P3715Impact of sleep-disordered breathing on short-term functional outcomes in ischemic stroke patients: a cardiopulmonary coupling analysis using holter-monitoring

Background

Sleep-disordered breathing (SDB) assessed by conventional polysomnography is reported to have close association with worsened clinical outcomes in patients with ischemic stroke. The cardiopulmonary coupling (CPC) analysis using Holter-monitoring is an easily assessable method to evaluate SDB. However, its prognostic impact needs to be investigated.

Purpose

The present study investigated the prognostic impact of SDB defined by CPC analysis using Holter-monitoring at early stage of ischemic stroke on the functional disability at 3-month follow-up.

Methods

Total 692 patients with acute ischemic stroke who underwent Holter-monitoring were enrolled. The CPC analysis was conducted and SDB was defined as the presence of narrow-band (NB) coupling during sleep time. We investigated the association between SDB and functional disability at 3-month measured by modified Rankin scale (mRS).

Result

The NB coupling was present in 216 (31.2%) of 692 patients with mean age of 64.2±12.8 years. The NB group showed significantly higher proportion of severe functional disability (mRS ≥3; 45.3% vs. 12.3%, p<0.001) and persistent disability (ΔmRS≤0; 42.6% vs. 56.4%, p<0.001) after 3-month. In multivariate analysis, the presence of NB coupling was an independent predictor of higher risk of both severe and persistent functional disability (HR: 3.97; 95% CI: 2.37–6.64; p<0.001; and HR 1.92; 95% CI: 1.34–2.77; p<0.001, respectively). The results were consistent after propensity-score matched analysis with 175 patient pairs (C-statistics=0.759).

Parameters of functional disability Overall population (n=692) PSM population (n=350) no NB (n=476) NB (n=216) OR (95% CI) p-value no NB (n=175) NB (n=175) OR (95% CI) p-value Initial NIHSS ≥5 89 (18.6) 81 (37.5) <0.001 52 (29.7) 52 (29.7) >0.999 Discharge mRS ≥3 146 (30.6) 126 (58.3) <0.001 90 (51.4) 89 (50.8) 0.915 3-month mRS ≥3 59 (12.3) 98 (45.3) 5.86 (4.00–8.60) <0.001 38 (21.7) 72 (41.1) 2.52 (1.57–4.02) <0.001 3-month ΔmRS ≤0 (persisent disability) 203 (42.6) 122 (56.4) 1.74 (1.26–2.41) 0.001 77 (44.0) 100 (57.1) 1.69 (1.11–2.58) 0.014 Data are expressed as n (%). mRS = modified Rankin's scale; NB = narrow-band; NIHSS = National Institutes of Health Stroke Scale; OR = odds ratio; PSM = propensity-score matched.

Functional disabilities after 3-month

Conclusion

SDB assessed by CPC analysis at early phase of ischemic stroke was able to predict both greater and persistent functional disability at 3-month. The CPC analysis using Holter-monitoring is a useful modality for predicting functional disabilities in acute ischemic stroke.

P1249Prognostic impact of sarcopenia on major adverse cardiovascular outcomes in coronary artery disease patients undergoing successful percutaneous coronary intervention

Background

Sarcopenia is an emerging marker of frailty. Its prognostic impact on atherosclerotic cardiovascular disease (ASCVD) requires further investigation.

Purpose

We investigated the long-term prognostic impact of computed tomography (CT)-determined sarcopenia in patients with coronary artery disease (CAD).

Methods

Total 475 CAD patients those who underwent successful percutaneous coronary intervention (PCI) and performed CT scan within 30 days of PCI were enrolled. The cross-sectional area of skeletal muscle at the first lumbar vertebra (L1) level was measured. Sarcopenia was defined as L1 skeletal muscle index of less than 34.60 cm2/m2 for men and of less than 25.90 cm2/m2 for women. Primary outcome was 3-year all-cause mortality and secondary outcome was 3-year major adverse cardiovascular event (MACE), a composite of all-cause mortality, any myocardial infarction, and repeat revascularization.

