Real-time monitoring of endogenous lipid peroxidation by exhaled ethylene inpatients undergoing cardiac surgery

Simona M. Cristescu,1* Rudolf Kiss,2,3* Sacco te Lintel Hekkert,4 Miles Dalby,5 Frans J. M. Harren,1

Terence H. Risby,6 and Nandor Marczin,2,3,7 on behalf of the Harefield BIOSTRESS study investigators1Department of Molecular and Laser Physics, Institute of Molecules and Materials, Radboud University, Nijmegen, theNetherlands; 2Department of Anaesthesia, Royal Brompton and Harefield NHS Foundation Trust, Harefield, UK; 3Section ofAnaesthesia, Pain Medicine and Intensive Care, Department of Surgery and Cancer, Faculty of Medicine, Imperial CollegeLondon, London, UK; 4Sensor Sense B.V., Nijmegen, The Netherlands; 5Department of Cardiology, Royal Brompton andHarefield NHS Foundation Trust, Harefield, UK; 6Department of Environmental Health Sciences, Bloomberg School of PublicHealth, The Johns Hopkins University, Baltimore, Maryland; and 7Centre of Anaesthesia and Intensive Care, SemmelweisUniversity, Budapest, Hungary

Submitted 30 June 2014; accepted in final form 10 August 2014

Cristescu SM, Kiss R, te Lintel Hekkert S, Dalby M, HarrenFJ, Risby TH, Marczin N; Harefield BIOSTRESS study investi-gators. Real-time monitoring of endogenous lipid peroxidation byexhaled ethylene in patients undergoing cardiac surgery. Am J PhysiolLung Cell Mol Physiol 307: L509–L515, 2014. First published Au-gust 15, 2014; doi:10.1152/ajplung.00168.2014.—Pulmonary andsystemic organ injury produced by oxidative stress including lipidperoxidation is a fundamental tenet of ischemia-reperfusion injury,inflammatory response to cardiac surgery, and cardiopulmonary by-pass (CPB) but is not routinely measured in a surgically relevant timeframe. To initiate a paradigm shift toward noninvasive and real-timemonitoring of endogenous lipid peroxidation, we have explored pul-monary excretion and dynamism of exhaled breath ethylene duringcardiac surgery to test the hypothesis that surgical technique andischemia-reperfusion triggers lipid peroxidation. We have employedlaser photoacoustic spectroscopy to measure real-time trace concen-trations of ethylene from the patient breath and from the CPBmachine. Patients undergoing aortic or mitral valve surgery-requiringCPB (n ! 15) or off-pump coronary artery bypass surgery (OPCAB)(n ! 7) were studied. Skin and tissue incision by diathermy causedstriking ("30-fold) increases in exhaled ethylene resulting in elevatedlevels until CPB. Gaseous ethylene in the CPB circuit was raised uponthe establishment of CPB ("10-fold) and decreased over time. Rep-erfusion of myocardium and lungs did not appear to enhance ethylenelevels significantly. During OPCAB surgery, we have observed in-creased ethylene in 16 of 30 documented reperfusion events associ-ated with coronary and aortic anastomoses. Therefore, novel real-timemonitoring of endogenous lipid peroxidation in the intraoperativesetting provides unparalleled detail of endogenous and surgery-trig-gered production of ethylene. Diathermy and unprotected regionalmyocardial ischemia and reperfusion are the most significant contrib-utors to increased ethylene.

cardiopulmonary bypass; oxidative stress; ischemia-reperfusion in-jury; off-pump coronary artery bypass grafting

CONTEMPORARY CARDIAC SURGERY provides symptom relief fromheart disease and improves patient prognosis. However, theoperation remains associated with significant myocardial, pul-monary, and systemic adverse outcomes (18, 25). Aortic cross-clamping and extracorporeal circulation and associated oxida-

tive stress and pulmonary and systemic inflammatory responseare considered major contributors.

Although the role of oxidative stress is clearly defined inanimal models of ischemia-reperfusion injury and extracorpo-real circulation, the nature, degree, and significance of oxida-tive stress in the setting of human cardiac surgery remain morecontroversial (3, 6, 10, 15). Liquid-phase biomarkers of oxi-dative stress have provided inconclusive evidence (1, 6, 7, 16,20). Parallel to these biomarker studies, pulmonary excretionand analysis of volatile organic compounds in exhaled breathhave also been considered. Interestingly, exhaled breath hy-drocarbons have emerged as gaseous footprints of in vivo lipidperoxidation, but studies have been limited to discrete sam-pling and offline analysis in a specialized laboratory settingwith results unavailable in a surgically relevant time frame (2,19, 23, 26).

