Bacterial Infections Associated with Hepatic Arteriography and Transarterial Embolization for Hepatocellular Carcinoma: A Prospective Study

Transcription

1 161 Bacterial Infections Associated with Hepatic Arteriography and Transarterial Embolization for Hepatocellular Carcinoma: A Prospective Study Chiher Chen, Yuk-Ming Tsang, Po-Ren Hsueh, Guan-Tarn Huang, Pei-Ming Yang, Jin-Chuan Sheu, Ming-Yang Lai, Pei-Jer Chen, and Ding-Shinn Chen From the Departments of Internal Medicine and Radiology and the Hepatitis Research Center, National Taiwan University Hospital, and the Graduate Institute of Clinical Medicine, National Taiwan University College of Medicine, Taipei, Taiwan Sepsis and liver abscess are serious complications following transarterial embolization (TAE) for hepatocellular carcinoma (HCC). However, the exact incidence and the necessity of antibiotic prophylaxis remain undetermined. Between November 1996 and November 1997, we prospectively studied bacterial infections in 231 HCC patients who underwent 287 angiographic procedures without antibiotic prophylaxis, including 176 TAEs and 111 hepatic arteriographies (HAs). Four of the 111 HAs were complicated by transient asymptomatic bacteremia. Of the 176 TAEs, 2 were associated with asymptomatic bacteremia, and 7 (4%) were associated with symptomatic bacterial infection, including 3 cases of sepsis, 2 of liver abscess, and 2 of infected biloma. For patients with HCC, TAE was associated with a higher risk of developing symptomatic bacterial infections than was HA (4% vs. 0, respectively; P.03). Previous gastrectomy was the only possible risk factor for liver abscess. Finally, early diagnosis and treatment of these infectious complications usually result in successful outcome. Hepatic arteriography (HA) and transarterial embolization (TAE) are important in the diagnosis and treatment of hepatic neoplasms, especially hepatocellular carcinoma (HCC). Common complications of TAE include fever, abdominal pain, nausea, and vomiting; other less frequent complications are gastroduodenal ulceration, splenic infarction, cholecystitis, sepsis, and liver abscess. Clinically, it is crucial to recognize bacterial infections to allow early, effective treatment. The incidence of transient bacteremia associated with angiography has been shown to be 4% [1]. For embolization procedures, an incidence of associated sepsis of 2.7% has been reported, even with antibiotic prophylaxis [2, 3]. In our previous study, the incidence of liver abscess after TAE for HCC was estimated to be 1% [4]. However, all these studies were retrospective in nature. To our knowledge, the spectrum and incidence of bacterial infections in HCC patients who have undergone HA and TAE have never been investigated by prospective studies. In addition, the necessity of antibiotic prophylaxis for infections associated with these procedures remains undetermined. In contrast to a predominance of gram-negative aerobes in sporadic pyogenic liver abscesses, the causative microorganisms were gram-positive in 60% of patients with liver abscess after TAE for HCC in two previous series, thus suggesting a different route of bacterial transmission [4, 5]. Whether this difference holds true Received 9 November 1998; revised 4 March Grant support: This work was supported by the Department of Health, Executive Yuan, Taiwan (grant DOH88-HR-804). Reprints or correspondence: Dr. Ding-Shinn Chen, Division of Gastroenterology and Hepatology, Department of Internal Medicine, National Taiwan University Hospital, 7 Chung-Shan South Road, Taipei 10016, Taiwan (dschen@ha.mc.ntu.edu.tw). Clinical Infectious Diseases 1999;29: by the Infectious Diseases Society of America. All rights reserved /99/ $03.00 in bacteremia or other infectious complications after angiographic procedures deserves further investigation, as it may affect strategy for antibiotic prophylaxis. In addition, risk factors for developing bacterial infections after TAE have not been determined prospectively, although diabetes mellitus, cholelithiasis, previous biliary