Portal de Periódicos da CAPES

OCHROBACTRUM ANTHROPI BACTEREMIA ASSOCIATED WITH A CATHETER IN AN IMMUNOCOMPROMISED CHILD AND REVIEW OF THE PEDIATRIC LITERATURE

1999; Lippincott Williams & Wilkins; Volume: 18; Issue: 7 Linguagem: Inglês

10.1097/00006454-199907000-00023

1532-0987

Jesús Saavedra, Carmen Garrido, María Dolores Folgueira, Maria Jose Torres, José Tomás Ramos,

Streptococcal Infections and Treatments

Ochrobactrum anthropi is a nonfermenting Gram-negative rod identified as a pathogenic microorganism during the last decade.1 Since then there has been an increase in the number of infections caused by this organism, mostly in immunocompromised hosts and patients with indwelling catheters. The first case described in pediatrics was osteochondritis of the foot in a 14-year-old boy in 1987.2 Cieslak et al.3 reported the first child with catheter-related bacteremia in 1992. Since then few infections have been described in children, with only two younger than 7 years. We describe a case of a catheter-related bacteremia in a child with abdominal neuroblastoma after an autologous bone marrow transplant. Case report. A 4-year-old boy developed O. anthropi catheter-related bacteremia in June, 1993, 6 months after autologous bone marrow transplant as consolidation treatment of a Stage III abdominal neuroblastoma. The patient was taking trimethoprim-sulfamethoxazole (TMP-SMX) as prophylaxis. He had had a Hickman-type central catheter since the transplant without problems. The child was evaluated in the clinic because of episodes of fever (up to 38.5°C) and vomiting. The physical examination was normal and the catheter site did not appear infected. The laboratory analysis showed mild neutropenia (1.09 × 109/l) with normal values in the rest of the hemogram. The absolute neutrophil count had been normal after recovery from neutropenia associated with the transplant. Blood cultures were obtained from central and peripheral veins. Therapy was started with ceftazidime and amikacin, and he defervesced 24 h later. O. anthropi resistant to both antibiotics was isolated from all of the blood cultures, and antibiotics were changed to imipenem. Despite this change O. anthropi continued to be isolated from several blood cultures for the next 2 days, and the Hickman catheter was removed. After that neither O. anthropi nor other pathogens were isolated from the blood cultures. The patient received 14 days of imipenem therapy with a good outcome. Discussion.O. anthropi is a Gram-negative, motile, aerobic, oxidase-positive and non-lactose-fermenting rod with strong urease activity. It grows well on McConkey agar medium. Previously known as CDC group Vd, in 1988 it was classified as an independent group and is the only species of this group.4 Its differentiation from other Gram-negative rods is difficult. The growth of Pseudomonas aeruginosa is similar to that of O. anthropi; however, unlike P. aeruginosa, O. anthropi does not produce pigment. Flavobacterium spp. also produce a pigment, thus distinguishing them from O. anthropi. In Acinetobacter and Flavimonas the positive reaction for oxidase in O. anthropi makes the distinction possible. Special tests are required to distinguish O. anthropi from other oxidase-positive species belonging to the former Achromobacter genus (Group B Achromobacter, Alcaligenes genus and Agrobacterium spp.). The differentiation is based on the O-nitrophenyl-D-galactosidase test.4 There are semiautomatic systems now available with the O-nitrophenyl-D-galactosidase test, and they have made O. anthropi detection and differentiation easier in recent years. This may also be one of the reasons for increased reporting of O. anthropi infection during the last few years. We used the PASCO system (Difco, Detroit, MI) for the identification of O. anthropi. O. anthropi is ubiquitous and has been isolated from water and soil, as well as several hospital environments.3-5 It has not been clearly identified as human saprophytic flora. In this regard O. anthropi is similar to Alcaligenes and Pseudomonas spp.3O. anthropi as a human pathogen has been found in clinical devices and has been isolated from blood, urine, wounds, stool, throat and vagina.5-10 The most frequent infection produced by O. anthropi is central venous catheter-related bacteremia.3, 5, 8, 11-14 Other infections described during the last few years are pancreatic abscess,1 osteochondritis,2 necrotizing fasciitis,15 meningitis due to a pericardial graft,6, 7 endophthalmitis secondary to intraocular lens implantation10 and thoracic wall abscess after pacemaker placement.9 O. anthropi infection is not very aggressive in the majority of cases; in general the outcome is good7 although