Abstract
Introduction
Isolated common femoral endarterectomy was recently reported to have a thirty-day mortality of 3.4%. The impact of adjunctive femoral endarterectomy at the time of lower extremity bypass is not well described and therefore the purpose of this study is to determine its associated perioperative and long-term risk.
Methods
Patients undergoing initial lower extremity bypass in the VSGNE from 2003–2015 were identified. After univariate analysis, multivariable logistic regression was used to identify the independent association of endarterectomy with adverse perioperative events. Kaplan-Meier and Cox hazard models were utilized for 1-year analysis.
Results
After exclusions 4496 patients were identified as undergoing infrainguinal bypass (33% with endarterectomy). There was no difference in proportion with chronic limb-threatening ischemia (CLI)(68% vs. 67%, P=.24), or tissue loss (of those with CLI: 65% vs. 63%, P=.34), between the adjunctive endarterectomy group and bypass alone, respectively. Patients undergoing adjunctive endarterectomy were older (mean 68yr vs. 67yr, P=.02), more likely white (95% vs. 93%, P=.02), smokers (91% vs. 87%, P=.001), and more often had prior CABG/PCI (34% vs. 31%, P=.02). The endarterectomy cohort had similar 30-day mortality (CLI: 2.6% vs. 2.9%, P=.60; Claudication: 0.2% vs. 0.4%, P=1.0) despite longer operative time (median 268min vs. 210min, P<.001) and increased blood loss (median 250cc vs. 180cc, P<.001). Patients with CLI undergoing adjunctive endarterectomy had more in-hospital MIs (6.2% vs. 3.8%, P=.003) and transfusions (11% vs. 6.8%, P<.001). At one-year this group had a suggestion of improved freedom from major amputation (91% vs. 87%, P=.049) and AFS (80% vs. 76%, P=.03) that did not reach significance after adjustment. For patients with claudication and adjunctive endarterectomy rates of MI (2.4% vs. 0.9%, P=.02), renal dysfunction (3.6% vs.1.4%, P=.01), SSI (5.0% vs. 2.6%, P=.02), and transfusion (4.6% vs. 1.8%, P=.002) were higher. After adjustment, all patients undergoing adjunctive endarterectomy were at increased risk of MI (OR 1.6, 95%CI 1.1–2.2), SSI (1.5, 1.1–2.0), and bleeding requiring transfusion (1.8, 1.4–2.3). There were no differences in one-year survival for either CLI or claudication groups and no difference in all one-year endpoints, for patients with claudication.
Conclusion
Adjunctive femoral endarterectomy with bypass is safe, with no difference in perioperative or 1-year mortality compared to bypass. However, surgeons should be aware that adjunctive endarterectomy is associated with increased risk of bleeding, surgical site infection, and myocardial infarction, likely from these patients’ disease burden and presumed more extensive atherosclerosis.
Introduction
Endarterectomy has been utilized to treat peripheral arterial disease for over 50 years and is a common adjunctive procedure during lower extremity bypass. The benefit of common femoral endarterectomy has been established, for both inflow and outflow improvements.[1–5] Some of these studies analyzed isolated common femoral endarterectomy and others reported on endarterectomy as part of a hybrid procedure. A recent study from the National Surgical Quality Improvement Program (NSQIP) evaluating isolated common femoral endarterectomy reported a 3.4% thirty-day mortality. [6] Siracuse et al, using NSQIP data from 2007–2010, found a thirty-day mortality rate of 1.5% for isolated common femoral endarterectomy but differed from the former in that it did not include emergent cases.[7] It is likely that many of the patients getting isolated common femoral endarterectomy, especially in the emergent setting, were too sick for a more extensive procedure, such as bypass. Neither NSQIP study analyzed the impact of adjunctive endarterectomy with bypass. Furthermore, bundled CPT coding could complicate this type of analysis using NSQIP data. However, the subset of patients undergoing adjunctive endarterectomy may be at increased perioperative risk in general given the presumably more extensive atherosclerosis that leads to the endarterectomy itself. It is therefore important to quantify this increased risk for appropriate risk-adjustment, especially as potential quality metrics for bypass are proposed.
Therefore, this study aimed to compare the perioperative and one-year morbidity and mortality associated with adjunctive femoral endarterectomy in lower extremity bypass using data from the Vascular Study Group of New England (VSGNE).
