Joost Swart

This thesis describes the results of epidemiological, preclinical and clinical research in juvenile idiopathic arthritis (JIA) aiming at the unmet needs in treatment.

In Chapter 1 I described how current treatment in JIA is performed and what issues patients may encounter due to the disease or its treatment. The unmet needs and necessity for tailoring the treatment in JIA are outlined.

PART 1 THERAPEUTIC STRATEGIES IN JIA

In Chapter 2 I reviewed the relevant time points in JIA (individual risk of disease, complications, damage, prediction of response to, and successful withdrawal of therapy) for which biomarkers may represent strong added value and which are already used or currently under study for clinical practice. For non-systemic JIA subtypes Human Leukocyte Antigen-B27, antinuclear-antibodies, rheumatoid factor, erythrocyte sedimentation rate and C-reactive protein are still used for classification, prognosis or active disease. Further immunological studies identified new immune markers (e.g. Myeloid Related Protein [MRP]-8, MRP-14 and S100-A12) and we describe the current status of immunological biomarkers used in diagnosis and treatment of JIA.

In Chapter 3 I described that in our center the worldwide accepted “American College of Rheumatology recommendations on the treatment of JIA” (ACR-CPG) are not strictly followed and that the implementation of the ACR-CPG would increase the current anti-Tumor necrosis factor (anti-TNF) use from 12% to 65% one year after the start of methotrexate (MTX). The decision not to escalate was correct in 70%–75% as shown by MTX response and the implementation of the ACR-CPG would lead to overtreatment. Physicians in our center escalate to anti-TNF in patients with significantly higher physician global assessment, clinical JADAS (cJADAS) and patient Visual Analogue Scale (VAS). The use of (c)JADAS in identifying patients in need of anti-TNF therapy outperforms the ACR-CPG with a much higher sensitivity, specificity and accuracy. The cJADAS threshold for treatment escalation at month 3 and 6 is >5 and >3 for oligoarticular JIA and >7 and >4 for polyarticular course JIA, respectively. The performance of the cJADAS decreases when the patient VAS contribution to the total score was restricted and overall did not improve by adding the erythrocyte sedimentation rate. For the first time it was shown that the cJADAS identifies patients in need of anti-TNF and is a user-friendly tool ready to be used for treat-to-target in JIA.

PART 2 PHARMACOVIGILANCE IN JIA

In Chapter 4 I reviewed the known immunological consequences of biological therapies used in JIA. For every frequently used biological agent their characteristics are clearly specified (molecular target, the isotype, registered indication for JIA, route of administration, half-life, contraindication, very common side effects, expected time of response and average cost in the first year). For every separate agent the adverse events have been calculated as incidence per 100 patient-years for the following categories: serious infections, tuberculosis, malignancies, response to vaccination, new-onset autoimmune diseases and development of anti-drug antibodies. There are large differences in side effects between various agents and there is a clear need for an international and standardized collection of post-marketing surveillance data of biologicals in the vulnerable group of JIA patients.

In Chapter 5 I presented the set up and the first results of the largest international pharmacovigilance JIA registry called Pharmachild. Sharing of data from national and international registries represents the most powerful tool for future analysis of safety and effectiveness of immunosuppressive therapies in JIA. The baseline characteristics of 15,284 patient’s are reported: 8,274 (54.1%) from the Pharmachild registry, 3,990 (26.1%) from the German (BiKeR) and 3,020 (19.8%) from the Swedish registry. Pharmachild patients showed a younger age (median of 5.4 years versus 7.6) at JIA onset and shorter disease duration (5.3 versus 6.1-6.8) when compared to the other registries. The most frequent JIA category was the rheumatoid factor negative polyarthritis (range 24.6-29.9%). MTX (61-84%) and etanercept (24%-61.8%) were the most frequently used synthetic and biologic disease-modifying anti-rheumatic drug (DMARD), respectively. There was a wide variability in glucocorticoid use (16.7-47.6%). Serious adverse events were present in 572 (6.9%) patients in Pharmachild versus 297 (7.4%) in BiKeR. Infection and infestations were the most frequent AE (29.4-30.1%) followed by gastrointestinal disorders (11.5-19.6 %). The most frequent predefined events of specific interest were infections (75.3-90.2%).

In Chapter 6 I analyzed and described the predictive risk factors for moderate, severe and serious infections in patients using drugs for JIA. The data of 7,884 unique JIA patients with 49,708 observation years were available. We excluded 915 patients who had no recordings of any drug use for their JIA. In 6,969 patients (2/3 with ever a biological) with a median follow-up of 5.3 years 9.9% had at least one moderate infection. Polyarticular course (OR 1.3) and systemic JIA (OR 1.7), younger age at diagnosis (OR 2.3) and antinuclear antibody (ANA) positivity (OR 1.6) are all risk factors of infection. Our study is the first observational study showing that MTX increases the risk of infection (OR 5.1) compared to patients only on NSAIDs or i.a. steroids. Biologics also increased the risk of infection (OR 2.7) and even higher (OR 3.7) when combined with MTX. The addition of steroids to both MTX and biologics increased the risk of infection even more significantly (OR 11.9 and 10.5, respectively). An enormous increase in the risk of infection (OR 112) was associated with rituximab with steroids and DMARDs.

