Actinomyces israelii is a gram positive, anaerobic filamentous bacteria often found in the normal oral cavity and upper respiratory tract of man. Under certain conditions, this commensal bacterium can become pathogenic, most often causing chronic abscesses in the cervicofacial region or in the abdominal cavity. It is an extremely fastidious organism which may be difficult to culture. Pelvic actinomycosis has been a recently recognized complication of intrauterine device placement.1'2 We report here such a case resulting in osteomyelitis of the sacrum, sepsis and complete destruction of the hip joint, and ultimately death from endocarditis and septic emboli.
M. H. was an asthenic 26-year old ballet student who was in good health until five months prior to admission. At that time, she came under the care of a nutritionist and naturopath for the treatment of diarrhea and was treated with dietary restriction and vitamin supplements over the ensuing four months. Anorexia and weight loss became manifest throughout this time.
The patient initially presented to the emergency room with a chief complaint of bilateral buttock pain. The patient's temperature was 1030F. Physical examination revealed bilateral, exquisitely tender buttock masses measuring 20 by 25 centimeters (R) and 15 by 10 centimeters (L). She was taken to the operating room where incision and drainage was performed with the expression of a large amount of creamy, purulent material and necrotic debris. Cultures obtained at this time later grew out Escherichia coli, Peptococcus asacrolyticus, Peptostreptococcus intermedius, and Bacteroides fragilis.
Gynecologic examination performed under anesthesia revealed a hard, "woody" pelvis and the presence of a Dalkon Shield intrauterine device which was removed. The patient was maintained on intravenous antibiotics and total parenteral nutrition.
Two weeks later an exploratory laparotomy was performed revealing a posterior perforation of the uterus and a large intrapelvic abscess. Postoperatively, the patient continued to spike fevers to 103°F. in spite oí intravenous Penicillin and Gentamicin therapy. Cultures obtained at the time of surgery revealed Bacteroides fragilis and Peptostreptococcus anaerobius. She required three subsequent debridements of the buttock and sacral region, culminating in subtotal excision of the sacrum. Gross inspection of the excised specimen revealed the presence of sulfur granules along its borders, cultures of which subsequently were positive For Actinomyces israelii. The patient was maintained on high dose parenteral Penicillin and additionally received courses of Chforamphenicol, CarbenicilUn and Tobramycin.
Five weeks subsequent to admission, the patient complained of left hip pain with adduction and internal rotation of the leg, Radiologie studies (Fig. 1) revealed posterior dislocation of the femoral head. Surgical exploration was performed, demonstrating a posterior perforation of the hip joint capsule, a large amount of necrotic material replacing the joint space and complete destruction of the femoral head. Fig. 2) Specimens sent for culture grew E. coli, and no sulfur granules were seen, The patient required three additional debridements of the hip with subsequent cultures positive for Bacteroides fragilis and Cryptococcus asacrolyticus. A complete diagnostic evaluation was undertaken for ongoing signs of sepsis; all tests, including an echocardiogram, were interpreted as normal.
Fig. 1: Five weeks after admission, the patient first complained of left hip pain. These films were obtained prior to surgical debridement.
The patient's clinical status steadily improved. Ten weeks after admission, all antibiotics were discontinued without recurrence of fever. She was discharged twenty-two weeks after admission in an anabolic state.
Fig. 2:This operative photograph was obtained at the time of initial hip joint debridement. Most of the femora/ head and a portion of the acetabulum have been destroyed. The probe is in the acetabufum.
Five weeks after discharge, the patient sustained a cardiopulmonary arrest. She was pronounced dead on arrival at the emergency room. Postmortem examination demonstrated numerous vegetations replacing the cusps of the tricuspid valve (Fig. 3) and a large embolus blocking the orifice of the pulmonary arterial main trunk. Cultures of the vegetations revealed Candida albicans. The remainder of pathologic examination showed no evidence of recurrent pelvic, abdominal or osseous disease and no additional areas of metastatic involvement.
