Results of studies using nonvisual method for iPET interpretation
| References . | mFU (mo) . | % of neg . | PPV . | NPV . | PFS according to iPET . | OS according to iPET . | Conclusion of the authors about iPET . |
|---|---|---|---|---|---|---|---|
| Lin et al20 | 42 | Visual: 63%; ΔSUVmax: 82.6% | For EFS visual: 50%, ΔSUVmax: 81.3%; for OS visual: 38.2%, ΔSUVmax: 73.3% | For EFS visual: 74%, ΔSUVmax: 75%; for OS visual: 86.2%, ΔSUVmax: 87% | 2-y EFS visual neg: 79%, visual pos: 51%, ΔSUVneg: 79%, ΔSUVpos: 21% | — | “SUV-based assessment … improves the prognostic value of early 18F-FDG PET compared with visual analysis” |
| Itti et al21 | 41 | Visual: 77%; ΔSUVmax:78.7% | For EFS visual: 77.8%, ΔSUVmax: 70.6%; for OS visual: 55.6%, ΔSUVmax: 53% | For EFS visual: 82.3%, ΔSUVmax: 79.4%; for OS visual: 88.7%, ΔSUVmax: 87.3% | 2-y EFS visual neg: 82%, visual pos: 25%, ΔSUVneg: 79%, ΔSUVpos: 32% | — | “semiquantification … performance is equivalent to visual analysis at 4 cycles” |
| Casasnovas et al13,25 | 19 | After 2 cycles visual: 34%, ΔSUV: 78%; after 4 cycles: visual: 51% ΔSUV: 88% | — | — | 2-y PFS (neg vs pos) visual after 2: 77 vs 73, visual after 4: 81% vs 73%; D5S after 2: 88% vs 79%; D5S after 4: 82% vs 69%; ΔSUV after 2: 77% vs 57%; ΔSUV after 4: 83% vs 40% | 2-y OS (neg vs pos) visual after 2: 93% vs 84%, visual after 4: 94% vs 83% ΔSUV after 2: 93% vs 60% ΔSUV after 4: 50% vs 94% | “These encouraging results suggest the use of ΔSUVmax in addition to visual analysis to interpret iPET … specifically when a therapeutic decision is to be guided by iPET results” |
| Lanic et al33 | — | Visual: 54.3%; ΔSUV: 63.1% | — | — | 2-y PFS ΔSUVmax > 70: 77% ΔSUVmax < 70: 15% estimated | 2-y OS ΔSUVmax > 70: 77% ΔSUVmax < 70: 33% | “Semiquantitative interim PET assessment was highly predictive of the outcome”; “Combination of GEP, aaIPI and interim PET more accurately predicts DLBCL prognosis and is therefore suitable for tailoring therapeutic strategies” |
| Safar et al22 | 38 | ΔSUV: 86% | — | — | 3-y PFS ΔSUV > 66%: 77% ΔSUV < 66%: 38% | 3-y OS ΔSUV > 66%: 82% ΔSUV < 66%: 64% | “Both visual and quantitative evaluations can, however, be improved”; “a centralized review of imaging can also improve evaluation of the PET response and should be encouraged in clinical trials” |
| Pregno et al50 | 26.2 | DS: 72%; ΔSUV: 84.1% | Deauville for PFS 36% | Deauville for PFS 82.5% | 2-y PFS Deauville: 85% vs 72% ΔSUV > 66%: 87% ΔSUV < 66%: 68% | — | “a negative iPET predicts a good outcome”; “positive I-PET is not predictive of a worse outcome in DLBCL” |
| Nols et al51 | 28 | Deauville: 72%; ΔSUVmax: 82% | Deauville for PFS and OS 55% and 50%; ΔSUV for PFS and OS 46% and 46% | Deauville for PFS and OS 81% and 87% ΔSUV for PFS and OS 81% and 87%; 75% and 82% | 2-y PFS ΔSUV > 66%: 78% ΔSUV < 66%: 50% | 2-y OS ΔSUV > 66%: 88% ΔSUV < 66%: 56% | “i-PET was highly and independently predictive of any outcome, and its negative predictive value was improved by combination with IPI” |
| Lee et al7 | 72 | 54% | 53.6% | 93.9% | 5-y PFS 52% vs 80.7% | 5-y OS 56.2% vs 81.5% | iPET “is a significant predictor of PFS and OS” |
| Mylan et al6 | 29 | IHP: 33%; DS: 46% | IHP: 20%; DS: 25% | IHP: 76%; DS: 83% | 2-y PFS: IHP: iPET1 neg: 81.9% iPET1 pos: 77.2% DS (pos if >3): iPET1 neg: 84% iPET1 pos: 77.1% DS (pos if =5): iPET1 neg: 84.8%, iPET1 pos: 50.9% | 2-y OS: IHP: no difference DS (pos if >3): no difference DS (pos if =5): iPET1 neg: 87% iPET1 pos: 57.8% | “PET after one course of chemotherapy was not able to safely discriminate PET-positive and PET-negative patients” |
| Mamot et al23 | 24 | Deauville: 54% | — | — | 2-y EFS Deauville iPET-2: 41% vs 76%; iPET-4: no difference ΔSUV iPET-0-2: ΔSUV > 66%: 61% ΔSUV < 66%: 42% | 2-y OS Deauville iPET-2: 84% vs 94%/ΔSUV iPET-0-3 ΔSUV > 66%: 91.3% ΔSUV < 66%: 73.7% | iPET “has limited prognostic value in patients with diffuse large B-cell lymphoma” and “is not ready for clinical use to guide treatment decisions in individual patients” |
