Pressure-controlled and Vo

Right Ventricular Function During One-lung Ventilation: Effects of

lume-controlled Ventilation

Abdullah M. Al Shehri, MD,* Mohamed R. El-Tahan, MD,† Roshdi Al Metwally, MD,† Hatem Qutub, MD,‡

Yasser F. El Ghoneimy, MD,§ Mohamed A. Regal, MD,§ and Haytham Zien, MD†

Objectives: To test the effects of pressure-controlled

(PCV) and volume-controlled (VCV) ventilation during one-

lung ventilation (OLV) for thoracic surgery on right ventric-

ular (RV) function.

Design: A prospective, randomized, double-blind, con-

trolled, crossover study.

Setting: A single university hospital.

Participants: Fourteen pairs of consecutive patients

scheduled for elective thoracotomy.

Interventions: Patients were assigned randomly to venti-

late the dependent lung with PCV or VCV mode, each in a

randomized crossover order using tidal volume of 6 mL/kg, I:

E ratio 1: 2.5, positive end-expiratory pressure (PEEP) of 5 cm

H2O and respiratory rate adjusted to maintain normocapnia.

Measurements and Main Results: Intraoperative changes

in RV function (systolic and early diastolic tricuspid annular

velocity (TAV), end-systolic volume (ESV), end-diastolic

Journal of Cardiothoracic and Vascular Anesthesia, Vol ], No ] (Month), 2

volume (EDV) and fractional area changes (FAC)), airway

pressures, compliance and oxygenation index were

recorded. The use of PCV during OLV resulted in faster

systolic (10.1 � 2.39 vs. 5.8 � 1.67 cm/s, respectively), dia-

stolic TAV (9.2 � 1.99 vs. 4.6 � 1.42 cm/s, respectively)

(p o 0.001) and compliance and lower ESV, EDV and airway

pressures (p o 0.05) than during the use of VCV. Oxygen-

ation indices were similar during the use of VCV and PCV.

Conclusions: The use of PCV offers more improved RV

function than the use of VCV during OLV for open thor-

acotomy. These results apply specifically to younger

patients with good ventricular and pulmonary functions.

& 2014 Elsevier Inc. All rights reserved.

KEY WORDS: Thoracic surgery, one-lung ventilation, pressure-

controlled ventilation, volume-controlled ventilation, right

ventricular function.

From the Departments of *Cardiology and †Anaesthesia and Sur-gical ICU, ‡Pulmonology and Intensive Care Unit, and §Cardiothor-acic Surgery, King Fahd Hospital of the University of Dammam, AlKhubar, Saudi Arabia.

Address reprint requests to Mohamed R. El-Tahan, University ofDammam, Department of Anesthesiology, Damam, Saudi Arabia, PO40289 Al Khubar 31952, Saudi Arabia. E-mail: mohamedrefaateltahan@yahoo.com© 2014 Elsevier Inc. All rights reserved.1053-0770/2601-0001$36.00/0http://dx.doi.org/10.1053/j.jvca.2013.09.012

ONE-LUNG VENTILATION (OLV) provides excellentoperative conditions for thoracic procedures but may

impair cardiac and right ventricular (RV) function indices,increasing postoperative morbidity and mortality rates.1–2 Inthe authors’ previous study, they demonstrated that RV ejectionfraction and cardiac index parameters were reduced signifi-cantly after the initiation of OLV due to concomitant increasesin right ventricular afterload, stroke work, and end-diastolicvolume augmented by increased airway pressures.2 This maybe harmful to patients with advanced obstructive lung diseasesand those with pulmonary hypertension. Therefore, loweringairway pressures may enhance RV function.

Volume-controlled ventilation (VCV) is less commonlyused for OLV during thoracic procedures because it mayincrease airway pressure, thus impeding RV function.3

Pressure-controlled ventilation (PCV) may be useful duringOLV in lowering airway pressure and intrapulmonary shunt.4–5

Some investigators demonstrated improved oxygenation duringthe use of PCV for OLV,4–5 whereas others found nosignificant difference in arterial oxygenation between PCVand VCV during OLV.6–8

The advantages of PCV over VCV during OLV includereductions in intrapulmonary shunts and mean and peak airwaypressures,9 with the latter limiting the risks of barotrauma andimpaired RV function. The effects of PCV and VCV on RVfunction during OLV for thoracic surgery have not beenstudied yet.

