Table 9: Study results for diagnostic accuracy of PET in predicting regional functional recovery after revascularization*.
| Author, Year | Viability Threshold | No. Dysfunctional Segments | Sensitivity (%) | Specificity (%) | PPV (%) | NPV (%) | Positive LR | Negative LR | Diagnostic Accuracy (%) |
|---|---|---|---|---|---|---|---|---|---|
| Kuhl et al., 2006 (30)† | Normal perfusion: Tc uptake >70% Flow/metabolism mismatch: ≤70% uptake of Tc-sestamibi, >70% uptake of FDG, and ≥20% difference between FDG and Tc-sestamibi uptake | 187 ‡ 29 (successfully revasc.) |
86.5 69.2 |
73.6 93.8 |
77.6 90.0 |
83.8 78.9 |
3.28 11.08 |
0.18 0.33 |
80.2 82.8 |
| Slart et al., 2006 a (31) | Flow/metabolism mismatch: a difference of 7% in relative segmental uptake (from ROC curve analysis) | 264 (successfully revasc.) | 90.9 | 86.8 | 89.0 | 89.0 | 6.9 | 0.10 | 89.0 |
| Slart et al., 2006 b (32) | FDG uptake ≥50% 10% WT (gated PET) |
213 (successfully revasc.) 213 (successfully revasc.) |
94.0 89.5 |
85.0 77.5 |
91.2 86.9 |
89.5 81.6 |
6.3 4.0 |
0.07 0.14 |
90.6 85.0 |
| Barrington et al., 2004 (33) | FDG uptake ≥ 68% Normal perfusion or flow-metabolism mismatch |
31 vascular territories 28 vascular territories |
100.0 100.0 |
92.0 86.4 |
75.0 66.7 |
100.0 100.0 |
12.5 7.3 |
0.00 0.00 |
93.5 89.3 |
| Schmidt et al., 2004 (34) | Normalized FDG uptake ≥50% of reference segment uptake | 40 patients (successfully revasc.) | 100.0 | 73.3 | 86.2 | 100.0 | 3.75 | 0.00 | 90.0 |
| Nowak et al., 2003 (35) | Normalized FDG uptake ≥ 70% | 72 (successfully revasc.) | 80.0 | 71.9 | 78.0 | 74.2 | 2.8 | 0.28 | 76.4 |
| Normalized FDG uptake ≥ 60% | 72 (successfully revasc.) | 90.0 | 62.5 | 75.0 | 83.3 | 2.4 | 0.16 | 77.8 | |
| Normalized FDG uptake ≥ 50% | 72 (successfully revasc.) | 95.0 | 21.9 | 60.3 | 77.8 | 1.2 | 0.23 | 62.5 | |
| Normalized FDG uptake ≥ 80% | 72 (successfully revasc.) | 55.0 | 84.4 | 81.5 | 60.0 | 3.5 | 0.53 | 68.1 | |
| Lund et al., 2002 (37) | Normalized FDG uptake > 55% | 34 patients (successfully revasc.) | 88.9 | 68.0 | 50.0 | 94.4 | 2.8 | 0.16 | 73.5 |
| Tani et al., 2001 (39) | FDG uptake ≥ 50% of the maximum uptake | 110 (successfully revasc.) 91 (successfully revasc)§ |
89.9 88.1 |
61.0 62.5 |
79.5 81.3 |
78.1 74.1 |
2.3 2.4 |
0.17 0.19 |
79.1 79.1 |
| Wiggers et al., 2000 (40) | FDG uptake ≥70% of that of the reference region for the patient | 314 (successfully revasc) | 79.2 | 52.1 | 25.1 | 92.5 | 1.7 | 0.40 | 56.7 |
| Fath-Ordoubadi et al., 1999 (41) | MRG ≥ 0.25 umol/g/min | 51 (successfully revasc.) 63 |
96.7 97.0 |
90.5 76.7 |
93.5 82.1 |
95.0 95.8 |
10.2 4.2 |
0.04 0.04 |
94.1 87.3 |
| Schoder et al., 1999 (42) | Relative FDG and ammonia uptake > 2 SD above the normal mean (>15 of the vascular territory must be hypoperfused) | 107 vascular territories (successfully revasc) | 93.0 | 81.3 | 87.0 | 89.7 | 5.0 | 0.09 | 88.0 |
