Advertisement

Editor's Choice – Protamine Reduces Serious Bleeding Complications Associated with Carotid Endarterectomy in Asymptomatic Patients without Increasing the Risk of Stroke, Myocardial Infarction, or Death in a Large National Analysis

Open ArchivePublished:October 27, 2020DOI:https://doi.org/10.1016/j.ejvs.2020.08.047

      Objective

      Controversy persists regarding the use of protamine during carotid endarterectomy (CEA), despite real world evidence to support its use. The purpose of this study was to determine the impact of protamine reversal of heparin anticoagulation on the outcome of CEA in the USA.

      Methods

      A prospective national registry (Society for Vascular Surgery Vascular Quality Initiative) of 72 787 patients undergoing elective asymptomatic CEA by 1879 surgeons from 316 centres in the USA and Canada from 2012 to 2018 was reviewed. Protamine use varied by both surgeon (20% rare use [< 10%], 30% variable use [11%–79%], 50% routine use [> 80% cases]) and geographical region (44% vs. 96%). Temporal trends in protamine use were also determined. End points included post-operative re-operation for bleeding, as well as potential protamine related thrombotic complications, including stroke, death, and myocardial infarction (MI). Predictors of end points were determined by multivariable logistic regression. Propensity matching was additionally used to control for differences between groups.

      Results

      Of the 72 787 patients who underwent CEA, 69% received protamine, while 31% did not. Protamine use increased over time from 60% (2012) to 73% (2018). In total, 378 patients (0.7%) in the protamine treated group underwent re-operation for bleeding vs. 342 patients (1.4%) in the untreated cohort (p < .001). Protamine use did not affect the rate of MI (0.7% vs. 0.8%; p = .023), stroke (1.1% vs. 1.0%; p = .20), or in hospital death (0.2% vs. 0.2%; p = 0.70) between treated and untreated patients, respectively. On multivariable analysis, protamine use was independently associated with reduced risk of re-operation for bleeding (odds ratio 0.5, 95% confidence interval 0.39–0.55; p < .001). Independent of protamine exposure, the consequences of a return to the operating room (RTOR) for bleeding were statistically significant, with a sevenfold increase in MI (RTOR 4.9% vs. no RTOR 0.7%; p < .001), an eightfold increase in stroke (RTOR 7.2% vs. no RTOR 0.9%; p < .001), and a 13 fold increase in death (RTOR 2.4% vs. no RTOR 0.2%; p < .001).

      Conclusion

      Protamine reduces serious bleeding complications at the time of CEA without increasing the risk of MI, stroke, or death, in this large North American analysis. Based on this and previous regional work regarding protamine use in CEA, it is believed that there is now sufficient evidence to support its routine use, and it should be considered as a benchmark for quality during CEA.

      Keywords

      This study documents the safety and efficacy of using protamine sulphate to reverse intra-operative heparin at the time of carotid endarterectomy (CEA) using a large national analysis. The work demonstrates a 50% reduction in re-operation for bleeding events with no attendant increased risk of stroke, myocardial infarction or death in patients who received protamine. Furthermore, this analysis highlights the precarious nature of operative re-exploration for bleeding in patients who did not receive protamine with substantially increased rates of thrombotic complications. The significance of this national real world data, which appears to be concordant with current European Society of Vascular Surgery guidelines for protamine administration during carotid endarterectomy, and reflects prior work, is that over one third of elective CEA procedures in North America fail to comply with treatment guidelines. Based on this analysis, surgeons should strongly consider using protamine at the time of CEA to optimise surgical outcomes.

