C-reactive protein and brain natriuretic peptide as predictors of adverse events after lower extremity endovascular revascularization

Abstract

Background

High-sensitivity C-reactive protein (hsCRP) and brain natriuretic peptide (BNP) have been shown to be independent predictors of adverse cardiovascular outcomes and increased risk of secondary interventions or limb loss in patients with peripheral arterial disease (PAD). To assist clinicians in decision-making about treatment approaches and predicting postprocedure mortality and morbidity, we retrospectively examined patients with preprocedure hsCRP and BNP levels who underwent elective angioplasty or stent placement for lower extremity PAD.

Methods

The study period was from January 1, 2007, to December 31, 2012, and patients were included who had angioplasty or stenting for PAD. Minimal required follow-up for study inclusion was at least one postoperative ankle-brachial index, contrast angiography, or duplex imaging of the treated limb. Events of interest included major adverse limb events (MALE), defined as target vessel revascularization, amputation, or disease progression by 1 year, and major adverse cardiovascular events (MACE; stroke, myocardial infarction, or death) by 2 years. Elevated/abnormal values for our biomarkers of interest were established by the upper limits of our institution's clinical laboratory reference range (hsCRP, >0.80 mg/dL; BNP, >100 pg/mL).

Results

A total of 159 limbs in 118 patients were included in analysis (42% men; median age [range], 64 [42-87] years). All limbs were symptomatic (Rutherford classification: 1-6). Iliac artery revascularization without other adjunct lower extremity intervention was performed in 60% of the limbs. High hsCRP levels (>0.80 mg/dL) were present in 32 patients (27%) and high BNP values (>100 pg/mL) in 24 patients (20%). Kaplan-Meier analysis with log-rank comparison demonstrated that elevated hsCRP levels were associated with MALE but only in limbs receiving interventions distal to the iliac arteries (P = .005). High BNP levels did not affect MALE rates (P = .821). Conversely, both elevated BNP levels (hazard ratio, 5.6; 95% confidence interval [CI], 2.0-5.8; P = .001) and hsCRP levels (hazard ratio, 2.9; 95% CI, 1.1-7.6; P = .034) predicted MACE at 2 years in the presence of confounders in Cox proportional hazards multivariate analysis. Patients with high preintervention values of hsCRP and BNP were 10.6 times (95% CI, 2.6-42.9; P = .001) more likely to experience MACE than were patients with normal hsCRP and BNP values.

Conclusions

After lower extremity endovascular interventions, elevated preprocedural hsCRP levels are associated with MALE (femoral-popliteal interventions), and elevated levels of hsCRP and BNP are associated with late cardiovascular events.

Because atherosclerosis is an inflammatory process involving the walls of the arterial system, a large body of research exists regarding the utility of inflammatory proteins as biomarkers of peripheral arterial disease (PAD).^1-3 Particular interest has been paid to high-sensitivity C-reactive protein (hsCRP), a marker of early inflammatory changes that is believed to be increased in PAD in the absence of cardiovascular disease, diabetes, or hypertension.^4,5 The hsCRP levels also carry predictive value for endovascular treatment of PAD. Elevated preintervention hsCRP has been shown to predict restenosis after below-knee percutaneous angioplasty,⁶ the need for reintervention, index limb amputation, and all-cause mortality in patients after endovascular therapy (EVT) for PAD.^7-9

Brain natriuretic peptide (BNP) is a well-characterized hormone important in fluid and blood pressure regulation.¹⁰ BNP is also elevated in patients with PAD, which is likely due to the correlation between PAD and ischemic heart disease.¹¹ Data have also shown that PAD patients with elevated BNP have increased all-cause mortality.¹² Studies also demonstrate that elevated preoperative BNP in patients undergoing aortic aneurysm repair, carotid endarterectomy, and peripheral vein bypass correlates with increased risk for postoperative cardiac events including nonfatal myocardial infarction and cardiovascular death.^13,14 However, the predictive value of BNP in PAD patients undergoing EVT remains unknown.

