ebm journal club sarah jean strube, d.o. resident physician st. mary medical center october 17, 2008

EBM Journal Club

Sarah Jean Strube, D.O.Resident Physician

St. Mary Medical Center

October 17, 2008

Overview Topic Background Defining the patient Article

Methods Results Statistics Author’s conclusion

Statistics discussion

Topic

Reperfusion injury status post PCI in acute myocardial infarction with and without cyclosporine injection on area of infarction.

Background

Myocardial infarction is a disabling disease and infarct size is considered a major determining factor for mortality. Limiting the size of an infarct through reperfusion therapy is an important strategy in decreasing morbidity whether through thrombolysis or PCI. However, reperfusion has its own detrimental effects through several mechanisms. One of which is via mitochondrial dysfunction.

Background

Mitochondrial dysfunction has been termed “permeability transition.” It is the opening of a nonspecific channel in the inner membrane of the mitochondria. This transition results in uncoupling of the respiratory chain and collapse of the inner mitochondrial membrane potential with subsequent efflux of proapoptotic factors causing myocardiocyte death.

BACKGROUND

Cyclosporine is mostly known for its immunosuppressive effects.

However, it has been found by several researchers in experimental models to have potent inhibiting effects on mitochondrial permeability transition and may prevent ischemia/reperfusion injury.

PICO Question

PatientInterventionControlOutcome

PICO question

Patient: A 65 YO male with a Hx of HTN, dyslipidemia and prior tobacco use presented to the Emergency Department with prolonged angina pectoris. Onset of his CP was six hours prior to admission and he was found to have an acute ST-segment elevation in two contiguous precordial leads along with elevation of CK and troponin I. An acute myocardial infarction was Dx and he was considered a candidate for urgent PCI.

PICO QUESTION

Intervention: Administration of cyclosporine via IV bolus at time of urgent PCI but prior to stenting of an occluded artery (TIMI flow 0) in an acute ongoing myocardial infarction

PICO Question

Control: NS bolus during PCI with stenting alone prior to the procedure in an occluded artery (TIMI flow 0).

PICO question

Outcome: did the intervention with pretreatment with cyclosporine vs normal saline decrease the area of myocardial infarction SP PCI?

Article selected

Effect of cyclosporine on reperfusion injury in acute myocardial infarction. New England Journal of Medicine

2008; 359: 473-481.

Research objective:

To evaluate the efficacy of a therapy

The article

Original research, pilot studyProspective, multicenter, randomized,

single blind, controlled trialJournal - peer reviewed, general Internal

Medicine, highly respected Sites – multicenterPatients – 58 randomly assigned, 30 CS

28 control

Patient Criteria INCLUSION

Male/female 18y or older present wi 12h of CP

STEMI 0.1 mV 2 cont leads

PCI eligible TIMI flow grade 0

at time of admission

EXCLUSIONCardiac arrestCardiogenic shockVent fibrillationStent thrombosisPrevious MIAngina wi 48hOccl LM/Circ or

collateralsHypersensitivity to

cyclosporine

Other exclusions:

Renal or liver failureUncontrolled hypertensionPregnancyWomen of childbearing age not on

contraceptionAny Ds of immunologic dysfunction; CA,

HIV, hepatitis

Study Population

July 2005 to October 2006 340 patients at three centers were hospitalized for management of acute MI

Approximately 80% men, mean age 58y 230 underwent PCI, 24 not enrolled -

inadequate manpower; 148 excluded, see below

Baseline characteristics of subjects were similar

Study Population

Similar in ischemia time, myocardium at risk and EF prior to PCI - MRI

Thombolytic therapy failed in 13 patients prior to PCI, 8 in control, 5 in CS

Culprit lesion stented in all patients and only infarct related lesions treated

Four patients, TIMI 2 flow was not achieved SP PCI

Study PopulationBaseline Characteristics

CSMen/women - 25/5 mean age 58 +/- 2yBMI mean 26 +/- 1, dyslipidemia 14, HTN 15,

DM 4HX CAD 4Tobacco 17

ControlMen/women – 21/7 mean age 57 +/- 2yBMI mean 27 +/- 1, dyslipidemia 12, HTN 13 ,

DM 4HX CAD 4Tobacco 16

Piot C et al. N Engl J Med 2008;359:473-481

Baseline Characteristics of the Study Population

Methods

Randomization: after coronary angio, before stent, a computer generated sequence assigned patients to receive placebo vs cyclosporine

Intervention: IV bolus of cyclosporine 2.5mg/kg of BW,

control given equivalent volume in NS

Blood Concentration of Cyclosporine during Reperfusion

Piot C et al. N Engl J Med 2008;359:473-481

Statistical Analysis

Calculated target sample size of 62 pts based on prior trial, 31 per group

Hypothesized that CS would reduce the AUC for CK release by 30% for a power of 80%

Probability of a type I error of 0.05 using a two sided test

Between group comparisons for AUC for trop, CK, area at risk, and infarct size by MRI evaluated with Wilcoxon rank-sum

Statistical Analysis

Analysis of covariance performed on the equality of slopes on the regression of infarct size on the area at risk in CS and control

Comparison of incidence of cumulative adverse events between groups using Fisher exact test

END POINTS

Primary : size of the infarct assessed by measurements of cardiac biomarkers

Secondary : size of infarct measured by area of hyper-enhancement seen on cardiac MRI, assessed day 5

END POINTS

Other: major adverse events first 48h including death, MI, heart failure, stroke, recurrent ischemia, renal/liver insufficiency, vascular complications, and bleeding

