Extensive Proof-of-Concept Studies in TNF-Alpha Antagonists might be Responsible for A Delay of Patient Access in Pediatric Rheumatology

in children could be extrapolated from the data in a corresponding adult disease, although, for chimeric or pegylated biologics, PK/PD parameter analysis would need to be investigated before selecting the effective dose in children. Conclusion: Use of this model has potential to enhance the clinical use of pediatric drugs, tailored to age, weight, and physiological development.


Introduction
In order to improve the route to market for approved pediatric therapeutics, the current Pediatric Regulation in the EU and the

Strategy
The review compared the effects of immune-modulatory drugs in adults and children, selected using the following criteria:

a)
Biologics in the same class to treat arthritis b) Clinically tested for the same or a similar indication in children and adults

c)
Subject to a PIP in children and approved for use in adults.
Drugs selected for this review are biologics targeting TNF-α including adalimumab, etanercept, golimumab, and infliximab.

TNF-α Inhibitors Tested in Adults and Children
Etanercept (Enbrel, Pfizer) is a soluble decoy receptor for TNF. It was the first TNF-α inhibitor launched for treatment of RA. The drug was FDA-approved in November 1998, and by the EMA in February 2000. It is approved for the treatment of RA, JIA, psoriatic arthritis, plaque psoriasis and ankylosing spondylitis [1] as the first biologic to treat JIA. Adalimumab (Humira, Abbot [now: AbbVie]) is a monoclonal anti-TNF-α inhibitor. It was the first fully human IgG1 protein to be approved by the FDA in December 2002.
It was approved by the EMA in September 2003. Adalimumab is indicated for the treatment of RA, JIA, psoriatic arthritis, ankylosing spondylitis, Crohn's disease, psoriasis and ulcerative colitis [2]. It was approved for JIA in 2008. Golimumab (Simponi, Janssen Biotech) is a human anti-TNFα IgG1κ monoclonal antibody.
Golimumab was approved in US and Canada as a treatment for RA, psoriatic arthritis, and spondylitis, and is undergoing regulatory review in the EU [3] for these indications. Golimumab missed the primary endpoint in JIA. Infliximab (Remicade, Janssen Biotech) is a chimeric monoclonal antibody directed against TNF-α which induces apoptosis in TNF-α-receptor + cells. Infliximab is only approved for RA. It failed to meet primary endpoint in JIA and therefore has not been approved by the FDA in children for JIA. A waiver for the PIP was agreed in the EU. Infliximab is used off label in JIA as it has not been approved for this indication.

Search Strategy
The search was focused on RA in adults and on JIA, prescribing information, clinical trials websites and the FDA and EMA websites in order to identify relevant study information [4][5][6][7][8][9][10][11][12][13][14][15][16]. Keywords adults. This comparison is quantified using the odds ratio with a 95% confidence interval. Additionally, the response probability adjusted for treatment and time is given with a 95% confidence interval for each group. Calculations were performed with prpc glimmix, SAS 9.4.

Results
A comparative analysis using clinical and pharmacokinetic data was performed, based on data obtained from pivotal studies of biologics for the treatment of inflammation in children vs. adults and evaluated in terms of efficacy, safety and dose used. In total, one or two pivotal pediatric trials, and four to seven pivotal studies in adults for all biologics were identified. All drugs were given as either monotherapy or in combination with methotrexate, and either placebo-controlled or without control. The following section summarizes results (Tables 1-6) obtained from meta-analyses.

Pharmacology
Etanercept is a dimeric soluble form of the p75 TNF receptor that can bind TNF molecules. Etanercept inhibits binding of TNF-α and TNF-beta (lymphotoxin α [LT-α]) to cell surface TNFRs, rendering TNF biologically inactive.
The dose is derived from the outcomes of controlled pivotal studies. Enbrel 25 mg administered twice weekly as a subcutaneous injection, is the recommended dose for optimal therapeutic response in rheumatoid arthritis; alternatively 50 mg administered once weekly was been shown to be safe and effective in a subsequent clinical study. 25 mg administered once weekly gives a slower response and maybe less effective.

Immunology
No patient had persistent elevations in autoantibodies or had signs or symptoms of another autoimmune disease. Two patients positive for non-neutralizing antibody to etanercept.
Non-neutralizing antibodies to etanercept were detected in one patient. Regarding to the levels of antinuclear antibodies and anticardiolipin antibodies, a small number of patients in each group shifted from positive to negative or from negative to positive test results, with no significant differences between the groups.
Less than 3% of etanercept-treated patients were positive intermittently on tests for serum non-neutralizing antibodies against etanercept, and the positive tests were not associated with a decrease in the clinical response or adverse effects.   Table 2b: Pharmacology data of clinical trials with adalimumab in children and adults.

