Efficacy of antibiotic penetration into pancreatic necrosis

Transcription

1 HPB, 2006; 8: 43/48 Efficacy of antibiotic penetration into pancreatic necrosis W. OTTO, K. KOMORZYCKI & M. KRAWCZYK Department of General, Transplantation & Liver Surgery, Medical University of Warsaw, Warsaw, Poland Abstract Background. Infection of pancreatic necrosis is the most dangerous complication of acute necrotizing pancreatitis. Infection is undoubtedly caused by the endogenous flora of the host. This is why prophylaxis with a broad-spectrum antibiotic is considered an effective procedure. However, two aspects should be taken into consideration when choosing the antibiotic; it should have the spectrum of action consistent with the pathogens and it should penetrate effectively to the necrotic tissue of the pancreas. The aim of the study was to estimate the efficacy of piperacillin/tazobactam penetration into pancreatic necrosis in patients who received intravenous infusion of piperacillin/tazobactam at a dose of 4.5 g every 8 h for 14/21 days, as the prophylaxis in the treatment of acute necrotizing pancreatitis. Patients and methods. Necrotic tissue of the pancreas and the inflammatory ascites surrounding the pancreas were derived from 15 patients (male, 10; female, 5; mean age 46 years), who underwent laparatomy due to pancreatic necrosis after treatment for 14 /21 days. Tissue/fluid samples were investigated for the concentration of the antibiotic by fluoroscopic/spectroscopic methods of registration in an HPLC system. Results. The mean concentration of piperacillin/tazobactam was established as 120 mg/kg (SD9/34) in the necrotic pancreatic tissue and as 183 mg/kg (SD9/37) in the inflammatory pancreatic ascites. Conclusions. In patients with acute necrotizing pancreatitis, the study indicates effective penetration of piperacillin/tazobactam to the necrotic pancreatic tissue and to the inflammatory ascites surrounding the pancreas. Key Words: Acute necrotizing pancreatitis, antibiotic prophylaxis, piperacillin/tazobactam, antibiotic penetration of pancreatic necrosis Introduction Infection of pancreatic necrosis is rightly considered a life-threatening complication in patients with acute necrotizing pancreatitis [1,2]. Many reports indicate that treatment with a broad-spectrum antibiotic, if applied at the initial stage of disease, is potent to counteract the endogenous infection of the necrosis and consequently improves the results of treatment. However, two aspects should be taken into consideration when choosing the antibiotic; it should have the spectrum of action consistent with the pathogens and it should penetrate effectively to the necrotic tissue of the pancreas [3 /6]. Piperacillin/tazobactam is a well known broad-spectrum antibiotic, in general use, with a spectrum of action consistent with most of the pathogens responsible for endogenous infection. It is indicated for most mixed hospital infections, serious intra-abdominal infections and sepsis, including infection of pancreatic necrosis [5,7 /10]. The aim of the study was to estimate the efficacy of piperacillin/ tazobactam penetration into pancreatic necrosis in patients who received intravenous infusion of piperacillin/tazobactam at a dose of 4.5 g every 8 h for 14/21 days, as prophylaxis in the treatment of acute necrotizing pancreatitis. Patients and methods There were 15 patients (male, 10; female, 5; mean age 46 years) with acute necrotizing pancreatitis treated at the department from 2000 to The criteria for including them in the study were established as follows: age over 18 years, pancreatic necrosis confirmed by USG/CT scanning, hospitalization not later than 72 h after the onset of disease, no antibiotic treatment in the 7 days prior to the symptoms of pancreatitis, level of C-reactive protein /120 mg/l at the time of hospitalization, patient s consent for investigation. Patients who did not comply with all points of the cited criteria were excluded from the evaluation. The diagnosis of pancreatic necrosis was established by clinical and biochemical evaluation. All patients presented on admittance with serious abdominal pain, vomiting, dehydration and oliguria, as well as circulatory and respiratory disorders such as tachycardia, low blood pressure, tachypnoea. There Correspondence: W. Otto, MD, PhD, Department of General, Transplantation & Liver Surgery, Medical University of Warsaw, Warsaw, Banacha 1a, Poland. Tel: /48 (22) , Fax: /48 (22) wotto@gsystem.pl ISSN X print/issn online # 2006 Taylor & Francis DOI: /

