Evaluation of amantadine hydrochloride in the treatment of A2 influenzal disease*

Bull. Org. mond. Sante 1969, 41, 671-676 Bull. Wld Hith Org. Evaluation of Amantadine Hydrochloride in the Treatment of A2 Influenzal Disease* RICHARD B. HORNICK, M.D., YASUSHI TOGO, M.D., SARA MAHLER, M.D. & DOMENIC IEZZONI, M.D. Amantadine hydrochloride is the first drug to show promise as a practical anti-influenza agent. Several studies demonstrating a prophylactic effect in volunteers with induced A2 disease as well as in patients during A2 outbreaks have created the impetus for therapeutic trials. The widespread A2 influenza epidemic that occurred early in 1968 provided the opportunity for therapeutic evaluation in the USA. A total of 197 prison inmates with proven influenza agreed to participate in the 10-day double-blind evaluation of amantadine (100 mg, twice daily). Onset of therapy was approximately 20 hours after first subjective awareness of illness. Assessment of drug effectiveness was based on rapidity of resolution of illness. There was a significant increase in the number of drug-treated as against placebo patients in the " rapid resolver " group whereas individuals receivingplacebo dominated the " slow resolver " group. Analysis of febrile responses indicated that amantadine-treated patients had significantly more rapid defervescence. Virus isolation studies revealed etiologically related virus in 90 % of all volunteers during the first 5 days of therapy. It was apparent that clinical improvement was not correlated with disappearance of the virus. Nevertheless, the trials conducted during the influenza season of early 1968 indicated the therapeutic effect of amantadine hydrochloride. Administration of 100 mg, twice daily, for 10 days did not cause any adverse effects. The therapeutic utilization of antibiotics has resulted in dramatic relief from the ravages of bacte- rial infections and we look forward to similar drugs for the control of viral infections. Few viral infec- tions presently prevalent in the USA are life- threatening; those previously dreaded, such as poliomyelitis and the childhood exanthemata (not necessarily benign), have been contained through the use of effective vaccines. This approach may be impracticable with the very common respiratory viral diseases which are initiated by a constellation of individual viral agents, although immunoprophylaxis may be useful for selected agents that cause some of the more serious infections. Unfortunately, vaccines developed against the influenza virus have several shortcomings and have not yet provided ultimate protection. It is against this array of respiratory viruses, responsible for significant morbidity, that a chemotherapeutic or chemoprophylactic drug would * From the University of Maryland School of Medicine, Baltimore, Md., and E. I. duPont de Nemours Company, Wilmington, Del., USA. be useful, and it is the purpose of this paper to summarize the data indicating a therapeutic effect of one drug, amantadine hydrochloride, in the manage- ment of infections caused by influenza viruses of the A2 variety. Promise of the potential usefulness of this antiviral agent was first entertained from results in tissue culture, eggs and mice some years ago (Davies et al., 1964; Grunert, McGahen & Davies, 1965). The information from these studies, plus many additional experiments, has been valuable in assessing whether an effect will be attained in man (lezzoni, D. G., unpublished data, 1969). Thus, there does appear to be a correlation between the sensitivity of particular influenza strains to amantadine, as demonstrated in eggs, tissue culture or mice, and the subsequent effectiveness of the drug in man. The prophylactic use of amantadine in volunteers with induced A2-strain infections demonstrated that the incidence and severity of illness were reduced and that sero- logical evidence of infection was impaired in drug- treated individuals. These studies implied that the 2422 671- R. B. HORNICK AND OTHERS drug was protecting susceptible cells from the virus. Amantadine does not have a viricidal action but prevents viral penetration through the cell membrane (Hoffman et al., 1965). The presence of the drug prior to infection could thus protect many cells, hence reducing the potential for viral replication and thereby aborting the infection and hindering the production of antibodies. Whether drug given after the infection had started could impede the progres- sion of disease remained to be elucidated. Early experiments conducted by Stanley et