Section Structure
This article is organized into the following sections: Highlights; Clinical Background; Study Design and Methods; Key Results; Clinical Interpretation; Strengths and Limitations; Practice Implications; Conclusion; Funding and Trial Registration; References.
Highlights
Adults with a first-time hospital-diagnosed seizure had a sharply increased 1-year risk of cancer compared with the general Danish population, with the strongest signal for neurological malignancies.
Within the first year after seizure diagnosis, the absolute risk of any cancer was 4.1%, including 2.4% for neurological cancers and 1.7% for nonneurological cancers.
The relative excess risk was especially striking for neurological cancer in the first year, with a standardized incidence ratio of 76.1, but a modestly elevated risk persisted for both neurological and nonneurological cancers over 5 to 20 years.
These findings support considering malignancy, not only primary brain tumors but also systemic cancers, in selected adults presenting with an unexplained first seizure.
Clinical Background
Seizures are a well-recognized manifestation of intracranial disease. In clinical practice, a first seizure often prompts evaluation for structural brain lesions, including stroke, infection, traumatic injury, and primary or metastatic tumors. Brain tumors are particularly relevant because seizures may be the initial presenting symptom, especially for cortical lesions. The relationship between seizures and occult nonneurological cancer is less intuitive but still biologically plausible. Systemic cancers may lead to seizures through brain metastases, paraneoplastic syndromes, metabolic disturbances, treatment effects, thromboembolic complications, or indirect inflammatory mechanisms.
Although clinicians are accustomed to searching for neurological causes after a first seizure, whether a seizure can function more broadly as an early marker of undiagnosed cancer has remained uncertain. This question matters because the differential diagnosis after a first seizure is wide, and extensive cancer workup for every patient would be impractical, expensive, and potentially harmful. Better risk estimates can help clinicians calibrate how aggressively to investigate for malignancy, and in which subgroups.
The study by Pedersen and colleagues addresses this evidence gap using nationwide Danish registry data spanning more than two decades. Its main contribution is to quantify both absolute risk and relative risk of cancer after a first-time seizure, while separating neurological from nonneurological cancers and examining short-, intermediate-, and long-term periods.
Study Design and Methods
Design and data source
This was a population-based cohort study using nationwide Danish medical registries from January 1996 through December 2022. Denmark’s linked health registries are particularly well suited to this question because they capture hospital diagnoses, cancer outcomes, migration, and mortality at the population level, reducing loss to follow-up and enabling robust comparisons with the general population.
Population
The investigators included all adults aged 18 years or older with a first-time hospital diagnosis of seizure and no prior cancer diagnosis. Excluding individuals with preexisting cancer was important because the study aimed to determine whether seizure could be an early sign of previously undiagnosed malignancy rather than a complication of known cancer.
The final cohort included 49,894 adults. Median age at seizure diagnosis was 51.5 years, with an interquartile range of 35.6 to 67.8 years. Women accounted for 20,648 patients, or 41.4% of the cohort.
Exposure and outcomes
The exposure was a first-time seizure diagnosis recorded in hospital-based care. The primary outcomes were incident cancers, excluding nonmelanoma skin cancer. Outcomes were evaluated as all cancers combined, neurological cancers, nonneurological cancers, and site-specific cancers.
Follow-up windows and measures
Follow-up extended until cancer diagnosis, emigration, death, or December 31, 2022. The authors reported cancer occurrence across three clinically meaningful periods: within 1 year; from 1 to less than 5 years; and from 5 to 20 years after the first seizure.
Two complementary measures were emphasized. Absolute risk quantifies the real-world probability of developing cancer over a given interval and is often most useful clinically. Standardized incidence ratios, or SIRs, compare observed cancer incidence in the seizure cohort with the expected incidence in the general Danish population, adjusted for relevant demographic factors. SIRs clarify the magnitude of excess relative risk.
Key Results
Overall cancer risk
The central finding was a pronounced increase in cancer diagnoses, especially during the first year after the seizure. Within 1 year, 4.1% of patients were diagnosed with any cancer, corresponding to an SIR of 5.30 (95% CI, 5.07-5.54). In other words, adults with a first-time seizure had more than five times the cancer incidence expected in the general population over the ensuing year.
From 1 to less than 5 years after seizure, the absolute risk of any cancer was 3.5%, with an SIR of 1.18 (95% CI, 1.12-1.25). From 5 to 20 years, the absolute risk was 13.4%, with an SIR of 1.34 (95% CI, 1.28-1.40). The long-term absolute risk rose, as expected, because cancer incidence accumulates with age and time, but the relative excess compared with the general population was modest rather than dramatic.
