CHAPTER XII
Bullis Fever
Dwight M. Kuhns, M.D., and Capt. Donald L. Learnard, MSC
CHARACTERISTICS AND CONTROL
Among military personnel, as among civilians, obscure fevers, or fevers of unknown origin, have occurred frequently. Notable advances have been made by discerning clinicians who have been able to see within this heterogeneous group certain distinctive cases which subsequently have become recognized as representative of a previously unrecognized or new disease. In modern times, the clinical differentiation has been followed by confirmatory epidemiological and laboratory studies which have rounded out the conception of the disease as an entity. This was essentially the course of events that were set in train by the original observations made in 1942 by Col. John C. Woodland, MC, Chief of Medicine, Station Hospital (later named BrookeGeneral Hospital), Fort Sam Houston, Tex. As a result, Bullis fever was recognized and was named for the camp and training area from which most of the cases came.
In April 1942, it became apparent to Colonel Woodland that, in the group of military patients suffering from various acute febrile diseases, there were a number who were ill with a clinical disease entity that defied definite identification. In June 1942, Colonel Woodland invited Dr. John R. Paul, Director, Commission on Neurotropic Virus Disease, Board for the Investigation and Control of Influenza and Other Epidemic Diseases in the Army, Preventive Medicine Service, Office of the Surgeon General, to visit the hospital as a consultant on the problem. One outcome of Dr. Paul's visit was a further request of the Preventive Medicine Service for assistance. As a consequence, a special advisory group of experts who accepted the invitations of the Surgeon, Eighth Corps Area, and of The Surgeon General went to Fort Sam Houston and Camp Bullis, Tex., in the middle of July 1942 to assist in the investigation. The members of this group were Dr. Kenneth F. Maxcy, consultant to the Secretary of War and a member of the Board for the Investigation and Control of Influenza and Other Epidemic Diseases in the Army; Dr. Norman H. Topping, Passed Assistant Surgeon, U.S. Public Health Service; and Dr. John C. Snyder, staff member, International Health Division, Rockefeller Foundation. In a preliminary report, dated 10 August 1942, to The Surgeon General, through Col. (later Brig. Gen.) James S. Simmons, MC, Chief, Preventive Medicine Service, this group expressed agreement with
344
the opinion of Colonel Woodland and his associates (especially Maj. M. M. McDowell, MC, and Capt. (later Maj.) John T. Richards, MC) that a typhus-like fever, possibly a tick fever, was probably a distinct disease among the patients suffering with acute febrile conditions who had been admitted to the station hospital at Fort Sam Houston.
Studies by other groups and individuals followed while Colonel Woodland and his associates continued their own studies of the disease. During 1943-44 a commission on Bullis fever was organized and administered by the Rocky Mountain Laboratory of the U.S. Public Health Service, Hamilton, Mont., under the direction of Dr. R. R. Parker. In addition, members of the staff of the Eighth Service Command Laboratory, Fort Sam Houston, Tex., the University of Texas, the Bureau of Entomology and Plant Quarantine of the U.S. Department of Agriculture, and various other organizations carried out clinical, experimental, and field investigations on the disease during the succeeding 5 years. The knowledge accumulated from these studies is the chief substance of this chapter.
In all cases studied, the onset was characterized by chills, headache, moderately high fever (102? to 105? F.) with nausea, vomiting, and lymphadenopathy.1 The fever lasted from 4 to 7 days and subsided by lysis. Pronounced lassitude, prostration, anorexia, and general weakness were noted during the febrile stage of the disease, and the majority of patients complained of severe pains in the postorbital and occipital regions. Upper respiratory symptoms were not consistently found.
Differential diagnosis presented many problems. Signs, symptoms, and laboratory findings indicated that the disease, later called Bullis fever, was similar to, but not identical with, infectious mononucleosis. The comparison is presented in table 32.