| References . | mFU (mo) . | % of neg . | PPV . | NPV . | PFS according to iPET . | OS according to iPET . | Conclusion of the authors about iPET . |
|---|---|---|---|---|---|---|---|
| Lin et al20 | 42 | Visual: 63%; ΔSUVmax: 82.6% | For EFS visual: 50%, ΔSUVmax: 81.3%; for OS visual: 38.2%, ΔSUVmax: 73.3% | For EFS visual: 74%, ΔSUVmax: 75%; for OS visual: 86.2%, ΔSUVmax: 87% | 2-y EFS visual neg: 79%, visual pos: 51%, ΔSUVneg: 79%, ΔSUVpos: 21% | — | “SUV-based assessment … improves the prognostic value of early 18F-FDG PET compared with visual analysis” |
| Itti et al21 | 41 | Visual: 77%; ΔSUVmax:78.7% | For EFS visual: 77.8%, ΔSUVmax: 70.6%; for OS visual: 55.6%, ΔSUVmax: 53% | For EFS visual: 82.3%, ΔSUVmax: 79.4%; for OS visual: 88.7%, ΔSUVmax: 87.3% | 2-y EFS visual neg: 82%, visual pos: 25%, ΔSUVneg: 79%, ΔSUVpos: 32% | — | “semiquantification … performance is equivalent to visual analysis at 4 cycles” |
| Casasnovas et al13,25 | 19 | After 2 cycles visual: 34%, ΔSUV: 78%; after 4 cycles: visual: 51% ΔSUV: 88% | — | — | 2-y PFS (neg vs pos) visual after 2: 77 vs 73, visual after 4: 81% vs 73%; D5S after 2: 88% vs 79%; D5S after 4: 82% vs 69%; ΔSUV after 2: 77% vs 57%; ΔSUV after 4: 83% vs 40% | 2-y OS (neg vs pos) visual after 2: 93% vs 84%, visual after 4: 94% vs 83% ΔSUV after 2: 93% vs 60% ΔSUV after 4: 50% vs 94% | “These encouraging results suggest the use of ΔSUVmax in addition to visual analysis to interpret iPET … specifically when a therapeutic decision is to be guided by iPET results” |
| Lanic et al33 | — | Visual: 54.3%; ΔSUV: 63.1% | — | — | 2-y PFS ΔSUVmax > 70: 77% ΔSUVmax < 70: 15% estimated | 2-y OS ΔSUVmax > 70: 77% ΔSUVmax < 70: 33% | “Semiquantitative interim PET assessment was highly predictive of the outcome”; “Combination of GEP, aaIPI and interim PET more accurately predicts DLBCL prognosis and is therefore suitable for tailoring therapeutic strategies” |
| Safar et al22 | 38 | ΔSUV: 86% | — | — | 3-y PFS ΔSUV > 66%: 77% ΔSUV < 66%: 38% | 3-y OS ΔSUV > 66%: 82% ΔSUV < 66%: 64% | “Both visual and quantitative evaluations can, however, be improved”; “a centralized review of imaging can also improve evaluation of the PET response and should be encouraged in clinical trials” |
| Pregno et al50 | 26.2 | DS: 72%; ΔSUV: 84.1% | Deauville for PFS 36% | Deauville for PFS 82.5% | 2-y PFS Deauville: 85% vs 72% ΔSUV > 66%: 87% ΔSUV < 66%: 68% | — | “a negative iPET predicts a good outcome”; “positive I-PET is not predictive of a worse outcome in DLBCL” |
| Nols et al51 | 28 | Deauville: 72%; ΔSUVmax: 82% | Deauville for PFS and OS 55% and 50%; ΔSUV for PFS and OS 46% and 46% | Deauville for PFS and OS 81% and 87% ΔSUV for PFS and OS 81% and 87%; 75% and 82% | 2-y PFS ΔSUV > 66%: 78% ΔSUV < 66%: 50% | 2-y OS ΔSUV > 66%: 88% ΔSUV < 66%: 56% | “i-PET was highly and independently predictive of any outcome, and its negative predictive value was improved by combination with IPI” |
| Lee et al7 | 72 | 54% | 53.6% | 93.9% | 5-y PFS 52% vs 80.7% | 5-y OS 56.2% vs 81.5% | iPET “is a significant predictor of PFS and OS” |
| Mylan et al6 | 29 | IHP: 33%; DS: 46% | IHP: 20%; DS: 25% | IHP: 76%; DS: 83% | 2-y PFS: IHP: iPET1 neg: 81.9% iPET1 pos: 77.2% DS (pos if >3): iPET1 neg: 84% iPET1 pos: 77.1% DS (pos if =5): iPET1 neg: 84.8%, iPET1 pos: 50.9% | 2-y OS: IHP: no difference DS (pos if >3): no difference DS (pos if =5): iPET1 neg: 87% iPET1 pos: 57.8% | “PET after one course of chemotherapy was not able to safely discriminate PET-positive and PET-negative patients” |
| Mamot et al23 | 24 | Deauville: 54% | — | — | 2-y EFS Deauville iPET-2: 41% vs 76%; iPET-4: no difference ΔSUV iPET-0-2: ΔSUV > 66%: 61% ΔSUV < 66%: 42% | 2-y OS Deauville iPET-2: 84% vs 94%/ΔSUV iPET-0-3 ΔSUV > 66%: 91.3% ΔSUV < 66%: 73.7% | iPET “has limited prognostic value in patients with diffuse large B-cell lymphoma” and “is not ready for clinical use to guide treatment decisions in individual patients” |