      Introduction

      In an attempt to minimise the incidence of peri-operative stroke, considerable attention has been devoted to optimising surgical technique regarding carotid endarterectomy (CEA).
      • Stone D.H.
      • Nolan B.W.
      • Schanzer A.
      • Goodney P.P.
      • Cambria R.A.
      • Likosky D.S.
      • et al.
      Protamine reduces bleeding complications associated with carotid endarterectomy without increasing the risk of stroke.
      • Weber C.F.
      • Friedl H.
      • Hueppe M.
      • Hintereder G.
      • Schmitz-Rixen T.
      • Zwissler B.
      • et al.
      Impact of general versus local anesthesia on early postoperative cognitive dysfunction following carotid endarterectomy: GALA Study Subgroup Analysis.
      • Lutz H.-J.
      • Michael R.
      • Gahl B.
      • Savolainen H.
      Local versus general anaesthesia for carotid endarterectomy – improving the gold standard?.
      • Aburahma A.F.
      • Stone P.A.
      • Elmore M.
      • Flaherty S.K.
      • Armistead L.
      • AbuRahma Z.
      Prospective randomized trial of ACUSEAL (Gore-Tex) vs Finesse (Hemashield) patching during carotid endarterectomy: long-term outcome.
      • Naylor R.
      • Hayes P.D.
      • Payne D.A.
      • Allroggen H.
      • Steel S.
      • Thompson M.M.
      • et al.
      Randomized trial of vein versus Dacron patching during carotid endarterectomy: long-term results.
      • Cao P.
      • Giordano G.
      • De Rango P.
      • Zannetti S.
      • Chiesa R.
      • Coppi G.
      • et al.
      Eversion versus conventional carotid endarterectomy: late results of a prospective multicenter randomized trial.
      • Hans S.S.
      • Jareunpoon O.
      Prospective evaluation of electroencephalography, carotid artery stump pressure, and neurologic changes during 314 consecutive carotid endarterectomies performed in awake patients.
      Nevertheless, persistent controversy and variation in practice persists surrounding technical nuances of the procedure. Specifically, the intra-operative administration of protamine sulphate to reverse heparin anticoagulation remains one such point of contention.
      Protamine advocates will champion reduced bleeding without an attendant increased risk of stroke and other perceived thrombotic complications. Surgeons opposed to its use will often cite a perceived increased risk of thrombotic events, including stroke or myocardial infarction (MI), or even anaphylaxis and pulmonary hypertension, with a low observed risk of bleeding complications. The historical literature devoted to this question was both conflicted and probably underpowered to detect definitively significant disparities in outcome for these low frequency events.
      • Dellagrammaticas D.
      • Lewis S.C.
      • Gough M.J.
      GALA Trial Collaborators
      Is heparin reversal with protamine after carotid endarterectomy dangerous?.
      • Levison J.A.
      • Faust G.R.
      • Halpern V.J.
      • Theodoris A.
      • Nathan I.
      • Kline R.G.
      • et al.
      Relationship of protamine dosing with postoperative complications of carotid endarterectomy.
      • Treiman R.L.
      • Cossman D.V.
      • Foran R.F.
      • Levin P.M.
      • Cohen J.L.
      • Wagner W.H.
      The influence of neutralizing heparin after carotid endarterectomy on postoperative stroke and wound hematoma.
      • Fearn S.J.
      • Parry A.D.
      • Picton A.J.
      • Mortimer A.J.
      • McCollum C.N.
      Should heparin be reversed after carotid endarterectomy? A randomised prospective trial.
      • Mauney M.C.
      • Buchanan S.A.
      • Lawrence W.A.
      • Bishop A.
      • Sinclair K.
      • Daniel T.M.
      • et al.
      Stroke rate is markedly reduced after carotid endarterectomy by avoidance of protamine.
      In 2010 a regional analysis of the use of protamine during CEA across New England demonstrated that its use significantly reduced serious peri-operative bleeding events without any observed increased risk of stroke.
      • Stone D.H.
      • Nolan B.W.
      • Schanzer A.
      • Goodney P.P.
      • Cambria R.A.
      • Likosky D.S.
      • et al.
      Protamine reduces bleeding complications associated with carotid endarterectomy without increasing the risk of stroke.
      At that time, protamine was used in just over 40% of CEAs performed across New England. A subsequent analysis by Patel et al.,
      • Patel R.B.
      • Beaulieu P.
      • Homa K.
      • Goodney P.P.
      • Stanley A.C.
      • Cronenwett J.L.
      • et al.
      Shared quality data are associated with increased protamine use and reduced bleeding complications after carotid endarterectomy in the Vascular Study Group of New England.
      focusing on the regional impact of this work, demonstrated that protamine use increased over time across New England and, consequently, bleeding events were reduced by 50% over the same time period. While these results are laudable, a significant percentage of CEAs are still performed without protamine. Moreover, it remains uncertain whether this regional analysis would accurately reflect outcomes more broadly across the USA.
      Accordingly, the purpose of this study was to determine the impact of protamine reversal of heparin anticoagulation on the outcome of asymptomatic CEA throughout the USA and Canada using the Society for Vascular Surgery (SVS) Vascular Quality Initiative (VQI). Specifically, the aim was to determine whether protamine had a significant effect on the incidence of bleeding complications, measured as re-operation for bleeding, as well as an impact on the incidence of thrombotic events, most importantly stroke, MI, and death.

      Materials and Methods

      Subjects and Database

      This study was approved by the research advisory committee and includes real world prospectively compiled registry data from all regional quality groups within the VQI. Hospitals participating in the CEA registry are required to enter all consecutive cases and undergo routine external audit to verify complete case capture. All data reported herein adhere to the ethical reporting guidelines as stipulated by the SVS patient safety organisation VQI governance.
      All CEAs that were submitted to the VQI database from 2012 to 2018 were identified (n = 95 930). Symptomatic and/or non-elective presentations were excluded (n = 12 368). Combined coronary artery bypass graft/CEA procedures was also excluded from the analysis (n = 4 951). All procedures with missing data and/or operations performed at the weekend (owing to the likelihood of these being non-elective procedures) were further excluded (n = 1 296), resulting in a total of 77 315 CEAs performed in 72 787 patients (Fig. S1; see Supplementary Material). Heparin and protamine use during surgery were at the discretion of the operating surgeon. All patient demographic and clinical variables (>100 in total) are prospectively aggregated in the VQI and were reviewed in this analysis.

      Outcomes and Variable Definitions

      The definitions for specific comorbidities, procedure related variables, and post-operative complications are available upon request online at www.vascularqualityinitiative.org/components/svs-pso.
      • Cronenwett J.L.
      • Kraiss L.W.
      • Cambria R.P.
      The society for vascular surgery vascular quality initiative.
      The end points of this study were post-operative bleeding requiring re-operation (to determine a potential benefit of protamine) and post-operative MI, stroke, or death (to reflect thrombotic complications potentially associated with protamine exposure). All endpoints were restricted to hospitalisation for CEA (median one day). The stroke end point included both major (defined as disability causing non-independent living status) and minor (defined as non-disabling) events. MI was defined by either troponin level or electrocardiogram changes detected during routine clinical care. All mortality events were verified by cross linking to the Social Security Death Index.

