Towards a targeted, risk‐based, antifungal strategy in neutropenic patientsreview
Аннотация: The major issues in the management of fungal infection are prevention, diagnosis and treatment. Our main goal must remain prevention, but for a number of reasons prophylaxis against invasive fungal infection (IFI), in the neutropenic patient, remains controversial. We consider this aim to be desirable, given the high mortality rates associated with established infection, due in part to inadequacies and substantial delays in diagnosis. In a meta-analysis Bow et al (1997) , found a mortality rate of 47% in patients with invasive fungal infection. In a more recent prospective EORTC survey, Denning et al (1999) found that the 3-month mortality rate in patients with invasive aspergillosis was 64%. However, clinical trials of antifungal prophylaxis have failed to show a reduction in overall mortality in almost every case, other than those conducted in the highest risk patients. Fuel was recently added to the fire of this debate by the publication of a further meta-analysis on the role of prophylaxis and empirical treatment ( Gotzsche & Johansen, 1997) . We have criticized this report for a number of reasons ( Kibber et al, 1997 ) but do support the contention that more studies are needed using modern diagnostic tools, including PCR, as end-points, given the inadequacy of conventional antemortem methods. In particular, we feel that consideration should be given to the study of targeted prophylaxis and treatment in defined risk groups. IFI is most commonly caused in Europe and North America by Candida spp. and Aspergillus spp.. Fusarium spp. and the mucorales are less common causes and cryptococcosis is seen almost exclusively in those with a T-cell deficiency. There has been a shift in the dominant species in most centres in Europe and N. America, where the use of prophylaxis by oral non-absorbable polyenes (amphotericin B or nystatin) has been replaced by the absorbable azoles and, more recently, the triazoles. The introduction of fluconazole led to a clear reduction in both colonization and invasive infection with Candida albicans. The reduction in candidosis due to fluconazole prophylaxis has been accompanied by an increase in invasive mould infections in some studies, e.g. 29% in bone marrow transplantation (BMT) patients given fluconazole vs. 18% in those not receiving it ( van Burik et al, 1998 ). What is consistently reported is the failure of fluconazole to prevent lower gastrointestinal tract (GIT) colonization in contrast to the unabsorbed polyenes. There have been few reports of fluconazole-resistant C. albicans infections occurring in neutropenic patients (in contrast to AIDS) but some major studies have documented an increased risk from C. krusei and C. glabrata, which are inherently resistant to this agent ( Wingard et al, 1991 ). By the use of the combination of fluconazole and oral amphotericin B, we have found a reduction in colonization with C. albicans in the absence of an increase in non-albicans species compared with our prior prophylaxis with ketoconazole and amphotericin B, in a longitudinal study, with well-documented cohorts separated by some years (C. albicans 48% to 23%, C. glabrata 18% to 18%, C. torulopsis 5% to 3%, C. krusei 5% to 1%). The overall rate of colonization by Candida species was therefore reduced (unpublished observations). The dominant species, in some bone marrow transplant centres, is now Aspergillus, for the prevention of which systemic agents such as itraconazole, voriconazole or amphotericin B lipid-associated formulations are used. Specific facilities including laminar airflow (LAF) or high-efficiency particulate air (HEPA) filtered rooms are also needed as acquisition is largely by the airborne route and > 95% of primary infections are in the lung. Studies in patients undergoing either chemotherapy or bone marrow transplantation have shown that the resulting neutropenia carries a risk for IFI of 2–40% ( Schwartz et al, 1984 ; Tollemar et al, 1989a , b; Goodrich et al, 1991 ; Verfaille et al, 1991; Goodman et al, 1992 ; McWhinney et al, 1993 ; Winston et al, 1993 ; Fetscher et al, 1999 ; Salazar et al, 1999 ). The risk is