Increased Pre-operative Pulse Pressure Predicts Procedural Complications and Mortality in Patients Undergoing Tibial Interventions for Critical Limb Ischemia

Abstract

Introduction

Pulse pressure is a non-invasive measure of arterial stiffness. Elevated pulse pressure is associated with an increased risk of cardiovascular events and death. The effects of pulse pressure on outcomes after endovascular interventions for critical limb ischemia (CLI), however, are unknown. We thus evaluated whether increased pre-operative pulse pressure was associated with adverse outcomes and mortality in patients undergoing endovascular tibial artery intervention.

Methods

All patients undergoing endovascular tibial intervention for CLI at a single institution from 2004 to 2014 were included in this study. Pre-operative pulse pressure was derived from measurements obtained in the holding area prior to the procedure. Patients were divided into 2 groups based on pulse pressure, < 80 or ≥ 80. Patient demographics and co-morbidities were documented, and outcomes including procedural complications, repeat intervention, amputation, and mortality were recorded. Multivariable logistic regression was utilized to account for patient demographics and comorbidities.

Results

Of 371 patients, 186 patients had a pre-operative pulse pressure <80 and 185 had a pre-operative pulse pressure ≥80. No significant differences in patient demographics or comorbidities were identified; however there was a trend toward older age in patients with elevated pulse pressure (70 vs. 72, P = 0.07). On univariate analysis, procedural complications (21% vs. 13%, P = 0.02), reinterventions (26% vs. 17%, P < 0.01), and restenosis (32% vs. 23%, P = 0.03) were more common among patients with pulse pressure ≥ 80. Procedural complications remained significant on multivariate analysis (OR 1.8, 95% CI 1.0-3.1, P = 0.04). There was no difference in 30-day mortality; however increased mortality was seen at 5 years of follow-up (OR: 1.6, 95% CI: 1.0-2.5, P = 0.04) following multivariable analysis.

Conclusions

Increased pre-operative pulse pressure is associated with procedural complications and increased mortality in patients undergoing endovascular tibial intervention for CLI. It is a marker of increased risk, and may be a suitable target for interventions aimed at improving outcomes in this high risk population.

INTRODUCTION

Pulse pressure, the difference between systolic and diastolic blood pressure, is used as a noninvasive marker of arterial stiffness. Increased pulse pressure reflects decreased compliance in large and medium sized arteries.¹ It is associated with increased cardiovascular and all-cause mortality in the general population.² It predicts adverse cardiovascular outcomes including cardiac events, stroke, congestive heart failure, and death following coronary artery bypass graft surgery.^{3, 4} And, it is a particularly notable marker of increased mortality in patients with end-stage renal disease.⁵ In patients undergoing lower extremity bypass, its effects on mortality have not been as consistent and this may be due to the late-stage at which such patients present.^6-8

Pulse pressure increases with age and is related to changes in the matrix composition of large and medium-sized elastic arteries.⁹ It is also thought to be associated with increased calcification in lower extremity vessels.^{10, 11} Clinical studies suggest that increased arterial stiffness may contribute to decreased pedal perfusion in patients with peripheral artery disease (PAD).^12-17

Endovascular procedures including angioplasty and stenting are commonly used for patients with tibial artery occlusive disease presenting with critical limb ischemia (CLI). However despite increased use, primary patency has been reported to be 33% at 1 year.¹⁸ For this reason, a better understanding of the factors that contribute to success of tibial artery interventions is needed. In the coronary arteries, extensive calcification has been associated with increased rates of procedural complications and restenosis.¹⁹ Because pulse pressure is related to altered matrix, we hypothesized that it might identify patients at increased risk for treatment failure after endovascular tibial interventions. Previous studies have evaluated the effect of elevated pulse pressure on outcomes following bypass among patients with critical limb ischemia; however the effect of elevated pulse pressure on outcomes following endovascular intervention remains unclear. We thus evaluated the effects of increased pre-operative pulse pressure on adverse outcomes and mortality in patients undergoing endovascular tibial artery procedures for CLI.

METHODS

Subjects

All patients undergoing angioplasty or stenting for tibial occlusive disease at a single institution from February 2004 to May 2014 were identified using Current Procedural Terminology (CPT) codes obtained through hospital billing, administration, and operative scheduling data. Patient charts were then reviewed to confirm the procedures performed. Only patients with critical limb ischemia as defined by Rutherford chronic ischemia categories 4 – 6, identified by exam and noninvasive testing, were included in this study. Patients requiring intervention on a previous bypass and patients with acute limb ischemia were excluded. Stents were placed at the discretion of the attending surgeon. Lesion anatomy was defined according to the modified Trans Atlantic Inter-society Consensus (TASC) classification described by Dormandy and Rutherford.²⁰ Multilevel intervention was defined as angioplasty or stenting of femoral or popliteal arteries at the time of tibial intervention. Unless a contraindication was present, patients were discharged on an antiplatelet agent and statin. The Institutional Review Board of Beth Israel Deaconess Medical Center approved this study and waived the need for informed consent due to the retrospective study design.

