MPI Risk Stratification Flashcards
MPI better for prognosis than diagnosis.
Allows you to estimate probability of CAD but cannot make definitive statement that there is or is not obstructive disease present.
Prognosis generally refers to death (all-cause, or cardiac death) or MI
K
Problem with studies assessing MPI sensitivity and specificity: Referral bias (aka Verification bias)
Only positive studies go to Cath. So you exclude most of the false-negative MPIs.
Net effect: artificially inflates sensitivity (makes sense…bc those tests would fail to pick up disease when it’s actually present).
And lowers specificity (why?)
Use “normalcy” metric to get around this–
I’m cohort that’s likely normal, how does test perform?
Meaning how reliably do you get negative studies in this group (which are presumably true-negatives).
This gets at accurate SPECIFICITY of the test (bc positives are presumably false-positives)
True statement: MPI has a very high normalcy rate (91%).
Better than sensitivity (89%) or specificity (75%)
K
The following all improve specificity (bc they minimize false positives):
Gating
Attenuation correction
Prone imaging (also upright imaging gets rid of diaphragm as well);
Using higher count isotope i.e. Tc instead of Thallium (less attenuation or scatter of photons)
Gating: to call something artifact, i.e. rule out scar, it needs to THICKEN in addition to seeing the wall move inward;
tethering can cause scar to move inward;
If it thickens, it’s alive.
Thickening is seen as BRIGHTENING and color images show this best
(Boards: AC improves SPECIFICITY, not sensitivity)
K
For DIAGNOSIS: stop meds
For PROGNOSIS: continue meds
K
MPI risk stratification
Low risk: <1%/year (death or MI)
Intermediate: 1-3%
High: >3%
Normal MPI scan (unless high-risk pt features present) confers <1%/year risk (not zero risk…approx 0.6%/year)
Follow-up in these studies was 2-3 years (“warranty period”)
Stress-only study (if normal) is just as reliable for low risk prognostication
Normal PHARM study has higher associated risk than this.
Those people are overall Low Risk but >1%/year (1.8%)
The following are subgroups where risk is >1%/year even if NORMAL study:
Pharm stress
Elderly (>85)…>3%/year
DM…1.9%/year (both men and women)
Maybe true in CKD pts as well…research not conclusive
K
Warranty period of normal MPI
We usually say risk remains low for 2-3 years
warranty period is only 1 year:
History of CAD (risk >1% in year 2 (approx 1.5%)
Also: the more exercised achieved (higher peak HR), the longer the warranty period (<130 bpm, warranty only around 1 year, maybe 2)
DM pts shorter warranty period
Boards: If no high-risk pt features, then warranty period is 2-3 years:
1) no DM
2) < 80 yo
3) NO CAD hx
4) achieved >80% MPHR
K
Predicting events based on extent of defects
Based exclusively on OBSERVATIONAL, retrospective studies
(except acute CP imaging in ER)
SSS: infarction and ischemia, predicts death (VF/CHF) AND MI
SDS: extent of ischemia, predicts MI
defect SEVERITY, not size, predicts prognosis
For normal EF pts, mild and moderate severity defects have similar prognosis, and only severe defect modified prognosis (eg SDS > 7)
If EF <50%, each step up in defect severity will sig alter prognosis
K
Duke Treadmill Score
Minutes - 5 x max ST depression - chest pain (0,1,2)
Chest pain:
0–none
1–present but not reason for stopping
2–present and reason for stopping
DTS 5+: low risk (99% survival at 4 yr)
DTS btw 4 and -10: intermediate (95% survival)
DTS < -10: high risk (75% survival)
Approx 50% of people are INTERMEDIATE DTS
Additive prediction of MPI over DTS is most useful in intermediate DTS:
significant diff in risk for normal MPI, vs mildly abnormal, vs severely abnormal in these pts
(and very low risk if normal scan, same as low risk DTS with normal scan)
If low risk DTS but severely abnormal MPI, that’s also where scans add a lot (they’re high risk)
It also adds, but less so, if mildly abnormal scan in this DTS group (1.8% risk)
THP: scans add a lot in low and intermediate DTS groups
But in high risk DTS, even a normal scan does not significantly reduce risk for events (approx 3%/year)
K
EF and SSS powerful predictors of DEATH (VF or CHF)
SDS best predictor of MI
SSS predicts both (makes sense), though not as powerful a predictor of MI as SDS
K
Predicting benefits from revascularization
Total ischemic myocardium:
>10-12.5%–revasc better than meds
<10-12.5%–meds better (mortality higher with revasc here–why?)
Different if pts have large scar in addition to ischemic territory (makes sense)
If scar <10% of myocardium, then those with >10-12.5% ischemia do better with revasc;
If scar >10% myocardium, they do t benefit from revasc over meds
K
COURAGE study
Had Cath first, and excluded ostial (vs ? prox) LAD pts–so few of them had large ischemic myocardium on MPI
Nuclear substudy
NOT randomized (who happened to have a pre and post MPI study for some reason
Findings:
1) More reduction in ISCHEMIA (as measured by % reduction in ischemic myocardium) with revasc vs meds, in those with moderate-severe ischemia on pre study
2) event rates were lower (including death) in those with greater reduction in % ischemic myocardium on 2nd scan vs those with greater reductions in ischemia
But didn’t PROVE that % reduction in ischemia predicts outcome, since not randomized study
And didn’t PROVE that revasc improves outcomes in these pts (i.e. With moderate to severe ischemia on baseline scan)
K
LBBB
Can see septal defect with vasodilator too but less often.
HR dep’t phenomenon–if tachy response to vasodilator you may see it.
LBBB artifact can involve AW but SPARES THE APEX
(If apex involved, it’s LAD ischemia)
Polar plots helpful
K
TID
Boards: cause is diffuse subendocardial ischemia
Ratio > 1.12 for single isotope, exercise
Ratio > 1.36 for vasodilator pharm stress
TID very predictive of extensive CAD and bad prognosis IF ALSO PERFUSION ABNORMAL
If normal perfusion, not specific for MVDz and not predictive of bad prognosis
TID in absence of MVDz:
Hypertension with LVH, esp if very high BP response to stress
Differences in isotope doses, acquisition parameters
Probably shouldn’t cath just bc TID (i.e. if perfusion is normal), unless high risk pt features
K
ST changes with vasodilator stress, normal perfusion scans
Uncommon finding
Typically see it in females
Only even consider it if normal baseline ECG
If nonspecific ST-T at rest, ignore it
Unclear if predicts CV risk–studies show disparate findings
K
Post-CABG <5 yrs,
Post-PCI < 2 yr:
AUC says “Rarely Appropriate” to do MPI in asymptomatic pts
K