There is extensive biochemical evidence linking productionof ethylene to lipid peroxidation. Lieberman and Hochstein(13) found that peroxidation of lipids was coupled to thegeneration of ethylene in rat liver microsomes in the presenceof cuprous ions, providing an original model for the biogenesisof ethylene in animal cells. These milestone observations wereconfirmed in different isolated preparations (9, 4) and extendedto animal models of endogenous or induced lipid peroxidation(12, 24, 27). Treating animals with established lipid peroxida-tion-inducing agents such as CCl4, #-aminolevulinic acid, orNi(II) caused a significant time-dependent and long-lastingproduction of ethylene, which was measureable in the breath ofmice and rats. Sagai and Ichinose (24) investigated age-relatedlipid peroxidation and found breath hydrocarbons as sensitiveindexes of peroxidation associated with aging with substantialincrease in exhaled ethylene in aged compared with younganimals. To explore the human relevance of these models, wehave conducted unique preclinical studies and demonstratedincreased ethylene release associated with lipid peroxidation inhealthy volunteers during UV irradiation (17, 11). In theseexperiments, 21 male volunteers were exposed to UV lightfrom a commercial solarium for 15 min. Individual data sug-gested that ethylene rises within 2 min of exposure, withethylene exhalation increasing to 17.2 $ 7.3 pmol·kg%1·min%1

from pre-UV exposure levels of 1.39 $ 0.38 pmol·kg%1·min%1. These responses were eliminated by shielding thesubjects from solarium by a UV protective Lexan barrier. On the

* S. M. Cristerscu and R. Kiss are joint first authors. T. H. Risby and N.Marczin are joint senior authors.

Address for reprint requests and other correspondence: N. Marczin, Dept. ofAnaesthetics, Harefield Hopsital, Hill End Rd., Harefield, Middlesex, UB96JH,UK (e-mail: n.marczin@imperial.ac.uk).

Am J Physiol Lung Cell Mol Physiol 307: L509–L515, 2014.First published August 15, 2014; doi:10.1152/ajplung.00168.2014. Rapid Reports

1040-0605/14 Copyright © 2014 the American Physiological Societyhttp://www.ajplung.org L509

basis of such substantial evidence, we have proposed utilizingreal-time laser photoacoustic detection of ethylene (8) to monitorin vivo oxidative stress associated with cardiac surgery.

To investigate the feasibility of continuous assessment of lipidperoxidation by expired ethylene in the clinical setting of cardiacsurgery, we have studied patients undergoing the most commonsurgical valvular and coronary artery bypass grafting procedures.There is tremendous scientific debate regarding the impact ofcardiopulmonary bypass, myocardial and pulmonary ischemia,and reperfusion on myocardial, pulmonary, and systemic lipidperoxidation. Furthermore, there are contradictory studies regard-ing the effect of electrocautery and sternotomy on gaseous foot-prints of lipid peroxidation. In our study, we addressed theseissues by conducting two related studies: one focusing on cardio-pulmonary bypass and the other on beating heart surgery.

METHODS

Patients

For the first series, 15 patients requiring aortic and mitral valverepair/replacement utilizing extracorporeal circulation were studied toevaluate the impact of 1) diathermy and sternotomy, 2) cardiopulmo-nary bypass (CPB), 3) cardioplegia-protected global myocardial isch-emia-reperfusion injury associated with aortic cross-clamping, 4)separation from CPB and pulmonary reperfusion, and 5) post-CPBperiod. In the second series, the influence of regional and unprotectedmyocardial ischemia-reperfusion was investigated in seven patientsundergoing beating heart off-pump coronary artery bypass grafting(OPCAB). The Regional Ethics Committee approved this study aspart of the Harefield BIOSTRESS program. All patients providedwritten, informed consent.

Specific anesthetic considerations included total intravenous anes-thesia (before, during, and after CPB), application of stable ventilationparameters throughout the study, and preparation of the anestheticmachines to avoid trace contamination with inhalational agents, whichcaused spectral interference with ethylene detection. Surgical conductwas according to routine practice. Standard CPB management in-cluded membrane oxygenators, arterial line filters, systemic hypother-mia down to 32°C, and nonpulsatile flow of 2.4 l·min%1·m%2, with amean arterial pressure greater than 50 mmHg. The myocardium wasprotected by use of intermittent antegrade cold blood cardioplegia (4:1blood to crystalloid ratio). Combined anterograde and retrogradecardioplegia was employed selectively, according to clinical need.

Measurement of Exhaled Ethylene

The ethylene detector. We used the ETD-300 ethylene detector (bySensor Sense, Nijmegen, The Netherlands), which is based on laserphotoacoustic spectroscopy principle and uses a CO2 laser emittinglight in the 10-&m region, where ethylene presents the strongestabsorption. Currently, ETD-300 is able of measuring ethylene onlinein the 300 pptv (pptv ! parts per trillion by volume, 1:1012) rangewithin a 5-s time scale offering wide dynamic range, operationalsimplicity, and relative portability.

The gas sampling system. The gas sampling system ensured unin-terrupted gas flow from the patient or CPB machine to the ethylenedetector. A small portion of the exhaled gas flow was continuouslysampled at a constant flow rate of 4 l/h with a mass flow controller(Brooks Instrument, Ede, The Netherlands) from the anesthesia/CPBmachine by using a diaphragm pump (Knf, type NMP830KTDC,Germany) through a Teflon PTFA sampling line of 1/16-in. diameterand '4 m long (Polyfluor Plastics, Oosterhout, The Netherlands). Thegas flow was passed through a CO2 scrubber containing 42 g of sodalime and a water trap with 40 g CaCl2 after the pump and priorentering the ethylene detector.

