Survivors of cancer diagnosed in childhood carry a dramatically elevated risk of developing a malignant tumor of the brain or other parts of the nervous system decades after their original diagnosis, according to a large population-based analysis published in the journal Cancer Causes & Control. The study, which drew on nearly five decades of cancer registry data from the United States, found that people who were diagnosed with cancer before the age of 15 were more than five times as likely as the general population to later develop a malignant nervous system tumor. By contrast, the excess risk among those diagnosed in adolescence or young adulthood was far more modest, and the absolute numbers involved remained small.
The research team, led by Xiaoxin Wu and Huadong Liang of Longyan First Affiliated Hospital of Fujian Medical University in China, set out to answer a question that has long troubled oncologists: how does the risk of a subsequent malignant brain tumor vary with both the age at which the first cancer was diagnosed and the length of time that has elapsed since then? Most previous studies have reported a single overall risk figure for survivors, which can obscure important patterns. By jointly analyzing age and latency, the researchers revealed structure that a summary statistic would have hidden.
To conduct the analysis, the investigators used the Surveillance, Epidemiology, and End Results, or SEER, program, a network of population-based cancer registries maintained by the National Cancer Institute. Specifically, they examined the SEER 8 registries, the original eight registries that have contributed data continuously since 1975, using the November 2025 submission covering the years 1975 through 2023. The study population consisted of 276,844 people who had been diagnosed with a first primary malignancy other than a brain or other nervous system cancer before reaching the age of 40. Together these individuals contributed 3,970,462 person-years of observation, an enormous cumulative window into the long-term health of cancer survivors.
The central statistical tool was the standardized incidence ratio, or SIR, a measure that compares the number of tumors actually observed in the survivor cohort with the number that would be expected if the cohort experienced the same age-, sex-, and calendar year-specific rates as the general population. An SIR of 1.0 indicates no excess; values above 1.0 indicate more tumors than expected. Over the entire follow-up period, 345 malignant brain and other nervous system tumors occurred in the cohort, against 236.97 expected, yielding an overall SIR of 1.46 with a 95 percent confidence interval of 1.31 to 1.62. The excess absolute risk, which translates the relative excess into additional cases per unit of population time, was 2.72 per 100,000 person-years.
That overall figure, however, concealed a steep age gradient. Among survivors whose first cancer was diagnosed between birth and age 14, the SIR was 5.20, meaning they experienced more than five times the expected number of malignant nervous system tumors. For those diagnosed at ages 15 to 19, the SIR fell to 1.84. Among survivors diagnosed at ages 20 to 29, the ratio was 1.17, and for those diagnosed at ages 30 to 39 it was 1.23, both only modest elevations. The pattern is striking: the younger the patient at the original cancer diagnosis, the greater the relative excess of later nervous system tumors, with the childhood group standing far apart from the rest.
Crucially, the researchers found that this age pattern itself changed depending on how much time had passed since the first diagnosis, a statistical interaction that was significant at a p-value of 0.0012. Latency, the interval between the first cancer and the subsequent tumor, matters because different mechanisms of tumor development unfold on different timescales. Radiation-induced tumors, for example, typically require years to decades to emerge, while other processes may act more quickly. When the analysis was restricted to intervals of ten or more years after the original diagnosis, the childhood group still showed an SIR of 4.64, with a 95 percent confidence interval of 3.22 to 6.49. In other words, the elevated relative risk among childhood cancer survivors did not fade away with time; it persisted well beyond a decade of survival.
To complement the relative measures, the team performed a supplementary competing-risk analysis estimating cumulative incidence, which accounts for the fact that survivors may die of other causes before a second tumor ever appears. Starting from a two-month landmark after the original diagnosis, the 20-year cumulative incidence of a malignant brain or other nervous system tumor was 0.257 percent among childhood cancer survivors, with a 95 percent confidence interval of 0.193 to 0.341 percent. This figure puts the risk in perspective: even in the highest-risk group, roughly one in four hundred childhood cancer survivors developed a malignant nervous system tumor within twenty years. The authors emphasized that while the relative excess was substantial, the absolute excess and cumulative incidence were small.
The findings align with a substantial body of prior research on subsequent malignant neoplasms, the medical term for new primary cancers arising in survivors. Landmark analyses from the Childhood Cancer Survivor Study, a cohort of more than five-year survivors of childhood cancer treated at institutions across North America, have documented elevated risks of central nervous system tumors, with cranial radiation therapy identified as a major driver. Population-based studies from the British Childhood Cancer Survivor Study and, more recently, work published in the Journal of the National Cancer Institute on temporal trends of subsequent central nervous system malignancies have similarly confirmed that survivors of childhood cancer remain at heightened risk for decades. Studies of adolescent and young adult survivors, including the Teenage and Young Adult Cancer Survivor Study in Britain and a recent population-based analysis from Alberta, Canada, have generally found lower relative risks in those diagnosed later in life, consistent with the age gradient observed here.
Why should age at diagnosis matter so much? Several biological explanations are plausible, though the registry design of this study cannot distinguish among them. Younger tissues are generally more susceptible to the carcinogenic effects of radiation and chemotherapy, and a diagnosis early in life leaves more remaining years for latent damage to manifest. The developing nervous system may be particularly vulnerable to treatment-related DNA damage during critical periods of growth. Inherited cancer predisposition syndromes, which are more likely to be diagnosed in childhood, may also contribute to second tumor risk. The SEER registry data do not include detailed treatment information, so the study cannot directly attribute the excess to radiation dose or specific agents, a limitation the authors implicitly acknowledge by framing the work as a descriptive, population-based analysis.
The practical implications reach into survivorship care. Long-term follow-up guidelines issued by the Children’s Oncology Group already recommend surveillance tailored to treatment exposure, and the new results reinforce the importance of sustained, decades-long follow-up for childhood cancer survivors specifically, since their excess risk remained evident beyond ten years. At the same time, the low absolute risk argues against indiscriminate intensive screening of all survivors, particularly those diagnosed in adolescence or adulthood whose excess was modest. The authors note that joint assessment of diagnosis age and latency characterized patterns that the overall estimate obscured, a methodological point with resonance for other survivorship research. The study’s analysis code has been publicly archived on Zenodo and maintained on GitHub, and the underlying SEER data remain available to eligible researchers through the National Cancer Institute, allowing independent verification of the findings. For the hundreds of thousands of people who survive cancer before age 40 each year, the message is one of calibrated vigilance: the risk of a later malignant brain tumor is real and highest for those diagnosed as children, but it remains, in absolute terms, uncommon.
Subject of Research: Age- and latency-specific risk of subsequent malignant brain and nervous system tumors among survivors of cancer diagnosed before age 40
Article Title: Age- and latency-specific patterns of subsequent malignant brain and other nervous system tumors after cancer diagnosed before age 40 years
Article References: Wu, X., Liang, H., Wu, Q., & Li, X. (2026). Age- and latency-specific patterns of subsequent malignant brain and other nervous system tumors after cancer diagnosed before age 40 years. Cancer Causes & Control, 37(10), Article 177. https://doi.org/10.1007/s10552-026-02265-7
Image Credits: AI Generated
DOI: 10.1007/s10552-026-02265-7
Keywords: childhood cancer, cancer survivors, brain tumor, subsequent malignant neoplasm, standardized incidence ratio, SEER program, cancer epidemiology, latency, adolescent and young adult cancer, survivorship, nervous system tumors, cumulative incidence
News Source: Nathaniel Bowman. (October 7, 2026). Childhood Cancer Survivors Face Fivefold Higher Risk of Later Brain Tumors, Population Study Finds. Scienmag.
