Diagnosis:

PIS is defined as a combination of constitutional symptoms, including fatigue and fever, with elevated inflammatory markers. There are several proposed diagnostic criteria. These result in discrepant rates, as low as 2% and up to 100%.^9,10 The common finding is the inclusion of fever as the symptom/sign for diagnosis. Regarding inflammatory markers, they are more variable. Most reports have used the systemic inflammatory syndrome criterion (SIRS) with a combination of fever and leukocytosis.^{1,3,4,5,6,12,13,14,15} Others have used a combination of fever coinciding with elevated C-reactive protein (CRP) levels.^2,7,16,17,18 Table 1 depicts diagnostic criteria used throughout the literature and corresponding PIS rates.

The first study reported was by Blum et al. in 1997. The authors prospectively analyzed the clinical outcomes after EVAR in 154 patients treated with polyester-covered nitinol stent grafts. Eighty-seven (56%) developed a fever (38.0-39.7°C) lasting 4-10 days. Most patients developed leukocytosis (range 9.800-29.500/μL) and CRP elevation (range 4 - 34.1 mg/dL). Diagnosis, however, was not based on specific markers.¹⁹

Two years later, Velazquez et al. described the use of specific markers for PIS diagnosis (leukocytosis and fever).¹¹ No cutoff values were reported. Within those with specified diagnostic criteria, incidence is similar (20-40%).

On the other hand, in the study by De la Motte et al.,¹⁰ the authors performed a randomized, double-blind, placebo-controlled trial that aimed to analyze the effect of a single preoperative dose of 30 mg/kg of methylprednisolone or placebo administered 2 hours before EVAR. For diagnosis of PIS, they used SIRS criteria (presence of at least two of the following criteria: temperature > 38°C or < 36°C; leukocytes >12.000/l, <4.000/ or > 10% bands; heart rate > 90; respiratory rate > 20; PaCO2 < 32 mm Hg), except the criterion of leukocytosis. Instead of leukocytosis, the criterion was elevation of CRP to >75 mg/L. With a single preoperative dose, they obtained a reduction of PIS from 100% in the placebo group to 27% in the treatment group. The 100% PIS rate in the placebo group is due to the addition of other clinical criteria that are common findings in EVAR patients, which are often associated with other cardiopulmonary diseases.

Many other studies did not specify diagnostic criteria.^9,19-22 In those with pre-specified criteria, there is a significant degree of variance in incidence. A universal consensual definition of this syndrome is urgently needed.

Clinical manifestations:

Symptoms associated with PIS are described above. The typical evolution is spontaneous resolution, although pharmacologic measures for symptom relief may be needed. These symptoms often resolve within two weeks; no significant permanent complications remain. However, kidney injury and impaired quality of life may ensue occasionally.^3,20 Conversely, a reduced risk for type 2 endoleak rates has been reported.^16,24 The impact on cardiovascular events remains to be determined.^16,25 The diagnosis will be frequently suspected in those with fever and inflammation and no infection source. Leukocytosis usually will peak on the first postoperative day, and CRP and temperature will peak on the second postoperative day. Most PIS cases will present up to the first 72 postoperative hours. A higher suspicion for infection should occur if inflammation and fever occur after this time.

Preditors of PIS:

The pathophysiology and etiology of PIS are partially clear. The mechanisms behind this condition seem multifactorial and mainly involve aneurysms and procedure-related characteristics.

1 - Stent graft material: Stent graft material is probably the most reliable predictor of PIS after EVAR. Multiple studies have compared the incidence of PIS and inflammatory response according to the implanted graft composition (polytetrafluoroethylene - PTFE vs Dacron), and the use of polyester-based grafts has been recurrently related to an increased odd of PIS, up to five-fold.^1-4,14 Vôute et al., in a retrospective analysis of 149 patients (82 polyester and 67 PTFE grafts), found a PIS rate of 56.1% vs 17.9%, P<.001 in polyester vs PTFE grafts, respectively. Almost every PIS case occurred during the first three postoperative days. The inflammatory response (temperature and CRP) presented a significantly higher magnitude after Dacron graft implantation, highest temperature rise (1.6 vs. 0.9°C, P<.001) and CRP level rise (154.8 vs. 38.0 mg/L, P<.001). Finally, a multivariable logistic regression model was constructed, and a woven polyester graft was the sole predictor of PIS (HR 5.58, 95% CI 1.60-19.42, P=.007).² Kakisis et al., in a retrospective analysis of 87 patients, found similar results when testing for risk factors for PIS.¹⁴ Chatzelas et al., in a prospective analysis of retrospectively collected data (191 patients), found that polyester graft composition was the single factor significantly associated with an increased PIS rate (HR 2.6, P=.043).²³

Contrary to these, other studies could not confirm a difference in PIS rates when stratifying for stent graft composition.^26-28 Gerasimidis et al. prospectively compared the degree of inflammatory response between endovascular aneurysm repair with polyester (12 patients) and ePTFE devices (10 patients). One patient in each group had PIS, according to SIRS criteria. Three patients in the polyester group had a fever (>38°C), and just one in the PTFE group (p < 0.005). There were no statistically significant differences between groups for all endpoints, possibly due to the sample size.²⁶ In a cohort reported by Sartipy et al., there were significant differences between the two types of graft material concerning fever and CRP, but there were no significant differences in the number of PIS events. It could also be related to sample size, with 32 patients treated with polyester grafts but only 13 with ePTFE grafts. They performed a sensitivity analysis showing that the results would have reached significance if three more patients in the polyester group had developed PIS (or none instead of one patient in the PTFE group).¹³ Lastly, Moulakakis et al. assessed the inflammatory and renal response after TEVAR, and no significant differences in inflammatory response between polyester and PTFE groups were noted. They attributed that to the small number of patients implanted with ePTFE stent grafts in their trial.²⁹

