Neuroimaging After Brain Radiotherapy and Radiosurgery

Radiation therapy is a cornerstone of modern neuro-oncology treatment protocols and is integral to the management of both primary brain tumors and metastases. Maximal surgical resection followed by a 6-week course of chemoradiation is standard of care for glioblastoma, and stereotactic radiosurgery (SRS) is widely used for brain metastases and certain benign lesions.¹ Advanced imaging, especially MRI, guides every stage of this process—from treatment planning (with MRI/CT fusion to define targets and spare normal structures) through follow-up—by precisely delineating targets and critical anatomy. In clinical practice, neuroimaging helps guide precise localization of treatment targets for radiation therapies and stereotactic interventions (eg, conformal intensity-modulated radiation therapy, proton therapy, stereotactic treatments) and is used repeatedly during follow-up to assess treatment response and detect complications.

Posttreatment MRI interpretation poses a fundamental challenge: therapy-induced changes can mimic tumor recurrence.² Conventional MRI criteria (lesion size and enhancement) are often misleading after radiotherapy. Nearly 50% of irradiated brain metastases transiently enlarge after SRS without true progression.³ Likewise, new or increased contrast enhancement may represent inflammation or necrosis rather than viable tumor. Pseudoprogression (an early transient increase in enhancement within ~3 months) typically stabilizes or resolves; radiation necrosis (a late delayed effect) may emerge many months later.⁴

To resolve these ambiguities, a multimodal imaging approach is often used: perfusion and permeability-based MRI sequences (dynamic susceptibility contrast [DSC] and dynamic contrast enhanced [DCE]) and diffusion imaging yield hemodynamic and cellularity metrics; magnetic resonance spectroscopy (MRS) probes metabolic changes; and positron emission tomography (PET)/single-photon emission CT with dedicated tracers (eg, amino acid, fluorothymidine F-18 [FLT], gallium-68–labeled DOTA-conjugated somatostatin receptor-targeting peptide [Ga-DOTA-SSTR]) can distinguish tumor from treatment effect.⁵ In practice, no single modality is definitive; therefore, radiologic findings are integrated with clinical data (eg, steroid response, timing) to guide management.

The following section addresses the spectrum of postradiation neuroimaging findings. Clinical integration is emphasized through standardized imaging protocols, multidisciplinary interpretation in equivocal cases, and imaging guidance for interventions such as biopsy and salvage therapy. Illustrative case examples are included. Future directions, including artificial intelligence–based analysis, imaging biomarkers, and personalized imaging strategies, are also reviewed.⁶

Brain Tumor Imaging After Radiation

Tumor response to radiation is traditionally assessed on MRI scans, using 2-dimensional (2D) measurements (eg, Macdonald criteria or Response Assessment Neuro-Oncology [RANO]) of contrast-enhancing lesions, using the sum of perpendicular diameters or product of diameters.^7,8 Volumetric (3-dimensional [3D]) segmentation of enhancing tumors may capture irregular shapes more accurately, but has traditionally been time-consuming. For glioblastoma, it has been thought that 2D and 3D measurements correlate strongly and lead to a similar response categorization.^9,10 A large study of recurrent glioblastoma cases treated with bevacizumab found no clear survival advantage to routine volumetric over 2D assessment.¹¹ In practice, therefore, the 2D RANO criteria (≥25% increase in enhancement to declare progression) remain commonly used. The modified RANO criteria allow for 3D volumetry data inclusion, and with the increasing availability of automatic tumor segmentation, 3D volumetry may become more routinely used in clinical practice.¹²

Pseudoprogression and Treatment Effects

Radiation often induces transient imaging changes that mimic tumor growth. Pseudoprogression is defined as an initial increase in enhancement or edema following therapy that stabilizes or regresses without new treatment.¹³ Pseudoprogression reflects treatment-induced processes (eg, gliosis, blood–brain barrier disruption) rather than true tumor proliferation. Pseudoprogression typically occurs shortly after therapy: in high-grade gliomas, it peaks ~3 months after radiation (almost always starting within ~3–6 months); in brain metastases, after SRS, it may appear later (median ~7–11 months).^13,14 By contrast, true progression is generally marked by continuously increasing enhancement after an interval of stability. Distinguishing between pseudoprogression and true progression is imperative. RANO guidelines for glioma allow suspect enhancement within 12 weeks of chemoradiation to be followed before declaring progression.

