Endovascular Repair of Ruptured Abdominal Aortic Aneurysms in a Rural Center Is Both Feasible and Associated with Reduced Blood Product Requirements

Method

Patients presenting to the Royal Hobart Hospital (RHH) with a diagnosis of rAAA between February 2004 and November 2008 were identified from the hospital discharge coding. Patients were also identified from a prospective database of a single consultant vascular surgeon. The RHH provides a statewide vascular surgery service for the state of Tasmania, Australia, with a population of approximately 500,000. The notes of these patients were reviewed retrospectively. Patients for whom the final diagnosis was other than rAAA (ie, coding errors, including ruptured thoracic aneurysms) were excluded. For the purpose of this study, a patient with an rAAA was defined as one who had had a computed tomographic scan diagnosis reported by a consultant radiologist as an rAAA or operative findings consistent with aneurysm rupture (ie, retroperitoneal hematoma or free intraperitoneal blood in the presence of an abdominal aortic aneurysm [AAA]). Demographic features were recorded, including comorbidities. Pulse and blood pressure at the time of admission to the RHH were recorded. Laboratory recordings of the following hematologic parameters were recorded at the time of admission and at the time of admission to the intensive care unit (ICU): hemoglobin (g/L), platelet count (/nL), hematocrit (%), international normalized ratio (INR), and activated partial thromboplastin time (APTT) (s). The volume of crystalloid, packed red blood cells, fresh frozen plasma (FFP), and platelets was recorded for two periods, prior to admission to the ICU (pre-ICU) and during ICU admission. Crystalloid and colloid were recorded only for the first 48 hours in the ICU as most patients will have been resuscitated by this time and will require only maintenance volumes of crystalloid thereafter.

The first EVAR of an rAAA at the RHH was performed in August 2004. From this time, all patients presenting with rAAA while one of the authors (S.R.W.) was on call (1:3) were considered for EVAR. Patients were excluded from EVAR using conventional anatomic criteria and as recommended by most stent graft manufacturers (proximal neck shorter than 15 mm, wider than 30 mm, or more than 60° angulation; iliac vessels less than 7 mm diameter). No patients in this report were treated for rAAA outside of these recommendations. All EVARs of rAAAs were performed by a consultant vascular surgeon and a consultant vascular radiologist in an angiography suite, under local anesthetic, using the bifurcated Zenith stent graft (Cook Medical, Brisbane, Australia), except for the first case, which was treated using a Zenith tube graft. All open rAAA repairs were performed by consultant vascular surgeons in an operating room, under general anesthetic, using a conventional transperitoneal approach and a Dacron (Vascutek, Renfrewshire, UK) tube or bifurcated graft. The surgeon performing EVAR of rAAA also performed nine ORs of rAAAs during the same time period.

Pre-ICU and in-ICU variables were compared between three groups (EVAR, OR and survived [ORS], and OR and died [ORD]) with a Mann-Whitney U test and p values ≤ .05 taken as significant.

Results

Between February 2004 and November 2008, 84 patients presented with a diagnosis of rAAA. One patient had undergone surgery in another hospital and was transferred with a resulting complication, and two patients were transferred to other hospitals for surgery. These three patients were excluded from subsequent analysis. Of the remaining 81 patients, 36 were deemed not to be operative candidates and were treated palliatively. This decision was always made by a consultant vascular surgeon and based on all of the information available at the time of admission, including the patient's age, comorbidities (eg, cardiac disease, renal failure, terminal cancer), the patient's wishes, and, where appropriate, the relatives' wishes. All of these patients died prior to discharge. Of the 45 patients who underwent intervention, 7 had EVAR, of whom all survived to discharge and 30 days (0% operative and 30-day mortality). Of the 38 patients who underwent OR, 16 died before discharge, for an operative mortality rate for OR of 42%. There were no further deaths between discharge and 30 days postoperatively. Of the nine cases of OR performed by the surgeon who also performed EVAR, two died. For all patients with rAAA undergoing intervention (OR plus EVAR), the overall operative mortality was 36%.

