BACKGROUND:The toxigenic mold Stachybotrys has controversially been linked to idiopathic pulmonary hemorrhage and "sick building syndrome." However, there are no previous clinical records of invasive stachybotryosis.METHODS:Sinus biopsy specimens from a 23-year-old male with refractory acute lymphocytic leukemia were obtained at 3 different time points during the patient's hospitalization (139 days) and examined by histopathology and immunohistochemistry (IHC). Antifungal susceptibility testing and fungal speciation using multilocus sequence typing were performed.RESULTS:Hemorrhage, fungal germination, and hyphal growth were observed in the first sinus biopsy tissues. Areas with fungal growth tested positive for Stachybotrys by IHC. Fungal isolates were genotyped and identified as Stachybotrys chlorohalonata. The patient was cured from Stachybotrys sinusitis following sinus surgery and antifungal treatment. While a subsequent second sinus biopsy and a bronchoscopy showed no signs of fungal infection, a later, third sinus biopsy tested positive for Aspergillus calidoustus, a rare human pathogen.CONCLUSIONS:Here, we report the first case of invasive S. chlorohalonata sinusitis that was surgically and medically cured but followed by invasive A. calidoustus sinusitis in the setting of refractory leukemia. Our findings emphasize the risk for unusual fungal infections in severely immunocompromised patients.

Ganciclovir (GCV) and foscarnet (FOS) are the most commonly used antivirals for preemptive treatment of cytomegalovirus (CMV) viremia in recipients of allogeneic hematopoietic cell transplantation (alloHCT). The current literature indicates similar efficacy between these agents. Thus, the primary consideration for choice of initial anti‐CMV treatment is the safety profile, time period after alloHCT, and concern of myelosuppression or renal dysfunction.

