ADVERTISEMENT
Predictors of In-hospital and 30-day Complications of Peripheral Vascular Interventions Using Bivalirudin as the Primary Anticoa
July 2005
Predictors of complications in peripheral percutaneous interventions (PPI) with bivalirudin as a base anticoagulant have not yet been defined. The Angiomax Peripheral Procedure Registry of Vascular Events (APPROVE)1 offers a unique opportunity to analyze predictors of complications with bivalirudin in renal, iliac and femoral interventions.
Multiple publications have addressed complications of PPI in the periphery with unfractionated heparin (UFH)
as a primary anticoagulant.2–10 These predictors have typically been the state of the limb (rest ischemia versus claudication), distal runoff, lesion characteristics (length, occlusion versus stenosis, location), intra-procedural thrombus, renal insufficiency, age, gender, smoking history, procedure length, number of lesions and vascular beds treated in one setting, and presence of various comorbidities (diabetes and coronary artery disease). We have recently reported predictors of complications with UFH as a base anticoagulant in PPI.11 Using Binary Logistic Regression analysis, current smoking (p = 0.014), recent onset of claudication (p = 0.052), and the presence of intra-procedural thrombus (p = 0.011) were strong predictors of the occurrence of serious adverse events. In this study, higher doses of UFH were clinically associated with higher complication rates.
In contrast to UFH, bivalirudin has a short half-life, provides predictable anticoagulation response, and inhibits free and bound thrombin, as well as thrombin-induced platelet activation, aggregation and granule release.12 Given the major pharmacologic differences between UFH and bivalirudin, traditional predictors of complications with UFH in PPI might differ with the use of bivalirudin as a primary anticoagulant. In this study, data from the APPROVE registry were analyzed for predictors of complications in PPI with bivalirudin. The registry was large, and prospective, with well-defined ischemic and bleeding endpoints.
Methods
The Angiomax Peripheral Procedure Registry of Vascular Events (APPROVE) trial was a prospective, open-label, multi-center clinical trial conducted at 26 centers in the United States. Details of this trial have been described previously.1 Patients undergoing interventions on their renal, iliac or femoral arteries were included. Patients were pre-treated with aspirin (300–325 mg p.o.) and clopidogrel (300 mg p.o.) up to 12 hours prior to the intervention. Bivalirudin was administered intravenously at a dose of 0.75 mg/kg bolus followed by an infusion of 1.75 mg/kg/hour for the duration of the procedure. Glycoprotein (GP) IIb/IIIa inhibitors were permitted at the discretion of the treating physician. With the exception of those patients who received a closure device, sheaths were removed 2 hours after discontinuation of the bivalirudin infusion.
In order to determine predictors of complications of PPI in this registry, the following variables were included in the model: lesion characteristics (occlusion versus stenosis and location of the lesion; length of lesion was not collected in this registry and occlusion > 15 cm were excluded), creatinine clearance, age, gender, smoking history, number of lesions and vascular beds treated in one setting, presence of various comorbidities (diabetes, coronary artery disease and congestive heart failure), weight, anemia at baseline (Hb 1
Statistical analysis. Descriptive analysis was performed using means for continuous variables and percentages for categorical ones. Logistic regression analysis with backward elimination was utilized to model the effect of the independent variables on outcome. A p-value of Results
Baseline demographics are shown in Table 1. Multiple comorbidities are noted, including hypertension (89.3%), congestive heart failure (15.6%) and diabetes (36.3%). There was also a higher prevalence of coexistent coronary artery disease. Table 2 shows the procedural details. Intra-procedural thrombus was present in 15 (3%) patients. Aspirin was given in 96.8% of the patients ( 2 stents.
Ischemic and bleeding complications are shown in Table 3. There was no death at discharge or at 30 days. Unplanned urgent revascularization, amputation or myocardial infarction remained at less than 1%. Major in-hospital and 30-day bleeding occurred in 2.2% and 3.8% of the patients, respectively.
