271
Vibrio vulnificus Infection in Taiwan: Report of 28 Cases and Review of Clinical Manifestations and Treatment Yin-Ching Chuang, Chung-Yin Yuan, Cheng-Yi Liu, Chi-Kin Lan, and Amy Hwei-Mei Huang
From the Department of Internal Medicine, National Cheng Kung University Hospital, Tainan; the Department ofSurgery, Yuan's General Hospital, Kaohsiung; the Department of Internal Medicine, Veterans General Hospital-Taipei; the Department of Internal Medicine, ChangGung Memorial Hospital-Kaohsiung; and the Department ofPathology, Father Fox Memorial Hospital, Tainan, Taiwan, Republic of China
There has been a dramatic increase in the number of reported cases of infection due to Vibrio vulnificus in Taiwan [1-4]. Although the organism has been etiologically implicated in a variety of clinical syndromes [5-7], most cases of V. vulnificus infection either are associated with wound infections or are categorized as primary septicemia, unlike ingestive infections, which are caused by other organisms including other vibrios [8]. Cases of infections such as gastroenteritis, pneumonia, meningitis, and endometritis that are due to V. vulnificus have also been reported [7, 911]. Most of the cases are associated with underlying diseases, and the clinical course is generally characterized by the acute onset of fever, the appearance of bullous skin lesions, and the development of hypotension and shock. The mortality rate for primary septicemia is as high as 56% [5]. A high index of suspicion for this relatively unusual disease plus aggressive appropriate antimicrobial and surgical therapy should result in better response rates.
Materials and Methods Isolates were identified with use of standard laboratory practices. Susceptibility testing was done by the Kirby-Bauer
Received 27 December 1991 ~ revised 2 March 1992. Reprints or correspondence: Dr. Yin-Ching Chuang, Section of Infectious Diseases, Department of Internal Medicine, National Cheng Kung University Hospital, Tainan, Taiwan 70428, Republic of China.
Clinical Infectious Diseases
1992;15:271-6
© 1992 by The University of Chicago. All rights reserved. 1058-4838/92/1502-0007$02.00
method. Bacteriologic reports were reviewed, and information on cases involving V. vulnificus infections was collected at five hospitals: Veterans General Hospital-Taipei (TVGH), National Cheng Kung University Hospital (CKUH), Yuan's General Hospital (YGH), Chang-Gung Memorial Hospital-Kaohsiung (CGMH), and Father Fox Memorial Hospital (FFMH). TVGH is in Taipei, the largest city and capital, which is in northern Taiwan. The other hospitals are in southern and more rural settings where fish aquaculture in ponds is common. Both inpatients and outpatients were included in the study, and no special methods were used in isolating V. vulnificus. From May 1985 through July 1990, 28 episodes of V. vulnificus infection were found and the hospital records of 27 patients were reviewed: one patient had two episodes of V. vulnificus infection.
Results Clinical manifestations. The age of the patients at the time of diagnosis ranged from 19 to 76 years (median, 61 years). The male-to-female ratio was 2: 1. There has been a steady increase in the number of cases encountered and reported in recent years (figure 1). All of the cases occurred during the summer (figure 2). Eleven episodes were categorized as primary bacteremia (table 1) with secondary skin manifestations; seven episodes were manifested as wound infections and bacteremia; eight episodes were wound infections without bacteremia; one was pneumonia; and one was gastroenteritis. One patient developed a primary skin lesion following injury in a fish pond and responded to therapy; another episode of skin disease occurred 1 year later that was
Downloaded from http://cid.oxfordjournals.org/ at University of Sussex on August 25, 2015
From May 1985 through July 1990, 28 episodes of Vibrio vulni.ficus infection in 27 patients were encountered in five major hospitals in Taiwan. The ages of patients ranged from 19 to 76 years; the ratio of male to female patients was 2:1. Eighteen episodes manifested as bacteremia and eight as wound infections alone. One patient each developed gastroenteritis and pneumonia after nearly drowning. Twenty-three patients exhibited skin manifestations. Twenty patients had underlying diseases. All patients were treated with antibiotics, and 14 also underwent some form of surgical treatment (incision and drainage, fasciotomy, debridement, or amputation). Thirteen of the 28 episodes were preceded by precipitating factors; most were due to ingestion of seafood or exposure of abraded skin to salt water. Ten of the 18 septicemic patients died-most within 48 hours of hospitalization. One patient without bacteremia who had a wound infection died. Results of in vitro susceptibility studies suggested that ampicillin or a third-generation cephalosporin would be effective. Susceptibility to aminoglycosides was observed for >90% of isolates. We recommend combined therapy with a third-generation cephalosporin or ampicillin and an aminoglycoside along with appropriate surgical therapy for the treatment of V. vulni.ficus infection.
ern 1992;15 (August)
Chuang et al.
272
10
8 rtI ' ~
' rt1
8
Figure 1. Number ofcases of Vibrio vulnificus infection encountered from May
to U
~
' 0
1985 through July 1990 in five major hospitals in Taiwan, Republic ofChina. There were a total of28 episodes in 27 patients.
4
0
I
z
I
2 -
not associated with injury. Twenty-three patients exhibited skin manifestations ranging from erythema and painful swelling to formation of hemorrhagic bullae, gangrenous changes , and necrotizing fasciitis or pyomyositis that resulted in amputation. Twenty patients had underlying diseases: 13 had chronic liver disease, two had chronic renal failure, three had gouty arthritis, and two had diabetes mellitus. Four had been taking corticosteroids. No apparent underlying diseases could be found in seven patients. Patients in 14 of the 28 episodes experienced overt precipitating event s; patients in 13 episodes had a history of ingestion of seafood or exposure ofopen wounds to salt water. Two ofthe four cases in Taipei, in which the causes could be traced through history, were associated with ingestion of raw seafood . One patient developed pneumonia after nearly drowning. Patients in 15 episodes had striking skin manife stations (figure 3); thus, V. vulnificus infection was suspected on emergency admission by the house officers.
