- Email: [email protected]
Cefamandole versus cefonicid prophylaxis in cardiovascular surgery: A prospective study
Cefamandole Versus Cefonicid Prophylaxis in Cardiovascular Surgery: A Prospective Study Michael S. Gelfand, MD, Bryan P. Simmons, MD, Philip Schoettle, MD, Oscar B. Harrington, MD, Frank Martin, MD, Edmond W. Owen, MD, Rebecca B. Craft, RN, and Naseem Amarshi, PharmD Departments of Medicine, Cardiovascular Surgery, and Pharmacy, Methodist Hospital, Memphis, Tennessee
We randomized 400 patients who were scheduled for an elective cardiovascular operation involving median sternotomy to receive cefamandole nafate or cefonicid in a prospective double-blind study. Three hundred fiftyseven patients were evaluable for prophylactic efficacy. Chest wound and donor site infections and early prosthetic valve endocarditis occurred more frequently with cefonicid (11 patients, 6.3%) than with cefamandole (4 patients, 2.2%) ( p = 0.05). Three patients, all in the cefonicid group, required sternal debridement to control postoperative deep wound infections. Twenty-five mis-
cellaneous postoperative infections (urinary tract infection, pneumonia, intravenous site infection, bacteremia, sepsis, Clostridium difficile diarrhea) occurred in 16 patients (9.19%) in the cefonicid group and four in 4 patients (2.19%)in the cefamandole group (p = 0.003). These data indicate that cefamandole is superior to cefonicid in preventing both surgical wound infections and miscellaneous nonsurgical infections after cardiovascular operations.
T
board. The patients scheduled for an elective cardiovascular procedure involving a median sternotomy were offered participation in the study. Written informed consent was obtained from all participants. Patients less than 18 years of age, pregnant women, and patients with an active infection, severe allergy to penicillins or cephalosporins, or history of antibiotic therapy within seven days before enrollment were excluded. A patient could participate in the study only once. The patients showered or bathed with chlorhexidine gluconate soap the night before operation. Also that night, the hair at the operative sites was clipped. The patients were assigned to one of two prophylactic regimens by a table of random numbers. All randomization was performed by one of us (N.A.). Regimen A consisted of cefamandole, 2 g intravenously at induction of anesthesia, 1 g every two hours during operation, and 1g every four hours after operation for 72 hours, based on the data of Kaiser and associates [4]. Regimen B included cefonicid, 2 g intravenously at induction of anesthesia and 1 g intravenously every 24 hours after operation for 72 hours, for a total of three postoperative doses, based on the data of Dudley and associates [7]. Patients on regimen B received a placebo (5% dextrose in water) between the doses of cefonicid. The placebo was given at the same time intervals as cefamandole in regimen A so that the investigators remained blinded as to the regimen the patient was receiving. We did not measure serum antibiotic levels in this study. All antibiotic doses were dispensed in minibags labeled "Moncef." The duration of antibiotic prophylaxis was based on previously published studies and on a widely accepted practice of continuing the antibiotic or antibiotics until removal of all invasive devices (except a peripheral
he use of prophylactic antibiotics is an established practice in cardiovascular surgery [l]. The most costeffective prophylactic regimen has not been determined [2, 31. First-generation cephalosporins, especially cefazolin sodium, are widely used in prophylaxis in cardiovascular surgery [l, 31. Kaiser [4], Slama [5], and their colleagues have demonstrated a superiority of secondgeneration cephalosporins, cefamandole nafate and cefuroxime sodium, over cefazolin. As a result of these studies, cefamandole has become the antibiotic of choice among cardiovascular surgeons in our hospital and elsewhere. The currently popular prophylactic regimens require additional doses of antibiotics during prolonged surgical procedures and multiple daily doses postoperatively [l, 41. Cefonicid is a second-generation cephalosporin with a long serum half-life of four to five hours [6]. High levels of cefonicid are achieved in the atrial appendage tissue and pericardial fluid [7]. In a study of patients undergoing coronary artery bypass grafting, Beam and co-workers [8] demonstrated equal surgical infection rates after prophylaxis with cefonicid and cefamandole. Because of the convenience of once-daily administration and the potential cost advantages of cefonicid, we wanted to compare cefonicid and cefamandole for prophylaxis in cardiovascular surgery.
Material and Methods The study was sponsored by Smith Kline & French Laboratories and approved by the institutional review Accepted for publication Nov 17, 1989 Address reprint requests to Dr Gelfand, 188 S Bellevue, Suite 420, Memphis, TN 38104.
0 1990 by The Society of Thoracic Surgeons
(Ann Thorac Surg 1990;49:435-9)
0003-4975/90/$3.50
436
GELFAND ET AL CEFAMANDOLE VERSUS CEFONICID
Ann Thorac Surg 1990;49:4359
intravenous line). However, all antibiotic prophylaxis was stopped at 72 hours regardless of the status of invasive devices. To be evaluable, the patient had to adhere to the protocol for a minimum of 24 hours. The study was conducted between February 1988 and March 1989. All patients were operated on by one group of cardiovascular surgeons (P.S., O.B.H., E.W.O.). The mediastinum was irrigated with normal saline solution before sternal closure. After the sternum was closed, the chest wound was irrigated with a solution of kanamycin sulfate in normal saline solution (1 g/L of saline solution). A chest wound or donor site was considered infected if purulence was observed during daily examination by one of us and confirmed by another. Patients were monitored prospectively during hospitalization for the presence of nonsurgical wound infections (eg, bacteremia, pneumonia, urinary tract infection, intravenous site infection, Clostridium dificile diarrhea) by us and the infection control nurses. Standard definitions of infections (Centers for Disease Control) were used. Sepsis was defined as an acute febrile episode with negative cultures that was treated with antibiotics and for which no other specific cause was found. All patients were telephoned 30 days after discharge and queried about signs and symptoms of wound infection and use of antibiotics. The following variables were recorded prospectively for all patients: age and sex; name of the operating surgeon; duration of hospitalization before operation; smoking history; presence of obesity (greater than 20% over ideal body weight), diabetes mellitus, hypertension, and renal insufficiency (serum creatinine level greater than 2 mg/ dL); use of internal mammary artery in coronary artery bypass grafting; perfusion time; and cross-clamp time. According to the protocol design, after the first 400 patients were enrolled into the study, the infection rate for the two regimens, identified with a letter code, was calculated. The study was ended and the code broken when a significant difference was detected between the two regimens. All statistical tests for association of frequency were performed by x2 analysis or by Fisher's exact test when expected frequencies were less than five.
