applied cardiopulmonary pathophysiology Vol. 11, No. 1 ...€¦ · ing, and pulmonary arterial hypertension, consecutive-ly resulting in a deterioration of arterial oxygenation. Therapeutic

Inhaled nitric oxide versus partial liquid ventilation in surfactant depleted pigsS. Wolf, T. Busch, M. Deja, H. Lohbrunner, S. Weber-Carstens, C. Sun, U. Kaisers . . . . . . . . . . . . . . . . . . . . . . . 3

Lung temperature during cardiopulmonary bypass and the effect of hypothermic ventilationJ. F. M. Bechtel, W. Eichler, H.-H. Sievers, C. Bartels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11

Thoracic epidural anaesthesia as part of a fasttrack-concept for cardiac surgeryC. D. Kratz, U. Schirmer, S. Weiss, H. Wulf, L. Eberhart, S. Vogt, G. Geldner . . . . . . . . . . . . . . . . . . . . . . . . . . 17

The inflated balloon - a possible cause of microvascular obstructionsF. Jung, J.-W. Park, R. P. Franke . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21

Fibrin pathways for precursor cell immigration from bone marrow into the osteotomy gapK. Wolf, J. Hamar, S. Moravec, T. Farkas, E. Höcherl, C. Pfister . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24

Morphometric analyses of the kinetics of vascular and neural ingrowth into the interfragmental gap following osteotomyK. Wolf, J. Hamar, E. Höcherl, T. Farkas, C. Pfister . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36

Laser-doppler-flowmetry and absorption-tissue-spectrometry of the transposed groin flap - A comprehensive and independent analysis of microcirculationK. Wolf, E. Höcherl, T. Derfuß, A. Krug . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44

Bronchoprovocation or reversibility in asthma patients with normal pulmonary function testsM. Erelel . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56

Induction of a reproducible, hypoxemic and hemodynamically compromised but stable early Acute Lung Injury (ALI) model with oleic acid in pigs - tips and pitfallsT. Iber, J. P. Roesner, C. Mutz, G. F. E. Nöldge-Schomburg, D. A. Vagts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60

applied cardiopulmonary pathophysiology Vol. 11, No. 1-2007

Contents

Subject Editors

AnesthesiologyW. Erdmann (Rotterdam, The Netherlands)F. Giunta (Pisa, Italy)

CardiologyE. O. McFalls (Minneapolis, MN, USA)P. Verdouw (Rotterdam, The Netherlands)

Intensive Care and Resuscitation B. W. Böttiger (Heidelberg, Germany)N. Mutz (Innsbruck, Austria)P. J. Papadakos (Rochester, NY, USA)

NeonatologyC. Speer (Tübingen, Germany)O. Saugstad (Oslo, Norway)

Nuclear MedicineF. Fazio (Milan, Italy)P. Wollmer (Malmö, Sweden)

Oxygen TransportN. S. Faithfull (San Diego, CA, USA)K. Reinhard (Jena, Germany)

PediatricsM. Silverman (London, Great Britain)H. Stopfkuchen (Mainz, Germany)

PneumologyR. G. Spragg (San Diego, CA, USA)N. F. Voelkel (Denver, CO, USA)

Pulmonary SurfactantL. M. G. van Golde (Utrecht, The Netherlands)E. V. Cosmi (Rome, Italy)

ShockH. Redl (Vienna, Austria)R. M. Strieter (Ann Arbor, MI, USA)

Editors:Uwe Schirmer, Dept. Cardiac Anesthesia, University Ulm, Steinhövelstr. 9, D-89075Ulm, E-mail: [email protected]

Matthias Heringlake, Dept. Anesthesia, University of Lübeck, Ratzeburger Allee160, D-23538 Lübeck, E-mail: [email protected]

B. Lachmann, Department of Anesthesiology, Erasmus University, P.O. Box 1738,NL-3000 DR Rotterdam, The Netherlands

Editorial/Publishers Office:Applied Cardiopulmonary Pathophysiology, Pabst Publishers, Eichengrund 28, D-49525 Lengerich, Germany, Phone: ++ 49 (0) 5484-97234, Fax ++ 49 (0) 5484-550,E-mail: [email protected], Internet: www.pabst-publishers.com

Subscription prices, per volume (= 4 issues): Institutions USD 212.00 / Euro 260.00 including postage and han-dling. Individuals (no institutions, libraries, etc.) may subscribe at the reduced rate of USD 16.00 / Euro 20.00 pervolume. They must declare that the subscription is for their own private use, it will not replace any institutionalsubscription and it will not be put at the disposal of any library. Subscription should be sent to Pabst Publishers,Eichengrund 28, D-49525 Lengerich, Germany, or to any subscription agent. Changes of mailing address shouldbe notified together with our latest label.

The journal Applied Cardiopulmonary Pathophysiology is indexed/abstracted in Current Contents (CC), ScienceCitation Index Expanded, Excerpta Medica/EMBASE, International Bibliography of the Social Sciences.

NO and partial liquid ventilation 3

Introduction

Acute lung injury (ALI) and the acute respiratory dis-tress syndrome (ARDS) are characterised by the pres-ence of alveolar collapse due to impairment of the pul-

monary surfactant system (1), a mismatch of ventila-tion to perfusion, intrapulmonary right-to-left shunt-ing, and pulmonary arterial hypertension, consecutive-ly resulting in a deterioration of arterial oxygenation.Therapeutic concepts, aimed at an improvement of ar-

Inhaled nitric oxide versus partial liquid ventilation in surfactant depletedpigs

S. Wolf1*, T. Busch2*, M. Deja1, H. Lohbrunner1, S. Weber-Carstens1, C. Sun1, U. Kaisers2

1Department of Anaesthesiology and Intensive Care Medicine, Campus Virchow Klinikum, Humboldt-University,Berlin, Germany; 2Department of Anesthesiology and Intensive Care Medicine, Medical Faculty, University ofLeipzig, Germany

Applied Cardiopulmonary Pathophysiology 11: 3-10, 2007

Keywords: acute lung injury, ARDS, nitric oxide, partial liquid ventilation, gas exchange, hemodynamics, animalmodel

Abstract

Objective: To determine the effects of inhaled NO (iNO) and partial liquid ventilation (PLV) on pulmonary gasexchange and hemodynamics in an experimental model of acute lung injury (ALI) in juvenile pigs.Design: Prospective, randomised, controlled study.Methods: 22 anaesthetised, tracheotomised and mechanically ventilated (FIO2 1.0) pigs underwent induction ofALI by repeated saline washout of surfactant. Animals were randomly assigned to receive either a continuouslyinhalation of 20 ppm NO (NO-group, n = 7), to receive PLV (PLV-group, n = 7) using 30 ml⋅kg-1 of perfluorocar-bons (PF 5080, 3M, Germany), or to receive no further intervention (Controls, n = 8).Measurements and Results: Measurements of pulmonary gas exchange and hemodynamics were performedhourly for a four hours period. Induction of ALI reduced PaO2 from 546 ± 19 to 61 ± 4 mmHg (mean ± SEM) inall animals. In the NO-group inhalation of NO induced an increase in PaO2 to 107 ± 12 mmHg that remained sta-ble throughout the experiment (p<0.05 compared to Controls) and intrapulmonary shunt ( ) decreased to 37± 4 %. In the PLV-group, PaO2 increased to 318 ± 41 mmHg four hours after intervention (p<0.05 vs. Controlsand vs. NO) and decreased to 26 ± 4 % (p<0.05 vs. Controls, two hours after induction of ALI). Inhalation of NO induced a significant and sustained reduction in mean pulmonary artery pressure (MPAP) (-4 ± 4 % p<0.05 vs. Controls and vs. PLV). In the PLV-group, treatment induced an increase in MPAP (+30 ± 9 %p<0.05 compared to NO). Cardiac output remained stable in the NO-group, whereas it was significantly decreasedin Controls (p < 0.05).Conclusion: In this surfactant washout model of ALI, full dose PLV was superior to improve oxygenation and todecrease intrapulmonary right-to-left shunt compared to inhalation of 20 ppm NO. However, inhalation of NO waseffective to reduce mean pulmonary artery pressure.

TS Q/Q

TS Q/Q

*S. Wolf and T. Busch contributed equally to this study.

4 S. Wolf, T. Busch, M. Deja, H. Lohbrunner, S. Weber-Carstens, C. Sun, U. Kaisers

terial oxygenation, produce their effects generallythrough two different strategies: One is alveolar re-cruitment and stabilisation of alveoli by means of ade-quate mechanical ventilation, and another is redistrib-ution of pulmonary blood flow towards regions withnormal matching, thereby decreasing .

In ARDS, inhalation of nitric oxide (iNO) has beendemonstrated to improve matching of ventilation toperfusion by a selective vasodilation of pulmonaryvessels in ventilated lung regions (2). However, anumber of patients with ARDS do not respond well toinhaled NO (3). It has been assumed that this might inpart be due to inhibition of hypoxic pulmonary vaso-constriction in non-ventilated lung areas (4). Accord-ingly, it has been demonstrated that NO-induced im-provements in arterial oxygenation were significantlyrelated to baseline levels of pulmonary vascular resist-ance (5).

A different experimental strategy to treatALI/ARDS is partial liquid ventilation (PLV), aimedat lung recruitment. Here, the lung is partially filledwith perfluorocarbons (PFCs) and then conventionalventilation is resumed (6-9). PFCs are liquid compo-nents with a high density, a low surface tension, and anexcellent solubility for oxygen and carbon dioxide.Due to their physico-chemical properties, intra-pul-monary applied PFCs induce a recruitment mainly independent lung regions providing improvements ingas exchange in ALI (10).

Recently, investigators addressed the questionwhether a combination of both treatments could en-hance the effects on gas exchange and pulmonary he-modynamics (11-16). These studies, mainly done inpremature animals with experimental acute lung in-jury, revealed positive results for the combined treat-ment. Since an exclusively, controlled comparison ofboth treatments has not been performed, the presentstudy aims at a comparison of iNO versus PLV andtheir effects on hemodynamics and gas exchange in anexperimental model of acute lung injury in pigs.

Methods

This study was approved by the Berlin Animal Protec-tion Committee in accordance with German AnimalProtection Law, and conforms with the Guide for theCare and Use of Laboratory Animals (DHHS, PHS,NIH Publication No. 85-23).

General experimental procedures

Twenty-two piglets with a body weight of 26 ± 2 kgwere studied. Anaesthesia was induced with thiopental(10 mg⋅kg-1 i.v.) and fentanyl (10 µg⋅kg-1 i.v.) followedby an infusion of thiopental (0.13 mg⋅kg-1) and fen-tanyl (0.05-0.08 µg⋅kg-1⋅min-1). Muscle relaxation wasobtained with pancuronium bromide (0.15 mg⋅kg-1 i.v.bolus, followed by a continuous infusion of 2.5µg⋅kg-1⋅min-1). Immediately after induction, pigletswere tracheotomized and intubated with a 9.0 mm out-er diameter tracheal tube, fitted with a heat moistureexchanger.

During the baseline and the induction of ALI, ani-mals were placed in supine position and ventilated vol-ume controlled (tidal volume 12±2 ml⋅kg-1, respiratoryrate 16 min-1, FIO2 1.0, inspiratory - expiratory ratio1:1, PEEP 5 cm H2O) using a Servo 300 ventilator(Siemens-Elema, Solna, Sweden). The core tempera-ture of the animals was maintained within ±0.5 °C ofthe pre-study value using a heating pad. Throughoutthe experiments no cardiotonic or vasoactive drugswere administered.

In each pig we percutaneously placed a pulmonaryartery catheter (model 93A-431-7.5Fr, Baxter Health-care Corporation, Irvine, CA, USA) via the femoralvein, and an arterial line (18 G; Vygon, Ecouen,France) was inserted into the femoral artery, for bloodsampling and hemodynamic measurements. Heart rate(HR), central venous pressure (CVP), mean arterialpressure (MAP), mean pulmonary artery pressure(MPAP) and pulmonary capillary wedge pressure(PCWP) were recorded using a Hewlett-Packard mon-itoring system (Model 66 S, Böblingen, Germany).Measurements were taken with pigs in the supine po-sition with a zero reference level at the midaxilla lev-el. Vascular pressures were the average taken at end-expiration of three successive respiratory cycles. Car-diac output (CO) was determined with thermodilutiontechnique and is expressed as the mean of four meas-urements using injections of saline (10 ml at 1-5 °C)arbitrarily performed during different phases of therespiratory cycle. Intrapulmonary shunt ( ), sys-temic vascular resistance (SVR), and pulmonary vas-cular resistance (PVR) were calculated using standardformula.

All blood samples were collected anaerobically,and analysed within 5 min (ABL 520, Radiometer,Copenhagen, DK). Arterial oxygen saturation (SaO2)and mixed venous oxygen saturation (SvO2) were

TS Q/Q

TS Q/Q


measured by spectrophotometry (OSM 3 Hemoxime-ter, Radiometer).

Induction of acute lung injury

A repeated lung lavage with isotonic saline was per-formed to produce lung surfactant depletion as report-ed and described in detail elsewhere (17). A PaO2/FIO2ratio persistently below 13 kPa (100 mmHg) for onehour was considered to be the onset of ALI.

Experimental protocol

After induction of ALI, animals were randomly as-signed to receive either a continuous inhalation of 20ppm NO (NO-group, n = 7), to receive PLV with 30ml⋅kg-1 of perfluorocarbons (PF 5080, 3M, Germany)(PLV-group, n = 7), or no further intervention (Con-trols, n = 8). Evaporative losses of PF 5080 were re-placed at a dose of 4±3 ml kg-1 every hour accordingto data obtained by our group previously (18). PF 5080(C8F18) is a non-ozone-depleting PFC with boilingpoint 102 °C, density 1.76 g ml-1 (at 25 °C), viscosity1.4 cp (at 25 °C), vapour pressure 6.8 kPa (at 37 °C),solubility of oxygen 49 ml 100 ml-1 (at 37 °C), solubil-ity of carbon dioxide 176 ml 100 ml-1 (at 37 °C) andsurface tension of 15 dynes cm-1 (at 25 °C) (3M datasheet).

NO was mixed to the inspired gas within a com-mercially available ventilator (Servo 300 A/NO,

Siemens-Elema, Solna, Sweden) from a stock of 1000ppm NO in nitrogen (AGA, Bottrop, Germany). ACLD 700 AL rapid responding chemiluminometer(ECO Physics, Duernten, Switzerland) was used fordetermination of the NO concentration applied in theinspiratory gas.

Statistical analysis

Results are expressed as mean ± SEM. The data wereobtained at baseline (pre-lavage), immediately afterthe induction of ALI (post-lavage) and at hourly inter-vals during four hours thereafter. Statistical analysiswas performed using SPSS for Windows 8.0 (SPSSInc., Chicago Illinois, USA). Differences betweengroups were evaluated with Kruskal-Wallis ANOVAfollowed by post-hoc comparisons using Mann Whit-ney U-test with Bonferroni correction. Statistical sig-nificance was assumed at p < 0.05.

Results

All animals were comparable with regard to bodyweight and pre-study conditions. Pre-lavage data ofpulmonary gas exchange and hemodynamics did notdiffer significantly between groups (Table 1, Table 2).

In all groups, induction of ALI resulted in an in-crease of concomitant with a decrease in PaO2.Cardiac output (CO), mean arterial blood pressure(MAP), central venous pressure (CVP), and pul-

TS Q/Q

Gas exchange parameters during the protocol. Baseline denotes pre-study conditions (not used for statistical comparison), ALI acute lung in-jury (post-lavage, prior to intervention), (* denotes p<0.05 compared to Controls, § p<0.05 compared to PLV)PaO2: arterial oxygen tension; SaO2: arterial oxygen saturation; AaDO2: alveolar-arterial oxygen difference, QS/QT: intrapulmonary right-to-left shunt

Group Baseline ALI 1h 2h 3h 4hPaO2 [mmHg] NO

PLVControls

512 ± 22572 ± 21555 ± 15

65 ± 553 ± 464 ± 4

107 ± 12*142 ± 5066 ± 5

108 ± 9*233 ± 54*56 ± 5

109 ± 10*§301 ± 36*57 ± 7

114 ± 15§318 ± 41*64 ± 9

SaO2 [%] NOPLVControls

98 ± 098± 098 ± 0

78 ± 575 ± 482 ± 5

92 ± 291 ± 282 ± 5

92 ± 2*97 ± 1*78 ± 4

92 ± 298 ± 0*76 ± 6

92 ± 398 ± 0*82 ± 7

AaDO2[mmHg] NOPLVControls

150 ± 2293 ± 22112 ± 12

550 ± 13§599 ± 8588 ± 7

519 ± 9*508 ± 50576 ± 9

511 ± 7*421 ± 53*589 ± 8

503 ± 8*§347 ± 34*577 ± 13

499 ± 15*§362 ± 39*580 ± 9

QS/QT [%] NOPLVControls

14 ± 210 ± 112 ± 2

51 ± 657 ± 347 ± 4

35 ± 342 ± 547 ± 5

36 ± 432 ± 4*50 ± 5

38 ± 429± 349 ± 6

37 ± 426 ± 444 ± 7

Table 1. Gas exchange parameters


monary capillary wedge pressure (PCWP) remainedstable, and the mean pulmonary artery pressure(MPAP) was increased in all animals (Table 1, Table2).

Gas exchange

In the NO-group an improvement in PaO2 from 65±5mmHg after induction of ALI to 114±15 mmHg after4 hours of treatment was found (p<0.05 vs. Controls at1-3 h; Figure 1, Table 1). In the PLV-group PaO2 in-creased from 53±4 mmHg after ALI to 318±41 mmHgfour hours after instillation of perfluorocarbons(p<0.05 vs. Controls and vs. NO; Figure 1, Table 1). Inthe NO-group decreased from 51±6 % after in-

duction of ALI to 37±4 % after four hours of treatment(Table 1). Decreases in shunt calculated as differencesfrom the onset of ALI were significant at 1-2 h (p <0.05 vs. controls, Figure 2). In the PLV-group changesin were significant when compared to Controls(p < 0.05 vs. Controls at 2-4 h after onset of ALI; Fig-ure 2, Table 1).

Hemodynamics

During the experiment no significant changes betweengroups were noted for HR, MAP, CVP, and PCWP(Table 2). In the NO-group MPAP was significantly re-duced (p<0.05 vs. Controls and vs. PLV,) (Fig. 3). Car-diac output remained stable in the NO-group, whereas

Table 2. Hemodynamics, oxygen delivery, an oxygen consumption

Group Baseline ALI 1h 2h 3h 4h

HR [bpm] NOPLVControls

83 ± 499 ± 491 ± 5

132 ± 12103 ± 11100 ± 13

121± 11100 ± 10106 ± 14

122 ± 1394 ± 8

109 ± 15

124 ± 1488 ± 6

110 ± 17

123 ± 1782 ± 688 ± 9

MAP [mmHg] NOPLVControls

103 ± 793 ± 591 ± 4

87 ± 489 ± 488 ± 5

90 ± 493 ± 779 ± 4

89 ± 595 ± 680 ± 4

85 ± 697 ± 785 ± 7

83 ± 691 ± 784 ± 8

MPAP[mmHg]

NOPLVControls

20 ± 224 ± 118 ± 2

36 ± 328 ± 231 ± 2

32 ± 327 ± 234 ± 2

34 ± 331 ± 136 ± 1

35 ± 435 ± 140 ± 3

35 ± 435 ± 139 ± 3

CVP [mmHg] NOPLVControls

9 ± 210 ± 18 ± 1

10 ± 214 ± 110 ± 1

11± 213 ± 110 ± 1

11 ± 214 ± 110 ± 2

12 ± 213 ± 111 ± 2

11 ± 213 ± 18 ± 1

PCWP[mmHg]

NOPLVControls

10 ± 113 ± 18 ± 2

10 ± 114 ± 110 ± 2

11 ± 216 ± 19 ± 2

11 ± 215 ± 010 ± 2

10 ± 214 ± 19 ± 2

11 ± 214 ± 17 ± 2

CO [l/min] NOPLVControls

4.1 ± 0.43.7 ± 0.53.7 ± 0.6

5.1 ± 0.44.8 ± 0.84.2 ± 0.8

4.6 ± 0.54.4 ± 0.64.4 ± 1

4.6 ± 0.54.0 ± 0.44.4 ± 0.9

5.1 ± 0.83.8 ± 0.53.8 ± 0.9

4.8 ± 0.7*3.2 ± 0.42.7 ± 0.1

PVR [dyn sec/cm5]

NOPLVControls

204 ± 14278 ± 67263 ± 43

418 ± 47279 ± 64484 ± 99

408 ± 64238 ± 43576 ± 114

436 ± 70335 ± 43621 ± 104

438 ± 67486 ± 51774 ± 138

439 ± 61*590 ± 74959 ± 90

DO2 [ml/min] NOPLVControls

513 ± 34409 ± 61424 ± 50

553 ± 83316 ± 49336 ± 73

588 ± 93*357 ± 49358 ± 71

594 ± 91*368 ± 47359 ± 85

634 ± 102*363 ± 38304 ± 42

617 ± 99*§311 ± 32262 ± 23

VO2 [ml/min] NOPLVControls

162 ± 22136 ± 15138 ± 13

212 ± 28130 ± 14148 ± 18

214 ± 29§128 ± 10156 ± 18

202 ± 26127 ± 11153 ± 10

213 ± 28§116 ± 6

151 ± 15

210 ± 38§108 ± 6128 ± 10

Hemodynamic parameters during the protocol. Baseline denotes pre-study conditions (not used for statistical comparison), ALI acute lung in-jury (post-lavage, prior to intervention),. MAP :mean arterial pressure, MPAP: mean pulmonary artery pressure, HR: heart rate, CVP: centralvenous pressure, PCWP: pulmonary capillary wedge pressure, CO: cardiac output, PVR: pulmonary vascular resistance, DO2: systemic oxy-gen delivery, VO2: oxygen uptake. (* denotes p<0.05 compared to Controls, § p<0.05 compared to PLV)

TS Q/Q

TS Q/Q


it was significantly decreased in Controls at4 h (p<0.05, Table 2).

Survival

In the PLV-group all animals survived until the end ofthe protocol. In the NO-group one animal died be-tween the third and the forth hour after ALI and in thecontrol-group, n = 4 animals died of acute hypoxemiaduring the study protocol, two between the second and

third hour, and two between the third and fourth hourafter induction of ALI, therefore n=4 controls werealive at the end of the protocol.

Discussion

In this study, we compared the effects of a continuousinhalation of 20 ppm nitric oxide with PLV on gas ex-change and hemodynamics in a model of ALI in juve-nile pigs. PLV was superior to iNO in improving gas

Figure 1. Values(mean SEM) ofPaO2 during theprotocol. ALI de-notes acute lunginjury (post-lavage, prior to in-tervention). (* denotes p<0.05compared to Con-trols, § p<0.05compared to PLV).

Figure 2. Changesin intrapulmonaryright-to-left shunt(Qs/Qt) (meanSEM) from onsetof ALI during theprotocol. ALI de-notes acute lunginjury (post-lavage, prior tointervention). (* denotes p<0.05compared to Con-trols).


exchange by reducing intrapulmonary right-to-leftshunt. However, only iNO was effective in reducingmean pulmonary artery pressure and in maintainingcardiac output after induction of ALI.

Inhaled NO and PLV have been evaluated com-bined or alone in different models of acute lung injury.Studies have been performed in surfactant deficientpreterm animals (14, 15), and in new-born animalswith ALI induced by surfactant wash-out (13, 16),meconium aspiration (12), and oleic acid induced lunginjury (11). Criteria for onset of ALI after surfactantwashout were only in two studies comparable to ourprotocol (13,16). The duration of PaO2/FIO2 < 100mmHg for one hour to indicate the induction of ALIwas specified only in the study by Houmes et al. (13).

Inhaled nitric oxide

Inhaled NO induces redistribution of pulmonary bloodflow due to a selective vasodilation of constricted pul-monary vessels in ventilated lung areas, thereby reduc-ing right-to-left shunt in experimental and humanARDS (19,20). Using MIGET in patients with severeARDS, Rossaint and co-workers demonstrated thatimprovement in arterial oxygenation is accompaniedby a reduction of blood flow in shunt regions duringinhalation of 18 ppm NO (19). Up to now, it remainsunclear which factors determine the NO response inacute lung injury. It has been speculated that oxygena-tion is improved by iNO merely if a certain vasocon-

striction is present in small arteries or veins of normal-ly ventilated lung areas (2). Recent investigations havedemonstrated that in early ARDS, the beneficial ef-fects of iNO on gas exchange are not sustained (21),and that a substantial number of patients do not re-spond favourably to inhaled NO (5).

During treatment relative changes in MPAP weresignificantly higher in the NO-group when comparedwith Controls and with the PLV-group. This finding isin accordance with clinical data (19). The decrease inMPAP in the NO-group was not accompanied by a de-crease in systemic blood pressure. This is explained bythe rapid inactivation of iNO by binding to hemoglo-bina (20).

Partial liquid ventilation with perfluorocarbons

The main mechanism by which PLV contributes to anincrease in PaO2 in ALI is alveolar recruitment and sta-bilisation of dependent lung regions (9). Furthermore,PLV might have vascular effects. The high densityPFCs, once filled into the lungs, might cause mechan-ical compression of shunt vessels, thereby causing aredistribution of blood flow away of dependent lungareas, towards better ventilated, non-dependent areas,as was suggested by Docter and colleagues (22). Thismight contribute to the increase in MPAP during PLV.In experimental studies from Curtis et al. in pigs (23)and from Hernan et al. in sheeps (24), an increase inMPAP and PVR was found during PLV.

Figure 3. Relativechanges in meanpulmonary arterypressure (MPAP)(mean SEM) fromonset of ALI dur-ing the protocol.ALI denotes acutelung injury (post-lavage, prior tointervention) (*denotes p<0.05compared to Con-trols, § p<0.05compared to PLV).


Morris et al. investigated PLV in healthy pigs with40 ml kg-1 of perfluorocarbons (25). They observedthat pulmonary blood flow was diverted from the de-pendent regions of the lung, associated with an in-crease in MPAP. The authors suggested that a greaterhydrostatic pressure gradient in PFC-filled alveoli,compared with the gradient in the blood vessels, couldcause these effects. In accordance to their findings, wefound a significant increase in oxygenation and a sig-nificant decrease in when compared to iNO andto Controls, concomitantly with an increase in MPAP.

Comparison of iNO and PLV

Barrington et al. (12) compared iNO and PLV aloneand combined in an experimental model of ALI withmeconium aspiration in newborn piglets. Inhalation of40 ppm NO alone was as effective as the combinedtreatment of PLV and NO in improving arterial oxy-genation during the first 120 minutes after induction ofALI. The treatment with PLV alone resulted in an im-provement in arterial oxygenation but this effect wasnot superior to iNO alone or the combined treatment.Regarding MPAP, no significant differences werefound between groups. The findings of Barrington etal. are difficult to interpret, since the iNO group andthe control group were ventilated with an FIO2 of 0.9,but for the other study groups no FIO2 was mentioned.Furthermore, in the PLV group and in the combinedtreatment group the PEEP levels were reduced from 4to 2 cmH2O.

Kinsella et al. (14) compared PLV (10 ml kg-1) andiNO (5 ppm), and conventional ventilation versus highfrequency oscillatory ventilation (HFOV) in surfactantdeficient premature lambs with ALI. The combinedtreatment was most effective to improve oxygenation.Comparing iNO and PLV, both treatments were effec-tive in improving arterial oxygenation when comparedwith the control group. However, between treatmentgroups no significant differences were found by theauthors (14).

In an oleic acid induced model of ALI, in rabbits,Uchida et al. (11), found no effect of 10 ppm iNOalone on oxygenation, even though pulmonary vascu-lar resistance was significantly decreased. During PLVa significant improvement in oxygenation was found,which significantly enhanced when combined withiNO. Therefore, in this specific animal model, a pre-conditioning for iNO might be neccessary to producebeneficial effects on arterial oxygenation. Houmes et

al. studied the effects of a combined treatment of iNOand PLV on gas exchange and hemodynamics in ex-perimental ALI (13). Animals received 5ml kg-1 PFCfollowed by incremental doses of 10 to 40 ppm iNOover 10 minutes. After receiving the maximum doseiNO, the gas was discontinued and animals receivedan additional bolus of 5 ml kg-1 PFC, after which iNOwas resumed. This procedure was repeated up to cu-mulative dosis of 20 ml kg-1 PFC. The authors foundthat a combination of 20 ml kg-1 PFC and 20 to 40 ppmiNO were most effective to increase oxygenation andto decrease MPAP. The lack of a control group and theshort time between interventions in this study providesno conclusive information whether a combination ofboth treatments might be more effective than eithertreatment alone.

