Outcomes of Femoropopliteal Interventions for Critical Ischemia in the Hemodialysis-Dependent Patient

Outcomes of Femoropopliteal Interventions for Critical Ischemia in the Hemodialysis-Dependent Patient Christopher J. Smolock, Hosam F. El-Sayed, and Mark G. Davies, Houston, Texas

Background: The number of patients maintained on hemodialysis is rising. There are limited data on the outcomes of femoropopliteal interventions, both open and endovascular, in this population. This report examines the anatomic and clinical outcomes in this population. Methods: A database of patients undergoing open (OPEN) and endoluminal (ENDO) intervention for femoropopliteal disease (2000 to 2010) was retrospectively queried. Patients on hemodialysis with critical ischemia at the time of surgery or intervention were selected. Patients who underwent tibial bypass or had concomitant tibial interventions were excluded. KaplaneMeier analyses were performed to assess time-dependent outcomes. Factor analyses were performed for time-dependent variables. Results: One hundred sixty-one hemodialysis-dependent patients underwent either OPEN or ENDO procedures for critical ischemia. Of these, 70 patients were treated with OPEN procedures and 91 with ENDO procedures. ENDO patients were more likely to present with a higher cardiac risk index (P ¼ 0.0001), metabolic syndrome (P ¼ 0.02), cerebrovascular disease (P ¼ 0.01), and a dependent living status preoperatively (P ¼ 0.04). ENDO patients presented with more rest pain and tissue loss (P ¼ 0.03). OPEN patients presented with more advanced lesions (P ¼ 0.04). Combined morbidity was higher in the OPEN group (P ¼ 0.05). Cumulative patency (P ¼ 0.04) and clinical efficacy (P ¼ 0.05) were higher in the OPEN group compared to those in the ENDO group. Conclusions: Hemodialysis patients undergoing femoralepopliteal endovascular interventions for symptomatic disease have a low cumulative patency and clinical efficacy. Although open surgical revascularization has higher perioperative morbidity and a trend toward higher perioperative mortality, it provides a superior 5-year cumulative patency and clinical efficacy and should be considered in this population subgroup.

INTRODUCTION

Presented at the Surgical Forum at the Annual Clinical Congress of the American College of Surgeons, San Francisco, CA, October 2011. Division of Vascular and Endovascular Surgery, Houston Methodist Hospital, Houston, TX. Correspondence to: Mark G. Davies, MD, PhD, MBA, Department of Cardiovascular Surgery, Houston Methodist Hospital, 6550 Fannin, Smith Tower, Suite 1401, Houston, TX 77030, USA; E-mail: [email protected] Ann Vasc Surg 2015; 29: 237–243 http://dx.doi.org/10.1016/j.avsg.2014.07.036 Ó 2015 Elsevier Inc. All rights reserved. Manuscript received: April 4, 2014; manuscript accepted: July 26, 2014; published online: October 6, 2014.

The number of patients maintained on hemodialysis is rising.1 There has been a substantial increase in patient comorbidities that lead to renal failure, such as hypertension and diabetes mellitus, in the last decade. Projections suggest that these patterns of disease will increase substantially in the next 2 decades.2 In addition, there has been a marked increase in and shift toward endoluminal therapy for superficial femoral artery (SFA) occlusive disease.3,4 The technology and technical skills in practice have improved and permitted increasingly more challenging lesions to be tackled.5,6 Multiple reports have demonstrated that renal insufficiency, particularly the need for 237

238 Smolock et al.

hemodialysis, negatively influences limb loss and mortality in such patients undergoing lower extremity surgical revascularization.7,8 The impact of chronic kidney disease on SFA endovascular interventions has also been reported.9 We have previously demonstrated the impact of various factors on SFA interventions and the disparities in anatomic outcomes and objective performance goals in patients undergoing lower extremity endovascular interventions versus open bypasses.10e12 The outcomes of SFA interventions compared with open bypasses in the hemodialysisdependent patient is unknown. The aim of this study was to compare the outcomes of endoluminal interventions and surgical revascularization for SFA disease in hemodialysis-dependent patients.