Although studies have demonstrated that hsCRP and BNP are independent predictors of adverse events in PAD patients, the two have not been studied together after EVT. Therefore, we retrospectively examined the relationship between preintervention hsCRP and BNP levels and the occurrence of adverse events in a cohort of patients who underwent elective angioplasty or stent placement for lower extremity PAD. By doing this, we hope to provide clinicians with additional tools to predict perioperative mortality and morbidity and to guide decisions about treatment approaches and postintervention surveillance after EVT for PAD.

Methods

We retrospectively examined patients who underwent elective EVT in the lower extremities by the primary author (P.A.S.) at Charleston Area Medical Center (CAMC), Charleston, West Virginia, between January 1, 2007, and December 31, 2012. The sample was identified by use of International Classification of Diseases, Ninth Revision (ICD-9) diagnostic codes for PAD in combination with ICD-9 procedure codes for EVT (Table I). Inclusion criteria included the presence of preintervention hsCRP and BNP levels obtained within 30 days before the procedure per standard protocol of the admitting physician (P.A.S.), elective EVT including angioplasty or stent placement, and at least one postoperative follow-up consisting of one or more of the following: duplex ultrasound examination, conventional contrast angiography, or ankle-brachial index (ABI) measurements. Exclusion criteria included patients younger than 18 years, emergency lower limb revascularization, lack of hsCRP or BNP levels before the procedure, lack of postoperative follow-up (ABI, angiography, or duplex ultrasound), and patients with concomitant procedures to treat arterial aneurysms. All aspects of the study were reviewed and approved by the CAMC/West Virginia University, Charleston Division, Institutional Review Board.

Table I. International Classification of Diseases, Ninth Revision (ICD-9) diagnosis and procedure codes used to identify patients receiving endovascular therapy (EVT) for peripheral arterial disease (PAD).

Code	Description
PAD ICD-9 diagnosis codes
250.7	Diabetes with peripheral circulatory disorders
440.2×	Atherosclerosis of native arteries of extremities
440.3×	Atherosclerosis of extremities of unspecified graft
440.4	Chronic total occlusion of artery of the extremities
443.9	Peripheral vascular disease, unspecified
444.22	Arterial embolism and thrombosis of lower extremity
in combination with
Endovascular ICD-9 procedure codes
39.50	Angioplasty or atherectomy of other non-coronary vessel(s)
39.90	Insertion of non-drug-eluting peripheral vessel stent(s)
00.55	Insertion of drug-eluting peripheral vessel stent(s)

Our two primary outcomes of interest included major adverse limb events (MALE) and major adverse cardiovascular events (MACE). We defined MALE as the composite end point encompassing target vessel revascularization (TVR), limb amputation, or disease progression by 12 months after the procedure. Disease progression was defined as the total occlusion of the treated vessel as determined by angiography or duplex ultrasound or a greater than 0.15 decrease in ABI in the treated limb.^15,16 MACE was the composite end point composed of stroke, myocardial infarction, and death occurring by 2 years after the procedure. Secondary end points included the individual components of the two primary end points. Follow-up for MALE continued from the time of the procedure until an event or the most recent angiography, duplex ultrasound, or ABI measurement (mean follow-up, 15 ± 12 months). Follow-up for MACE continued from the time of the procedure until the occurrence of an event or the most recent admission or visit at CAMC or an affiliated clinic or office (mean follow-up, 30 ± 17 months).