In conclusion, using this experimental model ofALI in surfactant depleted juvenile pigs, we found thattreatment with PLV alone was more effective to im-prove oxygenation and to reduce when com-pared to 20 ppm inhaled NO. However, iNO reducedMPAP and stabilised cardiac output. In view of the lit-erature, in particular the fact that either treatmentfailed to demonstrate improvements in mortality (21,26), it remains to be answered wether a combination ofboth treatments might improve clinical outcome.

References

1. Lewis JF, Jobe AH. Surfactant and the adult respiratory distresssyndrome. Am Rev Respir Dis 1993; 147: 218-233

2. Zapol WM, Rimar S, Gillis N, Marletta M, Bosken CH. Nitricoxide and the lung. Am J Respir Crit Care Med 1994; 149: 1375-1380

3. Manktelow C, Bigatello LM, Hess D, Hurford WE. Physiologicdeterminants of the response to inhaled nitric oxide in patientswith acute respiratory distress syndrome. Anesthesiology 1997;87: 297-307

4. Marshall BE, Hanson W, Frasch F, Marshall C. Role of hypoxicpulmonary vasoconstriction in pulmonary gas exchange andblood flow distribution. Intensive Care Med 1994; 20: 379-389

5. Puybasset L, Rouby JJ, Mourgeon E, Cluzel P, Souhil Z, Law-Koune JD, Stewart T, Devilliers C, Lu Q, Roche S, Kalfon P,Viscaut E, Viars P. Factors influencing cardiopulmonary effectsof inhaled nitric oxide in acute respiratory failure. Am J RespirCrit Care Med 1995; 152: 318-328

6. Hernan LJ, Fuhrman BP, Kaiser RE, Penfil S, Foley C, PapoMC, Leach CC. Perfluorocarbon-associated gas exchange innormal and acid-injured large sheep. Crit Care Med 1996; 24:475-481

7. Fuhrman BP, Paczan PR, DeFran M. Perfuorocarbon associatedgas exchange. Crit Care Med 1991; 19: 712-722

8. Tütüncü AS, Faithfull NS, Lachmann B. Comparison of ventila-tory support with intratracheal perfluorocarbon administration

TS Q/Q

TS Q/Q


and conventional mechanical ventilation in animals with acuterespiratory failure. Am Rev Respir Dis 1993; 148: 785-795

9. Papo MC, Paczan PR, Fuhrmann BP, Steinhorn DM, Hernan LJ,Leach CL, Holm BA, Fisher JE, Kahn BA. Perluorocarbon-as-sociated gas exchange improves oxygenation, lung mechanics,and survival in a model of adult respiratory distress syndrome.Crit Care Med 1996; 24: 466-474

10. Lim CM, Koh Y, Jung BO, Lee SD, Kim WS, Kim DS, KimWD. An optimal dose of perfluorocarbon for respiratory me-chanics in partial liquid ventilation for dependent lung-dominantacute lung injury. Chest 2000; 117: 199-204

11. Uchida T, Nakazawa K, Yokoyama K, Makita K, Amaha K. Thecombination of partial liquid ventilation and inhaled nitric oxidein the severe oleic acid lung injury model. Chest 1998; 113:1658-1666

12. Barrington KJ, Singh AJ, Etches PC, Finer NN. Ventilation withand without inhaled nitric oxide in a newborn piglet model ofmeconium aspiration. Am J Respir Crit Care Med 1999; 160:1922-1927

13. Houmes RJM, Hartog A, Verbrugge SJC, Böhm S, Lachmann B.Combining partial liquid ventilation with nitric oxide to improvegas exchange in acute lung injury. Intensive Care Med 1997; 23:163-169

14. Kinsella JP, Parker TA, Galan H, Sheridan BC, Abman SH. In-dependent and combined effects of inhaled nitric oxide, liquidperfluorochemical, and high-frequency oscillatory ventilation inpremature lambs with respiratory distress syndrome. Am JRespir Crit Care Med 1999; 159: 1220-1227

15. Göthberg S, Paker TA, Abman SH, Kinsella JP. High-frequencyoscillatory ventilation and partial liquid ventilation after acutelung injury in premature lambs with respiratory distress syn-drome. Crit Care Med 2000; 28: 2450-2456

16. Zobel G, Urlesberger B, Dacar D, Rodl S, Reiterer F, Friehs I.Partial liquid ventilation combined with inhaled nitric oxide inacute respiratory failure with pulmonary hypertension in piglets.Pediatr Res 1997; 41: 172-177

17. Kaisers U, Max M, Walter J, Kuhlen R, Pappert D, Falke K,Rossaint R. Partial liquid ventilation with small volumes of FC3280 increases survival time in experimental ARDS. Eur RespirJ 1997; 10: 1955-1961

18. Kaisers U, Max M, Schnabel R, Böhm S, Hendrik E, Rossaint R,Lachmann B. Partial liquid ventilation with FC 3280 in experi-mental lung injury: dose-dependent improvement of gas ex-change and lung mechanics. Applied Cardiopulm Pathophysiol1996; 6: 163-170

19. Rossaint R, Falke KJ, Lopez F, Slama K, Pison U, Zapol WM.Inhaled nitric oxide for the adult respiratory distress syndrome.N Engl J Med 1993; 328: 399-405

20. Rossaint R, Gerlach H, Falke KJ. Inhalation of nitric oxide – anew approach in severe ARDS. Eur J Anaesthesiology 1994; 11:43-51

21. Dellinger RP, Zimmerman JL, Taylor RW, Straube RC, HauserDL, Criner GJ, Davis K Jr, Hyers TM, Papadakos P. Effects ofinhaled nitric oxide in patients with acute respiratory distresssyndrome: results of a randomized phase II trial. Inhaled NitricOxide in ARDS Study Group. Crit Care Med 1998; 26: 15-23

22. Doctor A, Ibla JC, Grenier BM, Zurakowski D, Ferretti ML,Thompson JE, Lillehei CW, Arnold JH. Pulmonary blood flowdistribution during partial liquid ventilation. J Appl Physiol1998; 84: 1540-1550

23. Curtis SE, Fuhrman BP, Howland DF. Cardiac output during liq-uid (perfluorocarbon) breathing in newborn piglets. Crit CareMed 1991; 19: 225-230

24. Hernan LJ, Fuhrman BP, Kaiser RE. Perfluorocarbon-associatedgas exchange in normal and acid-injured large sheep. Crit CareMed 1996; 24: 475-481

25. Morris KP, Cox PN, Mazer CD, Frndova H, McKerlie C, WolfeR. Distribution of pulmonary blood flow in the perfluorocarbon-filled lung. Intensive Care Med 2000; 26: 756-763

26. www.allp.com, last entered July, 15.2001

Supported by a grant of Deutsche Forschungsgemein-schaft (KA 1212/3-1)

Address for corresponding: Udo Kaisers, M.D., Klinik und Polik-linik für Anästhesiologie und Intensivtherapie, UniversitätsklinikumLeipzig, Liebigstr. 20, D-04103 Leipzig, Germany, E-Mail: [email protected]

Matthias Glanemann, Thomas Henneberg

Therapiestandards der chirurgischen Intensivstation

160 Seiten, ISBN 978-3-89967-402-6, Preis: 20,- Euro

PABST SCIENCE PUBLISHERSEichengrund 28, 49525 Lengerich,Tel. ++ 49 (0) 5484-308, Fax ++ 49 (0) 5484-550,E-Mail: [email protected]: www.pabst-publishers.de

Das Buch beinhaltet neben intensivmedizinischenGrundlagen eine Ansammlung operationsspezifisch-er Behandlungspläne, die zu einem besseren Ver-ständnis und zur Vereinheitlichung der postoperativ-en Therapie führen sollen. Es richtet sich vor allem anNeuankömmlinge auf der Intensivstation, um die er-sten Tage und Wochen klarer und hoffentlich stress-freier zu erleben. So ist das Buch in drei Abschnittegegliedert. Im ersten Teil soll grundlegendes inten-sivmedizinisches Verständnis vermittelt werden,wohingegen im zweiten Teil besonderes Augenmerkauf die chirurgischen Besonderheiten und postopera-tiven Implikationen der einzelen Operationen gelegtwird. Im dritten Abschnitt werden die wichtigstenNotfälle beschrieben.

Lung temperature during CPB 11

Introduction

Respiratory dysfunction can be measured in virtuallyall patients undergoing cardiopulmonary bypass[CPB] [1]. Although usally mild [2], it accounts for se-vere postoperative morbidity in a significant numberof patients [3]. The etiology of post-CPB respiratorydysfunction is multifactorial [4], but ischemia of thelungs appears to contribute to its development [5].

Hypothermia decreases oxygen consumption, andis therefore an effective way to protect organs againstischemic damage. Likewise, lung hypothermia mayameliorate post-CPB lung function [6,7]. Hypothermic

CPB itself may be hypothesized to result in lung hy-pothermia because of the dense vascularisation of thelung, but no data are available on the characteristics ofthe lung temperature during CPB. However, these da-ta are needed to properly target interventions aimed ateither the induction or maintenance of lung hypother-mia. We have performed the following study firstly inorder to study the characteristics of the lung tempera-ture during systemic hypothermia; secondly we aimedto evaluate whether additional hypothermic ventilationshows any effect on the lung temperature.

Lung temperature during cardiopulmonary bypass and the effect of hypothermic ventilation

J. F. M. Bechtel1, W. Eichler2, M. Großherr2, H.-H. Sievers1, C. Bartels1

1Department of Cardiac Surgery and 2Department of Anesthesiology and Intensive Care, University HospitalLuebeck, Luebeck, Germany


Keywords: lung injury, hypothermia, cardiopulmonary bypass, helox

Abstract

Background: Lung hypothermia may be hypothesized to ameliorate lung function following cardiopulmonary by-pass [CPB]. Little is known about the characteristics of lung temperature during CPB. We therefore studied lungtemperature during hypothermic CPB. In addition, we investigated whether hypothermic ventilation has any in-fluence on the lung temperatureMethods: Pigs were submitted to moderate systemic hypothermia induced by cardiopulmonary bypass. Duringcardiac arrest (45 minutes), CPB-circuit temperature was gradually increased. During CPB, 5 pigs had hypother-mic ventilation with 80/20 helium/oxygen whereas 6 pigs had normothermic ventilation.Results: Lung temperature dropped to its nadir (normothermic ventilation group: 28.5 ± 0.9; hypothermic venti-lation group: 27.9 ± 0.9) in parallel to the decline of the CPB-circuit temperature. Rectal temperature decreasedsignificantly less than lung temperature (p = 0.001). Rewarming of the CPB-circuit induced almost immediate re-warming of the lungs. Hypothermic ventilation showed no effect on either lung or rectal temperature.Conclusions: The course of lung temperature during total cardiopulmonary bypass closely mirrors the tempera-ture in the CPB-circuit. These data might be helpful for future studies planning to evaluate the effect of hypother-mia on post-CPB lung function.

AbbreviationsACT: activated clotting time; bw: body-weight; CPB: cardiopulmonary bypass; NIH: National Institute of Health

12 J. F. M. Bechtel, W. Eichler, M. Großherr, H.-H. Sievers, C. Bartels

Methods

The experimental protocols used in the present studywere approved by the department for environment, na-ture, and forestry of the land Schleswig-Holstein, Ger-many. All animals received humane care in compli-ance with the „Guide for the Care and Use of Labora-tory Animals“ published by the National Institutes ofHealth [NIH publication 85-23, revised 1985].

The device for cooling of the inspired gas is shownin figure 1. In brief, a non-Fluor-Chlor-hydrocarboncoolant circulates through the device. Contact with theinspiration gas is maximized by helix-figuration of thegas mains. A Y-connector between the cooling deviceand the intratracheal tube separates in- and expiredgas. The ability of the cooling device to effectivelycool the gas was demonstrated by in vitro eperiments(data not shown). A 80% helium /20% oxygen mixture[Helox; Messer-Griesheim, Frankfurt/Mn., Germany]was chosen for hypothermic ventilation because it fea-tures an approximately 5 times higher thermal conduc-tivity than air [8].

Female pigs were used in a model of hypothermicCPB. In the control group, normothermic ventilationwas applied (n=6, 73.8 ± 14.1 kg). In the other group,hypothermic ventilation with Helox was used (n=5,76.2 ± 8.6 kg; p=ns). Premedication consisted of intra-muscular ketamine (15 mg/kg body-weight [bw]), 2mg/kg bw xylazine, and 0.5 mg atropine. Generalanesthesia was induced by an intravenous bolus injec-tion of sufentanil (0.3 µg/kg bw) plus propofol (3mg/kg bw) plus pancuronium (0.1 mg/kg bw), and waslater maintained with intravenous sufentanil (1 µg/kgbw/h) and propofol (3-5 mg/kg bw/h). The pigs wereintubated with an endotracheal tube through a tra-cheostomy. Tracheostomy was chosen in order to min-imize the length of the ventilation tubes. Mechanicalventilation [Dräger, Lübeck, Germany] was initiated(20 % oxygen, PEEP 5-7 mmHg), and the ventilationparameters (tidal volume 10 ml/kg bw; ventilation rate10-15 /min) were adapted to maintain physiologic pHand pCO2. Arterial blood gases were monitored peri-odically. A median thoracotomy was performed. Thepleurae were opened, and a temperature probe [My-ocardial temperature needle 15 mm; Medtronic, En-glewood, CO, USA] was inserted into the middle lobeof the right lung. Rectal temperature was measured.Heparin was given, initially 300 IU/kg bw, and the ac-tivated clotting time [ACT] was monitored periodical-ly. Cardiopulmonary bypass was initiated [membraneoxygenator: Cobe Laboratories, Planegg-Martinsried,Germany; roller pump: Stöckert, München, Germany]and conducted as follows: The temperature of the CPBheat-exchanger [System S3; Stöckert, München, Ger-many] initially was 30°C and lowered to 20°C afterfull pump flow was achieved (“cooling phase”). Onthe occurrence of cardiac fibrillation, cardiac arrestwas induced by use of cristalloid cardioplegia (St.-Thomas solution). After aortic clamping, the tempera-ture of the CPB heat exchanger was raised to approxi-mately 32°C and – in a stepwise fashion – furtherraised to 40°C shortly before declamping (“warmingphase”). Cardioplegia was repeated at 25 minutes.Pump flow was 2.0 - 2.5 l/min/m2. At the time of theinitiation of CPB, tidal volume was reduced to onethird in the normothermic ventilation group, whereashypothermic ventilation (5°C) with Helox was initiat-ed in the other group without change of other ventila-tion parameters. After 45 minutes aortic clamping wasreleased; CPB was discontinued after reperfusion, andthe pigs were sacrified shortly thereafter.

Figure 1. The cooling device is connected to the intra-tracheal tube. A non-Fluor-Chlor-hydrocarboncoolant circulates through the device and the heat ex-changer (not shown). Contact with the inspira-tion gasis maximized by helix-figuration of the gas mains. A Y-connector between the cooling device and the intratra-cheal tube separates in- and expired gas.



Continuous data are reported as mean ± standard devi-ation. Independent, not-normally distributed continu-ous data were compared using the Mann-Whitney Urank-sum test. Analysis of variance for repeated meas-ures was used to analyze the effect of group and timeon temperature. Post hoc analysis within groups wasperformed by paired t-test (temperature) or Wilcoxontest (time). A two-sided p < 0.05 was considered statis-tically significant. Statistical analysis was performedwith SPSS for Windows [Chicago, Il., USA].

Results

The animals in both groups did not differ significantlyin either initial lung (36.4 ± 1.0 vs. 35.8 ± 0.6 °C) orrectal (37.5 ± 0.8 vs. 36.8 ± 1.1 °C) temperature.

Lung temperature, effect of hypothermic ventilation

Figure 2 shows the change of the lung temperatureover time. There was no significant effect of hypother-mic ventilation (p = 0.97) on the pulmonary tempera-ture.

Lung temperature, effect of time

The lung temperature was significantly dependent ontime (p = 0.001). During the cooling phase a signifi-cant decline (p < 0.001) of the lung temperature wasobserved. Within 5 minutes (median 5 minutes) afterclamping of the aorta (i.e. end of the cooling phase)the lowest lung temperature (normothermic ventilationgroup, 28.5 ± 2.2; hypothermic ventilation group, 27.9± 2.0 °C) was measured. The lung temperature after 45minutes of CPB was significantly higher (normother-mic ventilation group, 32.5 ± 2.2; hypothermic venti-lation group, 31.7 ± 0.6 °C) than the temperature at thetime of aortic clamping (p < 0.001) and the lowest ob-served temperature (p < 0.001). The temperature in thearterial line of the CPB was recorded and is also de-picted in figure 2. The decline of the lung temperatureis strictly parallel to the decline of the perfusion tem-perature. Lung rewarming starts almost immediatelyafter raising of the CPB temperature.

Rectal temperature, effect of hypothermic ventilation

The rectal temperature is shown in figure 3. There wasno significant overall effect of hypothermic ventilation(p = 0.44) on the rectal temperature.

Rectal temperature, effect of time

A highly significant effect of time (p < 0.001) on therectal temperature was observed. The lowest rectaltemperature (normothermic ventilation group, 31.4 ±0.6; hypothermic ventilation group, 31.1 ± 0.8 °C) wasreached 5 to 30 (median 15) minutes after aorticclamping, which is significantly longer than the timeto the lowest lung temperature (p < 0.01). The rectaltemperature dropped significantly less than the lungtemperature (p = 0.001).

Figure 3 shows the CPB temperature, too. As withthe lung temperature, the decline of the rectal temper-ature is parallel to the decline of the perfusion temper-ature. However, rewarming of the rectal temperatureoccurs with a significant delay (p=0.003) as comparedto the CPB and the lung temperature.

Figure 2. Lung temperature in control animals [ ]and following hypothermic ventilation with Helox [ ],and CPB temperature [ ]. Hypothermic ventilationstarted with the induction of CPB and ended at t=45.A significant effect of time (p < .001) but not of hy-pothermic ventilation on the lung temperature is ob-served.


Discussion

Our study provides the first systematic description ofthe characteristics of the lung temperature during hy-pothermic CPB. It shows that the lung temperature isclosely linked to the perfusion temperature during bothcooling and rewarming. Therefore, lung temperaturestarts to rewarm as soon as the temperature of the CPBheat exchanger is raised in order to avoid excessivesystemic cooling. In addition, we failed to demonstratean effect of hypothermic ventilation on the lung tem-perature.

Lung dysfunction following CPB remains an im-portant clinical problem [3,9]. Its pathogenesis is mul-tifactorial [4] and includes reduced cardiac function[2], the inflammatory response caused by exposure ofthe blood to the artificial surface of the CPB circuit[10,11], the oxygen concentration used [12], as well asischemia-reperfusion injury [5,13].

Hypothermia, by decreasing oxygen consumption,is an effective way to protect organs against ischemicdamage. Hypothermia may also be advantagousagainst the inflammatory component of post-CPB lungdysfunction, because neutrophil sequestration wasshown to be involved in the pathogenesis of post-CPB

pulmonary dysfunction [7,14]. Our data, however,suggest that the lungs are not sufficiently cooled andmaintained cold during standard hypothermic CPB be-cause rewarming of the partly ischemic lungs startsimmediately after the temperature of the CPB heat ex-changer is raised. If one aims at lung hypothermia inorder to evaluate its effects on postoperative lungfunction, additional cooling seems therefore warrant-ed.

Due to the unique anatomy of the lung, hypother-mia can be applicated by either the airway or vesselsystem or by immersion in cold saline. The last, how-ever, cannot be expected to induce homogenous cool-ing of the lung and may cause phrenic nerve injury[15]. Selective pulmonary artery perfusion, as done bysome groups [6], has several disadvantages: it can on-ly be used intermittently, makes a rather small opera-tive field even smaller (due to the additional cannula-tion and tubing) and will hardly work with minimallyinvasive strategies. Induction or maintenance of pul-monary hypothermia via the airways is therefore of in-terest. However, the potential use of the respiratorytract as a heat exchanger is severely limited by power-ful physiologic mechanisms that maintain temperatureand moisture of the lung within narrow margins. Theinspired air is warmed and enriched with water, where-as the expired air is cooled, and water is withdrawn[16]. This mechanism can theoretically be influencedat different levels, e.g.: change of specific heat, ther-mal conductivity or water content of the inspired gas,change of the minute-ventilation and lowering of theinspiration temperature, or separation of in- and ex-pired gas.

Prior attempts of hypothermic ventilation havebeen largely disappointing. Harrison et al. andDougherty et al. demonstrated that lethal pulmonaryedema might develop following ventilation with cold,dry gases [17,18]. Whether direct thermal injury or de-hydration with consecutive cellular damage [19] orimpairment of surfactant [20] caused the pulmonarydamage was not determined. Zikria and co-workers[21], on the other hand, were able to induce a signifi-cant decline of the rectal temperature by means of hy-pothermic ventilation, but found it to be very ineffi-cient. The lungs did not suffer thermal injury, but lungtemperature was not reported.

Besides the systematic description of the lung tem-perature, the aim of our study was not induction of sys-temic hypothermia by hypothermic ventilation butcooling of the lung. We tried to overcome the physio-logic mechanisms that protect the lung from becoming

Figure 3: Rectal temperature in control animals [ ]and following hypothermic ventilation with Helox [ ],and CPB temperature [ ]. Hypothermic ventilationstarted prior with the induction of CPB and ended att=45. A significant effect of time (p < .001) but not ofhypothermic ventilation on the rectal temperature isobserved.


hypothermic by altering specific heat and thermal con-ductivity of the inspired gas. We could not demon-strate a significant effect of hypothermic ventilationwith Helox on the lung temperature. However, furthermodifications, e.g. exchanging gas for liquids [22,23],which feature a much higher thermal conductivity,might allow efficient maintenance of pulmonary hy-pothermia during systemic rewarming.

Limitations of the study

Lung temperature was measured by a single tempera-ture probe. The underlying assumption of homogenouscooling/rewarming may not be adequate, but evenmultiple probes could not have exterminated this prob-lem completely.

The reduction of the tidal volume in the controlgroup might have reduced the comparability of the twogroups since a recent paper suggests that bronchial ar-tery flow is decreased with decreased lung volumes[13]. However, the decision to decrease the tidal vol-ume in the control group was based on the fact that thisis commonly done in clinical practice. Because weaimed to compare “standard practice” to a potential in-tervention (hypothermic ventilation), we also had cho-sen not to administer Helox to the animals of the con-trol group.

No studies on lung function were performed. Wedecided to concentrate on the feasibility of lung cool-ing rather than its effect on lung function because ofthe known difficulties associated with performing hy-pothermic ventilation [17,18,21]. The cooling devicewas highly effective in an ex vivo model of lung cool-ing. We are unaware of studies showing a protectiveeffect of Helox on lung function. Given our negativefindings about maintaining lung hypothermia by hy-pothermic ventilation with Helox, an effect on lungfunction is highly unlikely.

In conclusion, we have shown that any change ofthe temperature of the CPB-circuit is closely mirroredby the lung temperature and that hypothermic ventila-tion with Helox had no significant effect on the lungtemperature. Our finding that rewarming of the lungsstarted immediately after the temperature of the CPBheat exchanger was raised may be helpful in targetingand evaluating strategies for the use of hypothermiafor lung protection during operations with CPB.

Perspective

Post-CPB lung dysfunction is a multifactorial prob-lem, and we believe that evidence is accumulating thatlung ischemia is contributing to its pathophysiology.Hypothermia has worked so well in protecting is-chemic organs that it is tentative to speculate that hy-pothermia might also protect the lungs. Unfortunately,almost nothing is known about the lung temperatureduring standard, mildly hypothermic CPB. Because ofthe dense vascularization of the lungs, some cooling ofthe lungs has to be expected during hypothermic CPB.However, without exact knowledge about the lungtemperature during CPB, targeting interventionsaimed at either induction or maintainance of lung hy-pothermia is impossible.

We hope that our manuscript promotes further re-search on the pathophysiology and prevention of post-CPB lung dysfunction. We believe that interventionsdesigned to target lung ischemia should aim at main-taining lung hypothermia during systemic rewarming.We did not observe any cooling of the lungs althoughwe were able to increase the thermal conductivity ofthe inspired air severalfold. We therefore speculatethat hypothermic ventilation with gases is no suitablemethod for lung cooling. However, we believe that theairways are an attractive candidate for interventionsbecause all local measures (e.g. selective hypothermicperfusion) increase the total perfusion time and de-crease the surgeon’s overview over the operative field.We speculate that liquid hypothermic ventilation willeasily induce and maintain lung hypothermia and lookforward to studies on its effect on post-CPB lung func-tion.

Acknowledgements

We are indebted to Elrina Joubert-Hübner, PeterHenke, Joachim Prosch, Ralf Lisson, and ChristianDörnbrack for conducting the cardiopulmonary by-pass; and to Wolfgang Kloess, MSc, for design andproduction of the cooling device.

References

1. Taggart DP, El-Fiky M, Carter R et al. Respiratory DysfunctionAfter Uncomplicated Cardiopulmonary Bypass. Ann ThoracSurg 1993; 56: 1123-8

2. Weiss YG, Merin G, Koganov E et al. Postcardiopulmonary by-pass hypoxemia: a prospective study on incidence, risk factors,


and clinical significance. J Cardiothorac Vasc Anesth 2000; 14:506-13

3. Hammermeister KE, Burchfiel C, Johnson R et al. Identificationof Patients at Greatest Risk for Developing Major Complica-tions at Cardiac Surgery. Circulation 1990; 82 [suppl IV]: IV-380-IV-389

4. Asimakopoulos G, Smith PLC, Ratnatunga CP et al. Lung injuryand acute respiratory distress syndrome after cardiopulmonarybypass. Ann Thorac Surg 1999; 68: 1107-15

5. Friedman M, Sellke FW, Wang SY et al. Parameters of Pul-monary Injury After Total or Partial Cardiopulmonary Bypass.Circulation 1994; 90 [suppl II]: II-262-II-268

6. Liu Y, Wang Q, Zhu X et al. Pulmonary artery perfusion withprotective solution reduces lung injury after cardiopulmonarybypass. Ann Thorac Surg 2000; 69: 1402-7

7. Le Deist F, Menasche P, Kucharski C et al. Hypothermia duringcardiopulmonary bypass delays but does not prevent neutrophil-endothelial cell adhesion. A clinical study. Circulation 1995; 92[suppl II]: II-354-II-358

8. Beran AV, Shinto RA, Proctor KG et al. Effect of inhalate ther-mal conductivity and high O2 in producing hypothermia. J Ap-pl Physiol 1979; 47: 228-32

9. Messent M, Sullivan K, Keogh BF et al. Adult respiratory dis-tress syndrome following cardiopulmonary bypass: incidenceand prediction. Anesthesia 1992; 47: 267-8

10. Morioka K, Muraoka R, Chiba Y et al. Leukocyte and plateletdepletion with a blood cell separator: effects on lung injury aftercardiac surgery with cardiopulmonary bypass. J Thorac Cardio-vasc Surg 1996; 111: 45-54

11. Gu YJ, deVries AJ, Boonstra PW et al. Leukocyte Depletion Re-sults In Improved Lung Function And Reduced InflammatoryResponse After Cardiac Surgery. J Thorac Cardiovasc Surg1996; 112: 494-500

12. Pizov R, Weiss YG, Oppenheim-Eden A et al. High oxygen con-centration exacerbates cardiopulmonary bypass-induced lung in-jury. J Cardiothorac Vasc Anesth 2000; 14: 519-23

13. Chai PJ, Williamson JA, Lodge AJ et al. Effects of ischemia onpulmonary dysfunction after cardiopulmonary bypass. Ann Tho-rac Surg 1999; 67: 731-5

14. Dreyer WJ, Michael LH, Millman EE et al. Neutrophil seques-tration and pulmonary dysfunction in a canine model of openheart surgery with cardiopulmonary bypass. Evidence for aCD18-dependent mechanism. Circulation 1995; 92: 2276-83

15. Tripp HF, Bolton JWR. Phrenic nerve injury following cardiacsurgery: a review. J Card Surg 2001; 13: 218-23

16. Walker JEC, Wells Jr. RE, with the collaboration of E.W.Merrill.Heat and Water Exchange in the Respiratory Tract. Am J Med1961; 30: 259-67

17. Harrison MR, Hysing ES, Bø G. Control of Body Temperature:Use of the Respiratory Tract as a Heat Exchanger. J Ped Surg1977; 12: 821-18

18. Dougherty JC, Sinha S, Kibble F et al. Intolerance of the is-chemic lung to hypothermic ventilation. J Appl Physiol 1972;32: 632-4

19. Marfatia S, Donahoe PK, Hendren WH. Effect of Dry and Hu-mified Gases on the Respiratory Epithelium in Rabbits. J PedSurg 1975; 10: 583-92

20. Erasmus ME, Petersen AH, Oetomo SB et al. The Function ofSurfactant is Impaired during the Reimplantation Response inRat Lung Transplants. J Heart Lung Transpl 1994; 13: 791-802

21. Zikria BA, Ferrer JM, Malm JR. Pulmonary hypothermia indogs. J Appl Physiol 1968; 24: 707-10

22. Forman DL, Bhutani VK, Tran N et al. A New Approach to In-duced Hypothermia. J Surg Res 1986; 40: 36-42

23. Cheifetz IM, Cannon ML, Craig DM et al. Liquid ventilationimproves pulmonary function and cardiac output in a neonatalswine model of cardiopulmonary bypass. J Thorac CardiovascSurg 1998; 115: 528-35

Address for correspondence: J. F. Matthias Bechtel, M.D., Klinik fürHerzchirurgie, Universitätsklinikum Schleswig-Holstein, CampusLübeck, Ratzeburger Allee 160, D-23538 Lübeck, Germany, E-Mail: [email protected]

I. Uhlenbusch-Körwer, E. Bonnie-Schorn, A. Grassmann, J. Vienken

Understanding Membranesand Dialysers616 pages ISBN 978-3-89967-005-9, Price: 48,- Euro


The dialyser and its centrepiece, the membrane,play a central role in chronic renal replacementtherapy. More than 1000 different types of dialy-sers are currently available on the market. Howare these characterised and how do they differ?What are their special features and how do theycontribute to acute and chronic effects in the pa-tient? Is there a link between morbidity and mor-tality and the use of a particular type of dialyser?This book addresses these questions and attemptsanswers based on current scientific knowledge. Inthis context, dialyser development and the basicsof filter performance and biocompatibility assess-ment are reviewed.

