Author + information
- Received June 14, 2004
- Revision received October 14, 2004
- Accepted October 18, 2004
- Published online February 1, 2005.
- H.M. Omar Farouque, MBBS, PhD, FRACP,
- Jennifer A. Tremmel, MD, SM,
- Farshad Raissi Shabari, MD, MPH,
- Meenakshi Aggarwal, MD,
- William F. Fearon, MD, FACC,
- Martin K.C. Ng, MBBS, FRACP,
- Mehrdad Rezaee, MD, PhD, FACC,
- Alan C. Yeung, MD, FACC and
- David P. Lee, MD, FACC* ()
- ↵*Reprint requests and correspondence:
Dr. David P. Lee, Stanford University Medical Center, Interventional Cardiology, Room H2103, 300 Pasteur Drive, Stanford, California 94304
Objectives We sought to determine the incidence, clinical features, and risk factors for retroperitoneal hematoma (RPH) after percutaneous coronary intervention (PCI).
Background Little is known about the clinical features, outcomes, and determinants of this serious complication in the contemporary era of PCI.
Methods A retrospective analysis yielded 26 cases of RPH out of 3,508 consecutive patients undergoing PCI between January 2000 and January 2004. Cases were compared with a randomly selected sample of 50 control subjects without RPH.
Results The incidence of RPH was 0.74%. Features of RPH included abdominal pain (42%), groin pain (46%), back pain (23%), diaphoresis (58%), bradycardia (31%), and hypotension (92%). The mean systolic blood pressure nadir was 75 mm Hg. The hematocrit dropped by 11.5 ± 5.1 points from baseline in RPH patients, as compared with 2.3 ± 3.3 points in controls (p < 0.0001). The mean hospital stay was longer in RPH patients (2.9 ± 3.8 days vs. 1.7 ± 1.5 days, p = 0.06). The following variables were found to be independent predictors of RPH: female gender (odds ratio [OR] 5.4, p = 0.005), low body surface area (BSA <1.73 m2; OR 7.1, p = 0.008), and higher femoral artery puncture (OR 5.3, p = 0.013). There was no association between RPH and arterial sheath size, use of glycoprotein IIb/IIIa inhibitors, or deployment of a vascular closure device.
Conclusions Female gender, low BSA, and higher femoral artery puncture are significant risk factors for RPH. Awareness of the determinants and clinical features of RPH may aid in prevention, early recognition, and prompt treatment.
Technical advances in percutaneous coronary intervention (PCI) have led to an overall improvement in short- and long-term outcomes. However, iatrogenic complications such as bleeding may still occur. Recent randomized studies indicate that vascular access site injury resulting in major bleeding remains a significant complication, with an incidence of ∼1% to 2% (1,2), although in unselected patients, the rate is considerably higher (3,4).
Retroperitoneal hematoma (RPH) is one of the most serious complications after PCI. Unlike other bleeding sites, the retroperitoneum can harbor a large volume of blood with few external manifestations until hypovolemia occurs, leading to delayed recognition, added morbidity, and a potentially fatal result. Although bleeding as a complication of PCI has been examined previously (4–8), there are few studies on RPH after PCI (9,10). These small studies report data from the late 1980s and early 1990s from a predominantly surgical perspective (9,10). Hence, the relevance of these studies to the current era remains uncertain. We therefore sought to analyze the occurrence and clinical and procedural determinants of RPH in a contemporary sample of patients from a single center using a case-control study design.
A retrospective analysis of the PCI database at Stanford University Medical Center between January 2000 and January 2004 was undertaken. Over a four-year interval, 3,508 PCIs were performed. The database was reviewed to identify definite cases of RPH associated with femoral approach PCI (n = 26). Twenty-five patients had computed tomographic scan documentation, and one patient had autopsy confirmation of RPH. To determine predictive factors of RPH, a case-control study was performed. Fifty control subjects were randomly selected from the overall group of patients who had PCI via the femoral approach uncomplicated by RPH. Matching of subjects was not performed so as not to preclude the estimation of the effect of matched variables on the outcome.
