Efficacy and safety of home-based subcutaneous immunoglobulin replacement therapy in paediatric patients with primary immunodeficiencies

Authors

  • M. Borte,

    Corresponding author
    1. Municipal Hospital St Georg Leipzig, Academic Teaching Hospital of the University of Leipzig, Germany
      M. Borte, Municipal Hospital St Georg, Academic Teaching Hospital of the University of Leipzig, Delitzscher Strasse 141, D-04129 Leipzig, Germany.
      E-mail: michael.borte@sanktgeorg.de
    Search for more papers by this author
  • E. Bernatowska,

    1. Child Health Institute, Warsaw, Poland
    Search for more papers by this author
  • H. D. Ochs,

    1. University of Washington School of Medicine, Department of Pediatrics, and Seattle Children's Research Institute, Seattle, WA, USA
    Search for more papers by this author
  • C. M. Roifman,

    1. The Hospital of Sick Children, Toronto, ON, Canada
    Search for more papers by this author
  • the Vivaglobin Study Group


M. Borte, Municipal Hospital St Georg, Academic Teaching Hospital of the University of Leipzig, Delitzscher Strasse 141, D-04129 Leipzig, Germany.
E-mail: michael.borte@sanktgeorg.de

Summary

Subcutaneous immunoglobulin infusions are effective, safe and well tolerated in the treatment of primary immunodeficiencies, but only limited data on the treatment of children are available. We investigated the efficacy, safety and pharmacokinetics of home therapy with a 16% liquid human immunoglobulin G preparation (Vivaglobin®) when administered subcutaneously in children with primary immunodeficiencies. Data were analysed from 22 children (2–<12 years) who participated in two prospective, open-label studies (one in Europe/Brazil, one in North America). All children had previously received intravenous immunoglobulins. They started weekly subcutaneous immunoglobulin infusions with an approximately 3-month wash-in/wash-out period, followed by a 6-month (Europe/Brazil) or 12-month (North America) efficacy evaluation period. In Europe/Brazil, subcutaneous doses generally equalled the previous weekly equivalent intravenous doses. In North America, subcutaneous doses during the efficacy evaluation period were 126% (median) of the previous weekly equivalent intravenous doses. Efficacy end-points in both studies included the occurrence of serious bacterial infections and any infections, and serum immunoglobulin G trough levels. Median serum immunoglobulin G trough levels exceeded those during previous intravenous therapy by 13% (North America) and 16% (Europe/Brazil). During the efficacy evaluation period of both studies, none of the children had a serious bacterial infection; the mean overall infection rate/patient year was 4·7 in Europe/Brazil and 5·6 in North America, concurring with previous reports in adults. The adverse event profile was comparable to previous reports in adults. Both studies confirmed the efficacy and safety of subcutaneous immunoglobulin therapy with Vivaglobin in children with primary immunodeficiencies.

Introduction

Patients with primary immunodeficiencies (PIDs) are susceptible to frequent, recurrent and severe infections, especially bacterial infections of the respiratory tract [1–3]. Immunoglobulin (Ig)G replacement therapy is standard practice for patients with primary antibody deficiencies.

Both intravenous immunoglobulin (IVIG) and subcutaneous immunoglobulin (SCIG) therapy effectively reduce the risk of serious infections in adults and children [3–7]. SCIG infusions are typically given weekly and at smaller doses [3–6,8,9], resulting in lower peak and higher trough levels of IgG compared to the large boluses given at 2-, 3- or 4-week intervals with IVIG infusions [3,9,10]. High and stable serum IgG trough levels are crucial to provide adequate protection against infections [7,11].

IVIG infusions can be problematic in some patients because they may be associated with recurrent systemic reactions [10,12], and administration can be difficult in patients with poor venous access, a frequent problem in children [9]. Because PIDs are diagnosed frequently in childhood, the number of children requiring regular immunoglobulin replacement therapy is relatively high. SCIG therapy may overcome some of the limitations of IVIG therapy in children, given that no venous access is needed and that SCIGs can be self-administered conveniently (or administered by a parent or guardian) at home [3–5,7], reducing the time off school or work for the children and their families. The benefits of home-based SCIG therapy are reflected in improved quality of life and treatment satisfaction reported by children and adults previously receiving IVIG therapy in hospitals [13–15].

Vivaglobin® (CSL Behring GmbH, Marburg, Germany) is the first drug to be approved specifically for SCIG therapy in the United States, in January 2006. It was first approved for this indication in Germany in December 2002.