Collected demographic data and health history evaluation included age, gender, race, body mass index, and current tobacco use. Comorbidity evaluation included obesity (body mass index >30), diabetes mellitus, hypertension, hyperlipidemia, coronary artery disease, chronic kidney disease, and heart failure. Recent events or conditions, with recent defined as occurring within 30 days before the procedure, that can induce a systemic inflammation response were examined and included the following: trauma or illness; recurring hospitalization; and the occurrence of stroke, transit ischemic attack, myocardial infarction, or liver failure. Preprocedure clinical data evaluated included ABI, home medication, history of prior vascular intervention, Rutherford classification of lower extremity ischemia, and laboratory values for hsCRP and BNP. The hsCRP level was measured by immunometric assay, with detection in the lower limit of 0.02 mg/dL and the upper limit of 10.0 mg/dL. Levels of BNP were measured by two-site sandwich immunoassay. Elevated/abnormal values for our biomarkers of interest were established by the upper limit of our institution's clinical laboratory reference range (high hsCRP, >0.80 mg/dL; high BNP, >100 pg/mL). In addition, information on the vessel treated, the presence of occlusion, whether the procedure was TVR, the type and number of stents, and the length of stent required were included in data collection. All-cause mortality after discharge was determined by the Social Security Death Index.

Data analysis

SPSS version 19.0 was used for data analysis. Descriptive statistics included frequencies and percentages for categorical variables; means ± standard deviations were used for continuous variables, except for age, for which median [range] was used. Univariate analysis to determine differences between groups (elevated vs normal preprocedure biomarker levels) for patient and limb characteristics was performed by χ² analysis (Fisher exact test when appropriate) for categorical data and Mann-Whitney U analysis for age The adverse outcome of MALE was analyzed on a per limb basis, whereas MACE was analyzed per patient. Kaplan-Meier curves with log-rank analysis were used to examine differences in time to events. We used univariate and multivariate Cox proportional hazards regression analyses to examine the relationship of risk factors, including elevated hsCRP and BNP levels, with the adverse outcomes of MALE and MACE. Criterion for inclusion in the multivariate model was a univariate analysis value of P < .10, with forward stepwise regression entry employed. A P level < .05 was considered statistically significant.

Results

A total of 163 patients having hsCRP and BNP testing before lower extremity EVT were identified. On chart review, 45 patients were excluded (14 for lacking follow-up, 12 for concomitant open vascular procedures, 12 for procedures outside the lower extremities, six for concomitant abdominal aortic procedures, and one patient for having an emergent procedure to treat acute limb ischemia). A total of 159 limbs in 118 patients undergoing endovascular interventions to the iliac or lower extremity arteries served as our study cohort. The study comprised 50 men and 68 women with a median age [range] of 64 [42-87] years. One patient had an inflammation-inducing event (myocardial infarction) in the 30 days previous to EVT. No other patients had a trauma or illness requiring hospitalization or an inflammation-inducing event in the 30 days before the procedure. The mean level of hsCRP in our study cohort was 0.98 ± 1.54 mg/dL, with elevated hsCRP levels (>0.80 mg/dL) present in 32 patients (27.1%). The mean level of BNP was 90.2 ± 133.4 pg/mL, with high BNP values (>100 pg/mL) in 24 patients (20.3%). Patient demographics are presented in Table II as a function of elevated hsCRP and BNP levels.

Table II. Patient demographics as a function of elevated high-sensitivity C-reactive protein (hsCRP) or brain natriuretic peptide (BNP) levels.

	hsCRP			BNP
	Normal ≤0.80 mg/dL (n = 86)	Abnormal/high >0.80 mg/dL (n = 32)	P	Normal ≤100 pg/mL (n = 94)	Abnormal/high >100 pg/mL (n = 24)	P
Male gender	44%	37%	.538	48%	2%	.031
Median age, years (range)	63 (42-87)	65 (45-83)	.670	61 (42-87)	69 (51-87)	.001
Nonwhite	4%	0%	.562	3%	0%	1.000
Obesity	28%	38%	.435	34%	17%	.161
Current tobacco use	57%	53%	.868	61%	38%	.071
Diabetes mellitus	41%	44%	.929	44%	33%	.496
Hypertension	85%	94%	.350	86%	92%	.733
Hyperlipidemia	65%	72%	.636	64%	79%	.237
Coronary artery disease	54%	59%	.716	49%	79%	.015
Chronic kidney disease	11%	16%	.523	7%	29%	.008
Heart failure	5%	3%	1.000	5%	0%	.582
Statin on admission	63%	50%	.295	57%	67%	.557

Bold entries indicate statistical significance.