2- Thrombus: The hypothesis that the amount of thrombus (pre-existing or new-onset) within the aneurysm sac could be related to PIS derives from the finding that mural thrombus of an aortic aneurysm contains high levels of interleukin-6.²⁷ It is hypothesized that thrombus manipulation during the procedure would be able to release this inflammatory mediator to the systemic circulation, contributing to the postoperative inflammatory response. Also, newly formed thrombus after aneurysm exclusion could be able to increase this response, ultimately resulting in PIS. Nano et al. reported an association between preoperative thrombus thickness and PIS with EVAR using the Anaconda^® stent graft (P=0.001).³ Kakisis et al. (87 patients) found a mean volume of chronic mural thrombus of 63.7 ± 73.6 mL, and a mean volume of new-onset thrombus was 55.1 ± 74.8 mL. A statistically significant correlation was found between the volume of new-onset thrombus and peak postoperative temperature (r=.379; P<0.01) as well as the increase in CRP level (r=.533; P<0.001) and IL-6 level (r=.380; P=0.01). On the contrary, the volume of chronic mural thrombus was not found to affect any parameter of PIS significantly.¹⁴ The volume of chronic mural thrombus, as well as the volume of new-onset thrombus, was calculated using imaging software. The closed polygon tool was used to outline the perimeter of the aneurysm sac on every transverse slice of the CTA, starting from the level of the renal arteries and ending at the aortic bifurcation. With this input, the software automatically calculated the aneurysm sac volume. In 2012, Voute et al., other than studying the influence of graft material, aimed to analyze the association between new-onset thrombus and the inflammatory response. They report no significant association between new-onset thrombus and temperature or C-reactive protein elevation.² Arnaoutoglou et al., in a prospective evaluation of 214 EVAR patients, found no significant association between preoperative and newly formed thrombus and increased inflammatory response.5 Chatzelas et al., in a cohort of 191 patients, found that new-onset thrombus did not significantly relate to an increased chance of developing PIS (OR 1.29, P=0.10).²³ Literature reports are variable regarding this specific feature, and although most discuss the association between mural thrombus (chronic or new-onset) and PIS, the real impact remains to be determined. Finally, as polyester stent grafts are associated with increased intraprosthetic mural thrombus, this raises the question of whether stent graft material does not act as a significant confounder of new-onset thrombus.²⁷

3 - Bacterial translocation: Transient bacterial translocation is another potential factor impacting the chance of developing PIS. This is thought to be due to the occlusion of a previously patent inferior mesenteric artery (IMA). The IMA is inevitably occluded during EVAR procedures, except in highly selected cases. This occlusion could result in transient colonic ischemia with the consequent risk of bacterial translocation, significantly contributing to PIS occurrence. Kakisis et al. analyzed the association between the patency of the IMA and the postoperative temperature and inflammatory markers. No significant correlation was found.¹⁴ Akin et al. performed a prospective single-center trial involving 40 patients. They aimed to study the effects of antibiotic therapy in PIS after thoracic aortic stent placement. Two groups were created, one receiving cefuroxime 20mg/kg tid during the first 24h postoperative hours and the other receiving the same dosing scheme during the first seven days. There were no significant differences in clinical or laboratory findings between treatment strategies. Therefore, the hypothesis of bacterial translocation as a cause for PIS seems remote, and there is no evidence to support it.

4 - Contrast media: Videm et al. suggested that the contrast medium iohexol provokes neutrophil degranulation, which is greatly enhanced when combined with stent graft material, contributing to PIS occurrence.³⁰ Other studies analyzed inflammatory response after endovascular aneurysm repair but did not find any correlation between contrast use or dosage and PIS parameters.^1,3,14,29 As such, this theory remains to be demonstrated.

5 - Other factors: The type of procedure performed can be independently associated with different chances of PIS. Using an additional iliac branch device or fenestrated graft for the visceral aorta results in additional procedural steps and stent graft material, increased branch manipulation, more lengthy operations, and more associated complications. These criteria have been scarcely studied, and fenestrated branched aortic grafts are frequent exclusion criteria throughout the literature. Ferreira et al., in a retrospective analysis of 52 patients, found a 0% PIS incidence after f/bEVAR. However, just four patients were included.¹⁵ In addition, Ribeiro et al. found an increased rate of PIS after complex EVAR when compared to standard infrarenal repair (34.1% vs. 18.2%, P=0.005; aOR 2.833, P=0.009).¹⁸ The impact of previous exposure to an aortic stent graft, particularly in cases of secondary interventions, has not been addressed. Other factors, such as age, gender, aneurysm diameter, length of operation, blood loss or blood product requirements, COPD, or heart disease, have been explored; however, no significant associations were noted.^1,3,13,20