To formulate consensus treatment plans (CTPs) for induction therapy of newly diagnosed proliferative lupus nephritis (LN) in juvenile systemic lupus erythematosus (SLE).
To formulate consensus treatment plans (CTPs) for induction therapy of newly diagnosed proliferative lupus nephritis (LN) in juvenile systemic lupus erythematosus (SLE).
A structured consensus formation process was employed by the members of the Childhood Arthritis and Rheumatology Research Alliance after considering the existing medical evidence and current treatment approaches.
After an initial Delphi survey (response rate = 70%), a 2-day consensus conference, and 2 followup Delphi surveys (response rates = 63–79%), consensus was achieved for a limited set of CTPs addressing the induction therapy of proliferative LN. These CTPs were developed for prototypical patients defined by eligibility characteristics, and included immunosuppressive therapy with either mycophenolic acid orally twice per day, or intravenous cyclophosphamide once per month at standardized dosages for 6 months. Additionally, the CTPs describe 3 options for standardized use of glucocorticoids, including a primarily oral, a mixed oral/intravenous, and a primarily intravenous regimen. There was consensus on measures of effectiveness and safety of the CTPs. The CTPs were well accepted by the pediatric rheumatology providers treating children with LN, and up to 300 children per year in North America are expected to be candidates for the treatment with the CTPs.
CTPs for induction therapy of proliferative LN in juvenile SLE based on the available scientific evidence and pediatric rheumatology group experience have been developed. Consistent use of the CTPs may improve the prognosis of proliferative LN, and support the conduct of comparative effectiveness studies aimed at optimizing therapeutic strategies for proliferative LN in juvenile SLE.
Despite advances in the understanding of the disease and research in therapeutic targets, children and adolescents with systemic lupus erythematosus (juvenile SLE) continue to experience frequent episodes of acute disease exacerbations and are at much higher risk than adults for permanent organ damage due to SLE or its treatments (1, 2). This is particularly true for lupus nephritis (LN), which affects up to 80% of children with juvenile SLE. Failure to achieve and maintain remission of juvenile SLE–associated LN (juvenile LN) reduces the overall 10-year survival by an estimated 15% (3).
To determine the severity of juvenile LN, a kidney biopsy is required, and the histologic findings are graded using the International Society of Nephrology/Renal Pathology Society (ISN/RPS) classification (4). The pathologic changes present on the kidney biopsy help guide treatment decisions and may be predictive of long-term kidney survival. Besides the presence of proliferative juvenile LN, defined as histologic findings compatible with ISN/RPS class 3 or 4, continuously active nonremitting juvenile LN and flares of juvenile LN are risk factors for the development of chronic kidney disease and poor juvenile SLE outcomes (5–9).
The treatment of proliferative juvenile LN is commonly divided into 2 distinct phases. The initial phase, induction therapy, is composed of intense immunosuppression aimed at achieving remission of juvenile LN with resolution of active inflammatory changes. Induction therapy is followed by a longer maintenance phase, during which less intense immunosuppressive drug regimens are used to sustain remission of juvenile LN, while attempting to minimize side effects associated with immunosuppressive therapy.
Head-to-head comparisons of treatments for proliferative juvenile LN in sufficiently large trials are lacking in the literature (10). Therefore, no drug to date has received Food and Drug Administration approval for the treatment of juvenile LN. Instead, the treatment of children with juvenile LN relies on the off-label use of medications that are approved for immunosuppression following pediatric kidney transplantation, treatment of solid organ tumors, or drugs that have been found effective in adults with LN. Despite clinical trials of cyclophosphamide (CYC), mycophenolate mofetil (MMF), and azathioprine for treatment of proliferative LN in adults with SLE, the optimal dosing, efficacy, and safety of these immunosuppressive agents when used in children and adolescents with juvenile LN remain to be determined (11–14).
