Presence of viremia during febrile neutropenic episodes in patients undergoing chemotherapy for malignant neoplasms

The importance of viral infections as a leading cause of morbidity and mortality is well documented in severely immunosuppressed patients undergoing allogeneic stem cell transplantation. By contrast, viral infections generally receive less attention in patients with malignant disorders undergoing chemotherapy, where the onset of neutropenic fever is mostly associated with bacterial or fungal infections, and screening for viral infections is not routinely performed. To address the occurrence of invasive viral infections in a clinical setting commonly associated with less pronounced immunosuppression, we have prospectively screened 237 febrile neutropenic episodes in pediatric (n = 77) and adult (n = 69) patients undergoing intensive chemotherapy, primarily for treatment of acute leukemia. Serial peripheral blood specimens were tested by RQ‐PCR assays for the presence and quantity of the clinically relevant viruses CMV, EBV, HHV6 and HAdV, commonly reactivated in highly immunocompromised patients. Viremia was documented in 36 (15%) episodes investigated, including the detection of HHV6 (n = 14), EBV (n = 15), CMV (n = 6), or HAdV (n = 1). While low or intermediate levels of viremia (<104 virus copies/mL) were commonly associated with bacterial or fungal co‐infection, viremia at higher levels (>104 copies/mL) was documented in patients without evidence for other infections, raising the possibility that at least in some instances the onset of fever may have been attributable to the virus detected. The observations suggest that viral infections, potentially resulting from reactivation, might also play a clinically relevant role in patients receiving chemotherapy for treatment of malignant neoplasms, and routine screening for viremia in this clinical setting might be warranted.

fever is mostly associated with bacterial or fungal infections, and screening for viral infections is not routinely performed. To address the occurrence of invasive viral infections in a clinical setting commonly associated with less pronounced immunosuppression, we have prospectively screened 237 febrile neutropenic episodes in pediatric (n = 77) and adult (n = 69) patients undergoing intensive chemotherapy, primarily for treatment of acute leukemia. Serial peripheral blood specimens were tested by RQ-PCR assays for the presence and quantity of the clinically relevant viruses CMV, EBV, HHV6 and HAdV, commonly reactivated in highly immunocompromised patients. Viremia was documented in 36 (15%) episodes investigated, including the detection of HHV6 (n = 14), EBV (n = 15), CMV (n = 6), or HAdV (n = 1). While low or intermediate levels of viremia (<10 4 virus copies/mL) were commonly associated with bacterial or fungal co-infection, viremia at higher levels (>10 4 copies/mL) was documented in patients without evidence for other infections, raising the possibility that at least in some instances the onset of fever may have been attributable to the virus detected. The observations suggest that viral infections, potentially resulting from reactivation, might also play a clinically relevant role in patients receiving chemotherapy for treatment of malignant neoplasms, and routine screening for viremia in this clinical setting might be warranted.

| INTRODUCTION
Owing to the clear predominance of bacterial and fungal infections in the context of chemotherapy for aggresive hematologic neoplasms, screening for systemic viral infections is rarely part of routine diagnostic surveillance. 1 Nevertheless, viral infections account for or contribute to a proportion of febrile neutropenic episodes in patients with hematological malignancies. [2][3][4][5] A study performed in neutropenic patients with hematologic disorders, including also patients who underwent hematopoietic stem cell transplantation (HSCT), indicated that virus-positive test results were more frequent in the nontransplant patient cohort. 5 We and others have shown that respiratory tract infections in immunocompromised patients with various malignancies, including also recipients of HSCT, display a high virusassociated morbidity. [6][7][8] Infections with community respiratory viruses (CRVs) have been recognized as a potential cause of pneumonia and death among patients with hematologic malignancies and HSCT recipients, 8,9 but reactivation of persistent viral infections might play a more prominent role. 10 In the past, transmission of viruses by various blood products was a serious problem, particularly in immunocompromised patients, and routine screening for a set of viral pathogens has become part of the standard safety precautions in transfusion medicine. 11,12 Nevertheless, transmission of viruses via blood products due to contamination with known or yet to be identified human viral pathogens cannot be completely excluded. 13 Protecting the blood supply from emerging infectious threats therefore remains a serious concern in the transfusion medicine community. 14-18 Moreover, highly immunosuppressive treatment approaches used in hematologic malignancies favor opportunistic infections including viral reactivation. 2,5,14 Persistent viral infections that are commonly reactivated during states of impaired immune surveillance include various members of the herpes virus family and different other viruses including particularly human adenovirus (HAdV) as a prominent example. 10,[19][20][21][22][23][24][25][26][27][28] The role of these viruses has been extensively studied in the context of HSCT, and we hypothesized that their contribution to infectious complications in patients undergoing intensive chemotherapy for hematologic neoplasms may be underestimated. To shed more light on the incidence and potential role of viral infection or reactivation in this setting, we have prospectively screened pediatric and adult patients displaying febrile neutropenic episodes during chemotherapy, primarily for malignant hematologic disorders, to assess the occurrence of viremia by targeting select viruses.

