Objective Pneumococcal infection is a leading cause of haemolytic uraemic syndrome (HUS) and is potentially vaccine preventable. Published data suggest high mortality and poor renal outcomes. The introduction of the 7-valent pneumococcal conjugate vaccine (PCV) has seen the emergence of disease caused by non-vaccine strains, particularly 19A. We sought to describe serotype prevalence and outcomes, particularly after the introduction of the 13-valent PCV.
Design and setting We performed a retrospective chart review, using hospital medical records to identify cases of HUS in a tertiary paediatric hospital in Australia over a 20-year period (January 1997–December 2016). Associated pneumococcal infection was identified, and serotype data were categorised according to vaccine era: prevaccine (January 1997–December 2004), PCV7 (January 2005–June 2011) and PCV13 (July 2011–December 2016).
Results We identified 66 cases of HUS. Pneumococcal infection was proven in 11 cases, representing 4% (1/26) of cases prior to the introduction of PCV7, 20% (3/15) in the PCV7 era and 28% (7/25) in the PCV13 era. Subtype 19A was the most prevalent pneumococcal serotype (6/11). All four patients who received PCV7 were infected with a non-vaccine serotype. Four of the five patients who received PCV13 were classed as vaccine failures. Median follow-up was 14 (range 1–108) months. Chronic kidney disease was the most common complication (4/7). We observed no mortality, neurological sequelae or progression to end-stage kidney disease.
Conclusions Serotype 19A is most commonly associated with pneumococcal HUS, despite the introduction of the 13-valent vaccine. Chronic kidney disease is a significant complication of pneumococcal HUS.
- general paediatrics
- infectious diseases
- intensive care
Statistics from Altmetric.com
What is already known on this topic?
Pneumococcal infection is a potentially vaccine preventable cause of haemolytic uraemic syndrome.
Published data suggest high mortality and poor renal outcomes.
With the introduction of the 7-valent pneumococcal conjugate vaccine, many authors have reported emergence of non-vaccine serotypes, particularly 19A.
What this study adds?
We observed an increasing proportion of HUS cases attributable to pneumococcal infection between 1996 and 2016.
Serotype 19A is the most commonly associated serotype, despite the introduction of the 13-valent vaccine.
While no mortality was observed, around half of our patients developed chronic kidney disease.
Haemolytic uraemic syndrome (HUS) is the most common cause of acute kidney injury in children.1 While Shiga toxin-producing Escherichia coli (STEC) infection remains the leading cause of HUS, the incidence of Streptococcus pneumoniae-associated HUS, a systemic complication of invasive pneumococcal disease (IPD), appears to be increasing.2 3 Presently, pneumococcal HUS represents approximately 5%–15% of all HUS cases, and over 200 cases of pneumococcal HUS in children have now been described in the literature.4
The introduction of pneumococcal vaccination has resulted in a significant reduction in the overall incidence of IPD.5 Since January 2005, the Australian National Immunisation programme has included universal vaccination of infants against pneumococcus, as well as catch-up vaccination for children less than 2 years of age. From January 2005 to July 2011, this consisted of the heptavalent pneumococcal conjugate vaccine (PCV) Prevenar 7, which includes serotypes 4, 6B, 9V, 14, 18C, 19F and 23F.6 Since July 2011, this has been replaced by Prevenar 13, which additionally includes serotypes 1, 3, 5, 6A, 7F and 19A.
Epidemiological data on pneumococcal HUS are limited in the postpneumococcal vaccination era. After the advent of universal PCV7 vaccination, a number of studies reported an evolving phenotype of pneumococcal HUS, with apparent serotype replacement with non-vaccine strains.3 7 8 The effect of universal PCV13 vaccination on pneumococcal HUS presentations has not been described.
We describe proven cases of pneumococcal HUS in a paediatric tertiary hospital over a 20-year period and examine the effect of universal pneumococcal vaccination with both PCV7 and PCV13 on the serotypes implicated in HUS.
Cases of HUS diagnosed by a paediatric nephrologist at the Royal Children’s Hospital in Melbourne, Australia, from January 1997 to December 2016, were identified through review of hospital medical records. Cases were included if they fulfilled the following diagnostic criteria:
Microangiopathic haemolytic anaemia (Hb <100 g/L) with fragmented erythrocytes.
Thrombocytopaenia (platelet count <130×109/L).
Acute renal impairment (oliguria and age-adjusted elevated plasma creatinine).
Proven or suspected pneumococcal infection.