The authors hypothesized that the use of PCV during OLVwould be associated with better-preserved RV function than theuse of VCV. The authors, therefore, compared the effects ofPCV and VCV during OLV on RV function, respiratorymechanics, arterial oxygenation, and postoperative complica-tions in patients scheduled for open thoracotomy.

METHODS

Following institutional review board (IRB) approval and writteninformed consent, 28 consecutive patients aged 18 to 65 years,

(American Society of Anesthesiologists physical class [II–III]) sched-uled for elective open thoracotomy, in which the duration of OLV wasexpected to exceed 1.5 hours, were included in this controlled,randomized, prospective, double-blind crossover study. The studywas registered with www.clinicaltrials.gov (ref.NCT01763879).

Patients were excluded if they had decompensated cardiac disease(New York Heart Association class 4II), forced vital capacity(FVC) or forced expiratory volume in 1 second (FEV1) o50% ofpredicted values, hepatic or renal disease, an absence of sinus rhythmsuch as atrial fibrillation and ventricular arrhythmias, mean pulmonaryartery pressure 430 mmHg, asthma, body mass index 435 kg/m2,or a previous history of pneumonectomy, bilobectomy, or lobectomy.

In all patients, standard monitors and state entropy (SE) andresponse entropy (RE) were applied. A thoracic epidural or para-vertebral catheter was inserted, but no local anesthetics were infusedduring the study to avoid their effects on hypoxic pulmonaryvasoconstriction.10

Anesthetic technique was standardized in all studied patients.Anesthesiologists who administered the anesthetic were not involvedin assessing the patient. General anesthesia was induced with propofol(2-3 mg/kg) and fentanyl (2-3 mg/kg). Cisatracurium (0.2 mg/kg) wasadministered to facilitate the placement of a left-sided double-lumentube, and the correct position of its tip was confirmed with a fiberopticbronchoscope. Anesthesia, consisting of a 0.7 to 1.5 minimum alveolarconcentration of sevoflurane, was administered to maintain SE valuesbelow 50 and the difference between RE and SE below 10. Fentanyl,0.5mg/kg increments, was administered when the SE values were 450,

014: pp ]]]–]]] 1

SHEHRI ET AL2

when the difference between RE and SE was 410, and when the meanarterial blood pressure (MAP) or heart rate was 420% of baselinevalues despite a target sevoflurane of MAC Z1.5. The radial artery wascatheterized, and cisatracurium increments were used to maintainsurgical relaxation.

Both lungs of each patient were ventilated mechanically using VCVmode, which was performed using a semi-closed circuit anesthesiaventilator, Aisys™ (Datex-Ohmeda, GE Healthcare, Helsinki, Finland),with 0.5 inspired oxygen (FIO2) in air and an actual delivered tidalvolume (VT) of 8 mL/kg predicted body weight (PBW). PBW wasestimated from the equation: PBW (kg) ¼ 50 þ 2.3 (height [in] � 60)for males and PBW (kg) ¼ 45.5 þ 2.3 (height [in] � 60) for females,an inspiratory-to-expiratory [I:E] ratio of 1:2.5 and a positiveend-expiratory pressure (PEEP) of 5 cmH2O, with respiratory rateadjusted to achieve a PaCO2 of 35 mmHg to 45 mmHg. Peak airwaypressure (Ppk) was limited to 35 cmH2O, and a fresh gas flow (FGF) of1.5 L/min to 1.8 L/min was used. A flow sensor (D-lite, GE, Helsinki,Finland) was connected between the tracheal tube and the “Y” piece ofthe respiratory circuit to measure the Ppk, plateau airway pressures(Ppl) and static compliance.

Transesophageal echocardiography was performed using a PhilipsiE33 Echocardiography System (Philips Technologies; Bothell, WA).RV function was evaluated by measuring both maximal systolic anddiastolic tricuspid annular velocities (TAV). These parameters wererecorded at the RV free wall from the deep transgastric long-axis apical4-chamber views using pulsed-wave Doppler tissue imaging. RV end-systolic volumes (ESV) and end-diastolic volumes (EDV) weremeasured with 3-dimensional (3D) TEE volumetric quantification.Later, an independent trained operator with more than 10 years’experience and who was blinded to the study groups reviewed theechocardiographic results.

