Severity of hypoxia predicts response to nitric oxide in a porcine pulmonary hypertension model

Severity

of Hypoxia

By Sherif Emil, Mahmoud

Predicts Pulmonary

Response to Nitric Oxide Hypertension Model

Kosi, Joel Berkeland, Sotaro Kanno, Christopher Los Angeles, California

l Although inhaled nitric oxide (NO) has been variably successful in resolving pulmonary hypertension in neonates, children, and adults, no parameters predictive of response to this therapy have been elucidated. We conducted an animal study to determine if severity of hypoxia can predict magnitude and sustenance of response to inhaled NO therapy. Seven Yorkshire swine weighing 11 to 20 kg underwent 16 experiments, each consisting of four phases: Phase 1: Control period of ventilation on Flop .3; phase 2: Hypoxic period on FIO~ .lO to .15, establishing pulmonary hypertension; phase 3: Treatment period with NO starting at five parts per million (ppm), doubling dose every 10 min to 80 ppm; phase 4: Posttreatment observation period after discontinuation of NO while maintaining hypoxia for 1 hour or until circulatory failure or pulmonary hypertension of pre-NO magnitude developed. Each animal underwent a maximum of three experiments in random order of hypoxia severity before sacrifice with pentobarbital overdose. Continuous hemodynamic parameters, intermittent cardiac output and pulmonary capillary wedge pressure, and intermittent arterial blood gas analyses were obtained through pulmonary and systemic artery catheters placed by femoral cutdown. Pulmonary and systemic vascular resistances (PVR and SVR) were calculated by standard formulas. Experiments were divided into two groups (n = 8 in each): group 1 with severe hypoxia (Paos, 25 to 35) and group 2 with moderate hypoxia (Paor, 36 to 65). Data for all hemodynamic parameters were expressed as mean percentage change from baseline (phase 1) + SEM under each set of conditions, and the two groups were compared by two-way analysis of variance and covariance adjusted for order of experimentation. The severely hypoxic group showed significantly less improvement than the moderately hypoxic group in mean pulmonary artery pressure during treatment with NO (17 f 3% versus -2 f 3%, P = .005) and after discontinuing the drug (55 f 11% versus 21 f 7%, P = .03). The severely hypoxic group also experienced a significant elevation in cardiac output during NO therapy when compared with the moderately hypoxic group (26 f 6% versus 3 + 4%, P = .02). Furthermore, there were no incidents of circulatory failure after discontinuing NO during moderate hypoxia versus three incidents (38%) during severe hypoxia, all of which were immediately reversed by restarting NO. In a porcine hypoxic pulmonary hypertension model, severe hypoxia predicts weaker improvement in the pulmonary hypertension during NO therapy and stronger recurrence after discontinuing the drug. Nitric oxide may also have a hemody-

Newth,

930

and James Atkinson

namic supportive role during severe hypoxia by enhancing cardiac output and preventing acute car pulmonale. Copyright o 1995 by W.B. Saunders Company INDEX poxia,

WORDS: Pulmonary car pulmonale.

hypertension,

nitric

oxide,

hy-

T

HE STORY of nitric oxide (NO) has been rapidly evolving since Palmer et al1 and Ignarro et al2 independently identified the molecule as endothelium-derived relaxing factor (EDRF) in 1987. Nitric oxide is synthesized from L-arginine in the vascular endothelium through the action of nitric oxide synthase (NOS), an enzyme that exists in constitutive as well as inducible forms.3-7 Once synthesized and released, the lipophilic NO molecule diffuses into vascular smooth muscle cells, where it activates soluble guanylate cyclase, resulting in production of cGMP from GTP. cGMP then initiates a cascade resulting in smooth muscle relaxation.4-6 Exogenous NO in gaseous form has been administered by inhalation to diverse groups of patients whose diseases included a major pulmonary hypertension component.8-20 Because it is quickly bound to hemoglobin and does not enter the systemic circulation, the drug acts as a true, selective, pulmonary vasodilator. It has been particularly advocated for neonates with persistent pulmonary hypertension of the newborn, a patient population that is frequently unable to tolerate the systemic effects of nonspecific vasodilators.11-14JsJ9 However, to our knowledge, no parameters predictive of response to this therapy have been delineated through laboratory or clinical studies. We used a swine hypoxic pulmonary hypertension model to test the following hypothesis: Increased hypoxia severity predicts weaker improvement in pulmonary hypertension during and after inhaled NO therapy. MATERIALS