Imaging clues of pseudoprogression include heterogeneous or patchy enhancement, concomitant edema, and lack of a new mass effect; by contrast, true recurrence often causes mass effect and shows progressive nodularity. However, conventional MRI cannot reliably differentiate between the pseudoprogression and true recurrence, especially in a postradiation or post-SRS setting.¹⁵ Advanced MRI techniques (eg, perfusion, spectroscopy) or PET may be useful, but a newer MRI technique—contrast clearance analysis (CCA), which is discussed later in the article—shows particular promise.

Pseudoprogression has been best documented in malignant gliomas, but benign tumors can also exhibit transient enlargement after radiosurgery. For example, post-SRS increases in the volumes of vestibular schwannomas has been described, with an incidence of 14%–33%.^16,17 Meningiomas rarely show such effects (only ~6% transient increase in one series).¹⁸ In practice, any enlarging lesion within ~6 months of radiosurgery should raise the possibility of pseudoprogression. Short-interval follow-up (eg, repeat MRI in 2–3 months) is often recommended before assuming true failure, as pseudoprogression usually regresses or stabilizes on serial imaging.

Patterns of Recurrence

Tumor recurrence after radiation therapy usually follows predictable spatial patterns. High-grade gliomas (eg, glioblastoma) usually recur within the high-dose region (near the surgical cavity or resection bed).¹⁹ Historical series report that >80%–90% of glioblastoma failures are in-field (within the 95%–100% isodose line).²⁰ Marginal failures (straddling the edge of the treated volume) are less common. True distant recurrences (outside the radiation field) can also occur, especially in infiltrative tumors or if target margins are tight.

In postoperative radiosurgery for brain metastases, recurrences are also usually local: one study found that expanding the SRS cavity margin by only ~3 mm would capture 90% of local recurrences.²¹ In that series, the original SRS target overlapped a median 70% of the eventual recurrence, but a uniform 2.8-mm margin expansion would have covered 90% of the recurrent tumor. These data suggest that most local failures lie very close to the treated volume. Overall, recognizing the pattern of recurrence (local vs marginal vs distant) informs diagnosis. For example, a focal enhancement at the margin of a previous treatment field might represent a marginal recurrence, whereas multiple new enhancing nodules elsewhere may indicate dissemination. Areas of edema or enhancement outside the high-dose region raise concern for disease spread rather than treatment effect.

Glioma Recurrence. Temozolomide use, especially in the setting of IDH sequence variation, greatly affects the risk of glioma recurrence. IDH-variant tumors treated with temozolomide often lead to distant, disseminated disease (leptomeningeal) and hypermutation, resulting in high-grade transformation and poor outcomes, whereas IDH–wild-type glioblastoma usually recurs locally (in field or marginally), but with varied patterns depending on biomarkers such as EGFR amplification (which predicts marginal recurrence) and MGMT methylation (which predicts better temozolomide response or survival).²² In contrast, metastatic disease can manifest new lesions that are spatially and temporally separated. Thus, follow-up imaging must survey the whole brain: local in-field regrowth must be differentiated from a new metastatic deposit or leptomeningeal spread. Specific imaging intervals are discussed later in the article.

Contrast Clearance Analysis

Conventional MRI scans often cannot be used to resolve the ambiguity between recurrence and radiation effect. CCA MRI has emerged as a powerful tool in this situation. CCA (also called delayed contrast enhancement MRI or treatment response assessment maps) involves acquiring early and late (often 1-hour postcontrast) T1-weighted sequences and automatically subtracting them. Viable tumor tends to wash out contrast quickly, whereas treatment-induced necrosis or edema retains contrast longer.²³ The resulting map highlights areas of persistent contrast (often color-coded) that correlate with active tumor (Figure 1).

CCA has shown excellent diagnostic performance. In one prospective series of 32 treated lesions, CCA correctly differentiated radiation necrosis or pseudoprogression from true progression with 93% sensitivity and 78% specificity (overall accuracy 84%).¹⁵ A recent systematic review of 9 studies (407 lesions) found a pooled CCA sensitivity of 91% and specificity of 92%.²⁴ These figures exceed those of standard MRI sequences and compare favorably with perfusion and spectroscopy. Histologic studies have validated CCA results: in mixed cohorts of gliomas, metastases, and atypical meningiomas, delayed enhancement patterns on CCA correlated with pathology in 100% of cases.²³ In practical terms, CCA is relatively easy to implement on modern scanners (ie, acquiring a second T1 sequence ≥60 minutes postcontrast) and can quickly produce heat maps of tumor activity without the need to use any radioactive isotopes.

