Arizona Hinshawii Osteomyelitis with Antecedent Enteric Fever and Sepsis A Case Report with a Review of the Literature
DAVID F. KEREN, M.D. WILLIAM RAWLINGS, Jr., M.D. HENRY W. MURRAY, M.D. WILLIAM R. LEONARD, B.S. Baltimore, Maryland
A case of Arizona osteomyelitis of the spine which occurred 11 months after an episode of gastroenteritis and enteric fever is presented. As a close biochemical and antigenk relative of Salmonella, Arizona infection produces a similar clinical course with gastrointestinal manifestations frequently precedtng localized infections by several months. The boney lesion in the present case and in three of the four other cases of Arizona osteomyelitis described in the literature was a chronic inflammation which may have a xanthomatous component. The bone destruction caused by Arizona infection is less severe than that of tubercuious or pyogenic osteomyelitis. Proposed treatment of Arizona osteomyelitis consists of debridement of the localized infection and long term antimicrobial therapy. Arizona hinshawii, a gram-negative bacillus, is a member of the family Enterobacteriacae and is antigenically closely related to the genus Salmonella [ 11. The clinical spectrum of Arizona infections varies from benign gastroenteritis to enteric fever and septicemia with localized infection and is similar to that caused by Salmonella species [2-131. We report a case illustrating the complete natural history of Arizona osteomyelitis. The causative organism was identified in the patient’s blood and stool during an episode of gastroenteritis and enteric fever 11 months prior to the diagnosis of vertebral osteomyelitis. CASE REPORT First Admission. A 53 year old black man was admitted to The Johns Hopkins Hospital (JHH) in September 1973 because of fever and obtundation. Two weeks and weakness.
From the Departments of Laboratory Medicine and Medicine, The Johns Hopkins Hospital, Baltimore, Maryland 21205. Requests for reprints should be addressed to Dr. David F. Keren, Departrnent of Laboratory Medicine, The Johns Hopkins Hospital, Baltimore, Maryland 21205. Manuscript accepted August 14, 1975.
prior to admission
he had a cough,
During the following week
vomiting and diarrhea.
pleuritic chest pain
he experienced
fever,
nausea,
There was no history of ingestion of contaminated
foods, exposure to sick animals or of contact with companions who were ill. Except for heavy alcohol intake, the patient’s past history was unremarkable. Pertinent physical findings on admission included a rectal temperature of 104’F, a pulse rate of 1 lO/min, basilar rales bilaterally, diffuse abdominal
tenderness
without
April 1976
organomegaly
or
masses,
The American Journal of Medicine
hypoactive
Volume 60
bowel
677
ARIZONA HINSHAWII OSTEOMYELlTlS WlTH FEVER AND SEPSIS-KEREN
Figure I. Lateral view of spine showing destruction of 12th lumbar vertebrae and first lumbar spine vertebral bodies. Soft tissue calcification (arrow) is seen anterior to second lumbar spine.
sounds and a negative stool guaiac. The laboratory data included a hematocrit value of 37 per cent, a leukocyte count of 2,700/mm3, a platelet count of 36,500/mm3, serum glutamic oxaloacetic transaminase @GOT) 250 IU/ liter, serum glutamic pyruvic transaminase (SGPT) 299 IU/ liter, albumin 2.8 g/d! and serum urea nitrogen 66 mg/dl. Chest roentgenogram showed a possible infiltrate in the lower lobe of the left lung and multiple calcified granulomas. The sputum contained gram-positive diplococci and gram-negative rods. Intravenous gentamicin and ampicillin therapy was begun. On the third hospital day, cultures of blood, stool and sputum grew a gram-negative rod that was subsequently identified as Arizona hinshawii serotype 7a,7b: 1,7,8. Cultures of cerebrospinal fluid and bone marrow aspirate were sterile. Since the organism was sensitive to both antimicrobials being given, the administration of gentamicin was discontinued and ampicillin (8 g, intravenously/day) was continued for 14 days. The patient became afebrile by the ninth hospital day. Other studies included a negative purified protein derivative, a normal hemoglobin electrophoresis, a negative glucose-6-phosphate dehydrogenase screen, a normal x-ray series of the upper gastrointestinal tract and small bowel, and a well-visualized gallbladder by intravenous cholangiogram. The patient was discharged on the 20th hospital day with a hematocrit value of 28 per cent and a leukocyte
578
April 1978
The American Journal of Medicine
Volume 80
ET AL.
count of 6,000/mm3. A repeat chest film showed fibrotic changes in the lower lobe of the left lung, x-ray films of the thoracolumbar spine were unremarkable. Six weeks after discharge stool cultures were negative for enteric pathogens. At that time the patient complained of mild low back pain and was given acetaminophen. Second Admission. The patient was readmitted 11 months later, in August 1974, with complaints of several weeks of malaise and a gradually enlarging, painful swelling in his right lower back. There had been no fever, chills, night sweats or back trauma. On physical examination he appeared emaciated with temporal wasting. Oral temperature was 99.4’F. There were two large, tender, fluctuant subcutaneous masses on his back. One was paraspinal, measured 8 by 3 cm and extended from the 12th thoracic vertebrae to the fourth lumbar spine, the other was located over the spinous processes of the 12th thoracic vertebrae and the first lumbar spine measured 2 by 3 cm. No microorganisms were seen on gram stain of the thick green fluid aspirated from the paraspinal abscess. Kinyoun stain of the aspirate on the admitting ward was initially thought to contain acid-fast bacilli although subsequent aspirates failed to confirm this finding. X-ray films of the spine showed a large soft-tissue mass and a marked gibbus deformity of the lower thoracic and upper lumbar spine with destruction of the 12th thoracic vertebrae and first lumbar vertebral bodies (Figure 1). Laboratory data included a hematocrit value of 36 per cent, a leukocyte count of 7, 100/mm3, a platelet count of 215,000/mm3, SGOT IMiter, albumin 3.4 gfdl, and a normal serum urea nitrogen. On the basis of the inltiil aspirate smear and the clinical picture, the patient was presumed to have Pott’s disease (tuberculous spondylitls), and isoniazid and ethambutol therapy was begun. A low grade fever developed on the second hospital day. On the fourth day cultures from the abscess aspirate grew an Arizona species subsequently identified as Arizona hinshawii 7a,7b: 1,7,8 sensitive to ampicillin. The significance of this isolate was not fully appreciated until later. Cultures of the blood and stool were negative. On the 10th hospital day, the patient underwent surgical decompression of the abscesses and anterior fusion of the 12th lumbar vertebrae and first lumbar spine wlth a rib graft. Extensive debridement, of the vertebral bodies was not necessary since relatively little necrotic bone was present. The abscess cavity contained 300 ml of pus from which Arizona hinshawii was again isolated. Histologicalty, the abscess wall showed acute and chronic inflammation with large masses of foam cells and gram-negative rods (Figure 2). No histologic evidence of tuberculosis was present, and all cultures were negative for mycobacteria. Because the patient was a poor historian and the medical records of the previous admission were unavailable, his physicians did not become aware of the prior Arizona hinshawii septicemia and its relationship to the present illness until the 1 lth hospital day. At that time, intravenous ampicillin therapy (8 g/day) was begun. One day later, the patient was afebrite and remained so for the rest of his 21 day hospitalization. Since discharge he has been main-
ARIZONA HINSHAWII OSTEOMYELITIS WITH FEVER AND SEPSIS-KEREN
ET AL.
