WSO - March 2024 - 361

Baik et al.
361
Figure 1. Proportion of patients with dual-antiplatelet
therapy after stent-assisted coil for unruptured intracranial
aneurysm.
for Statistical Computing). Statistical significance was set
at p < 0.05.
Results
SAC: stent-assisted coil.
Between January 2009 and December 2020, 115,761
patients with UIA were screened, and 32,198 underwent
SAC embolization (Supplemental Figure S1). After excluding
16,280 patients according to the study criteria, 15,918
patients with UIA treated with SAC were finally included.
Of the included patients, the mean age at SAC was
57.6 ± 10.8 years, and 3815 (24.0%) patients were men
(Supplemental Table S2). The proportion of patients with
DAPT was 79.4% at 90 days, 58.3% at 180 days, 28.9% at
1 year, and 6.9% at 2 years (Figure 1).
During the 2 year study period after SAC, 356 (2.2%)
patients had the primary composite outcome (229 (1.4%)
ischemic stroke and 127 (0.8%) major bleeding). Simon
and Makuch's plot showed no difference in the occurrence
of primary composite outcome according to DAPT
(Mantel-Byar test, p = 0.116, Figure 2(a)). Regarding the
secondary outcome, DAPT reduced the risk of ischemic
stroke (Mantel-Byar test, p < 0.001, Figure 2(b)) and
increased the risk of major bleeding (Mantel-Byar test,
p = 0.003, Figure 2(c)) after SAC.
that day; this was done to show a treatment trend with
DAPT after SAC for UIA. Considering the time-dependent
nature of DAPT, the estimated event-free survival curve
over the course of 2 years after SAC was constructed using
the Simon and Makuch method.14 Differences between the
survival curves according to DAPT over a 2-year period
were evaluated using the Mantel-Byar test for comparing
survival data with a time-dependent variable.15
We constructed a time-dependent Cox regression model
for the primary outcome and tested the proportional hazard
assumption with the Schoenfeld residual test.16 To elucidate
DAPT's potential effect on the risk of outcome occurrence
and explore the change over time, estimated unadjusted
time-varying hazard ratios (HRs) with confidence intervals
(CIs) were used.17 Since the proportional hazard assumption
of DAPT was violated in the Cox regression model, we
calculated the adjusted HR (aHR) and 95% CI for DAPT in
four periods after SAC. Adjustments were performed for
sex, age, insurance type, presence of hypertension, diabetes
mellitus, heart failure, chronic kidney disease, liver disease,
chronic obstructive pulmonary disease, malignancy, and
statin use. For the secondary outcome analysis, we constructed
two individual time-dependent Cox regression
models for ischemic stroke and major bleeding. The development
of competing risks was treated as censored at the
time. Statistical analyses were performed using SAS (version
9.4.2; SAS Institute) and R (version 3.5.1; R Foundation
When evaluating the proportional hazards assumption in
the Cox model with DAPT for the primary outcome, we
observed a violation through the Schoenfeld residual test
(Supplemental Figure S2). Hence, to explore the time-varying
effect of DAPT, we illustrated estimates of unadjusted
time-varying HR over the 2 years after SAC (Figure 3). The
time-varying effect of DAPT indicated a benefit for primary
composite outcome until approximately 180 days
after SAC, attenuating thereafter (Figure 3(a)). Dualantiplatelet
therapy had a protective effect for ischemic
stroke up to approximately 1 year after SAC; however, this
effect diminished over time (Figure 3(b)). Regarding major
bleeding, DAPT was associated with an increased risk of
major bleeding nearly throughout the observational period
(Figure 3(c)).
Regarding the time-varying effect of DAPT, we calculated
the aHR and event numbers in four different periods
after SAC (Supplemental Tables S1 and S3). The proportional
hazards assumption was not violated after dividing
time periods. Dual-antiplatelet therapy showed a 56%
reduced risk of primary composite outcome only within
90 days after SAC (aHR, 0.44; 95% CI 0.28-0.69;
p < 0.001; Table 1), with no significant difference in the
primary outcome beyond 90 days (p > 0.05; Table 1). The
primary composite outcome was also associated with male
sex, advanced age, medical aid (vs health insurance), and
history of diabetes and malignancy (all p < 0.05;
Supplemental Table S3). Regarding the occurrence of
ischemic stroke, DAPT was associated with 69% reduced
International Journal of Stroke, 19(3)
WSO - March 2024

Table of Contents for the Digital Edition of WSO - March 2024
