Transforming Performance of Clinical Trials in Pediatric Type 2 Diabetes: A Consortium Model

Abstract

Background: Completing phase 3 trials of new drugs for youth with type 2 diabetes is challenging. The Pediatric Diabetes Consortium (PDC) of U.S. pediatric treatment centers developed a Consortium model to improve the efficiency of successfully completing these trials.

Aims and Innovations: An aim of the PDC model is to utilize the resources of the PDC Coordinating Center and Executive Committee to improve study protocols, centralize interactions with sponsors, and oversee the performance of PDC Clinical Centers. Key features include a Consulting Group to improve protocol design; Master Service Agreements between the Coordinating Center and Clinical Centers covering confidentiality agreements and contract language; negotiation of a standard Site Budget with Contract Research Organizations (CROs)/Sponsors that reflect actual Clinical Center costs; Weekly Conference Calls with CROs/sponsors to track progress of Clinical Center launches, Monthly Oversight Calls with investigators and study Coordinators to track Clinical Center performance, discuss enrollment strategies, and identify emerging problems.

Successes and Challenges: The Consortium model played a key role in the completion of the pivotal trial of liraglutide for treatment of youth with type 2 diabetes. PDC centers also played a pivotal role in exceeding the projected number of randomized subjects needed by two ongoing studies that are nearing completion.

Conclusions: While the Consortium model is still a work in progress, PDC has assisted in the successful launch of new type 2 diabetes studies, and negotiations are in underway for PDC participation in pediatric type 1 diabetes and other diabetes-related studies.

Introduction

Prior to 2019, metformin was the only agent the U.S. Food and Drug Administration approved for use in the treatment of youth with type 2 diabetes based on the results of a pivotal (although small) phase 3 clinical trial that was completed in 1999.¹ Insulin is also approved for use in pediatric type 2 diabetes. However, this approval was not the result of a pediatric study but by extrapolation from the efficacy of insulin in adults with type 2 diabetes and youth with type 1 diabetes.

More recent pediatric clinical trials of new drugs for youth with type 2 diabetes have been nearly impossible to complete due to a variety of problems detailed in a white paper from the NIH Pediatric Diabetes Working Group as well as an American Diabetes Association Consensus Conference.^2,3 These challenges include large numbers of trials competing for a small number of potential participants; poorly designed protocols developed by sponsors and mandated by regulatory agencies; and administrative and financial burdens at clinical centers. As demonstrated in a survey of members of the Pediatric Endocrine Society, investigators in the United States have been reluctant to participate in pediatric type 2 diabetes trials due to availability of only a small number of patients; lack of interest by parents in participating in research; restrictive inclusion criteria; and inadequate reimbursement for the time and effort it takes to launch and carry out these trials.⁴

The Pediatric Diabetes Consortium (PDC) was initially established in 2008 to track clinical outcomes in a large cohort of youth with new-onset type 1 diabetes.⁵ However, the focus of the PDC evolved to the study of children and adolescents with type 2 diabetes over the past 9 years, resulting in the first U.S. Registry of pediatric patients with type 2 diabetes⁶ and a remarkable increase in the number of participating Clinical Centers that are widely distributed across the United States (Fig. 1).

FIG. 1.

Map of the Pediatric Diabetes Consortium Centers in the United States.

Recognizing the challenges associated with participation in pediatric type 2 diabetes clinical trials, the PDC has developed a Consortium model to improve the quality of study protocols, the efficiency of recruiting participants, the engagement of Clinical Centers, and the ability of Sponsors and Contract Research Organizations (CROs) to successfully carry out these trials. The basic elements of this model, early successes, continuing challenges, and recent learnings, are described below.

