J Neurooncol (2015) 122:357–366 DOI 10.1007/s11060-015-1722-4
CLINICAL STUDY
The role of diffusion tensor tractography in the surgical treatment of pediatric optic chiasmatic gliomas Ming Ge • ShaoWu Li • Liang Wang ChunDe Li • Junting Zhang
•
Received: 5 June 2014 / Accepted: 14 January 2015 / Published online: 24 January 2015 Ó Springer Science+Business Media New York 2015
Abstract Diffusion tensor tractography(DTT) can theoretically be used in assessing the optic chiasmatic glioma(OCGs),which are still in debate about optimal treatment. The purpose of this study was to investigate the role of this technology in offering more information about the tumor, assisting the debulking surgery, and helping to anticipate visual outcomes. As a prospective cohort study, the enrolled patients received routine pre- and postoperative neuro-ophthalmology, neuroimaging, and endocrine examinations. Fiber tractography was meanwhile performed based on diffusion tensor imaging examination. Identification of the position relationship between the lesions and residual optic path, and morphology analysis of them was done based on their DTT features. All the information was used for confirmation by the intraoperative findings. 11 pediatric patients were enrolled in this study. Most of them got subtotal resection of the tumors and stable postoperative visual outcomes. On the DTT imagings, the tumors were divided into infiltrative endophytic ones (TypeI) and inflated ones (TypeII), which can be subclassified as inferior and superior chiasmatic ones based on the positional relationships between the optic chiasm fibers and the tumors. These positional relationships were confirmed intraoperatively. The postoperative DTT images were quite different from preoperative ones. The application of DTT to children with OCGs is feasible, and valuable for getting more information about the
M. Ge S. Li L. Wang (&) C. Li J. Zhang Department of Neurosurgery, Beijing Tiantan Hospital, Capital Medical University, Tiantan Xili 6, Dongcheng District, Beijing 100050, People’s Republic of China e-mail: [email protected] M. Ge e-mail: [email protected]
disease, improving surgical techniques, and helping predict the overall and visual prognosis of the patients. The exact correlations of DTT features and visual outcomes need to be further verified. Keywords Diffusion tensor tractography Optic chiasmatic gliomas Oncology Visual outcome
Introduction Optic pathway gliomas (OPGs) are rare tumors that account for approximately 3–5 % of pediatric craniocerebral tumors [1], and 10–15 % of pediatric supratentorial tumors [2]. OPGs are more common in children, and rarely occur in adults [3]. For nearly 75 % of patients with an OPG, the disease onset occurs prior to age 10, and 90 % of OPG patients start to show clinical symptoms before reaching 20 years old [1, 3]. OPGs are predominantly slow-growing, low-grade astrocytomas, but their clinical characteristics vary substantially [2, 4, 5]. Most of OPGs are sporadic, but a small number are classified as neurofibromatosis type 1 (NF1)related, which eventually occur in approximately 7.5–20 % of NF1 patients [6, 7]. Most studies have suggested that sporadic OPGs exhibit more active clinical manifestations, increased invasiveness, and poor prognosis along with more frequent involvement of the optic chiasm [2, 8, 9]. In contrast, NF1-related OPGs typically have a relatively stable clinical manifestation and slow progression with more frequent involvement of the optic nerve than the optic chiasm [10, 11]. OPGs occur in different locations, including the optic nerve, optic chiasm, optic tract, and optic radiation, and are clinically categorized and named based on the location of
123
358
the primary pathological event. OPGs with primary lesions located in the optic chiasm are called optic chiasmatic gliomas(OCGs). The intervention timing and strategies for treating OPGs, including OCGs, are still controversial [9]. Most researches support that positive intervention is necessary to treat OCGs, especially sporadic chiasmatic gliomas that are not related to NF1 [10–12]. Although the current treatment strategies are more inclined to combine chemotherapy with radiotherapy as the first-line therapy [13], surgical treatment is still an important option for treating OPGs [10, 12, 14]. Magnetic resonance diffusion tensor imaging (MRDTI) was first used in the 1990s to clearly show the orientation of white matter fiber bundles [15–17]. In 2010, Nickerson et al. adopted MRDTI for preoperative examinations of pediatric sellar tumors to characterize optic nerve gliomas and other common tumors in the sellar region [15]. Diffusion tensor tractography (DTT) is based on MRDTI technology. DTT can more directly visualize the complete orientation of a target white matter fiber bundle and its spatial relationship with the pathological changes, offering clinical guidance. The optic nerve, optic chiasm, and optic tract are continuous and are composed of simple white matter fiber bundles that can be visualized using traditional MRI and can be distinguished from brain tissues [15]. Thus, theoretically, DTT technology can be employed to trace these structures. In 2008, Yamada et al. pioneered the use of DTT in imaging and reconstructing the optic pathway in animal models [18]. In 2012, Lober et al. conducted a retrospective study using this technology in pediatric OPGs to investigate the positional relationships between tumors and optic pathway fiber bundles [19]. As a prospective cohort study, this study integrated DTT technology with traditional MRI technology to explore the feasibility of utilizing DTT to guide the surgical treatment of patients with OCGs and to predict their prognosis.