All limbs were symptomatic and were categorized into clinical stages of PAD according to the Rutherford classification. A total of 133 limbs (84%) had reported claudication (Rutherford stages 1-3) and 26 (16.4%) had critical limb ischemia, with 17 limbs (11%) having rest pain (Rutherford stage 4) and nine limbs (6%) having tissue loss or ulceration (Rutherford stage 5-6). The distribution of vessels with treated lesions was iliac arteries in 111 limbs (70%), femoropopliteal artery in 58 limbs (37%), tibial/below-the-knee interventions in six limbs (4%), and previously placed bypass grafts in six limbs (4%), with 22 limbs (14%) having more than one vessel treated. A total of 96 limbs (60%) received EVT in the iliac region without concomitant procedures to distal arteries. The type of intervention consisted of angioplasty without stenting in 26 limbs (16%) and angioplasty with stenting in 133 limbs (84%). Intervention was performed secondary to occlusion in 57 limbs (35.8%). One stent was placed in 85 limbs (53.5%), with more than one stent required in 48 limbs (30%). In the 133 limbs receiving stents, bare metal stents were used in 118 limbs (89%), covered stents in three limbs (2%), and drug-eluting stents in 12 limbs (9%). Mean stent length was 101.2 ± 100.6 mm (range, 20-430 mm). Limb and procedural characteristics are presented as a function of elevated hsCRP and BNP levels in Table III.

Table III. Limb and procedural characteristics presented as a function of high-sensitivity C-reactive protein (hsCRP) or brain natriuretic peptide (BNP) levels.

		hsCRP			BNP
	Total	Normal ≤0.80 mg/dL (n = 115)	Abnormal/high >0.80 mg/dL (n = 44)	P	Normal ≤100 pg/mL (n = 124)	Abnormal/high >100 pg/mL (n = 35)	P
Indication for EVT
Claudication	84%	88%	73%	.039	86%	74%	.151
Critical limb ischemia	16%	12%	27%		14%	26%
Treated lesion location
Iliac arteries only	60%	60%	61%	1.00	62%	54%	.523
SFA, popliteal, tibial, or bypass	40%	40%	39%		38%	46%
Vessel occluded	36%	33%	43%	.314	34%	43%	.436
Stent use	84%	87%	75%	.113	85%	80%	.688

EVT, Endovascular therapy; SFA, superficial femoral artery.

Bold entries indicate statistical significance.

After endovascular intervention, the event-free survival rates from MALE were 91% at 6 months and 80% at 12 months. Kaplan-Meier analysis with log-rank analysis demonstrated that elevated hsCRP levels were not associated with MALE (P = .127; Fig 1, A). However, when limbs that received EVT distal to the iliac arteries (n = 63) were examined independently, hsCRP levels >0.80 mg/dL had significantly elevated rates of MALE compared with normal hsCRP levels (P = .005). There was no effect of elevated hsCRP levels on MALE in limbs with iliac arteries treated only (P = .438). High BNP levels (>100 pg/mL) were not associated with MALE in noniliac (P = .343) or iliac (P = .806) revasculariza-tions or in all treated limbs (P = .822; Fig 1, B). No limbs underwent amputation within 12 months of the procedure. However, 19 limbs required TVR (14 endovascular, five surgical reinterventions) within the initial year after the procedure. High hsCRP levels were associated with the need for TVR (P = .048), whereas elevated BNP levels were not associated (P = .633; Table IV). Neither hsCRP nor BNP elevated levels predicted MALE in univariate Cox proportional hazards analyses (hsCRP: hazard ratio [HR], 1.9; 95% confidence interval [CI], 0.8-4.2; P = .134; and BNP: HR, 0.9; 95% CI, 0.3-2.4; P = .823), with P values above the .10 threshold for inclusion in the multivariate analysis as described in the Methods section. Thus, multivariate analysis for predictors for MALE is not presented.

Fig 1.