Logistic regression analysis (Table 4) showed that the strongest predictors of in-hospital ischemic and major bleeding events were congestive heart failure (p = 0.0173) and ELS (p = 0.0045). Twenty-six patients had more than one sheath exchange, with 24 of these patients having an exchange to a larger sheath size (10 patients upsized by 1 size, 9 patients by 2 sizes, 5 patients by 3 sizes and 2 with no change in size, but a different access in 1 patient). When the number of sheath exchanges utilized is removed from the model as the number of these patients is small, congestive heart failure was the single most important predictor of ischemic/major bleeding (p = 0.031). At 30 days, the combined ischemic and bleeding events were predicted by the presence of lower lipid values (p = 0.034) and the number of ELS (p = 0.0113).
When logistic regression analysis was conducted with major bleeding alone as the dependent variable, the strongest predictor of major bleeding at discharge (p = 0.0041) and at 30 days (p = 0.0016) was the ELS. Also female gender (p = 0.08) and low weight (stratified by gender with 92 kg for males and 77 kg for females) (p = 0.096) showed a trend toward predicting major bleeding at 30 days. Finally, ELS (p = 0.0009) was also a significant independent in-hospital predictor of vascular complications (pseudoaneurysm and perforations). If the number of sheath exchanges is removed from the model, no significant predictor for major bleeding could be identified. The type of lesion (thrombotic versus atherosclerotic plaque), its location and severity, age and renal insufficiency were not predictors of ischemic and/or bleeding events.
Discussion
The APPROVE registry evaluated the feasibility of using bivalirudin in a relatively low-risk cohort of patients that included renal, iliac and femoral PPI, with only 3% of patients had identifiable thrombotic lesions. In this patient population, predictors of the combined bleeding and ischemic endpoints were congestive heart failure, ELS and lower lipid levels, whereas predictors of protocol-defined major bleeding alone were the ELS, female gender and low weight.
Although 22 patients received GP IIb/IIIa inhibitors, none of these patients had thrombotic lesions (data on 2 patients are missing). Both GP IIb/IIIa inhibitor use and thrombotic lesions were included in the logistic regression model, but none was a predictor of ischemic and/or bleeding complications. This is in contrast to data on UFH where intra-procedural thrombus was a predictor of adverse events.11 It is possible that the ability of bivalirudin to penetrate and inhibit clot-bound thrombin, provide reliable anticoagulation and its lack of platelet activation properties1 are responsible for this difference with UFH. Also, in contrast to UFH, diabetes, current smoking, lesion severity and renal insufficiency were not predictors of adverse events with bivalirudin use as a primary anticoagulant. A recent meta-analysis13 in patients undergoing percutaneous coronary interventions has shown that bivalirudin provides a greater absolute benefit in patients with chronic renal dysfunction compared to UFH.
In APPROVE, ELS appears to be a very strong predictor of major bleeding thus, keeping entry sites to a minimum with the use of smaller sheaths might reduce this complication. PPI is generally a longer procedure than PCI, and on occasion it requires multiple entry points with larger sheath insertion. In APPROVE, 26 patients required sheath exchanges, of which 24 were to larger sheath sizes. Despite this small number, these patients accounted for many of the major bleeding events that occurred (11.5% versus 1.7% in-hospital bleeding incidents for the patients with ELS versus single sheath usage, respectively, and 15.4% versus 3.1% for 30-day bleeding rates). In addition, female gender and low weight (adjusted for gender) had a trend toward being significant predictors of bleeding. In the coronary arteries, weight extremes were shown to be associated with an increase in major adverse events.14 Also, female gender has been associated with a higher rate of bleeding complications during PPI.15 The association of lower lipid values and an increase in major bleeding/ischemic complications is unclear. It is possible that these are high-risk patients who have been aggressively treated with statins, and therefore with lower lipid values. In APPROVE, the use of statins was not recorded.