Figure 2. A, photomicrograph of abscess wall with a predominance of “foamy” histiocyte? (xanthoma cells) (arrows). Hematoxylin and eosin stain, magnification X 270. B, photomicrograph of abscess wall showing gram-negative, rod-shaped organisms (arrow) and “foamy” histiocytes (H). Brown and Hopps bacterial stain, magnification X 1,100.
tained on oral ampicillin (2 g/day), and is ambulatoryand doing well eight months after discharge. Ampicillintherapy will be continuedfor one year. MlCftOBlOLOGY
by the agar dilution technic [ 141. The minimum inhibitory concentrations for the organism were tetracycline 4 pg/ml, chloramphenicol 4 pg/ml, cephalothin 8 pg/ml, ampicillin 2 fig/ml and kanamycin 4 &g/ml.
AND SEROLOGY COMMENTS
The cultures of Arizona hinshawii in the present case
from both hospital admissions (blood, stool, sputum, and paraspinous abscess) produced gram-negative, oxidase-negative bacilli on MacConkey agar (BBL). On triple sugar iron agar the Arizona produced an alkaline slant and an acid butt with hydrogen sulfide production. The organism was motile, indole negative and able to utilize citrate as a carbon source, but it did not produce urease. The preliminary identification of Arizona was made by using a modification of the biochemical tests recommended by Edwards and Ewing [I] for differentiating Arizona from Salmonella (Table I). This was confirmed by a strong agglutination of the isolated organism by Arizona polyvalent 0 antiserums (Difco). The anti-Arizona titer of the patient’s serum (obtained one month after hii second discharge) was 1: 1000. This was performed using the patient’s cultured organism. Serum from a noninfected person was used as a negative control. Antiiicrobial-susceptibility testing was performed
The Arizona organism, first isolated in 1939 from diseased reptiles in Arizona, was initially named “Salmonella Dar-es-Saalam variety from Arizona” because of biochemical similarities to Salmonella Dar-es-Saalam [15]. In 1941, Kauffman [IS] proposed classifying the organism as Salmonella Arizonae because of H (flagellar) antigens similar to those present in some Salmonella serotypes. SubsequentTABLE I
Arizona and Salmonella-Biochemical Differences
Biochemical Test
Arizona
Malonate utilization Gelatin liquification ONPGt (lactose fermentation) Dulcitol fermentation lnositol fermentation ‘(+)
refers to delayed
A* (+I -
v+ ”
positive reaction after three or more days.
+ONPG = 0 nitrophenyl-@-D tion of @-galactosidase. *v refers to variable reactions,
AprH1976
Salmonella
galactopyranoside i.e., +, -,
lhe&nerkanJaumalof&dktine
(BBL) for detec-
or (+).
vobne
69
579
ARIZONA HINSHAWII OSTEOMYELITIS WITH FEVER AND SEPSIS-KEREN
TABLE II
ET AL.
Arizona Osteomyelitis
Case1
Case2
Age, race and sex
2. 0. BJ
Serotype Underlying disease
5: 13-14
Signs and symptoms
Fever; pain and swelling at sites of bony involvement
Fever; pain and swelling of left knee
Sites of bone involvement
Skull, tibias, femurs, ribs, humerus, radius
Distal femur
Sites of positive cultures
Blood, tibia, radius
Knee joint fluid, femur
Serology
Anti-Arizona titer 1:40 Chronic osteomyelitis with low grade inflammation and lipid-filled
Anti-Arizona titer 1: 1280 Chronic osteomyelitis with low grade jnflammation
Pathology
63, W, F 7: 1,2,6 Thrombocytopenic purpura, splenectomy
?Letterer-Siwe (see text)
Case4
Case3 7: 1,2,6 -
-
“Bone lesions”
53, B, M 7a,i’b: 1.7,8 Alcoholic liver disease
Fever; swelling over involved bones
Fever; pain and swelling of involved bones, paraspinous abscess Spine; 12th
Chronic osteomyelitis
histiocytes Therapy
Course and outcome
Penicillin, chloramphenicol, streptomycin and tetracycline for several months; irradiation (200 RI Several months of low grade fever; clinical remission following radiation, chloramphenicol and tetracycline; well 5 mo later
NOTE:
Case 1 from
Fisher
[24j
;
-
Case 2 from
Tetracycline for several months
Ampicillin and gentamicin for six weeks
Afebrile after 3 wk of tetracycline; cultures continued positive from femur; died 5 mo later of bronchopneumonia
Afebrile after 4 wk; well at 5 mo follow-up
Krag and Shean [IO1 ; Case 3 from Edwards
[181
5
2%. t3. F 26: 23-30 Sickle cell anemia
Tibiae, fibulas, femurs, humeri, radius, ulna, metacarpals Blood, bone aspirate
Blood, “bone lesions”, joints, urine
Case
lumbar vertebrae and first lumbar spine Bone lesion, paraspinous, abscess Anti-Arizona titer 1: 1000 Abscess wall with acute and chronic inflammation with masses of lipid-filled histiocytes Ampicillin for 1 Year
Afebrile, well 8 mo follow-up after surgical fusion of spine
; Case4 from Hrubyet al.
