The increased risk of invasive pneumococcal disease (IPD) is 50-fold in children with hematological malignancies such as leukemia.1 IPD is a vaccine preventable disease, with both a polysaccharide vaccine (23-valent pneumococcal polysaccharide vaccine) and pneumococcal conjugate vaccine (PCV) available.2 The 23-valent pneumococcal polysaccharide vaccine is not efficacious in children <2 years of age and the immunogenicity response suboptimal in hematological malignancies.3–5 PCVs are therefore the vaccines of choice in high-risk immunosuppressed patients.6,7
In a UK study reviewing the pneumococcal serum antibody levels in a vaccine-naïve pediatric leukemia population, the majority did not have protective levels of antibody post completion of therapy, ranging from 5% (2/42) for serotype 9V to 31% (13/42) for serotype 19F.8 Lehrnbecher et al9 also identified low levels of pneumococcal antibody in vaccine-naïve childhood acute lymphoblastic leukemia (ALL) patients (n = 53), up to 9 months after completion of therapy. To try and optimize protection against IPD, the Australia Immunisation Handbook recommends patients with leukemia be vaccinated as soon as possible after diagnosis.10 The vaccine(s) recommended depends on the patients’ age and immunization history. If pneumococcal vaccines are only administered at the completion of chemotherapy, there is a window of approximately 3 years where a child with ALL is potentially not protected, despite being at high risk of IPD.
The PCV7 vaccine was introduced onto the Australian National Immunisation Program for infants in 2005, with a funded catch-up dose for children <2 years of age. At the time of this study (2010), a PCV10 vaccine was registered and available in Australia, but not included on the National Immunisation Program. The study aim was to determine in children and adolescents the immunogenicity of a PCV (PCV10) administered as soon as possible after the diagnosis of leukemia.
The study was conducted at the 2 tertiary Children’s Cancer Centres in Melbourne, Australia: The Royal Children’s Hospital, Parkville, and Monash Children’s Hospital, Clayton.
The PCV10 vaccine (Synflorix, GSK, Brentford, England) is a formulation of pneumococcal nontypeable Haemophilus influenzae protein D conjugate vaccine; a decavalent (10-antigen containing) vaccine composed of partially hydrolyzed capsular (outer coat) polysaccharide and oligosaccharide antigens from 10 Streptococcus pneumoniae bacteria serotypes (1, 4, 5, 6B, 7F, 9V, 14, 18C, 19F, 23F).11 Eight of the antigens are chemically conjugated to H. influenzae protein D from a nontypeable strain of H. influenzae, 1 conjugated to a Diphtheria toxoid carrier protein and another conjugated to a Clostridium tetani (tetanus) toxoid carrier protein.11
The ALL patients were stratified according to international criteria and Children’s Oncology Group study protocols: AALL0232, AALL033, AALL0434 and Study V11. All protocols have intensive induction and consolidation phases over the first 9 months of treatment, with chemotherapeutic agents (including corticosteroids) depending on the risk stratification protocol. Maintenance chemotherapy consisted of daily oral mercaptopurine and intermittent methotrexate and vincristine according to the participants protocol, with total treatment duration of approximately 3 years.
All children receiving therapy for leukemia, either ALL or acute myeloblastic leukemia (AML), were considered for inclusion in the study. Diagnoses were confirmed histopathologically according to national and international guidelines and International Classification of Diseases 10th Edition coding. Exclusion criteria included children who had a relapse of their leukemia, or a previous (or planned) hemopoietic stem cell transplant. If a study subject relapsed postenrollment, he or she was excluded from ongoing participation. Recent immunoglobulin therapy (within last 9 months) and previous anaphylaxis to a pneumococcal vaccine were additional exclusion criteria.12
Participants were divided into 2 groups depending on the previous pneumococcal immunization status. Group 1 had completed their primary PCV7 course before leukemia diagnosis, so received a single PCV10 dose. Group 2 had received none, or an incomplete PCV7 course, so were scheduled to receive 3 doses of PCV10 each 2 months apart. Patients did not receive PCV10 if their platelets were <50 × 109/L or neutrophils <500 × 106/L. It was clarified with the treating oncologist that participants were clinically stable before vaccine administration.