From the Divisions of Pediattic Surgev, Pediatric Critical Care, and Respiratory Therapy, Childrens Hospital, Los Angeles, CA. Presented at the 1994 Annual Meeting of the Section on Surgery of the American Academy of Pediatrics, Dallas, Texas, October 21-23, 1994. Address reprint requests to James Atkmson, MD, Dwzsion of Pediatric Surgery, Childrens Hospital Los Angeles, 4650 Sunset Blvd. Los Angeles, CA 90027. Copyright o 1995 by W B. Saunders Company 0022.3468/95/3007-0005$03.00/0

in a Porcine

AND METHODS

Animal Preparation Seven Yorkshire swine aged 12 to 16 weeks, weighing 11 to 20 kg, were used. The animals were first sedated with intramuscular atropine (.04 mgikg), xylazine (2 to 5 mg/kg), and ketamine (10 to 1.5 mgikg) before halothane induction and endotracheal intubation. A 12F temperature probe (Hewlett-Packard 21090A, Andover, MA) was inserted in the rectum, a pulse oximeter probe (Nellcor Oxysensor II. Hayward, CA) was placed on the elbow, tongue, or snout, and electrocardiogram limb leads were afhxed. Under continuous halothane anesthesia, femoral cutdowns were JournalofPedmtnc

Surgery,

Vol30,

No 7 (July),

1995:

pp 930-936

HYPOXIA

SEVERITY

PREDICTS

NITRIC

OXIDE

RESPONSE

performed bilaterally. A femoral artery catheter was used for continuous systemic blood pressure monitoring, arterial blood gas analyses, and methemoglobin sampling. A 5 to 7F pulmonary artery catheter (Biosensors, Singapore) was introduced through the femoral vein and final position confirmed by chest radiograph. A cutdown on the contralateral femoral vein allowed for largecaliber intravenous access and maintenance fluid infusion of normal saline at 4 ml/kg/h. Halothane anesthesia was then discontinued, and balanced continuous intravenous anesthesia was started with sodium pentobarbital (5 to 15 mg/kg/h) and pancuromum bromide (.2 mglkgih). The animal was then transferred to either continuous pressure-controlled mandatory conventional ventilation (Servo 9OOC; Siemens-Elema, Solna, Sweden) or highfrequency oscillatory ventilation (Sensormedics 3100; Sensormedits, Anaheim, CA). The protocol was approved by the Childrens Hospital Los Angeles Animal Care Committee.

Expetimental Design The animals underwent 16 experiments, each consisting of four sequential phases. Phase 1, the control phase, lasted 30 minutes and consisted of baseline ventilation (FIOJ = .30 and ventilator parameters set to maintain a Pacoz of 35 to 45 mm Hg). During phase 2, the animal received a hypoxic gas mixture (FIo~ = .lO to .15). This hypoxic insult resulted in a wide range of Paos (25 to 65), but consistently established pulmonary hypertension, defined as a minimum of 30% increase in mean pulmonary artery pressure over baseline. Hypoxia lasted for 30 minutes before treatment and the FIO~ and other ventilator parameters were not changed for the remainder of the experiment. Phase 3 constituted treatment with NO gas (Gilmore Liquid Air, San Gabriel, CA) starting at 5 parts per million (ppm) and doubling the dose every 10 minutes to 80 ppm. Nitric oxide and nitrogen dioxide levels were continuously monitored on line by chemiluminescence (42H; Therm0 Environmental Instruments, Franklin, MA). The treatment was then discontinued and the animal was observed for a maximum of 60 minutes or until circulatory failure (defined as a 35% or more acute drop in cardiac output and mean systemic arterial pressure from the previous measurement) or pulmonary hypertension of pretreatment magnitude developed and persisted for at least 20 minutes. Animals underwent a maximum of three consecutive experiments, in random order of hypoxia severity, separated by 30-minute recovery periods with normoxic ventilation, before being killed with sodium pentobarbital overdose (120 mgikg). The experiments were then divided into two groups of eight. based on severity of hypoxia, as discussed below.