Treatment and outcome. All of the 28 infections in 27 patients were treated with antibiotics, and 16 patients also underwent some form of surgical intervention. Fasciotomy was performed in five episodes, debridement in eight episodes, and incision and drainage in another two; two patients who manifested severe wound infection of the limbs underwent amputation in addition to receiving aggressive antimicrobial therapy. Ten of the 18 septicemic patients diedmost within 48 hours of hospitalization-whereas only one patient with advanced cirrhosis, who had a wound infection and severe fasciitis without bacteremia, died. He refused surgery and left the hospital against advice . Among the bactere mic patients, the skin manifestations appeared within 3-4 days of the onset of fever and within I day of hospitalization.
Discussion and Review Taiwan is a small island located in southeastern Asia with a population of more than 20 million people, among whom
~
to
U
~
o
.
o
z
Figure 2. Seasonal distribution of episodes of Vibrio vulnificus infection from selected hospitals in Taiwan, Republ ic of China, from May 1985 to July 1990. The total number of episodes was 28.
Downloaded from http://cid.oxfordjournals.org/ at University of Sussex on August 25, 2015
Year
Table 1. Summary of data for 28 cases of Vibrio vulnificus infection. Age
(y)/sex
Case no.
Skin manifestation
Wound infection with bacteremia Necrotizing fasciitis (both 63/F legs) 2 63/M Cellulitis with gangrenous change over left thigh Cellulitis. left lower leg 3 57/F 4 65/M Wound infection. left hand; bullae. necrosis Necrotizing fasciitis, left 5 51/M leg 6 76/M Necrotizing fasciitis, left hand 7
56/M
Pyomyositis, left leg, hemorrhagic bullae
13
73/F
14
231M
15
71/F
16 17 (see case 25)
65/M 62/M
18
351M
Ecchymosis. swelling, both legs Multiple bullae Hemorrhagic bullae. right leg Not evident Hemorrhagic bullae left hand. right leg Hemorrhagic bullae, swelling, both legs
Wound infection without bacteremia 19 75/M Bullae. right leg
Alcoholic liver disease
Antibiotics
Causes
Outcome
Duration of hospitalization
32
Moxalactam
Debridement
Recovered
Cephradine, amikacin, clindamycin Penicillin Ceftazidime, amikacin
None
Died Recovered Died*
Ceftazidime, amikacin
None Fasciotomy, debridement None
Died
Chloramphenicol
Amputation
Recovered
Alcoholic liver disease
Injured while working in fish pond! Fisherman
Ceftazidime. clindamycin
Debridement
Died
bacteremia) Chronic hepatitis
Not evident
Oxacillin. gentamicin
None
Died
Cirrhosis Cirrhosis Hepatoma Iatrogenic Cushing's syndrome
Not evident Not evident Ate raw seafood Not evident
None Debridement None Debridement
Died Died Recovered Recovered
Cirrhosis
Not evident
Ampicillin Penicillin Cefazolin Penicillin. gentamicin. clindamycin, cephalothin Penicillin
None
Died
Alcoholic liver disease
Ate raw oysters
Minocycline,
None
Recovered
34
None
Recovered
55
None Debridement. amputation
Recovered Died*
22 27
Not evident Not evident Steroid administration Cirrhosis, chronic renal failure Gout. chronic renal failure
Injury associated with seafood None
Surgical treatment
Not evident Injury associated with shrimp Not evident
amikacin Ampicillin. gentamicin Cefazolin, gentamicin Chloramphenicol. ceftazidime
Diabetes mellitus
Not evident
Chronic liver disease Gout. chronic renal failure. previous V. vulnificus infection Alcoholic liver cirrhosis
Not evident Not evident
Not evident
Cephalothin. gentamicin. piperacillin
None
Died
Not evident
Injury in fish
Chloramphenicol
Fasciotomy
Recovered
2 18 2 32
2
2 3 10 38
28
pond!
20
74/M
21
61/M
22
74/M
23
50/F
24
49/M
25 (see case 17) 26
61/M 60/M
Necrotizing fasciitis, left leg; hemorrhagic bullae Bullae. right forearm, right leg Cellulitis. left hand swelling
Cirrhosis
Injury by fish
Chloramphenicol
Fasciotomy
Died*
Gout, steroid administration Gout, steroid administration
Injury in fish
Ceftazidime
Debridement
Recovered
72
Debridement
Recovered
32
Cellulitis with lymphangitis, left foot Necrotizing fasciitis right hand. forearm; hemorrhagic bullae Necrotizing fasciitis right leg, forearm; bullae Necrotizing fasciitis left forearm; hemorrhagic bullae
Not evident
Injury by crab
Chloramphenicol. tobrarnycin, oxacillin Penicillin
Recovered
12
Diabetes mellitus. alcoholism
Injury by crab
Penicillin. gentamicin
Incision and drainage None
Gout, chronic renal failure Not evident
Injury in fish
Penicillin. oxacillin, tobramycin Chloramphenicol. tobramycin
Fasciotomy, debridement Fasciotomy, debridement
Recovered
72
Recovered
25
3
pond!
Not evident
pond!
Exposure of abraded skin to salt water
Recovered
Gastroenteritis 27
22/F
Not evident
Not evident
Not evident
None
None
Recovered
0
Pneumonia 28
341M
Not evident
Not evident
Near-drowning in salt water
Cefoxitin, tobramycin
None
Died
2
NOTE.
Cases 17 and 25 occurred in the same patient.
