From the *Department of Infectious Diseases, Monash Children’s Hospital, Monash Health; †Department of Paediatrics, Monash University; ‡Murdoch Childrens Research Institute; and §Department of Paediatrics, University of Melbourne, Royal Children’s Hospital, Victoria, Australia.
J.P.B. and J.E.B. have acted as chief investigators for epidemiologic studies sponsored by vaccine manufacturers (CSL, Merck) or serologic testing provided (J.P.B.: Merck). They have served on data safety monitoring boards for CSL (J.P.B.) and Merck (J.E.B.). All payments are to administrative funds held by Murdoch Childrens Research Institute.
The author have no other funding or conflicts of interest to disclose.
Address for correspondence: Jim Buttery, MB BS, MD, MSc, Department of Infectious Diseases, Monash Children’s Hospital, 246 Clayton Rd, Clayton, Victoria, 3168 Australia. E-mail: firstname.lastname@example.org.
Since licensure in 2006, the rotavirus vaccines, RV5 (pentavalent bovine-human reassortant vaccine, Rotateq, Merck) and RV1 (attenuated monovalent human rotavirus vaccine, Rotarix, GSK Biologicals), have reduced morbidity and mortality from acute gastroenteritis in young children. Consensus on these vaccines having a small increased risk of intussusception after the first 1 or 2 doses has slowly emerged, with the risk benefit equation strongly favoring benefit in both developed and developing settings.
INTUSSUSCEPTION: ETIOLOGY AND EPIDEMIOLOGY
Intussusception is an acute bowel obstruction because of the telescoping of 1 part of the bowel into an adjacent section, resulting in obstruction and reduced blood flow. Typically presenting in infants aged 4–9 months, intussusception can occur at any age. Most cases can be diagnosed by ultrasonography, contrast enemas and reduced using air enemas. Later presentations may require surgery for intussusception reduction or resection. Untreated, it can result in bowel necrosis, perforation and may be fatal. Outside the peak age, anatomic lead points such as polyps can be identified as the cause of intussusception. However, in infancy, idiopathic intussusception is most common. Multiple investigations have explored the etiology of idiopathic intussusception, with adenoviruses implicated in up to 40% of cases.1
Idiopathic intussusception is an uncommon condition, with marked international and ethnic variation in rates, as well as changing incidence over time in many countries. Rates in the first year of life vary from 9 per 100,000 in Bangladesh, up to 328 per 100,000 in South Korea.1 Recent US incidence estimates appear stable around 35 per 100,000 since 2000, with rates before this period 25% higher.2 Intussusception presentations are relatively constant throughout the year, with minimal seasonality demonstrated.1
INTUSSUSCEPTION AND WILD TYPE ROTAVIRUS
The role of natural rotavirus infection in intussusception was first proposed in the 1978 by Japanese investigators using electron microscopic examination of feces. These findings were not able to be replicated across a variety of case-series and case-control studies as well as epidemiologic studies in temperate settings, where the marked seasonal rotavirus gastroenteritis peak was not accompanied by similar rises in intussusception cases.1
RRV-TV AND INTUSSUSCEPTION
The first licensed rotavirus vaccine, rhesus rotavirus reassortant tetravalent vaccine (RRV-TV, Rotashield, Wyeth Vaccines) was introduced into the US infant immunization schedule in October 1998. Reassortant vaccines use the segmented rotavirus genome to incorporate human rotavirus genes coding for surface proteins into the genome of animal rotavirus strains, which are rarely pathogenic to humans, creating an animal-human reassortant strain. RRV-TV was a mix of four rhesus-human reassortant strains, expressing human rotavirus G1-4 surface proteins. Prelicensure trials identified more intussusception cases in the RRV-TV arm compared with placebo, which was not statistically significant. However, intussusception was included as a potential adverse event after immunization in the product information. By March 1999, the US Vaccine Adverse Event Reporting Scheme (VAERS) had received 15 reports of intussusception. After initial examination, the Centre for Disease Control recommended suspending further vaccinations in July 1999 pending further investigations, with the manufacturer voluntarily ceasing distribution. After review of several investigations, the US Advisory Committee on Immunization Practices withdrew its recommendation that RRV-TV be administered at 2,4 and 6 months, because of a significantly raised association with intussusception, especially after the first dose. Case-control and case-series studies indicated the odds of intussusception cases having received a first dose of RRV-TV 3–7 days prior were as high as 37.2, with the risk period extending from 3 to 14 days postvaccination. (Table 1) This corresponded to an estimated excess of 1 additional case per 4670–9474 infants vaccinated, and risk appeared increased with administration of first dose after 3 months of age.3 However, ecologic studies suggested a much lower population attributable risk.
SECOND-GENERATION VACCINES: RV5 AND RV1
The RRV-TV experience informed the clinical development programs of RV5 and RV1. Both underwent massive phase III safety trials, with sample sizes to detect an association with intussusception similar to that observed with RRV-TV. Trials of over 60,000 infants each demonstrated no association with intussusception, with both vaccines licensed in 2006. Over the next 2 years, rotavirus vaccination was implemented in many countries including the United States, Mexico, Brazil, Australia and some European nations, with strict guidelines around age at dosing. However, given the limitation of these studies to detect an association at a lower level than RRV-TV, post-licensure studies were conducted in multiple settings to ensure early detection of any association, across nations with varying background rates of intussusception. These included enhanced passive surveillance to regulatory and clinical reporting systems, usage of large population administrative datasets and active case-finding programs in pediatric hospitals.
