This review was written following PRISMA 2020 guidelines for scoping reviews (17). Databases searched included PubMed, SCOPUS, and Cochrane. The search terms included: “CEUS AND spine”, “CEUS AND spinal tumor”, “CEUS AND spinal cord”, “CEUS AND intramedullary spinal neoplasm”, “iCEUS AND spine”, “iCEUS AND spinal tumor”, “iCEUS AND spinal cord”, “iCEUS AND intramedullary spinal neoplasm”. The database search started in late December 2024 cand continued into late March 2025, with the last database search occurring on March 20^th, 2025. Articles were found using the keywords listed in each database, after which duplicates were excluded. Articles were then screened for their relevance to CEUS in spinal tumor resections. After the screening process, letters to the editor, conference comments, and systematic reviews were deemed ineligible. Due to the novelty of the subject, case reports and case series were included. English as a primary language was a requirement for selection. Eligibility criteria were not affected by the article’s publishing date. The PRISMA flowchart diagram illustrates our search, screening, and inclusion of reports (Figure 1).

Articles were screened for their relevance to the topic of CEUS use in spinal tumor resections regarding human or animal participants. Excluded articles predominately highlighted non-oncological applications of CEUS, systemic oncological applications of CEUS, and standard or doppler ultrasound imaging of spinal tumors. The screening process included two authors in an independent review. Disagreement on article selection was discussed and, if needed, a third author was consulted to mitigate bias in the final selection.

Data collection was performed by four authors. Articles were examined for the study type, tumor pathology, tumor location, and primary intraoperative CEUS findings. Enhancement characteristics, perfusion patterns, ease of delineation, and vascular quality for each tumor pathology were also recorded. Joanna Briggs Institute (JBI) assessment tools for case reports, case series, and cohort studies were used to evaluate the quality of studies and the risk of bias (18). Due to the limited quantitative data and heterogeneity of data reported in the included studies, the data analysis remained qualitative. No formal quantitative analysis was able to be performed.

Of the 150 articles found in our literature search, 71 of them were unique. Once the articles were assessed for relevance to CEUS and spinal tumors, 11 reports remained for screening. We were left with 8 papers – 3 case reports, 3 case series, 2 retrospective cohort studies – after systematic reviews and comments and letters to the editor were removed (Table 1). A total of 76 tumors were resection with CEUS guidance, 74 of which were intradural intramedullary and 2 were intradural extramedullary. No extradural tumors were reported in the literature. Tumor type varied among all studies. The three most common types were ependymomas (29), hemangiomas (10), and glioblastomas (8).

Enhancement characteristics, margin delineation, and perfusion pattern were summarized for each tumor pathology (Table 2). As illustrated in Figure 2, CEUS provides superior distinction of tumor boundaries compared to standard B-mode imaging. Well-defined borders were observed in the following pathologies: ependymomas, subependymomas, astrocytomas, pilocytic astrocytomas, neurocytoma, hemangioblastoma, hemangiopericytoma, and schwannomas. Poorly defined or blurry borders were seen with low grade gliomas, anaplastic gliomas, glioblastomas, cavernous hemangiomas, and anaplastic astrocytomas. Enhancement intensity ranged from homogenous in hemangioblastomas and hemangiopericytomas to variable in ependymomas, pilocytic astrocytomas, and subependymomas. Heterogenous enhancement was observed with hemangiopericytomas and glioblastomas, while cavernous hemangiomas showed minimal or absent enhancement. Dotted, speckled patterns were observed with ependymomas, subempendymomas, schwannomas, and low-grade gliomas. Perfusion time also varied by pathology, with rapid peak enhancement noted in hemangioblastomas, hemangiopericytomas, anaplastic astrocytomas, and glioblastomas. Slower perfusion time was observed anaplastic gliomas, pilocytic astrocytomas, subependymomas, schwannomas, neurocytomas, and low-grade gliomas. Notably, perfusion time within ependymomas varied. Vascular structures, including feeder arteries, draining veins and peritumoral cysts, were most frequently visualized in hemangioblastomas, hemangiopericytomas, ependymomas, pilocytic astrocytomas, anaplastic astrocytomas, and glioblastomas.

