MI myocardial infarction, UAP unstable angina pectoris, NSTEMI non-ST-segment elevation myocardial infarction, STEMI ST-elevation myocardial infarction, HbA1c Hemoglobin A1c, TC total cholesterol, LDL-C low-density lipoprotein cholesterol, TG triglyceride, ALT alanine aminotransferase, CK-MB creatine kinase-MB, NT-proBNP N-Terminal pro-brain natriuretic peptide, LVEF left ventricular ejection fraction, ACEI angiotensin-converting enzyme inhibitors, ARB angiotensin receptor blockers In clinical practice, there may be some factors that influence cardiologist's decision of PCI strategy. In the present study, more STEMI patients received PCI procedure, and patients with lower creatinine levels were more likely to experience PCI treatment. In addition, NT-proBNP was another important factor considered in PCI decision.

Limitations
The main limitation of the study is the retrospective design, single-centre data and experience. So, decision of therapy strategy reflected the convention and tendency of our single center, which may affect the objectivity of the conclusion. Furthermore, the sample size was relatively small after propensity matching (n = 129), especially the cases of female PCI patients with MACCE (n = 56), which may have some effect on the final clinical outcome analysis. We look forward to further studies with larger sample size, or a multicenter, prospective design to provide more information and understanding in this field.

Conclusion
In summary, PCI procedure attenuated the risks of MACCE and improved the long-term clinical outcomes in the very elderly female ACS patients in the era of drug-eluting stents. Aggressive PCI strategy should be considered as a treatment option in this population, and attentions should be paid to some factors that might have impacts on the clinical prognosis after PCI.