pulmonary edema related to changes colloid osmotic...

Pulmonary Edema Related to Changes in ColloidOsmotic and Pulmonary Artery Wedge Pressure in

Patients after Acute Myocardial Infarction

By PROTASIo L. DA Luz, M.D., HERBERT SHUBIN, M.D., MAx HARRY WEIL, M.D.,

EDWAIN JACOBSON, M.S., AND LEON STEIN, M.D.

SUMMARYPulmonary artery wedge and plasma colloid osmotic pressures and their relationship to pulmonary edema

were investigated in 26 patients with acute myocardial infarction of whom 14 developed pulmonary edema.In the absence of pulmonary edema, both the pulmonary artery wedge pressure and plasma colloid osmoticpressure were in normal range; after onset of pulmonary edema, a moderate increase in pulmonary wedgepressure and reduction in plasma colloid osmotic pressure were observed.When the gradient between the plasma colloid osmotic pressure and the pulmonary artery wedge

pressure was calculated, highly significant differences were demonstrated (P < 0.002). In the absence ofpulmonary edema, this gradient averaged 9.7 (+ 1.7 SEX1) torr; following appearance of pulmonary edema,it was reduced to 1.2 (+ 1.3) torr. During therapy with digoxin and furosemide, reversal of pulmonaryedema was closely related to a concomitant change in the colloid osmotic-hydrostatic pressure gradient.These observations indicate that both increases in pulmonary capillary pressure and decreases in colloid

osmotic pressure may follow the onset of pulmonary edema. Such decline in colloid osmotic pressure and es-

pecially the reduction in colloid osmotic-hydrostatic capillary pressure gradient may favor transudation offluid into the lungs.

Additional Indexing Words:Pulmonary congestion Pulmonary pressuresDiuretic action Colloid osmotic-hydrostatic pressure gradient

pULMONARY EDEMA after acute myocardialinfarction has been attributed to an increase in

the hydrostatic pressure in the pulmonary capillarybed due to left ventricular failure.1-4 The level of theleft ventricular filling pressure, however, does notalways correlate with radiographic signs of pulmonaryedema.5 6 Guyton and Lindsey7 demonstrated in dogsthat pulmonary edema occurred at left atrial pressurelevels of 24 torr when the albumin concentration in

From the Shock Research Unit and the Department of Medicine,University of Southern California School of Medicine, the LosAngeles Countv/USC Medical Center, and the Center for theCritically Ill, Hollywood Presbyterian Medical Center, Los Angeles,California.

Supported by USPHS Research Grants HL-05570 from theNational Heart, Lung and Blood Institute, GM-16462 from theNational Institute of General Medical Sciences, ROI HS 01474 fromHealth Resources Administration, and the Parker B. Francis Foun-dation.

Dr. Luz s present address is Cedars-Sinai Medical ResearchInstitute, 4751 Fountain Avenue, Los Angeles, California 90029.Address for reprints: Max Harry Weil, M.D., Center for the

Critically Ill, University of Southern California School of Medicine,1300 North Vermont Avenue, Los Angeles, California 90027.

Received June 20, 1974; revision accepted for publication Oc-tober 10, 1974.

350

Circulatory failure

the plasma was normal. After the albumin concentra-tion was reduced to 47% of the control values,pulmonary edema appeared when left atrial pressurewas increased above 11 torr. Levine et al,,8 also instudies on dogs, found no consistent relationshipbetween the gradient of capillary hydrostatic andcolloid osmotic pressure with the rate of fluid ex-travasation into the lungs. These investigators pointedout the role of tissue hydrostatic and oncotic pressuresas potentially important determinants of pulmonaryedema.

Current techniques for measurements of interstitialfluid hydrostatic and colloid osmotic pressure are not,as yet, applicable for clinical investigation. However,it is now possible to obtain an estimate of pulmonaryvenous pressure by the use of the flow-directedpulmonary artery catheter developed by Swan andGanz.9 Reliable measurements of plasma colloid os-motic pressure are also feasible by methods evolvedby Prather et al."0 and modified by Jacobson et al." inour laboratories. It is, therefore, possible to assesssimultaneous changes in the hydrostatic and colloidosmotic pressure in patients. Our investigation wasdirected to the practical issue of whether or not, in theabsence of quantitation of tissue hydrostatic and

Circulation, Volume 51, February 1975

by guest on June 20, 2018http://circ.ahajournals.org/

Dow

nloaded from

http://circ.ahajournals.org/

COLLOID OSMOTIC PRESSURE AFTER INFARCTION

colloid osmotic pressure, pulmonary edema could berelated to an altered plasma colloid osmotic-hydrostatic pressure gradient in the pulmonarycapillaries. The problem was specifically investigatedin 26 patients who had sustained an acute myocardialinfarction of whom 14 developed unequivocalradiographic signs of pulmonary edema.

