THE JOURNAL OF INFECTIOUS DISEASES. VOL. 150,.3. SEPTEMBER 1984 1984 by The University of Chicago. All rights reserved. Antitoxin l.evels in Botulism Patients Treated with Trivalent Equine Botulism Antitoxin to Toxin Types A, B, and E Charles H. Hatheway, John D. Snyder,* Jerry E. Seals.] Timm A. Edell,:j: and George E. Lewis, Jr. From the Centers for Disease Control, Atlanta, Georgia; and the United States Army Medical Research Institute of Infectious Diseases, Fort Detrick, Maryland Serum levels of equine-botulism antitoxin to toxin types A, B, and E were measured in four. type-a botulism patients who had received equine-botulism antitoxin. High circulating levels capable of neutralizing in excess of 1 x 10 8,9 x 107, and 6 x lq6 50070 mouse lethal doses of toxin of types A, B, and E, respectively, were detected. There was little depletion of type-a antitoxin even though two of the patients had circulating type-a toxin before treatment. The half-life for antitoxin persistence for one patient was calculated as being 6.5, 7.6, and 5.3 days for antitoxin types A, B, and E, respectively. Antitoxin levels were not proportionate to the amount (range, 2-4 vials) injected and did not appear to be affected by whether the route of administration was iv or im. Peak serum levels of antitoxin were 10-1,000 times higher than amounts needed to neutralize the toxin measured in the serum of these and other patients with botulism. Equine-botulism antitoxin for treatment of patients with botulism is available from the Centers for Disease Control [1]. The therapeutic product used in the United States in recent years has almost exclusively been trivalent antitoxin types A, B, and E (anti-abe). The potency for each vial of this product, as stated by the manufacturer (Connaught Laboratories, Willowdale, Ontario) on the package insert, is 7,500 IV for type A, 5,500 IV for type B, and 8,500 IV for type E. Despite a paucity of data on the efficacy of botulism antitoxin and despite the risks of anaphylaxis and serum sickness [I, 2], 830/0 of the patients with either foodborne or wound-related botulism reported in 1979 and 1980 were treated with this antitoxin [3]. Few data exist on the amount of antitoxin subsequently found to be circulating in recipients, the toxin neutralizing capacity, or the posttreatment half-life of Received for publication March 15, 1984, and in revised form May 9, 1984. Please address requests for reprints to Dr. Charles L. Hatheway, Centers for Disease Control (CID, DBD), Atlanta, Georgia 30333. * Present address: Pediatric GI and Nutrition Unit, Department of Pediatrics, Massachusetts General Hospital, Boston, Massachusetts 02114. t Present address: 5438 Smooth Meadow Way, Columbia, Maryland 21044. t Present address: 2773 Harris Street, Eureka, California 95501. Present address: US Army Medical Research Unit-M, Kuala Lumpur, c/o US State Department, Washington, DC 20520. the product [4]. In November 1978 an outbreak of type-a botulism occurred [5] that provided us with the opportunity to determine circulating-antitoxin levels, the half-life, the relation ofboth dose and route of administration to circulating-antitoxin levels, and the toxin neutralizing capacity of trivalent equine botulism antitoxin detected in the patients' serum. From these data we present a reappraisal of the recommendations on use and administration of botulinal antitoxin. Materials and Methods Study group. Four of seven persons with laboratory-confirmed type-a botulism acquired during a foodborne botulism outbreak in November 1978 [5] were treated with anti-abe. In patient I signs of botulism developed 27 hr after she had eaten the contaminated food, but she was not hospitalized until two days later, and an additional day had passed before she was treated with anti-abe. Type-A botulinal toxin was detected in two serum samples from this patient, one obtained three days after she ate the contaminated food and the second a day later, just before she was treated with anti-abe. Each of these samples of toxic sera contained <5 mouse LD so of toxin/rnl. This patient was the most severely affected person in the outbreak and required ventilator assistance for several months. She was given four vials of anti-abe beginning 103 hr after exposure at 1-2-hr 407