[251 ; Case 5.
present case.
ly, Edwards and others [3-5,171 demonstrated distinctive biochemical reactions (Table I) which differentiated Arizona from the genus Salmonella. The accepted name of the organism is now Arizona hinshawii for the single recognized species. Reptiles constitute the main natural reservoir of Arizona hinshawii [5,18-201, and its major economic importance lies in its pathogenicity for poultry with death of the animal or contamination of its eggs and egg products [5,18,20-221. Man has been infected after the ingestion of contaminated eggs, egg powder, ice cream, cream pies, chocolate eclairs and other custards [5,8,18,23] or after exposure to infected reptiles. Arizona organisms are infrequently isolated from the stools of asymptomatic carriers [ 181. Finding Arizona in the stools of patients with diarrhea, however, might not be considered abnormal by some laborato-
580
April 1976
The American Journal of Medicine
Volume 60
ries [5,18] since unlike the most common enteric pathogens (Salmonella and Shigella), Arizona ferments lactose [ 11. In an extensive review of Arizona infections, in which 229 isolates were reported from man [ 181, 59 per cent of these were from stools of patients with diarrhea, gastroenteritis and enteric fever; 15 per cent each were from blood and sites of localized infections: and 11 per cent were from stools-of asymptomatic persons. Clinical Syndromes. As a close biochemical and antigenic relative of Salmonella, it is not surprising that the clinical spectrum of Arizona infection in man closely resembles that produced by Salmonella infection [8,23]. The three most common clinical syndromes caused by Arizona infection are illustrated by the present case: gastroenteritis and enteric fever, septicemia and localized infection [8]. Gastroenteri-
ARIZONA
HINSHAWII OSTEOMYELITIS
WITH FEVER AND SEPSIS-KEREN
ET AL.
tis, which may be seen in any age group, is characterized by cramping abdominal pain, diarrhea (often watery), nausea, vomiting and low grade fever. Diarrhea is usually self-limited and lasts from one to five days [3-5,7,8,18,23]. Septicemia may follow, but also has been reported in the absence of gastroenteritis [4,5]. Localized Arizona infection in such diverse sites as brain, bone, liver, lung, joints and gallbladder is presumed, although not always documented, to follow bacteremia [2,4,5,8,10,13,24]. Arizona Osteomyelitis. The clinical and laboratory data of the three previously well-documented cases of Arizona osteomyelitis are summarized in Table II [ 10,24,25]. A fourth case for which scant clinical information was provided is also included [ 181. Blood cultures grew Arizona organisms in four of the five cases, but only in our case was antecedent Arizona gastroenteritis suggested by history and documented by stool cultures. In Cases 2, 3 and 4 the patients had underlying conditions which may have predisposed them to infection. The multiple sites of infection in Cases 1 and 4 (young, black children) resemble those reported for Salmonella osteomyelitis in patients with hemoglobinopathies [ 28-291, although only in Case 4 did the patient have sickle cell anemia. By demonstrating that the osteomyelitis in our patient was caused by the same Arizona hinshawii serotype which had caused enteric fever and septicemia eight months previously, it appears that the bone was “seeded” during the initial gastroenteritis and an exogenous reinfection mechanism need not be implicated. The 11 month interval between the gastroenteritis and the diagnosis of osteomyelitis illustrates the insidious development of the boney lesion. In this regard, our patient’s course also simulated Salmonella osteomyelitis which is often diagnosed months or years after a primary gastrointestinal infection [ 11,30,31]. Indeed, Salmonella osteomyelitis is often so remote from the episode of Salmonella gastroenteritis that other diagnoses such as Potts disease or pyogenic osteomyelitis are entertained as the primary diagnosis [ 11,3 11. Review of tissue sections from the infected bone in Cases 2” and 3”, together with the description of the histopathology in Case 1 [24], and examination of the present case (Case 5) allow a clear description of the histopathology of Arizona osteomyelitis. In all cases there was a chronic inflammatory response with reactive changes in the bone or surrounding soft tissue. In addition, in Cases 1 and 5 (present case) the patients displayed a xanthomatous
lesion with large numbers of foamy histiocytes, many lymphocytes, and a few polymorphonuclear leukocytes (Figure 2). The presence of lipid-filled histiocytes had prompted the diagnosis of LettererSiwe disease in Case 1. We believe, however, that upon review the histopathology in Case 1 is consistent with Arizona osteomyelitis. Furthermore, the complete recovery in Case 1 with antimicrobial therapy alone and the lack of hepatosplenomegaly make a diagnosis of Letterer-Siwe disease unlikely. Treatment. In Cases 1 and 4 the osteomyelitis resolved with either a combination of chloramphenicol and tetracycline or ampicillin and gentamicin. The role of the radiation therapy in Case 1 is not clear. Therapy in the present case consisted of abscess drainage, vertebral fusion and long-term ampicillin therapy. The antimicrobial susceptibility pattern suggested that a variety of drugs could inhibit in vitro growth of the Arizona organism. The 14 day course of intravenous ampicillin therapy, however, given during the first admission for gastroenteritis, was ineffective in preventing bone infection in its initial stages. Therefore, after the second hospitalization and three weeks of intravenous therapy, a prolonged course of oral ampicillin was selected. Although the patient has had no evidence of infection eight months after discharge, we plan to continue oral ampicillin therapy for one year. As illustrated by the positive postmortem bone lesion culture in Case 2, even a two to three month course with an antibiotic to which the organism is sensitive, may not eradicate the infection. Although there have been reports documenting an increase in osteomyelitis due to gram-negative organisms in recent years, these have not resulted in a consensus regarding the proper type, dose or duration of therapy [32,33]. Although our patient received intravenous ampicillin for three weeks, other investigators recommend treatment for at least four to six weeks with large parenteral doses of antibiotics [32]. We suggest that in cases of bacteremic Arizona infections, prolonged intravenous antimicrobial therapy should be considered initially, and that the patient be closely followed for delayed manifestations of a localized infection. Since it is a more chronic and less destructive disease than osteomyelitis caused by the typical pyogenic organism (staphylococci and streptococci), patients with Arizona osteomyelitis have a more favorable prognosis for recovery [ 301.