Actinomycosis is a chronic, suppurative infection caused by an opportunistic pathogen. While the mechanism of primary infection is often controversial, contraceptive devices have been implicated in playing a major role in pelvic disease.1'2 The type of device is of little importance. Schiffler, et al., have suggested that a pathway of infection is established from the anus across the perineum, and thus into the pelvis by virtue of endometrial erosion caused by the intrauterine device. Furthermore, the string may act as a wick. In none of their ten patients with intrapelvic disease did infection spread beyond the reproductive tract.
One of the hallmarks of anaerobic infections in general, and actinomycosis in particular, is superinfection with other organisms.4'5 This tendency may be abetted in the immunocompromised host, as in the present case, where we suspect sustained malnutrition may have played an important role in the spread of disease. Not only is the actinomycete fastidious in its culture requirements,3*6 but even when sought the recovery rate from an established, known infection may be quite low. In the present case, actinomycosis MÎIS cultured from the sacrum but noi the necrotic material removed at the time of hip joint debridement. It should be noted the patient was receiving high doses of Penicillin at the time. Sulfur granules need not be present in all foci of infection, and were not found in the patient's hip. We isolated E. coli, Bacteroides fragilis and Cryptococcus asacroiyticus as components of a typical "mixed" infection/'5 In the absence of a high index of suspicion, the isolation of concurrent microorganisms may serve to obfuscate the underlying, principal source of infection and so lead the clinician astray. Treatment of the secondary infections, which are often of the gram negative variety, is not likely to result in cure of the actinomycosis, This is especially true if an aminoglycoside agent is relied upon.
Fig. 3: The heart removed at the time of postmortem. /Vofe the exuberant, friable vegetations replacing the cusps of the tricuspid valve. Microscopìe and bactériologie evaluation revealed Candida albicans.
Even a known diagnosis does not guarantee successful treatment oí actinomycosis. We performed a total of nine debridements of the hip and pelvis and treated the patient for a total of ten weeks with antibiotic therapy aimed at both the actinomycete and the superinfecting organisms. By this time, however, the heart valves had already been seeded by another superinfecting agent (Candida albicane).
Actinomycotic osteomyelitis has been recognized in the past as affecting mainly the mandible and the vertebral column.8'9 In the preantìbiotic era, most such cases were felt to be secondary to contiguous spread from adjoining cervical, thoracic, or abdominal disease.4 As the advent of Penicillin therapy greatly reduced the frequency of bony involvement, so may the increased usage of intrauterine devices set the stage for intrapelvic disease with osseous sequelae. Physicians must be alert to the possibility of actinomycosis in superinfected cases and the propensity for both local and metas ta tic spread. Early surgical debridement10 and Penicillin therapy3'* remain the mainstays of treatment.
1. Laslo H. Actinomycosis of the small pelvis simulating rectal tumor, caused by intrauterine loop. Orvosi Helilap 118(24): 1415-1417, 1977.
2. Schiffler MD, Elguezabal A, Sultana M, et al. Actinomycosis infections associated with intrauterine contraceptive devices. Obstet Gynecol 45:67-72, 1975.
3. Beradi RS. Abdominal actinomycosis. Surg Gynecol Obstet 149:257-266, 1979.
4. Lewis RP, Sutter VL, Fînegold S. Bone infections involving anaerobic bacteria. Medicine 57:279-301, 1978.
5. Raff MJ, Melo JC. Anaerobic osteomyelitis. Medicine 57:83103, 1978.
6. Cuiten CH, Sharp ME. Infections of wounds with actirtorriyces. J Bone Joint Surg 338:221-227, 1951.
7. Proppe KH. Clinicopathologic conference. N Engl 1 Med 299(2):85-92, 1978.
8. Brett MD. Advanced actinomycosis of the spine. J Bone Joint Surg 23B-.215-220, 1951.
9. Cope VZ, Actinomycosis of bone with special reference to the vertebral column. J Bone Joint Surg 338:205-214, 1951.
10. Wangensteen OH. The role of surgery in the treatment of actinomycosis. Ann Surg 1936, 104:752-770.