The effect of ketamine on pulmonary vascular resistance (PVR) is controversial . It is possible that ketamine may cause a more pronounced increase in PVR in patients with preexisting pulmonary hypertension (PH) than in patients with normal pulmonary artery pressure (PAP) [2,3]. In infants and children with normal and increased PVR, the effect of ketamine on PVR was minimal when airway patency and ventilation were maintained . Patients with increased PAP develop changes in the pulmonary arteries, such as endothelial injury, medial hypertrophy of muscular arteries, and new muscularization of normally nonmuscular peripheral arteries . These structurally remodeled pulmonary arteries could respond to ketamine differently than normotensive pulmonary arteries.
In structurally remodeled hypertensive pulmonary arteries from humans with end-stage chronic obstructive lung disease, endothelium-dependent relaxation in response to acetylcholine is extensively impaired [6,7]. In the same arteries, maximal endothelium-independent relaxation in response to sodium nitroprusside (SNP) is not depressed, but the concentration required to produce 50% response of maximal relaxation was higher than control. In animal models of PH, both endothelium-dependent (acetylcholine, adenosine diphosphate [ADP]) and independent relaxation (SNP, isoproterenol) is impaired in isolated pulmonary arterial strips [8-10]. On the other hand, selective inhibition of endothelium-dependent nitric-oxide-mediated relaxation is observed without an effect on SNP relaxation in isolated perfused lung preparations from chronic hypoxic rats . Thus, several vasodilatory drugs are less potent in hypertensive pulmonary arteries than in those that are normotensive. The vasodilatory effect of ketamine has been observed in isolated preparations of rat aorta and portal vein , canine cerebral and mesenteric arteries , and rabbit mesenteric  and femoral arteries . However, there is no information in the literature concerning the direct effects of ketamine on structurally remodeled hypertensive pulmonary arteries. In the present study, we investigated the direct effect of ketamine on isolated pulmonary arteries obtained from normotensive rats and from rats with chronic hypoxic pulmonary hypertension. Because local anesthetics and volatile anesthetics inhibit endothelium-dependent vasodilation, we also studied the effect of ketamine on endothelium-dependent (acetylcholine) and -independent relaxation (SNP) in normotensive and hypertensive pulmonary arteries.
After institutional approval of this animal investigation was obtained, male Wistar rats (SLC, Shizuoka, Japan), weighing between 170 and 200 g, were randomly assigned to one of two groups: those kept in a hypobaric hypoxic chamber (air at 380 mm Hg) for 10 days (n = 24), and age-matched controls (n = 26) housed in room air at normal atmospheric pressure. Rats were housed under a 12-h light-dark cycle and were fed standard rat chow and water ad libitum throughout the duration of hypobaric hypoxia (air at 380 mm Hg). The pressure in the hypobaric chamber was reduced with an electrically driven vacuum pump. The chamber was opened for 15-30 min once a day so that the cages could be cleaned and food and water replenished.
Rats were anesthetized with intraperitoneal pentobarbital (50 mg/kg). The lungs and heart were removed en bloc and placed in modified Krebs-Henseleit solution at room temperature. The composition of this solution was (in mM): NaCl 115.0; KCl 4.7; CaCl2 2.5; MgCl2 1.2; NaHCO3 25.0; KH2 PO4 1.2; and dextrose 10.0. The right ventricle (RV) of the heart was dissected from the left ventricle plus septum (LV + S) and these cardiac portions were weighed separately. Then the value for RV/(LV + S) was calculated to determine whether right ventricular hypertrophy had developed. Two pulmonary artery segments, a left extrapulmonary artery (EPA, 1.4-1.6 mm outside diameter [OD]) and an intrapulmonary artery (IPA, 0.7-1.1 mm outside diameter), were isolated and gently cleaned of fat and connective tissue. Ring segments (2 mm) were cut (1-2 rings from the EPA and 2-4 rings from the IPA) and suspended vertically between hooks in organ baths (20 mL) containing modified Krebs-Henseleit solution; the solution was maintained at 37 degrees C and bubbled with a mixture of 95% air and 5% CO2.
In preliminary experiments, active and resting tension relationships were obtained by increasing the resting tension, using a range of force from 0.1 g to 2.0 g, according to the method of Toda and Hayashi . The optimal resting tension for vasodilation studies was 0.75 g for EPA rings and 0.5 g for IPA rings in control rats, and 1.0 g and 0.75 g in PH rats. At these resting tensions, the peak contraction was obtained in response to 70 mM KCl. In all experiments, changes in isometric force were measured with a force-displacement transducer (TB 612; Nihon Kohden, Tokyo, Japan) connected to a carrier amplifier (AP600G; Nihon Kohden) and were recorded on a pen recorder (MC 6622; Watanabe, Tokyo, Japan).