The pneumococcal immunogenicity enzyme-linked immunosorbent assay (ELISA) and opsonophagocytic assay (OPA) analysis was undertaken at the Murdoch Childrens Research Institute pneumococcal laboratory as per published methodology.13,14 Serotype-specific ELISA IgG were measured for the 10 serotypes in PCV10. Geometric mean concentrations (GMCs) were determined and antibody levels >0.35 μg/mL used as the primary serological correlate of protection.15–17 The primary immunogenicity endpoints were determined at >1 month post the PCV10 vaccine dose for those in group 1 and >5 months post study entry for group 2. An increase in serotype-specific IgG >2-fold rise for those positive at baseline was considered a significant immunogenic response to the PCV10 vaccine. A serologic response above this protective threshold in >60% of participants to 5 of the 10 serotypes in the PCV10 vaccine was considered protective.
The secondary endpoints included the following: (1) functional OPA response to 4 serotypes (1, 6A, 19F, 23F), with assay titers ≥1:8 considered significant; (2) immunogenicity sensitivity analysis of serotype-specific ELISA titers ≥0.20 μg/mL was used to compare study results to published immunogenicity data for PCV10 in infants11,18 and (3) safety of the vaccine, with a particular focus on febrile responses and tolerability in those aged ≥6 years in whom the PCV10 vaccine was not licensed at the time of the study.
In Australia, approximately 700 children and adolescents aged 0–18 years are diagnosed each year with cancer.19 One-third of these patients have leukemia (ALL or AML). In Victoria, approximately 80 children with leukemia are treated annually, managed at the 2 study sites. A recruitment rate of 60%, over a 12-month time frame, produced a study target of 50 participants, which is comparable with other vaccine immunogenicity studies in oncology cohorts.20,21
The ELISA serotype-specific antibody concentrations and OPA titers were log(e) transformed for statistical analysis to approximate normal distributions. The ELISA GMCs were determined by taking the antilogarithm of the means of the log-transformed values. The paired t test was used to compare the mean at postimmunization to preimmunization, with a P value <0.01 considered significant due to multiple comparisons. The ratios of the postimmunization to preimmunization geometric means were determined with 95% confidence intervals, where a ratio of 1 (null hypothesis value) indicates no change. For continuous nonparametric variable analysis, Wilcoxon rank (2 variables) and Kruskal–Wallis (3 variables) tests were used. Correlations between serotype-specific ELISA IgG concentrations and OPA titers were calculated using the Pearson correlation coefficient.
A standard postvaccine diary card was completed for each administered dose for 5 days [day of immunization (day 0) and next 4 days]. Information collected included: daily temperature, local reactions at the PCV10 site and any systemic reactions (eg, disrupted sleep, irritability, vomiting, diarrhea and rash not limited to the injection site). These data were correlated following a phone interview 5 days postimmunization and the diary forwarded back to the investigators.
The data were entered into an Access database (Microsoft, Redmond, WA) and analyzed using Stata 11.0 (College Station, StataCorp, TX). The Southern Health and Royal Children’s Hospital Human Research Ethics Committee approved the study (HREC # 28146). It was registered on the Australian and New Zealand clinical trials registry (No. ACTRN12609000515291).
We enrolled 39 participants between May 3, 2010, and January 13, 2011 (Fig. 1). All patients except 1 had baseline serology and 37 had at least 1 PCV10 vaccine administered. There were 69% (27/39) group 1 participants and 31% (12/39) PCV-naïve group 2 participants. Of the participants in group 1, 68% (17/25) had had 3 PCV7 infant doses, 20% (5/25) single dose and 12% (3/25) 2 doses. In participants receiving a PCV10 vaccine, 97% (36/37) had ALL and 1 participant AML. Two participants from group 2 withdrew from the study after receiving 2 of the scheduled 3 PCV10 doses (1 ALL relapse; 1 concerned regarding possible vaccine side effects).
The baseline demographics are detailed in Table 1. The median age was 6.2 years, and 62% were male. Fifty-five percent (21/38) were on maintenance chemotherapy at the time of baseline immunogenicity. The median length of time from diagnosis to baseline serology was 7.4 months and the median time between dose-1 PCV10 vaccine and immunogenicity serology was 1.9 months.