Here we report on the data obtained from 22 children <12 years of age enrolled in two multi-centre studies evaluating the efficacy, safety and pharmacokinetics of SCIG replacement therapy with Vivaglobin in patients with PID. Results from the overall study population (adults and children) have been reported previously [3,7].

Methods

Study design

Two prospective, open-label studies (one in Europe/Brazil and one in North America) investigated the efficacy, safety and pharmacokinetics of SCIG therapy with Vivaglobin in patients with PID. Baseline data, including steady-state serum IgG trough levels during previous IVIG therapy, were obtained 1–4 weeks before the first SCIG infusion. Weekly SCIG infusions during an approximately 3-month wash-in/wash-out period were started at the time the next IVIG infusion was scheduled (i.e. 3 or 4 weeks after the last IVIG infusion) in the European/Brazilian study, and 1 week after the last IVIG infusion in the North American study. After several SCIG infusions under supervision at the hospital, SCIG infusions were self-administered by the patient (or administered by a parent or guardian) at home. The wash-in/wash-out period was followed by an efficacy evaluation period of 28 weeks in Europe/Brazil and 52 weeks in North America, which included pharmacokinetic substudies.

Patients with PID were eligible for the studies if they required regular IgG replacement therapy and, in North America, weighed ≥10 kg. Before enrolment, patients had to have received IVIG therapy for at least 4 months and had to have a stable serum IgG trough level >5 g/l (or, in North America, ≥3·5 g/l above their IgG level before receiving IgG therapy). Relevant exclusion criteria included: evidence of current infection (North America only), bleeding disorders, requirement for immunosuppressive therapy, history of anaphylactic reactions to an IgG preparation, severe chronic diseases and known infection with hepatitis A, B or C, or human immunodeficiency virus. For inclusion in the pharmacokinetic substudy in Europe/Brazil, patients had to be ≥6 years of age. In North America, patients had to be ≥10 years of age and had to have been treated with Gamimune® N, a 10% IVIG preparation, for ≥3 months before enrolment into the study to be included in the pharmacokinetic substudy.

All children between 2 and <12 years of age enrolled in the two studies were included in the current analysis.

The protocols were approved by the independent ethics committee or institutional review board for each centre, and written informed consent was obtained from a legally acceptable representative of each child. The studies were performed in accordance with the Declaration of Helsinki and Good Clinical Practice Guidelines as well as any applicable local regulations.

Treatment

Vivaglobin was supplied as a ready-to-use pasteurized liquid containing polyvalent human IgG at a concentration of 160 mg/ml (16%). Subcutaneous infusions were administered at one or more injection sites using a portable infusion pump [3,7]. The rate of infusion was limited to ≤22 ml/h at each injection site in Europe/Brazil, and to ≤20 ml/h in North America. The maximum volume per injection site was 15 ml.

The dose regimens differed between the two studies: (i) in Europe/Brazil, patients received a weekly SCIG dose that generally equalled their previous weekly equivalent IVIG dose. During the wash-in/wash-out period, the dose could be adjusted once to a maximum of 150 mg/kg in case of an insufficient IgG trough level of <5 g/l. If the adjusted dose still yielded insufficient trough levels, the patient was discontinued from the study. (ii) In North America, the initial weekly SCIG dose required by the protocol was 120% of the previous weekly equivalent IVIG dose. At the end of the wash-in/wash-out period, the SCIG dose that was needed to attain a systemic IgG exposure [area under the concentration–time curve (AUC)] comparable to that measured during the previous IVIG therapy was determined for each patient in the pharmacokinetic substudy. At the beginning of the efficacy evaluation period, the SCIG doses were adjusted based on these determinations due to requirements of the Food and Drug Administration (FDA).

Study outcomes

Efficacy and pharmacokinetics.  Efficacy analyses were based on data obtained during the efficacy evaluation period. The primary efficacy variable was the serum IgG trough level in Europe/Brazil, and the number of clinically documented serious bacterial infections (defined as bacterial pneumonia, meningitis, sepsis, osteomyelitis or visceral abscess) in North America. Secondary efficacy variables included relevant bacterial infections (definition as above) in Europe/Brazil and serum IgG trough levels in North America. Thus, although different primary variables were used, the assessments were largely the same and therefore allowed for a combined evaluation of the children included in both studies. Other secondary efficacy variables in both studies included: non-serious infection and fever episodes, days missed school/kindergarten/work, days hospitalized due to infections and the use of antibiotics.

