Over the last several decades, cancer survival rates have improved significantly. This is particularly true in childhood cancer, in which the average overall survival has improved from an estimated 28% in the 1970s to nearly 80% today.1 These improvements are largely credited to the treatment of children in controlled clinical trials through cooperative groups.2–4 Greater than 90% of children younger than age 15 years who are diagnosed with cancer in the U.S. are treated at National Cancer Institute (NCI)-sponsored institutions,1 and 40% to 70% enroll in controlled clinical trials,2, 4 with similar statistics in parts of Europe.5, 6 For example, it is estimated that as many as 90% of children in the U.K. are treated on clinical trials.7, 8 These numbers far outpace historical and current enrollments of the next age group, adolescent and young adults (AYAs).9, 10
The AYA population has been variably defined internationally but is most commonly classified as patients ages 15 to 39 years of age at the time of cancer diagnosis.11, 12 This is a unique population for many reasons, including the types of cancer diagnoses, their access to care and clinical trials, health insurance coverage, as well as a multitude of other psychosocial factors.13–15 In the U.S., approximately 10% of patients ages 15 to 19 years and 1% to 2% of patients ages 20 to 39 years are enrolled in clinical trials.13 Importantly, this age group has seen little to no progress in cancer outcomes over the past quarter century, in dramatic contrast to children and older adults.13 A dearth of clinical trials in AYA-specific diseases and the lack of enrollment of AYAs on available clinical trials has resulted in a lack of survival progress and limited achievements in the knowledge of late effects. There are many factors contributing to these statistics and this article addresses the major factors limiting clinical trial enrollment of AYAs and ideas for addressing the problem.
Cancer in the AYA Population
The incidence of cancer in AYAs ages 15 to 39 years is 8 times that of cancer in children aged younger than 15 years12; in this population, cancer occurs in an estimated 68,000 people yearly in the U.S.12 Cancer is the leading cause of nonaccidental death among AYAs.13, 16, 17 However, the overall number of AYAs diagnosed with cancer is still relatively small compared with adult cancer because 1.3 million U.S individuals aged older than 40 years are diagnosed with cancer each year.
The types of malignancies in the AYA population differ from both the pediatric population, in which embryonal-type tumors predominate (neuroblastoma, Wilms tumor, hepatoblastoma, and retinoblastoma), and from the adult population, in which epithelial cancers are most prominent (breast, colon, prostate, and lung carcinoma). Instead, a unique mix of lymphomas (Hodgkin and non-Hodgkin lymphoma), germ cell tumors, melanoma, thyroid cancer, and breast cancer comprise the majority of cancers in the AYA population (Fig. 1).13 Certain tumors such as osteosarcoma, Ewing sarcoma, gonadal germ cell tumors, and Hodgkin lymphoma reach their peak incidence in the AYA period. Approximately 15% to 25% of the invasive cancers in patients ages 15 to 39 years are germ cell tumors and this is fairly consistent across the age range (Fig. 2).2 In comparison, sarcomas (bone, joint, and soft tissue) account for approximately 15% of all invasive cancers in patients ages 15 to 19 years but just 3% by those patients ages 30 to 39 years.13, 18, 19 This unique distribution of tumors with diseases that are fairly rare in the pediatric and/or the adult population has contributed to the lack of clinical trial development and the resultant lack of improvement in outcome for these patients.
Role of Clinical Trials in Oncology Progress: An Explanation for AYAs Being Left Behind
From the 1950s to the 1970s, as oncology cooperative groups were developing organized clinical research, AYAs were the most likely patients to survive their cancer. Such data likely contributed to the impression that this was not a group of patients in need of attention because they were doing so well already. Furthermore, medicine has been traditionally divided into pediatrics and internal medicine, with the AYA often falling into a gap in expertise for both groups. Because the population is divided between 2 groups of physicians with distinct medical research agendas, the AYAs have ended up underrepresented in the oncology progress of both groups. The definition and goals of oncology clinical trial research vary between the pediatric and adult oncology worlds.20 This division thwarts collaborative efforts and leaves the AYA patient in a gap that impedes progress.
