Pulmonary injury follows systemic inflammatory reaction in infrarenal aortic surgery*

Laboratory Investigations

Pulmonary injury follows systemic inflammatory reaction ininfrarenal aortic surgery*

Chiara Adembri, MD, PhD; Eleni Kastamoniti, MD; Iacopo Bertolozzi, MD; Simone Vanni, MD, PhD;Walter Dorigo, MD; Mirella Coppo, BSc; Carlo Pratesi, MD; A. Raffaele De Gaudio, MD;Gian F. Gensini, MD; Pietro A. Modesti, MD, PhD

Multiple complications occurafter repair of a thoracoab-dominal aortic aneurysm,the most common of which

is respiratory failure. The occurrence ofpulmonary dysfunction and respiratoryfailure is related to both the duration ofaortic occlusion and the level of clamping(1–3). A systemic inflammatory response

(SIR), characterized by a consistent in-crease in cytokine plasma concentration(tumor necrosis factor [TNF]-� and inter-leukin [IL]-6) (4–6), was reported to fol-low aortic surgery. In experimental stud-ies, inflammatory cytokines were foundto mediate neutrophil chemoattractionwithin the lung and the kidney with en-dothelial cell swelling, capillary leak, andedema (7). In humans, the increasedplasma concentrations of IL-6 and IL-8after aortic surgery were found to corre-late with increased protein permeabilityin lung (8). However, no clear relation-ship was observed between cytokineplasma concentrations and clinical out-come after aortic surgery (9), so that theclinical effects of a transient SIR remainunclear.

The maximal cytokine response oc-curs in thoracic aortic surgery (10), and

ischemic-reperfused abdominal visceraare considered the major source of cyto-kine release (11, 12). On the other hand,the participation of skeletal muscle in thesystemic inflammatory reaction remainselusive, although the lower limbs alwaysundergo ischemia reperfusion in aorticsurgery and represent a significant per-centage of body weight. Indeed, experi-mental studies performed in rats indicatethat lower limb reperfusion may inducedamage in remote tissue, causing alteredgut permeability (13) and gastrointestinalmucosal changes (14). In humans, aorticclamping during aneurysm repair in-duces a relevant local inflammatory re-sponse in lower limb skeletal musclecharacterized by complement activationand endothelial expression of adhesionmolecules (15) followed by granulocyterecruitment and local free radical pro-

*See also p. 1233.From Anesthesia and Intensive Care (CA, EK,

ARDG), Clinica Medica e Cardiologia (IB, SV, MC, GFG,PAM), and Vascular Surgery (WD, CP), Department ofCritical Care Medicine, University of Florence, Italy.

Supported, in part, by grant 9906108278 from theMinistero dell’Università e della Ricerca Scientifica eTecnologica, Roma, Italy, and by Prof. Soubrier, IN-SERM, Paris, who kindly donated the cDNA CE probe.

Copyright © 2004 by the Society of Critical CareMedicine and Lippincott Williams & Wilkins

DOI: 10.1097/01.CCM.0000124875.98492.11

Objective: To investigate whether an inflammatory responseoccurs in patients undergoing infrarenal aortic abdominal aneu-rysm repair, the localization and timing (ischemia and/or reper-fusion) of this activation, and finally whether it affects postoper-ative pulmonary function.

Design: Prospective, observational study.Setting: Academic referral center in Italy.Patients: We included 12 patients undergoing infrarenal aortic

abdominal aneurysm repair and 12 patients undergoing majorabdominal surgery.

Interventions: Timed measurement of gene activation (angio-tensinogen, angiotensin type 1 receptor, angiotensin-convertingenzyme, and interleukin-6 genes) in muscle biopsies by reversetranscriptase-polymerase chain reaction (RT-PCR), and prospec-tive assessment of interleukin-6 plasma concentration and pul-monary function (PaO2/FIO2 and PaO2/PAO2 ratios).

Measurements and Main Results: After 30 mins of aorticclamping, angiotensinogen, angiotensin type 1 receptor, angio-tensin-converting enzyme, and interleukin-6 genes were all over-expressed at RT-PCR studies in quadriceps muscle of patientsundergoing aortic abdominal aneurysm repair, and the overex-pression persisted after reperfusion. In situ hybridization and

immunohistochemistry revealed that the inflammatory responsewas localized in endothelial cells. A significant increase in plasmainterleukin-6 concentrations was then detectable at 6 and 12 hrsafter reperfusion in aortic abdominal aneurysm surgery comparedwith patients undergoing abdominal surgery (p < .05). The in-crease in interleukin-6 plasma concentration was then followed(12 and 24 hrs after surgery) by a significant reduction of PaO2/FIO2 and PaO2/PAO2 ratios (p < .05 vs. abdominal surgery).