Results Forty-three of the 400 patients enrolled in the study were excluded. Those excluded were evenly distributed between the two groups. The reasons for exclusion included cancellation of operation, administration of an antibiotic
Table 1 . Surgical Wound Infections and Early Prosthetic Valve Endocarditis" Infection Chest wound Donor site Endocarditis Totalb
Cefamandole 2 (1, debridement) 2 0 4183 (2.2%)
Cefonicid
5 (3, sternal debridement) 5 1 (valve replacement) 111174 (6.3%)
a All were diagnosed during initial hospitalization or resulted in readrnission. Significance: p < 0.05 between the two groups.
Table 2. Microbiology of Surgical Infections Antibiotic Cefamandole Chest
Cefonicid Prosthetic valve endocarditis (mitral) Chest Chest Chest Chest Chest
Isolates Methicillin-resistant Staphylococcus epidermidis None Enterobacter cloacae; enterococcus Pseudomonas aeruginosa; methicillinresistant Staphylococcus epidermidis Methicillin-resistant Staphylococcus epidermidis Staphylococcus aureus Methicillin-resistant Staphylococcus epidermidis None Staphylococcus epiderrnidis Methicillin-resistant Staphylococcus aureus None Methicillin-resistant Staphylococcus aureus Enterobacter cloacae; enterococcus None None
not covered in the study, or death unrelated to infection during the study period. Three hundred fifty-seven patients were evaluable for prophylactic efficacy. One hundred eighty-three received cefamandole and 174, cefonicid. We observed no antibiotic side effects in our study. An analysis of surgeon-specific infection rates revealed no significant differences. Demographic and risk factors did not differ between the groups. The rates of surgical wound infection and early prosthetic valve endocarditis detected during the initial hospitalization or resulting in hospital readmission and surgical procedures are listed in Table 1. One chest and one donor site infection in the cefamandole group and one chest and two donor site infections in the cefonicid group were detected after discharge of the patients and resulted in readmission; they are included in Table 1. There were significantly more surgical wound infections in the cefonicid group, and 4 patients required major surgical procedures for control of infection (sternal debridement with pectoral muscle flaps, prosthetic valve replacement). The only death, possibly related to infection, occurred in the cefonicid group (deep sternal wound infection and purulent pericarditis secondary to Staphylococcus aureus with a postoperative aortic hemorrhage). One patient in the cefamandole group had chest wound debridement. All of the patients with infection received local wound care and intravenous antibiotics. The microbiological isolates from the surgical wound infections are listed in Table 2. Although the small number of infections prevents any conclusions about the relative frequency of infections with specific microorgan-
GELFAND ET AL CEFAMANDOLE VERSUS CEFONICID
Ann Thorac Surg 1990;49435-9
Table 3. Nonsurgical lnfecfions Site of Infection Lower respiratory tract Urinary tract Bacteremia
Sepsis Intravenous site Clostridium dificile diarrhea
Total" a
Cefamandole 2 2 0 0 0 0 41183 (2.19%)
Cefonicid 7 9 4 2 1 2
251174 (14.37%)
Significance: p < 0.001, number of infections between the two groups.
isms, S aureus was isolated only from the cefonicid group (n = 3). Five patients, 4 in the cefonicid group, had no microorganisms isolated on culture. We did not test the isolates against the study antibiotics, but methicillinresistant staphylococci, enterococci, and Pseudomonas aeruginosa usually are resistant to both cefamandole and cefonicid. Infections that occurred after the patient was discharged and were reported on a 30-day telephone survey but did not require readmission to the hospital are as follows: leg infection, 5 patients in the cefamandole group and 3 in the cefonicid group; chest infection, 2 in the cefonicid group. These were superficial infections that were managed by each patient's private physician with oral antibiotics and local care. Cultures usually were not performed. We chose not to include these infections in the calculations of surgical wound infection rates for the following reasons: the infections had minimal morbidity, did not require readmission or intravenous antibiotics, and were not confirmed microbiologically. The nonsurgical infections detected during the initial hospitalization are listed in Table 3. Four patients in the cefamandole group had four nonsurgical infections, and 25 nonsurgical infections occurred in 16 patients in the cefonicid group ( p = 0.003 for patients with infection). The microbiological isolates from nonsurgical infections are listed in Table 4. Two patients in the cefamandole group had a culture-negative pneumonia. The difference between the groups in nonsurgical infections is highly significant and resulted in frequent use of therapeutic antibiotics (not presented).