Phase 1 and 2 trials develop new therapeutic agents
Medical oncologists have focused heavily on phase 1 and 2 clinical trials, as well as randomized phase 3 trials, with the goal of developing new therapies for diseases with poor prognoses such as nonsmall cell lung carcinoma, as well as metastatic breast, prostate, colorectal, and renal cell carcinoma. This research has led to the development of promising new agents such as imatinib in chronic myeloid leukemia,21 a strategy that has been challenging and constrained in pediatric oncology.
Phase 3 trials evaluate novel strategies to improve event-free and overall survival as well as increase knowledge regarding therapy toxicities
In contrast to medical oncology, phase 3 trials predominate in pediatrics, including both randomized and risk-based registration trials, with the underlying goal of improving survival rates and determining standards of care. There has historically been less emphasis on new therapy development partly due to the inherent limitations of research involving minors and the resultant difficulties obtaining new agents. Until recently, trials of novel agents were delayed until completed in adults, because industry had not been interested in pursuing the development of drugs that might only have a pediatric indication.
In diseases in which pediatric patients have achieved excellent outcomes, there has been more emphasis recently on developing trials that investigate the ability to decrease therapy and toxicity while maintaining excellent cure rates, perhaps best exemplified by the efforts of the National Wilms Tumor Study Group (NWTS). A current example of this strategy is observed in rhabdomyosarcoma (RMS), in which patients with low-risk RMS have attained cure rates in excess of 90% and intermediate-risk patients have cure rates greater than 70% through the efforts of the International Rhabdomyosarcoma Study Group (IRSG), but with considerable toxicity including infertility, risk for second malignancy, and neuropathy. The current Children's Oncology Group (COG) trials are investigating whether equivalent survival rates and lower toxicity can be obtained with less than 6 months of therapy in the best-risk group and the reduction of cumulative alkylating agent doses in intermediate-risk patients. Additional studies in intermediate-risk germ cell tumors, Hodgkin lymphoma, mature B-cell non-Hodgkin lymphomas, and acute myeloid leukemia in patients with Down syndrome are investigating decreasing chemotherapy and/or eliminating radiation with the goal of reducing long-term toxicity.
Establish tumor banks/perform biologic studies
In addition to prospectively studying therapeutic strategies in a controlled fashion, clinical trials are the most common mechanism by which tumor samples are collected and banked. The development of national tumor banks to provide specimens for basic and translational research has been critical for oncology progress. Nevertheless, AYA samples are underrepresented in national tumor banks, with only 57% of the expected number of samples actually present according to Cooperative Human Tissue Network (CHTN) data.13 This limits the ability of researchers to ask important questions such as why we have different outcomes for patients of different ages with the ‘same’ histologic diagnosis. Is there a basic biologic difference between a soft-tissue sarcoma in a child versus a young adult that explains the prognostic significance of age? Perhaps a change in the host biology or hormonal milieu plays a role? The lack of annotated tumor specimens obtained via clinical trials thwarts our efforts to answer such important questions.
Barriers in AYA Clinical Trials
Sarcomas as an example of lack of clinical trials resulting in lack of progress
Sarcomas are rare tumors that have suffered from the lack of available clinical trials and lack of improved outcomes. Overall, survival for patients with most sarcomas remains poor. Children with sarcomas have not benefited from the same improvements in survival in the last quarter century as children with other diagnoses and AYAs with sarcomas have had less improvement than children.19, 22 Because of the dramatic improvements observed with the introduction of adjuvant chemotherapy more than 20 years ago, the standard of care and survival rates for the 2 major bone tumors, osteosarcoma and Ewing sarcoma, have changed very little. Outcomes are particularly dismal (a 5-year survival rate of less than 30%) for patients with metastatic disease at the time of diagnosis.23, 24 Overall, to our knowledge there have been very few pediatric or adult cooperative group therapeutic clinical trials for sarcomas, with the exception of IRSG efforts in RMS. Those trials that have occurred have enrolled very few AYAs and these patients have seen the least improvement in survival; a review of NCI Surveillance, Epidemiology, and End Results (SEER) data found an age-dependent correlation between the average annual percent change in 5-year survival and clinical trial accrual patterns for both soft-tissue and bone sarcomas, suggesting that the low trial accrual may contribute to the stagnant survival trends in AYAs with sarcomas.22
RMS is the most common soft-tissue sarcoma diagnosed among children and adolescents1 but is uncommon in adults. Unlike other sarcomas prevalent in the AYA population, RMS has experienced significant improvement in survival rates over the last 25 years.25 Serial phase 3 prospective randomized clinical trials by pediatric cooperative groups have investigated RMS therapies. Although to our knowledge none of the experimental arms of these clinical trials have demonstrated superiority, survival results continue to improve when compared with historical rates. This suggests that the experience gained from systematic protocol-based therapy, rather than novel therapeutics, have contributed to this increase in survival.