Neurological cancers
The strongest association was with neurological cancers. Within the first year after seizure diagnosis, 1,172 neurological cancers were observed. The absolute risk was 2.4%, and the SIR was 76.1 (95% CI, 71.8-80.6). This is an extraordinarily high relative risk and strongly supports the idea that first seizure can be a sentinel clinical event for occult intracranial malignancy.
Even beyond the first year, the signal remained above baseline, though far weaker. During years 1 to less than 5, there were 87 neurological cancers, corresponding to an absolute risk of 0.2% and an SIR of 1.85 (95% CI, 1.48-2.28). During years 5 to 20, 112 neurological cancers were observed, for an absolute risk of 0.7% and an SIR of 1.46 (95% CI, 1.20-1.75).
These data suggest that the immediate post-seizure period is the key window for detecting occult brain tumors, but they also leave open the possibility that seizure marks an enduring vulnerability in some patients or that residual diagnostic surveillance continues to uncover cases later.
Nonneurological cancers
The study also found elevated risks for cancers outside the nervous system. Within 1 year, 850 nonneurological cancers were diagnosed, yielding an absolute risk of 1.7% and an SIR of 2.32 (95% CI, 2.17-2.48). This relative increase is much smaller than that seen for neurological cancers, but it is still clinically meaningful.
During 1 to less than 5 years, 1,226 nonneurological cancers occurred, with an absolute risk of 3.3% and an SIR of 1.15 (95% CI, 1.09-1.22). During 5 to 20 years, 2,120 nonneurological cancers occurred, with an absolute risk of 12.8% and an SIR of 1.33 (95% CI, 1.28-1.39).
The pattern is notable. For nonneurological cancer, the 1-year excess suggests that seizure may occasionally be an early sign of systemic malignancy, whether through unrecognized brain metastasis, paraneoplastic effects, or associated physiological disturbances. The modest long-term elevation may reflect shared risk factors, diagnostic intensity, comorbidity patterns, or persistent biological links rather than direct causation.
Interpreting absolute versus relative risk
One of the most clinically useful aspects of this study is its presentation of both absolute risks and SIRs. The first-year SIR for neurological cancer is very large, but the absolute risk was 2.4%. This means the association is strong and highly informative, yet most patients with a first-time seizure still did not receive a neurological cancer diagnosis within a year. Likewise, the first-year absolute risk of nonneurological cancer was 1.7%, which is elevated but not high enough to justify indiscriminate whole-body cancer screening in all patients.
For clinicians, this distinction matters. A first seizure should heighten awareness of malignancy, especially intracranial malignancy, but the response should remain targeted and guided by age, seizure semiology, imaging findings, neurological examination, systemic symptoms, and laboratory clues.
Clinical Interpretation
Why the short-term signal is so strong
The striking 1-year excess in neurological cancer risk is biologically and clinically plausible. Seizures are a common presenting symptom of gliomas and other cortical brain tumors, and the temporal clustering suggests that many of these cancers were already present, though undiagnosed, at the time of seizure. In effect, the seizure may have been an early symptom rather than a causal factor.
The 1-year excess in nonneurological cancer may arise through several pathways. Some patients may have had occult metastatic disease to the brain or leptomeninges before their systemic cancer was formally diagnosed. Others may have had cancer-related metabolic abnormalities, cerebrovascular events, or immune-mediated syndromes. Another explanation is diagnostic acceleration: once a patient enters hospital care for a first seizure, imaging, blood tests, and specialist assessments may uncover unrelated or indirectly related malignancies that would otherwise have been diagnosed later.
Implications for first-seizure evaluation
Current evaluation of first seizure typically includes a careful history, neurological examination, medication and substance review, laboratory testing, neuroimaging, and often electroencephalography. This study reinforces the importance of timely brain imaging, particularly MRI when available and clinically appropriate, because the strongest excess risk concerns neurological cancer. In adults with unexplained first seizure, focal deficits, persistent headache, cognitive change, or abnormal imaging, the threshold for expedited neuro-oncology evaluation should be low.
The findings do not necessarily imply that every patient with a first seizure needs systematic screening for extracranial malignancy. Rather, the data support a selective strategy. Red flags that may justify broader evaluation include constitutional symptoms, smoking history, anemia, unexplained weight loss, older age, imaging suggestive of metastasis, paraneoplastic features, or laboratory abnormalities indicating possible systemic disease.