JBI appraisal tools for case reports, case series, and cohort studies were used to evaluate the quality of the included studies (Table 3). The mean (SD) scores for the case reports and case series 7.0/8 (0.00) and 8.3/10 (0.47), respectively, indicating good methodologic quality. Case reports were limited in their inclusions of potential adverse effects, even though no adverse effects were reported. Case series were unclear regarding their description of patient inclusion, and some lacked quantitative analyses. Conversely, the mean (SD) score for the cohort studies was 5.0/11 (0.00), indicating low-moderately quality. These studies were restricted by their single-arm nature, which limited the applicability of group selection and controlling for confounding variables. Additionally, short follow-up lengths of less than 3 months limited the ability to evaluate long-term neurological recovery. Due to the limited reports of the CEUS for spinal tumor resection, the cohort studies were elected to be included.

US imaging is an indispensable tool in spinal surgery, applied across a wide range of conditions, including degenerative diseases, trauma, deformities, and tumors (1, 2). Building on this foundation, CEUS has emerged as a novel modality that not only improves the visualization of vascular structures and tumor margins by leveraging the unique echogenic properties of microbubble contrast agents, but also utilizes the differential echogenicity between tumor tissue and normal spinal cord—particularly in cases where extradural and intradural tumors lack distinct margins—to confirm adequate decompression and guide complete resection (12, 13, 19, 20).

Crucially, CEUS addresses several key limitations of iMRI. While iMRI remains the gold standard, it provides only a static snapshot of the anatomy, which can be rendered inaccurate by cerebrospinal fluid loss and tissue shift (‘brain shift’) after dural opening (6–8). Furthermore, iMRI significantly prolongs operative time and requires expensive, specialized infrastructure (7). In contrast, CEUS offers a real-time, dynamic assessment of the resection bed that is unaffected by tissue shift, easily repeatable throughout the surgery, and cost-effective to implement in standard operating suites.

CEUS enables detailed visualization of tissue perfusion by using microbubble contrast agents that are less than 10 microns in diameter—small enough to navigate capillaries without extravasating into surrounding tissues (21). When the ultrasound probe emits a frequency that resonates with these microbubbles, they cyclically expand and contract, generating echoes that are captured by the probe’s piezoelectric crystals (22). This process produces a non-linear echo signal that is distinct from the surrounding tissue, thereby enhancing the signal-to-noise ratio and delineating perfused soft tissue structures with precision (23). Such enhanced imaging is potentially instrumental in confirming adequate decompression and ensuring complete tumor resection (24).

This capability, combined with the established use of conventional ultrasound in managing degenerative, traumatic, and deformity-related spinal conditions, underscores the growing application and clinical impact of CEUS in spinal oncology over the past five years (1, 2, 9, 16). Thus, this narrative review of available literature details the studies that have utilized CEUS to aid in spinal tumor dissection. Specially, we highlight the advantages CEUS brought to each case and summarize tumor characteristics observed on CEUS for the various pathologies.

Vetrano et al. reported one of the first applications of CEUS in spinal tumor resection in a 14-year-old patient with suspected diffuse infiltrative intramedullary lesions (19). Preoperative MRI with gadolinium revealed three nodular lesions at C7, T1, and T4, all with cystic capsules and diffuse enhancement between them. The radiological findings initially suggested a tumor matching the characteristics of an astrocytoma. However, intraoperative CEUS revealed three cystic lesions without any enhancement in the intramedullary space between the lesions, differing from the preoperative MRI. CEUS also displayed a vascular peduncle located on the C7 lesion, protruding interiorly. These characteristics obtained by CEUS suggested a diagnosis of hemangioblastoma, which was later confirmed by pathology. Moreover, CEUS delineated the perfusion patterns of each lesion, allowing the surgical strategy to be modified for complete resection.

