RATE OF INDUCTION OF HYPOTENSION WITH TRIMETAPHAN MODIFIES THE INTRACRANIAL PRESSURE RESPONSE IN CATS

Br. J. Anaesth. (1988), 60, 161-166

RATE OF INDUCTION OF HYPOTENSION WITH TRIMETAPHAN MODIFIES THE INTRACRANIAL PRESSURE RESPONSE IN CATS

In 1954 Schallek and Walz [1] noted, in dogs, that the more rapidly hypotension was induced with trimetaphan, the more the EEG was depressed. Stullken and Sokoll [2] made the complementary finding that trimetaphan caused significant increases in the intracranial pressure (ICP) in cats. Conversely, Turner and colleagues [3] found a clinically significant increase in ICP in only two of 21 patients to whom trimetaphan was given to induce hypotension. As a result, there was no statistically significant mean increase for the group and the authors suggested that TMP does not increase ICP. Marsh and colleagues [4] reported that the technique by which nitroprusside is administered modified the ICP response in cats. This led us to wonder if different techniques of administration would cause different ICP responses to trimetaphan-induced hypotension also. If so, previously observed individual variability could be accounted for, and contradictory conclusions reconciled. To investigate the effect of technique of administration on ICP, we induced hypotension with trimetaphan in cats with induced intracranial hypertension using two different infusion rates. MATERIALS AND METHODS

SUMMARY An infusion of 0.1 % trimetaphan was administered to eight cats with artificially increased intracranial pressure (ICP) in order to decrease their mean arterial pressure (MAP) from 121 ± 9.5 (SEM) to 58±4.6 mm Hg in less than 1 min. All cats developed an increase in intracranial pressure (ICP) (from 16±1.4 to 23±3.2 mm Hg) accompanied by a partial rebound in MAP. Eight additional cats received 0.1 % trimetaphan to decrease their MAP from 128±13.4 to 52±8.1 mm Hg over more than 2 min. Four of these cats followed the same pattern, with ICP increases from 19±1.1 to 31 ±3.9 mm Hg, while in the other four ICP did not change. In nine of the 12 cats with an ICP increase, that increase was initiated before the partial MAP rebound. We conclude that trimetaphan causes clinically significant ICP increases in cats with increased ICP, that partial rebound in MAP frequently exacerbates these increases in ICP, and that rapid induction of hypotension tends to increase the frequency with which trimetaphan increases ICP.

placed in the supine position, the trachea intubated and the lungs ventilated with 30% oxygen in nitrous oxide. Physiological saline and drugs were infused via the cephalic vein, and arterial pressure monitored via a catheter in a femoral artery. Intracranial pressure was monitored, and adjusted through a 19-gauge needle placed in the cisterna magna and connected by a ANDREW KARLIN, M.D.; JOHN HARTUNG, PH.D. ; JAMES E. COTTRELL, M.D.; Department of Anesthesiology, State Uni- T-fitting to a bag of physiological saline (corrected versity of New York Health Science Center at Brooklyn, 450 to the pH of the cerebrospinal fluid) and to a Clarkson Avenue, Box 6, Brooklyn, New York 11203. transducer [5] (fig. 1). Fluid was added to, or Accepted for Publication: September 21, 1987. withdrawn from, the intracranial space to achieve Correspondence to A.K. After institutional review board approval for animal use had been received, 16 male mongrel cats were anaesthetized with pentobarbitone 32.5 mg kg"1 i.p., with 32.5-mg s.c. supplements as needed to prevent movement. The cat was

Downloaded from http://bja.oxfordjournals.org/ at University of Iowa Libraries/Serials Acquisitions on May 26, 2015

A. KARLIN, J. HARTUNG AND J. E. COTTRELL

BRITISH JOURNAL OF ANAESTHESIA

162 0.1'/. TMP infused to decrease MAP < 1 mln

0.17. TMP infused to decrease MAP > 2 mln

• Chart recorder

FIG. 1. Schematic representation of lay-out of experiment and of variables monitored.