Patients and methods Patient inclusion criteria Patients who were admitted to the Department of Pediatric Neurosurgery at Beijing Tiantan Hospital, affiliated with Capital Medical University, from July to December in 2013 and who were preoperatively diagnosed with optic pathway glioma were enrolled into this clinical study if they exhibited primary tumor location in the optic chiasm and were greater than 3 years old at the time of admission; informed signed consent was obtained from their parents. Patients were excluded from this study if they were not pathologically diagnosed with a neuroepithelial tumor or if
123
J Neurooncol (2015) 122:357–366
they could not cooperate and made obvious head movements in the procedure of DTT. This study protocol was approved by the Medical Ethics Committee of the participating hospital. Clinical assessment The enrolled patients received routine preoperative neuro-ophthalmology, neuroimaging, and neuroendocrine examinations. The ophthalmic examinations included a visual acuity test using a standard logarithmic visual acuity chart, an automated perimetry exam (Octopus 900, Switzerland), and color fundus photography. For ease of comparison, the visual acuity exam results using the logarithmic visual acuity chart were converted into a logarithmic expression of the minimum angle of resolution (logMAR). Based on previous literature, finger counting was defined as 1.6, perception of hand motion as 2, light perception as 2.5, and no light perception as 3 on the logMAR scale [6]. Fundus changes were classified by an independent neuro-ophthalmologist into three categories, papilledema, optic atrophy, and normal, based on the optic disc morphology. The neuroimaging examination included 3.0T head magnetic resonance imaging (plain, enhanced, DTI, and DTT).The neuroendocrine examinations included measuring serum PRL, GH, CRO, TT3, TT4, TSH, FT3, and FT4 levels as well as the serum levels of 3 tumor biomarkers, AFP, CEA, and b-HCG. One week after surgery, the enrolled patients received all the above routine neuro-ophthalmology, neuroimaging, and neuroendocrine examinations, except for the analysis of the tumor biomarkers. DTI examination parameters The enrolled patients underwent a DTI exam using 3.0T magnetic resonance imaging (Siemens Trio Tim, Germany) with an 8-channel head coil and sequence signal acquisition via interleaved echo planar imaging (IEPI) from 30 diffusion gradient directions. The specific parameters includes: FOV 256 mm,TR/TE 11000/94 ms, Slice thickness/gap 2.0/0 mm, EPI factor 128, voxel size 2.0 9 2.0 9 2.0 mm, Band width 1502 Hz/Px, and scan time duration 6 min 16 s. Image processing methods Image processing was conducted using manually outlined regions of interest (ROIs). A scanner workstation equipped with Neuro 3D image processing software (3.0T Siemens Trio Tim, Germany) was used to draw the bilateral optic nerves, optic chiasm, optic tract, and optic radiation.
J Neurooncol (2015) 122:357–366
Specifically, images were extracted from the T1-MPRAGE scanning sequence, and multiple seed regions were selected in both tumor tissue and adjacent normal tissue in the bilateral optic nerves and optic chiasm on the axial images. Fractional anisotropy threshold 0.05, curvature threshold 308. The image processing was performed by an independent neuroimaging technician and a neurosurgeon. Treatment All the enrolled patients received tumor resections by three senior physicians. Approaches were chosen based on their clinical and radiological manifestations, including the preoperative DTT results. The degree of resection was defined as: near-total resection(more than 90 % tumor resection), subtotal resection (50–90 % tumor resection), and partial resection (less than 50 % tumor resection).