Pharmacology
Blocks interaction of human TNF-α with receptors and modulates biological responses induced or regulated by TNF (changes in the levels of adhesion molecules responsible for leukocyte migration (ELAM-1, VCAM-1, and ICAM-1 with an IC50 of 1-2 X 10-10M). It does not bind or inactivate lymphotoxin (TNF-beta).

PD
-Decrease in levels of acute phase reactants of inflammation (C-reactive protein (CRPJ and erythrocyte sedimentation rate (ESRJ) and serum cytokines (IL-6).
-Decrease in serum levels of matrix metalloproteinases (MMP-1 and MMP-3) that produce tissue remodelling responsible for cartilage destruction. Source: PI -No gender-related pharmacokinetic differences were observed after correction for a patient's body weight.
-Healthy volunteers and patients with rheumatoid arthritis displayed similar adalimumab pharmacokinetics.
-Haematological events: Adverse reactions of the hematologic system, including medically significant cytopenia (e.g. thrombocytopenia, leukopenia) have been infrequently reported with HUMIRA.

AEs and SAEs
Most frequently

CHILDREN (JIA) (4 to 17 years of age) ADULTS (RA)
Pharmacology -Neutralizes the biological activity of TNFα by binding with high affinity to the soluble and transmembrane forms of TNFα and inhibits binding of TNFα with its receptors (it does not neutralize TNFβ (lymphotoxin-α)).
-Biological activities attributed to TNFα include: induction of pro-inflammatory cytokines such as interleukins (IL) 1 and 6, enhancement of leukocyte migration by increasing endothelial layer permeability and expression of adhesion molecules by endothelial cells and leukocytes, activation of neutrophil and eosinophil functional activity, induction of acute phase reactants and other liver proteins, as well as tissue degrading enzymes produced by synoviocytes and/or chondrocytes. -After treatment, patients exhibited decreased levels of serum IL-6 and C-reactive protein (CRP) compared to baseline. Peripheral blood lymphocytes showed no significant decrease in number or in proliferative responses. It is not known if there are differences in clearance or volume of distribution in patients with marked impairment of hepatic or renal function.
-Hematologic Events: Cases of leukopenia, neutropenia, thrombocytopenia, and pancytopenia, some with a fatal outcome, have been reported in patients receiving REMICADE.
-Autoimmunity: Treatment with REMICADE may result in the formation of autoantibodies and, rarely, in the development of a lupus-like syndrome.
C max : maximum serum concentration.   However, there is no statistically significant treatment difference effect in the between-age or study-duration group for the endpoints ACR 50 and ACR 70. We do not present forest plots for infliximab due to numerically instable results.

A single pediatric clinical trial (JIA-I [17]) in 2000 was
identified from the drug prescribing information (

Dosage and Study Duration
Children were dosed for three to four months with 0.4 mg/ kg bw etanercept, and a maximum of 25 mg per dose. Across all studies, adults received 10 or 25 mg etanercept over a period of six or twelve months. Only two trials [17,24] included a placebo in the control arm and etanercept only in the study drug arm. All other trials in adults were performed in combination with methotrexate in experimental and placebo groups (Table 1a).

ACR Response
Assessment of the ACR study data differed between children and adults. The pediatric studies used the ACR30, ACR50 and ACR70 criteria and the adult studies the ACR20, ACR50 and ACR70 criteria.
Thus, only the data for ACR50 and ACR70 could be considered for direct comparison (Table 1a and Figure 1). In addition, the selected time schedule for ACR assessment differed greatly between studies. While ACR50 and ACR70 were evaluated in week 12 or three studies in adults. Meta-analysis showed a treatment effect for etanercept in both, adults and children. However, no effect of age or study duration on the treatment effect could be measured (Table 4 and Figure 1). Thus, the results obtained on drug efficacy and dose showed no difference in adults and children.  [17][18][19][20].
The first column shows names of the covariates in the model. Odds ratios for dose levels are reported with placebo as the reference. Results are shown together with 95% confidence interval. The black dot on each line shows you the odds ratio for each variable.

Meta-Analysis Results of ACR50 AND ACR70
Results of mixed-effects logistic regression for adults and children concerning treatment effects, age group (adults vs children) and study duration (time in weeks) can be viewed in the following tables. numDF, degrees of freedom of term; denDF, degrees of freedom of error term; F, variance ratio; P, error probability; critical value of significance: p<0.05.  (Table 1a).