Conclusions: The present study shows that a) during aorticsurgery, the genes for interleukin-6 and for the components of thelocal renin-angiotensin system (angiotensinogen, angiotensin-converting enzyme, and angiotensin type 1 receptor subtype) areactivated early in the ischemic muscle, and activation persistsduring reperfusion; b) interleukin-6 plasma concentration in-creases only in patients with tissue ischemia (aortic abdominalaneurysm), whereas no changes are detectable in patients withabdominal surgery; and finally c) the occurrence of systemicinflammatory reaction with increased interleukin-6 plasma con-centrations is followed by impaired pulmonary function. (Crit CareMed 2004; 32:1170–1177)

KEY WORDS: ischemia reperfusion; aortic surgery; interleukin-6;renin-angiotensin system

1170 Crit Care Med 2004 Vol. 32, No. 5

duction (16). Oxidative stress (17) andangiotensin II (Ang II) via the angioten-sin type 1 receptor subtype (AT1) (17, 18)are known to stimulate IL-6 productionby endothelial cells. The clinical use ofAng II inhibitors also has been proposedto prevent ischemia-reperfusion syn-drome (19). However, cytokine releaseand overexpression of AT1 receptor (20)have been documented only in ischemicreperfused hearts of experimental ani-mals, whereas no studies were performedin human skeletal muscle. The absence ofdirect studies investigating the local in-flammatory reaction in skeletal musclealso leaves unresolved whether cytokineproduction in humans may be activatedduring aortic clamping (ischemia), assuggested by the effect of hypoxia on iso-lated endothelial cells (21), or duringreperfusion, as suggested by the contin-uous increase in plasma concentration ofproinflammatory cytokines throughoutthe period of reperfusion in experimentalmodels (22).

Therefore, the aim of the presentstudy was to prospectively investigate a)the activation of the components of thelocal renin-angiotensin system (RAS)(angiotensinogen [AGTN], angiotensin-converting enzyme [ACE], and AT1 re-ceptor subtype) and of interleukin-6 inischemic-reperfused muscle; b) the rela-tionship between lower limb ischemiaand increase of IL-6 plasma concentra-tion after surgery; and c) the relationshipbetween IL-6 plasma changes and pulmo-nary function. The study was performedin patients undergoing elective surgeryfor infrarenal aortic abdominal aneurysm(AAA) repair. The control group was rep-resented by patients undergoing abdom-inal surgery (AS) with expected compara-ble blood losses and duration ofanesthesia but no tissue ischemia duringsurgery.

MATERIALS AND METHODS

Subjects. The study was performed in twogroups of patients scheduled for elective aorticreconstructive surgery for infrarenal AAA re-pair or nonlaparoscopic elective abdominalprocedure (intestinal, gut, or kidney resec-tion) (AS), with comparable expected durationof anesthesia and blood losses.

Subjects were excluded from enrollmentfor the following reasons: diabetes, evidence ofperipheral vascular disease, recent history ofmyocardial infarction, intestinal chronic in-flammatory disease, widespread cancer, andinability to understand informed consent.

ACE inhibitors and AT1 receptor antago-nists were suspended �2 wks before surgery.Hypertensive patients received �1-adrenergicantagonists, adrenergic neuron-blockingdrugs, central nervous system acting antihy-pertensives, and direct-acting vasodilators.

Experimental protocols were approved bythe Institutional Review Board. All the pa-tients gave informed written consent to par-ticipate and to have their muscle and bloodsamples used for the study in strict compli-ance with the Helsinki Declaration (23).

Experimental Protocol. Patients under-went the same anesthetic procedure. After pre-medication with morphine sulfate (0.15 mg/kg), anesthesia was induced with proprofol(2.5 mg/kg) and muscle relaxation wasachieved with atracurium besilate (0.8 mg/kg).The trachea was intubated and patients wereconnected to a Cato ventilator (Draeger, Ger-many) in intermittent positive pressure venti-lation. A tidal volume of 8–10 mL/kg wasselected. Anesthesia was maintained withisoflurane in a mixture of oxygen/air; a con-tinuous infusion of atracurium (0.04mg·kg�1·hr�1) maintained muscle relaxation.End-tidal CO2 and oxygen saturation weremonitored to assess the adequacy of ventila-tion, and frequent arterial blood gas analysesallowed the maintenance of acid-base balance.Electrocardiogram and blood pressure weremonitored, as well as urine output. Bloodlosses were recorded and replaced with ade-quate amount of crystalloid and colloid solu-tions.

Infrarenal AAA repairs were performedthrough xiphopubic laparotomy. The aortaand the iliac vessels were prepared, isolated,and clamped below the renal arteries aftersystemic heparinization, reconstructed usingeither a tube or bifurcated Dacron graft de-pending on the extent of the aneurysm. Aorticclamping lasted 30–60 mins. The procedurewas always performed by the same surgeon. Allpatients undergoing AAA repair and AS re-ceived postoperative analgesia with morphineto prevent reduced ventilation consequent tolaparotomy.

Patients undergoing infrarenal AAA repairhad two muscle biopsies taken from the supe-rior third of femoral quadriceps of the rightleg at the three experimental times: a) beforeaortic clamping (baseline); b) immediately be-fore aortic declamping (ischemia); and c) 30mins after aortic declamping (reperfusion).Muscle biopsies were also taken from threepatients undergoing AS (control group) afterthe induction of anesthesia and at the end ofsurgery. Specimens for reverse transcriptase-polymerase chain reaction (RT-PCR) studieswere immediately frozen in liquid nitrogenand stored at �80°C until processing. Speci-mens for in situ hybridization and immuno-histochemistry were immediately fixed in 10%formalin solution.