METHODS Study Design A database of patients undergoing open or endovascular treatment of SFA disease for critical ischemia between 2000 and 2010 was retrospectively queried. The group receiving hemodialysis was identified and further stratified into those undergoing surgical revascularization and those receiving endoluminal therapy. Patients who underwent tibial bypass or had concomitant tibial interventions were excluded. Data utilization fell under the category of secondary use of pre-existing data and was approved by the institutional review board. Study Setting Academic Medical Center with 800 beds in a catchment area of 5 million people. It is a tertiary and quaternary referral facility for vascular surgery. Methodology For each patient captured, demographics, symptoms, existing comorbid conditions, and risk factors for atherosclerosis were identified. Therapy for individual patients was dictated by individual attending physician preference and was not regulated by unit guidelines. All patients received aspirin daily (81 mg or 325 mg) as a general cardiovascular protection agent. Noninvasive studies were performed initially on all patients receiving a work up for peripheral arterial disease. Patients with serious symptoms or signs of severe stenosis and/or occlusion based on the initial noninvasive tests received angiograms. Angiograms and angiographic reports were reviewed; lesions were described by length, calcification, and

Annals of Vascular Surgery

patency and then categorized under the TransAtlantic Inter-Society Consensus II (TASC-II) system.13 The preoperative distal runoff was scored by the number of patent tibial vessels and according to a modification of Society for Vascular Surgery (SVS) criteria employed for determining bypass runoff (using the cumulative score for the distal popliteal from knee joint to first tibial branch; maximum 9 + 1) and each of the tibial vessels (maximum 3 each) giving a maximum possible total score of 19.14 Angioplasty was performed with the patient under systemic heparin administration (40e60 U/kg), and completion angiography was performed to assess the technical result. Stents were used (at the discretion of the operator) primarily or as an adjunct for flow-limiting dissections, intimal flaps, or poor technical results (50% residual stenosis). No covered stents were used. The complexity of each endovascular intervention was scored according to the ad hoc system described by DeRubertis et al.15 in which 1 point was awarded for an intervention in the iliac, femoral, or tibial segments of the leg. Bypass surgery was performed from common femoral artery to above-knee or below-knee popliteal artery with conduit obtained via endoscopic vein harvest when a satisfactory vein was available. If satisfactory vein was not available, a prosthetic graft was chosen; either polytetrafluorethylene (6 or 8 mm) or Dacron (6 or 8 mm) was used. The post treatment regimen in bypass is ASA 81 mg and in ENDO Plavix 75 mg with ASA 81 mg for 1 month followed by continued therapy with ASA 81 mg. In early years, we did not emphasize statin therapy, but in later years, we make all efforts to discharge the patient on a statin. Patients underwent routine duplex ultrasound follow-up at 1, 3, and every 6 months after their procedure using criteria previously described.10 During follow-up, angiography was only performed if noninvasive studies suggested restenosis and/ or occlusion (positive duplex scan with a drop in ankle-brachial index [ABI] of >0.15 and toebrachial index of >0.1) and the patient had recurrent symptoms. For the OPEN group, median follow-up was 3.3 years (range, 0.1e13.6 years) with 15 patients lost to any follow-up (20%), whereas for the ENDO group, median follow-up was 1.4 years (range, 0e7.9 years) with 9 patients lost to any follow-up (19%). Statistical Analysis All statistical analyses were performed on an ‘‘intention-to-treat’’ basis. Measured values are reported as

Vol. 29, No. 2, February 2015

Intervention for SFA disease in HD patients 239

Table I. Characteristics of patients Open

Demographics Patients 70 Limbs treated (n) 75 Male (%) 62 66 ± 11 Average age (mean ± standard deviation), years Symptoms (%) Rest pain 57 Tissue loss 43 Comorbidities (%) Modified Cardiac 4.9 ± 1.6 Risk Index (mean ± standard deviation) Smoking history 79 Current smoker 10 Coronary artery disease 46 Hypertension 97 Diabetes 66 Hyperlipidemia 72 Statin 43 Metabolic syndrome 49 Cerebrovascular disease 14 Hypothyroidism 15 Hypercoagulability 4 Preoperative living status (%) Independent 84 Dependent 16 Preoperative ambulatory status (%) Ambulatory 84 Ambulatory/homebound 16 Nonambulatory/transfer 0 Nonambulatory/ 0 bedridden