However, RMS in adults has not had the same improvement in survival as younger children; in fact, the 5-year survival for patients ages 30 to 44 years with RMS have appeared to decrease from the period between 1975 and 1980 and the period between 1993 and 1998.13, 22 Increasing age is an adverse prognostic factor in RMS.26 However, RMS comprises several distinct histologies with variable prognoses. For example, alveolar RMS is more common in AYAs and carries a worse prognosis.1, 27 Therefore, age may be a surrogate for biology and account for the worse survival seen in adults. In 1 retrospective analysis of a single institution's experience, the survival of adults (stratified by histology and treated according to pediatric protocols) was similar to pediatric results.28 Similar to several other tumors that primarily occur in childhood, including Wilms tumor, to our knowledge, there have been no prospective clinical trials of RMS in adults.
In contrast to RMS, only 3 prospective trials—enrolling fewer than 200 patients—have ever been completed in children and adolescents with nonrhabdomyosarcoma soft-tissue sarcomas (NRSTS) despite an estimated 500 new cases diagnosed in people aged younger than 20 years old each year in the U.S..29–31 There have been more adult trials for NRSTS but they have usually included several histologies (variably common in AYAs) and even bone sarcomas. Although younger age is often noted to be a better prognostic factor,32 to our knowledge no studies to date have been powered to stratify by age. Taken together, this has led to limited knowledge regarding the biology and outcomes for AYA patients with these tumors.
Germ cell tumors as an example of age limiting access to available trials
Germ cell tumors (GCT) are another common AYA diagnosis, and along with non-Kaposiform soft-tissue sarcomas, were determined to be in the top tier of tumors in need of research focus to impact survival statistics by the NCI-Lance Armstrong Foundation Adolescent Young Adult Oncology Progress Review Group. This is despite the finding that there have been quite a few adult GCT trials, in contrast to sarcomas. GCTs have a bimodal age distribution, with the first peak in young children (predominantly sacrococcygeal tumors) and a second peak during adolescence and young adulthood (in which gonadal tumors predominate). Testicular GCTs are the most common solid tumor of young men. The incidence varies greatly around the world, with testicular GCTs accounting for up to 45% of all malignancies in males ages 15 to 29 years in Norway.33 Although ovarian germ cell tumors are less common, they account for up to 9% of all malignancies in the AYA age range. Unlike sarcomas, GCTs are among the most curable cancers.
The successful therapy given today for both adult and pediatric GCT is based on historical prospective adult clinical trials of gonadal GCTs.34 Cisplatin-based regimens developed in the 1970s dramatically improved the outcome for most patients with GCTs, but patients classified as having poor prognostic clinical features by International Germ Cell Consensus Classification (IGCCC) continue to only experience 50% to 60% survival.35, 36 This high-risk subset has had a fairly flat survival curve since the 1970s37 and, despite a number of trials, we have not successfully improved the outcomes for these patients.38–40 Dose-intensity strategies tested in adult trials have been unsuccessful in increasing survival, perhaps limited by increased toxicity (particularly gastrointestinal toxicity).35 However, such trials have included much older adults (the study by Motzer et al.35 had an age range of 17–60 years and a median age of 28 years) and pediatric trials have suggested that dose intensity might improve the outcome for high-risk pediatric GCT patients,41, 42 including patients with mediastinal tumors. It is therefore unclear whether young adults would tolerate dose intensity and gain survival benefits like children or that, regardless of tolerance, their biology is similar to that of older adults and dose intensity approaches are inappropriate.