Arterial rings under the optimal resting tension were washed every 15-20 min and allowed to equilibrate for 120 min. After the equilibration period, 70 mM KCl contraction curves were recorded routinely twice as a measure of maximal contractility with the contraction shown by the second curve considered as the maximal response. A cumulative concentration-response curve to prostaglandin F2 alpha (PGF2 alpha, 10-8 -10-5 M) was obtained, and the approximate concentration of PGF2 alpha needed to produce 50% of the maximal contraction induced by 70 mM KCl was determined. Then the pulmonary artery rings were washed, equilibrated for 60 min, and precontracted with PGF2 alpha (10-6 -3 times 10-6 M) to obtain 50%-70% of the maximal contraction induced by 70 mM KCl. After precontraction with PGF2 alpha, a cumulative concentration-response curve was obtained for ketamine (10-5 -10-3 M), SNP (10-9 -10-5 M), or acetylcholine (10-8 -10-5 M) in the presence of indomethacin (10-6 M). The responses to SNP and acetylcholine were then repeated in the presence and absence of ketamine (10-4 M). Only one drug was tested in each ring. In another set of rats the response to acetylcholine was examined in the presence and absence of a nitric oxide synthase inbititor, nitro-L-arginine (10-4 M, LNA). The results were expressed as percentage of relaxation of the precontraction induced by PGF2 alpha.
The following drugs were used: acetylcholine hydrochloride (Wako Pure Chemical Industries, Osaka, Japan); SNP, indomethacin, ketamine hydrochloride, and LNA (Sigma, St. Louis, MO); and FGF2 alpha (Ono Pharmaceutical Co., Osaka, Japan). The concentrations of the drugs were expressed as the final molar concentrations (M) in the organ bath.
Results were expressed as means +/- SD. Dose-response curves were analyzed by analysis of variance (ANOVA) with repeated measures. If a significant difference was found, comparisons between doses were made by Scheffe's F-test at the 95% significance level. Differences between control and experimental animals were determined by unpaired Student's t-test. When more than two means were compared, one-way ANOVA was used. If a significant difference was found, Scheffe's F-test was used to identify which groups were different. A P value less than 0.05 was considered statistically significant.
Control rats gained weight throughout the experiment. The experimental rats lost weight during the first 3 days of hypobaric hypoxia, but then began to slowly gain weight. In the experimental rats, the RV/LV + S ratio was greater than in the controls, indicating the occurrence of RV hypertrophy Table 1.
In EPA and IPA rings from control rats, ketamine produced relaxation in a concentration-dependent manner Figure 1. Ketamine induced significant relaxation at 10-4 in PH rats and 3 times 10-4 M in control rats (ANOVA with repeated measures). In rings obtained from PH rats, there was a greater relaxation in response to ketamine at 3 times 10-5 -3 times 10-4 M in the EPA and at 10-4 -10-3 M in the IPA than in control rings. We determined the concentrations of ketamine required to produce 50% relaxation of the precontraction induced by PGF (2) alpha (EC50) by normalizing the induced relaxations to a percentage of the precontraction induced by PGF2 alpha and plotting these percentages against the negative logarithmic values of the drug doses. EC50 (-Log[ketamine] M) was lower in PH rings than in control rings (EPA, 3.58 +/- 0.15 [n = 6] in PH rats vs 3.27 +/- 0.09 [n = 9] in control rats [P < 0.01]; IPA, 3.41 +/- 0.20 [n = 11] in PH rats vs 3.11 +/- 0.09 [n = 10] in control rats [P < 0.01]).
In EPA and IPA rings from control rats, acetylcholine Figure 2 and SNP Figure 3 induced relaxation in a concentration-dependent manner. Acetylcholine induced significantly less relaxation at concentrations of 10-7 -10-5 M in both EPA and IPA rings from PH rats compared with control rats Figure 2. In EPA rings from PH rats, the relaxation induced by SNP was markedly reduced at all concentrations Figure 3, while IPA rings from PH rats showed decreased relaxation at 10-7 -10-5 M Figure 3. The nitric oxide synthase inhibitor, nitro-L-arginine (10-4 M), markedly inhibited acetylcholine-induced relaxation in both control and PH rats Figure 4. Pretreatment with ketamine (10-4 M) had no effect on the relaxation response to any concentration of acetylcholine Figure 2 or SNP Figure 3 in either control or PH rats.
Chronic hypoxic PH and monocrotaline-induced pulmonary hypertension in rats have been used as animal models of PH. In both models, structural abnormalities of the pulmonary arteries occur. These include endothelial cell changes [6,17], muscularization of normally nonmuscular peripheral arteries, medial hypertrophy of proximal muscular arteries [6,18], and an increase of intercellular connective tissue proteins  consistent with the vascular structure in PH associated with congenital heart defects . In chronic obstructive lung disease, secondary PH develops, partly due to the prolonged alveolar hypoxia . A previous study  has shown that mean PAP was 17 mm Hg in control rats and 33 mm Hg in rats exposed to hypobaric hypoxia (air at 380 mm Hg for 10 days), conditions used in the present study.
