Hemodynamic Effects of Prolonged Hyperoxia JON F. MORAN, M.D.,* WALTER G. WOLFE, M.D.t

From the Department of Surgery, Duke University Medical Center, Durham, North Carolina

Experimental studies have consistently demonstrated the development of perivascular edema in the dog lung following prolonged exposure to 95% oxygen. This pathological change has been thought to result from capillary injury, but a direct effect secondary to left ventricular dysfunction has not yet been excluded. To evaluate the latter possibility, ten trained, awake dogs were prepared with monitoring of right and left atrial, systemic and pulmonary artery pressures, cardiac output, and mixed venous and arterial blood gases. Animals were exposed to an FIO2> 0.95 for 48-70 hours. Radioactive 8-10 ,u microspheres ('41Ce, 5'Cr, 85Sr, 46Sc) were injected into the left atrium at zero, six, 24, and 48 hours. Pao2 was 480 + 10 mm Hg during exposure, and the pulmonary shunt fraction increased from 11.3% to 16.9% (p < 0.0001) during 70 hours. Left atrial pressure fell from 9 + 2 mm Hg to 3 + 3 mm Hg (p < 0.0001), but cardiac output was constant at 2.7 + 0.1 I/min. Pulmonary arteriolar resistance increased from 183 + 20 dynes-secCM-5 to 791 + 30 at 70 hours (p < 0.0001). Histologic sections of the lungs demonstrated the characteristic perivascular edema. Of particular interest was the fact that myocardial perfusion was significantly increased to all three layers of the ventricular wall at 24 and 48 hours. These data indicate that perivascular edema developing after exposure to high concentrations of oxygen is secondary to pulmonary capillary endothelial damage with no evidence that myocardial dysfunction occurs during this period. D

ular failure in the pathogenesis of interstitial pulmonary edema during hyperoxia. 12'20'2' During the course of experimentally induced oxygen toxicity in mechanically ventilated and spontaneously breathing dogs, ventilatory dead space was increased following 24 and 36 hours hyperoxic exposure and decreased sharply after 48 hours.18 In the same study, pulmonary capillary filtration rate initially decreased and rose sharply after 48 hours exposure. These results are consistent with an initial vasoconstriction of the pulmonary vasculature in response to hyperoxia with a late, preterminal pulmonary vasodilatation. However, direct measurement of pulmonary vascular resistance at frequent intervals throughout the development of oxygen toxicity has not been previously reported. In order to assess the possible roles of left ventricular dysfunction, maldistribution of coronary blood flow, and pulmonary vasoreactivity in the pathogenesis of fluid accumulation within the lungs during prolonged hyperoxia, continuous monitoring of hemodynamic variables and serial measurement of regional coronary blood flow in spontaneously breathing dogs exposed to greater than 95% oxygen was undertaken.

URING THE PAST 20 YEARS, with the develop-

ment of efficient ventilators capable of delivering high concentrations of inspired oxygen, oxygen toxicity has become an increasingly well recognized clinical entity. Exposure of patients or experimental animals to greater than 95% oxygen for prolonged periods rapidly leads to diminished pulmonary function and perivascular interstitial pulmonary edema.1"3'7'9'19 The pathogenesis of oxygen toxicity and, in particular, the mechanisms of fluid accumulation within the oxygen-toxic lung have not been clearly delineated. Although ultrastructural studies have implicated direct damage to the pulmonary capillary endothelium with subsequent leakage of fluid into the interstitium during hyperoxic exposure,6'8 other evidence has implicated left ventric-

Methods

Ten mongrel dogs (13-18 kg) underwent sterile left thoracotomy under sodium pentobarbital (5 mg/kg) anesthesia. Sixteen gauge polyvinyl chloride catheters were implanted into the right and left atria, and the pulmonary and left subclavian arteries (Fig. 1). A silastic conduit (I.D. = 3 mm) was inserted through the anterior wall of the right ventricular outflow tract just beneath the pulmonic valve and sewn in place. The conduit and the four monitoring catheters were filled with heparin and their capped ends buried subcutaneously. Each animal was allowed to recover for a period of 3-6 weeks. During the recovery period, each animal was trained to stand and lie quietly while loosely tethered in an

* Research Fellow of the North Carolina and American Heart Associations. t Supported by USPHS Grant ROI HL 15877. Reprint requests: Walter G. Wolfe, M.D., P. 0. JBox 3507, Duke University Medical Center, Durham, North Carolina 27710. Submitted for publication: April 6, 1977. 0003-4932-78-0100-0073-0075 © J.B. Lippincott Company

