Clinical trials in youth with type 2 diabetes

Abstract

Type 2 diabetes in youth was almost unheard of only two decades ago. However, tracking the recent dramatic rise in childhood obesity, type 2 diabetes has become increasingly prevalent. Thus, there is an urgent need for high-quality clinical trials to increase in-depth knowledge about pathophysiology, optimal treatment, and prevention. We therefore systematically reviewed published and ongoing clinical trials of type 2 diabetes in children and adolescents. The results demonstrate that (i) few randomized clinical trials have been completed and published in children with type 2 diabetes; (ii) ongoing trials in type 1 diabetes clearly outnumber trials in type 2 diabetes; and (iii) recruitment and enrollment into the latter trials are challenging, however once achieved, drop-out rates are not excessively high. We conclude that type 2 diabetes in youth is an important but difficult new field of clinical research, and we discuss the existing barriers to successful recruitment, conduct, and support of these clinical trials.

The incidence of type 2 diabetes in adolescents and young adults continues to rise at an alarming rate, and much remains to be learned about etiology and pathophysiology in order to develop effective prevention and treatment approaches. Data from the SEARCH For Diabetes in Youth study showed that in the year 2002, 3700 youths were newly diagnosed with type 2 diabetes in the USA, approximately 25% of the annual incidence of type 1 diabetes in children and adolescents (1). Rates of type 2 diabetes were as high as 49.4 per 100 000 among 15- to 19-yr-old adolescent minority populations. In all ethnic groups, females were significantly more affected than males. Although a wide array of treatment options is available for type 2 diabetes in adults, only two medications, insulin and metformin, are approved for use in children and adolescents with this condition.

Clinical trials in children with type 1 diabetes were initiated soon after the discovery of insulin in 1921 (2, 3) and have markedly improved our knowledge about the immunologic and metabolic characteristics of this condition. However, interventional and prospective long-term clinical studies in pediatric type 2 diabetes are sparse. One reason is certainly the low prevalence of this disease until about 20 yr ago. Another issue is the fact that recruitment for these clinical trials remains challenging. An example is the necessity to prolong the predicted recruitment period for the large, multicenter Treatment Options for Type 2 Diabetes in Adolescents and Youth trial (TODAY) from 3 to 4 yr in order to complete enrollment of 750 children and adolescents nationwide (4). Our own experience has been similar: a study at the National Institutes of Health (5) has recruited only 11 subjects in a 2.5-yr period. At Texas Children’s Hospital, nine subjects have been recruited in 2 yr for another study (6), despite having a total population of about 300 children and adolescents with type 2 diabetes and diagnosing approximately 75 youths with new onset type 2 diabetes per year. Motivated by these difficulties, we reviewed the literature to examine the current status of completed and ongoing clinical trials in children and adolescents with type 2 diabetes and explored existing hurdles and obstacles to successful enrollment into these clinical study protocols.

Methods

Literature review

We conducted a systematic review of the literature using PubMed, EMBASE, and Web of Science to identify relevant reports about interventional trials involving children and adolescents with type 2 diabetes. Search criteria consisted of combinations of the key words ‘diabetes’, ‘type 2’, ‘clinical trials’, ‘pediatric’, ‘adolescent’, and ‘youth’. Inclusion criteria required that (i) articles be published in peer-reviewed journals in the English language; (ii) recruited subjects included children (ages 0–18 yr) with physician-diagnosed type 2 diabetes; and (iii) interventional clinical trials had a minimum of two observations or visits.

Abstracted data included method of diagnosis, intervention, number of patients per treatment group, whether randomization had taken place, outcome measures, total number of study visits, study duration, retention rate, and authors’ comments regarding recruitment and/or retention.

