Original Excluder component overlap from proximal or distal extension during initial repair not correlated with aneurysm sac shrinkage
Article Outline
Objectives
The original abdominal Excluder (W.L. Gore & Associates, Flagstaff, Ariz) endoprosthesis has been associated with late aneurysm sac expansion over time from transgraft ultrafiltration of serous fluid. This has been treated by relining the graft with original or low-permeability components. We asked whether additional component overlap of the original graft material resulting from proximal or distal extensions placed at the time of initial repair would influence the rate of late aneurysm sac expansion in the absence of endoleak.
Methods
Computed tomography (CT) scans from subjects (n = 120) receiving the original endoprosthesis from the Excluder pivotal trial were measured for total distance of original graft overlap (including contralateral gate, proximal extension, or distal extension overlap) based on reformatted CT scans. This was compared to change in aneurysm sac diameter and volume (as measured in independent laboratories) at the latest time point available. Patients were omitted if they were missing CT scan data (n = 10), their graft was explanted for endoleak (n = 2), they underwent an intervention for endoleak and did not have diameters available after their intervention (n = 3), or if they had a continued endoleak that could account for an increase in aneurysm sac diameter (n = 11). This left 27 patients with more overlapping components than the required contralateral limb/gate overlap (mean follow-up time 40.6 ± 17.0 months) and 67 patients with required gate overlap (mean follow-up time 46.2 ± 15.9 months).
Results
Subjects with increased component overlap (mean overlap 87.1 mm ± 57.4 mm) were not protected from aneurysm sac expansion when compared to those with the minimum required gate overlap (mean overlap 31.2 mm ± 3.4 mm). There was no association of total distance of overlap with aneurysm sac size change by diameter or volume (r2 = 0.00034, P = .86 for diameter and r2 = 0.0019, P = .68 for volume). Increasing percentage of overlap within the aneurysm sac was likewise not associated with aneurysm sac decrease in diameter (r2 = 0.0028, P = .61). Few patients had large percentages of original graft overlap (mean 26.2% ± 14.1% for the increased overlap group and 18.6% ± 5.5% for the required overlap group, P = .0097).
Conclusion
Partial graft overlap involving multiple original components from proximal and distal extensions is not protective against aneurysm sac expansion due to transgraft ultrafiltration. This suggests that transgraft ultrafiltration is not impeded by having partial double layers of original material. All patients who received the original Excluder and have late aneurysm sac expansion in the absence of endoleak should have as complete relining as feasible with low permeability components if sac shrinkage is the surrogate goal.
The original abdominal Excluder endoprosthesis (W.L. Gore & Associates, Flagstaff, Ariz) was initially approved by the Food & Drug Administration (FDA) in 2001 following a pivotal controlled trial demonstrating safety and efficacy in aneurysm exclusion while preventing aneurysm rupture.1 Longer term follow-up demonstrated that over a third of the 30,000 patients who received this endoprosthesis underwent late aneurysm sac expansion, most in the absence of any detectable endoleak.2, 3, 4 A better understanding of the factors that lead to delayed sac expansion would help management of the thousands of patients worldwide who received the original device.5
Explanted devices were found to be surrounded by proteinaceous acellular material,6 suggesting that ultrafiltration of serum was occurring through the original expanded polytetrafluroethylene (ePTFE) graft. These findings resulted in a modification of the endoprosthesis graft material to include a low-porosity film layer. The modified device was commercially introduced in July 2004 and appears in midterm studies to have less associated sac expansion than the original device.7
Goodney and Fillinger6 noted that no proteinaceous material was found around the contralateral gate graft overlap in explanted original material grafts and theorized that a double layer of graft material might be protective of the ultrafiltration effect. A recently described technique to detect endoleak using delayed magnetic resonance (MR) contrast imaging also did not demonstrate endoleak in regions of graft overlap.8 Relining either with original or low-porosity components has been employed successfully to decrease aneurysm sac expansion in patients who received the original endoprosthesis,6, 9 supporting the idea that increased graft-graft overlap within the aneurysm sac is protective against the ultrafiltration effect. It remains unclear why some patients develop delayed aneurysm sac expansion and others do not.
