Thirteen Years Cancer-Free, Then a Seizure: Rare Brain Metastasis Stuns Doctors

In September 2021, a 58-year-old man in Naples, Italy, arrived at the emergency department after a seizure involving his left upper limb. A brain scan revealed a 30-millimeter lesion in his right parietal region, surrounded by swelling and suggestive of a primary brain tumor. What followed was one of the most diagnostically challenging cases in modern oncology: the lesion turned out to be a metastasis from a gastroesophageal junction adenocarcinoma that had been surgically removed with curative intent thirteen years earlier, in 2008. The case, reported by a team at Ospedale del Mare and published in Frontiers in Oncology, offers a striking reminder that cancer can lie silent for more than a decade before reawakening in the most unexpected place.

The patient’s original cancer was a Siewert type II adenocarcinoma of the gastroesophageal junction, the anatomical crossroads where the esophagus meets the stomach. Preoperative staging with endoscopy, contrast-enhanced total-body computed tomography, and fluorodeoxyglucose positron emission tomography showed localized disease, and his tumor markers were normal. He underwent a total gastrectomy with D1+ lymphadenectomy, and pathological staging revealed a pT3N1 well-differentiated tumor with a single positive lymph node out of eighteen. From April to September 2009, he received twelve cycles of adjuvant modified FOLFOX-4 chemotherapy, which he tolerated without severe adverse events. For ten years afterward, he attended regular follow-up with no evidence of recurrence, and in 2019 his routine oncological surveillance was discontinued. By every conventional measure, he was cured.

When the seizure brought him back to the hospital, contrast-enhanced magnetic resonance imaging confirmed a solid lesion measuring 29 by 26 millimeters in the right parietal lobe, associated with extensive edema involving the frontoparietal and temporal regions and mild ventricular compression. The multidisciplinary brain team suspected a primary brain tumor, and in October 2021 the patient underwent craniotomy. Histological examination unexpectedly revealed metastatic adenocarcinoma, morphologically and immunophenotypically compatible with recurrence from his previous gastroesophageal junction primary. Immunohistochemistry showed positivity for cytokeratin 7, cytokeratin 20, CDX2, and epithelial membrane antigen, with focal positivity for the mucins MUC2 and MUC5, and negativity for TTF1 and GFAP. The TTF1 result argued against a lung or thyroid primary, while the GFAP negativity excluded a glial brain tumor. HER-2 was scored 2+ on immunohistochemistry, but in situ hybridisation showed no amplification.

The diagnostic work-up went further than morphology and immunostaining. Molecular analysis by EPIC array, performed according to the institutional protocol for brain tumors, showed no MGMT promoter methylation, IDH1 and IDH2 wild-type status, and no 1p/19q co-deletion. Copy-number alterations included partial chromosome 7 amplification, amplification of CDK4, MYC, NF1, and FGFR3, and loss of TP53 and MDM4. Methylome profiling supported a diagnosis of metastatic disease of probable gastric origin. The authors emphasized that attribution of the brain lesion to the prior gastroesophageal junction cancer rested on the integration of morphology, clinical history, immunohistochemistry, systemic staging, and methylome profiling rather than on any single assay. Methylation-based classifiers, they noted, depend on tumor purity, tissue quality, and the available reference classes, so the molecular result was interpreted as supportive rather than definitive.

Equally remarkable was what the systemic restaging did not show. Endoscopy, contrast-enhanced computed tomography, brain magnetic resonance imaging, and total-body positron emission tomography demonstrated no extracranial disease, and tumor markers remained normal. After discussion by the gastric multidisciplinary team, a strategy of close follow-up was adopted, given the exceptionally long disease-free interval and the absence of systemic disease. In February 2022, magnetic resonance imaging suggested local recurrence at the surgical cavity, and the patient received fractionated stereotactic radiotherapy, 25 Gy to the cavity and 30 Gy to an adjacent posterior nodule in five fractions. Post-treatment imaging showed a response.

The apparently isolated intracranial event did not remain isolated. In July 2022, the patient developed mandibular pain rated 6 on the numerical pain scale, and although imaging showed no recurrent systemic disease, CA 19-9 was elevated at 83.88 UI/ml. By October 2022, a cone-beam computed tomography scan revealed a bone gap in the lower right mandibular branch, and magnetic resonance imaging demonstrated increased dural thickening, meningeal infiltration, and progression around the surgical cavity. Colonoscopy and endoscopy were negative, but a biopsy of the right mandibular lesion confirmed metastatic gastric cancer. Immunohistochemistry on both specimens, the 2021 brain resection and the new bone biopsy, showed HER-2 negativity, PD-L1 combined positive score below 1, and microsatellite-stable disease. The HER-2 variability between the two metastases, 2+ without amplification in the brain and 1+ in the bone, underscores the potential for spatial and temporal intermetastatic heterogeneity and the need to interpret biomarker results in relation to assay limitations and specimen characteristics.