* Improved initially. but died of aspiration pneumonia-induced acute respiratory distress syndrome. All were from a saltwater pond. Improved initially. but died of complication of perforated peptic ulcer. § Shock on admission. Amputation was suggested. but patient's family refused to give consent. t
t
Downloaded from http://cid.oxfordjournals.org/ at University of Sussex on August 25, 2015
Bacteremia with secondary skin mainfestation (primary 8 69/M Toxic epidermal necrolysis, multiple bullae 9 311M Not evident 10 Bullae. both legs 36/F II 56/M Not evident 12 Bullae, right hand 71/F
Underlying disease
274
Chuang et al.
cm 1992;15 (August)
Figure 3. Marked erythematous swelling and bullous formation were not ed over the left forearm of the pat ient in case 17 on adm ission.
nal illness. characterized by vomiting, diarrhea, or abdom inal pain and a stool culture positive for V. vulnificus, negative blood cultures, and no evidence of wound infection. Cases of other infections, such as pneumonia (table I. patient 28) and endometritis, have also been reported [7, 9. 10]. Recurr ent skin lesions have not been described . On e of our patients (episodes 17 and 25) had an initial episode associated with a leg-wound infection following trauma in a fish pond . Th ere was also a bullous lesion on the arm , presumably due to bacteremia. One year later. bullo us lesion s occurred at the sites of previous lesion s of the arm and leg and resulted in the patient's death. The first episode was clearly associated with a wound infection. The second episode of infection could have occurred secondarily to ingestion ofcontaminated seafood and bacteremia. but this would be unlikely since the lesion s were at the same sites as were the original lesions. In reviewing the literature. we found 95 cases of prima ry bacteremia and 72 cases of wound infection caused by V. vulnificus. With the addition of II cases of prima ry bacteremia and 15 cases ofwound infection in this series, 193 cases were available for anal ysis. Full clinical details about the patients were not always available (table 2). For the patients with primary bacterem ia. 89 (94.7%) of 94 had underlying diseases. especially chronic liver diseases. Other und erlying conditions included chronic renal failure. gouty arthritis. diabetes mellitus, and chronic administration ofcorticosteroids. Cau ses of infection could be found in 41 (78. 8%)of 52 cases. All of them were related to seafood consumption. The incubation periods were very short. and the most striking clinical man ifestations were skin lesion s. The cutaneous lesions included ecchymosis, bullous for-
Downloaded from http://cid.oxfordjournals.org/ at University of Sussex on August 25, 2015
seafood is extremely popular. The island also has a very high rate of hepatitis B. and cirrhosis of the liver is often seen. Preexisting liver diseases have been considered risk factors associated with infections due to V. vulnificus. Seafood may be the most important vehicle for the entry of V. vulnificus into the human body. although exposure throu gh aquaculture in fish ponds is also a risk factor. In a number of instances. despite direct que stioning, the patients did not disclose a risk factor. Saltwater or freshwater fish prepared in home s and restaurants are eaten daily by most people living in Taiwan, and ingestion of raw or undercooked fish as well as minor trauma during food preparation could be overlooked by many patients. Consumption ofraw oysters is relatively rare, although other fish eaten raw as sashimi is common. The reason(s) for the increased incidence of V. vulnificus infections over the past decade is not clear. After the first report in 1987 [I] and subsequent reports [2, 3] in 1989, clinicians may have been alerted to the clinical syndrome, but laboratories should have isolated the organism in routine blood and wound cultures. Only a large-scale epidemiological study could clarity this issue. V. vulnificus characteristically produces three syndromes [5, 12]: ( I) prima ry sepsis with high fever and chills without an app arent focus of infection (usuall y, V. vulnificus is acquired through the gastrointestinal tract after ingestion of seafood contaminated by this organism); (2) wound infection with resulting cellulitis caused by direct inoculation of the microorganism. which may result in tissue necrosis and secondary bacteremia (usually associated with exposure of abraded skin to salt water or injuries associated with the cultivation and/or preparation of seafood ; and (3) gastrointesti-
ity rate tends to be higher if therapy is delayed in patients with septicemia. We have suggested that although the microorganisms and other halophilic vibrios may be sensitive to many antibiotics [13, 17-21], third-generation cephalosporins should be the drugs of choice for early antimicrobial therapy [2]. In a disease with such a high mortality rate (25%-56%), it would seem prudent to add an aminoglycoside, at least during the early phase of the infection. Ampicillin with or without an aminoglycoside seems a reasonable alternative. Trimethoprim-sulfamethoxazole (co-trimoxa-
Table 3. Summary of results ofdisk sensitivity tests for 28 clinical isolates of Vibrio vulnificus.
Antibiotic Penicillin Ampicillin Carbenicillin Piperacillin Tetracycline Chloramphenicol Trimethoprim-sulfamethoxazole Gentamicin Tobramycin Amikacin Colistin Cephalothin Cefamandole Cefoxitin Moxalactam Ceftriaxone Cefotaxime Ceftazidime
No. of sensitive isolates/ no. of tested isolates (%) 1/2(50) 25/26* (96) 9/9 (100) 12/12 (100) .20/22 (91) 24/26 (92) 8/8 (100) 25/26 (96) 21/23 (91) 21/23 (91)
0/7 (0) 18/26 22/24 9/15 10/11 22/22 28/28 18/18
(69) (92) (60) (91) (100) (100) (100)
NOTE. Isolates were tested at the time of isolation in the five different hospitals. * One isolate had intermediate susceptibility by the disk sensitivity test.