manipulation, enterobiliary anastomosis, and large tumor size have been proposed [4, 5]. Since TAE for unresectable HCC is common practice in our hospital, we conducted a prospective study to address these questions and to determine the exact incidence of and identify the possible risk factors for infectious complications after HA and TAE for HCC. Patients and Methods Patients. Between November 1996 and November 1997, all patients who had undergone HA or TAE for HCC in the medical wards of the National Taiwan University Hospital (Taipei) were prospectively enrolled in the study. The diagnosis of HCC was made by either histopathologic proof or clinical evidence based on elevated serum -fetoprotein (AFP) levels of 400 ng/ml with results of imaging procedures, including abdominal ultrasonography, CT, and MRI [4, 6]. The exclusion criteria were as follows: patients receiving emergent HA and TAE for ruptured HCC, those with evidence of localized or generalized bacterial infection before the procedure, those receiving endoscopic sclerotherapy for variceal bleeding in the preceding 3 days, any use of antibiotics within the preceding 7 days, treatment with corticosteroids or other immunosuppressive drugs, and those with a history of endocarditis or valvular heart disease. Evaluation of the patients. Demographic data including gender; age; history of previous abdominal surgery, biliary manipulation, and diabetes mellitus; and previous treatments

2 162 Chen et al. CID 1999;29 (July) for HCC were recorded. Biochemical parameters, hematologic profiles, serum AFP levels, and status of hepatitis virus infection were determined and documented. Appropriate imaging studies were done for the diagnosis and staging of HCC. Bacterial cultures before and after diagnostic HA or TAE. Prophylactic antibiotics were not given. Before the procedures, swabs of shaved skin over the inguinal area were obtained for culture with use of Culturette collection tubes and transport system (Becton Dickinson Microbiology Systems, Cockeysville, MD), and two sets of blood specimens were obtained in BACTEC culture vials (Becton Dickinson, Sparks, MD) for aerobic and anaerobic cultures immediately before angiography. Diagnostic HA was performed via the femoral artery by the Seldinger technique by qualified radiologists. Strict aseptic preparation of the skin over the catheter dwelling site was applied. A disposable catheter (Rosch Celiac 1; Wilson-Cook, Bloomington, IN) was used for both HA and TAE. TAE was performed with intraarterial injection of 40 mg of doxorubicin hydrochloride, 5 ml of Lipiodol (Laboratorie Guerbert, Aulnay-Sous-Bois, France), and 0.05 g of Gelfoam powder (Upjohn, Kalamazoo, MI). The withdrawn catheter tip was immediately cultured for aerobes and anaerobes. Blood specimens for culture were obtained immediately and 12 and 24 hours after the angiographic procedures. Management of the patients after the angiographic procedures. After HA and TAE, fever, chills, and other signs suggesting infection were carefully monitored. AFP levels, hematologic profiles, and biochemical parameters were measured routinely. If fever persisted for 3 days, complete workup for a possible septic condition, including urinalysis, two more sets of blood cultures, chest roentgenography, and abdominal ultrasonography, were done. Antibiotic therapy was not given unless sepsis was suspected, which was defined by a positive blood culture plus two or more of the following systemic inflammatory responses as the result of infection: body temperature of 38 C or 36 C; heart rate of 90; respiratory rate of 20 or PaCO 2 of 32 mm Hg; and WBC count of 12,000/mm 3 or 4,000/mm 3 or 10% immature forms [7]. Follow-up abdominal CT was performed regularly 1 week after TAE. Signs suggesting tumor necrosis and possible infections such as liquefaction and presence of air in the tumor were checked. If a liver abscess was suspected, the content was aspirated under ultrasonography guidance; gram staining of the aspirate was performed, as were cultures for aerobes and anaerobes. Treatment with