Why Pediatric Type 2 Diabetes Trials Were Failing

There were many reasons why pivotal phase 3 trials have been failing to enroll, randomize, and retain a sufficient number of participants to determine the efficacy and safety of new drugs for treatment of youth with type 2 diabetes.^2,3

An early obstacle was that regulatory agencies required initial pharmacokinetic and pharmacodynamic dose finding studies to ensure that subsequent phase 3 trials testing doses approved for use in adults did not expose youngsters to toxic levels of these new drugs. There were two main reasons why this requirement was problematic. First, even straight forward, single-dose pharmacokinetic and pharmacodynamic studies that required a relatively small number of subjects took 2–4 years to complete.⁷ Even worse, such studies were unnecessary because pediatric type 2 diabetes is restricted to older children and adolescents aged 10 to <18 years who are overweight or obese. Consequently, drug exposures in all these pediatric pharmacokinetic and pharmacodynamic studies were similar to or lower than drug exposures in adults because the body weight of pediatric participants was the same or more than adults with type 2 diabetes.^2,3 Based on this experience, preliminary pharmacokinetic and pharmacodynamic studies are no longer required.

A current problem is that there are simply too many studies required by the regulatory agencies in a population of patients that is still relatively small. Due to the epidemic of childhood obesity in the United States, America has one of the largest populations of youth with type 2 diabetes. However, as illustrated by patients enrolled in the PDC Registry, more than 90% of patients in the United States come from disadvantaged, low socioeconomic families⁶ and cannot afford to miss work to attend frequent study visits. Potential subjects for these clinical trials often miss scheduled clinic visits or are lost to follow-up,⁸ and many struggle with adherence to taking medications and carrying out other diabetes-related tasks. In addition, many patients are excluded from these studies due to behavioral problems, depression, other psychiatric disorders, and use of atypical antipsychotic medications. Moreover, minority parents are often leery of research.

Eligibility criteria mandated by regulatory authorities have compounded these challenges and made successful recruitment of subjects virtually impossible by excluding 50%–90% of potential participants.⁹ In addition, protocols developed by sponsors and endorsed by regulatory agencies have included unnecessary barriers for participation, such as frequent visits to measure morning fasting glucose levels. Finally, clinical trial budgets offered by CROs/sponsors to the Clinical Centers are often inadequate to cover the costs of carrying out the protocols, as well as the many activities required to launch and close out the study. Indeed, there are many other reasons why academic investigators in the United States were not interested in participating in these studies, especially working alone, as a single center with limited institutional support.⁴

PDC Consortium Model

One of the overarching goals of the PDC is to facilitate the completion of clinical trials that lead to the approval of new drugs for the treatment of youth with type 2 diabetes. To achieve this goal, the PDC targeted the following areas to develop novel approaches for industry-sponsored studies that would benefit study participants, clinical centers, and industry partners.

PDC Consulting Group

To address the issue of poorly designed protocols for the pediatric type 2 diabetes population, the PDC developed a PDC Consulting Group, consisting of experienced pediatric endocrinologists with expertise in planning and conducting clinical trials in children with type 2 diabetes. The goal of the PDC Consulting Group is to have our investigators actively involved early in the development of the study protocols.

Innovative approaches recommended by our consultants include using multi-arm study designs with a single placebo group to decrease sample size requirements; elimination of separate study arms for two different doses of the same drug; and establishment of an extended run-in period for enrolled subjects who fail initial screening due to a modifiable exclusion. As more than 50% of youth with type 2 diabetes have been excluded from previous studies because HbA1c levels were <6.5 or >10.5%,⁹ the extended run-in period has markedly increased the potential number of subjects screened for these trials who ultimately qualify for randomization.

The PDC Consulting Group has also reviewed drafts of study protocols to suggest ways to make them more “kid friendly.” Examples include reducing the number of study visits; eliminating all (or most) fasting visits in the morning; elimination of assessment of secondary outcomes that have limited importance; and maximizing reimbursement to parents and participants for attending study visits. Basing the need for rescue treatment with insulin on predefined and persistent increases in HbA1c, as in the TODAY study,¹⁰ rather than on increases in fasting glucose levels, works much better in this population. In addition, the PDC Consulting Group advises the PDC Executive Committee regarding the likelihood of successful completion of the trial, should the PDC elect to participate in the study.

Centralized support

All the PDC Clinical Centers have made a strong commitment to stick together as a group in participating in pediatric clinical trials rather than struggle separately. A key goal of the PDC Consortium model is to streamline regulatory, budgeting, and administrative processes for both Clinical Centers and CROs/sponsors through the resources of the PDC Coordinating Center.