Pseudoprogression: Definition and Time Course

Pseudoprogression is formally defined as treatment-related (nontumoral) enhancement that mimics growth, followed by stability or improvement. It is a subacute phenomenon, often occurring within 3–6 months after therapy. After chemoradiation for glioblastoma, up to ~30% of individuals may show new or increased enhancement at 1–3 months, which then regresses without additional therapy.¹⁸ Pseudoprogression is thought to reflect inflammation and vascular changes (eg, MGMT-methylated tumors show more pseudoprogression as a favorable response to therapy).¹³ By convention, any new enhancement within ~12 weeks of chemoradiotherapy is approached with caution and RANO criteria for high-grade glioma require confirmation of progression on a subsequent scan to avoid mislabeling pseudoprogression as true progression.

True progression implies active tumor proliferation, whereas pseudoprogression is a transient adverse radiation effect. Key differences include clinical context and time course: pseudoprogression usually occurs while therapy is ongoing or shortly after completion, and often in the absence of neurologic decline; true progression more often appears months later or is associated with clinical worsening. Perfusion MRI (in which relative cerebral blood volume is elevated in tumor vs low in radiation necrosis) and MRS (which shows an elevated choline peak in tumor) have been used to distinguish between the two, but are analysis- and threshold-dependent. CCA is a much more sensitive and specific technique that can be used in routine clinical neuro-oncology practice.

Clinical Implications

Accurate interpretation of postradiation imaging findings is crucial. Failing to identify pseudoprogression or overlooking true recurrence can have detrimental results (eg, premature cessation of effective therapy, unnecessary surgery or radiation, delay in second-line or salvage treatments). Advanced imaging techniques, such as CCA, improve diagnostic confidence and directly affect management (eg, individuals with confirmed treatment effect may be observed or given steroids or antiangiogenics; those with true recurrence may undergo re-resection, re-irradiation, or a change in systemic therapy). Owing to the high accuracy of CCA, its use may prevent invasive biopsies or needless treatment changes.^15,23

Imaging surveillance after brain tumor radiation must integrate both conventional and advanced techniques. Linear measures remain the mainstay for follow-up, but volumetric techniques using automated segmentation tools help identify tumor shifts and growth, especially for benign lesions. Pseudoprogression is a well-recognized, transient phenomenon (especially within 3–6 months), whereas local in-field recurrence is the norm for malignant gliomas. Emerging tools such as CCA greatly enhance our ability to distinguish these entities, optimizing subsequent clinical decisions.

Imaging Findings After Functional Radiosurgery

Trigeminal Neuralgia

Individuals with trigeminal neuralgia (TN) have a high incidence of vascular involvement, or vascular compression, of the trigeminal nerve (cranial nerve [CN] V). Bora et al²⁵ report the incidence to be 88.85% based on a systematic review of literature. Both TN itself²⁶ and previous treatments for TN that target the ganglion, such as rhizotomy,²⁷ can result in atrophy and thinning of the CN V trunk in the cisternal portion. The classic MRI finding after radiosurgery for TN, which involves the delivery of 70–90 Gy typically to the midcisternal portion of CN V, is focal enhancement on gadolinium administration (Figure 2).^28,29 Both the timing of onset of this enhancement and its persistence vary, with no clear dose dependency or clinical correlation being described to date.

Trigeminal nerve tractography with diffusion tensor imaging (DTI) was performed by Hodaie et al,³⁰ who found that DTI accurately detected the radiosurgical target. Radiosurgery resulted in a 47% drop in fractional anisotropy (FA) values at the target with no significant change in FA outside the target, demonstrating highly precise changes after treatment. Radial diffusivity but not axial diffusivity changed markedly, suggesting that radiosurgery primarily affects myelin. Tractography was more sensitive than conventional gadolinium-enhanced posttreatment MRI, as FA changes were detected regardless of trigeminal nerve enhancement. In participants with long-term follow-up, recovery of FA and radial diffusivity correlated with pain recurrence.

Thalamotomy and Brain Lesioning

Brain lesioning involves some of the highest prescribed doses in radiosurgery and radiation oncology in general, averaging ~130 Gy in a single session. The intent of these doses is to produce focal necrosis. Imaging findings can be divided into the expected effects of treatment and findings associated with adverse effects.

