In 2018, several of us reported a “survival cliff’ in the United States among older adolescent and young adult (AYA) patients with acute lymphoblastic leukemia (ALL) and lymphoblastic lymphoma (LBL) diagnosed during 2000–2004.¹ As shown in Figure 1, the cliff was an abrupt decrement between ages 16 and 19 in the 5-year survival rate of ALL and LBL patients diagnosed during 2000–2004. In just 3 years, the survival rate decreased from 73% to 52%, an absolute 21% and at an average percent change (APC) of −10.6.¹ In older patients, the survival rate declined at an APC of 1.7. Based on survival trends in ≤15 and 20- to 85-year-olds, AYAs 16 to 35 years of age during 2000–2004 had an estimated excess of total of 26,600 years of life lost (Fig. 1; red shading).

The survival cliff occurred simultaneously with lower accrual to clinical treatment trials registered with the U.S. National Cancer Institute (NCI) Cancer Therapy and Evaluation Program (CTEP) (Fig. 1). Joinpoint regression identifies an absolute 33% decrease accruals during similar years of the survival cliff, from 38% of 17-year-olds to 5% of 22-year-olds, an absolute percent decrease of 33% and APC of −32.6 (Fig. 1).¹ We concluded that the survival cliff was due in part to a sharp decline in clinical trial availability for, and participation by, AYAs. Others have since evaluated the problem, with the most recent publications emphasizing continued urgent need to solve it.^2,3 Six years after our report, we hoped that the cliff would be less steep and that AYAs with ALL and LBL in the United States would have a better survival. Herein, we report evidence for not only a favorable trend but also disappearance of the cliff and offer evidence suggesting why such has occurred.

Survival data up to the age of 84 for patients diagnosed annually during 2000–2019 were obtained with Surveillance Research Program, NCI SEER*Stat software version 5.0.1,⁴ from SEER 22 Registries that represent 48%−50% of the U.S. population. When Illinois and Massachusetts are excluded from SEER22 (designated SEER22⁻) in the Research Plus database, this represents 42%–44% of the U.S. population. The prior study¹ was based on SEER8 registries that represented 29% of the U.S. population. Joinpoint Regression Program v.5.0.1⁵ was used to identify trend changes and APC values, using weighted least squares, logarithmic transformation, and when standard errors were not available, constant variance. Average annual percent change (AAPC) was designated by the Joinpoint regression program. The Site and Morphology code definitions for ALL and LBL were AYA site recode 2020 Revision classifications 1.1 and 2.1.1., respectively.⁴

Accrual data up to the age of 84 for patients enrolled during 2000–2015 on NCI-sponsored Phase I, II, and III cancer treatment trials conducted by the NCI cooperative groups and NCI-designated cancer centers were provided by CTEP were provided in 2016 by CTEP, NCI.¹ The numerator of the accrual proportion was the number of patients entered onto the clinical trials selected for evaluation from the CTEP-provided data. The denominator was based on the SEER22 incidence of patients in the cancer and age category and population census data obtained from the U.S. government census website.⁶ No individual patients or individual clinical trials were identified, contacted, or reviewed. Institutional Review Board review was not required.

Among all ages in the United States, AYAs diagnosed with cancer have been reported to have had the least clinical trial participation and resultant least survival prolongation and mortality reduction.^1,7–9 During the last two decades, the survival “cliff’ in young American AYAs with ALL and LBL has decreased, overall and in all major races/ethnicities. Based on survival trends in younger and older patients, American AYAs diagnosed during 2000–2004 with ALL or LBL when 16–35 years of age can be estimated to have had an excess of 26,600 years of life lost (Fig. 1 red shading). Overall, in terms of the total U.S. benefit, the survival profiles in Figure 2 can be interpreted to indicate that approximately 26,600 lives of AYAs, 15 to 39 years of age with ALL or LBL were spared during 2003–2017 by the improved survival.