In Chapter 7 I analyzed and described 1,585 moderate or worse infections in 895 (10.8%) of 8,274 JIA patients during a median observation of more than 6 years. For adjudication by an infectious expert panel 772 events in 572 patients were eligible of which 335 as serious/very severe/severe non-opportunistic infections and 437 classified as opportunistic infections (OI) by the local pediatric rheumatologist. Of these 772 safety events the experts considered 682 (99.0%) as infections, 603 (88.4%) as common and only 119 (17.4%) as opportunistic. Therefore OI had an incidence rate of 2.4 per 1,000 patient-years. Of the cases in which consensus was reached, the experts considered in 77% the treatment of the adjudicated infection appropriate and in 76% the drug possibly related to the event. Herpes viral infections, respiratory tract infections and EBV were the most frequent infections, while the 119 events of adjudicated opportunistic infections (OI) consisted of many complicated herpes virus infections and mycobacterial infections. Because there was a great gap between local pediatric rheumatologists and the consensus expert opinion of what was considered as OI, we provided an expert panel approved list of definite and probable OI in children with JIA on immunosuppressive drugs. If this OI-list was used as diagnostic test for diagnosing OI (with the expert panel as the gold standard), the sensitivity of the OI-list would be 86% (19/22) and the specificity 98% (117/119). The consensus list on the definition of opportunistic infections in JIA patients makes future studies on this subject easier to compare.

PART 3 THE BURDEN OF COMORBID CONDITIONS

In Chapter 8 I analyzed and described the data of 8,309 patients with a total observation time of 50,767 patient-years. Only chronic diseases with a duration longer than 3 months were considered as comorbidity. We found 4,451 comorbidities in 3,059 patients. Therefore 36.8% of our JIA patients had a comorbidity with uveitis (17.6%), psoriasis (2.9%), macrophage activation syndrome (1.7%), asthma (1.5%) and thyroid disease (1.2%) forming the top 5 most prevalent diseases. Celiac disease, inflammatory bowel disease, depressive disorder and diabetes mellitus were also not uncommon (around 0.5%). Malignancies, demyelinating diseases and interstitial lung disease were not seen more often than in the general population. Having a comorbid condition negatively impacted our patients on pain, well-being, functioning and quality of life.

PART 4 CELLULAR THERAPIES FOR THERAPY-REFRACTORY ARTHRITIS

In Chapter 9 I reviewed autologous haematopoietic stem cell transplantation (HSCT) as the only treatment able to induce long-term, drug-free and symptom-free remission in several refractory autoimmune rheumatic diseases. Over 3,000 HSCT procedures for rheumatic and non-rheumatic severe autoimmune diseases have been performed worldwide. Specific conditioning regimens are currently used to eradicate the autoreactive immunological memory of patients. Although in vivo immune cell depletion with anti-thymocyte globulin or anti-CD52 is the norm for many regimens, ex vivo selection of CD4+ stem cells from the graft is controversial. Following the extensive immune depletion associated with serotherapy and chemotherapy, HSCT effectively resets the immune system by renewing the CD4+ T cell compartment, especially the regulatory T cell population. The risk of transplant-related mortality (TRM) within the first 100 days should be weighed against the risk of disease-related mortality, and the careful selection and screening of patients before transplantation is essential. I discuss the immunological mechanisms of HSCT in various autoimmune diseases and current HSCT regimens. After carefully taking into consideration the risks and benefits of HSCT and alternative therapies, I also discuss the efficacy, complications and proposed indications of this procedure.

In Chapter 10 I presented the results of an experimental arthritis model. I showed that both intraperitoneal (ip) and intraarticular (ia) mesenchymal stromal cell (MSC) injection resulted in a beneficial clinical and histological effect on established proteoglycan induced arthritis (PGIA) in mice. Bioluminescence imaging showed that MSC ip and ia in arthritic mice are largely retained for several weeks in the peritoneal cavity or injected joint respectively, without signs of migration. Following MSC treatment pathogenic PG-specific IgG2a antibodies in serum decreased. The Th2 cytokine Interleukin-4 (IL-4) was only upregulated in PG-stimulated lymphocytes from spleens in ip treated mice and in lymphocytes from draining lymph nodes in ia treated mice. An increase in production of IL-10 was seen with equal distribution. Although interferon-γ (IFN-γ) was also elevated, the IFN-γ/IL-4 ratio in MSC treated mice was opposite to the ratio in (untreated) active PGIA.

In Chapter 11 I reviewed the literature regarding MSC treatment of inflammatory arthritis; containing data of in vitro studies, animal studies and clinical studies. The properties of MSC, presence of MSC in the joint, intra-articular versus intravenous route, autologous versus allogeneic, ideal source of MSC, distribution, transdifferentiation, engraftment, rejection, efficacy and toxicology are all discussed in detail.