Table II. Lesion characteristics Endo

P value

91 99 48 67 ± 13

d d 0.05 0.60

47 53

0.20

5.0 ± 1.8 0.0001

76 16 56 98 78 74 70 68 31 13 2

0.71 0.25 0.27 1.00 0.11 1.00 0.0004 0.02 0.01 0.82 0.66

69 31

0.04

72 25 3 0

0.097

percentages or means ± standard deviation. Patency and limb salvage rates were calculated using KaplaneMeier analyses and reported using current SVS criteria and objective performance goals.16,17 Standard errors are reported in KaplaneMeier analyses. Analyses were performed using JMP software version 7.0 (SAS Institute, Cary, NC).

RESULTS Patient Population One hundred sixty-one hemodialysis-dependent patients underwent either OPEN or ENDO procedures for critical ischemia over the period in question, and this represented 6% of patients treated for symptomatic femoropopliteal disease. Of these, 70 patients were treated with OPEN procedures and

Open

SFA TASC-II category (%) A/B 34 C/D 66 Tibial runoff Number of tibial 2.0 vessels Modified SVS runoff 7.7 score Amputation risk scores Prevent III 6.3 Low (0e3; %) 0 Medium (4e7; %) 67 High (>7; %) 33 Finnvasc score 2.6 0 (%) 0 1 (%) 5 2 (%) 34 3 (%) 55 4 (%) 6

Endo

P value

51 49

0.04

± 0.8

1.7 ± 0.7

0.01

± 4.4

8.9 ± 3.7

0.08

± 1.6

6.9 ± 1.9 0 56 44 2.4 ± 0.8 2 13 32 50 3

0.03 0.28

± 0.7

0.04 0.14

91 with ENDO procedures. In the OPEN group, 62% were male, whereas in the ENDO group, 48% were male (Table I). Age was equivalent in both the groups (Table I). Past smoking history and current smoking were equal between the groups (Table I). There were no significant differences between the group with regard to the following comorbidities: coronary artery disease, hypertension, diabetes, hyperlipidemia, and hypercoagulability (Table I). However, there was a significant difference between the groups regarding metabolic syndrome (49% vs. 68%, OPEN vs. ENDO; P ¼ 0.02) and cerebrovascular disease (14% vs. 31%, OPEN vs. ENDO; P ¼ 0.01; Table I). Although there was a significantly higher number of patients whose living status was considered ‘‘dependent’’ in the ENDO group, the preoperative ambulatory status did not differ between groups (Table I). Patients receiving OPEN therapy tended to have more advanced lesions, that is, a greater number of TASC-II C and D lesions (66% vs. 49%; P ¼ 0.04). Scores used to grade the tibial vessel runoff showed a difference in the group averages between OPEN and ENDO in the number of tibial vessels present (2.0 ± 0.8% vs. 1.7 ± 0.7%, OPEN vs. ENDO; P ¼ 0.01), favoring better runoff in the OPEN group. However, the modified SVS runoff score (7.7 ± 4.4% vs. 8.9 ± 3.7%, OPEN vs. ENDO; P ¼ 0.08) was similar between the groups (Table II). When the amputation risk scores of PREVENT III and Finnvasc were reviewed for both the groups, the results were contradictory. PREVENT

240 Smolock et al.

Annals of Vascular Surgery

Table III. Mortality, morbidity, and objective performance goals Open

Mortality and morbidity (%) Mortality 4 Morbidity 20 Systemic 4 Local 11 Lesion n/a Objective performance goals (%) 30-Day MACE 8 30-Day MALE 21 30-Day major 8 amputations

Endo

P value

2 12 2 6 4

<0.05 <0.05 0.08 0.25 d

6 25 18

0.54 0.58 <0.05

Table V. Minor and major amputations

Amputation (%) No amputation Toe and/or forefoot Major amputation BKA AKA

Open

Endo

P value

66 10 25 44 56

60 10 30 37 63

0.76

0.43

BKA, below-knee amputation; AKA, above-knee amputation.

group to be at a higher statistical risk but with distributions of risk fairly equivalent to those predicted for the ENDO group (Table II).

n/a, not applicable; MALE, major adverse limb events.

