Determination of the RAdial
versus GrOiN coronary
angioplasty
The Result of DRAGON Trial
Shigeru Saito, MD
Department of Cardiology and Catheterization Laboratory
Shonan Kamakura General Hospital, Kamakura, Japan
on behalf of Dragon Trial investigators
Disclosure Statement of Financial Interest
• Grant/Research Support
• Consulting Fees/Honoraria
• Major Stock Shareholder/Equity
• Royalty Income
• Ownership/Founder
• Intellectual Property Rights
• Other Financial Benefit
• None
• TERUMO, Boston Scientific
• None
• None
• None
• None
• None
Within the past 12 months, I or my spouse/partner have had a financial
interest/arrangement or affiliation with the organization(s) listed below.
Affiliation/Financial Relationship Company
Background this study
• TRI is getting more popular all around the
world.
• However, it is not yet clear whether ad-hoc
TRI strategy can provide clinical outcomes
similar to TFI strategy.
• Thus, we initiated DRAGON trial to show the
effectiveness of ad-hoc TRI strategy in the
real world.
Objective of this study
• To determine the clinical benefit of the transradial
approach compared to the transfemoral approach in
patients, who receive percutaneous coronary
intervention in the real world practice.
• The benefit is shown based on the non-inferiority of
TRI in terms of clinical effectiveness at 1 year and on
the superiority of TRI to reduce the major bleeding
complications at 7 days compared to TFI.
Primary Investigators Shigeru SAITO, MD
Yong HUO, MD
Honorary Primary
Investigator
Runlin GAO, MD
Ferdinard KIEMENEIJ, MD
Steering Committee Shigeru SAITO
Junbo GE
Yujie ZHOU
Guosheng FU
Jian’an WANG
Bo XU
Yong HUO
Shubin QIAO
Yaling HAN
Jiyan CHEN
Haichang WANG
Wei LI
Biostatistic Analysis Medical Research & Biometrics Center
National Center for Cardiovascular Diseases
Cardiovascular Institute & Fuwai Hospital
Sponsor TERUMO
CRO/DM/CL CCRF
Study Organization
Endpoints
Primary Endpoint
 Major Adverse Cardiac or Cerebrovascular
Event (MACCE) free rate at 12 months
Major Secondary Endpoint
 Major bleeding complication (BARC definition
type 3 or 5) free rate at 7 days
Statistical considerations
• This study had 80% statistical powered to
demonstrate the non-inferiority (NI margin=5%) of
TRI on primary endpoint (MACCE-free rate at 1 year)
comparing to TFI
• And also powered (>85%) to demonstrate the
superiority of TRI on major secondary endpoint
(bleeding-free rate at 7 days) comparing to TFI
• Intention-To-Treat (ITT) principle had been used for
the primary analysis
Enrollments
Patients who had PCI were the target population
Kaplan-Meier curve
MACCE-free rate at 12 months
(PCI population before adjustment)
TRI 1212 1168 1147 1143 1133 1125 1114
TFI 527 494 483 475 467 462 459
HR = 0.707 (95% CI: 0.443-1.129)
P-value (log-rank test): 0.144
Pts. at risk
TRI
TFI
Primary Endpoint
MACCE-free rate at 12 months
The non-inferiority was met as the upper 95%
confidence bound is less than the non-
inferiority margin (5%)
Difference [95% CI]
TRI
(N=1,212)
TFI
(N=527)
Difference
[95% CI]
TFI-TRI
P-value
(non-
inferiority)
96.1% 94.4%
-1.7%
[-4.0%; 0.6%]
<0.001
TRI is better TFI is better
Non-inferiority
-5% 0% 5%
(PCI population before adjustment)
Major bleeding complication-free rate
at 7 days (major 2nd endpoint)
There was no significant difference between
the TRI and TFI group on the 2nd endpoint
among the unadjusted observed data
Difference [95% CI]