al. (1965) were designed to test the efficacy of drug given 1/2 hour prior to, or 3 hours after, virus inoculation. The attenuated vaccine strain of A2 virus employed in these studies was not inhibited in man when amanta- dine was administered at approximately the time of challenge. In contrast, when the drug was given 20 hours prior to challenge, it was extremely effective in preventing infection. Additional studies, conducted in volunteer males with low or absent titres of homologous circulating antibody to a virulent chal- lenge strain of A2 virus, confirmed the drug's prophylactic usefulness (Togo, Hornick & Dawkins, 1968). These encouraging results prompted further therapeutic trials to test the action of the drug once infection had already become manifest. EXPERIMENTAL METHODS Early in January 1968, A2 influenza was docu- mented in several areas of the USA. Following a uniform protocol, studies were conducted by various groups of investigators in Virginia, Maryland, Texas and Missouri in prisoners ill with an influenza-like illness. A small number of patients from a Masonic home in St. Louis, Missouri, were also enrolled. Febrile men who had been ill less than 48 hours were asked to volunteer. Treatment was assigned in a double-blind fashion and each volunteer received either 100 mg amantadine or lactose, as a placebo, twice a day for 10 days. Physicians admitted patients to a hospital ward where they were kept until discharge. Temperature recordings were obtained every 4 hours and findings were recorded on check sheets listing common signs and symptoms of influenza. Examinations were conducted twice daily, or at least once a day after the acute illness had abated. In each study, serum specimens were obtained on admission and on day 21. Throat and nasal swabs for virus isolation were obtained in all studies, at least twice during the first 3 hospital days. In the 2 Texas prisons we made a concentrated effort to determine the effect of amantadine on virus shedding and swabs were obtained daily for 5 days. RESULTS The analysis of the results of the combined studies involved mainly clinical responses, temperature, antibody response and virus shedding. Participants in the study comprised a drug and a placebo group that were comparable as to age, previous vaccination history and base-line antibody titres. Only individuals with influenza confirmed either by 4-fold or greater rise in antibody titre or by repeated virus isolation were included. Table 1 outlines the areas where TABLE I MEAN TIME OF STARTING MEDICATION AFTER ONSET OF INFLUENZA A2 ILLNESS Amantadine Placebo groups groups Time No. Time No.(h) N. (h) No Walls, Texas 24.2 23 22.0 20 Wynne, Texas 23.7 17 21.8 17 Jessup, Maryland 19.6 12 22.7 15 Richmond, Virginia 15.6 21 15.0 28 Missouri State 20.7 12 24.9 8 Masonic home (Mo.) 12.4 9 8.1 15 Total 94 103 these studies were conducted and also the time interval between the onset of illness and initiation of drug treatment. The figures represent only those men with confirmed influenza infection. Note the comparable time intervals for placebo and amanta- dine within the population groups. The groups were similar regarding mean temperature at the time of admission to the study. The greatest mean variation between the placebo and amantadine groups was 0.7 deg F (0.39 deg C). Clinical responses The subsequent clinical response of these groups has been analysed to determine whether amantadine had a beneficial effect. Studies by 2 of us (R. B. H. and Y. T.) in the 2 Texas prisons and the Maryland 672 AMANTADINE HYDROCHLORIDE IN THE TREATMENT OF A2 INFLUENZA CRITERIA FOR CLASSIFICATION OF CLINICAL RESPONSE FOLLOWING THE INITIATION OF THERAPY AND THE DISTRIBUTION OF PATIENTS AMONG THESE CLASSES TEMP. DROP TO 100°F OR WITHIN CLEARING OF SYMPTOMS TO 50% OR LESS WITHIN 100 so so OF 4a PATIENTS 20 0 NUMBER OF PATIENTS Amantadine, 100 mg, twice daily prison convinced us that post-treatment illness fell mainly into 2 categories: one in which there was a rapid clinical improvement and a second in which the response was delayed. Criteria were therefore established to categorize the majority of these cases and to account for those falling in between the 2 groups. The accompanying figure lists the criteria for rapid, medium and slow resolvers and shows the number of individuals from the 6 clinical trials in each group. There was a statistically significant difference between the rapid resolver group and the slow resolvers, favouring the effect of amantadine: 51 % of the patients receiving this drug had a rapid