Immediate Outcomes Table IV. Hemodynamic changes and immediate symptom relief Open

Endo

Hemodynamic changes Change in ABIa 0.31 ± 0.10 0.22 ± 0.13 ABI increase 73 76 >0.15 (%)a Change in TBIb 0.25 ± 0.11 0.19 ± 0.12 TBI increase 54 70 >0.1b (%) Immediate symptom relief (%) Resolved 14 14 Improved 72 72 No change 13 12 Deterioration 1 2 Postoperative ambulatory status (%) Ambulatory 60 60 Ambulatory/ 24 18 homebound Nonambulatory/ 15 21 transfer Nonambulatory/ 1 1 bedridden Discharge status (%) Home 70 62 Rehabilitation 12 22 facility 17 15 Skilled nursing facility Hospital 1 1

P value

0.0001 0.72 0.06 0.04

0.98

0.69

0.47

The technical failure rates were 0% and 4% for the OPEN and ENDO groups, respectively. However, no patient in the ENDO group required emergent conversion to open surgery. Eventual need for open bypass surgery in the ENDO group mirrored the technical failure rate previously mentioned. Thirty-day mortality was higher in the OPEN group (30-day mortality, 4% vs. 2%, OPEN vs. ENDO; P < 0.05). Total morbidity was higher in the OPEN group compared to that in ENDO (30-day morbidity, 20% vs. 12%, OPEN vs. ENDO; P < 0.05) with a trend toward an increase in the incidence of systemic complications and local wound or access site complications (Table III). Thirty-day major adverse cardiac events (MACE) were higher but not statistically significant in the OPEN group. Thirty-day major adverse limb events were similar between groups, but 30-day amputations were higher in the ENDO group (8% vs. 18%; P ¼ 0.05; Table III). Hemodynamically, there was a marked increase in ABIs in both groups, with w75% of all treated patients having a rise in both the OPEN and the ENDO groups (Table IV). However, the change in ABI was greater in the ENDO group (change in ABI, 0.69 ± 0.10 vs. 0.79 ± 0.23, OPEN vs. ENDO; P ¼ 0.0007). After intervention in the OPEN and ENDO groups, an equal percentage of patients, 86% and 86%, respectively, were considered to have improved or resolved symptoms. There was also no difference in postoperative ambulatory status and discharge status (Table IV).

TBI, toe-brachial index. a For nondiabetics and/or compressible vessels only. b For all the patients.

Long-Term Anatomic Outcomes

III scoring showed the ENDO group to be at a higher statistical risk but with distributions of risk roughly equivalent to those predicted for the OPEN group (Table II), whereas the Finnvasc showed the OPEN

There was no difference in overall major or minor amputations in either group but there was a trend toward more above-knee amputations in the ENDO group (Table V). There was no significant difference

Vol. 29, No. 2, February 2015

Intervention for SFA disease in HD patients 241

Fig. 1. Anatomic outcomes: Life-table analysis of hemodialysis patients with SFA disease treated with OPEN or ENDO procedures. Data are the mean ± SEM and number of limbs at risk shown in the table. No error bars are

shown if the SEM is >10%, and the data set terminates if the number at risk is <10. (A), Primary patency; (B), assisted primary patency; and (C), secondary patency.

in the primary patency between groups at 5 years (68 ± 6% vs. 54 ± 6%, OPEN vs. ENDO; P ¼ 0.07; Fig. 1A). There was a statistical difference in both assisted primary and secondary or cumulative patencies between the groups at 5 years (assisted primary patency, 81 ± 5% vs. 67 ± 5%, OPEN vs. ENDO; P ¼ 0.03; and secondary patency, 81 ± 5% vs. 68 ± 5% OPEN vs. ENDO; P ¼ 0.04; Fig. 1B, C).

PREVENT III scores did not predict amputation in this hemodialysis population. Finally, clinical efficacy, which includes amputations, major limb events, and ambulation and freedom from recurrent symptoms, was significantly different between groups favoring the OPEN group (clinical efficacy, 64 ± 7% vs. 49 ± 6%, OPEN vs. ENDO; P ¼ 0.05; Fig. 2D).