dependent on the underlying disease, the treatment given and the prophylaxis used and is higher in older patients ( Kiwan & Anaissie, 1999) . Historical data are likely to underestimate the risk of IFI during life because of inadequacies in diagnosis and, conversely, autopsy studies are likely to overestimate the risk given the likely recent antecedent immune status, e.g. very prolonged neutropenia. Within the neutropenic population, the risk of IFI ranges from 'low' in those with moderate (0·1–0·5 × 109/l) to 'high' in those with severe (< 0·1 × 109/l) neutropenia and, most substantially, with the length of neutropenia, less than 7 d carrying a minimal risk. However, this rises to a fourfold increased risk of invasive aspergillosis in those with > 21 d of profound neutropenia ( Gerson et al, 1984 ). Numerous studies have demonstrated additional factors associated with an increased risk of IFI and these are listed in Table I. We shall explore the evidence for some of these to assist in the construction of risk categories. Beyond the gastrointestinal mucosa, the upper airways and the epidermis, the next line of defence against IFI are the phagocytic elements of tissue macrophages and the blood neutrophils and monocytes. Fungal translocation is likely to be an occasional occurrence, as with transient bacteraemia from the oral flora and the lower GIT, yet those with normal numbers of functional phagocytic cells rarely suffer IFI. Macrophages are long lived and only defects in their function (see under corticosteroids) are likely to be of consequence. In contrast, the normal intravascular life span of the neutrophil is of the order of 8–12 h. Neutropenia is the dominant risk factor for IFI ( Goodrich et al, 1991; Goodman et al, 1992 ). Myelosuppressive chemotherapy, for example DAT (daunorubicin, cytarabine and 6-thioguanine) in acute myeloblastic leukaemia (AML), leads to an average 3 weeks of neutropenia, with 2 weeks being severe (< 0·1 × 109/l). Patients undergoing remission induction chemotherapy are thus in an intermediate risk group. That few infections are diagnosed probably reflects more the inadequacy of our tests and clearance of undetected IFI, with neutrophil recovery, than a true low incidence. It is reasonable to assume that those patients with relapsed and/or refractory disease who have prolonged moderate or severe neutropenia are at a greater risk. Shorter periods of neutropenia should carry minimal risk and may not require prophylaxis, such as for those with moderate (0·1–0·5 × 109/l) neutropenia for < 7 d, but even these patients can develop IFI. Fetscher et al (1999) found a 3% rate of invasive Aspergillus infection (with 85% mortality) in 683 patients receiving peripheral blood stem cell (PBSC) rescue after high-dose chemotherapy treatment for a range of solid tumours or lymphoma, although 9 of the 20 occurred at the time of disease relapse. Reliable rapid diagnostic methods would make targeted therapy more appropriate than prophylaxis for such patients. Within the allogeneic BMT group, higher donor stem cell dose has usually been found to improve survival. This could, in part, be explained by the significant (P = 0·01) impact of low cell dose on increasing the risk of IFI, found by Tollemar et al (1989a) . Acute myeloblastic leukaemia (AML) may be associated with a higher risk from invasive aspergillosis according to the EORTC study ( Denning et al, 1998 ) , which used retrospective reporting. Unfortunately, the denominator was not defined, making calculation of that risk impossible. Of interest, although more of the reported cases had AML [49% vs. 21% acute lymphoblastic leukaemia (ALL)], they had a better survival, perhaps reflecting the more common use of corticosteroids in the lymphoid malignancies. Within the AML category Bow et al (1995) found a higher risk in those with M0–2 AML with the greatest risk being in the M0 FAB subtype. Although the treatment-related mortality in children with AML undergoing remission induction, intensification and marrow transplant therapy is now modest, the single greatest cause of death was found to be infection (66%). Of the organisms identified, fungal infection was the most common, causing 23% of all