Laboratory findings in all cases showed a tendency toward leukopenia, with associated neutropenia. Normal results were obtained in erythrocyte counts, hemoglobin determinations, urinalyses, and cerebrospinal-fluid examinations. Blood cultures, heterophil antibody, Proteus, and cold agglutination tests were negative, as were agglutination tests for brucellosis, tularemia, typhoid and paratyphoid fevers. Biopsy of enlarged lymph nodes revealed lymphoid hyperplasia.
Mode of Transmission
In every case, there was evidence that the individual had been bitten by ticks, and in some instances the ticks were found clinging to the patient on admission. These ticks were all of the species Amblyomma americanum2
1(1) Woodland, J. C., McDowell, M. M., and Richards, J. T. : Bullis Fever (Lone Star Fever-Tick Fever). An Endemic Disease Observed at Brooke General Hospital, Fort Sam Houston, Tex. J.A.M.A. 122: 1156-1160, 21 Aug. 1943. (2) Livesay, H. R., and Pollard, M.: Laboratory Report on a Clinical Syndrome Referred to as "Bullis Fever." Am. J. Trop. Med. 23: 475-479, September 1943.
2Bishopp, F. C., and Trembley, H. L. : Distribution and Hosts of Certain North American Ticks. J. Parasitol. 31 : 1-54, February 1945.
345
(the Lone Star tick), which was found by later studies to be the predominant species of 12 species found in the area. None of the other species were known to attach themselves to man, although the rabbit tick (Haemaphysalis leporis-palustris), shown by Steinhaus and Parker3 to harbor a filter-passing agent, was present at Camp Bulbs in small numbers.
3Steinhaus, E. A., and Parker, R. R.: The Isolation of a Filter-passing Agent From the Rabbit Tick Haemaphysalis leporis-palustris Packard. Pub. Health Rep. 59: 1958-1959, November 1944.
346
In view of the ticks or tick bites found in all cases, the theory was advanced that the agent causing the disease was carried by the tick as a vector and by wild or domestic animals as hosts. Transferees from other nearby posts in this same area who had not come in contact with ticks did not show evidence of the disease.
Amblyomma americanumis a small, hard tick which varies considerably in size and also shows an unusual disparity between the smaller males and larger females. It is known to be a three-host tick found in abundance in the States bordering the Gulf of Mexico, as well as along the South Atlantic coast, and in parts of Oklahoma, Arkansas, and Missouri.4 It is also found in great numbers in Mexico and occasionally in other parts of Central America and in South America (maps 3 and 4).
MAP 3.-General distribution of Amblyomma americanum in the Western Hemisphere.
4Cooley, R. A., and Kohls, G. M.: The Genus Amblyomma (Ixodidae) in the United States. J. Parasitol. 30 : 77-111, April 1944.
347
MAP 4.-Areas of the United States where Amblyomma americanum is known to havebeen found.
A. americanum is important as a vector of Rocky Mountain spotted fever and of Q fever, as well as of Bullis fever. More complete studies of its prevalence were made, and investigations were initiated to determine effective methods of control.
Case report.5-A white soldier, 22 years of age, was admitted to Brooke General Hospital on 25 June 1942, complaining of severe generalized headache of 4 or 5 hours' duration. He appeared to be acutely ill. The onset of the illness was sudden, with nausea, vomiting, and severe headache. He had no prodromal symptoms, and his history was noncontributory. He had been at Camp Bullis,Tex., for 1 week, and had returned to Fort Sam Houston 5 days before the onset of the disease. While at Camp Bullis, he had suffered numerous tick and chigger bites. Physical examination revealed flushed skin, and evidence of numerous insect bites on the abdomen and legs. Moderate lymphadenopathy was found; otherwise, physical examination gave negative results. Laboratory findings were not indicative, hematological studies showed a range of 4,600 to 6,800 leukocytes over a 15-day period, erythrocyte and hemoglobin determinations remained normal throughout, and no malaria plasmodia were found. Urinalyses revealed only a slight trace of albumin during the febrile course of the disease, and the spinal-fluid cell count was normal. Cultures for bacteria and agglutination tests were negative.
