Six Years' Experience with Prostaglandin I2 Infusion in Elective Open Repair of Abdominal Aortic Aneurysm: A Parallel Group Observational Study in a Tertiary Referral Vascular Center
Article Outline
The prostaglandin I2 (PGI2) analogue iloprost, a potent vasodilator and inhibitor of platelet activation, has traditionally been utilized in pulmonary hypertension and off-label use for revascularization of chronic critical lower limb ischemia. This study was designed to assess the effect of 72
hr iloprost infusion on systemic ischemia post–open elective abdominal aortic aneurysm (EAAA) surgery. Between January 2000 and 2007, 104 patients undergoing open EAAA were identified: 36 had juxtarenal, 15 had suprarenal, and 53 had infrarenal aneurysms, with a mean maximal diameter of 6.9cm. The male-to-female ratio was 2.5:1, with a mean age of 71.9 years. No statistically significant difference was seen between the study groups with regard to age, sex, risk factors, American Society of Anesthesiologists (ASA) grade, or diameter of aneurysm repaired. All emergency, urgent, and endovascular procedures for aneurysms were excluded. Fifty-seven patients received iloprost infusion for 72hr in the immediate postoperative period compared with 47 patients who did not. Patients were monitored for signs of pulmonary, renal, cardiac, systemic ischemia, and postoperative intensive care unit (ICU) morbidity. Statistically significantly increased ventilation rates (p=0.0048), pulmonary complication rates (p=0.0019), and myocardial ischemia (p=0.0446) were noted in those patients not receiving iloprost. These patients also had significantly higher renal indices including estimate glomerular filtration rate changes (p=0.041) and postoperative urea level rises (p=0.0286). Peripheral limb trashing was noted in five patients (11.6%) in the non-iloprost group compared with no patients who received iloprost. Increased rates of transfusion requirements and bowel complications were noted in those who did not receive iloprost, with their ICU stay greater than twice that of iloprost patients. All-cause morbidity affected 67% of patients not receiving iloprost compared to 40% who did. Survival rates were significantly better with iloprost than without in both 30-day (p=0.009) and 5-year cumulative (p=0.0187) survival. Iloprost infusion for 72hr after open AAA repair was associated with improved systemic perfusion and decreased systemic ischemia. Patients had a significant survival benefit at 30 days and 5 years and significantly improved renal, cardiac, and respiratory function.
Introduction
A major objective post–abdominal aortic aneurysm (AAA) repair is to maintain systemic perfusion and minimize the risk of peripheral embolization in the immediate postoperative period. Due to the inherent risk of bleeding from the anastomotic site with the use of intravenous heparin, an alternative method has been sought to maintain optimal perfusion during this high-risk period.
Iloprost is a synthetic analogue of prostacyclin (Epoprostenol) with similar pharmacodynamic activity. It has been used successfully in the treatment of pulmonary arterial hypertension1, 2 and in patients with chronic thromboembolic pulmonary hypertension.3 As a potent vasodilator and effective inhibitor of platelet aggregation, iloprost has been utilized with varying degrees of success in the management of acute limb ischemia,4 advanced obliterative arteriosclerosis, Raynaud's phenomenon,5, 6 and congestive heart failure.7
This study assessed iloprost, a prostaglandin I2 (PGI2) analogue, as a potential adjunct to maintaining systemic perfusion while avoiding the postoperative hemorrhagic risk associated with heparin. This is a parallel group comparison and pivotal study to investigate the safety and efficacy of iloprost use in the postoperative period in patients undergoing elective AAA (EAAA) repair.
Methods
The aim of this study was to assess the effect of a 72hr iloprost infusion on ischemic injury post–open EAAA repair in a tertiary vascular referral center. Patients were included if their surgery occurred between January 1, 2000, and January 1, 2007. All emergency, urgent, and endovascular AAA repairs were excluded. The study protocol was approved by the pharmacological hospital ethical committee, for use in EAAA repair.