Histologic sections kindly supplied by the pathology departments of Santa Clara Valley Medical Center and the University of Illinois.
ADDENDUM Since this manuscript was submitted a case of Artzona hinshawii osteomyelitis occurring in a 23 year
l
April 1976
The American Journal of Medicine
Volume 60
581
ARIZONA HINSHAWII OSTEOMYELITIS WITH FEVER AND SEPSIS-KEREN
ET AL.
old black woman has been reported. This woman had sickle cell trait (hemoglobin AS) and a 10 year history of systemic lupus erythematosus. The osteomyelitis involved her proximal left tibia. In this patient, treatment consisted of surgical decompression and the prolonged administration of intravenous and
oral ampicillin
1.
2.
3. 4. 5.
6. 7.
6. 9.
10. 11. 12.
13.
14.
i5. 16. 17.
18.
582
Edwards PR, Ewing WH: The Genus Arizona. Identification of Enterobacteriaceae, 3rd ed, Minneapolis, Minn, Burgess Publishing Co., 1972. Butt E, Morris JF: Arizona paracolon recovered from middle ear discharge. J infect Dis 91: 283, 1952. Edwards PR: A paracolonlike bacillus isolated from colitis in an infant. J Bacterial 49: 513, 1945. Edwards PR, McWhotter AC, Fife MA: The occurrence of bacteria of the Arizona group in man. Can J Microbial 2: 281.1956. Edwards PR, McWhorter AC, Fife MA: The Arizona groups of enterobacteriaceae. Bull WHO 14: 511, 1958. Edwards PR, Kauffman F. Stuck P: A new Arizona serotype (27:17,20) isolated from man. Acta Pathol Microbial Stand 48: 265, 1960. Ferris AA, Hertzberg R, Atkinson N: An epidemic of diirrhea caused by a new strain of the Salmonella group. Med J Aust 2: 368, 1945. Guckian JC, Byers EH, Perry JE: Arizona infection of man. Arch Intern Med 119: 170, 1967. Hughes MH, Bartlett DI. Baker M, Draeper RE, Rowe B: Gastroenteritis due to Salmonella subgenus Ill (Arizona). J Hyg (Camb) 69: 507. 1971. Krag D, Shean DB: Serious human infections due to bacilli of the Arizona group. Calif Med 90: 230, 1959. Murphy JB: Bone and joint disease in relation to typhoid fever. Surg Gynecol Obstet 23: 119, 1916. Plows CD, Fretwell G. Gerichter CB: An Arizona serotype isolated from a case of gastroenteritis in Britain. J Hyg (Camb) 66: 109, 1968. Seiligmann E, Saphra I, Wassermann M: Occurrence of some unusual Salmonella types in man, including a new type Salmonella Georgia. Am J Hyg 40: 227, 1944. Ericsson HM, Sherris JC: Antibiotic sensitivity testing, report of an international collaborative study. Acta Pathol Microbial Stand [B], Suppl 217, p 1, 1971. Caldwell ME, Ryerson DL: Salmonellosis in certain reptiles. J Infect Dis 85: 242, 1939. Kauffmann F: Uber mehrere meue salmonella-typen. Acta Pathol Microbial Stand 18: 351, 1941. Edwards PR, Cherry WB, Bruner DW: Further studies on coliform bacjlli serologically related to the genus Salmonella. J Infect Dis 73: 229, 1943. Edwards PR, Fife MA, Ramsey CH: Studies on the Arizona group of enterobacteriaceae. Bacterial Rev 23: 155, 1959.
April 1976
The American Journal of Medlclne
Volume 60
for four and 12 weeks, respectively.
ACKNOWLEDGMENT The technical assistance of Raymond E. Lund and Nancy Warner is gratefully acknowledged.
21. 22. 23.
24.
25.
26.
27.
28. 29.
30. 31.
Iveson, JB, MacKay-Scollay EM, Bamford V: Salmonella and Arizona in reptiles and man in Western Australia. J Hyg (Camb) 67: 135, 1967. Sechter I, Altmann G. Cahan D, Gerichter CB: Arizona isolations in Israel (1957-1969). Ann lnst Pasteur (Paris) 119: 323, 1970. Solowey M: Paracolon organisms in spray-dried whole egg powder. J Bacterial 53: 667, 1947. West MG, Edwards PR, Bruner DW: A group of diphasic paracolon bacteria. J Infect Dis 81: 24, 1947. Murphy WJ, Morris JF: Two outbreaks of gastroenteritis apparently caused by a paracolon of the Arizona group. J Infect Dis 86: 255, 1950. Fisher RH: Multiple lesions of bone in Letterer-Siwe disease. Report of a case with culture of paracolon Arizona bacilli from bone lesions and blood, followed.by response to therapy. J Bone Joint Surg 35A: 445, 1953. Hruby MA, Honig GR, Lolekka S, Gotoff SP: Arizona hinshawii osteomyelitis in sickle cell anemia. Am J Dis Child 125: 867, 1973. Ellenbogen NC, Raim J, Grossman L: Salmonella sp. (type montevideo) osteomyelitis. Am J Dis Child 90: 275, 1955. Hook EW, Campbell CO, Weens HS, Cooper GR: Salmonella osteomyelitis in patients with sickle cell anemia. N Engl J Med 257: 403, 1957. Seidenstein H: Salmonella osteomyelitis. Bull Hosp Joint Dis 6: 126, 1945. Silver HK, Simon JL, Clement DH: Salmonella osteomyelitis and abnormal hemoglobin disease. Pediatrics 20: 439, 1957. Black PH. Kung LJ, Swartz MN: Salmonellosis-A review of some unusual aspects. N Engl J Med 262: 811,196O. Giaccia L, tdriss H: Osteomyelitis due to Salmonella infection. J Pediitr 41: 73, 1952.
32.
Wafdvogel FA, Medoff G, Swartz MN: Osteomyelitis: A review of clinical features, therapeutic considerations and unusual aspects. N Engl J Med 282: 198, 260, 316, 1970.
33.
Meyers BR, Berson BL, Gilbert M, Hirschman SZ: Clinical patterns of osteomyelitis due to gram-negative bacteria. Arch Intern Med 131: 228 1973.
34.
Smilack JD, Goldberg MA: Bone and joint involvement with Arizona Hinshawii. Report of a case and review of the literature. Am J Med Sci 270: 503, 1975.