In Chapter 12 I described that in a single-center Phase Ib/IIa, open label intervention study 6 JIA patients received 2 million/kg intravenous infusions of allogeneic bone-marrow derived MSC. In case of response but subsequent loss of response, one and maximal two repeated infusions were allowed. The 6 patients had 9.2 years median disease duration and all had failed methotrexate, corticosteroids and median 5 different biologicals. All had still active arthritis and damage. MSC were administered twice in 3 patients. No acute infusion reactions were observed and a lower post-treatment than pre-treatment incidence in AE’s was found. The one systemic JIA patient had again an evolving macrophage activation syndrome, 9 weeks after tocilizumab discontinuation and 7 weeks post-MSC infusion. Eight weeks after one MSC infusion, 4 patients showed less active joints, 5 patients improved in many clinical parameters and inflammatory parameters decreased in 3/4. After 1 year, we found significantly lower active joint counts, improved well-being scores, normalized median ESR- and CRP-levels. Inactive disease was reached by 3 patients at 1 year.

GENERAL DISCUSSION

In Chapter 13 I integrated the knowledge derived from this thesis and explored what can be implemented already in clinics. I discussed the remaining questions and presented what future studies are already planned or yet need to be set up in order to further improve the lives of children with JIA.

LIST OF ABBREVIATIONS

ACPA Anti-Citrullinated Peptide Antibody-Positive
ACR American College Of Rheumatology
ACR-CPG American College Of Rheumatology Clinical Practice Guideline
ADA Anti-Drug Antibodies
ADHD Attention Deficit Hyperactivity Disorder
AE Adverse Events
AJC Active Joint Count
ANA Antinuclear Antibodies
Anti-TNF Anti-Tumour Necrosis Factor
ASCT Autologous Haematopoietic Stem Cell Transplantation
ATG Antithymocyte Globulin
AUC Area Under The Curve
BLI Bioluminescence Imaging
BM-MSC Bone Marrow-Derived MSC
CD Celiac Disease
CFU Colony Forming Units
CHAQ Childhood Health Assessment Questionnaire
CIA Collagen-Induced Arthritis
CIBMTR Center For International Bone Marrow Transplant Registry
cJADAS Clinical Juvenile Arthritis Disease Activity Score
CRF Case Report Form
CRP C-Reactive Protein
DLCO Diffusing Capacity Of The Lung For Carbon Monoxide
DM Diabetes Mellitus
DMARD Disease-Modifying Anti-Rheumatic Drug
EBMT European Society For Blood And Marrow Transplantation
ESI Events Of Specific Interest
ESR Erythrocyte Sedimentation Rate
ETN Etanercept
EULAR European League Against Rheumatism
EULAR-RA EULAR Recommendations For Treatment Of RA
F Female
FDA Food And Drug Administration Of The USA
G-CSF Granulocyte Colony Stimulating Factor
GFP Green Fluorescent Protein
GVHD Graft-Versus-Host-Disease
HLA Human Leukocyte Antigen
HLT High Level Term
HR Hazard Ratio
HSCT Haematopoietic Stem Cell Transplantation
IA Intra-Articular
IBD Inflammatory Bowel Disease
IFN-γ Interferon-γ
IgG Immunoglobulin-G
IL Interleukin
ILAR International League Of Associations For Rheumatology
IQR Interquartile Range
IRR Incidence Rate Ratio
ISCT International Society For Cellular Therapy
IV Intravenous
JADAS Juvenile Arthritis Disease Activity Score
JADI Juvenile Arthritis Damage Index
JAFS Juvenile Arthritis Functionality Scale
JAMAR Juvenile Arthritis Multidimensional Assessment Report
JIA Juvenile Idiopathic Arthritis
LLT Lower Level Term
MAS Macrophage Activation Syndrome
MedDRA Medical Dictionary For Regulatory Activities
MHC Major Histocompatibility Complex
MRI Magnetic Resonance Imaging Scan
MRP Myeloid Related Protein
mRSS Modified Rodnan Skin Score
MS Multiple Sclerosis
MSC Mesenchymal Stromal Cells
MTX Methotrexate
NA Not Applicable
NSAIDs Non-Steroidal Anti-Inflammatory Drugs
OI Opportunistic Infections
OJIA Oligoarticular JIA
OR Odds Ratio
PBMC Peripheral Blood Mononuclear Cells
PBS Phosphate-Buffered Saline
PG Proteoglycan
PGA Physician Global Assessment
PGIA Proteoglycan-Induced Arthritis
PID Primary Immunodeficiency
PJIA Polyarticular Course Juvenile Idiopathic Arthritis
PRINTO Paediatric Rheumatology International Trials Organisation
PRO Patient Reported Outcome
PT Preferred Terms
RA Rheumatoid Arthritis
RCT Randomized Controlled Trial
RF Rheumatoid Factor
RR Relative Risk
SAC Safety Adjudication Committee
SAE Serious Adverse Events
SF Synovial Fluid
SIR Standardized Incidence Ratio
sJIA Systemic JIA
SLE Systemic Lupus Erythematosus
SOC System Organ Class
TB Tuberculosis
TCR T Cell Receptor
TRM Transplant-Related Mortality
ULN Upper Limit Of Normal
VAS Parent/Patient Visual Analogue Scale Of Well Being
VZV Varicella Zoster Virus