At the end, and at timed intervals (6, 12,24, and 48 hrs) after the end of surgery, arte-rial and vein blood samples were obtained

from each patient in the two groups to deter-mine IL-6 plasma concentration and to mea-sure arterial partial pressure of oxygen andCO2 by an automatic blood gas analyzer.

Pulmonary Outcomes. The primary out-come was pulmonary function as assessed bythe PaO2/FIO2 (Horowitz index) and PaO2/PAO2

ratios at the end, and at timed intervals (6, 12,24, and 48 hrs) after the end of surgery, wherePAO2 � FIO2 � (760 � 47) � PaCO2/0.8.

The secondary outcome was a multiple-criteria outcome of pulmonary complicationsduring the first six postoperative days (24) ofa) new cough and sputum production; b) ab-normal breath sounds compared with base-line; c) temperature �38°C; d) chest radio-graph documentation of atelectasis or newinfiltrate; and e) physician documentation ofatelectasis or pneumonia. The requirementcriteria present �48 hrs was used to focus onclinically significant pulmonary complica-tions.

Immunoassay of IL-6 Plasma Concentra-tion. Vein samples were collected with citrate,immediately placed in ice, and soon thereaftercentrifuged at 2700 rpm for 15 mins (4°C).Plasma samples were then aliquoted andstored at �80°C until analysis. The IL-6plasma concentrations were measured by asandwich enzyme-linked immunoassay by us-ing a commercially available kit (QuantikineHS IL-6 Immunoassay kit; R&D Systems, Min-neapolis, MN). The minimum detectable doseof IL-6 ranged from 0.016 to 0.110 pg/mL.

Quantification of AGTN, ACE, AT1 Recep-tor Subtype, and IL-6 mRNAs. Quantificationof messenger RNA (mRNA) concentrations inmuscle specimens was performed with RT-PCR. Glyceraldehyde-3-phosphate dehydroge-nase (GAPDH) mRNA expression was consid-ered as internal standard. Total mRNA wasisolated from homogenized frozen samples us-ing Trizol reagent (Life Technologies, Milan,Italy), as instructed by the manufacturer. To-tal RNA was separated from proteins and DNAby centrifugation after addition of chloroformand then recovered from the aqueous phase byprecipitation with isopropanol. For use inPCR, total RNA was reverse-transcribed tocomplementary DNA (cDNA) using oligo dT aspreviously reported (25). PCR reactions forinvestigation of AGTN, ACE, AT1, IL-6, andGAPDH gene expression were performed usingfive samples each containing 5 �L of reversetranscription reaction and 45 �L of a PCRmaster mix as previously described (25–27).

The primer sequences and thermal profiles(Table 1) were selected with the software Oligoon the basis of GenBank cDNA human se-quences. Melting-curve analysis was used toconfirm the specificity of the amplificationproducts. The relationship between PCR cy-cles and amplification products was tested inpreliminary experiments as previously re-ported (26), and we chose the numbers of PCRcycles in the linear region.

To ensure that different amounts of PCRson muscle biopsies were not due to markedly

1171Crit Care Med 2004 Vol. 32, No. 5

different mRNA starting concentrations, PCRanalysis was performed for the internal con-trol mRNA (GAPDH) on serial two-fold dilu-tions of cDNA for each sample. The last dilu-tion giving a positive reaction for GAPDH wasused to equalize the amount of cDNA used ineach PCR. PCR reactions were performed in aDNA Thermal Cycler (Perkin Elmer Cetus),band densities were analyzed using a com-puter image densitometer (Qwin, Leica), andthe densitometric ratio to GAPDH was thencalculated as previously reported (25).

In Situ Hybridization and Immunohisto-chemistry. The in situ hybridization proce-dure for AGTN and ACE was performed aspreviously described (28, 29), using specificcDNA photobiotin-labeled (Vector) probes.The cDNA CE probe was kindly donated byProf. Soubrier, INSERM, Paris. IL-6 proteinwas assayed by immunohistochemical analysisaccording to the avidin-biotin peroxidasemethod using primary monoclonal antibody(RD System) as previously described (27).

Statistical Analysis. Data are expressed asmean � SD. Comparisons between groupswere performed using one-way analysis ofvariance, chi-square test, and Student’s t-testfollowed by the Tukey multiple-range compar-ison test, as appropriate. Univariate linear re-lations were analyzed with the Pearson corre-lation. The significance level for univariateand multivariate testing was set at .05. Allcalculations were performed using BMDP Sta-tistical Software.

RESULTS

Clinical Outcomes. Patients undergo-ing infrarenal AAA repair and patientsundergoing AS (gastrectomy � 4; intes-tinal resection � 6; nephrectomy � 2)were comparable in terms of demograph-ics and preoperative risk factors (bloodpressure and glucose concentrations)measured the day before surgery (Table2). Likewise, the mean intraoperativeblood loss and duration of anesthesia

were comparable in the two groups (Ta-ble 2). The mean time of muscle ischemia(aortic clamping) in patients undergoingAAA was 40 � 9 mins (range, 30–60mins) with no visceral ischemia in theother group. No patients died in the op-erating room, had to return to the oper-ating room within 6 days, or had a post-operative stay �48 hrs.