Monitoting During the course of the experiment, systemic arterial pressure (SAP), pulmonary arterial pressure (PAP), central venous pressure (CVP), electrocardtogram, heart rate, and temperature were continuously monitored (Hewlett Packard 78534C) and recorded every 10 minutes (Hewlett Packard 78576A), coinciding with intermittent pulmonary capillary wedge pressure (PCWP) and duplicate thermodilution cardiac output (CO) measurements (9520A Cardiac Output Computer; American Edwards Laboratories, Irvine, CA). During the treatment period (phase 3) data were recorded 10 minutes after each new dose, just before the next one. Pulmonary and systemic vascular resistances (PVR and SVR) were calculated by the following formulas: PVR = (MPAP - PCWP) x 8OiCO; SVR = (MSAP - CVP) x 8O/CO, where MPAP and MSAP are the mean pulmonary and systemic arterial pressures, respectively. Hemoglobin saturation with oxygen (Saol) was continuously monitored (model N-100; Nellcor, Hayward, CA). Arterial blood gases were analyzed at least once during each phase

931

(model 170 pHiBlood Gas Analyzer; Corning, Mayfield, MA). Methemoglobm was measured by co-oximeter at baseline and at the termination of each experiment.

Statistical Methods The outcome variables of interest were MPAP, PVR, MSAP, SVR, and CO. The average of the measurements of each outcome variable taken 10 minutes apart during each phase was used as the data point for that phase. Baseline values (phase 1) between the two groups were compared using two-tailed f test. For phases 2,3, and 4, all outcome variables were expressed as mean percentage change from baseline [{(phase X - phase 1)iphase 1) x 1001 and compared between the two groups by two-way analysis of variance. Because some animals underwent more than one experiment and participated in both experimental groups, an order effect and interaction effect (between order of experiment and hypoxia severity) were tested for by analysis of variance. Where such an effect existed, analysis of covariance was used to adjust for it. Two-tailed t test was used to compare PVR, SVR, and CO outcomes at different phases within the same group. Bonferroni t test was used to compare MPAP and MSAP outcomes at different phases within the same group. Fisher’s exact test was used to compare categorical variables. A P value of 5 .05 was considered to represent statistical significance. All analyses were performed using SAS statistical software (SAS Institute Inc, Cary, NC). RESULTS

After the 16 experiments were completed, a Paoz of 35 mm Hg was chosen as the cutoff between severe and moderate hypoxia, dividing the experiments into two groups of eight. Group 1, severe hypoxia, included all experiments with a Pao? of 25 to 35 mm Hg. Group 2, moderate hypoxia, included all experiments with a Pao, of 36 to 65 mm Hg. Each group was composed of five experiments on conventional ventilation and three on high-frequency oscillatory ventilation. The eight experiments of group 1 were conducted on seven animals, whereas those of group 2 were conducted on five animals. This difference was adjusted for in the statistical analysis. The baseline characteristics of the two groups are shown in Table 1. The two groups were similar in all hemodynamic and respiratory parameters at baseline. Table

1. Baseline

Parameters Group

1

Unlt

Mean -+ SEM

P Value*

25.6 81.5

2 1.5 k 5 2

24.5 r 1.4 84.0t45

NS NS

2 35 +316 2 0.23

404 f 26 2826r357 2.28 2 0.18

NS NS NS

125+8 35.4 t 2.5 98.4 f 0.26 1.1 r 0.21

NS NS NS NS

MPAP

mm Hg

MSAP PVR SVR

mm Hg dyne sec. dyne. sec.

co

litersimin.