A, Kaplan-Meier curve showing event-free survival from major adverse limb event (MALE) at 1 year in limbs categorized by preprocedure levels of high-sensitivity C-reactive protein (hsCRP). B, Kaplan-Meier curve showing event-free survival at 1 year from MALE in limbs categorized by preprocedure levels of brain natriuretic peptide (BNP).

Table IV. Major adverse limb events (MALE) occurring in the initial year after the procedure.

	hsCRP			BNP
	Normal ≤0.80 mg/dL (n = 115)	Abnormal/high >0.80 mg/dL (n = 44)	P	Normal #100 pg/mL (n = 124)	Abnormal/high >100 pg/mL (n = 35)	P
One year free from
TVR	88%	71%	.048	85%	81%	.557
Amputation	100%	100%	—	100%	100%	—
Disease progression	84%	74%	.337	82%	81%	.924

BNP, Brain natriuretic peptide; hsCRP, high-sensitivity C-reactive protein; TVR, target vessel revascularization.

The rates of freedom from MACE in our patient sample at 12 months and 24 months were 92% and 81%, respectively. MACE events occurring in the 2-year period are presented in Table V. Patients with high hsCRP trended toward having decreased freedom from MACE (68% vs 86%), but the difference failed to meet statistical significance (P = .053) as shown in Fig 2, A. Conversely, elevated BNP values were associated with decreased event-free rates from MACE (P = .015), with a rate of 64% vs 86% in patients with normal BNP levels (Fig 2, B). In univariate Cox proportional hazards analyses, elevated BNP levels predicted MACE at 2 years (HR, 3.1; 95% CI, 1.2-8.0; P = .021), whereas high hsCRP values trended toward but failed to reach statistical significance for predicting MACE (HR, 2.4; 95% CI, 1.0-6.1; P = .061). Multivariate analysis demonstrated that both elevated BNP and hsCRP levels predicted MACE in the presence of confounders (BNP: HR, 5.6; 95% CI, 2.0-15.8; P = .001; and hsCRP: HR, 2.9; 95% CI, 1.1-7.6; P = .034; Table VI).

Table V. Major adverse cardiovascular events (MACE) occurring in 2 years after the procedure.

	hsCRP			BNP
	Normal ≤0.80 mg/dL (n = 86)	Abnormal/high >0.80 mg/dL (n = 32)	P	Normal ≤100 pg/mL (n = 94)	Abnormal/high > 100 pg/mL (n = 24)	P
Two years free from
Myocardial infarction	99%	79%	.004	93%	95%	.884
Stroke	94%	90%	.672	96%	76%	.032
All-cause death	94%	82%	.030	95%	72%	.003

BNP, Brain natriuretic peptide; hsCRP, high-sensitivity C-reactive protein.

Fig 2.

A, Kaplan-Meier curve showing event-free survival from major adverse cardiovascular events (MACE) at 2 years in patients categorized by preprocedure levels of high-sensitivity C-reactive protein (hsCRP). B, Kaplan-Meier curve showing event-free survival from MACE at 2 years in patients categorized by preprocedure levels of brain natriuretic peptide (BNP).

Table VI. Cox proportional hazards model analysis of predictors for major adverse cardiovascular events (MACE).

	HR (95% CI)	P value
Univariate analysis
Elevated hsCRP (>0.80 mg/dL)	2.4 (1.0-6.2)	.061
Elevated BNP (>100 pg/mL)	3.1 (1.2-8.0)	.021
Age >65 years	3.3 (1.2-9.3)	.023
Male gender	0.8 (0.3-2.0)	.613
Obesity (body mass index >30)	1.2 (0.5-3.2)	.703
Current tobacco use	1.0 (0.4-2.5)	.948
Diabetes mellitus	1.3 (0.5-3.3)	.578
Hypertension	2.9 (0.4-21.8)	.302
Hyperlipidemia	1.4 (0.5-3.8)	.564
Coronary artery disease	3.1 (1.0-9.6)	.043
Chronic kidney disease	2.9 (1.0-8.3)	.040
Heart failure	8.9 (1.8-43.9)	.007
No statin at discharge	2.6 (1.0-6.8)	.047
No clopidogrel at discharge	1.4 (0.5-3.6)	.541
Multivariate analysisa
Elevated hsCRP	2.9 (1.1-7.6)	.034
Elevated BNP	5.6 (2.0-15.8)	.001
Heart failure	27 (4.7-157)	<.001
No statin at discharge	3.2 (1.2-8.6)	.020