The superimposed “cliff’ patterns (Fig. 1) strongly suggest that the survival cliff was due in large part to the accrual cliff, and the correlation of survival decrement associated with the cliffs with accrual (Fig. 2) support a causal relationship, as previously implicated.¹ This progress is particularly noteworthy since the greatest effort to increase clinical trial accruals in AYAs was directed at ALL, the most common pediatric cancer. New clinical trials in ALL specifically designed for AYAs were launched,^10–12 particularly those that applied pediatric type of chemotherapy regimens had outstanding success in improving survival and cure rates in the 85%–95% range for children under 10 years of age (Fig. 1). Exemplary are those conducted by the Cancer and Acute Leukemia Group B and The Alliance,^13–15 Children’s Oncology Group (COG),^14,16,17,18 Dana–Farber Cancer Institute,^19,20 Nordic Pediatric Hematology Oncology and Baltic Centers,²¹ and Berlin–Frankfurt–Munster and Gruppo Italiano Malattie EMatologiche dell’Adulto.²² Another explanation for the improvement in younger AYAs is the cooperation between COG and the adult cooperative groups that study the overlapping AYA age range. Several of the adult cooperative groups created their own AYA committees focused on studying cancers prevalent in the AYAs. Also, the National Comprehensive Cancer Network released practice guidelines for ALL,²³ and an increasing number of presentations and publications on the topic occurred at national meetings and appeared in the peer-reviewed medical literature.¹ An effort in Ontario, Canada, which is to specifically study the 15- to 21-year age interval for determinants of outcome, including access to clinical trials, has also helped²⁴. The extra effort that had been expended successfully on increasing the clinical trial participation of adolescents with cancer^25–32 has apparently been successful in extending the trend to older AYAs. In 2014, the NCI organized the National Clinical Trial Network (NCTN) and National Community Oncology Research Program (NCORP) that include missives representing the AYA interests of AYAs and the pediatric and adult cooperative groups.³³ The timing of NCTN does not, however, explain the cliff erosion during 2002–2014 but has to some extent at least since. As of January 2024, the NCTN and NCORP had 18 cross-network treatment trials, 12 led by the COG and 6 by the adult NCTN.³⁴

Another possible contributing factor in the United States is the Patient Protection and Affordable Care Act (ACA) that was enacted nationally in 2010 and required insurance companies to allow 18 to 25 year olds to continue to be covered through their parents’ insurance plans, the Dependent Coverage Expansion (DCE). This should have allowed more patients of this age to be better able financially to participate in available clinical trials at major centers and further enabled survival improvement and cure. An early suggestion that the ACA would benefit young AYAs is the observation that uninsured AYAs were less likely to enroll in clinical trials, in 2006, before the ACA, but not in 2012/2013 after the ACA was implemented.³⁵ Also a prior report suggested that the survival of AYAs in the DCE age group improved after the ACA was enacted.³⁶ Current evidence suggests that survival has improved after a diagnosis of invasive cancer since the ACA DCE was implemented. For all invasive cancer, the AAPC in 5-year relative survival rate during the first 6 years of the ACA, 2010–2015, was greater in the ACA-DCE-impacted age range of 18 to 25 years than in any other age group between 9 and 54 years of age (Fig. 7 lower panel). It also contrasts with the age-dependent trends in younger and older patients and was also statistically significant, whereas it was not in any of the younger age groups above age 6 or in the next oldest age group (Fig. 7 lower panel). For ALL and LBL, a benefit is suggested by the switch in the negative trend change as a function of age in younger patients to a positive trend that started in the ACA age group (Fig. 7 upper panel). That ALL and LBL represent only 3% of all invasive cancers in 18 to 25 year olds may compromise the capability of their survival rate to detect an ACA impact. Also, the ALL-LBL survival cliff was shrinking before the ACA was enacted in 2010 (Fig. 2).

Another barrier to clinical trials accrual in the AYA years is the referral of patients to centers with, or access to clinical trials. In Utah, with one children’s hospital and one National Comprehensive Cancer Network center, the referral of cancer patients to either center dropped with each year of age from 14 at 92% to 17 at 44%.²⁸ By age 20, 2 of every 3 patients stayed in or were referred to private practices and by age 25 less than 30% of the patients were ever seen at an academic medical center.²⁸ Thus, a “referral cliff’ also exists in AYA oncology.