Downloaded from http://cid.oxfordjournals.org/ at University of Sussex on August 25, 2015
mation, necrotizing fasciitis, gangrenous changes, cellulitis, and pyomyositis, which in some cases required amputation. The skin manifestations usually developed at the time of admission or within 24 hours of hospitalization and became more severe by the hour. Secondary skin lesions were found on 65.1 %(69 of 106) ofthe patients with primary septicemia. The clinical course of septicemic patients was fulminant. Surgical procedures were performed on 37. 1% (13 of 35) of the patients. The mortality rate was as high as 54.2% (51 of 94). Among those patients who developed hypotension within 12 hours ofadmission, the mortality rate reached 90% [5, 13]. Most of our patients in this category died within 48 hours of hospitalization. More than one-half of cases of wound infection (44 of 80) were associated with underlying diseases. For all but seven cases (91 %) causes for the infection were determined. All of these patients had injuries due to fish bites or seafood handling or wounds that were exposed to salt water. In 30 of 69 cases, secondary bacteremia developed. Surgical procedures were done in more than half of the cases (19 of 37). The mortality rate was low (25.3%) compared with that for patients with primary bacteremia (54.2%). The sporadic occurrence of V. vulnificus infection in humans makes it highly unlikely that a clinical trial could be conducted for determining the optimal antimicrobial therapy for eradication of the organisms and the optimal therapeutic regimen for treatment of these patients. A review of our susceptibility studies and correlation of the results with clinical outcome reveal both successes with resistant organisms and failures with sensitive isolates (table 3). Surgical intervention confounds the interpretation of the results in these cases as well as the evaluation of the outcome in cases reported in the literature. Klontz et al. [5] emphasized the importance of beginning aggressive antibiotic treatment at the time of earliest suspicion ofinfection because the mortal-
276
Chuang et al.
zole) has excellent (100%) in vitro efficacy, as determined by the disk susceptibility test. Because of the widespread obliterative vasculitis and vascular necrosis seen in skin lesions, penetration of antibiotics to the affected area may be impaired. For this reason, early surgical debridement should be added to the therapeutic regimen [22]. In addition, since most patients died within 48 hours of hospitalization, a decision within the first 24 hours on removal of the infected areas (including limbs) that appear to be refractory to antimicrobial therapy is critical.
Acknowledgements
References I. Yuan CY. Yuan cc, Wei DC. Lee AM. Septicemia and gangrenous change of the legs caused by marine vibrio, Vibrio l'ulnificus-report of a case. J Formosan Med Assoc 1987;86:448-51. 2. Chuang yc. Young C. Chen CWo Vibrio vulnificus infection. Scand J Infect Dis 1989;21:721-6. 3. Chuang yc. Young CD. Vibrio vulnificus infections: clinical experience with 7 cases [abstract]. In: Proceedings of the Formosan Medical Association Annual Meeting 1989. Taipei, Taiwan: Formosan Medical Association, 1989: 114-5. 4. Chuang yc. Vibrio vulnificus infection in Taiwan [abstract]. In: Program and abstracts of the 2nd Western Pacific Congress on Infectious Diseases and Chemotherapy. Jomtien-Pattaya, Thailand: Infectious Disease Association ofThailand and Western Pacific Society of'Chernotherapy, 1990: 108. 5. Klontz KC. Lieb S, Schreiber M, Janowski HT, Baldy LM, Gunn RA. Syndromes of Vibrio vulnificus infections: clinical and epidemiologic features in Florida cases, 1981-1987. Ann Intern Med 1988;109: 318-23.
6. Bonner JR, Coker AS, Berryman CR, Pollock HM. Spectrum of Vibrio infections in a gulfcoast community. Ann Intern Med 1983;99:4649. 7. Blake PA, Merson MH, Weaver RE, Hollis DG, Heublein Pc. Diseases caused by a marine vibrio: clinical characteristics and epidemiology. N Engl J Med 1979;300: 1-5. 8. Hughes JM, Merson MH, Gangarosa EJ. The safety of eating shellfish. JAMA 1977;237:1980-1. 9. Kelly MT, Avery DM. Lactose-positive Vibrio in seawater: a cause of pneumonia and septicemia in a drowning victim. J Clin Microbiol 1980; II :278-80. 10. Tison DL, Kelly MT. Vibrio vulnificus endometritis. J Clin Microbiol 1984;20: 185-6. II. Johnston JM, Becker SF, McFarland LM. Gastroenteritis in patients with stool isolates of Vibrio vulnificus. Am J Med 1986;80:336-8. 12. Hoffmann TJ, Nelson B, Darouiche R, Rosen T. Vibrio vulnificus septicemia. Arch Intern Med 1988;148:1825-7. 13. Tacket CO, Brenner F, Blake PA. Clinical features and an epidemiological study of Vibrio vulnificus infections. J Infect Dis 1984; 149:55861. 14. Johnston JM, Becker SF, McFarland LM. Vibrio vulnificus: man and the sea. JAMA 1985;253:2850-3. 15. Howard RJ, Pessa ME, Brennaman BH, Ramphal R. Necrotizing softtissue infections caused by marine vibrios. Surgery 1985;98: 126-30. 16. Howard RJ, Lieb S. Soft-tissue infections caused by halophilic marine vibrios. Arch Surg 1988;123:245-9. 17. Joseph SW, DeBell RM, Brown WP. In l'itroresponse to chloramphenicol, tetracycline, ampicillin, gentamicin, and beta-lactamase production by halophilic vibrios from human and environmental sources. Antimicrob Agents Chemother 1978; 13:244-8. 18. Bowdre JH, Hull JH, Cocchetto DM. Antibiotic efficacy against Vibrio vulnificus in the mouse: superiority of tetracycline. J Pharmacol Exp Ther 1983;225:595-8. 19. Morris JG Jr, Tenney J. Antibiotic therapy for Vibrio vulnificus infection. JAMA 1985;253: 1121-2. 20. Morris JG Jr, Black RE. Cholera and other vibrioses in the United States. N Engl J Med 1985;312:343-50. 21. Arnold M, Woo ML, French GL. Vibrio vulnificus septicaemia presenting as spontaneous necrotising cellulitis in a woman with hepatic cirrhosis. Scand J Infect Dis 1989;21 :727-31. 22. Zielinski CJ, Bora FW Jr. Vibrio hand infections: a case report and review of the literature. J Hand Surg 1984;9A:754-7.
Downloaded from http://cid.oxfordjournals.org/ at University of Sussex on August 25, 2015
The authors express their gratitude to Dr. Kun-Yen Huang and Dr. Donald Armstrong for their critical reviewof this paper, to Dr. Ping-I Hsu for collecting patients' data, and to Miss JiaRong Shieh for preparing the manuscript.