| Patients
From 2015 to 2019, diagnostic surveillance of 237 febrile neutropenic episodes (defined by neutrophil counts < 500/mL blood and body temperature above 38°C) was performed in pediatric (n = 150) and adult (n = 87) settings. The prospective study was performed in patients undergoing intensive chemotherapy for different types of acute leukemia or solid tumors specified below, who carried a high risk for infectious complications. 29,30 The study was registered at clinicaltrials.gov (identifier: NCT02492594), and the protocol was

| Sample collection and microbial pathogen detection
Peripheral blood samples were collected at 0, 24 and 48 hours after onset of neutropenic fever. Bacterial and fungal culture were carried out in external certified laboratories, in line with the routine diagnostic work-up for septic patients. For molecular testing, the specimens were processed immediately or stored as full blood (for RQ-PCR) or plasma (for next generation sequencing -NGS) at −80°C prior to isolation of DNA. was assessed according to relevance scoring and the SIQ (sepsis indicating quantifier) score established previously. 33

| Statistical analysis
Differences between co-infections with viruses, fungi and bacteria were tested using an A/B test (based on N-1 chi-square test) by employing a calculator at MeasuringU.com. Two-tailed P values are indicated ( Figure 3). Note, p values <.05 were regarded as significant.

| Role of the funding source
The funding source had no involvement in the study design, in the collection, analysis and interpretation of data, in the writing of the report or in the decision to submit the paper for publication.

| RESULTS
While the presence of neutropenic fever in various earlier studies was   Indeed, high-level viremia was more frequently associated with virus detection already at the onset of fever (Figure 4(A)) and more common persistence of fever (Figure 4(B)). Because of the high prevalence of persistent infections involving the viruses analyzed, 42-45 reactivation rather than new infection may be a more probable cause of the infectious process, as described in recipients of allogeneic HSCT. 46 In contrast to patients undergoing HSCT, where concomitant reactivation of multiple viruses reflects the state of severe immunosuppression, 6,19 viral co-infection in the febrile neutropenic patients on chemotherapy was found in a single case only. This observation could be attributable to the lower severity of immunosuppression in the chemotherapy setting, although the degree of immunosuppression conveyed by chemotherapeutics can be considerable. 2,5,14,47 In contrast to earlier analyses in HSCT recipients performed at our center, 10,20,22,48 Table S2.
The virus-associated morbidity and mortality is well documented in allogeneic HSCT recipients, 10,19,20,49 (Table S1, Additional file 2), and detection of bacteria or fungi was generally associated with these clinical findings more frequently than viremia (Table S1, Additional file 2). We had the opportunity to employ unbiased NGS in a proportion of specimens, and the analyses revealed only a few additional viruses, some of which could not be confirmed by subsequently performed specific RQ-PCR tests. Since the sensitivity of well-established targeted PCR assays is not likely to be inferior to NGS analysis, the discrepant findings might be attributable to different technical approaches to nucleic acid isolation prior to the indicated molecular assays, which may have resulted in preferential detection of individual sequences. Moreover, unbiased NGSbased sequencing analysis may have permitted virus identification from any free circulating viral fragment DNA (cfDNA), which escaped detection by RQ-PCR tests relying on the amplifiability of a single defined viral fragment. It was somewhat surprising that the employment of unbiased NGS analysis did not permit identification of a greater number and variety of viruses that could have been related to febrile episodes in the patients investigated. It is conceivable, however, that the ongoing infectious process is not reflected by blood stream infections in many instances. Moreover, the NGS analyses performed were restricted to sequencing of DNA specimens, and the inclusion of RNA analysis might at least occasionally -have identified viral pathogens representing causative agents of the clinical symptoms observed.
Patients with febrile neutropenia not responding to different antibiotic treatment approaches for 72-96 hours commonly receive systemic antimycotic therapy without any evidence for the presence of fungal infection. This empirical treatment strategy is associated with conciderable toxicity and costs, and it is desirable therefore to greatly reduce or eliminate overtreatment by antifungal agents.