Comment The value of antibiotic prophylaxis in cardiovascular surgery has not been proved conclusively, but because of the high morbidity and mortality associated with sternal wound infections, mediastinitis, and prosthetic valve endocarditis, placebo-controlled trials are unlikely [2, 91. Furthermore, it is possible to infer the value of prophylaxis from comparative antibiotic trials because some antibiotics result in lower infection rates than others, and it is very unlikely that less-effective antibiotics cause infections [4, 5, 9, 101. Cardiovascular surgeons are, therefore, faced with a choice of the best and most cost-effective prophylactic
437
agents. The majority of surgical wound infections in cardiovascular surgery are caused by staphylococci (S aureus and Staphylococcus epidermidis), streptococci, and gram-negative aerobic rods [ll]. Until recently, firstgeneration cephalosporins (cephalothin sodium, cephapirin sodium, cefazolin) were the most popular agents. Several recent studies [4, 5, 121 showed a superiority of second-generation cephalosporins, cefuroxime and especially cefamandole, over cefazolin in cardiovascular surgery, although these findings have not been confirmed in other studies [13, 141. Several explanations have been offered for these results. Methicillin-resistant staphylococci (S epidermidis and S aureus) have increased in recent years and, as in our study, account for a high percentage of postoperative isolates [15]. Cefamandole is the most active of all cephalosporins in vitro against methicillin-resistant staphylococci and has been successfully used in clinical infections with methicillin-resistant S epidermidis and S aureus [15181. In our hospital, 25% of S aureus and 50% of S epidermidis infections are methicillin resistant. Kernodle and Kaiser [19] reported that surgical isolates of S aureus from St. Thomas Hospital in Nashville, TN, produced a p-lactamase capable of inactivating cefazolin but not cefamandole. These isolates were from the patients given prophylactic cefazolin. The activity of this plactamase against cefonicid has not been determined. Only one surgical wound infection with S aureus occurred in 814 patients given cefamandole in the studies by Kaiser and colleagues [4], Slama and co-workers [5], and ourselves [present study] and none in 705 patients given high-dose cefamandole prophylaxis [4, present study]. Cefamandole appears to be clearly superior to cefazolin in preventing cardiovascular surgical wound infections with S aureus ~51. Cefamandole does have a relatively short half-life (0.8 hour) and has to be administered every four hours (and more frequently intraoperatively) to maintain adequate trough levels [4]. Two second-generation cephalosporins with a prolonged half-life, ceforanide (half-life = three hours) and cefonicid (half-life = four to five hours), offer the conve-
Table 4 . Microbiology of Nonsurgical Infections Isolate
Cefamandole
Cefonicid
Staphylococcus aweus Staphylococciis epidermidis
Methicillin-resistant Staphylococcus aureus Enterococci Eschcrichia coli En terobncter cloacae Pseudomonas aeruginosn Miscellaneous
None
4 2 2 4 10" 0
They included the following: Klebsiella prieunioniae, Candida albicans, and Clostridiurn dificile, 2 patients each; and Acinetobucter unitratus, Branhamella cattnrrhnlis, group F streptococci, and viridans streptococci, 1 patient each.
438
GELFAND ET AL CEFAMANDOLE VERSUS CEFONICID
nience and potential cost advantage of less frequent administration [lo, 201. In a study published in 1984 [lo], ceforanide was shown to be superior to cephalothin in cardiovascular surgical prophylaxis. In 1987, Beam and colleagues [8] reported that the infection rate was similar in patients given cefonicid and cefamandole. A relatively small number of patients was studied, raising the question of a type I1 error. Our study was designed to compare the effectiveness of cefamandole and cefonicid for prophylaxis in cardiovascular surgery. In a 1986 review of antimicrobial prophylaxis in surgery, Kaiser [l]listed the criteria for designing studies of antimicrobial prophylaxis of surgical wound infections. The criteria included the following: prospective, randomized, double-blind design; written definitions of surgical wound infection; comparability of host risk factors; equal distribution of operative procedures; appropriate choice of antimicrobial agent or agents for the expected pathogens; adequate dosing throughout the operative procedure; controlled use of antibiotics or antiseptics not in the study; the use of antibiotics with a minimal risk of toxic or allergic reactions; sufficient numbers of patients to demonstrate significant differences (or lack thereof); and comparison of costs. All of these except the cost analysis were included in our study. The study revealed that the patients given cefamandole had significantly fewer surgical wound infections and nonsurgical infections than the cefonicid group. Therefore, we thought the cost analysis would be of secondary importance. Several possible explanations can be offered for these results. In vitro activity of cefonicid against staphylococci, both methicillin susceptible and methicillin resistant, is inferior to that of cefamandole, especially when tested with human serum [16,20,21]. Cefonicid is highly protein bound (98%) [20]. Serum levels of free (unbound) cefonicid during the second half of a 24-hour dosing interval may be too low to achieve adequate tissue levels [22]. Serum and tissue antibiotic levels during operation are known to be important in preventing surgical wound infections [23, 241. Low serum levels of free antibiotic can also result in suboptimal serum bactericidal activity, as reported by Chambers and colleagues [25] in patients failing once-daily cefonicid therapy for S aureus endocarditis. The high frequency of nonsurgical infections in patients given cefonicid is disturbing. These infections resulted in major morbidity and necessitated additional systemic antibiotic therapy. Although not specifically examined, the cost of hospitalization was undoubtedly higher in patients with infection. Krieger and associates [26] showed that urinary tract infections can result in postoperative surgical wound infections. We observed no such occurrence in our study. The design of the study does not allow us to determine whether cefonicid was less effective than cefamandole in preventing postsurgical infections or, although very unlikely, actually predisposed the patients to infections. A placebo group would have been needed to determine this. Flynn and co-authors [27] reported increased postoperative Enterobacter infections in cardiovascular surgical pa-
Ann Thorac Surg 1990;49:435-9
tients and related it to cefazolin prophylaxis. Kernodle and associates [28]demonstrated the emergence of methicillin-resistant coagulase-negative staphylococci in cardiac surgical patients given antibiotic prophylaxis. Only placebo-controlled studies can answer the question of whether antibiotic prophylaxis in cardiovascular surgery can actually predispose the patient to postoperative infections. Clostridiurn dificile diarrhea developed in 2 patients given cefonicid. Cannon and colleagues [29] reported C dificile diarrhea after perioperative antibiotic prophylaxis in orthopedic surgery. Surveillance after discharge has been emphasized recently [30]. We conducted a telephone survey 30 days after discharge and identified 10 patients with minor wound infections not necessitating rehospitalization or intravenous antibiotics. Our survey did not uncover any patients with serious infections who were readmitted to another hospital. All such patients returned to our hospital for care. Hospitals with different referral patterns may find postdischarge surveillance to be of greater value. In conclusion, we found that cefamandole is superior to cefonicid in preventing surgical wound infections and nonsurgical infections in patients after cardiovascular operations. Future studies of antibiotic prophylaxis in surgery should monitor both surgical wound and nonsurgical postoperative infections. We thank Smith Kline & French Laboratories for sponsoring this study.