The patient's progress in the hospital was rapid, and temperature receded by the 5th day, although general weakness, lassitude, and malaise were noted. He was ambulatory by the 10th day, and completely asymptomatic. There was still generalized glandular enlargement, the only constant physical derangement in this case, which disappeared by the 14th day after admission. Convalescence was complete after 22 days, except for a moderate lymphocytosis.
5See footnote 1(1), p. 344.
348
Treatment
Routine nursing and bed care were administered during the acute stage of the disease. Codeine sulfate, with acetyl salicylic acid, was given for relief of the severe headache, the intake of fluids was encouraged, dextrose and saline were given intravenously, and hydrotherapy was employed to combat the febrile reaction. Chemotherapy was used in a few cases but was observed to increase the toxicity of the disease. From clinical observation, it was deduced that the incubation period of this disease was from 3 to 10 days. The disease was self-limited in nature, and varied from a mild febrile illness of short duration, to a severe, debilitating, prolonged disease with a protracted convalescence.
MILITARY EXPERIENCE
At BrookeGeneral Hospital, during the months of May and June 1943, 485 patients with the syndrome were observed, and probably nearly as many had been observed the previous year. Epidemiologically, it appeared that this disease was more severe and more prevalent in 1943 than in 1942.6 It was postulated that either the agent had increased in virulence by 1943 as a result of repeated passage or the population exposed during 1943 was more susceptible. CampBullis had been used before World War II by National Guard troops composed largely of local individuals who perhaps possessed some degree of acquired immunity. Persons acquiring the disease in 1942 and 1943 were, for the most part, residents of relatively tick-free sections of the country, such as the northern part of the Midwest and the North Atlantic coastal regions. During this period, a division was in training at CampBullis. The incidence of Bullis fever in persons exposed to infection was about 1 in 30 (approximately 3 percent). During years following the war, cases of the disease continued to occur among military personnel engaged in testing tick repellents and among civilians who were exposed to ticks in this area.
Significance
Although no permanent disability was reported as a result of the disease, its frequent occurrence in individuals at Camp Bullis posed a distinct medical problem, and two known fatalities are attributed to Bullis fever, one in May 19437 and one in May 1946.8 Bullis fever is not listed as such by the Medical Statistics Division, Office of the Surgeon General of the Army, and breakdown of incidence and incidence rates has not been completed. The actual number of admissions is not known, as only preliminary tabulations have
6Livesay, H. R., and Pollard, M.: Serological Studies of Bullis Fever. Am. J. Trop. Med. 24: 281-284, September 1944.
7See footnote 1 ( 1 ), p. 344.
8Armed Forces Institute of Pathology, Accession No. 182614, Brooke Army Hospital No. A-59?46, May 1946.
349
been made.9 Military officials responsible for the health of troops finally decided not to use the area for training during the latter years of the war, despite the many control measures suggested and tested.
Research
The original description of Bullis fever as a clinical entity stimulated research in two fields of investigation. Clinical and laboratory studies were made at Brooke General Hospita1,10 the Eighth Service Command Laboratory (later known as the Fourth Army Area Medical Laboratory), the U.S. Public Health Service Rocky Mountain Laboratory, and the University of Texas Medical Branch, Galveston, Tex.11 Entomologic studies of methods of tick control were instituted at Camp Bullis by the Bureau of Entomology and Plant Quarantine, U.S. Department of Agriculture,12 and by the U.S. Army Medical Department, after the Lone Star tick was suspected as the arthropod vector.