parenteral antibiotics was given and drainage or surgery was performed as indicated. Patients were followed up every 2 4 weeks at our outpatient department after discharge. Admission to the hospital for evaluation was arranged if any infectious complications were suspected. Table 1. Clinical characteristics of patients with hepatocellular carcinoma who underwent hepatic arteriography or transarterial embolization and outcomes of bacterial infections. Characteristic Results Hepatic arteriography Value Transarterial embolization No. of patients No. of males:females 82:28 103:30 Mean age (y) of patients SD No. of patients with cirrhosis No. of HBV-related:HCV-related:HBV 72:24:6 86:33:7 and HCV related infections No. of procedures No. of indicated factors associated with procedures Diabetes mellitus Cholelithiasis 3 7 Tumor size of 5 cm Previous hepatobiliary surgery Previous gastroduodenal surgery 5 8 Previous transarterial embolization 3 58* No. of episodes of all bacterial infections/ 4/111 (3.6) 9/176 (5.1) total no. of procedures (%) No. of episodes of transient asymptomatic 4/111 (3.6) 2/176 (1.1) bacteremia/total no. of procedures (%) No. of episodes of symptomatic bacterial 0/176 7/176 (4.0) infections/total no. of procedures (%) Sepsis 0/176 3/176 (1.7) Liver abscess 0/176 2/176 (1.1) Infected biloma 0/176 2/176 (1.1) NOTE. HBV hepatitis B virus; HCV hepatitis C virus. * P.001, 2 test. P.03, Fisher s exact test. Both patients had previously undergone subtotal gastrectomy. During the study period, 248 patients who met the selection criteria were enrolled; after excluding 17 patients for whom collection of culture samples was incomplete, 231 patients were included in the final analysis (table 1). Altogether, 111 HAs for 110 patients and 176 TAEs for 133 patients, all without antibiotic prophylaxis, were studied. Twelve patients were enrolled in both groups because they received HA and TAE on different occasions. The demographic and clinical characteristics of all the study subjects and the total number of infectious complications are listed in table 1. The patients in the two groups were not different in terms of age, gender, underlying hepatitis virus infections, and the presence of cirrhosis. The frequencies of previously proposed risk factors, including diabetes mellitus, cholelithiasis, large tumor size ( 5 cm in diameter), and history of hepatobiliary surgery and gastrectomy, were similar among both groups. Patients in the HA group were significantly less likely to have a history of TAE than were patients in the TAE group (P.001, 2 test). The incidence of transient asymptomatic bacteremia was

3 CID 1999;29 (July) Infections After HA and TAE 163 Table 2. Clinical characteristics and microbiological features of patients with bacterial infections after HA or TAE for hepatocellular carcinoma. Procedure, patient no. Sex/ age (y) Complication Responsible microorganism Associated factor(s) HA 1 F/68 Bacteremia Bacillus cereus None 2 M/68 Bacteremia Propionibacterium acnes None 3 M/70 Bacteremia Klebsiella pneumoniae None 4 M/68 Bacteremia Staphylococcus aureus None TAE 5 M/61 Bacteremia Acinetobacter baumannii None 6 M/65 Bacteremia Bacteroides fragilis Diabetes mellitus 7 F/69 Sepsis A. baumannii None 8 M/67 Sepsis K. pneumoniae None 9 F/69 Sepsis Campylobacter fetus None 10 M/62 Liver abscess Streptococcus equinus Previous TAE, subtotal gastrectomy 11 M/69 Liver abscess Escherichia coli Diabetes mellitus, subtotal gastrectomy, cholelithiasis 12 M/47 Infected biloma Streptococcus salivarius Previous TAE 13 F/77 Infected biloma K. pneumoniae Previous hepatectomy and TAE NOTE. HA hepatic arteriography; TAE transarterial embolization. 