Abstract
This study was conducted to assess the transmissibility of rotavirus vaccine strains after rotavirus vaccination in a neonatal intensive care unit (NICU).
Pentavalent (RV5) or monovalent (RV1) rotavirus vaccine was administered to infants admitted to the NICU. Nineteen vaccinated infants and 49 unvaccinated infants whose beds were located in close proximity to the vaccinated infants were enrolled in this study. Dissemination and fecal shedding of vaccine viruses within the NICU were examined using real-time reverse transcription–polymerase chain reaction.
Shedding of the vaccine strain was detected in all 19 vaccinated infants. RV5 virus shedding started 1 day after the first vaccination and persisted for 8 days after the first vaccination, and viral shedding terminated by day 5 after administration of the second RV5 dose. The kinetics of RV1 virus shedding differed among vaccinated infants. The duration of RV1 virus shedding was longer after the first vaccination than after the second vaccination. In contrast to the vaccinated infants, no vaccine virus genomes were detected in any of the stool samples collected from the 49 unvaccinated infants.
This study is direct evidence of no transmission of rotavirus vaccine strains between vaccinated infants and unvaccinated infants in close proximity within a NICU.
Rotavirus is a leading cause of gastroenteritis in children, and it causes substantial morbidity and mortality worldwide [1]. To decrease rotavirus-associated morbidity, 2 live attenuated rotavirus vaccines are currently in use. Pentavalent rotavirus vaccine (RV5; RotaTeq, MSD) is a live, orally administered vaccine that has a backbone derived from bovine-origin rotavirus strains. Monovalent rotavirus vaccine (RV1; Rotarix, GlaxoSmithKline) is a live, attenuated, orally administered vaccine derived from a human-origin rotavirus strain. Owing to the use of live, attenuated rotavirus strains in the vaccines, vaccine-associated viruses are excreted in stools of the vaccine recipients [2–7]. Transmission of RV1 strains has been demonstrated between a vaccine recipient and their sibling without vaccination [8]. This is a major concern with regard to use of the vaccines in a neonatal intensive care unit (NICU).
It is well known that preterm infants admitted to the NICU are at increased risk of severe gastroenteritis due to rotavirus after discharge from a NICU, as a result of the relative immaturity of their immune responses and low levels of maternal antibodies [9]. Therefore, these infants should be vaccinated against severe rotavirus-associated gastroenteritis before discharge from the NICU. Studies have shown that rotavirus vaccination is safe and effective in preterm infants, and vaccination is recommended when infants are at least 6 weeks old. The Advisory Committee on Immunization Practices and the American Academy of Pediatrics recommends RV5 delivery to preterm infants who are at least 6 weeks old but only on the day of discharge, owing to concern about potential transmission of the vaccine virus to other vulnerable preterm infants in the unit due to shedding of virus in the stool, as documented elsewhere [10, 11].
The maximum age for starting the vaccine series is 14 weeks, after which the risk of intussusception increases. Many preterm infants with a birth weight of <1500 g have not been eligible to receive rotavirus vaccination because they were required to stay in the NICU beyond the upper age limit for initiating the vaccination series [12]. Although the use of rotavirus vaccine in the NICU is approved in the United Kingdom, only half of NICUs in the country regularly administer the vaccine [13–15]. To overcome this limitation, it is necessary to demonstrate the safety of rotavirus vaccination in the NICU.
The study conducted by Monk et al [16] suggested that use of RV5 was safe, with no documented cases of rotavirus transmission. However, this study did not specifically conduct surveillance testing of stool samples from vaccinated and unvaccinated infants who were in close proximity in the unit. Therefore, to determine whether rotavirus vaccine can be safely administered in a NICU, we prospectively elucidated fecal shedding of vaccine viruses and the dissemination of vaccine strains within the NICU, using highly sensitive real-time reverse transcription–polymerase chain reaction (RT-PCR) analysis.