| Zhang et al., 1999 (50) | Flow-metabolism mismatch (semi-quantitative assessment, no threshold reported) | 101 | 75.9 | 86.0 | 88.0 | 72.5 | 5.4 | 0.28 | 80.2 |
| Pagano et al., 1998 (43) | MRG ≥ 0.25 umol/min/g | 336 (successfully revasc.) | 99.0 | 33.3 | 66.4 | 96.0 | 1.5 | 0.03 | 70.8 |
| Maes et al., 1997 (44) | Ratio of metabolic and flow index > 1.2 | 23 patients | 83.3 | 90.9 | 90.9 | 83.3 | 9.2 | 0.18 | 87.0 |
| Baer et al., 1996 (45) | FDG uptake ≥ 50% of the maximal uptake | 371 (successfully revasc,)|| | 92.8 | 66.0 | 72.0 | 90.6 | 2.7 | 0.11 | 79.0 |
| ≥50% of akinetic/dyskinetic segments related to an infarct region had uptake ≥50% of maximal uptake | 42 patients (successfully revasc)|| | 96.2 | 68.8 | 83.3 | 91.7 | 3.1 | 0.06 | 85.7 | |
| Gerber et al., 1996 (46) | Flow-mismatch: ammonia uptake < 70% and ratio of FDG to ammonia > 1.2 | 39 patients (successfully revasc.) | 75.0 | 66.7 | 78.3 | 62.5 | 2.3 | 0.38 | 71.8 |
| vom Dahl et al., 1996 (47) | Flow-metabolism mismatch: Tc uptake ≤ 70%, FDG uptake > 50% and FDG minus Tc uptake > 20% | 48 vascular territories (successfully revasc.) | 92.0 | 34.8 | 60.5 | 80.0 | 1.4 | 0.23 | 64.6 |
| Grandin et al., 1995 (51) | Flow-metabolism mismatch: segmental FDG to ammonia activity ratio > 1.2 | 25 patients | 64.7 | 50.0 | 73.3 | 40.0 | 1.3 | 0.71 | 60.0 |
| Segments with flow <80% of maximal flow and flow-metabolism mismatch: segmental FDG to ammonia activity ratio > 1.2 | 17 patients | 88.9 | 50.0 | 66.7 | 80.0 | 1.8 | 0.22 | 70.6 | |
| Carrel et al., 1992 (48) | Flow-metabolism mismatch (no threshold reported) | 23 patients (successfully revasc.)¶ | 94.1 | 50.0 | 84.2 | 75.0 | 1.9 | 0.12 | 82.6 |
| Marwick et al., 1992 (49) | Flow-metabolism mismatch: avid FDG uptake in presence of hypoperfusion at rest | 73 segments 23 patients |
61.3 86.7 |
83.3 62.5 |
73.1 81.3 |
74.5 71.4 |
3.7 2.3 |
.46 0.21 |
74.0 78.3 |
*
FDG refers to F-18-fluorodeoxyglucose; LVEDV, left ventricular end-diastolic volume; LVESV, left ventricular end-systolic volume; LR, refers to likelihood ratio; MI, myocardial infarction; MRG, rate of metabolic glucose utilization; NPV, negative predictive value; PPV, positive predictive value; revasc., revascularized; Tc, technetium; WT, wall thickness
†
Sensitivity, specificity, PPV, and NPV values reported in this table vary slightly from those reported in the text of the published report because a different equation was used in the report to calculate these values.
‡
Only segments with severely reduced function (wall motion score ≥ 3) were included in the analysis
§
Results restricted to only those patients who had a previous MI (excluding acute MI patient population)
||
Results for dyskinetic/akinetic segments only
¶
Patient defined as viable or not viable based on 1 segment/patient that with abnormal contractility on preoperative ECHO and clearly PET-documented blood-flow metabolism match or mismatch