408 Hatheway et al. intervals, the first three of which were given im and the last of which was given iv. In patient 2 signs of botulism developed 43 hr after he had eaten the contaminated food. A serum sample obtained one day later, before anti-abe treatment, contained demonstrable botulinal toxin, quantitatively similar to that found in the serum of patient 1. This patient received three vials of anti-abe beginning at 78 hr after exposure, the first of which was given im and the other two of which were given iv. In patient 3 there was an interval of 92 hr between exposure and onset of illness. A serum sample was taken one day after onset, but no botulinal toxin was demonstrable. This patient was treated six days after onset with two vials of anti-abe, one given im and the other given iv. In patient 4 signs of botulism developed 30 hr after he had eaten the contaminated food. A sample of blood was taken for botulism testing approximately six days after onset, and the tests for toxin were negative. One day after sampling, this patient was treated with two vials of anti-abe the first being given iv and the second being given im 2 hr later. The state of paralysis of patients 2, 3, and 4 was much milder than that of patient 1 and primarily involved cranial nerves. Each patient was hospitalized less than one week. In addition to the postive tests for type-a botulinal toxin in the serum of patients 1 and 2, Clostridium botulinum type A was isolated from at least one stool specimen from each of the four patients. Therapeutic anti-abe. The therapeutic product administered to the four patients was lot 3001 of the anti-abe described above. The vials of each lot are filled with an amount of bulk anti-abe that will assure a minimal initial potency 20% in excess of the stated potency for each type of antitoxin as required by federal regulations [6]. The lot then has a five-year dating period. Assays on samples of the lot of the product administered to the patients in this study were performed in our laboratories. Thus, we were able to estimate how many IV were actually administered to each patient so that a valid comparison could be made with patient posttreatment serum levels, since identical procedures were used for all measurements. Patient posttreatment serum samples. At least four posttreatment serum samples were obtained from each of the four patients during the first five days. Seven additional samples were obtained from patient 2 at 6-10-day intervals up to 53 days posttreatment, and two additional samples were obtained from patient 3 on 20 and 27 days posttreatment. Antitoxin assays. Levels of antitoxin types A, B, and E were measured in samples of patient posttreatment sera and in samples of the therapeutic anti-abe by determining the neutralizing end points in titrations against botulinal-toxin types A, B, and E. Each toxin was standardized against its homologous type of World Health Organization International Standard antitoxin. The concentrations of toxins used in the tests were those that were neutralized by the requisite amount of antitoxin standard as described by Cardella [7] (i.e., 0.02, 0.005, and 0.0125 IV/ml for types A, B, and E, respectively). Doubling dilutions of the antitoxin samples were mixed with equal volumes of the standardized toxins and after being incubated for 1 hr at room temperature, 0.2 ml of each mixture was injected into each of four mice. In each test the World Health Organization International Standard was titrated against the test toxin to establish the precise IV amount for the end-point dilution, which was then translated into IV of antitoxin/ml of test sample. When more than one determination was made on a serum sample<or more than one samplewas obtained from a patient in a 24-hr period, the values were averaged. Half-life estimation. The half-lives for antitoxin types A, B, and E in the serum of patient 2 were calculated from the regression of the log of the antitoxin concentration plotted against the time of sampling. The following equation was used for half-life calculation: half-life = 0.693t/(log Co-log C t ), where t is the number of days after administration, Co is the peak concentration attained after administration of the antitoxin, and C, is the concentration measured at time t. Knowing the half-life and the initial (peak) concentration, we were able to estimate the concentration at time t from the following equation: C, = Co' e -(O.693t/half-life). The sampling period was too short to determine the antitoxin half-lives for the other three patients. Toxin-neutralizing capacity. One British Unit of botulism antitoxin was originally defined as that amount that would neutralize 10,000 mouse LD so of botulism toxin [8]. In the establishment of