Ongoing clinical trials

Registered ongoing and completed clinical trials in children with type 2 diabetes were identified through clinicaltrials.gov. Searches were conducted using the search phrases ‘type 2 diabetes and adolescent’ and ‘type 2 diabetes and pediatric’. The last search was performed on 17 November 2009. For comparison, additional searches for trials in children with type 1 diabetes were conducted using the same phrases, replacing the key words ‘type 2 diabetes’ with ‘type 1 diabetes’. Data were abstracted and categorized according to the patient enrollment criteria, study type (interventional vs. observational), status of study (e.g., completed, currently recruiting, and not yet open for recruitment), and whether results were published. Ongoing interventional trials in children with type 2 diabetes were further categorized according to the type of intervention (e.g., pharmacological vs. behavioral), the specific intervention being implemented, age range of study population, duration of study (e.g., start date and estimated end date), and number of study sites (single vs. multicenter).

Results

Literature review

One hundred and fifty journal articles were retrieved from our literature search. Only five published reports of prospective interventional trials were found [four randomized (7–10) and one non-randomized (11)] (Table 1).

Table 1.

Published interventional studies in adolescents with T2DM

Author, yr	Intervention	N	Study population	Randomized	Outcome measures	Study visits	Duration of study	Retention (%)	Comment on recruitment and/or retention
Jones et al., 2002 (7)	1–2 g metformin Placebo	42 40	All T2DM	Yes (double blind)	Mean change in FPG, HbA1c, weight, BMI, lipid, stimulated C-peptide	2	16 wk	75.7	481 screened; 82 randomized; 6 (14.3%) subjects on metformin and 4 (10.0%) on placebo discontinued during the double-blind period
Sellers and Dean, 2004 (11)	70/30 insulin*	18	All T2DM	No (open label)	Mean change in HbA1c	6	≤16 wk therapy and 12 months follow-up	Not reported	No data available on completion
FDA, 2004†	500 mg metformin 2.5 mg glyburide 1.25/250 mg Met/Gly	54 49 57	All T2DM	Yes (double blind)	Mean change in HbA1c	2	26 wk	Not reported	167 enrolled; 160 ITT
Gottschalk et al. 2007 (8)	1–8 mg glimepiride 1–2 g metformin	132 131	All T2DM	Yes (single blind)	Mean change in HbA1c and fasting SMBG	6	24 wk	80	536 screened; 285 randomized; 263 ITT; 210 patients completed the study
Gellar et al., 2009 (10)	1 high GI meal and 1 low GI meal	11	T2DM (5) IGT (6)	Yes (cross-over)	Mean blood glucose and blood glucose variability	2	4 d	100	Due to difficult recruitment, expansion of inclusion criteria to include IGT
Malloy et al., 2009 (9)	2.5 μg exenatide 5.0 μg exenatide Placebo	13	All T2DM	Yes (cross-over)	Pharmacokinetics and safety profile; glucose, insulin, and glucagon measurements	3	5 wk	92	13 randomized; 12 completed

T2DM, type 2 diabetes mellitus; FPG, fasting plasma glucose; HbA1c, hemoglobin A1c; BMI, body mass index; FDA, Food and Drug Administration; ITT, intention to treat; SMBG, self-monitored blood glucose; GI, glycemic index; IGT, impaired glucose tolerance.

^*Insulin was initiated at a starting dose of 0.5 U/kg/day in two doses. Insulin dose was titrated to maintain blood glucose in the target range of 4–7 mmol/L. Mean maximum insulin dose was 0.76 U/kg/day reached at 2.3 wk.

^†Unpublished data available from the FDA.

The first clinical trial in children and adolescents with type 2 diabetes was published by Jones et al. (7). In this randomized, double-blind, placebo-controlled, multicenter trial, metformin vs. placebo were administered for 16 wk and the effects on blood glucose control were examined. Recruitment for this study involved screening of 481 subjects in order to randomize 82 who met eligibility criteria. Of these 82 participants, only 21 (26%) completed the 16-wk double-blind treatment, although the majority of those who discontinued study drug did so due to hyperglycemia (30 subjects) or early study unblinding (20 subjects). Data from this study served as the basis for Food and Drug Administration (FDA) approval of metformin in the pediatric age range.