DAVID F. KEREN, M.D. WILLIAM RAWLINGS, Jr., M.D. HENRY W. MURRAY, M.D. WILLIAM R. LEONARD, B.S. Baltimore, Maryland
A case of Arizona osteomyelitis of the spine which occurred 11 months after an episode of gastroenteritis and enteric fever is presented. As a close biochemical and antigenk relative of Salmonella, Arizona infection produces a similar clinical course with gastrointestinal manifestations frequently precedtng localized infections by several months. The boney lesion in the present case and in three of the four other cases of Arizona osteomyelitis described in the literature was a chronic inflammation which may have a xanthomatous component. The bone destruction caused by Arizona infection is less severe than that of tubercuious or pyogenic osteomyelitis. Proposed treatment of Arizona osteomyelitis consists of debridement of the localized infection and long term antimicrobial therapy. Arizona hinshawii, a gram-negative bacillus, is a member of the family Enterobacteriacae and is antigenically closely related to the genus Salmonella [ 11. The clinical spectrum of Arizona infections varies from benign gastroenteritis to enteric fever and septicemia with localized infection and is similar to that caused by Salmonella species [2-131. We report a case illustrating the complete natural history of Arizona osteomyelitis. The causative organism was identified in the patient’s blood and stool during an episode of gastroenteritis and enteric fever 11 months prior to the diagnosis of vertebral osteomyelitis. CASE REPORT First Admission. A 53 year old black man was admitted to The Johns Hopkins Hospital (JHH) in September 1973 because of fever and obtundation. Two weeks and weakness.
From the Departments of Laboratory Medicine and Medicine, The Johns Hopkins Hospital, Baltimore, Maryland 21205. Requests for reprints should be addressed to Dr. David F. Keren, Departrnent of Laboratory Medicine, The Johns Hopkins Hospital, Baltimore, Maryland 21205. Manuscript accepted August 14, 1975.
prior to admission
he had a cough,
During the following week
vomiting and diarrhea.
pleuritic chest pain
he experienced
fever,
nausea,
There was no history of ingestion of contaminated
foods, exposure to sick animals or of contact with companions who were ill. Except for heavy alcohol intake, the patient’s past history was unremarkable. Pertinent physical findings on admission included a rectal temperature of 104’F, a pulse rate of 1 lO/min, basilar rales bilaterally, diffuse abdominal
tenderness
without
April 1976
organomegaly
or
masses,
The American Journal of Medicine
hypoactive
Volume 60
bowel
677
ARIZONA HINSHAWII OSTEOMYELlTlS WlTH FEVER AND SEPSIS-KEREN
Figure I. Lateral view of spine showing destruction of 12th lumbar vertebrae and first lumbar spine vertebral bodies. Soft tissue calcification (arrow) is seen anterior to second lumbar spine.
sounds and a negative stool guaiac. The laboratory data included a hematocrit value of 37 per cent, a leukocyte count of 2,700/mm3, a platelet count of 36,500/mm3, serum glutamic oxaloacetic transaminase @GOT) 250 IU/ liter, serum glutamic pyruvic transaminase (SGPT) 299 IU/ liter, albumin 2.8 g/d! and serum urea nitrogen 66 mg/dl. Chest roentgenogram showed a possible infiltrate in the lower lobe of the left lung and multiple calcified granulomas. The sputum contained gram-positive diplococci and gram-negative rods. Intravenous gentamicin and ampicillin therapy was begun. On the third hospital day, cultures of blood, stool and sputum grew a gram-negative rod that was subsequently identified as Arizona hinshawii serotype 7a,7b: 1,7,8. Cultures of cerebrospinal fluid and bone marrow aspirate were sterile. Since the organism was sensitive to both antimicrobials being given, the administration of gentamicin was discontinued and ampicillin (8 g, intravenously/day) was continued for 14 days. The patient became afebrile by the ninth hospital day. Other studies included a negative purified protein derivative, a normal hemoglobin electrophoresis, a negative glucose-6-phosphate dehydrogenase screen, a normal x-ray series of the upper gastrointestinal tract and small bowel, and a well-visualized gallbladder by intravenous cholangiogram. The patient was discharged on the 20th hospital day with a hematocrit value of 28 per cent and a leukocyte
578
April 1978
The American Journal of Medicine
Volume 80
ET AL.
count of 6,000/mm3. A repeat chest film showed fibrotic changes in the lower lobe of the left lung, x-ray films of the thoracolumbar spine were unremarkable. Six weeks after discharge stool cultures were negative for enteric pathogens. At that time the patient complained of mild low back pain and was given acetaminophen. Second Admission. The patient was readmitted 11 months later, in August 1974, with complaints of several weeks of malaise and a gradually enlarging, painful swelling in his right lower back. There had been no fever, chills, night sweats or back trauma. On physical examination he appeared emaciated with temporal wasting. Oral temperature was 99.4’F. There were two large, tender, fluctuant subcutaneous masses on his back. One was paraspinal, measured 8 by 3 cm and extended from the 12th thoracic vertebrae to the fourth lumbar spine, the other was located over the spinous processes of the 12th thoracic vertebrae and the first lumbar spine measured 2 by 3 cm. No microorganisms were seen on gram stain of the thick green fluid aspirated from the paraspinal abscess. Kinyoun stain of the aspirate on the admitting ward was initially thought to contain acid-fast bacilli although subsequent aspirates failed to confirm this finding. X-ray films of the spine showed a large soft-tissue mass and a marked gibbus deformity of the lower thoracic and upper lumbar spine with destruction of the 12th thoracic vertebrae and first lumbar vertebral bodies (Figure 1). Laboratory data included a hematocrit value of 36 per cent, a leukocyte count of 7, 100/mm3, a platelet count of 215,000/mm3, SGOT IMiter, albumin 3.4 gfdl, and a normal serum urea nitrogen. On the basis of the inltiil aspirate smear and the clinical picture, the patient was presumed to have Pott’s disease (tuberculous spondylitls), and isoniazid and ethambutol therapy was begun. A low grade fever developed on the second hospital day. On the fourth day cultures from the abscess aspirate grew an Arizona species subsequently identified as Arizona hinshawii 7a,7b: 1,7,8 sensitive to ampicillin. The significance of this isolate was not fully appreciated until later. Cultures of the blood and stool were negative. On the 10th hospital day, the patient underwent surgical decompression of the abscesses and anterior fusion of the 12th lumbar vertebrae and first lumbar spine wlth a rib graft. Extensive debridement, of the vertebral bodies was not necessary since relatively little necrotic bone was present. The abscess cavity contained 300 ml of pus from which Arizona hinshawii was again isolated. Histologicalty, the abscess wall showed acute and chronic inflammation with large masses of foam cells and gram-negative rods (Figure 2). No histologic evidence of tuberculosis was present, and all cultures were negative for mycobacteria. Because the patient was a poor historian and the medical records of the previous admission were unavailable, his physicians did not become aware of the prior Arizona hinshawii septicemia and its relationship to the present illness until the 1 lth hospital day. At that time, intravenous ampicillin therapy (8 g/day) was begun. One day later, the patient was afebrite and remained so for the rest of his 21 day hospitalization. Since discharge he has been main-
ARIZONA HINSHAWII OSTEOMYELITIS WITH FEVER AND SEPSIS-KEREN
ET AL.
