Author + information
- Received November 23, 2009
- Revision received April 29, 2010
- Accepted May 15, 2010
- Published online August 1, 2010.
- Santi Trimarchi, MD⁎,
- Dean E. Smith, PhD†,
- David Share, MD, MPH‡,
- Sandeep M. Jani, MPH§,
- Michael O'Donnell, MD∥,
- Richard McNamara, MD¶,
- Arthur Riba, MD#,
- Eva Kline-Rogers, RN†,
- Hitinder S. Gurm, MD†,
- Mauro Moscucci, MD⁎⁎,⁎ (, )
- BMC2 Registry
- ↵⁎Reprint requests and correspondence:
Dr. Mauro Moscucci, University of Miami, Miller School of Medicine, Clinical Research Building, 1120 NW 14th Street, Suite 1124, Miami, Florida 33136
Objectives This study sought to evaluate the prevalence, risk factors, outcomes, and predictors of mortality of retroperitoneal hematoma (RPH) following percutaneous coronary intervention.
Background Retroperitoneal hematoma is a serious complication of invasive cardiovascular procedures.
Methods The study sample included 112,340 consecutive patients undergoing percutaneous coronary intervention in a large, multicenter registry between October 2002 and December 2007. End points evaluated included the development of RPH and mortality.
Results Retroperitoneal hematoma occurred in 482 (0.4%) patients. Of these, 92.3% were treated medically and 7.7% underwent surgical repair. Female sex, body surface area <1.8 m2, emergency procedure, history of chronic obstructive pulmonary disease, cardiogenic shock, pre-procedural IV heparin, pre-procedural glycoprotein IIb/IIIa inhibitors, adoption of sheath size ≥8-F, and use of vascular closure devices were independent predictors of RPH, whereas the use of bivalirudin was associated with a lower risk. The development of RPH was associated with a higher frequency of post-procedure myocardial infarction (5.81% vs. 1.67%, p < 0.0001), infection and/or sepsis (17.43% vs. 3.00%, p < 0.0001), and heart failure (8.00% vs. 1.63%, p < 0.0001). In-hospital mortality was significantly higher in patients who developed RPH than in patients who did not (6.64% vs. 1.07%, p < 0.0001). Among patients with RPH, independent predictors of death were history of myocardial infarction, cardiogenic shock, pre-procedural creatinine ≥1.5 mg/dl, and left ventricular ejection fraction <50%.
Conclusions Retroperitoneal hematoma is an uncommon complication of contemporary percutaneous coronary intervention associated with high morbidity and mortality. The identification of risk factors for the development of RPH could lead to modification of procedure strategies aimed toward reducing its incidence.
The development of complications related to femoral artery puncture continues to be a frequent occurrence following contemporary percutaneous coronary intervention (PCI) (1–4). Whereas the detection and management of complications confined to below the inguinal ligament, such as hematoma, false aneurysm, lower extremity ischemia, and arteriovenous fistula, are usually rapid and with an uneventful course, a retroperitoneal hematoma (RPH) can harbor a large volume of blood with minimal external signs, and it is associated with a more serious clinical prognosis (1,3,4). In patients undergoing PCI, the necessity to take concomitant aggressive antiplatelet and/or anticoagulation therapy can further worsen post-procedural hemorrhage. The impact of glycoprotein (GP) IIb/IIIa inhibitors and vascular closure devices (VCDs) on the incidence of RPH has not been fully studied (5,6). In addition, predictors of mortality in patients developing RPH are currently unknown. The aim of this analysis was to identify the prevalence, risk factors, management, outcomes, and predictors of mortality of RPH following PCI.
The study sample included 112,340 consecutive patients undergoing contemporary PCI in a large, multicenter, PCI registry between October 2002 and December 2007 in Michigan (7,8). The registry was approved by the institutional review board of the University of Michigan and by local institutional review boards. The primary end point was the development of RPH. Data were prospectively collected using standardized definitions and a standardized data collection form. Each data form submitted to the coordinating center was assessed for completeness and for face validity. All participating centers were audited twice yearly. A random 2% sampling of all cases was selected for audit. In addition, all records of patients who died during the same hospitalization and of patients undergoing coronary artery bypass surgery were audited. The primary end point was the development of a RPH, defined as a large hematoma dissecting into the retroperitoneum.
All patients received aspirin unless an allergy was present. Ticlopidine, clopidogrel, GP IIb/IIIa receptor inhibitors, and heparin were used at the discretion of the operator.
Continuous variables are expressed as mean ± SD and discrete variables as frequency counts and percentages. Continuous variables were evaluated using the t test. Differences between discrete variables were analyzed using the chi-square test and, if needed, Fisher exact test. A p value <0.05 was considered significant. Logistic regression models were fitted to determine independent predictors of RPH and to identify independent predictors of death in the group of patients who developed RPH. Candidate variables were selected on the basis of clinical face value. Stepwise ascending regression was used to identify predictors with a pre-set cutoff p = 0.10 for retention of candidate variables. Statistical analysis was performed using SAS version 8.2 (SAS Institute, Inc., Cary, North Carolina).