Pharmacokinetics
The population pharmacokinetic analysis by Yim et al. Two compared adalimumab to placebo, and two were placebocontrolled plus methotrexate. One study compared adalimumab to methotrexate only, as well as to adalimumab plus methotrexate.

Dosage and Study Duration
The studies in children were carried out over 12 to 30 months with 24 mg/m² adalimumab, and a maximum of 20 or 40 mg per dose. Adults received 20, 40 or 80 mg over a period of six, six and a half or 13-24 months. The drug was given subcutaneously in all cases. The PI allows 10, 20 or 40 mg for children, depending on the body weight, and 40 mg is the approved dosage for adults as described in PI (Table 2a).

ACR response
Assessment of ACRs included were ACR30, ACR50, ACR70 and ACR90 for children, and ACR20, ACR50 and ACR70 for adults.
PJIA-II and RA-I, RA-III and RA-IV assessed ACR50 and ACR70 in week 24. However, these studies are not well comparable as their design differs considerably. PJIA-II was a placebo-controlled study, while RA-II and RA-III tested placebo plus methotrexate. RA-IV was also placebo-controlled, but allowed DMARDs during the study, whereas PJIA-II did not. Only studies with adalimumab in combination with methotrexate at week 24 were eligible for ACR50 and ACR70 comparative analyses. In the pediatric study PJIA-II, 83% of patients achieved ACR50, and in the adult RA-I study, 22% reached ACR50 at week 24 with 20 mg maximum dose treatment. In the PJIA-I study, ACR50 was achieved in 64% of the children at the 40 mg maximum dose compared with 37% in the RA-I study, 86% in the RA-III study and 59% in the RA-V study, respectively. 73% of children achieved ACR70 in the PJIA-II study, whereas only 7% and

Adverse Events
The most common event was injection site reactions. The most common AEs leading to discontinuation of adalimumab treatment were clinical flare reaction, rash and pneumonia. The rate of serious infections was 4.6 per 100 patients (Table 2a).

Pharmacokinetics
A higher apparent clearance of adalimumab in the presence of Neutralizing anti-adalimumab antibody (AAA) and lower clearance with increasing age in patients aged 40 to >75 years was observed in population pharmacokinetic analyses in patients with RA. No gender-related pharmacokinetic differences were observed after correction for a patient's body weight. Healthy volunteers and patients with rheumatoid arthritis displayed similar adalimumab pharmacokinetics. C max , T max , bioavailability and elimination values are only available for adults as described in the PI (Table 2b).

Golimumab
Golimumab has been confirmed to be an effective treatment to assess the efficacy and safety of golimumab in pediatric patients aged 2 to <18 years with active JIA with a polyarticular course (at least five joints) despite therapy with methotrexate (10 to 30 mg/m²/week) for at least 6 months [33]. The trial involved 173 patients (87.9% white, 75.7% female; median age 12 years, age 2 to 17 years) with moderately active disease. Nineteen (11%) patients discontinued in part 1 of the trial due to lack of efficacy (n=14), adverse effects (n=4), and withdrawal of consent (n=1).

Dosage and Study Duration
The drug (the usual adult dose for RA of an initial dose of 50 mg subcutaneously once a month or 2 mg per kg iv infusion over 30 minutes at weeks 0 and 4, then every 8 weeks thereafter. It should be given in combination with methotrexate. Corticosteroids, nonbiologic DMARDs, analgesics and/or NSAIDs may be continued during treatment with this drug [33].

ACR Response
During the first phase of the trial, 151 of the remaining 173

Adverse Events
Through week 48, adverse events, serious adverse events, and serious infections were reported in 87.9%, 13.3%, and 2.9% of all randomized patients, respectively. The most frequent serious adverse event was exacerbation of JIR. Death, active tuberculosis, or malignancy did not occur. Golimumab missed the primary endpoint in JIA. The reasons for the similarity in flare rates between the arms is unclear, and further study is needed if the regimen ultimately proves worthy of clinical use [33]. No Meta analysis for Golimumab on adult and pediatric data could be performed, as the data from the study in JIA is not publically available.

Infliximab
Study Description: A multicenter randomized doubleblind placebo-controlled trial of infliximab in 117 children with polyarticular JIA did not find a statistically significant effect of infliximab 3 mg/kg intravenous infusion therapy plus methotrexate on ACR-Pedi responses as compared with placebo at 14 weeks [34].
The open-label extension (OLE, 52-204 weeks) of the study involved 78 patients. However, 34% discontinued infliximab prematurely, mostly by withdrawing consent due to lack of efficacy [35]. Overall, 30% of the children continued the study up to week 204 (Table 3a).
Both trials were placebo-controlled and allowed methotrexate.
They worked with 3, 6 or 10 mg/kg i.v. application of infliximab.
ACR Response: After 14 weeks, following crossover from placebo to infliximab 6 mg/kg, ACR50 and ACR70 responses at week 52 were achieved in 70% and 52% of the children. However, there was no statistically significant difference between the placebo group and the treatment group. Meta analysis supports that study data on infliximab shows no statistically significant treatment effect in children compared to adults. Also, the impact of age and study duration did not play a significant role (Table 6).

