METABOLIC EFFECTS OF SODIUM NITROPRUSSIDE ON THE PIG KIDNEY IN VIVO: STUDIES BY PHOSPHORUS-31 MAGNETIC RESONANCE SPECTROSCOPY

British Journal of Anaesthesia 1990; 64: 710-716

METABOLIC EFFECTS OF SODIUM NITROPRUSSIDE ON THE PIG KIDNEY IN VIVO: STUDIES BY PHOSPHORUS-31 MAGNETIC RESONANCE SPECTROSCOPY F. PARIVAR, P. B. BARKER, L. W. JONES AND B. D. ROSS

Phosphorus-31 magnetic resonance spectroscopy was applied to the pig kidney in vivo to determine the metabolic effects of infusion of sodium nitroprusside, and of more severe hypotension produced by venesection or halothane. Sodium nitroprusside 7-20 ng kg~' min~' reduced mean systemic arterial pressure (AP) of the pig from 89 to 46 mm Hg. Glomerular filtration rate and total sodium reabsorption were reduced proportionally. Renal metabolism, assessed by the ratio [ATP]: [Pi] (inorganic phosphate), did not change. Adding halothane or venesection to nitroprusside hypotension caused a further reduction in systemic AP to 42 and 39 mm Hg, respectively. Renal [ATP] decreased in direct proportion to mean AP. It is concluded that sodium nitroprusside, even in high doses, does not reduce renal oxygen delivery as assessed by the ability of the kidney to maintain normal A TP content. However, haemorrhage or an increase in halothane dosage results in irreversible damage to the energy producing processes of the kidney. This may be relevant to the clinical use of the drug. KEY WORDS Anaesthetic techniques: hypotension. Complications: renal, adenosine triphosphate. Measurement techniques: magnetic resonance spectroscope. Pharmacology: sodium nitroprusside.

Sodium nitroprusside (SNP) is used to reduce arterial pressure (AP) in surgical procedures [1, 2]. Hypotension reduces glomerular filtration rate (GFR), which is virtually always recoverable [3]. On the other hand, it has been shown that SNP in clinical doses has some protective effect on renal blood flow (RBF) [4, 5]. However, the safe limits

of hypotension for renal metabolism are unclear. It is possible that the degree of hypotension observed clinically with SNP might reduce the effective RBF and therefore result in metabolic changes in the kidney, and that these may not be as readily reversible [6, 7]. Phosphorus-31 magnetic resonance spectroscopy (31P-MRS) has been used extensively to document changes in the kidney of rats in vivo during haemorrhagic hypotension. A significant reduction in ATP content and increases in inorganic phosphate concentration ([Pi]) and intracellular hydrogen ion concentration (acidosis) were observed [6]. The AP at which these changes were first observed was within the range of pressure used clinically with SNP [2]. Our own recent studies, using an isolated perfused pig kidney, showed whole kidney and renal cortical metabolite changes at an AP of 40 mm Hg [8]. These changes may be more extreme because of the non-physiological state of isolated perfused pig kidney. To establish the renal metabolic effects of hypotension caused by SNP in vivo, we monitored AP, renal function and 31P-MRS of the kidney of the pig during infusion of doses of the drug in excess of that used commonly in clinical practice. MATERIALS AND METHODS

We have developed a surgical preparation which permits repeated "P-MR spectra to be obtained for periods of up to 9 h from intact kidney of the FARHAD PARIVAR, M.D., F.R.C.S.; PETER B. BARKER, M.A., D.PHIL.(OXON); LAWRENCE W.JONES, M.D.; BRIAN D . R O S S , F.R.CS, F.R.C.PATH, D.PHIL.(OXON); Huntingdon Medical Re-

search Institutes, 660 South Fair Oaks Avenue, Pasadena, California 91105, U.S.A. Accepted for Publication: December 14, 1989. Correspondence to F. P.