Figure 2. A, photomicrograph of abscess wall with a predominance of “foamy” histiocyte? (xanthoma cells) (arrows). Hematoxylin and eosin stain, magnification X 270. B, photomicrograph of abscess wall showing gram-negative, rod-shaped organisms (arrow) and “foamy” histiocytes (H). Brown and Hopps bacterial stain, magnification X 1,100.
tained on oral ampicillin (2 g/day), and is ambulatoryand doing well eight months after discharge. Ampicillintherapy will be continuedfor one year. MlCftOBlOLOGY
by the agar dilution technic [ 141. The minimum inhibitory concentrations for the organism were tetracycline 4 pg/ml, chloramphenicol 4 pg/ml, cephalothin 8 pg/ml, ampicillin 2 fig/ml and kanamycin 4 &g/ml.
AND SEROLOGY COMMENTS
The cultures of Arizona hinshawii in the present case
from both hospital admissions (blood, stool, sputum, and paraspinous abscess) produced gram-negative, oxidase-negative bacilli on MacConkey agar (BBL). On triple sugar iron agar the Arizona produced an alkaline slant and an acid butt with hydrogen sulfide production. The organism was motile, indole negative and able to utilize citrate as a carbon source, but it did not produce urease. The preliminary identification of Arizona was made by using a modification of the biochemical tests recommended by Edwards and Ewing [I] for differentiating Arizona from Salmonella (Table I). This was confirmed by a strong agglutination of the isolated organism by Arizona polyvalent 0 antiserums (Difco). The anti-Arizona titer of the patient’s serum (obtained one month after hii second discharge) was 1: 1000. This was performed using the patient’s cultured organism. Serum from a noninfected person was used as a negative control. Antiiicrobial-susceptibility testing was performed
The Arizona organism, first isolated in 1939 from diseased reptiles in Arizona, was initially named “Salmonella Dar-es-Saalam variety from Arizona” because of biochemical similarities to Salmonella Dar-es-Saalam [15]. In 1941, Kauffman [IS] proposed classifying the organism as Salmonella Arizonae because of H (flagellar) antigens similar to those present in some Salmonella serotypes. SubsequentTABLE I
Arizona and Salmonella-Biochemical Differences
Biochemical Test
Arizona
Malonate utilization Gelatin liquification ONPGt (lactose fermentation) Dulcitol fermentation lnositol fermentation ‘(+)
refers to delayed
A* (+I -
v+ ”
positive reaction after three or more days.
+ONPG = 0 nitrophenyl-@-D tion of @-galactosidase. *v refers to variable reactions,
AprH1976
Salmonella
galactopyranoside i.e., +, -,
lhe&nerkanJaumalof&dktine
(BBL) for detec-
or (+).
vobne
69
579
ARIZONA HINSHAWII OSTEOMYELITIS WITH FEVER AND SEPSIS-KEREN
TABLE II
ET AL.
Arizona Osteomyelitis
Case1
Case2
Age, race and sex
2. 0. BJ
Serotype Underlying disease
5: 13-14
Signs and symptoms
Fever; pain and swelling at sites of bony involvement
Fever; pain and swelling of left knee
Sites of bone involvement
Skull, tibias, femurs, ribs, humerus, radius
Distal femur
Sites of positive cultures
Blood, tibia, radius
Knee joint fluid, femur
Serology
Anti-Arizona titer 1:40 Chronic osteomyelitis with low grade inflammation and lipid-filled
Anti-Arizona titer 1: 1280 Chronic osteomyelitis with low grade jnflammation
Pathology
63, W, F 7: 1,2,6 Thrombocytopenic purpura, splenectomy
?Letterer-Siwe (see text)
Case4
Case3 7: 1,2,6 -
-
“Bone lesions”
53, B, M 7a,i’b: 1.7,8 Alcoholic liver disease
Fever; swelling over involved bones
Fever; pain and swelling of involved bones, paraspinous abscess Spine; 12th
Chronic osteomyelitis
histiocytes Therapy
Course and outcome
Penicillin, chloramphenicol, streptomycin and tetracycline for several months; irradiation (200 RI Several months of low grade fever; clinical remission following radiation, chloramphenicol and tetracycline; well 5 mo later
NOTE:
Case 1 from
Fisher
[24j
;
-
Case 2 from
Tetracycline for several months
Ampicillin and gentamicin for six weeks
Afebrile after 3 wk of tetracycline; cultures continued positive from femur; died 5 mo later of bronchopneumonia
Afebrile after 4 wk; well at 5 mo follow-up
Krag and Shean [IO1 ; Case 3 from Edwards
[181
5
2%. t3. F 26: 23-30 Sickle cell anemia
Tibiae, fibulas, femurs, humeri, radius, ulna, metacarpals Blood, bone aspirate
Blood, “bone lesions”, joints, urine
Case
lumbar vertebrae and first lumbar spine Bone lesion, paraspinous, abscess Anti-Arizona titer 1: 1000 Abscess wall with acute and chronic inflammation with masses of lipid-filled histiocytes Ampicillin for 1 Year
Afebrile, well 8 mo follow-up after surgical fusion of spine
; Case4 from Hrubyet al.
