DIFFERENTIAL NERVE BLOCKADE: ESTERS V. AMIDES AND THE INFLUENCE OF pk a

Transcription

1 Br. J. Anaesth. (198), 59, DIFFERENTIAL NERVE BLOCKADE: ESTERS V. AMIDES AND THE INFLUENCE OF pk a J. A. W. WILDSMITH, A. J. GISSEN, B. TAKMAN AND B. G. COVINO All the clinically useful local anaesthetic drugs have a common basic molecular structure: an aromatic ring is connected to an amine group by an intermediate chain. The latter includes either an amide or an ester bond, and side chains may be added to any part of the molecule. The differences in drug profile which these variations in molecular structure produce may be related to their effects on physico-chemical properties notably lipid solubility and pk a. Absolute potency, for example, increases with lipid solubility (Covino, 1980). A study of the differential nerve blocking activity of a series of analogues of the ester-linked agent, procaine, showed (using an in vitro rabbit vagus nerve preparation) that an agent's lipid solubility affected also the relative rates at which different types of nerve fibre were blocked (Wildsmith et al., 1985). The small, slowly conducting C fibres were blocked at much the same rate by equipotent concentrations of all agents. The large, rapidly conducting A fibres were blocked at a faster rate than the C fibres by an agent of high lipid solubility, but more slowly than the Cfibresby an agent of low lipid solubility. The present study was undertaken to determine whether the same was true of a series of amide-linked agents, and to allow comparison of the in vitro activities of the two groups of drugs. MATERIALS AND METHODS Experiments were performed under exactly the same conditions using the same apparatus as in the J. A. W. WILDSMITH,* M.D., F.F.A.R.C.S.; A. J. GISSEN, MJJ.J B. TAXMAN, PH.D.; B. G. COVINO, M.D., PH.D.; Department of Anaesthesia, Harvard Medical School, Brigham and Women's Hospital, Boston, MA, U.S.A. Accepted for Publication: JuJy 24, 198. *Present address for correspondence: Department of Anaesthetics, The Royal Infirmary, Edinburgh EH3 9YW. SUMMARY The in vitro sensitivities to local anaesthetic blockade of A, B and C fibres in rabbit vagus nerves were examined using a series of structurally similar amide agents, which varied in lipid solubility and anaesthetic potency. The actions of these drugs were compared with one another, and with those of a series of amino-ester local anaesthetics studied previously. The results demonstrated that A fibres are the most, and C fibres the least, sensitive to blockade by local anaesthetic agents. As with the ester drugs, equipotent concentrations of the amides blocked C fibres at approximately the same rate, but the absolute and relative rates of development of A fibre blockade were related to lipid solubility. As the latter increased, so did the rate of A fibre blockade. Combining the results of the two studies suggests that an agent of low lipid solubility and high pk, might be used to produce differential C fibre blockade. Comparison of the results obtained with the two different classes of drug indicates that the ester structure may have an inherently more potent action than the amide. previous study of ester-linked agents. Full details of methodology have been published (Wildsmith et al., 1985); only a brief summary will be given here. The cervical vagus nerves of 2-kg albino rabbits were removed, desheathed and mounted in an airtight nerve chamber at room temperature. The apparatus allowed stimulation of the nerve and recording of the resultant compound action potentials for A, B and C fibres. A 1-cm section of the nerve was exposed to a submaximal blocking concentration of a drug (dissolved in a carbonated- Liley solution at ph.4 ±0.02) until the changes in the height of the action potentials were stable.