MATERIALS AND METHODS
Subjects and Sampling
This study was approved by the review board of our university (approval no. 14–140), and the guardians of the infants consented to their participation. Vaccinated and unvaccinated infants were recruited from NICUs at either Fujita Health University Hospital (37 beds) or Konan Kosei Hospital (18 beds) between 10 October 2014 and 25 December 2015. The study design was explained to the parents by using a document describing the risks and benefits of rotavirus vaccination, and then informed consent was obtained from the parents. Vaccinated infants were selected from infants without active diseases or a past history of gastrointestinal disease or gastrointestinal operation. Between 1 and 4 unvaccinated infants were selected from among neonates or infants admitted to the same pod with vaccinated infants within 8 days after vaccination. We selected unvaccinated infants whose beds were in closest proximity to those of the vaccinated infants. Each nursing staff took care of 3–7 patients, and all medical staff adhered to contact precautions, including donning of glove and gowns, while caring for patients. The same nurse took care of vaccinated infants and the unvaccinated infants in closest proximity during the same shift.
The distance between patients was approximately 0.6–1.5 m in the NICU. Demographic information was collected from medical records for all of the vaccinated and unvaccinated infants enrolled in the study. Clinical symptoms, including fever (temperature, >38.0°C), diarrhea, vomiting, abdominal distention, hematochezia, feeding intolerance, and intussusception, were prospectively monitored in all vaccinated and unvaccinated infants.
To determine the extent of rotavirus RNA shedding, stool samples were collected from unvaccinated infants and from vaccinated infants after vaccination. In some instances, rectal swab specimens were collected from infants without bowel evacuation. Stool samples were serially collected once a day between days 0 and 8 after vaccination. For vaccinated infants who were discharged from the NICU within 8 days after vaccination, we asked their parents to collect stool samples until 8 days after the vaccination. Stool samples were collected once a day from unvaccinated infants mainly between days 5 and 13 after vaccination.
RNA Extraction and Real-Time RT-PCR
Ten percent suspensions (1 mL) of each stool sample were prepared in physiological saline solution, and swab samples were rinsed in 200 μL of physiological saline solution. Then, each suspension was clarified by centrifugation for 20 minutes at 4000×g, and 140 μL of the supernatant was used for RNA extraction. RNA was extracted from stool samples, using the QIAamp viral RNA minikit (Qiagen, Chatsworth, CA).
All stool samples were analyzed by real-time RT-PCR for the presence of RV5, RV1, and wild-type strains targeting the VP6, NSP2, and NSP3 genes. Details of the real-time RT-PCRs were described elsewhere [17, 18]. All real-time RT-PCRs were performed on a Fast Optical 48-Well Reaction Plate, using a TaqMan RNA-to-Ct 1-Step kit (Thermo Fisher Scientific, Waltham, MA). Single-well denaturation, reverse transcription, and amplification were performed on a StepOne Real-Time PCR system in standard mode (Thermo Fisher Scientific). Thermocycling conditions consisted of a 15-minute hold at 48°C, a 10-minute cycle at 95°C, and 45 cycles of 15 seconds at 95°C and 1 minute at 60°C. Each rotavirus vaccine strain–specific real-time RT-PCR amplified only the designated vaccine strain, and no cross-reaction was demonstrated with any of the wild-type strains. Meanwhile, real-time PCR to detect NSP3 in wild-type strains could also detect the 2 vaccine strains. Serially diluted purified rotavirus virions were used to determine the lower detection limits of RV5 VP6 (10 copies/reaction), RV1 NSP2 (50 copies/reaction), and wild-type NSP3 (50 copies/reaction), using real-time RT-PCRs (Supplementary Figure). RNA extracted from RotaTeq and Rotarix were used as positive controls for vaccine virus strains, and the KU strain (G1P [8]) was used as positive control for the wild-type virus.