2 44 W. Otto et al. was inflammatory tumour palpable in the upper abdomen and regional peritonitis was also present in all of them. Increased white cell count, increased level of blood/urine amylase, and C-reactive protein /120 mg/l complemented the serious clinical status of each patient. The diagnosis was confirmed in every case by USG and spiral CT scanning. In patient nos 3, 7, 8, 11, 12 and 13, spiral CT indicated inflammatory process of pancreatic gland with necrosis of /30% but B/50% of the parenchyma. In patient nos 1, 2, 4, 5, 6, 9, 10, 14 and 15, spiral CT indicated inflammation of the pancreatic gland with necrosis of /50% of the parenchyma. Inflammatory ascites surrounding the pancreas was also present. The APACHE II scores amounted to at least 10 points in all patients and the Ranson score oscillated between 4 and 8 points, as established during 24 h after admittance (Table I). The treatment started immediately after admission and consisted of general and intensive care, TPN and intravenous infusion of piperacillin/tazobactam at a dose of 4.5 g infused over 30 min every 8 h. In patient nos 1, 3, 4, 6, 9, 10, 11 and 13 antibiotic infusion started before or simultaneously with the intravenous fluid resuscitation. In patient nos 2, 5, 7, 8, 12, 14 and 15 antibiotic infusion started after adequate fluid resuscitation. Intravenous infusion of the antibiotic was maintained up to 14/21 days of treatment / the day when the patients underwent operation. Surgical intervention was undertaken due to extended pancreatic necrosis. Deteriorating patient s status, growing intra-abdominal pressure ( B/20 cm H 2 O) and increasing toxaemia, as well as palpable mass and symptoms of regional peritonitis in the upper part of abdomen were the indications for the operation. The decision was supported by CT scanning indicating changes of the dense pancreatic inflammatory parenchyma into the liquid suppurative collection. Symptoms of infection before the operation were only evident in patient nos 5 and 10. Blood plasma concentration of piperacillin/tazobactam was evaluated in each patient while receiving three consecutive doses of the antibiotic at 1 /2 days before the operation. Blood samples were collected at the following sampling time points: T /0.5, i.e. 30 min after the dose infusion was completed; T /2, i.e. 2 h after the dose infusion was completed. Blood samples were centrifuged, separated and frozen at /808C until the time of analysis in an HPLC system. The operation comprised laparatomy, necrosectomy, peritoneal lavage and postoperative drainage. Shreds of the necrotic pancreatic tissue and the inflammatory ascites from omental bursa were collected for bacteriology, as well as for spectroscopic evaluation of antibiotic concentration. The spectroscopic procedure required approximately 1 g of necrotic pancreatic tissue and 1 ml of the ascites surrounding the pancreas. The samples were frozen at /808C until the time of analysis by fluoroscopic/ spectroscopic methods of registration in an HPLC system. The spectrometer was supplied with two independent monochromatic inductors and registration, xenon lamp and luminescence signal detector PMT with a spectrum range of 210/370 nm. The results were evaluated statistically. Results On day 1 /2 before the operation the mean blood plasma concentration of piperacillin/tazobactam in the 15 patients reached the level of 234 (SD9/11) to 240 (SD9/12) mg/l at 30 min after the dose infusion and 61 (SD9/7) to 64 (SD9/7) mg/l at 2 h after the infusion (Figure 1). Table I. The amount of pancreatic necrosis calculated by the initial spiral CTand the Ranson score calculated at 24 h after admittance in 15 patients with necrotizing pancreatitis. No. Initials Age Sex Amount of necrosis (%) Ranson score 1* J. Ch. 54 M B/50 6 pts 2 K. L. 30 M B/50 6 pts 3* D.T. 28 F /30 but B/50 4 pts 4* A. R. 48 M B/50 5 pts 5 T. D. 56 M B/50 7 pts 6* K. P. 32 M B/50 6 pts 7 G. S. 49 M /30 but B/50 4 pts 8 T. C. 52 F /30 but B/50 4 pts 9* B.K. 64 F B/50 5 pts 10* L. P. 49 M B/50 8 pts 11* B. J. 72 F /30 but B/50 4 pts 12 R. T. 47 M /30 but B/50 4 pts 13* H. M. 38 F /30 but B/50 4 pts 14 W. C. 42 M B/50 5 pts 15 E. S. 51 M B/50 6 pts *Patients receiving antibiotic before or simultaneously with fluid resuscitation.