3.6% (four of 111 procedures) after HA and 1.1% (two of 176 procedures) after TAE; in all six cases, bacteremia was selflimited and did not mandate treatment with antibiotics. The causative microorganisms were Staphylococcus aureus, Bacillus cereus, Propionibacterium acnes, Klebsiella pneumoniae, Bacteroides fragilis, and Acinetobacter baumannii (table 2). None of the patients developed sepsis and liver abscess after HA. On the other hand, of the 176 TAEs, 7 (4%) were associated with symptomatic bacterial infections, including 3 cases (1.7%) of sepsis, 2 (1.1%) of liver abscess, and 2 (1.1%) of infected biloma. The incidence of symptomatic bacterial infection among the TAE group was significantly higher than that among the HA group (4.0% vs. 0, respectively; P.03, Fisher s exact test). All three patients with sepsis were promptly and successfully treated with parenteral antibiotics, and the causative microorganisms were Campylobacter fetus, A. baumannii, and K. pneumoniae. The causative microorganisms of the liver abscesses were Streptococcus equinus and Escherichia coli; both patients were also successfully treated with parenteral antibiotics and percutaneous aspiration and drainage of the abscess (figure 1). Infected bilomas, specifically denoting tumorlike intrahepatic cystic collections of bile extravasated after bile duct injury and necrosis [8], were diagnosed for two patients in the TAE group. Both patients were enrolled in the study when they were undergoing TAE for the fifth time. They were discharged uneventfully 1 week after TAE but were admitted again because of both progressive jaundice and fever 5 and 7 weeks later. Abdominal CT revealed a cystic lesion together with hypodense branching from the hilum to the periphery along the biliary tree (figures 2A and 2B). One of the patients underwent endoscopic retrograde cholangiopancreatography that revealed a common hepatic duct stricture (figure 2C). Infected bilomas were confirmed after aspirates obtained by ultrasonographyguided aspiration were examined. Blood cultures yielded Streptococcus salivarius and K. pneumoniae. Both patients were treated with percutaneous aspiration and drainage, but the bilomas persisted despite subsidence of fever and jaundice. One patient recovered after surgical drainage, but the other died of repeated biliary tract infections and progressive hepatic failure 2 months later. Of 287 catheter tip cultures, only two (0.7%) yielded Staphylococcus epidermidis and an unidentified gram-negative bacillus. Skin cultures revealed that the most common microorganisms colonizing the inguinal areas in our patients were coagulase-negative Staphylococcus, S. aureus, S. epidermidis, Enterococcus, Corynebacterium, Propionibacterium, and yeastlike organisms. None of the skin swab and catheter tip cultures yielded the same microorganism responsible for the infectious complication (table 2). Of the proposed risk factors for bacterial infections after TAE, only previous subtotal gastrectomy was found to be significant for the development of liver abscess (P.01, Fisher s exact test). Of the 176 TAEs, only eight (4.5%) were performed on patients with a history of gastrectomy, while the two patients with liver abscesses both had a history of subtotal gastrectomy. Discussion Overall, without antibiotic prophylaxis, the incidences of transient bacteremia, sepsis, liver abscess, and infected biloma following TAE for HCC were 1.1%, 1.7%, 1.1%, and 1.1%, respectively, in our study. No symptomatic bacterial infections

4 164 Chen et al. CID 1999;29 (July) Figure 1. Scans of a 62-year-old man who had a liver abscess 10 days after transarterial embolization for hepatocellular carcinoma. A. An abdominal ultrasonogram shows a mass with anechoic content and linear hyperechoic marks in the right lobe. B. An abdominal CT also depicts the lesion clearly. developed after HA for HCC. Only four (3.6%) of 111 HAs were associated with transient asymptomatic bacteremia, and none of the patients needed antibiotic treatment. TAE for HCC was associated with a higher incidence of bacterial infections than was HA. This result is compatible with those of most previous retrospective studies and confirms that TAE indeed increases the frequency of symptomatic bacterial infections [2, 3, 6, 9]. In our study, only 0.7% of the catheter tip cultures were positive. This finding is different than that of the early report of Shawker et al. [1] who held the catheter responsible