The second study identified, published in 2004, was a non-randomized pilot study evaluating short-term insulin therapy in 18 adolescents with type 2 diabetes (11). The intervention consisted of up to 16 wk (mean 8.7 ± 4.3 wk) of insulin therapy and a minimum of 12-month follow-up off insulin. Mean change in hemoglobin A1c was measured at the start and end of insulin therapy and at 3, 6, 9, and 12 months off insulin. No information on recruitment or retention rates was available in this report.

In 2007, Gottschalk et al. (8) published a 24-wk, randomized, single-blind, comparative study evaluating glimepiride vs. metformin as monotherapy. Of 536 patients screened, 241 did not meet inclusion criteria. Of 285 subjects who were randomized to glimepiride or metformin therapy, 210 (80%) completed the study. Both drugs had a similar beneficial effect on hemoglobin A1c.

In 2009, Gellar et al. (10) published a small trial in five youths with type 2 diabetes. Though the intervention was safe and non-invasive (test meals with varying glycemic indices), recruitment was described as difficult and the inclusion criteria were adjusted to include six patients with impaired glucose tolerance (Nansel TR, personal communication). Also in 2009, Malloy et al. (9) published a placebo-controlled, cross-over trial assessing the pharmacokinetics, pharmacodynamics, and tolerability of single doses of exenatide (2.5 and 5.0 μg) vs. placebo. Thirteen patients were enrolled and 12 completed the study.

Review of ongoing clinical trials

One hundred and thirty-two registered studies were identified from the search for clinical trials in children and adolescents with type 2 diabetes in clinicaltrials.gov (Fig. 1). Of these, only 21 (16%) were actually trials involving children or adolescents with type 2 diabetes. Seventy studies involved patients with type 1 diabetes only; 22 involved obese and youth at risk for type 2 diabetes; 3 involved the screening and prevention of type 2 diabetes; 3 evaluated adults with type 2 diabetes; and 13 were judged to be irrelevant to type 2 diabetes in children.

Fig. 1.

ClinicalTrials.gov: Search results for trials in adolescents with type 2 diabetes.

Of the 21 trials in children and adolescents with type 2 diabetes, 16 were interventional and 5 were cross-sectional. Six of 16 interventional trials were classified as completed and 10 were ongoing. Of these 10 ongoing studies, 7 were actively recruiting, 2 were no longer recruiting, and 1 was not yet open for recruitment. Six of these 10 trials had a pharmacological (5, 6, 12–15) and 4 a behavioral intervention (16–19) (Table 2). Of these, one ongoing interventional trial has thus far resulted in the publication of a pilot study (20), but no results of the other clinical trials are presently available.

Table 2.

Ongoing clinical trials of type 2 diabetes in youth

Title	Intervention	Age (yr)	Start date of study	Duration of intervention	Multicenter
Pharmacological intervention
Treatment Options for Type 2 Diabetes in Adolescents and Youth (TODAY)*	Metformin 1000 mg bid Met + rosiglitazone 4 mg bid Met + intensive lifestyle intervention	10–17	May 2004		Yes
Effect of Short-Term Beta-Cell Rest in Adolescents and Young Adults with Type 2 Diabetes Mellitus	Metformin, Diet, and Exercise Insulin and Metformin, Diet, and Exercise	12–25	March 2007	12 Months	No
Safety and Efficacy of Exenatide as Monotherapy and Adjunctive Therapy to Oral Antidiabetic Agents in Adolescents with Type 2 Diabetes	Exenatide (5, 10 μg) Placebo	10–16	May 2008		Yes
A Study to Assess the Pharmacokinetics, Safety and Tolerability of Sitagliptin in Adolescents	Sitagliptin (50, 100, and 200 mg)	10–17	August 2008		Yes
Safety of Liraglutide in Pediatric Patients with Type 2 Diabetes	Liraglutide (0.3–1.8 or 0.6–1.8 mg)	10–45	July 2009		Yes
The Effect of Byetta and Symlin on Post-meal Meal Blood Sugar Levels in Children with Type 2 Diabetes	Exenatide 5 μg Pramlintide 60 μg	12–21	July 2009		No
Behavioral intervention
Situational Problem Solving in Adolescents with Type 2 Diabetes: Enhancing a Randomized Controlled Trial (RCT)	Participation in a comprehensive disease management program	12–19	June 2009		No
Study of Intensive, Home-Based Family Therapy to Improve Illness Management in Youth with Diabetes	60 min home-based, family psychotherapy sessions, two to three times a week for 6 months	10–17	March 2007		No
Personalized Exercise for Adolescents with Diabetes	2-h education session 6-wk personal exercise program	12–19	August 2007		No
Partners for Better Health in Adolescent Type 2 Diabetes: the Buddy Study	Weekly contact with volunteer health worker (buddy) × 6 months vs. conventional follow-up	12–20	October 2009	6 Months	No