The assessment of pulmonary func-tion revealed that in patients undergoingAAA, the Horowitz index was significantlyreduced vs. baseline at 6, 12, and 24 hrs

after surgery (p � .05 vs. baseline at alltimes), whereas no significant changeswere observed in patients undergoing ab-dominal surgery (p � .01 betweengroups, Fig. 1A). Similarly, the PaO2/PAO2

ratio at 12 hrs and 24 hrs after surgerywas significantly lower in patients whounderwent AAA than in patients who un-derwent abdominal surgery (p � .01 be-tween groups, Fig. 1B).

Pulmonary complications occurred intwo patients of the AAA group (abnormalbreath sounds compared with baseline

Table 2. Clinical characteristics, operative details, and postoperative outcome of patients undergoingabdominal surgery and abdominal aortic aneurysm repair

AbdominalSurgery

AbdominalAortic

Aneurysm p

Baseline characteristicsPatients, n 12 12 NSMen/women, n 8/4 10/2 NSAge, yrs 74 � 6 70 � 5 NSWeight, kg 76.5 � 8.6 77.5 � 7.2 NSSmoking history, n 6 8 NSGlycemia, mg/dL 95 � 8 98 � 5 NSBlood pressure, mm Hg

Systolic 135 � 7 138 � 7 NSDiastolic 86 � 8 87 � 5 NS

Hypertension, n 5 7 NSOperative details

Intraoperative blood losses, mL 250 � 80 300 � 75 NSFluids administered,

mL�kg�1�hr�110.6 � 2.2 11.5 � 2.6 NS

Colloids/crystalloids ratio 0.24 � 0.02 0.23 � 0.04 NSCross-clamping time, mins — 40 � 9 —Duration of surgery, mins 190 � 85 240 � 60 NSTidal volume, mL 666 � 46 654 � 61 NSMechanical ventilation in ICU, hrs 7.3 � 8.3 5.4 � 7.5 NSLength of stay in ICU, hrs 34.8 � 14.3 34.6 � 21.6 NS

Postoperative outcomesDeath, n 0 0 NSPulmonary complications, n 1 2 NS

ICU, intensive care unit.Data are expressed as mean � SD.

Table 1. Sequences and complementary DNA (cDNA) sizes of primers used for investigation of angiotensinogen (AGTN), angiotensin-converting enzyme(ACE), angiotensin type 1 receptor subtype (AT1), and interleukin (IL)-6 gene expression and their thermal profiles

Primers Sequence 5-3cDNA Size,

bpCycles

No.Denaturation

°C/min

Thermal Profiles AnnealingExtension

°C/min°C/min °C/min °C/min

GAPDH 5 TGAAGGTCGGAGTCAACGGA 987 35 94/5 94/1 58/1 72/1 72/73 CATGTGGGCCATGAGGTCCA

AGTN 5 CTGCAAGGATCTTATGACCT 238 40 95/4 95/1 52/1 72/1 72/73 TACACAGCAAACAGGAATGG

ACE 5 CAACATCACCACAGAGACCAGCAAGATTC 468 38 94/1 94/1 61/1 72/1 72/73 GGTGTCTCGTACATAGACCTCCACGAGTC

AT1 5 GATGATTGTCCCAAAGCTGG 255 37 94/7 94/1 51/1 72/1 72/73 TAGGTAATTGCCAAAGGGCC

IL-6 5 ATGAACTCCTTCTCCACAAGCGC 628 34 94/6 94/1 65/1 72/1 72/73 GAAGAGCCCTCAGGCTGGACTG

GAPDH, glyceraldehyde-3-phosphate dehydrogenase.


and chest radiograph documentation ofatelectasis) and in one patient of the con-trol group (new cough and sputum pro-duction; not significant between groups).

Plasma IL-6 Concentration. Patientsof the AS group had no changes in post-operative plasma IL-6 concentrations. Incontrast, a significant increase in plasmaIL-6 concentration was observed in pa-tients with AAA at 6 hrs (p � .04 vs. AS)and 12 hrs (p � .04 vs. AS) after surgery(Fig. 2). Circulating IL-6 peaked at 12 hrsand then decreased significantly ap-proaching baseline values by 24 hrs (Fig.2).

Expression Levels of AT1*, ACE,AGTN, and IL-6 mRNAs. RT-PCR showedthat the expression levels of mRNAs werenegligible or undetectable in biopsy spec-

imens taken before aortic clamping (Fig.3). A marked overexpression of the genesfor all the components of the local RASsystem (AT1*, ACE, and AGTN) and IL-6was detectable at the end of ischemia. Inparticular, muscle ischemia induced a 22� 8-fold increase in IL-6/GAPDH mRNAratio, with contemporary overactivationof AGTN (80 � 18-fold), ACE (3 � 1-fold),and AT1 genes (24 � 6-fold, Fig. 3). Geneoverexpression persisted at reperfusion(Fig. 3).

Localization of ACE, AGTN mRNAs,and IL-6 Protein. Negative and positivecontrols for in situ hybridization showedthat the signal was specific for mRNA andthat mRNA in the biopsies was intact. Atbaseline, mRNA for AGTN was undetect-able whereas ACE mRNA was weakly ex-

pressed in endothelial cells (Fig. 4). Atthe end of ischemia, AGTN and ACEgenes were markedly overexpressed onendothelial cells and more weakly on in-terstitial cells. Overexpression of allgenes investigated was still evident in thesame cell types during reperfusion. Nohybridization signal was detectable inmuscle fibers (Fig. 4).