Thoracic epidural anaesthesia as part of a fasttrack-concept for cardiac surgery 17

Background

In the last couple of years the importance of savingcosts and using existing resources to achieve the bestpossible effectiveness has increased drastically andhas led to new therapeutic concepts in all clinical dis-ciplines. Especially in cardiac surgery this develop-ment can be observed, since these procedures have al-ways been associated with high costs (1-4).

In the seventies anesthesia in cardiac surgery wasmainly performed as inhalative anesthesia. This would

principally have made an early extubation possible,but since saving costs was not a priority goal, early ex-tubation was rarely performed.

During the eighties a more opioid-based anesthesiawas practiced in cardiac surgery, since this concept ledto better hemodynamic stability, even in patients withpoor cardiac conditions (5, 6). However, this regimerequired postoperative assisted ventilation of 12 to 24h which led to an increased need for ICU beds. Thehigher costs that were associated with this strategywere not questioned, though there have been huge de-

Thoracic epidural anaesthesia as part of a fasttrack-concept for cardiacsurgery

C. D. Kratz1, U. Schirmer2, S. Weiss1, H. Wulf3, L. Eberhart3, S. Vogt4, G. Geldner1

1Clinic for Anaesthesia and Intensive Care, Klinikum Ludwigsburg, Germany; 2Clinic for Cardiothoracic Anaes-theology, University Ulm, Germany; 3Clinic for Anaesthesia and Intensive Care, Philipps-University Marburg,Germany; 4Clinic for Cardiac Surgery, Philipps-University Marburg, Germany


Keywords: fasttrack concept, fasttrack-protocol, cardiac anaesthesia, thoracic epidural catheters for ACB surgery

Abstract

Background and Goal of Study: This feasibility-study was performed to test the hypotheses that a fast-trackingprotocol including high thoracic epidural analgesia speeds discharge from ICU in patients undergoing aortic coro-nary bypass surgery (CABG).Materials and Methods: 38 patients treated according to a fast-tracking protocol were compared with 41 matchedcontrol patients. The latter received general anaesthesia with propofol, sufentanil and postoperative opioid anal-gesia on demand. The fast-tracking protocol includes a thoracic epidural catheter inserted at levels of Th1-Th4 theday before surgery. Anaesthesia was induced and maintained with propofol. Analgesia was provided by continu-ous epidural infusion of ropivacaine and sufentanil. Postoperative care was provided by the cardiac surgeons on aspecialized ICU. Times until tracheal extubation and discharge from ICU were defined as surrogate markers forfast-tracking eligibility.Results: Due to a strict matched pair’s technique, patients in both groups did not differ significantly with respectto any biometric or clinical variables. Times until tracheal extubation (6 hours for the fast-tracking group, 20 hoursfor the control group) was significantly shorter in patients receiving the fast-tracking protocol (p=0.0002; U-test).However, discharge from ICU was not faster in these patients (67 hours for the fast-tracking group, 85 hours forthe control group) (p=0.64). All data are shown as median and 25th-75th percentile (in brackets).Conclusion: A fast-tracking protocol including high thoracic epidural anaesthesia allows early tracheal extuba-tion. This advantage was well accepted by the attending cardiac surgeons but not transferred into faster dischargefrom ICU since the treatment pathways weren’t also changed.

18 C. D. Kratz, U. Schirmer, S. Weiss, H. Wulf, L. Eberhart, S. Vogt, G. Geldner

velopments in the areas of drugs, anesthetic techniquesand postoperative analgetic regimes.

During the last years different fasttrack-conceptsfor cardiac surgery have been developed, which aim atdecreasing the duration of endotracheal intubation andspeeding the discharge from the ICU. Those newregimes have been created to save costs and at thesame time lower the rate of nosocomial infections andventilator-associated pneumonias.

During the nineties epidural catheters became partfor analgesia during surgery and the postoperative pe-riod. This technique makes use of the advantageous ef-fects on coronary blood flow (7, 8), left ventricularfunction (9), relief of angina (10,11) hemodynamicstability (12,13), attenuation of stress response hor-mones (14-16), and surrogate markers of respiratoryfunction (17-19).

There has been a widespread trend toward morerapid recovery after CABG, with earlier extubationand shorter stays in intensive care and in the hospital(fast-track cardiac anesthesia and surgery), but onlylately concepts have been reviewed, that use epiduralanalgesia as a part of a fasttrack-concept.

Since then, several studies have established thatpatients who received an epidural catheter for cardiacsurgery could in many cases be extubated on the tableor at least an early extubation on ICU was possible(20-22).

However the general perioperative management ofmost studies did not attempt to fast-track patients.

Goal of this study

Our aim in this study was to determine the effect ofcervico-thoracic epidural analgesia not only on extu-bation time but also time until discharge from ICU us-ing a currently accepted fasttrack-concept for anesthe-sia for cardiac surgery.

Methods

38 patients treated according to a fast-tracking proto-col were compared with 41 historical control patientsundergoing surgery the year before the introduction ofthe fast-tracking protocol that were matched for bio-metric data, risk factors, preoperative health status,and type and duration of surgery, and the cardiac sur-geon performing the procedure . The fast-tracking pro-tocol includes a thoracic epidural catheter inserted at

levels of Th1-Th4 the day before surgery. The insertionwas performed by an experienced senior staff memberafter the appropriate amount of time since anticoagu-lation was stopped. The test dose was administeredright after catheter placement to rule out an intravascu-lar location of the catheter. In both groups generalanaesthesia was induced with etomidate and main-tained with propofol and sufentanil. Following the in-duction of anaesthesia a bolus of 10 ml ropivacaine0.375 % and 25 µg of sufentanil were administered viathe epidural catheter. The 41 patients in the controlgroup received postoperative opioid analgesia on de-mand (bolus injections of 3.75-7.5 mg piritramide un-til pain was lower than 5 on a numeric verbal ratingscale 0-11). Analgesia in the fasttrack-group of 38 pa-tients was provided by continuous epidural infusion ofropivacain 0.2 % and sufentanil 0.5 µg/ml in a dose of4 to 6 ml per hour.

Postoperative care was provided by the cardiacsurgeons on a specialized ICU. Times until trachealextubation and discharge from the ICU (both decisionswere based on the clinical judgement of the attendingsurgeon not involved in this study) were defined assurrogate markers for fasttracking eligibility.

Results

Due to a strict matched pair’s technique, patients inboth groups did not differ significantly with respect toany biometric or clinical variables. Times until tra-cheal extubation was significantly shorter in patientsreceiving the fasttracking protocol (p=0.0002; U-test)then in the control-patients who were treated accord-ing to standard procedure. However, this did not resultin an earlier discharge from ICU for the fasttrack-pa-tients (p=0.64) (see table 1). There were no differencesconcerning surgical complications between bothgroups. Major complications related to the epiduraltechnique were not observed.

time [hours] to fasttrack (n=38) control (n=41)

extubation

discharge from ICU

6 (1-14)

67 (45-135)

20 (11-18)

85 (45-122)

Table 1. All data are shown as median and 25th-75th percentile (inbrackets).

Thoracic epidural anaesthesia as part of a fasttrack-concept for cardiac surgery 19

Conclusions

Our findings support the theory that a fasttracking pro-tocol for coronary artery bypass grafts (CABG) in-cluding high thoracic epidural anaesthesia allows ear-ly tracheal extubation. This advantage was recognizedand well accepted by the attending cardiac surgeons.The successful early extubation however was nottransferred into faster discharge from ICU. Of coursethis may be due to the patients being deemed too unfitfor discharge or hospital logistics. Indeed, since thecardiac surgeons have 10 ICU beds for approximatelythree ACBG procedures per day there is rarely anypressure to transfer a patient more quickly to anotherward. However what seems most likely to be the maincause for the failure of the fasttracking protocol to re-sult in a faster ICU discharge is probably the fact thatthe involved ICU is managed by cardiac surgeons,who stuck to their customary postoperative routinesand showed a too inflexible approach to this new con-cept. Instead of considering the earlier extubation inthe patients with an epidural catheter as a chance for amore aggressive postoperative mobilisation they con-tinued to follow their usual clinical schemes. Habitualclinical pathways seemed to hinder early postoperativetransfer of patients to the ward before the third postop-erative day.

Priestly et al. (23) reached a similar conclusion intheir 2002 study. They analysed two groups of pa-tients, who underwent CABG surgery with one groupfollowing a fasttrack-protocol which employed anepidural catheter. Earlier extubation was achieved hereas well but time until hospital discharge stayed thesame as before.

Whether a thoracic epidural anaesthesia is reallyneeded for the implementation of a fasttrack concept isa question that has been discussed in several studieswithout reaching a definite conclusion (24). Severalstudies demonstrated that a fasttrack concept usinghigh thoracic epidural anaesthesia for intraoperativeand postoperative analgesia can be successful (19, 22,23, 26). At the same time several studies could showthat fasttrack concepts can be just as successful with-out the use of an epidural anaesthesia (1-3). There isstill a reluctance concerning the technique of high tho-racic epidural anaesthesia for cardiac surgery due tothe possible risks it presents such as intra- and postop-erative high dose anticoagulation and the consecutivefear of complications such as spinal or epiduralhaematoma. A recent survey in Germany concerningthe practice of high thoracic epidural anesthesia for

cardiac surgery showed that only 4 of the partaking 48anesthetic departments employed high thoracicepidural anesthesia for cardiac surgery on a regular ba-sis and has treated approximately 2500 patients withthis technique (25). None of these centers though hasencountered any severe neurologic or other complica-tions during this time. Pastor et al. analysed the data of714 patients with high thoracic epidural anaesthesiafor cardiac surgery over a course of seven years with-out the occurrence of any severe neurological compli-cations (26).

As a consequence of these results clinical imple-mentation of a fasttracking protocol must includebinding instructions for postoperative care. The per-sonnel involved (doctors and nurses) should be madefamiliar with the meaning and components of the fast-track concept and should receive schooling in the prac-tical management of such a regime. Physiotherapeuticpersonnel should receive special training on how tomobilise these patients if such a concept is to be suc-cessful. Only if all of these professional groups workseamlessly together can a fasttrack concept really leadto the desired success (20). Any fasttrack concept canonly be successful if the treatment pathways in thehospital are also changed.

References

1. Chong JL, Pillai R, Fisher A, Grebenik C, Sinclair M, WestabyS. Cardiac surgery: Moving away from intensive care. Br HeartJ 1992; 68: 430-433

2. Westaby S, Pillai R, Parry A, O’Regan D, Giannopoulos N,Grebenik K, Sinclair M, Fisher A. Does modern cardiac surgeryrequire conventional intensive care? Eur J Cardiothorac Surg1993; 7: 313-318

3. Cheng DCH. Fast track cardiac surgery pathways: Early extuba-tion, process of care, and cost containment. Anesthesiology1998; 88: 1429-1433

4. Cheng DC. Fast-track cardiac surgery: Economic implicationsin postoperative care. J Cardiothorac Vasc Anesth 1998; 12: 72-79

5. Lowenstein E, Hallowell P, Levine FH, Daggett WM, AustenWG, Laver MB. Cardiovascular response to large doses of intra-venous morphine in man. N Engl J Med 1969; 281: 1389-1393

6. Stanley TH, Webster LR. Anesthetic requirements and cardio-vascular effects of fentanyl-oxygen and fentanyl-diazepam-oxy-gen anesthesia in man. Anesth Analg 1978; 57: 411-416

7. Blomberg S, Emanuelsson H, Kvist H, Lamm C, Ponten J,Waagstein F, Ricksten SE. Effects of thoracic epidural anesthe-sia on coronary arteries and arterioles in patients with coronaryartery disease. Anesthesiology 1990; 73: 840-847

8. Kirno K, Friberg P, Grzegorczyk A, Milocco I, Ricksten SE,Lundin S. Thoracic epidural anesthesia during coronary arterybypass surgery: effects on cardiac sympathetic activity, myocar-

20 C. D. Kratz, U. Schirmer, S. Weiss, H. Wulf, L. Eberhart, S. Vogt, G. Geldner

dial blood flow and metabolism, and central hemodynamics.Anesth Analg 1994; 79: 1075-1081

9. Kock M, Blomberg S, Emanuelsson H, Lomsky M, StrombladSO, Ricksten SE. Thoracic epidural anesthesia improves globaland regional left ventricular function during stress-induced my-ocardial ischemia in patients with coronary artery disease.Anesth Analg 1990; 71: 625-630

10. Olausson K, Magnusdottir H, Lurje L, Wennerblom B,Emanuelsson H, Ricksten SE. Anti-ischemic and anti-anginal ef-fects of thoracic epidural anesthesia versus those of convention-al medical therapy in the treatment of severe refractory unstableangina pectoris. Circulation 1997; 96: 2178-2182

11. Blomberg SG. Long-term home self-treatment with high thoracicepidural anesthesia in patients with severe coronary artery dis-ease. Anesth Analg 1994; 79: 413-421

12. Liem TH, Booij LH, Hasenbos MA, van Egmond J. Coronaryartery bypass grafting using two different anesthetic techniques.I. Hemodynamic effects. J Cardiothorac Anesth 1992; 6: 148-155

13. Stenseth R, Bjella L, Berg EM, Christensen O, Levang OW,Gisvold SE. Thoracic epidural analgesia in aortocoronary by-pass surgery. I. Haemodynamic effects. Acta Anaesthesiol Scand1994; 38: 826-833

14. Moore CM, Cross MH, Desborough JP, Burrin JM, MacdonaldIA, Hall GM. Hormonal effects of thoracic extradural analgesiafor cardiac surgery. Br J Anaesth 1995; 75: 387-393

15. Stenseth R, Bjella L, Berg EM, Christensen O, Levang OW,Gisvold SE. Thoracic epidural analgesia in aortocoronary by-pass surgery. II. Effects on the endocrine metabolic response.Acta Anaesthesiol Scand 1994; 38: 834-839

16. Loick HM, Schmidt C, Van Aken H, Junker R, Erren M, Beren-des E, Rolf N, Meissner A, Schmid C, Scheld HH, Mollhoff T.High thoracic epidural anesthesia, but not clonidine, attenuatesthe perioperative stress response via sympatholysis and reducesthe release of troponin T in patients undergoing coronary arterybypass grafting. Anesth Analg 1999; 88: 701-709

17. Fawcett WJ, Edwards RE, Quinn AC, MacDonald IA, Hall GM.Thoracic epidural analgesia started after cardiopulmonary by-pass: adrenergic, cardiovascular and respiratory sequelae.Anaesthesia 1997; 52: 294-299

18. Liem TH, Hasenbos MA, Booij LH, Gielen MJ. Coronary arterybypass grafting using two different anesthetic techniques. II.Postoperative outcome. J Cardiothorac Anesth 1992; 6: 156-161

19. Stenseth R, Bjella L, Berg EM, Christensen O, Levang OW,Gisvold SE. Effects of thoracic epidural analgesia on pulmonaryfunction after coronary artery bypass surgery. Eur J Cardiotho-rac Surg 1996; 10: 859-865

20. Kehlet H. Multimodal approach to control postoperative patho-physiology and rehabilitation. Br J Anaesth 1997; 78: 606-617

21. Cheng DCH. Impact of early tracheal extubation on hospital dis-charge. J Cardiothorac Anesth 1998; 12 (Suppl 2): 35-40

22. Joachimsson P-O, Nystrom S-O, Tyden H. Early extubation af-ter coronary artery surgery in efficiently rewarmed patients: apostoperative comparison of opioid anesthesia versus inhala-tional anesthesia and thoracic epidural analgesia. J CardiothoracAnesth 1989; 3: 444-454

23. Priestley MC, Cope L, Halliwell R, Gibson P, Chard RB, Skin-ner M, Klineberg PL. Thoracic epidural anesthesia for cardiacsurgery: the effects on tracheal intubation time and length ofhospital stay. Anesth Analg 2002; 94: 275-282

24. Hansdottir V, Philip J, Olsen MF, Eduard C, Houltz E, RickstenSE. Thoracic epidural versus intravenous patient-controlledanalgesia after cardiac surgery: a randomized controlled trial onlength of hospital stay and patient-perceived quality of recovery.2006; 104 (1): 142-51

25. Kratz C, Eberhart L, Wulf H, Geldner G, Schirmer U. Use ofhigh thoracic epidural anaesthesia for cardiac surgery in Germanhospitals. Anästhesiologie und Intensivmedizin 2006; 12: 768-775

26. Pastor MC, Sanchez MJ, Casas MA, Mateu J, Bataller ML. Tho-racic epidural analgesia in coronary artery bypass graft surgery:seven years’ experience. J Cardiothorac Vasc Anesth 2003; 17:154-159

Address for correspondence: Caroline D. Kratz, Dep. for Anaesthe-sia and Intensive Care Medicine, Klinikum Ludwigsburg, Posili-postr. 4, 71640 Ludwigsburg, Germany, E-Mail: [email protected]

G. Seyffart

Poison IndexThe Treatment of Acute Intoxication

686 pages, 35 illustrations, hardcover Price: 97,50 EuroISBN 978-3-931660-46-8


– more than 350 types of poisoning– Poison-Index-List– rapid access to all appropriate treat-

ment methods– critical judgement of detoxification

measures– gastrointestinal decontamination– forced diuresis– artificial ventilation– administration of antidotes– intervention by medications– dialysis, hemoperfusion– supporting measures

The inflated balloon – a possible cause of microvascular obstructions 21

One of the most serious complications of the coronaryintervention is the thromboembolism. It may becaused by the detachment of thrombi, either from thevascular wall or from the catheter or the guide wire.Whereas a lot of these thrombi are inevitable at themoment, the inflated PTCA-balloon is a source ofsmall thrombogenic particles, which are – in our opin-ion – avoidable today.

All types of balloons are covered by a gliding lay-er, in order to pass also narrow stenoses. Figure 1shows as example a scanning electron microscopicphotograph (Zeiss DSM962) of this gliding layer on anoninflated brand-new balloon (type “RXELIPSE?014”, Guidant, Advanced Cardiovascular Systems,Temecula, Ca, USA). The gliding layer covered thewhole balloon continuously and evenly.

The average element composition of the balloonsurfaces was assessed by means of a backscatter elec-tron detector - in a scanning electron microscope at amagnification of 1:500 (EDAX 9800). The balloons

were unpacked from the original packings in a clean-roomworkbench. Afterwards the balloons were mount-ed with conductive tape on SEM sample holders. Thesamples were then covered with a thin carbon layer toincrease the conductivity. The carbon layer was ap-proximately 4 nm thick and it did not adhere verystrongly on the balloon surface.

On the new noninflated balloon a relative amountof oxygen of 8,48 percentage by weight and of siliconof 5,69 percentage by weight were detectable.

Obviously due to the different expansion coeffi-cients of the balloon material and the gliding layer, thegliding layer was damaged and became discontinuousin the course of the expansion of the balloon. Figure 2shows the balloon surface after 5 times inflation (pres-sure: 9 bar, duration: 1 minute each) against ambientair pressure under laminar-flow-conditions.

It is easy to observe, that the gliding layer on theballoon’s surface was discontinuous showing bigholes. It is possible, that parts of this gliding layer

The inflated balloon – a possible cause of microvascular obstructions

F. Jung1,2, J.-W. Park1, R. P. Franke3

1Centre for Biomaterial Development, GKSS Research Centre, Teltow, Germany; 2Department of Haemostasisand Transfusion Medicine, University of Saarland, Homburg/Saar, Germany; 3Central Institute of BiomedicalTechnology, Department of Biomaterials, University Ulm, Germany


Keywords: coronary intervention, balloon catheter, thromboembolism

Abstract

A serious complication of coronary interventions is thromboembolism which might be caused by the detachmentof platelet aggregates. One possible cause are small thrombogenic particles from the inflated balloon surface. The gliding layer of a balloon catheter was analyzed by means of a back scatter electron detector in a scanningelectron microscope. The average element composition of the gliding layer was assessed of a non-inflated and ainflated balloon. It could be shown that there was a continuous layer of silicon at the surface of the non-inflated balloon. After in-flation of a new balloon, however, there was only a very thin and discontinuous silicon layer showing big holes.It seems to be possible that parts of the gliding layer flaked off the balloon surface leaving only “plaques of sili-con” on the balloon surface.Detached silicon plaques surely would be thrombogenic and have therefore the potential to obstruct small myocar-dial arterioles and capillaries.

22 F. Jung, J.-W. Park, R. P. Franke

flaked off the surface leaving only ”plaques” on theballoon. Figure 3 shows such a “plaque” on the surfaceof the balloon. It turned out, that there were a lot ofsmall crevices in the remaining plaques. This mightlead in the end to the peeling of the gliding layer.

The gliding layer consisted substantially of silicon,but it contained also a great amount of oxygen. Com-

paring noninflated and inflated balloons, there weregreater amounts of oxygen (O: 8,48 w.- %) and of sili-con (Si: 5,69 w.- %) at the surfaces of the noninflatedballoons. The amount of these elements was substan-tially less at the surfaces of the repeatedly inflated bal-loons, where the oxygen amounted to 1,88 w.- % andsilicon to 0,20 w.- %.

The ratios of atomic weight percentages of oxygento silicon at noninflated balloon surfaces amounted toO/Si=2,6 and at 5 times inflated surfaces to O/Si=15,7.The data describing the average composition of the el-ements at the surfaces and the atomic weight ratios arecompatible with the hypothesis, that there was a con-tinuous layer of silicon at the surfaces of the noninflat-ed balloons. After 5 times inflating of a new balloon,however there was only a very thin and discontinuoussilicon layer.

In order to get more information about the thick-ness of the gliding layer, the element distribution waslocally assessed with the help of the energy dispersivex-ray analysis at a noninflated balloon surface.

To get this information the acceleration voltage ofthe electron beam was varied between 5 and 20 kV, sothat the penetration of the electron beam into the bal-loon surface increased from approximately 0,4 µm(5kV acceleration voltage) to about 2,8µm (20kV ac-celeration voltage) (table 1).

In the course of this the silicon-signal decreasedfrom 2,98 to 1,81 percentage by weight. This revealed,

Figure 1. Surface of a noninflated balloon (primarymagnification of 1:1000)

Figure 2. The middle section of the balloon after 5times inflation (primary magnification 1:9)

Figure 3. A section of the balloon surface, on whichthe gliding layer still adhered (magnification 1:100)

The inflated balloon – a possible cause of microvascular obstructions 23

that the volume fraction containing silicon decreasedwith an increase in the totally evaluated sample vol-ume. It can be assumed, that the thickness of the glid-ing layer was probably less than 3 µm. However, itmust be acknowledged, that the method used here toquantify the layer thickness needs intensive calibrationin order to evaluate the layer thickness more precisely.It would be necessary to make the same examinationon thin layers of the same material of defined thick-ness.

The size of the particles, which probably detachedfrom the balloon is not easy to determine. The forma-tion of crevices in the gliding layer as shown in fig. 3could be limited to the carbon layer and is typical ofsuch layers after deposition on polymeric substrates.The crevices in these carbon layers – however – usu-ally look different. In the case shown here the creviceswere deeper and revealed an upbulging quality, whichseems to be a hint, that the crevices continued into orstarted from the underlying silicon layer.

Detached silicon plaques surely would be throm-bogenic [3, 4, 5]. Plaque induced thrombi have the po-tential to obstruct small myocardial arterioles and cap-illaries and to start an ongoing thrombotic process atthese places.

Now, the dilating balloon catheters available todayhave considerably smaller diameters (starting from 1,5mm) and profiles and enable the passage of very smallstenoses. It is even possible to predilate stenotic re-gions with these catheters first and then to place thestent.

The gliding layer was necessary for the formerlyused balloon catheters with a bigger diameter in orderto pass narrow stenoses. From this point of view itseems to be unnecessary to use a gliding layer with thelatest developed balloon catheters equipped with aminute diameter. This can also be deduced from recentstudies, where resterilised balloon catheters were used.The gliding layer on these balloon catheters was re-moved by a thorough mechanical cleaning process be-fore the resterilisation. These catheters without thegliding layer revealed no deterioration neither in thehandling nor in the rate of success [1]. Catheters witha bigger diameter are of disadvantage economically,because for the passage of narrow stenoses the ad-vanced new balloon catheters with a narrow lumenhave not seldom to be used additionally [6].

From this point of view the gliding layer on the PT-CA-balloons isn’t necessary any longer. On the con-trary, the gliding layer constitutes a source for mi-crovascular obstructions - which are correlated withthe deterioration of the ventricular function [2, 7] - andshould therefore be applied no longer in order to avoidunnecessary risks for the patient.

References

1. Browne KF, Vlietstra RE, Brenner AS. The safe reuse of coronaryangioplasty catheters. Ptca aktuell 1999; 12: 5-10

2. Ito H, Maruyama A, Iwakura K et al. Clinical implications of the“no reflow” phenomenon: a predictor of complications and leftventricular remodelling in reperfused anterior wall myocardial in-farction. Circulation 1996; 93: 223-228

3. Lelah MD, Lambrecht LK, Young BR, Cooper SL. Physicochem-ical characterization and in vivo blood tolerability of cast and ex-truded Biomer J Biomed Mater Res 1983; 17: 1-22

4. Mrowietz C, Bach R, Seyfert U, Park J-W, Franke RP, Jung F.Haemocompatibility of coronary catheters. Biomed Technik2000; 45: 163-167

5. Mulvihill JN, Cazenave JP. Perfusion technique for in-vitro eval-uation of platelet interaction with biomaterials in catheter form.In: Lemm W (Ed.) The reference materials of the European com-munities (pp. 165-171). Amsterdam: Klouwer Academic Publish-ers, 1992

6. Plante S, Strauss BH, Goulet S, Watson RB, Chisholm RJ. Thereuse of balloon catheters for coronary angioplasty: a potentialcost-saving strategy? J Am Coll Cardiol 1994; 24: 1475-1481

7. Wu KC, Zerhouni EA, Judd RM et al. The prognostic significanceof microvascular obstruction by magnetic resonance imaging inpatients with acute myocardial infarction. Circulation 1998; 97:765-772

Address for correspondence: Prof. F. Jung, M.D., Centre for Bioma-terial Development, GKSS Research Centre, Kantstr. 55, 14513 Tel-tow, Germany, E-mail: [email protected]

Table 1. Results of the voltage dependent EDX- Analy-sis of the gliding layer

V[kV]

C[w %]

O[w %]

Si[w %]

Al[w %]

5 86,6 10,4 2,98 0,00

7 89,4 8,0 2,61 0,00

9 91,4 6,1 2,50 0,00

11 92,1 5,5 2,32 0,00

13 93,1 4,7 2,20 0,00

15 93,8 4,2 1,99 0,00

17 93,6 4,4 1,97 0,00

19 94,2 3,6 1,88 0,26

20 94,3 3,7 1,81 0,23

24 K. Wolf, J. Hamar, S. Moravec, T. Farkas, E. Höcherl, C. Pfister

Introduction

At the forefront of clinical wound treatment and re-search into bone tissue healing lies the goal of achiev-ing complete restoration of function and structure (9-12). Bioengineering opens up new opportunities in thisfield, where substitute bone tissue is created from thebody’s own cells. Its matrix contains messenger sub-

stances, and a pool of differently differentiated cells(1, 2).