Pulse Pressure Measurements

Preoperative systolic and diastolic blood pressures were recorded through review of the anesthesia evaluation in the pre-operative holding area. Pulse pressure was calculated as the difference between systolic and diastolic blood pressures. If both left and right brachial blood pressures were listed, the wider calculated pulse pressure was used. All other perioperative data values were collected through retrospective chart review. The mean pulse pressure in our study population was 80 mm Hg (Figure 1). Additionally, previous studies among lower extremity bypass patients as well as coronary revascularization patients utilized 80 mm/hg.^{3, 4, 6} We thus selected 80mmHg as the cutoff value for the present study.

Figure 1.

Distribution of Patient Pulse Pressures

Outcome variables

Technical success was defined as an intervention resulting in improved luminal contour at the time of completion angiography accompanied by appropriate documentation in the operative report. Procedural complications were also identified from the day of intervention by a review of the operative reports with focused searches for the terms dissection, spasm, embolism, rupture, or arteriovenous fistula. Documentation of a procedural complication did not exclude patients from consideration of technical success.

Standard follow up included duplex imaging at 3-month intervals for the first year following intervention. Restenosis was defined as a peak systolic velocity ratio >3.5 in surveillance post-procedural ultrasound or documentation in physician progress notes. Physician progress notes were included in an effort to include patients who had their follow-up imaging at another institution. Reintervention was defined as either bypass or repeat angioplasty after the initial procedure. Repeat angioplasty or stenting and bypass were defined as any ipsilateral intervention or operation related to the index procedure on the same lesion. Loss to follow up was defined by a lack of post-operative duplex imaging or subsequent vascular clinic evaluation within 12 months of the initial operation. Major amputation was defined as any amputation above the ankle of the extremity treated in the index procedure. Patients with no clinical documentation of an adverse event (procedural complication, restenosis, reintervention, major amputation) were documented as no adverse event. Mortality data was taken from the Social Security Death Index, and was documented as death within 30-day and 5-years of the intervention.

Statistical analysis

All statistical analyses were performed using Stata 12.0 (StataCorp, College Station, Texas). Pearson χ² and Fisher exact test were used for comparisons of categorical variables. Continuous variables were compared using Student t-test or Wilcoxon rank-sum test as appropriate. Univariate logistic regression was utilized to identify predictors of the outcomes studied. Those factors with p-value less than 0.1 were included in the multivariable model. Multivariate analysis was completed using logistic regression and Cox proportional hazards regression.

RESULTS

Baseline Characteristics

We identified 371 patients who underwent endovascular tibial intervention including 186 with a pre-operative pulse pressure less than 80 and 185 with a pre-operative pulse pressure greater than or equal to 80. 164 patients did not meet inclusion criteria due to endovascular intervention on a previous distal bypass graft, and 69 patients were excluded for a diagnosis of acute limb ischemia or claudication. The mean systolic blood pressure was 128 mm Hg among patients with a preoperative pulse pressure less than 80 and 162 mmHg among patients with pulse pressure greater than or equal to 80 mm Hg. Maximum systolic blood pressure and pulse pressure were moderately correlated (R=0.6). The mean arterial blood pressure (MAP) was 88 among patients without elevated preoperative pulse pressure and 97 among patients with elevated pre-operative pulse pressure. MAP was poorly correlated with pulse pressure (R=0.3). A total of 484 endovascular interventions were performed including 249 among patients with a pre-operative pulse pressure less than 80 and 235 in patients with a pulse pressure greater than or equal to 80. There were no significant differences between groups in age, and prevalence of cardiovascular risk factors including hypertension, diabetes, hyperlipidemia, and history of smoking. There were no significant differences in comorbid conditions including coronary artery disease, hemodialysis, and history of myocardial infarction, congestive heart failure, and stroke. There was no missing data for patient demographics, comorbid conditions, or mortality. There were no differences in lesion TASC classification; however patients with elevated pre-operative pulse pressure were more likely to have multilevel intervention (45% vs. 60%, P = 0.03)(Table I). The median post-operative follow-up time among living patients was 6 months: 5 months among patients with pulse pressure less than 80 and 10 months among patients with pulse pressure greater than or equal to 80 (P < .01). 320 (86%) patients had documented follow-up within 12 months of intervention.

Table I.