      Statistical Analysis

      The primary focus of the analysis was to determine whether protamine exposure increased or decreased the risk of post-operative return to the operating room for bleeding. To further understand the safety of protamine administration, a separate analysis was conducted to identify whether protamine administration was associated with increased risk of possible post-operative thrombotic complications such as MI, stroke, or in hospital death. Variables associated these end points were initially identified by univariable analysis using the Wilcoxon rank sum test and Pearson chi square test when comparing continuous and categorical variables, respectively. Owing to covariable imbalance between protamine treated and untreated groups, logistic regression was then used to assess the relative risk of post-operative return to the operating room for bleeding and/or the individual thrombotic complications (MI, stroke, and in hospital death) between groups. Variables found to be statistically significant at the p < .10 level on univariable analysis were then entered into a multivariable model using backwards stepwise logistic regression. Hazard ratios (HRs) or odds ratios (ORs) and 95% confidence intervals (CIs) were estimated for each risk factor, along with p values for Wald tests of HR = 1 or OR = 1.
      To further corroborate the multivariable logistic regression model, a separate analysis using inverse propensity weighting scores was employed to further address confounding. Overlap of predicted protamine use and non-protamine treatment probability distributions was sufficient for propensity score inverse weighting. In the inverse propensity scores (IPS) weighted models, each observation in the protamine group was weighted by 1/(probability of receiving protamine), and each observation in the no protamine group was weighted by 1/(1−probability of receiving protamine). Thus, observations of patients who received protamine despite a low probability of treatment, and those of patients who did not receive protamine despite a high probability of treatment, had more influence in the weighted models, thereby adjusting results for covariable imbalance. This method was successful in simulating relatively excellent covariable balance between protamine treated and non-treated groups (Table 1, Table 2; Tables S1 and S2 [see Supplementary Material]).
      • Austin P.C.
      An introduction to propensity score methods for reducing the effects of confounding in observational studies.
      Table 1Candidate predictors used in the protamine propensity score model development to evaluate the impact of protamine reversal of heparin anticoagulation on the outcome of carotid endarterectomy in the USA
      Surgery year
      Age
      Body mass index
      Sex
      Non-white ethnicity
      Any smoking history
      Transfer from nursing home
      Non-ambulatory pre-operative state
      ASA categories 3, 4, and 5 vs. 1 and 2
      Hypertension
      Coronary artery disease
      Congestive heart failure
      Pre-operative creatinine >1.8 g/dL
      Dialysis dependence pre-operatively
      Positive stress test pre-operatively
      Remote history of prior neurological event
      Remote history of neurological event means any stroke and/or transient ischaemic attack occurring >3 months pre-operatively.
      Prior history of non-specified event
      Pre-operative P2Y12 receptor antagonist
      Pre-operative statin
      Pre-operative beta blocker
      Pre-operative anticoagulation
      Prior history of coronary artery bypass graft
      Prior history of PCI/stent
      History of neck radiation exposure
      ASA = American Society of Anesthesiologists; PCI = percutaneous coronary intervention.
      Remote history of neurological event means any stroke and/or transient ischaemic attack occurring >3 months pre-operatively.
      Table 2Covariable balancing achieved by inverse propensity score weighting (IPW) for the impact of protamine reversal of heparin anticoagulation on the outcome of carotid endarterectomy in Vascular Quality Initiative centres in the USA
      VariableUnweighted before IPWWeighted after IPW
      No protamine (n = 23 966)Protamine (n = 53 349)No protamine (n = 23 966)Protamine (n = 53 349)
      Surgery year2015201520152015
      Body mass index28.5 ± 6.128.4 ± 5.928.4 ± 6.028.4 ± 5.9
      Non-white1 917 (8.0)4 801 (9.0)1 917 (8.0)4 267 (8.0)
      Any smoking17 734 (74.0)40 545 (76.0)17 974 (75.0)40 011 (75.0)
      ASA status 3, 4, 522 288 (93.0)50 681 (95.0)22 528 (94.0)50 148 (94.0)
      Non-ambulatory240 (1.0)534 (1.0)240 (1.0)533 (1.0)
      Hypertension21 329 (89.0)48 014 (90.0)21 329 (89.0)47 480 (89.0)
      Coronary artery disease6 231 (26.0)14 404 (27.0)6 470 (27.0)14 404 (27.0)
      PCI4 793 (20.0)12 270 (23.0)5 272 (22.0)11 736 (22.0)
      Congestive heart failure2 396 (10.0)5 868 (11.0)2 396 (10.0)10 669 (10.0)
      COPD5 512 (23.0)12 270 (23.0)5 512 (23.0)12 270 (23.0)
      Prior neurological event10 065 (42.0)22 406 (42.0)10 065 (42.0)22 406 (42.0)
      Pre-operative P2Y126 950 (29.0)18 138 (34.0)7 669 (32.0)17 071 (32.0)
      Pre-operative beta blocker13 181 (55.0)30 408 (57.0)13 660 (57.0)30 408 (57.0)
      Pre-operative anticoagulation2 396 (10.0)4 801 (9.0)2 396 (10.0)5 334 (10.0)
      Prior CABG4 794 (20.0)11 203 (21.0)4 794 (20.0)10 669 (2.00)
      Data are presented as n (%) or mean ± standard deviation. These numbers reflect the propensity scores. ASA = American Society of Anesthesiologists; PCI = percutaneous coronary intervention; COPD = chronic obstructive lung disease; CABG = coronary artery bypass graft.
      In addition, several procedural factors were incorporated into the analyses, including CEA type (eversion vs. standard), vein patch use, dextran exposure, surgical drain, re-exploration of the artery prior to leaving operating room, post-operative hypotension requiring intravenous (IV) medications, as well as arrival and highest intra-operative heart rate. To compare the observed effect of protamine use when accounting for these procedural factors, as well as pre-operative patient characteristics, four models were generated for each outcome, with the following sets of independent variables: (1) protamine + all pre-operative patient factors; (2) protamine + patient factor based IPS weight; (3) protamine + pre-operative patient factors + procedural factors; (4) protamine + IPS weight + procedural factors. For all outcomes, the effect of protamine was nearly unchanged, suggesting that the reported effects are largely independent of statistical method.
      A power calculation was performed and based on the sample size; there was 88% power to detect a true difference in return to operating room for bleeding rates as small as 0.007 (e.g., 0.94% vs. 0.87%). However, a twofold difference in the no protamine vs. the protamine treated group was seen, a very large effect size for an event with incidence of <1%. A p value < .05 was considered statistically significant for all analyses. Statistical analysis was performed using SAS version 9.4 (SAS Institute, Cary, NC, USA).