Our investigations focus on osteoneogenesis inbone fracture healing (8, 19, 20, 21, 26, 33, 34). Thedifferent control mechanisms involved with osteoneo-genesis are the subject of much discussion, and numer-ous other factors also exist which influence the osteo-neogenesis process, including impaired fragment vital-

Fibrin pathways for precursor cell immigration from bone marrow into theosteotomy gap

K. Wolf1, J. Hamar2, S. Moravec3, T. Farkas2, E. Höcherl1, C. Pfister1

1Department of Trauma, Hand and Reconstructive Surgery, Krankenhaus München-Schwabing, Teaching Hospital of Ludwig-Maximilians University, Munich, Germany; 2Central Institute for Traumatology, Budapest,Hungary; 3Klinikum Innenstadt, Chir. Klinik und Chir. Poliklinik, Ludwig-Maximilians-University Munich, Munich, Germany


Keywords: precursor cell immigration, platelet-rich plasma, retraction, osteoneogenesis, fibrin pathways

Abstract

The primary goal of scientific investigations into fracture healing is to achieve a complete structural and function-al restoration with research into control mechanisms and early biomechanical influences on the orientation of pre-cursor cell immigration pathways. Primary fibrin networks can exhibit isotropy or anisotropy as a result of retrac-tion triggered by messenger substances and energy carriers. This suggests the existence of a primary biomechan-ical component that is used in the immigration of precursor cells.The morphometry of fibrin structures gives a mean value of 1.12 ± 0.007 (mean ± SEM), which differs markedlyfrom the hypothetical value of zero, thus demonstrating anisotropy or alignment of the fibrin structures. The mor-phometric test of precursor cell counts to analyze the dynamics of precursor cell immigration, carried out in thebone marrow between the l0th and 25th days, reveals significant differences between the cell count in the marrowand the osteotomy gap. The greatest movement of mononuclear cells in the marrow takes place on the 15th day,with 4.9 ± 0.2 cells per counting field. In the osteotomy gap, the levels found on the 15th day were 3.1 ± 0.1, andremained the same until the 25th day, whereupon a figure of 2.9 ± 0.1 was returned.We interpret the alignment of the fibrin network as an early morphological correlate of a biomechanical compo-nent. The administration of platelet-rich plasma (PRP) in human patients seems sensible, and the industry alreadyoffers PRP methods for use on human patients (Curasan AG, Kleinostheim, Germany). Thrombocytic factors canbe obtained which boost the retraction process and pre-form network structures for precursor cell immigration.Nondifferentiated mononuclear cells migrate along both fine and thick fibrin structures. Growth factors, gene ther-apy and cell therapy may in future have a positive influence on the complex process of bone healing. Tissue en-gineering also offers other ways of restoring the original structure of bones in a short space of time.

Fibrin pathways for precursor cell immigration 25

ity, cortisone treatment, hyperglycemia and infection(23, 24). Bone induction is based on the immigrationof precursor cells and the differentiation of thesepluripotent cells, while the opposite of this is boneconduction, which is based on vital fragments at theedge of the fracture. The processes of bone inductionand bone conduction are essential for bone healing.There are however a large number of unansweredquestions with regard to precursor cell immigrationand how it works, and the answers to these questionsmay provide a pathway for developing early clinicalinfluences on fracture healing.

The question to be asked relates to the earliest bio-mechanical influences on the immigration of precursorcells (5). Our attention focuses on structures in the ear-ly stages, such as the role of fibrin septa as messengersfor biomechanical stimuli. Fibrin networks can exhib-it isotropic or anisotropic properties (27, 32), which in-dicates an early biomechanical function. Questions al-so need to be asked regarding potential messenger sub-stances or energy carriers, e.g. from disintegratingthrombocytes and their effects on the early stage offracture healing. Today, the industry is already offeringsets for PRP (platelet-rich plasma) methods (CurasanAG, Kleinostheim, Germany). More precise morpho-logical and experimental studies may provide addi-tional information on processes that have a positive in-fluence on bone healing.

Besides clarifying the morphological structure ofthe immigration paths, the quantity and quality of cellmigration also need to be examined. Previously, theperiosteum was believed to be the primary origin ofmigratory precursor cells, and the marrow was eitherdisregarded or declared insignificant. Detailed mor-phological pictures of the marrow adjacent to the frac-ture are not available for the early stage of fracturehealing. We must assume that an immense number ofmigrant precursor cells are subject to other controlmechanisms. In cell therapy, mesenchymal stem cellsare used that can differentiate into fibroblasts, en-dothelial cells, osteoblasts and adipocytes. These cellssynthesize - in different ways and according to theirdifferentiation - collagens, proteoglycans and cyto-kines.

Methods

Animals and surgery

White New Zealand female rabbits (2.2 - 2.5 kg bodyweight) were used. Surgical and animal preparationwere performed according to the NIH guidelines gov-erning the use of experimental laboratory animals. Therabbits were initially anesthetized with 0.4 mg/kg Ke-tamine HCL (Ketavet, Pharmacia Upjohn, Erlangen,Germany) and 5 mg/kg 5,6 Dihydro-2(2,6-xylidino)-4H-1,3-thiazine (Rompun, Bayer, Leverkusen, Ger-many), both administered intra-muscularly. Anesthesiawas maintained with intravenously administered Di-hydroxyxylidino-thiazine HCI, at 16 mg/hour (Nar-coren Rhone Merieux, Laubheim, Germany). A slowinfusion of Ringer lactate solution was injected intothe ear vein. At the end of surgery, the animals were al-lowed to wake up from anesthesia. They were kept in-dividually in their cages and had free access to tap wa-ter and standard rabbit chow.

Osteotomy model

Operative access was obtained between the M. tibialisanterior (peroneal and Tibialis anterior) and M. gas-trocnemius caput laterale. The tibia was exposed and a3 mm-thick ring was removed from its central sectionusing an oscillation saw. Osteosynthesis was per-formed with a 7-hole, 65 mm AO reconstruction DCplate (Synthes, Umkirch, Germany). Particular carewas taken to maintain the 3 mm distance. The woundwas sutured by layers.

Protocol

Following surgery, the animals were sacrificed by anoverdose of Narcoren on the 5th, l0th, 15th, 20th and 25th

post-operative days respectively. Each group consistedof seven rabbits. The plate was removed from the tib-ia. Tissue samples (Figure 1) were taken (50 mm3)from the fracture gap (site C) and also from two adja-cent sites (site A and site B). A thin, sharp scalpel wasused to remove pieces of marrow. Sampling site A was10 mm proximal from site C, and site B was in thetransition zone between the intact bone marrow andthe granulation tissue of the healing bone wound.Bone marrow samples were also collected from 7 un-


operated animals without osteotomy. These served ascontrol studies.

Histological analyses

The samples were put into Zamboni’s solution(buffered formaldehyde 4%, containing picric acid)and prepared as described by Bjurholm (4). Followingnormal processing (paraffin embedding, SP 56 °C,Histosec, Merck, Darmstadt, Germany), 8 micrometer-thick sections were cut (traversing microtome, Jung,Heidelberg, Germany) and methylene blue stainingwas applied. For cell counts, the sections were ana-lyzed with light microscopy (Leitz, Wetzlar, Germany)with a magnification power of 400x, 600x and 1000x.

Microscopy of Histology

The histological sections of the standard histologywere analyzed using a light microscope (Leitz, Wet-zlar, Germany) and a polarization device. For analysisof the fibrin structures, a 9 x 9 eye grid was used witha counting ring in the ocular of the microscope. Forprecursor cell counting, a 10 x 10 reticule was used in

the ocular. The micrometer was calibrated using astage micrometer (Leitz, Wetzlar, Germany).

Analysis of orientation of fibrin structures

In our present studies, we applied Oberholzer’smethod (17) to detect whether there are structural reg-ularities of fibrin structures in the healing bone in theearly phase of callus formation. Oberholzer’s formulafor calculating the orientation factor K* (32) was usedto analyze the direction of the fibrin structures. The K*factor is an expression of the isotropy or anisotropy ofstructures whose alignment or non-alignment isrecorded morphometrically. Fibrin structures of thehematoma in the gap were analyzed using polarizationmicroscopy. The numbers of crossing points per lengthof the vertical test lines, and the numbers of crossingpoints per length of the horizontal test lines, were in-cluded in the formula.

IL(x/B)1: Number of crossing points per length of verti-cal test lines

IL(x,B) 2: Number of crossing points per length of hori-zontal test lines

Figure 1. Model of dis-tance osteosynthesis.Localizations of the sam-pling sites for histologi-cal analyses. P = Os-teosynthesis plate, CB =Cortical Bone, BS =Bone screw, BM = BonemarrowA = Region (distal to theosteotomy) of precursorcell countingB = Region (proximal tothe osteotomy) of precur-sor cell countingC = Osteotomy gap ofthe distance osteosynthe-sis model and region ofprecursor cell counting


Formula for calculating IL(x/B)

I(x): Crossing pointsP(B): Hits in the structure aream: Magnificationd: Distance of test linesK1: Linear constant of screen = 2 (Oberholzer

1983)

Cell count of mononuclear cellsCell count of mononuclear cells

Only mononuclear cells (MNCs) were taken into con-sideration. The number of cells was assessed by using100 fields in three different sections (300 fields in to-tal). Cell counts were performed in the marrow in Re-gion A (distal to the osteotomy) and Region B (pro-ximal to the osteotomy), and in the interfragmentarygap. Different post-operative days were taken into ac-count (5th, 10th, 15th, 20th and 25th post-operative day).

Statistics

The extent of anisotropy of the fibrin structure was an-alyzed. Values of the K* factor were compared to atheoretically isotropic group having a K* of 0. If thevalue of K* is less than 0.5, there is isotropy - mean-ing that the analyzed structures are not oriented. If itsvalue is one or higher, there is anisotropy, and thestructures are oriented. The higher the value of K*, thegreater the probability of anisotropy. The orientationfactor K* was determined in the 5 day group becauseit was used for analyzing the very early stage of bonehealing.

For statistical evaluations of the numbers of pre-cursor cells, mean values were calculated, specifyingthe standard error of mean (SEM). A control group(samples taken from the marrow of rabbits which werenot operated on) was recorded and analyzed. Signifi-

cance calculations were performed using the t-test asdescribed by Student in comparison with the chrono-logical development in the different probe sample re-gions. The level of significance was set at p < 0.05.The SPSS computer program was used (SPSS-Corpo-ration, Chicago, USA).

Results

Histological evaluation of the osteotomy gap

In the initial stage of fracture healing, the morphologydemonstrated numerous erythrocytes (Fig. 2a) on the5th post-operative day. Granulocytes were isolated, andthrombocytes were also visible. In some places, in ad-dition to the uniform distribution of erythrocytes, atransition was noted towards a networked arrangement(Fig. 2b). Areas of greater or lesser development wereseen to lie close together. In between was a clear fluid,serum, containing defibrinated plasma. Disintegratingthrombocytes promoted the retraction process (Fig.2c) and caused the fibrin strands to nestle closer andmore densely together. Advancement of the retractionprocess caused the formation of fixed fibrin columnsas a result of shortening and folding (Fig. 2d). Bloodplatelets were visible in the middle of one of theprocesses involved in the retraction. At greater magni-fication, thick fibrin columns were seen in the histo-logical preparation (Fig. 2e). In this phase, there werestill no precursor cells, nor did the network-like struc-tures contain any mononuclear cells in the spaces be-tween them. It was only in the next phase (Fig. 2f) thatthe mesh structure was populated by precursor cells. Inthe serum between the mesh structure too, the matura-tion process led to the immigration of precursor cellsinto lacunar areas. Further development led to the net-work structures appearing packed with mononuclearcells. Examination at higher magnification (Fig. 2h)confirmed this. Fibroblasts had differentiated into fi-brocytes at this stage. Non-differentiated cells were al-so seen in large numbers.

Precursor cells differentiated themselves into an-gioblasts (Fig. 3a). Thick capillaries formed fromclosely-lying angioblasts (Fig. 3b). The accumulationof precursor cells was also seen in lacunar structures(Fig. 3c). Capillaries were also visible even in less ma-ture lacunar areas (Fig. 3d). Mononuclear precursorcells were now present in abundance, although serum-filled lacunae were still visible in this stage. The cap-illaries (Fig. 3e) appeared intertwined with rows of an-

Σ I(x) m I L(X/B) = • Σ P(B) d • K1

IL(x/B)l - 1 IL(x/B)2 K* = 3 • IL(x/B)1 + 1 IL(x/B)2


Figure 2a. Uniform distribution of erythrocytes with arrangement into a network-like configuration (*). Isolatedgranulocytes (↑), also includ-ing thrombocytes (↑↑). Magnification 630x, Hemalum stain.Figure 2b. Mesh structure of fibrin strands (↑) visible between the erythrocytes (retraction). Differently-developedareas lying close together. Between the erythrocytes is a clear fluid, the serum. Magnification l000x, Hemalumstain.Figure 2c. Granulocytes (↑) and blood platelets (↑↑) in the middle of a retraction process. Fixed fibrin columns(↑↑↑) as a result of shortening and folding Enlargement l000x, Hemalum stain.Figure 2d. Formation and densification of thick fibrin columns (↑). Magnification 400x, Hemalum stain.Figure 2e. Lacunar structures with defibrinated serum (↑), free of precursor cells. Fibrin column (↑↑) Magnifi-cation l000x, Hemalum stain.Figure 2f. Packing of mesh structure (↑) with mononuclear precursor cells (↑↑). Magnification 630x, Hemalumstain.Figure 2g. Mature tissue on the left-hand side of the image (*). Right-hand side of the image showing precursorcell packing of the lacunar structures (*). Magnification 100x, Hemalum stain.Figure 2h. Higher magnification of Figure 2g. Differentiation and proliferation of precursor cells. Fibroblasts (↑)between mononuclear cells (↑↑). Magnification l000x, Hemalum stain.


Figure 3a. Precursor cells (↑) differentiate into angioblasts (↑↑). Circular formation. Magnification 1000x,Hemalum stain.Figure 3b. Advancement of tissue maturation, capillaries (↑) among densely-packed precursor cells (↑↑) Magni-fication 630x, Hemalum stain.Figure 3c. Lacunar structures with tightly-packed precursor cells (↑). Magnification 630x, Hemalum stain.Figure 3d. Even less-mature lacunar areas contain capillaries (↑). Mononuclear precursor cells (↑↑) are presentin abundance, alongside serum-filled lacunae (↑↑↑). Magnification 630x, Hemalum stain.Figure 3e. Intertwined capillaries with rows of angioblasts (↑), surrounded by pericytes (↑↑) Magnification 630x,Hemalum stain.Figure 3f. Increase in precursor cell density in the tissue (*), Hemalum stain.Figure 3g. Wave-like immigration of precursor cells (↑). Magnification 40x, Hemalum stain.Figure 3h. Circular, dense waves of immigrant precursor cells (↑). Magnification 40x, Hemalum stain.


gioblasts, surrounded by pericytes. There were nolonger any lacunar structures in the mature tissue.There were areas of tissue with a large accumulationof precursor cells, which surrounded tightly-packedcapillaries. The immigration of precursor cells can bedescribed as a type of wave form (Fig. 3g). Corre-sponding sections, rotated around 90 degrees (Fig. 3h),showed circular, dense waves of immigrating precur-sor cells.

Orientation of the fibrin network

On the 5th post-operative day, the fibrin networkscould be clearly seen within the cell debris. The fibrinstrands had become more dense in the osteotomy gap.The mean value for the K* factor was significantly dif-ferent from the hypothetical value of zero, returning amean value of 1.12 ± 0.007 (mean ± SEM) (anisotropyof fibrin structures). If the value had been zero, thiswould have indicated complete isotropy of the fibrinnetworks (spherical form of fibrin structures). Themean value of 1.12, as a morphometric value, corre-sponded to anisotropy (alignment) of the fibrin net-works. Figure 4 shows the distribution of the K* fac-tor. The majority of values were found in the analyti-cal range between 0.5 and 1.99, corresponding to aproportion of 78 %. In the analytical range < 0.49,there were only 9 values for the K* factor, which cor-responded to 13%. In the analytical range 0.5 to 0.99,

there were 16 K* values, equating to 23 %. In the an-alytical range 1.00 to 1.49, there were 24 K* values,equating to 34 %. The analytical range 1.50 to 1.99 en-compassed 15 K* values, this equating to 21%. In theanalytical range > 2.00, there were 6 K* values, corre-sponding to a proportion of 9%.

We cannot define in which way these fibers are di-rected in the gap, however, there is strong evidencethat they are oriented. These fibers would lead the pre-cursor cells into the gap where the differentiating cellsproduce collagen fibers to replace those of fibrin. It ispossible that the orientation of the fibers is regulatedby biomechanical forces similar to those in the soft tis-sue (25). The rabbits were able to move freely and usetheir operated-on legs in their cages immediately afterthe operation, and the physiological load could be ap-plied to the leg in this way.

Counts of mononuclear cells (MNCs)

The counter results from the precursor cell counts inthe preparations taken from the 5th postoperative dayare shown in Figure 5. For Region A, distal to the os-teotomy, the mean value was found to be 3.1 ± 0.1(mean ± SEM) cells per counter field. Region B, prox-imal to the osteotomy, returned a counter result of 4.1± 0.1 (mean ± SEM) cells per field. The difference be-tween the regions distal and proximal to the osteotomywas not statistically significant. At this stage, no evi-

Figure 4. Test ofanisotropy or isotropy,Factor K*70 individual values weregrouped and presented intheir frequencies of occur-rence. The cut-off valuefor isotropy was < 0.49(K*). The cut-off value foranisotropy was higherthan 0.5 (K*). 70 meas-urements, taken on the 5th

post-operative day.


dence could be found of any precursor cells in the gapof the distance osteosynthesis model. On the l0th post-operative day, a slightly lower mean value of 3.4 ± 0.1(mean ± SEM) cells per counter field was found in Re-gion A, distal to the osteotomy. Region B, proximal tothe osteotomy, returned a counter result of 3.1 ± 0.2(mean ± SEM) cells per counter field. The differencebetween the regions distal and proximal to the osteoto-my was not statistically significant. In the gap, 0.9 ±0.0 (mean ± SEM) mononuclear cells were seen percounter field on the 10th post-operative day. The statis-tical test revealed significant differences between Re-gion A (distal to the osteotomy) and Region B (proxi-mal to the osteotomy) compared to the osteotomy gapin Region C. The counter results for the 15th post-op-erative day (Figure 5) returned a value for Region A(distal to the osteotomy) of 4.8 ± 0.2 (mean ± SEM)cells per counter field. Region B (proximal to the os-teotomy) returned a result for mononuclear cells of 4.9± 0.2 (mean ± SEM) cells per counter field. The differ-ence between Region A and Region B was not regard-ed as statistically significant. In the osteotomy gap, amean value of 3.1 ± 0.1 (mean ± SEM) cells per count-er field was returned. The statistical mean comparison,when comparing Region A (distal to the osteotomy)and Region B (proximal to the osteotomy) with the os-teotomy gap Region C, revealed a statistically signifi-cant difference (Figure 5).

Region A and Region B are shown in Figure 6 as asingle mean value, for standardization purposes (nostatistical difference). Following on from Figure 5,Figure 6 shows the long-term course of events. Com-bining the values described above for Region A (distalto the osteotomy) and Region B (proximal to the os-teotomy) resulted in a value of 3.6 ± 0.1 precursorcells per counter field on the 5th post-operative day(Figure 6). For the 10th post-operative day, the com-bined value (Region A, distal to the osteotomy and Re-gion B, proximal to the osteotomy) was 3.3 ± 0.2 pre-cursor cells (mean ± SEM). On the 15th post-operativeday, the value rose outside the osteotomy zone to 4.9 ±0.2 precursor cells per counter field (mean ± SEM).On the 20th post-operative day, the value outside theosteotomy fell to 2.4 ± 0.1 (mean ± SEM) precursorcells per counter field. Further on, on the 20th day, avalue of 2.5 ± 0.1 (mean ± SEM) was returned for theosteotomy gap - a value similar to that of regions A/Bwhich stood at 2.4±0.1 (mean ± SEM) cells per count-er field. On the 25th day, there was a further rise in theosteotomy gap, to 2.9 ± 0.1 (mean ± SEM) cells percounter field. The counter value in Regions A/B waslower, at 2.1 ± 0.1 cells per counter field.

Figure 5. Cell counts inbone marrow and osteoto-my siteChanges in cell counts atsampling site A and site B(bone marrow) and site C(interfragmental gap) as afunction of the healingtime.

significant changescompared to the value ofday 5. + significant differencesbetween A/B and C.

Region A, distalto osteotomyRegion B, proxi-mal to osteoto-myRegion C


Discussion

Morphology of fibrin structures in granulation tissue

The initial description and definition of the granulationtissue, as the first physiological response after a frac-ture, comes from Billroth (3). In spite of many inves-tigations, the early stage of bone tissue healing in theinitial hours and days after trauma has remained forthe most part a mystery (15, 16, 18). We interpret thetransformation into network structures, as we see themin our histological preparations, as a morphologicalcorrelate of the retraction process triggered by fibrin.This retraction leads to the organization of fine fibrinpillars and thick fibrin columns, and these are believedto pre-form the immigration paths for precursor cells.This is a previously unaccounted for metamorphosis inthe hematoma associated with the fracture. The defib-rinated plasma collects between the fibrin networkstructures (22). The retraction process for the fibrinstrands starts within just a few hours in the fracturehematoma, and changes the gelatinous mass into semi-solid blood clots. This finding tallies with what isknown about the underlying physiology. The stagesleading to the blood clotting process were first de-scribed by Morawitz in the year 1915 (22). The plas-ma protein prothrombin is converted to thrombin bythrombokinase released on the disintegration ofthrombocytes - in the presence of ionized calcium.Thrombin causes the dissolved plasma protein fibrino-

gen to become fibrin, which gives rise to the threadlikebasic structure of a blood clot. Under the influence ofelectrostatic forces, the fibrin monomers then laythemselves lengthwise in parallel to each other to formfibrin polymers. Only through the effects of fibrin-sta-bilizing Factor XIII, activated by thrombin in the pres-ence of Ca++, can covalent links be formed betweenthe fibrin monomers. These are what give the fibrinstrands their final chemico-physic properties.

The disintegrating thrombocytes release retrac-tozyme and thrombosthenin factors. These factorscause the fibrin strands to nestle closer together andbind more tightly, and to shorten after folding. Theblood clot shrinks, and quantities of ATP from disinte-grating thrombocyte fragments provide the energy forthis process. Retraction fixes the blood clot in placemechanically, the wound edges are brought togetherand this creates more favorable conditions for the for-mation of blood vessels.

The familiar phenomena of blood clotting, e.g. re-traction, are today exploited in the clinical use of PRPmethods (platelet-rich plasma). Thrombocytes are cen-trifuged out of the patient’s own blood and applied tothe bony defect in conjunction with substitute bonematerial. This enables thrombocyte factors to be usedthat trigger and boost the retraction process, so as topre-form the fibrin network structures for the migra-tion of precursor cells. The PRP method is already of-fered for clinical use by the industry (Curasan AG,Kleinostheim, Germany).

Figure 6. Cell counts duringhealing timeChanges of cell counts atsampling sites A/B, and Cas a function of the healingtime.* significant changes com-

pared to the control value. + significant differences be-

tween sites A/B and C


Isotropy or anisotropy of fibrin structures

The question relating to the aligned migration of pre-cursor cells is inextricably linked with the alignment(anisotropy or isotropy) of guiding structures. Themorphometric analysis of fibrin structures using analignment screen in the ocular of the microscope canhelp objectivize this. The result of the morphometricanalysis - isotropy or anisotropy (= K* factor as an ex-pression of the alignment of fibrin structures) - isshown in Figure 4. It shows clearly - with a figure of87% of all K* factor values above the significancethreshold (> 0.49) - an alignment (anisotropy) of thestructures, some of which are round, some of whichare oval. What we find most important is the findingthat the osteotomy gap features extremely dense fibrinstructures. We interpret this on the assumption that re-traction is being increased here, and supporting orstrengthening this mechanism, e.g. through the admin-istration of PRP at this point, would seem sensible. Inthe clinical use of the PRP method (platelet-rich plas-ma), the release of thrombocyte fragments must takeplace for the most part in the early stage (hours to ap-prox. 5 days after trauma), since this is when most ofthe retraction process occurs and ATP is available as anenergy donor. The retractozyme and thrombostheninfactors, released from the thrombocyte fragments, sup-port this process.

The fibrin retraction process in the fracture has sofar not been described. Papers by Sedlarik (25) are par-ticularly interesting in this regard. Sedlarik postulatesthat a biomechanical component exerts an influence onthe fibrin structures and the network-like arrangementof erythrocytes in the hematoma during soft tissuewound healing. Our experiments have confirmed, forthe first time, that this phenomenon is also true for thegranulation tissue of the osteotomy gap. The questionregarding a biometrically-triggered alignment of thehematoma in the fracture model, in the sense of de-fined micro-movement, can be answered with a clearaffirmative. The anisotropy discovered shows thatthere is alignment of the fibrin septa and columns. Mi-cro-movement is known to have a positive effect onsubsequent osteoneogenesis. Clinically-applied meth-ods of immobilization, be they conservative throughsplinting or invasive through osteosynthesis, take ac-count of this fact.

Morphology of precursor cells and osteoneogenesis

The bone marrow of long bones is filled up mostlywith adipose tissue, which is relatively poor in cells(14). In addition to the adipocytes, there are some un-differentiated cells and also representatives of the im-mune system. However, if the marrow is stimulatedthrough the increased need for hematopoesis follow-ing bone trauma, there is a rapid change in its cellularcomposition. The stimulus to produce undifferentiatedprecursor cells is very strong, and this is likely to bemediated by humoral pathways. This is supported byour earlier experimental results in which we analyzedthe cellular composition of the contralateral tibia (un-published results). There was a higher average numberof precursor cells compared to the non-operated con-trols, even on the 25’h post-operative day, indicatingthat the marrow is still being stimulated. The mole-cules bearing the humoral marker of this stimulationhave not been identified completely (28-31).

Undifferentiated mononuclear cells travel alongboth the fine and thick fibrin structures. The embryon-ic phases of tissue formation demonstrate a wave-likeinflux of cells. This phenomenon was also found in ourhistological preparations (Fig. 3g and 3h). These mi-gratory waves are associated with phase- and localiza-tion-oriented tissue maturation.

The intensive cellular immigration into the devel-oping callus is supported by the increased output of theadjacent marrow, and it levels off at the 15th post-oper-ative day. A one-way migration of the precursor cellsis suggested, and this is demonstrated in Figure 6. Thehighest cell number was found in the intermediaryzone on day 5 when we had the biggest difference be-tween the marrow and the gap. Later, this intermediarypart showed a homogenous structure together with therest of the marrow more proximal to the wound. Al-though we have emphasized the role of the bone mar-row in producing precursor cells, the surrounding con-nective tissue around the osteotomy site also has un-differentiated cells, such as endothelial cells and peri-cytes, which increase in population and transform intomesenchymal cells (7) mainly fibroblasts - and theseare also attracted to the site of the bone wound.

The 15th post-operative day, the time at which thecell counts of undifferentiated cells in the marrow startto decline, coincides very well with the start of cellu-lar maturation and differentiation in the callus site.This might indicate that the newly formed tissue in thefracture gap has achieved a certain degree of self-maintenance. Brighton and Krebs (6) found an in-


crease in the oxygen tension of the developing callus,starting in the third week of the healing process. Thisprocess was also paralleled by the maturation of thehealing bone. These results have shown that there is anincreased generation of undifferentiated precursorcells in the intact bone marrow, and that these cellsmove into the hematoma of the bone wound along thefibrin network present during the very early stages ofthe developing interfragmental tissue.

Current scientific opinion holds that the transfu-sion of stem cells is only clinically indicated in a fewcases of disease. It can also only be performed in spe-cial centers which fulfill the necessary staffing andtechnical criteria. The harvesting of stem cells throughbone marrow aspiration yields a mixture of indetermi-nate or already determined stem cells and cells with aclear differentiation. These latter represent the majori-ty of the cells harvested. Suitable separation methodscan be used to concentrate the stem cells. Blood cellseparation from peripheral blood also enables stemcells to be obtained in sufficient quantities. They arefound in the fraction of mononuclear blood cells.

Further thoughts

In addition to growth factors, gene and cell therapy(13) may well be of interest for exerting a positive in-fluence on the complex process of fracture healing.Tissue engineering harnesses these opportunities to re-store the original bone structure in a short space oftime, and is used as an adjunct to traditional treatmentmethods. The latest concepts involve bioactive con-structs based on a cell-packed, three-dimensional ma-trix used for bridging bone defects. They comprisevarious components including collagen, b-tricalciumphosphate, fibronectin or synthetic, biodegradablepolymers (polylactides and polyglycolic acid de-rivates). The kinetics of the constructs’ contraction de-pends on the initial cell density. If, for example, a col-lagen matrix is packed with a high density of autolo-gous cells, more aligned cells with an elongated nucle-us are found after 72 hours than with a low cell densi-ty.