Procedural antithrombotic therapy consisted of aspirin and unfractionated heparin. Glycoprotein (GP) IIb/IIIa receptor antagonists and coronary stents were used at the discretion of the operator. A thienopyridine was administered to patients receiving a stent. Femoral artery angiograms were obtained by hand injection of radiographic contrast through the side arm of the femoral sheath in a shallow (±20°) right or left anterior oblique projection without cranial or caudal angulation. A collagen (AngioSeal; St. Jude Medical, St. Paul, Minnesota) or suture-mediated (Perclose; Abbott Vascular, Redwood Shores, California) vascular closure device was used to close the femoral artery puncture based on operator preference. Sheaths left in situ after the procedure were removed when the activated clotting time was <180 s, and compression was applied to achieve hemostasis.
Detailed clinical and procedural data were obtained for all patients by reviewing database records, charts, and catheterization laboratory logs. Hypotension was defined as systolic blood pressure <90 mm Hg. Blood pressure was deemed to fluctuate if the observed hemodynamic response during the bleeding event cycled between hypotensive and normotensive values (systolic blood pressure >100 mm Hg) on at least two occasions.
An individual blinded to outcomes analyzed the femoral artery angiograms to determine the position of arterial puncture. Puncture site position was arbitrarily divided into two categories using the femoral head as a landmark: the proximal third of the femoral head or above (high) and the middle third of the femoral head or below (low). Digital quantitative angiography of the common femoral artery was performed with an automated edge-detection program (GE QCA version 3.1; GE Medical Systems, Milwaukee, Wisconsin), using the contrast-filled femoral sheath for calibration. Study approval was obtained from the Institutional Review Board at Stanford University Medical Center.
Data are expressed as the mean value ± SD, unless otherwise indicated. For categorical variables, intergroups comparisons were performed using the chi-square test. The Fisher exact test was used if the expected cell count for a 2 × 2 table was <5. The two-tailed unpaired Student ttest was used to analyze continuous variables, and the Mann-Whitney Utest was used if the assumption of normality was not met. A p value <0.05 was considered statistically significant. Univariate predictors with p < 0.05 were entered into a logistic regression model to determine independent factors associated with the development of RPH. The odds ratio (OR) and 95% confidence interval were calculated. Statistical analyses were performed using SPSS version 11.0 software package (SPSS Inc., Chicago, Illinois).
Baseline characteristics of the study sample are indicated in Table 1.Retroperitoneal hematoma occurred in 26 (0.74%) of 3,508 consecutive patients undergoing PCI.
Clinical features and in-hospital outcome
Clinical manifestations and the time course of RPH are indicated in Table 2and Figure 1.Anemia was a universal feature, with a mean fall in hematocrit of 11.5 ± 5.1 points from baseline in RPH patients, as compared with 2.3 ± 3.3 points in control subjects (p < 0.0001). Blood transfusions were given to 92% of RPH patients to correct the anemia (mean 2.7 ± 3.3 U [range 1 to 18 U]). Three patients (12%) required vascular surgical intervention due to persistent hypotension. Two of these patients had punctures in the distal external iliac artery, and one patient had an anterior wall common femoral artery puncture that was inadequately sealed by a suture-mediated closure device despite the appearance of external hemostasis. These patients had successful vascular repair and recovered uneventfully. One patient (4%) with RPH died from complications of retroperitoneal blood loss. All RPH episodes were ipsilateral to the femoral artery puncture, as determined by a computed tomographic scan. The duration of hospital stay after PCI was longer in RPH patients than in controls (2.9 ± 3.8 days vs. 1.7 ± 1.5 days, p = 0.06). There was one death (2%) in the control group due to an arrhythmic complication of acute myocardial infarction.
Predictors of RPH
Univariate and multivariate predictors of RPH are shown in Tables 3 and 4.⇓⇓Patients receiving a higher weight-adjusted heparin dose and those with smaller caliber femoral arteries were more likely to develop RPH, but these associations were of borderline statistical significance (Table 3). The use of a GP IIb/IIIa inhibitor or vascular closure device was not predictive of RPH. Three dichotomous variables were identified as independent predictors of RPH (Table 4). These factors included female gender (OR 5.4), BSA <1.73 m2(OR 7.1), and a higher femoral artery puncture site (OR 5.3). The relationship between BSA and femoral artery size was studied using simple linear regression, and this showed a modest correlation between the two variables (r = 0.40, p = 0.002) (Fig. 2).
