Effects of radiographic contrast media on the ... - BIR Publications

T he British Journal of Radiology, 70 (1997), 1229–1238 © 1997 The British Institute of Radiology

Effects of radiographic contrast media on the tension of isolated small pulmonary arteries 1Y X WANG, PhD, 1C J EMERY, PhD, 1E LAUDE, PhD and 2S K MORCOS, FRCS, FRCR 1Department of Experimental Medicine, Sheffield University Medical School, Sheffield S10 2JF and 2Department of Diagnostic Imaging, Northern General Hospital NHS Trust, Sheffield S5 7AU, UK Abstract. The aim of the study was to establish the direct effects of radiographic contrast media (RCM) on the tension of isolated small pulmonary arteries and to investigate any mediation by nitric oxide (NO) and endothelin (ET). Small pulmonary arteries (0.3–0.6 mm in diameter) from male Wistar rats were mounted in a Cambustion vessel myograph and vessel wall tension recorded. The effects of 10, 20, 40, 80, 150, 200 and 250 mgI ml−1 of diatrizoate, ioxaglate, iopromide and iotrolan and their mannitol osmolar control from basal condition, and when the vessels were preconstricted with prostaglandin F2a (PGF2a) either submaximally (10 mM) or maximally (100 mM), were studied. The constrictor response to diatrizoate (40 mgI ml−1) was tested in the presence of non-selective endothelin receptor antagonist (10 mM SB209670). The dilator response to ioxaglate (80 mgI ml−1) was tested in the presence of L-nitroarginine methyl ester (L-NAME, 100 mM). All RCM caused biphasic changes in tension, a small transient fall (dilatation) followed by a sustained rise (constriction). Mannitol caused constriction only. The potency order of constrictions at 10–40 mgI ml−1 was diatrizoate>iopromide>ioxaglate>iotrolan. When the vessels were preconstricted with PGF2a, RCM caused predominantly dilatation; ioxaglate produced the largest effect (−42.1±3.1%, n=12). Mannitol caused constriction only. SB209607 had no effect on the constrictor effect of diatrizoate [41.9±2.3 alone, 42.1±2.7 with SB209670, n=10]. L-NAME had no effect on the dilator response to ioxaglate [−38.2±1.6 alone, −43.6±2.2 with L-NAME, n= 8]. It is tempting to postulate that dimeric RCM may cause the least changes in the pulmonary circulation during angiography.

Introduction Administration of radiographic contrast media (RCM) into the pulmonary circulation, either for clinical purposes or under experimental conditions, causes an acute increase in pulmonary artery pressure (Ppa) [1–6]. This sudden rise in Ppa is thought to contribute to the morbidity and mortality associated with pulmonary angiography, particularly in patients with pulmonary hypertension [7, 8]. The pathophysiology of the increase in Ppa is not fully understood but it has been attributed to arterial vasoconstriction, venous spasm, nerve reflexes and increased resistance at the level of pulmonary capillaries caused by alterations in the red blood cells rheology [9–11]. RCM, including those with low osmolarity, can increase the rigidity and aggregation of the red blood cells thus impeding blood flow through the capillaries [10, 11]. RCM can also modulate the production of vasoactive agents that affect pulmonary haemodynamics, including histamine, bradykinin, serotonin, prostaglandins, antidiuretic hormone (ADH), atrionatriuretic peptide (ANP), Received 13 June 1997 and in revised form 23 July 1997, accepted 30 July 1997. Address correspondence to Dr S K Morcos. T he British Journal of Radiology, December 1997

angiotensin II and renin [12–16]. They may also influence the synthesis of the newly recognized biological mediators nitric oxide (NO) and endothelin (ET) which exert important effects on pulmonary vascular resistance [17, 18]. However, the importance of these substances in mediating the pulmonary haemodynamic effects of RCM is not known. It has been shown that RCM can induce a decrease rather than an increase in the pulmonary vascular resistance (PVR) and that observed increases in Ppa can be attributed to a rise in the cardiac output and alterations in blood rheology [2, 3]. RCM cause vasodilatation in most systemic vascular beds, except in the kidney where vasoconstriction predominates [19–21]. To our knowledge, no studies have investigated the direct effects of RCM on the pulmonary vasculature. Pulmonary arteries have comparatively little muscle, and thus possess low intrinsic tone. The response to vasoactive substances may be different for muscular systemic arteries and pulmonary arteries. Indeed, large conduit muscular pulmonary arteries behave differently to small poorly muscularized peripheral arteries [22]. We studied the direct effects of agents from the four main types of iodinated water soluble RCM; 1229

Y X Wang, C J Emery, E L aude et al

high osmolar ionic monomer, low osmolar nonionic monomer, low osmolar ionic dimer and isoosmolar non-ionic dimer, on the tension of isolated small peripheral pulmonary arteries of the rat using the tension myograph [22, 23]. We also investigated the possible role of the endogenous vasoactive mediators, NO and ET, in mediating the effects of RCM on pulmonary arteries.