infectious deaths with Aspergillus spp. being the usual isolate (MRC AML 10; Riley et al, 1999 ). The higher rate in AML could be due either to an intrinsic functional defect or to a reduction in the absolute numbers of neutrophils at the start of treatment. With the conventional chemotherapy used in the treatment of most solid tumours, the risk of IFI is presumed to be small and certainly the main risk factors of prolonged neutropenia and use of corticosteroids are uncommon. Thus, prophylaxis is not required, but any patient developing an antibiotic-resistant fever should be considered at risk. An informative study in allograft recipients randomized to receive or not methylprednisolone, in addition to standard cyclosporin A plus methotrexate, as graft-versus-host disease (GVHD) prophylaxis was performed by Sayer et al (1994) at Seattle. The corticosteroid was given during the first 35 d after transplant at a dose of 1 mg/kg until day 22 and then reduced. Both bacterial and fungal infection rates were increased, with Candida albicans being the most common fungal isolate. Acute GVHD was an independent risk factor in this study, providing evidence that the combination of these factors is of particular risk. In a study of GVHD prophylaxis by the City of Hope team, the introduction of corticosteroids at a dose of 0·5–1 mg/kg/d resulted in a sixfold increase in the incidence of IFI ( O'Donnell et al, 1994 ). Other than Gram-positive skin commensals, yeasts are the most frequent cause of line infections and a potential source of invasive disease ( Lecciones et al, 1992 ). Candida parapsilosis has a high affinity for plastics and is increasingly observed as a local and sometimes systemic cause of fungal infection ( Weems, 1992). Scrupulous catheter care is appropriate and the presence of such a device is not sufficient reason alone for antifungal drug prophylaxis ( Smith et al, 1997 ) The risk of IFI in allogeneic BMT recipients varies from 10% to 40% (for a review, see Bow et al, 1995 ). In those with an uncomplicated HLA-matched sibling transplant, it is as low as 10%, for those with an HLA-matched unrelated donor (MUD) the risk is increased ( Jantunen et al, 1997 ), whereas in those with a mismatched (MM) family donor it can be as high as 33% (F. Aversa, personal communication). In the last study, profound T-cell depletion was used as effective GVHD prophylaxis. What contribution is made by the HLA disparity is presently unknown. Several publications have confirmed an increased risk of IFI in those with GVHD, including extensive chronic GVHD (for example, see Tollemar et al, 1989a ; Sayer et al, 1994 ; Jantunen et al, 1997 ). It is important to be aware that there has been a shift to and an increase in the rate of late IFI, especially aspergillosis, associated with GVHD and its treatment, and is not always associated with neutropenia ( Wald et al, 1997 ). In fact, the authors of this report found a bimodal distribution. Graft rejection by increasing the period of neutropenia (and sometimes complicated by the inappropriate use of corticosteroids) substantially increases this risk. In our own (unpublished) series during the early 1980s, the risk of invasive aspergillosis was 39%, mainly in the recipients of unrelated donor transplants. This was probably an underestimate. TBI is a potent cause of gastrointestinal damage, the probable cause for its identification as a risk factor for a higher rate of IFI in BMT recipients ( Tollemar et al, 1989a ; Goodrich et al, 1991 ; Verfaille et al, 1991 ). Prior colonization is almost a prerequisite of invasive candidosis and is predictive of subsequent infection ( Sandford et al, 1980 ; Schwartz et al, 1984 ). Clear evidence exists for an increased risk of IFI if more than one site or heavy colonization at a single site is found ( Guiot et al, 1994 ). This group showed that fungal infection was infrequent (5%) in those not colonized compared with 35% in those heavily colonized at more than one site. Although Bow et al (1995) showed only a non-significant trend with multiple sites of colonization, they did show an increased risk of IFI (P = 0·032) in those patients with rectal colonization. F. Aversa (personal communication) showed, in