Results

Sarcopenia was present in 214 (45.1%) of 475 patients. The incidence of 3-year all-cause mortality and MACE was significantly higher in patients with sarcopenia than in those without sarcopenia (17.7% vs. 5.7%, p<0.001; and 35.0% vs. 11.2%, p<0.001, respectively). In the fully adjusted multivariable analysis, sarcopenia was an independent predictor of higher risk of 3-year all-cause mortality (odds ratio [OR]: 2.98; 95% confidence interval [CI]: 1.35 to 6.58, p=0.007) and MACE (OR: 4.39; 95% CI: 2.49 to 7.73, p<0.001). The results were consistent after propensity-score matched analysis with 100 pairs of study population (C-statistics = 0.868).

Kaplan–Meier analysis of 3-year outcomes Overall population PSM population Sarcopenia (n=214) No sarcopenia (n=261) Log-rank p-value Sarcopenia (n=100) No sarcopenia (n=100) Log-rank p-value All-cause mortality 36 (17.7) 14 (5.7) <0.001 19 (20.0) 7 (7.7) 0.013 Non-fatal MI 12 (6.6) 5 (2.0) 0.021 6 (7.0) 2 (2.3) 0.134 Repeat revascularization 32 (20.3) 14 (6.2) <0.001 17 (23.3) 8 (8.0) 0.027 Total MACEs 68 (35.0) 27 (11.2) <0.001 36 (39.3) 14 (15.4) 0.001 Data are expressed as n (%). MACE = major adverse cardiovascular event; MI = myocardial infarction; PSM = propensity-score matched.

Clinical impact of sarcopenia on CAD

Conclusion

Sarcopenia is a useful predictor of adverse clinical outcomes in patients with CAD undergoing PCI. CT-determined sarcopenia may further aid in risk stratification and decision-making for patients with established ASCVD.

Acknowledgement/Funding

National Research Foundation of Korea (NRF-2016R1A2B3013825), Ministry of Future Creation and Science of Korea (2018K000255)

P4339Reference values for cardiorespiratory fitness in healthy Koreans: compared to western nations and nomogram

Introduction

Cardiorespiratory Fitness (CRF), defined as the integrated ability to properly oxygenate skeletal muscles during physical activity, is associated with a high risk of cardiovascular disease and all-cause mortality. The reference range for CRF may differ among nations, with Asians under-represented in previous data.

Purpose

In this study, we sought to establish reference values of CRF for Asians using a recent Korean cohort.

Methods

We analyzed 2646 healthy Korean adults recently enrolled in the Korea Institute of Sports Science Fitness Standards (KISS FitS) project with estimated maximal oxygen uptake (VO2max) values during treadmill test. Patients with cardiovascular or renal disease, systemic infection, pregnant women and those with orthopedic injuries unable to measure physical fitness were excluded. Age-specific mean VO2max values were compared with those from recent American, Norwegian, Danish cohorts and old Korean data.

Results

Age-specific reference values for healthy Korean adults in this cohort were as shown (Table). We were able to draw a nomogram to predict exercise capacity for a given age and MET value (Figure). When compared to other countries, less CRF reduction by aging was seen in Asians than in other Westerners. When compared to old Korean data from the 1980s, values were similar after adjustment for difference in methods, except for those under 30 years old which were decreased.

Exercise capacity of healthy Korean Men Women Age VO2max (ml/kg/min) N P-value for trend Age VO2max (ml/kg/min) N P-value for trend 19–29 42.3±6.3± 209 <0.01 19–29 34.3±4.3± 110 <0.01 30–39 42.0±5.0± 170 39–39 32.2±4.5± 211 40–49 41.4±5.6± 238 40–49 30.8±4.6± 284 50–59 38.0±5.7± 274 50–59 28.3±4.6± 367 60–69 32.4±6.2± 134 60–69 26.0±5.7± 336 70–79 27.2±5.6± 83 70–79 23.9±4.4± 195 >80 24.1±4.0± 11 >80 21.0±3.7± 24 Total 38.6±7.4± 1119 Total 28.5±5.8 1527 Data are presented as mean ± standard deviation. VO2max, maximal oxygen uptake; N, number.

Nomogram of exercise capacity in Koreans

Conclusions

While there was no significant change in CRF over time in the same ethnic group, there was a clear inter-ethnic difference. CRF should be assessed according to ethnic or national standards, and it is necessary to establish a reference for each nation or ethnicity with periodic updates.

