STUDIES ON ANTICONVULSANT DRUG COMBINATIONS: PHENOBARBITAL AND DIPHENYLHYDANTOIN

¹ Departments of Pharmacology, University of Utah College of Pharmacy and College of Medicine, Salt Lake City, Utah

Dose-effect regression lines for phenobarbital sodium (5-ethyl-5-phenyl barbiturate sodium), diphenylhydantoin sodium (5 ,5-diphenythydantoinate sodium, Dilantin) and seven combinations of these two drugs were determined in rats by the maximal electroshock seizure (M. E. S.) test and minimal neurotoxicity test. The quantity of drug used in each combination was selected from a graph constructed from the regression lines for phenobarbital and Dilantin and in such a manner that the sum of the individual components of effectiveness would equal the 50 per cent level of response. The composition of the seven combinations varied from 3 per cent phenobarbital and 97 per cent Dilantin to 80 per cent phenobarbital and 20 per cent Dilantin. The dose of each drug or drug combination producing the desired endpoint (protection or neurotoxicity) in 50 per cent of animals (ED₅₀ or TD₅₀) was observed and protective indices (P.I. = TD₅₀/ED₅₀) were calculated. The results were statistically analyzed for type of joint action. The combinations were also tested for duration of anticonvulsant action. In addition, experiments were performed to determine the most suitable temporal sequence for giving the components of the drug mixtures and the optimal time at which the combinations should be tested. The conclusions reached may be summarized as follows:

1. The graphic method described for selecting combinations of two active drugs on the basis of components of effectiveness has general applicability to other pharmacological classes of drugs which have similar action and which yield parallel dose-response lines. It may also be used to evaluate the effects of combining more than two drugs with similar action.

2. It was shown that the graphically estimated ED₅₀ for a drug combination is reasonably close to the calculated ED₅₀. This makes the graphically estimated ED₅₀ a valuable basis of reference when large numbers of drug combinations require screening.

3. Data were presented which indicate that a period of coaction of the individual components of a mixture is essential for maximal anticonvulsant activity.

4. Combinations III (phenobarbital 37.5 per cent and Dilantin 62.5 per cent), VI (phenobarbital 67.5 per cent and Dilantin 32.5 per cent), and VII (phenobarbital 80 per cent and Dilantin 20 per cent) were 1.25, 1.72 and 1.35 times, respectively, more potent than calculated on the basis of similar joint action of the component parts of the mixture. The remaining four combinations exhibited no significant alteration in anticonvulsant potency from that predicted.

5. The duration of anticonvulsant action for combinations II (phenobarbital 7.3 per cent and Dilantin 92.7 per cent) and VI was not significantly different from that of the major component.

6. Minimal neurotoxicities of combinations I (phenobarbital 3 per cent and Dilantin 97 per cent) and II were decreased 26 and 23 per cent, respectively, below that calculated on the basis of similar joint toxicity of their component parts, whereas the neurotoxicity of combination V (phenobarbital 62.5 per cent and Dilantin 37.5 per cent) was increased 23 per cent. There was no significant deviation in the toxicity of any of the remaining combinations from that predicted

7. The results obtained were not compatible with the prediction that mixtures of phenobarbital and Dilantin exhibit either similar joint action or independent joint action. However, until additivity on the basis of all possible types of joint action has been similarly excluded, and because of the difficulty of defining such terms, the observed alterations in intensity of individual effects cannot be interpreted with certainty as representing either potentiation or antagonism of anticonvulsant activity.

8. The P. I.'s for combinations VI and VII were 8.80 and 6.72, respectively; the former value represents a significant increase in the margin of safety over that of the most effective component in the mixture. There was no significant alteration in the P. I.'s of the remaining 5 combinations.

The method employed herein for evaluating combinations of anticonvulsant drugs appears to offer a rational approach to the complex problem of combined drug action. The results presented indicate that increased anticonvulsant activity and enlarged margins of safety may be obtained with critical ratios of anticonvulsant drugs, and warrant a systematic evaluation of other anticonvulsant drug combinations.