Data were acquired from a prospectively maintained VascuBase® system (Consensus Medical, Vancouver, Canada), which maintains a database of patient demographics, disease pathology, operative and treatment details, vascular laboratory scanning records, and long-term follow-up. Patient demographics including age and sex were recorded, as were the aneurysmal dimensions including maximal diameter and location in relation to the renal vessels (Fig. 1). These demographics, illustrated in Table I, showed no statistically significant differences between the study groups.
Fig. 1
Aneurysm demographics.
Table I. Demographics
| PGI2 infusion | No PGI2 infusion | p | |
|---|---|---|---|
| Patient | 57 (45.2%) | 47 (54.8%) | n/a |
| Age (mean, years) | 71.9 | 72.4 | 0.345 |
| Male gender (%) | 67 | 79 | 0.09 |
| ASA grade (mean) | 3.4 | 3.4 | 0.96 |
| Smoking | 42 (40.4%) | 31 (29.8%) | 0.391 |
| Hypertension | 29 (27.9%) | 21 (20.2%) | 0.529 |
| Diabetes mellitus | 4 (3.8%) | 3 (2.9%) | 0.898 |
| Chronic airways disease | 3 (2.9%) | 3 (2.9%) | 0.807 |
| Hypercholesterolemia | 7 (6.7%) | 5 (4.8%) | 0.794 |
| Ischemic heart disease | 20 (19.2%) | 15 (14.4%) | 0.733 |
| Renal impairment | 2 (1.9%) | 2 (1.9%) | 0.844 |
Pulmonary function was assessed preoperatively by formal pulmonary function testing within the institution respiratory unit. The rate of postoperative ventilation and the duration thereof, including the need for tracheostomy, were recorded. Rates of pulmonary edema and effusion were recorded as an adjunct to assessment of pulmonary function. Clinically and radiologically confirmed pneumonia, embolism, and atelectasis were recorded as morbidity.
Indices used to assess renal function included serial pre- and postoperative urea and creatinine levels, pre- and postoperative estimated glomelular filtration rates (eGFR), creatinine clearance estimation, and postoperative dialysis rate.
Cardiac function with particular regard to ejection fraction, valvular defects, and cardiomyopathy were assessed by transthoracic echocardiography, along with electrocardiographic recordings and troponin levels. Any postoperative hemodynamic/clinical myocardial infarctions were recorded.
Bowel function was monitored using pH, lactate levels, confirmed mesenteric ischemia, and altered bowel function.
Lower limb complications including evidence of peripheral limb trashing was recorded.
Hematological changes were monitored by serial hemoglobin levels and platelet count analysis. Patients in the iloprost infusion and noninfusion groups had equal availability of cell-saver autotransfusion intraoperatively.
Iloprost was given as an infusion commenced immediately on arrival in the intensive care unit (ICU) and run at 100 μg diluted in 500mL normal saline at a rate of 20mL/hr continuously for 72hr. No dosage adjustment was made with regard to individual creatinine levels during this 72hr period. No patients received the iloprost infusion during the operative case, as per procedure protocol. Patient selection for administration of the PGI2 infusion was in accordance with the operating consultant vascular surgeon. Patients were monitored in the ICU for respiratory and/or hemodynamic side effects.
Patients were routinely prescribed aspirin 300mg once daily on postoperative day 7 and clopidogrel 75mg once daily on day 10 postoperatively. Patients also received a once-daily injection of low–molecular weight heparin (tinzaparin 3,500IU subcutaneously) from day 1 postoperatively until discharge from hospital.
End Points
Primary end points included all-cause mortality subdivided to 30 days and 5-year cumulative survival.
Secondary end points included major adverse clinical events such as cardiac, renal, respiratory, bowel, and peripheral ischemia.
Statistical Methods
Analysis was performed using one-way analysis of variance for independently weighted samples and t-test for comparison of means. p<0.05 was taken as significant. Kaplan-Meier survival analysis was used to assess 30-day and 5-year cumulative survival rates.