Methods Small pulmonary artery ring preparations Male Wistar rats (250–380 g) were anaesthetized with sodium pentobarbitone (60 mg kg−1, ip) and the lungs removed. Small pulmonary arteries (0.3–0.6 mm internal diameter) were dissected and mounted as ring preparations on a Cambustion biological automated small vessel myograph (Model AM10, Cambustion Ltd, Cambridge, UK) [22, 23]. Dissection was performed using a binocular operating microscope. The bronchial tree was dissected open along its course and then gently lifted and dissected free from the artery beneath. Lung tissue surrounding the remainder of the artery was then dissected free, being careful not to touch the artery. Four arteries were dissected out from each rat and were cut to produce four rings. Two inelastic stainless steel wires with a diameter of 40 mm were passed through the lumen of the artery in order to mount the vessels in the organ bath. The first wire was secured to a movable jaw connected to a micrometer and the second wire secured to a fixed jaw connected to a force transducer to measure tension (Figure 1). Two vessels were mounted per bath each of which contained 5 ml of physiological salt solution (PSS), was maintained at 37 °C and continuously bubbled with a gas mixture of 95% O and 5% CO . 2 2 After an equilibration period of 1 h, the vessels were normalized, i.e. stretching of the artery to

Figure 1. The vessel is mounted on two inelastic steel wires (40 mm diameter) and attached to the two steel jaws within the organ bath containing 5 ml physiological salt solution (PSS). One wire is screwed to a movable jaw connected to a micrometer and the other is screwed to a fixed jaw connected to a transducer for measurement of vessel tension. 1230

achieve maximum developed tension at minimum passive tension [22, 23]. The internal circumference characteristics of the blood vessel under different passive transmural pressures were determined with the help of computer software [22, 23]. Normalization allowed the myograph software to set an equivalent transmural pressure of 17.5 mmHg which imitated the stretch on the arterial wall at normal in vivo pulmonary arterial blood pressure by applying tension on the vessel wall without excessive stretching [23]. The vessel rings were then rested for a further hour, after which a reproducible maximal contractile response (two consecutive tests giving rises in tension within 10% difference) to a standard concentration of potassium chloride (KCl, 80 mM) was obtained. Vessels were washed between tests at least three times with PSS and allowed to rest at base line tension for 10 min before the next test. After the final KCl constriction the vessels were allowed to rest for half an hour before further testing, changing the PSS every 15 min. Changes in vessel tensions were recorded continuously onto a computer disk, and expressed both graphically and numerically on a computer monitor. The tension was measured in milliNewton (mN) per millimetre diameter of the blood vessel.

Experimental protocol Dose response to RCM from basal conditions Artery rings (n=12 per group), bathed in 5 ml of PSS, were exposed to varying concentrations (10, 20, 40, 80, 150, 200, 250 mgI ml−1) of either diatrizoate ( high osmolar ionic monomer), ioxaglate ( low osmolar ionic dimer), iopromide ( low osmolar non-ionic monomer), iotrolan (isoosmolar non-ionic dimer), or equivolume mannitol control solutions, for 20 min from basal conditions (PSS alone). The osmolality of the mannitol solutions was equal to that of RCM with the exception of diatrizoate which has a higher osmolality than a saturated solution of mannitol prepared at room temperature. Preliminary experiments had shown that alterations in vascular tension induced by RCM became almost stable within 20 min. Four rings were obtained per rat and two vessels were mounted per bath. One bath was exposed to RCM and the other to mannitol control solutions. Six of the 12 vessels were exposed to RCM or mannitol in ascending concentration order and the other six in descending order. After establishing the effects of each test solution, the vessels were rinsed with PSS three times and allowed to rest at baseline tension for at least 15 min before exposure to the next solution. The capacity of the organ bath allowed the neat commercial preparations of RCM to be added to give iodine concentrations up to T he British Journal of Radiology, December 1997