During the efficacy evaluation period, serum IgG trough levels were measured every 4 weeks in Europe/Brazil and approximately every 12 weeks in North America. Serum concentrations of IgG subclasses were measured at weeks 1, 16, 20 and at the final evaluation visit in Europe/Brazil, and when total IgG concentrations were measured in North America.

In Europe/Brazil, the pharmacokinetic substudy comprised determinations of serum IgG concentrations during one dosing interval between weeks 16 and 43 (samples taken before infusion, and at days 1, 2, 3, 5 and 7 after infusion).

In North America, for each patient in the pharmacokinetic substudy the mean of serum IgG trough levels at the end of the wash-in/wash-out period (i.e. before infusions in weeks 8–12) was compared to the mean IgG serum concentration during the previous IVIG treatment (the target level), which was derived from the AUC during IVIG treatment. Based on this comparison, individual dose adjustments that were needed to attain the target IgG serum level were calculated and implemented in the efficacy evaluation period. After at least 3 months of treatment with SCIG in the efficacy evaluation period, total IgG serum concentrations were measured every second day for 2 weeks for calculation of the AUC.

Safety.  Adverse events were documented throughout the study. Laboratory testing (haematology, serum chemistry and viral safety) was performed at several time-points. Vital signs and injection-site reactions were monitored before, during and 30 min after each infusion at the hospital. Physical examinations were performed at screening and at the final evaluation visit.

Statistical analysis

Due to differences in study design that reflect different regulatory guidance between Europe and North America (e.g. higher doses are requested by the FDA), the results from the two studies were analysed separately and are presented side by side. Data are presented using descriptive statistics. AUC values were obtained by the trapezoidal rule and standardized to a 7-day period. The data were processed using sas version 8·2 (SAS Institute Inc., Cary, NC, USA).

Results

Study populations and treatment

A total of 16 children (all male) in Poland, Germany and Brazil (European/Brazilian study) and six children (three female and three male) in North America were enrolled and treated with Vivaglobin (Table 1). All patients had previously received IVIG infusions at 3- or 4-week intervals (mean weekly equivalent doses of 93 mg/kg in Europe/Brazil and 132 mg/kg in North America; last IVIG dose of each patient). Two children discontinued their participation during the wash-in/wash-out period. A 5-year-old boy in North America was discontinued due to withdrawal of consent. The boy had serum IgG trough levels between 8·90 and 11·50 g/l and no infections during the wash-in/wash-out period; an adverse event of urine abnormality (abnormal microscopy finding possibly related to study drug) that was mild in intensity and resolved within 21 days was reported on the day of discontinuation. An 11-year-old boy in Europe/Brazil was withdrawn after his 11th infusion in the wash-in/wash-out period because of low IgG trough levels. He had received a previous IVIG dose of 405 mg/kg/month and started SCIG treatment at a weekly dose of 100 mg/kg. Serum IgG trough levels before infusions 1, 4 and 8 were 2·4, 3·1 and 3·2 g/l, respectively, and remained <5 g/l after dose adjustment to 150 mg/kg. The boy had ongoing bronchitis, pneumonia, otitis media, rhinitis and conjunctivitis at enrolment, and developed a symptomatic giardiasis infection with loose stools, which probably caused a substantial protein loss. He resumed IVIG therapy after withdrawal from the study, but his serum IgG trough levels remained low (<3 g/l). Thus, 20 of the 22 children were evaluated for efficacy. Twelve children in Europe/Brazil and four children in North America were included in the pharmacokinetic substudies.

Table 1.  Patient demographic and baseline characteristics.
CharacteristicEuropean/Brazilian study (n = 16)North American study (n = 6)
  1. BMI: body mass index; n: number of patients treated with Vivaglobin; s.d.: standard deviation; Ig: immunoglobulin.