Material and Methods

Clinical Material

Seventeen men and nine women, whose ages ranged from40 to 89 (median 65) years and who were admitted to theCoronary Care Unit of the University of Southern CaliforniaCenter for the Critically Ill between November 1971 andJune 1973, constitute the subjects of this report. Diagnosis ofacute myocardial infarction was based on compatibleclinical history, characteristic Q waves and/or ST segment-T wave changes, significant increases in the levels of serum

creatine phosphokinase, glutamic oxaloacetic transaminaseand lactic dehydrogenase. Seven patients presented withclinical signs of acute perfusion failure (shock) includingclammy skin, mental disorientation, reduction in intra-arterial systolic blood pressure to 90 torr or less, urine outputless than 30 ml/hr and arterial lactate concentration ex-

ceeding 1.4 mM/L. Only one of these patients with shocksurvived for more than 15 days. Pulmonary edema was

observed predominantly in older patients (table 1).

Methods

In nineteen patients, studies were performed during thefirst 24 hours following onset of severe chest pain and inseven patients, at intervals ranging from 2 to 10 days afterthe acute episode. Informed consent was routinely obtained.Diazepam, meperidine and/or oxygen were routinely ad-ministered prior to study; digoxin (0.5 mg) and furosemide(40 to 80 mg) had been administered to five patients(patients 1, 2, 7, 9, 19). Seven patients were mechanicallyventilated after endotracheal intubation (patients 1, 2, 5, 10,18, 23, and 25). Anteroposterior supine chest X-rays were

obtained within two hours of the hemodynamic studies.

Table 1

Clinical Data in 26 Patients with Acute MyocardialInfarction

Pulmonary edema Pulmonary edemaabsent present(12) (14) P value

Age (yr) 40 to 68 44 to 89 <0.001(median 53) (medial) 75)

SexMale 8 9 NSFemale 4 ,

LocationAnterior 4 10 NSInferior 8 4

OutcomeSurvivors 10 .) <0.025Died 2 9


Films were taken using parspeed screens and film (Kodak X-Omatic Cassette, screens and film) with 2/2 minute develop-ment. Tube-to-chest distance was 40 inches. Interpretationof the chest radiographs was independently reported by tworadiologists who had no prior knowledge of the clinicalstatus of the patients. Criteria for interpretation were thoseproposed by Turner et al. 12 adapted for reading supine chestfilms (table 2).

After local anesthesia utilizing lidocaine, 2%, a radioopa-que flow-directed catheter (Swan-Ganz catheter Model93/110/5F or 93/111/7F, Edwards Laboratories) was sur-gically inserted into a basilic vein. The tip of the catheterwas advanced to a position in the peripheral pulmonaryartery for wedge pressure recording as initiallyrecommended by Swan et al.9 The correct position wasdocumented by chest X-ray. Proper wedging of the catheterwas determined by: 1) immediate loss of typical arterialpulse morphology on inflation of the balloon, documentedby the pressure recording; 2) abrupt return of typical arterialpulse morphology on deflation of the balloon; 3) appearanceof wedge pressure morphology consistent with that of theleft atrial pressure pulse; 4) a mean pulmonary wedgepressure equal to or less than the pulmonary artery end-diastolic pressure. A thin walled teflon catheter (Longdwel,18 gauge, internal diameter 1.05 mm, length 20 cm, Becton,Dickinson and Company) was percutaneously inserted intoa femoral artery and advanced for a distance of ap-proximately 15 cm. A Statham P23Db pressure transducerwas used in conjunction with a two channel (Model 7702B)or eight channel (Model 7878A) Hewlett Packard medicalrecorder. The mid chest level was established as zeroreference for intravascular pressure measurements.

Cardiac output was measured by the dye-dilutiontechnique with injection of 2.5 or 5 mg of indocyanine greendye into the pulmonary artery and withdrawal of blood fromthe femoral artery utilizing either a Gilson (Model 17500A)or Waters (Model PR-5M) densitometer. The area undereach dye curve was calculated by numerical integration.`3

Table 2

Critera of Pulmonary Edema(adapted from Turner, Lau and Jacobson12)

Severity Radiographic criteria Number of patients

0 Normal pulmonary vessels with 6orderly t,apering toward theperiphery

1+ Equal perfusion of upper and lower 6lung zones

2+ a) Minimal interstitial prominence 2b) Perivascular edema insufficient

to obliterate vascular markings

3+ a) Prominent interstitial markings 6(Kerley's B lines)

b) Diffuse reticular pattern withobliteration of the vascularmarkings

4+ Bilateral confluent shadows of 6relatively uniform deinsity butill defined margins

351


Dow

nloaded from


LUZ ET AL.