The importance of mechanical stimuli for control-ling cell differentiation and matrix synthesis is alsogrowing. Future studies must demonstrate what influ-ence physiological stimuli such as cyclic stretchingand / or growth factors have on cell differentiation andproliferation. In this context, it is not just the origin ofthe cells and the matrix that matter, but also the initial

biomechanical properties of the construct, the rate ofmatrix degradation, the cell survival rate and theirability to respond adequately to physiological forces.Current research focuses on the influence of the three-dimensional structures of synthetic materials on cellorientation, cell proliferation and cell phenotypes.Clinically-oriented studies must in future investigatewhether biodegradable, synthetic polymers, packedwith autologous cells, can provide an adequate substi-tute for bone tissue.

Acknowledgements

This work was supported by grants from a) Wilhelm Sander Stiftung, Neustadt a.d. Donau, Pro-

ject Number 91.056.1. b) BMBF/DLR Bonn, Project Number UNG-054-96

and OMFB-TeT No D-69/1996.

The authors thank Mrs. Ellen Maier and Miss Is-abel Rosch for their support.

References

1. Aho AJ. Electron microscopic and histological observation onfracture repair in young and old rats. Acta Pathol MicrobiolScand Supplement 1966; 184

2. Aro H. Effect of nerve injury on fracture healing - Callus forma-tion studied in the rat. Acta Orthop Scand 1985; 56 (3): 233-237

3. Billroth T. Untersuchungen über die Entwicklung der Blutge-fäße. Berlin: Georg Reimer, 1856

4. Bjurholm A, Kreicbergs A, Brodin E, Schultzberg M. SubstanceP - and CGRP immunoreactive nerves in bone. Peptides 1988; 9:165-171

5. Bleiberg I. Colony forming cell-fibroblast development in extra-cellular matrix-induced bone and marrow formation in rat. Con-nect Tissue Res 1985; 14 (2): 121-127

6. Brighton C, Krebs AG. Oxygen tension of healing fractures inthe rabbit. J Bone Joint Surg 1972; 54-A: 323-333

7. Brighton CT, Hunt RM. Early histologic and ultrastructuralchanges in microvessels of periostal callus. Journal of OrthopTrauma 1997; 11: 244-253

8. Brighton CT, Hunt RM. Early histological and ultrastructuralchanges in medullary fracture callus. Journal of bone and jointsurgery 1991; 73 A: 832-847

9. Einhorn TA. Knochenumbau bei der Frakturheilung. Sandorama1993; 2: 15-23

10. Enderle A. Der metaphysäre Tibiadefekt der Ratte. Ein Modellzur Erzeugung eines rein endesmalen Knochens. Osteologie1992; 1 (1): 15-26

11. Enlow DH. Principles of Bone Remodeling. In: Evans FG,Charles E, Thomas C (Eds.) (pp. 1-30). Springfield Illinois, 1963


12. Frost HM. Bone development during Childhood. A tutorial(some insights of a new pardigm). In: Schonau E (Ed.) PediatricOsteology (pp. 3-38). 1996

13. Guerra FJD, Zaidi M, Bevis P, MacIntyre I and Emson PC. Ev-idence for release of Calcitonin gene-related peptide and Neu-rokinin A from sensory nerve endings in vivo. Neuroscience1988; 25 (3): 839-846

14. Hughes FJ and Aubin JE. Culture of cells of the osteoblast line-age. In: Arnett TR, Henderson B (Eds.) Methods in Bone Biolo-gy (pp. 1-39). London, Weinheim, New York, Tokyo, Mel-bourne, Madras: Chapman & Hall, 1998

15. Hult A, Olerud S. Early fracture callus in normal and cortisonetreated rats. A study by a combination of tetracycline labelling,microangiography and microradiology. Acta Orthop Scand1964; 34: 1-23

16. Iwaki A, Jingushi S, Oda Y, Izumi T, Shida JI, Tsuneyoshi M andSugioka Y. Localization and quantification of proliferating cellsduring rat fracture repair: detection of proliferating cell nuclearantigen by immunohistochemistry. J Bone Miner Res 1997; 12(1): 96-102

17. Oberholzer M. Quantitative Topographie. In: Oberholzer M(Ed.) Morphometrie in der Klinischen Pathologie, AllgemeineGrundlagen (S. 25-37). Berlin, Heidelberg, New York, Tokyo:Springer, 1983

18. Perren SM, Cordey J. Die Gewebsdifferenzierung in der Frak-turheilung. Unfallheilkunde 1977; 80: 161-164

19. Schenk RK, Willenegger R. Zur Histologie der primärenKnochenheilung. Modifikation und Grenzen der Spaltheilung inAbhängigkeit von der Defektgröße. Unfallheilkunde 1977; 80:155-160

20. Schenk RK. Die Histologie der primären Knochenheilung imLichte neuer Konzeptionen über den Knochenumbau. Unfall-heilkunde 1978; 81: 219-227

21. Schenk RK. Morphometrische Analyze der Umbauvorgänge inder Kompakta des Knochens. In: Weibel ER, Elias H (Eds.)Quantitative Methods in Morphology (pp. 199-217). Berlin,Heidelberg, New York: Springer, l967

22. Schmidt RF, Thews G. Gerinnungsfaktoren und Gerinnungs-ablauf. In: Schmidt RF, Thews G (Eds.) Physiologie des Men-schen (S. 335-337). Berlin, Heidelberg, New York: Springer-Verlag, 1977

23. Schweiberer L, Wolf K. Pathophysiologie der Knochenheilung.Hefte zur Unfallheilkunde. In: Habermeyer P, Schweiberer L(Eds.) (S. 26-31). Berlin, Heidelberg: Springer, 1992

24. Schweiberer L. Pathophysiologie der Frakturheilung. In: ZenkerR, Deucher F, Schink W (Eds.) Chirurgie der Gegenwart (S. 1-35). München, Wien, Baltimore: Urban & Schwarzenberg, 1976

25. Sedlarik KM. Primäre und sekundäre Wundheilung. In: SedlarikKM (Ed.) Wundheilung (S. 83-93). Jena, Stuttgart: Gustav Fis-cher, 1993

26. Sevitt S. Bone repair and fracture healing in man. In: Sevitt S(Ed.) Current Problems in Orthopedics (pp. 89-107). Edinburgh,London, Melbourne, New York: Churchill Livingstone, 1981

27. Underwood EE. Oriented structures. In: Underwood EE (Ed.)(pp. 48-73). London: Addison-Wesley Publishing Company,1969

28. Urist MR, Dowell TA, Hay PH, Strates BS. Inductive Substratesfor Bone Formation. Clin Orthop 1968; 59 (7): 59-96

29. Urist MR, Hay PH, Dubuc F, Buring K. Osteogenetic Compe-tence l969; 64 (5): 194-220

30. Urist MR, Mikulski J, Na-Ka Gawa M, Yen Ucla K. A bone ma-trix calcification-initiator non collagenous protein. J Physiol1977; 232: 115-127

31. Urist MR, Silverman BF, Buhring K, Dubuc FL, Rosenberg JM.The Bone Induction Principle. Clin Orthop 1967; 53 (7): 243-282

32. Weibel ER, Elias H. Measuring structures in space with randomprobes (planes, lines, points). In: Weibel ER, Elias H (Eds.)Quantitative Methods in Morphology (pp. 87-219). Berlin, Hei-delberg, New York: Springer-Verlag, 1967

33. Wolf K, Hamar J, Roesch I, Kalteis T, Schweiberer L. Die Dis-tanzosteosynthese der Kaninchentibia. Ein Beitrag zur Unter-suchung der Osteogenese im interfragmentären Spalt. Osteolo-gie 1997; 6 (4): 161-169

34. Wolff R. Knochenstabilitaet nach Kontakt und Spaltheilung.Eine tierexperimentelle Studie. In: Rehn I, Schweiberer L, Tsch-erne H (Eds.) Hefte zur Unfallheilkunde (S. 30-31). Berlin, Hei-delberg, New York, London, Paris, Tokyo: Springer, 1988

Address for correspondence: Konrad Wolf, M.D., Department ofTrauma, Hand and Reconstructive Surgery, Hospital MunichSchwabing, Teaching Hospital of Ludwig-Maximilians-UniversityMunich, Kölnerplatz 1, D-80804 Munich, Germany, E-Mail:[email protected]

36 K. Wolf, J. Hamar, E. Höcherl, T. Farkas, C. Pfister

Foreword

We think it is important to mention that all our inves-tigations into processes and substances which wouldaccelerate healing emanated from a profound concernand interest regarding the healing in bone. The startingpoint of all this research was an indepth discussion be-tween surgeons, a physiologist and a natural scientist.

Our collective research efforts constitute numeroushours of concentrated work which took place over aperiod of many years. This research paper was firstwritten in the German language and later for purposesof publication translated by the author into English.

Morphometric analyses of the kinetics of vascular and neural ingrowth intothe interfragmental gap following osteotomy

K. Wolf, J. Hamar*, E. Höcherl, T. Farkas*, C. PfisterDepartment of Trauma, Hand and Reconstructive Surgery, Hospital Munich Schwabing, Teaching Hospital ofLudwig-Maximilians-University Munich, Germany and *Central Institute of Traumatology, Budapest, Hungary


Keywords: Neuropeptide, Vessel growth, Nerve growth, Osteoneogenesis, Morphometry

Abstract

The transmitters and/or modulators Calcitonin gene-related peptide (CGRP), Substance P (SP), Neuropeptide Y(NPY) and Vasointestinal polypeptide (VIP) we posit are involved in bone growth, fracture healing and internalremodelling. We have immunohistochemical proof that neuropeptide positive fibers exist in normal bone and thisfinding convinced us that these substances affect the early phase of fracture healing. The exact timing of the ap-pearance of neuropeptide positive fibers, localisation in bone, chemospecifity and mode of genesis is, thus far, un-known.Extensive research was carried out on a model of distance osteosynthesis in the rabbit tibia. The osteotomy waslocated at the synostosis between the fibula and tibia. The periost and the surrounding soft tissue were removedsimultaneously. Primary and secondary antibodies were then used as an indirect immunohistochemical technique.Our findings are based on concrete stereological calculations of the length of nerve fibers rather than on conven-tional statistical evaluations.A 3 mm interfragmental gap was filled with vital granulation tissue, after distance osteosynthesis, with a plate. Af-ter histological preparation of tissue specimens from the interfragmental gap and the bone marrow beside the gap,the neuropeptides CGRP, Substance P and NPY, with the exception of VIP, were immunohistochemically ex-pressed. Evidence of CGRP, SP and NPY positive nerve fibers first appeared in the interfragmental gap on the 15thday postoperatively. At that point in time the length of CGRP positive nerve fibers measured 54 mm/mm3 in thegap with an increased growth in the bone marrow of 124 mm/mm3. Substance P also showed interfragmental ev-idence indicating the length of nerve fibers to be 48 mm/mm3 without having reached the length of nerve fibersin the bone marrow of 82 mm/mm3. Our findings indicate that the sprouting of CGRP- and SP-positive nerve fibers originates in bone marrow. A vas-cularisation in the early stage of osteoneogenesis, after fracture, appears to be unattainable without nerval pep-tidergic influence and transmission.

Morphometric analyses of the kinetics of vascular and neural ingrowth ... 37

Introduction and problems

The osteoneogenesis in the skeletal system of man is aprocess without interruption, from the embryonicphase to the death of an individual, and has its ownmomentum which cannot be influenced externally. Inspite of the numerous probabilities to have a lesion ina bone by old-age, disease, trauma or an operation, theregenerative process of osteoneogenesis occurs eachtime the bone is modelled. At the present time the in-fluence of the nervous system on vascularity and onosteoneogenesis is in question. In our summary of fac-tors which can influence fracture healing, the term in-nervation, is missing. A disturbed fracture healingwithin patients with severe problems allows us to pre-sume that the nerval factor will influence the course ofhealing significantly (2, 5, 22).

According to recent definitions innervation will al-so find consideration beside vascularity and biome-chanics (20, 21). Various factors were considered re-lating to the condition of primary and/or secondarybone healing: for example, wound infections, milieu oftissue and the effects of medications and hormones.These multifactorial components work in a supportingor inhibiting manner on the regulation mechanism anddetermine the result, which is osteoneogenesis. Frost(11, 12) described a multicellular mechanism on tissuelevel for the bone. He postulated that important „sup-porting“ cells exist together with effector cells (differ-enciated osteoblasts, osteoclasts, fibroblasts) evenwith ones own local blood supply and innervation (10,16).

Some questions and problems about innervationrelating to morphology, mediators and effects in themulticellular mechanism in tissue level have not beenresolved at the present time. Although much is still un-known concerning organotypical innervation of theskeletal system, we have proven that the supply ofbone and, in particular, the periost, definitely containsensorial and autonomous nerves. Alongside of theadrenergic components when all immunocytochemicalinvestigations were completed, we found evidence ofdifferent neuropeptides such as CGRP, Substance P,VIP and NPY. Transmitter and/or modulator sub-stances CGRP, SP, NPY and VIP are referenced by dif-ferent authors (6,7,14).

It is logical to assume that an effect in the earlystage of fracture healing occurs since we have im-munohistochemical proof of neuropeptide positivefibers in normal bone. The question then arises as tothe point in time of their appearance, localisation,

chemospecification and mode of creation, as all namedneuropeptides are known to have vasoactive effects invarious organ systems (6,7,10).

A model of distance osteosynthesis from the rabbittibia was developed in the course of investigating theimpact of the nerval factor in these processes and, inparticular, the clinical aspects. This model produced asufficient analysable quantity of vital granulation tis-sue without having any negative influence on fracturehealing (25).

We are of the opinion that immunohistochemicaland morphometrical methods should first be tried inorder to validate the existence of peptidergic nervefibers, with special attention on the role of the marrowspace in bone.

Material and methods

Animals

These investigations were carried out with thirty fe-male rabbits (Fa. Thomae GmbH, Biberach, Ger-many), with ages ranging from 3.5 to 4 months, andwith bodyweight of 3.9 ± 0.4 kg. During the entirespan of time necessary for our experiments the animalswere kept in single cages. „Principles of laboratory an-imal care“ were stringently followed (NIH publicationNo. 86-23, revised 1988).

Premedication and anaesthesia

Intragluteal mixed solutions comprising 0.4 mg/kgKetamin-HCl (Ketavet®, Pharmacia Upjohn GmbH,Erlangen, Germany), 5 mg/kg 5,6-Dihydro-2-(2,6-xy-lidino)-4H-1,3-Thiazin (Rompun®, Bayer, Lev-erkusen, Germany) and 16 mg Pentobarbital-Natrium(Narcoren®, Rhone Merieux GmbH, Laubheim, Ger-many) were injected as an initial dose into an ear veinper catheter (G22/24).

Distance osteosynthesis and operation technique

The operative approach was chosen between the ante-rior lodge of the tibia and the gastrocnemius caput lat-erale muscle. Muscles with periost were then pushedaside at the middle of the tibia and a seven-hole DCPplate (Maxillofacial product, 6 x 52 mm, Synthes,Umkirch, Germany) with 6 drill holes for AO cortical


screws was used for this procedure (2.7 mm, Synthes,Umkirch, Germany). With the help of an oscillatingsaw a 3 mm gap was produced by double osteotomyup to the height of the synostosis of both fibula andtibia. Deperiostation followed with the removal ofconnective tissue. The homicide of each group of rab-bits occurred on the 5th, 10th and 15th day postopera-tively (Narcoren®, 31.25 mg/kg).

Specimen taking and fixation for histology and immunohistochemistry

After removing the musculature and the plate, thegranulation tissue specimens were carefully taken outwith a scalpel, and directions were marked to identifythe borders of the interfragmental gap and the marrowspace. The fixation of specimens was performed witha buffered solution from Zamboni (26) for a durationof 48 hours at room temperature.

Embedding and histological cut

Specimens were taken in PBS and prepared with Di-methyl-Sulfoxid (Sigma-Aldrich Chemie GmbH, Mu-nich, Germany) in sacharose and aqua dest, and thenembedded in paraffin (SP 56 °C; Histosec, Merck,Darmstadt, Germany). Following these procedures thespecimens were placed into a graduated alcohol solu-tion. After these processes were completed 8 mm thickslides of paraffin were produced with a microtome(Jung, Heidelberg, Germany).

Indirect immunohistochemical techniques

Primary and secondary antibodies were specified fororigin and concentration. The secondary antibodiesshowed a definite diaminobenzidin reaction (9). Theconterstaining was prepared partially with hemalumand when the alcohol content reached its maximumlevel the specimens were embedded in Entellan® (Mer-ck, Darmstadt, Germany).

Morphometry and stereology

To produce density maps quantification of immunore-active structures was performed with the Overlap-Technique of Agnati (1) (density of immunoreactive

structures per unit volume). The method of Under-wood (23) called for calculating nerve length per vol-ume („length of line per unit volume - Lv“).

Statistics

Mean values per counting field were calculated for sta-tistical evaluation of all nerve fibers. A second controlgroup (specimens of rabbits without operation) was al-so evaluated. With the help of the known Students t-test additional significant calculations were madewhich we subsequently compared with both controlgroups at different postoperative points in time (day 5,day 10 and day 15 postoperatively). Different qualitiesof nerve fibers (CGRP, SP, VIP and NPY) were thenseparated. The test for connected random samples (17,19) was made by using the SPSS computer program(SPSS-Corporation, Chicago, USA). The three-dimen-sional calculation of nerve fiber length (mm/mm3) isbased on morphometrical and stereological formulas.Prevailing rules prevented statistical tests from beingperformed.

Applied devices

A lightmicroscope DIAPLAN (Leitz, Wetzlar, Ger-many) and an investigation microscope MICRO-PHOT-SA (Nikon, Düsseldorf, Germany) in combina-tion with a videocamera PROGRESSIVE 3 CCD(Sony, Tokyo, Japan) with the true color analysis sys-tem known as LUCIA G (Nikon, Düsseldorf, Ger-many) were used in our investigations. Photographswere produced by using the Stylos Photoprinter andphoto quality glossy film (Epson, Tokyo, Japan)

Results

Histology, model of distance osteosynthesis of therabbit tibia

The 3 mm wide gap was filled with vital granulationtissue, after osteosynthesis, with a plate. We studiedthe process of narrowly associated neuroneogenesisand osteoneogenesis, which are dependent upon thepostoperative space of time in the granulation tissueand in the marrow space near the osteotomy and pri-marily in the distal end of the tibia. After histologicalpreparation of these tissue specimens, as well as the


preparations of specimens from the interfragmentalgap of the distal marrow, it was possible to demon-strate CGRP, SP and NPY by using the indirect im-munohistochemical technique. The neuropeptide VIPcould not be demonstrated. The immunoreactive struc-tures were identified positively as nerve fibers, andwere partly visible as typical „string of pearl“-fibers(Fig. 1). These nerves were identified as having vesselconnections. Reactive structures and the immunoreac-tivity, as well as the quantity examined, proved to bedependent upon the point of postoperative time space.The first evidence of CGRP, SP and NPY appeared inthe interfragmental gap on the 15th postoperative day.In the control group all three qualities of nerve fiberswere definitely observed to exist in the tissue speci-mens obtained from the marrow space, while NPY im-munoreactive nerve fibers were found only in the mar-row space near osteotomy.

Statistical evaluation of neuropeptide positive fibersin the central blood vessel

The CGRP positive fibers of the marrow space in thecontrol group indicated a mean value of 0.14 ± 0.08SEM per vessel counting area (Tab. 1). A comparisonof the control group with the counting results of themarrow space produced from CGRP per counting fieldon the 5th postoperative day showed an adjusted meanvalue of 0.14 ± 0.10 SEM (p < 1.00). On the 10thpostoperative day the control group with CGRP posi-

tive nerve fibers indicated a significant increase of themean value (0.62 ± 0.16 SEM, p < 0.002). On the 15thpostoperative day another significant increase in themarrow space of the control group was evident (4.29 ±2.35 SEM, p < 0.09). On the same day (15th) we ob-served that the counting results (fibers per countingfield) of the interfragmental callus, when comparedwith the counting results of the marrow space, indicat-ed no significant difference (1.62 ± 0.37 SEM / 4.29± 2.34 SEM, p > 0.27, t-test for connected randomsamples) (Tab 1).

A mean value of 0.05 ± 0.03 SEM was establishedfor SP positive nerve fibers in the marrow space of thecontrol group (Tab 1). In a comparison with the con-trol group the countings indicated that Substance P(5th postoperative day) had no significant elevation inthe mean value (0.09 ± 0.06 SEM, p < 0.33). Howev-er, an impressive increase in the mean value appearedin the marrow space per counting field on the 10thpostoperative day (0.43 ± 0.15 SEM, p < 0.01). Equal-ly relevant, a significant elevation in the mean value(1.95 ± 0.45 SEM, p < 0.0001) was visible on the 15thpostoperative day. Within the interfragmental callus inthe marrow space on that same day no significant dif-ference at the vessel site per counting field was evident(1.62 ± 0.42 SEM /1.95 ± 0.42 SEM, p > 0.54, t-testfor the connected random samples) when comparedwith the neuropeptide positive fibers (Tab. 1).

In the marrow space of the control group, NPYpositive fibers reflected a mean value of 0.05 ± 0.03SEM per counting field. Countings from the control

Figure 1. Typical "string ofpearl"-fibers (↑) at a vessel(↑↑) in the interfragmentalgap on the 15th postopera-tive day. Magnification 630times, CGRP immunostain-ing, back dye with Hemalum


group showed no significant increase of the mean val-ue in the marrow space per counting field for NPYwhen taken on the 5th postoperative day (0.29 ± 0.17SEM, p < 0.20). On the 10th postoperative day we didnot see any significant increase in the mean value(0.19 ± 0.08 SEM, p < 0.08). Again, on the 15th post-operative day no significant increase was visible (0.24± 0.12 SEM, p < 0.10). When we compared the inter-fragmental callus with the marrow space on this dayno significant difference per counting area at the ves-sel site was seen (0.00 ± 0.00 SEM / 0.24 ± 0.12SEM, p > 0.056, t-test for connected random sample)(Tab. 1).

Stereological Calculation for Nerve Fiber Length

Data representing different fiber specifications wasposted on the density maps by using the overlap tech-nique (1). Our primary focus was directed to the ves-sel site in the counting area. The positive results ofthree ocular fields in specimens, which in every caseincluded seven animals, were added to the total roster.By using this particular method a numerical registra-tion of neuropeptide positive fibre calculations wasmade at different postoperative points in time. Mark-ings were made on the density maps to indicate the po-sition and number of CGRP positive nerve fibers. Themarrow space in the control group represented 3CGRP positive nerve fibers, and on the 15th postoper-ative day the number of CGRP positive nerves in-creased to 44 in the marrow space near the osteotomy.As previously stated, on the 5th postoperative day, noneuropeptide positive structures were found. However,we noticed that in the interfragmental gap the positiveCGRP fibers registered at 19. Regarding the immuno-histochemical specified nerve fibers, the development

in the postoperative course was measured with thehelp of stereological calculations of length in the vol-ume from Underwood (23). In the distal marrow spacethe nerve fibers on the 15th postoperative day showeda tangible increase in length when compared with thefirst control group (Fig. 2). It was observed that thelength of CGRP positive nerve fibers increased to 30(CGRP control group 4 mm per mm3, CGRP 15thpostoperative day 124 mm per mm3 marrow space).Substance P positive nerve fibers in the control grouphad a low starting value (1 mm per mm3 marrowspace), however, sometime later we noticed an in-crease in length measuring 80 (82 mm per mm3 mar-row space) (Fig. 2). The highest increase in length oc-curred in the interval between the 10th and 15th post-operative day (Fig. 2), at which time we noticed thatthe CGRP positive nerve fibers had impressivelylengthened from 19 to 124 mm per mm3.

Substance P positive nerve fibers increased from11 to 124 mm per mm3.. Only NPY (Fig. 2) showed adynamic which, by a decrease in value on the 10thpostoperative day (6 mm per mm3) revealed equallength on the 5th as well as on the 15th day (17 mm permm3). Observation of the length of CGRP positivenerve fibers on the 15th postoperative day in the inter-fragmental gap with the marrow space (Fig. 3) alsoshowed an increase in the length of nerve fibers in thegap (54 mm/mm3) and a definite growth in the marrowspace (124 mm pro mm3) (Fig. 3). At the same time wenoticed that Substance P increased in nerve fiberlength (48 mm pro mm3), however, Substance P didnot reach the values in the marrow space (82 mm permm3). Only NPY did not show any neuropeptide pos-itive fiber growth in the interfragmental gap. The mar-row space tissue at that point in time measured 17 mmnerve fibers per mm3.granulation tissue.

Table 1. Number of neuropeptide positive fibers in the bone marrow per counting field of the eye piece. Controlgroup and postoperative periods of time on the 5th to the 15th postoperative day. Mean values with standard er-ror of mean (SEM).

Group 1CGRP

2 Subst.P

3 NPY

1/2p-value

1/3 p-value

2/3 p-value

Control 0.14 ± 0.08 0.05 ± 0.01 0.05 ± 0.01 p<0.33 p<0.16 p<1.00

Day 5 0.14 ± 0.10 0.10 ± 0.07 0.29 ± 0.17 p<0.58 p<0.09 p<0.21

Day 10 0.62 ± 0.16 0.43 ± 0.15 0.19 ± 0.09 p<0.36 p<0.009 p<0.17

Day 15 4.29 ± 1.35 1.95 ± 0.45 0.24 ± 0.12 p<0.31 p<0.10 p<0.002


Discussion

Our model of distance osteosynthesis (25) is essential-ly different from other animal models which studyfracture healing processes. We found that the healing

of bone fractures exists without foremost participationof the periost up to the 25th postoperative day. As wasevident, 40 mm3 of granulation tissue grew and thisfact enabled us to continue with our investigations ofhistological, osteoneogenetic and immunocytochemi-

control

day 05 10 15

days

length of

nerve fibers

(mm/mm3)

50

100

10

20

30

40

60

70

80

90

110

120

CGRP

Substance P

NPY

Figure 2. Developmentof nerve fiber length ofthe bone marrow in thepostoperative course(mm per mm3) in com-parison with the controlgroup

interfragm. gap

length of

nerve fibers

(mm/mm3)

50

100

10

20

30

40

60

70

80

90

110

120

CGRP

Substance P

NPY

bone marrow

Figure 3. Comparison ofnerve fiber length fromthe interfragmental gapand the bone marrownear osteotomy at the15th postoperative day(mm per mm3)


cal processes. We are convinced that the vitality ofbone tissue from the fracture gap has been impressive-ly demonstrated by our research (24,25).

Further, we observed that the vessel sproutings inthe interfragmental gap, consistently and fundamental-ly were accompanied by a peptidergic nerve as seen inour histological photographs. We affirm that CGRPand Substance P positive nerve fibers are of primaryrelevance (10) and we are convinced that the nerval in-fluence in processes of fracture healing is vessel asso-ciated.

All our research efforts demonstrate that nervefibers emanating from the marrow space are of differ-ent chemospecifities. CGRP and Substance P in-creased, as indicated in Figure 2, from the 5th up to the15th postoperative day. This result provided us withconvincing and documented proof that CGRP im-munoreactive nerve fibers do most certainly increaseand that they are an inherent process in the fracturegap. We differ with Hukkanen et al. (14) that the es-sential origin of peripheral nerves can only be ob-served in the periost. In distinct contrast, our researchhas proven that immunocytochemical results are veri-fiable and this fact is validated by our findings thatCGRP and Substance P are processes of osteo- andvasculoneogenesis which have been demonstrated inthe evolution of our research. We restate our convic-tion that CGRP is one of the most powerful vasodilata-tory substances: CGRP stimulates the proliferation ofendothelial cells (13) and the osteogenic bone forma-tion in vitro (7), CGRP increases the bone mineralisa-tion in vivo (3,4) and it increases the capillary perme-ability (8) which is induced from the Bradykinin andPlatelet Activating Factors, CGRP also increases theforming of c-AMP as well as the formation of os-teoblasts in certain species (15). A similiar but weakerdevelopment of Substance P positive nerve fibers cor-responds with the established point of view: that thisoctapeptide plays a major and profound role in angio-genesis (27).