[251 ; Case 5.
present case.
ly, Edwards and others [3-5,171 demonstrated distinctive biochemical reactions (Table I) which differentiated Arizona from the genus Salmonella. The accepted name of the organism is now Arizona hinshawii for the single recognized species. Reptiles constitute the main natural reservoir of Arizona hinshawii [5,18-201, and its major economic importance lies in its pathogenicity for poultry with death of the animal or contamination of its eggs and egg products [5,18,20-221. Man has been infected after the ingestion of contaminated eggs, egg powder, ice cream, cream pies, chocolate eclairs and other custards [5,8,18,23] or after exposure to infected reptiles. Arizona organisms are infrequently isolated from the stools of asymptomatic carriers [ 181. Finding Arizona in the stools of patients with diarrhea, however, might not be considered abnormal by some laborato-
580
April 1976
The American Journal of Medicine
Volume 60
ries [5,18] since unlike the most common enteric pathogens (Salmonella and Shigella), Arizona ferments lactose [ 11. In an extensive review of Arizona infections, in which 229 isolates were reported from man [ 181, 59 per cent of these were from stools of patients with diarrhea, gastroenteritis and enteric fever; 15 per cent each were from blood and sites of localized infections: and 11 per cent were from stools-of asymptomatic persons. Clinical Syndromes. As a close biochemical and antigenic relative of Salmonella, it is not surprising that the clinical spectrum of Arizona infection in man closely resembles that produced by Salmonella infection [8,23]. The three most common clinical syndromes caused by Arizona infection are illustrated by the present case: gastroenteritis and enteric fever, septicemia and localized infection [8]. Gastroenteri-
ARIZONA
HINSHAWII OSTEOMYELITIS
WITH FEVER AND SEPSIS-KEREN
ET AL.
tis, which may be seen in any age group, is characterized by cramping abdominal pain, diarrhea (often watery), nausea, vomiting and low grade fever. Diarrhea is usually self-limited and lasts from one to five days [3-5,7,8,18,23]. Septicemia may follow, but also has been reported in the absence of gastroenteritis [4,5]. Localized Arizona infection in such diverse sites as brain, bone, liver, lung, joints and gallbladder is presumed, although not always documented, to follow bacteremia [2,4,5,8,10,13,24]. Arizona Osteomyelitis. The clinical and laboratory data of the three previously well-documented cases of Arizona osteomyelitis are summarized in Table II [ 10,24,25]. A fourth case for which scant clinical information was provided is also included [ 181. Blood cultures grew Arizona organisms in four of the five cases, but only in our case was antecedent Arizona gastroenteritis suggested by history and documented by stool cultures. In Cases 2, 3 and 4 the patients had underlying conditions which may have predisposed them to infection. The multiple sites of infection in Cases 1 and 4 (young, black children) resemble those reported for Salmonella osteomyelitis in patients with hemoglobinopathies [ 28-291, although only in Case 4 did the patient have sickle cell anemia. By demonstrating that the osteomyelitis in our patient was caused by the same Arizona hinshawii serotype which had caused enteric fever and septicemia eight months previously, it appears that the bone was “seeded” during the initial gastroenteritis and an exogenous reinfection mechanism need not be implicated. The 11 month interval between the gastroenteritis and the diagnosis of osteomyelitis illustrates the insidious development of the boney lesion. In this regard, our patient’s course also simulated Salmonella osteomyelitis which is often diagnosed months or years after a primary gastrointestinal infection [ 11,30,31]. Indeed, Salmonella osteomyelitis is often so remote from the episode of Salmonella gastroenteritis that other diagnoses such as Potts disease or pyogenic osteomyelitis are entertained as the primary diagnosis [ 11,3 11. Review of tissue sections from the infected bone in Cases 2” and 3”, together with the description of the histopathology in Case 1 [24], and examination of the present case (Case 5) allow a clear description of the histopathology of Arizona osteomyelitis. In all cases there was a chronic inflammatory response with reactive changes in the bone or surrounding soft tissue. In addition, in Cases 1 and 5 (present case) the patients displayed a xanthomatous
lesion with large numbers of foamy histiocytes, many lymphocytes, and a few polymorphonuclear leukocytes (Figure 2). The presence of lipid-filled histiocytes had prompted the diagnosis of LettererSiwe disease in Case 1. We believe, however, that upon review the histopathology in Case 1 is consistent with Arizona osteomyelitis. Furthermore, the complete recovery in Case 1 with antimicrobial therapy alone and the lack of hepatosplenomegaly make a diagnosis of Letterer-Siwe disease unlikely. Treatment. In Cases 1 and 4 the osteomyelitis resolved with either a combination of chloramphenicol and tetracycline or ampicillin and gentamicin. The role of the radiation therapy in Case 1 is not clear. Therapy in the present case consisted of abscess drainage, vertebral fusion and long-term ampicillin therapy. The antimicrobial susceptibility pattern suggested that a variety of drugs could inhibit in vitro growth of the Arizona organism. The 14 day course of intravenous ampicillin therapy, however, given during the first admission for gastroenteritis, was ineffective in preventing bone infection in its initial stages. Therefore, after the second hospitalization and three weeks of intravenous therapy, a prolonged course of oral ampicillin was selected. Although the patient has had no evidence of infection eight months after discharge, we plan to continue oral ampicillin therapy for one year. As illustrated by the positive postmortem bone lesion culture in Case 2, even a two to three month course with an antibiotic to which the organism is sensitive, may not eradicate the infection. Although there have been reports documenting an increase in osteomyelitis due to gram-negative organisms in recent years, these have not resulted in a consensus regarding the proper type, dose or duration of therapy [32,33]. Although our patient received intravenous ampicillin for three weeks, other investigators recommend treatment for at least four to six weeks with large parenteral doses of antibiotics [32]. We suggest that in cases of bacteremic Arizona infections, prolonged intravenous antimicrobial therapy should be considered initially, and that the patient be closely followed for delayed manifestations of a localized infection. Since it is a more chronic and less destructive disease than osteomyelitis caused by the typical pyogenic organism (staphylococci and streptococci), patients with Arizona osteomyelitis have a more favorable prognosis for recovery [ 301.
Histologic sections kindly supplied by the pathology departments of Santa Clara Valley Medical Center and the University of Illinois.