Methods
We performed a retrospective cohort study of all patients undergoing lower extremity bypass in the VSGNE from 2003 through 2015 (N= 9144 operations). VSGNE is a prospectively collected clinical registry created by vascular surgeons in New England that collects both short and long-term data on all patients from participating hospitals with the aim or improving regional outcomes. Data in the VSGNE are collected by both the surgeon and trained nurses or clinical abstractors, as described previously, on over 100 clinical and demographic variables.[8]
Patients and Cohorts
Of the 9144 operations identified only a patients’ initial procedure in the registry was used to create our cohorts (excluding 1398 subsequent contralateral and ipsilateral operations). From the 7746 patients we excluded all emergent (N=276), asymptomatic or acute limb ischemia indications (N=960), and any bypasses that did not originate in the common femoral or profunda femoris arteries (N=1950). The VSGNE has a surgeon-entered variable specifying whether a concurrent proximal ipsilateral endarterectomy was performed with the bypass, which was used to create our cohorts. There was a clear reporting change in 2009, with 1.6% (N = 49) of patients missing data on adjunctive endarterectomy in the later period, compared to 64% before 2009. Those missing endarterectomy data before 2009 were presumed to have not undergone an adjunctive endarterectomy. Figure 1 illustrates the steady rate of adjunctive endarterectomy after this assumption was made. After exclusion of the 49 patients missing this variable from 2009 to 2015 we were left with 4496 patients undergoing a bypass. We performed a sensitivity analysis on years 2009–15 utilizing the same models for adjustment at both thirty-day and one-year for all endpoints discussed and found similar results. The purpose of our analysis was to evaluate the associated risk in patients undergoing adjunctive endarterectomy and so we felt it more important to keep patients prior to 2009 in this study, and perform a sensitivity analysis, in order to have as much power as possible to detect a difference, rather than just focus the entire study on the 2009 – 2015 cohort.
Figure 1.
Proportion of patients undergoing adjunctive endarterectomy with bypass by year
A case was considered urgent in the VSGNE if non-elective surgery occurred 12 to 72 hours after admission. Outside of mortality all perioperative outcomes reported are for inhospital adverse events due to the limits of the registry and, similarly, all long-term outcomes represent 1-year results. Renal function deterioration was defined as a > 0.5mg/dl increase from baseline creatinine or new dialysis. Prolonged postoperative length of stay was defined as discharge greater than 7 days after the bypass. Major adverse limb event (MALE) was defined as subsequent thrombectomy, lysis, or revision of bypass, or major ipsilateral amputation. Major ipsilateral amputation includes both above-knee and below-knee major amputation. The registry does not distinguish between these at 1-year.
Statistical Analysis
Categorical variables were presented as counts and percentages. Continuous variables were presented as mean ± standard deviation, or as median and interquartile range as appropriate. Univariate differences between cohorts were assessed using χ2 and Fisher’s exact tests for categorical variables and Student’s t-test and Mann Whitney U test for continuous variables, where appropriate. Univariate outcomes were stratified by claudication and CLI. Adjusted analyses for in-hospital adverse events were performed using multivariable logistic regression. Due to limited event numbers multivariable models were combined, and adjusted for, patients with claudication and CLI symptoms. Purposeful selection was used to populate the above models (using significance of P < .10 for initial inclusion as well as important covariates from prior studies) to test for the independent association of adjunctive endarterectomy with outcomes of interest.[9] Event rates were listed for both perioperative and 1-year outcomes to demonstrate each model had an adequate number of adverse events to allow for inclusion of adjusted for confounders.[10] Our sensitivity analysis used all variables identified in the overall models in order to adjust for the same potential confounders. Operative time was not included in the models because of collinearity with surgery type. Only patients eligible for 1-year follow-up were included for long-term analysis; eligibility was defined as a patient with an operative date 9 or more months prior to date of data extraction, which for our analysis was August 2015. For 1-year endpoints we used the previously published definitions from the Society for Vascular Surgery and performed time to event analysis using Kaplan Meier and Life Table analysis with the log-rank test for univariate testing and Cox Proportional Hazards for adjusted analysis.[11] For Cox analysis visual inspection of the log-log plot was used to verify that the baseline proportional hazards assumption did not cross, meaning change over time, between comparison groups. For Kaplan Meier and Life Table analysis a standard error of >.10 was used to truncate reporting of data. Mortality data in the VSGNE are verified using the Social Security Death Index.
The Beth Israel Deaconess Medical Center Institutional Review Board approved this study and waived informed consent due to the use of de-identified data.
Results
Of the total 4496 patients undergoing lower extremity bypass, 1464 (33%) had an adjunctive endarterectomy. The proportion of patients undergoing adjunctive endarterectomy over time was stable (see Figure 1).
Patient Characteristics
There was no difference in urgent vs. elective cases (urgent: 20% vs. 20%, P=.99), proportion for chronic limb-threatening ischemia (CLI)( 68% vs. 67%, P = .24), or proportion of CLI with tissue loss (65% vs. 63%, P = .34), between those undergoing adjunctive endarterectomy and bypass alone, respectively (Table I). Patients undergoing adjunctive endarterectomy were older (68 vs. 67, P = .02), more likely to be white (95% vs. 93%, P = .02), smokers (91% vs. 87%, P = .001), and have chronic obstructive pulmonary disease (COPD)(31% vs. 26%, P = .001). The endarterectomy group was also more likely to have undergone a prior coronary bypass or percutaneous coronary intervention (34% vs. 31%, P = .02) and prior endovascular lower extremity endovascular intervention (37% vs. 33%, P = .004), but were less likely to have had a prior lower extremity bypass (22% vs. 28%, P < .001) or an elevated (>1.78mg/dl) preoperative creatinine (5.7% vs. 7.4%, P = .04).
Table I.