Results
5-Year Cumulative Survival
A significant 5-year cumulative survival rate was noted in patients who had received an iloprost infusion compared to those who did not. A 95.4±10% rate of survival with iloprost compared with 76.7±7.89% 5-year survival without iloprost (p=0.0187, hazard ratio=0.20, 95% confidence interval 0.05-0.76) (Fig. 2).
Fig. 2
Kaplan-Meier 5-year survival curve.
30-Day Mortality
Figures for 30-day mortality were calculated for both groups excluding any mortality occurring in the first 72hr postoperatively. Ten patients who did not receive iloprost died (22%) during this period compared with two patients (4%) who had received an iloprost infusion, showing a significant benefit in mortality during the early postoperative period (p=0.009).
Pulmonary Morbidity
Significantly higher rates of ventilation were noted postoperatively in patients not receiving iloprost, with 22 patients (48%) requiring mechanical support compared to 10 patients (18%) in the infusion group (p=0.002).
In concordance with this, a higher rate of pulmonary complications including pneumonia, edema, effusions, atelectasis, embolism, and respiratory failure was noticed in the non-iloprost group, with 48.8% of patients experiencing morbidity related to pulmonary function compared to 17.8% of patients in the iloprost cohort (p=0.001).
Renal Morbidity
eGFR was measured preoperatively along with the minimum postoperative value. Mean values were compared between the iloprost infusion and noninfusion groups. A significantly greater decline in mean eGFR was observed in the noninfusion group, 15mL/min/1.73m2, compared with the PGI2 infusion group, 9.4mL/min/1.73m2 (p=0.041) (Table II). Urea and creatinine were recorded in all patients preoperatively, 1 day postoperatively, and as a serial measurement up to discharge from the ICU, with maximal levels recorded for both parameters. Creatinine clearance was estimated from preoperative creatinine and maximal ICU creatinine levels. A significant rise from preoperative urea levels occurred in patients not receiving iloprost, rising an average of 7.39mmol/L during the postoperative ICU course compared with 2.47mmol/L in patients receiving iloprost (p=0.0286).
Table II. Renal morbidity
| PGI2 infusion | No PGI2 infusion | p | |
|---|---|---|---|
| Post-operative decline in eGFR (mL/min/1.73m2) | 9.4 | 15 | 0.041 |
| Mean postoperative urea rise (mmol/L) | 2.47 | 7.39 | 0.0286 |
| Mean postoperative creatinine rise (μmol/L) | 25.4 | 59 | 0.119 |
| Mean creatinine clearance reduction (mL/min) | 4.1 | 10.4 | 0.08 |
| Acute renal failure (%) | 13.3 | 25.6 | 0.072 |
| Renal dialysis (%) | 8.9 | 18.6 | 0.216 |
The rise in creatinine levels from preoperative samples was more than double in patients not receiving iloprost, with an average rise of 59μmol/L compared to 25.4μmol/L in patients on the infusion (p=0.119).
Creatinine clearance was calculated and the drop in clearance from preoperative levels averaged 10.4mL/min in patients not receiving iloprost and 4.1mL/min in those who did (p=0.08).
Eleven patients (24%) in the non-iloprost group developed acute tubular necrosis compared with six receiving iloprost (11%) (p=0.07), and six of 11 patients required dialysis compared to two of six iloprost patients.
Cardiac Morbidity
Myocardial ischemia was observed with a significantly greater frequency in patients who did not receive iloprost in the postoperative course, with a significant troponin rise in 17 of these patients (37%) compared with 10 patients (18%) who received 72hr iloprost infusion (p=0.0223) (Table III).
Table III. Systemic morbidity
| PGI2 infusion | No PGI2 infusion | p | |
|---|---|---|---|
| Bowel complications | 1 | 8 | 0.06 |
| Mean ICU stay (days) | 6.1 | 13.2 | <0.05 |
| Peripheral trashing | 0 | 5 | 0.0219 |
| Myocardial ischemia | 10 | 17 | 0.047 |
There were eight hemodynamically/clinically relevant myocardial infarctions reported. In the non-iloprost group six patients died from the event, while only two patients out of 10 died in the iloprost group.