CID 1992; 15 (August)
Vibrio vulnificus Infection in Taiwan: Report of 28 Cases and Review of Clinical Manifestations and Treatment Yin-Ching Chuang, Chung-Yin Yuan, Cheng-Yi Liu, Chi-Kin Lan, and Amy Hwei-Mei Huang
From the Department of Internal Medicine, National Cheng Kung University Hospital, Tainan; the Department ofSurgery, Yuan's General Hospital, Kaohsiung; the Department of Internal Medicine, Veterans General Hospital-Taipei; the Department of Internal Medicine, ChangGung Memorial Hospital-Kaohsiung; and the Department ofPathology, Father Fox Memorial Hospital, Tainan, Taiwan, Republic of China
There has been a dramatic increase in the number of reported cases of infection due to Vibrio vulnificus in Taiwan [1-4]. Although the organism has been etiologically implicated in a variety of clinical syndromes [5-7], most cases of V. vulnificus infection either are associated with wound infections or are categorized as primary septicemia, unlike ingestive infections, which are caused by other organisms including other vibrios [8]. Cases of infections such as gastroenteritis, pneumonia, meningitis, and endometritis that are due to V. vulnificus have also been reported [7, 911]. Most of the cases are associated with underlying diseases, and the clinical course is generally characterized by the acute onset of fever, the appearance of bullous skin lesions, and the development of hypotension and shock. The mortality rate for primary septicemia is as high as 56% [5]. A high index of suspicion for this relatively unusual disease plus aggressive appropriate antimicrobial and surgical therapy should result in better response rates.
Materials and Methods Isolates were identified with use of standard laboratory practices. Susceptibility testing was done by the Kirby-Bauer
Received 27 December 1991 ~ revised 2 March 1992. Reprints or correspondence: Dr. Yin-Ching Chuang, Section of Infectious Diseases, Department of Internal Medicine, National Cheng Kung University Hospital, Tainan, Taiwan 70428, Republic of China.
Clinical Infectious Diseases
1992;15:271-6
© 1992 by The University of Chicago. All rights reserved. 1058-4838/92/1502-0007$02.00
method. Bacteriologic reports were reviewed, and information on cases involving V. vulnificus infections was collected at five hospitals: Veterans General Hospital-Taipei (TVGH), National Cheng Kung University Hospital (CKUH), Yuan's General Hospital (YGH), Chang-Gung Memorial Hospital-Kaohsiung (CGMH), and Father Fox Memorial Hospital (FFMH). TVGH is in Taipei, the largest city and capital, which is in northern Taiwan. The other hospitals are in southern and more rural settings where fish aquaculture in ponds is common. Both inpatients and outpatients were included in the study, and no special methods were used in isolating V. vulnificus. From May 1985 through July 1990, 28 episodes of V. vulnificus infection were found and the hospital records of 27 patients were reviewed: one patient had two episodes of V. vulnificus infection.
Results Clinical manifestations. The age of the patients at the time of diagnosis ranged from 19 to 76 years (median, 61 years). The male-to-female ratio was 2: 1. There has been a steady increase in the number of cases encountered and reported in recent years (figure 1). All of the cases occurred during the summer (figure 2). Eleven episodes were categorized as primary bacteremia (table 1) with secondary skin manifestations; seven episodes were manifested as wound infections and bacteremia; eight episodes were wound infections without bacteremia; one was pneumonia; and one was gastroenteritis. One patient developed a primary skin lesion following injury in a fish pond and responded to therapy; another episode of skin disease occurred 1 year later that was
Downloaded from http://cid.oxfordjournals.org/ at University of Sussex on August 25, 2015
From May 1985 through July 1990, 28 episodes of Vibrio vulni.ficus infection in 27 patients were encountered in five major hospitals in Taiwan. The ages of patients ranged from 19 to 76 years; the ratio of male to female patients was 2:1. Eighteen episodes manifested as bacteremia and eight as wound infections alone. One patient each developed gastroenteritis and pneumonia after nearly drowning. Twenty-three patients exhibited skin manifestations. Twenty patients had underlying diseases. All patients were treated with antibiotics, and 14 also underwent some form of surgical treatment (incision and drainage, fasciotomy, debridement, or amputation). Thirteen of the 28 episodes were preceded by precipitating factors; most were due to ingestion of seafood or exposure of abraded skin to salt water. Ten of the 18 septicemic patients died-most within 48 hours of hospitalization. One patient without bacteremia who had a wound infection died. Results of in vitro susceptibility studies suggested that ampicillin or a third-generation cephalosporin would be effective. Susceptibility to aminoglycosides was observed for >90% of isolates. We recommend combined therapy with a third-generation cephalosporin or ampicillin and an aminoglycoside along with appropriate surgical therapy for the treatment of V. vulni.ficus infection.
ern 1992;15 (August)
Chuang et al.
272
10
8 rtI ' ~
' rt1
8
Figure 1. Number ofcases of Vibrio vulnificus infection encountered from May
to U
~
' 0
1985 through July 1990 in five major hospitals in Taiwan, Republic ofChina. There were a total of28 episodes in 27 patients.
4
0
I
z
I
2 -
not associated with injury. Twenty-three patients exhibited skin manifestations ranging from erythema and painful swelling to formation of hemorrhagic bullae, gangrenous changes , and necrotizing fasciitis or pyomyositis that resulted in amputation. Twenty patients had underlying diseases: 13 had chronic liver disease, two had chronic renal failure, three had gouty arthritis, and two had diabetes mellitus. Four had been taking corticosteroids. No apparent underlying diseases could be found in seven patients. Patients in 14 of the 28 episodes experienced overt precipitating event s; patients in 13 episodes had a history of ingestion of seafood or exposure ofopen wounds to salt water. Two ofthe four cases in Taipei, in which the causes could be traced through history, were associated with ingestion of raw seafood . One patient developed pneumonia after nearly drowning. Patients in 15 episodes had striking skin manife stations (figure 3); thus, V. vulnificus infection was suspected on emergency admission by the house officers.