In this study, both endothelium-dependent (acetylcholine) relaxation and -independent relaxation (SNP) was impaired in rings from PH rats, while relaxation in response to ketamine was not impaired. Since acetylcholine produced relaxation in the pulmonary artery rings obtained from control rats, the depressed relaxation of rings obtained from the PH rats was not due to endothelial injury or denudation during vessel preparation. It has been shown that isolated pulmonary artery rings from patients with Eisenmenger's syndrome and chronic obstructive lung disease exhibit diminished endothelium-dependent relaxation in response to acetylcholine [6,7]. These findings were also shown in isolated pulmonary arteries such as rats exposed to chronic hypoxia [9,10] and rats injected with monocrotaline , consistent with the present results.
The data available on response to endothelium-independent dilator (SNP) in PH is inconsistent. Both endothelium-dependent (acetylcholine) and -independent relaxation (SNP) was impaired in isolated pulmonary arteries from chronic hypoxic rats [9,10]. In contrast, only endothelium-dependent relaxation to acetylcholine, but not endothelium-independent relaxation to SNP, was impaired in isolated perfused lung from chronic hypoxic rats  and isolated pulmonary arteries from pulmonary hypertensive calves . Thus relaxation response to SNP in PH may vary depending on the species and preparations studied.
There are two mechanisms by which ketamine prevents increases in intracellular free Ca2+ in vascular smooth muscle cells (i) Ketamine, like organic Ca2+ channel blockers, competes with Ca2+ at Ca2+-binding sites associated with L-type Ca2+ channels [12-15]. (ii) Ketamine reduces phospholipase C activation  and inhibits the agonist-induced synthesis of inositol 1,4,5-triphosphate in vascular smooth muscle cells . In general, inositol 1,4,5-triphosphate induces Ca2+ release from intracellular Ca2+ storage sites, inducing contraction through an increase in myoplasm free Ca2+ concentration. Inhibition of transmembrane calcium-ion influx by ketamine may explain the current results of enhanced relaxation to ketamine in rings from PH rats, since it has been shown that relaxation in response to a calcium-channel blocker was potentiated in hypertensive pulmonary arteries from rats with monocrotaline-induced PH  and hypoxia-induced pulmonary hypertension . Relaxation in response to biogenic amines has been shown to be enhanced in hypertensive pulmonary arteries , a finding that appears to support the present result of enhanced relaxation in response to an amine-type anesthetic, ketamine, in PH rats. Rodman  showed that short-term (3 days) hypoxic exposure induced augumentation of Ca2+ channel-dependent relaxation with a further potentiation after more prolonged 4-wk exposure. We studied the hypertensive rats after only 10 days of hypoxia. It is possible that with an increased time of exposure to hypoxia, the pulmonary hypertensive changes may have had a more prominent effect on ketamine-induced relaxation.
The peak plasma concentration of ketamine was 1.1 times 10-4 after intravenous administration (2.0-2.2 mg/kg) [13,22]; accordingly, we selected the concentration of 1.0 times 10-4 M to test the effect of ketamine on endothelium-dependent and -independent relaxation. Unlike local anesthetics and volatile anesthetics, ketamine, at concentrations used clinically, had no effect on either endothelium-dependent or -independent relaxation in rings from normal and pulmonary hypertensive rats. Ketamine induced significant relaxation at 10-4 M in PH rats and 3 times 10 (-4) M in control rats (ANOVA with repeated measures). The concentration producing significant relaxation in control rings are higher than the peak plasma concentration in clinical setting and of questionable physiologic significance. Although biphasic responses are not uncommon, lower concentration of ketamine did not cause a vasoconstriction.
Ketamine should be administered with caution in patients with preexisting PH . Morray et al.  reported that patients with increased baseline mean PAP had a greater PAP response to ketamine than did patients with normal PAP. They suggested that the thickened medial layer of the hypertensive pulmonary vascular bed has accentuated response to any direct pulmonary vasoconstrictive effects that ketamine might have . On the other hand, changes in PAP due to ketamine would appear to be governed mainly by cardiac output, and not by primary vasoconstriction or vasodilation of the pulmonary vasculature . Thus, several mechanisms may affect hemodynamic responses to ketamine, i.e., sympathetic nervous stimulation, a vagolytic effect, and a direct effect on the vasculature. There are limitations of this study especially with regard to clinical extrapolation of the ketamine-induced relaxation in hypertensive pulmonary artery because the data was obtained in conduit pulmonary arteries and not pulmonary resistance vessels. Although the clinical implications of the present study are speculative, we suggest that mechanisms other than ketamine-induced direct pulmonary vasoconstriction or depressed vasocilation could explain the ketamine induced-increase in PVR in patients with PH.