We hypothesized that patients who received components in addition to the main body and contralateral limb (either proximal or distal extensions at the time of initial implantation), and thereby had a higher percentage of graft-graft overlap would have a decreased risk for aneurysm sac expansion. In order to test this hypothesis, we analyzed computed tomography (CT) scans from patients who received the original device in the Excluder pivotal trial for length of graft overlap and compared this to the change in diameter or volume of the aneurysm sac.
Methods
We analyzed orthogonal slice reformations at 1 mm intervals (Preview M2S, previously Medical Metrix Solutions, West Lebanon, NH) of CT scans from 120 of the 273 patients who had the original Excluder implanted in the Excluder pivotal trial. This was a randomly chosen subset used as the control group in the controlled trial against the low-porosity graft, and therefore had reformatted CT scans available for analysis. Overlap was defined as at least two pieces of graft within one another. All patients had overlap between the contralateral gate of the main body device and the contralateral docking limb. Additional overlap occurred from implanted components such as any proximal extension or distal limb extension placed within the main body/ipsilateral limb component or within the contralateral limb. Each patient was measured for total distance of graft overlap in millimeters of the entire treated segment. If the length could not be directly measured from the reformatted CT scans (secondary to poor image quality, generally resulting from reformatted CT scans obtained from scanned cut films or from large slice intervals), the overlap length was derived from the devices known to be implanted into that patient and the measured overall length covered by stent graft.
Graft overlap length for the entire treated segment was compared to change in maximal aneurysm sac diameter (as measured by the core laboratory during the original trial) at the latest time point available (see Table I, Table II for the follow-up time points for each patient) or aneurysm sac volume in cubic centimeters (as measured from three-dimension [3-D] CT scan reformations) at the latest time point available (Table I, Table II). If a patient had undergone a secondary intervention which resolved an endoleak, the change in aneurysm sac diameter was calculated as the latest time point diameter minus the sac diameter on the scan immediately preceding the secondary intervention (n = 15).
Table I. Increased overlap group data
| Subject | Age | Antiplt/anticoag | Devices | Interventions/other features | Overlap | % Sac overlap | Orig sac size (mm) | Diam change (mm) | F/U (mo) | Vol change (cc) | F/U (mo) |
|---|---|---|---|---|---|---|---|---|---|---|---|
| 1 | 74 | No | R ipsi, R extend, prox extend | Extend | 104 | 34.3% | 58.4 | 6 | 27.9 | 34.1 | 0 |
| 2 | 69.7 | No | L ipsi, L extend × 3 | 140 | 8.4% | 54.4 | 8.2 | 60.4 | 19.3 | 23.8 | |
| 3 | 54.9 | Yes | R ipsi, R extend | 67 | 24.5% | 54.8 | −2.3 | 59.9 | −29.6 | 23 | |