Treatment and outcome. All of the 28 infections in 27 patients were treated with antibiotics, and 16 patients also underwent some form of surgical intervention. Fasciotomy was performed in five episodes, debridement in eight episodes, and incision and drainage in another two; two patients who manifested severe wound infection of the limbs underwent amputation in addition to receiving aggressive antimicrobial therapy. Ten of the 18 septicemic patients diedmost within 48 hours of hospitalization-whereas only one patient with advanced cirrhosis, who had a wound infection and severe fasciitis without bacteremia, died. He refused surgery and left the hospital against advice . Among the bactere mic patients, the skin manifestations appeared within 3-4 days of the onset of fever and within I day of hospitalization.
Discussion and Review Taiwan is a small island located in southeastern Asia with a population of more than 20 million people, among whom
~
to
U
~
o
.
o
z
Figure 2. Seasonal distribution of episodes of Vibrio vulnificus infection from selected hospitals in Taiwan, Republ ic of China, from May 1985 to July 1990. The total number of episodes was 28.
Downloaded from http://cid.oxfordjournals.org/ at University of Sussex on August 25, 2015
Year
Table 1. Summary of data for 28 cases of Vibrio vulnificus infection. Age
(y)/sex
Case no.
Skin manifestation
Wound infection with bacteremia Necrotizing fasciitis (both 63/F legs) 2 63/M Cellulitis with gangrenous change over left thigh Cellulitis. left lower leg 3 57/F 4 65/M Wound infection. left hand; bullae. necrosis Necrotizing fasciitis, left 5 51/M leg 6 76/M Necrotizing fasciitis, left hand 7
56/M
Pyomyositis, left leg, hemorrhagic bullae
13
73/F
14
231M
15
71/F
16 17 (see case 25)
65/M 62/M
18
351M
Ecchymosis. swelling, both legs Multiple bullae Hemorrhagic bullae. right leg Not evident Hemorrhagic bullae left hand. right leg Hemorrhagic bullae, swelling, both legs
Wound infection without bacteremia 19 75/M Bullae. right leg
Alcoholic liver disease
Antibiotics
Causes
Outcome
Duration of hospitalization
32
Moxalactam
Debridement
Recovered
Cephradine, amikacin, clindamycin Penicillin Ceftazidime, amikacin
None
Died Recovered Died*
Ceftazidime, amikacin
None Fasciotomy, debridement None
Died
Chloramphenicol
Amputation
Recovered
Alcoholic liver disease
Injured while working in fish pond! Fisherman
Ceftazidime. clindamycin
Debridement
Died
bacteremia) Chronic hepatitis
Not evident
Oxacillin. gentamicin
None
Died
Cirrhosis Cirrhosis Hepatoma Iatrogenic Cushing's syndrome
Not evident Not evident Ate raw seafood Not evident
None Debridement None Debridement
Died Died Recovered Recovered
Cirrhosis
Not evident
Ampicillin Penicillin Cefazolin Penicillin. gentamicin. clindamycin, cephalothin Penicillin
None
Died
Alcoholic liver disease
Ate raw oysters
Minocycline,
None
Recovered
34
None
Recovered
55
None Debridement. amputation
Recovered Died*
22 27
Not evident Not evident Steroid administration Cirrhosis, chronic renal failure Gout. chronic renal failure
Injury associated with seafood None
Surgical treatment
Not evident Injury associated with shrimp Not evident
amikacin Ampicillin. gentamicin Cefazolin, gentamicin Chloramphenicol. ceftazidime
Diabetes mellitus
Not evident
Chronic liver disease Gout. chronic renal failure. previous V. vulnificus infection Alcoholic liver cirrhosis
Not evident Not evident
Not evident
Cephalothin. gentamicin. piperacillin
None
Died
Not evident
Injury in fish
Chloramphenicol
Fasciotomy
Recovered
2 18 2 32
2
2 3 10 38
28
pond!
20
74/M
21
61/M
22
74/M
23
50/F
24
49/M
25 (see case 17) 26
61/M 60/M
Necrotizing fasciitis, left leg; hemorrhagic bullae Bullae. right forearm, right leg Cellulitis. left hand swelling
Cirrhosis
Injury by fish
Chloramphenicol
Fasciotomy
Died*
Gout, steroid administration Gout, steroid administration
Injury in fish
Ceftazidime
Debridement
Recovered
72
Debridement
Recovered
32
Cellulitis with lymphangitis, left foot Necrotizing fasciitis right hand. forearm; hemorrhagic bullae Necrotizing fasciitis right leg, forearm; bullae Necrotizing fasciitis left forearm; hemorrhagic bullae
Not evident
Injury by crab
Chloramphenicol. tobrarnycin, oxacillin Penicillin
Recovered
12
Diabetes mellitus. alcoholism
Injury by crab
Penicillin. gentamicin
Incision and drainage None
Gout, chronic renal failure Not evident
Injury in fish
Penicillin. oxacillin, tobramycin Chloramphenicol. tobramycin
Fasciotomy, debridement Fasciotomy, debridement
Recovered
72
Recovered
25
3
pond!
Not evident
pond!
Exposure of abraded skin to salt water
Recovered
Gastroenteritis 27
22/F
Not evident
Not evident
Not evident
None
None
Recovered
0
Pneumonia 28
341M
Not evident
Not evident
Near-drowning in salt water
Cefoxitin, tobramycin
None
Died
2
NOTE.
Cases 17 and 25 occurred in the same patient.