Based on experience in other chronic pediatric diseases, consensus treatment plans (CTPs) represent an alternative approach to determine the preferred treatment in settings where clinical trial data are not available or the conduct of large-scale trials is too difficult due to population size or lack of funding (15, 16). The reduction of clinical practice variability through the use of a limited set of CTPs can allow for future comparison of outcomes and standardization of therapy. CTPs are best developed by organizations that represent the health care providers that most frequently determine the treatments prescribed. Therapeutic decisions for juvenile SLE are generally made by pediatric rheumatologists, and the majority of pediatric rheumatologists in North America are members of the Childhood Arthritis and Rheumatology Research Alliance (CARRA), a research network whose mission is “to prevent, treat, and cure rheumatic diseases in children and adolescents through fostering, facilitating, and conducting high quality research.”
The current project was part of a larger CARRA initiative aimed at the development of CTPs for various pediatric rheumatic diseases. With respect to juvenile SLE, there was close to unanimous support among CARRA members to develop CTPs for proliferative juvenile LN. Therefore, the objectives of this project were to obtain insight into and describe the current treatment practices for juvenile LN among pediatric rheumatologists in North America, and to develop CTPs for induction therapy of proliferative juvenile LN that are consistent with both current practice and the best available evidence in the literature.
Treatment approaches that are based on scientific evidence and clinical experience and endorsed by the majority of pediatric rheumatology professionals in North America are presented.
Induction therapy of proliferative, newly diagnosed lupus nephritis in children includes treatment with glucocorticoids and either oral mycophenolic acid or intravenous cyclophosphamide for 6 months.
Consistent use of consensus treatment plans may improve the prognosis of proliferative lupus nephritis, and support the conduct of comparative effectiveness studies aimed at optimizing therapeutic strategies for proliferative lupus nephritis in juvenile systemic lupus erythematosus.
Fundamental to this project was an in-depth review of the medical literature addressing the current scientific evidence for the treatment of juvenile LN (see Supplementary Appendix A and Supplementary Appendix B, both available in the online version of this article at http://onlinelibrary.wiley.com/journal/10.1002/(ISSN)2151-4658). The literature search included studies of adults with SLE because of the limited published data specifically addressing treatment of juvenile LN. Levels of evidence were assigned based on Oxford Centre for Evidence-Based Medicine criteria (www.cebm.net) and were as follows: A (supported by randomized clinical trials [RCTs]), B (supported by nonrandomized controlled studies or extrapolations from RCTs), C (supported by uncontrolled studies, extrapolations from nonrandomized controlled studies, or marked extrapolations from RCTs [e.g., inference and adaptation of results in pediatrics from a study with adult patients]), and D (based on expert opinion).
Members of CARRA employed consensus formation techniques with the goal of developing CTPs for proliferative juvenile LN (Figure 1) (17). The group had experience in using these techniques when developing outcome measures of several pediatric rheumatologic diseases (18–20). The level of consensus for the Delphi surveys and the consensus conference was set at 80%.
The first phase of the project consisted of an online Delphi survey that was sent to the 103 members of the CARRA SLE Disease Specific Committee. The questionnaire included 20 questions designed to delineate current clinical practice in the management of proliferative juvenile LN. We used a clinical case of a prototypical patient with proliferative juvenile LN to guide survey responses (see Supplementary Appendix C, available in the online version of this article at http://onlinelibrary.wiley.com/journal/10.1002/(ISSN)2151-4658).
The second phase of the project consisted of a formal face-to-face consensus meeting held for 2 days in April 2010. The goal of the consensus conference was to develop a limited set of CTPs for induction therapy of proliferative juvenile LN that was consistent with current practice and informed by the best available evidence (see Phase I above). Participants included 32 voting members of the CARRA SLE Disease Specific Committee who were pediatric rheumatologists with substantial experience in the assessment and treatment of juvenile SLE. Twelve trainees supported the consensus meeting activities, but they were not involved in the nominal group deliberations or the voting process.