407 3081 1.86

Pa02 Pace,

mm Hg mm Hg % % of total

1202 11 38.5 f 2.3 98.0 2 0.5 99 2 0.11

SSO>

MetHgb

cm-5 cm-5

Hgb

Abbrewations: MPAP, mean mean systemic arterial pressure; *P value for group differences

1)

Group

2 In = 8) Mean ? SEM

In = 8)

Parameter

(Phase

pulmonary CO, cardiac by two-tailed

arterial output. ttest.

pressure;

MSAP,

932

EMIL

The differences in hemodynamic outcomes between the severely and moderately hypoxic groups during each phase are shown in Table 2. Cardiac output was not consistently measured during phase 4 (posttreatment observation), and PVR and SVR values are therefore not reported for this phase. Whereas NO succeeded in returning MPAP to baseline during moderate hypoxia, it resulted in large but partial improvement during severe hypoxia, remaining at approximately 17% over baseline. Furthermore, there was a complete recurrence of pulmonary hypertension within 60 minutes of discontinuing NO in the severely hypoxic group versus a sustained therapeutic benefit in the moderate group. Changes in PVR parallel the changes in MPAP and closely approach statistical significance (I’ = .057). Severe hypoxia resulted in a significantly larger drop in SVR, but treatment with NO did not have further effect on these parameters in either group. However, the hypoxia-induced increase in cardiac output was more prominent in the severely hypoxic group and reached statistical significance with NO therapy. Discontinuation of treatment resulted in cardiovascular collapse in three experiments (in three separate animals) in the severely hypoxic group versus none in the moderately hypoxic group. Figure 1 shows the typical course of an animal manifesting this “NO dependence” to maintain cardiac output and systemic blood pressure. In those three experiments, phase 4 MPAP values were always depressed because of a low output state coupled with severely increased PVR, not because of resolution of pulmonary hypertension. These values were therefore excluded from analysis of MPAP after NO in group 1. There was no significant change in Paoz, Paco2, or pH during NO treatment or withdrawal. Finally, there were no significant increases in methemoglobin, with a maximum of 1.1% and 1.4% of total hemoglobin after NO therapy in groups 1 and 2, respectively. Nitrogen dioxide levels did not rise over 4 ppm. Table

2. Changes

in Pulmonary

Severe

66 121 -16 -32 142

(n = 8)

2 a i f

o( 100 I g 8.

< .05 when

compared

with

(n = 8)

51 -+6 98k 11 125 -3 k 6 323

phase

.-

-

.-

6oTi,,,;p”, loo

Severe

2 (hypoxia)

1; _--

1 \\ MSAP

60

I..

‘\-.

--..

group,

--“.:,

7

-,’

I :I

\

MpAp

‘\ :, .. . .

~.

4o I----.,.

B O

*O

.i

4o

.“\k

/’

: /’

“,’

6oTime(;;) loo

140

120

Fig 1. Hemodynamic parameters for an animal manifesting NO dependence during severe hypoxia (Pao,, 29). The increase in mean pulmonary artery pressure (MPAP) and PVR with hypoxia and the therapeutic effect of NO (shaded areas) are evident. (A) On discontinuing NO, the animal experiences an acute and severe drop in CO (41%). inversely related to the acute increase in PVR. (B) The consequence is severe systemic hypotension, with near equalization of systemic and pulmonary arterial pressures. The increase in MPAP is blunted secondary to the low output state. Circulatory failure is immediately reversed by 80 ppm NO. The phenomenon is duplicated once again by discontinuing NO and then reinstituting it at 40 ppm.

DISCUSSION

Our results indicate that NO is more effective in the treatment of pulmonary hypertension during moderate hypoxia versus severe hypoxia. This pertains both to the extent of improvement in elevated pulmonary vascular pressure and resistance during treatment and to the sustenance of the therapeutic benefit after treatment is discontinue{. Hypoxia resulted in an acute decrease in systemic vascular resistance, significantly higher in the severely hypoxic group, that was not exacerbated by NO. Commonly Systemic

(n = 8)

17k33t 12 f 6t -16~9 -36 Y!Z8” 26 + 6*

in the same

140

120

-\,\ -.