Contrast media and pulmonary arteries

80 mgI ml−1. For higher concentrations, it was necessary to replace the organ bath fluid with solutions prepared by mixing PSS with either RCM or equivolume of mannitol control solution. The osmolality of the diluted RCM solutions was measured using an osmometer (Advanced Instruments Inc., Massachusetts, USA). Osmolality above 1300 mosmol kg−1 H O, could not be accu2 rately measured by this instrument. EVect of RCM after pre-constriction with prostaglandin F2a (PGF2a) Vessels pre-constricted with either 10 mM of PGF2a (submaximal constriction, approximately 40% K-Emax) or with 100 mM PGF2a which produces a maximal constriction (approximately 70% K-Emax), were exposed to either RCM (80 mgI ml−1) or equivolume mannitol control solution (n=12 per group). The dose of 80 mgI ml−1 was selected as being vasoactive without causing a significant dilution of the PSS of the bath solution. Each vessel was exposed to all RCM in a random order. EVect of endothelin (ETA/B) receptor blockade The constrictor response to diatrizoate (40 mgI ml−1), which produced the maximum vasoconstriction, and its equivolume mannitol control were tested from basal conditions in the presence and absence of a non-selective endothelin (ET)A/B receptor antagonist (10 mM SB209670). EVect of nitric oxide synthase (NOS) blockade on the dilator response The effect of L-nitroarginine methyl ester [L-NAME] (100 mM), a nitric oxide synthase inhibitor, was tested on the dilatation to ioxaglate 80 mgI ml−1, the most potent dilator. Vessels were pre-constricted with 100 mM PGF2a. As L-NAME caused further constriction, the dose of PGF2a was reduced to 7.5 mM in order to achieve a tension similar to that produced by 100 mM PGF2a alone. The dilator response to 80 mgI ml−1 ioxaglate was tested in the presence of 100 mM PGF2a alone and then, after washing with PSS and allowing the blood vessel to rest for 1 h, tested against 7.5 mM PGF2a+100 mM L-NAME. The dose of L-NAME is known to produce a maximum inhibition of NO synthesis under the employed experimental conditions [24]. In a control group the reproducibility of the vasodilatory response to ioxaglate 80 mgI ml−1 from a pre-constricted state (100 mM PGF2a) was assessed.

Assessment of viability The vasoreactivity to KCl (80 mM), as a test of vessel viability was repeated at the end of all T he British Journal of Radiology, December 1997

experiments. The preparation was considered viable if the constrictor response was within 10% of the response at the beginning of the experiment. Only results from viable preparations were included. In protocols 3 and 4, only observations from preparations with an intact endothelium were included in the results analysis. This was assessed by establishing a vasodilator response, after preconstriction with 100 mM PGF2a, to the endothelial dependent dilator acetylcholine (1 mM) at the beginning of each experiment.

Materials The physiological salt solution (300 mosmol kg−1 H O) was prepared fresh each day. 2 The composition of PSS in 1 l of deionized water was: 120 mmol NaCl; 4.7 mmol KCl; 1.17 mmol MgSO .7H O; 25 mmol NaHCO ; 1.18 mmol 4 2 3 KH PO ; 26.9 mmol EDTA; 2.5 mmol 2 4 CaCl .2H O; 5.5 mmol glucose. The mannitol 2 2 osmolar control solutions were prepared by dissolving mannitol in deionized water to match the RCM osmolality. As it was not possible to supersaturate mannitol in order to achieve the high osmolarity of diatrizoate 370 (2070 mosmol kg−1 H O), a saturated solution of mannitol at room 2 temperature was used as the control of diatrizoate 370. The chemicals used were obtained from the following sources: diatrizoate (Urografin 370 mgI ml−1, 2070 mosmol kg−1 H O, Schering 2 AG, Berlin, Germany); ioxaglate (Hexabrix 320 mgI ml−1, 580 mosmol kg−1 H O, May & 2 Baker Ltd, Dagenham, England); iopromide (Ultravist 300 mgI ml−1, 610 mosmol kg−1 H O, 2 Schering AG, Berlin, Germany); iotrolan (Isovist 300 mgI ml−1, 320 mosmol kg−1 H O, Schering 2 AG, Berlin, Germany); D-Mannitol (Sigma Chemical Co., St Louis, USA); PGF2a (Prostin F2a, Upjohn, Crawley, UK); endothelin 1 (Sigma, UK); SB209670 was kindly donated by Dr Ohlestine of Smith Kline Beecham, USA; L-nitroarginine methyl ester (L-NAME) (Sigma, UK); acetylcholine (Sigma, UK).

Analysis of results Rises in tension (constriction) are expressed as a percentage of K-Emax (% K-Emax=rise in tension caused by the tested drug (mN mm−1 ×100/rise in tension caused by KCl 80 mM (mN mm−1). The measurements were taken at the end of the observation period (20 min) when the constrictor response became stable. When vessels were pre-constricted by PGF2a, the vasorelaxation caused by the RCM was expressed as a percentage fall from the raised tension. The maximum fall in 1231

Y X Wang, C J Emery, E L aude et al

tension was used to measure the dilatory response. All results are presented as mean±standard error of the mean (SEM). An unpaired t-test or a one way analysis of variance (ANOVA) was used for comparisons between the groups. A p-value of less than 0.05 was considered to be significant.

Results Dose response to RCM from basal conditions All four RCM (n=12 per group) produced a biphasic response, a small transient dilatation lasting approximately 2 min, followed by a vasoconstriction which was maintained over the rest of the observation period (20 min) (Figure 2a). Constriction was observed only with mannitol control. Constriction with RCM was dependent on type and dose (Figure 3). Potency, as defined by the dose producing the maximum constriction (Cmax), was in the order: diatrizoate (40 mgI ml−1) >ioxaglate (150 mgI ml−1)>iotrolan (200 mgI ml−1) ( p