In a case series of 14 patients, Han et al. used CEUS to aid in the resection of 4 astrocytomas, 5 ependymomas, 1 anaplastic astrocytoma, and 4 cavernous hemangiomas (25). The authors stated that CEUS greatly improved visualization of tumor margins and perfusion patterns, which limited the extent of the initial durotomy. Such findings were especially true for the ependymomas and astrocytomas, which are generally poorly delineated. Of note, the authors found that there were varying degrees of enhancement of cavernous hemangiomas on CEUS; however, no further details of tumor characterization on CEUS were reported. Regardless, CEUS still allowed for the differentiation between cystic and solid tissues, displaying as hypoechoic and hyperechoic, respectively. After each resection, the authors performed CEUS imaging a second time to identify any residual tumors, confirming GTR along with favorable neurological outcomes at a mean follow-up of 15.9 months. These findings underscore CEUS’s additional value in detecting residual tumor. While long-term recurrence data in these specific studies is limited, the ability of CEUS to facilitate GTR, a well-established prognostic factor for progression-free survival, suggests a theoretical potential to reduce recurrence risk (1, 2, 4, 5).

Barkley et al. characterized several types of intramedullary tumors on CEUS in a seven patient case series consisting of three hemangioblastomas, two ependymomas, one subependymoma, and one pilocytic astrocytoma (9). Both hemangioblastomas and ependymomas displayed considerable contrast enhancement on CEUS, while there was a lack of contrast enhancement when imaging the pilocytic astrocytoma and subependymoma. However, the authors noted that even without contrast enhancement of the tumors, specifically within the pilocytic astrocytoma, the physiological uptake of contrast agent in the surrounding healthy tissue alone, which was starkly different from the tumor, sufficiently delineated the tumor margins. Additionally, in the hemangiomas, identifying the vascular supply with CEUS allowed for early coagulation and minimized intraoperative blood loss. CEUS also improved the delineation of the vascular pedicle of the ependymoma from the adjacent spinal cord tissue. Similar to Han et al., CEUS imaging was performed a second time, after each resection, indicating total resection was achieved (25). Overall, the authors demonstrated that intraoperative CEUS aided in resection of all of the reported tumors, especially the more vascularized tumor types, even without full contrast enhancement present.

In 2021, Vetrano and colleagues published an additional case series that further characterized intramedullary tumors on CEUS (3). Over a four-year period, the authors used CEUS to aid in the resection of two pilocytic astrocytomas, one anaplastic astrocytoma, one glioblastoma, four ependymomas, one subependymoma, two hemangioblastomas, and one neurocytoma. All tumors, except for the glioblastoma and anaplastic astrocytoma, displayed sharp margins on CEUS that delineated the tumor from surrounding spinal cord tissue. Additionally, the hemangioblastomas, glioblastoma, and anaplastic astrocytoma all rapidly enhanced upon contrast administration, whereas the pilocytic astrocytomas, subependymoma, and neurocytoma were slower to reach peak enhancement. The enhancement time of the four ependymomas varied from slow to rapid. Of note, a hypoechoic peritumor cyst in one of the ependymomas was found on CEUS, which contradicted the preoperative MRI findings and confirmed a syringomyelic dilation. In one pilocytic astrocytoma, CEUS findings suggested an expansion of the initial durotomy, which the authors noted was an influential step to achieve total resection. The authors were also able to capture tumor vascular patterns on CEUS that did not display any obvious patterns. This study suggested that the peak constant enhancement measured by CEUS may be related to the degree of vascularity within the tumor or arteriovenous shunts and venous engorgements that are uniquely influenced by mass effect and pathological angiogenesis. Furthermore, this study reinforces the concept that tumor delineation is inversely related to the invasive nature of the tumor type.