A retrospective chart review of all cases of pneumococcal HUS identified patient demographics, pneumococcal bacteriology, clinical course and follow-up. Vaccination status was determined via the Australian Childhood Immunisation Register. Vaccine failure was defined as proven pneumococcal infection with a vaccine serotype despite receiving a three-dose schedule of that vaccine. Long-term renal sequelae were defined as presence of proteinuria, hypertension (systolic blood pressure (SBP) >90th centile), use of antihypertensive agents, impaired estimated glomerular filtration rate (eGFR) or requirement for renal replacement therapy (RRT). eGFR was calculated using a modified Schwartz formula.9 Chronic kidney disease (CKD) was defined, according to the National Kidney Foundation’s Kidney Disease Outcomes Quality Initiative, as either kidney damage or GFR <60 mL/min/1.73 m2 for ≥3 months, and classified from stage I (mild disease) through stage V (end-stage kidney disease (ESKD)).10 Missing height data were estimated from individual proceeding height centile from WHO growth charts.11 SBP centile was calculated using National Heart Lung and Blood Institute’s Fourth Report on the Diagnosis, Evaluation and Treatment of High Blood Pressure in Children and Adolescents, May 2005.12
Sixty-six patients were diagnosed with HUS between January 1997 and December 2016. Of these, 22 (33%) had proven STEC, 19 (29%) had probable STEC, 11 (17%) had proven pneumococcal HUS, 5 (8%) had another associated infection (Streptococcus pyogenes bacteraemia, Streptococcus pyogenes pharyngitis, Clostridium difficile colitis, Enterococcus faecalis and Stenotrophomonas maltophilia urinary tract infections) and 9 (14%) had proven or suspected atypical HUS. Pneumococcal HUS represented 4% (1/26) of all cases of HUS between January 1997 and December 2004, 20% (3/15) between January 2005 and June 2011 (PCV7 era) and 28% (7/25) between July 2011 and December 2016 (PCV13 era) (see figure 1).
Patients with pneumococcal HUS were a median age of 12 (range 7–20) months (table 1). All patients were born in Australia and were non-Indigenous. More specific ethnicity data were not available. The median duration of hospitalisation was 17 (range 9–36) days, median nadir haemoglobin was 59 (range 35–69) g/L, median nadir platelet count was 16 (range 7–38) ×109L, and median lowest eGFR was 10 (range 5–35) mL/min/1.73 m2. Two patients (cases 9 and 10) had disseminated intravascular coagulation, while in the remaining nine patients, the INR ranged from 1.1 to 1.3 (normal 0.9–1.1), activated partial thromboplastin time 35–72 (normal 23–35) s and fibrinogen 5.0–10.8 (normal 0.8–3.8) g/L. Coombs test was positive in four of nine patients tested. Nine patients required RRT (seven peritoneal dialysis (PD) and two continuous veno-venous haemofiltration) for a median of 7 (range 2–428) days. All patients received blood products, including packed red blood cells, platelets, albumin and fresh frozen plasma.
S. pneumoniae was isolated from sterile biological fluid in all 11 patients. Nine patients were culture positive, and three were detected by pneumococcal PCR. Primary sites of infection included pneumonia in nine patients (4/9 with bacteraemia, 8/9 with empyema) and meningitis with associated bacteraemia in two patients. Pneumococcal serotype was identified in all patients. Subtype 19A was the most prevalent (6/11), followed by subtype 3 (2/11) and one case each of 7F, 10A and 14 (1/11). All isolates were susceptible to penicillin except for three isolates of serotype 19A (cases 4, 8 and 11), which were found to have intermediate penicillin susceptibility. Median duration of antibiotic therapy was 38 (range 16–65) days.
Nine patients had received a complete three-dose schedule of a PCV. The shortest time interval between completing the series and presentation to hospital was 26 days. Two patients were unvaccinated. All four patients who received PCV7 were infected with a non-vaccine serotype. Four of the five patients who received PCV13 were infected with a vaccine serotype.
Outcome data were available for all patients (table 2). Median follow-up was 14 (range 1–108) months. Of seven patients with follow-up data available beyond 1 month, three had a normal eGFR and four had CKD. Six of 11 patients were hypertensive or on antihypertensive medication. We observed no neurological sequelae and no mortality. Case 6 had a single, isolated seizure during their admission. Case 10 had MRI findings of cerebellar hyperintensity; however, subsequent imaging was normal.
This case series of pneumococcal HUS is the largest in the post-PCV13 era and demonstrates an increasing proportion of HUS cases attributable to pneumococcal infection between 1996 and 2016. This may be attributable to greater awareness and identification of pneumococcal HUS but could reflect increasing incidence. While we cannot determine incidence through this case series, our findings are aligned with data reported internationally. A number of large, population-based studies that sought to identify cases of HUS among patients with IPD found that the proportion of HUS cases secondary to pneumococcal infection is apparently increasing over the last two decades. A review of children with IPD and HUS admitted to hospitals in the USA between 1997 and 2009 reported that the number of hospital admissions with pneumococcal HUS doubled in this period.2 Similarly, a Utah study of paediatric HUS and IPD admissions found an increased proportion, from 0.3% to 5.6%, of IPD complicated by HUS since the introduction of PCV7 in 2000.3 Pneumococcal HUS is most commonly associated with complicated pneumonia.7 8 13 14 A review of children with pneumococcal empyema, who were admitted to a tertiary referral hospital in the UK, found 7.5% (6/80) developed HUS, with most cases occurring in children under 2 years old.15
Clinical course, treatment and outcomes
Pneumococcal HUS is associated with a high morbidity and mortality, with reported complications including CKD, proteinuria, hypertension, the need for ongoing RRT including dialysis and transplantation and neurological sequelae.4Among our cohort, 82% (9/11) of patients required acute RRT, predominantly PD (7/9), which compares with a recent case series in New Zealand, where all nine cases received PD.16 Little data exist to allow comparison of outcomes between different means of RRT.2 7 8 17 PD is comparatively simple and a potential therapy for children with pneumococcal HUS in developing countries.18 Certainly, among our cohort, PD in the acute setting did not result in progression to ESKD or death.