At the Eighth Service Command Laboratory, Livesay and Pollard, in 1943, demonstrated that the disease had no immunologic relationship to Rocky Mountain spotted fever and that guinea-pig reaction, Proteus agglutinations, and complement-fixation tests indicated that it was not typhus fever. The clinical syndrome and complement-fixation tests also served to separate it from Q fever. They succeeded in passing through guinea pigs a rickettsialike agent (later believed to be a virus), obtained from clinical cases, and in inducing febrile reactions. Rickettsialike bodies were observed in guinea pigs killed during the febrile stage of the disease and in hyperplastic lymph nodes from human patients. During the same year, Anigstein and Bader, at the University of Texas Medical Branch, recovered a rickettsialike organism from ground, filtered suspensions of Amblyomma americanum ticks collected at CampBullis and later reported on its specificity.
9The Medical Statistics Division, SGO, advised on 4 December 1952, "except for 1942, * * * admission data have been obtained from sample tabulations of individual medical records. The 1942 data are based on a complete file of records; data for 1943 are based on a 20 percent sample of admissions. All data are to be considered preliminary pending publication of final tabulations * * * Bullis fever was first included in an alphabetical list by the Medical Statistics Division in 1943, but, in that year, Bullis fever was coded to a diagnostic group which also included agranulocytosis and infectious mononucleosis. From 1944-48, under a revised coding system, Bullis fever was coded to a residual category, 'other infectious and parasitic diseases.' Thus, it should be noted that through 1948, no specific code for Bullis fever existed in the coding systems used by the Medical Statistics Division. Beginning with 1949, diagnostic classification and coding were revised,
* * * [and] since a card for Bullis fever does not appear in this file, it appears unlikely that,from 1949 to date, medical records have been received with the diagnosis of Bullis fever.''
10Pollard, M., Livesay, H. R., Wilson, D. J., and Woodland, J. C.: Experimental Studies With Bullis Fever. Am. J. Trop. Med. 26 : 175-187, March 1946.11(1) Anigstein, L., and Bader, M. N.: Preliminary Report on Investigations of Bullis Fever. Texas Rep. Biol. & Med. 1: 298, 1943. (2) Anigstein, L., and Bader, M. N.: Investigations on Rickettsial Diseases in Texas. Part 4. Experimental Study of Bullis Fever. Texas Rep. Biol. & Med. 1 : 389-409, 1943. (3) Bader, M. N., and Anigstein, L.: Specificity of Bullis Fever Rickettsia. Texas Rep. Biol. & Med. 2: 405-412, 1944. (4) Blair, R. K., and Bader, M. N.: Observations on Experimental Bullis Fever in Man. Texas Rep. Biol. & Med. 3: 105-111, 1945.
12Smith, C. N., and King, W. V.: Field Studies of Tick Repellents. Am. J. Trop. Med. 30 : 97-102, January 1950.
350
In further studies, Livesay and Pollard, during 1944, concluded that the Bullis fever syndrome is not characterized by significant Proteus OX-19, OX-K, or OX-2 agglutination reactions. They also demonstrated that the complement-fixation test did not reveal a serologic relationship between the agents of Bullis fever and Q fever. At the University of Texas Medical Branch, Blair and Bader, in 1945, demonstrated the infectivity of guinea-pig-propagated strains of Bullis fever for human beings.
Control Measures
During the course of World War II, little was done to control conditions which led to outbreaks of Bullis fever except to close the post for training activities of troops. Attempts were made to destroy ticks on the ground by burning grass around campsites, but few troops were allowed in the area after 1943. Because of the prevalence of Bullis fever, the Air Forces placed the CampBullis and Stanley reservations off limits for the training of troops.
POSTWAR RESEARCH
Laboratory Studies
After actual cessation of World War II, many studies were conducted in relation to this disease. In 1946, Pollard, Livesay, Wilson, and Woodland13 reported their success in reproducing the Bullis fever syndrome in human volunteers by inoculation of whole blood from febrile cases (table 33). In this work, they stated that the size of the agent appeared to be nearer to the size of the elementary body agents such as ornithosis than to that of the typical rickettsiae. On one occasion, they reproduced the syndrome by using the filtrate passed through a Seitz E.K. filter (chart 2). In this instance, all human volunteers had been found previously to be seronegative for murine typhus, Rocky Mountain spotted fever, Q fever, lymphogranuloma venereum, Proteus OX-19, Proteus OX-2, and heterophil.
