Infections Due to Non-O1 Vibrio cholerae in Southern Taiwan: Predominance in Cirrhotic Patients

Wen-Chien Ko 0 1 2 Yin-Ching Chuang 0 1 2 Guang-Chang Huang 0 1 2 Shiu-Yuan Hsu 0 1 2 0 To identify cases of vibrio infections at the National Cheng Kung University Hospital (NCKUH) , Tainan , Taiwan , we re- viewed the records of the microbiological laboratory from Oc- tober 1989 to October 1997. The medical records of patients with non - O1 V. cholerae infections were screened. Clinical data including demographic characteristics, underlying dis- eases, results of laboratory studies, treatment course, and clini- cal outcome were collected. Because of the retrospective nature of this study, culture of seafood or environmental water was not performed 1 Received 29 January 1998; revised 26 May 1998. Medicine, National Cheng Kung University Hospital , 138 Sheng Li Road, Tainan , Taiwan 2 From the Department of Internal Medicine, National Cheng Kung University Hospital, and the Department of Medicine, National Cheng Kung University Medical College, Tainan; and the Southern Branch Laboratory, National Institute of Preventive Medicine, Department of Health , Kaohsiung , Taiwan Although Taiwan is not an area where cholera is endemic, from October 1988 to October 1997 30 episodes of non - O1, non - O139 Vibrio cholerae infection were noted at the National Cheng Kung University Hospital in Taiwan. Infections generally occurred in hot seasons, and two episodes were concomitant with Vibrio vulnificus infection. Three major clinical presentations were found: bacteremia with concurrent spontaneous bacterial peritonitis or invasive soft-tissue infections that occurred solely in cirrhotic patients; self-limited acute febrile gastroenteritis that occurred in patients with no underlying medical disease; and necrotizing fasciitis or cellulitis that often resulted from a wound on extremities. Other manifestations included fatal pneumonitis in a drowned man and acute pyosalpinx. The differential diagnosis of invasive infections in cirrhotic patients should include infections due to non-O1 V. cholerae or V. vulnificus, and a third-generation cephalosporin and a tetracycline analogue or a fluoroquinolone alone is recommended for treatment of severe vibrio infections. - Currently, there are at least 139 serogroups of Vibrio cholerae [1]. V. cholerae O1 has been well known to be the etiologic agent of cholera, an epidemic or pandemic diarrheal disease. Only the cholera toxin producing bacteria, including classical and El Tor biotypes of serogroup O1 or O139, have the potential to cause cholera, and they rarely cause extraintestinal infection. Strains of V. cholerae not agglutinating with O1 antiserum are referred to as non-O1 or nonagglutinating V. cholerae. They are morphologically and biochemically indistinguishable from serogroup O1 [2]. Isolates of V. cholerae have been recognized to be heterogeneous for pathogenicity for humans. As is typical for serotype O1, non-O1 V. cholerae may cause sporadic cases and occasional outbreaks of diarrheal diseases [2], but non-O1 V. cholerae can lead to invasive extraintestinal diseases, often occurring in immunocompromised persons [3, 4]. Suggestive clinical clues for V. cholerae infections include travel, seawater or freshwater exposure, or ingestion of raw seafood [5]. Because Taiwan is a subtropical island, a natural environment suitable for the growth of V. cholerae, and an area where chronic hepatitis B virus infection is hyperendemic, there are many people with postnecrotic cirrhosis who are susceptible to invasive infections due to V. cholerae. To elucidate the clinical presentations of and the morbidity and mortality caused by non O1 V. cholerae infections in Taiwan, a retrospective case survey in a university hospital was conducted. Twenty-six isolates of V. cholerae were available for in vitro antibiotic susceptibility testing by means of the Kirby-Bauer method with use of Mueller-Hinton agar. Drugs tested included the following (mg per disk): gentamicin (10), amikacin (30), ampicillin (10), piperacillin (100), ampicillin/sulbactam (10/ 10), ticarcillin/clavulanate (75/10), piperacillin/tazobactam (100/10), cephalothin (30), cefuroxime (30), cefixime (5), ceftriaxone (30), cefotaxime (30), ceftazidime (30), imipenem (10), norfloxacin (10), ofloxacin (5), ciprofloxacin (5), tetracycline (30), minocycline (30), chloramphenicol (30), and trimethoprim-sulfamethoxazole (1.25/23.75). The guidelines of the performance procedures and the breakpoint diameters for interpretation of the National Committee for Clinical Laboratory Standards were used [6]. The identification of V. cholerae was as follows. Curved gram-negative bacteria with positive oxidase reaction, b-hemolysis on a blood agar plate, growth of yellow colonies on thiosulfate citrate bile salts sucrose agar, susceptibility to 10 mg and 150 mg of vibriostatic agent O-129 (Rosco, Tastrup, Denmark), tolerability to 1% salt solution, and intolerability to 10% salt solution were identified as V. cholerae. Typical biochemistry profiles for V. cholerae included positive nitrate reduction, indole production, catalase production, citrate utilization, Voges-Proskauer test, ornithine decarboxylase production, and lysine decarboxylase production and negative arginine dihydrolase production and urea hydrolysis [7, 8]; these characteristics were noted by the API 20E System (bioMerieux Vitek, Hazelwood, MO) or by traditional biochemistry reactions in test tubes. Serotyping was determined with use of O1 antiserum (Difco Laboratories, Detroit) and O139 antiserum (Denka Seiken, Tokyo). Serotyping of all strains with use of O139 antiserum was performed at the Southern Branch Laboratory, National Institute of Preventive Medicine, Department of Health, Kaohsiung, Taiwan, a reference laboratory for cholera toxin producing V. cholerae. Definitions of Terminology Leukocytosis was defined as a total leukocyte count of 12,000/mm 3, and thrombocytopenia was defined as a platelet count of 100,000/mm 3. Early mortality was defined as death within 96 hours after visiting the hospital, and late mortality was defined as death occurring after at least 96 hours of hospitalization and before discharge from the hospital. The degree of decompensation of hepatic cirrhosis was assessed by the scoring system proposed by Pugh et al. [9] that is based on serum albumin and bilirubin levels, prolongation of prothrombin time, amount of ascites, and degree of hepatic encephalopathy. Results During an 8-year period at NCKUH, 30 V. cholerae isolates from 30 different patients were identified. All isolates were non-O1, non-O139 serotypes; these identifications were confirmed by the reference laboratory. Infections due to these isolates generally occurred in hot-weather months, especially from May to September (figure 1). In vitro antibiotic susceptibility testing by the disk diffusion method showed that non-O1 V. cholerae was susceptible to many drugs, including combinations of a b-lactam agent and a b-lactamase inhibitor, br (...truncated)