We did not find immunoreactivity forIL-6 in muscle specimens taken beforeaortic clamping (Fig. 5). Conversely, atthe end of ischemia, immunoreactivityfor IL-6 was highly expressed and pre-dominantly localized in microvascularendothelial cells (Fig. 5) but not in mus-cle fibers. Markedly positive immunore-activity for IL-6 was detectable in endo-thelial cells at reperfusion (Fig. 5).

DISCUSSION

The present study shows that a) dur-ing aortic surgery, an early inflammatoryresponse (with overexpression of IL-6,AGTN, ACE, and AT1 receptor subtype) isactivated by ischemia in lower limb skel-etal muscles; b) gene activation persistsduring reperfusion and is followed by anincrease of IL-6 plasma concentration;and finally c) the transient systemic in-flammatory reaction is followed by im-paired pulmonary function (PaO2/FIO2

�300). These alterations were absent inpatients who underwent abdominal sur-gery with no aortic clamping.

Gene Activation Occurs During Isch-emia. The expression of genes investi-gated was already activated in endothelialcells of ischemic muscles at the end ofaortic clamping. The absence of gene ac-tivation in the three patients with abdom-inal surgery indicates that the activationof RAS and IL-6 is essentially due to localfactors that are not related to the sys-temic hemodynamic changes commonlyassociated with anesthesia and surgery.In situ hybridization studies showed thatACE and AGTN mRNAs were overex-pressed almost exclusively on the endo-thelial cells of microvessels and on theinterstitial cells, whereas mRNA was notdetectable in myocytes. Although thelower sensitivity of in situ hybridizationin comparison with quantitative analysisdoes not rule out the possibility thatAGTN and ACE genes may be increasedalso in myocytes, our present findingsindicate the main involvement of mi-crovessel endothelial cells. In the samecondition, these cells were previously

Figure 1. PaO2/FIO2 and PaO2/PAO2 in patients who underwent abdominal surgery (AS, n � 12) orinfrarenal aortic aneurysm repair (AAA, n � 12). Each point of different experimental times (0, 6, 12,24, and 48 hrs) represents the mean � SD. Values measured at different experimental times arecompared with both baseline values measured after surgery and values measured at the same time inpatients of the other group. A, in patients undergoing AAA, the Horowitz index was significantlyreduced vs. baseline at 6, 12, and 24 hrs after surgery, whereas no significant changes were observedin patients undergoing abdominal surgery (p � .01 between groups at analysis of variance). B, thePaO2/PAO2 ratio at 12 and 24 hrs after surgery was significantly lower in patients who underwent AAAthan in patients who underwent abdominal surgery. *p � .05 vs. AS.

Figure 2. Plasma interleukin (IL)-6 concentrations in patients who underwent abdominal surgery (AS,n � 12) or infrarenal aortic aneurysm repair (AAA, n � 12). Each point of different experimental times(0, 12, 24, and 48 hrs) represents the mean � SD. Values measured at different experimental times arecompared with both baseline values measured after surgery and with values measured at the same timein patients of the other group. *p � .05 vs. AS.


found to overexpress adhesive molecules(15).

Besides the enhanced mRNA for AGTNand ACE, we found a selective increase inAT1 mRNA concentrations. A selective in-crease in the number and gene expres-sion of AT1 receptors was found in hu-man heart after acute myocardialinfarction (30), and an enhanced myocar-dial expression of AT1 was reported inrats following experimental ischemia/reperfusion (20). Although Ang II hasbeen reported to be a potent negativeregulator of AT1 receptor gene in isolatedrat myocytes (31), Ang II overexpressionin pressure overloaded rat heart wasfound to be directly related with AT1 pro-moter activity (32). Thus, Ang II couldprovide a positive feedback loop for po-tentiation of its effects.

Which Role for Local RAS Activation?At least three lines of evidence support arole of local activation of the RAS systemin the ischemia-reperfusion syndrome oflower limb. First, Ang II was reported tosupport leukocyte transendothelial mi-gration via an AT1 dependent mechanism(33), and local recruitment and migra-tion of inflammatory cells are a key pro-cess of ischemia-reperfusion damage (15,16). Second, the local generation of AngII might increase local vasoconstrictionthus contributing to the “no-reflow phe-nomenon.” Third, Ang II might also par-ticipate in cytokine overactivation vianicotinamide adenine dinucleotide phos-phate-oxidase-dependent free radical

Figure 3. Expression of glyceraldehyde-3-phosphate dehydrogenase (GAPDH), interleukin (IL)-6, angiotensin-converting enzyme (ACE), angiotensin type1 receptor subtype (AT1), and angiotensinogen (AGTN) genes, in skeletal muscle before aortic clamping (Bas.), after 30 mins of ischemia (Isch.), and after2 hrs of reperfusion (Reper.). Representative reverse transcriptase-polymerase chain reaction of three different experiments. Bp, base pairs.

Figure 4. Localization of genes for glyceraldehyde-3-phosphate dehydrogenase (GAPDH), angiotensin-converting enzyme (ACE), and angiotensinogen (AGTN, red signal), at baseline (before aortic clamp-ing), after 30 mins of ischemia, and after 2 hrs of reperfusion.


overproduction (17). However, a majorrole for Ang II in IL-6 overexpression, asobserved in isolated cells (17) and exper-imental studies (18), can hardly be sus-tained in this clinical condition whereeven the reported anti-inflammatory ef-fects of heparin (34–36) were overriddenand different mediators could be involved(37).