Left ventricular stroke work index (SWI) was calculated bythe formula:

SWI = SI X(MAP - PAWP) X 0.0136

in which SI = stroke work index, MAP = mean arterialpressure and PAWP = pulmonary artery wedge pressure.

Samples of blood were obtained from the femoral arterywithin an interval of less than three minutes following in-scription of the indicator dilution curve. Samples were sub-merged in iced water and blood gas analyses were per-formed within 30 min after blood withdrawal. The partialpressure of oxygen (P02), carbon dioxide (PCO2) and pHwere measured at 37°C with the standard electrode techni-que utilizing a radiometer system (Radiometer Company,Denmark, Model PHA927). The output of the pH, P02, andPCO2 amplifiers were recorded on a strip chart recorder(Hewlett Packard, Model 680) and the appropriate meterwas read after the slope of the recording was zero for 30 sec.Plasma volume was measured using radioiodinated humanserum albumin (RI 125SA). Isotope activity was determinedat intervals of 15, 25, 35, and 45 min with subsequent ex-trapolation of the values to zero time.

Colloid osmotic pressure was measured with a transducermembrane system, utilizing a modification of methodsproposed by Prather et al.`0 A Diaflow R ultrafilter IOPM30(Amicon, Lexington, Massachusetts) was rigidly mountedabove an airtight saline chamber. The membrane isspecified to be impermeable to molecules larger than 30,000MW. The oncometer incorporated a Statham P23Db straingauge pressure transducer in conjunction with a differentialamplifier of our own design and Hewlett Packard Model7100 BM ten inch recorder. The transducer was calibratedfrom 0 to 40 torr by use of a saline column and preciselyspaced platforms; the fluid column was lowered in steps of13.6 cm (10 torr). Zero reference was obtained by pipeting0.3 ml of saline solution into the chamber above the mem-brane. Since the lower chamber also contained saline therewas no osmotic gradient across the membrane. The salinewas then absorbed with a pledget of gauze and 0.3 ml of thepatient's plasma was delivered into the chamber; the plasmawas obtained from blood anticoagulated with sodiumheparin in amounts of less than 100 units/ml. Allmeasurements were made at room temperature, rangingfrom 25 to 28°C.The validity of the method was confirmed by

measurements of colloid osmotic pressure of crystalinebovine albumin in concentrations of 1 to 7C and bymeasurements on normal ambulatory and supine subjects.Within the range of 3 to 6 gm % of albumin concentration,the colloid osmotic pressures were plotted against thealbumin concentrations. This demonstrated a linear regres-sion line. The values of the small F-statistic for testing thehypothesis of linearity [F(2, 4) = 1,349] and the large Pvalue (close to 0.50) incidate that the data do not contradictthe hypothesis of a linear regression model. In 19 am-bulatory subjects, colloid osmotic pressure was 25.4 ± 2.3torr and in 14 subjects who had been at bed rest for morethan 12 hours, colloid osmotic pressure was 21.6 ± 3.6 torr.These results were comparable to those previouslyreported.'4 Refrigerated 5% human serum albumin (HylandLaboratories, Costa Mesa, California) was used as areference prior to each measurement. Stability andreproducibility of the reference measurements weredemonstrated over a 10 month period; the mean colloid os-motic pressure of 45 reference measurements was 19.3 torrand the maximum deviation was less than ± 1 torr.

Duplicate measurements on 40 samples in which colloid os-motic pressure ranged from 4 to 29 torr, revealed an averagedifference of less than + 0.1 torr and a correlation coefficientof 0.999."5The two-tailed Student's t-test was used for statistical

comparison of the data.

Results

1. Radiographic Interpretation

Chest roentgenograms were classified in severitygroups 0 and 1 in 12 cases, and for purposes ofstatistical analysis these were designated as cases ofuncomplicated acute myocardial infarction. "Early"signs of interstitial edema were observed in the twocases classified in category 2 and more advancedstages of pulmonary edema were observed in twelveadditional patients. For purposes of statisticalanalysis, these 14 patients were designated as cases ofacute myocardial infarction complicated bypulmonary edema (table 2).

2. Hemodynamic Measurements

The mean arterial pressure and heart rate werewithin normal range in both groups of patients (table3). Although the cardiac index and stroke work indexwere reduced in both groups of patients, no significantdifference was observed after the onset of pulmonaryedema.

Pulmonary artery systolic, mean, diastolic andwedge pressures were slightly but significantly higherin patients with pulmonary edema (table 4). Thewedge pressures averaged only 15.7 (+ 1.5 SEMI) torrin patients with pulmonary edema and 11.0 (± 1.2)torr in the absence of pulmonary edema. In theabsence of pulmonary edema, the arterial oxygen ten-sion was 124 torr and in patients with pulmonaryedema, it was 96 torr. Although the oxygen content ofinspired gas was an uncontrolled variable, anoxemiawas excluded as a mechanism accounting for therelatively higher pulmonary artery pressures valuesobserved among patients with pulmonary edema.