.-c e 3 I b E

Note. All data reflect mean percent change from baseline [(phase X - phase l)/phase “P < .05 when compared with the moderate hypoxia group dung the same phase order and interaction effects. tf

4o

Hemodynamics

Phase 3 Hypom + NO Moderate

9 20 5 4, 9

2o

120

and

Phase 2 Hypom

MPAP PVR MSAP SVR co

A O

ET AL

by Bonferroni

Phase 4 Hypom After NO

Moderate

(n = 8)

-2 -7

2 3t c 8t l&6 -3 + 5 324

(ll = 8)

Moderate

55 f 11* -20

(n = 8)

21 e 7t

z!z 16

226 -

11 x 100 f SEM. by two-way analysis t test (MPAP,

Severe

MSAP),

of variance

and covarlance,

or two-tailed

t test (PVR,

adjusted SVR, CO).

for

HYPOXIA

SEVERITY

PREDICTS

NITRIC

OXIDE

RESPONSE

seen in patients, this hypoxic SVR decrease is secondary to vasodilation of systemic vascular beds. An increase in cardiac output, significantly accentuated by NO during severe hypoxia, ensued in order to support systemic blood pressure. The ability of exogenous, inhaled, pharmacological doses of NO to resolve pulmonary hypertension has been shown in several animal models,Z1-30 including models of hypoxic pulmonary vasoconstriction in the newborn piglet,‘3 newborn lamb,29,30 and adult sheep.21J2 Furthermore, inhibition of NOS has been found to augment pulmonary hypertension induced by in vitro hypoxia,31 oleic acid lung injury,32 and sepsis,33 pointing to endogenous NO as an important modulator of pulmonary hypertension in these states. There is evidence that NO-dependent pulmonary vasodilation is distinct from vasodilation induced by hyperoxia and alkalosis, the two major interventions available at this time for selectively dilating the pulmonary vasculature.34 In the last 2 years, multiple case reports and small case series have appeared in the literature citing NO therapy for several cardiopulmonary diseases with a major pulmonary hypertension component.8-20 The drug has created special excitement in neonatology, where it was heralded as a breakthrough in the treatment of persistent pulmonary hypertension of the newborn (PPHN), a drug that would hopefully decrease mortality and the need for extracorporeal membrane oxygenation (ECM0).11-14JsJ9 This excitement has been somewhat tempered by clinical experience showing a wide spectrum of responses among infants with PPHN. One of the major unanswered questions is: Which infants are most likely to benefit from this innovative therapy? Because severity of hypoxia is a major factor in the clinical decisions regarding the course of therapy, we chose to investigate whether this is also a factor in predicting the degree of response to NO. As in patients, a spectrum of responses was seen in our animals. However, two major patterns emerged. When an animal suffered a moderate hypoxic insult, it usually showed a complete resolution of pulmonary hypertension and increased pulmonary vascular resistance during NO treatment, sustaining significant therapeutic benefit after discontinuation of treatment. On the other hand, severe hypoxia was associated with significantly less improvement during NO treatment and a recurrence of the pulmonary hypertension after discontinuing NO. Although this is a model of healthy lungs vasoconstricted by hypoxia, it should be remembered that hypoxic pulmonary vasoconstriction is the common denominator of the vari-

933

ety of pathologies resulting in PPHN and other pulmonary hypertensive diseases. Some data in the literature allude to the observations made here. Previous models of in vivo hypoxic vasoconstriction, using similar degrees of hypoxia to those of our severely hypoxic group, all showed an immediate return to the same magnitude of pulmonary hypertension on discontinuing N0.21-23 In two studies, it seemed that some increase of pulmonary artery pressure persisted during NO therapy (at least at lower dosages). 22,23In a severe pulmonary hypertension model created by infusion of U46619, a thromboxane AZ analog, DeMarco et aP” noted that NO concentrations up to 80 ppm improved, but did not return PVR to normotensive values. Immediately on discontinuing NO, we observed a very predictable, transient rebound in the pulmonary hypertension. This lasted for 2 to 3 minutes and often resulted in even higher pulmonary artery pressures than those observed before initiating NO. After this transient rebound period, pulmonary artery pressures once again normalized (in the moderately hypoxic group) or remained elevated to a lesser degree (in the severely hypoxic group). The relationship of hypoxia to endogenous NO production has received conflicting conclusions. In vascular endothelial cells, decreasing the oxygen concentration is associated with deceased endotheliumderived NO activity.35 However, as mentioned earlier, inhibition of NOS augments hypoxic pulmonary vasoconstriction, suggesting that endogenous production of NO is increased during hypoxic exposure. All these observations can be accommodated by the following theory. Pharmacological doses of NO may inhibit NOS, both in its inducible and constitutive forms, by typical negative feedback regulation. In addition, the inducible enzyme may react differently to different degrees of hypoxia, being induced by moderate hypoxia but inhibited, or at best unaffected, by severe hypoxia. Therefore, during moderate hypoxia, pharmacological doses of NO completely reverse the pulmonary vasoconstriction independent of endogenous NO. On discontinuation of NO, there is a transient rebound in hypertension caused by the lack of enough endogenous NO, but soon after, the enzyme is induced by the moderate degree of hypoxia and the removal of the negative feedback signal. This results in persistent vasodilation. It is known that, once induced, NOS produces NO continuously, resulting in sustained elevation of soluble guanylate cyclase and leading to prolonged smooth muscle relaxation.ll Under severely hypoxic conditions, exogenous NO is unable to result in maximal vasodilation, perhaps