Downloaded from http://bja.oxfordjournals.org/ at RMIT University Library on July 18, 2015

SUMMARY

SODIUM NITROPRUSSIDE AND KIDNEY METABOLISM

Abdominal wall

Kidney dish

Ureter

Artery Ver

FIG. 1. Diagram of the isolated kidney inside the coil. (See methods for details.)

tinuously inside the bore of the magnet to maintain body temperature. With this arrangement, bore temperature was maintained at 32 ± 1 °C. AP was monitored by a Honeywell Simultrace Recorder (Model AR-6). To maintain fluid balance and continuous urine flow, physiological saline was infused i.v. at a mean rate of 1.2 ml min"1 and urine and blood samples were taken at 30-min intervals for measurement of serum electrolyte and creatinine concentrations. After 1 h of normotension, hypotension was induced by administration of SNP in doses of 7-20 ug kg"1 min"1 (mean 12 ^g kg"1 min"1) to obtain maximal hypotensive effects. After infusion of SNP, further hypotension was produced by one of two methods: increase in inspired halothane concentration from 0.5% to 2.0% (n = 5), or venesection by withdrawal of 200-400 ml of blood (n = 3). Hypotension was maintained for a mean of 70 min (30-135 min), after which AP was restored to normal (by autotransfusion in the case of venesection). The period of hypotension was ' different in the 11 pigs in the study because of the different dose response to SNP, halothane and venesection in the animals. 31P-MRS was carried out during normotension, hypotension and recovery. Spectra were recorded for up to 30 min after reversal of hypotension (n = 4). MRS methods All experiments were performed on a CSI-II spectrometer from General Electric Instruments (Fremont, California), equipped with a 4.7 T, 33-cm bore magnet manufactured by Oxford Instruments (Oxford, U.K.). The kidney was placed at the centre of an oval two-turn Helmholz coil, 6.5 x 4 cm diameter, the turns being 3 cm apart. The coil was doubly tuned to 81 MHz for

Downloaded from http://bja.oxfordjournals.org/ at RMIT University Library on July 18, 2015

anaesthetized pig. Eleven mature male pigs (Great White strain, mean weight of 15.2 kg) were studied and allowed free access to water before operation. In eight pigs, 31P-MRS was performed, and in three the dose response of systemic AP to SNP was established, but MRS was not performed. Anaesthesia was induced with ketamine hydrochloride (Ketaset, Bristol Laboratories) 22 mg kg"1 and' xylazine (Rompum, Haver) 2 mg kg"1 i.m. The trachea was intubated with a size 6 Portex tracheal tube and lungs were ventilated with a Harvard ventilator (Harvard Apparatus-Model 613) at a frequency of 16-20 b.p.m. and tidal volume of 200 ml. Anaesthesia was maintained with 0.5-1 % halothane in oxygen 4 litre min"1. Right nephrectomy was performed through a subcostal incision. An arterial cannula was placed in the aorta by direct puncture at or about the level of the renal arteries, to monitor AP and obtain samples at 30-min intervals. A second cannula was placed in the inferior vena cava (by direct puncture) for infusion of fluids and venous blood sampling. With the catheters in place, the incision was closed and the other kidney was exposed through a left subcostal incision and the left ureter cannulated using an 8-French gauge Argyle catheter. The kidney was placed in a 10-cm diameter dish with a notch on the medial aspect to accommodate the renal vessels and the ureter. A 9.5-cm diameter acrylic cylinder (Plexiglass) of 6 cm height was lowered over the kidney so that the kidney lay within the cylinder, and the dish occluded one end. Both cylinder and dish were covered with copper foil, acting as a Faraday shield, to exclude magnetic resonance signals from the adjacent muscle masses. The muscles of the abdominal wall, subcutaneous fat, skin and other extraperitoneal tissues were also excluded from the sensitive volume of the coil by this device. To avoid drying and to prevent shorting of electrical capacitors, the kidney was wrapped in a plastic bag and positioned anatomically inside the radio frequency (RF) coil (fig. 1). Finally, the pig was placed inside a cradle designed to fit the bore of the magnet. The RF coil was fixed to the cradle to prevent respiratory movements of the kidney. The cradle, with the animal in it, was placed in the magnet. Core body temperature was recorded electronically via an intra-oral probe (Digi-Sense, Cole Parmer, model No. 8523-00) and warm air was blown con-

711

BRITISH JOURNAL OF ANAESTHESIA

712

peak-height, referred from the corrected baseline of the Fourier-transformed spectrum. In the absence of phosphocreatine (PCr), pH was determined from the differences in the chemical shift of Pi and y-ATP. Previous authors have demonstrated that the chemical shift of a- and y-ATP are insensitive to pH [12-14], and either resonance may therefore be used as a reference. Good correlation between this and the standard pH assay (a = PCr — Pi) was obtained, where PCr was still observed in the "renal" spectrum. RESULTS