BNP, Brain natriuretic peptide; CI, confidence interval; HR, hazard ratio; hsCRP, high-sensitivity C-reactive protein.

^aMultivariate model included variables having P < .10 by univariate analyses and employed forward stepwise entry.

When hsCRP and BNP levels were examined in combination, we found that having elevated preintervention levels of both hsCRP and BNP did not predict MALE at 1 year (HR, 2.0; 95% CI, 0.4-8.7; P = .377) in comparison to those with normal biomarker levels (Fig 3, A). However, patients having both elevated hsCRP and BNP values were 10.6 times (95% CI, 2.6-42.9; P = .001) more likely to experience the adverse outcome of MACE within 2 years of the procedure in comparison to patients with normal hsCRP and BNP values (Fig 3, B).

Fig 3.

A, Hazard ratios for major adverse limb event (MALE) at 1 year for preprocedure high-sensitivity C-reactive protein (hsCRP) levels in combination with brain natriuretic peptide (BNP) levels. B, Hazard ratios of major adverse cardiovascular events (MACE) at 2 years for preprocedure hsCRP levels in combination with BNP levels. High hsCRP equaled levels >0.80 mg/dL; high BNP equaled levels >100 pg/mL.

Discussion

Biochemical markers assist in the diagnosis and management of vascular patients. D-dimer levels are used to help determine risk of deep venous thrombosis and in some settings have eliminated the need for duplex ultrasound examinations on the basis of the biomarker's high negative predictive value. Postoperatively, troponin levels aid in identifying patients with non-ST myocardial infarctions. BNP has been used to quantify degrees of heart failure and as a predictor for long-term mortality.¹⁷ In the cardiac literature, preprocedural hsCRP levels have been associated with an increased risk of coronary restenosis after intervention. Buffon et al,¹⁸ more than a decade ago, reported more than twice the incidence of clinical coronary restenosis at 1 year after angioplasty in patients with elevated levels of preprocedural hsCRP vs those possessing levels within the reference range. Aligned with the Achilles' heel of peripheral revascularization (ie, development of recurrent restenosis or progression of disease), elevated hsCRP has been associated with increased risk of amputation or reintervention of the treated limb.^8,9 Thus, we believe that if preprocedure biochemical markers in endovascular interventions for PAD predict future limb-related complications and late cardiovascular events, the information may improve outcomes by directing aggressive secondary prevention in patients with high markers.

In our current study, the patient cohort included those receiving elective iliac, femoropopliteal, or tibial interventions during a 6-year period performed by a single practitioner. The majority of lesions treated were in the iliac territory, and elevated hsCRP levels were not associated with an increased risk of repeated intervention or subsequent ipsilateral limb loss. However, in limbs treated below the inguinal ligament, high hsCRP levels revealed an increased risk of secondary procedures or intervention failure. This is comparable to the results of Ishii et al,⁹ who evaluated the risk of secondary interventions and major amputation by quartiles of preoperative hsCRP levels and demonstrated by multivariate analysis that the risk of ipsilateral limb events was related to hsCRP, elevated TransAtlantic Inter-Society Consensus lesion class, and presence of ulcer or gangrene. Similarly, Schillinger et al,¹⁹ in evaluating patients with angioplasty of the femoropopliteal segment, found that baseline CRP levels predicted 12-month restenosis.