UK data report high early mortality (11%, 5/43), while in the Asia-Pacific region a cross-sectional study of children with pneumococcal HUS in New Zealand found 9% mortality (1/11).7 16 A 2013 review of available literature estimated that long-term renal and neurological sequelae occur in 26%–40%, ESKD occurs in 10%–16% and that mortality is between 2% and 12%.4 Comparatively, among those patients with adequate follow-up data, approximately half (4/7) of our patients had CKD. Adequate control of blood pressure and proteinuria is important in slowing the progression of CKD.19 Despite this, we note that 5 of 11 patients had prehypertension, hypertension or significant proteinuria at follow-up.
The true prevalence of CKD may be underestimated due to limited follow-up data. There were no cases of ESKD, no mortality and no neurological sequelae, even among those patients with pneumococcal meningitis.
Serotype and vaccination status
The advent of pneumococcal vaccination has resulted in a decline in IPD; however, the impact of vaccination on the nature of pneumococcal HUS is unclear.5 6 14 19 A retrospective cohort study showed that the introduction of PCV7 vaccination reduced systemic complications of pneumococcal pneumonia including HUS by 35% in children under 1 year old but also found increasing rates of local complications for pneumococcal pneumonia in all age groups, suggesting more invasive disease.14
Prior to the era of pneumococcal vaccination, serotypes most likely to cause HUS were 3, 6B, 8, 9V, 14, 19 and 23F.4 20 21 After the introduction of PCV7, published data suggested the emergence of non-vaccine serotypes 19A, 1, 3, 6A and 7F, with many authors advocating for these strains to be included in subsequent vaccines.3 4 7 8 17 19 In retrospective reviews of pneumococcal HUS reporting serotypes, coverage of PCV13 would have been 11/12 (92%) from the UK data, and 22/24 (92%) and 12/12 (100%) from North American data.7 8 17 Subtype 19A has emerged as a major cause of HUS is the post-PCV7 era, potentially related to enhanced expression of neuraminidase.8 22 Serotype 19A was the most common in our cohort, even in those vaccinated with PCV13. Vaccine failure was observed in patients infected with serotypes 7F and 19A, despite data that suggest that PCV13 is effective in preventing infection with these serotypes.23 24
Few studies have examined the effect of PCV 13 vaccination on serotype replacement and disease severity in pneumococcal HUS.4 While a number of case reports of pneumococcal HUS have been published in the post-PCV13 era, there are limited data pertaining to vaccine status and pneumococcal serotype.25–32 Although many authors have advocated higher valency conjugate vaccines in order to prevent emerging serotypes of IPD, four of the five children in our cohort, who were vaccinated with PCV13, were infected with a vaccine serotype. While these cases were classed as vaccine failure, it would appear that these children had a reduced requirement for dialysis compared with the PCV7 cohort.33 Two of four cases of vaccine failure did not require RRT, while the remaining two required only short-term dialysis. By comparison, renal replacement therapy was universally required in both unvaccinated patients and patients infected with a non-vaccine serotype. This change may reflect earlier diagnosis and advances in treatment, including critical care; however, we speculate it may also suggest reduced disease severity.
Pneumococcal infection remains a major cause of HUS despite improved vaccine coverage and a reduction in the prevalence of IPD. Pneumococcal serotype 19A is most commonly associated with pneumococcal HUS despite the introduction of the PCV13 vaccine. Among our case series, pneumococcal HUS tended to be severe, with most children developing renal impairment in the acute phase, requiring RRT and prolonged hospital admission. PD was used in the majority of our patients with no mortality, ESKD or neurological sequelae. Nevertheless, most patients require long-term follow-up due to the high chance of progression to CKD, which was observed in around half of our cases. Future research is imperative to examine the impact of serotype switching on disease severity, clinical course and outcomes in pneumococcal HUS.
Contributors JL: conceptualised and designed the study, collected data, carried out the initial analyses, drafted the initial manuscript and approved the final manuscript as submitted. AG and CQ: conceptualised and designed the study, critically reviewed and revised the manuscript and approved the final manuscript as submitted. All authors approved the final manuscript as submitted and agree to be accountable for all aspects of the work.
Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.
Competing interests None declared.
Patient consent Not required.
Ethics approval Ethics approval was obtained from the Royal Children’s Hospital Human Research Ethics Committee (HREC 36185A).
Provenance and peer review Not commissioned; externally peer reviewed.
Data sharing statement All data collected for this case series are summarised in the article.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.