OLV was initiated after pleurotomy, and the patients were allocatedrandomly to 2 groups by drawing sequentially numbered, sealed,opaque envelopes containing a computer-generated randomizationcode. In the PCV-VCV group (n ¼ 14), the dependent lung wasventilated with PCV for 30 minutes followed by the VCV, while in theVCV-PCV group (n ¼ 14), PCV and VCV were applied inreverse order.

During the PCV period, the inspiratory pressure was adjusted todeliver an actual TV of 6 mL/kg PBW to the patient’s dependent lungwithout exceeding a Ppk of 35 cmH2O. During the VCV period, thepatient’s dependent lung was ventilated with an actual TV of 6 mL/kgPBW. FIO2, I:E ratio, PEEP, respiratory rate, and FGF (1.5-1.8 L/min)were maintained as during TLV, and the lumen of the nondependentlung was left open to air. Recruitment maneuvers for the dependentlung were repeated at 30-minute intervals by increasing the inspiratorypressure to 35 cmH2O for 10 seconds.2

All surgical procedures were performed by the same surgeons.Intraoperative hypoxemia was defined as a decrease in arterial oxygensaturation to less than 90% and was treated by increasing FIO2 to 1.0.An addition of 2 cmH2O of CPAP was considered if increasing FIO2

failed to correct hypoxemia.2

Patients were administered lactated Ringer’s solution at a rate of2 mL/kg/h during surgery. If MAP decreased below 20% of baselinevalues, 250 mL of a 5% plasma protein fraction were administered; ifinsufficient, patients were administered repeated doses of intravenousephedrine (5 mg) or norepinephrine (5 mg) to maintain urine out-put Z0.5 mL/kg/hour. Hemoglobin concentration Z8 g/dL was main-tained by administering red blood cell concentrates.

At the end of surgery, the nondependent lung was re-expanded,TLV was resumed as before surgery, sevoflurane was discontinued, theresidual neuromuscular block was antagonized, and the patient wasextubated. Postoperative analgesia consisted of continuous epidural orparavertebral infusion of bupivacaine, 0.125%, and fentanyl, 2 mg/mL.

Computerized data were collected by an independent investigatorblinded to patient allocation and not involved in patient management.Primary outcomes included changes in RV function (systolic and earlydiastolic TAV, EDV, ESV, and fractional area changes [FAC]).Secondary outcome variables included hemodynamic parameters (heartrate and MAP); Ppk; Ppl; static compliance of the respiratory systemcalculated as exhaled TV/Ppl (the sum of externally applied andintrinsic PEEP); the ratios of arterial tension to inspired fraction ofoxygen (PaO2/FIO2), arterial oxygen saturation (SaO2) and carbondioxide tension (PaCO2); intraoperative need for CPAP during OLV;use of fentanyl, ephedrine, and norepinephrine; perioperative hypoxe-mia (SaO2 o90%); and the rates of respiratory and circulatory failureand arrhythmias. Secondary outcomes also included length of hospitalstay and postoperative complications such as the need for ICUadmission, acute lung injury (ALI), pneumonia, atelectasis, and re-thoracotomy for air leakage as well as a 30-day mortality rate. RVfunction and hemodynamic, oxygenation, and ventilation variableswere recorded after induction of anesthesia (postinduction), 30 minutesafter the use of PCV and VCV during OLV (PC-OLV and VC-OLV,respectively), and 15 minutes after resuming TLV.

A previous study showed that the normally distributed mean systolicTAV in anesthetized patients was 7 cm/s (standard deviation [SD], 1.4cm/s).11 A priori power analysis indicated that a sample size of 13 pairswas sufficiently large to detect a 20% difference in mean systolic TAVafter the start of OLV with a type-I error of 0.05 and a power of 90%.To compensate for patients dropping out during the study, more patients(10%) were added for a final sample size of 28 patients.

The carryover effect (ie, the persistence of the effect of the firstintervention on the operative conditions into the second period) wasavoided by comparing the effects of period (time effect) and theorder of treatment using independent t-tests. Data were tested fornormality using the Kolmogorov-Smirnov test. Fisher’s exact testwas used to compare categoric data. Repeated two-way ANOVA andpaired t-tests were used to study changes in primary and secondaryendpoints during each intervention. The Wilcoxon rank sum test wasused to compare nonparametric values. Data were expressed asmean � SD or number (%). A p value o 0.05 was consideredstatistically significant.