3. Colloid Osmotic Pressure and Its Relationships toPulmonary Artery Wedge Pressure

In the absence of pulmonary edema, colloid os-motic pressure was 20.8 (± 1.0) torr but it wasreduced to 16.9 (± 1.1) torr in patients withpulmonary edema (P < 0.02).Among the 14 patients who had pulmonary edema,

six also were in shock and plasma colloid osmoticpressure averaged 15.5 (± 1.40) torr; in the remainingeight patients, in whom pulmonary edema wasobserved in the absence of shock, the plasma colloidosmotic pressure averaged 18.0 (+ 1.59) torr, astatistically insignificant difference for this smallgroup.


352


Dow

nloaded from



Table 3

Hemodynamic and Respiratory Measutrements

HR MAP CI SWI Pao2 pHa Paco2Patient (beat/min) (torr) (L/min/M2) (gm-m/mi) (torr) (units) (torr) Fio,

AMAI, uncomplicated1 68 76 1.6 18.6 .50 7.38 37 12 100 63 1.9 14.2 a3 7.41 47 13 120 97 3.6 34.7 72 7.45 38 14 103 86 1.0 9.5 132 7.47 28 25 60 85 2.8 43.1 64 7.45 39 16 116 110 1.6 18.6 83 7.36 46 17 68 62 1.7 17.7 65 7.38 43 18 102 85 2.0 19.2 108 7.45 34 29 73 104 2.3 41.6 3,58 7.40 43 310 120 80 3.2 27.9 103 7.46 43 211 69 73 1.4 18.2 323 7.51 28 312 77 80 2.2 25.6 74 7.38 39 1

Mean 90 83 2.1 24.0 123.7 7.42 38.7SEM -6.5 =4.3 .0.2 -3.1 -30.1 30.01 0. 8AMI + pulmonary edema

13 103 62 1.6 8.2 261 7.46 41 214 107 62 1.5 9.5 76 7.30 .54 315 98 83 1.3 11.7 59 7.47 34 116 68 70 1.4 17.0 74 7.46 36 217 83 85 3.8 46.0 80 7.43 42 218 84 64 1.6 13.5 78 7.34 36 219 75 85 1.4 15.2 74 7.34 47 220 76 64 2.6 21.9 61 7.33 42 121 90 57 1.1 8.3 136 7.49 30 222 70 64 1.8 17.5 139 7.50 30 223 92 120 3.0 46.6 64 7.44 42 224 126 .54 2.4 7.5 82 7.33 31 225 107 67 1.8 11.7 97 7.33 37 226 68 85 2.1 29.0 68 7.40 43 1

Mean 89.0 73.0 1.9 18.8 96.3 7.40 38.9SEM 4.7 4.6 =40.2 - 3. 14.3 -0.02 1.8AMI, uncomplicated vsAMI + pulmonary edema

P value NS NS NS NS NS NS NS

Abbreviations: HR = heart rate; MAP = mean arterial pressure; CI = cardiac index; SWI = stroke workindex; pHa = arterial pH; PaO2 = arterial P02 ; F102 = inspired oxygen fraction (1-room air; 2-room air+ 02 by mask ranging from 2 to 10 1/min; 3-100% 02).

When both colloid osmotic pressure and pulmonaryartery wedge pressure were taken into account for es-timation of the net effects of these forces on fluidtransfer between the intravascular and extravascularcompartments, highly significant differences weredemonstrated. In 13 normal supine individualspreviously studied in this laboratory, the plasmacolloid osmotic pressure averaged 20.6 (± 3.9 SD) torr;the pulmonary artery wedge pressure normally rangesfrom 4 to 12 torr. 16 Based on these values, and withoutreference to changes in interstitial colloid andhydrostatic pressure or alterations in membranepermeability, a gradient of 9 to 17 torr normally favorsthe retention of fluid within the intravascular com-partment. Among the patients without pulmonaryCirculation, Volume 51, February 1975

edema, the colloid osmotic-hydrostatic pressuregradient averaged 9.7 (± 1.7) torr, a low normalvalue. In 14 patients with pulmonary edema,however, this gradient was reduced to 1.2 (+ 1.3) torr(P < 0.002) and it exceeded 3 torr in only three cases(fig. 1).