there have been cases of septic shock.15 Our patient had mild clinical findings as previously described in a child with a catheter-related infection.12 Few cases have been described in children (Table 1), of whom 5 (>50%) had a catheter-related bacteremia. In the pediatric literature reviewed all patients recovered from the illness regardless of the clinical appearance or the source of the infection.TABLE 1: Pediatric cases of Ochrobactrum anthropi infections Outbreaks have been described related to contaminated conservative liquids6 and iv solutions.14 Three of 8 cases of O. anthropi infection in children in the literature we reviewed were produced by a contaminated conservative liquid of a pericardial graft.6, 7 O. anthropi generally affects immunocompromised hosts3-5, 11, 12, 14 and patients with clinical devices. In more than one-half of the cases described in children, the patients were immunocompromised. One patient had osteomyelitis of the foot from a puncture wound. The rest of the children had an indwelling device in place. The most common devices involved are central catheters; however, grafts, pacemakers, intraocular lenses and others may become infected. O. anthropi has a capacity to adhere to various materials used frequently in clinical settings. Staphylococcus spp. have similar characteristics,5 which may explain the unusually high association of these organisms with foreign bodies in humans. There is no consensus about which is the best treatment for infections caused by O. anthropi. Some authors have stated that it can be cured by the withdrawal of the catheter or device without any other treatment,8, 13 and in other cases the outcome has been favorable with antibiotic treatment alone without catheter removal.3, 11 It is probable that every situation must be individualized with regard to the clinical situation and the device the patient has in place. In our case as in others previously described5-7, 9, 10, 12 antibiotic treatment alone was not sufficient, and we had to withdraw the device before the patient could be cured. In the 60% of the pediatric cases in which the catheter was the cause of the infection it was removed. In one child removal of the catheter was the only intervention performed.8 Our patient along with the child reported by Klein et al.12 were the only children who were treated with both antibiotics and catheter removal. Contrary to other Gram-negative rods (including Achromobacter xylosoxidans) O. anthropi is usually resistant to beta-lactams, such as broad spectrum penicillins and third generation cephalosporins3, 5, 7, 8, 10-14, as well as aztreonam.3, 5, 8, 12, 14 In general O. anthropi is susceptible to imipenem and aminoglycosides,3, 5, 7-10, 12, 14, 15 TMP-SMX,3, 8, 9, 12, 13, 15 ciprofloxacin5, 7-10, 12, 15 and tetracyclines.5, 10, 11 Nonetheless several cases have been published of O. anthropi infections with resistance to aminoglycosides3, 7, 10, 11 and TMP-SMX.5, 7, 10, 11 Combination of these antibiotics has been suggested in the management of this infection.3, 8, 12 Some authors have recommended ciprofloxacin as the best treatment5 and others suggested TMP-SMX.3 The real importance of the susceptibility data in the treatment of this infection is unclear. In the case presented by Haditsch et al.11 the child was cured by antibiotic treatment with piperacillin, flucloxacillin and gentamicin without removal of the catheter, even when the organism was not susceptible to the antibiotics that were used. That situation may indicate a low virulence of O. anthropi or that the susceptibility patterns cannot predict the outcome in patients. The authors comment that the good outcome could have been partly due to the broad spectrum antibiotic treatment used, which was also appropriate for P. aeruginosa.11 In our case O. anthropi was resistant to TMP-SMX most likely because of the long duration of prophylaxis this patient was receiving with this antibiotic. Resistance to TMP-SMX has been reported by Hadistch et al.11 and other authors. In contrast Klein et al.12 reported that the organism was susceptible to this antimicrobial even when the patient was receiving TMP-SMX prophylaxis. In our case O. anthropi was also resistant to amikacin, which is very uncommon although it has been described previously.7 In children imipenem, gentamicin and TMP-SMX were the most frequent antibiotics used. Acknowledgment. We thank Dr. Cynthia McCoig for her help with preparation of the manuscript. Jesus Saavedra, M.D. Carmen Garrido, M.D. Dolores Folgueira, M.D. Maria Jose Torres, M.D. Jose Tomas Ramos, M.D. Departments of Pediatrics (JS, CG, MJT, JTR) and Microbiology (DF); Hospital Materno-Infantil Doce de Octubre; Madrid, Spain