Necrotising fasciitis: clinical features in patients with liver cirrhosis

British Journal of Plastic Surgery (2005) 58, 702–707

Necrotising fasciitis: clinical features in patients with liver cirrhosis Nai-Chen Chenga, Hao-Chi Taia, Yueh-Bih Tanga,*, Shan-Chwen Changb, Jann-Tay Wangb a

Division of Plastic Surgery, Department of Surgery, National Taiwan University Hospital, 7 Chung-Shan South Road, Taipei 100, Taiwan, ROC b Division of Infectious Diseases, Department of Medicine, National Taiwan University Hospital, Taiwan, ROC Received 18 April 2004; accepted 11 January 2005

KEYWORDS Necrotising fasciitis; Liver cirrhosis; Diabetes mellitus

Summary Necrotising fasciitis is a fulminant and life-threatening infection. It is associated with a high mortality rate and is often seen in the aged and immunocompromised patients. Liver cirrhosis is regarded as a risk factor of necrotising fasciitis. From January 1995 to December 2003, 17 cirrhotic patients who had been admitted to our hospital for necrotising fasciitis were identified. The infection all developed in the lower extremities. Only six patients survived, and the overall case fatality rate was 64.7%. The cases were divided into two groups: survivors and nonsurvivors. Comparisons were made on age, gender, presenting symptoms, underlying medical diseases, laboratory data and clinical course. Underlying diabetes mellitus and grade C liver cirrhosis were the only statistically significant factors that led to poor prognosis (p!0.05). q 2005 The British Association of Plastic Surgeons. Published by Elsevier Ltd. All rights reserved.

Necrotising fasciitis is a rare but fatal soft tissue infection, which is characterised by widespread necrosis of the superficial fascia and the subcutaneous fat. The causative organisms are usually toxin-producing, virulent bacteria, including Streptococcus, Staphylococcus, or a combination of Gramnegative bacilli and anaerobes. This condition is primarily found in the elderly and in the immunocompromised patients. Previous surgery, trauma, * Corresponding author. Tel.: C886 2 23562500; fax: C886 2 23934358. E-mail address: [email protected] (Y.-B. Tang).

diabetes and arterial insufficiency are common predisposing factors.1 Liver cirrhosis was associated with necrotising fasciitis in several reports,2–4 but the characteristics of this specific group of patient have not been elucidated. Therefore, we retrospectively analysed 17 cirrhotic patients who presented to our hospital with necrotising fasciitis from 1995 to 2003. The demographics, underlying diseases, clinical courses, and outcome were illustrated and analysed. The patients were divided into survivors and nonsurvivors. Several parameters were compared between the two groups to identify possible prognostic factors.

S0007-1226/$ - see front matter q 2005 The British Association of Plastic Surgeons. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.bjps.2005.01.019

Necrotising fasciitis in cirrhotic patients

Patients and methods We retrospectively reviewed the medical records of patients who were treated for necrotising fasciitis in National Taiwan University Hospital between January 1995 and December 2003. Patients were identified by a computer-generated search for patients who had been diagnosed with necrotising fasciitis in this 9-year period. Among the 136 identified patients, 17 who had been diagnosed with liver cirrhosis based on clinical, laboratory and echographic findings were included in the study. Definitive diagnosis of necrotising fasciitis depends on operative findings that include a lack of resistance of normally adherent fascia to blunt dissection, the presence of necrotic fascia and purulent discharge with foul ‘dish-water’ odour. Histopathological examination of surgical specimens was used to confirm the diagnosis when available. There were two cases that did not undergo surgery, but demonstrated rapid spreading topical skin changes and systemic toxicity. Computed tomography also revealed fascial plane dissection with fluid accumulation. So these two cases were also included for analysis. These cases were divided into two groups. Those who survived were classified as group I (nZ6), and those who did not survive were classified as group II (nZ11). The demographic characteristics, clinical presentations, severity of liver cirrhosis, laboratory data and hospital courses were analysed. Statistical analysis was carried out using the software STATA (Stata Inc., College Station, TX, USA). Mann– Whitney U-test and and Fisher’s exact test were used to compare the statistically significant difference between the two groups. The difference was considered significant if p value !0.05.