3 Concentration mg/l Time (hr) : T +0.5 T +2 T +0.5 T +2 T+0.5 T +2 Figure 1. Blood serum concentrations of piperacillin/tazobactam in 15 patients with necrotizing pancreatitis while receiving 3 consecutive doses of the antibiotic at 1/2 days before the operation. Concentrations were measured from blood samples collected at the following time points: T/0.5, 30 min after the antibiotic infusion; T/2, 2 h after the antibiotic infusion. Mean values for T/0.5/ 234 mg/l (SD9/11) to 240 mg/l (SD9/12); T/2/61 mg/l (SD9/7) to 64 mg/l (SD9/7). Bacteriological examination of the necrotic tissue samples collected during the operation indicated aseptic necrosis in all patients except patient nos 5 and 10, in whom infection with Enterococcus faecium and methicillin-resistant Staphyloccoccus aureus (MRSA) was recognized, respectively (Table II). The concentration of piperacillin/tazobactam differed from 67 mg/kg to 102 mg/kg in the necrotic pancreatic tissue and from 101 mg/kg to 190 mg/kg in the ascites surrounding the pancreas in patients who received antibiotic after sufficient fluid resuscitation. The concentration of piperacillin/tazobactam was higher in patients receiving antibiotic before or simultaneously with the fluid resuscitation and differed from 102 mg/kg to 176 mg/kg in the necrotic pancreatic tissue and from 190 mg/kg to 261 mg/kg in the ascites surrounding the pancreas. The differences in the concentration of piperacillin/tazobactam in the necrotic pancreatic tissue and in the ascites surrounding the pancreas between these two groups were significant (p B/0.01). The mean concentration of piperacillin/tazobactam in the necrotic tissue of pancreas was established at 120 mg/kg (SD9/34) and in the ascites surrounding the pancreas at 183 mg/kg (SD9/37). The results are indicated in Table III / patients receiving antibiotic before or simultaneously with the fluid resuscitation are indicated with an asterisk. Efficacy of antibiotic penetration into pancreatic necrosis 45 Discussion The study indicates that intravenous infusion of piperacillin/tazobactam at a dose of 3/4.5 g per 24 h, administered to the patient at the initial stage of acute necrotizing pancreatitis, allowed the antibiotic to reach high and effective concentrations in the necrotic pancreatic tissue and in the inflammatory ascites surrounding the pancreas. Experimental studies in rats, as well as pharmacodynamic studies performed on healthy pancreas, indicate sufficient penetration of piperacillin/tazobactam to the pancreatic parenchyma [4,6,10/13]. This opinion is distinctly supported by Bassi and his co-workers, who concluded that the antibiotics which proved capable of penetrating into the pancreas in physiological conditions maintained this ability in the course of acute disease with a necrotic component. Their use in the prophylaxis of necrotizing pancreatitis appears useful [14]. So, it was a challenge to prove whether the antibiotic could reach high (i.e. therapeutic) concentrations in the necrotic pancreatic tissue in clinical conditions, in patients with advanced hypovolaemia and pancreatic toxaemia. The data achieved in the present study fulfilled the expectations. Antibiotic prophylaxis has been shown in many reports as the therapeutic modality able to improve the results of treatment and to increase chances of the patient for a positive outcome in acute necrotizing pancreatitis [7,8,11,15,16]. Despite the adequate fluid resuscitation, effective intensive care and total parenteral nutrition, antibiotic prophylaxis became the gold standard of the inceptive therapy for serious pancreatic necrosis. Opponents of this approach do not see the advantages of antibiotic prophylaxis, as shown by Isenmann in his latest double-blind, placebo-controlled study with ciprofloxacin/metronidazole [17]. The policy of the inceptive antibiotic therapy is obvious / to keep necrosis sterile for as long as possible [1,3,7,8]. Infection of the necrosis is usually initiated by the endogenous pathogens of the host, mainly by enterococci and bacilliform Gram-negative bacteria from the alimentary tract [18,19]. Two aspects should be taken into consideration when choosing the antibiotic; it should have the spectrum of action consistent with the pathogens and it should penetrate effectively to the necrotic tissue of the pancreas. Table II. Bacteriological evaluation of the necrotic pancreatic tissue samples collected during the operation in 15 patients. No. 1* 2 3* 4 5 6* 7 8 9* 10 11* 12 13* Result / / / / E / / / / S / / / / / /, aseptic; E, Enterococcus faecium; S, Staphylococcus aureus (MRSA). *Patients receiving antibiotic before or simultaneously with fluid resuscitation.