Patient Characteristics
| Bypass w/Endarterectomy | Bypass only | ||
|---|---|---|---|
|
|
|||
| N | 1464 | 3032 | P-value
|
| Age (years, mean ±SD) | 68 ±10.5 | 67 ±10.9 | .02 |
| Elective (vs. urgent), %(N) | 80 (1173) | 80 (2430) | .99 |
| CLI (vs. claudication), %(N) | 68 (1000) | 67 (2018) | .24 |
| CLI-tissue loss (vs. rest pain), %(N) | 65 (648) | 63 (1272) | .34 |
| Female, %(N) | 34 (490) | 35 (1071) | .22 |
| White, %(N) | 95 (1387) | 93 (2819) | .02 |
| Diabetes, %(N) | 45 (664) | 43 (1292) | .08 |
| Obesity (BMI >30), %(N) | 27 (393) | 28 (828) | .73 |
| Smoker, %(N) | 91 (1326) | 87 (2642) | .001 |
| CAD, %(N) | 36 (530) | 34 (1041) | .22 |
| Prior CABG/PCI, %(N) | 34 (497) | 31 (927) | .02 |
| CHF, %(N) | 17 (251) | 16 (470) | .16 |
| COPD, %(N) | 31 (452) | 26 (794) | .001 |
| Abnormal stress test, %(N) | 11 (164) | 11 (340) | .99 |
| Dialysis dependent, %(N) | 4.6 (67) | 5.1 (154) | .47 |
| Elevated Preop Creatinine, %(N) | 5.7 (79) | 7.4 (211) | .04 |
| H/o Lower Ext. Open Bypass, %(N) | 22 (322) | 28 (835) | <.001 |
| H/o Lower Ext. Endovascular, %(N) | 37 (541) | 33 (988) | .004 |
| H/o Major Amp. , %(N) | 3.4 (50) | 3.4 (103) | .98 |
| Preop Medications, %(N) | |||
| Aspirin | 80 (1176) | 77 (2335) | .01 |
| Clopidogrel | 18 (259) | 16 (486) | .16 |
| Antiplatelet and Statin | 68 (997) | 60 (1805) | <.001 |
CLI: Chronic Limb-threatening Ischemia, BMI: body mass index, CAD: coronary artery disease, CABG/PCI: coronary artery bypass graft/percutaneous coronary intervention, CHF: congestive heart failure, COPD: chronic obstructive pulmonary disease, H/o: history of..
Operative details
Only lower extremity bypasses originating from the common femoral and profunda femoris were included. There was no difference in proportion with profunda origin between groups (Table II). Total operative time was longer with adjunctive femoral endarterectomy (median 268 minutes, [Interquartile Range 200–338] vs. 210 minutes, [160–283], P < .001), although 43% of patients were missing this information. When restricted to 2009–2015 only 15% of patients were missing operative time and the same difference remained. This operative time difference persisted when patients undergoing any concurrent endovascular or suprainguinal bypass procedures were excluded (overall: 256min vs. 206min, P < .001; CLI: 266min vs. 223min, P < .001; Claudication: 233min vs. 184min, P < .001). Bypass with endarterectomy also had higher estimated blood loss (median 250ml vs. 180ml, P < .001) and a greater proportion undergoing concurrent endovascular intervention (14% vs. 6%, P < .001, although 21% were missing these data).
Table II.
Operative details
| Bypass w/Endarterectomy | Bypass only | P-value | |
|---|---|---|---|
|
|
|||
| N | 1464 | 3032 | |
| OR time (min, median (IQR))[43%] | 268 (200–338) | 210 (160–283) | <.001 |
| OR time w/o concomitant procedures | 256 (194–326) | 206 (158–279) | <.001 |
| Claudication | 233 (180–305) | 184 (140–244) | <.001 |
| Fem to AK-pop with PTFE | 180 (138–224) | 142 (118–180) | <.001 |
| CLI | 266 (204–336) | 223 (168–294) | <.001 |
| Chlorhexidine Prep [42%], %(N) | 77 (669) | 78 (1365) | .37 |
| Estimated Blood Loss (ml, median (IQR)) | 250 (150–450) | 180 (100–300) | <.001 |
| Claudication | 200 (100–350) | 150 (100–250) | <.001 |
| CLI | 288 (150–500) | 200 (100–300) | <.001 |
| Prostethic (vs. vein), %(N) | 39 (568) | 38 (1154) | .63 |
| Origin-Common Femoral (vs. Profunda), %(N) | 96 (1411) | 96 (2896) | .18 |
| Graft Target, %(N) | .01 | ||
| SFA/Profunda/CFA | 3.3 (48) | 2.3 (69) | |
| Popliteal | 70 (1018) | 67 (2012) | |
| TP trunk/Tibials | 25 (359) | 28 (832) | |
| Pedal | 2.5 (37) | 3.6 (110) | |
| Concomitant Procedures, %(N) | |||
| Proximal Endovascular Int. [21%] | 14 (210) | 6.4 (135) | <.001 |
| Suprainguinal Bypass [40%] | 2.9 (26) | 2.8 (50) | .87 |
[] indicates % with datapoint missing, all variables with > 5% missing data reported in table
Fem to AK-pop = femoral to above-knee popliteal bypass
Perioperative Outcomes
The overall 30-day mortality rate was no different between the adjunctive endarterectomy and bypass only cohorts (1.8% vs. 2.1%, P = .60). In patients with CLI there was no difference in in-hospital mortality or 30-day mortality between bypass with endarterectomy and without (2.0% vs. 1.5%, P = .32 and 2.6% vs. 2.9%, P =. 61, respectively)(Table IIIA). Patients with CLI undergoing adjunctive endarterectomy had higher rates of postoperative myocardial infarction (MI) (6.2% vs. 3.8%, P = .003) and transfusion of more than 2 units of packed red blood cells (11% vs. 6.8%, P < .001). They were also more likely to be discharged on an antiplatelet and statin regimen (76% vs. 71%, P = .01). There was no difference in renal function deterioration, major ipsilateral amputation, surgical site infections, or primary patency at discharge. Despite higher rates of transfusion there was also no difference in return to the operating room (12% vs. 14%, P = .44) or in prolonged postoperative length of stay (7.5% vs. 5.2%, P = .08) between bypass with adjunctive endarterectomy and without, respectively.