Adverse Events:
The pediatric trial demonstrated that infliximab was safe, though the 3 mg/kg group had a less favorable safety profile, with a higher incidence of injection-site reactions and more serious infections. As the efficacy of infliximab in a pivotal study has not revealed a superior effect compared with placebo [34], the FDA did not approve infliximab for JIA, although it is still used in children. It is recommended as backup drug to treat JIA in the guidelines for JIA treatment [38] at the usual pediatric dose for JIA: 10 years or older: 3 mg/kg via iv infusion at weeks 0, 2, and 6, followed by infusions every 8 weeks [39]. Moreover, infliximab is approved for the therapy of refractory Crohn's disease in children over 6 years (Table 3a).

Pharmacokinetics:
The childrens' trial observed formation of antibodies to infliximab, antinuclear antibody or anti-dsDNA antibodies in greater proportion in the 3 mg/kg group [34,35]. This confirmed results from one adult study [37], although other studies could not detect anti-chimeric antibodies, or only below detection limit [36,40,41] (Table 3b).

Discussion
The introduction of PIPs aimed to initiate a formal approval Similar findings were true for the comparison of dosage between adult and children for adalimumab. The PI shows a dose of 10, 20 or 40 mg for children, depending on the body weight; 40 mg is the approved dosage for adults. Thus, the pediatric study results did not lead to any significant differences in dosage that could not have been predicted from the adult studies. ACR response data was also only partially directly comparable due to difference in assessment values and schedules. In addition, no new safety and efficacy data was obtained by these studies. However, no statistically significant treatment difference effect in the beween-age or study-duration group for the endpoints ACR 50 and ACR 70 (Table 5) [44]. However, 88% of these children were also receiving other immunosuppressive drugs, including corticosteroids, azathioprine and methotrexate. Approximately 50% of the malignancies reported were lymphomas, leukemia and melanoma. The FDA and EMA added a boxed warning with regard to the possible increased risk of malignancy, especially lymphomas, in children treated with anti-TNF-α agents. Despite this, a recent summary of worldwide pediatric malignancies in children treated with etanercept did not find an overall increased risk. However, the authors acknowledged that it is difficult to assess the actual risk due to the rarity of malignant events, the underlying higher risk of lymphomas and leukemias in children with JIA and the confounding use of other immunosuppressants [45].
The time before the marketing approval of drugs is particularly important, as the overall aim of drug development in clinical trials should focus on patient benefit to make sure that access of drugs to patients is as simple and fast as possible. However, the studies performed to support marketing approvals in children does not seem to support this overall aim, as shown in the model based on TNF-α blocking agents. Therefore, prolonging the drug approvals process does not benefit children, and promotes off-label pediatric used as these drugs are already marketed for use in adults. All four TNF-α blocking agents discussed here are approved in adults for RA, and all have been tested in children for JIA. The results broadly confirm the findings in adult studies other than infliximab, which has not been approved at a dose of 3 mg/kg for JIA (although it is continued to be used in children). Based on the similarity of dose administered in adults and children, the assumption is that the key parameters are likely to be similar across age groups for a range of biologics. Therefore, the question arises -is it important to carry out confirmatory studies in children? Are these studies really necessary or can the data for biologics be extrapolated if the expression of the respective target is the same in adults and children? The data reviewed suggests that the results obtained in adult RA studies are likely to be useful in predicting the dose, efficacy and safety for children with JIA. It therefore does not support further performance of extensive proof-of-concept studies in children.

Conclusions
The overall aim of drug development in clinical trials should focus on patient benefit, making sure that patient access to drugs is as simple and fast as possible. However, the studies performed to support marketing approvals in children do not seem to support this overall aim, and actually prolong the approvals process. They also promote off-label paediatric use as the drugs are already marketed for use in adults. All four of the TNF-α blocking agents discussed here are approved in adults for RA, and all have been tested in children in JIA. Based on the similarity of dose administered in adults and children for the two biologics approved in children, the assumption is that the key parameters are likely to be similar across age groups for a range of biologics. Therefore, the question arises -is it important to carry out confirmatory studies in children? Are large pivotal studies really necessary or can the data for biologics be extrapolated if the expression of the respective target is the same in adults and children? Our review of the data suggests that the results