References 1. Kaiser AB. Antimicrobial prophylaxis in surgery. N Engl J Med 1986;315:1129-38. 2. Farrington M. Antimicrobial prophylaxis. J Antimicrob Chemother 1986;18:656-8. 3. Gorbach SL. The role of cephalosporins in surgical prophylaxis. J Antimicrob Chemother 1989;23(Suppl D):61-70. 4. Kaiser AB, Petracek MR, Lea JW, et al. Efficacy of cefazolin, cefamandole, and gentamicin as prophylactic agents in cardiac surgery. Ann Surg 1987;206:791-7. 5. Slama TG, Sklar SJ, Misinski J, Fess SW. Randomized comparison of cefamandole, cefazolin, and cefuroxime prophylaxis in open-heart surgery. Antimicrob Agents Chemother 1986;29:7447. 6. Barriere SL, Hatheway GJ, Gambertoglio JG, Lin ET, Conte JE. Pharmacokinetics of cefonicid, a new broad-spectrum cephalosporin. Antimicrob Agents Chemother 1982;21: 935-8. 7. Dudley MN, Nightingale CH, Drezner AD, Low HB, Quintiliani R. Comparative penetration of cefonicid and cefazolin into the atrial appendage and pericardial fluid of patients undergoing open-heart surgery. Antimicrob Agents Chemother 1984;26:347-50. 8. Beam TR, Gordon G, Tuazon C. Cefonicid versus cefamandole prophylaxis in patients undergoing coronary artery bypass surgery (CABG) [Abstract 10591. Presented at the 27th Interscience Conference on Antimicrobial Agents and Chemotherapy, New York, NY, Oct 4-7, 1987. 9. Fong IW, Baker CB, McKee DC. The value of prophylactic antibiotics in aorta-coronary bypass operations. J Thorac Cardiovasc Surg 1979;78:90%13.
GELFAND ET AL CEFAMANDOLE VERSUS CEFONICID
Ann Thorac Surg 1990:49435-9
10. Platt R, Munoz A, Stella J, VanDevanter S, Koster JK. Antibiotic prophylaxis for cardiovascular surgery. Ann Intern Med 1984;101:77M. 11. Brook I. Microbiology of postthoracotomy sternal wound infection. J Clin Microbiol 1989;27806-7. 12. Meszaros R, Windisch M, Koltoi C, Lucacs L. Comparative study of cefamandole versus cephalothin as antibiotic prophylaxis for open heart surgery. Cor Vasa 1986;28:61-6. 13. Gentry LO, Zeluff BJ, Cooley DA. Antibiotic prophylaxis in open-heart surgery: a comparison of cefamandole, cefuroxime, and cefazolin. Ann Thorac Surg 1988;46:167-71. 14. Doebbeling BN, Massanari RM, Pfaller MA, Wenzel RP. Prophylaxis in open-heart surgery: a randomized, controlled comparison of cefazolin and cefuroxime [Abstract 11941. Presented at the 28th Interscience Conference on Antimicrobial Agents and Chemotherapy, Los Angeles, CA, Oct 2S26, 1988. 15. Sabath LD. Reappraisal of the antistaphylococcal activities of first-generation (expanded-spectrum) cephalosporins. Antimicrob Agents Chemother 1989;33:407-11. 16. Frongillo RF, Bianchi P, Moretti A, Pasticci BM, Ripa S, Pauluzzi S. Cross-resistance between methicillin and cephalosporins for staphylococci: a general assumption not true for cefamandole. Antimicrob Agents Chemother 1984;25:66&8. 17. Frongillo RF, Donati L, Federico G, et al. Clinical comparative study on the activity of cefamandole in the treatment of serious staphylococcal infections caused by methicillin-susceptible and methicillin-resistant strains. Antimicrob Agents Chemother 1986;29:789-96. 18. Coppens L, Hanson B, Klastersky J. Therapy of staphylococcal infections with cefamandole or vancomycin alone or with a combination of cefamandole and tobramycin. Antimicrob Agents Chemother 1983;23:3641. 19. Kernodle DS, Kaiser AB. Staphylococcus aureus (SA) infections failing cefazolin (CZ) or cefamandole (CM) prophylaxis (PX) in cardiac surgery (CS): clinical relevance of B-lactamase (BL) production [Abstract 871. Presented at the 27th Inter-
20. 21.
22.
23. 24. 25.
26. 27.
28.
29. 30.
439
science Conference on Antimicrobial Agents and Chemotherapy, New York, NY, Oct 4-7, 1987. Pontzer RE, Kaye D. Cefonicid: a long-acting, second-generation caphalosporin. Pharmacotherapy 1984;4:32.533. Dudley MN, Nightingale CH, Quintiliani R, Tilton RC. In vitro activity of cefonicid, ceforanide, and cefazolin against Staphylococcus aureus and Staphylococcus epiderrnidis and the effect of human serum. J Infect Dis 1983;148:178. Dudley MN, Shyu WC, Nightingale CH, Quintiliani R. Effect of saturable serum protein binding on the pharmacokinetics of unbound cefonicid in humans. Antimicrob Agents Chemother 1986;30:565-9. Goldmann DA, Hopkins CC, Karchmer AW, et al. Cephalothin prophylaxis in cardiac valve surgery. J Thorac Cardiovasc Surg 1977;73:470-9. Bergamini TM, Polk HC. The importance of tissue antibiotic activity in the prevention of operative wound infection. J Antimicrob Chemother 1989;23:301-13. Chambers HF, Mills J, Drake TA, Sande MA. Failure of a once-daily regimen of cefonicid for treatment of endocarditis due to Staphylococcus aureus. Rev Infect Dis 1984;6(Suppl 4): 87M. Krieger JN, Kaiser DL, Wenzel RP. Nosocomial urinary tract infections cause wound infections postoperatively in surgical patients. Surg Gynecol Obstet 1983;1563134. Flynn DM, Weinstein RA, Nathan C, Gaston MA, Kabins SA. Patients' endogenous flora as the source of "nosocomial" Enterobacter in cardiac surgery. J Infect Dis 1987;156:3634. Kernodle DS, Barg NL, Kaiser AB. Low-level colonization of hospitalized patients with methicillin-resistant coagulasenegative staphylococci and emergence of the organisms during surgical antimicrobial prophylaxis. Antimicrob Agents Chemother 1988;32:202-8. Cannon SR, Dyson PHP, Sanderson PJ. Pseudomembranous colitis associated with antibiotic prophylaxis in orthopaedic surgery. J Bone Joint Surg [Br] 1988;70-B:600-2. Reimer K, Gleed C, Nicolle LE. The impact of postdischarge infection on surgical wound infection rates. Infect Control 1987;8:237-40.