the desired baseline ICP. Needle placement was considered correct when the ICP fluctuated with ventilation. A rectal probe was placed to monitor temperature. Arterial blood-gas tensions and pH were measured: in all experiments pH was 7.35±0.03; PaOi was greater than 10.7 kPa; P3LCOI was 4.7 + 0.3 kPa; and temperature was 37± 1.5 °C. Mean arterial pressure (MAP) was calculated as the diastolic pressure plus one-third of the pulse pressure. An infusion of 0.1 % trimetaphan (Arfonad) in 5% dextrose in water was titrated at one of two rates in order to decrease MAP to approximately 55 mm Hg. A "rapid" rate of induction of hypotension (n — 8) decreased MAP to a minimum in less than 1 min. A " slow " rate of induction (M = 8) brought the MAP to a minimum over more than 2 min. At least 30 min passed between any pentobarbitone administration and an infusion of trimetaphan, and ICP was increased for 30-60 min before each administration of trimetaphan.

All of the eight cats receiving the rapid induction experienced a partial rebound of arterial pressure during the infusion, as did four of the cats receiving the slow induction. For all cats in each treatment group, and for the subset of cats without a rebound in arterial pressure in the slow induction group, MAP, ICP and heart rate (HR) at maximum achieved hypotension were compared with control values by paired two-tail t tests. In the cats in which there was a rebound of arterial pressure, values at maximum hypotension and maximum rebound were compared with control values by paired two-tail t tests if a oneway ANOVA for repeated measures proved significant at P < 0.05. RESULTS

The results are summarized in table I. When hypotension was induced rapidly with a 0.1 % solution of trimetaphan, such that MAP decreased to 58 + 4.6 mm Hg in less than 1 min,

Downloaded from http://bja.oxfordjournals.org/ at University of Iowa Libraries/Serials Acquisitions on May 26, 2015

o

TRIMETAPHAN AND ICP

163

TABLE I. Effects of rapid v. slow induction of hypotension with 0.1% trimetaphan on mean arterial pressure (MAP) (mm Hg), intracranial pressure (ICP) (mm Hg) and heart rate (HR) (beat min'1) (mean ± SEAT). * P < 0.05 compared with control

Withirebound during infusion

All

Control

Hypotension

Control

Hypotension

Without rebound

Rebound

Control

Hypotension

Fast technique 8

8

121+9.5 16±1.4 160+14.5

58 + 4.6* 17 + 6.0 187+10.3*

0

102 + 6.4* 23 + 3.2* 223 + 9.2*

(Same as "all")

Slow technique n

8

MAP ICP HR

4

128+13.4 18+1.1 197 + 19.4

52 + 8.1* 19 + 1.8 192 + 20.5

4

136+14.1 19 + 1.1 209 ±13.1

50 + 8.5* 21 + 1.8 197 + 10.3

112±12.7 31+3.9* 246+10.3

119± 14.5 16 + 0.4 185 + 25.5

56 + 8.8* 17 + 0.7 187 + 29.3

200 n MAP

, 150

x E

100

a. 50-

30

90

120

150

180

210

240

270

300

Time after start of T M P infusion ( s )

FIG. 2. Tracing from one study during which hypotension was induced rapidly.

this decrease was followed by a partial rebound of arterial pressure in all cats. The rebound was transient, occurred during the infusion, and was followed by a decrease approaching 55 mm Hg. Figure 2 shows a typical tracing. In each of these cats there was an increase in ICP, from a control mean of 16 ± 1.14 to 23 + 3.2 mm Hg. The duration of the increase in ICP and the increase in in arterial pressure paralleled each other with a Pearson correlation coefficient between the two durations of 0.98 (P < 0.001). The average duration of the increase in ICP was 324 ± 113 (SEM) s. The average duration of partial rebound in arterial pressure was 301+97 s. Despite the close relation of the durations of the increases in arterial pressure and ICP, the increase in ICP preceded the increase in arterial pressure in six cats. When hypotension was induced more slowly the results were more variable. Induction of hypotension was smooth in four of eight cats, with