A single study recently revealed favor-able effects of ACE inhibitors in the re-duction of ischemia-reperfusion injuryafter lung transplantation (19). However,more definite insights into the effect ofAng II antagonism on muscle gene acti-vation and SIR after aortic surgery will beprovided only by specifically designedprospective studies.

Pulmonary Effects of Systemic In-flammatory Response. An increase inplasma IL-6 concentration, a marker ofsystemic inflammatory response, wasconstantly shown after aortic surgery (9).IL-6 is a multifunctional proinflamma-tory cytokine that mediates both immuneand inflammatory responses (38). IL-6 isproduced not only by endothelial cells(33) but also by different cell types in-cluding leukocytes, vascular smoothmuscle cells, fibroblasts, and respiratoryepithelial cells (38). In thoracoabdominalaortic surgery, visceral ischemia was con-sidered the major source of cytokine pro-duction (10). However, the present studyindicates that IL-6 also can be generatedin the ischemic muscle. The role of mus-cle ischemia in IL-6 production is indi-rectly confirmed by the absence ofchanges of cytokine plasma concentra-tions in patients undergoing abdominalsurgery with no visceral ischemia.

IL-6 produced in ischemic musclemay be responsible for the occurrence ofcomplications in remote districts. IL-6has the ability to increase inducible nitric

oxide synthase along with the release ofnitric oxide, which leads to vasodilationmediated through cyclic guanosinemonophosphate in the vascular bed (39)and may modulate the pulmonary vascu-lar tone. IL-6 is able to modulate epithe-lial permeability in the airways (40). Inhumans, plasma IL-6 concentrations af-ter aortic surgery were recently reportedto correlate with increased protein per-meability in lung (8). Aortic aneurysmrepair was recently reported to be fol-lowed by a reduced pulmonary elastance,which was independent from the effectsof surgical approach on the respiratorymechanics (41). In the present study,AAA-operated patients had a significantreduction in pulmonary function withimpaired oxygen diffusion (PaO2/FIO2

�300), different from patients undergo-ing AS with no tissue ischemia-reperfu-sion (42). Both groups had the sametransperitoneal approach and both re-ceived postoperative analgesia, but a re-duced pulmonary function was detectableonly in patients with lower limb ischemiareperfusion and increased IL-6 plasmaconcentrations, thus revealing the pri-mary role of aortic clamping and SIR inaltered ventilation consequent to surgicalapproach. We did not attempt to explorethe relationship between IL-6 and oxy-genation because the peak of plasma IL-6increase (12 hrs) preceded the peak ofoxygen diffusion index alteration (24hrs). This pattern suggests two possibili-ties (that other cytokines might play arole or that the pulmonary injury mayrequire a time delay to be assessed bychanges in PaO2/FIO2 ratio) which cannotbe discriminated by searching for a cor-relation at selected times. In any case,time-related changes of PaO2/FIO2 mightrepresent a useful surrogate end point toinvestigate the efficacy of new anti-

inflammatory strategies in patients un-dergoing aortic aneurysm repair.

In our study, elective infrarenal aorticsurgery was not associated with mortalityor increased incidence of pulmonarycomplications. However, pulmonary con-sequences of infrarenal aortic surgerymight be harmful when other risk factorsfor postoperative complication coexist(advanced age or chronic obstructive pul-monary disease), or when surgery is per-formed in an emergency (43).

Limitations of the Study. As with anyclinical study, this too has limitations.First, we did not measure the transla-tional products of assayed genes becauseof the small size of biopsies. No assay ofAng II in plasma was performed becauseit might have not revealed the activationof the local RAS system (25, 29). Second,we did not directly investigate the role ofAng II by the administration of Ang IIantagonists (ACE inhibitors or AT1 recep-tor antagonists). However, the main goalof the present study was to investigate thesource of cytokine production and theactivation of local RAS in the ischemic-reperfused muscle. The role of Ang II maybe investigated only in a specially de-signed, prospective, randomized clinicalstudy. Third, it cannot be excluded thatother cytokines may be involved in isch-emia-reperfusion syndrome and espe-cially in lung injury (44). Likewise, localproduction in the lung of proinflamma-tory cytokines (IL-6 and others) was notspecifically examined. Although it is rel-

Figure 5. Immunoreactivity for interleukin-6 (red signal) in muscle specimens taken before aorticclamping (A), at the end of ischemia (B), and at reperfusion (C).

T his study clearly

demonstrates that

systemic inflam-

matory response, with in-

creased interleukin-6 plasma

concentrations, which fol-

lows infrarenal aortic sur-

gery is preceded by activa-

tion of the renin-angiotensin

system and cytokine produc-

tion in ischemic-reperfused

lower limbs.


atively easy to sample bronchoalveolar la-vage fluid before and after surgery, we didnot perform such an investigation forethical reasons because patients wereusually extubated early after the proce-dure. In addition, the type and amount ofanesthetic given intraoperatively may af-fect BAL cytokine concentrations, thusproducing conflicting results (45).

