Arterial Spin-Labeling Perfusion Lightbulb Sign: An Imaging Biomarker of Pediatric Posterior Fossa Hemangioblastoma

DISCUSSION

Hemangioblastoma is a highly vascular benign tumor.^6,7 Although more common in adults, it also occurs in children.⁶ The spine and posterior fossa are relatively common locations.^2,4,6 The 2021 WHO Classification of Tumors of the CNS does not distinguish “pediatric hemangioblastoma” from hemangioblastoma. However, it is noteworthy that the differential diagnoses for hemangioblastoma vary between adults and children. The specific characteristics and distinctive features considered when diagnosing hemangioblastoma in adults are not entirely applicable to the pediatric population. Our goal was to identify imaging markers in differentiating hemangioblastoma from other pediatric posterior fossa tumors due to the difference in preoperative management of children with hemangioblastoma, including embolization. We clinically observed a distinct ASL perfusion pattern in hemangioblastomas. In our daily practice, we noticed a pattern of intensely and homogeneously high perfusion within all solid tumor components without any low perfusion area, which we refer to as the ASL lightbulb sign. This study investigates the validity and diagnostic value of the ASL lightbulb sign compared with other MRI findings. All the hemangioblastomas showed peripheral edema. While hemangioblastoma is a low-grade benign tumor, it is known that peritumoral edema is a common presentation in the posterior fossa.^17⇓-19 Furthermore, profound surrounding edema disproportionately greater than expected for tumor size is an important feature of hemangioblastoma.^8,20,21 Profound peritumoral edema and neurologic symptoms associated with hemangioblastoma typically warrant surgical resection, irrespective of solitary or multiple lesions.^8,22 Our series consisted of only hemangioblastoma with surgery indication and showed 100% peritumoral edema; thus, our findings were in line with the previous literature. Another feature that we observed in all hemangioblastomas was T2-flow void. T2-flow void results from intravascular flow and intratumoral demonstration of this phenomenon is expected in highly vascular tumors, such as hemangioblastoma.⁷ Hemangioblastomas did not show reduced diffusion in our series and this finding was in line with previous literature.^9,11 It is known that hemangioblastoma usually shows high ADC values.^9,11 Diffusion characteristics can help to differentiate hemangioblastoma and metastatic masses but these values are not always useful in distinguishing it from certain posterior fossa tumors.^9,11 Although peripheral edema, lack of reduced diffusion, and T2-flow void were observed in all of the hemangioblastomas in our series, almost one-half of the control group also demonstrated these features. While these findings are statistically significant and might help differentiate hemangioblastoma from other posterior fossa tumors, none is stand-alone sufficient to distinguish hemangioblastoma completely. Still, we find it valuable to define the imaging characteristics of hemangioblastoma in children as enhancing solid parts with high ADC values, T2-flow voids, and peripheral edema. In our cohort the top 3 posterior fossa tumors were medulloblastoma, pilocytic astrocytoma, and posterior fossa ependymoma type A. Because posterior fossa ependymoma type A is mostly expected in infantile age groups, a dilemma between hemangioblastoma and posterior fossa ependymoma type A is not a common clinical scenario.²³ DWI can help differentiate medulloblastoma from other common posterior fossa tumors. Still there are many overlapping imaging characteristics between hemangioblastoma and other common posterior fossa tumors.²⁴ Previous literature showed the importance of perfusion imaging in diagnosis of hemangioblastoma and based on our study ASL MRI can improve to distinguish hemangioblastoma from other posterior fossa tumors in children.^10,11,25

It should be noted that DSC/DCE MR perfusion imaging has limited results in children due to difficulties in reproducing technical standards derived from adults and minimum use in children.²⁶ Age-related normative values for relative CBV derived from MR perfusion techniques are not yet fully established for children. ASL offers many advantages that make it suitable for children, less time than DCE/DSC, as well as elimination of IV access, contrast, or radiotracer agent, in addition to decreased susceptibility artifact at skull base.^27,28 These advantages of ASL over MR perfusion in children further highlight the diagnostic value of our proposed ASL lightbulb sign for accurate diagnosis of pediatric hemangioblastoma.

ASL perfusion imaging was available in 4 patients with hemangioblastoma and 38 patients with nonhemangioblastoma. Previous studies showed the high perfusion characteristics of hemangioblastoma due to its highly vascular nature.^10,11,14 High signal on ASL perfusion was observed in all hemangioblastomas but we also noticed this feature in 13 other posterior fossa tumors. Our raters grouped hemangioblastomas with some other tumors when they analyzed ASL perfusion characteristics superficially as just high or not high. Interrater agreement was not high in this binary evaluation due to subjectivity associated with slight heterogeneous lesional perfusion near or just above the normal cerebellar cortex and perfusion related to nearby or coursing vessels within the tumor. We concluded that high perfusion on ASL was a typical feature of hemangioblastoma but was not specific. As we hypothesized, both blinded reviewers marked as ASL lightbulb sign all the hemangioblastomas and none of the others in the 42 ASL study pool. Our results showed detailed pattern analyses of ASL perfusion, giving a better insight into the differential diagnoses. Furthermore, intense homogeneous high signals of all the solid parts of the tumor were as precise as pathology, so-called pathognomonic, in our series. Thirteen tumors showed high perfusion but no lightbulb signs on ASL. These tumors were broad including 4 posterior fossa group A ependymoma and 3 non-WNT, non-SHH medulloblastoma. Our results suggest some tumors could mimic hemangioblastoma on ASL, but detailed analyses of the perfusion pattern to check ASL lightbulb sign can help in such cases. Moreover, ASL lightbulb sign evaluation showed higher interrater agreement than superficial qualitative ASL review.

This study has several limitations. Posterior fossa hemangioblastoma is a rare entity in children and usually this low-grade tumor is not going to resection. Because we aimed to discuss a potential pathognomonic value of a feature we had to include only pathology-proved cases. All these restrictions resulted in a small sample size. Although we believe our findings strongly indicate the pathognomonic value of the ASL lightbulb sign, we could overestimate its importance due to the small sample size. Another point is that we included only hemangioblastoma with surgery indication. Although our pathology results suggested that these tumors are classic hemangioblastoma, we still do not know whether those have specific features. Additionally, while most of our patients had 3D pulsed ASL, our cohort includes different ASL perfusion techniques. Because our analyses are qualitative, we believe the effect of technical differences is minimal, but such variations are still a factor in interpretation of perfusion. All the hemangioblastomas in our cohort had either 2D or 3D pulsed ASL. The high signal of ASL is a representation of the blood flow and is mostly independent of the technique. Still, we are not able to conclude the validity of the lightbulb sign with different ASL techniques.

We did not notice any mimic of the lightbulb sign in our cohort. On the other hand, posterior fossa lesions are not limited to tumors, and focal high perfusion is also noted in some nontumoral lesions, such as high-flow arteriovenous malformation.²⁹ Nontumoral lesions should be excluded before implementing a lightbulb sign as a diagnostic tool. Because the tumor biology of hemangioblastoma does not change by patient age, hemangioblastomas potentially might represent lightbulb signs in different age groups. Moreover, previously published ASL figures of adult hemangioblastomas resemble our lightbulb sign.^14,30 However, further studies in the adult population are necessary to confirm this postulate, and we are not able to generalize our results for the adult population. Quantitative analyses of ASL perfusion are also another common analysis of this technique but in daily practice such quantitative analyses have relatively limited usage and we wanted to provide a methodology without necessity of any measurements. Our study proved that morphologic analyses of the perfusion pattern are superior to common qualitative analyses, but we still do not know whether morphologic analyses is superior to quantitative analyses.