CMVi leads to significant morbidity in R+ HCT and preemptive therapy (PET) has been the standard of care. LTV was FDA-approved in Nov. 2017 for CMVi prevention in R+ HCT. Besides the clinical trials, there is are real world data on LTV.Upon IRB approval, we retrospectively studied consecutive R+ HCT pts with their first HCT between 1/1/2017 and 6/30/2018. LTV group included pts with HCT between 2/20/18 and 6/30/2018, who had LTV Px started within 28 days of HCT (n=59), and R+ HCT between 1/1/2017 and 2/19/2018 (n=307) served as control (ctrl). We compared CMVi rates in first 100 days of HCT, and time to engraftment between the 2 groups. Risk stratification: high risk - haplo/cord HCT & ATG use, low risk - all others. CMVi was defined as viral load (VL) of 625 IU/ml to 1250 IU/ml or higher (CMV assay conversion factor of 1 genomic copy/ml = 2.5 IU/ml). CMV VL less than 625 IU/ml is reported negative, VL between 625 and 1250 IU/ml is qualitative positive but numeric value is provided only for VL ≥ 1250 IU/ml. PET was recommended for VL >1250 IU/ml (500 copies/ml) in high risk and > 3750 IU/ml (1500 copies/ml) in low risk HCT (including those on LTV requiring PET). Descriptive statistics was done for baseline characteristics. Cumulative incidence curves were generated for CMVi within 100 days post-HCT and Gray's test was used to compare the difference between the 2 groups.In both groups, median age was 54 years and HCT indications were similar. Donor type in LTV/ Ctrl groups: MRD (27 vs 36%), MUD (44 vs 47%), haplo (27 vs 15%), cord (1.7 vs 1.6%). PBMC was graft source in 90% of both groups. Myeloablative conditioning: 40.7% in LTV and 35% in Ctrl. GVHD Px: Cellcept (32 vs 25%), methotrexate (7 vs 9%), tacro/siro (58 vs 65), and others (3.4 vs 0.3%) in LTV & Ctrl respectively. A majority (n=36) received both intravenous & oral LTV formulation while 23 received oral only. Median time from HCT to LTV start was 13 days (range: 4-26). LTV group had significant reduction in CMVi rate (22.4% [95%CI: 12.7-34.0]) compared with ctrl group (41.1% [95%CI: 35.4-46.7], p=0.008, Fig 1). In a subgroup of high-risk HCT LTV was significantly reduced CMVi rate (22.2%) compared with ctrl (62.8%, p=0.004, Fig 2) while statistical difference was not reached in low-risk HCT pts (22.8% vs. 35.6%, p=0.11). In the LTV Px, clinically significant (CS)-CMVi requiring PET occurred in 8.4% (n=5) and on excluding 2 pts who were not on LTV at the time of CMVi, the rate was 5% (Fig 1). CMVi cleared without PET in 8/13 LTV pts and LTV was continued; all pts had VL < 2500 IU/ml with 50% having VL <1250 IU/ml. There was no difference in time to WBC and platelet engraftment in 2 groups (Fig 3).LTV use in a real world setting is associated with significant reduction of CMVi and CS-CMVi without any discernible myelosuppression. The low level CMVi resolved spontaneously in majority with continued LTV Px and PET was not necessary. The high-risk HCT had most benefit with LTV Px. CMVi leads to significant morbidity in R+ HCT and preemptive therapy (PET) has been the standard of care. LTV was FDA-approved in Nov. 2017 for CMVi prevention in R+ HCT. Besides the clinical trials, there is are real world data on LTV. Upon IRB approval, we retrospectively studied consecutive R+ HCT pts with their first HCT between 1/1/2017 and 6/30/2018. LTV group included pts with HCT between 2/20/18 and 6/30/2018, who had LTV Px started within 28 days of HCT (n=59), and R+ HCT between 1/1/2017 and 2/19/2018 (n=307) served as control (ctrl). We compared CMVi rates in first 100 days of HCT, and time to engraftment between the 2 groups. Risk stratification: high risk - haplo/cord HCT & ATG use, low risk - all others. CMVi was defined as viral load (VL) of 625 IU/ml to 1250 IU/ml or higher (CMV assay conversion factor of 1 genomic copy/ml = 2.5 IU/ml). CMV VL less than 625 IU/ml is reported negative, VL between 625 and 1250 IU/ml is qualitative positive but numeric value is provided only for VL ≥ 1250 IU/ml. PET was recommended for VL >1250 IU/ml (500 copies/ml) in high risk and > 3750 IU/ml (1500 copies/ml) in low risk HCT (including those on LTV requiring PET). Descriptive statistics was done for baseline characteristics. Cumulative incidence curves were generated for CMVi within 100 days post-HCT and Gray's test was used to compare the difference between the 2 groups. In both groups, median age was 54 years and HCT indications were similar. Donor type in LTV/ Ctrl groups: MRD (27 vs 36%), MUD (44 vs 47%), haplo (27 vs 15%), cord (1.7 vs 1.6%). PBMC was graft source in 90% of both groups. Myeloablative conditioning: 40.7% in LTV and 35% in Ctrl. GVHD Px: Cellcept (32 vs 25%), methotrexate (7 vs 9%), tacro/siro (58 vs 65), and others (3.4 vs 0.3%) in LTV & Ctrl respectively. A majority (n=36) received both intravenous & oral LTV formulation while 23 received oral only. Median time from HCT to LTV start was 13 days (range: 4-26). LTV group had significant reduction in CMVi rate (22.4% [95%CI: 12.7-34.0]) compared with ctrl group (41.1% [95%CI: 35.4-46.7], p=0.008, Fig 1). In a subgroup of high-risk HCT LTV was significantly reduced CMVi rate (22.2%) compared with ctrl (62.8%, p=0.004, Fig 2) while statistical difference was not reached in low-risk HCT pts (22.8% vs. 35.6%, p=0.11). In the LTV Px, clinically significant (CS)-CMVi requiring PET occurred in 8.4% (n=5) and on excluding 2 pts who were not on LTV at the time of CMVi, the rate was 5% (Fig 1). CMVi cleared without PET in 8/13 LTV pts and LTV was continued; all pts had VL < 2500 IU/ml with 50% having VL <1250 IU/ml. There was no difference in time to WBC and platelet engraftment in 2 groups (Fig 3). LTV use in a real world setting is associated with significant reduction of CMVi and CS-CMVi without any discernible myelosuppression. The low level CMVi resolved spontaneously in majority with continued LTV Px and PET was not necessary. The high-risk HCT had most benefit with LTV Px. Figures 1, 2 and 3.Figure 2View Large Image Figure ViewerDownload Hi-res image Download (PPT)Figure 3View Large Image Figure ViewerDownload Hi-res image Download (PPT)

The prevalence of vancomycin-resistance Enterococci colonization (VRE-C) in patients undergoing allogeneic hematopoietic cell transplantation (aHCT) is between 23-40%. Pre-HCT VRE-C is shown to be associated with high risks of VRE bloodstream infection (VRE-BSI), non-relapse mortality (NRM) and lower overall survival. Recent studies investigating the association between VRE-C and risk of acute graft-versus-host disease (aGVHD) after aHCT has demonstrated conflicting results, possibly due to the heterogeneous transplant conditioning and GVHD prophylactic regimens. Here, we sought to examine the VRE-C prevalence and determine its impact on aHCT outcomes, in patients receiving tacrolimus and sirolimus (T/S) as aGVHD prophylaxis.