Demographics, co-morbidities, and TASC Class for patients with pulse pressure < 80 and ≥ 80

Demographics and Co-morbidities, No. (%)	Pulse pressure < 80 N=186 (%)	Pulse pressure ≥ 80 N=185 (%)	P-value
Mean Age (SD)	69.9 (13.4)	72.3 (11.0)	0.07
Male sex	90 (48)	102 (55)	0.19
Hypertension	150 (81)	160 (87)	0.12
Diabetes	147 (79)	146 (79)	0.94
Dialysis dependence	37 (20)	26 (14)	0.13
Hyperlipidemia	107 (58)	115 (63)	0.36
Coronary artery disease	91 (49)	94 (51)	0.72
History of myocardial infarction	36 (19)	38 (21)	0.78
Congestive heart failure	55(30)	46 (25)	0.31
History of cerebrovascular accident	25 (14)	26 (14)	0.86
Chronic Obstructive Pulmonary Disease	13 (7)	13 (7)	0.99
History of smoking	76 (51)	75 (50)	0.95
Symptoms
Rest Pain	22 (12)	24 (13)	0.72
Multilevel Intervention	84 (45)	111 (60)	0.03
TASC Class
TASC A	24 (18)	23 (15)	0.51
TASC B	36 (27)	33 (22)	0.28
TASC C	43 (32)	52 (33)	0.81
TASC D	59 (42)	67 (42)	0.99

SD: standard deviation; TASC: TransAtlantic Inter-Society Consensus

Outcomes

On univariate analysis, the technical success rates did not differ between groups (95% vs. 94%). However, we found a higher incidence of procedural complications in patients with a pre-operative pulse pressure greater than or equal to 80 mmHg (21% vs. 13%, P = 0.02) (Table II). The most commonly reported procedural complication in both groups was angioplasty induced arterial dissection followed by spasm. Rupture, distal embolization, and arteriovenous fistula were less common (Table III). In analysis of long-term results, restenosis (32% vs. 23%, P = 0.03) and reintervention (26% vs. 17%, P < 0.01) were more common in patients with elevated pre-operative pulse pressure. In subgroup analysis, subsequent endovascular intervention was significantly more common among patients with elevated pre-operative pulse pressure (16% vs. 9%, P = 0.02); however, the rates of subsequent bypass (11% vs. 8%, P = 0.15) although trending towards a difference did not achieve significance. There was no significant difference in the rate of major amputation (11% v. 13%, P = 0.45), however, there was a trend toward increased long-term mortality in patients with increased pre-operative pulse pressure P=0.07 in unadjusted results (Table II).

Table II.

Univariate effect of pulse pressure on outcomes for tibial interventions for CLI

Outcomes, No. (%)	Pulse pressure < 80 N= 249 (%)	Pulse pressure ≥ 80 N= 235 (%)	P-value
Technical success	234 (94)	223 (95)	0.52
Procedural complications	32 (13)	49 (21)	0.02
Restenosis	58 (23)	75 (32)	0.03
Reintervention	41 (17)	62 (26)	<0.01
Repeat Angioplasty or Stenting	23 (9)	38 (16)	0.02
Bypass	19 (8)	27 (11)	0.15
Major amputation	32 (13)	25 (11)	0.45
Mortality
30-day Mortality	4 (2)	7 (4)	0.35
5-year Mortality	86 (46)	103 (56)	0.07

Table III.

Detailed procedural complications following endovascular tibial intervention

Procedural Complications Number (%)	Total	Pulse pressure < 80	Pulse pressure ≥ 80	P-Value
Total	80	32 (13)	49 (21)	0.04
Dissection	51	20 (8)	31 (13)	0.07
Spasm	17	7 (3)	10 (4)	0.39
Vessel rupture	1	0 (0)	1 (0.4)	0.30
Distal embolization	7	3 (4)	4 (3)	0.91
Arteriovenous Fistula	5	2 (0.8)	3 (1.3)	0.61

We next performed multivariate analysis. In the final model, restenosis and re-intervention both lost their significant association with increased pre-operative pulse pressure; however, procedural complications (OR: 1.8, CI 1.0-3.1, P = 0.04) and 5-year mortality remained significant (OR: 1.6, CI 1.0-2.5, P = 0.04) (Table IV).

Table IV.

Multivariate odds of adverse event with pulse pressure greater than 80

Outcome	Hazard Ratio	95% Confidence Interval	P-value
Procedural complicationsa	1.8	1.0-3.0	0.04
Restenosisb	1.2	0.8-1.7	0.45
Reinterventionc	1.2	0.7-1.8	0.50
Major Amputationd	0.9	0.5-1.7	0.77
Mortalitye
30-day	2.1	0.5-8.1	0.29
5-year	1.6	1.0-2.5	0.04

^aAdjusted for Age, TASC Classification, Multilevel Intervention.

^bAdjusted for Multilevel Intervention.

^cAdjusted for Age, TASC Classification.

^dAdjusted for Age, Dialysis Dependence, Congestive Heart Failure, TASC Classification.