At baseline, the percentage of participants overall with protective GMCs above the protective threshold (>0.35 μg/mL) ranged from 5.3% for serotype 4 to 71% for serotype 19F. Following immunization, there was a significant increase in GMC from baseline for all serotypes (P < 0.01). Following a single dose of PCV10, the percentage of all participants with protective GMC above the protective threshold ranged from 33.3% for serotype 4 to 81.1% for serotype 14. The primary study endpoint of >60% of participants above the protective threshold following a single PCV10 dose was achieved in 5 of the 10 PCV serotypes (6B, 9V, 14, 19F, 23F; Fig. 2). In those participants who were seropositive at baseline, the percentage who had a >2-fold rise in GMC ranged from 0% to 63.6%. A sensitivity analysis for the combined results (groups 1 and 2) post a single PCV10 dose using the GSK 0.20 μg/mL protective threshold cutoff,11 compared with infant historical data, was undertaken (data not shown). Using this lower threshold (0.20 μg/mL) for our oncology cohort, 8 of the PCV10 serotypes had 60% or more participants above this cutoff.
A subgroup analysis of immune response by subgroups of intensive early-phase chemotherapy and maintenance chemotherapy is detailed in Table 2. It identified that the maintenance chemotherapy subset in group 1 (single dose) was more likely to have a significant increase above baseline, with 9 of the 10 serotypes having a P value <0.01. Only 1 of the 10 serotypes (serotype 14) had an increase above baseline in the intensive therapy subgroup. There was no statistical difference in the group 2 analysis by subgroup. The percentage of participants above the protective threshold (>0.35) for group 1 ranged from 45% to 82% (maintenance subgroup) and 38% to 86% (intensive). For group 2 (dose 1) it ranged from 45% to 82% (maintenance subgroup) and 38% to 86% (intensive chemotherapy subgroup).
The overall immunogenicity data by group are detailed in Tables 3 and 4. The primary endpoint of >60% of participants above the threshold following a single PCV10 dose in group 1 was obtained in 7 of the 10 PCV serotypes (4, 6B, 9V, 14, 19F, 18C, 23F). These are all serotypes in the PCV7 vaccine, which by inclusion criteria had been received before diagnosis. In group 2 (Table 4) at baseline, the percentage with GMC above the protective threshold (>0.35 μg/mL) ranged from 9.1% for serotype 4 to 54.6% for serotype 19F. The GMC only increased significantly from baseline (P value <0.01) following dose-1 PCV10 in 1 serotype (23F). There was not a significant difference directly comparing GMC between PCV10 doses 2 and 3. The primary endpoint in group 2, of >60% of participants being above the threshold following a PCV10 course (2 or 3 doses), was obtained in 7 of the 10 PCV serotypes (6B, 7F, 9V, 14, 19F, 18C, 23F).
At baseline, the percentage above the OPA threshold titer (≥1:8) to 4 serotypes (1, 6b, 19F, 23F) ranged from 0% (serotype 1, a non-PCV7 serotype) to 40% for serotype 6B. Following a single dose of PCV10 vaccine this increased to 39.4% (serotype 1), 83.8% (serotype 6B), 86.5% (serotype 19F) and 70.3% (serotype 23F). ELISA and OPA showed good correlation for the serotypes in both the 7-valent and 10-valent PCVs (6B, 19F and 23F), with the Pearson coefficient (r) ranging from 0.81 to 0.84 (Fig. 3). It was lower for the non-PCV7 serotype 1, with r = 0.51.
There was 1 participant with a diagnosis of AML (myeloid leukemia of Down syndrome), receiving intensive chemotherapy over 6 monthly cycles. The individual was 23 months of age at diagnosis and in group 1, having received 3 previous PCV7 doses at the routine 2, 4 and 6 months of age. At study baseline, 4.5 months since diagnosis, he was seropositive to 4 serotypes (6B, 9V, 19F, 23F). At 1 month post the vaccine, GMC ELISA antibody levels were seroprotective to 6 of the PCV10 serotypes (1, 5, 6B, 7F, 14, 23F).