Immune checkpoint inhibitors (ICIs) are innovative cancer immunotherapies used for solid-organ and hematologic malignancies. ICIs are known for their immune-related adverse events (irAE) but there are limited reports on infectious complications of immunosuppression for these complications. The purpose of this study was to describe the spectrum of infections in patients with melanoma, renal cell carcinoma or non-small cell lung cancer receiving ICI. Retrospective review of City of Hope patients with melanoma, renal cell carcinoma or non-small cell lung cancer on nivolumab, pembrolizumab, and/or ipilimumab from January to November 2017 and received two or more doses of ICI. Pt characteristics assessed: age, sex, prior chemotherapy, steroid use, and type of immunosuppression for irAE. Microbiology records were used to identify infections. Thirty-nine infectious episodes (35 bacterial, four viral) were identified among 111 patients. Four bacteremia (two B. cereus, coagulase-negative staphylococcus, 1 S. aureus), 12 urinary tract (10 Gram-negative rods, 2 Gram-positive cocci), one intra-abdominal, eight skin and soft-tissue infections (one S. aureus, one Actinomyces radinge, one E. faecalis, and one E. cloacae). There were two probable viral pneumonias (two rhinovirus, two enterovirus) and no fungal infections. Fourteen (12.6%) infections were defined as serious (requiring intravenous antimicrobials and/or hospitalization). There was no association between the specific malignancy or ICI used and risk of infection. Steroid use was significantly associated with serious infections: 12/14 (85.7%) vs. 27/95 (28.4%); P = 0.0003), and no patients had received infliximab or other immunosuppressant. Bacterial infections were most common, and the only risk factor associated with serious infections in our study was steroid use. Type of ICI did not impact the rate of infection. S. Dadwal, Ansun Biopharma: Investigator, Research grant.

Abstract Background Isavuconazole (ISV) is an antifungal approved for treatment of invasive aspergillosis and mucormycosis. Although data correlating response to serum levels is lacking, we found in a previous report a trend towards increased side effects with elevated serum trough levels. This study expands to a larger population to evaluate ISV trough levels in relation to clinical outcomes and side effects (AE). Methods Patients who received ISV >/ = 7 days with serum trough levels from April 2015 to September 2016 were included in AE analysis. Patients with proven or probable invasive fungal infection (IFI) were evaluated for response. Demographics, ALT, total bilirubin, serum creatinine, QTc, 14-, 30-, and 90-day response (complete, partial or no response) were collected. Results 94 patients were evaluated for response. At baseline, 39% were female, 70% had leukemia, 55% had hematopoietic cell transplant (87% were allogeneic), and 73% had active disease (relapsed, progressing, newly diagnosed). Of the IFIs, 64% were due to Aspergillus species, 12% Mucorales, and 4% Fusariumspecies. Response rates for patients by trough levels <3 vs. 3 to 5 vs. >5 mcg/mL are as follows: 14-day was 32% vs. 32% vs. 25%, 30-day was 48% vs. 52% vs. 44%, and 90-day was 39% vs. 50% vs. 33%, all P = NS. 205 patients were evaluated for AE in relation to trough levels. In comparing patients with trough levels <5 vs. >/ = 5 ug/mL, ALT levels 3 times upper limit was 18% vs. 36% (P = 0.032), total bilirubin 3 times upper limit was 22% vs. 32% (P = 0.79), serum creatinine 3 times upper limit was 7.8% vs. 12% (P = 0.47), and QTc interval was 1% vs. 2% mean decrease from baseline (P = 0.9). Conclusion Higher ISV levels were associated with increased ALT. ISV levels did not show apparent correlation with response rates, but there is a trend towards improved response rates with ISV levels between 3 - 5 ug/mL. Disclosures S. Dadwal, Merck: Investigator, Research support. GlaxoSmithKline: Investigator, Research support. Ansun Biopharma: Investigator, Research support. Oxford Immunotec: Investigator, Research support. Gilead Sciences: Investigator, Research support. J. Ito, Astellas: Speaker’s Bureau, Speaker honorarium.