Botulism Antitoxin 409 Table 1. Antitoxin type A B E Results Potency of therapeutic anti-abe. Stated 7,500 5,500 8,500 IV per vial Assayed* 9,890 ± 2,860 9,090 ± 4,950 27,860 ± 1,430 * Seven determinations each were done for antitoxin types A and B, and six determinations were done for antitoxin type E. the World Health Organization International Standards for Botulism Antitoxin, the British Units were accepted, with the exception that the IV for type E was made one-tenth the size of the British Vnit [9]; therefore, for type E, 1 IV should neutralize 1,000mouse LD so, and for all other types, 1 IV should neutralize 10,000 mouse LD so To estimate the toxin-neutralizing capacity conferred by antitoxin treatment of a patient, it is necessary to multiply the IVIml assayed by the plasma volume times either 10,000mouse LD so (for antitoxin types A or B) or times 1,000 mouse LD so (for antitoxin E). For simplicity 3,000 ml was assumed to be the plasma volume for all four patients in our study. The IV per vial of each type of antitoxin as determined in our assay are listed in table 1. For each type the amount exceeds the amount stated in the package insert. This is especially true for type E, which has an assay value three times that of the stated value. Table 2 lists the antitoxin levels determined for 28 different patient-day intervals posttreatment. Pretreatment samples from patients 1 and 2 were found to contain circulating type-a toxin. Most of the samples were taken during the first six days posttreatment. The data for each patient regarding onset of illness, diagnostic testing of serum for toxin, time of treatment with and amount of antitoxin given, and the peak amount of antitoxin measured in the circulating blood are summarized in table 3. The type-a antitoxin levels were always the highest, whereas the type-e levels were either the lowest or comparable to the type-b levels. Sufficient serum samples were obtained from patient 2 to determine the half-lives of antitoxin types A, B, and E (figure 1). The half-life approximated one week for each type; the time in days being 6.5, 7.6, and 5.3 for antitoxin types A, B, and E, respectively. The level of circulating antitoxin did not vary with the number of vials administered to the patients; higher percentages of circulating antitoxin were found in the patients given two vials than in those given three or four vials. No difference in circulating-antitoxin levels was evident between those patients whose injections were predominantly iv and those whose injections were mostly im. Discussion Several limitations in methodology must be considered when reviewing the data reported in the present study. Because the mouse-toxicity assay Table 2. Levels of antitoxin types A, B, and E in the serum of patients with type-a botulism treated with anti-abe. Type-A/type-B/type-E antitoxin levels in IV/ml in patient Days --- posttreatment 2 3 4 I 2 3 4 5 6 12 20 22 27 31 38 45 53 5.35/1.7312.02 4.40/3.2012.52 3.90/1.3012.52 5.89/1.30/1.26 4.43/1.6012.00 4.7812.15/1.80 4.04/1.55/ 3.28/1.02l 3.91/1.81/1.42 2.00/1.60/0.71 1.26/0.80/0.086 0.40/0.40/0.14 0.18/0.14/0.04 0.10/0.07/NM NM/0.04/NM NM/ 4.00/3.8012.09 4.41/2.8211.60 5.36/3.17/1.22 3.94/1.36/0.80 2.86/3.20/1.13 0.20/0.10/ND NM// 4.5412.27/1.21 3.60/1.82/1.26 1.80/3.20/1.13 3.20/3.1612.00 TE. = not done; NM = not measurable.
410 Hatheway et al. Table 3. Details of four type-a botulism cases. Antitoxin treatment Time in hr after exposure of Amount in IV Serum Detected Patient no. Onset of toxin Antitoxin Total (010 of (sex, age, weight in kg) Illness testing treatment (route) Type InjectedII Peak level amount injected)# 1 (F, 55, 59)* 27 72,96t 103 (im), 105 (im), A 39,560 5.89 17,670 (45) 107 (im), 108 (iv) B 36,360 3.20 9,600 (26) E 111,440 2.52 7,560 (7) 2 (M, 46, 65)* 43 48t 78 (im), 88 (iv), A 29,670 4.78 14,330 (48) 93 (iv), B 27,270 2.15 6,450 (24) E 83,580 2.00 6,000 (7) 3 (M, 37, 75) 92 144t 240 (im), 240 (iv) A 19,780 5.36 16,080 (81) B 18,180 3.80 11,400 (63) E 55,720 2.09 6,270 (11) 4 (M, 27, 71) 30 168t 189 (iv), 191 (im) A 19,780 4.54 13,620 (69) B 18,180 3.20 9,600 (53) E 55,720 2.00 6,000 (11) * Patients with circulating toxin. t Botulinal toxin detected. t Botulinal toxin not detected. Each entry indicates one vial of antitoxin. II No. of vials given x IV per vial. Assay values for lot 3001 are reported in table 1. # Assuming a plasma volume of 3000 ml for each patient. was used, doubling dilutions of the test sera were employed, and differences as great as twofold were sometimes observed when repetitive tests were performed on the same sample. To give more precise values, it would be necessary to test from six to seven incremental amounts of serum over a twofold-dilution range bracketing the 50% end point; however, in the present study this would have been both prohibitively expensive (in terms of time, effort, and number of mice) and impossible (due to limitations on sample quantities); nevertheless, sufficient consistency in the data was present 8.0 4.0 2.0 1.0 0.5 0.25 0.125 0.0625 0.0312............ Figure 1. Levels of antitoxin types... "-.s: 'it... A (0), B (0), and E (.6.) in the serum...... 0 of patient 2. -- a~ e-, 'A. e-, 10 15 20 25 30 35 40 45 50 DAYS