Long-Term Functional Outcomes

DISCUSSION

Within 5-year follow-up, 9 patients (9.9%) in the OPEN and 27 patients (29.7%) in the ENDO groups were dead. Overall mortality was higher in the OPEN group, with patient survival rates of 64 ± 5% and 73 ± 6% at 5 years in the OPEN and ENDO groups, respectively, but this difference was not statistically significant (P ¼ 0.11; Fig. 2A). Thus, 26 and 15 patients were alive in follow-up in the OPEN and ENDO groups, respectively, at 5 years. Five-year limb salvage was 89 ± 4% in the OPEN group and 84 ± 4% in the ENDO group (P ¼ 0.11; Fig. 2B). Amputation-free survival was not statistically significant between the groups at 5 years (64 ± 6% vs. 49 ± 6%, OPEN vs. ENDO; P ¼ 0.77; Fig. 2C). In those patients undergoing ENDO, the Finnvasc score was a significant predictor of limb salvage (risk ratio, 4.3; 95% confidence interval, 1.5e14.6; P ¼ 0.006). Finnvasc did not predict amputation after OPEN.

This study analyzed the influence of hemodialysis on the short- and long-term outcomes between groups of patients undergoing either SFA endovascular intervention or surgical revascularization for symptomatic disease. There was no difference in mortality and MACE, but overall morbidity and combined complications were significantly higher in the OPEN group. Immediate post procedural hemodynamic and/or symptom improvement and restoration of impaired ambulation were equivalent in both groups. At 5 years, there was either a trend toward better patency (primary) or a significant difference (assisted primary and secondary) in patency that favored the OPEN group. This was consistent with clinical efficacy being superior in the OPEN group. This report found that the patients selected for OPEN surgical revascularization were as likely to present with more advanced symptoms of rest pain

242 Smolock et al.

Annals of Vascular Surgery

Fig. 2. Functional outcomes: Life table analysis of hemodialysis patients with SFA disease treated with OPEN or ENDO procedures. Data are the mean ± SEM and number of patients or limbs at risk shown in the table. No

error bars are shown if the SEM is >10%, and the data set terminates if the number at risk is <10. (A), Survival; (B), limb salvage; (C), amputation-free survival; and (D), clinical efficacy.

or tissue loss than patients undergoing ENDO intervention, despite the OPEN group having more advanced anatomic disease (TASC C and D lesions). This was perhaps a function of patients with less cardiac risk and lower body mass index being selected for OPEN procedures rather than ENDO (Table I). We have previously shown that the presence of metabolic syndrome, diabetes alone, and chronic renal insufficiency will affect the outcomes after SFA intervention.9e12 Several other authors have demonstrated that the severity of presenting symptoms,18 lesion severity,15,19 and the poor runoff will affect the outcomes of SFA interventions.14,19 Technical success was achieved in all OPEN procedures. However, 4 patients undergoing ENDO intervention did not achieve technical success. This is interesting because it does not reflect the presence of less advancedelesion anatomy in the ENDO group. It is also important to note that many patients

with HD do have infrapopliteal disease, which was not the subset of patients studied in this report. Functional outcomes, namely reduction in symptoms, preservation of limb, and maintenance of ambulation must remain the ultimate goal of all vascular interventions, including those in the SFA. There has been a decrease in major amputations reported in the Nationwide Inpatient Sample associated with an increase in endoluminal interventions and a decrease in surgical procedures over the past few decades for a variety of reasons.20 We were unable to determine a predictor for limb loss in this group on HD. The most significant finding in this study is that the group receiving OPEN therapy, in addition to having a superior cumulative patency at 5 years, also has a superior clinical efficacy at 5 years despite the perioperative period associated with this trend toward greater mortality. This durable clinical effect occurred despite an equivalent

Vol. 29, No. 2, February 2015

number of patients showing an initial ABI rise >0.15 and most patients showing initial improvement or resolution of symptoms in both groups. This study is retrospective in nature, and the clinical decision making was individualized, not driven by a standard protocol. This study was also gathered from data over a 10-year period during which patterns and trends can change. Specific to this 10year time period, ENDO interventions became increasingly more favored for SFA and popliteal disease for all ranges of clinical severity and anatomic complexity. We cannot discount completely the influence of triage for high cardiac risk on the results, as patients treated with ENDO interventions had significantly higher cardiac risk. The superior ENDO results likely reflect selection bias as a result of the ability to offer surgery to the patients.