RESULTS
Nineteen vaccinated infants (9 RV5 recipients and 10 RV1 recipients) and 49 unvaccinated infants (29 in contact with RV5 recipients and 20 in contact with RV1 recipients) were enrolled during the observation period. A total of 676 stool specimens were collected: 199 samples were collected from the 19 vaccinated infants, and 477 samples were collected from the 49 unvaccinated infants. The median number of stool samples collected per infant in the RV5 and RV1 groups was 10 (minimum–maximum, 1–18 samples) and 10 (minimum–maximum, 4–20 samples), respectively. Demographic characteristics of the vaccinated and unvaccinated infants are shown in Table 1. Among the vaccinated infants, the median gestational age and the median chronological age at the time of first vaccination was 27 weeks (interquartile range, 26.5–29 weeks) and 10 weeks after birth (interquartile range, 9.5–13 weeks after birth), respectively. Among the unvaccinated infants, the median gestational and chronological ages were 34 weeks (interquartile range, 32–37 weeks) and 3 weeks after birth (interquartile range, 2–5 weeks after birth), respectively. Although 1 vaccinated infant had severe underlying disease (thanatophoric dysplasia), the rest of the vaccinated infants had no underlying disease or had mild complications, such as chronic lung disease (9 cases) and meconium ileus (1 case). Meanwhile, some of the unvaccinated infants had severe underlying diseases, such as chromosomal anomalies and immunodeficiency, as shown in Table 1. Detailed information of each vaccinated infant, including birth weight, gestational age, feeding, and underlying diseases, is summarized in Supplementary Table 1. Vaccinated infants 8–10 were exclusively breast-fed; vaccinated infants 2–4, 7, and 11–13 were formula-fed; and the remaining 9 were mixed-fed.
Clinical Characteristics of Infants Hospitalized in the Neonatal Intensive Care Unit Who Did or Did Not Receive Rotavirus Vaccination
| Characteristic | Pentavalent Vaccine | Monovalent Vaccine | ||
|---|---|---|---|---|
| Vaccinated (n = 9) | Unvaccinated (n = 29) | Vaccinated (n = 10) | Unvaccinated (n = 20) | |
| Male sex | 4 | 17 | 5 | 11 |
| Birth weight, g | 974 (585–1176) | 1794 (1616–2065) | 1033 (894–1250) | 1344 (1196–2069) |
| Gestational age, wk | 28 (27–29) | 35 (33–37) | 27 (26–29) | 33 (29–35) |
| Chronological age, wk | 11 (10–11) | 3 (2–4) | 10 (9–14) | 3 (1–5) |
| Exclusively breast-feeding | 2 | 11 | 1 | 0 |
| Exclusively formula-feeding | 4 | 5 | 3 | 7 |
| Mixed feeding | 3 | 13 | 6 | 13 |
| Clinical eventa | ||||
| Diarrhea or liquid stool | 1 | 2 | 2 | 0 |
| Vomiting | 0 | 0 | 0 | 0 |
| Underlying disease | CLD, 5; meconium ileus, 1 | CLD, 1; Wiskott-Aldrich syndrome, 1; pulmonary stenosis, 1; Down syndrome, 1; TD, 1; Hirschsprung disease/trisomy 21, 1; multiple-malformation syndrome, 1; tracheal ring, 1 | CLD, 4; TD, 1 | CLD, 4; monosomy 13, 1; intraventricular hemorrhage, 1; tracheal ring, 1 |
| Characteristic | Pentavalent Vaccine | Monovalent Vaccine | ||
|---|---|---|---|---|
| Vaccinated (n = 9) | Unvaccinated (n = 29) | Vaccinated (n = 10) | Unvaccinated (n = 20) | |
| Male sex | 4 | 17 | 5 | 11 |
| Birth weight, g | 974 (585–1176) | 1794 (1616–2065) | 1033 (894–1250) | 1344 (1196–2069) |
| Gestational age, wk | 28 (27–29) | 35 (33–37) | 27 (26–29) | 33 (29–35) |
| Chronological age, wk | 11 (10–11) | 3 (2–4) | 10 (9–14) | 3 (1–5) |
| Exclusively breast-feeding | 2 | 11 | 1 | 0 |
| Exclusively formula-feeding | 4 | 5 | 3 | 7 |
| Mixed feeding | 3 | 13 | 6 | 13 |
| Clinical eventa | ||||
| Diarrhea or liquid stool | 1 | 2 | 2 | 0 |
| Vomiting | 0 | 0 | 0 | 0 |
| Underlying disease | CLD, 5; meconium ileus, 1 | CLD, 1; Wiskott-Aldrich syndrome, 1; pulmonary stenosis, 1; Down syndrome, 1; TD, 1; Hirschsprung disease/trisomy 21, 1; multiple-malformation syndrome, 1; tracheal ring, 1 | CLD, 4; TD, 1 | CLD, 4; monosomy 13, 1; intraventricular hemorrhage, 1; tracheal ring, 1 |
Data are no. of infants or median value (interquartile range).