Sex – no. (%)
 Female03 (50%)
 Male16 (100%)3 (50%)
Age (years)
 Mean (± s.d.)7·6 (±2·2)9·5 (±2·3)
 Range3–115–11
Race or ethnic group – no. (%)
 White14 (88%)5 (83%)
 Black01 (17%)
 Other2 (13%)0
Weight (kg)
 Mean (± s.d.)32 (±11)35 (±15)
 Range10–5320–59
BMI (kg/m2)
 Mean (± s.d.)18 (±3)18 (±3)
 Range15–2415–23
Diagnosis – no. (%)
 Common variable immunodeficiency2 (13%)3 (50%)
 Congenital hypo- or agammaglobulinaemia11 (69%)3 (50%)
 IgG subclass deficiency2 (13%)0
 Severe combined immunodeficiency1 (6%)0

In Europe/Brazil, the median weekly SCIG dose was close to the intended 100% of the previous weekly equivalent IVIG dose (Table 2). In North America, the children had dose adjustments of 127–142% of their previous IVIG dose at the beginning of the efficacy evaluation period; due to body weight changes during the efficacy evaluation period, the median weekly dose in mg/kg over the complete efficacy evaluation period was somewhat lower (126% of the previous weekly equivalent IVIG doses, Table 2).

Table 2.  Administered SCIG doses and serum IgG trough levels during the efficacy evaluation period.
 European/Brazilian study (n = 15)North American study (n = 5)
  1. Analysed variables were first aggregated to medians for each patient. n: number of patients with available data; SCIG: subcutaneous immunoglobulin; IVIG: intravenous immunoglobulin; Ig: immunoglobulin.

SCIG dose  
 Median (range) (mg/kg/week)89 (61–128)144 (92–326)
 Median % of previous IVIG dose97%126%
IgG trough level  
 Median (range) (g/l)8·5 (6·5–16·6)11·5 (6·2–18·1)
 Median % of IgG trough level during previous IVIG treatment116%113%

Outcomes

Infections.  There were no serious bacterial infections during the efficacy evaluation period. During the wash-in/wash-out period, a 7-year-old boy in Europe/Brazil experienced pneumonia that required hospitalization, but resolved within 10 days.

The annual rate of any infection was 4·7 per patient in Europe/Brazil and 5·6 in North America (Table 3). All but one child experienced at least one infection during the efficacy evaluation period; the most common infections were upper respiratory infections (Table 3).

Table 3.  Type of infections reported.
InfectionNumber of infection episodes
European/Brazilian study (n = 15)North American study (n = 5)
  • *

    Episode per patient-year (mean observation periods were 196 days in the European/Brazilian study and 368 days in the North American study). n: number of patients with available data.

Any infection (annualized rate*)38 (4·7)28 (5·6)
Upper respiratory infection108
Conjunctivitis61
Rhinitis52
Sinusitis15
Bronchitis41
Unspecified infection32
Otitis media23
Pharyngitis32
Gastroenteritis11
Urinary tract infection02
Enteritis10
Herpes simplex10
Laryngitis10
Periodontal abscess01

None of the children in Europe/Brazil and only one child in North America was hospitalized (for 2 days) due to infections during the efficacy evaluation period. The annual rate of treatment with antibiotics per patient was lower in Europe/Brazil (63 days) compared to North America (100 days), while the number of fever episodes was higher in Europe/Brazil (11 episodes, annual rate of 2·1 days with fever) than in North America (one episode, annual rate of 0·6 days with fever). However, fever was defined as a body temperature >38°C in Europe/Brazil and >38·5°C in North America. The number of days per year children missed school or kindergarten due to infections was lower in Europe/Brazil (5 days) compared to North America (11 days). Nine children (60%) in Europe/Brazil did not miss any days at school or kindergarten due to infections.

Serum IgG trough levels and other pharmacokinetic parameters.  In both studies, the serum IgG trough levels during the efficacy evaluation period increased compared to those during the previous IVIG therapy (Table 2). The serum IgG trough levels remained stable throughout the efficacy evaluation period (Fig. 1). Except for one child in North America, with an isolated serum IgG trough level of 4·4 g/l before the first infusion in the efficacy evaluation period, individual serum IgG trough levels exceeded 5 g/l during the efficacy evaluation period of both studies. Furthermore, the mean values of the individual median trough concentrations of IgG subclasses (data not shown) were generally consistent with the physiologic distribution in children [16].

Figure 1.

Mean (±standard deviation) serum immunoglobulin (Ig)G trough levels during the efficacy evaluation period. (a) European/Brazilian study; maximum of 15 patients per time-point. (b) North American study; four patients per time-point; one patient with missing values for weeks 29 and 53 was excluded from this analysis.

In Europe/Brazil, the mean [±standard deviation (s.d.)] serum IgG concentrations measured during one dosing interval in the 12 children included in the pharmacokinetic substudy ranged between 8·8 (±1·2) g/l on day 1 and 8·4 (±1·1) g/l on day 7.