2 380 BRITISH JOURNAL OF ANAESTHESIA Drug Lignocaine Etidocaine W 301 Procalnamlde Procaino Structure CH3 o CjH (os-n-c-c-ic _.Q<3 0 H CH3 CiHc Dt3 H 2N^O)~ C "' t " <Clt 2 ) 2 H '>. Chloroprocaine HJl(O)-C-<HBU -H Amethocalne dl-mapa «<OVc-0-{CHjljfl 12 N CH 3 P*a...4 C'logP FIG. 1. Chemical formulae of the drugs used, their pjc m values and the calculated indices of their base forms (Clog P). The drugs below the dotted line were examined in the previous study. Photographs of the action potentials, as displayed on oscilloscope screens, were taken onto polaroid film to allow measurement of the degree of blockade (denned as percentage decrease in the height of the action potential) at appropriate intervals. Dose-response relationships were constructed by plotting the maximal effect on each fibre type against drug concentration using logarithmic-probit graph paper. In addition, linear plots of the changes in the amplitudes of the compound action potential with time (rate of block) were constructed. The drugs studied were the hydrochloride salts of lignocaine, etidocaine and W301 (fig. 1). Linear regressions of the log of drug concentration against probit score were derived for the effect of each drug on each fibre type and solved for the point of 50 % depression of the action potential amplitude (ED 0 ). The standard error (SE) of that value was also obtained. The relationship between potency (ED 0 ) and lipid solubility was examined by linear regression analysis on a log-log plot. Clog P, a calculated octanol/water partition coefficient for the base form of the drug (Hansch and Leo, 199), was used as the index of lipid solubility and was plotted against the concentration of the base form of the drug which would be present at its ED 0 at ph.4. This was derived using the formula: ED M [base] = ED 0 [total] 1 + lph. pjr Statistical analyses were performed using analysis of covariance. RESULTS Figure 2 illustrates the dose response relationships for the three drugs on A, B and C fibres drawn on a log-probit plot. Table I shows the ED 0 for each drug on each fibre type, together with the SE and the number of experiments used to obtain these data. For each drug, the ED 0 for A fibres was less than that for B fibres, which was lower than that for C fibres. The slopes of the dose response relationship were steeper for A than B, and steeper for B than C fibres. Figure 3 shows the percent changes in the compound action potentials for A and C fibres plotted against time in two individual investigations involving approximately equipotent concentrations of etidocaine and lignocaine. Etidocaine blocked A fibres more rapidly than C fibres in this, and every other, investigation. Lignocaine blocked A and C fibres at much the same rate. In the example shown, the initial rate of C fibre blockade was very slightly faster than A, but the reverse was true in other investigations. In most studies with W301, the initial rate of C TABLE I. ED W (mmol litre' 1 ) (and its standard error) far the effect of each drug used on each fibre type, n = Number of experiments Fibre type Etidocaine Lignocaine W301 A n B n C n 0.019(0.004) 0.028(0.010) 0.054(0.024) 0.108(0.039) 0.15(0.0) 0.2(0.13) 0.5(0.111) 0.95(0.29) (0.8)

3 DIFFERENTIAL NERVE BLOCKADE is ro & «0 E ro o ~ "in ^ m S 40 gg 30 IS A fibres Bfibres C fibres Etldocain* Lignocaine W Concentration (mmol litre' 1 ) o 5.0 Z Q_ FIG. 2. Dose-response relationships for the effect of three amide local anaesthetics (lignocaine and two analogues) on three fibre types. Desheathed rabbit vagus in vitro; carbonated Liley solution; 23-2 C; ph.4. fibre blockade was faster than A (see figure 5), but this was not a consistent finding. Figure 4 shows the plot of Clog P against ED M [base] for Afibreson a log scale. Data for both the amides studied here, and for the procaine analogues studied previously, are included. 0.5 n Procainamide s»w Etidocarne \,y Lignocaine 0.3 mmol litre' Duration of drug application (min) 0.00O5-I C'logP FIG. 3. Percent decreases with time in the compound action potentials for A and C fibres when exposed to etidocaine FIG. 4. Relationship between ED u [Base] and calculated lipid 0.03 mmol litre" 1 and lignocaine 0.3 mmol litre" 1. solubility.