Bowel Morbidity
No significant differences were recorded with regard to overall bowel complications, although persistent diarrhea (>3 days) was noted in five patients not receiving iloprost compared with only one in the infusion group.
Raised lactate levels were noted with a higher frequency in patients not receiving iloprost (n=12, 26%) compared with seven (12.5%) of those who did (p=0.07).
Other bowel morbidity noted in the non-iloprost group included a perforated duodenal ulcer (n=1), upper gastrointestinal bleed (n=1), and pseudomembranous colitis (n=1).
Peripheral Ischemia
Five patients (11%) had peripherally trashed lower limbs, and all were in the non-iloprost arm of the study (p=0.018).
ICU Stay
Patients had an ICU stay 50% shorter if they had iloprost infusion postoperatively, with an average of 6.3 days compared to 13.1 days if they did not receive iloprost (p=0.05).
Blood Parameters
Blood transfusion requirements were greater in the noninfusion group, with a mean of 4.15 units of packed red blood cells tranfused compared to 3.6 units in the iloprost infusion group (p=0.09).
Patients were appraised with regard to blood parameters including serial hemoglobin and platelet counts. A significant drop in platelet count was noted in patients receiving iloprost, with a mean decrease of 46g/dL during the ICU stay compared to a mean rise of 25g/dL (p=0.006) in the non-iloprost group. Between presurgery and ICU discharge, mean hemoglobin count fell by 2.2g/dL in patients on iloprost infusion compared to 3g/dL in patients not on an infusion (p=0.006).
Iloprost-Related Adverse Events
Three patients of the 45% who received an iloprost infusion experienced an adverse event. One developed severe hypotension requiring discontinuation of the iloprost infusion after 36hr. Two developed pulmonary edema, one toward the latter part of the infusion; and no change to infusion protocol was required.
Patients Excluded from the Study
Any mortalities occurring within the first 72hr were excluded as patients had not received a full course of iloprost during this period. These data were recorded and subanalyzed. Two mortalities occurred within the first 72hr postoperatively, one in each group; and both were excluded from subsequent analysis. Both mortalities were secondary to cardiac arrest within the first 24hr postoperatively.
Discussion
In major complex operative procedures such as open EAAA repair it remains imperative to maximize patients both preoperatively and postoperatively. Given the nature of this type of surgery, the potential for ischemic, embolic, and inflammatory complications remains high and almost one-third of patients who undergo open AAA procedures will experience a morbidity of varying degree. In an effort to minimize these potential complications, we developed a novel regimen of iloprost infusion.
The potential role for iloprost in a broad range of clinical conditions has not gone unnoticed, with studies purporting the benefits of such a medication in an array of conditions associated with vascular insufficiency. These have included treatment of digital ischemia in combination with dextran,8 long-lasting clinical improvement in patients with occlusive peripheral vascular disease,9 prolonged exercise tolerance in patients with stable angina pectoris,10 effective treatment of vasospastic peripheral vascular diseases such as Raynaud's phenomenon,6 inhibition of platelet aggregation associated with heparin-induced platelet activation, and allowing cardiovascular surgical procedures to be performed in those patients with decreased risk of thrombotic complications.11, 12, 13, 14 Case studies have also reported benefit in renal failure secondary to vasospasm,15 thromboangitis obliterans, ischemia of the digits, and renal failure in cholesterol emboli syndrome.16
Many of these publications have shown benefit in conditions that patients post–open AAA repair are liable to suffer from. As a logical progression of ideas, our development of an iloprost infusion in the early postoperative setting was aimed at prophylactically preventing patients developing such ischemic injury.
