A Strategy For Early Diagnosis Of Invasive Mold Infections In Children With Hematologic Malignancies

Infections with invasive molds are an important cause of morbidity and mortality with published mortality rates of 21-48% among pediatric cancer patients infected by these organisms. Diagnosing invasive fungal infections is difficult because signs and symptoms are non-specific, and delays in diagnosis limit successful debridement of infected tissues. Despite this, there are no uniform guidelines for the diagnosis of invasive fungal infections. The deaths of three teens, at our institution, with leukemia and widespread mold, lead to the creation of a screening protocol for invasive fungal infection in November 2006. Neutropenic patients with persistent fever at 5 days or recurrent fever after defervescing were evaluated with a non-contrast computed tomography (CT) of the chest, abdominal ultrasound, and nasal endoscopy, performed at the bedside, by an otorhinolaryngologist. Initially the screen included CT of the abdomen, but the proclivity of mold for solid organ involvement supported the use of ultrasound as a screening tool. Additional studies were obtained as clinically indicated.

To determine the impact of this screening protocol on mortality associated with invasive mold, we performed a retrospective chart review of patients receiving intensive therapy for hematologic malignancies from 2004-2011 who were diagnosed as having proven, probable, or possible invasive mold (candida excluded) infections (N=52) using the European Organization for Research and Treatment of Cancer and the National Institute of Allergy and Infectious Diseases Mycoses Study Group (EORTC/MSG) criteria. Of the 20 mold infections in the pre-protocol group, 14 were classified as proven, 6 as possible. Among the 32 infections in the post-protocol group, 22 were proven, 4 probable, and 6 possible. Organisms included Aspergillus, Bipolaris, Curvalaria, Exserohilum, Fusarium, Rhizopus, and Scedosporium. Clinically indicated evaluations among the 20 patients in the pre-protocol group included, 16 chest CTs, 20 abdominal CTs or ultrasounds, and 7 underwent evaluation of their sinuses by direct nasal endoscopy. The lungs were the most common site of infection, with involvement detected in 15/20 patients (75%). Five patients (25%) had sinus involvement; in 1 patient this was the only site of disease. All 5 were symptomatic with rhinorrhea, congestion, facial pain, or facial numbness. Of the 32 patients in the post-protocol group, 30 had chest CTs, 32 had abdominal imaging (5 CT, 27 ultrasound), and 31 had direct nasal endoscopy. One patient did not have an ENT evaluation or chest CT because fungal disease was only detected post-mortem, and in 1 patient, cardiovascular instability precluded CT imaging. The lungs were again the most common site affected, with fungal pneumonia seen in 23/32 patients (72%). Fourteen patients (44%) had sinus involvement; in 4 patients this was the only site of disease. Nine patients with sinus involvement had no nasal symptoms or findings on routine physical exam.

Mortality specifically associated with invasive mold infection decreased significantly after initiation of the screening protocol. Before implementing the screening protocol, 8/20 patients (40%) who developed invasive mold infections died from the infection; afterward 4/32 (12.5%), (Fisher's exact p=0.04). Prior to routine evaluation of the sinuses by direct nasal endoscopy, 5/20 patients with mold infections had demonstrable disease in the sinuses and all had symptoms referable to sinus disease prior to evaluation. Once direct nasal endoscopy was implemented as part of the screening protocol, 14/32 patients with invasive mold infection were found to have sinus disease; 9 had no symptoms other than fever. Age, gender, race and length of hospital stay did not differ significantly before and after implementing the screening protocol. Before implementation, 8/20 patients (40%) died from all causes; afterward 6/32 (19%), (Fisher's exact p=0.12).

A screening protocol for the evaluation of neutropenic patients with persistent or recurrent fever led to early detection of invasive fungal infections in patients with hematologic malignancies and a significant decrease in infection associated mortality. Non-invasive, direct nasal endoscopy, performed at the bedside, is an effective tool for diagnosis of invasive fungal sinusitis and often detects fungal sinusitis before specific symptoms develop.