We randomized 400 patients who were scheduled for an elective cardiovascular operation involving median sternotomy to receive cefamandole nafate or cefonicid in a prospective double-blind study. Three hundred fiftyseven patients were evaluable for prophylactic efficacy. Chest wound and donor site infections and early prosthetic valve endocarditis occurred more frequently with cefonicid (11 patients, 6.3%) than with cefamandole (4 patients, 2.2%) ( p = 0.05). Three patients, all in the cefonicid group, required sternal debridement to control postoperative deep wound infections. Twenty-five mis-
cellaneous postoperative infections (urinary tract infection, pneumonia, intravenous site infection, bacteremia, sepsis, Clostridium difficile diarrhea) occurred in 16 patients (9.19%) in the cefonicid group and four in 4 patients (2.19%)in the cefamandole group (p = 0.003). These data indicate that cefamandole is superior to cefonicid in preventing both surgical wound infections and miscellaneous nonsurgical infections after cardiovascular operations.
T
board. The patients scheduled for an elective cardiovascular procedure involving a median sternotomy were offered participation in the study. Written informed consent was obtained from all participants. Patients less than 18 years of age, pregnant women, and patients with an active infection, severe allergy to penicillins or cephalosporins, or history of antibiotic therapy within seven days before enrollment were excluded. A patient could participate in the study only once. The patients showered or bathed with chlorhexidine gluconate soap the night before operation. Also that night, the hair at the operative sites was clipped. The patients were assigned to one of two prophylactic regimens by a table of random numbers. All randomization was performed by one of us (N.A.). Regimen A consisted of cefamandole, 2 g intravenously at induction of anesthesia, 1 g every two hours during operation, and 1g every four hours after operation for 72 hours, based on the data of Kaiser and associates [4]. Regimen B included cefonicid, 2 g intravenously at induction of anesthesia and 1 g intravenously every 24 hours after operation for 72 hours, for a total of three postoperative doses, based on the data of Dudley and associates [7]. Patients on regimen B received a placebo (5% dextrose in water) between the doses of cefonicid. The placebo was given at the same time intervals as cefamandole in regimen A so that the investigators remained blinded as to the regimen the patient was receiving. We did not measure serum antibiotic levels in this study. All antibiotic doses were dispensed in minibags labeled "Moncef." The duration of antibiotic prophylaxis was based on previously published studies and on a widely accepted practice of continuing the antibiotic or antibiotics until removal of all invasive devices (except a peripheral
he use of prophylactic antibiotics is an established practice in cardiovascular surgery [l]. The most costeffective prophylactic regimen has not been determined [2, 31. First-generation cephalosporins, especially cefazolin sodium, are widely used in prophylaxis in cardiovascular surgery [l, 31. Kaiser [4], Slama [5], and their colleagues have demonstrated a superiority of secondgeneration cephalosporins, cefamandole nafate and cefuroxime sodium, over cefazolin. As a result of these studies, cefamandole has become the antibiotic of choice among cardiovascular surgeons in our hospital and elsewhere. The currently popular prophylactic regimens require additional doses of antibiotics during prolonged surgical procedures and multiple daily doses postoperatively [l, 41. Cefonicid is a second-generation cephalosporin with a long serum half-life of four to five hours [6]. High levels of cefonicid are achieved in the atrial appendage tissue and pericardial fluid [7]. In a study of patients undergoing coronary artery bypass grafting, Beam and co-workers [8] demonstrated equal surgical infection rates after prophylaxis with cefonicid and cefamandole. Because of the convenience of once-daily administration and the potential cost advantages of cefonicid, we wanted to compare cefonicid and cefamandole for prophylaxis in cardiovascular surgery.
Material and Methods The study was sponsored by Smith Kline & French Laboratories and approved by the institutional review Accepted for publication Nov 17, 1989 Address reprint requests to Dr Gelfand, 188 S Bellevue, Suite 420, Memphis, TN 38104.
0 1990 by The Society of Thoracic Surgeons
(Ann Thorac Surg 1990;49:435-9)
0003-4975/90/$3.50
436
GELFAND ET AL CEFAMANDOLE VERSUS CEFONICID
Ann Thorac Surg 1990;49:4359
intravenous line). However, all antibiotic prophylaxis was stopped at 72 hours regardless of the status of invasive devices. To be evaluable, the patient had to adhere to the protocol for a minimum of 24 hours. The study was conducted between February 1988 and March 1989. All patients were operated on by one group of cardiovascular surgeons (P.S., O.B.H., E.W.O.). The mediastinum was irrigated with normal saline solution before sternal closure. After the sternum was closed, the chest wound was irrigated with a solution of kanamycin sulfate in normal saline solution (1 g/L of saline solution). A chest wound or donor site was considered infected if purulence was observed during daily examination by one of us and confirmed by another. Patients were monitored prospectively during hospitalization for the presence of nonsurgical wound infections (eg, bacteremia, pneumonia, urinary tract infection, intravenous site infection, Clostridium dificile diarrhea) by us and the infection control nurses. Standard definitions of infections (Centers for Disease Control) were used. Sepsis was defined as an acute febrile episode with negative cultures that was treated with antibiotics and for which no other specific cause was found. All patients were telephoned 30 days after discharge and queried about signs and symptoms of wound infection and use of antibiotics. The following variables were recorded prospectively for all patients: age and sex; name of the operating surgeon; duration of hospitalization before operation; smoking history; presence of obesity (greater than 20% over ideal body weight), diabetes mellitus, hypertension, and renal insufficiency (serum creatinine level greater than 2 mg/ dL); use of internal mammary artery in coronary artery bypass grafting; perfusion time; and cross-clamp time. According to the protocol design, after the first 400 patients were enrolled into the study, the infection rate for the two regimens, identified with a letter code, was calculated. The study was ended and the code broken when a significant difference was detected between the two regimens. All statistical tests for association of frequency were performed by x2 analysis or by Fisher's exact test when expected frequencies were less than five.