Abstract Background Invasive aspergillosis (IA) causes significant morbidity and mortality in patients with hematologic malignancies (HM). Azole resistance has emerged as a therapeutic challenge in managing IA. The aim of this study was to investigate Aspergillus susceptibility to antifungals over the past decade among HM patients, and correlate susceptibility to clinical outcomes. Methods All Aspergillusisolates banked from 2002 to 2014 isolated from HM patients with probable/proven IA were tested for antifungal susceptibility. Patients with hematopoietic cell transplant, duplicate and non-viable isolates were excluded. Data were collected on demographics and clinical factors that could affect the treatment response, antifungal susceptibility (MICs/MECs), and treatment response at 14, 30, and 90 days. Results Forty patients were identified. MICs for amphotericin B slightly increased over the past decade (R = 0.32, P = 0.09), but were stable for voriconazole (R = −0.08, P = 0.61). The MIC50 during the first 3 years (2002–2004) and last 3 years (2012–2014) for amphotericin B were 0.5 and 1 mg/l, and for voriconazole 0.5 and 1. Mean age 56 years, 48% male, 82% had active HM and 45% had received chemotherapy within 14 days of IA. 50% were neutropenic and 30% had circulating blasts. Forty percent were on antifungal prophylaxis. Seventy-five percent of isolates were A. fumigatus. Fourteen responded to treatment (TR) and 26 were non-responders (NTR), and they did not differ in baseline characteristics. However, neutropenia (14% TR vs. 58%, NTR, P < 0.017) and circulating blasts (0% TR vs. 35% NTR, P < 0.02) at 14 days differed. The MIC50 for voriconazole was 0.5 mg/l in both groups, and for amphotericin B was 0.25 in TR vs. 1 mg/l in NTR. Fourteen-day response correlated with 90-day response (R = 0.74, P < 0.01) which validated the use of 14-day response for clinical outcome. All responders on amphotericin B at 14, 30, and 90 days had isolates with MIC < 1, whereas no apparent MIC-response correlation was found for voriconazole. Conclusion Although not statistically significant, a trend of increasing Aspergillus amphotericin B MICs was observed over the past decade. Neutropenia and persistent disease correlated with treatment failure. Clinical response was not affected by the azole or polyene MICs. Disclosures J. Ito, Astellas: Speaker’s Bureau, Speaker honorarium. S. Dadwal, Merck: Investigator, Research support. GlaxoSmithKline: Investigator, Research support. Ansun Biopharma: Investigator, Research support. Oxford Immunotec: Investigator, Research support. Gilead Sciences: Investigator, Research support

We are presenting a quantitative proteomics tally of the most commonly expressed conserved fungal proteins of the cytosol, the cell wall, and the secretome. It was our goal to identify fungi-typical proteins that do not share significant homology with human proteins. Such fungal proteins are of interest to the development of vaccines or drug targets. Protein samples were derived from 13 fungal species, cultured in rich or in minimal media; these included clinical isolates of Aspergillus, Candida, Mucor, Cryptococcus, and Coccidioides species. Proteomes were analyzed by quantitative MSE (Mass Spectrometry—Elevated Collision Energy). Several thousand proteins were identified and quantified in total across all fractions and culture conditions. The 42 most abundant proteins identified in fungal cell walls or supernatants shared no to very little homology with human proteins. In contrast, all but five of the 50 most abundant cytosolic proteins had human homologs with sequence identity averaging 59%. Proteomic comparisons of the secreted or surface localized fungal proteins highlighted conserved homologs of the Aspergillus fumigatus proteins 1,3-β-glucanosyltransferases (Bgt1, Gel1-4), Crf1, Ecm33, EglC, and others. The fact that Crf1 and Gel1 were previously shown to be promising vaccine candidates, underlines the value of the proteomics data presented here.

Infectious diseases are important causes of morbidity and mortality in patients with cancer. The NCCN Clinical Practice Guidelines in Oncology (NCCN Guidelines) for Prevention and Treatment of Cancer-Related Infections characterize the major pathogens to which patients with cancer are susceptible, with a focus on the prevention, diagnosis, and treatment of major common and opportunistic infections. This portion of the guidelines highlights the sections on antifungal and antiviral prophylaxis. Antifungal and antiviral prophylaxis recommendations have expanded over the past few years. New agents for the treatment of fungal infections and incorporation of therapeutic drug monitoring are presented. Antiviral prophylaxis for hepatitis B and management considerations for hepatitis C and HIV have been further developed.