We employed a modified nominal group technique using an approach similar to that previously used by CARRA investigators (Figure 2) (18–20). Conference participants were assigned to 1 of 5 subgroups that were led by moderators trained in nominal group technique. Table 1 shows the key questions addressed by these subgroups. Evidence from the literature and summaries of current clinical practice obtained by the initial Delphi survey were referenced during the discussions to ensure consistency of the final agreed-upon CTPs with these practices. All subgroups achieved consensus about the patient population to which the CTPs were to be applied. Each subgroup focused on discussing a specific question pertaining to treatments. Subgroup leaders then reported back to all members of the CARRA SLE Disease Specific Committee who were present at the consensus conference for further discussion and final voting. Complex issues, for which no large-group consensus was achieved during the face-to-face meeting, were addressed in a subsequent Delphi survey (see Phase III below).
|1. What would be the patient characteristics that are appropriate for use of these treatment plans?|
|2. What should be the patient characteristics that would indicate that these treatment plans would not be appropriate to use?|
|3. What should be the dose/route/frequency for use of cyclophosphamide in the treatment plans?|
|4. What should be the dose/route/frequency for use of mycophenolic acid in the treatment plans?|
|5. What should be the dose/route/frequency for use of glucocorticoids in the treatment plans?|
|6. What outcome measures should be used to define response to treatment and how often should they be measured?|
|7. How should quality of life and medication toxicity be measured?|
Through the third Delphi survey, we presented the CTPs formulated by the CARRA SLE Disease Specific Committee to the entire CARRA voting membership (n = 216) to achieve CARRA-wide consensus. We also collected information about the potential impact of applying the CTPs in actual clinical practice.
The key findings of the initial Delphi survey (response rate 71/103 = 70%) included that monthly intravenous (IV) CYC (56/71 = 79%) was the most commonly prescribed immunosuppressive medication for the induction therapy of proliferative juvenile LN, followed by MMF (12/71 = 17%). Rituximab and azathioprine were rarely prescribed for induction therapy of juvenile LN (0–1% of 71). Not surprisingly, the survey revealed striking variability in the dosing, route of administration, tapering schedule, and duration of glucocorticoid therapy among pediatric rheumatology providers, despite its almost universal use for juvenile LN therapy (69/70 = 99%).
The majority of providers (62/71 = 87%) prescribed IV CYC in accordance with the National Institutes of Health protocol (level C) (21). The choice of the immunosuppressive medication was influenced by coexistent morbidities, especially lupus cerebritis (55/68 = 81%), and the perceived risk of nonadherence of the patient to oral medications (58/67 = 87%). Almost all pediatric rheumatologists described themselves as the prescribing physician who selected the immunosuppressive therapy, although patient care was often shared with a pediatric nephrologist (65/66 = 98%) whose focus was typically the management of coexisting proteinuria and/or hypertension.
We achieved consensus on adapting the criteria for flare (60/66 = 91%) and response to therapy (64/67 = 96%) that were developed for LN by the American College of Rheumatology (ACR) and the European League Against Rheumatism (level D) (22, 23).
We achieved consensus on characteristics of the patients for whom the CTPs are suitable (Table 2). As is summarized in Figure 3, either MMF or CYC will be prescribed concomitantly with the chosen glucocorticoid dosing regimen. This will result in 6 potential treatment approaches.
|Fulfill at least 4 of 11 criteria of the American College of Rheumatology classification for SLE or fulfill 3 of the 11 criteria, provided 1 is histologic evidence of lupus nephritis (1)|
|Be ≤16 years of age at SLE diagnosis|
|Be currently ≤20 years of age|
|Have new diagnosis of class III or IV lupus nephritis, with or without class V lesions, as defined by the International Society of Nephrology Classification System/Renal Pathology Society Classification of Lupus Nephritis (4)|
|Patients should not have/be|
|Pregnant or nursing|
|Requirement for intensive care unit support|
The dosing of the glucocorticoids reflected physician experience and was informed by regimens used in previous studies of juvenile LN (level D) (24, 25). Recent evidence suggests that high-dose IV methylprednisolone pulses, but not oral glucocorticoids, have the potential to eliminate the interferon-α gene expression signature in juvenile SLE by reducing the number of plasmacytoid dendritic cells (26). Therefore, all 3 glucocorticoid regimens allowed for the use of up to 3 high-dose methylprednisolone pulses (30 mg/kg/dose up to 1,000 mg/dose) at the time of induction therapy initiation (level C) (27). Despite dramatic variability of glucocorticoid prescribing practices, we achieved consensus to treat juvenile LN with 1 of the 3 glucocorticoid regimens (primarily oral, primarily IV, and mixed oral and IV). The common goal for these 3 glucocorticoid regimens is to achieve a daily dosage of oral glucocorticoids between 10 and 20 mg upon completion of the induction therapy after 24 weeks (or 6 months; level D) (28).
