Molecular characteristics and successful management of a respiratory syncytial virus outbreak among pediatric patients with hemato-oncological disease

An outbreak case is a patient with a positive RSV laboratory testing in samples from the upper or lower respiratory tract and a definite or possible nosocomial onset. A definite nosocomial case was defined as a positive RSV laboratory testing on day 5 or later of the hospital stay. A possible nosocomial case was defined as a positive RSV laboratory testing on day 2 to 4 of hospital stay. Patients who were admitted to the ward with a new positive RSV laboratory testing, and had been on the ward within 8 days prior to admission were also considered a possible nosocomial case. All patients that were accommodated in the same room with cases were considered contacts.

A combined nose/throat swab was taken for routine viral diagnostics. Material from the lower respiratory tract was suitable as well.

Samples taken for diagnostic purposes were processed at the Institute of Virology using real-time RT-PCR or direct fluorescent antibody (DFA) staining. RNA was extracted from the specimens using a QiaAmp Viral RNA Mini Kit in a QIAcube according to the manufacturer’s instruction (Qiagen, Hilden, Germany). cDNA synthesis, amplification and detection of nucleic acid were performed in an Applied Biosystems® 7500 Real-Time PCR System (Life Technologies, Carlsbad, California) by a commercially available one-step real-time RT-PCR kit (RSV/hMPV r-gene® PCR Kit, bioMérieux, Nürtingen, Germany) according to the manufacturer’s instructions. For DFA staining a ready to use FITC (fluorescein isothiocyanate) -labeled monoclonal RSV antibody (LIGHT DIAGNOSTICS, Merck, Darmstadt, Germany) was used according to a protocol described before [22]. PCR and DFA did not differentiate between RSV A and B. One diagnostic specimen was tested using a point-of-care test (POCT) system (Sofia, Quidel, Kornwestheim, Germany), which is available in the pediatric emergency room.

Nasopharyngeal aspirates from 6 outbreak patients (case 1, 2, 3, 5, 6, and 7) were taken only for strain typing purposes on one occasion (March18^th), which were exclusively processed at the Institute for Experimental Virology, Twincore - Centre for Experimental and Clinical Infection Research. In addition, selected archived (frozen) samples taken for diagnostic purposes from outbreak patients (case 2, 3, 4, 5, 7, 8) were provided by the Institute of Virology and processed at the Institute for Experimental Virology, Twincore - Centre for Experimental and Clinical Infection Research.

Linearized acrylamide (Ambion, Thermo Fisher Scientific; 35 μg/ml final concentration) was added to the sample and total RNA was extracted from 140 μl of aspirate according to the manufacturer’s description (QiaAmp Viral RNA Mini Kit, Qiagen, Hilden, Germany). cDNA synthesis was performed using the Superscript III kit from Invitrogen (Invitrogen, Darmstadt, Germany) and random hexamer primers. Next, a nested PCR was performed first amplifying the RSV-G and F protein coding region and in a second round amplifying the G protein gene. PCR products were sent for Sanger sequencing (GATC, Konstanz, Germany) and the sequences were analyzed using MEGA software and the Highlighter analysis tool [23].

We obtained ethical approval for this study from the ethics committee of the Hannover Medical School.

The outbreak occurred in the Clinic for Pediatric Hematology and Oncology which is a tertiary referral center for children from 0-18 years with hematologic and solid neoplasia. The affected ward harbors 5 single- and 5 two-bed rooms. Each single room has an anteroom and a high-efficiency particulate air filtration with the air flow directed to the hallway. The same floor houses the outpatient clinic and a recreation room for the hemato-oncological pediatric patients. Exchange from patients between the ward and the outpatient clinic occurs regularly. Moreover two recreation rooms for social activities are part of the ward. Most patients receive antineoplastic therapy during their stay. A substantial number of patients have a neutrophil count below 500 cells / μl. The ward is serviced by permanent health care workers (HCWs) and house-keeping staff. External personnel (such as medical consultants or physiotherapists) enter the ward when necessary. Autologous and allogenic HSCT are performed on a separate ward with 6 single rooms. Parents are allowed to stay overnight with their children.

Patients with a positive RSV test are electronically marked in the hospital alert system. Besides single room accommodation, contact and droplet precautions (surgical mask, gown, gloves) are used at any time by visitors and HCWs when entering the room of a RSV positive patient. Affected patients are encouraged to stay in their room and are trained in hygienic hand washing to minimize spread via contact. When leaving the room becomes necessary (e.g. for examination), patients wear a surgical mask. These measures also apply for patients with typical respiratory symptoms (e.g. cough or sneezing) before a pathogen is identified. Measures are usually suspended when there are two negative RSV PCR-based test results at a minimum 2-day interval and therefore patients are not anymore considered infectious. HCWs with symptoms of URTI are suspended from direct patient care and wear a surgical mask while on the ward. Visitors with symptoms of acute RTI are not permitted to enter the ward. Strict hand hygiene following WHO guidelines is implemented. Targeted RSV diagnostics are performed in case of suspected viral RTI. Positive testing for respiratory viruses is regularly reported for epidemiologic and infection control reasons to the infection control staff.

In March a total of 8 patients (cases 1 to 8) were tested RSV-positive in respiratory samples. Patient characteristics are shown in Table 1. 6 patients fulfilled criteria for nosocomial acquisition. The remaining 2 patients had a possible nosocomial acquisition, as they had been on the ward within 8 days prior to admission. Epidemic curve and an outbreak timeline with the diagnostic results can be seen in Figs 1 and 2.