Effects of hypotension on renal physiological function

One hour was allowed for equilibration of AP, urine flow and arterial blood-gas tensions of the pig within the bore of the magnet; thereafter the pigs maintained a constant state for up to 9 h while anaesthetized with halothane. Table I shows AP, GFR and urine flow of the pigs in the present study, and published normal values [15]. In this study mean GFR was within physiological limits. Urine output was higher than published values for male pigs of same age and weight, but the high percentage of sodium reabsorption indicated that tubular function in this preparation was within normal physiological limits. The effect of nitroprusside on systemic AP was highly variable, and the dose required to produce

TABLE I. Renal function in the in vivo pig kidney (mean (SEMj). Hypotension was produced by i.v. infusion of SNP with or without halothane or venesection. In two animals used as controls, AP was unchanged over the period of the entire experiment (108 (3) mm Hg). n = No. animals, except *n = 9-13 individual measurements in 3-4 animals. Statistics: Student's t test compared with control; ns = not significant (P > 0.05). n/a = Not available

Published normal values Normo tension Hypotension SNP P SNP + venesection P SNP + halothane

P Recovery

P

Mean AP (mmHg)

GFR (ml min"1 g"1)

(ml min"1)

Urine

Reabsorption ofNa(%)

TNa (jimol min"1)

64-145

0.31-0.75

0.025-0.125

n/a

n/a

89 (6.5) ( i = 11)

0.41 (0.07) (i = 11)

0.38 (0.08) (i=H)

98.5 (0.28) (i=H)

28 (4.4) (n = 11)

46 (6.9) (n = 7) 0.0002 39 (2.0) (i = 3) 0.03 42 (2.3)

0.21 (0.09) (i = 4)* 0.01 0.04 (0.04) (i = 3)* 0.02 0

0.20 (0.03) (i = 4)* 0.007 0.04 (0.04) (i = 3)* 0.01 0

98.6(0.57) (i = 4) ns 97.9 (0.2) (n = 3) ns 0

5.62 (2.9) (n = 4) 0.003 3.33 (3.33)

0.001 77 (5.7) (i = 6) ns

0.005 0.32(0.18) (i = 6)

0.007 0.31 (0.12)

0 24.33(11) (i = 6)

ns

ns

0.0001 98.5(1.0) ( i = 5) ns

d = 6)

d = 3) 0.01 0

ns

Downloaded from http://bja.oxfordjournals.org/ at RMIT University Library on July 18, 2015

phosphorus-31 and 200 MHz for protons using the tuning circuit described [9]. The coil was used both to transmit and to receive. After shimming to a proton line width of 50 Hz or less, a hydrogen1 image of the kidney was obtained using a standard spin-echo imaging sequence (TR 2 s, TE 40 ms). The image was used to check the position of the kidney within the coil. The coil was then tuned to phosphorus-31 and spectra were recorded at 5-min intervals using the lowing spectral parameters: pulse width 150ns, spectral width 5 kHz, 1 K complex data points, 2.2 s interpulse delay and 128 averages. A digital exponential filter corresponding to a line broadening of 15 Hz was applied before Fourier transformation, followed by a frequency domain baseline deconvolution. Ratios of the metabolites were determined from peak height, since no absolute quantitation of metabolites was attempted. Ratios of peak heights are approximately proportional to the concentration of metabolites if the T1 relaxation times are similar [10]. The variation in Tx between P-ATP and Pi is 0.99-1.08 s in rat kidney [11]. Therefore, using a repetition rate of 2.2 s, the relative saturation of the P-ATP and Pi would be similar. In a subsequent study, absolute quantitation of porcine renal metabolites by means of 31P-MRS was undertaken [paper in preparation]. To overcome the difficulty of overlapping peaks, we have used the computer-determined

SODIUM NITROPRUSSIDE AND KIDNEY METABOLISM

713

SNP 5 pg kg"1 mln"1

0.55 ~

60

80 120 150 180 Elapsed time (min)