4. Fluid Balance and Intravascular VolumePatients with pulmonary edema had a positive fluid

balance of 598 ± 231 ml in the eight hours precedingthe study. In three of these patients, the positive fluidbalance was not associated with a reduction in colloidosmotic pressure. A positive fluid balance of116 ± 183 ml also was documented among thepatients without pulmonary edema. There was no

353


Dow

nloaded from


LUZ ET AL.

Table 4

Pressures and Plasma Osmotic

Patient PPAS PPA PPAD PPAW PCoP

A MI, uncomplicated1 36 23

2 20 12

3 29 224* 245 18

5 45 30

6 20 14

7 28 16

8 42 24

9 22 13

10 17 12

11 16 10

12 24 18

Mear1 27.0 17.6

SEM *2.7 1.7

AMI + pulmonary edema13* 33) 27

14* 38 28

15* 42 30

16 26 16

17 35 21

18*19 40 30

20 38 25

21* 23 16

22* 35 22

23 32 22

242526

MeanSEM

46

30

34.9*1.9

301923.8*1.5

AMI, uncomplicated vs

AMI + pulmonary edemaP value <0.05 <0.02

178

171525111015

98

8

1513.2

*1.5

23

23

171114

251912

15

18

24

13

17.8*1.4

178

12141711101373

714

11.0*1.2

23

12

18

91112t25

17

714

1,25t161613.7-1.5

21182118

142125222319202820.8*1.0

1915201614131418

1 .5

111928191616.91.1

<0.05 <0.03 <0.02

Abbreviations: AMI = acute myocardial infarction; PPAS= systolic pulmoinary artery pressure; PAP = mean pul-monary artery pressure; PPAD = diastolic pulmonary arterypressure; PPAW = pulmoinary artery wedge pressure; Pcop= plasma colloid osmotic pressure.*Further complicated by acute circulatory failure (shock).tLeft ventricular end-diastolic pressuire (torr).

significant difference in the positive fluid balancebetween the two groups of patients. In addition, no

significant differences in plasma volume and plasmasodium were detected (table 5).

5. Effects of TherapyIn nine patients with pulmonary edema,

measurements were repeated at an average interval of26 hours for the purpose of evaluating changes incolloid osmotic pressure and pulmonary artery wedgepressure following treatment with digoxin and