Results Clinical features This study enrolled five male and six female patients with an average age of 8.1 years. Two patients had previously received ventricular-peritoneal shunt, and one patient had undergone frontotemporal craniotomy for biopsy. The remaining patients were not treated with any interventions, including radio- or chemotherapy. The preoperative symptoms, signs, and neuroendocrine results are summarized in Table 1. In this study, seven patients underwent tumor resection via the interhemispheric approach with a right frontal coronal incision, whereas the transcallosal interforniceal approach was selected with a right frontal incision for three patients, and with a right frontotemporal approach for one patient. The definitive pathological diagnoses were attained postoperatively. The operation and pathological results are shown in Table 1. Preoperative MRI features This cohort mostly presented with equal T1 and long T2 weighted images in the preoperative MRI with heterogeneous enhancement in the center of the lesions, as shown in Fig. 1a–c. However, one patient exhibited equal T1 and equal T2 weighted images without significant enhancement. Another lesion presented with cystic changes, but no significant enhancement was observed in the cyst wall. Three patients had lesions restricted to the suprasellar region, while 5 lesions protruded upward into the third ventricle. One patient had lesion (Fig. 2a–c) growing toward the anterior skull base and protruding upward into the third ventricle. Two patients had lesions with lateral
359
growth toward the medial temporal lobe through the lateral side of the optic nerve, as demonstrated in Fig. 3a–c. The tumor origin was determined based on preoperative imaging and intraoperative microscopy; none of the patients had tumors that originated solely from the optic nerve. Six patients presented with lesions originating from the optic chiasm, three patients had lesions originating from both the optic chiasm and the optic nerve, and two patients had lesions from both the optic chiasm and the optic tract. Based on the modified Dodge classification criteria [20], Dodge III were most frequently observed in nine patients (81.8 %), and the other two patients presented with Dodge II lesions. Preoperative DTT features Preoperative DTT examinations were utilized to evaluate the positional relationships between the lesions and the visual pathway structures. The optic nerves were visualized using DTT in 10 patients, with a visualization rate of 90.9 %; the optic nerves were frequently pushed to the temporal side of the tumor. The DTT indicated that the optic nerve fibers were predominantly located in the bilateral walls of the tumor (eight cases, 72.7 %). Among these eight patients, two exhibited an interruption in the continuity of the unilateral optic nerve fiber bundles in the lesioned area, as shown in Fig. 1d and 4b. In the two patients with tumors showing lateral growth, the DTT images revealed that one side of the optic nerve fiber bundle traveled through the tumor rather than being pushed to the lateral wall of the tumor. The optic chiasm was the predominant origin of the primary lesions in the patients in this study and therefore exhibited a closer positional relationship with the tumors, which resulted in more complicated DTT imaging. The optic chiasm fiber bundles visualized by DTT exhibited different positional relationships with the tumors, which demonstrates their different growth patterns. Based on the DTT images, the patients were divided into the following categories, which has been summarized in Table 1. Type I comprised the infiltrative endophytic tumors, and there was only 1 patient in this category, which was just the NF-1 related one. In this type, the optic chiasm was fused to the tumor, which could not be distinguished from normal structures using either traditional MRI or DTT images. Patients with expanded growth were categorized into Type II, the inflated type. 10 patients were in this category. These lesions all can be distinguished from the remaining optic chiasm by DTT. DTT of five patients showed the residual optic chiasm fiber bundles diverged around the lesions, as shown in Fig. 1d, e. The visualized optic chiasm fiber bundles in the other five patients exhibited trajection in the tumors and interrupted continuity in certain fiber bundles, as demonstrated in Figs. 2d, 3d, and 4b.