210

0.15 240

FIG. 2. Effect of increasing doses of SNP on arterial pressure (AP) and glomerular filtration rate (GFR). Seventy-five minutes was allowed for equilibration of AP and GFR. In this experiment, increasing the dose of SNP had a graded but non-linear effect on AP. Note the well preserved but reduced GFR during 130 min of hypotension.

significant hypotension varied widely. SNP 7.0-20 ng kg"1 min"1 i.v. (the recommended clinical dose is 3-5 jig kg"1 min"1), resulted in a reduction in mean AP of 28-43 % from control values, to 46 (7) mm Hg (P < 0.0002). Increasing doses of SNP caused further hypotension. Figure 2 shows a representative experiment in which the dose response of AP to SNP was established. Doses of SNP in excess of 15 ug kg"1 min"1 i.v. did not result in further hypotension. The dose required to produce hypotension differed from one animal to the next and therefore the unifying point in these experiments was the smallest AP attained and not the dose of SNP. For this reason, data from the 11 pigs in the study shown in table I are pooled. Increasing halothane or venesection reduced AP further, by 9% and 15% (P < 0.05 and 0.07), respectively. On cessation of SNP infusion, significant recovery (P < 0.7 vs control) occurred in the majority of animals. In five of the I1 pigs, AP did not recover, presumably because of prolonged hypotension with a combination of SNP and halothane or venesection. Hypotension produced the anticipated effects on renal function (table I), with significant reductions in GFR, urine flow and total sodium reabsorption (TNa). "P-MRS Figure 3A shows a representative phosphorus31 spectrum of the pig kidney under normotensive conditions. Resonances are identified from chemical shift of (from left to right): phosphomonoester

(PME); inorganic phosphate (Pi); phosphodiester (PDE), including glycerophosphoryl choline (GPC); Y-, a- and P-phosphates of ATP. Tissue pH in this spectrum was 7.41 and the corresponding blood pH was 7.40. The mean [ATP]: [Pi] in this representative experiment was 1.01 ±0.05 and the tissue pH was 7.41 ±0.02. The ratios of metabolites calculated from peak heights were similar to those reported previously for rat kidney in vivo [13, 16] and showed improved overall renal oxygenation for this in vivo pig kidney compared with the isolated perfused pig kidney [8]. Over the course of 3 h, [ATP]: [Pi] did not decrease in two kidneys in which no drug was administered (final [ATP]: [Pi] = 1.01). Administration of SNP did not produce any change in the ratio [ATP]:[Pi], and intra-renal pH remained unchanged (table II), even at a mean AP of 46 mm Hg. However, even a modest further reduction in AP, produced by the addition of halothane or by venesection, resulted in a significant reduction in renal [ATP]:[Pi] (fig. 4, table II). The [ATP]:[Pi] observed was reduced by 25% with halothane and 51% with venesection. Renal [ATP]:[Pi] appeared to be correlated significantly with systemic AP (r = 0.50) below the inflexion point which occurred at 46 ± 7 mm Hg. Recovery of [ATP]: [Pi] was noted in only four of eight animals, and in those was significantly less than the normotensive controls (table II).

Downloaded from http://bja.oxfordjournals.org/ at RMIT University Library on July 18, 2015

30

BRITISH JOURNAL OF ANAESTHESIA

714

TABLE II. Metabolite ratios and tissue pH {mean (SEM)) of in vivo pig kidney during normotension, hypotension and recovery measured by peak heights. Control animals that did not receive any SNP showed no significant decline of [ATP]:[Pi] during the course of the experiment, n = No. animals except *n = 10 individual measurements in two animals. Statistics: Student's t test compared with control; ns = not significant (P > 0.05)

ATP PME

10

Normotension

0

-5

-10

-15

-20

p.p.m.

Hypotension SNP

[ATP]: [Pi]

Tissue pH

0.95 (0.05) (n = 8)

7.41 (0.02)

0.94(0.15) 0.46 (0.32)

7.41 (0.04) (n = 5) ns 7.18(0.21)

(i = 2)*

(i = 2)*

(i=-5)

ns

P SNP + halothane P Recovery 10

-5

-10

-15

"20

P

0.02 0.71 (0.08)

0.02 7.38 (0.02)

0.008 0.52(0.13)

ns 7.46 (0.03)

(i = 4)

(i = 4)

0.03

ns

1.20 1.00080CL ol Orat-

or 10

-5 -10 P.P.ITI.