response compared with 13.6% of the placebo group. Temperature responses Analysis of fever, the objective parameter of this clinical complex, was carried out in several ways. When the symptoms were not used as criteria in the preceding analysis no alteration in results was apparent. The drug-treated group was found to have a significantly greater incidence ofmen who had fever of 100°F (37.8°C) or more, for 24 hours or less (52% versus 17 Y.). Those men with temperatures of Placebo, twice daily 100°F or higher that persisted beyond 24 hours were mainly in the placebo group (48% with placebo versus 22% with drug). Duration of temperature of 99°F (37.2°C) or higher from onset of illness was compared in each group and the results are shown in Table 2. The expected 3-day period of fever due to influenzal disease was apparent in the control group. Treatment with amantadine appeared to shorten the temperature course by about 1 day. The differences between TABLE 2 DURATION (h) OF FEVER OF >990F FROM ONSET OF INFLUENZA ILLNESS Amantadine Control P-value Richmond prison 60.9 80.1 <0.05 Wynne prison 49.8 82.1 <0.02 Walls prison 65.1 88.3 <0.01 Jessup prison 66.0 92.0 <0.05 23 *1- I _nw __ S i , *|u_ I 673 R. B. HORNICK AND OTHERS amantadine and placebo groups were significantly different in the 4 prisons analysed using 99°F as the base-line. When 100°F was established as the base-line for fever, however, the results were less clear-cut (Table 3). The trend in all 4 was for less fever in the TABLE 3 DURATION (h) OF FEVER >100°F FROM ONSET OF INFLUENZA ILLNESS Amantadine Control Pgroup group P-value Richmond prison 41.7 63.3 <0.02 Wynne prison 38.5 68.8 <0.01 Walls prison 53.5 68.4 ns Jessup prison 48.0 ns amantadine groups but the results were only signi- ficant in 1 of the Texas prisons and in the Virginia prison. Additional confirmatory evidence demonstrating the effectiveness of amantadine in shortening the febrile response in influenza was obtained by compar- ing moving group averages of body temperature. Table 4 outlines the time intervals at each location where the drug-treated groups had significantly less fever than the placebo groups. In the Jessup prison group only at 1 period late in the trial were the TABLE 4 TIMES OF SIGNIFICANT DIFFERENCES BETWEEN MOVING GROUP AVERAGES OF BODY TEMPERATURE DURING THE 5 DAYS FROM START OF MEDICATION Study units Periods (h) of significant Walls prison 33-90 Wynne prison 0-85 Jessup prison 80 Richmond prison 24-90 Missouri prison 28-108 Masonic home 17-70 a Between the drug-dosed and placebo-dosed groups for subjects with serologically confirmed influenza, e.g., in Walls prison, temperatures were significantly different in the period from 33 hours after the start of medication until 90 hours after the start. group averages significantly different. The reasons for this exception to the trend in the other study locations are unknown. Symptoms and signs Symptoms and signs were recorded twice daily during the early phases of each study. In a prisoner population such subjective reactions are at times difficult to interpret and quantitation of symptoms is frequently difficult in any population group. It has been shown that the duration of symptoms following influenzal disease is often a function of the personality involved and not of the illness itself (Cluff, Canter & Imboden, 1966). Nevertheless, when symptoms and signs were analysed in the individual groups, differ- ences were seen. As shown previously, a 50% reduction in symptom score was incorporated in the criteria for clinical response-rapid, medium and slow resolution. In these prison populations the trend was for recorded signs and symptoms to persist for a shorter period of time in the amantadine- treated than in the placebo-treated groups. In each location a few signs or symptoms were found to be significantly more persistent in the placebo group. Antibody response The effect of prophylactic amantadine on antibody response has been to reduce the number ofindividuals with 4-fold or greater antibody responses and also to reduce the geometric mean neutralizing antibody titre. In the latter instance this was assumed to be due to reduction in antigenic mass or perhaps prevention of viral multiplication which is apparently needed for antibody stimulation. In the therapeutic studies included in this report no significant differ- ences in antibody measurements were noted between the placebo- and amantadine-treated groups. It would appear that by the time treatment was started and became effective, antibody stimulation