123
123
9.0, M
15.0, F
9.0, F
6.0, M
3.6, M
13.0, M
4.0, F
11.0, F
5.5, F
6.0, F
7.0, M
1
2
3
4
5
6
7
8
9
10
11
Binocular visual deterioration & precocious puberty for 1 year
Binocular visual deterioration for 2 years
Visual deterioration & limb weakness for 12 ms
Right visual deterioration for 2 years & left visual deterioration for 2 ms
Headache & nausea for 2 ms
Headache & polyuria for 2ws
Accidental finding for 3 ms
Binocular visual deterioration for 2 ms
Visual deterioration for 1 y, headache for 1 m
Headache & right visual deterioration for 3 ws
Headache & left visual deterioration for 2 ms
Symptoms
no
yes
no
no
no
no
no
no
no
no
no
NF1
Interhemispheric
Interhemispheric
Frontotemporal
Interhemispheric
Transcallosal
Interhemispheric
Transcallosal
Interhemispheric
Transcallosal
Interhemispheric
Interhemispheric
Surgical approach
Subtotal
Partial
Subtotal
Subtotal
Subtotal
Subtotal
Subtotal
Subtotal
Subtotal
Neartotal
Subtotal
Resection degree
Astrocytoma
PMA
Astrocytoma
PMA
PMA
PMA
Astrocytoma
Astrocytoma
PMA
PMA
PA
Pathological result
IIb
I
IIb
IIb
IIb
IIa
IIb
IIa
IIb
IIb
IIb
DTT type
TT3, TSH;/ FT3, TT3, FT4, TT4;
OD 0.9/1.2,
N/N
FT3, TT3:, TSH;/N
N/N
N/N
T, E2:/N
FT3, TT3:/N
N/N
N/N
N/N
PRL;/N
Pre-/postoperative neuroendocrine results
OS0.8/1.2,
OS0.7/0.7,
OD 0.7/0.7,
OS1.0/1.0
OD 1.0/1.0,
OD 3/3, OS 0.2/0.3
Do not cooperate
OS 0.6/2
OD -0.1/3,
Do not cooperate
OS 0.6/3,
OD 0.1/0.2,
OS 0.9/0.9,
OD 0.7/0.7,
OS -0.1/-0.12
OD 0.7/0.4,
OS -0.2/-0.2
OD 0.4/0.4,
Pre-/postoperative ophthalmic findings
PA pilocytic astrocytoma, PMA pilocytic mucinous astrocytoma, m month, y year, w week, N normal, PRL serum prolactin, FT3 free triiodothyronine, TT3 total triiodothyronine, T testosterone, E2 estradiol, TSH thyroid stimulating hormone, : higher than normal level, ; lower than normal level
Age(years), sex
Case no.
Table 1 Clinical characteristics of 11 patients with OCG
360 J Neurooncol (2015) 122:357–366
J Neurooncol (2015) 122:357–366
361
Fig. 1 a–c Preoperative MR imagings and f–h postoperative MR imagings of Case NO.1. DTT imagings d–e show left optic nerve fiber was interrupted and optic chiasm fiber diverged around the lesion
For the cases of inflated type (Type II), the relative positional relationships between the tumors and the remaining optic chiasm fiber bundles on these DTT images were generally divided into two subgroups, Type IIa, inferior to the optic chiasm(IOC), and Type IIb, superior to the optic chiasm(SOC). Depending on whether the tumor was below or above the visualized optic chiasm fiber bundles, as shown in Table 1. Among the patients in this study, two had tumors that were inferior to the optic chiasm (Fig. 4b) and eight had tumors that were superior to the optic chiasm (Fig. 3d). The optic tract and optic radiation fiber bundles were not visualized using DTT imaging in four and five patients,
respectively, with lower visualization rates (63.6 and 54.5 %, respectively) than those for the optic nerve and the optic chiasm. One patient had lesions originating from the optic chiasm and the right optic tract, and the right optic tract fibers were located lateral to the inner side of the tumor on the DTT images. The optic tract fibers were all pushed to the side of the tumors in the remaining six patients. In the six patients with visualized optic radiation fibers, the optic radiation fibers were of significantly different thickness on the two sides in three patients (Fig. 2d). The positional relationships between the visual pathway structures and the tumors in the preoperative DTT images
123
362
J Neurooncol (2015) 122:357–366
Fig. 2 a–c Preoperative MR imagings and f–h postoperative MR imagings of Case NO. 2. DTT imaging d shows the lesion was superior to the residual optic chiasm fiber, and left visual radiation
fibers were more thick than right ones. The postoperative DTT imaging e was quite different from the preoperative ones
were confirmed by the intraoperative microscope and neuronavigation findings in eight patients, with a coincidence rate of 100 % for the clinical and imaging findings, as shown in Figs. 3e, f and 4c, d. The DTT results were not intraoperatively confirmed for the 3 patients who underwent the transcallosal interforniceal procedure with a frontal incision due to insufficient exposure.
thinner in most of the patients, with poor imaging of the visual pathway in certain patients. The postoperative DTT images clearly illustrated the optic nerve and optic radiation in particular patients, but the optic chiasm was poorly visualized (Fig. 2e).