Table IIIA.
Perioperative Outcomes for patients with CLI
| Bypass w/Endart. % (N) | Bypass only % (N) | ||
|---|---|---|---|
|
|
|||
| N | 1000 | 2018 | P-value
|
| In-hospital mortality | 2.0 (20) | 1.5 (30) | .30 |
| 30-day mortality | 2.6 (26) | 2.9 (59) | .61 |
| Myocardial Infarction (MI) | 6.2 (62) | 3.8 (76) | .003 |
| Dysrhythmia | 5.8 (58) | 4.2 (85) | .05 |
| Renal Function Deterioration | 4.7 (46) | 4.6 (92) | .92 |
| Ipsilateral Major Amputation | 1.2 (12) | 1.3 (26) | .84 |
| Above-knee | 0.2 (2) | 0.2 (5) | 1.0 |
| Below-knee | 1.0 (10) | 1.0 (21) | .92 |
| Return to OR | 12 (124) | 14 (271) | .44 |
| SSI | 5.6 (56) | 4.7 (95) | .29 |
| Transfusion >2units | 10.9 (104) | 6.8 (129) | <.001 |
| Patency at discharge | .38 | ||
| Primary | 96.2 (955) | 95 (1907) | |
| Primary assisted | 1.4 (14) | 2.1 (41) | |
| Secondary | 0.9 (9) | 1.4 (27) | |
| Occluded | 1.5 (15) | 1.2 (24) | |
| Antiplatelet and Statin at discharge | 76 (661) | 71 (1200) | .01 |
| Prolonged postop LOS (>7days) | 24 (240) | 22 (444) | .22 |
For patients with claudication there were few deaths overall and again no difference between patients undergoing bypass with endarterectomy and without for in-hospital (0.2% vs. 0.2%, P = 1.0) or 30-day mortality (0.2% vs. 0.4%, P = 1.0)(Table IIIB). Those undergoing adjunctive endarterectomy had higher rates of myocardial infarction (2.4% vs. 0.9%, P = .02), renal function deterioration (3.6% vs. 1.4%, P = .01), surgical site infections (5.0% vs. 2.6%, P = .02), and transfusions of more than 2 units (4.6% vs. 1.8%, P = .002). There was no difference in major ipsilateral amputation, primary patency at discharge, prolonged postoperative length of stay, or antiplatelet and statin combination at discharge.
Table IIIB.
Perioperative Outcomes for patients with Claudication
| Bypass w/Endart. % (N) | Bypass only % (N) | ||
|---|---|---|---|
|
|
|||
| N | 464 | 1014 | P-value
|
| In-hospital mortality | 0.2 (1) | 0.2 (2) | 1.0 |
| 30-day mortality | 0.2 (1) | 0.4 (4) | 1.0 |
| Myocardial Infarction (MI) | 2.4 (11) | 0.9 (9) | .02 |
| Dysrhythmia | 3.5 (16) | 2.9 (29) | .54 |
| Renal Function Deterioration | 3.6 (16) | 1.4 (14) | .01 |
| Ipsilateral Major Amputation | 0.0 (0) | 0.3 (3) | .24 |
| Above-knee | 0.0 (0) | 0.1 (1) | 1.0 |
| Below-knee | 0.0 (0) | 0.2 (2) | 1.0 |
| Return to OR | 6.3 (29) | 4.1 (42) | .08 |
| SSI | 5.0 (23) | 2.6 (26) | .02 |
| Transfusion >2units | 4.6 (21) | 1.8 (18) | .002 |
| Patency at discharge | .59 | ||
| Primary | 95.7 (444) | 97 (998) | |
| Primary assisted | 2.6 (12) | 1.7 (17) | |
| Secondary | 1.3 (6) | 0.9 (9) | |
| Occluded | 0.4 (2) | 0.4 (4) | |
| Antiplatelet and Statin at discharge | 75 (325) | 78 (696) | .36 |
| Prolonged postop LOS (>7days) | 7.5 (35) | 5.2 (53) | .08 |
Utilizing multivariable analysis, bypass with endarterectomy was associated with an increased risk of MI (Odds Ratio [OR] 1.6, 95% Confidence Interval [CI] 1.09 – 2.2)(Table IV). This same model was run without adjusting for transfusion to demonstrate the degree of interaction between transfusion and MI with a similar independent association between adjunctive endarterectomy and MI (OR 1.6, 95%CI 1.1–2.2). Bypass with endarterectomy was also independently associated with an increased risk for surgical site infections (OR 1.5, 95% CI 1.1–2.0) and transfusion of more than 2 units (1.8, 1.4 – 2.3). There was no difference in risk for 30-day mortality (0.87, 0.54 – 1.4) or renal dysfunction (1.2, 0.83 – 1.7) between groups.