Relation to broader evidence
Prior literature has established seizures as common in patients with primary brain tumors and brain metastases, and practice guidelines already emphasize structural neuroimaging after an unprovoked first seizure in adults. What this study adds is the population-level estimate of subsequent cancer risk, including nonneurological cancer, across both short and long intervals. The magnitude of short-term neurological cancer risk is consistent with existing clinical intuition, but the quantified excess of nonneurological cancer is a more novel and practice-relevant observation.
Strengths and Limitations
Strengths
The study’s major strengths are scale, completeness, and real-world applicability. Nearly 50,000 adults were included, follow-up was nationwide and prolonged, and registry linkage minimized attrition. The use of both absolute risks and SIRs strengthens clinical interpretation. Separating neurological from nonneurological malignancies also improves actionability for frontline clinicians.
Limitations
As with all registry-based observational studies, several limitations deserve attention. First, seizure diagnoses were based on hospital coding rather than adjudicated clinical review, so some misclassification is possible. Second, surveillance bias is a real concern, especially in the first year, because a seizure often triggers diagnostic workup that can accelerate cancer detection. Third, the abstract does not provide detailed subgroup analyses by seizure type, imaging findings, age strata, or cancer stage, all of which would help refine risk stratification.
Fourth, residual confounding is likely. Shared determinants such as smoking, alcohol use, vascular disease, socioeconomic status, and other comorbidities may contribute to both seizure risk and cancer risk. Fifth, the findings arise from the Danish healthcare context and may not translate perfectly to settings with different cancer epidemiology, access to neuroimaging, or thresholds for hospital diagnosis of seizure.
Finally, elevated long-term SIRs should be interpreted cautiously. They may represent shared risk factors or persistent surveillance rather than a direct pathophysiologic link between seizure and future malignancy.
Practice Implications
For neurologists, emergency physicians, internists, and primary care clinicians, the practical message is nuanced. A first-time seizure should prompt careful evaluation for structural brain disease, including malignancy, because the short-term neurological cancer risk is substantial. This is especially true when the seizure is unprovoked and accompanied by focal signs or suspicious symptoms.
At the same time, the study does not support indiscriminate cancer screening in all patients with first seizure. Instead, it argues for heightened clinical vigilance and broader assessment in select patients whose history, examination, laboratory profile, or neuroimaging suggests systemic disease. A measured approach is likely to maximize diagnostic yield while minimizing unnecessary testing.
From a health-system perspective, these findings may also inform diagnostic pathways. Protocols for adults with first seizure should ensure access to timely neuroimaging and a clear mechanism for escalation when imaging or systemic features raise concern for malignancy. Future research should aim to identify clinical predictors that distinguish patients who need only standard seizure workup from those who warrant expanded oncologic evaluation.
Conclusion
This nationwide Danish cohort study provides compelling evidence that a first-time seizure can be an early clinical marker of occult cancer, particularly neurological cancer. The first year after seizure diagnosis is the critical period, with a 4.1% absolute risk of any cancer and a more than fivefold overall relative increase compared with the general population. The neurological cancer signal is especially strong, with an SIR of 76.1 in the first year. Nonneurological cancers also occur in excess, though at lower absolute and relative frequencies.
The results strengthen the case for thorough etiologic evaluation after a first seizure, with prompt attention to possible intracranial malignancy and selective consideration of systemic cancer when clinical clues are present. The key translational point is not that every first seizure should launch a broad cancer search, but that malignancy should remain firmly on the differential diagnosis, particularly in the early diagnostic window.
Funding and Trial Registration
The provided abstract and citation do not report funding details or a ClinicalTrials.gov registration number. As an observational registry-based cohort study, formal trial registration may not be applicable.
References
Pedersen AL, Farkas DK, Fuglsang CH, Henderson VW, Al-Mashhadi SK, Elser H, Sørensen HT. Risk of Cancer in Patients With First-Time Seizure. JAMA Neurology. 2026 Apr 27. PMID: 42043826. Available at: https://pubmed.ncbi.nlm.nih.gov/42043826/
Krumholz A, Wiebe S, Gronseth GS, et al. Evidence-based guideline: Management of an unprovoked first seizure in adults. Report of the Guideline Development Subcommittee of the American Academy of Neurology and the American Epilepsy Society. Neurology. 2015;84(16):1705-1713.
National Institute for Health and Care Excellence. Epilepsies in children, young people and adults. NICE guideline NG217. Updated 2025.
Weller M, van den Bent M, Hopkins K, et al. EANO guideline for the diagnosis and treatment of anaplastic gliomas and glioblastoma. Lancet Oncology. 2014;15(9):e395-e403. This reference remains relevant for the clinical context of tumor-associated seizures, although newer disease-specific updates should be consulted for current management decisions.