The pharmacokinetics of iloprost are multifactorial: (1) PGI2 stimulates adenylate cyclase and causes an increase in circulating cyclic adenosine monophosphate (cAMP), which is modulated by the activation of calcium and potassium channels in endothelial smooth muscle cells, leading to hyperpolarization and subsequent inhibition of myosin kinases. This effect is propagated by iloprost's inhibitory action on leukotrienes, thromboxane A2, endothelin-1, and endothelium-derived constricting factor.5 (2) Iloprost is a potent inhibitor of platelet activation and aggregation, and studies have shown its anti-aggregatory effect to be 10 times that of endogenous prostacyclin and over 100 times that of alprostadil (PGE2).17 (3) Anti-inflammatory and antifibrotic activities are mediated through inhibition of proinflammatory and profibrotic cytokines and inhibition of leukocyte activation and adhesion. (4) The half-life of iloprost given by intravenous administration is 30min, with steady-state distribution of active substance in organs and tissues achieved 3-4min from commencement of infusion. Seventy percent is eliminated by the kidney, with a smaller percentage eliminated with the feces.18 Severe hepatic and renal dysfunction influence the plasma concentrations of iloprost, but mild and moderate renal impairment did not appear to impede clearance.19
Major drawbacks to the use of iloprost infusion relate to its effects on hemodynamic stability, with dose-related reductions in vascular resistance and arterial pressure with or without increases in heart rate.10, 14, 20, 21 Such changes can lead to marked hypotension; hence, a systolic pressure less than 85mm Hg is a relative contraindication to its use. Serious adverse reactions have been documented including pulmonary edema as a consequence of pulmonary venous hypertension and other side effects such as persistent dry cough, headaches, and trismus. No such hemodynamic instability was observed in our study in any of the patients who had perioperative epidural anesthesia.
Renal and hepatic failure must be taken into account with iloprost. Drug clearance can be reduced by half in patients with impaired liver function (Child Pugh class B). One study notes that in patients with severe liver cirrhosis the terminal half-life of the drug remained the same but that after an infusion of 1mg/kg/min the steady-state concentrations were up to twice as high as in healthy volunteers.22
Severe renal failure has also been shown to affect the clearance of iloprost postinfusion, with Hildebrand et al.22 showing a threefold reduction in clearance postinfusion in end-stage renal failure patients compared to patients with normal renal function. Dose adjustment is not required in patients not on dialysis.
As primary end points, 30-day and 5-year cumulative survival rates were significantly improved using an iloprost infusion; and this alone suggests a significant role for such an infusion as an adjunct in early postoperative open EAAA repairs.
We subsequently looked at a number of secondary end points related to systemic ischemia to assess the overall benefit of the infusion regimen. Iloprost has been shown to be of benefit in the treatment of cardiac ischemia, embolic conditions, peripheral vascular disease, diabetic angiopathy, and Raynaud's disease. Our data assessed its protective role with regard to systemic circulation and the maintenance of organ perfusion. We reported the macrocirulatory clinical findings of peripheral limb trashing; however, future assessment of the microcirculation, with regard to transcutaneous oxygen measurement, may supplement the actions of the infusion. To varying degrees, cardiac, pulmonary, and renal functions showed significant benefit related to the infusion. Those results that failed to show statistical significance did, however, show a trend toward better outcome. Due to the power of our study, it was not possible to significantly identify any subgroups of patients with preexisting comorbidities who benefited from the iloprost infusion. However, with a larger prospective study, these higher-risk patients may be recognized for advantageous infusion.
No other study has assessed the use of iloprost in such a cohort of patients, and the benefits suggested by our data lead us to believe that, although further studies are needed in relation to this therapy, iloprost infusion is an exciting and potentially revolutionary new adjunct to the management of open AAA repair in the early postoperative course. As a result of this study, all of our patients who undergo open elective AAA repair now routinely receive this novel iloprost infusion regimen postoperatively.
We believe that the favorable trends in improved systemic perfusion which we demonstrated may indeed show significance with a larger study cohort. Unfortunately, given the nature of the procedure and the advent of endovascular techniques, especially in EAAA repair, a large enough cohort remains elusive and beyond the scope of this study. Any potential selection bias may be avoided by randomizing patients, and thus preventing self-selection. The hope would be for larger multicenter trials to supplement our initial findings and to prospectively assess a larger cohort in a randomized manner.