+16

+ 12

+ 8 -

+ 4

o+

4±

12

~~~~~~~~~~~~~~~~~~.......

........-.,.,

............ NORMAL.RANG-.--,~ ~ ~S...... .....~~~~~~~~~~~~~~~~~.. .. .. .. .. .. .. ----.

0

Scm

0 0

0

0

UNCOMPLICATED PULM. EDEMA

p < 0.002

Figure 1

Plasma colloid osmotic-pulmonary artery wedge pressure gradientsin 26 patients.

furosemide. In three patients in whom pulmonaryedema was reversed, the plasma colloid osmoticpressure remained in the normal range or increased. Adecline in pulmonary wedge pressure was alsoobserved (fig. 2). In six cases in which the clinicaltreatment failed to reverse pulmonary edema, plasmacolloid osmotic pressure was either unchanged orfurther reduced. The effects on pulmonary arterywedge pressure were variable. Nevertheless, a closerelationship between the course of pulmonary edemaand changes in the colloid osmotic-hydrostaticpressure gradient was demonstrated (fig. 3). Thisgradient increased in each of the patients in whompulmonary edema was reversed. However, the

Table 5Intravascular Volume and Fluid Balance

Positive Plasmafluid voluime Plasma

balance (mi/kg) Nal (mEq/1)

AMI, uncomplicated 4/8 43.5* 134.7-2.0 *1.3

(8) (4) (9)AMI + pulmoniary edema 10/14 39.0 135.5

*3.5 *1.8(14) (8) (14)

Abbreviation: AMI = acute myocardial infaretion.Nuimbers in parentheses represents total of observations

in each group.*Mean - SEM.


Pulmonary ArteryPressures

354


Dow

nloaded from



30 0-

0

O --- -

0

o 0D

10 O. h O

4 4 26hrs_-*

Figure 2

Changes in pulmonary artery wedge pressure (left) and colloidosmotic pressure (right) in relation to the course of pulmonaryedema in 9 patients.

gradient remained below 4 torr or was further reducedwhen treatment proved ineffective.

Discussion

Experimental ligation of a coronary artery"7 18 oracute myocardial infarction in patients'9-22 is followedby increases in left ventricular end-diastolic pressure.The fact that these increases account for elevation inleft atrial pressure and in turn pulmonary venouspressure is also well established. This increase inhydrostatic pressure is generally regarded as the maincause of pulmonary edema after acute myocardial in-farction.However, increases in left ventricular filling

pressure do not always explain the presence ofpulmonary edema. Forrester et al.'23 in their study offifty patients with acute myocardial infarction, con-

REVERSAL OFPULM. EDEMA

PERSISTENCE OFPULM. EDEMA

< 26 hrs

Figure 3

Changes in the colloid osmotic-pulmonary artery wedge pressure

gradient following treatment with digoxin and furosemide in 9patients.

sistently observed radiographic signs of pulmonaryedema when pulmonary artery wedge pressures ex-ceeded 17 torr. McHugh et al.,5 in a subsequentreport, proposed 22 to 25 torr as the critical range ofpulmonary artery wedge pressure for appearance ofalveolar pulmonary edema. In the majority of patientsreported by Lassers et al.24 a good correlation betweenthe level of pulmonary artery wedge pressure and thepresence or absence of pulmonary edema was alsofound. However, these authors were unable to explainpulmonary edema in three patients in whom thepulmonary artery wedge pressure was lower than 16torr. A recent report by Kostuk et al.6 included 86patients who had sustained acute myocardial infarc-tion. The level of pulmonary capillary wedge pressurecorrelated with the radiographic findings in fewerthan one-half of their cases. Of 32 patients with in-terstitial pulmonary edema, wedge pressure was equalto or less than 18 torr in 18 cases. On the other hand,in 17 patients, the chest X-ray was normal despite per-sistent elevation of the wedge pressure for six to 24hours.When the left ventricular end-diastolic pressure was

measured directly following retrograde catheteriza-tion of the aorta, similar discrepancies were noted.Hamosh and Cohn25 studied 40 patients with acutemyocardial infarction; in six of these cases, elevationsin the left ventricular end-diastolic pressure to levelsranging from 25 to 40 torr were observed, and yetclinical signs of pulmonary edema were absent. Nix-on,26 on the other hand, reported pulmonary edema inthe absence of an increase in left ventricular end-diastolic pressure in one patient with acute myocardialinfarction.

In our series a similar discrepancy was observed.Pulmonary artery wedge pressure averaged only 15.7(+ 1.5) torr among the 14 patients with pulmonaryedema. However, when both plasma colloid osmoticpressure and pulmonary wedge pressure were takeninto account by computing an oncotic-hydrostaticgradient, a more sensitive index for separatingpatients with and without pulmonary edemaemerged. All patients with pulmonary edema had agradient of less than 9 torr, and in only three instanceswas it less than 8 torr in the absence of pulmonaryedema. The same applied when pulmonary edemawas reversed by therapy. When pulmonary edemawas reversed, the gradient increased to 7 torr or more.When treatment failed to reverse pulmonary edemathe gradient remained less than 2 torr.The possibility that increases in pulmonary

capillary pressures preceded the onset of pulmonaryedema and that lowering of colloid osmotic pressurewas the result rather than the cause of pulmonaryedema can not be excluded. Patients with pulmonary

30

20 -

10-

00

\11o con0 \

0 1

PERSISTENCEOF PULM. EDEMA

REVERSALOF PULM. EDEMA