Botulism Antitoxin 411 to give us confidence that they present an accurate picture of the passive immunity conferred on the patients by the anti-abe. There was little difference in the antitoxin levels seen in the four patients in the present study, even though different amounts of antitoxin were administered. The percentage of injected antitoxin in the circulating blood of the patients who had received three and four vials was lower than that in those who had received two vials. Patients 1 and 2 did have demonstrable toxin in their serum, which might have consumed a portion of the type-a antitoxin but would not have accounted for similar deficits in the amounts of antitoxin types Band E in their circulatory systems. An upper limit of 5 mouse LDso of toxin/ml for patients 1 and 2 was estimated, since 0.8 ml of serum was required for consistent killing of unprotected mice; 0.4 ml (the normal test dose) was not sufficient. (The minimum demonstrable concentration of toxin in serum is f\jl mouse LDso of toxin/ml, which probably will not cause death after injection of 0.8 ml but may cause clear signs of botulism in the mouse.) When quantitation of toxin in the serum of patients with botulism has been done in the CDC laboratory, a level of 32 mouse LDso of toxinlml has never been exceeded. This amount of toxin would consume less than 0.01 IU of antitoxin from each milliliter of serum; thus, detectable levels of circulating toxin should have little effect on the administered antitoxin. We have no explanation for the failure to achieve higher antitoxin levels with higher rates of administration and suggest that further investigation is needed in this area. The half-life values calculated for antitoxins types A, B, and E were similar to those estimated by Lewis and Metzger [4] and those calculated for types A and B in patient 3 (in whom these values were determined largely on the basis of a sample taken 20 days posttreatment). With 5 IU of type A/ml, 3 IU of type B/ml, and 2 IU of type E/ml as the approximate average initial peak values for patients 3 and 4 (who received two vials each of antitoxin) and the half-life values from figure 1, one can project the antitoxin levels for any subsequent time. After six weeks the levels would be f\j0.11, 0.11, and 0.02 IU/ml for antitoxin types A, B, and E, respectively. When we compared the peak antitoxin levels of patient 1, who had received four vials, with those of patient 3, who had received only two vials, we saw no difference. In assessing the relative efficacies of iv and im injection, we can compare the percentage of injected antitoxin detectable in the serum of patient 1, who had received threequarters of the antitoxin im, with that in the serum of patient 2, who had received two-thirds of the antitoxin iv; the percentages of each of the three types are almost identical (table 3).. The neutralizing capacities of the peak antitoxin levels measured in the patients in the present study were 50,000 mouse LDso of type-a toxin, 30,000 mouse LDso of type B, and 2,000 mouse LDso of type E per milliliter of patient serum. When multiplied by the plasma volume of the treated patient (conservatively estimated to be 3,000 ml), these amounts will give the total neutralizing capacity for the respective type of toxin. The amount of ingested toxin that comprises a human lethal dose is largely a matter of speculation [10]. The portion of the ingested toxin absorbed into the circulatory system is also unknown, although it is apparent both that the absorption is incomplete (since toxin is often demonstrable in the feces [11]) and that much of it may be inactivated or destroyed during passage through the digestive tract. The amount of toxin measured in the serum of patients before antitoxin treatment probably represents only a portion of the absorbed toxin, since some of the toxin is removed as it binds to the receptors on nerve endings. The serum of most patients examined in the Centers for Disease Control laboratory fails to show any demonstrable toxin. In one published study [11] only 35070 of the patients had sera positive for toxin. As stated above, the highest concentration of toxin that we have detected in human serum is f\j32 mouse LDso/ml. The peak amounts of antitoxin measured in the patients in the present study will neutralize 1,500, 1,000, and 60 times the 32-mouse-LDso/ml toxin levels of toxin types A, B, and E, respectively. The highest amount of circulating toxin in human serum that we know to have been reported is f\j160-mouse-ldso of type E in a case that occurred in England in 1978 [12]. The levels of antitoxin type E found in patients in the present study would have neutralized f\j 12 times that amount of toxin. One unexplained observation is that although the therapeutic product contains more units of antitoxin type E than of antitoxin types A or B, the amount of the former measured in the circulating