MAP decreasing from 119 + 14.5 to 56 + 8.8 mm Hg and ICP remaining essentially unchanged at 17 mm Hg. The other four cats receiving the slow induction followed the same pattern as the cats which received the rapid induction. That is, during the infusion of the trimetaphan MAP decreased initially from 136+14.1 to 50 + 8.5 mm Hg. Shortly thereafter, but still during the infusion, MAP rebounded to 112+ 12.7 and then returned to about 50 mm Hg. The ICP increased roughly in parallel, from 21 + 1.8 to 31+3.9 mm Hg. As the MAP returned toward 50 mm Hg the ICP also decreased toward 17 mm Hg. In three of these four cats the ICP started to increase before the arterial pressure increased. The duration of the rebound in arterial pressure was 121 +24 s. The duration of the rebound in ICP was 122 + 28 s. The Pearson correlation coefficient for the duration of rebound in arterial pressure and the duration of the increase in ICP was 0.92 (P < 0.0001).

Downloaded from http://bja.oxfordjournals.org/ at University of Iowa Libraries/Serials Acquisitions on May 26, 2015

n MAP ICP HR

164

BRITISH JOURNAL OF ANAESTHESIA

DISCUSSION

All eight cats in which hypotension was induced in less than 1 min had significant increases in ICP. Although the effect of the partial rebound in arterial pressure is implicated as a confounding variable, in six cats the increase in ICP began 20 s or more before the rebound in arterial pressure. The increase in ICP before the rebound in arterial pressure may be the result of either autoregulation or direct cerebral vasodilatation, that is, loss of autoregulation. If autoregulation were maintained, we would expect to see an increase in ICP as the cerebral vessels dilated in response to a decreased cerebral blood flow. On the other hand, if trimetaphan had a direct cerebrovascular dilating action which abolished autoregulation, ICP would also increase. If autoregulation remained intact, we should see a decrease in ICP during the rebound in arterial pressure after 0.5-2 min [6]. The fact that ICP remained increased for the entire duration of the rebound (about 5 min) argues that autoregulation was lost and that trimetaphan given to decrease arterial pressure in less than 1 min acts as a cerebrovascular dilator. These findings are in agreement with those of Stullken and Sokoll [2]. They also may corroborate the work by Ishikawa and co-workers [7] in dogs with normal ICP. Although they induced hypotension more slowly than we, they found that ICP doubled, from 6 to 12 mm Hg, within 5 min of starting the infusion of trimetaphan, and returned to control within 20 min. There is other evidence that trimetaphan is a direct vasodilator. Siems and Rottstein [8] demonstrated, in dogs, that tetraethylammonium chloride (TEAC), the parent compound of trimetaphan, had a direct vasodilating action that was at least as important as its effects on the

vasomotor nerves via ganglionic blockade. McCubbin and Page [9] found that the hypotensive action of trimetaphan in dogs was not decreased by section of the spinal cord at C6, removal of the paravertebral sympathetic chains, ganglionic blockade with TEAC or hexamethonium, or the administration of histamine, indicating that trimetaphan-induced hypotension in dogs is independent of sympathetic block or histamine release. The action of trimetaphan on isolated cerebral arteries was examined by Harioka and associates [10], who showed that trimetaphan was a direct arterial vasodilator and alpha-adrenoceptor antagonist. However, the vasodilatation was more prominent in extracerebral arteries than in cerebral arteries. If this finding can be extended to intact animals, perhaps during slow induction of hypotension, this differential action is more prominent, with blood pooling extracerebrally; during more rapid induction a higher concentration of trimetaphan is delivered to the cerebral arteries, causing them to dilate more rapidly. Additionally, trimetaphan may increase intracranial blood volume by blocking sympathetic innervation of cerebral vessels [2]. Although an increase in arterial pressure can cause an increase in ICP when cerebral autoregulation is lost, we may, perhaps, differentiate the effect of trimetaphan on ICP from the effect of the rebound in arterial pressure by looking only at those cats which experienced an ICP increase before the rebound in arterial pressure. Although the rebound in arterial pressure was a common occurrence in our study, its cause was not clear. It is tempting to suppose that, in the observations in which a partial rebound followed an increase in ICP, the partial rebound occurred because of the ICP increase; that is, the rebound may have been a Cushing response. In man the Cushing response is inconsistent [11], but our experience in several other experiments using this model [5, 12-16] suggests that the Cushing response is more uniform in cats: on a few occasions during previous investigations we have inadvertently increased ICP rapidly and on most of those occasions arterial pressure has increased as if to maintain cerebral perfusion pressure. However, direct investigation of the Cushing response in cats is beyond the scope of the present paper. A related question is whether acute increases in ICP itself, affect autoregulation. We have not investigated this, but others, using models similar