Many other factors contribute to lack of participation in clinical trials. In addition to economic and insurance-based factors, these can be classified as issues of continuity of care and philosophy, provider bias, patient/family preferences, and cooperative group and cancer center limitations.³⁸ Psychosocial factors that are particularly challenging for AYAs with cancer include identity³⁹, relationships³⁹, mental health³⁹, autonomy³⁹, occupation,³⁹ sexuality,^40,41 and fertility.^42,43 Another factor is race/ethnicity, particularly in view of evidence that participation of Hispanic, Black, and Asian AYAs with cancer worsened from 2006 to 2013.^35,44,45 For ALL specifically, the focus of the next stage for AYAs should not only involve novel treatment approaches but also standardization and optimization of supportive care measures, psychosocial support, adherence interventions, oncofertility treatment, and survivorship care. All these efforts should simultaneously work to address health disparities to ensure that a future of improved outcomes is experienced equitably for all AYA patients.⁴⁶

Participation in clinical trials also provides biospecimens for AYAs that are under-represented in biobank repositories and have thereby restricted translational research.¹ Also as mentioned above, the array of cancers in AYAs is distinctly different from that at any other age, and those that appear to be the same as in other age groups are often of a different biology. The need for translational research is greater and more challenging in AYAs than in any other age group.⁴⁷ Increased availability of treatment trials with laboratory correlates and accrual to them are essential to improving the outcome of cancer in AYAs.

Recent studies have suggested that effective evidence-based treatment strategies generated in AYA treatment trials may not be rapidly adopted by oncologists.⁴⁸ ALL, where the evidence is most abundant, only 31% of AYA patients in the greater San Francisco Bay area received a demonstrably superior pediatric type of regimen during 2008–2012, and the rate declined thereafter to 21%.⁴⁸ That the pediatric-inspired regimens for ALL therapy have been demonstrated to be more effective than most regimens originally developed in adult patients is well established.⁴⁹ Adult facilities treating ≥2 AYA ALL patients per year captured in the region were statistically significantly more likely to administer a pediatric regimen than lower volume centers, further indicating the importance of referral patterns.⁴⁸ The need for continued improvement of AYAs to clinical trial accrual remains a significant challenge, not only in community settings and low AYA volume centers but also among minorities.⁵⁰ Fortunately, NCI-designated cancer centers are evaluating their own AYA referral patterns and clinical trial determinants,⁵¹ and intergroup efforts are under way within the current organizational structure of the federal clinical trials enterprise, including the NCTN, to create novel opportunities for collaborative AYA oncology research among the pediatric and adult NCTN groups.^52,53

Limitations of the results presented herein include the lack comparing survival rates of AYAs who participated in clinical trials with those of AYAs who did not participate. Also, the clinical trial accrual data analyzed include phase I, II, and III trials combined, and since phase III trials are most meaningful in comparing new versus older therapies, the accuracy of the results may be limited. Also, the clinical trial data do not include trials sponsored by the pharmaceutical industry that can predominate at certain age levels. Among AYAs however, these trials are less common since AYAs have different cancers than do older adults and, since they represent a much smaller population than those of older age, fewer drugs are developed for them by the industry. Also, the extent to which elimination of the survival cliff by itself helped increase the survival rate in older AYAs is uncertain. Nonetheless, the causes of its abatement undoubtedly contributed to the overall AYA benefit.

A previous decrement in survival of older adolescents and young adults with ALL and LBL in the United States is no longer apparent, resulting in 1 thousands of fewer premature deaths and tens of thousands of years of life saved, and impacting all major races in the country. Primary reasons include increased clinical trials available to, and participation by, 15- to 29-year-old patients, the application of pediatric-inspired treatment regimens to AYAs, recently available expanded national health insurance coverage for 18- to 25-year-old people improved AYA cancer services, and a national focus on AYA oncology. Older AYAs with ALL or LBL have not had a similar increase in clinical trial accrual, however, and their survival improvement is largely due to the elimination of the survival gap in younger AYAs. They need more of what younger AYAs have benefitted from, including health insurance for all AYAs and not just for those less than 25 years of age who are able to continue on the parents’ insurance plan. Altogether, the recently established National Cancer Treatment Network, National Community Oncology Research Program, and the FDA may help eliminate the clinical trial accrual gap.