Abbreviations: CLD, chronic lung disease; TD, thanatophoric dysplasia.
Clinical events occurring <28 days following rotavirus vaccination.
Clinical Characteristics of Infants Hospitalized in the Neonatal Intensive Care Unit Who Did or Did Not Receive Rotavirus Vaccination
| Characteristic | Pentavalent Vaccine | Monovalent Vaccine | ||
|---|---|---|---|---|
| Vaccinated (n = 9) | Unvaccinated (n = 29) | Vaccinated (n = 10) | Unvaccinated (n = 20) | |
| Male sex | 4 | 17 | 5 | 11 |
| Birth weight, g | 974 (585–1176) | 1794 (1616–2065) | 1033 (894–1250) | 1344 (1196–2069) |
| Gestational age, wk | 28 (27–29) | 35 (33–37) | 27 (26–29) | 33 (29–35) |
| Chronological age, wk | 11 (10–11) | 3 (2–4) | 10 (9–14) | 3 (1–5) |
| Exclusively breast-feeding | 2 | 11 | 1 | 0 |
| Exclusively formula-feeding | 4 | 5 | 3 | 7 |
| Mixed feeding | 3 | 13 | 6 | 13 |
| Clinical eventa | ||||
| Diarrhea or liquid stool | 1 | 2 | 2 | 0 |
| Vomiting | 0 | 0 | 0 | 0 |
| Underlying disease | CLD, 5; meconium ileus, 1 | CLD, 1; Wiskott-Aldrich syndrome, 1; pulmonary stenosis, 1; Down syndrome, 1; TD, 1; Hirschsprung disease/trisomy 21, 1; multiple-malformation syndrome, 1; tracheal ring, 1 | CLD, 4; TD, 1 | CLD, 4; monosomy 13, 1; intraventricular hemorrhage, 1; tracheal ring, 1 |
| Characteristic | Pentavalent Vaccine | Monovalent Vaccine | ||
|---|---|---|---|---|
| Vaccinated (n = 9) | Unvaccinated (n = 29) | Vaccinated (n = 10) | Unvaccinated (n = 20) | |
| Male sex | 4 | 17 | 5 | 11 |
| Birth weight, g | 974 (585–1176) | 1794 (1616–2065) | 1033 (894–1250) | 1344 (1196–2069) |
| Gestational age, wk | 28 (27–29) | 35 (33–37) | 27 (26–29) | 33 (29–35) |
| Chronological age, wk | 11 (10–11) | 3 (2–4) | 10 (9–14) | 3 (1–5) |
| Exclusively breast-feeding | 2 | 11 | 1 | 0 |
| Exclusively formula-feeding | 4 | 5 | 3 | 7 |
| Mixed feeding | 3 | 13 | 6 | 13 |
| Clinical eventa | ||||
| Diarrhea or liquid stool | 1 | 2 | 2 | 0 |
| Vomiting | 0 | 0 | 0 | 0 |
| Underlying disease | CLD, 5; meconium ileus, 1 | CLD, 1; Wiskott-Aldrich syndrome, 1; pulmonary stenosis, 1; Down syndrome, 1; TD, 1; Hirschsprung disease/trisomy 21, 1; multiple-malformation syndrome, 1; tracheal ring, 1 | CLD, 4; TD, 1 | CLD, 4; monosomy 13, 1; intraventricular hemorrhage, 1; tracheal ring, 1 |
Data are no. of infants or median value (interquartile range).
Abbreviations: CLD, chronic lung disease; TD, thanatophoric dysplasia.
Clinical events occurring <28 days following rotavirus vaccination.