CONCLUSIONS

This study clearly demonstrates thatSIR, with increased IL-6 plasma concen-trations, which follows infrarenal aorticsurgery is preceded by activation of theRAS system and cytokine production inischemic-reperfused lower limbs. Al-though the direct role of IL-6 on thepathophysiology of respiratory dysfunc-tion is not well understood, the transientSIR is followed by impaired pulmonarygas exchange, which appears to be inde-pendent of postoperative surgical ap-proach-related changes in ventilation.Future studies are needed to investigatethe role of the activation of the local RASsystem in determining SIR and impairedpulmonary function.

REFERENCES

1. Golden MA, Donaldson MC, Whittemore AD,et al: Evolving experience with thoracoab-dominal aortic aneurysm repair at a singleinstitution. J Vasc Surg 1991; 13:792–796

2. Huber TS, Harward TR, Flynn TC, et al: Op-erative mortality rates after elective infrare-nal aortic reconstructions. J Vasc Surg 1995;22:287–293

3. Money SR, Rice K, Crockett D, et al: Risk ofrespiratory failure after repair of thoracoab-dominal aortic aneurysms. Am J Surg 1994;168:152–155

4. Heimer RG, Gross J, Schein M: Pro- andanti-inflammatory cytokine-response in ab-dominal aortic aneurysm repair: a clinicalmodel of ischemia-reperfusion. Shock 1999;11:305–310

5. Rocker GM: Ischaemia/reperfusion, inflam-matory responses and acute lung injury.Thorax 1997; 52:841–842

6. Syk I, Brunkwall J, Ivancev K, et al: Postop-erative fever, bowel ischaemia and cytokineresponse to abdominal aortic aneurysm re-pair—a comparison between endovascularand open surgery. Eur J Vasc Endovasc Surg1998; 15:398–405

7. Seekamp A, Warren JS, Remick DG, et al:Requirements for tumor necrosis factor-alpha and interleukin-1 in limb ischemia/reperfusion injury and associated lung in-jury. Am J Pathol 1993; 143:453–463

8. Raijmakers PG, Groeneveld AB, RauwerdaJA, et al: Acute lung injury after aortic sur-

gery: The relation between lung and leg mi-crovascular permeability to 111indium-labelled transferrin and circulatingmediators. Thorax 1997; 52:866–871

9. Swartbol P, Truedsson L, Norgren L: Theinflammatory response and its consequencefor the clinical outcome following aortic an-eurysm repair. Eur J Vasc Endovasc Surg2001; 21:393–400

10. Welborn MB, Oldenburg HS, Hess PJ, et al:The relationship between visceral ischemia,proinflammatory cytokines, and organ injuryin patients undergoing thoracoabdominalaortic aneurysm repair. Crit Care Med 2000;28:3191–3197

11. Fantini GA, Conte MS: Pulmonary failurefollowing lower torso ischemia: clinical evi-dence for a remote effect of reperfusion in-jury. Am Surg 1995; 61:316–319

12. Harward TR, Welborn MB III, Martin TD etal: Visceral ischemia and organ dysfunctionafter thoracoabdominal aortic aneurysm re-pair. A clinical and cost analysis. Ann Surg1996; 223:729–734

13. Corson RJ, Paterson IS, O’Dwyer ST, et al:Lower limb ischaemia and reperfusion altersgut permeability. Eur J Vasc Surg 1992;6:158–163

14. Yassin MM, Barros D’Sa AA, Parks TG, et al:Lower limb ischaemia-reperfusion injury al-ters gastrointestinal structure and function.Br J Surg 1997; 84:1425–1429

15. Formigli L, Manneschi LI, Adembri C, et al:Expression of E-selectin in ischemic andreperfused human skeletal muscle. Ultra-struct Pathol 1995; 19:193–200

16. Novelli GP, Adembri C, Gandini E, et al:Vitamin E protects human skeletal musclefrom damage during surgical ischemia-reperfusion. Am J Surg 1997; 173:206–209

17. Schieffer B, Luchtefeld M, Braun S, et al:Role of NAD(P)H oxidase in angiotensin II-induced JAK/STAT signaling and cytokine in-duction. Circ Res 2000; 87:1195–1201

18. Ruiz-Ortega M, Ruperez M, Lorenzo O, et al:Angiotensin II regulates the synthesis ofproinflammatory cytokines and chemokinesin the kidney. Kidney Int 2002; 82:12–22

19. Fischer S, Maclean AA, Liu M, et al: Inhibi-tion of angiotensin-converting enzyme bycaptopril: A novel approach to reduce isch-emia-reperfusion injury after lung transplan-tation. J Thorac Cardiovasc Surg 2000; 120:573–580

20. Xu Y, Kumar D, Dyck JR, et al: AT(1) andAT(2) receptor expression and blockade afteracute ischemia-reperfusion in isolated work-ing rat hearts. Am J Physiol Heart CircPhysiol 2002; 282:H1206–H1215

21. Karakurum M, Shreeniwas R, Chen J, et al:Hypoxic induction of interleukin-8 gene ex-pression in human endothelial cells. J ClinInvest 1994; 93:1564–1570

22. Yassin MM, Harkin DW, Barros D’Sa AA, etal: Lower limb ischemia-reperfusion injurytriggers a systemic inflammatory responseand multiple organ dysfunction. WorldJ Surg 2002; 26:115–121