934

EMIL

secondary to saturation of the receptor sites on guanylate cyclase or depletion of cGMP. In addition, on terminating therapy, pulmonary hypertension rapidly recurs as endogenous NO production remains inhibited by the severe hypoxia. This theory can be readily tested in future studies. We believe that this also explains why in some severely hypoxic animals, heart failure ensues immediately on terminating NO therapy (Fig 1). As our data clearly display, the physiological increase in cardiac output during severe hypoxia is further enhanced by NO. This hemodynamic effect is not seen under moderate hypoxia. The right heart, which has become dependent on exogenous NO for afterload reduction, immediately fails during the critical rebound period when endogenous NO is maximally depressed by the additive effects of feedback suppression and severe hypoxia. Reinstituting exogenous NO immediately reverses this situation. Interestingly, case reports have recently appeared where NO was used for this purpose.15J7 The observations made here have important clinical consequences. There is a trend at the moment to treat severely hypoxic children who have qualified for

ET AL

ECMO with NO instead.13J8J9 Although some success has been reported, this approach must remain in question until further studies are done to evaluate the role of NO in treating very severe hypoxia and pulmonary hypertension. Furthermore, it may be wiser to initiate NO therapy at an earlier point in the course of the disease before the onset of very severe hypoxia. Finally, NO may have to be discontinued gradually and with caution when used as a right ventricular afterload reducer. In conclusion, in a swine hypoxic pulmonary hypertension model, severe hypoxia predicts weaker improvement in the hypertension during inhaled NO therapy and stronger recurrence after discontinuing the drug. Nitric oxide augments the increase in cardiac output during severe hypoxia, and may have a hemodynamic supportive role by preventing acute or pulmonale. ACKNOWLEDGMENT The authors are grateful for the statistical assistance offered by Dr Linda Chan and MS Carla Rother of the Biostatistics Program, Childrens Hospital Los Angeles.

REFERENCES 1. Palmer RMJ, Ferrige AG, Moncada S: Nitric oxide release accoupts for the biological activity of endothelium-derived relaxing factor. Nature 327:524-526,1987 2. Ignarro LJ. Buga GM, Wood KS, et al: Endothelium-derived relaxing factor produced and released from artery and vein is nitric oxide. Proc Nat1 Acad Sci U S A 84:9265-9269,1987 3. Palmer RMJ, Ashton DS, Moncada S: Vascular endothelial cells synthesize nitric oxide from L-arginine. Nature 333:664-666, 1988 4. Ignarro L: Biological actions and properties of endotheliumderived nitric oxide formed and released from artery and vein. Circ Res 65:1-Q 1989 5. Moncada S, Palmer RMJ, Higgs EA: Nitric oxide: Physiology, pathophysiology, and pharmacology. Pharmacol Rev 43:109-142, 1991 6. Moncada S, Higgs A: The L-arginine-nitric oxide pathway. N Engl J Med 329:2002-2012,1993 7. Gaston B, Drazen JM, Loscalzo J, et al: The biology of nitrogen oxides in the airways. Am J Respir Crit Care Med 149:538-551,1994 8. Pepke-Zaba J, Higenbottam TW, Dinh-Xuan At, et al: Inhaled nitric oxide as a cause of selective pulmonary vasodilation in pulmonary hypertension. Lancet 338:1173-1174, 1991 9. Girard C, Lehot JJ, Pannetier JC, et al: Inhaled nitric oxide after mitral valve replacement in patients with chronic pulmonary artery hypertension. Anesthesiology 77:880-883,1992 10. Rossaint R, Falke KJ, Lopez F, et al: Inhaled nitric oxide for the adult respiratory distress syndrome. N Engl J Med 328:399-405, 1993 11. Roberts JD, Polaner DM, Lang P, et al: Inhaled nitric oxide in persistent pulmonary hypertension of the newborn. Lancet 340:818-819, 1992 12. Kinsella JP, Neish SR, Shaffer E, et al: Low-dose inhala-