4 382 BRITISH JOURNAL OF ANAESTHESIA Analysis of covariance showed no significant difference in the slopes of the fitted lines for esters and amides, and a model with parallel lines was, therefore, fitted. This showed a significant difference in the intercepts of the two lines (P < 0.01), indicating that esters are more potent than amides at any given value of lipid solubility. Analysis of covariance was also used to test whether the figure for procainamide differed significantly from the relationship for amides (P < 0.95) or esters (P < 0.1). This confirms the impression from the figure that procainamide fits the amide line well, but the ester line poorly. Similar plots may be obtained using the ED 0 figures for B and C fibres. DISCUSSION The results of this study confirm previous work (Gissen, Covino and Gregus, 1980) that the in vitro sensitivity to local anaesthetic blockade of the fibres in the rabbit vagus nerve is in the order A > B > C (fig. 2, table I) once stable equilibrium blockade has been established. The results also indicate that the potency/lipid solubility of amide local anaesthetic drugs had a similar effect on the rate of development of in vitro differential fibre blockade, as it did with the ester-linked agents (Wildsmith et al., 1985). The very potent agent, etidocaine, clearly blocked A fibres at a faster rate than C fibres. This suggests that such a highly lipid soluble agent was able to penetrate easily the very considerable lipoprotein coverings around A fibres to demonstrate rapidly their greater basic sensitivity to blockade. The other two amide agents blocked both types offibreat approximately the same rate. Because of their lower lipid solubility, they would have taken longer than etidocaine to penetrate the A fibre diffusion barriers, so that the onset of blockade of those fibres was delayed somewhat. C fibres are surrounded by little in the way of diffusion barriers and, consequently, they were blocked by equipotent concentrations of each drug at much the same rate. As with the ester agents, the absolute potency of the amides correlated well with lipid solubility. However, comparison of the results of the two studies (fig. 4) has suggested that there is a difference in basic potency between the two types of drug. Calculations based on the 95 % confidence Umits of the regression lines for the two types of drug indicate that an ester will have an ED M between one-half and one-sixth that of an amide of equal lipid solubility. In fact, it is often believed that the amide drugs are the more potent, but this is the result of the differences in pk a between the two groups (fig. 1). At physiological ph, much greater proportions of the un-ionized, lipid soluble form of an amide will be present in solution than for a similar ester. Unless the concentrations of the base form (the species of the drug that actually diffuses into the axonal membrane) are calculated, the amides will appear to be more potent. This difference in potency between esters and amides has been noted before. In his original work on the amide drugs, Lofgren (1948) related the potency of a large series of lignocaine analogues to lipid solubility. Although there was a very good correlation, procaine did not fit the relationship. Similarly, Courtney (1980) found a good correlation between the lipid solubility of a series of amide drugs and their frequency-dependent blocking characteristics, but neither amethocaine or procaine fitted this relationship. In both studies the points for the ester drugs lay "below and to the left" of the amide regression line, just as they do in figure 4. Lofgren (1948) also pointed out that the slope of the regression between potency and lipid solubility was less than unity (0.88 in figure 4). He concluded that the Meyer-Overton rule does not apply to local anaesthetics so that, while lipid solubility is important, it is not the only factor affecting their potency. He considered that, if two agents had the same lipid solubility but one was more potent than the other, then it had a more "specific" action. Possibly, there is something about the ester structure that allows a closer interaction with the local anaesthetic receptor than an amide. Although the local anaesthetic receptor has yet to be clearly located, let alone characterized, much is known about the structure activity relationships of local anaesthetic drugs and two possible explanations for the greater potency of the esters have been put forward (Courtney and Strichartz, 198). First, the amide bond is much "bulkier" and may physically restrict the access of the drug molecule to its binding site. Second, the o-methyl groups present on the aromatic ring of lignocaine and its analogues (fig. 1) may also interfere with binding to a receptor by causing the adjacent carbonyl oxygen to project away from the molecule. Procainamide contains an amide link-