MRI-guided radiosurgical thalamotomy is not new,³¹ but quantification of FLAIR changes and response to treatment has been documented inconsistently. We have observed that FLAIR changes at the site of lesioning among our patients undergoing thalamotomy reach a peak at an average of 8.9 months after treatment, and then slowly dissipate in all cases (unpublished data). FLAIR changes measuring 4 mm in diameter represented a threshold for tremor response; smaller FLAIR changes were not associated with satisfactory tremor response.

Ohye et al³² described 2 types of thalamic lesions after SRS: ring enhancement (a clearly defined lesion with ring-shaped contrast uptake) and a more diffuse lesion with some extension into parts of the internal capsule. Imaging changes were seen at an average of 3–5 months after thalamotomy and seemed to bear an inconsistent relationship to delivered dose, tremor response, and the onset of complications.

Friehs et al³³ described a tendency toward smaller lesions in women.

Duma et al^34,35 found more signal changes on T2-weighted sequences with higher doses. Friehs et al³³ observed this association as well, with doses >160 Gy associated with lesions >2000 mm^3.

Dose rate has been related to lesion volume: before replenishment of Co-60 sources, lesions averaged 311 mm³ vs 175 mm³ after replenishment (P=.041).³²

As observed among our patients, extensive changes in T2-weighted and FLAIR sequences affecting the thalamus or internal capsule and adjacent brain have been associated with higher complication rates^31,36-38; these complications may be serious,^39-41 or asymptomatic.^42,43 Young et al⁴⁴ reported larger lesions (871 mm³) in individuals with complications compared with the overall mean of 188 mm³ in 157 cases (P<.001). MRI findings may be present in the absence of complications.^43,44 Among our patients, lesion sizes in individuals with response as well as adverse effects were larger in volume than those reported by Young et al⁴⁴ (unpublished data).

DTI provides an elegant demonstration of the dentato-rubro-thalamic tract (Figure 3), but it remains a subjective technique that varies based on the parameters used to generate the images. In the authors’ experience, fiber tracts were consistently identified in individuals before tremor and were found to be diminished or absent after a successful thalamotomy. The extent of disruption of the dentato-rubro-thalamic tract has correlated with clinical tremor response among our patients.

Radiation-Induced Changes in Normal Brain Tissue

Photon beam irradiation of intracranial targets inevitably results in exposure of adjacent normal brain tissue. This exposure leads to a spectrum of structural, vascular, and metabolic changes that evolve over time and can be effectively characterized using advanced neuroimaging techniques.

Pathophysiology of Radiation-Induced Brain Injury

Radiation injury to normal brain tissue is multifactorial. Endothelial damage plays a pivotal role, leading to blood–brain barrier disruption, ischemia, and subsequent possible chronic hypoxia. These vascular insults trigger inflammatory cascades that exacerbate tissue injury. White matter is particularly vulnerable due to the radiosensitivity of oligodendrocytes, resulting in impaired myelin maintenance and axonal degeneration.^45,46

Temporal Evolution of Radiation Injury

Radiation-induced injury follows a characteristic temporal course, broadly classified into acute, early delayed, and late delayed effects. Acute and early delayed effects occur within days to months following radiation exposure and are primarily driven by vascular permeability alterations and cellular injury. Vasogenic edema is typically demonstrated as hyperintensity on T2-weighted and FLAIR sequences, whereas radiation-induced endothelial injury may lead to cytotoxic edema, which is detectable on diffusion-weighted imaging (DWI). DWI is particularly valuable in this context, as it provides sensitive insight into alterations in tissue microstructure associated with cellular injury.⁴⁷

Neuroimaging Features of Radiation Injury

DTI provides quantitative markers of microstructural integrity. Radiation-induced white matter damage is reflected by decreased FA and increased mean diffusivity, consistent with demyelination and axonal loss. Perfusion-weighted imaging and susceptibility-weighted imaging demonstrate radiation-related vascular changes, including reduced cerebral blood volume and microhemorrhages, respectively.⁴⁸

Metabolic alterations are evident on MRS, with reductions in N-acetylaspartate (NAA) serving as a marker of neuronal dysfunction or loss. These advanced imaging techniques offer complementary insights into the structural, vascular, and metabolic consequences of radiation therapy.⁴⁹

Radiation Necrosis

Radiation necrosis represents a severe late complication of cranial irradiation and is characterized by irreversible injury to normal brain parenchyma. The underlying pathophysiology involves progressive endothelial damage, resulting in ischemia, disruption of the blood–brain barrier, and a sustained inflammatory response, ultimately culminating in coagulative necrosis.