for transient bacteremia after angiography. Their rate of positive catheter cultures was 12%. The difference could perhaps be explained by the fact that the angiographic catheters were sterilized and reused in their study, while we used disposable catheters. In our study, none of the skin swab and catheter tip cultures yielded the same microorganism responsible for the infectious complication. Therefore, if a disposable catheter and routine sterilization procedures are strictly applied, the infectious complications following HA and TAE are unlikely to occur after radiological manipulations. Instead, the infection was more likely related to the changes in the host caused by embolization, such as alterations of the gastrointestinal barrier and damages to intrahepatic vascular supply and defense mechanisms. In contrast to the predominance of gram-negative bacteria in cases of pyogenic liver abscess, gram-positive bacteria, mainly Staphylococcus, Streptococcus, Enterococcus, and Clostridium, have been reported before as causative microorganisms in approximately one-half of the cases of TAErelated liver abscesses [3, 5]. Our prospective study confirms this finding. S. equinus (group D Streptococcus) and S. salivarius (viridans Streptococcus) accounted for two of the four cases of liver abscess and infected biloma after TAE; these two bacteria usually inhabit the oral mucous membrane and the gastrointestinal tract [10]. Of the three pathogens that caused sepsis after TAE, K. pneumoniae and A. baumannii are well-known pathogens of nosocomial infections. The other case of sepsis was caused by C. fetus; despite the ubiquitous nature of this gram-negative rod, many manifestations of infection with this organism are not well recognized. This organism tends to inhabit the gastrointestinal tract [11]. It is proposed that the bacterium may remain quiescent in the gut but cause generalized infection in patients with HCC who have decreased host resistance and immunity caused by TAE. The risk factors proposed by previous retrospective studies, such as diabetes mellitus, cholelithiasis, and history of hepatobiliary surgery, gastroduodenal surgery, and enterobiliary anastomosis, were analyzed in our study. However, a history of

5 CID 1999;29 (July) Infections After HA and TAE 165 subtotal gastrectomy was the only significant risk factor for liver abscess after TAE. The stomach, with its strong acidity, provides an important barrier to regulate the number of ingested bacteria. Many of these organisms are promptly killed in the acidic environment. Gastrectomy removes this barrier and increases the number of bacterial flora accessing the compromised gastrointestinal mucosa after TAE. In addition, retrograde bacterial contamination of the biliary tree from anastomosis and of the gastrointestinal tract also increase. Both Figure 2. Scans of a 47-year-old man who had an infected biloma following transarterial embolization (TAE) for hepatocellular carcinoma; he had elevated serum levels of alkaline phosphatase and -glutamyltransferase and then had jaundice and fever 5 weeks after TAE. A. An abdominal CT shows a hypodense mass (arrowhead) away from the tumor (arrow). Examination of an aspirate proved that the lesion was an infected biloma. B. Another abdominal CT reveals a cross section of the biloma that appears as a prominent branching structure (arrowhead) extending from the hepatic hilum to the periphery along the biliary tree. The bright sparkles are the iodized oil retained in the tumor (arrow) after TAE. C. An endoscopic retrograde cholangiopancreatogram demonstrates a cystic lesion with a collection of contrast medium in the biloma (arrow) and a concomitant stricture over the common bile duct (arrowhead). factors might contribute to bacterial infection and liver abscess after TAE [9, 12, 13]. In addition, we noted two cases (1.1%) of infected biloma after TAE for HCC. Biloma mainly results from traumatic blunt biliary injuries in traffic accidents and iatrogenic injuries due to abdominal surgery and percutaneous and endoscopic