Real-time ethylene measurement. While the patient was anesthe-tized in the anesthetic induction room, the sampling line was con-nected to the expiratory limb of ventilation circuit in the operatingroom to monitor baseline ethylene levels present in the ventilationcircuit at a ventilation setting mimicking the predicted ventilationstrategy for the given patient. The sampling line was positioned distalto the bacterial filter in the expiratory limb of the ventilation circuitwith a clinically approved tube attachment.

Following transfer of the patient to the operating room, ventilationcommenced with this setting, allowing determination of the patientbaseline ethylene concentrations. Particular attention was given to com-mencement of surgery and the temporal relationship of diathermy appli-cation for skin incision and subsequent sternotomy. For the prebypassperiod, the sampling gas line remained connected to the ventilation circuitto monitor the real-time production of ethylene. Following heparinadministration, the sampling line was transferred to the exhaust of theoxygenator part of the cardiopulmonary bypass machine. Gaseous ethyl-ene was then continuously measured during the recirculating primingphase, upon commencement of cardiotomy suction and establishment offull cardiopulmonary bypass. Special consideration was given to mea-suring ethylene concentration at the time of the release of aortic cross-clamp and myocardial reperfusion. After rewarming and commencementof mechanical ventilation, the sampling line was returned to the ventila-tion circuit with an aim to compare exhaled ethylene just prior to and afterseparation of cardiopulmonary bypass. In addition, ethylene was con-stantly monitored in the postbypass period.

Data Analysis

The main advantage of the real-time analysis is that changes inmeasured concentrations can be directly linked to clinical events andmanagement changes. Thus we have made serious efforts to liaisewith clinical staff to document surgical events, anesthetic measuresand perfusion procedures and to identify their impact and relationshipwith dynamic ethylene changes. To demonstrate these relationships,we have presented a series of representative traces obtained at specificphases of surgery or in response to particular clinical events.

Because there are no similar real-time breath analysis studies on thefield, it is important to compare our experiments with previous studiesthat used offline analysis at discrete and predetermined sampling timepoints. To do this, we have averaged our real-time ethylene data for 3min at different time points during surgery. These data would beequivalent to ethylene concentrations in a breath sample collected inbags over a similar time period.

The distribution of quantitative data was checked by the Shapiro-Wilk or Kolmogorov-Smirnow test. Normally distributed data arepresented as means $ SD. Data exhibiting nonnormal distribution arepresented as a median and lower and upper quartile. Qualitative dataare presented as percentage of whole analyzed group. Repeated-measures analysis of variance or variance on ranks was used tocompare experimental groups with appropriate post hoc tests. Valueof P ( 0.05 was considered statistically significant. The analysis wasperformed with the SigmaPlot version 12.0 statistical software pack-age (Systat Software 2011–12).

RESULTS

Clinical Summary

Patient demographics and clinical outcomes of the twopatient populations are summarized in Table 1. Both groupsincluded mainly elderly patients. All patients survived in theOPCAB group whereas one patient died after valve surgery.Patients in both groups required a few hours of postoperativemechanical ventilation and stayed in the intensive care unit fora few days. Vasoactive and inotrope requirement was higher inthe valve patients and both groups exhibited systemic inflam-

Rapid Reports

L510 EXHALED ETHYLENE IN CARDIAC SURGERY

AJP-Lung Cell Mol Physiol • doi:10.1152/ajplung.00168.2014 • www.ajplung.org

matory response as evidenced by laboratory measures of whitecell blood counts and C-reactive protein levels.

Exhaled Ethylene During Valve Surgery and CPB

Figure 1A shows a representative example of continuousmonitoring of ethylene produced spanning the entire durationof cardiac surgery. This experiment clearly demonstrates dy-namic changes in expired ethylene concentrations at specificphases of surgery and the unique release of ethylene from theexhaust of CPB circuit. Figure 1B summarizes averaged breathethylene data of all valve surgery patients at various stages ofsurgery.

Sternotomy and prebypass period. Valvular surgery patientsexhibited 1.200 (0.7; 2.0) parts per billion by volume (ppbv)baseline levels of ethylene prior to commencement of surgery.Surgical skin and tissue incision using diathermy resulted instriking, robust, and persistent increases in ethylene [peakconcentrations of 40.8 (7.4; 85.2) (1:109), Figs. 1 and 2A]. Theconcentrations of breath ethylene remained higher comparedwith baseline levels for the remaining of the prebypass period[11.5 (4.1; 25.4) ppbv at the end of the pre-CPB periodrepresenting an area under curve (AUC) of 7.7 $ 1.6 ppbv )s for the entire pre-CPB episode].

Bypass period. Compared with the recirculating primingphase (in the absence of blood return), there was an immediaterise in ethylene upon return of the cardiotomy suction blood tothe CPB circuit following full heparinization. In addition, therewas a significant rise in ethylene upon achievement of full CPB[9.75 (7.0; 16.3) vs. 0.7 (0.5; 1) ppbv; Figs. 1 and 2B]. Averageethylene concentrations were significantly less at later timepoints during CPB such as prior to removal of aortic cross-clamp and at the end of rewarming period. We noted signifi-cant changes in measured ethylene release in response tovarious changes in perfusion management such as changes insweep gas flow (Fig. 2B) or application of hemofiltration (notshown).