4 L 226 lhrs --

+ 14 :+122

+ 8-

+ 4-

0 a

- 4.

355

1

2

.111


Dow

nloaded from


LUZ ET AL.

edema were older and more often presented withclinical signs of circulatory shock; the low perfusionstate itself may increase permeability of the alveolar-capillary membrane with consequent leakage ofplasma fluid into pulmonary interstitium.27 This couldalter both hydrostatic and colloid osmotic interstitialpressures; in the absence of these measurements,however, the role of these variables remainsspeculative.

Nevertheless, the reduction of colloid osmoticpressure regardless of cause would of itself facilitatetransudation of fluid into the pulmonary interstitiumand alveoli at relatively low left ventricular fillingpressure.7 8 Since in this series measurements wereperformed when the patients were in pulmonaryedema rather than preceding it, a cause-effectrelationship between decreased colloid osmoticpressure and the development of pulmonary edemawas not demonstrated. However, the data are une-quivocal with respect to changes that follow treatmentof pulmonary edema. It is the combined effects ofchanges in pulmonary hydrostatic pressure andplasma colloid osmotic pressure that correlate withreversal or persistence of pulmonary edema. Theseinitial observations make it likely that substantial in-creases in precision for quantifying the risk ofpulmonary edema will emerge from the concurrentmeasurement of colloid osmotic pressure andpulmonary artery wedge pressure. Of equal impor-tance, these measurements provide opportunity forfurther studies of the mechanisms of pulmonaryedema both in the presence and absence of heartfailure.The mechanism by which the colloid osmotic

pressure decreased in patients with pulmonary edemaremains unexplained. Among the possible causes, in-creases in capillary permeability to albumin havealready been cited. Experimentally,28 the infusion ofcrystalloids may provoke pulmonary edema withoutconsistent elevation in pulmonary artery wedgepressure. In patients without heart failure thedevelopment of pulmonary edema after fluid ad-ministration in association with lowering of colloid os-motic pressure, has been recently reported by ourgroup.29 Although no consistent relationship betweenthe volumes of fluid administered and reductions incolloid osmotic pressure could be demonstrated in thisseries, we cannot exclude the possibility that the infu-sion of crystalloids may have contributed to loweringof osmotic pressure on a dilutional basis.No clinical evidence of malnutrition or renal losses

of protein were observed in our patients and thereforechronic protein deficiency cannot be implicated as afactor in the genesis of pulmonary edema.The present data should not distract from the fact

that left ventricular dysfunction is the major andpossibly initial factor accounting for pulmonaryedema after acute myocardial infarction. What isknown from this study is that after acute myocardialinfarction, pulmonary edema develops in close quan-titative relationship with simultaneous changes inboth colloid osmotic and hydrostatic pressures.

Therapeutic Implications

The administration of fluids using a well regulatedtechnique of fluid challenge has been demonstrated tobe an effective method for reversing shock followingacute myocardial infarction. Admittedly, it reversedshock in only a minority of patients.30 The pulmonaryartery wedge pressure serves as the primary monitorfor assessing the response to volume loading. McHughet al.5 suggested the level of 18 torr as a safe level forfluid challenge. On the basis of the present study, wewould regard concomitant measurements of colloidosmotic pressure and pulmonary artery wedgepressures as a more reliable basis for guiding fluidchallenge. Optimally, the colloid osmotic-wedgepressure gradient should be calculated. When thisgradient declines below 9 torr, there is a substantialrisk of pulmonary edema if fluid infusion is continued.

Acknowledgment

We are grateful to Doctors John Cerrone, James O'Dea andDalvin Smolin for interpretation of radiographs; Mrs. Vinnie Liu forassistance in statistical analyses and Mrs. Linda Bowen and MissMary Armato for assistance in preparation of this report. The colloidosmometer (Oncometer) was developed with the aid of Mr. JoeBisera and Mr. Michael Chaffee.

References

1. WEL(H WH: Zur pathologie des lugennodems virchows. ArchPath Anat 72: 375, 1878

2. FISHMIAN P: Pulmonary edema - the water exchangingfunction of the lung. Circulation 46: 390, 1972

3. ABEnRANIAN A, FULOP M: The metabolic and respiratory acidosisof acute pulmonary edema. Ann Intern Med 76: 173, 1973

4. FL LOP M, HOROWITZ M, ABERMAN A, JOFFE ER: Lactic acidosisin pulmonary edema due to left ventricular failure. AnnIntern Med 79: 180, 1973

5. McHU(GH TJ, FORRESTER JS, ADLER L, ZION D, SWAN HJC:Pulmonary congestion in acute myocardial infarction:Hemodynamic and radiologic correlations. Ann Intern Med76: 29, 1972

6. KOSTUK W, BARR JW, SIMON AL, Ross J JR: Correlationsbetween the chest film and hemodynamics in acute myocar-dial infarction. Circulation 48: 624, 1973

7. GuCi-o'. A, LINDSEY AW: Effect of elevated left atrial pressureand decreased plasma protein concentration on the develop-ment of pulmonary edema. Circ Res 7: 649, 1959