Table IV.
Multivariable Models for Perioperative Outcomes
| SSI | MI | Transfusion > 2 units | |||||||
|---|---|---|---|---|---|---|---|---|---|
|
| |||||||||
| OR | 95%CI | P-value | OR | 95%CI | P-value | OR | 95%CI | P-value | |
| Bypass only | ref | ref | ref | ||||||
| Bypass w/ endarterectomy | 1.46 | 1.06–2.02 | .02 | 1.55 | 1.09–2.21 | .01 | 1.8 | 1.4–2.3 | <.001 |
| Urgent (vs. elective) | 1.03 | 0.69–1.53 | .89 | 0.95 | 0.64–1.42 | .81 | 1.59 | 1.2–2.1 | .001 |
| Female | 1.08 | 0.77–1.50 | .67 | 0.87 | 0.60–1.26 | .47 | 0.75 | 0.57–0.98 | .03 |
| Age (decades) | 0.89 | 0.76–1.03 | .12 | 1.11 | 0.92–1.34 | .28 | 1.3 | 1.1–1.5 | <.001 |
| Prosthetic (vs. vein) | 0.75 | 0.52–1.08 | .12 | 0.73 | 0.50–1.07 | .10 | 0.93 | 0.71–1.2 | 0.59 |
| Obesity (BMI >30) | 1.31 | 0.93–1.83 | .12 | * | * | * | * | * | * |
| Transfusion > 2units | 2.04 | 1.22–3.41 | .01 | 3.52 | 2.29–5.40 | <.001 | * | * | * |
| Symptoms | |||||||||
| Claudication | ref | ref | ref | ||||||
| CLI-rest pain | 1.13 | 0.71–1.80 | .62 | 1.80 | 0.95–3.42 | .07 | 1.49 | 0.96–2.32 | .08 |
| CLI-tissue loss | 1.47 | 0.97–2.23 | .07 | 3.34 | 1.91–5.84 | <.001 | 2.66 | 1.8–4.0 | <.001 |
| COPD | 1.28 | 0.91–1.80 | .16 | 1.76 | 1.21–2.56 | .003 | 0.88 | 0.67–1.17 | .39 |
| H/o bypass | 0.70 | 0.47–1.05 | .09 | * | * | * | 1.95 | 1.5–2.6 | <.001 |
| Chlorhexadine prep | 0.93 | 0.58–1.51 | .78 | * | * | * | * | * | * |
| Common Femoral (vs. Profunda) origin | 0.91 | 0.43–1.91 | .80 | 1.23 | 0.59–2.59 | .58 | 1.45 | 0.86–2.4 | .16 |
| Preop Antiplatelet and Statin | * | * | * | 2.15 | 1.38–3.33 | .001 | * | * | * |
| Preop Aspirin | * | * | * | * | * | * | 1.48 | 1.1–2.1 | .03 |
| Preop P2Y12 Antagonist | * | * | * | * | * | * | 1.86 | 1.4–2.5 | <.001 |
| Smoking | * | * | * | ||||||
| Previous | 1.37 | 0.77–2.43 | .29 | 1.66 | 1.1–2.6 | .02 | |||
| Current | 0.68 | 0.36–1.34 | .27 | 1.23 | 0.75–2.0 | .41 | |||
| Diabetes | * | * | * | 1.86 | 1.26–2.75 | .002 | 1.18 | 0.89–1.6 | .25 |
| On dialysis preop | * | * | * | 1.04 | 0.55–1.98 | .90 | 1.53 | 0.95–2.5 | .08 |
| CAD | * | * | * | 1.53 | 1.02–2.29 | .04 | 0.88 | 0.67–1.2 | .39 |
| Prior CABG/PCI | * | * | * | 1.23 | 0.81–1.85 | .33 | * | * | * |
| CHF | * | * | * | * | * | * | 1.52 | 1.1–2.1 | .01 |
Not part of final model for outcome
For SSI model: H and L = .16, 164 wounds in model; Also adjusted for graft target (jump, Pop-ref, tibial, pedal)
For MI model: H and L = .99, 142 MIs in model; When above model run without post-op transfusion, given association with MI, bypass w/ endarterectomy still associated with MI (OR 1.59 95%CI 1.14–2.21)
For Transfusion model: H and L = .54, 264 events in model
One-year Outcomes
Overall there was no difference in proportion eligible for 1-year follow-up between the bypass with endarterectomy cohort and bypass without (83% vs. 84%, P = .30) and no difference in median follow-up time (median 314 days [86–391] vs. 312 [74–391], P = .30 respectively). Table V reports the unadjusted kaplan-meier event rates, and associated log-rank p-values, stratified by symptom status, as well as the unadjusted and adjusted Cox hazard ratios without stratification and for CLI alone. There were too few events in the claudication group to adjust for differences separately. One-year survival was similar among patients with claudication (97% vs. 96%, P = .36) and CLI (85% vs. 83%, P = .15). Patients undergoing a bypass with endarterectomy for CLI had improved 1-year freedom from major amputation (91% vs. 87%, P = .049) and 1-year amputation free survival (80% vs. 76%, P = .03) compared to those not undergoing endarterectomy.