We have also observed characteristic biphasic ethylene re-sponses (consisting of an initial decrease followed by return tobaseline) associated with the deairing protocol (filling the heartand commencement of ventilation) and release of aortic cross-

clamp (Fig. 2B). Compared with the immediate predeairingperiod (5 min average), ethylene concentrations were similarafter release of the aortic cross-clamp [4.4 (2.1; 6.55) vs. 4.4(2.9; 6.85) ppbv, respectively, Fig. 1B].

Postbypass period. The process of separation from CPB,rapidly changing hemodynamics, protamine administration,and occasional breath holding resulted in dynamically chang-ing expired ethylene levels during the initial postbypass period.Average concentrations of ethylene in the first 5 min after CPBwere not different from the preceding period just prior toseparation from CPB [3.2 (1.9; 5.1) ppbv vs. 3.85 (1.975; 6.50)ppbv]. Subsequently, ethylene levels remained largely un-changed apart from chest closure, where repeated applicationof diathermy for hemostasis was associated with higher breathethylene [7.4 (2.3; 8.2) ppbv, Fig. 1B].

Table 1. Patient demographics and outcomes

Patient Variable CPB OPCAB

Age, yr 74.8 $ 6.7 73.3 $ 6.6Sex, % male 66% 100%Body mass index 27.8 $ 3.7 27.0 $ 3.1Surgery time, min 286 $ 90 209 $ 25CPB time, min 82 (74;106)AXC time, min 62 (43;69)Clinical outcomesSurvival 14/15 7/7Duration of mechanical ventilation, h 9 (4;72) 5 (3;5)Intensive care unit stay, days 3 (2;10) 4 (2;4)Inotrope requirement, &g·kg%1·72 h%1

Epinephrine 0.0 (0.0;16.4) 0.0 (0.0;0.0)Norepinephrine 130 (0.0;599) 3.5 (0.0;16.2)Milrinone 242 (63;393) 0.0 (0.0;0.0)SIRS responsePeak white blood cell count, 109/l 15.2 $ 4.8 11.6 $ 3.6Peak CRP, mg/l 184 $ 112 179 $ 73

CPB, cardiopulmonary bypass; OPCAB, off-pump coronary artery bypassgrafting; SIRS, systemic inflammatory response; CRP, C-reactive protein.

Ethy

lene

[ppb

v]

0

20

40

60

80

100

120

140

160Pre CPB CPB Post CPB

*

*

* * * # #

0 60 120 180

0

10

20

30

40

50

Ethy

lene

(ppb

v)

Time (min)

A

B

Pre CPB CPB Post CPB

Fig. 1. Intraoperative real-time monitoring of ethylene measured in patientsundergoing cardiac surgery with cardiopulmonary bypass (CPB). Samplingline was positioned in the expiratory limb of the ventilator circuit prior to andat the end of CPB and following CPB and to the exhaust of the oxygenator ofCPB machine during CPB. Box plots represent 3–5 min ethylene averages.*P ( 0.05 compared with patient baseline for ventilated groups. #P ( 0.05among CPB groups.

Rapid Reports

L511EXHALED ETHYLENE IN CARDIAC SURGERY

AJP-Lung Cell Mol Physiol • doi:10.1152/ajplung.00168.2014 • www.ajplung.org

Exhaled Ethylene During OPCAB

Increases in the concentrations of ethylene as a result of skinincision and sternotomy during OPCAB surgery was similar tovalvular surgery (Fig. 3). Notably, there was more prolongedapplication of diathermy during the preparatory phase of sur-gery (prior to grafting) resulting in more prolonged increases inethylene (AUC of 14.6 $ 0.7 ppbv ) s) in this period. Wehave noted 22 reperfusion events associated with clamping andrelease of main coronary arteries during distal anastomoses. Inaddition we have documented eight reperfusion events follow-ing proximal aortic anastomoses. Fourteen of these eventsproduced no change in breath ethylene concentrations. Sixteenreperfusion events were associated with typical increases inethylene release; however, the magnitude of these increasesappeared smaller than the effect of diathermy. Four of theseincreases may have been confounded by concomitant surgicaland anesthetic factors such as breath holding or major manip-ulation of heart and hemodynamic changes (Fig. 3). Summarydata are presented in Fig. 4 showing the effect of diathermy,significantly higher than baseline levels of ethylene at the end

of harvesting the graft conduits and at the time of heparinadministration and during the distal anastomoses. Exhaledethylene remained high throughout surgery although later timepoints did not show statistical significance by multiple pairwisecomparisons.