Acknowledgement/Funding

National Sports Promotion Fund of the Korea Sports Promotion Foundation in 2015

P5744Sleep-disordered breathing assessed by holter-monitoring is associated to worsened one-year clinical outcomes in ischemic stroke patients: a cardiopulmonary coupling analysis

Background

Sleep-disorder breathing (SDB) using polysomnography is closely associated to poor functional and clinical outcomes in ischemic stroke patients. The cardiopulmonary coupling analysis using Holter-monitoring (CPC-Holter analysis) is an emerging feasible modality to investigate SDB.

Purpose

We investigated the association between SDB defined by CPC-Holter analysis and one-year clinical outcome in patients with acute ischemic stroke.

Methods

Total 666 patients with acute ischemic stroke who underwent Holter-monitoring were enrolled. The CPC-Holter analysis was conducted and SDB was defined as the presence of narrow-band (NB) coupling during sleep time. Primary outcome was recurrent ischemic stroke, and secondary outcome was major adverse cerebrovascular event (MACE), a composite of recurrent ischemic stroke, transient ischemic attack, and all-cause mortality within one year since discharge.

Result

The NB coupling was present in 205 (30.8%) of 666 patients with mean age of 64.1±12.8 years. The NB group showed significantly higher incidence of both recurrent ischemic stroke (8.3% vs. 1.4%, p<0.001) and MACE (14.9% vs. 3.0%, p<0.001) within one-year. In multivariate analysis, presence of NB coupling remained as an independent predictor of both recurrent ischemic stroke and MACE (HR: 4.81; 95% CI: 1.73–13.4; p=0.003; and HR 4.17; 95% CI: 1.74–10.0; p<0.001, respectively). The results were consistent after propensity-score matched analysis with 164 patient pairs (C-statistics=0.757).

One-year clinical outcomes Overall population (n=666) PSM population (n=328) no NB (=461) NB (n=205) Log-rank p-value OR (95% CI) no NB (n=164) NB (n=164) Log-rank p-value OR (95% CI) Recurrent ischemic stroke 6 (1.4) 14 (8.3) <0.001 5.73 (2.20–14.9) 3 (2.0) 11 (8.1) 0.026 3.85 (1.07–13.8) Transient ischemic attack 3 (0.7) 3 (1.7) 0.275 2 (1.3) 3 (2.1) 0.633 Hemorrhagic stroke 0 (0.0) 2 (1.2) 0.027 0 (0.0) 2 (1.5) 0.148 Total death 3 (0.7) 9 (4.8) 0.001 2 (1.3) 3 (1.9) 0.641 MACEs 12 (3.0) 25 (14.9) <0.001 4.63 (2.06–10.4) 7 (5.2) 17 (13.1) 0.030 2.95 (1.06–8.21) Data are expressed as n (%). CI = confidence interval; MACE = major adverse cardiovascular event; NB = narrow-band; OR = odds ratio.

One-year clinical outcomes

Conclusion

SDB assessed by CPC-Holter analysis at early phase of ischemic stroke is a powerful prognostic marker for predicting one-year adverse clinical outcomes. The CPC analysis using Holter-monitoring is a useful modality and could be easily applied to predict clinical outcomes in acute ischemic stroke patients.

P2501Validation of FRIEND and ACSM equations for cardiorespiratory fitness: comparison to direct measurement in male CAD patients

Introduction

Cardiorespiratory fitness (CRF) is associated with a high risk of cardiovascular disease and all-cause mortality. The regression equation of American College of Sports Medicine (ACSM) was a preferred method for estimating maximal oxygen consumption (VO2max). It is well-known that CRF is overestimated in ACSM equation. Recently, Kokkinos reported more precise equation from the Fitness Registry and the Importance of Exercise National Database (FRIEND). Both equations were made from western healthy people.

Purpose

In this study, we compared VO2max estimated by ACSM and FRIEND equations to VO2max directly measured in male coronary artery disease (CAD) patients.

Methods

We analyzed 103 male CAD patients who underwent percutaneous coronary intervention and who participated in cardiac rehabilitation between June 2015 and December 2018. VO2max was directly measured by the gas exchange analysis during treadmill test with modified Bruce protocol. Exclusive criteria were pulmonary disease, chronic kidney disease on hemodialysis, malignancy, peripheral artery disease, insufficient cardiopulmonary exercise test and orthopedic injuries. Directly measured VO2max were compared to ACSM and FRIEND equations.