The first suggestion of an association was reported in Australia, with both RV5 and RV1 showing an excess of observed intussusception cases after dose 1 compared with expected number of cases based on historical rates. The small numbers and potential bias in the study relating to use of historical rates necessitated further confirmation. Subsequently, case-control studies from Mexico and Brazil confirmed an association with RV1 dose 1 and dose 2. However, despite using multiple data sources and methodologies, an association in the United States with RV5 was not able to be confirmed until 2013, when VAERS reports were examined using the self-controlled risk interval (SCRI) methodology where the patient is used as their own control, comparing the risk of intussusception in the postvaccine period with other periods.4 SCRI and case-control methodologies from a larger dataset have since confirmed the original Australian findings (Table 1).5
BENEFIT VERSUS RISK
Although the evidence for the risk of intussusception has evolved, the benefits of rotavirus immunization have been rapidly and consistently demonstrated in developed and developing settings. Despite reduced vaccine efficacy in developing setting trials, reductions in mortality in children <5 years from acute “all-cause” gastroenteritis of 43–55% were demonstrated after RV1 implementation in Mexico.6 In developed settings, where mortality is less common, hospital admissions for gastroenteritis in the United States and Australia, respectively, have reduced by 80% and 71% for rotavirus-coded gastroenteritis and 48% and 38% for non-rotavirus–coded admissions.7,8 Evidence of herd immunity has also been demonstrated, with unimmunized children experiencing decreased risk of hospital admission and death.
These are balanced with a small increased risk of intussusception (Table 1). In the US context, the benefit risk ratio for deaths prevented versus intussusception deaths precipitated favors vaccination by 71:1.9 Although this equation will vary in different settings depending upon gastroenteritis mortality, background rate of intussusception and access to health care, the balance strongly favors rotavirus vaccination in all settings it has been considered.
A CLASS EFFECT?
All rotavirus vaccines that have been licensed to date have demonstrated a temporally increased intussusception risk of varying degree. This includes 2 different reassortant vaccines as well as an attenuated human monovalent strain. This suggests all rotavirus vaccines may carry some risk, and clinical trials may be insufficiently powered to detect a small risk. This is especially relevant for new rotavirus candidates, including the Indian 116E human monovalent strain, the UK bovine-human reassortant vaccine and the Australian human monovalent candidate RV3, with the first dose intended at birth, when intussusception risk is lowest. Post-licensure risk management plans that incorporate surveillance strategies to detect and quantify intussusception risk will be important, both for new rotavirus vaccines, and countries introducing rotavirus vaccines. The presence of a class effect, if proven, should restimulate investigations into whether, despite the weight of negative evidence to date, low level wild type rotavirus infection may indeed be associated with intussusception.
LESSONS FOR VACCINE SAFETY
The rotavirus vaccine intussusception experience has illustrated the importance of a number of features required for investigating rare adverse event after immunization: National and international collaborations; use and combination of large administrative datasets and the use of case only methodologies including SCRI. Additionally, the interplay of immunization coverage and background rates of the underlying health event all impact upon our ability to detect or confidently refute adverse associations as early as possible.
1. Jiang J, Jiang B, Parashar U, et al. Childhood intussusception: a literature review. PLoS One. 2013;8:e68482
2. Tate JE, Simonsen L, Viboud C, et al. Trends in intussusception hospitalizations among US infants, 1993–2004: implications for monitoring the safety of the new rotavirus vaccination program. Pediatrics. 2008;121:e1125––1132
3. Murphy TV, Gargiullo PM, Massoudi MS, et al.Rotavirus Intussusception Investigation Team. Intussusception among infants given an oral rotavirus vaccine. N Engl J Med. 2001;344:564–572
4. Haber P, Patel M, Pan Y, et al. Intussusception after rotavirus vaccines reported to US VAERS, 2006–2012. Pediatrics. 2013;131:1042–1049
5. Carlin JB, Macartney K, Lee KJ, et al. Intussusception risk and disease prevention associated with rotavirus vaccines in Australia’s national immunisation program. Clin Infect Dis. 2013;57:1427–1434
6. Gastañaduy PA, Sánchez-Uribe E, Esparza-Aguilar M, et al. Effect of rotavirus vaccine on diarrhea mortality in different socioeconomic regions of Mexico. Pediatrics. 2013;131:e1115–e1120
7. Dey A, Wang H, Menzies R, et al. Changes in hospitalisations for acute gastroenteritis in Australia after the national rotavirus vaccination program. Med J Aust. 2012;197:453–457
8. Desai R, Curns AT, Steiner CA, et al. All-cause gastroenteritis and rotavirus-coded hospitalizations among US children, 2000–2009. Clin Infect Dis. 2012;55:e28–e34
9. Desai R, Cortese MM, Meltzer MI, et al. Potential intussusception risk versus benefits of rotavirus vaccination in the United States. Pediatr Infect Dis J. 2013;32:1–7
10. Yih WK, Lieu T, Kulldorff M, et al. Intussusception Risk After Rotavirus Vaccination in US Infants. 2013 Boston, MA Food and Drug Administration, Rockville, MD
11. Patel MM, López-Collada VR, Bulhões MM, et al. Intussusception risk and health benefits of rotavirus vaccination in Mexico and Brazil. N Engl J Med. 2011;364:2283–2292
12. Buttery JP, Danchin MH, Lee KJ, et al.PAEDS/APSU Study Group. Intussusception following rotavirus vaccine administration: post-marketing surveillance in the National Immunization Program in Australia. Vaccine. 2011;29:3061–3066