Conclusion
Iloprost infusion for 72hr after open AAA repair is associated with improved systemic perfusion and decreased systemic ischemia. Iloprost, as an adjunct to surgery, offers a significant survival benefit at 30 days and 5 years with significantly decreased renal, cardiac, and respiratory complications.
References
- . Prostanoid therapy for pulmonary arterial hypertension. Clin Chest Med. 2007;28:127–142
- . Inhaled iloprost to control residual pulmonary hypertension following pulmonary endarterectomy. Eur J Cardiothorac Surg. 2005;28:882–888
- . Acute improved hemodynamics following inhaled iloprost in chronic thromboembolic pulmonary hypertension. Respiration. 2008;76(2):154–159
- ILAILL Study Group. Acute limb ischemia in elderly patients: can iloprost be useful as an adjuvant to surgery? Results from the ILAILL study. Eur J Vasc Endovasc Surg. 2007;34:194–198
- Iloprost treatment in patients with Raynaud's phenomenon secondary to systemic sclerosis and the quality of life: a new therapeutic protocol. Rheumatology (Oxford). 2006;45:999–1004
- Oral iloprost in Raynaud's phenomenon secondary to systemic sclerosis: a multicentre, placebo-controlled, dose-comparison study. Br J Rheumatol. 1998;37:952–960
- Iloprost improves hemodynamics in patients with severe chronic cardiac failure and secondary pulmonary hypertension. Can J Anaesth. 2002;49:1076–1080
- . Acute hand ischemia secondary to intraarterial buprenorphine injection: treatment with iloprost and dextran-40: a case report. Angiology. 1999;50:605–606
- Clinical benefits of iloprost, a stable prostacyclin (PGI2) analog, in severe peripheral arterial disease (PAD). In: Samuelsson B, Wong YK, Sun FF editor. Advances in Prostaglandin, Thromboxane, and Leukotriene Research, vol. 19. New York: Raven Press; 1989;p. 311–316
- Effects of iloprost, a stable prostacyclin analog, on exercise capacity and platelet aggregation in stable angina pectoris. Am J Cardiol. 1986;58:453–459
- Intraoperative management of patients with heparin-induced thrombocytopenia. Ann Thorac Surg. 1990;49:714–723
- Surgical management of heparin-associated thrombocytopenia. J Vasc Surg. 1988;8:395–401
- Efficacy of iloprost (ZK36374) versus aspirin in preventing heparin-induced platelet activation during cardiac operations. J Thorac Cardiovasc Surg. 1987;94:405–413
- Prevention of heparin-induced thrombocytopenia during open heart surgery with iloprost (ZK36374). Surgery. 1987;102:796–807
- Acute oliguric renal failure associated with unilateral renal embolism: a successful treatment with iloprost. Am J Nephrol. 1998;18:444–447
- . Improvement in cholesterol emboli syndrome after iloprost therapy. BMJ. 2002;324:268–269
- . Comparison of equimolar concentrations of iloprost, prostacyclin, and prostaglandin E1 on human platelet function. J Lab Clin Med. 1987;109:184–190
- . Pharmacokinetics and pharmacodynamics of radio-labeled iloprost in elderly volunteers. Eur J Clin Pharmacol. 1987;32:597–605
- Pharmacokinetics of iloprost in patients with hepatic dysfunction. Int J Clin Pharmacol Ther Toxicol. 1990;28:430–434
- Central hemodynamic and antiplatelet effects of iloprost—a new prostacyclin analogue—in acute myocardial infarction in man. Eur Heart J. 1987;8:362–368
- Myocardial ischemia induced by prostacyclin and iloprost. Clin Pharmacol Ther. 1985;38:101–108
- . Pharmacokinetics of iloprost in patients with chronic renal failure and on maintenance haemodialysis. Int J Clin Pharmacol Res. 1990;10:285–292
PII: S0890-5096(08)00340-3
doi:10.1016/j.avsg.2008.08.036
© 2008 Annals of Vascular Surgery Inc. Published by Elsevier Inc All rights reserved.