Results

Age-specific VO2max values, which were directly measured from male CAD patients, were shown in Table. Smaller CRF difference was shown in FRIEND equation than ACSM equation. Compared to the measured value, CRF estimated by ACSM equation was overestimated by 22%, but the one estimated by FRIEND equation had only 2% gap. Figure presents Bland-Altman plots. ACSM equation had the higher bias (5.52ml/kg/min) compared with FRIEND equation (0.200ml/kg/min).

Comparison table of VO2max estimated by ACSM and FRIEND equations with directly measured VO2max in male CAD patients Age Number VO2max (ml/kg/min) Measured ACSM % predicted FRIEND % predicted 30–39 4 29.5 ± (6.6) 35.7 ± (6.1) 122.3 ± (8.5) 29.1 ± (4.3) 100.1 ± (8.7) 40–49 20 29.1 ± (5.1) 35.7 ± (5.4) 123.6 ± (11.1) 29.1 ± (3.8) 101.2 ± (9.5) 50–59 32 25.6 ± (4.3) 31.5 ± (5.1) 123.8 ± (10.1) 26.2 ± (3.6) 103.2 ± (8.5) 60–69 27 26.0 ± (5.1) 31.1 ± (2.6) 120.5 ± (12.6) 25.9 ± (4.0) 100.8 ± (10.5) 70–79 18 21.2 ± (6.0) 26.8 ± (5.4) 123.2 ± (14.8) 22.8 ± (3.9) 105.2 ± (12.0) >80 2 25.0 ± (10.1) 22.8 ± (2.6) 97 ± (28.9) 19.9 ± (1.9) 85.1 ± (27.0) Total 103 25.7 ± (5.6) 31.4 ± (6.0) 122.2 ± (12.4) 26.1 ± (4.3) 102.1 ± (10.4) Data are presented as mean ± (standard deviation).

Bland-Altman plots

Conclusions

FRIEND equation can estimate CRF more accurately than ACSM equation, even in Asian patients with CAD.

4098Reduced skeletal muscle mass is associated to worsened long-term clinical outcomes in patients with coronary artery disease: a quantitative analysis by computed tomography

Background

Sarcopenia is closely associated to poor clinical outcomes in patients with atherosclerotic cardiovascular disease (ASCVD). However, it is unclear whether the skeletal muscle mass at baseline has quantitative effect on future cardiovascular outcomes.

Purpose

We investigated the quantitative effect of skeletal muscle mass on future cardiovascular outcomes in patients with coronary artery disease (CAD).

Methods

Total 475 patients those who underwent successful percutaneous coronary intervention (PCI) for CAD and performed computed tomography (CT) scan within 30 days of PCI were enrolled. The cross-sectional area of skeletal muscle at the first lumbar vertebra (L1) level was measured. Whole study population was divided into 4 groups according to the sex-specific quartiles of skeletal muscle index (SMI). Primary outcome was all-cause mortality and secondary outcome was major adverse cardiovascular event (MACE) within 3 years of follow-up.

Results

Mean follow-up duration was 4.11±3.02 years and average time period from the date of PCI to CT scan was −3.33±11.72 days. The incidence of 3-year all-cause mortality (23.2% vs. 9.9% vs. 6.6% vs. 4.4%, p<0.001) and MACE (42.9% vs. 24.0% vs. 14.3% vs. 6.2%, p<0.001) was significantly higher in the group of lower quartiles of L1-SMI. In the fully adjusted multivariable analysis, lower quartiles of L1-SMI was an independent predictor of higher risk of all-cause mortality and MACE (lowest vs. highest quartile; OR: 4.90, 95% CI: 1.54 to 15.5, p=0.007; and OR: 12.3, 95% CI: 4.99 to 30.4, p<0.001, respectively).