In North America, the four children included in the pharmacokinetic substudy similarly showed a low intrapatient variability of serum IgG levels during a dosing interval (means between 14·8 g/l on day 2 and 14·1 g/l on day 6). The median ratio of the AUC values (normalized to 1 week) during SCIG therapy compared to the previous IVIG therapy was 98% (range: 90–110%). At the time the AUC with SCIG was determined, the children had had at least 24 weeks of SCIG therapy and the median ratio of the weekly SCIG dose compared to the previous IVIG dose was 128% (range: 114–137%).

Adverse events.  Adverse events are shown in Table 4. Except for two events in each study (pneumonia and rash in Europe/Brazil; headache and gastrointestinal disorder in North America), all adverse events were mild or moderate in intensity. Serious adverse events were reported in two children in Europe/Brazil (pneumonia and salmonella enteritis in one child and acute sinusitis in another child) and in one child in North America (gastrointestinal disorder); they were considered to be unrelated to the study drug.

Table 4.  Adverse events.
Adverse eventNumber of patients with adverse events, irrespective of causalityNumber of patients with adverse events possibly or probably related to study drug
European/Brazilian study (n = 16)North American study (n = 6)European/Brazilian study (n = 16)North American study (n = 6)
  1. Table summarizes adverse events occurring in two or more children in either study. In addition, the following possibly or probably related adverse events occurred in one subject each in the North American study: nervousness, rash, syncope and urine abnormality. n: number of patients treated with Vivaglobin.

Injection-site reaction156156
Upper respiratory infection104
Rhinitis92
Gastrointestinal disorder641
Headache643
Conjunctivitis63
Fever631
Infection53
Bronchitis41
Otitis media41
Pharyngitis41
Sinusitis23
Accidental injury3
Laryngitis3
Leucopenia2
Lymphadenopathy2
Asthma2
Nausea22
Skin disorder2

As expected, the most common adverse event related to study drug in both studies was injection-site reaction, which was experienced at least once by 21 of the 22 children. The overall rate of injection-site reactions per infusion, pooled for both studies, was 0·40. Among all injection-site reactions, 91% were mild in intensity, none were severe and the majority persisted only for a short period of time (median duration: 1 day). The incidence of injection-site reactions decreased over time, as shown in Fig. 2.

Figure 2.

Number of children with injection-site reactions. (a) European/Brazilian study; maximum of 16 patients per time-point. (b) North American study; maximum of six patients per time-point.

Systemic adverse events such as headache, fever and chills that are often associated with the high bolus doses of IVIG infusions [10] occurred at very low rates per infusion (rates of 0·033, 0·026 and 0·001, respectively).

No clinically relevant changes in any of the haematology or serum chemistry tests were observed. Furthermore, there were no clinically relevant changes from baseline in vital signs or physical examinations, and no indications for any infection with human immunodeficiency virus 1 or 2, hepatitis A, B or C virus or parvovirus B19.

Discussion

Both studies demonstrated that SCIG therapy with Vivaglobin in children with PID is effective in preventing bacterial infections; infection rates were similar to those reported previously for IVIG therapy [17,18]. None of the children experienced a serious bacterial infection during the efficacy evaluation period, concurring with previous reports for SCIG therapy in children with PID [19–21].

High IgG trough levels with low oscillations between consecutive infusions are known beneficial effects of SCIG therapy in patients with PID [5,9,10], and were confirmed by the stable IgG levels observed in both studies (variability typically <1 g/l during one dosing interval). Thus, SCIG therapy with Vivaglobin provides children with more constant IgG levels than IVIG therapy, as reported previously for the overall populations in these studies [3,7].

While the mean IgG trough levels increased during SCIG therapy in both studies compared to the levels measured during previous IVIG therapy, the children in North America had generally higher trough levels than the children in Europe/Brazil, as was also the case during their previous IVIG therapy. However, because the FDA requested an SCIG dose that provided a similar systemic IgG exposure to the previous IVIG dose, the doses in North America were higher than in Europe/Brazil, where the SCIG dose was equal to the previous IVIG dose. Given that no serious bacterial infections occurred during the efficacy evaluation period in both studies, and that the median increase in IgG trough level compared to the previous IVIG therapy was even slightly higher in Europe/Brazil (+16%) than in North America (+13%), the requested SCIG dose increase in North America does not appear to have been necessary to maintain the efficacy of the IgG replacement therapy. A good safety and efficacy profile at 100% of the previous IVIG dose has also been reported recently for SCIG therapy in children with PID in Sweden [21].