-15

-20

T

020 . 0

20

40 60 80 100 120 Mean AP (rranHg)

FIG. 4. Effect of mean systemic arterial pressure (AP) on intra-renal [ATP]:[Pi] in the pig. [ATP]:[Pi] of the in situ porcine kidney was monitored continuously by "P-MRS. AP was reduced by sodium nitroprusside (SNP) ( • ) (i = 5), followed by venesection ( x ) (i = 2) or halothane (O) ( i = 5). Bars represent SEM. Kidneys examined in control pigs ( # ) (n = 2) over an equivalent time period showed no significant change in [ATP]: [Pi] compared with normotensive pigs (A) (n = 8) during the experiment. Differences from control: not significant for SNP; P < 0.05 for venesection and for halothane. 10

-5 P.P.m.

-10

-15

-20

FIG. 3. Normal 31 P-MR spectrum of the pig kidney in vivo (A). Resonances from left to right are: phosphomonoester (PME), inorganic phosphate (Pi), phosphodiester (PDE) and three large peaks for y-, a- and [5-ATP, respectively. [ATP]:[Pi] in this spectrum was 1.06. During SNP infusion (B) there was virtually no change in [ATP] and [Pi] ([ATP]: [Pi] = 1.05). Increasing halothane (c) and venesection (D) produced significant reduction of [ATP]: [Pi] in these experiments, to 0.65 and 0.24, respectively.

DISCUSSION

Hypotension produced by SNP results in a reduction in GFR and urine flow which is virtually always recoverable [3]. On the other hand, it has been shown that SNP in clinically recommended doses has some protective effect on RBF [4, 5]; in high doses RBF, cardiac output

Downloaded from http://bja.oxfordjournals.org/ at RMIT University Library on July 18, 2015

P SNP + venesection

(i = 8)

SODIUM NITROPRUSSIDE AND KIDNEY METABOLISM

rection of this systemic acidosis prevented observation of any significant change in intra-renal pH. During clinical use of SNP, maintenance of effective RBF is of utmost importance. In this study, although SNP, even at high doses, did not affect the energy producing processes of renal cells, in combination with halothane or haemorrhage it may result in irreversible damage to these processes. The importance of this reduction in [ATP] for tubular function is not clear, but reduced renal [ATP] has been thought to induce acute renal failure [11]. The advent of practical clinical 31P-MRS may make similar investigations feasible in man. ACKNOWLEDGEMENTS This work was supported by grants from the General Fund of Huntingdon Medical Research Institutes and from the Richard M. Lucas Cancer Foundation. F. P. is grateful to the latter for a Research Fellowship. Electrolyte measurements were performed by Dr P. T. Pinto.

REFERENCES 1. Taylor TH, Styles M, Lamming AJ. Sodium nitroprusside as a hypertensive agent in general anaesthesia. British Journal of Anaesthesia 1979; 42: 859-860. 2. Miller RD (ed). Anaesthesia. New York: Churchill Livingston, 1986; 1954. 3. Behnia R, Signera EB, Brunner EA. Sodium nitroprusside induced hypotension: effect on renal function. Anesthesia and Analgesia 1978; 57: 521-526. 4. Leighton KM, Bruce C, MacLead BA. Sodium nitroprusside-induced hypotension and renal blood flow. Canadian Anaesthetists Society Journal 1977; 24: 637-640. 5. Ohmura A, Wang KC, Pace NL, Johansen RK. Effects of halothane and sodium nitroprusside on renal function and autoregulation. British Journal of Anaesthesia 1982; 54: 103-107. 6. Chan L, Ledingham JGG, Dixon JA, Thulborn KR, Waterton JC, Radda GK, Ross BD. Acute renal failure: A proposed mechanism based upon " P magnetic resonance studies in the rat. In: Eliahou HE, ed. Acute Renal Failure. London: J. Libbey, 1982; 35-41. 7. Chan L, Ledingham JGG, Clarke J, Ross BD. The importance of pH in acute renal failure. In: Eliahou HE, ed. Acute Renal Failure. London: J. Libbey, 1982;58-61. 8. Freeman DM, Barker PB, Parivar F, Benninghoven S, Jones LW, Moress EA, Ross BD. Differentiation between renal cortex and medulla in the response to hypotension using localised "P magnetic resonance spectroscopy. Transplantation 1989; 48: 202-209. 9. Schnall MD, Harihara Subramanian V, Leigh JS jr, Chance B. A new double-tuned probe for current JH and " P N M R . Journal of Magnetic Resonance 1985; 65:

122-129. 10. Gadian DG. Nuclear Magnetic Resonance and Its Applications to Living Systems. Oxford: Clarendon Press, 1982;

Downloaded from http://bja.oxfordjournals.org/ at RMIT University Library on July 18, 2015

and total peripheral resistance are reduced [5] and heart rate is increased [17]. We have also noted an increase in heart rate and a small (6 %) increase in RBF during infusion of SNP (results not presented in full). It is possible that this was responsible for maintaining and preserving [ATP] during SNP-induced hypotension. Another possible explanation is that [ATP] was affected differently in various kidney regions under the influence of SNP. In this study, a 49 % decrease in AP produced by SNP resulted in a 49 % decrease in GFR. This was accompanied by an 80% change in TNa, from 28 to 5.6 (imol min"1. This reduction in TNa may have caused some conservation of renal ATP in the relevant tubular structures—viz. thick ascending limb in the outer medulla. It is postulated, therefore, that [ATP] in this region increased during hypotension, that medullary [ATP] increased by virtue of a decreased demand for sodium transport, and that this off-set any decrease in cortical [ATP] caused by reduced RBF. The net effect is an unchanged [ATP]: [Pi] for the kidney as a whole. Separate analysis of cortical and medullary [ATP] is now feasible in vivo using a technique of spectroscopic localization [8], and this hypothesis is the subject of a current study. The further reduction in AP of 9-15% produced by venesection or increase in dose of halothane had profound effects on renal energy metabolism. TNa was reduced further; this should have resulted in a further increase in [ATP], as it was not used for Na transport. However, we must presume that a critical reduction in RBF occurred at these low values of AP, resulting in a decrease in [ATP] which was not reversible. We have noted that, whilst the functional recovery of the kidney was almost complete, metabolic recovery was not and recovery of [ATP] was always incomplete in those animals that survived. Intra-renal pH is normally a sensitive indicator of renal hypoxia [7]; this was not altered significantly by SNP or halothane in this model. We presume that RBF after SNP and halothane was sufficient to wash out any excess lactate formed. Systemic lactic acidosis, normally observed with excess doses of SNP [1], was constantly corrected in these studies by intermittent administration of sodium bicarbonate and by ventilation, maintaining arterial pH close to 7.4, and arterial Pco, at 4-5.3 kPa. It is possible therefore, that cor-

715

716 11. Freeman DM, Bartlett S, FUdda GK, Ross BD. Energetics of sodium transport in the kidney: saturation transfer " P N M R . Biochimica Biophysica Ada 1983; 762: 325336. 12. Moon RB, Richards JH. Determination of intra-cellular pH by " P magnetic resonance. Journal of Biological Chemistry 1973; 248: 7276-7278. 13. Freeman DM, Chan L, Yahaya H, Holloway P, Ross BD. Magnetic resonance spectroscopy for the determination of renal metabolic rate in vivo. Kidney International 1986; 30: 35-42. 14. Malloy CR, Cunningham CC, Radda GK. The metabolic state of the rat liver in vivo measured by 31P NMR

BRITISH JOURNAL OF ANAESTHESIA spectroscopy. Biochimica Biophysica Acta 1986; 885: 1-11. 15. Airman PL, Dittmer DS. Biology Data Book, 2nd Edn. Bethesda: Federation of American Societies for Experimental Biology, 1974; 1722. 16. Shapiro JI, Chan L. P-31 Nuclear magnetic resonance spectroscopic study of obstructive uropathy in the rat. Journal of Climcal Investigation 1987; 80: 1422-1427. 17. Norlen K. Central and regional haemodynamicj during controlled hypotension produced by adenosine, sodium nitroprusside and nitroglycerin. British Journal of Anaesthesia 1988; 61: 186-193.

Downloaded from http://bja.oxfordjournals.org/ at RMIT University Library on July 18, 2015