On contrast-enhanced MRI, radiation necrosis most commonly presents as heterogeneous or irregular peripheral enhancement with central necrosis and surrounding vasogenic edema; however, enhancement patterns are variable, and may also appear nodular, patchy, or ill-defined.⁵⁰

These imaging features substantially overlap with those of tumor recurrence, posing a considerable diagnostic challenge. In this context, the individual’s treatment history and the temporal relationship between lesion development and previous radiation exposure serve as key diagnostic indicators, as radiation necrosis typically manifests months to years following therapy. Advanced imaging techniques such as perfusion MRI and MRS may improve diagnostic confidence, but no single imaging modality is definitive.⁵¹

Advanced and Ancillary Imaging Modalities and Techniques

Advanced MRI: Perfusion Images for Vascularity

Imaging is an important part of defining normal and pathologic tissue microvasculature, including aberrant vessels and defects in the blood–brain barrier.⁵² The 3 main measurable parameters are flow (F), contrast transfer coefficient (ktrans), and relative cerebral blood volume.⁵³ These parameters are assessed on perfusion MRI studies, such as DSC, DCE, and arterial spin labeling. Regional blood flow in the brain (F) is typically measured with arterial spin labeling, ktrans with DCE, and relative cerebral blood volume with DSC. These methods are often helpful in differentiating benign extra-axial tumors from intra-axial malignancies and in assessing treatment response.^53-55

Advanced MRI: Diffusion (DWI, ADC) for Cellularity

DWI is able to provide functional and ultrastructural data on tumor cellularity and the tumor microenvironment by measuring water mobility through tissues.⁵⁶ The restriction to water movement between densely packed tumor cells results in a low apparent diffusion coefficient (ADC); this often helps determine the grade and type of tumor.⁵⁶ Logically, therefore, higher-grade tumors would have a lower ADC signal. DWI and ADC also correlate with the IDH variation status of gliomas.⁵⁶ These studies are helpful in analyzing primary brain tumors, differentiating gliomas from CNS lymphoma, and are useful in the setting of benign tumors, such as meningiomas and schwannomas.⁵⁶ For brain metastases, DWI restriction can vary by primary tumor type. For example, adenocarcinomas are typically hypointense, whereas small cell neuroendocrine tumors can be hyperintense.⁵⁶

In a posttreatment scenario, an increase in diffusion restriction will favor increased cellularity, representing recurrence of the tumor. In combination with MRS, DWI and ADC increase the sensitivity and specificity for differentiating true tumor recurrence from pseudoprogression.⁵⁷

Advanced MRI: MRS for Metabolic Profile

MRS is a useful tool for evaluating the presence and concentration of different metabolites within the tumor and normal brain tissue and serves as an auxiliary technique to other MRI sequences. Choline (Cho) is involved in the synthesis and degradation of the cell membrane; an elevated Cho peak represents increased membrane turnover, such as would occur in rapidly dividing brain tumor cells. In contrast, NAA is present in normal neuronal tissue.⁵⁸

In gliomas, due to neuronal loss and high tumor cell turnover, the NAA peak will usually be diminished and the Cho level increased. Lipid and lactate levels also increase due to anerobic metabolism in the tumor cells. This results in high Cho/NAA and Cho/creatinine ratios, with elevated peaks for lipids and lactate. In contrast to this scenario, in tissue that is exhibiting postradiation effects, NAA, Cho, and creatinine all will be low.⁵⁸

Positron Emission Tomography

FDG-PET: Metabolic Activity. [¹⁸F]fluorodeoxyglucose (FDG) is a well-established and the most widely used tracer for PET imaging, with tracer uptake corresponding to the level of glucose metabolism. Uptake is elevated in tumors due to overexpression of the glucose transporter-1.⁵ Because the main metabolic substrate for neurons is glucose, brain tissue is normally FDG-avid, interfering with metabolic signaling of lesions and limiting its interpretation and utility.⁵ The development of amino acid tracers has helped overcome this limitation and increase the capacity for interpretation in neuro-oncology, and amino acid tracers are recommended for use in assessing brain tumors by RANO.