Table V.
One-year Outcomes
| Bypass w/Endarterectomy | Bypass only | Log-Rank P-value | Unadjusted, HR (95%CI) | Adjusted, HR (95%CI)* | |
|---|---|---|---|---|---|
| Survival | 0.86 (0.70–1.1) | 0.84 (0.66–1.1) | |||
|
| |||||
| Claudication only | 97% | 96% | 0.36 | ||
| CLI only | 85% | 83% | 0.15 | 0.86 (0.69–1.1) | 0.86 (0.67–1.1) |
|
| |||||
| Amputation Free Survival (AFS) | 0.83 (0.70–0.98) | 0.83 (0.68–1.01) | |||
|
| |||||
| Claudication | 96% | 94% | 0.24 | ||
| CLI | 80% | 76% | 0.03 | 0.82 (0.68–0.98) | 0.84 (0.68–1.04) |
|
| |||||
| Freedom from Major Amputation | 0.75 (0.56–1.02) | 0.78 (0.56–1.1) | |||
|
| |||||
| Claudication | 99% | 98% | 0.44 | ||
| CLI | 91% | 87% | 0.049 | 0.73 (0.53–1.0) | 0.79 (0.56–1.1) |
|
| |||||
| Freedom from MALE | 0.84 (0.84–1.2) | 0.95 (0.84–1.2) | |||
|
| |||||
| Claudication | 78% | 81% | 0.29 | ||
| CLI | 71% | 68% | 0.26 | 0.9 (0.75–1.1) | 0.91 (0.75–1.1) |
Overall median 313 days (IQR 76–391); Number of events at 1-year: 338 deaths, 175 major amputations, 501 major amputation/death, and 622 MALE
Too few events in claudication group to stratify analysis, so symptom status included in adjustment and overall Hazard Ratio (HR) with respect to adjunct endarterectomy represented above. CLI only Hazard Ratios therefore similar to overall. Adjusted for: age (decades), gender, race (white vs. non-white), preop renal dysfunction (Creatinine>1.78 mg/dl), smoker, diabetes, CAD, prior CABG/PCI, CHF, COPD, h/o lower extremity bypass, symptoms (claudication, rest pain, tissue loss), postop in-hospital ipsilateral major amputation (only included in Survival model), discharged on Antiplatelet and Statin
When adjusting for all relevant preoperative differences, there was no association between adjunctive endarterectomy and one-year mortality (Hazard Ratio [HR] 0.9, 95% CI 0.7 – 1.1), one-year amputation free survival (AFS) (HR 0.8, 95% CI 0.7 – 1.01), one-year major amputation (HR 0.8, 95% CI 0.5 – 1.1), or one-year MALE (HR 1.0, 95% CI 0.9 – 1.2)(Table V and Supplemental Table I for full Cox proportional hazard models). When these same endpoints were limited to patients with CLI only there again was no difference between our cohorts at one-year. Our sensitivity analysis, limiting years to 2009–15, had no impact on our adjusted findings although the univariate benefit to endarterectomy seen in all years for patients with CLI, with respect to one-year freedom from major amputation (91% vs. 87%, P = .09) and AFS (81% vs. 77%, P = .12), did not reach significance.
Discussion
This study demonstrates that patients undergoing adjunctive femoral endarterectomy with lower extremity bypass have similar mortality to those undergoing bypass alone, both in the perioperative period and at one-year. Performance of an endarterectomy with bypass does, however, appear to be independently associated with postoperative myocardial infarction, need for transfusions of more than 2 units, and surgical site infections. However, there was a suggestion of potential limb benefit with endarterectomy in terms of one-year freedom from major amputation and AFS.
Our thirty-day mortality rate of 1.8% for those undergoing bypass with endarterectomy was similar to the 1.2% reported by Malgor et al, which consisted of 85 patients with adjunctive femoral endarterectomy, 40% of whom had claudication.[5] This same study also reported a similar one-year survival to ours (84% vs. 89%).