      Results

      During the 77 315 CEAs performed, 69.0% (n = 53 349) received intra-operative protamine, while 31.0% (n = 23 966) did not. Procedures were performed electively, using general anaesthesia, with the majority undergoing patch angioplasty and intra-operative shunting. Eighty-seven per cent of the asymptomatic cohort had >70% internal carotid artery stenosis. The majority of patients in both the protamine treated and untreated groups were taking antiplatelet therapy (84% vs. 83%; p = .024) at the time of CEA. Additional demographic details, comorbidities, and pre-operative medications of the study cohort are given in Table 3.
      Table 3Demographics and comorbidities of patients treated with or without protamine during elective asymptomatic carotid endarterectomy in Vascular Quality Initiative centres in the USA
      VariableProtamine (n = 53 349)No protamine (n = 23 966)p value
      Age – years70.0 ± 9.169.1 ± 9.2.20
      Male sex32 235 (60)14 314 (60).071
      Diabetes19 129 (36)8 632 (36).70
      Hypertension47 844 (90)21 192 (88)<.001
      Coronary artery disease14 470 (27)6 336 (26).044
      Congestive heart failure5 883 (11)2 357 (10)<.001
      Current or former smoker40 382 (76)17 768 (74)<.001
      COPD12 129 (23)5 490 (23).60
      Creatinine >1.8 gm/dL2 243 (4)1 030 (4).50
      Aspirin44 642 (84)19 903 (83).021
      Clopidogrel17 838 (33)6 964 (29)<.001
      Statin43 969 (82)19 490 (81)<.001
      Beta blocker30 525 (57)13 169 (55)<.001
      General anesthesia48 749 (91)22 173 (93)<.001
      Patch arterioplasty48 856 (92)20 495 (86)<.001
      Data are presented as n (%) or mean ± standard deviation. COPD = chronic obstructive pulmonary disease.
      On the one hand, over time, protamine use at the time of CEA increased throughout the USA, reaching rates of nearly 70%. On the other hand, nearly one third of CEAs performed in the USA are currently performed without protamine (Fig. 1A). When stratified by frequency of use, 50% of VQI surgeons used protamine routinely (defined as used in >80% of their CEAs), while 20% used it rarely (defined as used in <10% of their CEAs) and 30% used protamine selectively during surgery (defined as used in 10%–70% of CEAs performed). Moreover, when protamine use was stratified by region, there was similar heterogeneity in practice. Specifically, some regions used protamine almost routinely, achieving >90% utilisation, while other geographical areas used protamine selectively, using it in <50% of recorded CEAs (Fig. 1B).
      Figure 1
      Figure 1Variation in protamine use during elective carotid endarterectomy (CEA). (A) Temporal trends in protamine use across Vascular Quality Initiative (VQI) centres in the USA during elective asymptomatic CEA from 2012 and 2018. Red lines denote mean percentage of protamine use for designated time intervals. (B) Regional variation in protamine use during elective CEA across VQI designated regions. VQI mean protamine use was 69% (dashed line).
      Re-operation for bleeding was performed in 1.4% (n = 342) of patients who did not receive protamine vs. 0.7% (n = 378) of those who did (p < .001; Fig. 2). The observed length of stay was over twofold greater in patients who underwent re-operation for bleeding vs. those who did not (median 1.0 [interquartile range {IQR} 1–2] vs. 3.0 [IQR 2–6] days; p < .001). Conversely, thrombotic complications were not significantly affected by protamine administration. There were no observed statistically significant differences in the incidence of stroke (0.96% vs. 1.1%), death (0.2% vs. 0.21%), or clinically apparent MI (0.68 vs. 0.83%) in patients who received protamine vs. those who did not (Fig. 3). However, irrespective of protamine treatment the observed consequences of re-operation for bleeding were highly significant. Specifically, there was a sevenfold observed increase in the incidence of clinically statistically significant MI in patients requiring re-operation compared with those patients who did not (4.9% vs. 0.7%; p < .001). In addition, there was an eightfold increase in the incidence of stroke in patients who required re-operation for bleeding (7.2% vs. 0.9%; p < .001). Finally, there was a 13 fold increase in the incidence of in hospital death among patients who required re exploration for bleeding vs. patients who did not require re-operation for bleeding (2.4% vs. 0.2%; p < .001) (Fig. 4).
      Figure 2
      Figure 2Re-operation for bleeding rates with 95% confidence intervals in asymptomatic carotid endarterectomy (CEA) in 53 349 protamine treated and 23 965 untreated patients in Vascular Quality Initiative centres in the USA. The figure reflects non-risk adjusted data.