Results Seventeen patients (15 men and two women) were enrolled with a mean age 58 years (range, 35–82 years). The age and gender distribution between group I and group II revealed no statistical difference. Demographic, clinical and laboratory details of the patients were summarised in Table 1. The aetiology of liver cirrhosis was hepatitis B infection in eight cases, hepatitis C infection in three, coinfection of hepatitis B and C in two, and alcoholism in four. Twelve patients had grade C cirrhosis according to Child–Turcotte’s classification with Pugh’s modification.5 Another three patients had grade B cirrhosis, and the rest two had grade A. The distribution of severe cirrhosis (grade C) between

703 the two groups was significantly different by Fisher’s exact test (pZ0.028). All but one case in our study had comorbidities other than liver cirrhosis. Six patients had diabetes mellitus, five had malignancy, five had oesophageal varices, two suffered from chronic renal insufficiency and one had spinal cord injury with paraplegia. Pulmonary tuberculosis, hypertension, pneumonia and gouty arthritis were also sporadically noted (Table 1). Four patients in group II had underlying malignant diseases, including two with hepatocellular carcinoma, one with nonHodgkin lymphoma, and one with myelodysplastic syndrome. One in group I had hepatocellular carcinoma. The distribution of malignancy between the two groups was not significantly different (pZ0.6). All the six diabetic patients belonged to group II. There was a statistically significant difference between the two groups (pZ0.043). The infection involved the lower extremities in every case. Twelve of them started in unilateral lower leg, three in bilateral lower legs and two in unilateral thigh. The majority of patients presented with exquisite pain (16 patients), and local swelling/erythema (17 patients). Fever was noted in only eight patients. The possible portal of entry of microorganisms could be identified in seven patients, including five minor trauma, one insect bite and one burn site. Crepitus was noted in only two patients. Seven patients presented with septic shock in the emergency department (Table 2). Leukocytosis (white blood cell count O 10 000/ mm3) was found in 10 of 17 patients. The mean white blood cell count was 12 500/mm3 (range, 2400–25 480/mm3). A trend of neutrophilia was noted with the mean proportion of segment from 87.1% (range, 63–96.2%). The mean level of serum albumin was 2.2 g/dl (range, 0.6–3.1 g/dl). For these continuous variables, there was no statistically significant difference between the two groups by Mann–Whitney U-test. Blood were collected in the emergency department before the application of antibiotic, and the positive rate was 58.8% (10 in 17 patients). Cultures of tissue specimens were obtained at the time of the first operation. Excluding the two patients who did not undergo any operation, the total positive rate of tissue culture was 86.7% (13 in 15 patients). Organism isolated included Escherichia coli (three cases), Aeromonas hydrophila (two cases), Klebsiella pneumoniae (two cases), Enterobacter cloacae (two cases), Pseudomonas aeruginosa (two cases), Viridans streptococcus (one case), Staphylococcus aureus (one case), Vibrio vulnificus (one case), and Vibrio cholerae (one case) (Fig. 1). Eight patients visited our emergency department

704

N.-C. Cheng et al.

Table 1

Details of the 17 patients with necrotising fasciitis and liver cirrhosis

Case

Sex

Age

Initial site of infection

Other underlying diseasesa

Child’s grade of cirrhosis

Cause of cirrhosisb

Causative organismc

Entry port found

Outcome

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17

M M M M F M M M M M M M M M M M F

81 48 76 35 60 50 68 82 66 53 73 51 53 38 44 46 61

Right leg Left leg Both legs Left foot Right leg Left leg Both legs Left leg Left thigh Both legs Left leg Left leg Right leg Right leg Left leg Left leg Right thigh

COPD EV HCC SCI CRF – DM HCC DM, HCC EV DM, MDS DM, EV Pneumonia NHL DM DM, EV EV

B C B A C A B C C C C C C C C C C

Alcoholism HBV HBV, HCV Alcoholism HBV Alcoholism HCV HCV HBV Alcoholism HCV HBV HBV HBV HBV HBV HBV, HCV