Results Forty-three of the 400 patients enrolled in the study were excluded. Those excluded were evenly distributed between the two groups. The reasons for exclusion included cancellation of operation, administration of an antibiotic
Table 1 . Surgical Wound Infections and Early Prosthetic Valve Endocarditis" Infection Chest wound Donor site Endocarditis Totalb
Cefamandole 2 (1, debridement) 2 0 4183 (2.2%)
Cefonicid
5 (3, sternal debridement) 5 1 (valve replacement) 111174 (6.3%)
a All were diagnosed during initial hospitalization or resulted in readrnission. Significance: p < 0.05 between the two groups.
Table 2. Microbiology of Surgical Infections Antibiotic Cefamandole Chest
Cefonicid Prosthetic valve endocarditis (mitral) Chest Chest Chest Chest Chest
Isolates Methicillin-resistant Staphylococcus epidermidis None Enterobacter cloacae; enterococcus Pseudomonas aeruginosa; methicillinresistant Staphylococcus epidermidis Methicillin-resistant Staphylococcus epidermidis Staphylococcus aureus Methicillin-resistant Staphylococcus epidermidis None Staphylococcus epiderrnidis Methicillin-resistant Staphylococcus aureus None Methicillin-resistant Staphylococcus aureus Enterobacter cloacae; enterococcus None None
not covered in the study, or death unrelated to infection during the study period. Three hundred fifty-seven patients were evaluable for prophylactic efficacy. One hundred eighty-three received cefamandole and 174, cefonicid. We observed no antibiotic side effects in our study. An analysis of surgeon-specific infection rates revealed no significant differences. Demographic and risk factors did not differ between the groups. The rates of surgical wound infection and early prosthetic valve endocarditis detected during the initial hospitalization or resulting in hospital readmission and surgical procedures are listed in Table 1. One chest and one donor site infection in the cefamandole group and one chest and two donor site infections in the cefonicid group were detected after discharge of the patients and resulted in readmission; they are included in Table 1. There were significantly more surgical wound infections in the cefonicid group, and 4 patients required major surgical procedures for control of infection (sternal debridement with pectoral muscle flaps, prosthetic valve replacement). The only death, possibly related to infection, occurred in the cefonicid group (deep sternal wound infection and purulent pericarditis secondary to Staphylococcus aureus with a postoperative aortic hemorrhage). One patient in the cefamandole group had chest wound debridement. All of the patients with infection received local wound care and intravenous antibiotics. The microbiological isolates from the surgical wound infections are listed in Table 2. Although the small number of infections prevents any conclusions about the relative frequency of infections with specific microorgan-
GELFAND ET AL CEFAMANDOLE VERSUS CEFONICID
Ann Thorac Surg 1990;49435-9
Table 3. Nonsurgical lnfecfions Site of Infection Lower respiratory tract Urinary tract Bacteremia
Sepsis Intravenous site Clostridium dificile diarrhea
Total" a
Cefamandole 2 2 0 0 0 0 41183 (2.19%)
Cefonicid 7 9 4 2 1 2
251174 (14.37%)
Significance: p < 0.001, number of infections between the two groups.
isms, S aureus was isolated only from the cefonicid group (n = 3). Five patients, 4 in the cefonicid group, had no microorganisms isolated on culture. We did not test the isolates against the study antibiotics, but methicillinresistant staphylococci, enterococci, and Pseudomonas aeruginosa usually are resistant to both cefamandole and cefonicid. Infections that occurred after the patient was discharged and were reported on a 30-day telephone survey but did not require readmission to the hospital are as follows: leg infection, 5 patients in the cefamandole group and 3 in the cefonicid group; chest infection, 2 in the cefonicid group. These were superficial infections that were managed by each patient's private physician with oral antibiotics and local care. Cultures usually were not performed. We chose not to include these infections in the calculations of surgical wound infection rates for the following reasons: the infections had minimal morbidity, did not require readmission or intravenous antibiotics, and were not confirmed microbiologically. The nonsurgical infections detected during the initial hospitalization are listed in Table 3. Four patients in the cefamandole group had four nonsurgical infections, and 25 nonsurgical infections occurred in 16 patients in the cefonicid group ( p = 0.003 for patients with infection). The microbiological isolates from nonsurgical infections are listed in Table 4. Two patients in the cefamandole group had a culture-negative pneumonia. The difference between the groups in nonsurgical infections is highly significant and resulted in frequent use of therapeutic antibiotics (not presented).
Comment The value of antibiotic prophylaxis in cardiovascular surgery has not been proved conclusively, but because of the high morbidity and mortality associated with sternal wound infections, mediastinitis, and prosthetic valve endocarditis, placebo-controlled trials are unlikely [2, 91. Furthermore, it is possible to infer the value of prophylaxis from comparative antibiotic trials because some antibiotics result in lower infection rates than others, and it is very unlikely that less-effective antibiotics cause infections [4, 5, 9, 101. Cardiovascular surgeons are, therefore, faced with a choice of the best and most cost-effective prophylactic
437
agents. The majority of surgical wound infections in cardiovascular surgery are caused by staphylococci (S aureus and Staphylococcus epidermidis), streptococci, and gram-negative aerobic rods [ll]. Until recently, firstgeneration cephalosporins (cephalothin sodium, cephapirin sodium, cefazolin) were the most popular agents. Several recent studies [4, 5, 121 showed a superiority of second-generation cephalosporins, cefuroxime and especially cefamandole, over cefazolin in cardiovascular surgery, although these findings have not been confirmed in other studies [13, 141. Several explanations have been offered for these results. Methicillin-resistant staphylococci (S epidermidis and S aureus) have increased in recent years and, as in our study, account for a high percentage of postoperative isolates [15]. Cefamandole is the most active of all cephalosporins in vitro against methicillin-resistant staphylococci and has been successfully used in clinical infections with methicillin-resistant S epidermidis and S aureus [15181. In our hospital, 25% of S aureus and 50% of S epidermidis infections are methicillin resistant. Kernodle and Kaiser [19] reported that surgical isolates of S aureus from St. Thomas Hospital in Nashville, TN, produced a p-lactamase capable of inactivating cefazolin but not cefamandole. These isolates were from the patients given prophylactic cefazolin. The activity of this plactamase against cefonicid has not been determined. Only one surgical wound infection with S aureus occurred in 814 patients given cefamandole in the studies by Kaiser and colleagues [4], Slama and co-workers [5], and ourselves [present study] and none in 705 patients given high-dose cefamandole prophylaxis [4, present study]. Cefamandole appears to be clearly superior to cefazolin in preventing cardiovascular surgical wound infections with S aureus ~51. Cefamandole does have a relatively short half-life (0.8 hour) and has to be administered every four hours (and more frequently intraoperatively) to maintain adequate trough levels [4]. Two second-generation cephalosporins with a prolonged half-life, ceforanide (half-life = three hours) and cefonicid (half-life = four to five hours), offer the conve-
Table 4 . Microbiology of Nonsurgical Infections Isolate
Cefamandole
Cefonicid
Staphylococcus aweus Staphylococciis epidermidis
Methicillin-resistant Staphylococcus aureus Enterococci Eschcrichia coli En terobncter cloacae Pseudomonas aeruginosn Miscellaneous
None
4 2 2 4 10" 0
They included the following: Klebsiella prieunioniae, Candida albicans, and Clostridiurn dificile, 2 patients each; and Acinetobucter unitratus, Branhamella cattnrrhnlis, group F streptococci, and viridans streptococci, 1 patient each.