      Figure 3
      Figure 3Thrombotic complication (myocardial infarction [MI], stroke, and death) event rates with 95% confidence intervals after elective carotid endarterectomy for asymptomatic carotid artery stenosis in 53 349 protamine treated and 23 965 untreated patients at the time of surgery in Vascular Quality Initiative centres in the USA. This figure reflects non-risk adjusted data. NS = not significant.
      Figure 4
      Figure 4Consequences of Reoperation for bleeding. Clinical sequelae (myocardial infarction [MI], stroke, and death) with 95% confidence intervals associated with re-operation for bleeding after elective carotid endarterectomy for asymptomatic carotid artery stenosis in 53 349 protamine treated and 23 965 untreated patients in Vascular Quality Initiative centres in the USA. p < .001 for all intergroup comparisons.
      Multivariable logistic regression confirmed that protamine administration was a statistically significant independent predictor of reduced risk of re-operation for bleeding, after accounting for other variables, including centre variation, surgical technique, and antiplatelet therapy (OR 0.46, 95% CI 0.39–0.55; p < .001). In addition, carotid eversion appeared to demonstrate a statistically significant association with increased re-operation for bleeding (OR 1.4, 95% CI 1.1–1.7; p = .002) (Table 4).
      Table 4Multivariable independent predictors of return to the operation room for bleeding after elective carotid endarterectomy in Vascular Quality Initiative centres in the USA
      PredictorContrastOR (95% CI)p value
      ProtamineY: N0.46 (0.39–0.55)<.001
      Body mass indexPer unit increase0.98 (0.96–0.99)<.008
      SexM: F1.3 (1.1–1.5).002
      Congestive heart failureY: N1.5 (1.2–1.9)<.001
      Percutaneous coronary interventionY: N1.3 (1.1–1.6).002
      Remote history of prior neurological eventY: N1.2 (1.0–1.4).014
      Pre-operative P2Y12 antagonistY: N1.6 (1.4–1.9)<.001
      Pre-operative anticoagulationY: N1.3 (1.0–1.7).032
      Type of carotid endarterectomyEversion: Patch1.4 (1.1–1.7).002
      Re-exploration of arteryY: N15.1 (12.2–18.8)<.001
      Hypotension requiring intravenous medicationsY: N1.8 (1.4–2.2)<.001
      High intra-operative heart rate (bpm)(3 knot RCS)<.001
      OR = odds ratio; CI = confidence interval; Y = yes; N = no; M = male; F = female; bpm = beats per minute; RCS = restricted cubic spline.
      In order to identify any other potential complications of protamine use, several additional variables were examined. Protamine use was associated with a slightly higher risk of hypotension requiring IV medication by univariable analysis (10% vs. 9.0%; p < .001). However, this association was not statistically significant in the multivariable model when adjusted for potentially confounding variables such as the factors that predicted a return to the operating room for bleeding. In fact, under multiple different modelling frameworks using IPS weighting, protamine administration was consistently identified to be associated with a statistically significantly decreased risk of a return to the operating room for bleeding (Table 5).
      Table 5Protamine association with estimated risk of return to the operation room for bleeding in Vascular Quality Initiative centres in the USA under different modelling frameworks
      Effect or confounderContrastOR (95% CI)p value
      Protamine unweighted + pre-operative patient factorsY: N0.49 (0.42–0.58)<.001
      Protamine IPS weighted
      Procedural factors included carotid endarterectomy type (eversion vs. standard), vein patch use, dextran exposure post-operatively, drain left in surgical wound, re-exploration of artery prior to leaving the operating room with index procedure, hypotension requiring IV medications, high intra-operative heart rate (beats per minute), and arrival heart rate.
      Y: N0.47 (0.40–0.55)<.001
      Protamine IPS weighted + procedural factorsY: N0.47 (0.40–0.56)<.001
      Protamine + pre-operative patient factors + procedural factorsY: N0.46 (0.39–0.55)<.001
      OR = odds ratio; CI = confidence interval; Y = yes; N = no; IPS = inverse propensity scores.
      Procedural factors included carotid endarterectomy type (eversion vs. standard), vein patch use, dextran exposure post-operatively, drain left in surgical wound, re-exploration of artery prior to leaving the operating room with index procedure, hypotension requiring IV medications, high intra-operative heart rate (beats per minute), and arrival heart rate.