A key element of this commitment is the completion of a Master Service Agreement between each Clinical Center and the PDC Coordinating Center. This Master Service Agreement includes a confidentiality agreement with respect to all new industry-sponsored studies, which eliminates the great expense of time that is typically needed for separate Confidential Disclosure Agreements between each Clinical Center and CRO/sponsor. Only requiring one Confidential Disclosure Agreement between the sponsor and the PDC Coordinating Center, which flows down to the PDC Clinical Centers, allows for a much more rapid review of study protocols and site budgets by our investigators.

The PDC Executive Committee (consisting of the Steering Committee Chair, Vice Chair, and Directors of the Coordinating Center) also negotiates a standard study budget for all PDC sites with CROs/sponsors that more accurately reflect the real time and effort required by each of the centers to participate in the study. An early survey of PDC Clinical Centers demonstrated that the “Fair Market Value” template budget that was being offered to PDC Clinical Centers grossly underestimated the centers' true costs of launching and carrying out the study protocol.

In adult type 2 diabetes trials, low start-up and closeout payments could be made up by the income derived from enrolling and randomizing a large number of participants. In contrast, the number of subjects randomized into pediatric trials has averaged fewer than five participants per Clinical Center. Consequently, we developed our own list of template budget items based on information from our Clinical Centers, which is being used as the starting point for negotiations for each new study proposal. The PDC Clinical Center budget also includes increased stipends to more adequately cover the costs faced by patients and parents when participating in a clinical trial, including travel, work, and childcare reimbursements. The standard Clinical Center study budget also saves the CRO/sponsor time in negotiating separate budgets with each individual clinical site.

In most of the early trials in which the PDC was engaged, CROs/sponsors underestimated the number of PDC Clinical Centers that were needed to complete the studies in a timely manner. Thus, the PDC provides guidance regarding the feasibility of recruitment at sites, resulting in negotiations with CROs/sponsors to determine the number of PDC Clinical Centers the CROs/sponsors seek to participate in their study. The PDC Coordinating Center commits to assist in identifying appropriate PDC Clinical Centers to meet recruitment targets for each study.

Once all agreements are in place, the Executive Committee works closely with CROs/sponsors through weekly teleconferences to track the progress of each PDC Clinical Center in completing administrative tasks required for participation in each study. It is our experience that PDC Clinical Centers that have lagged behind in launching a trial respond better to encouragement from their peers in the PDC than to e-mails and calls from CROs/sponsors.

An important element of the Consortium model is to instill a feeling of ownership and commitment by investigators and coordinators in studies in which they are participating by improved communications between PDC Clinical Centers, CROs/sponsors, and the PDC Coordinating Center. This is accomplished, in large part, by monthly “Oversight” teleconferences that are modeled after similar conference calls carried out in the TODAY study.¹⁰ It was not an accident that the healthy competition that developed between TODAY centers in enrolling subjects during similar calls contributed to the ability of 15 centers to enroll >1200 adolescents with type 2 diabetes in the study and to randomize a total of 699 patients.

A major purpose of monthly calls is tracking of each Clinical Center's performance in pre-screening, enrollment, randomization, and retention of participants. Best practices, as well as novel approaches in overcoming obstacles to enrollment and follow-up, are also discussed. The calls also provide opportunities for investigators and coordinators to discuss emerging challenges with CROs/sponsors to identify trends which are consequently addressed. Between teleconferences, the Executive Committee serves as a buffer between performance sites and CROs/sponsors to resolve emerging problems.

The relationships developed by regular interactions among PDC investigators and study coordinators have been further enhanced by face-to-face meetings at the annual scientific sessions of the American Diabetes Association and Investigator Meetings organized by the CRO/sponsor. The opportunities that these meetings provide to investigators and coordinators to network with others should not be underestimated. Young investigators, in particular, may benefit from opportunities to establish national reputations for excellence in clinical research to advance up the academic ladder.

Initial Successes

The PDC had little input in the development of the first three studies in which it participated because they were already ongoing. Nevertheless, the strategies employed by the PDC to facilitate the participation of our Clinical Centers have been highly successful: 41 PDC Clinical Centers randomized 97 participants in three recent pediatric type 2 diabetes clinical trials that have now completed enrollment.