The majority of adverse events following immunization reported were tenderness at the injection site in 73% (27/37). Systemic symptoms were reported in 41% (15/37) overall, with diarrhea being the most common (4 participants). There was 1 reported fever >38°C in a non-neutropenic patient, who had some associated diarrhea and required hospital admission for 48 hours of intravenous fluids. These symptoms while temporally related to vaccine timing may have been chemotherapy-related side effects.
This Australian study describes the serum immune response to PCV10 immunization in children and adolescents receiving chemotherapy for leukemia. The majority had received a previous PCV (PCV7) in infancy, but did not have protective antibody levels above the >0.35 μg/mL threshold at baseline, with 6 serotypes <20% protective coverage and only 1 (serotype 19F) having >60% of participants seroprotected. In group 1, the primary study endpoint was obtained in all of the PCV7 serotypes, indicating a satisfactory immunogenicity “booster” response. In the subgroup analysis by therapy intensity, those on maintenance therapy (n = 11) were more likely to have a significant increase in GMC from baseline in 9 of 10 serotypes. The primary endpoint in group 2, following a PCV10 course (2 or 3 doses), was obtained in 7 of the PCV10 serotypes. There was no difference in subgroup analysis by chemotherapy intensity following dose-1 PCV10 vaccine.
The results of this study can be compared with that of a PCV7 immunogenicity study in a Hong Kong pediatric oncology cohort.22 Cheng et al22 reviewed the immunogenicity post 2 doses of PCV7 vaccine administered a month apart in 44 patients [29 hematological (25 ALL); 15 solid tumors]. Pneumococcal serology was taken at baseline and 1 month post the 2nd PCV7 dose. At the study endpoint, the protective antibody GMC above threshold to the PCV7 serotypes ranged from 86% to 100%. Directly correlating our PCV10 study with the Hong Kong PCV7 data is difficult, as a large number of patients in the study by Cheng et al22 were completely off therapy (median 6 months; range 1–12 months). The high levels of baseline seropositive results may reflect pneumococcal carriage incidence and the levels of IPD in Hong Kong, calculated as 18.3 per 100,000 in children aged ≤2 years (1995–2004).23
A limitation of both this PCV10 study and the PCV7 study by Cheng et al22 is the small sample size. This was particularly in group 2 (12 participants), with recruitment hampered in adolescents when they were advised that 3 doses of PCV10 were required. The 2 participants who withdrew before the 3rd dose also affected the comparison between 2 and 3 doses in group 2. Both withdrawal participants had very robust responses to 2 doses, and it is encouraging that despite being PCV naïve, the study endpoint was obtained in 7 of the 10 PCV serotypes. It was also a limitation of subgroup analysis by intensive therapy, with only 4 participants in this group. This study also did not assess pneumococcal antibody status at the end of therapy, and these data would help in formulating the optimal guideline for long-term IPD protection. With only 1 AML participant, more data are required to determine the serum immune responses in this disease.
A sensitivity analysis reviewed how the findings would vary depending on the serum immune response threshold. Using the lower threshold of 0.20 μg/mL, the number of serotypes that reached the primary study endpoint increased to 8 serotypes.11 This highlights a requirement to standardize threshold cutoffs for PCV studies, particularly those in immunocompromised hosts. As expected, the lowest immune response was in the non-PCV7 serotypes (1, 5 and 7F) indicating that more than 1 dose of PCV may be required to optimize protection for naïve pneumococcal serotypes.
A satisfactory correlation between OPA and ELISA was obtained for 3 of the 4 serotypes. The lower correlation for serotype 1 most likely relates to it being a non-PCV7 serotype, so it was not a “booster” dose for the majority of group 1 participants. These findings are in keeping with the moderate correlation between IgG ELISA and OPA (r values, 0.41–0.70) found in a PCV7 study of children and young adults with sickle cell disease.24
The main solicited adverse event following immunization reported in the study was pain or tenderness at the injection site in 73% (27/37). Adolescents and older children often have higher reports of pain at the injection site, as seen with the 81% reporting rate following a quadrivalent human papillomavirus vaccine.25 It was reassuring that there were few systemic adverse events following immunization reported and no admissions for febrile neutropenia postimmunization, which reflected our strict protocol of delaying vaccine administration if neutropenic or any concerns from the participant or treating oncologist.