DISCUSSION

Here we present the results of the first real-time assessmentof lipid peroxidation in cardiac surgery and perioperativemedicine utilizing pulmonary excretion and accumulation ofethylene in exhaled breath. Such application represents the truetranslation of several decades of substantive basic research onethylene and lipid peroxidation to biochemical and molecularmonitoring of inflammation at the bedside at the point ofpatient care. The ability of monitoring ethylene at trace gaslevels noninvasively, in real time, and in the real-world clinicalenvironment without significant sample preparation and pro-cessing has provided us with unprecedented resolution of thedynamism of lipid peroxidation during the entire course ofsurgery. This novel information has major implications for thecentral debate regarding the status of oxidative stress in cardiacsurgery and for the innovation of breath analysis to perioper-ative medicine and beyond.

Several authors have found significant rise of lipid peroxi-dation products such as malondialdehyde-conjugated dienes,and thiobarbituric acid reactive substances in the venous cor-onary sinus blood following myocardial reperfusion in thesetting of cardiac surgery (1, 10, 15, 20). Yet there aresignificant inconsistencies in the literature regarding the tim-ing, magnitude, and significance of such changes. Furthermore,recent investigation found no increase in lipid peroxidationproducts in the coronary effluent or myocardial homogenatesor subcellular fractions following aortic cross-clamp and rep-erfusion and no detectable alterations in myocardial ultrastruc-tures (6, 7, 16).

On the basis of short-chain hydrocarbon production duringdecomposition of omega fatty acids, the pulmonary excretionof volatile gases ethane and pentane have been considerednoninvasive exhaled breath biomarkers of total body and myo-cardial lipid peroxidation. The milestone study of Andreoni etal. (2) and subsequent studies (14, 19, 26) utilized discretebreath collections and laboratory analysis and provided in-sights into these breath biomarkers. However, there are signif-icant inherent limitations of such methodology and significantcontroversy remains regarding the triggers, timing, and rele-vance of ethane kinetics.

Our measurements evidently demonstrate the feasibility ofcontinuous measurement of exhaled ethylene in mechanicallyventilated patients throughout surgery and in the exhaust of thecardiopulmonary bypass circuit. The ethylene monitor can belocated adjacent to the operating room. Moreover, samplingand analysis can be performed with only minor modification ofthe ventilation circuit. However, total intravenous anesthesiahad to be employed to avoid spectral interference from volatileanesthetic agents.

Our data clearly demonstrate substantial increases in ethyl-ene release at critical early stages of surgery and persistentlyelevated levels for most part of the intraoperative period. Theexperiments provide conclusive evidence pertinent to the de-bate on the substantial contribution of electrocautery to overall

A

B

0 10 20 30 400

20

40

60

80

100

Ethy

lene

(ppb

v)

Time (min)

DT

P

V

V +ST

DTon

0 12 24 36 48 60 720

10

20

30

40

Ethy

lene

(ppb

v)

Time (min)

CPB

Gas flow 2l/min

RewarmingXC

T = 34°C

T = 32°CGas flow 3 l/min

Fig. 2. Impact of sternotomy (A) and cardiopulmonary bypass (B) on gaseousethylene. A: typical example of real-time monitoring of ethylene followinginitial diathermy (DT) for skin incision and sternotomy (ST). P, patientbaseline; V, ventilation hold. B: example of real-time monitoring of ethyleneat onset of CPB showing influence of changing fresh gas flow, temperature (T),rewarming, and aortic cross-clamp release (XC).

Rapid Reports

L512 EXHALED ETHYLENE IN CARDIAC SURGERY

AJP-Lung Cell Mol Physiol • doi:10.1152/ajplung.00168.2014 • www.ajplung.org

lipid peroxidation (2, 19, 26). Firstly, our experiments revealpreviously unrecognized details and magnitude of early hydro-carbon release associated with diathermy and skin incision. Inour experiments, this significant rise in ethylene clearly fol-lowed the application of diathermy and preceded events asso-ciated with sternotomy such as breath holding and splitting thesternum with the surgical saw. Secondly, although we typicallyobserved a washout curve, ethylene levels did not return tobaseline levels, suggesting the possibility of longer-term im-pact of diathermy and sustained ethylene release from cauter-ized tissue. Thirdly, lesser applications of diathermy prior tochest closure also produced smaller but significant increases inbreath ethylene, thereby potentially contributing to overall

oxidative stress burden. Finally, such effect of diathermy wasalso pronounced in OPCAB procedures. Indeed the prolongedapplication of diathermy to prepare the vascular conduitsresulted in greater releases of ethylene (doubling the ethyleneAUC) in the period prior to revascularization. There is evi-dence that OPCAB is not free from surgical inflammation andsuch contribution of diathermy should be taken into consider-ation when interpreting fluid phase biomarkers of lipid peroxi-dation and oxidative stress in beating heart surgery (22, 21).

The contribution of extracorporeal circulation to oxidativestress is less conclusive. We observed increased ethylenerelease by the oxygenator of the CPB circuit at the beginningof CPB (compared with background levels during priming),which is consistent with washout of ethylene from the bloodcompartment. This may relate to earlier diathermy applica-tions, but we cannot rule out potential triggers associated withcommencement of bypass such as temporary activation of bloodelements, partial lung ischemia, and systemic effects of nonpul-satile blood flow. Nevertheless, steady-state levels of ethylenedecreased to baseline rather than increased with progression ofCPB, suggesting that these mechanisms were not the majorsource of lipid peroxidation during surgery.