| 4 | 70.6 | No | L ipsi, prox extend × 2 | 63 | 21.6% | 67.5 | 5.4 | 57.8 | −45.1 | 25.3 | |
| 5 | 58 | No | R ipsi, R extend | Embolized | 75 | 28.2% | 70.3 | 5.2 | 40 | 68.7 | 3.2 |
| 6 | 78.8 | No | R ipsi, prox extend | Embolized | 63 | 43.0% | 71.2 | 1.8 | 12.9 | 92.4 | 0 |
| 7 | 75.8 | No | R ipsi, L extend | Embolized | 39 | 13.2% | 56.5 | 14.5 | 24.1 | 73.8 | 24.1 |
| 8 | 73.3 | Yes | R ipsi, prox extend | 61 | 24.0% | 55.6 | 11.9 | 59.5 | 21.8 | 23.6 | |
| 9 | 71.8 | No | R ipsi, L extend | 75 | 34.8% | 48.8 | −8.1 | 23.5 | −26.6 | 23.5 | |
| 10 | 63.2 | No | L ipsi, R extend | 83 | 18.3% | 51.0 | −7.7 | 60.5 | −21.1 | 23.9 | |
| 11 | 71.1 | No | L ipsi, B extend (R 14 × 12, L 10 × 7) | Branch lumbar | 152 | 47.1% | 57.0 | −2.8 | 60.6 | −34.2 | 24.1 |
| 12 | 79.1 | Yes | R ipsi, L extend | 71 | 26.0% | 53.1 | 2.7 | 60.2 | 13.5 | 23.3 | |
| 13 | 72.7 | Yes | R ipsi, R extend | Unknown type | 62 | 22.6% | 53.7 | −6.1 | 48.1 | no data | 24 |
| 14 | 73.7 | No | R ipsi, R extend | 68 | 20.9% | 49.2 | −3.5 | 61 | −2.7 | 24.7 | |
| 15 | 75 | No | R ipsi, prox extend | 62 | 25.5% | 59.9 | 6.8 | 23.4 | 125.8 | 23.4 | |
| 16 | 83 | No | R ipsi, prox extend | 60 | 17.3% | 58.7 | 4.2 | 35.7 | 7.7 | 24.8 | |
| 17 | 71.1 | No | R ipsi, R extend | Embolized | 69 | 15.2% | 53.1 | 11.9 | 41.2 | 27.2 | 5.7 |
| 18 | 72.8 | Yes | R ipsi, prox extend | Explant (no leak) | 55 | 12.9% | 57.0 | 3.6 | 36.1 | 10 | 24.2 |
| 19 | 60.7 | No | L ipsi, prox extend | Explant (no leak) | 55 | 27.5% | 54.5 | 20.5 | 26.7 | 184.4 | 26.7 |
| 20 | 65.3 | Yes | L ipsi, R extend | 79 | 33.8% | 47.7 | −1.4 | 49.1 | 1.6 | 23.5 | |
| 21 | 77.9 | No | R ipsi, R extend | 73 | 14.8% | 74.4 | −18.7 | 35.8 | −92.2 | 23.4 | |
| 22⁎ | 74.6 | Yes | L ipsi, limb extend | 35 | 22.6% | 68.5 | 1.9 | 60.9 | −6.6 | 23.9 | |
| 23 | 72.4 | Yes | L ipsi, R extend | Embolized | 50 | 19.0% | 59.9 | 3 | 13.6 | 16.2 | 0 |
| 24 | 80 | No | R ipsi, R extend | 79 | 24.3% | 55.3 | −17.7 | 36.4 | −77.6 | 24.5 | |
| 25 | 69 | No | R ipsi, L extend | 105 | 35.8% | 51.7 | 7.6 | 59.3 | −4.6 | 24.2 | |
| 26 | 65.1 | No | R ipsi, R extend | Embolized | 75 | 13.3% | 48.9 | −0.9 | 10.9 | 15.1 | 10.9 |
| 27 | 64.4 | No | R ipsi, prox extend, R extend × 2, L extend | 342 | 80.8% | 57.2 | 0 | 23.9 | −93.9 | 23.9 | |
| 28 | 78.8 | No | L ipsi, L extend × 2 | Embolized | 124 | 24.1% | 51.6 | 1.2 | 48.1 | −8.2 | 11.7 |
| Mean | 71.2 | 87.1 | 26.2% | 57.1 | 1.7 | 41.3 | 10.0 | 19.2 | |||
| St Dev | 6.9 | 57.4 | 14.1% | 7.2 | 8.6 | 17.1 | 61.4 | 9.0 |
⁎Unable to determine limb overlap due to incomplete CT scan, but aneurysm sac overlap percentage available. |
Table II. Required overlap group data
| Subject | Age | Antiplt/anticoag | Devices | Interventions/other features | Overlap | % Sac overlap | Orig sac size (mm) | Diam change (mm) | F/U (mo) | Vol change (cc) | F/U (mo) |
|---|---|---|---|---|---|---|---|---|---|---|---|
| 29 | 69 | No | R ipsi | 29 | 20.1% | 47.2 | −7.5 | 60 | −27 | 24.2 | |
| 30 | 65.1 | Yes | R ipsi | 33 | 22.0% | 52.1 | 0 | 60.1 | 16.2 | 24.3 | |
| 31 | 84.3 | No | R ipsi | Embolized | 30 | 14.2% | 67.6 | −2.8 | 13.2 | 31.8 | 13.2 |