MRSA Ec Pa NA Ah NA Vc Vv Ec Ah Vs Kp E. coli Pa E. coli E. coli Kp

Yes Yes Yes Yes Yes No Yes Yes Yes No No No No No No No No

Survived Survived Survived Survived Survived Survived Diseased Diseased Diseased Diseased Diseased Diseased Diseased Diseased Diseased Diseased Diseased

a COPD, chronic obstructive pulmonary disease; EV, oesophageal varicies; HCC, hepatocellular carcinoma; SCI, spinal cord injury; CRF, chronic renal failure; DM, diabetes mellitus; MDS, myelodysplastic disease; NHL, nonHogkin lymphoma. b HBV, hepatitis B virus; HCV, hepatitis C virus. c E. coli, Escherichia coli; Ah, Aeromonas hydrophila; Kp, Klebsiella pneumoniae; Ec, Enterobacter cloacae; Pa, Pseudomonas aeruginosa; Vs, Viridans streptococcus; MRSA, methicillin-resistant Staphylococcus aureus; Vv, Vibrio vulnificus; Vc, Vibrio cholerae, NA, not available.

directly, while nine patients were referred from other hospitals. All patients acquired the infection in the community. Twelve patients received their first surgery within 24 h after admission. During hospitalisation, the patients received an average of 1.7 operations (range, 0–4). Two patients did not undergo any surgical procedure. The period of

Table 2

Initial clinical presentations in the 17 cirrhotic patients with necrotising fasciitis

Clinical manifestations Symptoms and signs Fever Local swelling/erythema Pain Shortness of breath Crepitus Jaundice Conscious disturbance Bullous lesion Blood analysis Leukocytosis Hypoalbuminemiaa Shock in emergency departmentb a b

hospital stay ranged from hours to 73 days. The mean was 23.9 days and the median was 22 days. Amputation was performed to control the infection in five patients. Eleven patients died, and were classified into group II. The survivors were classified into group I. The case fatality rate was 64.7%. Several

Defined as serum albumin !2.5 g/dl. Defined as systolic blood pressure !90 mmHg.

No. of patients (%) 8 (47.1) 17 (100) 16 (94.1) 4 (23.5) 2 (11.8) 12 (70.6) 3 (17.6) 8 (47.1) 10 (58.8) 13 (76.5) 7 (41.2)

Necrotising fasciitis in cirrhotic patients

705

Figure 1 Distribution of the results of bacterial culture in the 17 cirrhotic patients with necrotising fasciitis.

parameters, including the distribution of gender and age, initial presentations, comorbidities, laboratory data, and clinical managements were compared between the two groups (Table 3). The only two variants that existed a significant difference between these two groups were underlying diabetes mellitus and grade C cirrhosis.

Discussion Meleney6 reported the landmark study of necrotising fasciitis in 1924, and found b-haemolytic Streptococcus to be the major causative agent. In 1952, Wilson7 named the disease ‘necrotising fasciitis’, which adequately describes a rapidly spreading necrosis of superficial fascia and subcutaneous tissue, with relative sparing of muscle and skin tissue. The reported mortality associated with Table 3

Statistical analysis for possible prognostic factors

Cases (n) Male/female (n) Mean age (years) Referred cases (n) Fever (n) Mean leukocyte count (/mm3) Segment (%) Serum albumin level (g/dl) Diabetes mellitus (n) Malignancy (n) Grade C cirrhosis (n) Shock in emergency department (n) Operation within 24 h after admission (n) Amputation (n) a