Downloaded from http://bja.oxfordjournals.org/ at University of Iowa Libraries/Serials Acquisitions on May 26, 2015

The association between the increase in ICP and the partial rebound in arterial pressure during infusion of trimetaphan was statistically significant (P < 0.05, Fisher Exact). In nine of the 12 cats in which ICP increased during the infusion of trimetaphan (six of eight from the fast group, three of four from the slower group), that increase began before the rebound. The association between the increase in ICP and initiation of change before partial rebound was also significant (crosstabulation of increase v. no increase by prior v. not prior or no rebound, P < 0.05, Fisher Exact).

TRIMETAPHAN AND ICP

165

thesia and Analgesia 1984; 63: 1005-1008.

Downloaded from http://bja.oxfordjournals.org/ at University of Iowa Libraries/Serials Acquisitions on May 26, 2015

to ours, have found no effect on autoregulation ditions which were least likely to cause an increase in ICP in our study. until ICP approaches MAP [17-19]. Evidence in man [3] with intracranial hyperIt should be pointed out that decreases in cerebral perfusion pressure (CPP = MAP-ICP) tension and our evidence from cats with artificially secondary to decreases in MAP can cause loss of increased baseline ICP suggest that trimetaphan autoregulation. In fact, loss of autoregulation can cause clinically significant increases in ICP caused by hypotension with normal ICP occurs at when intracranial hypertension is present. Our a higher CPP than does loss of autoregulation results also suggest that the risk of this increase is caused by increased ICP during normotension reduced if hypotension is induced slowly. [18, 20]. However, as we noted earlier, loss of autoregulation during hypotension should not ACKNOWLEDGEMENT lead to increased ICP. It is also true that brain ischaemia can cause We thank Ms Christine Capuano for indispensable laboratory increased ICP through cytotoxic oedema, fol- assistance, Bartol Matanic, D.V.M., for professional courtesies, lowed by hyperaemia as the arterial pressure and Ms Ellen L. Jackson for her assistance in manuscript recovers, at times aggravated by vasogenic preparation. oedema if MAP increases rapidly to hypertensive REFERENCES values. However, oedema formation and reso1. Schallek W, Walz D. Effects of drug-induced hypotension lution take considerably more time than did the on the electroencephalogram of the dog. Anesthesiology onset and resolution of the increase in ICP in our 1954; 15: 673-680. cats. Accordingly, we think that oedema and 2. Stullken EH, Sokoll MD. Intracranial pressure during hyperaemia secondary to ischaemia were not hypotension and subsequent vasopressor therapy in anesthetized cats. Anesthesiology 1975; 42: 425-431. implicated in our results, and that our observa3. Turner JM, Powell D, Gibson RM, McDowall DG. tions should be attributed to changes in cerebral Intracranial pressure changes in neurosurgical patients blood volume. during hypotension induced with sodium nitroprusside or trimetaphan. British Journal of Anaesthesia 1977; 49: Do our findings have clinical relevance ? Turner 419-425. and colleagues [3] reported that ICP was increased 4. Marsh ML, Aidinis SJ, Naughton BS, Marshall LF, in 10% of their patients during trimetaphanShapiro HM. The technique of nitroprusside admininduced hypotension. Ten percent is a high rate of istration modifies the intracranial pressure response. untoward reactions, but it appears to be much Anesthesiology 1979; 51: 538-541. lower than the response found by Stullken and 5. Cottrell JE, Hartung J, Griffin JP, Shwiry B. Intracranial and hemodynamic changes after succinylcholine adminisSokoll [2] in cats. In fact, we cannot compare tration in cats. Anesthesia and Analgesia 1983; 62: percentages because Stullken and Sokoll reported 1006-1009. group averages, not individual responses. Turner 6. McDowall, DG. Cerebral circulation during induced and colleagues [3] suggested the difference behypotension. In: Enderby GEH, ed. Hypotensive Anesthesia. New York: Churchill Livingstone, 1985; 30-44. tween the two studies was attributable to a species 7. Ishikawa T, Funatsu N, Okamota K, Takeshita H, difference in sympathetic control of cerebral McDowall DG. Cerebral and systemic effects of hypovasomotor tone: "alterations in intracranial blood tension induced by trimetaphan or nitroprusside in dogs. volume produced by sympathetic blockade are Ada Anaesthesiologica Scandinavica 1982; 26: 643. relatively much greater in the cat than in man, 8. Siems LL, Rottstein HS. The locus of the peripheral except when patients with critical intracranial vasodilator action of ethyl alcohol, tetraethylammonium and priscoline. American Journal of Medical Science 1950; compression are studied." However, only six of 220: 649-654. Turner and colleagues' trimetaphan group began McCubbin JW, Page IH. Nature of the hypotensive with an ICP greater than 14 mm Hg. If we take 9. action of a thiophanium derivative (RO 2-2222) in dogs. these six as the clinically relevant sample of Journal of Pharmacology and Experimental Therapeutics 1952; 105: 437-442. patients with critical intracranial compression, then two of six (33 %) had significant increases in 10. Harioka T, Hatano Y, Mori K, Toda N. Trimethaphan is a direct arterial vasodilator and an a-adrenoceptor anICP; that is, as implied by Turner and colleagues tagonist. Anesthesia and Analgesia 1984; 63: 290-296. [3], the response to trimetaphan of different 11. Little N. Head injury. In: Tintinalli JE, Rothstein RJ, species with initially increased ICP may not be so Krome RL, eds. Emergency Medicine. New York: McGraw different. Also, all of the patients of Turner and Hill, 1985; 788-800. colleagues [3] had a slow induction of hypotension 12. Cottrell JE, Hartung J, Giffin JP, Shwiry B. Intracranial pressure during tetrodotoxin-induced hypotension. Anesand a steady reduced arterial pressure—the con-