tional nitric oxide in persistent pulmonary hypertension of the newborn. Lancet 340:819-820,1992 13. Kinsella JP, Neish SR, Dunbar-Ivy D, et al: Clinical responses to prolonged treatment of persistent pulmonary hypertension of the newborn with low doses of inhaled nitric oxide. J Pediatr 123:103-1081993 14. Kinsella JP, Toews WH, Henry D, et al: Selective and sustained pulmonary vasodilation with inhalational nitric oxide therapy in a child with idiopathic pulmonary hypertension. J Pediatr 122:803-806, 1993 ‘, 15. Allman KG, Young JD, Stevens JE, et al: Nitric oxide treatment for fulminant pulmonary hypertension. Arch Dis Child 69:449-450, 1993 16. Roberts JD, Lang P, Bigatello LM, et al: Inhaled nitric oxide in congenital heart disease. Circulation 87:447-453,1993 17. Sellden H, Winberg P, Gustafsson LE, et al: Inhalation of nitric oxide reduced pulmonary hypertension after cardiac surgery in a 3.2 kg infant. Anesthesiology 78:577-580,1993 18. Kinsella JP, Abman SH: Efficacy of inhalational nitric oxide therapy in the clinical management of persistent pulmonary hypertension of the newborn. Chest 105:92S-94S, 1994 19. Finer NN, Etches PC, Kamstra B, et al: Inhaled nitric oxide in infants referred for extracorporeal membrane oxygenation: Dose response. J Pediatr 124:302-308,1994 20. Wessel DL, Adatia I, Thompson JE, et al: Delivery and monitoring of inhaled nitric oxide in patients with pulmonary hypertension. Crit Care Med 22:930-938. 1994 21. Frostell C, Fratacci MD, Wain JC, et al: Inhaled nitric oxide: A selective pulmonary vasodilator reversing hypoxic pulmonary vasoconstriction. Circulation 83:2038-2047, 1991 22. Pison U, Lopez FA, Heidelmeyer CF, et al: Inhaled nitric oxide reverses hypoxic pulmonary vasoconstriction without impairing gas exchange. J Appl Physiol74:1287-1292, 1993

HYPOXIA

SEVERITY

PREDICTS

NITRIC

OXIDE

935

RESPONSE

23. Etches PC, Finer NN, Barrington KJ, et al: Nitric oxide reverses acute hypoxic pulmonary hypertension in the newborn piglet. Pediatr Res 35:15-19,1994 24. Butler MW, Lazar EL, Smerling AJ, et al: Differential effects of inhaled nitric oxide on normoxic and hypoxic isolated in situ neonatal pig lungs perfused by extracorporeal membrane oxygenation. J Pediatr Surg 29:275-279, 1994 25. Zayek M, Cleveland D, Morin FC III: Treatment of persistent pulmonary hypertension in the newborn lamb by inhaled nitric oxide. J Pediatr 122:743-750,1993 26. Fratacci MD, Frostell CG, Chen TY, et al: Inhaled nitric oxide: A selective pulmonary vasodilator of heparin-protamine vasoconstriction in sheep. Anesthesiology 75:927-931,199l 27. Shah NS, Nakayama DK, Jacob TD, et al: Efficacy of inhaled nitric oxide in a porcine model of adult respiratory distress syndrome. Arch Surg 129:158-164,1994 28. Berger JI, Gibson RL, Redding GJ, et al: Effect of inhaled nitric oxide during group B streptococcal sepsis in piglets. Am Rev Respir Dis 147:1080-1086,1993 29. Zapol WM, Falke KJ, Hurford WE, et al: Inhaling nitric