Morphometrical analyses of our immunohisto-chemical specimens were performed to indicate, ob-jectively, the increase of peptidergic structures in thepostoperative sequence. Results charted on the densitymaps will illustrate the time-dependent distribution ofCGRP positive fibers, which show a progressivesprouting of nerve fibers from the osteotomy near themarrow space into the granulation tissue. With thehelp of the existing overlay techniques, Agnati andcolleagues (1) have led the way to the still unsolvedproblems of how to quantify immunocytochemical re-

actions with mathematical verification of structures inthe area per unit. Agnati developed this method prima-rily to analyse neuronal tissue in the brain. His workinspired and helped us to solve various problems inour research particularly because of the fact that, in thecentre of any region of interest, a vessel diameter ex-ists. Our findings indicate that nerve association to avessel site is an inherent as well as an essential factorin osteoneogenesis after any bone fracture.

The establishing of Hukkanen (14) allows us toagree that: „CGRP is most probably continuously andincreasingly released from stimulated nerve terminalsat the site of the trauma and inflammation. Therefore,we are of the opinion that this peptide will have directeffects on bone derived and other surrounding cells“.And it is evident from our extensive research that bonemarrow is a most important area for the sprouting ofvasculo- and osteoneogenetic processes.

We wish to point out that our investigations withLaser-Doppler-Flowmetry (characterization of vari-ables of microcirculation) revealed the influence of ex-ogen supplied neuropeptides. CGRP and Substance Pproduced measurable effects in the vessels exactly onthe 10th and 15th postoperative days. All results fromour research prove the existence of immunoreactivestructures. Our findings indicate a progressive influ-ence in fracture healing when CGRP is applied local-ly, and above all, to patients who have critical boneconditions. These findings consistently prove that ameasurable vascular effect is attainable when usingextremely low concentrations of neuropeptides (18).

At the present time we cannot verify that VIP im-munoreactive structures exist in the postoperativecourse. It is a well known fact however that VIP posi-tive nerve fibers exist in the periost and in the neigh-boring connective tissue, and we know that NPY im-munoreactivity can only be demonstrated up to the25th postoperative day, and only in the marrow spacenear osteotomy. As we did not find any fibers in thetissue of the interfragmental gap, we concluded thatthe known osteotrophic effect of NPY after osteosyn-thesis will only be effective at the level of remodellingof the de novo created bone.

We arrived at the following conclusions:a) The „sprouting out“ of CGRP and SP positive nerve

fibers definitely emanates from bone marrow, andthat specific care is essential and should be given tothe marrow space during surgical interventions afterfracture.


b) In the early phase of osteoneogenesis the role ofvascularisation after fracture is unlikely withoutnerval peptidergic influence or procurement.

c) We recommend that further clinical investigationsand basic research into innervation and fracturehealing be carried out in the not too distant future.All our research findings strongly indicate thatCGRP and Substance P have major significance inprocesses of fracture healing. We have proven thatthese substances most clearly affect cellular com-partments of bone and/or granulation tissue.

Acknowledgements

We express our appreciation to Mrs. Ellen Maier forher excellent production of histological specimens andtechniques of dye.

We also express our appreciation for the support ofour project to the Wilhelm Sander Foundation, Neu-stadt a.d. Donau, file 91.056.1 and the BMBF at theDLR Bonn, project number UNG-054-96 and project-number UNG 99/001.

References

1. Agnati LF, Fuxe K, Janson AM, Zoli M, Härfstrand D. Quanti-tative analysis: Computer-assisted morphometry and microden-sitometry to immunostained neurons. Immunochemistry 1991;13: 205-223

2. Amare SG, Jonas V, Rosenfeld MG, Ong ES, Evans RM. Alter-native RNA Processing in calcitonin gene expression generatesmRNAs encoding different polypeptide products. Nature 1982;298: 240-244

3. Ancill AK, Bascal ZA, Whitaker G, Dacke CG. Calcitonin gene-related peptide promotion transient radiocalcium uptake intochick bone in vivo. Expl Physiol 1991; 76: 143-146

4. Ancill AK, Bascal ZA, Whitaker G, Dacke CG. Effects of rat andchicken calcitonin gene-related peptides (CGRP) upon calciummetabolism in chicks. Regul Pept 1990; 30: 231-238

5. Bernhard GW, Shih C. The osteogenetic stimulating effect ofneuroactive calcitonin gene-related peptide. Peptides 1990; 11:625-632

6. Bjurholm A, Kreicbergs A, Brodin E, Schultzberg M. SubstanceP- and CGRP-immunoreactive nerves in bone. Peptides 1988; 9:165-171

7. Bjurholm A, Kreicbergs A, Terenius L, Goldstein M,Schultzberg M. Neuropeptide Y-, Tyrosin Dehydroxylase andvasoactive intestinal polypeptide-immunoreactive nerves inbone and surrounding tissues. J auton Nerv Syst 1990; 25: 119-125

8. Brain SD, Williams TJ. Substance P regulates the vasodilatoryactivity of calcitonin gene-related peptide (CGRP) and media-tors of increased vascular permeability. Br J Pharmacol 1985;86: 855-860

9. Cote´ SL, Ribeiro DA, Silva A, Cuello AC. Methods in the Neu-roscience. IBRO-Handbook Series 1993; 14: 147-168

10. Duncan CP, Shim SS. The autonomic nervous supply of bone. JBone and Joint Surg 1977; 53-B: 323-330

11. Frost HM. The Biology of Fractur Healing. J Clinical Or-thopaedics and Related Research 1989; 11: 283-293

12. Frost HM. Bone development during childhood. A tutorial(some insights of a new paradigm). In: Schönau E (Ed.) Pedi-atric osteology: New developments in diagnostics and therapy(pp. 1-39). 1996

13. Haegerstrand A, Dalsgaard CJ, Jonsson B, Larsson O, Nilsson J.Calcitonin gene-related peptide stimulates proliferation of hu-man endothel cells. Proc Natl Acad Sci 1990; 87: 3299-3303

14. Hukkanen M, Konttinen YT, Santavirta S, Paavolainen P, GuXH, Terenghi G, Polak JM. Rapid proliferation of calcitoningene-related peptide immunoreactive nerves during healing ofrat tibial fracture suggests neural involvement in bone growthand remodeling. Neuroscience 1993; 54 (4): 969-979

15. Michelangeli VP, Fletcher AE, Allan EH, Nicholson GC, MartinTJ. Effects of calcitonin gene-related peptide on cyclic AMP for-mation in chicken, rat and mouse bone cells. J Bone Miner Res1989; 4: 269-272

16. Parfitt AM. Bone Remodeling. Journal of Bone and Mineral Re-search 1990; 5 (5): 533

17. Ramm B, Hofmann G. Spezielle Tests. In: Ramm B, Hofmann G(Hrsg.) Biomathematik und medizinische Statistik (S. 305-312).Stuttgart: Ferdinand Enke

18. Rosenfeld MG, Mermod JJ, Amara SG, Swanson LS, SawchenkoPE, Rivier J, Vale WW, Evans RM. Production of a novel neu-ropeptide encoded by the calcitonin gene via tissue-specificRNA processing. Nature 1983; 304 (7): 129-135

19. Sachs L. Statistische Prüfverfahren. In: Sachs L (Hrsg.) Ange-wandte Statistik (S. 242-244). Berlin, Heidelberg, New York,Tokyo: Springer, 1984

20. Schweiberer L, Wolf K. Die Pathophysiologie der Knochen-heilung. Hefte zur Unfallheilkunde 1992; 222: 26-31

21. Schweiberer L, Wolf K. Pathophysiologie der Knochenbruch-heilung. Osteologie, Knochentumoren 1995; 4 (Suppl. 1): 5

22. Sevitt S. Control of bone repair. In: Sevitt S (Ed.) Bone Repairand Fracture Healing in Man (pp. 41-51). Edinburgh, London,Melbourne, New York: Churchill Livingstone, 1981

23. Underwood EE. Length of Line per Unit Volume. In: Under-wood EE (Ed.) Quantitative Stereology (pp. 43-44). Reading-Massachusetts, London, Don Mills-Ontario: Addison-WesleyPublishing Company, 1969

24. Wolf K, Hamar J, Höcherl E, Farkas T, Pfister C. Knochenneu-bildung-Osteokonduktion-Osteoinduktion. Unter besondererBerücksichtigung des interfragmentären Raumes. TraumaLinc2000; 1: 54-60

25. Wolf K, Hamar J, Rösch I, Kalteis T, Schweiberer L. Die Distan-zosteosynthese der Kaninchentibia. Ein Beitrag zur Unter-suchung der Osteogenese im interfragmentären Spalt. Osteolo-gie 1997; 6: 161-169

26. Zamboni L, DeMartino C. Buffered picric acid formaldehyde: Anew, rapid fixative for electron microscopy. J Cell Biol 1967; 35:148

27. Ziche M, Morbidelli L, Masini E, Amerini S, Granger JH, Mag-gi CA, Geppetti P, Ledda F. Nitric oxide mediates angiogenesisin vivo and endothelial cell growth and migration in vitro pro-moted by substance P. J Clin Invest 1994; 94: 2036-2044

Address for correspondence: Konrad Wolf, M.D., Department ofTrauma, Hand and Reconstructive Surgery, Hospital MunichSchwabing, Kölnerplatz 1, D-80804 München, Germany, E-Mail:[email protected]

44 K. Wolf, E. Höcherl, T. Derfuß, A. Krug

Laser-doppler-flowmetry and absorption-tissue-spectrometry of thetransposed groin flap – A comprehensive and independent analysis of microcirculation

K. Wolf1, E. Höcherl1, T. Derfuß2, A. Krug2

1Department of Trauma, Hand and Reconstructive Surgery, Hospital Munich Schwabing, Teaching Hospital ofLudwig-Maximilians-University Munich, Germany; 2LEA Medizintechnik GmbH, Gießen, Germany


Keywords: laser-doppler-flowmetry, microcirculation, absorption-tissue-spectrometry, oxygen-saturation, hemo-globin concentration

Abstract

Our actual surgical experience with the transposed groin flap provides valid proof that an autonomous blood cir-culation is derived from the recipient site before disconnection of the pedicle of the flap. The tangible benefits ob-tained from the use of Laser-Doppler-Flowmetry (LDF) and the Absorption-Tissue-Spectrometry (ATS) were ap-parent during two operations as outlined in this report. Case reports have been published which claim that an au-tonomous blood supply from the recipient site failed to appear. We think, therefore, that the autonomy of bloodvessels, from the groin flap pedicle, can safely be excluded so that surgeons can determine the exact point in timefor a disconnection of the groin flap when using guidelines from numerical values. Standardized procedures andrules need to be established for the application of LDF and ATS as we are convinced that these systems can be thebasis for training broad groups of potential users worldwide. These methods have been used on two patients andwe documented the preoperative conditions of the pedicle disconnections. The LDF system represents a physio-logical term which is synonymous with „microcirculation“. The numerical values represent the conditions of theexisting blood volume in the capillaries as well as the velocity of blood particles which float throughout the ves-sel system. The ATS system measures the oxygen (O2) saturation of the hemoglobin in the analysed tissue volumeas well as in the local concentration. While studying the groin flap transposition to the upper extremity of the in-jured hand we made a comparison with the contralateral hand as well as with the contralateral groin region. Onthe day prior to the operative separation we analysed the clamping of the pedicle during a two hour period. Wewere faced with specific demands which were strictly followed prior to the operative disconnection of the pedicleof the flap. We recognised the importance that there be a continuous flow of blood in the flap without any decreaseof counts after the clamping procedure ends. The tolerable decrease of oxygen saturation of the hemoglobin meas-ured slightly below 25 % of the starting value. We realized that for this procedure to be successful the hemoglo-bin concentration should not decrease once the clamping procedure begins and that this procedure requires an in-terface with granulation tissue. After a period of three weeks we discovered that the transposed groin flapmatched the tissue proportionality of the upper extremities. The LDF and the ATS systems are excellent methodswhich have helped us to resolve the question of how to monitor a sufficient blood supply flowing out of the re-cipient site. Having utilized these methods we have proven that the retrograde blood supply after clamping of theflap pedicle from the upper extremity is measurable. Also by using the channel system of the flexible LaserDoppler Opto Flow it was possible for us to obtain a full spectrum of safety. These results confirmed and rein-forced our decision to proceed with the pedicle disconnection.

Laser-doppler-flowmetry and absorption-tissue-spectroemtry 45

Foreword

A group of surgeons, technicians and natural scientistsshared ideas, experiences and observations concerningmicrocirculation, hence the origin of this report.

Numerous indepth, yet informal discussionsbrought an awareness that these experiences and ob-servations would be helpful to other investigators. Theinformation contained in our report delineates the ben-efits obtained when the Laser-Doppler-Flowmetry sys-tem is used in combination with conventional opera-tive procedures.

The LDF system with all its components was de-veloped and manufactured in Germany. This remark-able system provides accurate measurements duringcomplex operative procedures thus allowing surgeonsto monitor bodily processes while patients undergo se-rious operations following catastrophic accidentswhich, of course, necessitate skillful surgical interven-tion.

We are convinced that the LDF system can andshould be widely used and as we have seen, in the longrun, patients can reenter their daily lives with an abili-ty to resume activities and duties.

As a collective effort, this report was written in theGerman language and translated by Dr. K. Wolf intoEnglish for publication.

Introduction

The operative treatment of a traumatised hand can becharacterized as the most important challenge for thesurgical community today. The acute situation as wellas the reconstructive phase require transposed flaps orfree flaps (26). Although these operative proceduresrequire careful monitoring of the microcirculation nostandardised procedures have been established for thisnew and innovative technology (2,3,11,19,23). Morerecent experimental research into the field of Laser-Doppler-Flowmetry (27) provides promising and tan-gible applications. We have seen that by using Laser-Doppler-Flowmetry (LDF) and the Absorption-Tissue-Spectrometry (ATS) a unique and unparalled service isavailable to patients who suffer from serious trauma.Unfortunately, prescriptions and formulas for this LDFsystem which accurately measures index values for pa-tients who suffer catastrophic accidents to the handand who require immediate surgery as well as tissueflaps are not available at this time. Our experience hasconvinced us that there is need for further research!

Problems

As is generally known the primary autonomy of theblood supply is present in the vessel anastomoses offree flaps as well as in the vessels of the pedicle of atransposed flap. This autonomy can be lost as a resultof the angio and vasculoneogenesis in the granulationtissue in the recipient site. Case reports have been pub-lished which state that there is a persistent autonomyof the primary blood supply. We have a conflicting ob-servation: operative damage to the pedicle of the flapcan, and often does, lead to necrosis of the flap (11).

The primary objective for a transposition of thegroin flap is to cover tissue defects in the hand. Dis-connection of the vessel pedicle occurs approximatelythree weeks later. A retrograde blood supply at thisjuncture flows through the de novo growing granula-tion tissue which can be explained by angioneogenesisand vasculoneogenesis (1).

A transposition of the groin flap is generally per-formed when there are serious injuries to the hand andreplantation is essential. Surgeons are often obliged tooperate on damaged and underperfused tissue. Todaysurgeons are faced with a huge challenge: how to findthe exact point in time for the disconnection of thepedicle. The importance of the time factor is vastly dif-ferent from other recommendations which encouragedisconnection of the pedicle of the flap after threeweeks. We recognise therefore that a sufficient retro-grade blood supply between the transponated flap andthe surface of the recipient site is of major importancefor transponate survival. An unrecognised demarcationfrom the necrotic tissue layer can become a seriousrisk factor and this could delude even the most experi-enced investigator and thus lead to the flap being loos-ened.

The primary existing autonomy of a vessel supplycoming out from out of the groin site generally dis-solves after this point in time when the disconnectionof the flap pedicle occurs. For the purpose of continu-ity of the blood flow two procedures at two differentpoints in time are necessary. During our first operationwhen the disconnection of the main vessels in thepedicle of the flap took place (26), a thin skin and tis-sue bridge remained. During the second operation thepedicle of the flap was totally cut out.

The successful use of standardized LDF wasdemonstrated in our analysis of microcirculation byour two patients. It is important to mention that duringthese operations a misjudgement would have resultedin a loss of the flap for both patients, culminating in


the loss of some segments of the hand. Our surgicaltask was to exclude a persistent primary autonomy ofthe supplying vessels in the pedicle of the groin flap,and then to prove that a newly growing angioneogen-esis with vasculoneogenesis is possible in the recipientsite. We restate our opinion that guidelines need to beassembled which relate to all applications of LDF andATS. When industrially produced equipment is madeavailable to a broad group of investigators we are con-vinced that this will be of invaluable help in demon-strating the benefits derived from obtaining patho-physiological numerical values which will assist sur-geons during operative procedures.

Patients and methods

Casuistry

Case Report 1

A 20 year old man sustained multiple injuries when asa passenger in the front seat of a car an accident thrusthim out of the vehicle which turned over. The youngman struggled to release himself, however his righthand was caught in parts of the car and in the ground.

Ninety five minutes later he was flown to our hos-pital by helicopter. The intraoperative situation re-vealed a subtotal amputation of the right hand with aglove like avulsion injury and dismemberment at fourdifferent levels to his upper extremity.

The skin and the soft tissue layer was glove likeavulsed (Fig. 1a) outward from the carpus. The handwas axial devided in two parts: the thumb with thethenar eminence and the palm with four long fingers.Fractures were sustained in numerous parts of the ex-tremity. There was a fracture of the first metacarpalbone as well as fractures in the palm of metacarpal III,IV and V (level 1). Additionally, a fracture of thescaphoid (level 2) and a fracture of the ulna 7 cm prox-imal of the wrist (level 3), and a supracondylar fracture(level 4). The radial artery with the superficial palmararch was totally torn out. The ulnar artery of the fore-arm was disconnected. The tendons of the extensorpollicis longus muscle, abductor pollicis longus mus-cle and flexor pollicis longus muscle were pulledaway. At the dorsal site the extensor tendons of longfingers in the forearm were also pulled out. Our surgi-cal sequence began with the pin osteosynthesis ofmetacarpal bones I, III, IV und V for fixation of thepartially amputated hand (Fig 1b). The next step began

with the revascularisation of the midhand and the longfingers with the suture of the ulnar artery. We contin-ued with a preparation of the pulled out vessels of thesuperficial palmar arch and the radial artery for therevascularisation of the thumb and the thenar emi-nence. Additionally we prepared the proximal stumpsof the medianus nerve which had segmental defects. Asplitting of the lodges was performed at the forearmand when this procedure ended we used superficialskin to cover the damaged skin.

Three days later necrosectomie of the thumb mus-cles was performed. A venous bypass was given to thethumb to improve the blood circulation of the thumbeminence. Two venous anastomoses followed. Whileperforming the operation on the ulnar shaft fracture weinserted a low contact DC plate (Synthes, Umkirch,Germany). A supracondylar fracture of the same ex-tremity was prepared with cannulated screws. Six dayslater (day 9 posttrauma) the major step was taken forthe transposition of a groin flap (Fig. 1c, 1d) for tissuecovering in the palm and on the dorsum of the hand.We first performed a radical debridement with resec-tion of the differentiated thumb. Additionally we per-formed a skin graft transplantation at the forearm. Af-ter a continuous training time, which started at the be-ginning of the second week and continued until theend of the third week, the retrograde vascularisationproved to be successful again with the support of theLDF technology. The disconnection took place on thefollowing day. An analysis followed with the help ofLDF before cutting through the pedicle of the flap(Fig. 1e, 1f). Disconnection of the flap occured on the27th day after the transposition.

Case report 2

While working a night shift at a newsprint company a38 year old woman inadvertently placed her left handunder a moving mechanical assembly drive. Completeamputation resulted and a glove like avulsion injuryoccured in the soft tissue cover of the thumb. Only theskeletal and tendon parts were sustained (Fig. 2a). Theamputated part of the thumb was totally disturbed andonly soft tissue in one third of the circumference couldbe seen. All the arteries and nerves were pulled out.We determined that an immediate operation was es-sential in order to help this patient (Fig. 2b). One weekafter the postoperative training began, we clamped thepedicle of the groin flap. Until the end of the thirdpostoperative week the clamping training increased to


Figure 1. Case Report 1. The 20 year old patient sustained multiple injuries when as a passenger in the front seatof a car an accident thrust him out of the car which turned over.Figure 1a. Intraoperative findings of a patient with glove like avulsed soft tissue. The hand was axial devided intwo parts: the thumb and the palm with the four long fingers.Figure 1b. Pin osteosynthesis of the metacarpal bones for fixation of the partially amputated hand.Figure 1c. Transposition of a groin flap to the hand, nine days posttraumatic.Figure 1d. Tissue covering in the palm and on the dorsum of the hand. Radical debridement with resection of thedamaged thumb.Figure 1e. Analysis with the help of Laser-Doppler-Flowmetry.Figure 1f. Retrograde vascularisation proved to be successful. LDF numeric values without decrease of the meas-uring level after clamping.


Figure 2. Case report 2. A 38 year old women inadvertently placed her left hand under a moving mechanical as-sembly drive.Figure 2a. Complete amputation with glove like avulsion injury. Destruction of the amputated thumb with soft tis-sue in several levels.Figure 2b. Immediate operation and covering with a transponated groin flap for transposition to the left thumb.Figure 2c. Measuring system at the groin flap transponate. Figure 2d. Retrograde blood supply was secured with help of Laser-Doppler-Flowmetry and Absorption-Tissue-SpectrometryFigure 2e. Radial sight after modified Wrap-around-transfer of the ipsilateral big toe to the hand.Figure 2f. Wrap-around-Transfer, palmar sight of the hand.


two hours. With help from the LDF (Fig. 2c) and theATS systems the retrograde blood supply flowingthrough the forearm was secured before cutting thegroin flap (Fig.2d). We disconnected the pedicle of thegroin flap one day later. The operative result showeda vital groin flap. A wrap around flap from the ipsilat-eral big toe was transplanted after 9 months (Fig. 2e,2f).

Laser-Doppler-Flowmetry

The LDF system demonstrates and validates the neces-sity for accurate measurements of blood movementduring operative procedures which measure the in-creasing and/or decreasing effects of procedures andthe accurate measurements of microcirculation (7).Laser-Doppler-Flowmetry reveals, in special cases,the relevance of blood flow in a noninvasive and re-producible manner and these systems assist surgeonsby evaluating the effects of procedures, and at thesame time numeric values of increased and/or de-creased microcirculation. The physiological term „cir-culation“ is valid and determines two important pa-rameters: first the volume of blood in the capillariesand second, the velocity of the blood flow within theblood vessels.

The LDF system produces an output signal whichis proportional to the blood cell perfusion. This is veryimportant as microcirculation measures the product ofthe mean flow velocity as well as the mean count ofblood cells in the defined tissue volume which ismeasured by the laser beam.

Formula:

Microcirculation number of mean(Blood cell perfusion) = blood cells x velocity

circulating of the cellsin the tissue specimens

The LDF system provides a highly sensitivemethod by which measuring values can be correlatedwith real values of bloodperfusion. It searches out themeasured numeric counts which are relative values. Itis essential to consider the fact that in different tissuequalities the factor of proportionality varies. The LDFsystem allows a continuous measurement without in-fluencing the circulation of blood cells. Not to be over-looked is the fact that the blood perfusion units repre-sent relative values of a motility standard.

Utilisation of the LDF equipment (Opto Flow,LEA, Gießen, Germany) scans and measures the sur-face of the skin down into a tissue depth of 14 mm.The relative blood flow can be measured in a real timeprocedure. The time factor allows a concomitant ob-servation: first, the changes of blood flow in differenttissue depths and second, at different measuring pointssuch as extremities, the groin, the abdomen, etc. Theblood flow measurement is not restricted to the totalvolume alone but parallels a deep selective real timeanalysis of the blood flow in different tissue layers.

The measuring volume can be calculated up to3,600 mm3. The laser beam frequently shifts and fo-cuses on the streaming blood particles. The LDF sys-tem analyses the optical doppler frequency with FastFourier Transformation in real time. The higher thespeed of the doppler shift the more erythrocytes canbe seen in the tissue.

Absorption-Tissue-Spectrometry

The Opto Flow of ATS provides a parallel benefit. TheATS system is suitable for a non invasive determina-tion of the oxygen supply in the blood circulation ofthe tissue. Serious scientific questions arise: for exam-ple, analyses of angiogenesis. With precision one ob-tains the value of oxygen saturation in the hemoglobin.At a parallel point in time the local concentration ofhemoglobin will be evaluated in the tissue being inves-tigated. Hemoglobin of the arterial circulation contains100 % oxygen. In the peripheral tissue the oxygen willremain alongside the capillaries in the cells. There isan adequately low saturation of the hemoglobin in thetissue parts at the venous ends of the capillaries whichkeeps on flowing into the „venules“ and veins.

The equipment is a two channel system. It straight-ens and allows a determination of the oxygenation ofhemoglobin (SO2 %) as well as in the hemoglobin con-centration (HB conc) into two tissue depths.

The laser light of the ATS system penetrates tissueand displays the colour of blood. The system alsoanalyses the visible light and evaluates the oxygen sat-uration via a colour spectrum. Whenever white lightenters tissue a spectral change occurs. The saturationof hemoglobin obtained from a red coloured dye in theblood produces different colours: the saturated arterialblood appears as light-red, while the venous blood ap-pears as a dark-red or violet colour.

With the help of the ATS system the local concen-tration of hemoglobin can be calculated. The mecha-


nism operates with minimal sound while the absorp-tion values of the hemoglobin are extracted from thetotal wave length area. The level of hemoglobin valuescorrelates exponentially with the calculated integralvalues. The basic scatter spectrum can evaluate thenorm of the tissue spectrum with the help of the ATSsystem. This system allows one to quantify the localtissue value of hemoglobin concentration in a repro-ducible and verifiable manner.

The measurement procedure can be used with asonde measurement in combination with the LDF sys-tem. Thus the energetic cell metabolism which is con-nected to the receptivity of oxygen can be evaluated.The level of the oxygen saturation correlates exponen-tially with all calculated integral values.

Measurement table

The counts of LDF were measured, as were the oxy-gen and hemoglobin concentrations. Measurementswere taken at the same locations parallel to the on lineprocedure and in real time. The measurement table included three parts:1. The LDF and ATS values were compared with those

of the contralateral thumb as well as with the con-tralateral region of the groin.

2. A clamping of the groin flap was analysed over atime space of 2 hours (Abb. 1e, 1f), and betweeneach measurement there was an interval of 15 min-utes.

3. The development of a transposed groin flap wasrecorded: at rest, at a clamping phase, and at thetime when the clamp was opened.

Statistics

For a statistical evaluation changes in oxygen satura-tion were calculated in mean values with Standard Er-ror of Mean (SEM). In the control group various meas-urements were taken: at the contralateral thumb and inthe region of the contralateral groin. Additional andsignificant calculations were also taken with the helpof the Students t-test for purposes of comparison. Thetest for connected tissue samples (15, 17) from thecomputer program SPSS was used (SPSS-Corpora-tion, Chicago, USA).

We required specific proof of reactions concerningthe comparability with all contralateral body areas.This necessitated an indepth investigation of the non-

operated dorsum of the hand at the contralateral side,as well as in the forearm of the operated side.

We chose two levels of tissue for measurement outof all possible depths. The first was at a level of 2 mmwhich corresponded with the microcirculation of theskin. The second level corresponded with a depth of 8mm from the subcutaneous tissue of the groin flap.

Results

Groin flap in comparison with thumb and groin flapregion of the contralateral side

Different depths of measurement indicated diverse re-sults in measurement of the microcirculation (Fig.3a).From the first measurement level (2 mm depth) thegroin flap measured a value of 1.1 Counts. The meas-urement level of the second depth (8 mm depth) indi-cated 2.3 Counts (Fig. 3a). In the deeper tissue layerwe noticed that the microcirculation revealed a highercount value, and at the contralateral thumb the meas-urements were 0.8 Counts (level 1) and 2.4 Counts(level 2). It was remarkable that the value from thecontralateral thumb of 2.4 Counts was similar to thetransposed groin flap which measured 2.3 Counts. Wewere truly astonished when the microcirculation of thecontralateral groin area in the measuring level 2showed a decreased value from one third to 1.4Counts.

The ATS system indicated a high oxygen saturationin the transposed groin flap (Fig. 3b) with 93.1 ± 1.4%. This value was statistically significantly higher incomparison with the mean value of the contralateralthumb (60.6 ± 0.6 % oxygen saturation; p < 0.000) andthe contralateral groin region (55.3 ± 0.4 % oxygensaturation; p < 0.000). In the transposed groin flap theATS system showed a hemoglobin content of 64.5 ±0.3 (Fig. 3c). A comparison with the measuring valuesof the thumb of the contralateral side revealed a signif-icantly lower need for hemoglobin (47.3 ± 0.2; p <0.000), and the reaction from the contralateral groinarea registered lower (34.4 ± 0.6, p < 0.000).