RESULTS

All 28 consecutive patients enrolled and scheduled for openthoracotomy (14 in the PCV-VCV group and 14 in the VCV-PCV group) completed the study. Their demographic andclinical characteristics, all of which were similar in the 2groups, are presented in Table 1, including FEV1, FVC, side ofthoracotomy, underlying pathology, type of surgery and intra-operative fentanyl use, administration of ephedrine and nor-epinephrine, and durations of surgery and anesthesia.

The PCV-VCV and VCV-PCV groups also did not differsignificantly in post-induction systolic and early diastolic TAV,ESV, EDV, FAC, HR, MAP, Ppk, Ppl, static compliance,PaO2/FIO2 ratio, SaO2, and PaCO2.

Compared with the postinduction, VC-OLV and two-lungventilation periods, the use of PC-OLV resulted in significantlyfaster systolic (7.1 � 1.4, 5.8 � 1.7, and 6.9 � 1.5 v10.1 � 2.4 cm/s, respectively) and early diastolic (5.4 � 1.4,4.6 � 1.4, and 5.7 � 1.5 v 9.2 � 1.9 cm/s, respectively), TAVvalues (p o 0.001 each) and significantly lower RV EDVvalues (p o 0.005) (Table 2).

Compared with the postinduction, PC-OLV, and two-lungventilation periods, the use of VC-OLV resulted in significantly

Table 1. Patients’ Characteristics

PCV-VCV

(n ¼ 14)

VCV-PCV

(n ¼ 14)

Age (yrs) 37.4 � 11.51 39.1 � 13.93

Sex (Male : Female) 10 : 4 8 : 6

Weight (kg) 76.9 � 10.27 80.3 � 8.18

Height (cm) 170.2 � 6.04 168.9 � 8.12

FEV1 (% of the predicted value) 83 � 9.9 86 � 11.6

FVC (% of the predicted value) 91 � 7.4 93 � 9.0

Side of thoracotomy (Right:Left) (n) 5 : 9 7 : 7

Primary diagnoses (n [%])

Bronchiectasis 5 (35.7%) 4 (28.6%)

Lung abscess 4 (28.6%) 7 (50%)

Solitary lung nodule 2 (14.3%) 1 (7.2%)

Tuberculosis 3 (21.4%) 1 (7.2%)

Bronchogenic carcinoma 0 (0%) 1 (7.2%)

Type of Surgery (n [%])

Lobectomy 2 (14.3%) 1 (7.2%)

Wedge resection 4 (28.6%) 2 (14.3%)

Segmentectomy 6 (42.8%) 10 (71.4%)

Repair of bronchopleural fistula 2 (14.3%) 1 (7.2%)

Intraoperative fentanyl use (mg) 260 � 58.1 242 � 57.6

Patients received ephedrine (n) 3 (21.4%) 2 (14.3%)

Patients received norepinephrine (n) 0 (0%) 1 (7.2%)

Duration of surgery (min) 125.6 � 26.7 118.3 � 37.9

Duration of anesthesia (min) 145.0 � 15.6 135.9 � 20.0

NOTE. Data are presented as mean � standard deviation and

number (%).

Abbreviations: FEV1, forced expiratory volume in 1 second; FVC,

forced vital capacity; PCV, pressure-controlled ventilation; VCV,

volume controlled ventilation.

ONE LUNG VENTILATION MODE AND RIGHT VENTRICULAR FUNCTION 3

higher RV ESV values (p o 0.001). Moreover, compared withthe postinduction and two-lung ventilation periods, the use ofVC-OLV had significantly lower FAC values (p o 0.01)(Table 2).

The peak and plateau airway pressures were significantlylower (p o 0.001) and static lung compliance was significantlyhigher (p o 0.001) during the use of PC-OLV than during theuse of VC-OLV (Table 3). Intraoperative heart rate, MAP,

Table 2. Right Ventricular Function Parameters

Postinduction

(n ¼ 28)

PCV

(n ¼ 28)

VCV

(n ¼ 28)

TLV

(n ¼ 28)

Systolic TAV

(cm/s)

7.1 � 1.4 10.1 � 2.4* 5.8 � 1.7 6.9 � 1.5

Diastolic TAV

(cm/s)

5.4 � 1.4 9.2 � 1.9* 4.6 � 1.4 5.7 � 1.5

ESV (mL/m2) 31.6 � 7.7 30.2 � 8.4 46.0 � 10.9† 33 � 6.0

EDV (mL/m2) 62.6 � 25.4 57.3 � 22.4* 79.3 � 28.4 68.5 � 25.5

FAC (%) 44.7 � 16.6 43.4 � 14.5 38.3 � 15.5‡ 45.4 � 23.2

NOTE. Data are presented as mean � standard deviation.

