Outcomes Using Aprotinin Therapy

1
Outcomes Using Aprotinin Therapy
Randomized Trials vs. Observational Studies

• Linda Shore-Lesserson, M.D.
• Associate Professor
• Department of Anesthesiology
• Montefiore Medical Center
• Bronx, New York

2
Disclosure

• Worked with aprotinin since 1993 (compassionate
  use protocol)
• Clinical research- Bayer support
• Consultant- TMC, Abbott, AZ, Bayer
• Speakers Bureau- TMC, Abbott, Bayer, ARC
• McSPI member
• PI for McSPI project on antifibrinolytic Rx

3
Major Points

• Cardiac surgery becoming ?? complex
• Preponderance of RCTs show reduced blood loss and
  transfusion and reops
• STS/SCA Guidelines
• Clinical experience agrees with RCT
• Observational studies cannot control for ALL
  confounders, esp if unmeasured or unknown.

4
Meta-Analysis

• Randomized trials, double blind CABG
• 72 trials by Med Search
• 35 trials (n3887) listed adverse events
• 29 used full dose aprotinin

Sedrakyan et al JTCVS 2004128442-8
5
Events/Number
Transfusion 793/1966 936/1464 Mortality
53/2149 39/1630 MI
96/2024 77/1531
Renal Failure 26/1755 16/1248 Stroke
19/1714 28/1262 A Fib
320/1408 263/1052
Sedrakyan et al JTCVS 2004128442-8
6
Rise of Catheter Based Therapies
7
Meta-Analysis in CPB

• Medline and Cochrane database
• 1999-2004
• Clopidogrel within 7days vs. control
• Blood loss, transfusions, AEs, vent time, LOS,
  death, reops
• (n4002), n605 clop, n3397 control

Purkayastha S Heart 200692531-532
8
Clopidogrel Group

• ? Blood loss
• ? Transfusion requirements
• ? AEs
• ? LOS
• ? Re-exploration
• ? Ventilator time

Purkayastha S Heart 200692531-532
9
Meta-Analysis in CPB
WMD/OR Heterog p
CTD (ml) 323ml 137-510 Y lt0.01
Tx req (U) 1.36U 0.8-1.92 Y lt0.01
Tx risk () 4.9 2.79-8.59 Y lt0.01
Reop () 6.76 3.37-13.56 N lt0.01
LOS (d) 1.18d 0.24-2.12 Y 0.01
Purkayastha S Heart 200692531-532
10
Blood Loss in CABG
n73, plt0.001
Van der Linden Circulation 2005112I276-80
11
Transfusion in CABG
Plt0.04
Van der Linden Circulation 2005112I276-80
12
CABG After ACS

• ACS- unstable angina or NSTEMI
• CABG performed after at least 5 days
• Randomization (n49)
• Continue clop and ASA, HD aprotinin (mean drug Rx
  time 17d)
• D/C drugs 5d prior, heparin IV until 30mins

Akowuah et al Ann Thorac Surg 200580149-52
13
Bleeding and Transfusion
Aprotinin Clop Heparin Rx no Clop p
8H CTD (ml) 265169 385273 0.036
Total CTD (ml) 447287 702495 0.004
CT time (hr) 19.66.8 22.87.7 0.004
RBC (U) 0.321.2 1.01.7 0.03
Akowuah et al Ann Thorac Surg 200580149-52
14
Clinicians Do Not Agree With Findings of this
Observational Study

• Unusual circumstances of authorship given the
  history of McSPI

15
(No Transcript)
16
(No Transcript)
17
(No Transcript)
18
(No Transcript)
19
CABG-Surgery in Europe and North America
Timelines and Outcomes

• Ott E, Moehnle P, Tudor C, Hsu PH, Mangano DT

Anesthesiology 200399A254 (Submitted, in
revision, JTCVS)
20
Outcome UK (n 619) Canada (n 444) USA (n 1,283) Germany (n 8 34) p value
In-hospital Mortality 1.5 2.0 2.7 3.8 0.034
Cardiac morbidity 9.2 12.4 13.6 18.5 lt 0.001
Morbidity and/or mortality 12.4 15.5 18.0 23.9 lt 0.001
Aprotinin () 23.1 5.7 19.5 69.3 lt0.001
FFP in OR () 2.4 1.4 8.4 10.6 lt0.001
Ott E et al Anesthesiology 200399A254
21
(No Transcript)
22
Renal Composite Outcomes
Risk Factor OR OR p propensity P value
Aprotinin 2.52 2.411.49-3.90 lt0.001
Hx renal dz 2.5 2.531.70-3.75 lt0.001
Creat gt 1.3 2.71 3.122.11-4.60 lt0.001
CHF adm 2.33 2.641.84-3.80 lt0.001
FFP admin 2.51 2.401.58-3.66 lt0.001
Hx liver dz 0.35 0.280.13-0.61 lt0.001
Mangano DT et al NEJM 200634353-65
23
(No Transcript)
24
Summary

• Cardiac surgery has become increasingly complex.
• Hemostasis improved by aprotinin
• RCTs and clinical impression of clinicians
  support that end-organ outcomes are either
  unchanged or improved with aprotinin.

25
Summary

• Observational study captures that aprotinin is
  selected in higher risk patients. (Mangano 2006)
• There is a true association of morbid outcomes
  with aprotinin because they are co-linear.
• Even the best propensity matching cannot control
  for this selection bias.

26
Conclusion

• Covariates not evaluated (or not published) in
  the NEJM study
• Country, FFP use, CPB time, aspirin use
• Covariates not measured in NEJM 2006
• Surgeon expertise, regional techniques
• To suggest cause and effect (ischemic events)
  from observational data is irresponsible
• To suggest use of cheaper alternatives unstudied
  and not labelled for this use is irresponsible