Table 1 shows the demographics of the patients who underwent intervention for rAAA. There were no significant differences in the age and admission pulse rate of those who had EVAR, those who had OR and survived, and those who had OR and died. Despite there appearing to be a higher admission systolic blood pressure and admission mean systolic pressure in those who underwent EVAR, this difference was not statistically significant (median admission systolic pressure EVAR [132 mm Hg] versus ORS [103 mm Hg], p = .21; EVAR [132 mm Hg] versus ORD [102 mm Hg], p = .8; ORS [103 mm Hg] versus ORD [102 mm Hg], p = .42; admission mean arterial pressure EVAR [90 mm Hg] versus ORS [75 mm Hg], p = .41; EVAR [90 mm Hg] versus ORD [72 mm Hg], p = .54; and ORS [75 mm Hg] versus ORD [72 mm Hg], p = .92).

Table 2 shows the admission hematologic variables for patients who underwent intervention. Again, there were no significant differences in the admission hemoglobin level, hematocrit, platelet count, INR, and APTT between the three groups.

Table 3 shows the intraoperative crystalloid and blood product requirements for patients undergoing intervention. Only one patient received perioperative Prothrombinex (a concentrate of factors II, IV, VII, and X used for acute warfarin reversal). This patient presented with an INR of 5.1 as he had been on warfarin. The patient underwent EVAR. Patients undergoing EVAR, when compared with those undergoing OR, received significantly less pre-ICU crystalloid (3 L vs 7.5 L, p = .001), less packed red blood cell transfusion (1 unit vs 6.5 units, p = .0006), and less colloid (0 L vs 0.5 L, p = .008).

Table 4 shows hematologic values on admission to the ICU and how on admission patients undergoing EVAR had lower hemoglobin levels (90 vs 106 g/L) and a lower hematocrit (27% vs 31%).

Table 5 shows the blood products and fluid requirements during the first 48 hours following admission to the ICU. There were no significant differences between the three groups.

Table 6 shows the total volumes of crystalloid, blood products, and colloid for those patients who had EVAR, OR and survived, and OR and died.

Discussion

The reduction in blood product requirements for patients undergoing elective EVAR of AAA has been previously documented. ⁶ Despite the small number of EVARs of rAAA reported in this article (type II error), there is a trend for a reduction in blood product requirements when compared to those undergoing conventional OR. It is interesting to note that patients undergoing EVAR of rAAA in our center arrive in the ICU with a lower hemoglobin level and hematocrit compared with those who undergo conventional OR. This may reflect a perception by both anesthetists and surgeons that the patient has undergone a less invasive procedure that does not require more aggressive resuscitation. It may also reflect that the blood loss is underestimated by the nondirect visualization of the retroperitoneal hematoma. The reduction in blood product requirement for EVAR of rAAA may have implications in smaller centers, where rapid access to large volumes of blood products may be difficult.

Some authors have suggested that EVAR of rAAA requires a large inventory of equipment and organizational design suited to large-volume tertiary hospitals. We have found that by working in close collaboration with stent graft manufacturers, a limited on-the-shelf stent graft stock can allow EVAR of rAAA in a small remote hospital. By rotating stock with elective cases, expiry dates of prosthesis are avoided. An inventory of our stock to deal with emergencies is shown in Table 7. All vascular surgeons at the RHH perform elective EVAR of AAA either alone or in combination with a vascular radiologist. Our elective experience of EVAR of AAA has enabled anesthetists, scrub nurses, and ward nurses to become confident in the management of patients following EVAR, and there have been no issues with those undergoing EVAR of rAAA. Similar comments have been made in publications from other small centers. ⁷ To put our aneurysm activity into perspective, over the same time period, 140 patients underwent elective infrarenal AAA repair, 31 using an endovascular technique. This reflects our initially conservative use of this technology.

This observational study lacks any great power to change practice in those not already offering EVAR of rAAA. Given the local population and resources, our numbers of both EVAR and OR of rAAA are low compared with those of larger centers. Therefore, there is a strong possibility of a type II error in some of the presented results. For example, the mean admission systolic pressure appears lower in the OR groups, but this was not statistically significant. Despite a Cochrane review concluding that there is no high-quality evidence to support the use of EVAR in the treatment of rAAA, large-volume centers continue to report encouraging results. ^8,9

Conclusion

We conclude that EVAR of rAAA is feasible in small rural vascular units and that this treatment modality may have favorable consequences for the local blood bank.