* Improved initially. but died of aspiration pneumonia-induced acute respiratory distress syndrome. All were from a saltwater pond. Improved initially. but died of complication of perforated peptic ulcer. § Shock on admission. Amputation was suggested. but patient's family refused to give consent. t
t
Downloaded from http://cid.oxfordjournals.org/ at University of Sussex on August 25, 2015
Bacteremia with secondary skin mainfestation (primary 8 69/M Toxic epidermal necrolysis, multiple bullae 9 311M Not evident 10 Bullae. both legs 36/F II 56/M Not evident 12 Bullae, right hand 71/F
Underlying disease
274
Chuang et al.
cm 1992;15 (August)
Figure 3. Marked erythematous swelling and bullous formation were not ed over the left forearm of the pat ient in case 17 on adm ission.
nal illness. characterized by vomiting, diarrhea, or abdom inal pain and a stool culture positive for V. vulnificus, negative blood cultures, and no evidence of wound infection. Cases of other infections, such as pneumonia (table I. patient 28) and endometritis, have also been reported [7, 9. 10]. Recurr ent skin lesions have not been described . On e of our patients (episodes 17 and 25) had an initial episode associated with a leg-wound infection following trauma in a fish pond . Th ere was also a bullous lesion on the arm , presumably due to bacteremia. One year later. bullo us lesion s occurred at the sites of previous lesion s of the arm and leg and resulted in the patient's death. The first episode was clearly associated with a wound infection. The second episode of infection could have occurred secondarily to ingestion ofcontaminated seafood and bacteremia. but this would be unlikely since the lesion s were at the same sites as were the original lesions. In reviewing the literature. we found 95 cases of prima ry bacteremia and 72 cases of wound infection caused by V. vulnificus. With the addition of II cases of prima ry bacteremia and 15 cases ofwound infection in this series, 193 cases were available for anal ysis. Full clinical details about the patients were not always available (table 2). For the patients with primary bacterem ia. 89 (94.7%) of 94 had underlying diseases. especially chronic liver diseases. Other und erlying conditions included chronic renal failure. gouty arthritis. diabetes mellitus, and chronic administration ofcorticosteroids. Cau ses of infection could be found in 41 (78. 8%)of 52 cases. All of them were related to seafood consumption. The incubation periods were very short. and the most striking clinical man ifestations were skin lesion s. The cutaneous lesions included ecchymosis, bullous for-
Downloaded from http://cid.oxfordjournals.org/ at University of Sussex on August 25, 2015
seafood is extremely popular. The island also has a very high rate of hepatitis B. and cirrhosis of the liver is often seen. Preexisting liver diseases have been considered risk factors associated with infections due to V. vulnificus. Seafood may be the most important vehicle for the entry of V. vulnificus into the human body. although exposure throu gh aquaculture in fish ponds is also a risk factor. In a number of instances. despite direct que stioning, the patients did not disclose a risk factor. Saltwater or freshwater fish prepared in home s and restaurants are eaten daily by most people living in Taiwan, and ingestion of raw or undercooked fish as well as minor trauma during food preparation could be overlooked by many patients. Consumption ofraw oysters is relatively rare, although other fish eaten raw as sashimi is common. The reason(s) for the increased incidence of V. vulnificus infections over the past decade is not clear. After the first report in 1987 [I] and subsequent reports [2, 3] in 1989, clinicians may have been alerted to the clinical syndrome, but laboratories should have isolated the organism in routine blood and wound cultures. Only a large-scale epidemiological study could clarity this issue. V. vulnificus characteristically produces three syndromes [5, 12]: ( I) prima ry sepsis with high fever and chills without an app arent focus of infection (usuall y, V. vulnificus is acquired through the gastrointestinal tract after ingestion of seafood contaminated by this organism); (2) wound infection with resulting cellulitis caused by direct inoculation of the microorganism. which may result in tissue necrosis and secondary bacteremia (usually associated with exposure of abraded skin to salt water or injuries associated with the cultivation and/or preparation of seafood ; and (3) gastrointesti-
ity rate tends to be higher if therapy is delayed in patients with septicemia. We have suggested that although the microorganisms and other halophilic vibrios may be sensitive to many antibiotics [13, 17-21], third-generation cephalosporins should be the drugs of choice for early antimicrobial therapy [2]. In a disease with such a high mortality rate (25%-56%), it would seem prudent to add an aminoglycoside, at least during the early phase of the infection. Ampicillin with or without an aminoglycoside seems a reasonable alternative. Trimethoprim-sulfamethoxazole (co-trimoxa-
Table 3. Summary of results ofdisk sensitivity tests for 28 clinical isolates of Vibrio vulnificus.
Antibiotic Penicillin Ampicillin Carbenicillin Piperacillin Tetracycline Chloramphenicol Trimethoprim-sulfamethoxazole Gentamicin Tobramycin Amikacin Colistin Cephalothin Cefamandole Cefoxitin Moxalactam Ceftriaxone Cefotaxime Ceftazidime
No. of sensitive isolates/ no. of tested isolates (%) 1/2(50) 25/26* (96) 9/9 (100) 12/12 (100) .20/22 (91) 24/26 (92) 8/8 (100) 25/26 (96) 21/23 (91) 21/23 (91)
0/7 (0) 18/26 22/24 9/15 10/11 22/22 28/28 18/18
(69) (92) (60) (91) (100) (100) (100)
NOTE. Isolates were tested at the time of isolation in the five different hospitals. * One isolate had intermediate susceptibility by the disk sensitivity test.