| 32 | 54.5 | No | R ipsi | 33 | 19.2% | 63.4 | 0 | 61.2 | −27.1 | 25 | |
| 33 | 71.4 | No | R ipsi | Explant (no leak) | 28 | 13.1% | 58.0 | 7.5 | 49.6 | 4.8 | 24.1 |
| 34 | 82.5 | Yes | R ipsi | 33 | 18.3% | 62.4 | 24.1 | 59.3 | 10.7 | 12.5 | |
| 35 | 70 | Yes | R ipsi | Type II resolved | 29 | 14.7% | 56.4 | −0.1 | 58 | 2.9 | 23.2 |
| 36 | 78.3 | No | R ipsi | 32 | 23.8% | 82.3 | −5.7 | 23.4 | −5.1 | 23.4 | |
| 37 | 84.5 | Yes | R ipsi | 22 | 12.4% | 52.7 | 13.2 | 60.9 | −85.3 | 24.6 | |
| 38 | 72.3 | Yes | R ipsi | Embolized | 30 | 17.3% | 69.9 | 5.4 | 45.5 | 10.7 | 0 |
| 39 | 69.2 | No | R ipsi | 28 | 24.1% | 50.7 | 5.8 | 60.7 | 11.8 | 24.4 | |
| 40 | 54.3 | Yes | R ipsi | 31 | 27.3% | 48.5 | 17.2 | 63.1 | −17.5 | 24.7 | |
| 41 | 80.5 | Yes | R ipsi | 35 | 20.7% | 54.5 | −1.4 | 52.8 | 1.9 | 11.8 | |
| 42 | 85.7 | No | R ipsi | Embolized | 29 | 17.2% | 69.0 | 17.9 | 38.7 | 40 | 13.9 |
| 43 | 71.6 | Yes | L ipsi | 28 | 19.2% | 55.6 | −20.9 | 60.9 | −33.9 | 24.6 | |
| 44 | 91.3 | Yes | L ipsi | 30 | 22.7% | 54.5 | 0.5 | 61.3 | 5.2 | 24.8 | |
| 45 | 83.8 | No | R ipsi | 32 | 15.7% | 66.9 | −2.1 | 24.3 | −10.5 | 24.3 | |
| 46 | 76 | No | R ipsi | 32 | 15.3% | 50.2 | 3.4 | 23.8 | 0.2 | 23.8 | |
| 47 | 74.1 | No | L ipsi | 30 | 21.3% | 50.4 | −6 | 48.5 | −7.3 | 23.8 | |
| 48 | 57.5 | No | R ipsi | 31 | 17.0% | 49.0 | 7.9 | 64.8 | 9.4 | 24.2 | |
| 49 | 73.7 | No | R ipsi | 34 | 26.3% | 54.0 | −5.2 | 48.6 | 0.5 | 24.6 | |
| 50 | 76 | No | L ipsi | 34 | 37.8% | 53.3 | −3.1 | 60.1 | 2.7 | 23.7 | |
| 51 | 67.5 | No | R ipsi | 32 | 16.1% | 50.2 | 3.3 | 61.7 | −24.6 | 24.5 | |
| 52 | 75.5 | No | R ipsi | 36 | 22.5% | 58.0 | 14.5 | 60.1 | 22.5 | 24.9 | |
| 53 | 64.4 | No | R ipsi | 27 | 12.3% | 59.9 | 11.2 | 62.6 | 9.1 | 24.6 | |
| 54 | 68 | No | R ipsi | 35 | 18.6% | 60.4 | −0.8 | 60.1 | −26.1 | 23.7 | |
| 55 | 78.8 | Yes | R ipsi | 35 | 21.7% | 62.6 | 1.6 | 60 | 2.8 | 24.4 | |
| 56 | 71.4 | No | R ipsi | 31 | 16.7% | 48.3 | 3.4 | 27.1 | 8.8 | 27.1 | |
| 57 | 66.5 | Yes | R ipsi | 35 | 22.9% | 61.8 | 6.7 | 60 | 77.6 | 23.9 | |
| 58 | 69.2 | Yes | R ipsi | 38 | 24.4% | 45.7 | 2.8 | 60.1 | −0.9 | 23.9 | |
| 59 | 65.6 | No | R ipsi | 30 | 19.9% | 48.5 | 9.1 | 60.1 | 0.2 | 23.6 | |
| 60 | 74.9 | Yes | R ipsi | 32 | 10.9% | 44.2 | −24.3 | 60.3 | 5.7 | 23.8 | |
| 61 | 71.7 | No | R ipsi | 26 | 13.0% | 68.8 | −39 | 59.8 | −53.3 | 23.7 | |
| 62 | 83.2 | Yes | R ipsi | 28 | 14.7% | 65.1 | 7.6 | 64.6 | −3 | 25.4 | |
| 63 | 62.4 | Yes | R ipsi | Type II resolved | 26 | 13.4% | 51.3 | −8.5 | 23.9 | −70.5 | 23.9 |
| 64 | 78.5 | Yes | R ipsi | Embolized | 26 | 10.9% | 49.6 | −11.1 | 22.9 | 50.9 | 0 |
| 65 | 62.7 | No | L ipsi | 32 | 18.6% | 64.4 | −5.8 | 26.2 | −44.9 | 26.2 | |
| 66 | 77.2 | Yes | R ipsi | Embolized | 27 | 4.8% | 60.0 | −8.8 | 33.3 | 11.5 | 0 |
| 67 | 66.8 | No | R ipsi | Embolized | 23 | 4.4% | 71.4 | 4.6 | 28.5 | 10.1 | 16.1 |
| 68 | 85.7 | Yes | R ipsi | Embolized | 32 | 20.4% | 72.0 | 7.1 | 28.3 | 16.8 | 15.4 |
| 69 | 66.1 | No | R ipsi | 33 | 18.6% | 62.3 | −1.6 | 23.3 | 5.9 | 23.3 | |
| 70 | 75.2 | No | R ipsi | 27 | 14.5% | 73.1 | −1.6 | 23.9 | 3 | 23.9 | |