Our ethylene data should be considered as part of the widerinflammatory response to CPB. Similar to others, we havedemonstrated increased neutrophil activation during CPB;therefore, neutrophil-derived oxidants could contribute to over-all lipid peroxidation. Analysis of proinflammatory cytokineresponse demonstrated small increases in some proinflamma-tory cytokines (IL-8, IL-6, IL-1) and parallel large increases inthe appropriate anti-inflammatory cytokines indicating cyto-kine balance in the plasma compartment (unpublished obser-vations). Beyond ethylene, we have also investigated otherredox gases including exhaled nitric oxide (NO) and carbonmonoxide (CO) release in patients undergoing cardiac surgery.There was no evidence for increased NO and CO in our routinepatient populations.

0 60 120 180 240

0

10

20

30

40

Ethy

lene

(ppb

v)

Time (min)

P

ST

DT

DTDT on

120 150 180

2

4

6

8

10

Ethy

lene

(ppb

v)

Time (min)

LAD

PDA

V

VOM

ASC for OM

Fig. 3. Example of real-time monitoring of ethyleneduring off-pump coronary artery bypass surgery (OP-CAB) grafting showing the reperfusion events associ-ated with completing of distal and proximal aorticanastomoses. LAD, left anterior descending, PDA, pos-terior descending; CFX, circumflex coronary artery;ASC, aortic side clamp; OM, obtuse marginal.

0

20

40

60

80

100

Ethy

lene

[ppb

v]

*

** *

Fig. 4. Summary data of real-time monitoring of ethylene during OPCABgrafting showing the major clinical and reperfusion events associated withcompleting of distal anastomoses and proximal aortic anastomoses. Box plotsrepresent 3–5 min ethylene averages. *P ( 0.05 compared with patientbaseline.

Rapid Reports

L513EXHALED ETHYLENE IN CARDIAC SURGERY

AJP-Lung Cell Mol Physiol • doi:10.1152/ajplung.00168.2014 • www.ajplung.org

These results may indicate that current conduct of CPBresults in lesser inflammatory and oxidative stress response inroutine and low-risk patients. We have to realize that perfusiontechnologies have undergone tremendous developments andnow are extensively utilized for days and weeks to facilitateorgan recovery during extracorporeal mechanical support andmembrane oxygenation. Perfusion technologies are also beingrecognized as some of the most revolutionary progresses incurrent practice of cardiothoracic transplantation such as dur-ing ex vivo perfusion and short-term recovery and recondition-ing of injured hearts and lungs.

In addition to events during CPB, separation from CPB wasnot associated with significant immediate increases in breathethylene suggesting that pulmonary reperfusion and restorationof systemic pulsatile flow was not associated with increasedoxidative stress. However, at later stages exhaled ethyleneremained significantly higher than baseline measurements andongoing pulmonary and systemic inflammation may have con-tributed to such responses.

Global myocardial ischemia-reperfusion during valvularprocedures was not associated with increased ethylene release,suggesting that there was effective protection against lipidperoxidation-mediated damage. Although the measurementsreveal exceptional details of ethylene production during reper-fusion that may be confounded by deairing events and tempo-rary reduction of perfusion flow during release of aortic cross-clamp, average ethylene prior to and after reperfusion did notchange. Blood cardioplegia, moderate degree of hypothermia,and antioxidant effects of propofol may have contributed tolack of lipid peroxidation in our patients (5, 7, 15).

Ethylene responses to regional myocardial ischemia duringOPCAB were variable, and we have identified three prototyperesponses. In '50% of reperfusion events, restoration of nativecoronary flow or full bypass grafting was associated with nochanges in ethylene release. However, the remaining reperfu-sion events produced a detectable rise in exhaled ethylene. Ina few cases we cannot rule out confounding factors associatedwith necessary management of the anastomoses (stoppingventilation, major manipulation of the heart, short burst ofdiathermy), but in the rest of the events we conclude a truereperfusion-related increase in ethylene. Our real-time analysisindicates that these reperfusion events occur on the downslopeof the diathermy-induced ethylene release and that the contri-bution of reperfusion to measured concentrations of ethylene isvariable (3–44%).

The determinants of lipid peroxidation and ethylene releaseduring OPCAB surgery remain to be established, but collateralcirculation, antioxidant treatment, hemodynamic managementof heart displacement, and coronary occlusion may play a role.

In conclusion, our original studies suggest that sensitive,totally noninvasive, and real-time analysis of lipid peroxidationusing breath ethylene is feasible in the clinical and periopera-tive setting. This innovative approach has irrefutably demon-strated the significant contribution of diathermy to overalloxidative stress and identified increased ethylene associatedwith regional myocardial ischemia on an individual basis.Clearly, our observations challenge the oxidative stress con-cept of injury associated with aortic cross-clamping and CPB.They also highlight complexities of breath analysis in theperioperative environment and strongly support the application

of real-time methods as frontiers of breath biomarker researchin the dynamic surgical setting.