In addition to the high and stable serum IgG trough levels, another advantage of SCIG home therapy is the reduced frequency of hospital visits, which in turn reduces the number of days that children with PID have to miss school. During SCIG treatment with Vivaglobin, children missed school for a mean of only 5 days/year in Europe/Brazil and 11 days/year in North America.

Compared to IVIG therapy, SCIG therapy generally results in a lower frequency of systemic adverse events such as headache, fever and chills [10], which can be intolerable for some patients and may require premedication with corticosteroids [12]. The results observed in both studies confirm that SCIG therapy in children is associated with a low rate of systemic adverse events, with the majority of adverse events being mild in intensity. However, almost all children reported injection-site reactions, which thus have to be expected with SCIG administration [6,7]. Local tissue reactions are known to occur in many patients, especially those who just recently started with SCIG therapy [3], and are usually not perceived to cause any discomfort [22]. Most of the injection-site reactions were short in duration and mild in intensity, and the incidence decreased over time. These observations are in agreement with the favourable safety profile of SCIG therapy reported previously [3,6,7].

In conclusion, the paediatric data from two prospective studies conducted in Europe/Brazil and North America confirm the efficacy, safety and tolerability of SCIG home therapy with Vivaglobin in children with PID requiring immunoglobulin replacement therapy.

Acknowledgements

We are grateful to all the nurse staff members of the included study centres for their devotion in teaching and caring of patients. We are indebted to Stephan M. Borte (Leipzig, Stockholm) for critical reading of the manuscript. We sincerely thank Sylvia Herget for clinical trial management on behalf of CSL Behring. We also thank Cordula Massion and Dirk Spruck for their careful data management and statistical support on behalf of Accovion GmbH, and Christina Wendel for providing medical writing services to CSL Behring on behalf of Trilogy Writing & Consulting GmbH.

The Vivaglobin Study Group: Andreas Böck MD, Vienna, Austria; Beatriz T. Costa-Carvalho MD, São Paolo, Brazil; Donald Stark MD, Vancouver, Canada; Chaim M. Roifman MD, FRCP, Ontario, Canada; Peter Vadas MD, PhD, Toronto, Canada; Michael Borte MD, PhD, Leipzig, Germany; Stefan Haag MD, PhD, Marburg, Germany; Tim Niehues MD, PhD, Düsseldorf, Germany; Sigune Schmidt MD, Freiburg, Germany; Ilka Schulze MD, PhD, Berlin, Germany; Ewa Bernatowska MD, PhD, Warsaw, Poland; Jan Kus MD, PhD, Warsaw, Poland; Oscar Asensio MD, Sabadell, Spain; Dolores Hernández MD, PhD, Valencia, Spain; Nuria Matamoros MD, PhD, Palma de Mallorca, Spain; Ann Gardulf RN, PhD, Stockholm, Sweden; Carl Granert MD, PhD, Stockholm, Sweden; Uwe Nicolay Dipl Math, Stockholm, Sweden; Arthur Althaus MD, Louisville, KY, USA; Pedro C. Avila MD, San Francisco, CA, USA; Melvin Berger MD, PhD, Cleveland, OH, USA; Sudhir Gupta MD, PhD, MACP, Irvine, CA, USA; Harry Hill MD, Salt Lake City, UT, USA; Robert W. Hostoffer DO, South Euclid, OH, USA; Lisa J. Kobrynski MD, FRCP, Atlanta, GA, USA; Robyn Levy MD, Atlanta, GA, USA; Laurie Myers MD, Durham, SC, USA; Hans D. Ochs MD, Seattle, WA, USA; G. Wendell Richmond MD, Oak Brook, IL, USA; Robert L. Roberts MD, PhD, Los Angeles, CA, USA; Ralph Shapiro MD, Plymouth, MN, USA; Suzanne Skoda-Smith MD, Gainesville, FL, USA; Mark Stein MD, North Palm Beach, FL, USA; S. Bobo Tanner MD, Nashville, TN, USA.

Trial registration

This is a secondary analysis of children included in two studies that were completed before 2005 and are therefore not registered at http://www.clinicaltrials.gov.

Disclosure

Drs Borte, Bernatowska, Ochs and Roifman all received research support as investigators in these studies sponsored by CSL Behring. Dr Ochs reports having served as a consultant for CSL Behring and Baxter. Drs Borte, Bernatowska, and Roifman report no additional sources of funding.

Ancillary