412 Hatheway et al. blood was consistently lower. Our results were similar to those reported by Lewis and Metzger [4] in their investigation of the levelsof circulating antitoxins in a treated patient with type-a botulism, since in both studies the level of antitoxin type A was greater than that of antitoxin type B and that of antitoxin type B was greater than that of antitoxin type E. The half-life calculations of from five to seven days for the three antitoxin types as we determined them in patient 2 are also in agreement with the estimates of Lewis and Metzger. From the data obtained in the present study, it appears that there may not be any advantage in giving more than two vials of antitoxin. Given both the initial high levels of circulating antitoxin detected in the present study and the low levels of toxin found in the serum of human patients with botulism, one vial may suffice. However, since the half-life may depend on the possibly slower release of antitoxin administered im, the recommendation on the package insert that one vial be given iv and one vial be given im should be followed. Limitation to two vials is important in the context of the report of Black and Gunn [2], who found that the incidence of serum sickness in patients treated with botulism antitoxin was highest for persons who received more than 40 ml (four vials). The data obtained in the present study show that administration of two vials of anti-abe to human subjects results in high levels of circulating antitoxins that are capable of neutralizing from 12 to 1,500 times the amount of toxin that is found in the serum of patients with botulism. The route of administration (iv or im) does not appear to affect the amount of anti-abe that is incorporated into the circulatory system within the first day, although the iv route would assure the highest initial serum levels. References 1. Center for Disease Control. Botulism in the United States 1899-1977. Handbook for epidemiologists, clinicians, and laboratory workers. Atlanta: Centers for Disease Control 1979 2. Black RE, Gunn RA. Hypersensitivity reactions associated with botulinal antitoxin. Am J Med 1980;69:567-70 3. Morris JG Jr. Current trends in therapy of botulism in the United States. In: LewisGE Jr, ed. Biomedical aspects of botulism. New York: Academic Press, 1981;317-26 4. Lewis GE, Jr, Metzger JF. Studies on the prophylaxis and treatment of botulism. In: Eaker D, Wadstrom T, eds. Natural toxins. New York:Pergamon Press, 1980;601-6 5. Seals JE, Snyder JD, Edell TA, Hatheway CL, Johnson CJ, Swanson RC, Hughes JM. Restaurant-associated type A botulism: transmission by potato salad. Am J Epidemiol 1981;113:436-44 6. Code of Federal Regulations Title 21: Food and Drugs, Chapter 1, part 610.50F. Dating period limitations. Food and Drug Administration. Washington, DC: Government Printing Office, 1983;52-3 7. Cardella MA. Botulinum toxoids. In: Lewis KH, Cassel K Jr, eds. Botulism: proceedings of a symposium. PHS Publication No. 999-FPl, Washington, DC: Government Printing Office 1964:113-130 8. Bowmer EJ. Antitoxins of Clostridium difficile types A, B, C, D, and E: Preparation and assay of proposed international standards. MD thesis. Liverpool, England: University of Liverpool, 1962 9. Bowmer EJ. Preparation and assay of the International Standards for Clostridium botulinum types A, B, C, D, and E antitoxins. Bull. WHO 1963;29:701-9 10. Smith LDS. Botulism: The organism, its toxins, the disease. Springfield, Ill: Charles C Thomas, 1977:177 11. Dowell VR Jr, McCroskey LM, Hatheway CL, Lombard GL, Hughes JM, Merson MH. Coproexamination for botulinal toxin and Clostridium botulism. JAMA 1977;238:1829-32 12. Ball AP, Hopkinson RB, Farrell ID, Hutchinson JOP, Paul R, Watson RDS, Page AJF, Parker RGF, Edwards CW, Snow M, Scott DK, Leone-Ganado A, Hastings A, Ghosh AC, Gilbert RJ. Human botulism caused by Clostridium botulinum type E: the Birmingham outbreak. Q J Med 1979;48:473-91