Results of 3-year clinical outcomes SMI Q1 (n=124) SMI Q2 (n=116) SMI Q3 (n=112) SMI Q4 (n=123) Log-rank p-value All-cause mortality 27 (23.2) 11 (9.9) 7 (6.6) 5 (4.4) <0.001 Non-fatal MI 9 (8.7) 3 (3.0) 2 (2.0) 3 (2.6) 0.038 Repeat revascularization 20 (24.9) 15 (15.2) 7 (7.1) 4 (3.8) <0.001 Total MACEs 47 (42.9) 26 (24.0) 15 (14.3) 7 (6.2) <0.001 Data are expressed as n (%). MACE = major adverse cardiovascular event; MI = myocardial infarction; SMI = skeletal muscle index; Q = quartile.

Impact of reduced skeletal muscle on CAD

Conclusion

Skeletal muscle mass at baseline is a powerful predictor of future adverse clinical outcomes in patients with CAD undergoing successful PCI. Quantitative assessment of skeletal muscle mass at L1 level by CT scan provides prognostic implication for future cardiovascular risk stratification.

Acknowledgement/Funding

National Research Foundation of Korea (NRF-2016R1A2B3013825), Ministry of Future Creation and Science of Korea (2018K000255)

P2523Hand grip strength as a predictor of exercise capacity in coronary heart disease

A recent study has shown that quadriceps strength can be used to predict the level of exercise capacity in patients with coronary heart disease (CHD). We investigated whether the relationship between muscular strength and exercise capacity is also observed with hand grip strength (HGS). We studied 443 participants (age, 61.8±11.2 y; 77.7% male) who underwent coronary intervention and participated in cardiac rehabilitation between 2015 and 2018. Participants were assessed for grip strength, measured using a Jamar dynamometer. Logistic regression was used to assess the relationship between various clinical measures (HGS, age, sex, etc) with the distance walked on a 6-minute walk test (6MWT) and maximal oxygen uptake (VO2max). HGS was significantly related to distance walked on the 6MWT (r=0.435, p<0.001). It was the only predictor of all exercise capacity categories, and one of the strongest predictors of each exercise capacity category. A HGS of 25.5% of body weight predicted an achievement of a 200 m walk on the 6MWT (positive predictive value = 0.95). However, HGS less than 35.5% of body weight predicted that 500m could not be done in 6 minutes (negative predictive value = 0.97). This trend was also observed in the subgroups in which VO2max was measured. This study demonstrates that HGS is associated with exercise capacity in CHD and can be used to predict the level of exercise capacity. These findings may contribute to setting the recommended level of daily activity as well as the level of cardiac rehabilitation in CHD.

Logistic regression models for different levels of exercise capacity Level of exercise capacity B±S.E p-value Odd ratio 95% CI Distance of 6MWT   200 m Grip strength 0.054±0.014 <0.001 1.056 1.027–1.086   300 m Grip strength 0.042±0.009 <0.001 1.042 1.024–1.062   400 m Grip strength 0.047±0.011 <0.001 1.048 1.026–1.070   500 m Grip strength 0.051±0.016 0.001 1.053 1.021–1.086 VO2max level   4 METs Grip strength 0.054±0.010 <0.001 1.056 1.036–1.076   6 METs Grip strength 0.059±0.011 <0.001 1.061 1.039–1.083   8 METs Grip strength 0.081±0.015 <0.001 1.085 1.053–1.117   10 METs Grip strength 0.113±0.049 0.019 1.12 1.019–3.232 Data are presented as mean ± standard deviation (SD). 6MWT, 6-minute walk test; STEMI, ST-Elevation Myocardial Infarction; SE, standard error; CI, confidence interval; VO2max, Maximal Oxygen uptake; METs, Metabolic equivalents.