| 71 | 74.2 | Yes | L ipsi | 33 | 20.4% | 52.6 | −18.2 | 23.9 | −35.8 | 23.9 | |
| 72 | 66.5 | Yes | L ipsi | 35 | 24.0% | 68.5 | −2.6 | 48.6 | 8.1 | 13.1 | |
| 73 | 67.1 | No | R ipsi | 28 | 15.2% | 52.0 | −2.6 | 25.6 | 4.2 | 25.6 | |
| 74 | 74.9 | Yes | L ipsi | 37 | 24.5% | 53.8 | −9.5 | 61.3 | −25 | 28 | |
| 75 | 79.1 | No | L ipsi | 36 | 23.1% | 59.0 | −7.5 | 49.2 | −9.6 | 25.2 | |
| 76 | 74.8 | Yes | R ipsi | 28 | 16.1% | 69.8 | 3.1 | 24.6 | 10.9 | 24.6 | |
| 77 | 74 | Yes | R ipsi | 39 | 26.9% | 52.0 | −13.2 | 46.8 | −0.3 | 21 | |
| 78 | 68.2 | Yes | L ipsi | 32 | 21.0% | 52.1 | 3.9 | 46.7 | 8.2 | 22.9 | |
| 79 | 77.7 | No | R ipsi | Embolized | 30 | 23.4% | 54.3 | −4.8 | 31.5 | −4.4 | 6.4 |
| 80 | 87.1 | No | R ipsi | Embolized | 32 | 18.0% | 54.2 | 18.5 | 59.4 | 32.8 | 20.9 |
| 81 | 48.9 | Yes | R ipsi | 31 | 15.6% | 47.0 | −1.3 | 59.4 | −44.9 | 11.7 | |
| 82 | 78.4 | Yes | R ipsi | 29 | 19.1% | 51.9 | −1.1 | 23.4 | 6.3 | 23.4 | |
| 83 | 69.8 | No | L ipsi | 29 | 13.7% | 64.3 | −1.7 | 61.2 | −28.9 | 24.4 | |
| 84 | 83.2 | Yes | R ipsi | Explant (no leak) | 33 | 15.9% | 57.1 | 13.4 | 24.2 | 64.2 | 24.2 |
| 85 | 80.8 | Yes | 29 | 8.3% | 72.1 | 1.5 | 38.2 | 25.9 | 24.6 | ||
| 86 | 68.9 | Yes | R ipsi | 30 | 18.4% | 61.7 | 4.5 | 59.9 | −3 | 24 | |
| 87 | 78.9 | No | R ipsi | Embolized | 33 | 22.9% | 47.1 | 1.6 | 46.9 | 4.8 | 11.1 |
| 88 | 71.4 | Yes | R ipsi | 29 | 14.5% | 49.8 | 4.5 | 59.9 | −1.2 | 24.4 | |
| 89 | 71.5 | No | R ipsi | 33 | 18.2% | 61.8 | −3.9 | 59.7 | −48.5 | 24.1 | |
| 90 | 70.3 | Yes | R ipsi | 31 | 16.0% | 50.0 | 0.7 | 23.1 | −9.4 | 23.1 | |
| 91 | 76.6 | Yes | L ipsi | 34 | 25.8% | 46.9 | −11.9 | 61 | −33.9 | 22.4 | |
| 92 | 52 | Yes | R ipsi | 32 | 13.4% | 57.8 | 1.7 | 24.3 | −1.1 | 24.3 | |
| 93 | 80.2 | Yes | R ipsi | 34 | 24.3% | 61.5 | 20.3 | 50.4 | 43.3 | 24 | |
| 94 | 69.8 | No | R ipsi | Unknown | 32 | 23.2% | 58.3 | 14.8 | 36.3 | 3.3 | 23.7 |
| 95 | 70.1 | No | R ipsi | Type II resolved | 37 | 22.3% | 50.5 | 17.9 | 44.7 | 28.2 | 24.2 |
| Mean | 69.8 | 33.3 | 20.8% | 54.2 | 7.3 | 40.0 | 5.1 | 23.6 | |||
| St Dev | 9.7 | 2.2 | 5.0% | 5.8 | 12.5 | 14.9 | 27.4 | 0.7 |
Additionally, the percentage of overlap within the aneurysm sac was calculated by measuring the length of overlap within the aneurysm sac and dividing the result by the total length of device within the aneurysm sac (comprised of the length of main body within the sac added to the length of each leg within the sac). The total surface area of the graft within the aneurysm sac was estimated by adding p*d*h (surface area of a cylinder where d = diameter of the graft and h = length of each component) for the main body and each limb. Surface area of overlap within the aneurysm sac was estimated by adding p*d*h of each region of graft overlap. The percentage of surface area overlapped within the aneurysm sac was then estimated by dividing the surface area of overlap by the total surface area of the graft.