Abbreviations: EDV, right ventricular end-diastolic volume; ESV,

right ventricular end-systolic volume; FAC, right ventricular fractional

area changes; PCV, pressure-controlled ventilation; TLV, two-lung

ventilation; TAV, tricuspid annular velocity; VCV, volume-controlled

ventilation.

*p o 0.001 significant compared with the postinduction, VCV

and TLV.

†p o 0.001 significant compared with the postinduction, PCV

and TLV.

‡p o 0.01 significant compared with the postinduction and TLV.

PaO2/FIO2 ratio, SaO2 and PaCO2 values were similar duringthe PC-OLV and VC-OLV periods (Table 3).

Statistical analysis showed no significant influence ofmeasured systolic and early diastolic TAV, ESV, EDV, orFAC during one period on the other. None of these patientsexperienced complications, including respiratory and circula-tory failure, ALI, pneumonia or death within 30 days. Rates ofCPAP during OLV, perioperative hypoxemia, and arrhythmiaswere similar in the PCV-VCV and VCV-PCV groups, as werelengths of hospital stay and postoperative need for ICUadmission, atelectasis, and reintervention (Table 4).

DISCUSSION

OLV has been associated with significant reductions in theRV ejection fraction2,12 and myocardial performance index(MPI).13 OLV concomitantly may increase airway pressures,resulting in increases in the right ventricular afterload, EDV,and stroke work index.2,14

The authors have shown that, when compared with VC-OLV, the use of PC-OLV, when the ventilator was set todeliver the same tidal volumes, resulted in higher systolic anddiastolic TAV and static compliance and lower ESV, EDV, andairway pressures in patients undergoing open thoracotomyunder sevoflurane anesthesia. To the authors’ knowledge, thisstudy is the first to assess the effects of ventilation mode on RVfunction during OLV.

Although right ventricular systolic function has beenevaluated echocardiographically, there is no gold standard forits assessment because of its complex crescent-shaped structurewrapped around the left ventricle and its incomplete visual-ization on any single 2D echocardiographic view.15–16 Max-imal tricuspid annular plane systolic excursion has been foundto correlate well with right ventricular function but also appearsto depend on left ventricular systolic function.17 Two-dimensional FAC has been shown to be significantly correlatedwith right ventricular ejection fraction measured by magneticresonance imaging (r ¼ 0.88).18 However, it is important to

Table 3. Hemodynamic and Respiratory Data

Postinduction

(n ¼ 28)

PCV

(n ¼ 28)

VCV

(n ¼ 28)

TLV

(n ¼ 28)

Heart rate (bpm) 71 � 10.3 73 � 14.7 69 � 11.5 83 � 10.7

MAP (mmHg) 85 � 9.6 89 � 10.3 91 � 10.5 93 � 9.4

Ppk (cm H2O) 18 � 3.1 21 � 3.0* 28 � 2.2 17 � 3.5

Ppl (cm H2O) 12 � 2.9 14 � 3.0* 18 � 2.4 11 � 2.4

Static

compliance

(mL/cm H2O)

37 � 5.1 31 � 5.4* 24 � 3.7 35 � 4.5

PaO2/FIO2 ratio 298 � 78.4 147 � 50.2 119 � 37.6 328 � 81.2

SaO2 (%) 97.2 � 1.4 94.8 � 1.2 94.2 � 1.5 97.1 � 11.5

PaCO2 (mmHg) 37.4 � 3.4 41.4 � 3.3 42.0 � 3.3 38.3 � 3.0

NOTE. Data are presented as mean � standard deviation.

Abbreviations: bpm, beats per minute; MAP, mean arterial blood

pressure; PaCO2, carbon dioxide tension; PaO2/FIO2 ratio, arterial

tension to inspired fraction of oxygen ratio; PCV, pressure-controlled

ventilation; Ppk, peak airway pressure; Ppl, plateau airway pressure;

SaO2, arterial oxygen saturation; TLV, two-lung ventilation; VCV,

volume controlled ventilation.

*p o 0.001 significant compared with VCV.