oxide: A selective pulmonary vasodilator and bronchodilator. Chest 105:87S-91S, 1994 30. DeMarco V, Skimming J, Ellis TM, et al: Nitric oxide inhalation: Effects on the ovine neonatal pulmonary and systemic circulations. Chest 105:91S-92S, 1994 31. Archer SL, Tolins JP, Raij L, et al: Hypoxic pulmonary vasoconstriction is enhanced by inhibition of the synthesis of an endothelium-derived relaxing factor. Biochem Biophys Res Commun 164:1198-1205,1989 32. Leeman M, De Beyl VZ, Gilbert E, et al: Is nitric oxide released in oleic acid lung injury? J Appl Physiol74:650-654, 1993 33. Robertson FM, Offner PJ, Ciceri DP, et al: Detrimental hemodynamic effects of nitric oxide synthase inhibition in septic shock. Arch Surg 129:149-156,1994 34. Fineman JR, Wong J, Soifer SJ: Hyperoxia and alkalosis produce pulmonary vasodilation independent of endotheliumderived nitric oxide in newborn lambs. Pediatr Res 33:341-346,1993 35. Rodman DM, Yamaguchi T, Hasunuma K, et al: Effects of hypoxia on endothelium-dependent relaxation of rat pulmonary artery. Am J Physiol258:L207-L214,1990

Discussion D. CoZn (Dallas, TX): Dr Emil, I think it’s reasonable to postulate that the degree of hypoxia can be sufficiently severe to blunt the vasodilatory effect of nitric oxide on the pulmonary vasculature. However, the design and the data collection in your experiment are sufficiently flawed in some areas to question the validity of your conclusion about the weaker response in hypoxemia. There are many variables in your experiment that appear not to be adequately controlled. You have seven animaIs in your experimental group, whose weights vary by nearly lOO%, and ages by 25%. There are no mean values given for these parameters. It seems appropriate to question the influence of such large differences in size in these groups in which there are multiple experiments performed for as long as 450 minutes, involving multiple measurements with blood draws. You also let these animals cross over from one group to another, which is a little bit disturbing. In your 16 experiments on these seven animals developing pulmonary hypertension, they were divided into severe and moderate hypoxia based on PO*. The data are not given regarding the distribution in the two groups of the number of experiments on each animal. It is possible that all seven animals could be in one group, plus one of them being there twice, and that only three animals could be in the other group with this crossover in your experiment. There’s also wide variation and no statistical difference in the degree of hypoxia between the two groups. I think to answer your basic question it would be

better to separate groups without the crossover and with adequately and targeted controlled Pozs. I also have a little bit of concern about using the mean percent change from the baseline rather than the difference in the means themselves. There are three instances of circulatory failure in the severe hypoxic group, and none in the moderate group. You have shown, however, that there is tremendous efficacy for the administration of nitric oxide to these animals because of the resolution of their circulatory failure. And I think that’s the main message of your presentation and that’s an excellent result. You have done a lot of good work with the experiment and have given us useful information about the beneficial effect of nitric oxide on the failing heart, but I think the effect on the degree of hypoxia and the response to nitric oxide is still not resolved. S. End (response): Thank you for those comments. The best way I can respond is by expanding on the design of the experiment. I agree that the experiment would have been a better design to have one animal per experiment. Each experiment was separated by a l-hour baseline period where the animal was recovered. When we compared the data at baseline of the two groups of experiments, not the two groups of animals, those data were identical. There was no significant difference. The reason I used percent above baseline is that within adolescent swine there was quite a variation of

936

pulmonary pressure. So what would be pulmonary hypertension for one animal may actually be the baseline for an animal just 2 or 3 kg larger. But again, when I compared the baseline values, those were actually the raw numbers, and they were similar. With respect to the oxygenation point, oxygenation in our experiment was the predictor factor, not the outcome factor. We simply used that as a factor to

EMIL

ET AL

look at what happens to the hemodynamic parameters once we achieved this degree of oxygenation. Seven animals were involved in group 1 and five in group 2. This difference, as well as differences in the order of experimentation and number of experiments performed on each animal, were adjusted for by an extremely thorough statistical analysis, including analysis of covariance.