Downloaded from http://cid.oxfordjournals.org/ at University of Sussex on August 25, 2015
mation, necrotizing fasciitis, gangrenous changes, cellulitis, and pyomyositis, which in some cases required amputation. The skin manifestations usually developed at the time of admission or within 24 hours of hospitalization and became more severe by the hour. Secondary skin lesions were found on 65.1 %(69 of 106) ofthe patients with primary septicemia. The clinical course of septicemic patients was fulminant. Surgical procedures were performed on 37. 1% (13 of 35) of the patients. The mortality rate was as high as 54.2% (51 of 94). Among those patients who developed hypotension within 12 hours ofadmission, the mortality rate reached 90% [5, 13]. Most of our patients in this category died within 48 hours of hospitalization. More than one-half of cases of wound infection (44 of 80) were associated with underlying diseases. For all but seven cases (91 %) causes for the infection were determined. All of these patients had injuries due to fish bites or seafood handling or wounds that were exposed to salt water. In 30 of 69 cases, secondary bacteremia developed. Surgical procedures were done in more than half of the cases (19 of 37). The mortality rate was low (25.3%) compared with that for patients with primary bacteremia (54.2%). The sporadic occurrence of V. vulnificus infection in humans makes it highly unlikely that a clinical trial could be conducted for determining the optimal antimicrobial therapy for eradication of the organisms and the optimal therapeutic regimen for treatment of these patients. A review of our susceptibility studies and correlation of the results with clinical outcome reveal both successes with resistant organisms and failures with sensitive isolates (table 3). Surgical intervention confounds the interpretation of the results in these cases as well as the evaluation of the outcome in cases reported in the literature. Klontz et al. [5] emphasized the importance of beginning aggressive antibiotic treatment at the time of earliest suspicion ofinfection because the mortal-
276
Chuang et al.
zole) has excellent (100%) in vitro efficacy, as determined by the disk susceptibility test. Because of the widespread obliterative vasculitis and vascular necrosis seen in skin lesions, penetration of antibiotics to the affected area may be impaired. For this reason, early surgical debridement should be added to the therapeutic regimen [22]. In addition, since most patients died within 48 hours of hospitalization, a decision within the first 24 hours on removal of the infected areas (including limbs) that appear to be refractory to antimicrobial therapy is critical.
Acknowledgements
References I. Yuan CY. Yuan cc, Wei DC. Lee AM. Septicemia and gangrenous change of the legs caused by marine vibrio, Vibrio l'ulnificus-report of a case. J Formosan Med Assoc 1987;86:448-51. 2. Chuang yc. Young C. Chen CWo Vibrio vulnificus infection. Scand J Infect Dis 1989;21:721-6. 3. Chuang yc. Young CD. Vibrio vulnificus infections: clinical experience with 7 cases [abstract]. In: Proceedings of the Formosan Medical Association Annual Meeting 1989. Taipei, Taiwan: Formosan Medical Association, 1989: 114-5. 4. Chuang yc. Vibrio vulnificus infection in Taiwan [abstract]. In: Program and abstracts of the 2nd Western Pacific Congress on Infectious Diseases and Chemotherapy. Jomtien-Pattaya, Thailand: Infectious Disease Association ofThailand and Western Pacific Society of'Chernotherapy, 1990: 108. 5. Klontz KC. Lieb S, Schreiber M, Janowski HT, Baldy LM, Gunn RA. Syndromes of Vibrio vulnificus infections: clinical and epidemiologic features in Florida cases, 1981-1987. Ann Intern Med 1988;109: 318-23.
6. Bonner JR, Coker AS, Berryman CR, Pollock HM. Spectrum of Vibrio infections in a gulfcoast community. Ann Intern Med 1983;99:4649. 7. Blake PA, Merson MH, Weaver RE, Hollis DG, Heublein Pc. Diseases caused by a marine vibrio: clinical characteristics and epidemiology. N Engl J Med 1979;300: 1-5. 8. Hughes JM, Merson MH, Gangarosa EJ. The safety of eating shellfish. JAMA 1977;237:1980-1. 9. Kelly MT, Avery DM. Lactose-positive Vibrio in seawater: a cause of pneumonia and septicemia in a drowning victim. J Clin Microbiol 1980; II :278-80. 10. Tison DL, Kelly MT. Vibrio vulnificus endometritis. J Clin Microbiol 1984;20: 185-6. II. Johnston JM, Becker SF, McFarland LM. Gastroenteritis in patients with stool isolates of Vibrio vulnificus. Am J Med 1986;80:336-8. 12. Hoffmann TJ, Nelson B, Darouiche R, Rosen T. Vibrio vulnificus septicemia. Arch Intern Med 1988;148:1825-7. 13. Tacket CO, Brenner F, Blake PA. Clinical features and an epidemiological study of Vibrio vulnificus infections. J Infect Dis 1984; 149:55861. 14. Johnston JM, Becker SF, McFarland LM. Vibrio vulnificus: man and the sea. JAMA 1985;253:2850-3. 15. Howard RJ, Pessa ME, Brennaman BH, Ramphal R. Necrotizing softtissue infections caused by marine vibrios. Surgery 1985;98: 126-30. 16. Howard RJ, Lieb S. Soft-tissue infections caused by halophilic marine vibrios. Arch Surg 1988;123:245-9. 17. Joseph SW, DeBell RM, Brown WP. In l'itroresponse to chloramphenicol, tetracycline, ampicillin, gentamicin, and beta-lactamase production by halophilic vibrios from human and environmental sources. Antimicrob Agents Chemother 1978; 13:244-8. 18. Bowdre JH, Hull JH, Cocchetto DM. Antibiotic efficacy against Vibrio vulnificus in the mouse: superiority of tetracycline. J Pharmacol Exp Ther 1983;225:595-8. 19. Morris JG Jr, Tenney J. Antibiotic therapy for Vibrio vulnificus infection. JAMA 1985;253: 1121-2. 20. Morris JG Jr, Black RE. Cholera and other vibrioses in the United States. N Engl J Med 1985;312:343-50. 21. Arnold M, Woo ML, French GL. Vibrio vulnificus septicaemia presenting as spontaneous necrotising cellulitis in a woman with hepatic cirrhosis. Scand J Infect Dis 1989;21 :727-31. 22. Zielinski CJ, Bora FW Jr. Vibrio hand infections: a case report and review of the literature. J Hand Surg 1984;9A:754-7.
Downloaded from http://cid.oxfordjournals.org/ at University of Sussex on August 25, 2015
The authors express their gratitude to Dr. Kun-Yen Huang and Dr. Donald Armstrong for their critical reviewof this paper, to Dr. Ping-I Hsu for collecting patients' data, and to Miss JiaRong Shieh for preparing the manuscript.
CID 1992; 15 (August)