TRI
(N=1212)
TFI
(N=527)
Difference
[95% CI]
TFI-TRI
P-value
(superiority)
99.8% 99.4%
-0.4%
[-1.1%; 0.3%]
0.131TRI is better TFI is better
-3% 0% 3%
(PCI population before adjustment)
TRI
(N=1,212)
TFI
(N=527)
P-value
Male (%) 72.5% 64.7% 0.001
Age (yrs) 61.1±10.8 62.9±10.9 0.001
Height (cm) 168.6±6.9 166.9±7.5 <0.001
Weight (Kg) 71.2±10.6 69.0±10.8 <0.001
BMI (Kg/m2) 25.0±3.1 24.6±3.1 0.049
Heart Rate (Beats/m) 72.1±11.8 72.4±13.0 0.626
Current Smoker (%) 37.8% 31.3% 0.033
Baseline characteristics
(PCI population before adjustment)
Disease history
(PCI population before adjustment)
TRI
(N=1,212)
TFI
(N=527)
P-value
Hypertension (%) 58.7% 60.9% 0.380
Hyperlipidemia (%) 15.2% 11.8% 0.043
Diabetes (%) 25.0% 24.9% 0.950
Prior MI (%) 11.4% 13.7% 0.185
Prior PCI (%) 12.1% 14.2% 0.231
Prior CABG (%) 0.2% 3.2% <0.001
Cerebrovascular Disease (%) 11.0% 12.3% 0.415
Peripheral vascular Disease(%) 1.4% 0.9% 0.424
Family history of CAD (%) 8.8% 8.3% 0.744
Lesion characteristics
(PCI population before adjustment)
TRI
(N=1,611)
TFI
(N=729)
P-value
Lesion type (%)
A 1.9% 2.3% 0.780
B1 25.5% 24.0%
B2 7.2% 7.8%
C 65.4% 65.8%
Bifurcation lesion (%) 8.9% 12.1% 0.018
CTO (%) 5.3% 8.9% 0.001
TIMI 0&1 (%) 22.7% 35.1% <0.001
Number of stents 1.2±0.6 1.2±0.6 0.7007
Stent dia. (mm) 3.0±0.4 3.0±0.4 0.9629
Stent length (mm) 25.3±7.4 25.3±7.5 0.9850
Lesion success (%) 99.7% 99.2% 0.1078
• The 23 baseline covariates had been included in the PS
model were pre-specified without any knowledge of clinical
outcomes (both efficacy & safety)
• HL test had a P=0.857
• IPW (inverse probability
weighting) had been
carried out
Since we unexpectedly found imbalanced
randomized allocation, we did propensity score
analysis based on 23 baseline covariates.
Weintraub WS, et al. Comparative effectiveness of revascularization strategies. N
Engl J Med. 2012; 366: 1467-76.
TRI
(N=1,212)
TFI
(N=527)
P-value
Male (%) 70.4% 70.5% 0.928
Age (yrs) 61.54 ±12.85 61.15 ±20.62 0.467
Height (cm) 168.09 ±8.34 167.71 ±13.65 0.349
Weight (Kg) 70.74 ±12.82 70.39 ±21.40 0.573
BMI (Kg/m2) 24.99 ±3.78 24.97 ±5.83 0.926
Heart Rate (Beats/m) 72.12 ±13.97 72.65 ±24.25 0.378
Current Smoker (%) 37.1% 33.3% 0.012
Baseline characteristics
(After IPW adjustment)
TRI
(N=1,212)
TFI
(N=527)
P-value
Hypertension (%) 59.0% 56.1% 0.084
Hyperlipidemia (%) 13.9% 13.1% 0.458
Diabetes (%) 25.4% 24.4% 0.492
Prior MI (%) 12.2% 11.1% 0.301
Prior PCI (%) 12.3% 13.4% 0.319
Prior CABG (%) 0.8% 1.1% 0.323
Cerebrovascular Disease (%) 10.7% 12.9% 0.039
Peripheral vascular Disease(%) 1.2% 1.2% 0.894
Family history of CAD (%) 7.9% 7.2% 0.482
Disease history
(After IPW adjustment)
TRI
(N=1,212)
TFI
(N=527)
P-value
Silent Myocardial
Ischemia (%)
1.7% 1.8% 0.797
Angina (%) 61.0% 62.7% 0.283
MI (%) 37.4% 35.5% 0.248
LVEF (%) 58.63 ±11.24 58.61 ±16.92 0.973
NYHA class (%)
Ⅰ 63.3% 64.5% 0.010
Ⅱ 33.0% 29.7%
Ⅲ 3.6% 5.8%
Ⅳ 0.2% 0.0%
Baseline diagnosis
(After IPW adjustment)
TRI
(N=1,611)
TFI
(N=729)
P-value
TYPE (%)
A 1.8% 2.0% 0.939
B1 25.2% 25.6%
B2 7.7% 7.4%
C 65.3% 65.0%
Bifurcation lesion (%) 9.4% 9.6% 0.864
CTO (%) 6.1% 6.3% 0.754
Pre-op TIMI 0&1 (%) 23.9% 30.9%
Number of
stents/lesion
1.2±0.7 1.2±1.1 0.515
Lesion characteristics
(PCI population after IPW adjustment)
TRI
(N=1,212)
TFI
(N=527)
P-value
Lesion number 1.3±0.7 1.4±1.2 0.054
Procedural success (%) 99.6% 98.7% 0.007
Operative complication (%) 1.6% 3.3% 0.001