438
GELFAND ET AL CEFAMANDOLE VERSUS CEFONICID
nience and potential cost advantage of less frequent administration [lo, 201. In a study published in 1984 [lo], ceforanide was shown to be superior to cephalothin in cardiovascular surgical prophylaxis. In 1987, Beam and colleagues [8] reported that the infection rate was similar in patients given cefonicid and cefamandole. A relatively small number of patients was studied, raising the question of a type I1 error. Our study was designed to compare the effectiveness of cefamandole and cefonicid for prophylaxis in cardiovascular surgery. In a 1986 review of antimicrobial prophylaxis in surgery, Kaiser [l]listed the criteria for designing studies of antimicrobial prophylaxis of surgical wound infections. The criteria included the following: prospective, randomized, double-blind design; written definitions of surgical wound infection; comparability of host risk factors; equal distribution of operative procedures; appropriate choice of antimicrobial agent or agents for the expected pathogens; adequate dosing throughout the operative procedure; controlled use of antibiotics or antiseptics not in the study; the use of antibiotics with a minimal risk of toxic or allergic reactions; sufficient numbers of patients to demonstrate significant differences (or lack thereof); and comparison of costs. All of these except the cost analysis were included in our study. The study revealed that the patients given cefamandole had significantly fewer surgical wound infections and nonsurgical infections than the cefonicid group. Therefore, we thought the cost analysis would be of secondary importance. Several possible explanations can be offered for these results. In vitro activity of cefonicid against staphylococci, both methicillin susceptible and methicillin resistant, is inferior to that of cefamandole, especially when tested with human serum [16,20,21]. Cefonicid is highly protein bound (98%) [20]. Serum levels of free (unbound) cefonicid during the second half of a 24-hour dosing interval may be too low to achieve adequate tissue levels [22]. Serum and tissue antibiotic levels during operation are known to be important in preventing surgical wound infections [23, 241. Low serum levels of free antibiotic can also result in suboptimal serum bactericidal activity, as reported by Chambers and colleagues [25] in patients failing once-daily cefonicid therapy for S aureus endocarditis. The high frequency of nonsurgical infections in patients given cefonicid is disturbing. These infections resulted in major morbidity and necessitated additional systemic antibiotic therapy. Although not specifically examined, the cost of hospitalization was undoubtedly higher in patients with infection. Krieger and associates [26] showed that urinary tract infections can result in postoperative surgical wound infections. We observed no such occurrence in our study. The design of the study does not allow us to determine whether cefonicid was less effective than cefamandole in preventing postsurgical infections or, although very unlikely, actually predisposed the patients to infections. A placebo group would have been needed to determine this. Flynn and co-authors [27] reported increased postoperative Enterobacter infections in cardiovascular surgical pa-
Ann Thorac Surg 1990;49:435-9
tients and related it to cefazolin prophylaxis. Kernodle and associates [28]demonstrated the emergence of methicillin-resistant coagulase-negative staphylococci in cardiac surgical patients given antibiotic prophylaxis. Only placebo-controlled studies can answer the question of whether antibiotic prophylaxis in cardiovascular surgery can actually predispose the patient to postoperative infections. Clostridiurn dificile diarrhea developed in 2 patients given cefonicid. Cannon and colleagues [29] reported C dificile diarrhea after perioperative antibiotic prophylaxis in orthopedic surgery. Surveillance after discharge has been emphasized recently [30]. We conducted a telephone survey 30 days after discharge and identified 10 patients with minor wound infections not necessitating rehospitalization or intravenous antibiotics. Our survey did not uncover any patients with serious infections who were readmitted to another hospital. All such patients returned to our hospital for care. Hospitals with different referral patterns may find postdischarge surveillance to be of greater value. In conclusion, we found that cefamandole is superior to cefonicid in preventing surgical wound infections and nonsurgical infections in patients after cardiovascular operations. Future studies of antibiotic prophylaxis in surgery should monitor both surgical wound and nonsurgical postoperative infections. We thank Smith Kline & French Laboratories for sponsoring this study.