Postoperative DTT features
Within the patients in this study, three patients had worse visual acuity, and one exhibited improved visual acuity after surgery, as listed in Table 1. The postoperative fundus examination showed that there were no notable changes by comparison with their preoperative results.
Morphological changes in the visual pathway were observed by postoperative DTT. The posterior visual pathway (the optic tract and optic radiation) became
123
Variations of clinical results pre-& postoperatively
J Neurooncol (2015) 122:357–366
363
Fig. 3 a–c Preoperative MR imagings and g–j postoperative MR imagings of Case NO. 8. DTT imaging d shows the residual optic chiasm fiber transversed through the lesion. The intraoperative
imaging e verified the superior chiasmatic subtype (Type IIb), and f after the tumor resection, the residual optic pathway structures was demonstrated
Among the patients in this study, four patients experienced a postoperative onset of non-ophthalmologic complications, with an incidence of 36.4 %. One patient had abnormal postoperative hormone levels, specifically reduced thyroid hormone levels. In the two patients with preoperative polydipsia and polyuria, one had a significant improvement postoperatively, whereas the symptom remained in the other patient. One patient experienced transient diabetes insipidus symptoms after surgery, and two patients had transient electrolyte disturbances postoperatively(elevated serum sodium levels), which were all corrected by the time of discharge. None of the patients experienced a postoperative fever, coma, hydrocephalus, cerebral infarction, or death. The results can be seen in Table 1.
Discussion The strategy for treating OPGs, especially OCGs, has gradually, yet controversially, evolved. Surgery combined with postoperative radiotherapy used to be the primary treatment strategy [10, 14, 21]. However, with the progress of European clinical trials of chemotherapy for OPGs, most researchers use chemotherapy and/or radiotherapy as the first-line treatment [13, 22], and surgery is only utilized for the unilateral optic nerve tumors that cause severe monocular vision loss, or the exogrowth and cystic lesions [4]. Nevertheless, Goodden et al. [23] recently claimed that surgery has a clear role for diagnosis, tumor control, and relief of mass effect in children with OCGs. By comparing the treatment effects in 53 children with OPGs, our project
123
364
J Neurooncol (2015) 122:357–366
Fig. 4 a Preoperative MR imaging and e postoperative MR imaging of Case NO. 4. DTT imaging b shows left optic nerve fiber was interrupted and the lesion was inferior to the optic chiasm fiber. The intraoperative imaging c, d verified the inferior chiasmatic subtype
(Type IIa), and his left optic nerve was thinner than right one. The fact that the patient’ left visual outcome turned to worse were consistent with his preoperative DTT imagings and intraoperative findings
team found that surgery combined with stereotactic radiotherapy significantly prolonged overall survival [24]. Therefore, for OCGs with an onset after 4 years old, our project team employs surgical resection as the first-line treatment. OPGs can be divided into the Dodge [20] and McCullough [21] subtypes, and most OCGs are types II–III within the Dodge subtype. In 1974, Miller et al. [25] divided OCGs into pre- and post-chiasmatic types according to the positional relationship between the lesion and the optic chiasm. In 1990, Wisoff et al. [21] divided OCGs into ingrowth and outgrowth types based on whether the tumor broke through the perineurium into the subarachnoid space. In this study, the application of DTT to the perioperative assessment of OCGs provides a more adequate and intuitive basis for classifying these tumors. The high rebuilding imaging rate for the optic nerve and optic chiasm structures in this study provides a premise for the appropriate clinical classification. Moreover, with the knowledge of previous classifications, certain lesions were classified differently when imaged from a different angle. Using DTT images, we revised the typing in Wisoff et al. [21] as infiltrative