During the observation period, vaccinated infants 1–9 received a first dose of RV5, vaccinated infant 4 received a second dose of RV5, and none received a third dose of RV5 (Figures 1A and 2A). Vaccinated infants 10–19 received a first dose of RV1, and vaccinated infants 10, 11, and 13 received a second dose. The mean duration of contact between vaccinated and unvaccinated infants after RV5 and RV1 receipt was 12 and 13 days, respectively (Figure 1A and 1B). Neither vomiting nor fever were observed in any vaccinated or unvaccinated infants during the 28 days after vaccination. Diarrhea or liquid stools were seen in 3 vaccinated infants (2 RV1 recipients and 1 RV5 recipient) and 2 unvaccinated infants. No other abnormal findings, including fever (temperature, >38.0°C), abdominal distention, hematochezia, feeding intolerance, and intussusception, were observed in any of the vaccinated and unvaccinated infants during the observation period.
Stool sampling schedules and associations between vaccinated infants (VIs) and unvaccinated infants (UVIs) after receipt of pentavalent (A) or monovalent (B) rotavirus vaccine. Black bars indicate the duration of stool sample collection from VIs. White bars indicate the duration of stool sample collection from UVIs. The first doses are indicated by black arrowheads, and the second doses are demonstrated by the gray arrowheads. The shaded areas indicate the contact period between VIs and UVIs.
Kinetics of viral shedding in vaccinated infants (VIs) after receipt of pentavalent (A) or monovalent (B) rotavirus vaccine. Viral RNA load was examined by real-time reverse transcription–polymerase chain reaction (RT-PCR) analysis and is represented by the Ct value. A high cycle threshold (Ct) value indicates a low copy number of viral genomes in the sample. The lack of a bar indicates that copy numbers were below the detection limit of the real-time RT-PCR. Gray bars indicate Ct values after receipt of the second vaccine dose, delivered at the time specified by gray arrowheads.
The quantity of viral RNA was based on the cycle threshold. Shedding of the RV5 strain was detected in 71 of 80 stool samples (88.7%) collected after the first dose. After the second dose of RV5, vaccine strain shedding was detected in 6 of 9 stool samples (66.7%; Figure 2A). Meanwhile, shedding of the RV1 strain was detected in 68 of 89 stool samples after the first dose (76.4%) and in 11 of 21 (52.4%) after the second dose (Figure 2B). RV5 virus shedding started 1 day after the first vaccination and persisted for 8 days after vaccination in all vaccinated infants except 2 and 7. Meanwhile, vaccine virus shedding started on day 0 and continued to day 5 after the second RV5 dose in vaccinated infant 4. The kinetics of virus shedding was unremarkable in these vaccinated infants (Figure 2A). In contrast, the kinetics of RV1 virus shedding differed among vaccinated infants. The detection rate and quantity of viral genomes were remarkably low in vaccinated infants 10, 17, and 18. Following the second vaccination with RV1, virus shedding was terminated by day 2 in vaccinated infants 10 and 11 and persisted for 8 days in vaccinated infant 13 (Figure 2B). In contrast to vaccinated infants, no vaccine virus genomes were detected in any stool samples collected from the unvaccinated infants (data not shown). Additionally, no wild-type rotavirus was detected in any of the stool samples collected from vaccinated and unvaccinated infants (data not shown).
DISCUSSION
Highly sensitive real-time RT-PCR analysis, which can detect 50 copies/reaction of the RV1 genome and 10 copies/reaction of the RV5 genome, was used to monitor viral shedding following rotavirus vaccination while in the NICU (Supplementary Figure). Similar to our previous study, conducted in foster homes [19], vaccine virus shedding persisted for a defined period in vaccinated infants, but no rotavirus genomes were detected in any stool samples collected from unvaccinated infants in this study. To the best of our knowledge, this is the first study yielding direct evidence of the lack of transmission of the vaccine virus strain within the NICU regardless of which strain was used for vaccination (RV1 or RV5). Furthermore, our current findings on the safety of rotavirus vaccine in the NICU are in line with the strategy proposed by Monk et al [16].