Mazzapicchi et al. used CEUS for resection of four intramedullary lesions as part of a comprehensive retrospective cohort on cranial and spinal resection imaging techniques for hemangioblastomas (16). All four hemangioblastomas were characterized as nodular and homogenously hyperechoic, with cystic components. Two of the tumors were noted to be deeply embedded. In such cases, CEUS helped with localization and surgeons were able to adapt the extent of the myelotomy in real-time. The CEUS complemented well with standard B-mode and Doppler imaging to provide dynamic guidance for resections. However, the authors note CEUS may be limited by a considerable learning curve for anatomical orientation, as well as the possibility of artifact creation by post-resection bleeding and/or hemostatic agents.

Lastly, an additional retrospective cohort study was performed by Han et al. that evaluated contrast perfusion patterns in 36 high grade gliomas: one pilocytic astrocytoma, seven low-grade gliomas, three anaplastic gliomas, seven glioblastomas, and 18 ependymomas (26). Overall, GTR was achieved with CEUS in a majority of cases, which was improved upon compared to previous resection reports at the author’s center (27). The visualization of tumor margins and cystic components was greatly enhanced by CEUS in the ependymoma and pilocytic astrocytoma cases. CEUS was also able to delineate low-grade gliomas, which generally have poor margins and enhancement on MRI. Additionally, while the parenchyma of the glioblastomas and anaplastic astrocytomas did not show obvious boundaries on CEUS, the high-grade gliomas displayed a higher degree of vascularization than surrounding tissues. High-grade gliomas also displayed a greater peak contrast intensity and were faster to reach peak intensity compared to low grade gliomas. While these findings were not statically significant, they agree with the diffusion patterns presented by Vetrano et al. and thus suggest a potential link between contrast perfusion patterns and tumor vascularity (3). CEUS was ultimately able to visualize the vascularization of the gliomas to guide resection and avoid unexpected blood loss or operative complications.

Vetrano et al. demonstrated the advantages of CEUS compared to MRI and standard B-mode ultrasound imaging in the case of a 33-year-old female patient with dorsal Schwannoma (20). Although preoperative MRI suggested an intramedullary lesion, standard B-mode imaging showed a hyperechoic mass with poorly defined borders that did not clearly separate the tumor from the spinal cord. However, upon CEUS imaging, an extramedullary tumor was revealed, distinct from the surrounding perilesional cysts and the edematous spinal cord. Techniques were adjusted, and complete resection was confirmed by CEUS. The authors theorized that the combination of the perilesional cysts, considerable edema, and cord compression may be responsible for the intramedullary preoperative diagnosis. This case highlights the ability of CEUS to improve surgical planning and outcomes by characterizing spinal tumors to a high degree of specificity and by protecting against MRI misinterpretations.

Della Pepa et al. described the utility of using CEUS during resection of highly vascularized tumors. In one case specifically, a 61-year-old male with recurrent dorsal hemangiopericytoma, confirmed as an intradural extramedullary lesion, underwent a T12 laminectomy for resection (13). Here, real-time CEUS was pivotal in revealing the vasculature architecture of the tumor before durotomy. With visualization of the tumor’s feeder arteries and venous drainage, the authors reported a reduction blind surgical maneuvering, avoided damage to the displaced nerves, and achieved GTR. Furthermore, additional CEUS imaging following the initial dissection identified a residual tumor, which was removed and later confirmed to be pathologic. This case demonstrates how the visualization of the vasculature by CEUS pre- and post-operatively can be a useful tool to ensure safe total resections in vascularized intradural extramedullary tumors.