Some major differences between our study and that of Ischii et al are that their series included a much larger cohort with 234 patients, nearly twice the sample size of our group, and all patients in the study of Ischii et al had end-stage renal disease (ESRD) requiring hemodialysis compared with <10% of our patients having end-stage renal disease. In addition, their mean follow-up was longer at 33 months compared with our mean follow-up of 15 months. It is possible that at a longer follow-up period, we will be able to demonstrate an increased risk for either MALE or MACE with elevated hsCRP values for the entire cohort. Last, our group was composed of lesions predominantly treated in the iliac region, 60% in our patient cohort compared with 31% in that of Ischii et al, which could strongly affect results because restenosis and reinterventions are much less common in the iliac arteries compared with the femoropopliteal segment and tibial arteries. Other reports examining significance of hsCRP in patients with PAD have focused on the relationship between inflammation and disease progression. Vainas et al²⁰ measured baseline hsCRP levels and ABI at initial evaluation and subsequently at 12-month clinical follow-up. Patients with elevated hsCRP statistically were more likely to show a reduction in ABI at their 12-month follow-up compared with those with lower hsCRP values. Furthermore, Owens et al,²¹ in evaluating prospectively 91 patients undergoing infrainguinal vein bypass surgery, found that preoperative elevated hsCRP values showed a strong correlation with adverse conduit-related and cardiovascular outcomes.

Plasma BNP levels, on the other hand, have been primarily used as a method to monitor intravascular fluid levels and to drive therapy for congestive heart failure and as a predictor of adverse events after surgery. In our series, patients with baseline BNP values above normal levels demonstrated higher rates of major adverse events, including mortality during the follow-up period. Limited published information is available on mortality and whether BNP values are predictive of outcome for those undergoing EVT for PAD. BNP has been demonstrated to be an independent predictor of postoperative myocardial injury in patients undergoing abdominal aortic aneurysm surgery or in patients with subcritical limb ischemia.²² In addition to perioperative cardiac events, BNP has been associated with major adverse cardiac events and mortality for years after major vascular surgery as demonstrated by the same author in a later publication²³ as well as by others.^12-14

The most impressive finding in our series is the combined effect of an increased hsCRP and BNP on MACE and mortality. The combination of elevated levels of both biomarkers poses a 10 times hazard risk compared with individuals with normal values of both markers. The dramatic risk in this subset may help further target medical strategies for those at highest risk for poor long-term survival. Future studies may look specifically at statin therapy to lower hsCRP values in patients with normal lipid profiles. This would include assessing hsCRP before elective endovascular interventions and maximizing statin medication therapy to assess improvement in limb outcomes and mortality. The JUPITER study, with a planned 5-year follow-up, randomized patients with normal low-density lipoprotein and elevated hsCRP levels (>2.0 mg/L) to either placebo or rosuvastatin.²⁴ The trial was stopped at <2 years secondary to outcomes. There were 142 major cardiovascular events in the treated arm vs 251 in the placebo cohort, resulting in a 44% reduction of MACE in those medicated. Rosuvastatin reduced hsCRP levels by 37%, but without a control group with normal hsCRP levels, there was no clear indication that elevated hsCRP is required for rosuvastatin to provide a benefit regarding cardiovascular adverse events. Further research is needed to clarify the role of statins as a method of reducing hsCRP levels in PAD patients to improve outcome.

Limitations

The main limitation of the study is that the sample size was small, and outcomes may have been affected if a larger sample size were included. The main example of this was that patients with a high hsCRP trended toward having decreased freedom from MACE (68% vs 86%) but failed to meet statistical significance (P = .053). A larger sample size may have resulted in this demonstrating a statistically significant difference. Last, this was a retrospective study, and only patients were included who had appropriate follow-up. Furthermore, whereas our vascular practice advocates that patients follow up with color duplex ultrasound or ABI at 1 and 6 months and then every 12 months after EVT, patient compliance is imperfect. Patients who are doing well tend to be less likely to follow up, and this may have affected the outcomes of the study.

Conclusions

Biochemical markers in patients undergoing lower extremity intervention can serve as a predictor of both limb-related events (femoral-popliteal interventions) and MACE, especially when hsCRP and BNP are used in combination.