Although it has been demonstrated that rotavirus vaccine strains can be transmitted between siblings [8, 20], no vaccine virus transmission was demonstrated in either foster homes [19] or the NICU. Standard precautions were followed in the foster home, and contact precautions were practiced in the NICU. Therefore, precautions against infectious diseases play an important role in the prevention of viral spreading in close-contact environments such as foster homes and NICUs. Patient density is associated with an increased risk of virus transmission in close-contact environments. One of the 2 NICUs in this study was built 20 years ago, and the distance between patients is considerably shorter than that of standard NICUs in the United States. To limit the risk of nosocomial infections, some NICUs in Japan generally adhere to contact precautions for all patients. Our current findings support the notion that rotavirus vaccination in conjunction with contact precautions may be safe even in NICUs with a high density of patients. Guidelines from the American Academy of Pediatrics and the Advisory Committee on Immunization Practices indicate the need for contact precautions for any infant admitted to the NICU within 2–3 weeks of vaccination [10, 21]. Meanwhile, the previous retrospective analysis highlighted the safety of rotavirus vaccination in NICUs practicing standard contact precautions [16]. Further studies are needed to elucidate whether rotavirus vaccination within NICUs and standard contact precautions are sufficient to prevent the spread of vaccine virus strains.
Although no remarkable adverse effects in association with rotavirus vaccination were demonstrated in this study, these data are not sufficient for confirming the safety of rotavirus vaccination in preterm infants. However, it is important to note that previous reports demonstrate the safety and effectiveness of rotavirus vaccination in preterm infants [22–24]. RV5 was well tolerated in hospitalized infants admitted to an NICU in a previous retrospective study [16, 25]. These data suggest that rotavirus vaccine strains can be safely administered to hospitalized preterm infants. Precise criteria for selecting an eligible infant for rotavirus vaccination are still required because hospitalized infants typically have various types of complications or underlying diseases.
It was difficult to measure the precise amount of stool samples for quantitative analysis of the rotavirus genome by real-time RT-PCR analyses because some of the samples were rectal swab specimens. However, stool samples collected after the second vaccination demonstrated lower copy numbers in comparison to the stool samples collected after the initial vaccination (Figure 2A and 2B). Similar findings were also demonstrated in previous studies [26–28]. Thus, these findings suggest that suppression of viral replication by the initial rotavirus vaccination was due to immune responses. In this study, remarkably lower concentrations of rotavirus genomes were intermittently detected in stool samples collected from vaccinated infants 10, 17, and 18. Various factors, such as breast-feeding [29–31], vitamin A deficiency [31, 32], and antibiotics administration [33], may be associated with rotavirus vaccine replication in intestinal tissues. Additional studies including a larger number of cases is required to elucidate mechanisms for controlling viral shedding in neonates after vaccination.
Nosocomial infection of wild-type rotavirus in the NICU was reported from several institutes [34–36]. It has been suggested that wild-type rotavirus infection can cause severe clinical symptoms, such as necrotizing enterocolitis, in high-risk neonates [37]. Although some of the vaccinated and unvaccinated infants demonstrated gastrointestinal symptoms during the observation period, no nosocomial infections with wild-type virus were demonstrated in this study.
Herein, we provide direct evidence of the lack of viral spreading from rotavirus vaccine recipients in the NICU when standard contact precautions are practiced. With the understanding that it would be challenging to examine a large number of patients, a retrospective analysis that is similar to the previous study [16] should be conducted in the United Kingdom and Australia, because rotavirus vaccination in the NICU has already been recommended in these countries. These findings would further validate the use of rotavirus vaccines in NICUs.
Supplementary Data
Supplementary materials are available at The Journal of Infectious Diseases online. Consisting of data provided by the authors to benefit the reader, the posted materials are not copyedited and are the sole responsibility of the authors, so questions or comments should be addressed to the corresponding author.
Notes
Acknowledgments. We thank Mrs Akiko Yoshikawa, Chieko Mori, Yoko Osakabe, Honami Okada, and Nozomi Koshiyama, of Fujita Health University School of Medicine (Toyoake, Japan), for helping with collection and processing of samples.
Financial support. This work was supported by the Ministry of Education, Culture, Sports, Science, and Technology of Japan (Kiban-C: 22461610) and by the Ministry of Health, Labor, and Welfare of Japan (grant 27270201) for Research Promotion of Emerging and Reemerging Infectious Diseases.
Potential conflicts of interest. All authors: No reported conflicts of interest. All authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Conflicts that the editors consider relevant to the content of the manuscript have been disclosed.