Baseline demographics, clinical characteristics, and pre-procedural treatments
Among 112,340 patients enrolled in the registry, RPH occurred in 482 (0.4%). Patients who developed RPH were of similar ages to those who did not. Retroperitoneal hematoma occurred more frequently in women (68.1% of patients with RPH vs. 34.1% of patients without RPH, p < 0.0001), in patients with a body surface area (BSA) <1.8 m2, and in patients with chronic obstructive pulmonary disease (Table 1). Patients with RPH were more likely to have undergone an emergency procedure and to have received pre-procedural fibrinolytic therapy (p = 0.01), GP IIb/IIIa inhibitors, and IV heparin (both p < 0.0001) (Table 2). There were no significant differences in pre-procedural use of aspirin or warfarin. We found that VCDs, and in particular Angio-Seal (St. Jude Medical, St. Paul, Minnesota), were more frequently used in patients who developed RPH than in those who did not (p = 0.0006 for VCDs and p = 0.0004 for Angio-Seal) (Table 2). In addition, RPH was more frequently detected following procedures performed using a sheath size ≥8-F (Table 2).
Of the patients who developed RPH, 92.3% were treated medically and 7.7% underwent surgical repair. The development of RPH was associated with a higher frequency of post-procedure myocardial infarction (5.81% vs. 1.67%, p < 0.0001), infection and/or sepsis (17.43% vs. 3.00%, p < 0.0001), and congestive heart failure (8.00% vs. 1.63%, p < 0.0001) (Table 3). In-hospital mortality was significantly higher in patients who developed RPH than in those who did not (6.64% vs. 1.07%, p < 0.0001). In addition, a nonsignificant trend toward higher in-hospital mortality was observed in patients with RPH treated surgically when compared with those treated medically (8.11% vs. 6.52%, p = 0.73). Length of in-hospital stay was significantly higher for patients with RPH than for patients without (5 days vs. 2 days, p < 0.0001) (Table 3).
Multivariable predictors of RPH
Logistic regression analysis identified female sex, BSA <1.8 m2, history of chronic obstructive pulmonary disease, emergency procedure, cardiogenic shock, use of pre-procedural IV heparin, use of pre-procedural GP IIb/IIIa inhibitors, use of a VCD versus manual compression, and sheath size ≥8-F as independent predictors of RPH (Table 4). The use of bivalirudin was associated with a lower risk for the development of RPH (Table 4). There was good model calibration for the outcome of RPH (Fig. 1) with Hosmer-Lemeshow p = 0.63.
Multivariable predictors of in-hospital death
Among those patients (n = 482) who experienced RPH, independent predictors of death in a sex-controlled model were history of myocardial infarction, cardiogenic shock, pre-procedural creatinine level ≥1.5 mg/dl, and left ventricular ejection fraction <50% (Table 5). The model was well calibrated for the outcome of death with Hosmer-Lemeshow p = 0.17.
The prevalence of RPH after contemporary PCI was 0.4% in this analysis and is the largest experience of RPH reported in the literature. This finding agrees with previous literature (1–4) suggesting that RPH is an uncommon complication in patients undergoing PCI. Although RPH can also develop as a subclinical hemorrhage, without specific signs such as hypotension, abdominal distension or fullness, and pain (2,3,9), our findings confirm the association between the development of RPH and a significantly higher risk of morbidity and mortality.
Several risk factors emerged as independent predictors of RPH in our analysis. The adoption of VCDs (vs. manual compression) was associated with an increased risk for the development of RPH. This association was significant for Angio-Seal (odds ratio [OR]: 1.68, p < 0.0001), whereas there was a nonsignificant trend with Perclose (Abbott Laboratories, Abbott Park, Illinois) (OR: 1.29, p = 0.13). In contemporary practice, arterial access sites are commonly managed with percutaneous VCDs, although variable results as to their benefits and risks have been reported (4–6,10,11). In particular, VCDs, when compared with manual compression, are associated with a higher incidence of large pseudoaneurysms not amenable to ultrasound compression, greater blood losses requiring transfusion, a higher incidence of arterial stenosis and/or occlusion, and the need for more extensive surgical repair (12,13). A recent meta-analysis on vascular complications associated with the use of VCDs after PCI has suggested that in the setting of PCI, the rate of complication appeared higher with VasoSeal (Datascope Corporation, Montvale, New Jersey) (6). In our analysis, VasoSeal or other VCDs were used only in 0.3% and 3.2% of cases, respectively, and were excluded from the RPH prediction model for their statistical inconsistency.
The analysis of different sheath size showed that the use of catheters ≥8-F was a predictor of RPH. This data suggests per se that the use of small catheters is preferable, regardless the presence of other risk factors for RPH.
