Antimicrobial regimens prescribed by Canadian physicians for chemotherapy-induced febrile neutropenic episodes

Investigators and participating centres in the Canadian Fluconazole Prophylaxis Study Group include: Dr EJ Bow, Health Sciences Centre, Winnipeg, Manitoba; Dr G Evans, Kingston General Hospital, Kingston, Ontario; Dr IW Fong, St Michael’s Hospital, Toronto, Ontario; Dr G Garber, Ottawa General Hospital, Ottawa, Ontario; Dr D Haase, Queen Elizabeth II Health Sciences Centre, Halifax, Nova Scotia; Dr R Horn, Royal Victoria Hospital, Montreal, Quebec; Dr M Laverdière, Hôpital Maisonneuve-Rosemont, Montréal, Québec; Dr P Phillips, St Paul’s Hospital, Vancouver, British Columbia; Dr C Rotstein, Henderson Site, Hamilton, Ontario; Dr R Saginur, Ottawa Civic Hospital, Ottawa, Ontario; Dr I Salit, The Toronto Hospital, Toronto, Ontario; Dr C Sinave, Centre U Sante I’Estrie-Fleurimont, Sherbrooke, Québec; Dr F Smaill, McMaster Site, Hamilton, Ontario Presented in part at the 10th International Symposium on Infections in the Immunocompromised Hosts, June 21-24, 1998 Davos, Switzerland (Abstract 22) Department of Microbiology and Infectious Diseases, Hôpital Maisonneuve-Rosemont, Montréal, Québec; Department of Medicine and Medical Microbiology, Health Sciences Centre, The University of Manitoba and The Manitoba Cancer Treatment and Research Foundation, Winnipeg, Manitoba; Division of Infectious Diseases, Department of Medicine, Henderson Site, Hamilton Health Sciences Corporation, McMaster University, Hamilton, Ontario; Pharmaceuticals Group, Pfizer Canada, Limited, Pointe-Claire, Québec Correspondence and reprints: Dr M Laverdière, Department of Microbiology and Infectious Diseases, Hôpital Maisonneuve-Rosemont, 5415 boul L’Assomption, Montréal, Québec H1T 2M4. Telephone 514-252-3817, fax 514-252-3898, e-mail laverdim@ere.umontreal.ca Received for publication September 28, 1998. Accepted January 2, 1999 M Laverdière, EJ Bow, C Rotstein, S Ioannou, D Carr, N Moghaddam and the Canadian Fluconazole Study Group. Antimicrobial regimens prescribed by Canadian physicians for chemotherapy-induced febrile neutropenic episodes. Can J Infect Dis 1999;10(5):353-357.


OBJECTIVE:
To study the antimicrobial management of cancer patients with chemotherapy-induced neutropenia by Canadian physicians.SETTING: A cohort of 274 cancer patients with severe neutropenia (ie, less than 0.5×10 9 neutrophils/L) who participated in a prospective double-blind, placebo controlled study on antifungal prophylaxis conducted in 14 Canadian university-affiliated centres.Antifungal prophylaxis (oral fluconazole 400 mg daily) was administered to 153 of 274 (56%) patients.RESULTS: Antibacterial prophylaxis with a quinolone was given to 87 patients (32%) at the onset of chemotherapy whereas trimethoprim/sulphamethoxazole was given to 56 (20%) patients.Fever (ie, 38°C or over) occurred in 216 (79%) patients after a median duration of neutropenia of four days (range one to 31 days).Empirical antibacterial antibiotics were administered in 214 febrile patients.In 164 (77%) patients antibiotics were started during the first 24 h of fever.Monotherapy with a third generation cephalosporin and duotherapy with a antipseudomonal beta-lactam and an aminoglycoside were prescribed in 69 (32%) and 61 (28%) of the febrile patients, respectively.Inclusion of vancomycin in the initial empirical regimen was noted in 32 (15%) patients.Modifications of the initial regimen occurred in 187 (87%) pa-

See next page
O ver the past three decades, important advances in bio- medical and pharmacological technology have had a dramatic impact on the care of patients with malignancies.Antibiotics have significantly improved the outlook for cancer patients.Improved preventive and therapeutic strategies for infections have contributed to reduce the morbidity and mortality observed in severely immunocompromised patients (1).Recommendations on the initial selection and subsequent management of empirical antibiotics in profoundly neutropenic patients with unexplained fever have been written and recently updated (2).Although these recommendations, based on scientific publications and peer reviewed information, are general and must be applied wisely with respect to individual situations, adherence to them has rarely been evaluated.We have recently conducted a prospective double-blind, placebo controlled study on antifungal prophylaxis in 14 Canadian university-affiliated centres (3).In this study's protocol, there were provisions for antibiotic management, but physicians were permitted to apply their own centres' usual standard of care.We, therefore, had the opportunity to monitor the use of antimicrobial agents in cancer patients with unexplained fever and report an observational study on the antimicrobial management of cancer patients with chemotherapy-induced neutropenia by Canadian physicians.

