Fever of unknown origin is a common presentation in children with an extensive differential diagnosis that encompasses multiple specialties. From a hematologic standpoint, the differential includes hyperinflammatory syndrome, such as hemophagocytic lymphohistiocytosis (HLH), among others. Due to the rarity of HLH and nonspecific symptoms at initial presentation, specialists are often consulted later in the disease progression, which complicates disease evaluation further. Cook Children’s Medical Center (CCMC) has recently developed a multidisciplinary histiocytic disorder group that is often consulted on cases presenting with fever of unknown origin to increase awareness and potentially not miss new HLH cases. In this study, we examine the clinical presentation and workup of 13 patients consulted by the HLH work group at a single institution and describe the clinical course of 2 patients diagnosed with HLH. The goal of this project was to describe the formation of a disease-specific team and the development of a stepwise diagnostic approach to HLH. A review of the current diagnostic criteria for HLH may be warranted given findings of markers such as soluble IL2 receptor and ferritin as nonspecific and spanning multiple disciplines including rheumatology, infectious disease, and hematology/oncology.

The application of molecular tumor profiles in clinical decision making remains a challenge. To aid in the interpretation of complex biomarkers, molecular tumor boards (MTBs) have been established worldwide. In the present study, we show that a multidisciplinary approach is essential to the success of MTBs. Our MTB, consisting of pediatric oncologists, pathologists, and pharmacists, evaluated 115 cases diagnosed between March 2016 and September 2021. If targetable mutations were identified, pharmacists aided in the evaluation of treatment options based on drug accessibility. Treatable genetic alterations detected through molecular testing most frequently involved the cell cycle. For 85% of the cases evaluated, our MTB provided treatment recommendations based on the patient’s history and results of molecular tumor testing. Only three patients, however, received MTB-recommended targeted therapy, and only one of these patients demonstrated an improved clinical outcome. For the remaining patients, MTB-recommended treatment often was not administered because molecular tumor profiling was not performed until late in the disease course. For the three patients who did receive MTB-recommended therapy, such treatment was not administered until months after diagnosis due to physician preference. Thus, the education of healthcare providers regarding the benefits of targeted therapy may increase acceptance of these novel agents and subsequently improve patient survival.

Hematologic tumors are mostly treated with chemotherapies that have poor toxicity profiles. While molecular tumor profiling can expand therapeutic options, our understanding of potential targetable drivers comes from studies of adult liquid tumors, which does not necessarily translate to efficacious treatment in pediatric liquid tumors. There is also no consensus on when profiling should be performed and its use in guiding therapies. We describe a single institution’s experience in integrating profiling for liquid tumors. Pediatric patients diagnosed with leukemia or lymphoma and who underwent tumor profiling were retrospectively reviewed. Ten (83.3%) patients had relapsed disease prior to tumor profiling. Eleven (91.7%) patients had targetable alterations identified on profiling, and three (25%) received targeted therapy based on these variants. Of the three patients that received targeted therapy, two (66.7%) were living, and one (33.3%) decreased. For a portion of our relapsing and/or treatment-refractory patients, genetic profiling was feasible and useful in tailoring therapy to obtain stable or remission states. Practitioners may hesitate to deviate from the ‘standard of therapy’, resulting in the underutilization of profiling results. Prospective studies should identify actionable genetic variants found more frequently in pediatric liquid tumors and explore the benefits of proactive tumor profiling prior to the first relapse.