Table 4. Secondary Outcomes

PCV-VCV

(n ¼ 14)

VCV-PCV

(n ¼ 14)

Intraoperative use of CPAP during OLV

(n [%])

2 (14.3%) 1 (7.15%)

Length of hospital stay (days) 8.51 � 3.45 10.05 � 5.5

Postoperative complications (n [%])

Perioperative hypoxemia (n [%]) 2 (14.3%) 1 (7.15%)

Respiratory and circulatory failure 0 (0%) 0 (0%)

Arrhythmias 0 (0%) 1 (7.15%)

The need for ICU admission 1 (7.15%) 0 (0%)

Acute lung injury 0 (0%) 0 (0%)

Pneumonia 0 (0%) 0 (0%)

Atelectasis 1 (7.15%) 1 (7.15%)

Redo thoracotomy for air leakage 1 (7.15%) 0 (0%)

30 days mortality (n [%]) 0 (0%) 0 (0%)

NOTE. Data are presented as number and mean � standard devia-

tion.

Abbreviations: CPAP, continuous positive airway pressure; ICU,

intensive care unit; OLV, one-lung ventilation; PCV, pressure-con-

trolled ventilation; VCV, volume-controlled ventilation.

SHEHRI ET AL4

exclude trabeculations while tracing the right ventricular area.15

Tissue imaging of TAV often is used as a surrogate for regionalright ventricular work and depends on both the loadingconditions and contractility.15 Pulsed-wave Doppler techniquesmeasure the “peak” TAV, whereas color tissue Dopplermeasures the “average” velocity over the sample area.19 Rightventricular function may be assessed more accurately by 3-than by 2-dimensional TEE because the right ventricularoutflow tract geometry is generally oval rather than circular.20

Three-dimensional TEE measurements of the right ventricularESV and EDV have been found to correlate well with strokevolume (r ¼ 0.85), validating its use in the present study.21 Theauthors found that PC-OLV was superior to VC-OLV forthoracic procedures, resulting in significantly faster systolic andearly diastolic TAV, lower ESV and EDV, and higher FAC,because of the concurrent decreases in airway pressures and,hence, the RV afterload. The initiation of positive-pressureventilation after induction of anesthesia usually is associatedwith a 5% to 10% decrease in RV FAC.22 Using VC-OLV, theauthors observed an additional 14% decrease in right ventric-ular FAC values than during the postinduction period.

OLV has been associated with increased peak airwaypressures and reduced static lung compliance.23 The authorsfound that, when compared with VC-OLV, PC-OLV wasassociated with 30% lower airway pressures and lower staticlung compliance. Similarly, others have reported 4% to 30%reductions in peak airway pressure during PC-OLV.5–6 Incontrast, other investigators found that during PC-OLV, theassociated decrease in peak airway pressures was observedmainly in the respiratory circuit (28.5%) rather than in thebronchus of the dependent lung (6%), which is essentiallylinked to the resistive pressure originating from the smallinternal diameter of the endobronchial tube.9 Therefore, the useof PC-OLV and VC-OLV may have comparable effects onalveolar distention, the compression of intra-alveolar vessels,and resistance to pulmonary blood flow in the dependent lungand, hence, on RV function, which may affect the validity ofthe authors’ findings. That study, however, included fewer

patients and a non-blinded, non-randomized design, and acarryover effect could not be excluded.9

Similar to previous results,6–7 the authors observed com-parable arterial oxygenation during the use of PC-OLV andVC-OLV. Although another study reported that PC-OLVimproved arterial oxygenation during PC-OLV,24 that studyincluded fewer patients.

The authors found that PC-OLV was of clinical importancebecause, compared with VC-OLV, it was associated withsignificant improvements in tricuspid annular systolic(þ72%) and early diastolic (þ100%) velocities, lower ESV(�53%) and EDV (�38.5%), and lower airway pressures(�30%). RV function declined when changing from PCV toVCV and improved when changing from VCV to PCV. Thismay be beneficial in patients with right ventricular dysfunctionand pulmonary hypertension in whom the use of VC-OLV maybe associated with a clinically relevant hemodynamic responseto RV dysfunction.

Although this may not be clinically relevant, varying FGFmay result in variations in the total delivered tidal volumes.The product of FGF and inspiratory time yields the volume offresh gas to be added to the set tidal volume for ventilators thatare not FGF-decoupled, such as Aisys™.25 Thus, to overcomethis problem, the authors targeted the delivery of an actualrather than a set tidal volume of 8 mL/kg during two-lungventilation and 6 mL/kg during OLV in both groups.