Invasive mold infections (IMI) are life‐threatening complications of allogeneic hematopoietic stem cell transplantation (HSCT) and are mostly caused by Aspergillus species and Mucorales. We examined whether elevated serum ferritin prior to HSCT was associated with increased risk of IMI after allogeneic HSCT. Elevated serum ferritin was defined as values ≥1000 ng/mL. Pretransplant ferritin levels were available for 477 transplants. Nine developed IMI at day 30 and 21 had IMI at day 100 for a cumulative incidence of 1.9% and 4.4%, respectively. Among the high ferritin group, eight of 220 transplant cases (3.6%) developed an IMI within 30 d after HSCT compared with one of 257 (0.4%) in the low ferritin group ( P = 0.01). Fourteen of 220 (6.4%) and seven of 257 transplant cases (2.7%) in the high and low ferritin groups, respectively, had developed an IMI by day 100 after HSCT ( P = 0.07). Nine of 53 (17%) patients with grades III and IV acute GVHD and iron overload experienced IMI, when compared to three of 37 (8.1%) with high‐grade aGVHD , but no iron overload. Among patients without aGVHD , those with elevated ferritin had a 2.7% incidence of IMI compared with 0.9% for patients without elevated ferritin. There was a marginally significant difference in cumulative incidence function between high and low ferritin groups for IMI ( P = 0.06). However, elevated serum ferritin (≥1000 ng/mL) was not a significant risk factor for IMI in a multivariate competing risk regression model after adjusting for aGVHD .

BACKGROUND:Invasive fungal infections are a major cause of morbidity and mortality among solid organ transplant (SOT) and hematopoietic cell transplant (HCT) recipients, but few data have been reported on the epidemiology of endemic fungal infections in these populations.METHODS:Fifteen institutions belonging to the Transplant-Associated Infection Surveillance Network prospectively enrolled SOT and HCT recipients with histoplasmosis, blastomycosis, or coccidioidomycosis occurring between March 2001 and March 2006.RESULTS:A total of 70 patients (64 SOT recipients and 6 HCT recipients) had infection with an endemic mycosis, including 52 with histoplasmosis, 9 with blastomycosis, and 9 with coccidioidomycosis. The 12-month cumulative incidence rate among SOT recipients for histoplasmosis was 0.102%. Occurrence of infection was bimodal; 28 (40%) infections occurred in the first 6 months post transplantation, and 24 (34%) occurred between 2 and 11 years post transplantation. Three patients were documented to have acquired infection from the donor organ. Seven SOT recipients with histoplasmosis and 3 with coccidioidomycosis died (16%); no HCT recipient died.CONCLUSIONS:This 5-year multicenter prospective surveillance study found that endemic mycoses occur uncommonly in SOT and HCT recipients, and that the period at risk extends for years after transplantation.

Study ObjectiveTo evaluate the relationship between voriconazole dose and corresponding serum concentrations in obese and overweight immunocompromised patients.DesignRetrospective medical record review.SettingNational Cancer Institute–designated comprehensive cancer center.PatientsA total of 92 patients with hematologic malignancies and/or hematopoietic stem cell transplants who received voriconazole and had reported steady‐state serum concentrations (peak, random, or trough) during 2005–2010; 124 serum concentrations were available for analysis.Measurements and Main ResultsData on patient demographics, voriconazole concentrations, and other clinical and safety data were collected. Patients were stratified based on body mass index (BMI). Patients with higher BMIs tended to have significantly higher median random voriconazole concentrations with intravenous administration (6.4 mg/L for BMI ≥ 25 kg/m2 vs 2.8 mg/L for BMI < 25 kg/m2, p=0.04). This trend was more notable with the intravenous than the oral formulations. With the oral formulation, patients with a BMI of 25 kg/m2 or greater had a median random concentration of 2.8 mg/L compared with 2.0 mg/L in patients with a BMI less than 25 kg/m2 (p=0.18). Patients with a BMI of 25 kg/m2 or greater also had a higher median daily voriconazole dose (640 vs 400 mg, p<0.001). No significant differences were noted in factors that would affect oral absorption of voriconazole (e.g., graft‐versus‐host disease) among BMI groups. When comparing all voriconazole concentrations, higher concentrations were associated with a greater percentage of patients who had alanine aminotransferase levels of more than 3 times the upper limit of normal. Patients with voriconazole random concentrations of 2 mg/L or greater had higher response rates (50%) than patients with concentrations lower than 2 mg/L (33%).ConclusionStandard voriconazole dosing using actual body weight in obese and overweight patients resulted in higher associated serum concentrations. Dosing using adjusted body weight may be necessary in this population in order to achieve optimal concentrations while preventing the potential for increased toxicity.