a very high-risk population having mismatched T-cell-depleted PBSC transplants, that colonization was the most significant factor in predicting fungal related death with 68% of patients dying compared with 7% in those not colonized. Colonization with Candida tropicalis is highly predictive of invasion, such that we, and others, will treat as if there is an established infection on discovering colonization at a single site ( Wingard et al, 1979 ; Sandford et al, 1980 ). Lass-Florl et al (1999) found 63% of patients having biopsy material taken at thoracic surgery or autopsy, after sudden death, had fungal colonization (37% Aspergillus) of the lung. However, this figure seems extremely high and far in excess of the experience of most other authors. An informative study from Aisner et al (1979) showed, in a cohort of 125 patients with AML, that 10 of 11 who had nose cultures positive for A. flavus or A. fumigatus developed IFI, compared with only 8 of 114 with sterile cultures (P = < 0·000001). Thus, a positive culture is predictive for IFI but a negative culture does not preclude such a development. More recently, Einsele et al (1998) have demonstrated the value of sampling from the lower respiratory tract and using PCR as a more sensitive and specific test. Bow et al (1995) showed more IFI in those with bacteraemia (59%) compared with those without (32%; P = 0·03). Sparrelid et al (1998) reported that of those patients with proven bacteraemia, who had a mortality rate of 27%, there were 21 deaths out of a total of 45 which were due to fungal infection compared with 6 of 25 who had negative bacterial blood cultures and 5% mortality. Recent bacteraemia has also been shown to be a risk factor by other groups ( Guiot et al, 1994 ). Hence, it would seem that these patients are candidates for pre-emptive therapy. The alteration of the natural gastrointestinal flora by the administration of broad spectrum antibiotics is a well-established cause of fungal colonization which, in turn, increases the risk for IFI ( Schwartz et al, 1984 ; Richet et al, 1991 ; Bow et al, 1995; Cole et al, 1996 ; Walsh et al, 1996 ). Several studies have suggested an increased risk from IFI in CMV seropositive patients undergoing BMT. In the study from Huddinge, this was found to be a significant (P = 0·01) factor ( Tollemar et al, 1989b ). Gastrointestinal damage caused by high-dose (HD) Ara-C plus etoposide was implicated as the principal factor leading to a substantially increased rate of IFI by Bow et al (1995) . In this retrospective study, IFI was seen in 36% of patients receiving HD Ara-C in contrast to 6% and 2·6% seen in previous and concurrent chemotherapies used in induction treatment for AML, which did not include HD Ara-C. Although no formal reports are as yet available, our extensive local experience suggests that the combination of fludarabine with moderate to high-dose Ara-C with/without idarubicin (FLAG ± Ida) is one with a particularly high risk, perhaps because of a combination of a T-cell deficiency and neutropenia. In 76 patients having 112 courses, we saw 121 episodes of fever, 45% were classified as pyrexia of unknown origin (PUO) but 18 were due to proven or probable pulmonary aspergillosis and one sinus infection was seen with the same organism giving an incidence of 16% (unpublished observations). Fludarabine is a powerful immunosuppressant with a prolonged length of effect (months) on lymphopoiesis and a medium-term (3–5 weeks) myelosuppression, which suggests that this agent even when used alone might place patients in a high-risk category. A global approach, i.e. prophylaxis for all neutropenic patients, is not appropriate. Although the risk is probably small, drug-related toxicity and emergence of resistance could outweigh any benefit for those at low risk of infection. Evidence for the benefit of modern absorbed triazoles comes from a number of randomized studies in different risk groups. It is inappropriate to include older less effective drugs such as nystatin ( De Gregorio et al, 1982 ) , except perhaps by the parenteral route in a liposomal formulation which is presently undergoing evaluation, or the older azoles such as ketoconazole in an assessment of