Addressing the knowledge gap in the genomic landscape and tailored therapeutic approaches to adolescent and young adult cancers

Abstract

Background: Adolescents and young adults (AYAs) represent a small proportion of patients with cancer. The genomic profiles of AYA patients with cancer are not well-studied, and outcomes of genome-matched therapies remain largely unknown.

Patients and methods: We investigated differences between Japanese AYA and older adult (OA) patients in genomic alterations, therapeutic evidence levels, and genome-matched therapy usage by cancer type. We also assessed treatment outcomes.

Results: AYA patients accounted for 8.3% of 876 cases. Microsatellite instability-high and/or tumor mutation burden was less common in AYA patients (1.4% versus 7.7% in OA; P = 0.05). However, BRCA1 alterations were more common in AYA patients with breast cancer (27.3% versus 1.7% in OA; P = 0.01), as were MYC alterations in AYA patients with colorectal cancer (23.5% versus 5.8% in OA; P = 0.02) and sarcoma (31.3% versus 3.4% in OA; P = 0.01). Genome-matched therapy use was similar between groups, with overall survival tending to improve in both. However, in AYA patients, the small number of patients prevented statistical significance. Comprehensive genomic profiling-guided genome-matched therapy yielded encouraging results, with progression-free survival of 9.0 months in AYA versus 3.7 months in OA patients (P = 0.59).

Conclusion: Our study suggests that tailored therapeutic approaches can benefit cancer patients regardless of age.

Keywords: adolescent; genome-matched therapy; genomic profiling; older adult; treatment outcome; young adult.

Figure 1

Genomic landscape. (A) Flowchart of the overall patient disposition in this study. A total of 4498 genomic alterations were detected in 876 patients, of which 3132 (69.6%) were considered pathogenic. Thirty-seven genes were selected for further investigation. (B) The frequency of driver gene alterations by cancer type in AYA and OA patients. Thirty-three genes listed in all three comprehensive genomic profiles and meeting one of the following criteria were selected: (i) mutation frequency of ≥5% or (ii) a cancer-type-specific mutation frequency of ≥10% in this cohort. AYA and OA patients had similar frequencies of MSI-H/TMB-H. However, AYA patients with breast cancer had significantly higher rates of BRCA1 mutations (P = 0.01), and MYC alterations were significantly more common in AYA patients with colorectal cancer and sarcoma (P = 0.02 and 0.01, respectively). AYA, adolescent and young adult; CUP, cancer of unknown primary origin; HCC, hepatocellular carcinoma; MSI-H, microsatellite instability-high; NE, not evaluated; NS, not significant; OA, older adult; TMB-H, tumor mutation burden-high.

Figure 2

Patient-based analysis. (A) Distribution of OncoKB therapeutic evidence levels in AYA and OA patients. There were no significant differences in the distribution of OncoKB therapeutic evidence levels assigned between AYA and OA patients, neither overall nor by cancer type. The numbers represent the following therapeutic evidence levels: 1, FDA-approved drugs; 2, standard care; 3, clinical evidence; and 4, biological evidence. (B) Use of genome-matched therapy in AYA and OA patients. AYA patients had a tendency to more likely receive genome-matched therapy (20.5% and 12.9%, P = 0.07). Regarding cancer type, there was no significant difference between AYA and OA patients. (C) Overall survival according to the use of genome-matched therapy in AYA and OA patients. Overall survival was 33.0 months (95% CI 6.8-not achieved) in patients with genome-matched therapy compared with 13.0 months (95% CI 8.8-17.8) in those with nongenome-matched therapy in AYA patients (P = 0.10), while the corresponding values were 14.7 months (95% CI 11.8-20.9) and 12.5 months (95% CI 11.0-14.2) in OA patients (P = 0.08), respectively. AYA, adolescent and young adult; CUP, cancer of unknown primary; HCC, hepatocellular carcinoma; MSI-H, microsatellite instability-high; OA, older adult; TMB-H, tumor mutation burden-high.

Figure 2

Patient-based analysis. (A) Distribution of OncoKB therapeutic evidence levels in AYA and OA patients. There were no significant differences in the distribution of OncoKB therapeutic evidence levels assigned between AYA and OA patients, neither overall nor by cancer type. The numbers represent the following therapeutic evidence levels: 1, FDA-approved drugs; 2, standard care; 3, clinical evidence; and 4, biological evidence. (B) Use of genome-matched therapy in AYA and OA patients. AYA patients had a tendency to more likely receive genome-matched therapy (20.5% and 12.9%, P = 0.07). Regarding cancer type, there was no significant difference between AYA and OA patients. (C) Overall survival according to the use of genome-matched therapy in AYA and OA patients. Overall survival was 33.0 months (95% CI 6.8-not achieved) in patients with genome-matched therapy compared with 13.0 months (95% CI 8.8-17.8) in those with nongenome-matched therapy in AYA patients (P = 0.10), while the corresponding values were 14.7 months (95% CI 11.8-20.9) and 12.5 months (95% CI 11.0-14.2) in OA patients (P = 0.08), respectively. AYA, adolescent and young adult; CUP, cancer of unknown primary; HCC, hepatocellular carcinoma; MSI-H, microsatellite instability-high; OA, older adult; TMB-H, tumor mutation burden-high.

Figure 3

Outcomes of genome-matched therapy guided by comprehensive genomic profiling. (A) Flowchart of patient disposition according to the use of genome-matched therapy in AYA and OA patients. Forty-eight patients received genome-matched therapy guided by comprehensive genomic profiling. Six patients were excluded owing to being treated elsewhere. (B) Objective response and progression-free survival in AYA and OA patients. Arrows indicate continued response; disease control for >6 months was achieved in 3 AYA patients (60.0%) and 10 OA patients (27.0%). The table details each cancer type and treatment. ALK, anaplastic lymphoma kinase; AYA, adolescent and young adult; BRAF, v-raf murine sarcoma viral oncogene homolog B1; CDK, cyclin-dependent kinase; CGP, comprehensive genomic profiling; ERBB2, v-erb-b2 avian erythroblastic leukemia viral oncogene homolog 2; FGFR, fibroblast growth factor receptor; KRAS, Kirsten rat sarcoma virus; MEK, mitogen-activated protein kinase kinase; OA, older adult; PARP, poly(ADP-ribose) polymerase; PI3K, phosphoinositide 3-kinase.