Interestingly, Malgor et al also found that isolated common femoral endarterectomy had a 1.4% thirty-day mortality rate (67% of whom had claudication), compared to the 3.4% reported by Nguyen et al, although the later study had a larger sample size, 13% emergent cases, and 29% with rest pain, but overall proportion with CLI could not be determined in the later study due to limits of the registry used.[6] For comparison, a second study using a similar NSQIP sample to Nguyen et al, but excluding emergent cases, reported a 1.5% thirty-day mortality following isolated common femoral endarterectomy.[7] The later study did not identify proportion with rest pain and as mentioned above could not identify the proportion with CLI. We think it is the emergent cases that carry higher mortality risk, and which led Nguyen et al to report a 3.4% mortality rate for isolated femoral endarterectomy, whereas our study, which excluded emergent cases, is more consistent with prior literature using similar populations.
Patients with CLI undergoing adjunctive endarterectomy had improved one-year amputation free survival and freedom from major amputation on Kaplan-Meier analysis in our study. It is possible that the addition of an endarterectomy helps keep the foot perfused through the profunda and collaterals when the bypass goes down, as compared to those patients who did not undergo an adjunctive endarterectomy, and that this results in better limb salvage. Antiplatelet and statin were prescribed more commonly in the endarterectomy group, both preoperatively and at discharge. When this was accounted for in our adjusted Cox regression there was a dampening of the effect of endarterectomy. Despite this our confidence interval only just crosses 1.0 for AFS and freedom from major amputation, making the significant results on univariate testing less likely to be a false positive. Therefore further investigation is warranted.
There were significant differences in operative time, estimated blood loss, and transfusions administered between patients undergoing adjunctive endarterectomy and bypass alone in our study. After excluding additional concurrent procedures adjunctive endarterectomy added approximately 45 minutes to a lower extremity bypass, regardless of symptoms. Using the Vascular Quality Initiative national registry, Kalish et al reported a surgical site infection (SSI) rate of 4.8% for lower extremity bypass, and found that ABI <0.35, operative time >220min, estimated blood loss >100ml, and transfusion of more than two units were predictive of SSIs.[12] Given these data, it is not surprising that we saw a higher rate of surgical site infections in the endarterectomy group. Our SSI rate of 5.4% was slightly higher than reported by Kalish for patients with groin incisions, likely because we included only patients with bypasses originating in the CFA and profunda but not the SFA.
We also found a nearly two-fold increased rate of cardiac events in the endarterectomy group that persisted even after adjustment for preoperative differences. Malgor et al also found similar rates of MI (5%) and dysrhythmia (5%) in their 85 patients who underwent endarterectomy with bypass.[5] This could be related to the higher estimated blood loss and prolonged operative time in those undergoing adjunctive endarterectomy. In addition, vascular patients in general are at increased risk for MI postoperatively but the difference reported in our analysis could be explained further by the increased atherosclerotic burden and likely increased coronary vascular disease for those selected to undergo an adjunctive endarterectomy, presumably for plaque-laden arteries. In addition, we report a significantly higher rate of prior CABG or percutaneous coronary intervention in patients undergoing adjunctive endarterectomy, which further supports the likelihood of more clinically aggressive atherosclerotic disease in the adjunctive endarterectomy group. There were no differences in reported rates of CAD or CHF, although we do know from prior reports that there are large proportions of vascular patients with undiagnosed but significant coronary disease, which may further explain the high rates of cardiac complications in the adjunctive endarterectomy group which we presume to have greater atherosclerotic burden, despite similar CAD and CHF rates.[13]
Both cohorts in our study had similarly small proportions of concurrent suprainguinal bypasses but a large difference in concurrent proximal endovascular interventions, with higher rates in the adjunctive endarterectomy group. Prior studies have proven the efficacy of iliac angioplasty and stenting in patients with inflow disease and we believe the higher rate of proximal endovascular intervention is a reflection of the increased atherosclerotic burden in patients undergoing adjunctive endarterectomy.[3]
Limitations of this study include those inherent in any retrospective analysis, such as misreporting and missing data. However, a major strength of the VSGNE data is that the surgeon fills out the operative details for each patient making these data reliable and specific. Also, in terms of missing data, we focused our one-year outcomes discussion on time-to-event analysis using an accepted standard error rate of <. 10, which adequately accounts for missing data by censoring and indicates the time point our sample size becomes too small to perform reliable analyses, although this does not account for non-random censoring. Furthermore, we were unable to distinguish between the simple endarterectomy performed for debulking at the time of anastomosis and the more extensive endarterectomy that results in patch repair before graft anastomosis. It is likely that this latter group drove the increased risk for perioperative events and longer operative time but this could not be proven. We were also unable to distinguish whether the SSI was at the level of the endarterectomy or at other incisions related to the bypass. Theoretically endarterectomy could protect against SSI below the groin by improving blood flow but could also increase SSI, both at the groin and more distally, through interruption of lymphatics. In addition, this study does not have the appropriate data to imply a causal relationship between adjunctive endarterectomy and higher rates of MI and surgical site infections. We believe a surgeons decision to perform an adjunct endarterectomy is likely a marker of a patient with more systemic disease, and it is important to demonstrate that adjunctive femoral endarterectomy with lower extremity bypass is safe despite a recent report of high perioperative mortality in isolated common femoral endarterectomy.