To our knowledge, this study is the first to examine characteristics and predictors of RPH occurring after PCI in the current era. Our findings indicate that RPH is an uncommon albeit potentially serious complication of PCI. Moreover, female gender remains an independent predictor of the occurrence of RPH, in addition to other patient- and procedure-related factors.
The reported incidence of RPH after cardiac catheterization varies. Sreeram et al. (10) documented an incidence of 0.15% in patients undergoing diagnostic and interventional cardiac catheterization between 1990 and 1991. However, this figure is likely to be an underestimate, as RPH cases were selected based on referral for vascular surgical intervention. Others have reported an RPH incidence of 0.5% after PCI during a similar era (5). In one of the larger series, Kent et al. (9) observed RPH in 0.47% of all patients undergoing cardiac catheterization between 1988 and 1993. In the subgroup that had PCI, the incidence of RPH was 1% and was more frequent in patients receiving first-generation coronary stents than in those who had balloon angioplasty alone (3% vs. 0.8%). Our findings are consistent with a subset analysis of the Do Tirofiban And ReoPro Give Similar Efficacy Outcomes Trial (TARGET), in which RPH was seen in 0.7% of patients (7). The variation in frequency of RPH after PCI between these studies may reflect differences in periprocedural antithrombotic regimens and evolution of vascular access site management over the last two decades.
The early diagnosis of RPH is elusive due to its concealed nature. Typically, the diagnosis is made only after significant blood loss has occurred, as in our study. Asymptomatic RPH was rare, with 96% of cases having at least one symptom. However, some symptoms of RPH in the setting of PCI are nonspecific, such as back or groin discomfort. Back pain is often regarded as a common feature of RPH, but in this series, it was an infrequent symptom. In contrast, lower abdominal pain and diaphoresis were more commonly observed. These are unexpected occurrences after uncomplicated PCI and should lead the physician to consider concealed bleeding as a potential cause.
The vast majority of RPH patients had evidence of hypovolemic shock, with pronounced hypotension. In contrast, previous studies of RPH in both the non-PCI and PCI settings suggest a lower frequency of hypotension occurring in 43% to 64% (10,11). A frequent observation in our study among hypotensive patients was a fluctuating blood pressure related to the balance between the rate of ongoing bleeding and intravenous fluid resuscitation. Moreover, several patients exhibited clinically significant bradycardia in association with hypotension. The low cardiac output and impaired coronary flow that may result from these hemodynamic perturbations can lead to myocardial ischemia and expose the patient to the potential risk of stent thrombosis. Global hypoperfusion may result in deleterious effects in patients with preexisting disease in other arterial beds (4).
Lower abdominal tenderness ipsilateral to the puncture site was another common finding and underscores the importance of a careful physical examination of the abdomen in evaluating suspected cases of RPH. Previous studies have documented clinical evidence of femoral neuropathy in 23% to 54% of RPH patients, caused by compression of the femoral nerve, which was treated with surgery in some cases (9–11). However, overt manifestations of femoral neuropathy were absent in our study. This discrepancy may be due to differences in the size and location of the hematoma in relation to the femoral nerve or variability in reporting subtle lower limb neurologic changes. It is instructive to examine the timing of the initial manifestation of RPH in relation to PCI. Almost 75% of cases presented within the first 3 h after conclusion of the procedure, with a rapid decline in frequency after this time period, suggesting that increased vigilance during the first few postprocedural hours may be important.
It has been suggested that some cases of RPH after cardiac catheterization may be unrelated to femoral artery puncture and are more likely due to altered hemostasis. Using computed tomographic imaging, Quint et al. (12) found that 25% of retroperitoneal bleeds were remote from the site of femoral artery puncture, with the majority of these being on the contralateral side to the puncture site. However, we found that all instances of RPH were ipsilateral to the femoral puncture site and contiguous with the presumed site of vessel puncture in the inguinal region. Our observations imply that femoral artery puncture was an integral element to the formation of RPH in all cases.