By January 2023, systemic progression involving lymph nodes, liver, and lungs was documented. Based on international guidelines available at the time, the patient’s tumor profile, and his clinical status, he received modified FOLFOX-4 chemotherapy plus denosumab from February to March 2023. Treatment initially produced symptomatic improvement, but after four cycles he developed pulmonary embolism and rapidly deteriorated. He was transitioned to palliative care at home and died in May 2023. The subsequent course illustrates how an apparently brain-only relapse may precede overt extracranial dissemination, a pattern with direct implications for how clinicians interpret solitary intracranial lesions in long-term cancer survivors.

To place the case in context, the authors performed a focused narrative review of the literature up to June 2026. Brain metastases from gastric cancer are historically reported in fewer than 1 percent of patients, while contemporary surgical cohorts after esophagectomy have described brain tumors or solitary brain metastases in approximately 2 to 3.4 percent of patients. These figures should not be directly equated, the authors caution, because the studies differ in tumor location, histology, stage, treatment, surveillance practices, and whether pathological confirmation was obtained. The distinction is particularly relevant for a Siewert type II tumor, which sits at the interface between gastric and esophageal classifications. A recent genomic analysis of 350 surgically treated gastroesophageal junction adenocarcinomas showed that Siewert types I and II are predominantly characterized by chromosomal-instability features and overlap more with lower esophageal tumors, whereas Siewert type III tumors share more features with gastric cancer. The literature confirms that solitary or isolated brain relapse is not unprecedented: the gastric literature includes two patients with disease-free intervals exceeding 120 months, a case of isolated brain metastasis 22 years after surgery, and a recent series of 12 patients with solitary intracranial relapse.

What could explain a thirteen-year silence? The authors point to tumor dormancy as one biologically plausible explanation. Dormant tumor cells can remain viable in a quiescent state, evading immune surveillance and resisting cytotoxic chemotherapy because they are not actively proliferating. Microscopic residual disease may also remain below the detection threshold of currently available imaging techniques. Reactivation of dormant cells may occur through incompletely understood mechanisms, including genetic and epigenetic alterations, angiogenic switching, immune escape, and interactions with the tumor microenvironment. The authors stress that this interpretation remains speculative in a single case and should not be considered proof of dormancy, but it provides a framework for understanding the unusually long interval between curative-intent treatment and metastatic relapse. Brain metastases are believed to spread predominantly through the bloodstream, although the underlying mechanisms remain incompletely understood, and evidence specifically on circulating tumor cells in gastric cancer brain metastases remains limited.

The practical message of the case is carefully calibrated. The authors argue that late metastatic recurrence should be retained in the differential diagnosis of a new symptomatic intracranial lesion in a long-term survivor, regardless of how many years have passed since curative treatment, and that tissue confirmation and integrated pathological assessment are essential when a new brain lesion appears many years later. At the same time, they are explicit that this observation does not imply a role for routine neuroimaging in asymptomatic long-term survivors, and that the available evidence does not support prolonged intensive surveillance. Because no specific guidelines exist for brain metastases from gastric cancer, multidisciplinary collaboration between neuro-oncology, neurosurgery, radiation oncology, pathology, radiology, and gastrointestinal oncology remains essential. For symptomatic solitary lesions, local therapy, whether surgery or stereotactic radiotherapy, remains central, while systemic therapy should be individualized according to extracranial disease status and biomarkers, with drug delivery across the blood-brain barrier and efflux transporters such as P-glycoprotein posing persistent challenges. The main limitations of the report are its single-case nature, the lack of archival primary-tumor tissue for repeat biomarker and genomic concordance testing, and the absence of paired comprehensive genomic profiling across all tumor sites. Further research, the authors conclude, is needed to define the biology, risk factors, and optimal management of brain metastases across the heterogeneous spectrum of gastric and gastroesophageal junction cancers.

Subject of Research: Oncology

Article Title: A solitary brain metastasis from gastroesophageal junction adenocarcinoma 13 years after curative multimodality treatment: a case report and review of the literature

Article References: Di Lorenzo, S., Giordano, P., Catapano, G., Giugliano, F. M., Federico, P., Frio, F., D’Antonio, A., Daniele, B., & Petrillo, A. (2026). A solitary brain metastasis from gastroesophageal junction adenocarcinoma 13 years after curative multimodality treatment: a case report and review of the literature. Frontiers in Oncology, 16, Article 1892637. https://doi.org/10.3389/fonc.2026.1892637

Image Credits: AI Generated

DOI: 10.3389/fonc.2026.1892637

Keywords: brain metastasis, gastroesophageal junction adenocarcinoma, tumor dormancy, late recurrence, methylome profiling, immunohistochemistry, gastric cancer, case report, solitary, brain, metastasis, gastroesophageal