RESULTS

The cyclosporine and the control group were similar in ischemia time, area of myocardium at risk and EF prior to PCI

RESULTS

Assessment of infarct size by biomarkers:CK release sig decreased in CS group

vs control group over time (P=0.04)

Trop I not sig decreased in CS group vs control group over time (P=0.15)

RESULTS

Infarct size as a function of area at riskFor any given area at risk CS

administration was associated with a reduction infarct size as measured by CK/trop I release (P=0.006) / (P=0.002)

Assessment of Infarct Size by Biomarker Measurement

Piot C et al. N Engl J Med 2008;359:473-481

Infarct Size as a Function of the Area at Risk

Piot C et al. N Engl J Med 2008;359:473-481

MRI or CMR

MRI has been used since 1984 on imaging of the heart and recently improved technology with contrast enhancement improves delineation of hyper-enhanced regions (acute MI)

MRI

Bright is dead

RESULTS

Subgroup analysis 27 patients:

Infarct size (absolute mass) decreased on MRI day 5 in CS group 37 g vs 46 g control group (P=0.04)

Area of infarction: E A X slice thick X M sd

Kim R et al. N Engl J Med 2000;343:1445-1453

Typical Cine Image and Contrast-Enhanced Image Obtained by MRI before Revascularization

Kim R et al. N Engl J Med 2000;343:1445-1453

Typical Contrast-Enhanced Images Obtained by MRI in a Short-Axis View (Upper Panels) and a Long-Axis View (Lower Panels) in Three Patients

Assessment of Infarct Size by Magnetic Resonance Imaging (MRI)

Piot C et al. N Engl J Med 2008;359:473-481

Author’s Conclusion

The effect of CS in this small pilot study of patients having an acute myocardial infarction undergoing PCI, showed a decrease in infarct size as measured by release of CK and delayed hyper-enhancement on MRI.

Trop I was not significantly reduced by CS

Evaluation

Methods: randomized, but unclear if truly similar in

ischemia timeBlinded to full extent allowableAll patients accounted forSubgroup was noted beforehand (MRI) and

not added later because of Tx effectsSmall population but was a pilot study

Evaluation

OutcomesResults definitely applied to my patientResults are meaningful, however difficult to

know if truly affected mortality CS did show Tx effect especially in MRI

group with proven validity of acute MI hyper-enhancement

Statistics Discussion

To reviewBetween group comparisons for AUC for trop,

CK, area at risk, and infarct size by MRI evaluated with Wilcoxon rank-sum

AUC for normal data is a Gaussian distribution and the usual parametric stats can be used

With non normal data (continuous or ordinal data) nonparametric stats like Wilcoxin rank sum can be used

Wilcoxin Rank Sum

A descriptive nonparametric statistic using non normal data. Similar to performing a two sample t testWhy use Wilcoxin?

Appropriate for small populationEasier to interpret ordinal or continuous dataNo assumption of population distributionMore robust

Wilcoxin Rank Sum

DisadvantagesLess sensitiveLess powerNot appropriate for large N

Wicoxin Rank Sum

The procedure:1. Arrange observations for both groups into a

single rank series

2. Add up the ranks for both series

3. The rank sum is then divided by the number of observations

4. Observe the rank sum difference, as the magnitude tells you how close the groups are

Wicoxin Rank Sum

Example Imagine choosing an

Olympic Team of Karate experts from two states, CA and NV. Your decision is based on how many boards each athlete can break in 5 minutes

Statistics in a Nutshell

CA NV4 25 36 36 47 48 59 109 109 11

CA NV Rank

2 1

3 2

3 3

4 4

4 5

4 6

5 7

5 8

6 9

CA NV Rank

6 2 10

7 11

8 3 12

9 13

9 4 14

9 10 15

5 10 16

10 17

11 18

CA NV Rank

2 1

3 2.5

3 2.5

4 5

4 5

4 5

5 7.5

5 7.5

6 9.5

CA NV Rank

7 11

8 12

9 14

9 14

9 14

10 16.5

10 16-5

11 18

Wilcoxin Rank Sum

Sum the ranks

E (CA) = 5 + 7.5 + 9.5 + 9.5 +11 +12 +14 + 14 +14 = 96.5

E (NV) = 1 + 2.5 +2.5 +5 + 5+ 7.5 +16.5 + 16.5 +18 = 74.5

So I choose the California team to go to the Olympics

Statistics in a Nutshell

Questions?

Original Article

Effect of Cyclosporine on Reperfusion Injury in Acute Myocardial Infarction

Christophe Piot, M.D., Ph.D., Pierre Croisille, M.D., Patrick Staat, M.D., Hélène Thibault, M.D., Gilles Rioufol, M.D., Ph.D., Nathan Mewton, M.D., Rachid Elbelghiti,

M.D., Thien Tri Cung, M.D., Eric Bonnefoy, M.D., Ph.D., Denis Angoulvant, M.D., Christophe Macia, M.D., Franck Raczka, M.D., Catherine Sportouch, M.D., Gerald Gahide, M.D., Gérard Finet, M.D., Ph.D., Xavier André-Fouët, M.D., Didier Revel,

M.D., Ph.D., Gilbert Kirkorian, M.D., Ph.D., Jean-Pierre Monassier, M.D., Geneviève Derumeaux, M.D., Ph.D., and Michel Ovize, M.D., Ph.D.

N Engl J MedVolume 359(5):473-481

July 31, 2008