8. LEVINE OR, MELLINS RB, SENIOR RM, FISHMAN AP: Theapplication of Starling's law of capillary exchange to thelungs. J Clin Invest 46: 934, 1967

9. SWs ANN HJC, GANZ W, FORRESTER JS, MARCuS H, DIAMOND G,

356


Dow

nloaded from



CHONETTE D: Catheterization of the heart in man with useof a flow directed balloon-tipped catheter. N Engl J Med283: 447, 1970

10. PRATHER JW, GAAR KA JR, GuYTON AC: Direct continuousrecording of plasma colloid osmotic pressure of whole blood.J Appl Physiol 24: 602, 1968

11. JA(OBSON EJ, WEIL MH, RITZ R, SHUBIN H: Clinicalmeasurement of plasma colloid osmotic pressure. (abstr)Physiologist 14: 167, 1971

12. TuRNEER AF, LAu FV, JACOBSON G: A method for the estimationof pulmonary venous and arterial pressure from the routinechest roentgenogram. Am J Roentgenol 116: 97, 1972

13. HAMXlILTON WF, MOORE JW, KINSMAN JM, SPUIRLINC RG:Studies on the circulation. IV. Further analysis for the injec-tion method and of changes in hemodynamic underphysiological and pathological conditions. Am J Physiol 99:534, 1932

14. LADECAARD1-PEDERSEN HJ: Measurements of colloid osmoticpressure in patients. Scand J Clin Lab Invest 20: 79, 1967

15. WEIL MH, MORISSETTE M, MICHAELS S, BISERA J, BOYCKS E,SHUBIN H, JACOBSON E: Routine plasma colloid osmoticpressure measurements. Critical Care Medicine, In press

16. F.ALIcox RE, RESNEKOv L: Relationship of pulmonary arteryend-diastolic pressure to the left ventricular end-diastolicand mean filling pressures in patients with and without leftventricular dysfunction. Circulation 42: 65, 1970

17. REGAN TJ, PASSANNANTE AJ, KHAN MI, OLDEWURTEL HA,JESRANI MU: Influence of scar on left ventricular perfor-mance at the onset of myocardial ischemia: Shock versusheart failure. J Clin Invest 50: 534, 1971

18. HOOD WB JR, MCCARTHY B, LoWN B: Myocardial infarctionfollowing coronary ligation in dogs. Hemodynamic effects ofisoproterenol and acetylstrophantidin. Circ Res 21: 191,1967

19. PARNILEY WW, DIAMOND G, TOMODA H, FORRESTER J, SWAN

HJC: Clinical evaluation of left ventricular pressures inmyocardial infarction. Circulation 45: 358, 1972

20. RA-1IHSHIN RA, RACKLEY CE, RUSSELL RO JR: Hemodynamicevaluation of left ventricular function in shock complicatingmyocardial infarction. Circulation 45: 127, 1972

21. KIRBY BJ, McNICHOL MW, TATTERSFIELD AE: Left ventriculardiastolic pressure in acute myocardial infarction. Lancet 1:944, 1968

22. RASHI-ITOOLA SH, LOEB HS, EHSANI A, SIMM MZ, CHUQU.IMIAR, LOL R, ROSEN KM, GUNNAAR RM: Relationship ofpulmonary artery to left ventricular diastolic pressures inacute myocardial infarction. Circulation 46: 283, 1972

23. FORRESTER JS, DIAMOND G, McHUGH TJ, SWAAN HJC: Fillingpressures in the right and left sides of the heart in acutemyocardial infarction. A reappraisal of central venouspressure monitoring. N Engl J Med 285: 190, 1971

24. LASSERS BW, GEORGE M, ANDERSON JL, HIGGCINS MR, PHILIP T:Left ventricular failure in acute myocardial infarction. Am JCardiol 25: 511, 1970

25. HAMIOSH P, COHN JN: Left ventricular function in acutemyocardial infarction. J Clin Invest 50: 523, 1971

26. NIXON PGF: Pulmonary edema with low left ventriculardiastolic pressure in acute myocardial infarction. Lancet 2:146, 1968

27. ROBIN ED, CAREY LC, GRENVIK KA, GLANSER F, GANDio R:Capillary leak syndrome with pulmonary edema. ArchIntern Med 130: 66, 1972

28. LEVINE OR, MELLINS RB, FISHMAN AP: Quantitativeassessment of pulmonary edema. Circ Res 17: 414, 1965

29. STEIN L, BERAUD JJ, Luz PL, WEIL MH, SHUBIN, H: Pulmonaryedema during fluid infusion in the absence of heart failure.JAMA 229: 65, 1974

30. Luz PL, WEIL MH, Liu VY, SHUBI.N H: Plasma volume prior toand following volume loading during shock complicatingacute myocardial infarction. Circulation 49: 98, 1974

357


Dow

nloaded from


P L Luz da, H Shubin, M H Weil, E Jacobson and L Steinpressure in patients after acute myocardial infarction.

Pulmonary edema related to changes in colloid osmotic and pulmonary artery wedge

Print ISSN: 0009-7322. Online ISSN: 1524-4539 Copyright © 1975 American Heart Association, Inc. All rights reserved.

is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231Circulation doi: 10.1161/01.CIR.51.2.350

1975;51:350-357Circulation.

http://circ.ahajournals.org/content/51/2/350the World Wide Web at:

The online version of this article, along with updated information and services, is located on

http://circ.ahajournals.org//subscriptions/

is online at: Circulation Information about subscribing to Subscriptions:

http://www.lww.com/reprints Information about reprints can be found online at: Reprints:

document.

Permissions and Rights Question and Answer Further information about this process is available in therequested is located, click Request Permissions in the middle column of the Web page under Services.the Editorial Office. Once the online version of the published article for which permission is being

can be obtained via RightsLink, a service of the Copyright Clearance Center, notCirculationpublished in Requests for permissions to reproduce figures, tables, or portions of articles originallyPermissions:


Dow

nloaded from

http://circ.ahajournals.org/content/51/2/350

http://www.ahajournals.org/site/rights/

http://www.lww.com/reprints

http://circ.ahajournals.org//subscriptions/


pulmonary edema related to changes colloid osmotic...

Documents