Procedure time (min) 59.1±35.3 60.1±62.3 0.516
Procedural information
(After IPW adjustment)
HR = 0.928 (95% CI: 0.672-1.280)
P-value (log-rank test): 0.822
Kaplan-Meier curve
MACCE-free rate at 12 months
(After IPW adjustment)
TRI 1212 1168 1147 1143 1133 1125 1114
TFI 527 494 483 475 467 462 459
Pts. at risk
TRI
TFI
Difference [95% CI]
TRI is better TFI is better
Non-inferiority
-5% 0% 5%
TRI
(N=1,212)
TFI
(N=527)
Difference
[95% CI]
TFI-TRI
P-value
(non-
inferiority)
95.8% 95.5%
-0.3%
[-1.7%; 1.0%]
<0.001
Primary Endpoint
MACCE-free rate at 12 months
(After IPW adjustment)
The non-inferiority was met as the upper 95%
confidence bound is less than the non-
inferiority margin (5%)
The superiority is met (after propensity score-
IPW adjustment) as the upper 95% confidence
bound is less than 0%
Difference [95% CI]
TRI
(N=1212)
TFI
(N=527)
Difference
[95% CI]
TFI-TRI
P-value
(superiority)
99.9% 99.0%
-0.9%
[-1.4%; -0.4%]
<0.001
TRI is better TFI is better
-3% 0% 3%
Major bleeding complication-free rate
at 7 days (major 2nd endpoint)
(After IPW adjustment)
Limitation
• Since randomization between TRI and
TFI was made before assessing the
feasibility of CABG or medical therapy,
there might be some bias favoring for
TFI.
Conclusion
• In real world PCI situation with ad-hoc
PCI strategy, TRI is as effective as TFI
in  terms  of  12  months’  MACCE,  and  TRI  
brings less incidence of bleeding
complications at 1 week after PCI.

Saito S DRAGON trial

  • 1.
    Determination of theRAdial versus GrOiN coronary angioplasty The Result of DRAGON Trial Shigeru Saito, MD Department of Cardiology and Catheterization Laboratory Shonan Kamakura General Hospital, Kamakura, Japan on behalf of Dragon Trial investigators
  • 2.
    Disclosure Statement ofFinancial Interest • Grant/Research Support • Consulting Fees/Honoraria • Major Stock Shareholder/Equity • Royalty Income • Ownership/Founder • Intellectual Property Rights • Other Financial Benefit • None • TERUMO, Boston Scientific • None • None • None • None • None Within the past 12 months, I or my spouse/partner have had a financial interest/arrangement or affiliation with the organization(s) listed below. Affiliation/Financial Relationship Company
  • 3.
    Background this study •TRI is getting more popular all around the world. • However, it is not yet clear whether ad-hoc TRI strategy can provide clinical outcomes similar to TFI strategy. • Thus, we initiated DRAGON trial to show the effectiveness of ad-hoc TRI strategy in the real world.
  • 4.
    Objective of thisstudy • To determine the clinical benefit of the transradial approach compared to the transfemoral approach in patients, who receive percutaneous coronary intervention in the real world practice. • The benefit is shown based on the non-inferiority of TRI in terms of clinical effectiveness at 1 year and on the superiority of TRI to reduce the major bleeding complications at 7 days compared to TFI.
  • 5.
    Primary Investigators ShigeruSAITO, MD Yong HUO, MD Honorary Primary Investigator Runlin GAO, MD Ferdinard KIEMENEIJ, MD Steering Committee Shigeru SAITO Junbo GE Yujie ZHOU Guosheng FU Jian’an WANG Bo XU Yong HUO Shubin QIAO Yaling HAN Jiyan CHEN Haichang WANG Wei LI Biostatistic Analysis Medical Research & Biometrics Center National Center for Cardiovascular Diseases Cardiovascular Institute & Fuwai Hospital Sponsor TERUMO CRO/DM/CL CCRF Study Organization
  • 6.