Clamping procedure at the groin flap and progressduring 2 hours

The clamping procedure at the base of the transposedgroin flap (Fig. 4a) showed a decrease in the microcir-culation with a low count value. During the clamping


Figure 3. Development of Microcirculation, O2-Saturation of Hemoglobin, Concentration of Hemoglobin.Figure 3a. Microcirculation. Different depths of measurement indicated diverse results of the microcirculation.Figure 3b. O2-Saturation of Hemoglobin.(*) = statistical significant difference of the groin flap numeric values to the contralateral thumb and to the con-tralateral groin region.(#)(+)(0) = statistical significant difference of the groin flap numeric values. Comparison of the measured numer-ic values with the starting value. Figure 3c. Concentration of Hemoglobin.(*) = statistical significant difference of the groin flap numeric values to the contralateral thumb and the con-tralateral groin region.(+)(0) = statistical significant difference of the groin flap numeric values. Comparison of the measured numericvalues from the clamping procedure with the starting value.

1

2

3

50

100

10

20

30

40

50

60

70

80

a) Laser-Doppler-flow

Without clamping

15 minutes after start of clamping

End of clamping

b) 02-Saturation of Hemoglobin

c) Concentration of Hemoglobin

15 minutes after the

end of clamping

Without clamping


End of clamping

15 minutes after the

end of clamping

Groin flap without clamping

End of clamping

Groin flap 15 minutes after

end of clamping

Groin flap without clamping



End of clamping

Groin flap 15 minutes after

end of clamping

Flow

[AU]

Hb SO2

[%]

rHb

[AU]

*

#+

o

+o

Groin flap, Level 2

Skin of the flap, Level 1

*


Figure 4. Development of Microcirculation, O2-Saturation of Hemoglobin, Concentration of Hemoglobin over atime period of 2 hours.Figure 4a. Microcirculation. Time point 0 (min) shows the measured numeric value of the groin flap before the be-ginning and at the start of the clamping procedure.o o = Level 1 = Level 2Figure 4b. O2-Saturation of Hemoglobin.(*) = significant decrease of O2-saturation during the clamping procedure compared to the numeric value of thestarting point.Figure 4c. Concentration of Hemoglobin.(*) = significant difference to the measured numeric value of the starting point. Slight but significant increase ofhemoglobin concentration

1

2

3

100

61

62

63

64

65

66

67

68

a) Laser-Doppler-Flow

b) 02-Saturation of Hemoglobin

c) Concentration of Hemoglobin

0 15 30 45 60 75 90 105

0 15 30 45 60 75 90 105

0 15 30 45 60 75 90 105

Flow

[AU]

Hb SO2

[%]

rHb

[AU]

60

70

80

90

* * * **

* *

*

*

* *

*

120

120

120

*

*


procedure, which lasted two hours, the microcircula-tion in the measuring level 2 at the groin flap was rel-atively constant with values ranging from 2.3 to 2.5Counts (Fig. 3a). In measuring level 1 a temporary andstaggered decrease of microcirculation from 1.1Counts to 0.5 Counts was visible. The measurement ofoxygen saturation during the clamping procedure (Fig.3b) showed a base value of 93.1 ± 1.4 %. After a 15minute delay there was a significant statistical de-crease to 76.8 ± 0.2 % of oxygen saturation (p <0,000). An evaluation of the hemoglobin concentra-tion followed (Fig. 3c) and in a comparison with thestarting point of 64.5 ± 0.3, we found no significantdifference in value (63.0 ± 0.0; p < 0.51). After remov-ing the clamp the numeric value increased significant-ly (66.6 ± 0.2). Fifteen minutes later we revealed a sig-nificant increase to 71.0 ± 0.1 Counts.

General view of tissue reactions from the groin flap

In general the LDF results (Fig. 4a) obtained from theclamping procedure (level 1) of the transposed groinflap indicated a flow of 1.1 Counts. The microcircula-tion increased during the clamping procedure to a val-ue of 1.2 Counts and decreased during a period of twohours to 0.8. Towards the end of the measuring proce-dure we observed a further decrease to 0.5 Counts.

A totally different situation occurred when wemeasured level 2. Prior to the clamping procedure themicrocirculation in the measuring level 2 began with apoint value of 2.3 Counts. Fifteen minutes later we no-ticed a slight increase to 2.4 Counts. This elevated lev-el remained constant for a duration of 2 hours. We re-moved the clamp 15 minutes later at which point thecount value registered 2.3.

The ATS system (Fig. 4b) showed, in comparisonwith the starting point (93.1 ± 1.14 % oxygen satura-tion), a slight nonsignificant increase (96.0 ± 1,1) andthen a significant decrease to 76.8 ± 0.2 % oxygen sat-uration and this decrease occurred 30 minutes after theclamping procedure (p < 0.000). At the end of theclamping procedure the oxygen saturation measuredwas 79.2 ± 0.3 %. This was significantly lower in com-parison with the starting point level (93.1 ± 1.4; p <0,000). Once the clamping procedure ended we ob-served a significant increase (80.8 ± 0.6 % O2-satura-tion) in comparison with the end value of the clampingprocedure (79.2 ± 0,3; p < 0.02).

The measurement of the hemoglobin concentration(Fig. 4c; after 15 minutes clamping time) showed no

significant change (64.3 ± 0.2; p < 0.51) in comparisonwith the situation at the starting point (64.5 ± 0.3). Af-ter 30 minutes a significant increase (64,3 ± 0.2) wasrevealed. A two hour clamping time was necessary.When it ended we observed a slight but significant in-crease of hemoglobin concentration to 66.6 ± 0.2 % (p<0,000). The final removal of the clamp led to a sig-nificant increase in the hemoglobin concentration witha value of 71.0 ± 0.1 % (p 0.000).

Discussion

The Laser-Doppler-Flowmetry is a preferable methodto all other systems because this system enables one toeffectively answer a vital question: whether or notthere is sufficient blood supply in a recipient site. Byusing this method we have proven that a retrogradeblood flow from the flap of the upper extremity doesoccur. An equal niveau of the microcirculation (Fig.3a-c) was present and hence measurable. We haveproven that there is a retrograde autonomy in the mi-crocirculation.

It was remarkable that once the clamping of thepedicle occured the oxygen saturation of hemoglobinreached a plateau and remained constant with all othervalues. This is clear evidence that a retrograde auton-omy of the oxygen saturation of hemoglobin from theupper extremity takes place. Measurements indicated aconstant value of 80 % oxygen saturation and this pre-sented percentage validated our opinion that there issufficient hemoglobin in the groin flap.

A new and recognisable factor arose when wemeasured the value of hemoglobin concentration in theinvestigated groin flap. We observed that no decreaseoccurred once the clamping procedure ended. At thetime of clamping an increased concentration of hemo-globin was observed. We found this to be phenominal.We conducted statistical analyses which indicated sig-nificant differences. And the increase of hemoglobinconcentration prooves that there is a transfer of a spe-cific tissue proportionality from the recipient site.

Before clamping the pedicle, the transposed groinflap measurement remained at the same level in the ip-silateral hand. The flow was 40 % lower than on thecontralateral side. This fact is a hint to all researcherssuggesting that there is a need to adopt different andspecific tissue proportionalities. We have observedthat the groin flap transponate possesses an increasedoxygen saturation of hemoglobin (Fig. 1b), which issimilar to the contralateral upper extremity (93.1%).


This mechanism needs to be established as a require-ment which can function as an interface for granula-tion tissue. The newly formed layer of granulation tis-sue with its numerous capillaries decreases the veloci-ty of erythrocytes and forms a strongly filled networkwith new capillaries having oxygen saturated hemo-globin. The LDF values of the transposed groin flapand the hand of the contralateral side revealed a bal-anced level of counting values. We have included thisformula in our report on microcirculation.

A dimensional shift occurs in the count value ofmicrocirculation because there are two impacting fac-tors: the number of blood cells and the velocity ofblood cells. The number of blood cells increases andthe velocity decreases. The stimuli for new formationsof granulation tissue is related to the development ofminute capillaries and the release of mediators. Theoxygen saturation of hemoglobin is based on erythro-cytes increasing in the groin flap and when this occursthe measured value is significantly higher than in thecontralateral groin area.

We also found that the measuring values of the Op-to Flow system (LEA, Gießen, Germany) contains re-markable prognostic parameters. Our recommenda-tions for assessment of prognosis are listed in Table 1.The flow from the flap indicates a constant value with-out any decrease in the curve of the measuring values.The microcirculation in the measurement level of theskin can decrease by 20 % in comparison with thestarting value. The change in oxygen saturation of thehemoglobin should remain below 25 % as it is veryimportant that the concentration of hemoglobin re-mains constant.

On the tenth postoperative day a training of thegroin flap began for autonomisation in the circulationof the flap. The pedicle was compressed with a rubbercovered clamp. The lengthening of compression time(3 x 5 minutes to two hours a day) led to an increasedsupport of the flap from the upper extremity. Afterclamping the pedicle a continuous capillary pulse wasa manifestation indicating increased vascularisationout of the upper extremities. The cutting through of theflap pedicle followed after three weeks. Vital questionsarise: Is the pedicle needed for covering the donorsite? Is it possible for the artery to be ligated and cuton the 21th day? Does the skin flap totally cover thedefect? Should the flap remain and the skin bestemmed for more than three weeks?

Although flap transplantation is connected with thefixed parts of surgical reconstructive covering of tissuedefects, it is unclear whether or not the muscle flap

will be postoperative over a long span of time(months, years) and remain autonomously perfusedover a vessel site from the microsurgical reconstructedvessel pedicle. Also we question whether or not thepedicle will be independent from the process of neo-vascularisation which emanates from out of the recip-ient site (autonomous flap perfusion by the recipientsite). In various publications case studies have includ-ed statements verifying the fact that a flap necrosis ap-peared after a disconnection and/or injury thus damag-ing the vessel pedicle even after eight years (16, 18,22).

A truly new spectrum of safety and confidence ap-pears when the Laser-Doppler-Flowmetry Opto Flowis utilised. This system is a valuable tool which can berelied upon as it allows surgeons, scientists and techni-cians to accurately decide when there is sufficient mi-crocirculation.

References

1. Billroth T. Untersuchungen über die Entwicklung der Blutge-fäße. Berlin: Georg Reimer Verlag, 1856

2. Carpenter GK, Swiontkowski MF. An In-Vitro Analysis of TwoLaser Doppler Flowmetry Systems for Evaluation of Bone Per-fusion. Calcified Tissue International 1991; 48: 414-420

3. Dunlap JN, Brinker MR, Cook SD. A New In Vivo Method forthe Direct Measurement of Nutrient Artery Blood Flow. Ortho-pedics 1997; 20 (7): 613-619

4. Höper J, Plasswilm L. Micro-Light Guide SpectrophotometricMeasurement of Changes in Local Haemoglobin Oxygenationand Concentration in the Rabbit Lung Induced by Hypoxia andHyperoxia. International Journal of Microcirculation. 1994; 14:282-288

5. Kakihana Y, Kessler M, Douplik AJ, Krug A. Stable and Reli-able Measurement of Intracapillary Hemoglobin-Oxygenationin Human Skin by Empho II. Abstract book: International Bio-medical Optics Symposium. San Jose, California, 1997

Table 1. Prognostic Parameters and a Tabular Gener-al View

Microcirculation constant flow in the flap with-out decrease of count values inreal time

Oxygen Saturation Hemoglobin

tolerable oxygen saturation ofhemoglobin < 25 %

Concentration of Hemoglobin hemoglobin concentrationshould not decrease after theclamping procedure


6. Kessler M, Höper J, Cristea P, Albert-Hehn B, Krug A. Func-tional Images of Local O2 Contents and O2 Gradients in IntactHuman Skin. SPIE 1996; 2926: 84-93

7. Krogstadt AL, Elam M, Karlsson T, Wallin G. Arteriovenousanastomoses and the thermoregulatory shift between cutaneousvasoconstrictor and vasodilator reflexes. Journal of AutonomicNervous System 1995; 53: 215-222

8. Krug A, Kessler M, Höper J, Batz M, Otto A, Zellner S, GärtnerD. Simultaneous Monitoring of NAD(P)H, Cytochromes, pO2and Hb O2 in Liver Tissue. SPIE 2324: 155-165

9. Krug A, Kessler M, Khuri R, Lust R, Chirwood R. Investigationof local heterogeneity of hbO2 an hb in working dog heart in situunder isovolemic hemodilution and critical coronary stenosis.SPIE 2927: 119-129

10. Machens HG, Mailänder P, Brenner P, Pasel J, Funke M, SiemersF, Berger A. Bleibt die Gewebeperfusion nach freiermikrovaskulärer Gewebetransplantation autonom? HandchirMikrochir Plast Chir 2000; 32: 193-196

11. Machens HG, Mailänder P, Kremer M, Reimer R, Berger A.Techniken zur postoperativen Überwachung der Gewebedurch-blutung nach freier mikrovaskulärer Gewebetransplantation.Handchir Mikrochir Plast Chir 1999; 31: 107-112

12. Mc Gregor IA, Jackson IT. The groin flap. Br J Plast Surg 1972;25: 3-16

13. McDonald F, Pitt Ford TR. Blood Flow Changes in the TibiaDuring External Loading. Journal of Orthopaedic Research1993; 11: 36-48

14. Racz IB, Sarkadi L, Hamar J. The functional damages of is-chemic/reperfused skeletal muscle. Acta Physiologica Hungaria1996; 84 (3): 205-216

15. Ramm B, Hofmann G. Biomathematik und medizinische Statis-tik. Stuttgart: Ferdinand Enke Verlag, 1976

16. Rothaus KO, Acland RD. Free flap neo-vascularisation: Case re-port. Br J Plast Surg 1983; 36: 348-349

17. Sachs L. Angewandte Statistik. Anwendung statistischer Metho-den. Berlin, Heidelberg, New York: Springer Verlag, 1984

18. Sadove RC, Kanter MJ. Absent neovascularization in a lowerextremity free flap; A case report. J Reconstr Microsurg 1993; 9:5-9

19. Schemitsch EH, Kowalski MJ, Swiontkowski MF, Senft D. Cor-tical Bone Blood Flow in Reamed and Unreamed Locked In-

tramedullary Nailing: A Fractured Tibia Model in Sheep. Journalof Orthopaedic Trauma 1994; 8 (5): 373-382

20. Schemitsch EH, Kowalski MJ, Swiontkowski MF. Evaluation ofa Laser Doppler Flowmetry Implantable Fiber System for Deter-mination of Threshold Thickness for Flow Detection in Bone.Caldif Tissue Int 1994; 55: 216-222

21. Schwarz Lausten G, Kiar T, Dahl B. Laser Doppler Flowmetryfor Estimation of Bone Blood Flow: Studies of Reproducibilityand Correlation with Microsphere Technique. Journal of Ortho-peadic Research 1993; 11 (4): 573-580

22. Serafin D, Shearin JC, Georgiade NG. The vascularization offree flaps: A clinical and experimental correlation. Plast Recon-str Surg 1977; 60: 233-241

23. Swiontkowsky MF, Tepic S, Perren SM, Moor R, Ganz R, RahnBA. Laser Doppler Flowmetry for Bone Blood Flow Measure-ment: Correlation with Microsphere Estimates and Evaluation ofthe Effect of Intracapsular Pressure on Femural Head BloodFlow. Journal of Orthopaedic Research 1986; 4: 362-371

24. Swiontkowsky MF. Laser Doppler Flowmetry for Osteomyelitis.J Orthop Trauma 1989; 3 (1): 1-5

25. Tran MA, Geral JP. The Influence of some Vasoactive Drugs onBone Circulation. European Journal of Pharmacology 1987; 52:109-114

26. Wilhelm K, Putz R, Hierner R, Giunta RE. Lappenplastiken inder Handchirurgie. Angewandte Anatomie, Operationstech-niken, Differentialtherapie. Wien, Baltimore: Urban undSchwarzenberg Verlag, 1997

27. Wolf K, Hamar J, Höcherl E, Pfister C. Development and reag-ibility of the micro-circulation in the granulation tissueafter distance osteosyntheses: An experimental analysis of laserdoppler flowmetry under the influence of CGRP. Applied Car-diopulmonary Pathophysiology 2000; 9: 333-342

28. Zoltan J. Hautersatzverfahren. Akademiai Kiado, Budapest,1984

Address for correspondence: Konrad Wolf, M.D., Department ofTrauma, Hand and Reconstructive Surgery, Hospital MunichSchwabing, Teaching Hospital of Ludwig-Maximilians-UniversityMunich, Kölnerplatz 1, D-80804 Munich, Germany, E-Mail:[email protected]

B. W. Urban, M. Barann (Eds.)

Molecular and Basic Mechanisms of Anesthesia

588 pages, Hardcover, colour picturesISBN 978-3-936142-41-9, Price: 95,- Euro


More than 70 condensed papers, written byleading experts in this field have been arrangedfollowing the hierarchical order of the centralnervous system: Molecular Basis of AnestheticInteractions, Anesthetic Interactions with IonChannels and Other Proteins, Integration ofAnesthetic Actions (In-Vitro) and, finally, In-Vi-vo-Networks. In addition to insights at the mo-lecular level, exciting results have been ob-tained from the various in-vitro slice prepara-tions and from in-vivo electrophysiology as wellas from in-vivo functional imaging, indicating amuch higher level of complexity than originallyappreciated.

56 M. Erelel

Introduction

Asthma is an obstructive lung disease characterizedwith eosinophilic bronchial inflammation and relatedbronchial hyperreactivity. Although bronchial inflam-mation still remains, airway obstruction may improvewith drugs or spontaneously. This feature differenti-ates asthma from chronic obstructive pulmonary dis-ease (1).

Apart from patients with chronic obstructive pul-monary disease (COPD), in asthma patients 15-20minutes after 200 microgram salbutamol inhalation,12% or more increase in forced expiratory volume inone second (FEV1) or 200 ml or more increase in ini-tial value is a reliable marker of reversible airway ob-struction (reversibility test).

It is known that inflammation and related hyperre-activity still continue even if obstruction cannot be de-termined. Thus 20 % or more decrease in initial FEV1value is also a characteristic feature for asthma (non-specific bronchoprovocation test) (NSBPT) (1-5).

NSBPT, which is performed to show hyperreactiv-ity in cases with normal pulmonary function tests, re-quires fresh solution and dosimeter. All the equip-ments are used for standardizing tests and it takesmore time than reversibility. Stability of the solutionsthat are used for tests does not last long. NSBPT, dis-tinct from the reversibility test that could be done inevery centre that performs pulmonary function test,cannot be performed in every centre easily (6-9).

In this study, we compared NSBPT and reversibil-ity test and reversibility test’s diagnostic value in the

Bronchoprovocation or reversibility in asthma patients with normal pulmonary function tests

M. ErelelDepartment of Pulmonary Disease, Istanbul Faculty of Medicine, Istanbul University, Istanbul, Turkey


Keywords: non-specific provocation, reversibility, diagnosis of asthma

Abstract

Asthma is a reversible obstructive lung disease, which can be symptom free spontaneously or with the treatment.Thus, no obstruction can be determined in pulmonary function tests though patients are admitted to the physicianwith clinical symptoms. For the diagnosis of asthma, reversibility and bronchoprovocation tests are used. Howev-er patients have symptoms but not physical findings or obstruction in pulmonary function tests, they get betterwith medication. So, it is thought that there may be reversibility in these cases. In this study 5% or more increasein the initial value of FEV1 following salbutamol inhalation in pulmonary function tests of asthma patients withno obstruction but with clinical findings was considered positive reversibility test (modified reversibility test).Modified reversibility test’s specificity and sensitivity was 54% and 55% respectively. As a result, it was concluded that modified reversibility test is not as useful as a diagnostic test as bronchoprovo-cation in this group of patients.

Abbreviations:COPD: chronic obstructive pulmonary disease; NSBPT: non-specific bronchoprovocation test; PD20: provocativedose; FEV1: forced expiratory volume in one second; FVC: forced vital capacity; FEF25-75%: forced expiratoryflow rate; PEF: peak expiratory flow

Bronchoprovocation or reversibility in asthma patients with normal pulmonary function tests 57

patient population with normal pulmonary functiontests but with pulmonary complaints and in a healthycontrol group.

Material and methods

Flow and volume measurements were performed 3times with Sensormedics Vmax 229 spirometry in allcases that had been admitted to the outpatient clinic ofIstanbul Faculty of Medicine with recurrent shortnessof breath and wheezing. Forced vital capacity (FVC),FEV1, FEV1/FVC, forced expiratory flow rate(FEF25-75%), peak expiratory flow (PEF) measure-ments were evaluated. The biggest sum of FEV1 andFVC was recorded. Forty-five NSBPT positive cases(group A), which had 85% or more in predicted valueof FEV1, FVC, FEV1/FVC, 75% or more in FEF25-75% and PEF value and diagnosed asthma clinically,were enrolled to the study. Smoker cases stoppedsmoking at least 48 hours before the tests.

Eleven healthy volunteers who didn’t have any up-per or lower respiratory tract infection for 6 monthsand symptoms related to the respiratory system, werechosen as the control group (Group C).

Any patient who has another condition that resultswith bronchial hyperreactivity was excluded.

Performed tests and followed approaches areshown in Figure 1 for all cases. 200 microgram salbu-

tamol metered dose inhaler was used for reversibilitytest. Test was repeated 20 minutes after salbutamol in-halation. FVC, FEV1, FEV1/FVC, isoFEF25-75%,PEF were measured on the basis of reversibility initialvalue and absolute value, FEV1, FEV1/FVC,isoFEF25-75% were measured on the basis of predict-ed value.

Non-specific bronchoprovocation test was done 24hours after reversibility test. Histamine was given inincreasing doses with dosimeter in NSBPT. Test wasstopped when 20% or more decrease (PD20) in initialFEV1 was determined or cumulative histamine dosewas more than 8 mg/ml. PD20 value was calculatedwith Sensormedics Vmax 229spirometry and “Bron-cho Challenge” database programme.

Results

45 asthma patients (group A) and 11 voluntary cases(group C) were enrolled to the study. The distributionof cases in A and C groups according to the gender (fe-male/male and respectively) were 39/6 and 6/5 and themean age was 33 ± 12 and 26 ± 8 respectively. 2 cas-es (18%) in group C and 10 cases (22%) in group Awere smokers.

FVC, FEV1, FEF25-75%, PEF and FEV1/FVCvalues of the control group were 4368 ± 932, 3868 ±957, 4556 ± 1537, 7242 ± 2682 ml and 88 ± 1% re-

Suspicion of Patients Anamnesis/Physical Examination Asthma? Chest X-ray (Group A)

Suspicion of

Asthma

Another Disease Pulmonary Function Tests FEV1-FVC-FEV1/FVC < %85

FEF25-75%, PEF < % 75

No Yes

Exclude

Reversibility Following Day Nonspecific

Bronchoprovocation Test No Yes Control Cases Group C Diagnosed Pulmonary Disease History

and/or Respiratory tract infection last 6 months Figure 1. Inclusion criteria

58 M. Erelel

spectively. The measured values of asthma patientswere 3578 ± 886, 2900 ± 651, 3918 ± 719, 5487 ±1562ml and 81 ± 0.7% respectively (Table 1).

NSBPT with histamine was performed to all cases.NSBPT ’s cumulative dose threshold value (8 mg/ml)was exceeded in group C. When FEV1 value droppedless than 20% or the threshold value was exceeded thetest was considered negative. PD20 value determinedin Group A was positive (1.24 ± 1.52 mg/ml) (Table 2).This value was significantly less in asthma group thancontrol group (p<0.05).

Baseline reversibility results in group C were 2% ±2 for FVC, 3% ± 3 for FEV1/FVC, 17%±10 forisoFEF25-75% and 13% ± 17 for PEF. In group Athese results were 3% ± 5, 3% ± 4, 19% ± 15, 6% ± 14respectively. There wasn’t any statistically significantdifference among these values (p>0.05). Initial FEV1reversibility was 5% ± 2 and predicted reversibilitywas 6% ± 4. When 5% reversibility value was consid-ered threshold value in initial reversibility the differ-

ence was found significant. But specificity and sensi-tivity were 54% and 55% respectively (p<0.05).

Predicted reversibility results in group C were 2%± 2 for FVC, 5% ± 3 for FEV1, 18% ± 12 forisoFEF25-75% and 11% ± 11 for PEF. In group Athese results were 3% ± 5, 6% ± 3, 14% ± 11 and 4%± 11 respectively. The differences between groupswere not statistically significant (p>0.05).

Absolute reversibility results in group C were91 ± 117ml for FVC, 188 ± 111ml. for FEV1, 850 ±656ml for isoFEF25-75% and 880 ± 999ml PEF. Ingroup A these results were 101 ± 180ml, 181 ± 112ml,542 ± 417ml and 299 ± 854ml respectively. The differ-ences between groups were not statistically significant(p>0.05). The reversibility values of both groups areshown in Table 3.

Table 2. Bronchoprovocation Test Results (PD20) (mg/ml)

Minimum Maximum Mean ± SD

Group C 8.11 10.20 9.89 ± 0.71

Group A 0.30 5.75 1.24 ± 1.52

Table 1. Initial Pulmonary Function Tests

FVC(ml) FEV1(ml) FEV1/FVC(%) isoFEF25-75%(ml) PEF(ml)

Group C 4368 ± 932 3868 ± 957 88±1 4556 ± 1537 7242 ± 2682

Group A 3578 ± 886 2900 ± 651 81±0,7 3918 ± 719 5487 ± 1562

Table 3: Reversibility Results in GroupsC

FVC FEV1 FEV1/FVC IsoFEF25-75% PEF

A

I (%)2 ± 2 5 ± 2 3 ± 3 17 ± 10 13 ± 17

3 ± 5 6 ± 4 3 ± 4 19 ± 15 6 ± 14

II (%)2 ± 2 5 ± 3 18 ± 12 11 ±11 2 ± 2

3 ± 5 6 ± 3 14 ± 11 4 ±11 3 ± 5

III (ml)91 ± 117 188 ± 111 850 ± 656 880 ± 999 91 ± 117

101 ± 180 181 ± 112 542 ± 417 299 ± 854 101 ± 180

C: Healthy Volunteers; A: Asthma Cases; I: Initial Reversibility (%); II: Predicted Reversibility (%); III: Absolute Reversibility (ml)

Bronchoprovocation or reversibility in asthma patients with normal pulmonary function tests 59

Discussion

Reversibility used in the diagnosis of obstructive lungdiseases is related to the degree of obstruction. Themore obstruction means the more reversibility and theless obstruction means the less reversibility (10).There are no detailed studies on the degree of re-versibility in asthma patients with normal pulmonaryfunction tests.

Spontaneous daily variability in FEV1 may occurin asthma patients. This variability is not a marker forreversibility (11).

In the present study no significant differences weredetermined in predicted and absolute reversibility inFEV1 and FVC. There was no significant initial re-versibility in FVC, baseline and predicted reversibilityin FEV1/FVC.

In baseline FEV1 reversibility, statistically signifi-cant 5% increment was found in 27 (60%) of asthmapatients. But we didn’t diagnose 18 (40%) asthma pa-tients with the criteria. On the other hand it also caused6 cases (54%) to have false positive diagnosis. Fivepercent in baseline FEV1 reversibility’s specificityand sensitivity were 54% and 55% respectively.

We concluded that reversibility in baseline FEV15% increase is not as sufficient as NSBPT test. AndNSBPT is the necessary method for diagnosis of asth-ma patients when obstruction is not proved.

Speculations

Diagnostic tools improved lastly however diagnosis ofasthma still has some controversies. Especially thisproblem arose in patients with symptomatic but withnormal pulmonary function tests. In this group of pa-tients NSBPT is necessary. But it needs special equip-ments, time, drugs (methacholine, provacholine, hista-mine etc.) and trained technicians. Drug stability is al-so an additional problems. If special precautions don’thave been taken, drugs pH decrease and become acidicsolution. And it causes cough related threshold insteadof PD20. Problems as listed previously must besolved.

When searched Medline we didn’t find any articlewhich compares NSBPT and reversibility. We see inclinical practice some asthma patients with sympto-matic but normal pulmonary function tests. If any an-ti- asthmatic treatment is given to them, they get betterand become symptom free. In some of these patientsreversibility tests are positive too.

We thought whether this procedure should be usedfor diagnosis of asthma or not. We didn’t approve it.

We used histamine for NSBPT. Methacholine orprovocholine may be used instead of histamine.

When planned new studies;1. Additional study with methacholine or provocholine

is necessary.2. Late reversibility and daily peak expiratory flow

measurement may be included.3. A greater number than our cases may be enrolled to

the study.