endophytic type (Type I) and the inflated type (Type II), and we revised the typing in Miller et al. [25] as inferior chiasmatic ones(Type IIa) versus suprachiasmatic ones(Type IIb). There are two significant reasons for classifying OCGs based on DTT imagings. (1) DTT classification provides a deeper understanding of OCGs. There were significantly more cases of the suprachiasmatic type than the inferior chiasmatic type in this study because most OCGs gradually push the remaining optic chiasm structure downward as the tumor grows. In addition, residual optic chiasm fiber bundles have two arrangements with the relationship of the lesion, either diverge around it or transversed through it. All of these informations can only be able to be captured by DTT imagings. (2) DTT contributes to improving surgical strategy. To retain vision and to reduce the complications of OCGs, it is important to maintain the integrity of the normal optic chiasm structure [26, 27]. We can clearly distinguish the optic nerve, the optic chiasm, and the main tumor body on a preoperative DTT image. Most tumors pushed the bilateral optic nerves to each side; therefore we chose the midline approach (including the interhemispheric or transcallosal
123
J Neurooncol (2015) 122:357–366
365
Table 2 Comparision of clinical results between two series of OPGs from the authors’ group Period range
Num
Sex ratio (M:F)
Mean age (years)
NF1 ratio
Resection degree ratio (subtotal:partial)
Ophthalmologic complications
Non-ophthalmologic complications
Death rate within 1 m
2003–2011 [24]
53
1.5:1
7.6
2/53 (3.8 %)
0:53
–
23/53 (43.4 %)
2/53 (3.8 %)
2013
11
0.8:1
8.1
1/11 (9.1 %)
10:1
3/11 (27.3 %)
4/11 (36.4 %)
0
Num number, M male, F Female, – not mentioned, m month
interforniceal approaches) rather than the lateral approach (such as the frontotemporal approach) to better protect the displaced bilateral optic nerves, which was consistent with Goodden et al. [23]. In addition, for inferior chiasmatic tumors, the transcallosal interforniceal approach and the incision in the perineurium should both be carefully selected to minimize the interference with the remaining optic chiasm. Compared with the previous series from the same group [24], most cases in this series gained more aggressive tumor resection, and better postoperative outcomes with the help of DTT images, as shown in Table 2. It should be noted that DTT imaging and the subsequent classification depend on intraoperative validation and verification. In this study, the intraoperative microscope findings were consistent with the preoperative DTT-determined positional relationships between the tumors and the visual pathway for most patients, which proved the technology in this study was feasible and reliable. The only one previous study of utilizing DTT with OPGs reported that there are no correlation between vision changes and the ability to image the optic fiber bundles [19]. In this series, the pre- and postoperative DTT images of the same patients most had changed substantially. These amplitude of variations are more obvious than their visual changes before and after the operations. These phenomenons may be related to the hemosiderosis and edema in the neighboring brain tissue postoperatively. It remains undetermined whether the sensitivity and specificity of DTT are sufficient to detect changes in vision, which need to be confirmed in the future. There is no doubt that this study has its limitations. Firstly, the DTT technology itself relies greatly on the subjective thinking of examiners, as the choice of ROIs depends on the operators’ understanding of brain anatomy and this disease [15–17, 19]. Secondly, long-term followup results including their DTT images and visual outcomes have not been able to be gathered right now, which needs more continuity of this cohort research.
Conclusion It is feasible and will be beneficial to utilize DTT to diagnose and treat children with OCGs. The application of
DTT to the peri-operative care of children with OCGs is valuable for increasing our understanding of the disease, developing reasonable surgical strategies, and helping determine the overall and visual prognosis of a patient. But, the exact relationship between DTT features and visual outcomes are still uncertain.
Funding This work was supported by the Beijing Nova Program (No. 2011086), which is a chinese official foundation. Conflict of interest The authors have declared no conflicts of interest. No financial interests or affiliations with institutions, organization, or companies are mentioned or have impacts on the views expressed in the article.