    Endpoints Primary Endpoint  MajorAdverse Cardiac or Cerebrovascular Event (MACCE) free rate at 12 months Major Secondary Endpoint  Major bleeding complication (BARC definition type 3 or 5) free rate at 7 days
  • 7.
    Statistical considerations • Thisstudy had 80% statistical powered to demonstrate the non-inferiority (NI margin=5%) of TRI on primary endpoint (MACCE-free rate at 1 year) comparing to TFI • And also powered (>85%) to demonstrate the superiority of TRI on major secondary endpoint (bleeding-free rate at 7 days) comparing to TFI • Intention-To-Treat (ITT) principle had been used for the primary analysis
  • 8.
    Enrollments Patients who hadPCI were the target population
  • 9.
    Kaplan-Meier curve MACCE-free rateat 12 months (PCI population before adjustment) TRI 1212 1168 1147 1143 1133 1125 1114 TFI 527 494 483 475 467 462 459 HR = 0.707 (95% CI: 0.443-1.129) P-value (log-rank test): 0.144 Pts. at risk TRI TFI
  • 10.
    Primary Endpoint MACCE-free rateat 12 months The non-inferiority was met as the upper 95% confidence bound is less than the non- inferiority margin (5%) Difference [95% CI] TRI (N=1,212) TFI (N=527) Difference [95% CI] TFI-TRI P-value (non- inferiority) 96.1% 94.4% -1.7% [-4.0%; 0.6%] <0.001 TRI is better TFI is better Non-inferiority -5% 0% 5% (PCI population before adjustment)
  • 11.
    Major bleeding complication-freerate at 7 days (major 2nd endpoint) There was no significant difference between the TRI and TFI group on the 2nd endpoint among the unadjusted observed data Difference [95% CI] TRI (N=1212) TFI (N=527) Difference [95% CI] TFI-TRI P-value (superiority) 99.8% 99.4% -0.4% [-1.1%; 0.3%] 0.131TRI is better TFI is better -3% 0% 3% (PCI population before adjustment)
  • 12.
    TRI (N=1,212) TFI (N=527) P-value Male (%) 72.5%64.7% 0.001 Age (yrs) 61.1±10.8 62.9±10.9 0.001 Height (cm) 168.6±6.9 166.9±7.5 <0.001 Weight (Kg) 71.2±10.6 69.0±10.8 <0.001 BMI (Kg/m2) 25.0±3.1 24.6±3.1 0.049 Heart Rate (Beats/m) 72.1±11.8 72.4±13.0 0.626 Current Smoker (%) 37.8% 31.3% 0.033 Baseline characteristics (PCI population before adjustment)
  • 13.
    Disease history (PCI populationbefore adjustment) TRI (N=1,212) TFI (N=527) P-value Hypertension (%) 58.7% 60.9% 0.380 Hyperlipidemia (%) 15.2% 11.8% 0.043 Diabetes (%) 25.0% 24.9% 0.950 Prior MI (%) 11.4% 13.7% 0.185 Prior PCI (%) 12.1% 14.2% 0.231 Prior CABG (%) 0.2% 3.2% <0.001 Cerebrovascular Disease (%) 11.0% 12.3% 0.415 Peripheral vascular Disease(%) 1.4% 0.9% 0.424 Family history of CAD (%) 8.8% 8.3% 0.744
  • 14.
    Lesion characteristics (PCI populationbefore adjustment) TRI (N=1,611) TFI (N=729) P-value Lesion type (%) A 1.9% 2.3% 0.780 B1 25.5% 24.0% B2 7.2% 7.8% C 65.4% 65.8% Bifurcation lesion (%) 8.9% 12.1% 0.018 CTO (%) 5.3% 8.9% 0.001 TIMI 0&1 (%) 22.7% 35.1% <0.001 Number of stents 1.2±0.6 1.2±0.6 0.7007 Stent dia. (mm) 3.0±0.4 3.0±0.4 0.9629 Stent length (mm) 25.3±7.4 25.3±7.5 0.9850 Lesion success (%) 99.7% 99.2% 0.1078
  • 15.
    • The 23baseline covariates had been included in the PS model were pre-specified without any knowledge of clinical outcomes (both efficacy & safety) • HL test had a P=0.857 • IPW (inverse probability weighting) had been carried out Since we unexpectedly found imbalanced randomized allocation, we did propensity score analysis based on 23 baseline covariates. Weintraub WS, et al. Comparative effectiveness of revascularization strategies. N Engl J Med. 2012; 366: 1467-76.
  • 16.