The development of RPH has also been associated with the use of GP IIb/IIIa inhibitors, particularly in conjunction with other antithrombotics such as heparin (14–16). Our analysis confirms previous reports demonstrating that the use of both pre-procedural GP IIb/IIIa inhibitors and IV heparin were independently predictive of the development of RPH. In addition, similarly to previous experiences (1,5), female sex was an independent risk factor for RPH. The reasons for this association remain elusive, although several hypotheses have been suggested, including differences due to estrogen-related arterial structures (9,17) and smaller common femoral artery dimensions in women (18). Furthermore, women have different arterial mechanical properties than men do, such as increased likelihood of arterial mobility (10,19), that may increase the need for multiple arterial punctures and which, in turn, contribute to increased bleeding risk. We also observed that BSA <1.8 m2 was a risk factor for RPH. Smaller arteries could complicate a PCI and lead to RPH, and not surprisingly, women and other patients with a lower BSA are at higher risk.
In addition, our analysis showed that emergency procedures, cardiogenic shock, and the use of antithrombotic therapy were all independently associated with an increased risk for the development of RPH. It remains to be determined whether this was due to differences in vascular access techniques in emergency situations and/or to a more aggressive anticoagulation protocol in these patient subgroups.
The development of a RPH can lead to severe morbidity and increased risk of death. We observed in-hospital mortality of 6.64% in patients who developed a RPH, which was significantly higher than in patients who did not develop RPH (1.07%, p < 0.0001). After PCI, the presence of RPH was associated not only with a higher frequency of post-procedure cardiac complications, including myocardial infarction and congestive heart failure, but also with a higher frequency of infection and/or sepsis, gastrointestinal bleeding, and contrast nephropathy.
Of the patients who developed RPH, 92.3% were treated medically and 7.7% underwent surgical repair. A trend toward a higher in-hospital mortality was observed in patients with RPH treated surgically than in those treated medically, possibly reflective of the fact that a surgical approach might be performed in more unstable patients in whom fluid resuscitation and blood transfusions are inadequate in re-establishing a stable hemodynamic status. Importantly, our analysis of factors associated with mortality revealed the presence of myocardial infarction, shock, and renal insufficiency as independent correlates of mortality in this patient group.
In view of its observational, nonrandomized design, the trend identified toward higher in-hospital mortality in patients treated surgically does not necessarily support a management strategy actively avoiding surgery for the treatment of RPH. In addition, we did not have specific data on the femoral artery puncture site in patients who developed RPH and who underwent surgical management. Thus, we were unable to further evaluate the relationship between the development of RPH and the arterial puncture site. It has been suggested that femoral arterial access will sometimes be gained above the inguinal ligament, where manual compression of the artery to stop bleeding is less effective. However, although this technique has been advocated as a potential cause of RPH (10,14), adoption of common femoral artery puncture does not necessarily avoid the occurrence of RPH (10), given that previous studies have shown that the site of arterial bleeding is often just distal to the inguinal ligament (1–3). In addition, we were unable to evaluate other potential predictors of bleeding such as peak activated clotting time, or heparin dose, and we cannot exclude that other unknown confounders could have modified the observed association between risk factors and the development of RPH.
Retroperitoneal hematoma is an uncommon complication of contemporary PCI, and it is associated with high morbidity and mortality. Although it remains to be determined, the identification of risk factors for the development of RPH could lead to modification of procedure strategies aimed toward reducing its incidence, such as, for example, the adoption of a radial artery approach in high-risk patients.
The authors thank Sophie Rushton-Smith, PhD, for editorial assistance in the preparation of this manuscript. Her work was funded by the University of Michigan Medical Center.
Supported by a grant from Blue Cross Blue Shield of Michigan. Dr. Moscucci has received grant and salary support from Blue Cross Blue Shield of Michigan. Dr. Gurm receives grant and salary support from Blue Cross Blue Shield of Michigan. All other authors report that they have no relationships to disclose.
- Abbreviations and Acronyms
- body surface area
- percutaneous coronary intervention
- retroperitoneal hematoma
- vascular closure devices
- Received November 23, 2009.
- Revision received April 29, 2010.
- Accepted May 15, 2010.
- American College of Cardiology Foundation
- Farouque H.M.,
- Tremmel J.A.,
- Raissi Shabari F.,
- et al.
- Nikolsky E.,
- Mehran R.,
- Halkin A.,
- et al.
- Moscucci M.,
- Share D.,
- Kline-Rogers E.,
- et al.,
- Blue Cross Blue Shield of Michigan Cardiovascular Consortium (BMC2)
- Kline-Rogers E.,
- Share D.,
- Bondie D.,
- et al.,
- Blue Cross Blue Shield of Michigan Cardiovascular Consortium (BMC2)
- Blankenship J.C.,
- Hellkamp A.S.,
- Aguirre F.V.,
- Demko S.L.,
- Topol E.J.,
- Califf R.M.
- Mandak J.S.,
- Blankenship J.C.,
- Gardner L.H.,
- et al.
- Lincoff A.M.,
- Bittl J.A.,
- Harrington R.A.,
- et al.,
- REPLACE-2 Investigators
- Celermajer D.S.,
- Sorensen K.E.,
- Spiegelhalter D.J.,
- Georgakopoulos D.,
- Robinson J.,
- Deanfield J.E.
- Smulyan H.,
- Asmar R.G.,
- Rudnicki A.,
- London G.M.,
- Safar M.E.