      Discussion

      This study represents the largest and most definitive analysis to date demonstrating a reduction in serious bleeding complications at the time of CEA associated with protamine administration. Moreover, this large national prospective registry study did not reveal any increase in the incidence of in hospital stroke, MI, or death in patients receiving protamine. Specifically, patients who did not receive protamine were documented to experience more than a twofold increased risk of re-operation for bleeding. In addition, this large national analysis confirms the results of an initial regional analysis from 2010, justifying the routine use of protamine during CEA, and appears to be concordant with current European Society of Vascular Surgery practice guidelines (recommendation 52, class IIA, Level B).
      • Stone D.H.
      • Nolan B.W.
      • Schanzer A.
      • Goodney P.P.
      • Cambria R.A.
      • Likosky D.S.
      • et al.
      Protamine reduces bleeding complications associated with carotid endarterectomy without increasing the risk of stroke.
      ,
      • Naylor A.R.
      • Ricco J.-B.
      • de Borst G.J.
      • Debus S.
      • de Haro J.
      • Halliday A.
      • et al.
      Editor's choice - Management of atherosclerotic carotid and vertebral artery disease: 2017 clinical practice guidelines of the European Society for Vascular Surgery (ESVS).
      It should be noted that while temporal trends pertaining to protamine administration during CEA suggest that its use has increased over time, 30% of CEAs in VQI participating centres are still performed without it. While the root cause for this observation remains unknown, one can hypothesise that this ongoing variation reflects a historically conflicted and underpowered literature addressing this topic and engrained preconceived surgeon bias largely influenced by the posture of mentors and historical training paradigms. Given the observed overall low peri-operative event rates surrounding CEA, surgeons may feel falsely validated in their respective surgical approach, unable to detect statistically significant disparities among low frequency event signals.
      Indeed, as highlighted previously, there has only been one randomised trial addressing the use of protamine reversal of heparin at the time of CEA. Unfortunately, the study documented 64 patients of whom only 31 received intra-operative protamine. Most notably, however, the authors highlighted that two protamine treated patients acutely thrombosed their respective operated internal carotid arteries and died. Based on these findings, the authors concluded that protamine use may predispose to internal carotid artery thrombosis and subsequent stroke. Given the extremely small sample size, these results are certainly susceptible to error, making it difficult to extrapolate these results to guide practice.
      • Fearn S.J.
      • Parry A.D.
      • Picton A.J.
      • Mortimer A.J.
      • McCollum C.N.
      Should heparin be reversed after carotid endarterectomy? A randomised prospective trial.
      It should be further emphasised that given the low frequency of these end points in real world practice, it is unlikely that this trial could sufficiently answer the intended question for which it was designed.
      Conversely, a post-hoc sub-analysis from the GALA trial (General Anaesthetic vs. Local Anaesthetic for Carotid Surgery Trial) examined outcomes of patients treated with protamine vs. those who were untreated.
      • Dellagrammaticas D.
      • Lewis S.C.
      • Gough M.J.
      GALA Trial Collaborators
      Is heparin reversal with protamine after carotid endarterectomy dangerous?.
      Specifically, the authors documented a 4.4% stroke rate in untreated patients vs. 2.9% in protamine treated patients (p = .10). Furthermore, the study demonstrated a statistically lower incidence of associated neck haematoma in the protamine treated group compared with the untreated cohort (7.4% vs. 10.4%; p = .040). The authors concluded that their findings refuted the perception that protamine is associated with deleterious outcomes during CEA. It should be noted that the GALA trial was not designed to study the effects of protamine during CEA and only 28% of the study cohort received protamine.
      • Weber C.F.
      • Friedl H.
      • Hueppe M.
      • Hintereder G.
      • Schmitz-Rixen T.
      • Zwissler B.
      • et al.
      Impact of general versus local anesthesia on early postoperative cognitive dysfunction following carotid endarterectomy: GALA Study Subgroup Analysis.
      Nevertheless, this analysis demonstrates the historical conflict within the literature thereby perpetuating practice variation.
      These, now repeated, findings of reduced re-operation for bleeding risk associated with protamine use in conjunction with the sub-analysis from the GALA trial of a lower incidence of neck haematomas offers significant clinical insight. Surgeons, who forego protamine use, espousing a more liberal approach to managing bleeding or haematomas post-operatively, are exposing patients to a second operation, which appears to be associated with an increased risk of MI, stroke, and death. However, it cannot be discerned whether this observation is directly related to protamine use during the index operation. This finding is consistent with the 2010 study. Specifically, in this new national analysis of >70 000 CEAs, patients who returned to the operating room for bleeding experienced a sevenfold increased risk of MI, an eightfold increased risk of stroke, and a 13 fold increased risk of death vs. patients who did not return to the operating room. Although the underlying aetiologies for the increased incidence of these complications cannot be stipulated, they demonstrably refute the notion that returning to the operating room for bleeding is a clinically trivial event, and, in part, further justifies the routine administration of protamine during elective CEA. It is important to emphasise that protamine administration did not protect against stroke, but rather from re-operation for bleeding, which is associated with additional risk of post-operative stroke.