Consensus was reached to administer a total of 6 monthly IV CYC dosages. The initial dose of CYC is 500 mg/m2 body surface area, and subsequent doses are to be increased, but will not exceed a maximum monthly dosage of 1,500 mg (level C) (21, 29, 30). The group recommended adjusting the CYC dose for renal insufficiency and for a low white blood cell nadir, which is anticipated 7–10 days after the infusion of CYC (level C) (21).
Extensive discussion during the consensus conference focused on concomitant therapies to address and avoid CYC side effects, including the use of 2-mercaptoethane sulfonate sodium for the prevention of hemorrhagic cystitis (level C) (31), antiemetics for nausea, gonadotropin antagonist for ovarian protection (level C) (32), and prophylaxis for Pneumocystis jiroveci infections (level D) (33). There was consensus that concomitant therapies will not be included in the CTPs and that use of these concomitant therapies should be left to the discretion of the treating provider.
MMF will be administered at 600 mg/m2/dose twice daily with a maximum dosage of 1,500 mg taken 2 times per day. A lower dose of MMF could be used at initiation of treatment, but the dose should be escalated to the target dose within 4 weeks of starting therapy, thereby allowing for dose titration according to side effects and tolerability (level C) (12, 34, 35).
Mycophenolate sodium (MPA) may be used as an alternative to MMF for the treatment of juvenile LN with a target dosage of 400 mg/m2/dose twice daily and a maximum dosage of 1,080 mg twice daily. Monitoring of complete blood count, liver function tests, and serum creatinine every 2 weeks during the first month after the initiation of MMF/MPA followed by monthly testing during the induction therapy for juvenile LN were considered sufficient for the surveillance of MMF/MPA side effects (level C) (12, 26, 27). MMF/MPA should be held or discontinued for serum creatinine levels that increase more than 30% above baseline or for leukopenia of <1.3 × 109/liter (level C) (12). Rechallenge, dose adjustments, and interruption of MMF/MPA intake for suspected drug toxicity are to be left to the judgment of the treating physician. No consensus was reached on whether therapeutic drug monitoring of MMF/MPA should be utilized (level C) (36, 37), or if random mycophenolate levels should be obtained to screen for patient adherence (level C).
Table 3 summarizes definitions of response to induction therapy and flare of juvenile LN as adapted from adult LN (22, 23). There was agreement to assess effectiveness of treatment at 1 month, 3 months, and 6 months after initiation of induction therapy (level C) (12). No consensus was reached about the timing of or indication for repeat renal biopsy for routine monitoring or assessment of flare (level C) (38).