benefit. We have therefore concentrated on the triazoles and take evidence principally from the larger randomized studies, including those in which polyenes were used as the control agent. Fluconazole has been compared with polyenes in three important studies in adults, first at a 'low' dose of 50 mg/d ( Philpott-Howard et al, 1993 ). The control group received either amphotericin B and/or nystatin and 110 of 536 entered were BMT recipients. There was a significant reduction in the risk of fungal infection (3·9% vs. 12·2%; P = 0·001) in 511 evaluable patients randomized, although the difference in proven invasive infection did not reach significance. Although colonization rates were similar, there was significantly more colonization with fecal isolates in the fluconazole arm. Included in this study was a subgroup of 50 patients ( Rosenberg-Arska et al, 1991 ) who received fluconazole at 50 mg/d compared with amphotericin B, given as suspension plus tablets (each 200 mg four times per day) . Although local and invasive infection rates were equivalent, amphotericin B was far more effective in preventing colonization in the lower GI tract, with 4% compared with 54% persistent positive stool cultures in the fluconazole arm, which considering the almost total absorption of fluconazole in the upper GI tract is not surprising. Thus the benefit is modest, at best, in these low-dose studies. In 154 episodes of chemotherapy-induced neutropenia, Schaffner & Schaffner (1995) reported that 400 mg/d fluconazole compared with placebo significantly increased the time to use of i.v. amphotericin B (P = 0·003) and reduced the incidence of fever by 20%. Although oropharyngeal candidiasis was significantly reduced, there was no effect on invasive disease. This study showed an adverse effect on neutropenia, not reported elsewhere. A further recent small study failed to show any benefit for fluconazole prophylaxis at 400 mg/d in patients with relapsed/refractory AML ( Kern et al, 1998 ). A randomized, placebo-controlled, double-blind, multicentre trial in 257 adults undergoing chemotherapy for acute leukaemia compared fluconazole (400 mg orally once daily or 200 mg intravenously every 12 h) with placebo ( Winston et al, 1993 ). Fluconazole decreased fungal colonization [83 of 122 (68%) in placebo patients compared with 34 of 119 (29%)] at the end of prophylaxis (P < 0·001) and proven fungal infections with 27 of 132 (21%) seen in placebo patients compared with 11 of 123 (9%) in the fluconazole patients (P = 0·02). Superficial fungal infections occurred in 20 of 132 (15%) placebo patients but in only 7 of 123 (6%) fluconazole patients (P = 0·01), whereas invasive fungal infections developed in 10 of 132 (8%) placebo patients and in 5 of 123 (4%) fluconazole patients (P = 0·3). Fluconazole was effective in eliminating colonization and infection by Candida species other than Candida krusei with 66 of 122 (64%) placebo patients colonized at the end of prophylaxis compared with 11 of 119 (9%) fluconazole patients (P < 0·001) and 22 of 132 (17%) placebo patients infected compared with 7 of 123 (6%) fluconazole patients (P = 0·005). There was no difference seen in Aspergillus infections (fluconazole and placebo three cases each). The use of amphotericin B, the incidence of drug-related side-effects and overall mortality were similar in both study groups ( Winston et al, 1993 ). Fluconazole was compared with oral amphotericin B in a large (820 patients) randomized multicentre Italian study for the prevention of fungal infection in neutropenic patients with acute leukaemia. This compared 150 mg fluconazole, as a once daily capsule, with oral amphotericin B suspension (500 mg every 6 h). Definite systemic fungal infection occurred in 2·6% of fluconazole recipients and 2·5% of amphotericin B recipients; suspected systemic fungal infection requiring the empirical use of intravenous amphotericin B occurred in 16% of fluconazole recipients and 21% of oral amphotericin B recipients (P = 0·07). Superficial fungal infection was found in 1·7% of fluconazole recipients compared with 2·7% of amphotericin B recipients (P > 0·2). Side-effects and compliance were less satisfactory with the amphotericin ( Menichetti et al, 1994 ). In a substantial multicentric study in 502 children, Ninane (1994) showed a significant reduction in IFI, with 2·1% in the fluconazole arm compared with 8·4% with the polyenes (P = 0·002). The dose of fluconazole required to achieve an appropriate MIC for each of the candida species is presently unknown (Rex et al, 1997). Itraconazole in the original capsule formulation suffered from poor bioavailability of the drug, which has been overcome by the introduction of an itraconazole–cyclodextrin complex in solution. Pharmacokinetic studies of this solution have shown high Cmin values with 1468 ± 306 ng/ml at day 15 in chemotherapy patients ( Prentice et al, 1995 ), 845 ± 221 in autograft recipients at day 15 ( Prentice et al, 1994 ) and 300 ± 90 ng/ml in allograft recipients at day 7 ( Michallet et al, 1998 ). In the last study, the level of hydroxyitraconazole was 550 ± 120 ng/ml. This first pass metabolite achieves a steady-state concentration twice that of itraconazole and has equivalent antifungal activity. All data were generated in adult patients. Unpublished data in children show a day 15 Cmin of 711 ± 251 at age 1·7–4·9 years and 1072 ± 408 at age 6·2–14·3 years (S. Lemerle; Janssen Clinical Research Report) . These levels are well above the minimum inhibitory concentrations (MICs) for Aspergillus spp. reported as required for effective prophylaxis by Boogaerts et al (1989) and for mucosal Candida species infections (C. albicans and non-albicans spp.). It is apparent that higher blood levels are, in general, better ( Menichetti et al, 1999 ), although there are as yet insufficient data available and the incidence of resistant isolates requires enumeration. Three large randomized controlled trials have compared itraconazole solution (5 mg/kg/d) with a variety of agents. In one study of 405 chemotherapy patients, cyclodextrin was the placebo and there was an unexpected, but not significant, excess of proven deep Aspergillus spp. infections with itraconazole use (1 vs. 4), one being fatal in each arm. This trial, which was conducted in over 30 centres, used a placebo with known gut toxicity and therefore, surprisingly, reported very high compliance rates ( Menichetti et al, 1999 ). The UK study ( Morgenstern et al, 1999 ) of itraconazole solution vs. fluconazole solution (100 mg/d) included chemotherapy, autograft and allograft patients, some of whom were entered more than once during repeated neutropenic episodes to the same arm (581 episodes). Up to 100 d after the start of prophylaxis, no proven, probable or possible deep Aspergillus spp. infections were seen with itraconazole, with only one C. albicans proven fungaemia and no fungal deaths. In comparison, with fluconazole, there were seven fungal deaths [proven: five Aspergillus spp., one C. tropicalis; probable: one Aspergillus spp. (P = 0·024)]. In the first randomized episodes, there were six proven deep infections with fluconazole and none with itraconazole (P = 0·016). All proven infections were in chemotherapy patients. Compliance was a major problem because of cyclodextrin gut toxicity. In a French study of 557 chemotherapy patients, amphotericin B capsules (2 g/d) were used as a comparator ( Harousseau et al, 1998 ). There were more proven IFI (13 vs. 8), proven Aspergillus spp. (9 vs. 5) and fungal deaths (5 vs. 1) in the amphotericin arm than with itraconazole, but none of these differences were statistically significant. This trial also reported significant problems with compliance. Despite the differences between their designs and outcomes, all three trials reported a lower overall incidence of fungal infection, a lower death rate from fungal infection and a lower use of i.v. amphotericin B for suspected IFI in patients given itraconazole solution. A double-blinded study that compared itraconazole 200 mg/d with fluconazole 100 mg/d, as capsules, in patients having chemotherapy with or without autologous stem cell transplantation showed no benefit for the itraconazole in any of the relevant parameters, despite achieving blood levels of 1·04 mg/l of itraconazole plus metabolite. The active metabolite accounted for about 66% of this ( Huijgens et al, 1999 ). This