CONCLUSIONS Hemodialysis patients undergoing femoralepopliteal endovascular interventions for critical ischemia have a low cumulative patency and clinical efficacy. Although open surgical revascularization has higher perioperative morbidity and a trend toward higher perioperative mortality, it provides a superior 5year cumulative patency and clinical efficacy and should be considered in this population subgroup.

The authors thank Daynene Vykoukal, PhD for critical reading of the article. REFERENCES 1. Robinson BM, Fuller DS, Bieber BA, et al. The DOPPS practice monitor for US dialysis care: trends through April 2011. Am J Kidney Dis 2012;59:309e12. 2. Maggio CA, Pi-Sunyer FX. Obesity and type 2 diabetes. Metab Clin North Am 2003;32:805e22. 3. Davies MG, Waldman DL, Pearson TA. Comprehensive endovascular therapy for femoropopliteal arterial atherosclerotic occlusive disease. J Am Coll Surg 2005;201:275e96. 4. Lee LK, Kent KC. Infrainguinal occlusive disease: endovascular intervention is the first line therapy. Adv Surg 2008;42:193e204.

Intervention for SFA disease in HD patients 243

5. Lumsden AB, Das TS. Endovascular management of infrainguinal disease. J Endovasc Ther 2006;13(Suppl 2):II1e2. 6. Lumsden AB, Davies MG, Peden EK. Medical and endovascular management of critical limb ischemia. J Endovasc Ther 2009;16(Suppl II):II-31e62. 7. O’Hare AM, Sidawy AN, Feinglass J, et al. Influence of renal insufficiency on limb loss and mortality after initial lower extremity surgical revascularization. J Vasc Surg 2004;39: 709e16. 8. O’Hare AM, Bertenthal D, Sidawy AN, et al. Renal insufficiency and use of revascularization among a national cohort of men with advanced lower extremity peripheral arterial disease. Clin J Am Soc Nephrol 2006;1:297e304. 9. Bakken AM, Protack CD, Saad WE, et al. Impact of chronic kidney disease on outcomes of superficial femoral artery endoluminal interventions. Ann Vasc Surg 2009;23:560e8. 10. Bakken AM, Palchik E, Hart JP, et al. Impact of diabetes on the outcomes of superficial femoral artery endoluminal interventions. J Vasc Surg 2007;46:946e58. 11. Smolock CJ, Anaya-Ayala JE, Bismuth J, et al. Impact of metabolic syndrome on the outcomes of superficial femoral artery interventions. J Vasc Surg 2012;55:985e93. 12. Smolock CJ, Anaya-Ayala JE, Bismuth J, et al. Current outcomes for superior femoral artery interventions: role of diabetes and renal insufficiency. JACC 2011;57:E1580. 13. TASC II Working Group. TASC-II: Inter-Society Consensus for the management of peripheral vascular arterial disease. J Vasc Surg 2007;45(Suppl S):S1e67. 14. Davies MG, Saad WE, Peden EK, et al. Impact of runoff on superficial femoral artery endoluminal interventions for rest pain and tissue loss. J Vasc Surg 2008;48:619e26. 15. DeRubertis BG, Pierce M, Chaer RA, et al. Lesion severity and treatment complexity are associated with outcome after percutaneous infra-inguinal intervention. J Vasc Surg 2007;46:709e16. 16. Rutherford RB, Baker JD, Ernst C, et al. Recommended standards for reports dealing with lower extremity ischemia: revised version. J Vasc Surg 1997;26:517e38. 17. Goodney PP, Schanzer A, Demartino RR, et al., Vascular Study Group of New England. Validation of the Society for Vascular Surgery’s objective performance goals for critical limb ischemia in everyday vascular surgery practice. J Vasc Surg 2011;54:100e8. 18. Ryer EJ, Trocciola SM, DeRubertis B, et al. Analysis of outcomes following failed endovascular treatment of chronic limb ischemia. Ann Vasc Surg 2006;20:440e6. 19. Surowiec SM, Davies MG, Lee D, et al. Percutaneous angioplasty and stenting of the superficial femoral artery. J Vasc Surg 2005;41:269e78. 20. Goodney PP, Beck AW, Nagle J, et al. National trends in lower extremity bypass surgery, endovascular interventions, and major amputations. J Vasc Surg 2009;50:54e60.