had already begun and subsequently was not significantly altered. Stanley et al. (1965) have shown the beginning of antibody rise within 3 days of challenge in their volunteer studies: this was usually 48 hours after the onset of illness. Assuming that naturally acquired disease has a similar incubation period, i.e., approximately 24 hours, and that amantadine therapy began to hasten recovery after another 24-hour period, it appears plausible to assume that antibody synthesis begins during these early days of infection and is not influenced by amantadine. The antigenic stimulation for these antibodies is obviously the virus. One might guess, since no 674 AMANTADINE HYDROCHLORIDE IN THE TREATMENT OF A2 INFLUENZA TABLE 5 PERSISTENCE OF A2 INFLUENZA VIRUS IN THE INMATES OF WALLS AND WYNNE PRISONS a Amantadine-treated group Placebo-treated group No. positive/no. tested No. positive/no. tested Day no. (and % positive) (and % positive) Rapid Medium Slow Total Rapid Medium Slow Total| resolvers resolvers resolvers resolvers resolvers resolvers 1 19/25 4/5 10/10 33/40 4/6 10/12 16/19 30/37 (Before treatment) (69 %) (80 %) (100 %) (83 %) (67 %) (83 %) (84 %) (81 %) 2 22/25 5/5 9/10 36/40 6/6 10/12 12/19 28/37 (85 %) (100 %) (90 %) (90 %) (100 %) (83 %) (63 %) (76 %) 3 23/25 5/5 10/10 38/40 5/6 12/12 16/19 33/37 (85 %) (100 %) (100 %) (95 %) (83 %) (100 %) (84 %) (89 %) 4 24/25 5/5 9/10 38/40 6/6 11/12 16/19 33/37 (92 %) (100 %) (90 %) (95 %) (100 %) (92 %) (84 %) (89 %) 5 23/23 4/5 9/10 36/38 6/6 11/11 16/19 33/36 (100 %) (80 %) (90 %) (95 %) (100 %) (100 %) (84 %) (92 %) Total 92/98 19/20 37/40 148/158 23/24 44/47 60/76 127/147 (93%) (95%) (93%) (93%) (96%) (94%) (79%) (86%) a Presence of virus determined with throat and nasal swabs. difference in antibody synthesis occurred between the 2 groups, that the antigen persisted in both: such indeed was the case. Table 5 illustrates this point. The data in this table were obtained in the Walls and Wynne prisons in Texas. Virus recoveries were attempted during the first 5 days of participation in the study and the table shows that amantadine treatment did not eliminate A2 influenza virus from the pharyngeal area. Secondly, virus disappearance was not associated with recovery from the disease, since controls as well as amantadine-treated patients were well by day 4 and most by day 3. The duration of viral carriage was not determined beyond 5 days but no evidence of lessening incidence of virus isolation by this day was evident: this suggested prolonged persistence. DISCUSSION The results summarized here demonstrate a thera- peutic efficacy of amantadine. Previous studies proved a prophylactic effect. However, several questions remain to be answered. Of prime interest is the mechanism of therapeutic action. Amantadine does not have antipyretic or antihistamine activities. Perhaps the prevention of viral spread from the initial foci in the superficial layers of the respiratory epithelium, by prevention of cell penetration, allows the host to recover more readily from this somewhat limited infection. The drug does not eliminate the etiological agent from its residence in the throat but this temporary asymptomatic convalescent carrier state is not a unique situation. Other respiratory viruses, mycoplasma organisms and causative bac- teria can be isolated from similar areas during or beyond convalescence. Additional information on this problem is needed. Throughout our investigations and those of others reported here, no toxic effects of amantadine were noted. In most instances a total daily dose of 200 mg was administered. Recent evidence by Smorodincev et al. (1969) suggests that this dose may suffice to prevent illness. It is to be hoped that the encouraging results obtained with amantadine herald even more potent antiviral drugs in the future. 675 676 R. B. HORNICK AND OTHERS REFERENCES Cluff, L. E., Canter, A. & Imboden, J. (1966) Arch. intern. Med., 117, 159 Davies, W. L., Grunert, R. R., Haff, R. F., McGahen, J. W., Neumayer, E. M., Paulshock, M., Watts, J. C., Wood, T. R., Hermann, E. C. & Hoffmann, C. E. (1964) Science, 144, 862 Grunert, R. R., McGahen, J. W. & Davies, W. L. (1965) Virology, 26, 262 Hoffmann, C. E., Neumayer, E. M., Haff, R. F. & Goldsby, R. A. (1965) J. Bact., 90, 623 Smorodincev, A. A., Zlydnikov, D. M. & Kiseleva, A. M. (1969) Ann. N. Y. Acad. Sci. (in press) Stanley, E. E., Muldoon, R. L., Akers, L. W. & Jackson, G. G. (1965) Ann. N. Y. Acad. Sci., 130, 44 Togo, Y., Hornick, R. B. & Dawkins, A. T., Jr (1968) J. Amer. med. Ass., 203, 1089