Statistical analysis was performed using t test for proportions for the analysis of patients on antiplatelet or anticoagulation therapy with increasing or decreasing sac diameters and for the comparison of percentage of graft overlap within the aneurysm sac between the required overlap and increased overlap groups. Spearman rank correlations were performed for comparisons between groups with sac overlap and diameter (or volume) change as the variables. An exploratory P value of less than .05 was considered to be significant.
Results
Patients were chosen for analysis if they had reformation CT scan follow-up for at least 12 months. Patients were excluded from the analysis if they were missing CT scan data (n = 7) or if the reformation CT scan did not include the distal end of the stent graft (n = 2), their graft was explanted for endoleak (n = 2), they underwent an intervention for an endoleak and did not have diameters available after their intervention (n = 3), or if they had a continued endoleak of any type that could account for an increase in aneurysm sac diameter (n = 11). This left 27 patients with more overlapping components than the required contralateral limb/gate overlap (increased overlap group) and 67 patients with the minimum required overlap (required gate overlap group). Nine patients had proximal extensions (9.6% of the total 94 patients analyzed) and 19 patients had limb extensions (20.2% of the analyzed patients).
An increase in graft overlap length was not correlated with a decrease in maximal aneurysm sac diameter, r2 = 0.00034, P = .86, degrees of freedom = 92 (Fig 1). The mean graft overlap length for the increased overlap group was 87.1 ± 57.4 and for the required gate overlap group was 31.2 ± 3.4. As aneurysm sac volume has been found to be more sensitive than sac diameter for changes in aneurysm sac morphology,4, 10 we also compared length of graft overlap with aneurysm sac volume which was available at 24 months for the majority of patients. One patient with more than the required overlap had diameter but not volume measurements available and was omitted from the volume analysis, leaving 25 patients with increased overlap and 66 patients with required overlap. An increased length of graft overlap was likewise not correlated with a decrease in aneurysm sac volume, r2 = 0.0019, P = .68, degrees of freedom = 91 (Fig 2).
Fig 1.
Original Excluder aneurysm sac diameter change not correlated with increased overlap. Length of overlap of patients (n = 67) with baseline overlap from the contralateral gate (white circles) and patients (n = 27) with additional overlap from extra components (black squares) plotted against change in aneurysm sac diameter does not show any correlation between increased length of overlap and decrease in aneurysm sac diameter.
Fig 2.
Original Excluder aneurysm sac volume change not correlated with increased overlap. Length of overlap of patients (n = 67) with baseline overlap from the contralateral gate (white circles) and patients (n = 26) with additional overlap from extra components (black squares) plotted against change in aneurysm sac volume does not show a correlation between increased length of overlap and decrease in aneurysm sac volume.