5 DIFFERENTIAL NERVE BLOCKADE 383 age, but no o-methyl groups (fig. 1). Since it fits the regression for the amides very closely (fig. 4), this would suggest that it is indeed the " bulkiness " of the amide bond which is responsible for the lower potency of this group of agents. The results for procainamide are interesting in another way. On the plot of ED 50 [base] against Clog P (fig. 4), its point lies very close to that for W301. However, a marked difference was noted between them in the rate of development of differential nerve blockade. This is illustrated in figure 5, which shows the onset of A and C fibre blockade with approximately equipotent concentrations of the two drugs. The explanation for this would again seem to be related to pk &. At physiological ph, approximately 50% of W301 would be in the un-ionized, lipid soluble base form, but only 1 % of procainamide. This would allow more rapid penetration of W301 through the lipid diffusion barriers around A fibres. After the earlier study of procaine analogues, we felt that the rate of development of differential nerve blockade was not related to pk a, but rather to lipid solubility. The compounds examined had similar dissociation constants, but very different lipid solubilities and the effect of the latter probably obscured any effect of p*t a. The rate of penetration of a drug through lipid membranes will depend on its pk a, as well as its lipid solubility, since the former will affect the amount present in the lipid permeant form. The above analysis is very dependent on the use 100 r W301 1 mmol litrē r- Procainamide 4.0 mmol litre* 1» A fibres C fibres Duration of drug application (h) FIG. 5. Percent decreases with time in the compound action potentials for A and C fibres when exposed to W mmol litre"' and procainamide 4.0 mmol litre" 1. of Clog P as an index of lipid solubility. Ideally, potencies should be related to measured partition coefficients, but this would require figures that had been obtained under constant laboratory conditions. Such data are not available for all the compounds studied here. When the figures that are available are compared, considerable variation is apparent, not just in absolute terms but, occasionally, even in relative values. This is because different groups of workers have examined only a few compounds and have used different analytical methods and experimental conditions. There is a great need for definitive measurements, but as a substitute the calculation of Clog P is relatively straightforward and the method is internally self-consistent, so that using the figures to compare compounds should be a valid exercise. Perhaps the most extensive data on partition coefficients were obtained by Lofgren (1948) and, as mentioned above, his results gave the same relative difference in potency between procaine and the amides as is shown in figure 4. CONCLUSION This study has examined the differential nerve blocking activity of three amide-linked local anaesthetics. The results have been compared with those obtained in a similar study of a series of ester-linked agents and procainamide. There is a major difference between the clinical situation and the conditions under which these studies were performed, but the implication of the results is that C fibres would be blocked differentially by an agent of low lipid solubility and high p/c a. A compound with these properties might produce blockade of C fibres relatively quickly, but in vivo would be removed by the circulation before it could penetrate the greater diffusion barriers around A fibres. If it is assumed that pain is mediated by C fibres and motor function by A fibres, the type of blockade outlined above would produce analgesia without motor paralysis a desirable aim in several clinical situations. This simple definition of fibre function is almost certainly an oversimplification, but it is a useful working hypothesis. Using cocaine, an agent that has a relatively low lipid solubility and a high pat a, Bahar and Rosen (193) were able to manipulate its concentration until they produced analgesia without motor paralysis after subarachnoid injection in rats.

6 384 BRITISH JOURNAL OF ANAESTHESIA ACKNOWLEDGEMENTS We would like to thank Astra Pharmaceuticals Inc., Worcester, Mass, for donating the compounds used in the study; Dr M. S. Tute of Pfizer Central Research, Sandwich, Kent for performing the calculations of Clog P; and Dr R. Elton, Medical Computing and Statistics Unit, University of Edinburgh for help with statistical analysis of the results. REFERENCES Bahar, M., and Rosen, M. (1983). Cocaine, lignocaine and bupivacaine for spinal analgesia without muscle paralysis in the rat. Br. J. Anaesth,, 55, 243. Courtney, K. R. (1980). Structure-activity relations for frequency-dependent sodium channel block in nerve by local anesthetics. J. Pharmacol. Exp. Ther., 213, 114. Strichartz, G. R. (198). Structural elements which determine local anesthetic activity; in Handbook of Experimental Pharmacology, Volume 8: Local Anesthetics, Ch. 3 (ed. G. R. Strichartz). Heidelberg: Springer-Verlag. Covino, B. G. (1980). The mechanisms of local anaesthesia; in Topical Reviews in Anaesthesia, Volume 1 Ch. 2 (eds J. Norman and J. G. Whitwam). Bristol: John Wright & Sons. Gissen, A. J., Covino, B. G., and Gregus, J. (1980). Differential sensitivity of mammalian nerves to local anesthetic drugs. Anesthesiology, 53, 4. Hansch, C, and Leo, A. (199). Substituent Constants for Correlation in Chemistry and Biology. New York: Wiley-Intertcience. Lofgren, N. (1948). Studies on local anesthetics. Xylocaine a new synthetic drug. Ph.D. thesis, University of Stockholm. Reprinted Worcester, Ma: The Morin Press. Wildsmith, J. A. W., Gissen, A. J., Gregus, J., and Covino, B. G. (1985). Differential nerve blocking activity of amino-ester local anaesthetics. Br. J. Anaesth., 5, 12.