^*Ongoing, not recruiting.

Published results were found for five of six completed interventional trials (Table 1) (7–11). For the study without published results, results had been submitted to the FDA (21). The study was a pharmaceutical company-sponsored trial and enrolled 167 patients aged 9–16 yr with type 2 diabetes. It was a randomized, controlled, double-blind, 26-wk trial with three treatment arms: glyburide/metformin (1.25/250 mg), glyburide (2.5 mg), and metformin (500 mg) titrated for glucose control. Results showed that the combination of metformin and glyburide produced an average decrease in hemoglobin A1c of 0.86%, which was not statistically superior to either metformin or glyburide alone.

Clinicaltrials.gov was also searched for registered studies in children and adolescents with type 1 diabetes, and a total of 213 ongoing clinical trials were found (Fig 2). Of these, 118 (55%) were actually studies in children with type 1 diabetes. Fifteen were trials in children with type 2 diabetes, 47 evaluated adults with type 1 diabetes, 6 involved obese and at risk youth, 7 were prevention trials, and 20 were not relevant to type 1 diabetes. Of 118 clinical trials involving children with type 1 diabetes, 96 were interventional and 22 were observational trials.

Fig. 2.

ClinicalTrials.gov: Search results for trials in adolescents with type 1 diabetes.

Ongoing trials in children with type 1 diabetes exceeded the number of studies in type 2 diabetes over fivefold. Of the identified studies in clinicaltrials.gov, there were nearly eight times as many completed and five times as many ongoing interventional trials in type 1 compared to type 2 diabetes.

Discussion

A number of informative reports concerning pediatric patients with type 2 diabetes have been published in the form of case reports or as cross-sectional and epidemiological studies (22–30). However, only 381 patients have been described in five published interventional studies to date. Once these patients were enrolled in clinical trials, retention rates were satisfactory, ranging from 75 to 100% (7–10). Thus, recruitment of adolescents with type 2 diabetes into existing studies appears to be the most challenging predicament.

The challenge begins before study recruitment, with up to 60% of patients dropping out of routine clinical care within 2 yr (31). The demographics of type 2 diabetes in youth may explain why so many patients fail to participate in both clinical and research-oriented medical care: its highest incidence occurs in 15- to 19-yr-old adolescents, disproportionately affecting minority populations. Furthermore, a positive family history of obesity and diabetes may influence the reaction to diagnosis and perceptions of illness.

Adolescents as a group are generally less adherent than younger pediatric and adult cohorts to oral treatment regimens and study visits. This has been documented for adolescent patients enrolled in cancer and HIV trials (32–34). Self-reported reasons for non-compliance included forgetfulness, non-availability of medication (32), jobs and busy schedules (32, 34) as well as less developed concepts of illness, less perceived vulnerability, higher levels of denial, and less cohesive future orientation (33). Many adolescents may also leave home to attend college or live independently while others may have parents who are absent or apathetic to their child’s condition. Patients report that major facilitators to research participation are positive peer and family influences, program incentives (e.g., money and school credit), spending time with friends, commitment, and personal gain (34).