    TRI (N=1,212) TFI (N=527) P-value Male (%) 70.4%70.5% 0.928 Age (yrs) 61.54 ±12.85 61.15 ±20.62 0.467 Height (cm) 168.09 ±8.34 167.71 ±13.65 0.349 Weight (Kg) 70.74 ±12.82 70.39 ±21.40 0.573 BMI (Kg/m2) 24.99 ±3.78 24.97 ±5.83 0.926 Heart Rate (Beats/m) 72.12 ±13.97 72.65 ±24.25 0.378 Current Smoker (%) 37.1% 33.3% 0.012 Baseline characteristics (After IPW adjustment)
  • 17.
    TRI (N=1,212) TFI (N=527) P-value Hypertension (%) 59.0%56.1% 0.084 Hyperlipidemia (%) 13.9% 13.1% 0.458 Diabetes (%) 25.4% 24.4% 0.492 Prior MI (%) 12.2% 11.1% 0.301 Prior PCI (%) 12.3% 13.4% 0.319 Prior CABG (%) 0.8% 1.1% 0.323 Cerebrovascular Disease (%) 10.7% 12.9% 0.039 Peripheral vascular Disease(%) 1.2% 1.2% 0.894 Family history of CAD (%) 7.9% 7.2% 0.482 Disease history (After IPW adjustment)
  • 18.
    TRI (N=1,212) TFI (N=527) P-value Silent Myocardial Ischemia (%) 1.7%1.8% 0.797 Angina (%) 61.0% 62.7% 0.283 MI (%) 37.4% 35.5% 0.248 LVEF (%) 58.63 ±11.24 58.61 ±16.92 0.973 NYHA class (%) Ⅰ 63.3% 64.5% 0.010 Ⅱ 33.0% 29.7% Ⅲ 3.6% 5.8% Ⅳ 0.2% 0.0% Baseline diagnosis (After IPW adjustment)
  • 19.
    TRI (N=1,611) TFI (N=729) P-value TYPE (%) A 1.8%2.0% 0.939 B1 25.2% 25.6% B2 7.7% 7.4% C 65.3% 65.0% Bifurcation lesion (%) 9.4% 9.6% 0.864 CTO (%) 6.1% 6.3% 0.754 Pre-op TIMI 0&1 (%) 23.9% 30.9% Number of stents/lesion 1.2±0.7 1.2±1.1 0.515 Lesion characteristics (PCI population after IPW adjustment)
  • 20.
    TRI (N=1,212) TFI (N=527) P-value Lesion number 1.3±0.71.4±1.2 0.054 Procedural success (%) 99.6% 98.7% 0.007 Operative complication (%) 1.6% 3.3% 0.001 Procedure time (min) 59.1±35.3 60.1±62.3 0.516 Procedural information (After IPW adjustment)
  • 21.
    HR = 0.928(95% CI: 0.672-1.280) P-value (log-rank test): 0.822 Kaplan-Meier curve MACCE-free rate at 12 months (After IPW adjustment) TRI 1212 1168 1147 1143 1133 1125 1114 TFI 527 494 483 475 467 462 459 Pts. at risk TRI TFI
  • 22.
    Difference [95% CI] TRIis better TFI is better Non-inferiority -5% 0% 5% TRI (N=1,212) TFI (N=527) Difference [95% CI] TFI-TRI P-value (non- inferiority) 95.8% 95.5% -0.3% [-1.7%; 1.0%] <0.001 Primary Endpoint MACCE-free rate at 12 months (After IPW adjustment) The non-inferiority was met as the upper 95% confidence bound is less than the non- inferiority margin (5%)
  • 23.
    The superiority ismet (after propensity score- IPW adjustment) as the upper 95% confidence bound is less than 0% Difference [95% CI] TRI (N=1212) TFI (N=527) Difference [95% CI] TFI-TRI P-value (superiority) 99.9% 99.0% -0.9% [-1.4%; -0.4%] <0.001 TRI is better TFI is better -3% 0% 3% Major bleeding complication-free rate at 7 days (major 2nd endpoint) (After IPW adjustment)
  • 24.
    Limitation • Since randomizationbetween TRI and TFI was made before assessing the feasibility of CABG or medical therapy, there might be some bias favoring for TFI.
  • 25.
    Conclusion • In realworld PCI situation with ad-hoc PCI strategy, TRI is as effective as TFI in  terms  of  12  months’  MACCE,  and  TRI   brings less incidence of bleeding complications at 1 week after PCI.