4 46 W. Otto et al. Table III. Piperacillin/tazobactam concentration in the necrotic pancreatic tissue and in the ascites surrounding the pancreas in the patients operated on after treatment for 14/21 days for necrotizing pancreatitis. Patient no. Antibiotic concentration 1* 2 3* 4* 5 6* 7 8 9* 10* 11* 12 13* Mean SD Necrosis (mg/kg) /34 Ascites (mg/kg) /37 *Patients receiving antibiotic before or simultaneously with fluid resuscitation. The systemic use of imipenem-cilastatinum, meropenem, fluoroquinolones and metronidazole is currently recommended for the management of pancreatic endogenous infection [1,3,10,13]. All indicated agents have a suitable spectrum of action and seem to penetrate well to the pancreatic juice and pancreatic parenchyma. Studies have indicated that meropenem, cefepim and imipenem are potent enough to reach high concentrations in the tissue of the pancreatic gland [6,16,20]. High concentrations of meropenem and cefepin in pancreatic parenchyma were also observed by Saglamakaya in rat experimental pancreatitis [10]. However, some of these studies were performed on pancreas unchanged by the inflammatory process. On the other hand, the infection of pancreatic necrosis could develop despite the fact that the antibiotic spectrum is suitable for the pathogens responsible for the infection, as shown by the study with meropenem and imipenem, and by many other clinical observations [6,9,10,12,15,21,22]. The rate of necrotic area in the pancreas corresponds statistically with the rate of inflammatory complications in the patient. In patients with necrosis exceeding /50% of the pancreas, as evaluated on the Baltazar scale by spiral CT, infection is the rule [22]. Perez and his co-workers argue for such a correlation and Bassi et al. [15] indicate that the concentration of imipenem in the pancreatic necrotic tissue does not very often exceed the minimal inhibitory concentration (MIC) that could be pernicious for the endogenous bacteria [7,15,17]. This is presumably due to poor penetration of the antibiotic to necrotic tissue, so the MIC of the antibiotic could not effectively prevent the infection of the expansive necrosis. The failure of the antibiotic action could be also influenced by the patient s hypovolaemia, the general level of toxicity and the time evolved from the onset of disease. The effectiveness of antibiotic penetration to the area of pancreatic necrosis could be increased by changes in dosage schedules and by the use of multiple doses, as well as by using more effective ways of administration. The intravenous route appears to be more effective than the intramuscular route, especially in the repeated infusion regimen of administration [5,14,15]. Piperacillin/tazobactam is a very well known broadspectrum antibiotic indicated for most mixed hospital infections, serious intra-abdominal infections and sepsis, including infection of pancreatic necrosis. It has a spectrum of action consistent with most of the pathogens responsible for endogenous infection of pancreatic necrosis. The evaluation of piperacillin/ tazobactam concentration by fluoroscopic/spectroscopic methods in patients with pancreatic necrosis presented in this study shows that the antibiotic is potent enough to reach levels as high as 176 mg/kg in necrotic tissue and 261 mg/kg in ascites surrounding the pancreas. The tissue/ascites concentration levels of the antibiotic were almost as high as the