Future Directions

These pilot studies in low-risk cardiac surgical patients pavethe way for future studies to study oxidative stress in higherrisk populations and in those individuals who develop postop-erative complications. This technology could advance the fieldby correlating ethylene levels with broader measures of inflam-mation in both preclinical and clinical settings and with organdysfunction or outcomes in future large-scale clinical trials.We need to better understand mechanisms of release of ethyl-ene, and single organ contributions should be evaluated by useof ex vivo human lung or heart perfusions. Studies should alsofurther address physiological and physical determinants ofexhaled ethylene such as the influence of cardiac output,ventilation, and temperature. Once these aspects are estab-lished, real-time monitoring of ethylene may have the potentialof becoming the “capnography” of perioperative inflammationand oxidative stress. Ultimately such monitoring could enabletesting pharmacological interventions or changing surgicalpractices in reducing lipid peroxidation and surgical compli-cations. Our innovative method may also offer novel opportu-nities in molecular monitoring for other fields of pulmonaryand circulation research and medicine.

ACKNOWLEDGMENTS

We are grateful for Harefield Hospital Theatres perfusionists and nurses forvaluable collaboration, help, and discussions during these studies.

Collaborators of the Harefield BIOSTRESS study include Mohamed Am-rani, Toufan Bahrami, Fabio De Robertis, Jullien Gaer, Shahzad Raja, andAndre R. Simon from the Department of Surgery, Royal Brompton andHarefield NHS Foundation Trust, Harefield, UK; Uta Bellin, John Farrimond,Isabelle Fonteyne, Andrew Gaunt, Donna Hall, Paul Harris, Sundeep Kaul,Lakshmi Kuppurao, Nicholas Lees, Simon Mattison, Grainne McDermott,Ian McGovern, Jerry Mitchell, Royston David, Anne Sigel, ChristopherWalker, and Gavin Wright from the Department of Anaesthesia, RoyalBrompton and Harefield NHS Foundation Trust, Harefield, UK; PaulaRogers from the Department of Cardiology, Royal Brompton and HarefieldNHS Foundation Trust, Harefield, UK; and Rosalba Romano and RosadaDavey from the Section of Anaesthesia, Pain Medicine and Intensive Care,Department of Surgery and Cancer, Faculty of Medicine, Imperial CollegeLondon, London, UK.

GRANTS

These studies were supported by the National Institute of Health Research,The Royal Brompton and Harefield NHS Foundation Trust Biomedical Re-search Unit; the EU-FP6-Infrastructures-5, FP6-026183, Life Science TraceGas facility; the GO-EFRO Ultragas 2009-010034, Ultra sensitive gas analysissystems for medical diagnostic and quality monitoring of horticultural prod-ucts; and Erasmus Programme Mobility 2011/2012 for staff training.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the author (s).

AUTHOR CONTRIBUTIONS

S.M.C., S.t.L.H., M.D., F.J.H., T.H.R., and N.M. conception and design ofresearch; S.M.C., R.K., T.H.R., and N.M. performed experiments; S.M.C.,R.K., T.H.R., and N.M. analyzed data; S.M.C., S.t.L.H., T.H.R., and N.M.interpreted results of experiments; S.M.C. and N.M. prepared figures; S.M.C.and N.M. drafted manuscript; S.M.C., R.K., M.D., T.H.R., and N.M. editedand revised manuscript; S.M.C., R.K., S.t.L.H., M.D., F.J.H., T.H.R., and N.M.approved final version of manuscript.

Rapid Reports

L514 EXHALED ETHYLENE IN CARDIAC SURGERY

AJP-Lung Cell Mol Physiol • doi:10.1152/ajplung.00168.2014 • www.ajplung.org

REFERENCES

1. Akila, D’souza B, Vishwanath P, D’souza V. Oxidative injury and antiox-idants in coronary artery bypass graft surgery: off-pump CABG significantlyreduces oxidative stress. Clin Chim Acta 375: 147–152, 2007.

2. Andreoni KA, Kazui M, Cameron DE, Nyhan D, Sehnert SS, RohdeCA, Bulkley GB, Risby TH. Ethane: a marker of lipid peroxidationduring cardiopulmonary bypass in humans. Free Radic Biol Med 26:439–445, 1999.

3. Bell RM, Yellon DM. There is more to life than revascularization:therapeutic targeting of myocardial ischemia/reperfusion injury. Cardio-vasc Ther 29: e67–e79, 2011.

4. Bousquet JF, Thimann KV. Lipid peroxidation forms ethylene from1-aminocyclopropane-1-carboxylic acid and may operate in leaf senes-cence. Proc Natl Acad Sci USA 81: 1724–1727, 1984.

5. Brown RH, Wagner EM, Cope KA, Risby TH. Propofol and in vivooxidative stress: effects of preservative. J Breath Res 3: 016003, 2009.

6. Chambers DJ. Oxidative stress injury during cardiac surgery: howimportant is it? Cardiovasc Res 73: 626–628, 2007.

7. Clermont G, Vergely C, Jazayeri S, Lahet J, Goudeau J, Lecour S,David M, Rochette L, Girard C. Systemic free radical activation is amajor event involved in myocardial oxidative stress related to cardiopul-monary bypass. Anesthesiology 96: 80–87, 2002.