References 1. Shapey J, Danesh-Meyer HV, Kaye AH (2011) Diagnosis and management of optic nerve glioma. J Clin Neurosci 18:1585–1591 2. Kornreich L, Blaser S, Schwarz M et al (2001) Optic pathway glioma: correlation of imaging findings with the presence of neurofibromatosis. Am J Neuroradiol 22:1963–1969 3. Shofty B, Constantini S, Bokstein F, Ram Z, Ben-Sira L, Freedman S, Vainer G, Kesler A (2014) Optic pathway gliomas in adults. Neurosurgery 74:273–280 4. Binning MJ (2007) Optic pathway gliomas: a review. Neurosurg Focus 23(5):E2 5. Yeom KW, Lober RM, Andre JB, Fisher PG, Barnes PD, Edwards MS, Partap S (2013) Prognostic role for diffusionweighted imaging of pediatric optic pathway glioma. J Neurooncol 113:479–483 6. de Blank PM, Berman JI, Liu GT, Roberts TP, Fisher MJ (2013) Fractional anisotropy of the optic radiations is associated with visual acuity loss in optic pathway gliomas of neurofibromatosis type 1. Neuro Oncol 15:1088–1095 7. Lama G, Esposito SM, Grassia C, Calabrese E, Grassia MG, Bismuto R, Melone MA, Russo S, Scuotto A (2007) Neurofibromatosis type 1 and optic pathway glioma. A long-term followup. Minerva Pediatr 59:13–21 8. Avery RA, Liu GT, Fisher MJ, Quinn GE, Belasco JB, Phillips PC, Maguire MG, Balcer LJ (2011) Retinal nerve fiber layer thickness in children with optic pathway gliomas. Am J Ophthalmol 151:542–549 9. Andrew GL (2007) Neuroophthalmological management of optic pathway gliomas. Neurosurg Focus 23:E1 10. Wong JYC, Uhlv Wara WM, Sheline GE (1987) Optic gliomas. Cancer 60:1847–1855
123
366 11. Hiyama H, Kubo O, Nakajima H, Hayashi M, Kagawa M (1992) A case of cystic optic glioma involving chiasma and bilateral posterior optic pathway. No Shinkei Geka 20:1199–1204 12. Ahn Y, Cho BK, Kim SK, Chung YN, Lee CS, Kim IH, Yang SW, Kim HS, Kim HJ, Jung HW, Wang KC (2006) Optic pathway glioma: outcome and prognostic factors in a surgical series. Childs Nerv Syst 22:1136–1142 13. Dalla Via P, Opocher E, Pinello ML, Calderone M, Viscardi E, Clementi M, Battistella PA, Laverda AM, Da Dalt L, Perilongo G (2007) Visual outcome of a cohort of children with neurofibromatosis type 1 and optic pathway glioma followed by a pediatric neuro-oncology program. Neuro Oncol 9:430–437 14. Kanamori M, Shibuya M, Yoshida J, Takayasu M, Kageyama N (1985) Long-term follow-up of patients with optic glioma. Child’s Nerv Syst 1:272–278 15. Nickerson JP, Salmela MB, Koski CJ, Andrews T, Filippi CG (2010) Diffusion tensor imaging of the pediatric optic nerve: intrinsic and extrinsic pathology compared to normal controls. J Magn Reson Imaging 32:76–81 16. Nucifora PP, Verma R, Lee SK, Melhem ER (2007) Diffusiontensor MR imaging and tractography: exploring brain microstructure and connectivity. Radiology 245:367–384 17. Mukherjee P, Berman JI, Chung SW, Hess CP, Henry RG (2008) Diffusion tensor MR imaging and fiber. Am J Neuroradiol 29:632–641 18. Yamada M, Momoshima S, Masutani Y, Fujiyoshi K, Abe O, Nakamura M, Aoki S, Tamaoki N, Okano H (2008) Diffusiontensor neuronal fiber tractography and manganese-enhanced MR
123
J Neurooncol (2015) 122:357–366
19.
20.
21.
22.
23.
24.