5 concentration levels established in the patient s blood plasma during the infusion of three consecutive doses at 1/2 days before the operation. These values exceeded three- to five-fold the usual concentration of the antibiotic indicated for pancreatic parenchyma as therapeutic. The rate of antibiotic concentration was significantly higher in patients receiving the antibiotic before or simultaneously with the fluid resuscitation (p B/0.01). It seems to be a little curious but it could not be explained at that stage of the study. Bacteriological examination of the tissue samples collected at the time of operation indicated the effective antimicrobial activity of the antibiotic. In all patients bacteriological seeds were negative, except for patients 5 and patient 10, in whom infection with E. faecium and MRSA was recognized, respectively. Of course, these two bacteria are resistant to piperacillin/tazobactam. Patient nos 5 and 10 developed infected necrosis and undoubtedly required operation. Why then, were the patients who remained aseptic operated upon? It must be emphasized that surgical intervention was undertaken in these patients because of deteriorating general status due to growing intra-abdominal pressure (B/20 cm H 2 O) and increasing toxaemia. CT scans indicated changes in the dense pancreatic inflammatory parenchyma into a liquid suppurative collection corresponding with the palpable mass and symptoms of regional peritonitis in the upper part of abdomen. The increasing risk of multiple organ failure forejudged the decision, although there were no evident symptoms of infection. As shown in a previous clinical study that was conducted from 1999 to 2003 in the Department of General, Transplantation and Liver Surgery, Medical University of Warsaw, prophylactic use of piperacillin/ tazobactam in our patients significantly increased the rate of aseptic necrosis and significantly reduced the rates of inflammatory complications and mortality. Thus, the antibiotic prophylaxis with piperacillin/ tazobactam appeared to be beneficial for our patients from a clinical point of view [23,24]. In contrast, some other studies do not confirm these effects. In the latest double-blind, placebo-controlled study with ciprofloxacin and metronidazole, Insenmann and his co-workers found no benefits from the antibiotic prophylaxis. They included 114 patients in the intention-to-treat analysis. Patients were recruited on the basis of an elevated serum CRP in combination with pancreatic necrosis on CT. In patients with necrotizing pancreatitis, no differences between the verum and the placebo group were observed [17]. The results of this study add fuel to the discussion as to whether prophylactic antibiotic therapy in patients with acute necrotizing pancreatitis is a beneficial approach. Unfortunately, neither the data presented in this paper nor our previous experiences allow us to draw firm conclusions. Efficacy of antibiotic penetration into pancreatic necrosis 47 Despite the substantial doubts as to the effectiveness of antibiotic prophylaxis in the treatment of acute necrotizing pancreatitis, it remains acceptable and in general use as a modality of inceptive therapy. In most of the published studies, meropenem or imipenem is widely recommended. Our study shows that although the spectrum of action of piperacillin/tazobactam is not as broad as the spectrum of meropenem, imipenem-cilastatinum and fluoroquinolones, it could be of greater clinical value. Piperacillin/tazobactam shows extremely good penetration into pancreatic necrosis, it is cost-effective and does not increase problems with antibiotic resistance and fungal infection. Such requirements could not be fulfilled by imipenemcilastatinum and meropenem or by fluoroquinolones [1,6,12,10,13]. Conclusions In patients with acute necrotizing pancreatitis treated with intravenous infusion of piperacillin/tazobactam, concentrations of the antibiotic reached effective, therapeutic levels in the necrotic pancreatic tissue and in the inflammatory ascites surrounding the pancreas. References [1] Butturini G, Salvia R, Bettini R, Falconi M, Peterzoli P, Bassi C. Infection