The most serious complication secondary to PCI-related RPH is death. In our study the mortality rate as a consequence of RPH was 4%, which is in accordance with previous reports (9,10). A frequent requirement was the transfusion of blood to correct anemia. Although it is regarded as a safe and often life-saving therapy, blood transfusion is not without risks. Moreover, a recent study of 10,974 patients indicated that major bleeding and blood transfusion after PCI were both independent predictors of in-hospital and one-year mortality (4). Another contributor to morbidity in our study was the need for vascular surgical intervention, which was required in 12% of RPH cases. Our data are consistent with the findings of Kent et al. (9), who found that 16% of cases in their series required an operation. These observations indicate that the majority of patients can still be managed conservatively without resorting to surgical intervention. Due to these factors, there was a strong trend toward increased length of hospital stay in patients with RPH. In addition, the occurrence of RPH has significant economic implications. Bleeding after PCI is known to be an important contributor to hospital costs. A recent analysis of a large U.S. hospital claims database between 1995 and 1997 indicated that bleeding complications resulted in the addition of over $10,000 to the baseline cost of PCI (3).
The findings of this study underscore the importance of female gender, even after correction for BSA, as an independent predisposing factor for the occurrence of femoral vascular access site complications (8,13). Only one study has previously addressed risk factors for the occurrence of RPH. Kent et al. (9) found that women were more likely to have RPH after coronary stenting, but only 13 RPH cases were analyzed, and details of the multivariate analysis were not reported. Our results corroborate and extend this finding to the current era, in addition to providing an estimate of risk. The reason for the intriguing female predisposition to RPH after PCI is unclear. Several possibilities exist, including gender-related differences in arterial structure and function related to the vascular effects of estrogen (14,15). Although menopausal status was not ascertained in this study, it is likely that the majority of women were postmenopausal given their age range. It is conceivable that such alterations may predispose to arterial fragility and vascular complications such as RPH. Women are also known to have smaller diameter and shorter length common femoral arteries than men (16), which may make access more challenging. This anatomic difference may lead to a greater likelihood of multiple arterial and posterior wall punctures. Gender-specific differences and menopause-related changes in coagulation and fibrinolysis have also been documented (17), but these alterations are complex and may not predispose to an increase in bleeding risk. There are also gender differences in arterial mechanical properties, such as increased pulsatility, possibly related to body size (18), which could increase bleeding risk.
A low BSA was another independent predictor of RPH. This may be partly due to its association with a smaller femoral artery diameter, resulting in arterial access issues, as alluded to earlier. The use of a micropuncture technique to access the femoral artery may reduce the impact of multiple arterial punctures, although this hypothesis is untested. Access via the radial artery could also be considered in order to minimize access site bleeding.
A higher femoral arterial puncture site, defined as above the middle third of the femoral head on fluoroscopy, was an important procedure-related risk factor for RPH. Puncture and sheath insertion above the inguinal ligament, and hence into the external iliac artery, may predispose to uncontrolled bleeding due to the difficulty in achieving compression of this vessel. Meticulous attention to puncture technique with fluoroscopic guidance may minimize the occurrence of RPH. However, it is important to realize that a lower arterial puncture does not eliminate the risk of RPH. In our study, 45% of patients with RPH had sheath insertion in the common femoral artery well below the presumed site of the inguinal ligament. The occurrence of RPH in these cases was most likely related to the spread of bleeding from the puncture site along well-defined anatomic fascial planes (19). Specifically, bleeding from the femoral vascular structures contained within the femoral sheath can extend into the retroperitoneum by virtue of its superiorly directed communication with this space. Because the vascular structures at this location are accessible to effective manual compression, we advocate applying this simple measure at an early stage to all cases of presumed RPH until stabilization of the patient is achieved and the diagnosis confirmed.