73

74

MORAN AND WOLFE

FIG. 1. Schematic drawing showing positioning of chronic hemodynamic monitoring catheters in the central circulation. Position of thermistor-tipped thermodilution catheter is shown after insertion through chronically implanted Silastic conduit.

environmental chamber (Fig. 2). The ends of the previously implanted catheters were exteriorized under local anesthesia following the completion of training. A thermistor-tipped 7F thermodilution cardiac output catheter was introduced through the Silastic conduit and the tip positioned in the main pulmonary artery under fluoroscopy. After a baseline chest x-ray was taken, each animal was loosely restrained within the

Ann. Surg. * January 197R

environmental chamber and the four monitoring cathe ters connected to strain-gauge manometers through shielded fluid-filled 16 gauge connecting tubes for continuous monitoring of right and left atrial, pulmonary and systemic arterial pressure on an eight-channel recorder. Strain-guage manometers were rezeroed and calibrated against a mercury manometer every two hours. The thermistor-tipped pulmonary artery catheter was conncected to a thermodilution cardiac output computer. After baseline measurements were completed, the chamber was sealed and flushed with 100% oxygen. All vascular pressures and core temperatures were recorded continuously. At two hour intervals arterial and mixed venous blood gases (Po2, Pco2, and pH) were analyzed by standard techniques, hemoglobin concentration was determined by the cyanmethemoglobin method, and cardiac output was measured in triplicate after the injection of a five per cent dextrose in water solution (40) through the right atrial catheter. Oxygen tension within the chamber was maintained greater than 700 mm Hg and each animal remained in the chamber until death or 70 hours of exposure. Prior to sealing the chamber and after six, 24, and 48 hours exposure to hyperoxia, radioactively labelled ("1Ce, 51Cr, 85ST 46Sc) 7-10 ,u microspheres were injected throught the left atrial catheter with reference sampling through the left subclavian artery catheter for later calculation of quantitative regional myocardial blood flow.5 Animals surviving 70 hours exposure had chest x-rays prior to being sacrificed. At autopsy multiple lung sections were taken for histologic examination and the left ventricular free wall was carefully sectioned into subepcardial, midmyocardial, and sub-

Thermal dilution cardiac 1 output computer a:

000

1:1 0

Oxygen

0 C02~ °°° orber

r A

_5

FIG. 2. Schematic diagram of the experimental preparation showing the dog within the sealed oxygen chamber with attached

monitoring equipment.

Vol. 187 . No. 1

75

EFFECTS OF HYPEROXIA

endocardial layers for determination of regional myocardial perfusion. Pulmonary shunt fraction (QSQT) was calculated from arterial and mixed venous blood gas values and hemoglobin concentration using a Severinghaus slide rule and the standard shunt formula. Systemic (SVR) and pulmonary (PVR) vascular resistances were calculated from the vascular pressures and the measured cardiac output (C.O.) using standard formulae. Blood gas and hemodynamic variables were correlated with length of exposure time to hyperoxia using a multiple linear regression computer program. The significance of changes in mean and regional myocardial blood flows during exposure was assessed using an analysis of variance computer program.

Pe 02

0

*a-mle p :

*

* -.

FI:I"

*

0

1

9

(mmHg) 360

*

S

P.02a 501-0.9XTIME

240[

-0.25

ra

0~~~~~

120[

Results All animals remained quiet in the chamber without agitation throughout the exposure period. After 40-48 hours of hyperoxic exposure, the breathing pattern gradually changed to slower and deeper inspiration. The respiratory rate fell to less than 10/min during the preterminal two or three hours. Mean survival for the ten animals was 57 hours, with survival times ranging from 46 hours to 70 hours. Only two dogs survived the entire 70 hour period. All pre-exposure chest x-rays were normal except for minor postoperative blunting of the left costophrenic angle. Postmortem and postexposure chest x-rays showed a uniform ground-glass interstitial edema pattern throughout both lung fields. Histologic sections from all lobes of each animal showed the hallmarks of oxygen toxicity with perivascular cuffing, increased cellularity of alveolar septa, swelling of alveolar septa, and fluid accumulation within alveoli. Core temperatures of all animals remained relatively normal (

Hemodynamic effects of prolonged hyperoxia.

Hemodynamic Effects of Prolonged Hyperoxia JON F. MORAN, M.D.,* WALTER G. WOLFE, M.D.t From the Department of Surgery, Duke University Medical Center...
917KB Sizes 0 Downloads 0 Views