diagnostic procedures [8]. The mechanism of bile duct necrosis and biloma formation after TAE is believed to be due to ischemic damages of the peribiliary plexus, which uniquely supplies blood to the bile ducts [14, 15]. Both of our patients had undergone multiple TAEs before the development of bilomas. Repeated damage by multiple and superselective embolization might be related to the occurrence of this complication. Finally, the necessity of preventing bacterial infections after TAE by antibiotic prophylaxis needs to be discussed. To our knowledge, it has never been concluded that without antibiotic prophylaxis, there would be more fatal bacterial infections after TAE, and most researchers recommend a conservative strategy and antibiotic prophylaxis is not helpful [3, 16]. Only a randomized, prospective study can resolve this problem. However, because the incidence of clinical infections following TAE for HCC without antibiotic prophylaxis is only 4%, a study to determine the possible effects of prophylaxis might have to enroll 1,000 patients. Unless justified by cost-benefit analysis, such a large-scale clinical study seems impractical. At present, most importantly clinicians must be fully aware that bacterial infections occur after

6 166 Chen et al. CID 1999;29 (July) TAE for HCC, and appropriate effective treatment should be given as early as possible in cases of infection. References 1. Shawker TH, Kluge RM, Ayella RJ. Bacteremia associated with angiography. JAMA 1974;229: Hemingway AP, Allison DJ. Complications of embolization: analysis of 410 procedures. Radiology 1988;166: Reed RA, Teltelbaum GP, Daniels JR, Pentecost MJ, Katz MD. Prevalence of infection following hepatic chemoembolization with cross-linked collagen with administration of prophylactic antibiotics. J Vasc Interv Radiol 1994;5: Chen C, Chen P-J, Yang P-M, et al. Clinical and microbiological features of liver abscess after transarterial embolization for hepatocellular carcinoma. Am J Gastroenterol 1997;92: De Barre T, Roche A, Amenabar JM, et al. Liver abscess formation after local treatment of liver tumors. Hepatology 1996;23: Beppu T, Ohara C, Yamaguchi Y, et al. A new approach to chemoembolization for unresectable hepatocellular carcinoma using aclarubicin microspheres in combination with cisplatin suspensed in iodized oil. Cancer 1991;68: Bone RC, Fein AM, Balk RA, et al. Definitions for sepsis and organ failure and guidelines for the use of innovative therapies in sepsis. Chest 1992;101: Vazquez JL, Thorsen MK, Dodds WJ, et al. Evaluation and treatment of intraabdominal bilomas. Am J Roentgenol 1985;144: Ishikawa H, Kanai T, Ono T, et al. Analysis of cases with liver abscess following transcatheter arterial embolization (TAE) for malignant hepatic tumors (in Japanese). Japanese Journal of Cancer and Chemotherapy 1994;21: Awada A, van der Auwera P, Meunier F, Daneau D, Klastersky J. Streptococcal and enterococcal bacteremia in patients with cancer. Clin Infect Dis 1992;15: Morrison VA, Lloyd BK, Chia JKS, Tuazon CU. Cardiovascular and bacteremic manifestations of Campylobacter fetus infection: case report and review. Rev Infect Dis 1990;12: Giannella RA, Broitman SA, Zamcheck N. Influence of gastric acidity on bacterial and parasitic enteric infections. A perspective. Ann Intern Med 1973;78: Carter AO, Borczyk AA, Carlson JA, et al. A severe outbreak of Escherichia coli O157:H7 associated hemorrhagic colitis in a nursing home. N Engl J Med 1987;317: Kobayashi S, Nakanuma Y, Terada T, Matsui O. Postmortem survey of bile duct necrosis and biloma in hepatocellular carcinoma after transcatheter arterial chemoembolization therapy: relevance to microvascular damages of peribiliary capillary plexus. Am J Gastroenterol 1993;88: Makuuchi M, Sukigara M, Mori T, et al. Bile duct necrosis: complication of transcatheter hepatic artery embolization. Radiology 1985;156: Castells A, Bruix J, Ayuso C, et al. Transarterial embolization for hepatocellular carcinoma. Antibiotic prophylaxis and clinical meaning of postembolization fever. J Hepatol 1995;22:410 5.