      Interestingly, this analysis demonstrated that protamine was an independent predictor of reduced re-operation for bleeding by both multivariable analysis, as well as propensity matching to control for differences between the large cohorts of treated and untreated patients. While the true incidence of anaphylactic reactions to protamine in this study cannot be commented on, it has been documented elsewhere to be <1%.
      • Levison J.A.
      • Faust G.R.
      • Halpern V.J.
      • Theodoris A.
      • Nathan I.
      • Kline R.G.
      • et al.
      Relationship of protamine dosing with postoperative complications of carotid endarterectomy.
      Moreover, a prior report demonstrated that protamine use was not associated with hypotension requiring IV medication.
      • Stone D.H.
      • Nolan B.W.
      • Schanzer A.
      • Goodney P.P.
      • Cambria R.A.
      • Likosky D.S.
      • et al.
      Protamine reduces bleeding complications associated with carotid endarterectomy without increasing the risk of stroke.
      It remains difficult to comment further on protamine associated hypotension as the threshold for treatment varies by provider. Nevertheless, given the low frequency of a clinically severe allergic reaction and seeming benignity of any medication associated hypotension, it would appear that the benefits of protamine outweigh the potential risks.
      Additionally, concordant with the repeated findings surrounding the beneficial impact of intra-operative protamine use, a meta-analysis by Newhall et al.
      • Newhall K.A.
      • Saunders E.C.
      • Larson R.J.
      • Stone D.H.
      • Goodney P.P.
      Use of Protamine for anticoagulation during carotid endarterectomy: a meta-analysis.
      documented that protamine was not associated with an increased risk of stroke, while decreasing the observed rates of re-operation for bleeding. It is important to note that this comprehensive analysis of 12 observational studies and >10 000 patients re-affirmed the current study findings with no increased risk of MI, stroke, or death, and a concordant reduction in significant bleeding events in protamine treated patients. Moreover, this overview of applicable literature offers a sample size sufficiently powered to detect subtle disparities in low frequency events unlike many historical series that were unable to do so. These findings were re-affirmed by a meta-analysis by Kakisis et al.,
      • Kakisis J.D.
      • Antonopoulos C.N.
      • Moulakakis K.G.
      • Schneider F.
      • Geroulakos G.
      • Ricco J.B.
      Protamine reduces bleeding complications without increasing the risk of stroke after carotid endarterectomy: a meta-analysis.
      which concluded that the use of protamine for heparin reversal during CEA reduced the risk of haematoma without increased stroke risk.
      This study had several limitations. Firstly, it was unable to provide level I evidence commensurate with a randomised controlled trial. The analysis of end points only reflected in hospital events and 30 day event rates cannot be commented on. Adverse post-operative events occurring after hospital discharge also cannot be commented on, but protamine administration is done intra-operatively, so the analysis focused on in hospital outcomes owing to the clinical likelihood of a corresponding neurological event. However, it does provide the largest experience to date reflecting real world practice in both academic and private hospitals across the USA and Canada.
      Despite its inherent strengths, the limitations of registry data are recognised. Regarding this, cross validation of VQI outcomes to claims data has been performed.
      • Columbo J.A.
      • Kang R.
      • Hoel A.W.
      • Kang J.
      • Leinweber K.A.
      • Tauber K.S.
      • et al.
      A comparison of reintervention rates after endovascular aneurysm repair between the Vascular Quality Initiative registry, Medicare claims, and chart review.
      ,
      • Hoel A.W.
      • Faerber A.E.
      • Moore K.O.
      • Ramkumar N.
      • Brooke B.S.
      • Scali S.T.
      • et al.
      A pilot study for long-term outcome assessment after aortic aneurysm repair using Vascular Quality Initiative data matched to Medicare claims.
      All cases are audited for consecutive case capture, and outcomes linkage to alternative data sources has validated the quality of the data captured in VQI. Given the non-likelihood of a randomised trial to address this subject, it is believed that this report, in conjunction with a prior regional study, reproducibly justifies the routine use of protamine for heparin anticoagulation reversal during CEA. Importantly, the dosing of both heparin and protamine, and the potential impact this could have on measured end points, could not be identified. Similarly, activated clotting times could not be commented on as this is not recorded in the registry. It is important to note that strokes reported in the CEA VQI registry are not neurologist adjudicated. Therefore, the event rate for neurological outcomes reported herein may be underestimated. Furthermore, stroke severity and type (e.g., haemorrhagic vs. ischaemic) is not well characterised in the registry and thus clinical differences related to the stroke events between groups cannot be commented on.
      Finally, the overwhelming majority of patients were taking antiplatelet therapy at the time of surgery, which could confound the proclivity for bleeding. Furthermore, the VQI does not record if patients are on dual antiplatelet therapy, so the additive impact of this therapy on subsequent risk of re-operation for bleeding could not be determined. However, neither aspirin nor clopidogrel use were independent predictors of a return to the operating room for bleeding. Nevertheless, it is believed that these results are sufficiently powered to detect the low frequency events sought in this analysis and furthermore incorporate the practice patterns and inherent variation among regions, centres, and surgeons included in the VQI. It is important to note that clinical adherence to ESVS guidelines could not be commented on as this data set does not reflect practice patterns in Europe.