prevention in necrotising pancreatitis: an old challenge with new perspectives. J Hosp Infect 2001;/48:/4/ 8. [2] Steinberg W, Tenner S. Acute pancreatitis. N Engl J Med 1994;/330:/1198/210. [3] Isenmann R. Natural history of acute pancreatitis and the role of infection. Baillieres Clin Gastroenterol 1999;/13:/291/301. [4] Sainio V, Kemppainen E, Puolakkainen P, Taavitsainen M, Kivisaari L, Valtonen V, et al. Early antibiotic treatment in acute necrotising pancreatitis. Lancet 1995;/346:/663/6. [5] Bassi C. Infected pancreatic necrosis. Int J Pancreatol 1994;/ 16:/1/10. [6] Pederzoli P, Bassi C, Visentini E. A randomized multicentre clinical trial of antibiotic prophylaxis of septic complications in acute necrotizing pancreatitis with imipenem. Surg Gynecol Obstet 1993;/176:/480/3. [7] Delcenserie R, Yzet T, Ducroix JP. Prophylactic in treatment of severe acute alcoholic pancreatitis. Pancreas 1996;/13:/198/ 201. [8] Gloor B, Schmidtmann AB, Worni M, Ahmed Z, Uhl W, Buchler MW. Pancreatic sepsis: prevention and therapy. Best Pract Res Clin Gastroenterol 2002;/16:/379/90. [9] Perez A, Whang EE, Brooks DC. Is severity of necrotizing pancreatitis increased in extended necrosis and infected necrosis? Pancreas 2002;/24:/258/64. [10] Saglamakaya N. Penetration of meropenem and cefepim into pancreatic tissue during the course of experimental acute pancreatitis. Pancreas 2002;/24:/258/64. [11] Ho HS, Frey CF. The role of antibiotic prophylaxis in severe acute pancreatitis. Arch Surg 1997;/132:/487/93. [12] Brattstroem C, Malmborg AS, Tyden G. Penetration of clindamycin, cefoxin and piperacillin into pancreatic juice in man. Surgery 1988;/103:/563/7. [13] Dietrich ES, Schubert B, Ebner W, Daschner F. Cost efficacy of tazobactam/piperacillin versus imipenem/cilastatin in the

6 48 W. Otto et al. treatment of intra-abdominal sepsis. Pharmacoeconomics 2001;/19:/79/94 [14] Bassi C, Pederzoli P, Vesentini S, Falconi M, Bonora A, Abbas H, et al. Behavior of antibiotics during human necrotizing pancreatitis. Antimicrob Agents Chemother 1994;/38:/830/6 [15] Bassi C, Falconi M, Talamini G, Uomo G, Papaccio G, Dervenis C, et al. Controlled clinical trial of pefloxacin versus imipenem in severe acute pancreatitis. Gastroenterology 1998;/ 115:/1513 /7. [16] Buchler MW, Gloor B, Muller CA. Acute necrotising pancreatitis: treatment strategy according to the status of infection. Ann Surg 2000;/232:/619/36 [17] Isenmann R, Runzi M, Kron M, Kahl S, Kraus D, Jung N, et al., and the ASAP Study Group. Prophylactic antibiotic treatment in patients with predicted severe acute pancreatitis: a placebo-controlled, double blind trial. Gastroenterology 2004;126:997/1004. [18] Ammori BJ. Role of the gut in the course of severe acute pancreatitis. Pancreas 2003;/26:/122/9. [19] Wiest R, Rath HC. Bacterial translocation in the gut. Best Pract Res Clin Gastroenterol 2003; /17:/397/425. [20] Insenmann R, Rau B, Beger HG. Bacterial infection and extent of necrosis are determinants of organ failure in patients with acute necrotizing pancreatitis. Br J Surg 1999;/86:/ 1020/4. [21] Sitges-serra A, Lopez MJ, Girvent M, Almirall S, Sancho JJ. Postoperative enterococcal infection after treatment of complicated intra-abdominal sepsis. Br J Surg 2002;/89:/ 361/7. [22] Balthazar EJ, Robinson DL, Megibow AJ. Acute pancreatitis: value of CT in establishing prognosis. Radiology 1990;/174:/ 331/6. [23] Otto W, Paluszkiewicz R, Suchowera D, Jeziorek J, Krawczyk M, Karwowski A. Infected necrosis / the fatal outcome of acute pancreatitis. Proceedings of the 3rd World Congress of IHPBA 1999;/1:/332/5. [24] Otto W, Komorzycki K, Krawczyk M. The results of antibiotic therapy and surgery in patients with necrotizing pancreatitis. Proceedings of the 5th European Congress IHPBA, Istanbul, Turkey, 28/31 May 2003.