^eAdjusted for Age, Gender, Dialysis Dependence, Congestive Heart Failure, Chronic Obstructive Pulmonary Disease

DISCUSSION

Our study shows that in patients undergoing tibial artery interventions for critical limb ischemia, a pre-operative pulse pressure of 80 mm Hg or higher is independently associated with increased procedural complications and 5-year mortality, but not restenosis or major amputations. These data suggest that procedural complications are more likely during endovascular interventions on poorly compliant, calcified arteries, and the survival of this patient population is lower. Elevated pulse pressure is associated with adverse outcomes even after adjustment for age, pre-operative medical conditions such as diabetes and renal disease, and the extent of tibial occlusive disease.

Increased pulse pressure is strongly associated with arterial stiffness and vascular calcification, which may explain the increased frequency of procedural complication in this subgroup of patients. ^{10, 11} Calcified arteries may provide increased technical challenges for endovascular procedures. In the heart, calcified arteries have been associated with the adverse procedural complications including dissection and embolization due to increased pressure required for successful balloon angioplasty and ineffective device deployment. ^21-24 Among patients with peripheral arterial disease, calcification has also been associated with increased risk of dissection as well as major adverse limb events.^{23, 25} Furthermore, calcification may worsen the effects of critical limb ischemia and contribute to poor outcomes including amputation.^{13, 14, 25}

The increased risk of mortality among patients with elevated pre-operative pulse pressure may also reflect its association with decreased arterial compliance. Such factors decrease the ability of arteries to expand and contract normally leading to disturbed heart-vessel coupling and an increased cardiac work load.²⁶ Several studies have confirmed the association between pulse pressure and excess mortality in patients undergoing coronary artery bypass.^3,4 Increased pulse pressure is also associated with lower survival in the general population and in patients undergoing hemodialysis for chronic kidney disease.^{2, 5}

While our results concur with those of several larger trials, they differ from those of two investigations on patients undergoing lower extremity bypass and this may be related to different study populations.^{6, 7} All patients in the present study underwent endovascular tibial intervention for critical limb ischemia, whereas Mazzeffi et al. studied a population of patients undergoing bypass for a variety of indications.⁷ Asopa et al. also studied a mixed population of claudication and CLI patients.⁶ Such studies may have included patients with fewer comorbid conditions who were able to tolerate open bypass surgery. This is suggested by the higher rates of diabetes (79% vs. 66%), dialysis (17% vs. 10%) and CHF (28% vs. 18%) in our patients compared to those studied by Mazzeffi et al.⁶

These results may have important clinical implications for the treatment of patients with critical limb ischemia. Because pulse pressure is easy to calculate, it may provide a rapid noninvasive tool for risk prediction in the pre-operative period. This differs from other measures of arterial stiffness currently available, including augmentation index and carotid-femoral wave velocity, which are complex tools and not readily available to most surgeons. At present, our ability to decrease pulse pressure remains limited because most antihypertensive agents are unable to decrease systolic blood pressure while maintaining diastolic blood pressure.⁹ However, new efforts aimed at increasing arterial compliance may improve outcomes in our sickest patients. In a study by Cushman et al., hydrochlorothiazide was found reduce pulse pressure significantly more than other medications or placebo at one year; however, this study evaluated only patients treated with a single antihypertensive medication and thus the results may not be generalizable to patients with lower extremity arterial disease.²⁷ Despite these data, it remains to be determined whether reduction in pulse pressure through medical management will result in improved outcomes. Given the limited treatment options available to treat arterial stiffness, efforts to better understand its causes and effects are needed.

Our study has several limitations. First, it is based on a retrospective chart review of patient records from a single institution. It is possible that repeat intervention rates are underestimated due to underreporting or failure to capture those events that occurred at other institutions. Additionally, pulse pressure was taken from a single pre-operative measurement and may not represent patients’ average blood pressure readings or ambulatory pulse pressure. This study did not evaluate postoperative pulse pressure readings, which may have also affected patient outcomes. The impact of heart rate on pulse pressure was also not evaluated in this study based on previous studies, which have not found an association between these hemodynamic parameters.²⁸ Due to the high rate of non-compressible tibial arteries in our population, we were unable to use ABIs in our analysis. While our one year follow up of 86% was reasonable for this patient population, restenosis rates may have been underestimated due to limited follow-up beyond this time frame. Lastly, medication use and compliance may also have affected long-term mortality and we were unable to include this information due to incomplete data. Despite these limitations, the strengths of this data set are the detailed clinical variables and long term mortality data.

Conclusions

Elevated pre-operative pulse pressure is associated with an increased risk of procedural complications and late mortality among patients undergoing endovascular tibial intervention for CLI. This noninvasive test may provide valuable data for risk estimation. Investigations aimed at understanding the effect of pulse pressure on procedural and long-term outcomes are needed, as are efforts to improve morbidity and mortality in this high-risk patient population.