23. World Medical Association Declaration ofHelsinki: Ethical principles for medical re-search involving human subjects. Bull WorldHealth Organ 2001; 79:373–374

24. Seymour DG, Pringle R: Post-operative com-plications in the elderly surgical patient.Gerontology 1983; 29:262–270

25. Neri Serneri GG, Boddi M, Cecioni I, et al:Cardiac angiotensin II formation in the clin-ical course of heart failure and its relation-ship with left ventricular function. Circ Res2001; 88:961–968

26. Vanni S, Polidori G, Cecioni I, et al: ET(B)receptor in renal medulla is enhanced bylocal sodium during low salt intake. Hyper-tension 2002; 40:179–185

27. Neri Serneri GG, Boddi M, Modesti PA, et al:Immunomediated and ischemia-independentinflammation of coronary microvessels in un-stable angina. Circ Res 2003; 92:1359–1366

28. Modesti PA, Cecioni I, Migliorini A, et al:Increased renal endothelin formation is as-sociated with sodium retention and in-creased free water clearance. Am J Physiol1998; 275:H1070–H1077

29. Neri Serneri GG, Modesti PA, Boddi M, et al:Cardiac growth factors in human hypertro-phy. Relations with myocardial contractilityand wall stress. Circ Res 1999; 85:57–67

30. Tsutsumi Y, Matsubara H, Ohkubo N, et al:Angiotensin II type 2 receptor is upregulatedin human heart with interstitial fibrosis, andcardiac fibroblasts are the major cell type forits expression. Circ Res 1998; 10:1035–1046

31. Everett AD, Heller F, Fischer A: AT1 receptorgene regulation in cardiac myocytes and fi-broblasts. J Mol Cell Cardiol 1996; 28:1727–1736

32. Herzig TC, Jobe SM, Aoki H, et al: Angioten-sin II type1a receptor gene expression in theheart: AP-1 and GATA-4 participate in theresponse to pressure overload. Proc NatlAcad Sci U S A 1997; 94:7543–7545

33. Pastore L, Tessitore A, Martinotti S, et al:Angiotensin II stimulates intercellular adhe-sion molecule-1 (ICAM-1) expression by hu-man vascular endothelial cells and increasessoluble ICAM-1 release in vivo. Circulation1999; 100:1646–1652

34. Benador NM, Girardin EP: Influence of hep-arin and type-IV collagen on IL-6 synthesisby rat glomerular mesangial cells. Nephron1997; 77:219–224

35. Attanasio M, Gori AM, Giusti B, et al: Cyto-kine gene expression in human LPS- andIFNgamma-stimulated mononuclear cells isinhibited by heparin. Thromb Haemost 1998;79:959–962

36. Pepe G, Giusti B, Attanasio M, et al: Tissuefactor and plasminogen activator inhibitortype 2 expression in human stimulatedmonocytes is inhibited by heparin. SeminThromb Hemost 1997; 23:135–141

37. Yan SF, Tritto I, Pinsky D, et al: Induction ofinterleukin 6 (IL-6) by hypoxia in vascularcells. Central role of the binding site fornuclear factor-IL-6. J Biol Chem 1995; 270:11463–11471


38. Akira S, Taga T, Kishimoto T: Interleukin-6in biology and medicine. Adv Immunol 1993;54:1–78

39. Kanno K, Hirata Y, Imai T, et al: Induction ofnitric oxide synthase gene by interleukin invascular smooth muscle cells. Hypertension1993; 22:34–39

40. Persson CG: Plasma exudation in the air-ways: Mechanisms and function. Eur Respir J1991; 4:1268–1274

41. Volta CA, Ferri E, Marangoni E, et al: Respi-ratory function after aortic aneurysm repair:

A comparison between retroperitoneal andtransperitoneal approaches. Intensive CareMed 2003; 29:1258–1264

42. Abraham E, Matthay MA, Dinarello CA, et al:Consensus conference definitions for sepsis,septic shock, acute lung injury, and acuterespiratory distress syndrome: Time for a re-evaluation. Crit Care Med 2000; 28:232–235

43. Chang JK, Calligaro KD, Lombardi JP, et al:Factors that predict prolonged length of stayafter aortic surgery. J Vasc Surg 2003; 38:335–339

44. Yassin MM, Barros D’Sa AA, Parks TG, et al:Lower limb ischaemia-reperfusion injurycauses endotoxaemia and endogenous anti-endotoxin antibody consumption but notbacterial translocation. Br J Surg 1998; 85:785–789

45. Kotani N, Hashimoto H, Sessler DI, et al:Supplemental intraoperative oxygen aug-ments antimicrobial and proinflammatoryresponses of alveolar macrophages. Anesthe-siology 2000; 93:15–25

ACCM Guidelines on SCCM Website

The Guidelines and Practice Parameters developed by the American College of Critical CareMedicine are now available online at http://www.sccm.org/professional_resources/guidelines/index.asp. The printed version of the Guidelines, provided in a binder, is alsoavailable through the SCCM Bookstore, located at http://www.sccm.org/pubs/sccmbookstore.html. Please watch the Website to stay updated on the ACCM Guidelines and PracticeParameters.


Documents

Pulmonary injury follows systemic inflammatory reaction in infrarenal aortic surgery*