Another reason for the limited knowledge regarding how tumors in teens respond and what their bodies can tolerate pertains to trial eligibility and referral patterns. Specifically, several ongoing adult studies include patients aged 16 years and older but the number of adolescents (ages 16–19 years) in those trials appears limited due to the discord between referral patterns and sites in which the trials are open. Patients with testicular and extragonadal GCTs who are older than 14 years are excluded from pediatric trials because their tumors are biologically similar to those of young adults, not younger children (Table 1). Although this acknowledges the biologic variability across the age spectrum, the use of an age limit is somewhat artificial because the intent is to separate biology, not age; this can serve as a barrier to clinical trial access. This excludes 15-year-old male patients from all ongoing clinical trials (albeit a small percent of the total GCT population). Males older than 14 years who are referred to pediatric institutions, in which the adult clinical trials will not be readily available, will not enroll in clinical trials. Interestingly, girls up to age 21 years who have an ovarian GCT are eligible for pediatric trials.
Table 1. Current Germ Cell Tumor Trials
GCT clinical trials by cooperative group
GCT indicates germ cell tumor; EORTC, European Organization for Research and Treatment of Cancer; FRE-FNCLCC-GETUG, Federation Nationale des Centres de Lutte Contre le Cancer; MRC-TE23, UK Medical Research Council; MSKCC, Memorial Sloan-Kettering Cancer Center.
Nontesticular tumors, <22 y
For gonadal and extragonadal GCTs
Testicular tumors, <15 y
For male gonadal and extragonadal GCTs
For gonadal and extragonadal GCTs
For male nonseminomatous GCTs
For gonadal and extragonadal GCTs
The equivalent barrier applies on the other end: some AYAs will be referred to adult institutions in which pediatric trials are unavailable. Between ages 15 and 18 years, many adolescents transition from pediatric to adult medical management.3, 11, 43–45 Some pediatric centers have upper age limits that restrict access to adults over certain ages. Unlike pediatric oncology, which is almost exclusively practiced in academic settings, the medical oncology community in the U.S. is weighted heavily in community practice, with only a minority of medical oncologists practicing in a tertiary NCI center. As a result, the majority of AYAs in the U.S. are treated by community oncologists, with limited numbers receiving care in a pediatric or adult academic center in which clinical trials are available.3, 10, 11 This leads to very few AYA patients having the opportunity for clinical trial enrollment even if a trial is available to them.
Insurance issues may impede clinical trial enrollment
Further complicating issues with access to clinical trials, and more broadly access to care, is the finding that AYAs are the largest and fastest-growing uninsured or underinsured population in the U.S.46 Young adults between the ages of 18 and 24 years are the least likely of any age group to have health insurance. Greater than 30% of this group is uninsured, the highest percent of any age group.47 Young adults are also the fastest-growing age group among the uninsured, accounting for 40% of the increase in the uninsured under age 65 years.46 Many of these patients lose health coverage under their parents' policies or public programs at age 19 years, or when they graduate from high school or college, and do not have a secure job that supplies insurance or the assets to buy private insurance. Although to our knowledge there is no published data to support a direct link between lack of insurance and lack of clinical trial enrollments, many experts believe there is a correlation.
Solutions to Overcome the Limitations of AYA Oncology
Increase the number of trials with emphasis on obtaining biologic samples and prospective data
The perspective that AYAs with cancer have good survival rates and that other age groups would benefit more from clinical research may have led to lack of clinical trials and thus lack of progress. But although breast cancer and pediatric acute lymphoblastic leukemia have overall cure rates greater than 80%, research and clinical trials continue because of the acknowledgment that there is still room for improvement and knowledge. Many AYA cancers, including sarcomas and advanced GCTs, continue to have inferior outcomes and the lack of prospective trials in this age group has certainly contributed to our limited knowledge regarding the biology and/or outcomes of such diseases. In addition, there is room to improve the quality of that cure such that the large number of cancer survivors in our society will experience fewer long-term effects resulting from their disease and/or its therapy. Clinical trials need to be developed to prospectively study treatment for these diseases with emphasis on new agents, the evaluation of toxicity, and links to the biology of each disease.
Change age eligibility limits
The Children's Oncology Group (COG) has worked to take away limitations to clinical trial access for AYAs by increasing the upper age limits for many of their clinical trials; there are currently almost 30 COG trials with upper age limits between 25 and 50 years. However, the challenge does not end there. These older patients will only be enrolled if their treating physicians are aware of these options and the patient can be treated at an institution that has the trial open or is willing to travel to such an institution.