8. Cristescu SM, Persijn ST, Hekkert STL, Harren FJM. Laser-basedsystems for trace gas detection in life sciences. Appl Phys B 92: 343–349,2008.

9. Dumelin EE, Tappel AL. Hydrocarbon gases produced during in vitroperoxidation of polyunsaturated fatty acids and decomposition of pre-formed hydroperoxides. Lipids 12: 894–900, 1997.

10. Ferrari R, Alfieri O, Curello S, Ceconi C, Cargnoni A, Marzollo P,Pardini A, Caradonna E, Visioli O. Occurrence of oxidative stressduring reperfusion of the human heart. Circulation 81: 201–211, 1990.

11. Harren FJM, Berkelmans R, Kuiper K, te Lintel Hekkert S, Scheep-ers P, Hollander P, Dekhuijzen R, Parker DH. On-line laser photoa-coustic detection of ethene in exhaled air as biomarker of ultravioletradiation damage of the human skin. Appl Phys Lett 74: 1761–1763, 1999.

12. Knight JA, Hopfer SM, Reid MC, Wong SH, Sunderman FW, Jr.Ethene (ethylene) and ethane exhalation in Ni[II]-treated rats, using animproved rebreathing apparatus. Ann Clin Lab Sci 16: 386–394, 1986.

13. Lieberman M, Hochstein P. Ethylene formation in rat liver microsomes.Science 152: 213–214, 1966.

14. Mendis S, Sobotka PA, Leja FL, Euler DE. Breath pentane and plasmalipid peroxides in ischemic heart disease. Free Radic Biol Med 19:679–684, 1995.

15. Mezzetti A, Calafiore AM, Lapenna D, Deslauriers R, Tian G, SalernoTA, Verna AM, Bosco G, Pierdomenico SD, Caccurullo F. Intermittentantegrade warm cardioplegia reduces oxidative stress and improves me-tabolism of the ischemic-reperfused human myocardium. J Thorac Car-diovasc Surg 109: 787–795, 1995.

16. Milei J, Forcada P, Fraga CG, Grana DR, Iannelli G, Chiariello M,Tritto I, Ambrosio G. Relationship between oxidative stress, lipid per-oxidation, and ultrastructural damage in patients with coronary arterydisease undergoing cardioplegic arrest/reperfusion. Cardiovasc Res 73:710–719, 2007.

17. Moeskops BW, Steeghs MM, van Swam K, Cristescu SM, ScheepersPT, Harren FJM. Real-time trace gas sensing of ethylene, propanal andacetaldehyde from human skin in vivo. Physiol Meas 27: 1187–1196,2006.

18. Nalysnyk L, Fahrbach K, Reynolds MW, Zhao SZ, Ross S. Adverseevents in coronary artery bypass graft (CABG) trials: a systematic reviewand analysis. Heart 89: 767–772, 2003.

19. Pabst F, Miekisch W, Fuchs P, Kischkel S, Schubert JK. Monitoring ofoxidative and metabolic stress during cardiac surgery by means of breathbiomarkers: an observational study. J Cardiothorac Surg 2: 37, 2007.

20. Pantke U, Volk T, Schmutzler M, Kox WJ, Sitte N, Grune T. Oxidizedproteins as a marker of oxidative stress during coronary heart surgery.Free Radic Biol Med 27: 1080–1086, 1999.

21. Raja SG, Berg GA. Impact of off-pump coronary artery bypass surgeryon systemic inflammation: current best available evidence. J Card Surg22: 445–455, 2007.

22. Rasmussen BS, Laugesen H, Sollid J, Gronlund J, Rees SE, Toft E,Gjedsted J, Dethlefsen C, Tonnesen E. Oxygenation and release ofinflammatory mediators after off-pump compared with after on-pumpcoronary artery bypass surgery. Acta Anaesthesiol Scand 51: 1202–1210,2007.

23. Risby TH, Sehnert SS. Clinical application of breath biomarkers ofoxidative stress status. Free Radic Biol Med 27: 1182–1192, 1999.

24. Sagai M, Ichinose T. Age-related changes in lipid peroxidation asmeasured by ethane, ethylene, butane and pentane in respired gases of rats.Life Sci 27: 731–738, 1980.

25. Selim M. Perioperative stroke. N Engl J Med 356: 706–713, 2007.26. Stenseth R, Nilsen T, Haaverstad R, Vitale N, Dale O. Frequent

sampling allows detection of short and rapid surges of exhaled ethaneduring cardiac surgery. Perfusion 22: 391–396, 2007.

27. Stolik S, Ramon-Gallegos E, Pacheco M, Tomas SA, Cruz-Orea A,Pérez-Zapata AJ, Gaebler R, Sánchez-Sinencio F. Photoacoustic mea-surement of ethylene as a real time biomarker of lipid peroxidationprocesses in mice. Anal Sci 17: s365–s367, 2001.

Rapid Reports

L515EXHALED ETHYLENE IN CARDIAC SURGERY

AJP-Lung Cell Mol Physiol • doi:10.1152/ajplung.00168.2014 • www.ajplung.org