References 1. Kaiser AB. Antimicrobial prophylaxis in surgery. N Engl J Med 1986;315:1129-38. 2. Farrington M. Antimicrobial prophylaxis. J Antimicrob Chemother 1986;18:656-8. 3. Gorbach SL. The role of cephalosporins in surgical prophylaxis. J Antimicrob Chemother 1989;23(Suppl D):61-70. 4. Kaiser AB, Petracek MR, Lea JW, et al. Efficacy of cefazolin, cefamandole, and gentamicin as prophylactic agents in cardiac surgery. Ann Surg 1987;206:791-7. 5. Slama TG, Sklar SJ, Misinski J, Fess SW. Randomized comparison of cefamandole, cefazolin, and cefuroxime prophylaxis in open-heart surgery. Antimicrob Agents Chemother 1986;29:7447. 6. Barriere SL, Hatheway GJ, Gambertoglio JG, Lin ET, Conte JE. Pharmacokinetics of cefonicid, a new broad-spectrum cephalosporin. Antimicrob Agents Chemother 1982;21: 935-8. 7. Dudley MN, Nightingale CH, Drezner AD, Low HB, Quintiliani R. Comparative penetration of cefonicid and cefazolin into the atrial appendage and pericardial fluid of patients undergoing open-heart surgery. Antimicrob Agents Chemother 1984;26:347-50. 8. Beam TR, Gordon G, Tuazon C. Cefonicid versus cefamandole prophylaxis in patients undergoing coronary artery bypass surgery (CABG) [Abstract 10591. Presented at the 27th Interscience Conference on Antimicrobial Agents and Chemotherapy, New York, NY, Oct 4-7, 1987. 9. Fong IW, Baker CB, McKee DC. The value of prophylactic antibiotics in aorta-coronary bypass operations. J Thorac Cardiovasc Surg 1979;78:90%13.
GELFAND ET AL CEFAMANDOLE VERSUS CEFONICID
Ann Thorac Surg 1990:49435-9
10. Platt R, Munoz A, Stella J, VanDevanter S, Koster JK. Antibiotic prophylaxis for cardiovascular surgery. Ann Intern Med 1984;101:77M. 11. Brook I. Microbiology of postthoracotomy sternal wound infection. J Clin Microbiol 1989;27806-7. 12. Meszaros R, Windisch M, Koltoi C, Lucacs L. Comparative study of cefamandole versus cephalothin as antibiotic prophylaxis for open heart surgery. Cor Vasa 1986;28:61-6. 13. Gentry LO, Zeluff BJ, Cooley DA. Antibiotic prophylaxis in open-heart surgery: a comparison of cefamandole, cefuroxime, and cefazolin. Ann Thorac Surg 1988;46:167-71. 14. Doebbeling BN, Massanari RM, Pfaller MA, Wenzel RP. Prophylaxis in open-heart surgery: a randomized, controlled comparison of cefazolin and cefuroxime [Abstract 11941. Presented at the 28th Interscience Conference on Antimicrobial Agents and Chemotherapy, Los Angeles, CA, Oct 2S26, 1988. 15. Sabath LD. Reappraisal of the antistaphylococcal activities of first-generation (expanded-spectrum) cephalosporins. Antimicrob Agents Chemother 1989;33:407-11. 16. Frongillo RF, Bianchi P, Moretti A, Pasticci BM, Ripa S, Pauluzzi S. Cross-resistance between methicillin and cephalosporins for staphylococci: a general assumption not true for cefamandole. Antimicrob Agents Chemother 1984;25:66&8. 17. Frongillo RF, Donati L, Federico G, et al. Clinical comparative study on the activity of cefamandole in the treatment of serious staphylococcal infections caused by methicillin-susceptible and methicillin-resistant strains. Antimicrob Agents Chemother 1986;29:789-96. 18. Coppens L, Hanson B, Klastersky J. Therapy of staphylococcal infections with cefamandole or vancomycin alone or with a combination of cefamandole and tobramycin. Antimicrob Agents Chemother 1983;23:3641. 19. Kernodle DS, Kaiser AB. Staphylococcus aureus (SA) infections failing cefazolin (CZ) or cefamandole (CM) prophylaxis (PX) in cardiac surgery (CS): clinical relevance of B-lactamase (BL) production [Abstract 871. Presented at the 27th Inter-
20. 21.
22.
23. 24. 25.
26. 27.
28.
29. 30.
439
science Conference on Antimicrobial Agents and Chemotherapy, New York, NY, Oct 4-7, 1987. Pontzer RE, Kaye D. Cefonicid: a long-acting, second-generation caphalosporin. Pharmacotherapy 1984;4:32.533. Dudley MN, Nightingale CH, Quintiliani R, Tilton RC. In vitro activity of cefonicid, ceforanide, and cefazolin against Staphylococcus aureus and Staphylococcus epiderrnidis and the effect of human serum. J Infect Dis 1983;148:178. Dudley MN, Shyu WC, Nightingale CH, Quintiliani R. Effect of saturable serum protein binding on the pharmacokinetics of unbound cefonicid in humans. Antimicrob Agents Chemother 1986;30:565-9. Goldmann DA, Hopkins CC, Karchmer AW, et al. Cephalothin prophylaxis in cardiac valve surgery. J Thorac Cardiovasc Surg 1977;73:470-9. Bergamini TM, Polk HC. The importance of tissue antibiotic activity in the prevention of operative wound infection. J Antimicrob Chemother 1989;23:301-13. Chambers HF, Mills J, Drake TA, Sande MA. Failure of a once-daily regimen of cefonicid for treatment of endocarditis due to Staphylococcus aureus. Rev Infect Dis 1984;6(Suppl 4): 87M. Krieger JN, Kaiser DL, Wenzel RP. Nosocomial urinary tract infections cause wound infections postoperatively in surgical patients. Surg Gynecol Obstet 1983;1563134. Flynn DM, Weinstein RA, Nathan C, Gaston MA, Kabins SA. Patients' endogenous flora as the source of "nosocomial" Enterobacter in cardiac surgery. J Infect Dis 1987;156:3634. Kernodle DS, Barg NL, Kaiser AB. Low-level colonization of hospitalized patients with methicillin-resistant coagulasenegative staphylococci and emergence of the organisms during surgical antimicrobial prophylaxis. Antimicrob Agents Chemother 1988;32:202-8. Cannon SR, Dyson PHP, Sanderson PJ. Pseudomembranous colitis associated with antibiotic prophylaxis in orthopaedic surgery. J Bone Joint Surg [Br] 1988;70-B:600-2. Reimer K, Gleed C, Nicolle LE. The impact of postdischarge infection on surgical wound infection rates. Infect Control 1987;8:237-40.