Increase collaboration between cooperative groups
Collaboration between pediatric and adult medical oncologists, as well as gynecologic, surgical, and radiation oncologists, will improve the ability to accrue as many patients as possible and promote progress in AYA oncology research. One such collaboration would result in intergroup trials without age limits. However, the challenges continue. Which group should take the lead with any given trial? Do we develop trials based on histology with the group that has the greatest experience running the trial? In certain tumors the pediatric and adult oncologists may use different staging, grading, or classification systems and must reach consensus before opening a joint trial. Finally, implementation of a common protocol does not assure common implementation or outcomes. On the European protocols CESS81-EICESS 92, patients ages 15 to 19 years with Ewing sarcoma who were treated at pediatric institutions fared better than those on the same protocol at adult institutions.48 There is also a requirement to open the trials at as many treating facilities/cooperative groups as possible to allow for increased capture of patients.
In a time of limited resources, this can be a major challenge. Grant funding for oncology research has become increasingly scarce. Possible ways to approach this situation would be to encourage collaboration between cooperative groups to minimize competing protocols and maximize opportunities to enroll as many patients as possible. The Intergroup Coalition Against Sarcomas (ICAS) was such an endeavor, with opportunity for collaboration across all cooperative groups interested in studying sarcomas. This group developed several histology-specific protocols and enrolled patients across groups. Before meeting study accrual goals, the Southwest Oncology Group (SWOG) opened the COG Ewing sarcoma study AEWS0031 and enrolled several patients, proving that it is possible to cross the traditional age barrier between pediatrics and internal medicine. Unfortunately, SWOG budget cuts resulted in termination of ICAS research, another setback in the attempt to develop strategies leading to progress for AYAs.
Another example of an effort to cross traditional barriers is the ongoing osteosarcoma study, EURAMOS1, a collaborative trial between cooperative groups in North America, Australia, New Zealand, and Europe, investigating resectable osteosarcoma. Such a huge international collaboration in pediatric oncology opens up possibilities to learn more regarding the biology and response to therapy of this challenging disease.49 If successful, it could serve as a model for other intergroup collaborations.
Advocate for education initiatives, insurance coverage, and research funding
Outside of working to increase the number and quality of clinical trials focusing on the AYA population, there are several significant logistical and psychosocial factors that will also need to be addressed if we are to optimize enrollment in clinical trials for this group. These include poor insurance coverage, the use of community rather than academic medical care, and potential problems with medical adherence that may be aggravated by the lack of a primary caregiver to help with transportation, bills, and emotional support. All the parties involved—patient, family, community primary care providers, and oncologists—need to be convinced of the value of the trial for these patients to overcome such challenges.
Education of the general public as well as integration of increased medical school education
Heightening awareness of AYA oncology is needed. Collaboration between community oncologists and academic cancer centers is necessary if we are to successfully increase the number of tumor samples obtained for important biologic research as well as for accrual to clinical therapeutic trials. Finally, the oncology community will have to advocate for improved funding for vital projects to government agencies. An example of such an effort is the COG “Reach The Day” effort in June 2007, which had as its goal to have representatives from every COG institution in the U.S. travel to Washington DC to meet with congressional members and march for heightened awareness of cancer research needs.
The AYA cancer population has suffered from a lack of progress in outcomes for decades due to the lack of recognition of need, the lack of access to and availability of prospective clinical trials, the lack of biologic samples for basic and translational research, and the lack of psychosocial support for patients—all of which were hampered by the fragmentation of care between pediatrics and internal medicine. The international oncology community needs to acknowledge this underserved population and work together to bring about positive change (Table 2). This will involve a concerted effort to obtain tissue samples for researchers, to eliminate barriers to clinical trial enrollment, and to educate the medical community as well as the general public concerning the importance of clinical trials to erase this deficiency in our field.
Table 2. Limitations to and Opportunities for Clinical Trials in the AYA Population
Limitations to AYA clinical trials
Opportunities for progress
AYA indicates adolescent and young adult.
• Separation of pediatric and adult medicine
• Collaboration between cooperative groups
• Age barriers
• Removal of age limits
• Rare diagnoses
• Histology based eligibility
• Lack of biologic samples
• Develop AYA specific trials
• Limited number of trials
• Encourage tumor banking/ increase national tumor bank access
• Poor grant funding
• Advocacy for an underserved population
• Lack of insurance coverage
• Redefine referral patterns
• Referral patterns not reflective of expertise or trial availability
• Increase education of the medical community about AYA characteristics and needs