Case 1 was neutropenic and developed severe LRTI with a RSV positive bronchoalveolar lavage and a requirement for oxygen. Case 2 also suffered from LRTI, which was less severe. Cases 3 to 8 had respiratory symptoms of an URTI such as cough, sneeze and a positive RSV testing in secretions from the upper respiratory tract. Case 6 was co-infected with influenza A virus .

4 patients received oral Ribavirin therapy (case 1, 2, 5, 8) and 5 patients (case 1 to 5) temporarily required supportive oxygen administration via nasal cannula. IVIG or Palivizumab was not administered. No direct RSV-associated mortality was observed. In addition, all patients were empirically treated by antibiotics presuming bacterial co- respectively superinfection according to in-house standards (Table 1).

Viral persistence (viral shedding) is defined as the time period from first positive diagnostic test to sustained negativity. Duration of viral persistence was minimally 4 days (case 7) and maximally at least 63 days (case 1) – see also Table 1.

Active outbreak management was started after detection of 3 new RSV infected patients in calendar week 10 (see Fig. 1). An outbreak control team consisting of the infection control unit and the physicians in charge was established. The head of the Clinic for Pediatric Hematology and Oncology, the medical director of our institution, and the public health authority were informed. In addition to the existing standard infection control measures described above, interventional measures were introduced. All HCWs (including permanent staff members and external personnel), visitors and outpatients were required to wear a surgical mask at any time (patient care and non-patient care activities) when on the ward and in the outpatient clinic (preemptive barrier precaution). Moreover, roommates of patients tested positive for RSV in their clinical course were moved to single rooms for 8 days (typical maximum incubation period; so called quarantine). They were repeatedly tested for RSV using PCR. All newly admitted patients were tested for RSV (admission screening). Twice weekly PCR RSV prevalence screening for all patients on ward was established (prevalence screening). If possible, elective patient admissions were delayed to reduce patient-to-nurse ratio. Moreover, only parents were allowed as visitors. All social activities for the patients and relatives were suspended. During outbreak, all two bed-rooms were occupied by one patient only, meaning single room isolation for all patients on the ward. Repeated training sessions for staff were provided by the infection control team. They addressed RSV transmission pathways and underlined the importance of droplet precautions (e.g. masks and cough etiquette for HCWs and visitors) as well as hand hygiene.

Intervention measures were fully implemented on 22^nd of March. The last nosocomial case occurred on 29^th of March (case 8). However, the patient had been discharged from the ward on 22^nd of March, and readmitted on 29^th of March (Fig. 2). Thus, after intervention measures had been in place no further nosocomial RSV cases occurred. At the beginning of May the last positive patients were discharged respectively sustainably tested negative and all outbreak control measures were suspended.

For phylogenetic analysis of the RSV genome we focused on the viral glycoprotein G of RSV as this gene is highly variable and shows the highest sequence variance between the RSV subgroups A and B (53% amino acid sequence identity [24]). From the total of 8 patients, five of them were infected with a RSV A strain (Case 1, 3, 4, 5, 6), one was tested positive for RSV B (case 8) and two patients were infected with RSV A and B (case 2 and 7) dependent on the time point of sampling. Despite the coexistence of genetically definite genotypic strains with many nucleotide exchanges especially in the C-terminal variable region of RSV-G [25‐27], we detected the very same nucleotide sequence for the coding region of the RSV-G protein for all patients infected with RSV A (Fig. 3a). Only one single nucleotide exchange was detected in the intergenic region of the virus infecting patient C3 and only at one time point of sampling (Fig. 3a; C3_13_3). In the specimen taken 5 days later (C3_18_3), this variation was no longer detectable. Among the three patients infected with RSV B, we observed 4 nucleotide differences in the coding region of the G protein (Fig. 3b). Two additional variations were observed in the intergenic region of the G gene (Fig. 3b). The RSV B viruses infecting patients C2 and C7 were almost identical with merely three nucleotide differences between them. Notably, the chromatograms of the Sanger sequencing from these patients revealed sequence ambiguity at exactly these three positions: between G/A at position 907, between T/A at position 974 and G/A at residue 979. While in case C7, residues G, T and G were dominant, in patient C2, residues A, A, A predominated. This result suggested that these two patients were infected with an essentially identical viral quasispecies and that merely the relative number of viruses with G, T, G residues compared to viruses with A, A, A nucleotides at these positions varied between these two patients. In contrast, patient C8 carried a RSV B virus with an unambiguous sequence at these three positions G, T, G (Fig. 3d). Moreover, it displayed three additional polymorphisms in the G protein coding region relative to the virus infecting patients C2 and C7, thus supporting the conclusion that this patient was infected with a different RSV B virus. Noticeably, case 2 was tested RSV negative in two samples taken and processed for routine diagnostic on March 15^th and 18^th. RSV strain sequences available from other, non-outbreak pediatric patients (same season as the outbreak) were used for comparison. These strains show a predominantly polyclonal pattern for RSV A (Fig. 3c). Interestingly, one patient (RSV_02_1) was infected with the very same RSV B strain as cases 2 and 7 (Fig. 3c, lower part), whereas there were clear sequence differences for the other RSV B strains isolated from non-outbreak pediatric patients.