Patient selection may have played a major role in thedifferences between the 2 ventilatory techniques in RVfunction, especially in older patients with chronic restrictiveor obstructive lung disease who may have some degree ofpulmonary hypertension and right ventricular dysfunction.26

Obese patients with a BMI 430 kg/m2 were excluded fromthis study to avoid any effects of obesity on pulmonarymechanics. PCV and VCV have shown comparable effectson hemodynamics and oxygenation in morbidly obese patientsduring pneumoperitoneum for laparoscopic gastric banding,27

which may preclude extending the benefits of PC-OLV on RVfunction in obese patients.

The potential problem in any crossover study is that carry-over effects may bias the direct effects of each intervention onRV function. The authors assumed that a 30-minute washoutperiod would eliminate these effects. They observed no carry-over effect on TAV and RV volumes because statistical analysisshowed no evidence that one period influenced the other.

Further multicenter studies are needed to address the clinicalimportance of the use of PC-OLV and of various I:E ratiosduring the use of VC-OLV, especially in elderly and obesepatients with severe pulmonary and RV dysfunction.

The present study had several limitations. First, the patientsin the present study were extremely young, not obese, and hadvery good baseline pulmonary and cardiac functions. Thecommon underlying pathologies of the authors’ patients werequite different from those of elderly patients with lung cancerand possibly impaired right ventricular function. Second,although a pulmonary artery catheter with a rapid-responsethermistor may be a more reliable tool for assessment of RVfunction and afterload,28 this type of monitor was not availableat the authors’ center during the performance of this study.Third, measuring echocardiographically-derived stroke volume

ONE LUNG VENTILATION MODE AND RIGHT VENTRICULAR FUNCTION 5

or MPI may have been informative in assessing changes in RVfunction with the different modes of ventilation.13 Fourth, theauthors used only 1 VCV setting, and their I:E ratio of 1:2.5,although common, was low. Changing the I:E ratio and theFGF in VCV may have mimicked PCV. However, otherinvestigators also have used a fixed I:E ratio when comparing

PCV and VCV during OLV to avoid any possible confoundingeffects on outcome variables.4–9

In conclusion, the authors found that, compared with VCV,PCV during OLV improves RV function during open thor-acotomy. These results apply specifically to younger patientswith good ventricular and pulmonary functions.

REFERENCES

1. Diaper J, Ellenberger C, Villiger Y, et al: Comparison of cardiacoutput as assessed by transesophageal echo-Doppler and transpulmo-nary thermodilution in patients undergoing thoracic surgery. J ClinAnesth 22:97-103, 20102. El Tahan MR, El Ghoneimy Y, Regal M, et al: Effects of

nondependent lung ventilation with continuous positive-pressure ven-tilation and high-frequency positive-pressure ventilation on right-ventricular function during 1-lung ventilation. Semin CardiothoracVasc Anesth 14:291-300, 20103. Shelley B, Macfie A, Kinsella J: Anesthesia for thoracic surgery:

A survey of UK practice. J Cardiothorac Vasc Anesth 25:1014-1017, 20114. Sentürk NM, Dilek A, Camci E, et al: Effects of positive end-

expiratory pressure on ventilatory and oxygenation parameters duringpressure-controlled one-lung ventilation. J Cardiothorac Vasc Anesth19:71-75, 20055. Tuğrul M, Camci E, Karadeniz H, et al: Comparison of volume

controlled with pressure-controlled ventilation during one-lung anaes-thesia. Br J Anaesth 79:306-310, 19976. Pardos PC, Garutti I, Piñeiro P, et al: Effects of ventilatory mode

during one-lung ventilation on intraoperative and postoperative arterialoxygenation in thoracic surgery. J Cardiothorac Vasc Anesth 23:770-774, 20097. Unzueta MC, Casas JI, Moral MV: Pressure-controlled versus

volume-controlled ventilation during one-lung ventilation for thoracicsurgery. Anesth Analg 104:1029-1033, 20078. Choi YS, Shim JK, Na S, et al: Pressure-controlled versus

volume-controlled ventilation during one-lung ventilation in the proneposition for robot-assisted esophagectomy. Surg Endosc 23:2286-2291, 20099. Rozé H, Lafargue M, Batoz H, et al: Pressure-controlled ven-