Estrogen Selectively Enhances TMJ Disc but Not Knee Meniscus Matrix Loss

The preponderance of temporomandibular joint (TMJ) degenerative disorders in women and their early onset during reproductive years have implicated female sex hormones, particularly 17-β estradiol (E2), in the pathogenesis of these disorders. Nevertheless, the mechanisms by which E2 contributes to TMJ degenerative disorders and the reasons for its targeted effects on the TMJ but not other joints remain poorly understood. Here, we developed an ovariectomized mouse model in which systemic E2 concentrations mimicked those in cycling women, and we determined the effect of E2 on the targeted turnover of TMJ fibrocartilage matrix via E2-induced matrix metalloproteinases MMP9 and MMP13. Infusion of E2 and progesterone (P4; hormone control) over 7 d resulted in 5- and 8-fold greater serum E2 and P4 levels relative to controls, respectively, achieving systemic hormone levels similar to high baseline levels in cycling women. Administration of E2 but not P4 caused a significant loss of TMJ collagen and glycosaminoglycans, which was accompanied by amplification of ERα and specific increases in MMP9 and MMP13 expression. This dose of E2 had no effect on knee meniscus fibrocartilage, demonstrating the specificity of the degradative effect of E2. Dose-response experiments showed a greater sensitivity and a higher peak induction of MMP9 and MMP13 in TMJ fibrocartilaginous cells than knee meniscus cells to E2, providing an explanation for the differential responses of these tissues to E2. Using MMP9- and MMP13-null mice, we observed no discernible effects of each proteinase individually to E2-mediated TMJ matrix loss but noted a significant compensatory reciprocal induction of each MMP by E2 in the absence of the other. The redundancy in E2's induction of MMP9 and MMP13 suggests that the proteinases may together contribute to E2-mediated TMJ fibrocartilage loss. These results advance our understanding of E2-mediated upregulation of MMP9 and MMP13 on fibrocartilage matrix turnover targeted to the TMJ.

Compressive stress effect on the loss mechanism in a soft piezoelectric Pb(Zr,Ti)O3

High power piezoelectric devices are often subjected to external mechanical biases, in applications such as underwater transducers. While the performance of these devices under external pressure has been rather extensively studied, there is a lack of study on the loss mechanism in terms of three dielectric, elastic, and piezoelectric losses. Thus, in this paper, we study the mechanical bias stress dependence of the loss mechanism in a soft piezoelectric Pb(Zr,Ti)O₃ (PZT) from a scientific viewpoint, using an equivalent circuit methodology based on the fundamental longitudinal mode. In order to measure the loss behavior, a modified bolt-clamped Langevin transducer was designed and optimized using finite element analysis in order to facilitate the analysis easier. We present the preliminary experimental part of our project on the design of the proposed structure/methodology, material creep behavior, stress relaxation, and uniform stress distribution, in order to minimize the experimental errors. We also introduce a six terminal equivalent circuit analysis in order to determine three losses in the PZT specimen. The resonance/antiresonance frequencies and quality factors showed monotonous increase under compressive stress. Loss factors for one PZT composition are reported in this paper to show the feasibility of our methodology for measuring the uniaxial compressive stress dependence.

Abstract P1-14-07: Randomized phase 3 study of anthracycline-containing regimens versus S-1 as first-line treatment for metastatic breast cancer (SELECT BC-CONFIRM)–A combined analysis of two randomized phase 3 studies (SELECT BC-CONFIRM and SELECT BC)–

Background: Anthracycline-containing regimens and taxane have been standard as the first-line chemotherapy for metastatic breast cancer (MBC). We conducted SELECT BC (randomized phase 3 study of taxane versus S-1 as first-line treatment for MBC) for evaluating the efficacy of S-1 for patients with HER2-negative MBC from 2006 to 2010 in Japan. This study demonstrated non-inferiority of S-1 in overall survival (OS) (median OS was 37.2 months in taxes group and 35.0 months in S-1 group (HR 1.05, 95% CI 0.86–1.27, p=0.015)), and superiority in health-related quality of life (HRQOL) to taxanes. S-1 was also shown as less toxic than taxane (Lancet Oncol 2016; 17: 90-98). S-1 might provide clinical benefit as first-line treatment for patients with HER2-negative MBC. To confirm this suggestion, we have conducted further study (randomized phase 3 study of anthracycline-containing regimens versus S-1 as first-line treatment for HER2-negative MBC: SELECT BC-CONFIRM) from 2011 to present, and a combined analysis of two randomized studies (SELECT-BC CONFIRM and SELECT-BC).

Methods: In SELECT BC-CONFIRM, 230 patients receiving first-line treatment for MBC were randomly assigned to either anthracycline group (n=115) or S-1 group (n=115). Anthracycline group patients received anthracycline-containing regimens (AC, EC, FAC, FEC, q3w) at the discretion of the treating physician. S-1 group patients received S-1 40–60 mg twice daily based on the patient's body surface area for 28 days on, 14-day off. The primary endpoint was OS, and secondary endpoints were progression-free survival (PFS), time to treatment failure (TTF), adverse events, HRQOL, and cost-effectiveness. The results were combined with SELECT-BC, to confirm the hypothesis that S-1 treatment is not inferior to the standard therapy (taxanes / anthracycline) for HER2-negative MBC.