ADDENDUM Since this manuscript was submitted a case of Artzona hinshawii osteomyelitis occurring in a 23 year
l
April 1976
The American Journal of Medicine
Volume 60
581
ARIZONA HINSHAWII OSTEOMYELITIS WITH FEVER AND SEPSIS-KEREN
ET AL.
old black woman has been reported. This woman had sickle cell trait (hemoglobin AS) and a 10 year history of systemic lupus erythematosus. The osteomyelitis involved her proximal left tibia. In this patient, treatment consisted of surgical decompression and the prolonged administration of intravenous and
oral ampicillin
1.
2.
3. 4. 5.
6. 7.
6. 9.
10. 11. 12.
13.
14.
i5. 16. 17.
18.
582
Edwards PR, Ewing WH: The Genus Arizona. Identification of Enterobacteriaceae, 3rd ed, Minneapolis, Minn, Burgess Publishing Co., 1972. Butt E, Morris JF: Arizona paracolon recovered from middle ear discharge. J infect Dis 91: 283, 1952. Edwards PR: A paracolonlike bacillus isolated from colitis in an infant. J Bacterial 49: 513, 1945. Edwards PR, McWhotter AC, Fife MA: The occurrence of bacteria of the Arizona group in man. Can J Microbial 2: 281.1956. Edwards PR, McWhorter AC, Fife MA: The Arizona groups of enterobacteriaceae. Bull WHO 14: 511, 1958. Edwards PR, Kauffman F. Stuck P: A new Arizona serotype (27:17,20) isolated from man. Acta Pathol Microbial Stand 48: 265, 1960. Ferris AA, Hertzberg R, Atkinson N: An epidemic of diirrhea caused by a new strain of the Salmonella group. Med J Aust 2: 368, 1945. Guckian JC, Byers EH, Perry JE: Arizona infection of man. Arch Intern Med 119: 170, 1967. Hughes MH, Bartlett DI. Baker M, Draeper RE, Rowe B: Gastroenteritis due to Salmonella subgenus Ill (Arizona). J Hyg (Camb) 69: 507. 1971. Krag D, Shean DB: Serious human infections due to bacilli of the Arizona group. Calif Med 90: 230, 1959. Murphy JB: Bone and joint disease in relation to typhoid fever. Surg Gynecol Obstet 23: 119, 1916. Plows CD, Fretwell G. Gerichter CB: An Arizona serotype isolated from a case of gastroenteritis in Britain. J Hyg (Camb) 66: 109, 1968. Seiligmann E, Saphra I, Wassermann M: Occurrence of some unusual Salmonella types in man, including a new type Salmonella Georgia. Am J Hyg 40: 227, 1944. Ericsson HM, Sherris JC: Antibiotic sensitivity testing, report of an international collaborative study. Acta Pathol Microbial Stand [B], Suppl 217, p 1, 1971. Caldwell ME, Ryerson DL: Salmonellosis in certain reptiles. J Infect Dis 85: 242, 1939. Kauffmann F: Uber mehrere meue salmonella-typen. Acta Pathol Microbial Stand 18: 351, 1941. Edwards PR, Cherry WB, Bruner DW: Further studies on coliform bacjlli serologically related to the genus Salmonella. J Infect Dis 73: 229, 1943. Edwards PR, Fife MA, Ramsey CH: Studies on the Arizona group of enterobacteriaceae. Bacterial Rev 23: 155, 1959.
April 1976
The American Journal of Medlclne
Volume 60
for four and 12 weeks, respectively.
ACKNOWLEDGMENT The technical assistance of Raymond E. Lund and Nancy Warner is gratefully acknowledged.
21. 22. 23.
24.
25.
26.
27.
28. 29.
30. 31.
Iveson, JB, MacKay-Scollay EM, Bamford V: Salmonella and Arizona in reptiles and man in Western Australia. J Hyg (Camb) 67: 135, 1967. Sechter I, Altmann G. Cahan D, Gerichter CB: Arizona isolations in Israel (1957-1969). Ann lnst Pasteur (Paris) 119: 323, 1970. Solowey M: Paracolon organisms in spray-dried whole egg powder. J Bacterial 53: 667, 1947. West MG, Edwards PR, Bruner DW: A group of diphasic paracolon bacteria. J Infect Dis 81: 24, 1947. Murphy WJ, Morris JF: Two outbreaks of gastroenteritis apparently caused by a paracolon of the Arizona group. J Infect Dis 86: 255, 1950. Fisher RH: Multiple lesions of bone in Letterer-Siwe disease. Report of a case with culture of paracolon Arizona bacilli from bone lesions and blood, followed.by response to therapy. J Bone Joint Surg 35A: 445, 1953. Hruby MA, Honig GR, Lolekka S, Gotoff SP: Arizona hinshawii osteomyelitis in sickle cell anemia. Am J Dis Child 125: 867, 1973. Ellenbogen NC, Raim J, Grossman L: Salmonella sp. (type montevideo) osteomyelitis. Am J Dis Child 90: 275, 1955. Hook EW, Campbell CO, Weens HS, Cooper GR: Salmonella osteomyelitis in patients with sickle cell anemia. N Engl J Med 257: 403, 1957. Seidenstein H: Salmonella osteomyelitis. Bull Hosp Joint Dis 6: 126, 1945. Silver HK, Simon JL, Clement DH: Salmonella osteomyelitis and abnormal hemoglobin disease. Pediatrics 20: 439, 1957. Black PH. Kung LJ, Swartz MN: Salmonellosis-A review of some unusual aspects. N Engl J Med 262: 811,196O. Giaccia L, tdriss H: Osteomyelitis due to Salmonella infection. J Pediitr 41: 73, 1952.
32.
Wafdvogel FA, Medoff G, Swartz MN: Osteomyelitis: A review of clinical features, therapeutic considerations and unusual aspects. N Engl J Med 282: 198, 260, 316, 1970.
33.
Meyers BR, Berson BL, Gilbert M, Hirschman SZ: Clinical patterns of osteomyelitis due to gram-negative bacteria. Arch Intern Med 131: 228 1973.
34.
Smilack JD, Goldberg MA: Bone and joint involvement with Arizona Hinshawii. Report of a case and review of the literature. Am J Med Sci 270: 503, 1975.
