Objectives: Pattern of respiratory viral infections (RVI) in children with cancers is not well characterized in India. We conducted this study to determine the spectrum, incidence of fever-neutropenia, need for hospital admission, economic burden, and outcome (concomitant bacterial infections, need for ICU care, and death) of RVI in our pediatric-oncology unit.
Methods: Data was collected retrospectively from July 2015-April 2016. All children (<18-years) with symptoms of an upper respiratory tract infection (fever/cough/rhinorrhea) were included. Nasal and throat swabs were collected in Hi-Viral transport medium and analyzed using duplex Real Time PCR with Taqman Probes-Vi. Children with Influenza A/B received oseltamivir; those with symptomatic Respiratory Syncitial Virus (RSV) received ribavirin. All children with fever-neutropenia received intravenous antibiotics as per unit protocol.
Results: 104 viral isolates were identified in 89 patients. Median age was 5.5-years (range: 1.4-17.8); 65% were males. Majority of the episodes (91%) were in haemato-lymphoid malignancies. For children on treatment for ALL, 35 (42%) were in intensive and 48 (58%) in non-intensive phase of chemotherapy. Clinical features included: fever (89.5%), cough (71%), rhinorrhea (25%), chest-signs (25%), vomiting (12.5%), diarrhea (5.5%), rash (2%). Commonest isolate was Influenza-A (36.5%) followed by RSV (22.1%), Coronavirus (9.6%), Metapneumovirus (8.7%), Parainfluenza (8.7%), Influenza-B (6.7%), Adenovirus (1.9%) and Rhinovirus (1.9%). Multiple viruses were found in 4 (3.8%). Hematological parameters at presentation included: hemoglobin (median:9.9g%, IQR 8.7;11.1), platelet (median:1,41,605/mm3, IQR 60,000;1,95,750), neutrophil count (median:1962/mm3; IQR 390;2446), lymphocyte count (median:1277/mm3, IQR 300;1155), Neutrophil:lymphocyte ratio (NLR) (median:3.5; IQR0.3-4.2).
Overall, 50% of episodes warranted admission; majority (88.5%) with fever-neutropenia. Concomitant bacterial infection was documented in 11 (10.5%) episodes, with Pseudomonas (5;45%) being the commonest isolate. One died (0.9%) following Acinetobacter sepsis with RSV infection. Twenty-five (54%) required empirical escalation of antibiotics due to persistent fever-neutropenia; 7 (28%) had a proven bacterial isolate. Six/104 (5.7%) children needed ICU admission; 5 (83%) had concomitant bacteremia (p<0.001).Median duration of neutropenia was 12-days (range:1-47). Median duration of hospital-stay was 8-days (range:1-35).Median cost of admission was INR 22,466 (range:2,176-1,15,160).Cost and duration of stay was significantly more in children with concomitant bacteremia (p=0.002). Median cost for children with RVI and no bacteremia, whose antibiotics were escalated empirically (median INR 60,217, IQR 41,131; 1,24,037), was significantly higher than those who stayed on 1st-line antibiotics (median INR 11,503, IQR 2,699;24,868) (p<0.001).
On univariate analysis, significant predictor of bacteremia in RVI was low hemoglobin (p=0.01). Median hemoglobin in children with bacteremia (8.4g%, IQR 6.6;9.8) was significantly lower than those without bacteremia (10g%; IQR 9;11.3; p=0.06). Age (p=0.8), gender (p=0.5), underlying malignancy (p=0.6), intensive chemotherapy (p=0.1), neutrophil count (p=0.1), duration of neutropenia (p=0.72), lymphocyte count (p=0.9), platelet count (p=0.1) and NLR (p=0.1), failed to predict bacteremia.
Conclusion: Documented RVI in children with cancers does not preclude bacteremia. However, cost of care gets significantly increased for children with RVI and persistent fever-neutropenia, even with no bacteremia, when empirically escalated to higher antibiotics as per protocol (p<0.001). A prospective study is needed to analyze whether the protocol for antibiotic-escalation can be better rationalized in this cohort.
TM_SC-1_V1.13