References

1. Eisen EA, Docrey DW, Speizer FE et al. The association be-tween health status and the performance of excessively variablespirometry tests in population-based study in six U.S. cities. AmRev Respir Dis 1987; 136/137: 1-6

2. No authors listed. Standards for the diagnosis of patients withchronic obstructive pulmonary disease (COPD) and asthma.This official statement of the America Thoracic Society wasadopted by the ATS Board of Directors, November 1986. AmRev Respir Dis 1987; 136: 225-44

3. Crii E, Balbo A, Lazzaroti M. Bronchial provocation test withallergen: comparison between two different techniques. Monal-di Ach Chest Dis 1994; 49: 447-450

4. Kersten W. Indication for inhalation provocation test. ArchMaragliano Patol Clin 1978; 34: 27-31

5. Gardners RM, Hankinson JL, Clausen JL. ATS statement onstandardization of spirometry-1987 Update. Am Rev Respir Dis1987, 136: 1285-98

6. Marshik P, Moghaddam S, Tebbet I et al. Degradation of hista-mine solution used for bronchoprovocation. Chest 1999; 115:194-9

7. Ulrik CS, Backer V, Skov PG. Usefulness of repeated measure-ments of bronchial hyperresponsiveness for the diagnosis of oc-cupational asthma. J asthma 1994; 31: 35-42

8. David PJ, Kenneth RC, Steven K. Prior diagnosis and the treat-ment of patients with normal results of methacoline challengeand unexplained respiratory symptoms. Chest 1196: 109: 697-701

9. Wits M, Hop WC, Van der Heyden GH, Kerrebijn KF, JongsteJC. Measurement of bronchial responsiveness in young chil-dren: comparison of transcutaneous oxygen tension and func-tional residual capacity during induced bronchoconstruction anddilatation. Pediatr Pulmonol 1992; 12: 181-185

10. Turki J, Green SA, Mewman KB et al. Human lung cell beta-2adrenergic receptors desensitize in response to in vitro adminis-tered beta agonist. Am J Physiol 1995; 269 (5 pt:1): L709-714

11. Guclu G, Sayiner A. Effects of spontaneous FEV1 fluctuation onairway reversibility estimation. Monaldi Arch Chest Dis 1994;49: 466-9

Address for correspondence: Mustafa Erelel, M.D., SenesenevlerKocayol sok.N:11 Cigdem ap D:29 Bostanci-81110 Istanbul,Turkey, E-Mail: [email protected]

60 T. Iber, J. P. Roesner, C. Muth, G. F. E. Nöldge-Schomburg, D. A. Vagts

Introduction

The acute respiratory distress syndrome (ARDS) is asevere pathological condition with a high mortalityrate of more than 30% (15). Therefore, extensive re-search is performed on various animal models of acutelung injury (ALI) and ARDS. The four common waysof inducing ALI are repeated bronchioalveolar lavagewith saline, intrabronchial instillation of hydrochloricacid, intravenous infusion of endotoxin and the intra-venous infusion of oleic acid. All techniques are easi-ly achieved since they are well described in the litera-ture (10,11,16). A comparison of these ALI models re-vealed that none is superior to the other because it isdifficult to compare them (19). Each one has a differ-ent impact on hemodynamics and oxygenation: i.e.saline bronchioalveolar lavage and hydrochloric acidinstillation models produce acute hypoxemia in a he-modynamically stable animal, while a brief intra-

venous endotoxin infusion produces a hemodynamicinstable animal with pulmonary hypertension but failsto produce profound hypoxemia. Last but not least ole-ic acid infusion produces a hemodynamically compro-mised animal with pulmonary hypertension and pro-found hypoxemia (19).

We wanted to establish a hypoxemic and hemody-namically compromised but stable and reproduciblemodel of ALI in pigs for an investigation period of fivehours. Because the effects described in the literaturecame nearest to our needs we have chosen to use ole-ic acid as induction agent. Oleic acid induced ALI re-sembles to a great extent the pathological features ofARDS, which are summarized in an excellent reviewby Schuster (23). But unfortunately the majority of au-thors does not specify how and in which form oleicacid is infused. Hence we tested different approachesof ALI-induction. However, the injection of pure oleicacid is followed by an unfavorable hemodynamic col-

Induction of a reproducible, hypoxemic and hemodynamically compromised but stable early Acute Lung Injury (ALI) model with oleicacid in pigs - tips and pitfalls

T. Iber*, J. P. Roesner*, C. Mutz, G. F. E. Nöldge-Schomburg, D. A. VagtsDepartment of Anesthesiology and Intensive Care Medicine, University of Rostock, Germany*Both authors had equal contribution to this manuscript


Keywords: acute lung injury, oleic acid, hypoxia, atelectasis

Abstract

The acute respiratory distress syndrome (ARDS) is still associated with a high mortality rate. Therefore, extensiveresearch is performed in various animal models of acute lung injury (ALI) and ARDS. Our aim was to establish ahypoxemic and hemodynamically compromised, but stable and reproducible model of ALI in pigs over an inves-tigation period of five hours. Among the four common ways of inducing an ALI, we decided to use oleic acid dueto its special effects on hemodynamic variables and gas exchange. Screening the literature we found no entiremethod of ALI-induction by oleic acid. Therefore, we performed the two most frequently described methods ofpure bolus application of oleic acid and the application of oleic acid dissolved in isotonic saline. Both methodsfailed to induce the required pathological, but stable condition. We describe a new detailed procedure to induce anALI by oleic acid with impaired pulmonary function and compromised, but stable hemodynamics over an inves-tigation period of 5 hours. All previously described experimental models provide lack of at least one desirableproperty.

Induction of a reproducible, hypoxemic and hemodynamically compromised ... 61

lapse due to right heart failure. Dilution of oleic acidin watery solutions is almost inhomogeneous and dif-ficult, while solving oleic acid in ethanol seemed to bedisadvantageous because of the known modulatory ef-fects of alcohol on immune responses occurring evenunder small amounts of alcohol intake (26,27).

Because all previously described experimentalmodels provide lack of at least one desirable property,we describe a new procedure to induce an ALI by ole-ic acid with impaired pulmonary function and compro-mised hemodynamics (hypoxemia, increased staticpulmonary compliance, pulmonary hypertension, re-duced cardiac output and only mild alterations in meanarterial pressure), which is stable for at least 5 hours.

Materials and Methods

Anesthesia

After approval by the local Ethics Committee on Ani-mal Research and in concordance with the Helsinkiconvention for the care and use of animals, 5 femaleGerman domestic pigs with a median body weight of35.0 kg were premedicated i.m. with flunitrazepam 0.2mg kg-1 body weight (bw) (Rohypnol®, Hoffmann-LaRoche, Grenzach-Wyhlen, Germany) and ketamine 15mg kg-1 bw (Ketanest®, Parke-Davis, Freiburg, Ger-many) after overnight fasting and receiving water adlibitum. Anesthesia was induced i.v. via an ear veinwith ketamine 2.0 mg kg-1 bw and fentanyl 3 µg kg-1

bw (Fentanyl-Janssen®, Janssen-Cilag, Neuss, Ger-many). The trachea was intubated after injection of ve-curonium 0.3 mg kg-1 bw (Norcuron®, Organon Tekni-ka, Eppelheim, Germany). Anesthesia was maintainedby continuous i.v. infusion of flunitrazepam 0.1 mg kg-1 h-1, ketamine 8.0 mg kg-1 h-1 and vecuronium0.7 mg kg-1 h-1. Pressure-controlled mechanical venti-lation was provided with a Servo900® ventilator(Siemens Erlangen, Germany).

Respiratory rate and pressure were adjusted tomaintain arterial carbon dioxide tension (PaCO2) be-tween 4.8 and 5.6 kPa. Inspired oxygen fraction wasadjusted to maintain arterial oxygen partial pressure(PaO2) of 12-15 kPa. A positive end expiratory pres-sure (PEEP) of 5 cmH2O was applied.

Instrumentation

After induction of anesthesia, the animals were placedin supine position on a heating pad to keep body tem-perature constant.

A double-lumen catheter (7 F Two-Lumen CentralCatheterization Set; Arrow, Reading, PA, USA) and an8.5 F introducer (Arrow Percutaneous Sheath Intro-ducer Set; Arrow) were inserted into the right internaljugular vein. Both catheters were advanced 11-13 cmto guarantee correct position of the tip in the superiorvena cava. A Swan-Ganz thermodilution catheter(model 93A-131-7F; 7 F Swan-Ganz ThermodilutionsCatheter; American Edwards Laboratories, Irvine, CA,USA), was introduced into the pulmonary artery. Bodytemperature was monitored continuously with a ther-mistor in this flow-directed catheter. The right femoralartery was cannulated with an arterial catheter(Leader-Cath, 18 G, Vygon, Aachen, Germany).

For maintenance of normovolemia, all animals re-ceived full-electrolyte solution 12 ml kg-1 h-1 i.v.(Jonosteril®; Fresenius-Klinik, Bad Homburg, Ger-many) to maintain blood pressure, central venous pres-sure, and hematocrit values at the level measuredshortly after insertion of the femoral artery and centralvenous catheters.

Measurements and calculations

Arterial blood gas values were taken to adjust inspiredoxygen concentration and ventilation to maintain pre-determined PaO2 and PaCO2 values, and to measureinitial hematocrit values. Blood gases, and serum elec-trolytes were measured using an ABL615 Autoanalyz-er (Radiometer, Copenhagen, Denmark).

Hemodynamics

All intravascular catheters were connected to pressuretransducers. Arterial blood pressure, central venouspressure and pulmonary artery pressure were recordedby a digital data acquisition device (PO-NE-MAH(Digital Acquisition analysis and Archive Systems,Simsbury, USA)). Heart rate was derived from thespike-interval of the continuous arterial blood pressuremeasurement.

Cardiac output was determined by thermodilutiontechnique (Baxter Explorer CO-computer, Unter-schleissheim, Germany). The mean value of three in-


jections of 10 ml ice-cold saline was considered to es-timate actual cardiac output if the measurements werewithin a range of ±10% of the calculated mean.

Static compliance of the respiratory system

The static compliance of the respiratory system (Cstat)and the static inspiratory (Pin) and expiratory pressures(PEEP) were measured using the airway occlusiontechnique described by Sly (1987) as Cstat = Vt * (Pin -PEEP)-1 (22).

Experimental protocol

Before intravenous injection of oleic acid we increasedthe fractional inspiratory oxygen concentration to 50% and applied a positive end-expiratory positive air-way pressure (PEEP) of 8 cm H2O to minimize the for-mation of atelectasis with regard to the stability of themodel.

According to the literature, we initially performedtwo different methods of oleic acid application:1. Dissolving oleic acid in isotonic saline: We dis-

solved 0.08 ml*kg-1 oleic acid in 20 ml isotonicsaline and infused the suspension over 10 minutes.

2. Bolus application of oleic acid: We injected the ole-ic acid with a dosage of 0.08 ml*kg-1 with a syringeinto the central venous catheter and flushed thecatheter with saline afterwards.

After observing the frustrating results of bothmethods, we tried out a third, novel and simple methodof application:

3. Mixing oleic acid in glucose 5 % by aid of two sy-ringes:Two 5 cc syringes (Luer-Lock, Braun®) were con-nected to a three-way valve linked to the central ve-nous catheter, thus, allowing intensive mixing ofoleic acid (0.08ml*kg-1) and 3 ml of glucose 5 %between the two syringes. After emulsificationrepetitive injections of 0.5-1 ml of the mixture, ac-cording to PAP and PaO2 were feasible.

All measurements were performed at baseline, 2, 3,4 and 5 hours after the induction of ALI. At the end ofthe observation period of 5 hours all animals werekilled in deep anesthesia with a potassium chloride

overdose, according to German laws for animal stud-ies.

CT-Scan:5 hours after the induction of ALI one anesthetized

pig was transported to the CT-Scan of our universityhospital and scans of the thorax were performed dur-ing halted inspiration.


Statistical analysis was carried out with JMP® softwarepackage (SAS, Cary, NC, USA). Medians are giventhroughout with interquartile range (25th-75th per-centiles) (5,2). Differences between experimental peri-ods were analyzed using Friedman rank-sum analysisfollowed by the Tukey-Kramer HSD test. The level ofsignificance was set at p<0.05. All results were in-dexed to compensate for differences in body weight.

Results

Dissolving oleic acid in isotonic saline:The attempt to dissolve oleic acid in isotonic salinefailed, because oleic acid precipitated and clottedwithin the syringe and the line of the central venouscatheter and the infusion lines.

Bolus application of oleic acid: The injection of a bolus lead to a dramatic decrease inblood pressure and an increase in the pulmonary arterypressure (PAP) with systolic values being higher thansystolic arterial blood pressure (Figure 1). This in-duced a instable and life threatening condition.

Mixing oleic acid in glucose 5 % using two syringes:Two 5 cc syringes (Luer-Lock, Braun®) were connect-ed to a three-way valve linked to the central venouscatheter, thus allowing intensive mixing of oleic acid(0.08 ml*kg-1) and 3 ml glucose 5 % between the twosyringes. After emulsification of the mixture repetitiveinjections of 0.5 - 1 ml, according to PAP and paO2were feasible and induced an ALI as defined by aPaO2/FiO2-ratio < 300 mmHg after 30 min. With addi-tional small injections up to the total throughout thefollowing 90 minutes, we achieved a hypoxic and he-modynamically compromised but stable ALI for a fur-ther observational period of three hours. This patho-


logical, but stable condition was characterized by thefollowing physiological and morphological changes:

Pulmonary functionOur oleic acid induced ALI model fulfilled the cri-

teria of ALI (PaO2/FiO2 < 300 mmHg) and even ofARDS with (PaO2/FiO2 < 200 mmHg) after initiationof lung injury according to the definition of the Amer-ican-European Consensus Conference in 1994 (3). Theextent of ALI remained unchanged for 5 hours withoutspontaneous improvement or deterioration (Figure 2).

Static ComplianceAfter induction of ALI the static compliance of the

respiratory system decreased significantly by 25%from 21 to 16 ml*cm H2O-1 and remained compro-mised until the end of the observation period (Table 1).

HemodynamicsAfter induction of ALI we observed a moderate de-

crease of MAP to 78 mmHg, which recovered slowlyduring the observational period to 90 mmHg reachingbaseline values (Table1). Mean PAP increased dramat-ically from 18 mmHg to 35 mmHg and remained ele-vated at this level throughout the observation period(Figure 3).

Cardiac output decreased from a baseline value of139 ml*min-1*kg-1 by almost 50 % to 72 ml*min-1*kg-1. After ALI was induced cardiac output remaineddepressed at this level until the end of the observation-al period (Table 1).

CT-ScanEven though pigs were ventilated in a pressure-

controlled mode with positive end-expiratory airwaypressure bilateral atelectasis, infiltrations and a dra-matic reduction in normal, aerolized lung area were

0

10

20

30

40

50

60

70

80

90

100

0 5 10 15

min

MAP MPAP

Figure 1. Hemodynamic effects of bolus application ofoleic acid, mean arterial pressure (MAP) and meanpulmonary artery pressure (MPAP) in mmHg.

time (h)

54320

500

400

300

200

100

0 #

Figure 2. PaO2-FiO2-ratio (mmHg), # p<0.05 for 2,3,4and 5 hours versus baseline.

Table 1: Hemodynamic and pulmonary variables, # p<0.05 versus baseline.

0h 2h 3h 4h 5h

MAP(mmHg)

102 (92-111)

78 # (68-104)

88 (76-104)

86 (78-101)

93 (81-106)

CO_i (ml/kg-1*min-1)

138 (113-145)

72 # (67-85)

71 # (64-86)

75 # (62-84)

79 # (74-88)

Static Compliance (ml/cmH2O)

21 (19-24)

16 # (14-17)

16 # (13-17)

14 # (13-16)

14 # (12-16)


visible on the CT-scan (Figure 4) five hours after ALI-induction.

Discussion

Screening the literature the following approaches toinduce an ALI by oleic acid have been described: 1. ALI-induction in pigs: “Oleic acid 0.1 ml*kg-1 sus-

pended in 20 ml isotonic saline was slowly injectedvia the central venous catheter.” (14)

2. ALI-induction in pigs: “After a baseline determina-tion, 10 ml of saline solution containing 0.1 ml*kg-1 of oleic acid was infused into the right atriumvia the internal jugular catheter for a period of 10min.” (6)

3. ALI-induction in adult mongrel dogs: “In group“Oleic acid only” the only intervention was 0.08ml*kg-1 oleic acid given into the central venouscatheter.” (8)

4. ALI-induction in mongrel dogs: “After control datawere recorded, a bolus of 0.08 ml*kg-1 of oleic acidwas administered through the proximal port of thepulmonary artery catheter.” (21)

5. ALI-induction in sheep: “Lung injury was inducedby injecting 0.12 ml oleic acid per kg body weightemulsified in 15-20 ml of previously extractedblood over 30 min into the right atrium.”(17)

6. ALI-induction in pigs: “Oleic acid (dissolved in96% alcohol) was administered in a series of 0.1 mluntil paO2 was lower than 8 kPa.” (7)

7. ALI-induction in rats: “Acute lung injury was in-duced by intravenous administration of 100 mg*kg-1 of oleic acid. Oleic acid was initially diluted inethanol ...” (9)

8. ALI-induction in ex vivo perfused New Zealandwhite rabbits heart lung blocks: “Oleic acid (OA)injury was induced with 0.4 cc of 50% solution ofOA and 95% ethanol infused over 20 min directlyinto the perfusion circuit.” (30)

As expected by its physico-chemical propertiesoleic acid cannot be dissolved in watery solutions.This attempt was followed by precipitation and oc-cluding lines and catheters. Intensive stirring andshaking improved solubility but precipitation occurredafter terminating the mixing process. Additionally,once clot-formation had occurred, the lines were im-mediately congested and could not be cleared after-wards by flushing them with saline or any other fluid.

The bolus injection caused – at least in pigs – a dra-matic increase in PAP despite normoxia. The increasein PAP could not be ameliorated by increasing the in-spiratory oxygen fraction and PEEP. Therefore, we as-sume that the increase in PAP must be due to a conges-

time (h)

54320

50

40

30

20

10

0

#

Figure 3. Mean pulmonary arterial pressure (mmHg),# p<0.05 for 2,3,4 and 5 hours versus baseline.

Figure 4. CT-scan 5 hours after ALI-induction. LowerLobes are non-aerated with extensive atelectasis for-mation.


tion of pulmonary microvessels similar to pulmonaryembolism leading to right heart failure, which is thesyndrome the model originally was adopted for (23).

The use of ethanol (96 %) as a solvent seems prom-ising (7,9,21) but we did not favor this method sinceethanol is known to modulate immune responses, i.e.alterations in cytokine pattern, inhibition of leukocyterecruitment and endothelial cell activation and inhibi-tion of transcriptional factors such as NF-kappa B(13,26,27). This effect is not only seen in chronic alco-hol abuse but in acute moderate alcohol dosages in vi-vo (4,13,20,26) and in vitro (12,13,20,26), too. ALI isconsidered to be a pathological condition associatedwith multiple inflammatory pathways and therefore,any interaction with immunological responses shouldbe excluded. Because the application of oleic acidsolved in ethanol via a central venous or pulmonarycatheter is very close to the organ to be affected, theremight be a good chance of having disturbing interac-tions as mentioned above.

In our model of oleic acid induced ALI weachieved an impaired pulmonary function measuredby a PaO2/FiO2-Ratio < 200 mmHg which fulfils thecriteria for ALI and ARDS according to the American-European Consensus Conference (AECC) in 1994 (3).The pulmonary and circulatory alterations wereachieved after 30 minutes but another 90 minutes withoccasional injections (2-4 x 0.2 ml) according to thePaO2 were necessary to maintain stability. Afterwardsthe impaired pulmonary function was unaltered anddid not show any improvement over the observationalperiod of 5 hours even though a moderate PEEP wasapplied.

We decided to apply a moderate PEEP in compari-son to other investigators (7,8,19) to gather normoxiawithout application of high inspiratory oxygen con-centrations, which further lead to the formation of at-electasis (18). In order to prevent other causes of pul-monary deterioration from mechanically ventilationwe used a pressure-controlled mode according to theevidence-based criteria’s (1).

Hemodynamic stability

Our goal was to achieve a hemodynamically compro-mised animal that did not show changes such as im-provement or deterioration during the observationalperiod. The MAP decreased to a moderate extent,guaranteeing a sufficient perfusion pressure, thus notcontributing to multiple organ failure. The mild in-

crease of MAP after 5 h is not due to an increase incardiac function since cardiac output stayed depressedby roughly 50 % until the end of the observation peri-od. This is consistent with the data of other researchgroups (21,24,25). However, the mechanism by whicholeic acid affects the myocardium remains unclear. Aninterference with the anesthetics being used can be ex-cluded, since we have proven the hemodynamic stabil-ity of our model in the past (28,29).

Conclusion

In order to establish a hypoxemic and hemodynamical-ly compromised but stable porcine model of ALI weused oleic acid to induce lung injury. However, screen-ing the literature we realized that there is obviously nostandardized method of inducing ALI with oleic acid.The only feasible way of injecting oleic acid is by dis-solving it in ethanol, thus hazarding the consequencesof having any alcohol-induced modulation of immuneresponses. Therefore, we established a simple methodof mixing oleic acid in glucose 5 % in order to receivean injectable emulsion causing an ALI that fulfilledour goal criteria of being hypoxemic and hemodynam-ically compromised but stable throughout an observa-tion period of 5 hours.

References

1. Amato MB, Barbas CS, Medeiros DM, Magaldi RB, SchettinoGP, Lorenzi-Filho G, Kairalla RA, Deheinzelin D, Munoz C,Oliveira R, Takagaki TY, Carvalho CR. Effect of a protective-ventilation strategy on mortality in the acute respiratory distresssyndrome. N Engl J Med 1998; 338: 347-354

2. Altman DG, Gardener MJ. Confidence intervals for researchfindings. Br J Obstet Gyn 1992; 99: 90-91

3. Bernard GR, Artigas A, Brigham KL, Carlet J, Falke K, HudsonL, Lamy M, Legall JR, Morris A, Spragg R. The American-Eu-ropean Consensus Conference on ARDS (1994) Definitions,mechanisms, relevant outcomes, and clinical trial coordination.Am J Respir Crit Care Med 1994; 149: 818-824

4. Boe DM, Nelson S, Zhang P, Quinton L, Bagby GJ. Alcohol-in-duced suppression of lung chemokine production and the hostdefense response to Streptococcus pneumoniae. Alcohol ClinExp Res 2003; 27: 1838-1845

5. Bulpitt CJ. Confidence intervals. Lancet 1987; 1: 494-4976. Colmenero-Ruiz M, Pola-Gallego de Guzman D, del Mar

Jiminez-Quintana M, Fernandez-Mondejar E. Abdomen releasein prone position does not improve oxygenation in an experi-mental model of acute lung injury. Intensive Care Med 2001; 27:566-573

7. Grotjohan HP, van der Heijde RMJL, Jansen JRC, WagenvoortCA, Verspille A. A stable model of respiratory distress by small


injections of oleic acid in pigs. Intensive Care Med 1996; 22:336-344

8. Gust R, Kozlowski J, Stephenson AH, Schuster DP. Synergistichaemodynamic effects of low-dose endotoxin and acute lung in-jury. Am J Respir Crit Care Med 1998; 157: 1919-1926

9. Hilker-Silva M, Leme AS, Arantes-Costa FM, Saldiva PHN,Lichtenstein A, Martins MA. Effects of mannitol in the acutelung injury induced by oleic acid in rats. Shock 2001; 16: 415-416

10. Jacobs RF, Kiel DP, Balk RA. Alveolar macrophage function ina canine model of endotoxin-induced lung injury. Am RevRespir Dis 1986; 134: 745-751

11. Lachmann B, Robertson B, Vogel J. In vivo lung lavage as an ex-perimental model of the respiratory distress syndrome. ActaAnaesth Scand 1980; 24: 231-236

12. Mandrekar P, Dolganiuc A, Bellerose G, Kodys K, Romics L,Nizamani R, Szabo G. Acute alcohol inhibits the induction of nu-clear regulatory factor kappa B activation through CD14/toll-like receptor 4, interleukin-1, and tumor necrosis factor recep-tors: a common mechanism independent of inhibitory kappa Balpha degradation? Alcohol Clin Exp Res 2002; 26: 1609-1614.

13. Nelson S, Bagby GJ, Bainteon G, Warren WR. The effects ofacute and chronic alcoholism on TNF? and inflammatory re-sponses. J Inf Dis 1989; 89: 422-429

14. Neumann P, Berglund JE, Mondejar EF, Magnusson A, Heden-stierna G. Effect of different pressure levels on the dynamics oflung collapse and recruitment in oleic-acid induced lung injury.Am J Respir Crit Care Med 1998; 158: 1636-1643

15. Pesenti A, Patroniti N. Update on the treatment of ARDS. Euro-pean Journal of Anesthesiology 2005; 22: S34, 61-65

16. Rabinovici R, Neville LF, Abdullah F, Phillip DR, Vernick J,Fong KL, Hillegas L, Feuerstein G. Aspiration-induced lung in-jury: Role of complement. Crit Care Med 1995; 23: 1405-1411

17. Ragaller M, Bleyl J, Tschö U, Winkler T, Regner M, Rasche S,Koch T, Albrecht M. Effects of inhalation of perfluorcarbonaerosol on oxygenation and pulmonary function compared toPGI2 inhalation in a sheep model of oleic acid-induced lung in-jury. Intensive Care Med 2001; 27: 889-897

18. Rigg JR. Pulmonary atelectasis after anaesthesia: pathophysiol-ogy and management. Can Anaesth Soc J 1981; 28: 305-313

19. Rosenthal C, Caronia C, Quinn C, Lugo N, Sagy M. A compar-ison among animal models of acute lung injury. Crit Care Med1998; 26: 912-916

20. Saeed RW, Varma S, Peng T, Tracey KJ, Sherry B, Metz CN.Ethanol blocks leukocyte recruitmant and endothelial cell acti-vation in vivo and in vitro. J Immunol 2004; 173: 6376-6383

21. Scillia P, Kafi SA, Melot C, Keyzer C, Naeije R, Gevenois PA.Oleic acid-induced lung injury: thin-section CT evaluation indogs. Radiology 2001; 219: 724-731

22. Sly PD, Bates JHT, Milic-Emili J. Measurement of respiratorymechanics using the siemens servo ventilator 900C. Pediatr Pul-monol 1987; 3: 400-405

23. Schuster DP. ARDS. Clinical lessions from the oleic acid modelof acute lung injury. Am J Resp Crit Care Med 1994; 149: 245-260

24. Schuster DP, Perez JE, Trulock EP, Williamson JR, Biello DR,Kenzora JL, Amundsen T, Lange LG. Cardiac dysfunction dur-ing acute lung injury induced by oleic acid in dogs. Am RevRespir Dis 1986; 133: 519-525

25. Schuster DP, Trulock EP. Correlation of changes in oxygenation,lung water and haemodynamics after oleic acid-induced acutelung injury in dogs. Crit Care Med 1984; 12: 1044-1048

26. Szabo G. Consequences of alcohol consumption on host defence.Alcohol & Alcoholism 1999; 34: 830-841

27. Szabo G, Mandrekar P. Ethanol-mediated regulation of transcrip-tion factors in immunocompetent cells. Front Biosci 2002; 7: 80-89

28. Vagts DA, Iber T, Szabo B, Haberstroh J, Reising K, Puccini M,Geiger K, Noldge-Schomburg GF. Effects of epidural anaesthe-sia on intestinal oxygenation in pigs. Br J Anaesth 2003; 90:212-220

29. Vagts DA, Iber T, Puccini M, Szabo B, Haberstroh J, Villinger F,Geiger K, Noldge-Schomburg GF. The effects of thoracicepidural anesthesia on hepatic perfusion and oxygenation inhealthy pigs during general anesthesia and surgical stress.Anesth Analg 2003; 97: 1824-1832

30. Young JS, Rayhrer CS, Edmisten TD, Cephas GA, Tribble CG,Kron IL. Sodium Nitroprusside mitigated oleic acid inducedacute lung injury. Ann Thorac Surg 2000; 69: 224-227

Address for correspondence: Thomas Iber, M.D., Department ofAnesthesiology and Intensive Care Medicine, University of Ros-tock, Schillingallee 35, D-18055 Rostock, Germany; E-Mail:[email protected]

S. Klaus, L. Bahlmann (Eds.)

Microdialysis - Monitoring Tissue Chemistry in Intensive Care MedicineExperimental Results and Clinical Application of Biochemical Tissue Monitoring

in Critical Illness

140 pages, ISBN 978-3-89967-124-7, Price: 18,- Euro

PABST SCIENCE PUBLISHERSEichengrund 28, 49525 Lengerich, Tel. ++ 49 (0) 5484-308, Fax ++ 49 (0) 5484-550,

E-Mail: [email protected] – Internet: www.pabst-publishers.de

Documents

applied cardiopulmonary pathophysiology Vol. 11, No. 1 ...€¦ · ing, and pulmonary arterial hypertension, consecutive-ly resulting in a deterioration of arterial oxygenation. Therapeutic