Glycoprotein IIb/IIIa inhibitors are potent blockers of the final common pathway of platelet aggregation and have been shown to increase the risk of vascular access site complications (6,8). Our findings indicate that GP IIb/IIIa inhibitor use did not predispose to the occurrence of RPH; however, this conclusion needs to be tempered by the small number of RPH cases. Before using these agents in any individual patient, an assessment of bleeding risk in relation to the potential benefits of therapy should be obtained. Early studies had also suggested that higher doses of heparin in conjunction with GP IIb/IIIa use increased vascular access site bleeding (8). In the present study, there was a trend toward an increase in RPH in patients receiving larger weight-adjusted heparin doses. Several measures have been suggested to minimize the risk of femoral access site complications when using GP IIb/IIIa inhibitors, including early sheath removal, avoiding the use of venous sheaths, ensuring reduced weight-adjusted heparin dosing, and optimal post-PCI nursing care (6,8,20).
Alternative antithrombotic regimens should be considered in patients deemed to be at higher risk of RPH. In the Randomized Evaluation in PCI Linking Angiomax to reduced Clinical Events (REPLACE-2) randomized trial, bivalirudin with provisional GP IIb/IIIa blockade was associated with fewer major bleeding events, as compared with unfractionated heparin and planned GP IIb/IIIa blockade in patients with low- to moderate-risk characteristics (21). Specifically, there was a 60% reduction in RPH (p = 0.06), without a significant change in ischemic end points. The use of preprocedural high-dose clopidogrel loading without GP IIb/IIIa antagonists may be another option (2).
Vascular closure device deployment is relatively common in the current era of PCI. Potential advantages include improved patient comfort by avoiding manual compression and reduced time to ambulation. However, there is some evidence to suggest that vascular closure device use may be associated with an increased risk of access site bleeding (22). In contrast, our findings and those of others suggest that bleeding may not be increased in patients receiving closure devices (7). The discrepant findings between these nonrandomized studies of vascular closure device use could be attributed to differences in the characteristics of the respective study populations and their predisposition to the development of vascular complications. To adequately assess the impact of closure devices on vascular complications, a randomized trial assessing outcomes would be required. Nevertheless, in the PCI setting, these devices are typically deployed in highly anticoagulated patients, and their use may engender a perception of lower bleeding risk among staff involved in early post-PCI management. Patients receiving closure devices should therefore be closely observed to ensure early detection of bleeding complications such as RPH.
This study is a retrospective case-control review, and although it represents the largest published series to date reporting the manifestations and risk factors of PCI-related RPH, the number of cases (n = 26) limits the analysis. The case-control design is well suited to studying an uncommon outcome such as RPH. A random sample of 50 patients from the overall group of 3,482 patients who did not have RPH was selected as the control group for the case-control analysis. This method permits an unbiased estimation of parameters and, for practical considerations, enabled a complete and detailed examination of clinical, procedural, and outcome variables.
Retroperitoneal hematoma is a serious vascular access site complication of PCI. Several factors may predispose to RPH in the current era of PCI, including female gender, low BSA, and higher femoral artery puncture. Awareness of the clinical features and determinants of RPH may aid in prevention, earlier recognition, and prompt therapy of this complication.
Dr. Farouque was supported by a Neil Hamilton Fairley Fellowship of the National Health and Medical Research Council of Australia.
- Abbreviations and acronyms
- body surface area
- odds ratio
- percutaneous coronary intervention
- retroperitoneal hematoma
- Received June 14, 2004.
- Revision received October 14, 2004.
- Accepted October 18, 2004.
- American College of Cardiology Foundation
- Popma J.J.,
- Satler L.F.,
- Pichard A.D.,
- et al.
- 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.
- Kent K.C.,
- Moscucci M.,
- Mansour K.A.,
- et al.
- Ling S.,
- Dai A.,
- Dilley R.J.,
- et al.
- 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.
- Juran N.B.,
- Rouse C.L.,
- Smith D.D.,
- O'Brien M.A.,
- DeLuca S.A.,
- Sigmon K.,
- the Standards of Angioplasty Nursing techniques to Diminish Bleeding Around the Groin (SANDBAG) Nursing Coordinators