study emphasized the need for better absorbed preparation and clarification of the required blood levels for effective prophylaxis. The recent availability of both the oral solution and an intravenous preparation of itraconazole has allowed flexibility in prophylaxis suitable for most situations ( Prentice et al, 1999 ). The first randomized trial in allogeneic BMT by Goodman et al (1992) compared high-dose (400 mg/d) fluconazole with placebo. This blinded study, conducted during the neutropenic period, found a significant reduction in superficial infection (P = 0·001), invasive infection (P = 0·001) (15·8% vs. 2·8%) and fungal-related deaths (1/179 vs. 10/177; P = 0·001) by fluconazole. Only one each of C. krusei, Aspergillus spp. and a zygomycete were isolated as causative organisms for IFI in the fluconazole arm. Slavin et al (1995) conducted a double-blinded, randomized, placebo-controlled trial of fluconazole at 400 mg/d against placebo in recipients of allogeneic or autologous marrow transplant. Fungal infection in 10 of 152 (7%) patients was reduced when compared with 26 of 148 (18%) seen in those receiving placebo (P = 0·004). No Candida albicans was seen in the fluconazole group and superficial infection, fungal colonization and the empirical use of amphotericin B were all significantly reduced. Survival at day 110 was significantly improved in the fluconazole arm. Prophylactic intravenous amphotericin B has generally been disappointing when used at a low dose as a single agent for prophylaxis ( Perfect et al, 1992 ), although one study, using 5 or 10 mg/d, showed a reduction in fungal infection from 30% to 9% in recipients of allogeneic BMT ( O'Donnell et al, 1994 ). Patients having haploidentical peripheral blood stem cell transplants with profound T-cell depletion of the PBSC grafts for GVHD prevention are at very high risk of developing IFI. Long-term (d−10 to 5 months after BMT) liposomal amphotericin B has been used with apparent benefit in decreasing the fungal-related mortality from 33% to 23% compared with an historical cohort (F. Aversa, personal communication). Other studies in lower-risk groups support a possible role for this non-toxic preparation of amphotericin B, ranging from reduced colonization to reduced IFI ( Tollemar et al, 1993; Kelsey et al, 1999 ). Comparison with an historical control group has suggested a substantial reduction in risk of invasive fungal infection in children undergoing intensive induction therapy for leukaemia using 1 mg/kg AmBisome on alternate days. There was a reduction from an historical incidence of 15% (30 of 198) to 0% in 88 ( Ritter et al, 1995 ). Of interest, in this study, the incidence in AML was 47% in contrast to 8·5% in ALL and the incidence in those with indwelling central catheters was 39%. Whereas Candida spp. were the most common cause of infection, Aspergillus spp. infection was seen only in 4·5%. The potential benefit of AmBisome is the subject of an ongoing randomized trial by the same group (J. Ritter, personal communication). Additional support for a role in prophylaxis comes from the observation of a reduction in the emergence of fungal infection in a large North American randomized trial of empirical therapy ( Walsh et al, 1999 ). Other than some 'negative' studies mentioned already, some studies have suggested adverse consequences of triazole prophylaxis. Kern et al (1998) conducted a randomized study of the addition of high-dose (400 mg/d) fluconazole to 'standard' cotrimoxazole, colistin and amphotericin B suspension in 68 patients. They were considered to be at high risk because of GI-toxic drugs (intermediate/HD Ara-C) used for the treatment of refractory/relapsed disease and an anticipated prolonged period of neutropenia. The incidence of PUO was similar, but the rate of microbiologically defi
Год издания: 2000
Авторы: H. G. Prentice, C Kibbler, Archie Prentice
Издательство: Wiley
Источник: British Journal of Haematology
Ключевые слова: Antifungal resistance and susceptibility, Fungal Infections and Studies, Neutropenia and Cancer Infections
Другие ссылки: British Journal of Haematology (HTML)
PubMed (HTML)
PubMed (HTML)
Открытый доступ: closed
Том: 110
Выпуск: 2
Страницы: 273–284