Brain Tumors That Follow Radiation Therapy Come Back Sooner, Study Finds

Meningiomas are the most common primary tumors of the central nervous system, accounting for up to 40 percent of all primary brain and spinal tumors, with a population-based incidence of roughly 8 to 9 cases per 100,000 person-years. Most arise sporadically, their causes still poorly understood, though female sex hormones and long-term use of certain synthetic progestogens such as cyproterone acetate have been consistently linked to their development and growth. But a smaller and clinically distinct subset of these tumors has a known trigger: ionizing radiation. These radiation-induced meningiomas typically emerge 15 to 35 years after cranial radiotherapy given for childhood cancers, pituitary disease, or other tumors of the nervous system, and as survival after radiotherapy-treated malignancies continues to improve, the population at risk keeps growing.

For decades, descriptive studies have suggested that these radiation-induced tumors behave worse than their sporadic counterparts. They appear to present at younger ages, more often as multiple synchronous lesions, with a higher share of atypical WHO grade 2 histology, and with higher recurrence rates after surgery. The problem with those earlier comparisons, however, is that they were largely unmatched. Radiation-exposed patients and historical sporadic controls differed systematically in age, tumor grade, and extent of surgical resection, all of which independently influence the odds of a tumor coming back. Without controlling for those factors, no one could say whether the worse outcomes reflected an intrinsically more aggressive tumor biology or simply a skewed comparison.

A new study from the University Medical Center Mainz in Germany, published in the Journal of Neuro-Oncology, set out to answer that question with a methodological upgrade. Led by neurosurgeons Leon Schmidt and Dragan Jankovic, with Darius Kalasauskas as senior author, the team identified every radiation-induced intracranial meningioma resected at their institution between 2007 and 2023, then applied propensity score matching to construct a fair comparison group. To qualify as radiation-induced, tumors had to satisfy Cahan’s criteria, the classic set of conditions used for radiation-associated second cancers: the tumor had to arise within a previously irradiated field, at least five years had to have elapsed since irradiation, the meningioma diagnosis had to be histologically confirmed, and the tumor could not have been present at the time of the original treatment.

From an initial pool of 817 eligible historical sporadic cases, the researchers matched 195 controls to 39 radiation-induced meningioma patients at a one-to-five ratio, balancing the groups on age, WHO grade, and Simpson grade, the standard measure of how completely a meningioma has been resected. Propensity scores were calculated with a logistic regression model, and nearest-neighbor matching without replacement was performed using a caliper width of 0.2 standard deviations of the logit of the score. After matching, standardized mean differences confirmed that the groups were well balanced on the key prognostic factors, with most values falling below the conventional threshold of 0.10, though some residual imbalance remained for certain Simpson grade levels.

The results were striking. Even after this careful balancing, radiation-induced meningiomas recurred significantly more often than matched sporadic tumors: 28 percent versus 17 percent, a difference that reached statistical significance. Median progression-free survival, the time from surgery to the first radiographically documented recurrence or progression, was 66 months in the radiation-induced group compared with 88 months in controls. In total, 44 relapse events were observed across the matched cohort, 11 among the 39 radiation-exposed patients and 33 among the 195 controls.

The most rigorous test came from multivariable Cox proportional hazards regression, which adjusted simultaneously for complete resection and for surgically complex tumor location. In that model, radiation-induced tumor status remained an independent predictor of shorter progression-free survival, with a hazard ratio of 2.1, meaning the exposed tumors carried roughly twice the risk of relapse at any given time. Complete resection, defined as Simpson grade I or II, was strongly protective, cutting the hazard of recurrence to about 0.29 of that seen after incomplete removal. Surgically complex location, by contrast, did not achieve a significant association. The proportional hazards assumption was verified using Schoenfeld residuals and was not violated for any covariate.

Yet the study is notable as much for its caution as for its findings. The authors emphasize that matching did not erase every clinically relevant difference between the groups. Radiation-induced patients carried a higher comorbidity burden, measured by the Charlson Comorbidity Index, with a median score of 2 versus 0 in controls. Nearly 39 percent of them had undergone prior cranial surgery, compared with just over 4 percent of controls, likely reflecting interventions related to the original condition for which they were irradiated. Their preoperative Karnofsky Performance Status was modestly lower, and adjuvant radiotherapy after meningioma resection was administered far more often in the exposed group, 33.3 percent versus 11.3 percent. Because these variables were not included in the propensity score model, residual measured and unmeasured confounding cannot be excluded, and the authors explicitly warn that the results should not be interpreted as evidence of intrinsically more aggressive tumor biology.

The biological question remains open. Previous work has suggested that therapeutic radiation drives distinct molecular changes in meningiomas, including radiation-induced DNA damage, chromosomal instability, and structural aberrations of the NF2 gene, a key tumor suppressor frequently altered in these tumors. A 2017 study in Nature Communications showed that radiation for childhood cancer can produce structural rearrangements of NF2 specifically in radiation-associated meningiomas, and other groups have reported epigenetic signatures such as loss of H3K27me3. But molecular data were not systematically available in the Mainz cohort, so the study cannot determine whether the earlier relapses reflect these biological differences or the lingering effects of confounding. The authors call for further molecular studies to settle the question.

The limitations of the work are worth taking seriously. It is a retrospective, single-center analysis with only 39 radiation-induced cases, which limits statistical power, and the median follow-up of 20 months is short for a tumor whose late recurrences beyond five years are well described. Follow-up duration also differed between groups, with longer surveillance in the radiation-induced patients, and differences in imaging frequency between the two populations cannot be ruled out in a retrospective design. Historical pathology specimens were not systematically rereviewed under the 2021 WHO classification, so some degree of grade misclassification in older cases is possible. The authors also note that because Simpson grade was included as a matching covariate, the similar rates of complete resection between groups, 71.8 percent versus 76.4 percent, are expected by design and should not be read as evidence of comparable resectability.

Still, the methodological contribution is real. By applying propensity score matching to this question for the first time, the Mainz team has strengthened and refined an association that earlier unmatched studies had only suggested, showing that radiation-induced meningioma status predicts earlier relapse even when age, grade, and extent of resection are held constant. The practical messages for clinicians are twofold. First, patients with a history of cranial irradiation deserve long-term surveillance, since their tumors may return sooner after surgery. Second, complete surgical resection remains the single most powerful modifiable predictor of progression-free survival, supporting maximal safe resection whenever clinically feasible. As the ranks of childhood cancer survivors continue to swell, understanding and managing these delayed consequences of life-saving radiation will only become more important.

Subject of Research: Recurrence risk and progression-free survival of radiation-induced cranial meningiomas compared with sporadic meningiomas using propensity score matching

Article Title: Radiation-induced cranial meningiomas are associated with earlier relapse: a propensity score–matched cohort study

Article References: Schmidt, L., Jankovic, D., Thavarajasingam, S., Fakak, R., Kosterhon, M., Dauth, A., Brockmann, M., Sommer, C., Ringel, F., & Kalasauskas, D. (2026). Radiation-induced cranial meningiomas are associated with earlier relapse: a propensity score–matched cohort study. Journal of Neuro-Oncology, 179(3), Article 97. https://doi.org/10.1007/s11060-026-05809-8

Image Credits: AI Generated

DOI: 10.1007/s11060-026-05809-8

Keywords: radiation-induced meningioma, meningioma recurrence, propensity score matching, progression-free survival, extent of resection, cranial radiotherapy, WHO grade, Simpson grade, neurosurgery, brain tumors, NF2, ionizing radiation