1. Burrows FA, Klinck JR, Rabinovitch M, et al. Pulmonary hypertension in children: perioperative management. Can Anaesth Soc J 1986;33:606-28.
2. Gassner S, Cohen M, Aygen M, et al. The effect of ketamine on pulmonary artery pressure. Anaesthesia 1974;29:141-6.
3. Morray JP, Lynn AM, Stamm SJ, et al. Hemodynamic effects of ketamine in children with congenital heart disease. Anesth Analg 1984;63:895-9.
4. Hickey PR, Hansen DD, Cramolini GM, et al. Pulmonary and systemic hymodynamic responses to ketamine in infants with normal and elevated pulmonary vascular resistance. Anesthesiology 1985;62:287-93.
5. Rabinovitch M, Keane JF, Norwood WI, et al. Vascular structure in lung tissue obtained at biopsy correlated with pulmonary hemodynamic findings after repair of congenital heart defects. Circulation 1984;69:655-67.
6. Dinh Xuan AT, Higenbottam TW, Clelland CA, et al. Impairment of endothelium-dependent pulmonary artery relaxation in chronic obstructive lung disease. N Engl J Med 1991;324:1539-47.
7. Dinh Xuan At, Higenbottam TW, Clelland CA, et al. Impairment of pulmonary endothelium-dependent relaxation in patients with Eisenmenger's syndrome. Br J Pharmacol 1989;99:9-10.
8. Altiere FJ, Olson JW, Gillespie MN. Altered pulmonary vascular smooth muscle responsiveness in monocrotaline-induced pulmonary hypertension. J Pharmacol Exp Ther 1985;236:390-5.
9. Rodman DM, Chronic hypoxia selectively augments rat pulmonary artery Ca2+
channel-mediated relaxation. Am J Physiol 1992;263:L88-94.
10. Crawley DE, Zhao L, Ciembycz MA, et al. Chronic hypoxia impairs soluble guanylyl cyclase-mediated pulmonary arterial relaxation in the rat. Am J Physiol 1992;263:L325-32.
11. Adnot S, Raffestin B, Eddahibi S, et al. Loss of endothelium-dependent relaxation activity in the pulmonary circulation of rats exposed to chronic hypoxia. J Clin Invest 1991;87:155-62.
12. Altura BM, Altura BT, Carella A. Effects of ketamine on vascular smooth muscle function. Br J Pharmacol 1980;70:257-67.
13. Fukuda S, Murakawa T, Takeshita H, et al. Direct effects of ketamine on isolated canine cerebral and mesenteric arteries. Anesth Analg 1983;62:553-8.
14. Kanmura Y, Kajikuri J, Itoh T, et al. Effects of ketamine on contraction and synthesis of inositol 1,4,5-trisphosphate in smooth muscle of the rabbit mesenteric artery. Anesthesiology 1993;79:571-9.
15. Ratz PH, Callahan PE, Lattanzio FA, et al. Ketamine relaxes rabbit femoral arteries by reducing [Ca2+
] and phospholipase C activity. Eur J Pharmacol 1993;236:433-41.
16. Tota N, Hayashi S. Age-dependent alteration in response of isolated rabbit basilar arteries to vasoactive agents. J Pharmacol Exp Ther 1979;211:716-21.
17. Ilkiw R, Maruyama K, Rabinovitch M. Intralipid effect on normal and hypoxic remodeled rat pulmonary vasculature. Am J Physiol 1988;246:H1499-1508.
18. Maruyama K, Ye C, Rabinovitch M, et al. Chronic hypoxic pulmonary hypertension in rats and increased ealstolytic activity. Am J Physiol 1991;261:H1716-26.
19. Orton EC, Reeves JT, Stenmark KR. Pulmonary vasocilation with structurally altered pulmonary vessels and pulmonary hypertension. J Appl Physiol 1988;65:2459-67.
20. Wanstall JC, O'Donnell SR, Kay CS. Increased relaxation by felodipine on pulmonary artery from rats with monocrotaline-induced pulmonary hypertension does not reflect functional impairment of the endothelium. Pulm Pharmacol 1991;4:60-6.
21. Porcelli RJ, Bergman MJ. Effects of chronic hypoxia on pulmonary artery vascular responses to biogenic amines. J Appl Physiol 1983;55:534-40.
22. Domino EF, Zsigmoid EK, Domino LE, et al. Plasma levels of ketamine and two metabolites in surgical patients using a gas chromatographic mass fragmentographic assay. Anesth Analg 1982;61:87-92.