Of particular importance, 17 PDC Clinical Centers played a key role in the successful completion of Novo Nordisk's pivotal, phase 3, pediatric trial of liraglutide, the results of which were recently reported in the New England Journal of Medicine.¹¹ As shown in Figure 2, the reduction in HbA1c (primary outcome) and fasting glucose (a secondary outcome) was significantly greater in the liraglutide compared with the placebo group after 26 weeks of treatment. This randomized clinical trial also demonstrated the safety of liraglutide as add-on therapy for youth with type 2 diabetes, which resulted in the Food and Drug Administration approval of the drug in youth with type 2 diabetes.

FIG. 2.

Results of the Ellipse liraglutide versus placebo study: changes from baseline during 52-week trial are shown for the primary (A) and the major secondary endpoint (B). Means are estimated from mixed model of repeated measurements (MMRM) that included treatment, sex, and age group as fixed effects and baseline value as a covariate. Results from a pattern-mixture model (PMM) are shown at weeks 26 and 52. Figure from: Tamborlane et al. (2019). Copyright 2019 Massachusetts Medical Society. Reprinted with permission. https://www.nejm.org/doi/full/10.1056/NEJMoa1903822

It is also noteworthy that recruitment was completed well ahead of schedule in two ongoing studies in which the PDC is participating. In one of these studies, 15 PDC Centers randomized 64% of the 84 subjects randomized and the randomization to screening ratio was better than that among non-PDC sites. The monthly oversight calls for these studies were well received and resulted in sharing of strategies and best practices to recruit participants, optimize clinical trial execution, and improve data quality.

Recent Successes in Implementing the Consortium Model

The PDC has been successful in fully implementing the Consortium model in the two most recent trials in which it is participating. In both studies, members of the PDC Consulting Group recommended some modifications that greatly improved the study protocol. Of these, the ability to enter enrolled subjects into an extended run-in period if they failed initial screening due to a modifiable exclusion increased the pool of potentially eligible subjects.

The PDC Coordinating Center streamlined the clinical trials agreement process by signing single contract agreements with CROs/sponsors for the number of PDC centers that would participate and helped identify appropriate centers for each study. The contracts have included a single budget agreement between the CRO/sponsor and the PDC Coordinating Center, as the prime contractor. The PDC Coordinating Center is, in turn, managing study-related payments to the Clinical Centers, which serve as subcontractors. A separate contract between the Coordinating Center and CROs/sponsors also covered the costs of monthly oversight teleconferences and other support provided to the PDC Clinical Centers.

Learning the Hard Way

We also learned from recent experience that some of the processes that we incorporated into the PDC Consortium model are not always seen as being more efficient by some PDC Clinical Centers and CROs/sponsors. These include issues related to the following:

Master Service Agreements

Many of the PDC's academic institutions already have Master Service Agreements with some industry sponsors that cover testing of new drugs for pediatric patients with type 2 diabetes. Consequently, having a separate Master Service Agreement with the PDC Coordinating Center is redundant and less efficient because wording of the trial-specific Work Order for a particular trial has to be congruent with both the sponsor's and PDC Coordinating Center's Master Service Agreements. The solution has been to allow institutions with pre-existing Master Service Agreements with industry sponsors to negotiate their own Clinical Trial Agreements directly with industry sponsors while still receiving support through the PDC.

Direct clinical trial agreements between CROs/sponsors and PDC Clinical Centers

Some sponsors and some PDC Clinical Centers have preferred having a direct Clinical Trial Agreement with each other rather than using the PDC Coordinating Center as a go between. The main concern here is that the Clinical Trial Agreement clearly defines the potential exposure of both the sponsor and the Clinical Center in case of malpractice litigation due to serious adverse events occurring in study subjects.

Negotiating a single trial budget for each PDC Clinical Center

While the template budget that we have negotiated is based on the high-end costs most often received by our experienced clinical research centers, one size does not fit all the items in every budget for every center. For example, one of our centers was pleased with the per patient and invoiceable costs being offered, but their Grants and Contracts Office would not sign off on the agreement because their lowest indirect cost rate was 5% higher than the indirect cost rate in the PDC negotiated budget. The solution here is to have the Executive Committee develop a template budget with the CRO/sponsor that will be offered to each of the PDC Clinical Centers. The Clinical Centers will be able to negotiate center-specific requirements with CROs/sponsors. In such cases, the PDC Coordinating Center will not manage the study payments.