Amino Acid PET (FET, MET, FDOPA): Tumor vs Necrosis Differentiation

Radiolabeled amino acids, such as [¹⁸F]fluorodopa (FDOPA), [11C]methionine (MET), and [¹⁸F] fluoroethyl (FET), target L-amino-acid transport receptors and are used for imaging gliomas. Both [¹⁸F]FDG and amino acids are able to cross the blood–brain barrier and can be used to image brain metastases.5 [¹⁸F]FET-PET can provide reliable response assessment, whereas [¹⁸F]FDOPA has greater value in differentiating recurrence from radiation necrosis.⁵

68Ga-DOTATATE PET for Atypical Meningioma Recurrence Detection

In 68Ga-DOTATATE, the radionuclide gallium-68 binds to a somatostatin receptor analog, resulting in physiologic uptake in the pituitary gland. This technique is exquisitely sensitive for the imaging of meningiomas due to the high expression of somatostatin receptor subtype 2 (SSTR2), and is especially useful for identifying residual or recurrent tumor after surgery or radiation.^5,59,60 The SUVmax level has been correlated with meningioma growth rate and with biologic response to therapy.⁶¹ Despite its high sensitivity, DOTATATE PET has low specificity, showing uptake in other pathologies, including metastatic disease. The SUVmax level, however, is often lower in metastases than in meningiomas, with the exception being renal cell carcinoma, which has uptake similar to meningiomas.^62,63

PSMA PET

Prostate-specific membrane antigen–targeted molecular imaging with PET (PSMA PET) targets PSMA produced during the neovascularization of tumors. Despite being commonly used to identify recurrence or metastasis of prostate cancer,⁶⁴ it is also useful to identify other pathologies, including high-grade gliomas and meningiomas.^65,66 PSMA PET can be used not only for diagnosis but also for treatment planning and the evaluation of treatment response.⁶⁷

Radiomics and Machine-Learning Approaches

With the expansion of artificial intelligence into routine medical practice, radiomics has become part of radiology assessments. Radiomics models are able to extract and collate data from multiple imaging modalities, going beyond human analysis of scans. The workflow includes several steps that obtain information by mathematically predefined (feature-based radiomics) or automatically extracted (learned) parameters from the input images (ie, deep learning–based radiomics). Radiomics can be used to identify biomarker expression (such as IDH and MGMT methylation status), assess response, and help differentiate true progression from pseudoprogression in neuro-oncology, supporting clinical decision-making.⁶⁸

Hybrid PET/MRI

Although not widely used, a hybrid system combining PET with MRI can produce images with higher anatomic resolution, including high soft-tissue resolution and functional MRI capability, providing both anatomic and molecular information. These characteristics make hybrid PET/MRI suitable for neuro-oncologic applications.⁶⁹ The advancement of molecular analysis in oncologic diagnosis and classification has also enabled its use for tailored targeted therapy and radiologic scans. Targeted imaging allows clinicians to predict response early after treatment with a greater level of precision.⁷⁰

Clinical Integration

Effective postradiation management requires integrating imaging with clinical context. Standardized MRI protocols and RANO criteria guide surveillance. For high-grade gliomas, the first posttreatment MRI is typically performed at ~2–6 months; if early enhancement appears, a short-interval rescan (4–8 weeks) is used to distinguish true progression from pseudoprogression.

• Glioma surveillance: Glioblastoma is monitored with MRI every 2–6 months after therapy. Multidisciplinary teams integrate clinical data (eg, neurologic examination, steroid use) and advanced imaging (eg, perfusion, spectroscopy, CCA, PET) to distinguish recurrence from treatment effect.
• Metastasis surveillance: After SRS for brain metastases, whole-brain MRI is recommended. National Comprehensive Cancer Network and National Institute of Clinical Excellence guidelines advise MRI every 2–3 months for the first 1–2 years, then every 4–6 months thereafter. Enhancing nodules at the margin of the treated area are evaluated with contrast clearance, perfusion MRI, or amino acid PET to differentiate marginal recurrence from radionecrosis.
• Benign tumor surveillance: Vestibular schwannomas are followed with annual MRI for ~5 years, then every 2–3 years if stable. Individuals should be informed that transient enlargement (ie, pseudoprogression) occurs in ~20%–50% of cases. Any suspected growth should be reimaged in 3–6 months with thin-slice volumetric MRI before considering further treatment.

Conclusions

Postradiation neuroimaging is essential for guiding neuro-oncologic care. Serial MRI (anatomic, perfusion, diffusion, and spectroscopy) integrated with clinical findings enables differentiation among recurrence, pseudoprogression, and radiation necrosis. Standardized protocols and multidisciplinary review optimize the timing of additional treatment or observation. Tailored imaging follow-up and collaborative interpretation improve decision-making and ultimately enhance patient outcomes. Recognizing pseudoprogression can spare individuals from unnecessary changes in therapy.