PATIENTS AND METHODS
This was a descriptive study on the use of antimicrobial agents in a cohort of severely neutropenic patients who participated in a prospective double-blind, placebo controlled study on antifungal prophylaxis.The use of antibiotics was assessed from the induction of chemotherapy to the initiation of parenteral antifungal agents (amphotericin B [Fungizone, Bristol-Myers Squibb Canada Inc, Montreal, Quebec]).Eligible patients were to undergo cytotoxic chemotherapy for acute leukemia or conditioning therapy for autologous bone marrow transplantation (aBMT).Patients with acute myeloid leukemia (AML) or acute lymphoblastic leukemia (ALL) undergoing remission induction, reinduction therapy after primary relapse or postremission consolidation therapy, and patients undergoing aBMT were included if they were expected to remain neutropenic (less than 0.5×10 9 neutrophils/L ) for more than seven days.Occurrence of fever (38.0°C or greater) related to the onset of neutropenia, and the use of antimicrobial agents for prophylaxis and treatment of febrile neutropenia were recorded.

RESULTS
The clinical characteristics of the 274 evaluable patients are shown in Table 1.The study was conducted during a 23month period (January 1994 to November 1995).Antifungal prophylaxis (oral fluconazole, 400 mg once daily) was administered to 153 (56%) patients according to the drug randomization protocol.Antibacterial prophylaxis was added at the onset of chemotherapy at the attending physicians' discretion.In 87 (32%) patients, a quinolone was used, whereas trimethoprim/sulphamethoxazole was used in 56 (20%) patients.Fever occurred in 216 (79%) patients (Figure 1) after a median of 9.5 days (range zero to 42 days) following the first day of chemotherapy and a median of three days (range one to 31 days) following the onset of neutropenia.Among the 216 patients

354
Can J Infect Dis Vol 10 No 5 September/October 1999 who became febrile, 50 (23%) received a quinolone and 46 (21%) trimethoprim/sulphamethoxazole.Febrile neutropenia occurred in 100 of 120 aBMT (83%, 95% CI 77% to 90%) compared with 116 of 154 non-BMT patients (75%, 95% CI 69% to 82%).The 95% CI for the difference between the aBMT and non-BMT patients was -1.5% to 17.5% (P=0.144,not significant).Empirical antibacterial antibiotics were administered in 214 febrile patients.In 164 (77%) patients, antibiotics were started during the first 24 h of fever and in 91% after 48 h of fever.Monotherapy with a third generation cephalosporin or a carbapenem and duotherapy with a antipseudomonal betalactam and an aminoglycoside were prescribed to 69 (32%), five (2%) and 61 (28%) febrile patients, respectively (Table 2).Vancomycin was included in the initial empirical regimen in 32 (15%) patients.The first modifications of the initial regimen occurred in 187 (87%) patients after a median of five days (range one to 28 days) (Figure 2).Systemic amphotericin B was added to the antibacterial regimen in 102 of the 214 (48%) patients (Figure 3).Its addition occurred after a median of nine days of empirical antibiotics and 13 days after the first day of cytotoxic therapy.A total of 18 patients (6.5%) died during the study.