|Core renal parameters†||Proteinuria (spot urine protein/creatinine ratio) Renal function (creatinine clearance or serum creatinine)‡ Urine sediment (urine WBCs, RBCs, and casts)|
|Renal response definitions†|
|Substantial response (complete remission)||Normalization of renal function, inactive urine sediment (<5 WBCs/hpf, <5 RBCs/hpf, and no casts), plus spot protein/creatinine ratio <0.2 or age appropriate|
|Moderate response||At least 50% improvement in 2 core renal parameters (with maximum spot protein/creatinine ratio ≤1.0) without clinically relevant worsening of the remaining renal core parameter|
|Mild response||30–50% improvement in 2 core renal parameters without clinically relevant worsening of the remaining renal core parameter|
|No response||Patient does not qualify for any improvement as above|
|Renal flare definitions†|
|Proteinuric/nephrotic renal flare||A persistent increase in proteinuria to values >0.5 after achieving complete response, or a doubling of proteinuria with values >1.0, after achieving a partial response|
|Nonproteinuric/nephritic renal flare||Increase or recurrence of active urinary sediment (increased hematuria with or without reappearance of cellular casts), with or without a concomitant increase in proteinuria|
|Other measures of safety, effectiveness, and extrarenal SLE|
|Physician rating of disease activity†||Physician global assessment of disease activity as measured on a 10-cm visual analog scale (0 = inactive, 10 = very active)|
|Patient well-being†||Parent/patient global assessment of well-being as measured on a 10-cm visual analog scale (0 = very poor, 10 = very well)|
|Physician estimated adherence †||10-cm visual analog scale of adherence (0 = nonadherent, 10 = complete adherence)|
|Disease activity†||Systemic Lupus Erythematosus Disease Activity Index (44)|
|Disease damage§||Systemic Lupus International Collaborating Clinics/American College of Rheumatology Damage Index (45)|
|Safety†||Moderate or serious adverse events at grade ≥3 according to the National Cancer Institute Common Terminology Criteria for Adverse Events|
|Quality of life§||Pediatric Quality of Life Inventory Generic Core Scale 4.0 (38)|
|Kidney biopsy||Results of initial and any subsequent renal biopsy pathology reports|
Recommendations were made (based on consensus) to also monitor extrarenal disease activity, using the Systemic Lupus Erythematosus Disease Activity Index, physician global assessment of disease activity, parent/patient global assessment of well-being, and disease damage using the Systemic Lupus International Collaborating Clinics/ACR Damage Index (level D) (39). For patient-reported outcome measures, the Pediatric Quality of Life Inventory Generic Core Scale version 4.0 (40) will be completed by patients at least every 6 months (level C) (41).
The National Cancer Institute's Common Terminology Criteria for Adverse Events version 3.0 (http://ctep.cancer.gov) was selected as the taxonomy of choice for monitoring treatment side effects. Moderate to severe (grade 3 or higher) adverse events (level D) will be recorded.
A second Delphi survey was sent to 103 CARRA SLE Disease Specific Committee members shortly after the consensus conference to address unresolved issues related to steroid dosing, measurement of health-related quality of life, and assessment of adherence (response rate = 79%).
The third Delphi survey was sent to all voting members of CARRA at 92 pediatric rheumatology centers in the US and Canada (response rate 137/216 = 63%). Based on Delphi questionnaire responses, there was consensus among all CARRA voting members regarding the usefulness, compatibility with daily clinical practice, and scientific soundness of the CTPs for the induction therapy for proliferative juvenile LN that had been developed by the CARRA SLE Disease Specific Committee.
Approximately 90% of the respondents stated that they would prescribe CYC (rather than MMF) as per the CTPs to their patients with proliferative juvenile LN. There was also consensus that they would utilize 1 of the 3 glucocorticoid regimens when treating patients with proliferative juvenile LN, with 39% of the survey respondents indicating a preference for the primarily oral regimen, 15% for the primarily IV regimen, and 46% for the mixed IV/oral regimen.
An additional focus of the third Delphi survey was the assessment of the impact of implementing the CTPs developed by the CARRA SLE Disease Specific Committee at pediatric rheumatology centers in North America. Estimates provided by the survey respondents suggest that there are at least 300 patients with juvenile SLE with proliferative juvenile LN receiving treatment at CARRA sites who could be candidates for these CTPs over a 12-month period.
Consensus has been achieved by the majority of pediatric rheumatologists in North America for a limited set of treatment regimens for the induction therapy of newly diagnosed proliferative juvenile LN. These CTPs, which are consistent with the best available scientific evidence, specify the use of either MMF or IV CYC in combination with 1 of 3 standard glucocorticoid regimens. The CTPs reflect common current treatment practices employed by the pediatric rheumatology community and provide agreement regarding definitions of response to therapy, dosing of medications, and patient monitoring.