necrotising fasciitis ranged from 6 to 76%, with a cumulative mortality rate of 34%.8 Early and adequate surgical debridement and fasciotomy have been associated with improved survival.9 Early surgical treatment requires prompt diagnosis in the early stage of the disease. However, diagnosis of necrotising fasciitis is often difficult and relies on a high index of suspicion. Patients usually present with severe pain, swelling and fever. The early cutaneous signs are nonspecific, including oedema, erythema, local anaesthesia and occasional crepitus. Despite the minimal local manifestations, the patients usually complain of exquisite pain. Pain out of proportion of physical findings in a patient with systemic toxic signs should raise the clinical suspicion of necrotising fasciitis.1 In the late stage of the disease, the infection is disseminated through the relatively avascular fascial planes. It causes thromosis of the affected blood vessels. The overlying skin is devascularised, resulting in patchy discolourisation and hemorrhagic bullae formation. As organisms and toxins are released into the bloodstream, sepsis invariably develops.10 The susceptibility of patients with liver cirrhosis to infection has been emphasised. The abnormal defensive mechanisms in cirrhotic patients include impaired monocyte function, decreased phagocytic activity of the reticuloendothelial system, defective chemotaxis and low serum levels of complements. 11–13 Necrotising fasciitis has been previously divided into two distinct groups according to bacteriologic cultures. Type 1 is polymicrobial, synergistic infection that usually involves nongroup A streptococci, aerobic and anaerobic organisms. Type 2 infection is usually caused by group A b-haemolytic streptococci alone or in

Statistically significant, p!0.05.

Group I

Group II

p value

6 5/1 58.3 3 4 14540G7130 87.8G12.3 2.48G0.47 0 1 2 1 3 2

11 10/1 57.7 6 6 11380G6960 86.8G4.1 2.03G0.74 6 4 10 6 7 3

1.0 0.91 1.0 0.62 0.42 0.056 0.314 0.043a 0.6 0.028a 0.304 0.64 1.0

706 combination with staphylococci. However, a third type of necrotising fasciitis, caused by Gramnegative bacillus of the Vibrionaceae family, has been recognised in recent years. The pathogens included Vibrio and Aeromonas species.10,14 Chronic liver diseases, particularly liver cirrhosis, is the frequent predisposing factor of this type of infection.15 In our study, two cases were caused by Aeromonas hydrophila, and the other two were caused by vibrios (Vibrio vulnificus and Vibrio cholerae, serogroup non-O1), which constituted the most common aetiological agent. Necrotising fasciitis frequently develops in the extremities, the perineum, and the abdominal wall.10 The infections in our 17 cases all involved the lower limbs, with bilateral involvement in three of them. To induce infection of the skin and soft tissue, the causative organism may invade through a traumatic wound and causes primary infection of the soft tissue. Cirrhotic patients usually have chronic oedema of lower limbs, which may predispose to minor trauma and served as an entry port of bacteria. However, in immunocompromised or cirrhotic patients, necrotising fasciitis can occur without an obvious lesion on the body surface to allow bacterial invasion.9 Obvious portal of bacterial entry could be identified in only seven cases in our study. On the other hand, bacteraemia may first occur via the intestinal–portal route because liver cirrhosis weakens the barrier to the passage of bacteria from the intestine to the systemic circulation. The bacteria in the bloodstream may subsequently seed in the oedematous soft tissue of the lower limbs and cause necrotising fasciitis.16 One patient (case 2) reported a history of ingestion of raw seafood 1 day before the onset of symptoms, and blood and tissue cultures both yielded V. vulnificus. It has been found that patients with severe cirrhosis had higher rates of infection.12,17 Kuo et al.12 reported that the frequencies of bacteraemia were 1.0% for grade A, 4.8% for grade B, and 17.1% for grade C. A significant increase in mortality of grade B and C cirrhotic patients after onset of bacteraemia was also noted.12 The results of our study are consistent with these findings. Twelve of the 17 patients were classified as grade C liver cirrhosis, which was subsequently found to be an associating factor of mortality (Fisher’s exact test, pZ0.028). Diabetes mellitus has been reported as the most common underlying disease in patients with necrotising fasciitis, ranging from 72.3 to 57.1%.9,18,19 Our study also showed that diabetes is a poor prognostic factor for cirrhotic patients with necrotising fasciitis (Fisher’s exact test, pZ 0.043).