DISCUSSION
The management of infection in neutropenic patients requires careful decision-making.Since the early 1970s, prompt empirical antimicrobial therapy has been advocated as a way to treat febrile episodes in neutropenic patients with cancer (4).Our observations show that, in general, the participating physicians in our study used prompt empirical antibacterial therapy.Empirical antibiotics were given to 99% of our febrile neutropenic patients, and, in 77% of these febrile patients, antimicrobial agents were started during the first 24 h of fever.
Antimicrobial regimens available for the treatment of febrile episodes in neutropenic cancer patients initially were recently reviewed by the Infectious Diseases Society of America's expert panel (2).There is a general consensus that when the etiology is unknown, there are very little differences between monotherapy and multidrug combinations in the treat-Can J Infect Dis Vol 10 No 5 September/October 1999 355 Antimicrobial regimens for febrile neutropenic episodes  ment of uncomplicated episodes of fever in neutropenic patients.Third generation cephalosporins (ceftazidime or cefepime) or a carbapenem (imipenem/cilastatin or meropenem) can be considered standards of monotherapy, whereas antipseudomonal cephalosporins or penicillins combined with aminoglycosides represent usual combination therapy (5)(6)(7).In this population of patients, monotherapy was used in 90 of 214 (42%) patients, with ceftazidime or imipenem selected 77% of the time.Combination therapy was used in more than half (58%) of the patient population.In most cases, antipseudomonas cephalosporins or penicillins combined with aminoglycosides were used.Such combinations have the advantage to be potentially synergistic against some Gram-negative bacilli (8), and may also prevent emergence of resistance during treatment (9,10).
Vancomycin was included in the initial empirical regimen by our attending physicians in 32 of 214 (15%) patients.Controversy surrounds whether vancomycin should be part of the initial antimicrobial regimen in febrile neutropenic patients.Emerging resistance of enterococci to vancomycin and failure to improve the overall mortality associated with Gram-positive infections have been the major arguments against systematically adding vancomycin to the initial empirical regimen.However, at times, it is warranted because some Grampositive infection, particularly viridans streptococcal infections, may be fulminant and associated with a high mortality (11).To reduce the excessive use of vancomycin and to prevent the spread of vancomycin resistance, it is recommended that institutions with a low incidence of fulminant Gram-positive infections avoid using vancomycin systematically as part of the initial antimicrobial regimen in febrile neutropenia (12).However, it would be prudent to include vancomycin in the regimen of selected patients with the following: serious catheterrelated infections (13); intensive chemotherapy that causes substantial mucosal damage such as high dose cytarabine; hypotension or other evidence of cardiovascular impairment; Gram-positive blood stream infection before final identification and susceptibility testing; colonization with penicillin-resistant pneumococci or methicillin-resistant staphylococci and in patients who have received quinolone prophylaxis.
Antibiotic strategy, after the empirical phase often requires modification if fever persists.In our observations, modification of the antibiotic regimens was documented in 87% of our patient population, with the first modification occurring after a median of five days (range one to 28 days) of antibiotic therapy.This is in keeping with previous reports that the time-todefervescence for febrile neutropenic patients who receive antibiotic regimen was two to seven days (median time five days) (14).Therefore, at least three days of antibiotic treatment are usually required to determine the efficacy in clinically stable patients.
There is no consensus on the optimal timing for addition of amphotericin B in persistently febrile neutropenic patients despite the administration of broad-spectrum antibiotics in adequate dosage.Most experts agree that, after one week of profound neutropenia and fever, the addition of systemic antifungal is warranted due to the increased incidence of candida and aspergillus infections in this patient population (15).Amphotericin B remains the preferred selection in such situations.In our study, amphotericin B was added to the antibacterial regimen in 102 of 214 (48%) initially febrile patients.Although approximately half of our cohort of patients received prophylactic fluconazole, it proved no more successful than placebo in obviating the need for parenteral amphotericin B (3).  cians follows the current suggested guidelines to treat febrile patients promptly with broad-spectrum antibiotics, avoid the systematic addition of vancomycin in the initial regimen and allow, in clinically stable patients, two to seven days before modifying the initial regimen.

Figure 3 )
Figure 3) Administration of amphotericine B relative to the first day of chemotherapy and onset of empirical initial antibiotics

Figure 2 )
Figure 2) Time to the first modification of the antimicrobial regimen relative to the first day of empirical antibiotics
Laverdière et altients after a median of five days (range one to 28 days).Empirical systemic amphotericin B was added after a median duration of nine days (range one to 34 days) of the empirical antibacterial regimen.CONCLUSIONS:Key Words: Antimicrobial management; Cancer patients; Febrile neutropenia Antibiothérapie CONCLUSIONS : De façon globale, chez les patients cancéreux atteints de neutropénie induite par la chimiothérapie, l'antibiothérapie prescrite par les médecins canadiens est fidèle aux directives actuelles préconisées par l'Infectious Disease Society of America.

TABLE 1 Characteristics of the 274 evaluable patients Characteristic Value
BMT Bone marrow transplantFigure 1) Proportion of febrile patients relative to the first day of chemotherapy and the first day of neutropenia