This study overcame some of the logistical difficulties in administering vaccines to children and adolescents with cancer, highlighted in both Australia and the United Kingdom.26–28 Bate et al28 in a survey of principal treating pediatric oncologists found PCV was recommended post completion of chemotherapy by only 58%, reflecting the lack of specific guidelines in the United Kingdom regarding this vaccine in pediatric cancer patients. The administration of vaccine only at completion of therapy means that they are still vulnerable when most at risk of IPD, that is, when immunosuppressed during therapy.1
Post this study, in mid-2011, Australia’s National Immunisation Program decided to transition from the PCV7 vaccine to PCV13.29 This vaccine is made by the same manufacturer as PCV7 and has 6 additional serotypes (1, 3, 5, 6A, 7F, 19A). This includes 19A, identified as the most prevalent cause of IPD in most Australian states following PCV7 introduction.30 With this change in the Australian National Immunisation Program, PCV13 is now the funded “booster” vaccine for special-risk groups and recommended during chemotherapy, as well as a “booster” 6 months post completion of therapy.10 It remains crucial to have ongoing national IPD surveillance and capture medical risk factors, such as chemotherapy for pediatric oncology. This will help monitor disease prevalence in special-risk populations and formulation of strategies to administer additional doses required, both during and after therapy.
In conclusion, this study identified that it is possible and safe to administer PCV10 vaccine during therapy for pediatric leukemia. It provided a satisfactory serum immune response for the majority of vaccine serotypes. A single-dose “booster” in those who had received PCV7 before diagnosis produced a significant increase in response, particularly in those participants on maintenance chemotherapy. Multiple doses may be required during therapy and for vaccine-naïve serotypes.
The authors thank the patients and families involved in the study and SAEFVIC research staff: Annette Alafaci, Ainsley Gillies, Thao Nguyen and Julie Quinn. N.W.C. acknowledges support from an NHMRC postgraduate PhD scholarship. The authors also thank the treating pediatric oncologists at the Children’s Cancer Centre Royal Children’s Hospital and Monash Children's Hospital (alphabetical: David Ashley, John Heath, Lisa Orme, Elizabeth Smibert, Leanne Super, Karin Tiedemann, Keith Waters).
1. Hjuler T, Wohlfahrt J, Staum Kaltoft M, et al. Risks of invasive pneumococcal disease
in children with underlying chronic diseases. Pediatrics. 2008;122:e26–e32
2. Rijkers GT, van Mens SP, van Velzen-Blad H. What do the next 100 years hold for pneumococcal vaccination
? Expert Rev Vaccines. 2010;9:1241–1244
3. Moberley S, Holden J, Tatham David P, et al. Vaccines for preventing pneumococcal infection in adults. Cochrane Database Syst Rev. 2013:CD000422
4. Goldblatt D. Bacteria, polysaccharides, vaccines and boosting: measuring and maintaining population immunity. Arch Dis Child. 2008;93:646–647
5. Spickermann D, Gause A, Pfreundschuh M, et al. Impaired antibody levels to tetanus toxoid and pneumococcal polysaccharides in acute leukemias. Leuk Lymphoma. 1994;16:89–96
6. Centers for Disease Control and Prevention. . Use of 13-valent pneumococcal conjugate vaccine and 23-valent pneumococcal polysaccharide vaccine for adults with immunocompromising conditions: recommendations of the Advisory Committee on Immunization Practices (ACIP). MMWR Morb Mortal Wkly Rep. 2012;61:816–819
7. Orthopoulos GV, Theodoridou MC, Ladis VA, et al. The effect of 23-valent pneumococcal polysaccharide vaccine on immunological priming induced by 7-valent conjugate vaccine in asplenic subjects with [beta]-thalassemia. Vaccine. 2009;27:350–354