Conclusion
Adjunctive femoral endarterectomy with bypass does not lead to increased perioperative or one-year mortality risk despite increased in-hospital myocardial infarctions, transfusions, and surgical site infections. The practicing surgeon should not hesitate to perform an adjunctive endarterectomy in appropriate patients but should be aware of the higher perioperative risk associated with patients undergoing this procedure and take precautions to minimize these complications. The possibility of a benefit for one-year AFS and freedom from major amputation, as reported in this analysis, with adjunctive endarterectomy when deemed clinically appropriate warrants further study.
Supplementary Material
Acknowledgments
Supported by grant 5R01HL105453-03 from the NHLBI and the NIH T32 Harvard-Longwood Research Training in Vascular Surgery grants HL007734.
Footnotes
Poster Presentation at the Society for Vascular Surgery 2016 Vascular Annual Meeting (VAM), National Harbor, MD June 8-11, 2016
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
References
- 1.Springhorn ME, Kinney M, Littooy FN, Saletta C, Greisler HP. Inflow atherosclerotic disease localized to the common femoral artery: treatment and outcome. Ann Vasc Surg. 1991;5(3):234–40. doi: 10.1007/BF02329379. [DOI] [PubMed] [Google Scholar]
- 2.Kang JL, Patel VI, Conrad MF, Lamuraglia GM, Chung TK, Cambria RP. Common femoral artery occlusive disease: contemporary results following surgical endarterectomy. J Vasc Surg. 2008;48(4):872–7. doi: 10.1016/j.jvs.2008.05.025. [DOI] [PubMed] [Google Scholar]
- 3.Chang RW, Goodney PP, Baek JH, Nolan BW, Rzucidlo EM, Powell RJ. Long-term results of combined common femoral endarterectomy and iliac stenting/stent grafting for occlusive disease. J Vasc Surg. 2008;48(2):362–7. doi: 10.1016/j.jvs.2008.03.042. [DOI] [PubMed] [Google Scholar]
- 4.Kechagias A, Ylonen K, Biancari F. Long-term outcome after isolated endarterectomy of the femoral bifurcation. World J Surg. 2008;32(1):51–4. doi: 10.1007/s00268-007-9309-7. [DOI] [PubMed] [Google Scholar]
- 5.Malgor RD, Ricotta JJ, 2nd, Bower TC, Oderich GS, Kalra M, Duncan AA, et al. Common femoral artery endarterectomy for lower-extremity ischemia: evaluating the need for additional distal limb revascularization. Ann Vasc Surg. 2012;26(7):946–56. doi: 10.1016/j.avsg.2012.02.014. [DOI] [PubMed] [Google Scholar]
- 6.Nguyen BN, Amdur RL, Abugideiri M, Rahbar R, Neville RF, Sidawy AN. Postoperative complications after common femoral endarterectomy. J Vasc Surg. 2015;61(6):1489–94 e1. doi: 10.1016/j.jvs.2015.01.024. [DOI] [PubMed] [Google Scholar]
- 7.Siracuse JJ, Gill HL, Schneider DB, Graham AR, Connolly PH, Jones DW, et al. Assessing the perioperative safety of common femoral endarterectomy in the endovascular era. Vasc Endovascular Surg. 2014;48(1):27–33. doi: 10.1177/1538574413508827. [DOI] [PubMed] [Google Scholar]
- 8.Cronenwett JL, Likosky DS, Russell MT, Eldrup-Jorgensen J, Stanley AC, Nolan BW, et al. A regional registry for quality assurance and improvement: the Vascular Study Group of Northern New England (VSGNNE) J Vasc Surg. 2007;46(6):1093–1101. doi: 10.1016/j.jvs.2007.08.012. discussion 1101–2. [DOI] [PubMed] [Google Scholar]
- 9.Bursac Z, Gauss CH, Williams DK, Hosmer DW. Purposeful selection of variables in logistic regression. Source Code Biol Med. 2008;3:17. doi: 10.1186/1751-0473-3-17. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Vittinghoff E, McCulloch CE. Relaxing the rule of ten events per variable in logistic and Cox regression. Am J Epidemiol. 2007;165(6):710–8. doi: 10.1093/aje/kwk052. [DOI] [PubMed] [Google Scholar]
- 11.Conte MS. Understanding objective performance goals for critical limb ischemia trials. Semin Vasc Surg. 2010;23(3):129–37. doi: 10.1053/j.semvascsurg.2010.06.001. [DOI] [PubMed] [Google Scholar]
- 12.Kalish JA, Farber A, Homa K, Trinidad M, Beck A, Davies MG, et al. Factors associated with surgical site infection after lower extremity bypass in the Society for Vascular Surgery (SVS) Vascular Quality Initiative (VQI) J Vasc Surg. 2014;60(5):1238–46. doi: 10.1016/j.jvs.2014.05.012. [DOI] [PubMed] [Google Scholar]
- 13.Steg PG, Bhatt DL, Wilson PW, D’Agostino R, Sr, Ohman EM, Rother J, et al. One-year cardiovascular event rates in outpatients with atherothrombosis. JAMA. 2007;297(11):1197–206. doi: 10.1001/jama.297.11.1197. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.