TABLE 33.-Transmission of chick-embryo-propagated Bullis fever agent from a febrile case
Human volunteer | Generations of passage | Inoculations of yolk-sac emulsion | Clinical syndrome reproduced |
No. 1 | 2 mouse and 4 chick embryo generations | 1 cc. | After 6 days |
No. 2 | ...do... | 1 cc. | After 4 days |
No. 3 | 2 mouse and 11 chick embryo generations | 1 cc. | Do. |
No. 4 | 2 mouse and 20 chick embryo generations | 1 cc. | After 3 days |
13See footnote 10, p. 349.
351
CHART 2.-Human whole blood transmission experiments
Pollard and coworkers found that blood-propagated yolk sac strains, after 20 serial transfers, and tick-propagated yolk sac strains, after 12 transfers, reproduced the disease in human beings.
In further studies, they showed that the immunologic responses induced by natural cases of the disease, by the blood-propagated yolk sac strain, and by the tick-propagated yolk sac strain, were the same.
From work done at this time, it was concluded that there was no immunologic relationship between this agent and the agent of Colorado tick fever. Livesay, Wilson, Pollard, and Woodland, in 1946, demonstrated that there is no immunologic relationship between Bullis fever and dengue fever.14
Tick Control Studies
As a result of a conference held in May 1946, at Fort Sam Houston, where representatives of the Bureau of Entomology and Plant Quarantine, U.S. Department of Agriculture, and the Office of the Chief of Engineers, U.S. Army, were in attendance, investigations into tick control at Camp Bullis were instigated. An informal report was submitted by the Bureau of Entomology and Plant Quarantine in June 1946 stating that: Large numbers of ticks inhabited thick bushes and the underside of leaves, making control problems difficult; pastured areas did not possess the heavy undergrowth
14(1) Livesay, H. R., Wilson, D. J., Pollard, M., and Woodland, J. C. : Experimental Studies of Bullis Fever and Dengue Fever. Am. J. Trop. Med. 26: 379-381, July 1946. (2) Pollard, M., Livesay, H. R., Wilson, D. J., and Woodland. J. C.: Immunological Studies of Dengue Fever and Colorado Tick Fever. Proc. Soc. Exper. Biol. & Med. 61 : 396-398, April 1946.
352
required by ticks; deer were the principal host of ticks; preliminary tests indicated that good (over 95 percent) control of ticks could be obtained by application of DDT in pyrophyllite at the rate of 40 pounds per acre; and power dusters were recommended in lieu of spray because of the thick undergrowth. The report also disclosed that nymph and adult ticks were present in considerable numbers early in May, reaching a peak late in May and the first week in June. The decline was rapid during the next 8 weeks, and activity was practically ended by the middle of October.
Department of Agriculture workers carried out investigations on different insecticides and methods of application, including airplane and ground sprays. Sprays of nicotine, lethane, thanite, and pyrethrin did not prove satisfactory, but mixtures of DDT and benzene hexachloride gave highly effective control for a period of 4 or 5 weeks. An experiment using the tick-infested area as pasturage for cows and periodically eliminating the ticks from their bodies did nothing to reduce the tick population.
In 1947, 20 volunteer soldiers from the Field Medical Service School, Brooke Army Medical Center, under Dr. J. M. Brennan, U.S. Public Health Service, Hamilton, Mont., underwent a series of investigations in which uniforms impregnated with various chemical agents were tested.15 Numerous chemicals were used, including butylacetanilide, benzyl cyclohexanol, 2-phenylcyclohexanol, benzyl bonzoate, dimethylphthalate, dibutylphthalate, a mixture of the last three, and phthalic acid-hexahydro-di-ethyl ether. Technical assistance and materials were supplied by the U.S. Department of Agriculture. Results of the experiments showed that butylacetanilide afforded complete protection against nymphs and adults, and phenylcyclohexanol provided good protection properties. All others proved to be either ineffective or toxic. Three of these twenty volunteers developed a syndrome identical to that described as Bullis fever. Attempts at human transmissions were unsuccessful.