166

BRITISH JOURNAL OF ANAESTHESIA

Downloaded from http://bja.oxfordjournals.org/ at University of Iowa Libraries/Serials Acquisitions on May 26, 2015

13. Giffin JP, Hartung J, Cottrell JE, Capuano C, Shwiry B. Effects of varied cerebrospinal fluid pressure on cerebral Effect of vecuronium on intracranial pressure, mean blood flow in dogs. Acta Physiologica Scandinavica 1970; arterial pressure and heart rate in cats. British Journal of 79: 262-271. Anaesthesia 1986; 58: 441^143. 18. Grubb RL, Raichle ME, Phelps ME, Ratcheson RA. Effects of increased intracranial pressure on cerebral 14. Giffin JP, Hartung J, Cottrell JE, Litwak B, Capuano C. blood volume, blood flow, and oxygen utilization in Intracranial pressure, mean arterial pressure and heart monkeys. Journal of Neurosurgery 1975; 43: 385-398. n.te after rapid paralysis with atracurium in cats. Canadian Anaesthetists Society Journal 1985; 32: 618-621. 19. Harris AP, Street WM, Jones MD, Koehler RC, Trayst15. Mazzoni P, Giffin JP, Cottrell JE, Hartung J, Capuano C, man RJ. The effect of raised intracranial pressure on Epstein JM. Intracranial pressure and cardiac rhythm regional bloodflowin fetal sheep. Anesthesiology 1985; 63: disturbances during diltiazem-induced hypotension. AnesA354. thesia and Analgesia 1985; 64: 1001-1004. 20. Miller JD, Stanek A, Langfitt TW. Concepts of cerebral 16. Giffin JP, Cottrell JE, Hartung J, Shwiry B. Intracranial perfusion pressure and vascular compression during pressure during nifedipine-induced hypotension. Anesintracranial hypertension. Progress in Brain Research thesia and Analgesia 1983; 62: 1078-1080. 1972; 35: 411-432. 17. Haggendal EJ, Lofgren J, Nilsson NJ, Zwetnow NN.