Results: A combined analysis of the two studies showed that HR was 1.06, 95%CI 0.90-1.253, and p=0.0071 between the standard therapy group and S-1 group. In addition, the Bayesian posterior probability for which HR would be less than 1.333 was about 99.6%.

Conclusions: A combined analysis of SELECT BC-CONFIRM and SELECT BC clearly demonstrated that OS with S-1 was not inferior to that with the standard therapy in patients receiving first-line treatment for HER2-negative MBC. S-1 could become a standard therapy for this patient population.

Citation Format: Park Y, Akabane H, Watanabe T, Takahashi M, Sagara Y, Nishimura R, Tsurutani J, Takashima T, Fujisawa T, Hozumi Y, Uemura Y, Mukai H. Randomized phase 3 study of anthracycline-containing regimens versus S-1 as first-line treatment for metastatic breast cancer (SELECT BC-CONFIRM)–A combined analysis of two randomized phase 3 studies (SELECT BC-CONFIRM and SELECT BC)– [abstract]. In: Proceedings of the 2018 San Antonio Breast Cancer Symposium; 2018 Dec 4-8; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2019;79(4 Suppl):Abstract nr P1-14-07.

Abstract P4-14-09: A nationwide data on the cardiovascular protective effect of tamoxifen and aromatase inhibitor in postmenopausal women with breast cancer

A large proportion of breast cancer patients receive hormonal therapy as their adjuvant treatment options. For postmenopausal women, the initial choice for the hormonal therapy is aromatase inhibitor (AI), and tamoxifen (TM) is reserved for women experiencing severe side effects against AI or having low bone density. An important but unresolved clinical question regarding the use of AI in postmenopausal women is the safety of AI regarding the risk cardiovascular events. Studies have shown inconsistent results over the cardiovascular safety of AI and TM. In this study, we investigated the risk of developing cardiovascular and cerebrovascular events in women with breast cancer who receive hormonal therapy using AI, TM, or both.

To this end, we used the National Health Insurance Sharing Service in Korea which is provided by National Health Insurance Service. The database provides anonymized insurance data for research purposes after the approval of the review committee. In the database, we identified 47,569 women with the age older than 55 who were diagnosed with breast cancer. Patients were classified as no hormonal treatment group (n=18,807), AI group (n=19,584), TM group (n=7,081), or Switch group (n=2,097). The Switch group was defined as the women with history of both AI and TM prescriptions. During the studied period, a total of 2,032 cardiovascular or cerebrovascular events (CVE) were recorded.

Overall, the women prescribed with TM had significantly less hazard ratio for developing CVE when compared to the women who did not receive any hormonal treatment (HR 0.809 95% C.I. 0.706-0.928). However, this protective effect of tamoxifen was not observed in either AI or Switch group (HR 0.917 95% C.I. 0.833-1.010, and HR 0.856 95% C.I. 0.695-1.053, respectively). The protective effect of TM was also similar in women older than 60 (HR 0.808 95% C.I. 0.696-0.938). The cardiovascular and cerebrovascular protective effects of tamoxifen was also substantial in high risk women defined by their family history of cardiovascular diseases and the diagnosis of hypertension or diabetes.

Our results suggest that the use of TM is associated with a substantial protective effect against developing cardiovascular or cerebrovascular events in women with breast cancer. However, the protective effect was not observed for women receiving AI. Our data suggest the potential tailored approach in hormonal treatment in breast cancer patients who are at high risk of cardiovascular of cerebrovascular events.

Citation Format: Moon H-G, Choi SH, Park Y, Jung JG, Ju YW, Kim KE, Kim Y, Lee E, Lee H-B, Han W, Noh D-Y, Yoon H-J. A nationwide data on the cardiovascular protective effect of tamoxifen and aromatase inhibitor in postmenopausal women with breast cancer [abstract]. In: Proceedings of the 2018 San Antonio Breast Cancer Symposium; 2018 Dec 4-8; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2019;79(4 Suppl):Abstract nr P4-14-09.