Immoveable objects

There are at least two requirements which some CROs/sponsors have not been amenable to adapt: (1) CROs/sponsors continue to require on-site qualification visits at each potential PDC Clinical Center, despite being provided with data from PDC surveys regarding the center's personnel, infrastructure, and past research experience. (2) Although sponsors cover the expenses for investigators and coordinators to participate in Investigator Meetings, most sponsors do not reimburse institutions for workdays and clinical revenue that are lost to attend these meetings. While the Sunshine Act precludes them for paying physicians to attend meeting, paying the institution for lost time should be considered.

Considerations from Other Consortium Models

Since most childhood diseases in the United States can be classified as rare,¹² Consortium models offer a practical approach to study and improve care for children. Varying models exist in the pediatric literature, with those described here all successfully completing clinical trials. We have highlighted below key characteristics of the consortia that differ from the PDC structure.

The Children's Oncology Group, a large consortium representing over 200 institutions, abides by a formalized Constitution and Bylaws.¹³ To streamline study approval, other models have utilized an Institutional Review Board Reliance Agreement, allowing the review board at the study principal investigator's institution to be a common, primary review board for the other institutions.¹⁴ Central institutional review boards are also utilized in some Consortium models,¹³ which can mitigate administrative burden on both local review boards and study sites.

While the PDC Consulting Group assesses the feasibility of studies, the Institute for Advanced Clinical Trials for Children (I-ACT) takes this a step further and provides a Feasibility Assessment Report prepared before protocol development.¹⁵ Patient and caregiver engagement in clinical development are key parts of the Cystic Fibrosis Foundation as well as I-ACT and should be considered in consortium design. Furthermore, educational missions are common features of pediatric research consortia.

Next Steps

We strongly believed and now have direct evidence that a consortium of leading pediatric diabetes treatment centers working together as a group and managed by its own Coordinating Center can help transform the treatment of pediatric type 2 diabetes. In addition to contributing to successful completion of pivotal clinical trials of new drugs for youth with type 2 diabetes (Table 1), this approach could also be readily expanded to future pivotal clinical trials of new therapies for other pediatric endocrine disorders such as testing weight loss drugs for childhood obesity and studies of adjunctive treatments for youth with type 1 diabetes.

Table 1.

Pediatric Diabetes Consortium Participation in Pediatric Type 2 Diabetes Studies

Drug	Sponsor	Status	Clinicaltrials.gov identifier
Liraglutide	Novo Nordisk	Completed	NCT01541215
Dapagliflozin	AstraZeneca	Ongoing	NCT02725593
Exenatide extended-release	AstraZeneca	Ongoing	NCT01554618
Alogliptin	Takeda	Ongoing	NCT02856113
Empagliflozin/linagliptin	Boehringer Ingelheim	Ongoing	NCT03429543

We will continue to improve the flexibility of our model to meet varying needs for trial participation by different institutions, learning lessons from other well-established consortia. The PDC has already collaborated with the German/Austrian Diabetes Patienten Verlaufsdokumentation registry to compare data in our two diabetes registries,¹⁶ and a similar international collaboration might be able to improve study designs that better align with distinct challenges in different countries and regions, as well as satisfy regulatory requirements for both European Medicines Agency and the U.S. Food and Drug Administration.

Author Disclosure Statement

G.K. reports consulting for AstraZeneca, Boehringer Ingelheim, and Takeda. K.W. reports research funding from Boehringer Ingelheim and AstraZeneca. W.V.T. reports personal fees from AstraZeneca, Boehringer Ingelheim, Janssen, Novo Nordisk, Sanofi, Takeda, and Medtronic Diabetes and has grant funding from Novo Nordisk, Boehringer Ingelheim, and Takeda. M.V., B.N., J.M., and K.N. report no conflict of interest.

Funding Information

The Pediatric Diabetes Consortium and its activities are supported by the Jaeb Center for Health Research Foundation through unrestricted grants from Novo Nordisk, Boehringer Ingelheim, and Takeda.