N.-C. Cheng et al. By analysing 24 patients with necrotising fasciitis of a limb, Tang et al.3 found no correlation between amputation and survival. It is consistent with the result of our study. However, it is noteworthy that one patient (case 1) presented with septic shock fully recovered after a single operation of aboveknee amputation. Amputation is usually a shorter procedure than radical debridement of the necrotic skin and soft tissue, and it requires less, if any, reconstructive operations. For patients with liver cirrhosis, bleeding tendency due to defective coagulative system may cause substantial blood loss when performing radical debridement. With cautious clinical judgments, amputation may be a better alternative for treating cirrhotic patients with necrotising fasciitis of limbs. Although rare, necrotising fasciitis is a lifethreatening disease. Early diagnosis and aggressive surgical debridement remains the cornerstone in the management of necrotising fasciitis. In view of our study, cirrhotic patients with necrotising fasciitis have a poor prognosis with the overall case fatality rate 64.7%. The presence of grade C liver cirrhosis and underlying diabetes mellitus are associated with mortality in these patients. When cirrhotic patients present with soft tissue infection, a high suspicion for necrotising fasciitis is recommended, especially when the infection involves the lower limbs. If the diagnosis is suspected, early involvement of experienced surgeons and a low threshold of surgical exploration are mandatory.

References 1. Meltzer DL, Kabongo M. Necrotizing fasciitis: a diagnostic challenge. Am Fam Physician 1997;56:145–9. 2. Lin CS, Cheng SH. Aeromonas hydrophila sepsis presenting as meningitis and necrotizing fasciitis in a man with alcoholic liver cirrhosis. J Formos Med Assoc 1998;97:498–502. 3. Tang WM, Ho PL, Fung KK, Yuen KY, Leong JC. Necrotising fasciitis of a limb. J Bone Joint Surg Br 2001;83:709–14. 4. 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. 5. Pugh RN, Murray-Lyon IM, Dawson JL, Pietroni MC, Williams R. Transection of the oesophagus for bleeding oesophageal varices. Br J Surg 1973;60:646–9. 6. Meleney F. Hemolytic Streptococcus gangrene. Arch Surg 1929;9:317–29. 7. Wilson B. Necrotizing fasciitis. Ann Surg 1952;18:416. 8. McHenry CR, Piotrowski JJ, Petrinic D, Malangoni MA. Determinants of mortality for necrotizing soft-tissue infections. Ann Surg 1995;221:558–63. 9. Wong CH, Chang HC, Pasupathy S, et al. Necrotizing fasciitis: clinical presentation, microbiology, and determinants of mortality. J Bone Joint Surg Am 2003;85A:1454–60. 10. Green RJ, Dafoe DC, Raffin TA. Necrotizing faciitis. Chest 1996;110:219–29.

Necrotising fasciitis in cirrhotic patients 11. Foreman MG, Mannino DM, Moss M. Cirrhosis as a risk factor for sepsis and death: analysis of the National Hospital Discharge Survey. Chest 2003;124:1016–20. 12. Kuo CH, Changchien CS, Yang CY, Sheen IS, Liaw YF. Bacteremia in patients with cirrhosis of the liver. Liver 1991;11:334–9. 13. Wyke RJ. Problems of bacterial infection in patients with liver disease. Gut 1987;28:623–41. 14. Cheng NC, Horng SY, Chang SC, Tang YB. Nosocomial infection of Aeromonas hydrophila presenting as necrotizing fasciitis. J Formos Med Assoc 2004;103:53–7. 15. Howard RJ, Pessa ME, Brennaman BH, Ramphal R. Necrotizing soft-tissue infections caused by marine vibrios. Surgery 1985;98:126–30.

707 16. Corredoira JM, Ariza J, Pallares R, et al. Gram-negative bacillary cellulitis in patients with hepatic cirrhosis. Eur J Clin Microbiol Infect Dis 1994;13:19–24. 17. Bleichner G, Boulanger R, Squara P, Sollet JP, Parent A. Frequency of infections in cirrhotic patients presenting with acute gastrointestinal haemorrhage. Br J Surg 1986;73: 724–6. 18. Lin C, Yeh FL, Lin JT, et al. Necrotizing fasciitis of the head and neck: an analysis of 47 cases. Plast Reconstr Surg 2001; 107:1684–93. 19. Hung CC, Chang SC, Lin SF, et al. Clinical manifestations, microbiology and prognosis of 42 patients with necrotizing fasciitis. J Formos Med Assoc 1996;95:917–22.