Complications in the periphery have been strongly linked to the state of the limb (ischemic versus claudicants) and to vessel runoff. However, in the APPROVE registry, no data were collected on limb status and the protocol mandated at least one good runoff to the lower extremities. Further data in chronic limb ischemia and high-risk patients are needed to determine whether the state of the limb remains a predictor of complications with bivalirudin.
Study limitations. APPROVE was a registry inolving a predominantly low-risk patient population with a low incidence of thrombotic lesions. Therefore, we cannot extrapolate these data to a higher risk cohort of patients with chronic or acute limb ischemia. Also, these data cannot be extrapolated to predictors of complications in patients undergoing infra-popliteal interventions as these patients were excluded from this registry. This registry, however, provides data for the first time on predictors of complications in PPI with bivalirudin as a base anticoagulant and allows the hypothesis that these predictors might be different from traditional ones seen with UFH.
Acknowledgment. The authors are indebted to Weihong Fan, MSc, for statistical analysis and to Meredith Todd in helping in the preparation of this manuscript
shammas@mchsi.com
1. Allie DE, Hall P, Shammas NW, et al. The Angiomax Peripheral Procedure Registry of Vascular Events Trial (APPROVE): In-hospital and 30-day results. J Invasive Cardiol 2004;16:651–656.
2. Becker GJ, McClenny TE, Kovacs ME, et al. The importance of increasing public and physician awareness of peripheral arterial disease. J Vasc Interv Radiol 2002;13:7–11.
3. McDaniel MD, Cronenwett JL. Basic data related to the natural history of intermittent claudication. Ann Vasc Surg 1989;3:273–277.
4. Matsi PJ, Manninen HI. Complications of lower-limb percutaneous transluminal angioplasty: A prospective analysis of 410 procedures on 295 consecutive patients. Cardiovasc Intervent Radiol 1998;21:361–366.
5. Gutteridge W, Torrie EP, Galland RB. Cumulative risk of bypass, amputation or death following percutaneous transluminal angioplasty. Eur J Vasc Endovasc Surg 1997;14:134–139.
6. Axisa B, Fishwick G, Bolia A, et al. Complications following peripheral angioplasty. Ann R Coll Surg Engl 2002;84:39–42.
7. Morse MH, Jeans WD, Cole SE, et al. Complications in percutaneous transluminal angioplasty: Relationships with patient age. Br J Radiol 1991;64:5–9.
8. Hasson JE, Acher CW, Wojtowycz M, et al. Lower extremity percutaneous transluminal angioplasty: Multifactorial analysis of morbidity and mortality. Surgery 1990;108:748–752.
9. Boyer L, Therre T, Garcier JM, et al. Infrapopliteal percutaneous transluminal angioplasty for limb salvage. Acta Radiol 2000;41:73–77.
10. Greenfield AJ. Femoral, popliteal, and tibial arteries: percutaneous transluminal angioplasty. AJR Am J Roentgenol 1980;135:927–935.
11. Shammas NW, Lemke JH, Dippel EJ, et al. In-hospital complications of peripheral vascular interventions using unfractionated heparin as the primary anticoagulant. J of Invas Cardiol 2003;15:242–246.
12. Ogilby JD, Kopelman HA, Klein LW, Agarwal JB. Adequate heparinization during PTCA: Assessment using activated clotting times. Cathet Cardiovasc Diagn 1989;18:206–209.
13. Chew DP, Bhatt DL, Kimball W, et al. Bivalirudin provides increasing benefit with decreasing renal function: a meta-analysis of randomized trials. Am J Cardiol 2003;92:919–23.
14. Minutello RM, Chou ET, Hong MK, et al. Impact of body mass index on in-hospital outcomes following percutaneous coronary intervention (report from the New York State Angioplasty Registry). Am J Cardiol 2004;93:1229–1232.
15. Krikorian RK, Kramer PH, Vacek JL. Percutaneous revascularization of lower extremity arterial disease in females compared to males. J Invasive Cardiol 1997;9:333–338.