Based on explant analysis, ultrafiltration appears to occur primarily when the original graft material is within the aneurysm sac and not when the original graft material is opposed to the vessel wall. Therefore, scans were also analyzed for the length of graft overlap in contact with the aneurysm sac as a percentage of total length of graft within the aneurysm sac to derive a percentage of aneurysm sac overlap. One of the 2 patients whose reformatted CT scan did not include the distal end of the stent graft was measured for the percentage of overlap within the aneurysm sac, although the total length of overlap was not able to be measured or derived. The mean percentage of aneurysm sac overlap for the increased overlap group was 26.2% ± 14.1% and the required gate overlap group was 18.6% ± 5.5% (P = .0097). Only 3 patients had more than 40% graft overlap within the aneurysm sac. The results are presented in Fig 3; there is no correlation with decrease in aneurysm sac diameter (r2 = 0.0028, P = .61, degrees of freedom = 93). The percentage of aneurysm sac overlap derived from estimating surface areas was highly correlated with that calculated using lengths, r2 = 0.87, suggesting that using overlap length is a reasonable surrogate measure for vulnerable graft surface.
Fig 3.
Percentage aneurysm sac overlap not correlated with sac diameter change. Patients with additional overlap from baseline (n = 26, black squares) who had higher percentage of graft overlap within the aneurysm sac were not protected from aneurysm sac expansion. They did have a significantly higher percentage of graft overlap (26.2% ± 14.1% vs 18.6% ± 5.5%, P = .0097) than patients with baseline overlap from the contralateral gate (white circles). Few patients had extensive graft overlap.
Anticoagulation or antiplatelet therapy at the time of discharge was recorded for the trial patients. There was no significant difference between the percentage of patients on antiplatelet/anticoagulation therapy at the time of discharge whose aneurysm increased in diameter (49.0%) or decreased in diameter (41.5%, P = .47). Table I includes details of the patients who had additional overlap. Table II includes details of the patients who had the required overlap.
Discussion
Partial graft overlap from multiple original Excluder components placed during initial endovascular aortic aneurysm repair did not protect patients from aneurysm sac expansion, suggesting that other patient-related factors are responsible for which patients will develop delayed aneurysm sac expansion. Neither aneurysm sac diameter nor volume was associated with length of graft overlap. The percentage of overlap of total graft length within the aneurysm sac was likewise not correlated with change in aneurysm sac diameter. The hypothesis that graft overlap should be protective against aneurysm sac expansion was not fully tested, however, as the sample lacked a significant proportion of patients with large lengths of overlap.
A probable explanation for these findings is that the majority of the overlap in this group of patients was not in contact with the aneurysm sac where the ultrafiltration effect may be assumed to be the greatest. The average percentage of actual overlap within the aneurysm sac was only 26.2% ± 14.1% for the group of patients with more than gate overlap, with 10 of 27 patients having no additional overlap beyond the contralateral gate within the aneurysm sac. This was because these patients received additional components as proximal or distal extensions. Therefore, these findings do not contradict that graft relining, where more than 90% of the portion of the graft within the aneurysm sac is overlapped, may be an effective strategy for preventing or treating ultrafiltration.
Graft relining strategies have included use of an aortic cuff and two iliac limbs, two iliac limbs deployed up into the main body of the original endoprosthesis, and two iliac limbs deployed into the two limbs of the original endoprosthesis.6, 9 Kougias et al9 described 3 patients who had aneurysm sac expansion in the absence of demonstrable endoleak after implantation of the original Excluder device. Sac pressure measurements from a translumbar approach were equivalent to systemic pressures. All 3 patients were treated by relining using an aortic cuff and two iliac limbs, 2 with low-porosity components and 1 with original components. All 3 had sac shrinkage with 12-month follow-up.9 Goodney and Fillinger6 described their experience of relining 9 patients with aneurysm sac expansion in the absence of demonstrable endoleak after implantation of the original Excluder endoprosthesis. All three of the above strategies were used. They reported >6-month follow-up for 5 patients, all of whom had aneurysm sac shrinkage or stabilization in size after relining. The remaining 4 patients had less than 6-month follow-up, and the effect of relining on their aneurysm sac size was not reported. Taken together with the data from the present study, there is compelling rationale that, when indicated, relining should be performed with the goal to reline as much endograft surface within the aneurysm sac as is safely feasible.