In contrast to individuals with type 1 diabetes, patients with type 2 diabetes frequently have a strong family history of the disease and share many high-risk features. Recently, Pinhas-Hamiel et al. (35) reported that first-degree family members of adolescent patients (aged 11–17 yr) with type 2 diabetes, including healthy siblings, shared many or all the following high-risk features: marked central obesity, poor dietary habits, and a sedentary lifestyle. The strong hereditary component of type 2 diabetes may present a significant barrier to participation in clinical research, particularly if relatives with the disease are not sufficiently concerned about their own disease management and are not supportive in the management of the affected adolescent. Furthermore, unless a high degree of independence has been achieved, lifestyle interventions of diet and exercise often need to target the entire family in order to be effective in the adolescent.

Historically, racial and ethnic minority populations have been underrepresented in clinical research, and individuals with type 2 diabetes, who often belong to minority groups, have lower odds of research participation compared to age-matched individuals with type 1 diabetes (36). Individuals who consent to clinical research tend to be more trusting of the medical system and possess a greater understanding of the research conducted (37). Mistrust of medical research generated by the Tuskegee syphilis experiments endures among some African American communities today and permeates into other racial and ethnic minority groups’ perceptions of clinical research. Furthermore, decisions regarding Hispanic and Native American youth are often made by the family and tribal elders, respectively. An acute awareness of the dynamics of daily life, the distinct subculture to which these groups subscribe (e.g., movies, songs, and advertisements), is thus expected to increase interest and facilitate involvement of minority adolescents in clinical research.

Lower socioeconomic status and lack of a powerful political and medical lobby are additional challenges affecting youth with type 2 diabetes. In contrast, children with type 1 diabetes benefit from support by organizations such as the Juvenile Diabetes Research Foundation (JDRF), which is the largest charitable funder and advocate of type 1 diabetes research in the world (38). Since its founding in 1970, JDRF has awarded over 1.3 billion dollars to research, including more than 156 million dollars in the 2008 fiscal year. No comparable private funding organization exists to promote research in youth with type 2 diabetes.

Finally, several studies in adolescent type 2 diabetes report that a large percentage of screened patients do not meet eligibility criteria. This may be due in part to the high prevalence of obesity-related comorbidities, such as hypertension, hyperlipidemia, non-alcoholic fatty liver disease, menstrual irregularities, and obstructive sleep apnea, which result in failure to meet strict inclusion or exclusion criteria (39, 40).

How might we improve recruitment and retention of adolescent patients in clinical studies? Studies outside the field of pediatric diabetes offer some promising strategies. One group demonstrated that widely dispersed advertisements in community-specific newspapers, using culturally sensitive language and graphics, captured higher response rates than did flyers send to churches and beauty salons (41). Villarruel et al. (34) showed that a combination of incentives (money and school credit for participation), flexible program start times, continued contact with project staff including more frequent reminders, increasing the interactive components of follow-up, and recognizing the importance of potential mobility limitations among adolescents and their families contributed to a reduction in risky sexual behavior in Latino youth. Finally, we suggest that more should be done to find young patients in the places where they are most comfortable – at their local clinics, in their neighborhoods, and at their schools. Rather than asking patients to come to us, perhaps it is time that we, as physicians and investigators, go to them.

In summary, many demographic and disease-specific variables represent barriers to active participation in much-needed clinical research. Ongoing clinical studies, such as the TODAY trial (14, 42), are expected to shed light on important questions, such as whether the natural history and treatment response of youths with type 2 diabetes is different compared to the well-characterized condition of type 2 diabetes in older adults. Specifically, the time course of deterioration of beta-cell function (27, 29) and the regenerative capacity of beta-cells in these young patients needs to be studied (43). Efforts taking into account the specific psychological and socioeconomic situation of affected children and adolescents may lead to improved patient recruitment and the successful conduct of clinical research in the field.