Invasive aspergillosis is one of the most devastating opportunistic infections in immunocompromised hosts, such as those with hematologic malignancies and those undergoing hematologic cell transplantation (HCT). Despite the recent availability of potent antifungal agents, mortality rates remain unacceptably high for invasive aspergillosis in these 2 patient groups, at 42% and 58%–72%, respectively [1–3]. One of the reasons for this lack of improvement in outcomes with current therapeutic regimens may lie in the pathogenesis of invasive aspergillosis. The major risk factors for invasive aspergillosis are neutropenia (ie, chemotherapy-induced neutropenia in patients with hematologic malignancies and HCT recipients in the early, preengraftment phase) and corticosteroid treatment (ie, for graft-versus-host disease during the postengraftment phase of HCT). The immunosuppression induced by these agents and the resulting pathogenesis of invasive aspergillosis may differ between these 2 immunosuppressive states. Aspergillus conidia are inhaled through the sinopulmonary tract and land in the pulmonary alveoli. In the immunocompetent host, the first lines of defense are the alveolar macrophage and the neutrophil. If the patient is immunosuppressed, either with cytotoxic agents (inducing neutropenia) or corticosteroids (impairing macrophage function), conidia survive and go on to germinate into hyphae. These hyphae then proceed to invade the alveolar endothelial cells and extend into the pulmonary arterioles, the final act of angioinvasion. Penetration of endothelial cells results in endothelial cell damage, proinflammatory cytokine release, activation of the coagulation cascade, and intravascular coagulation [4]. These events, in turn, lead to tissue hypoperfusion and tissue necrosis. Intravascular coagulation, tissue hypoperfusion, and necrosis lead to an ideal state for the survival and progression of Aspergillus hyphae, a sequestered infection in a privileged site without access to fungicidal macrophages or neutrophils or antifungal agents. There are differences, however, between the responses in neutropenic and corticosteroid-treated hosts. While the pulmonary lesions in neutropenic patients consist predominantly of angioinvasion and intravascular hemorrhage, the lesions in corticosteroid-treated patients consist mainly of neutrophilic and monocytic infiltrates and inflammatory necrosis [5]. Thus, although the result, tissue necrosis, is the same in both types of immunosuppression, the mechanisms, coagulative vs inflammatory necrosis, differ. Angiogenesis, the formation of new blood vessels from existing blood vessels, is a physiologic response to tissue inflammation and ischemia [6]. Thus, it is not surprising that Aspergillus infection induces angiogenesis. It was demonstrated that Aspergillus fumigatus hyphae stimulate the production of the proinflammatory cytokines tumor necrosis factor α (TNF-α) and interleukin 8 (IL-8) by endothelial cells [7] and that these are potent inducers of proangiogenic signaling pathways involving vascular endothelial growth factor (VEGF) and basic fibroblast growth factor (bFGF) [8]. A. fumigatus, however, produces many metabolites, including fumagillin and gliotoxin, that demonstrate antiangiogenic activity [9–12]. Thus, to determine the balance between proangiogenic and antiangiogenic factors in invasive aspergillosis, Ben-Ami et al [13] studied angiogenesis in a neutropenic model of cutaneous invasive aspergillosis. They noted that angiogenesis was suppressed by Aspergillus-infected mice, when compared to uninfected mice. The antiangiogenic effect was completely abolished Received and accepted 10 December 2012; electronically published 9 January 2013. Correspondence: James I. Ito, MD, Division of Infectious Diseases, City of Hope Comprehensive Cancer Center, 1500 E Duarte Rd, Duarte, CA 91010 ( jito@coh.org). The Journal of Infectious Diseases 2013;207:1031–3 © The Author 2013. Published by Oxford University Press on behalf of the Infectious Diseases Society of America. All rights reserved. For Permissions, please e-mail: journals. permissions@oup.com. DOI: 10.1093/infdis/jis944