Antiplatelet or anticoagulation therapy at the time of discharge did not correlate with aneurysm sac expansion, suggesting that this is not a factor affecting ultrafiltration. Unfortunately, data regarding anticoagulation alone was not available for these patients, preventing us from analyzing this as an independent variable. Another possible patient-related factor would be dynamic motion at the proximal neck, although this would result in an intermittent type I endoleak with thrombus rather than proteinaceous material found within the sac.
A possible hypothesis is that fluid exchange across the original graft membrane is a dynamic equilibrium, with more fluid exchanged out into the sac in the patients with aneurysm sac expansion then returned into the systemic circulation. Patients who had sac regression would have more fluid return than egression. Patients with stable sac size would have equivalent amounts exchanged in and out. In this hypothesis, having additional graft overlap might bias the fluid exchange rate towards absorption over time. The results presented here, however, suggest that small amounts of original graft overlap are not sufficient to cause sac regression, and that small amounts of increased overlap was not among the patient-related factors which promoted sac regression in this cohort.
Although sac expansion due to ultrafiltration has for the most part been benign,11 several complications are theoretically possible. The most significant is missing another type of endoleak as the true cause for aneurysm sac expansion. Nevertheless, aneurysm sac rupture, venous compression, ureteral obstruction, bowel obstruction, and loss of proximal or distal seal zone4 are also potential complications from continued sac expansion.
Short areas lined by double layers of original material do not impede transgraft ultrafiltration sufficiently to prevent aneurysm sac expansion. As with all patients who undergo endovascular aneurysm repair (EVAR), those who received the original Excluder need to have continued long term surveillance. Patients who have late aneurysm sac expansion that is deemed to require reintervention and have no detectable endoleak by thorough imaging investigation should have as complete relining as feasible with low permeability components.
Author contributions
The authors would like to thank Erin Tims, Bonny Tab, Phillip Cancino, and Jeffrey Bell of W.L. Gore & Associates for their assistance in obtaining reformatted CT scans and core lab data.
References
- . A multicenter controlled clinical trial of open versus endovascular treatment of abdominal aortic aneurysm. J Vasc Surg. 2003;37:262–271
- Abdominal aortic aneurysm size regression after endovascular repair is endograft dependent. J Vasc Surg. 2003;37:716–723
- . Late abdominal aortic aneurysm enlargement after endovascular repair with the Excluder device. J Vasc Surg. 2004;39:1236–1241discussion 2141-2
- . Three-dimensional analysis of enlarging aneurysms after endovascular abdominal aortic aneurysm repair in the Gore Excluder Pivotal clinical trial. J Vasc Surg. 2006;43:888–895
- Associates WL Gore Excluder bifurcated endoprosthesis annual clinical update. Flagstaff, Ariz: WL Gore & Assoc.; February 2006;
- . The effect of endograft relining on sac expansion after endovascular aneurysm repair with the original-permeability Gore Excluder abdominal aortic aneurysm endoprosthesis. J Vasc Surg. 2007;45:686–693
- . Outcomes of original and low-permeability Gore Excluder endoprosthesis for endovascular abdominal aortic aneurysm repair. J Vasc Surg. 2007;45:243–249
- . Visualizing type IV endoleak using magnetic resonance imaging with a blood pool contrast agent. J Vasc Surg. 2008;47:861–864
- . Successful treatment of endotension and aneurysm sac enlargement with endovascular stent graft reinforcement. J Vasc Surg. 2007;46:124–127
- . Maximal aneurysm diameter follow-up is inadequate after endovascular abdominal aortic aneurysm repair. Eur J Vasc Endovasc Surg. 2000;20:177–182
- . Nonoperative approach to endotension. J Vasc Surg. 2005;42:194–199
Competition of interest: Dr Fillinger has grant, research, or consulting relationships with W.L. Gore & Associates, Medtronic, and Preview M2S. Dr Matsumura has grant, research, or consulting relationships with W.L. Gore & Associates, Medtronic, Cook, Cordis, ev3, Lumen, Bard, and Abbott Vascular.
PII: S0741-5214(09)00491-1
doi:10.1016/j.jvs.2009.02.209
Published by Elsevier Inc.