Extradural extramedullary tumors are outside of the dural layer, and to our knowledge, no published reports have described the use of CEUS for resection of extradural extramedullary tumors. However, based on insights gained from these intradural cases, we speculate that CEUS could be a valuable adjunct for visualizing vascular architecture and delineating tumor margins in extradural lesions, potentially aiding in achieving complete resection. For extradural masses, such as meningiomas, schwannomas, and chordomas, while often not involving the spinal cord, they often lead to nerve root compression. Like intradural masses, CEUS could allow valuable imaging that can help with safe resection without injuring adjacent nerve roots. Moreover, assessment of perfusion parameters acquired from CEUS can provide valuable insight into characteristics of the tumor and assist in diagnosis and prognosis. Much like its use in hepatology, where CEUS has been shown to have high specificity for diagnosing liver nodules as hepatocellular carcinoma, CEUS has the potential to provide diagnostic value to undiagnosed extradural masses such as metastatic processes and primary cancers of the spine. As we continue to utilize both ultrasound and CEUS for the spine and spinal cord, its application to extradural masses will surely be realized.

While such studies highlight the potential of CEUS within spine tumor resections, there are still limitations of the technique. First, despite the high resolution provided by modern high-frequency transducers, the physics of ultrasound waves results in limited penetration depth. This can impede the visualization of deeply seated ventral tumors or lesions obscured by complex bony anatomy, necessitating careful preoperative planning regarding the surgical window. The imaging scope is further restricted by the initial approach; the spinal cord must be exposed from the overlying laminae to provide access for the ultrasound probe. Therefore, an insufficient laminectomy may prevent full tumor visualization by CEUS.

Secondly, the application of intraoperative transducers entails technical challenges and potential risks related to their introduction into the surgical cavity. The physical footprint of the probe requires a sufficiently large laminectomy for adequate contact (3, 20), and the manipulation of the device in close proximity to the exposed spinal cord carries a risk of mechanical injury, requiring a steady hand and significant operator experience.

A third significant drawback is the occurrence of acoustic artifacts. Acoustic enhancement artifacts, often caused by the difference in signal attenuation between the physiological saline in the resection cavity and the surrounding spinal tissue, can create hyperechoic noise. As Mazzapicchi et al. noted, these artifacts can be observed from post-resection bleeding or with the presence of hemostatic agents (16). These artifacts may obscure residual neoplastic fragments, particularly along the resection cavity walls, thereby complicating the assessment of GTR.

Furthermore, standard CEUS relies on the interpretation of 2D cross-sectional images, which can be difficult to mentally reconstruct into a 3D volume, especially within a confined surgical approach. Unlike iMRI, there is currently a lack of commercially available 3D neuronavigation systems fully integrated with CEUS. This absence diminishes the ability to precisely map the volumetric extent of diffuse or infiltrative tumors in stereotactic space.

Moreover, CEUS is highly operator-dependent. The acquisition of high-quality images and the real-time interpretation of perfusion patterns require a specialist proficient in ultrasound diagnostics. This learning curve regarding anatomical orientation and image interpretation may hinder the widespread adoption of the technique in centers without dedicated training or access to an experienced sonographer within the surgical team (3, 16, 26).

Finally, the benefits of CEUS for tumor margin visualization may be limited by tumor pathology. As previously mentioned, margins of high-grade glioblastomas and anaplastic astrocytomas were reported to be poorly visualized in some studies (3, 25, 26). However, recent publications assessing the efficacy of transcutaneous CEUS to evaluate the spinal cord have shown promise, suggesting a future strategy that could allow for serial assessment of tumor recurrence without undergoing expensive MRI assessment (28, 29).

To our knowledge, this is the first review to examine the use of CEUS in spinal tumor resections. However, several study limitations should be acknowledged. The review included a majority of case reports and case series, which limited the population size and increases risk of selection bias. The non-random nature of case reports and series also limits this study’s ability to form comparative conclusions. Furthermore, the included studies were limited in data on patient outcomes. Most follow-up lengths were either not reported or less than 3 months, arguably an inadequate amount of time to observe long-term adverse outcomes. There were no objective measures, such as tumor reoccurrence or development of neurological injury, provided to quantify the success of tumor resection. Without comprehensive individual patient data, this study was limited to a narrative review.