A similar series of tests was made in 1948 under Drs. Willard V. King and Carroll N. Smith,16 Department of Agriculture Experiment Station, Orlando,Fla. Among many repellents tested were Indalone, benzyl benzoate, benzobenzoate, dimethyl-carbate, 2-ethyl, 2-butyl, 1,3-propanediol, and n-hexyl mandelate. Of these agents, Indalone appeared to hold the best promise for protection against ticks.
Treatment Studies
As noted in earlier clinical studies, treatment of Bullis fever with chemotherapeutic agents, except for the relief of headache and fever, either in‑
15(1) Brennan, J. M.: Field Investigations Pertinent to Bullis Fever. Texas Rep. Biol. & Med. 3: 112-121 and 204-226, 1945. (2) Brennan, J. M.: Preliminary Reports on Some Organic Materials as Tick Repellents and Toxic Agents. Pub. Health Rep. 62 : 1162-1165, August 1947. (3) Brennan, J. M.: Field Tests With Tick Repellents. Pub. Health Rep. 63: 339-346, March 1948.
16(1) See footnote 12, p. 349. (2) Smith, C. N., and Burnett, D., Jr.: Laboratory Evaluation of Repellents and Toxicants as Clothing Treatments for Personal Protection From Fleas and Ticks. Am. J. Trop. Med. 28: 599-607, July 1948.
353
creased the toxicity of the disease or contributed nothing to the welfare of the patient. Penicillin and sulfonamide drugs were used in a number of cases without apparent beneficial effect.
At least one case responded to treatment with PABA (para-aminoben?zoic acid), according to a report in 1949 by Arnold and Van Noate.17 Because of the dramatic response of certain rickettsia' diseases (typhus fever and Rocky Mountain spotted fever) to PABA, a patient at Walter Reed Army Hospital, who demonstrated all of the symptoms of Bullis fever, was given a therapeutic trial by oral administration of this drug. A rapid response was accomplished by the return to normal temperature and general clinical improvement within a few hours. Although there appear to be no published data, other cases of Bullis fever are reported to have been successfully treated with PABA, further strengthening the theory that the disease is rickettsial in origin.
SUMMARY
An apparently new clinical entity with fever, lymph node involvement, and evanescent rash, and with absence of sore throat, occurred in 1,000 or more military personnel stationed at Camp Bullis, Tex., during and following World War II. Laboratory findings were negative for known infectious diseases of this area, including infectious mononucleosis, Colorado tick fever, dengue fever, and Rocky Mountain spotted fever.
The disease was attributed to a filterable organism believed to be of the rickettsiae, although differential leukocyte counts were similar to those found in diseases of viral origin. Laboratory workers during the period reported that they were able to isolate causative organisms from emulsions of the tick species Amblyomma americanum believed to be the vector of the agent.
A. americanum was found in large numbers in this area as ectoparasites of wild and domestic animals. Control measures tested at Camp Bullis included evaluation of insecticides and tick repellents. Good results were obtained from use of DDT with benzene hexachloride and Indalone respectively.
Successful treatment of the infection has been reported in a few cases by oral administration of PABA, but other chemotherapeutic agents were of no apparent value.
It is evident that this reservation should not be used again by great numbers of troops unless every effort is made to confirm the etiology and the mode of transmission of this disease. However, it is believed that, if investigations were carried out, this syndrome would be discovered in many other areas where A. americanum is found in abundance.
17Arnold, W. T., and Van Noate, H. F.: Bullis Fever : Report of a Case Treated With Para?Aminobenzoic Acid. Bull. U.S. Army M. Dept. 9: 218-223, March 1949.