tilation and intrabronchial pressure during one-lung ventilation. Br JAnaesth 105:377-381, 201010. Silva-Costa-Gomes T, Gallart L, Vallès J, et al: Low- vs. high-

dose almitrine combined with nitric oxide to prevent hypoxia duringopen-chest one-lung ventilation. Br J Anaesth 95:410-416, 200511. Michaux I, Filipovic M, Seeberger M, et al: Are normal

echocardiographic values obtained by transthoracic echocardiographyin awake patients suitable for evaluation of cardiac function inanesthetized and mechanically ventilated patients? Anesth Analg 109:1701, 200912. Kellow NH, Scott AD, White SA, et al: Comparison of the

effects of propofol and isoflurane anaesthesia on right ventricularfunction and shunt fraction during thoracic surgery. Br J Anaesth 75:578-582, 199513. Matyal R, Mahmood F, Hess P, et al: Right ventricular

echocardiographic predictors of postoperative supraventricular arrhyth-mias after thoracic surgery: A pilot study. Ann Thorac Surg 90:1080-1086, 2010

14. Kozian A, Schilling T, Fredén F, et al: One-lung ventilationinduces hyperperfusion and alveolar damage in the ventilated lung: Anexperimental study. Br J Anaesth 100:549-559, 200815. Lang RM, Bierig M, Devereux RB, et al: Recommendations for

chamber quantification. Eur J Echocardiogr 7:79-108, 200616. Gopal AS, Chukwu EO, Iwuchukwu CJ, et al: Normal values of

right ventricular size and function by real-time 3-dimensional echo-cardiography: Comparison with cardiac magnetic resonance imaging.J Am Soc Echocardiogr 20:445-455, 200717. López-Candales A, Rajagopalan N, Saxena N, et al: Right

ventricular systolic function is not the sole determinant of tricuspidannular motion. Am J Cardiol 98:973-977, 200618. Khattab K, Schmidheiny P, Wustmann K, et al: Echocardiogram

versus cardiac magnetic resonance imaging for assessing systolicfunction of subaortic right ventricle in adults with complete trans-position of great arteries and previous atrial switch operation. Am JCardiol 111:908-913, 201319. Tumuklu MM, Erkorkmaz U, Ocal A: The impact of hyper-

tension and hypertension-related left ventricle hypertrophy on rightventricle function. Echocardiography 24:374-384, 200720. Izumo M, Shiota M, Saitoh T, et al: Non-circular shape of right

ventricular outflow tract: A real-time 3-dimensional transesophagealechocardiography study. Circ Cardiovasc Imaging 5:621-627, 201221. Karhausen J, Dudaryk R, Phillips-Bute B, et al: Three-

dimensional transesophageal echocardiography for perioperative rightventricular assessment. Ann Thorac Surg 94:468-474, 201222. Boldt J, Müller M, Uphus D, et al: Cardiorespiratory changes in

patients undergoing pulmonary resection using different anestheticmanagement techniques. J Cardiothorac Vasc Anesth 10:854-859, 199623. Ye FF, Li LW: Effects of different ventilation modes for one-lung

ventilation anesthesia on respiratory function and F(A)/F(I) changesduring sevoflurane inhalation. Nan Fang Yi Ke Da Xue Xue Bao 31:714-717, 201124. Heimberg C, Winterhalter M, Strüber M, et al: Pressure-

controlled versus volume-controlled one-lung ventilation for MIDCAB.Thorac Cardiovasc Surg 54:516-520, 200625. Wallon G, Bonnet A, Guérin C: Delivery of tidal volume from

four anaesthesia ventilators during volume-controlled ventilation: Abench study. Br J Anaesth 110:1045-1051, 201326. Andersen CU, Mellemkjær S, Nielsen-Kudsk JE, et al: Pulmo-

nary hypertension in chronic obstructive and interstitial lung diseases.Int J Cardiol (in press)27. De Baerdemaeker LE, Van der Herten C, Gillardin JM, et al:

Comparison of volume-controlled and pressure-controlled ventilationduring laparoscopic gastric banding in morbidly obese patients. ObesSurg 18:680-685, 200828. Zink W, Nöll J, Rauch H, et al: Continuous assessment of right

ventricular ejection fraction: New pulmonary artery catheter versustransoesophageal echocardiography. Anaesthesia 59:1126-1132, 2004