      Conclusion

      Based on this national analysis of nearly 73 000 asymptomatic CEAs, protamine use was associated with a 50% reduction in serious bleeding events, measured by re-operation for bleeding. Furthermore, there was no association with an increased incidence of MI, stroke, or peri-operative mortality. The consequences of re-operation for bleeding were substantial, with demonstrably higher rates of MI, stroke, and death. This analysis confirms the similar findings of a previous regional study. Given these reproducible, large, real world studies, it is believed that the data surrounding protamine use at the time of CEA justifies its routine use and should be considered as a quality benchmark during carotid surgery.

      Conflicts of interest

      None.

      Funding

      None.

      References

        • Stone D.H.
        • Nolan B.W.
        • Schanzer A.
        • Goodney P.P.
        • Cambria R.A.
        • Likosky D.S.
        • et al.
        Protamine reduces bleeding complications associated with carotid endarterectomy without increasing the risk of stroke.
        J Vasc Surg. 2010; 51: 559-564
        • Weber C.F.
        • Friedl H.
        • Hueppe M.
        • Hintereder G.
        • Schmitz-Rixen T.
        • Zwissler B.
        • et al.
        Impact of general versus local anesthesia on early postoperative cognitive dysfunction following carotid endarterectomy: GALA Study Subgroup Analysis.
        World J Surg. 2009; 33: 1526-1532
        • Lutz H.-J.
        • Michael R.
        • Gahl B.
        • Savolainen H.
        Local versus general anaesthesia for carotid endarterectomy – improving the gold standard?.
        Eur J Vasc Endovasc Surg. 2008; 36: 145-149
        • Aburahma A.F.
        • Stone P.A.
        • Elmore M.
        • Flaherty S.K.
        • Armistead L.
        • AbuRahma Z.
        Prospective randomized trial of ACUSEAL (Gore-Tex) vs Finesse (Hemashield) patching during carotid endarterectomy: long-term outcome.
        J Vasc Surg. 2008; 48: 99-103
        • Naylor R.
        • Hayes P.D.
        • Payne D.A.
        • Allroggen H.
        • Steel S.
        • Thompson M.M.
        • et al.
        Randomized trial of vein versus Dacron patching during carotid endarterectomy: long-term results.
        J Vasc Surg. 2004; 39: 985-993
        • Cao P.
        • Giordano G.
        • De Rango P.
        • Zannetti S.
        • Chiesa R.
        • Coppi G.
        • et al.
        Eversion versus conventional carotid endarterectomy: late results of a prospective multicenter randomized trial.
        J Vasc Surg. 2000; 31: 19-30
        • Hans S.S.
        • Jareunpoon O.
        Prospective evaluation of electroencephalography, carotid artery stump pressure, and neurologic changes during 314 consecutive carotid endarterectomies performed in awake patients.
        J Vasc Surg. 2007; 45: 511-515
        • Dellagrammaticas D.
        • Lewis S.C.
        • Gough M.J.
        • GALA Trial Collaborators
        Is heparin reversal with protamine after carotid endarterectomy dangerous?.
        Eur J Vasc Endovasc Surg. 2008; 36: 41-44
        • Levison J.A.
        • Faust G.R.
        • Halpern V.J.
        • Theodoris A.
        • Nathan I.
        • Kline R.G.
        • et al.
        Relationship of protamine dosing with postoperative complications of carotid endarterectomy.
        Ann Vasc Surg. 1999; 13: 67-72
        • Treiman R.L.
        • Cossman D.V.
        • Foran R.F.
        • Levin P.M.
        • Cohen J.L.
        • Wagner W.H.
        The influence of neutralizing heparin after carotid endarterectomy on postoperative stroke and wound hematoma.
        J Vasc Surg. 1990; 12: 440-445
        • Fearn S.J.
        • Parry A.D.
        • Picton A.J.
        • Mortimer A.J.
        • McCollum C.N.
        Should heparin be reversed after carotid endarterectomy? A randomised prospective trial.
        Eur J Vasc Endovasc Surg. 1997; 13: 394-397
        • Mauney M.C.
        • Buchanan S.A.
        • Lawrence W.A.
        • Bishop A.
        • Sinclair K.
        • Daniel T.M.
        • et al.
        Stroke rate is markedly reduced after carotid endarterectomy by avoidance of protamine.
        J Vasc Surg. 1995; 22: 264-269
        • Patel R.B.
        • Beaulieu P.
        • Homa K.
        • Goodney P.P.
        • Stanley A.C.
        • Cronenwett J.L.
        • et al.
        Shared quality data are associated with increased protamine use and reduced bleeding complications after carotid endarterectomy in the Vascular Study Group of New England.
        J Vasc Surg. 2013; 58: 1518-1524
        • Cronenwett J.L.
        • Kraiss L.W.
        • Cambria R.P.
        The society for vascular surgery vascular quality initiative.
        J Vasc Surg. 2012; 55: 1529-1537
        • Austin P.C.
        An introduction to propensity score methods for reducing the effects of confounding in observational studies.
        Multivariate Behav Res. 2011; 46: 399-424
        • Naylor A.R.
        • Ricco J.-B.
        • de Borst G.J.
        • Debus S.
        • de Haro J.
        • Halliday A.
        • et al.
        Editor's choice - Management of atherosclerotic carotid and vertebral artery disease: 2017 clinical practice guidelines of the European Society for Vascular Surgery (ESVS).
        Eur J Vasc Endovasc Surg. 2018; 55: 3-81
        • Newhall K.A.
        • Saunders E.C.
        • Larson R.J.
        • Stone D.H.
        • Goodney P.P.
        Use of Protamine for anticoagulation during carotid endarterectomy: a meta-analysis.
        JAMA Surg. 2016; 151: 247-255
        • Kakisis J.D.
        • Antonopoulos C.N.
        • Moulakakis K.G.
        • Schneider F.
        • Geroulakos G.
        • Ricco J.B.
        Protamine reduces bleeding complications without increasing the risk of stroke after carotid endarterectomy: a meta-analysis.
        Eur J Vasc Endovasc Surg. 2016; 52: 296-307
        • Columbo J.A.
        • Kang R.
        • Hoel A.W.
        • Kang J.
        • Leinweber K.A.
        • Tauber K.S.
        • et al.
        A comparison of reintervention rates after endovascular aneurysm repair between the Vascular Quality Initiative registry, Medicare claims, and chart review.
        J Vasc Surg. 2019; 69: 74-79
        • Hoel A.W.
        • Faerber A.E.
        • Moore K.O.
        • Ramkumar N.
        • Brooke B.S.
        • Scali S.T.
        • et al.
        A pilot study for long-term outcome assessment after aortic aneurysm repair using Vascular Quality Initiative data matched to Medicare claims.
        J Vasc Surg. 2017; 66: 751-759

      Linked Article

      • Protamine Reduces Dangerous Reoperations After Asymptomatic Carotid Surgery
        European Journal of Vascular and Endovascular SurgeryVol. 60Issue 6
        • Preview
          Surgery for asymptomatic carotid disease can only harm the patient in the short term. In addition to statin and antiplatelet use, it is accepted routine practice to administer unfractionated heparin during carotid surgery. The safest dose and activity of heparin has not been clearly established, and it remains unclear whether its activity should be reversed at the end of the operation. Protamine sulphate may be used to reverse the action of heparin, and it does so effectively.
        • Full-Text
        • PDF
        Open Archive
      • Protamine and Myocardial Infarction Risk after Carotid Endarterectomy: A New Finding
        European Journal of Vascular and Endovascular SurgeryVol. 61Issue 4
        • Preview
          We read with great interest the work by David H. Stone and colleagues1 on the benefits of protamine reversal of heparin anticoagulation after carotid endarterectomy. This paper reports a large, registry based analysis of > 70 000 asymptomatic patients from over 300 centres in the United States and Canada. Interestingly, the use of protamine increased in the temporal trend, for a total 69% of subjects receiving it after surgery. This correlated with a 50% reduction in the re-operation for bleeding rate (1.4% vs.
        • Full-Text
        • PDF
        Open Archive

      Comments

      Commenting Guidelines

      To submit a comment for a journal article, please use the space above and note the following:

      • We will review submitted comments as soon as possible, striving for within two business days.
      • This forum is intended for constructive dialogue. Comments that are commercial or promotional in nature, pertain to specific medical cases, are not relevant to the article for which they have been submitted, or are otherwise inappropriate will not be posted.
      • We require that commenters identify themselves with names and affiliations.
      • Comments must be in compliance with our Terms & Conditions.
      • Comments are not peer-reviewed.