8. Patel SR, Bate J, Borrow R, et al. Serotype-specific pneumococcal antibody concentrations in children treated for acute leukaemia. Arch Dis Child. 2010;97:46–48
9. Lehrnbecher T, Schubert R, Behl M, et al. Impaired pneumococcal immunity in children after treatment for acute lymphoblastic leukaemia. Br J Haematol. 2009;147:700–705
13. Balloch A, Licciardi PV, Leach A, et al. Results from an inter-laboratory comparison of pneumococcal serotype-specific IgG measurement and critical parameters that affect assay performance. Vaccine. 2010;28:1333–1340
14. Russell FM, Carapetis JR, Burton RL, et al. Opsonophagocytic activity following a reduced dose 7-valent pneumococcal conjugate vaccine infant primary series and 23-valent pneumococcal polysaccharide vaccine at 12 months of age. Vaccine. 2011;29:535–544
16. Siber GR, Chang I, Baker S, et al. Estimating the protective concentration of anti-pneumococcal capsular polysaccharide antibodies. Vaccine. 2007;25:3816–3826
17. Plotkin SA. Correlates of protection induced by vaccination
. Clin Vaccine Immunol. 2010;17:1055–1065
18. Romanowski B, de Borba PC, Naud PS, Roteli-Martins CM, De Carvalho NS, Teixeira JC, Aoki F, Ramjattan B, Shier RM, Somani R, Barbier S, Blatter MM, Chambers C, Ferris D, Gall SA, Guerra FA, Harper DM, Hedrick JA, Henry DC, Korn AP, Kroll R, Moscicki AB, Rosenfeld WD, Sullivan BJ, Thoming CS, Tyring SK, Wheeler CM, Dubin G, Schuind A, Zahaf T, Greenacre M, Sgriobhadair AGlaxoSmithKline Vaccine HPV-007 Study Group. . Sustained efficacy and immunogenicity of the human papillomavirus (HPV)-16/18 AS04-adjuvanted vaccine: analysis of a randomised placebo-controlled trial up to 6·4 years. Lancet. 2009;374:1975–1985
20. Patel SR, Ortin M, Bernard JC, et al. Revaccination of children after completion of standard chemotherapy for acute leukaemia. Clin Infect Dis. 2007;44:635–642
21. van Tilburg CM, Sanders EA, Rovers MM, et al. Loss of antibodies and response to (re-)vaccination
in children after treatment for acute lymphocytic leukemia: a systematic review. Leukemia. 2006;20:1717–1722
22. Cheng FWT, Ip M, Chu YYL, et al. Humoral response to conjugate pneumococcal vaccine in paediatric oncology patients. Arch Dis Child. 2012;97:358–360
23. Ho PL, Chiu SS, Cheung CH, et al. Invasive pneumococcal disease
burden in Hong Kong children. Pediatr Infect Dis J. 2006;25:454–455
24. Vernacchio L, Romero-Steiner S, Martinez JE, et al. Comparison of an opsonophagocytic assay and IgG ELISA to assess responses to pneumococcal polysaccharide and pneumococcal conjugate vaccines in children and young adults with sickle cell disease. J Infect Dis. 2000;181:1162–1166
25. Block SL, Brown DR, Chatterjee A, et al. Clinical trial and post-licensure safety profile of a prophylactic human papillomavirus (types 6, 11, 16, and 18) L1 virus-like particle vaccine. Pediatr Infect Dis J. 2010;29:95–101
26. Crawford NW, Heath JA, Buttery JP. Immunisation practices of paediatric oncologists: an Australasian survey. J Paediatr Child Health. 2007;43:593–596
27. Crawford NW, Heath JA, Ashley D, et al. Survivors of childhood cancer
: an Australian audit of vaccination
status after treatment. Pediatr Blood Cancer. 2010;54:128–133
28. Bate J, Patel SR, Chisholm J, et al. Immunisation practices of paediatric oncology and shared care oncology consultants: a United Kingdom survey. Pediatr Blood Cancer. 2010;54:941–946
30. Williams SR, Mernagh PJ, Lee MH, et al. Changing epidemiology of invasive pneumococcal disease
in Australian children after introduction of a 7-valent pneumococcal conjugate vaccine. Med J Aust. 2011;194:116–120
Keywords:© 2015 by Lippincott Williams & Wilkins, Inc.
childhood cancer; invasive pneumococcal disease; vaccination