The results of RCTs are often used as the foundation of treatment standards in medicine in an effort to improve disease prognosis. Unfortunately, as shown in our review of the literature, there is limited high-quality evidence from clinical trials to guide therapeutic decisions in juvenile LN because of challenges inherent to the design and conduct of clinical trials for relatively rare and highly complex pediatric diseases, like juvenile SLE. As an alternative approach, in an effort to improve the outcome of juvenile LN, members of CARRA developed CTPs that allow for the standardization of juvenile LN therapy. The experience from other adult and pediatric diseases suggests that standardization of care using CTPs has the potential to improve disease outcomes (42–44). This is because CTPs promote the provision of therapies using the best available scientific knowledge, and allow for the conduct of comparative effectiveness analyses that can further elucidate the most effective treatment approach.
It is important to emphasize that these CTPs are not meant to serve as treatment guidelines since sufficient evidence regarding the best treatment for juvenile LN is not available. In contrast to treatment guidelines, these CTPs do not address all treatment and management issues for patients with juvenile LN. Further, these CTPs focus on the treatment of prototypical patients with newly diagnosed proliferative juvenile LN and not those with longstanding proliferative juvenile LN. Therefore, these CTPs do not address treatment strategies for patients with intractable extrarenal or multisystemic disease. Nevertheless, we believe that these CTPs are highly relevant because they are applicable to a large proportion of patients with juvenile LN. These CTPs might also be used by other providers, such as adult rheumatologists or pediatric nephrologists, who care for patients with juvenile LN in areas where there is limited access to pediatric rheumatology.
Although we are optimistic about the current applicability of these CTPs for the induction therapy of proliferative juvenile LN, we are cognizant that the formulation of CTPs is a dynamic process. New drugs and biomarkers for juvenile SLE are in development, and the CTPs will need to be modified and updated to incorporate such new medical evidence. Additionally, there is interest in developing consensus around the use of alternate-day steroid dosing and the use of low-dose CYC regimens similar to those tested in the Euro-Lupus Nephritis Trial for adults with LN (13).
The development of CTPs for the induction therapy of proliferative juvenile LN represents a major cornerstone for the systematic evidence-based evaluation of current treatment strategies of children with juvenile SLE. Once widely utilized, these CTPs will allow for the accumulation of data that, when analyzed for comparative effectiveness, will have the potential to pave the way for the identification of therapies that have a high potential for achieving remission of juvenile LN. Complete treatment of these patients requires ongoing maintenance therapy following induction, and therefore the development of CTPs addressing maintenance therapy of proliferative juvenile LN is currently in process.
All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be published. Dr. von Scheven had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study conception and design. Von Scheven, Wallace, Brunner.
Acquisition of data. Mina, von Scheven, Ardoin, Eberhard, Punaro, Ilowite, Hsu, Klein-Gitelman, Moorthy, Muscal, Radhakrishna, Wagner-Weiner, Adams, Blier, Buckley, Chalom, Chédeville, Eichenfield, Fish, Henrickson, Hersh, Hollister, Jones, Jung, Levy, Lopez-Benitez, McCurdy, Miettunen, Quintero-Del Rio, Rothman, Rullo, Ruth, Schanberg, Silverman, Singer, Soep, Syed, Vogler, Yalcindag Yildirim-Toruner, Wallace, Brunner.
Analysis and interpretation of data. Von Scheven, Ardoin, Eberhard, Punaro, Klein-Gitelman, Muscal, Wallace, Brunner.
The authors acknowledge Lena Das, Ofra Goldzweig, Michael Blakley, Amanda Brown, Ioannis Kalampokis, Arzu Soybilgic, Geraldina Lionetti, Annette Lopez-Martinez, Anjalii Patwardhan, Caitlin Sgarlat, Karen Peterson, Anna Huttenlocher, and Jenny Palter.