25. 26. 27.
imaging of primate visual pathway in the common marmoset: preliminary results. Radiology 249:855–864 Lober RM, Guzman R, Cheshier SH, Fredrick DR, Edwards MS, Yeom KW (2012) Application of diffusion tensor tractography in pediatric optic pathway glioma. J Neurosurg Pediatr 10:273–280 Dodge HW, Love JG, Craig WM, Dockerty MB, Kearns TP, Holman CB, Hayles AB (1958) Gliomas of the optic nerves. AMA Arch Neurol Psychiatry 79:607–621 Wisoff JH, Abbott R, Epstein F (1990) Surgical management of exophytic chiasmatic-hypothalamic tumors of childhood. J Neurosurg 73:661–667 Kell JP, Leary S, Khanna P, Weiss AH (2012) Longitudinal measures of visual function, tumor volume, and prediction of visual outcomes after treatment of optic pathway gliomas. Ophthalmology 119:1231–1237 Goodden J, Pizer B, Pettorini B, Williams D, Blair J, Didi M, Thorp N, Mallucci C (2014) The role of surgery in optic pathway/hypothalamic gliomas in children. J Neurosurg Pediatrics 13:1–12 Tian YJ, Li DL, Jia G (2012) Clinical characteristics and treatment of pediatric optic pathway gliomas. Chin J Neurosurg 28:1136–1140 Miller NR, Iliff WJ, Green WR (1974) Evaluation and management of gliomas of the anterior visual pathways. Brain 97:743–754 Astrup J (2003) Natural history and clinical management of optic pathway glioma. Br J Neurosurg 17:327–335 Kelly JP, Weiss AH (2013) Detection of tumor progression in optic pathway glioma with and without neurofibromatosis type 1. Neuro Oncol 15:1560–1567
CLINICAL STUDY
The role of diffusion tensor tractography in the surgical treatment of pediatric optic chiasmatic gliomas Ming Ge • ShaoWu Li • Liang Wang ChunDe Li • Junting Zhang
•
Received: 5 June 2014 / Accepted: 14 January 2015 / Published online: 24 January 2015 Ó Springer Science+Business Media New York 2015
Abstract Diffusion tensor tractography(DTT) can theoretically be used in assessing the optic chiasmatic glioma(OCGs),which are still in debate about optimal treatment. The purpose of this study was to investigate the role of this technology in offering more information about the tumor, assisting the debulking surgery, and helping to anticipate visual outcomes. As a prospective cohort study, the enrolled patients received routine pre- and postoperative neuro-ophthalmology, neuroimaging, and endocrine examinations. Fiber tractography was meanwhile performed based on diffusion tensor imaging examination. Identification of the position relationship between the lesions and residual optic path, and morphology analysis of them was done based on their DTT features. All the information was used for confirmation by the intraoperative findings. 11 pediatric patients were enrolled in this study. Most of them got subtotal resection of the tumors and stable postoperative visual outcomes. On the DTT imagings, the tumors were divided into infiltrative endophytic ones (TypeI) and inflated ones (TypeII), which can be subclassified as inferior and superior chiasmatic ones based on the positional relationships between the optic chiasm fibers and the tumors. These positional relationships were confirmed intraoperatively. The postoperative DTT images were quite different from preoperative ones. The application of DTT to children with OCGs is feasible, and valuable for getting more information about the
M. Ge S. Li L. Wang (&) C. Li J. Zhang Department of Neurosurgery, Beijing Tiantan Hospital, Capital Medical University, Tiantan Xili 6, Dongcheng District, Beijing 100050, People’s Republic of China e-mail: [email protected] M. Ge e-mail: [email protected]
disease, improving surgical techniques, and helping predict the overall and visual prognosis of the patients. The exact correlations of DTT features and visual outcomes need to be further verified. Keywords Diffusion tensor tractography Optic chiasmatic gliomas Oncology Visual outcome
Introduction Optic pathway gliomas (OPGs) are rare tumors that account for approximately 3–5 % of pediatric craniocerebral tumors [1], and 10–15 % of pediatric supratentorial tumors [2]. OPGs are more common in children, and rarely occur in adults [3]. For nearly 75 % of patients with an OPG, the disease onset occurs prior to age 10, and 90 % of OPG patients start to show clinical symptoms before reaching 20 years old [1, 3]. OPGs are predominantly slow-growing, low-grade astrocytomas, but their clinical characteristics vary substantially [2, 4, 5]. Most of OPGs are sporadic, but a small number are classified as neurofibromatosis type 1 (NF1)related, which eventually occur in approximately 7.5–20 % of NF1 patients [6, 7]. Most studies have suggested that sporadic OPGs exhibit more active clinical manifestations, increased invasiveness, and poor prognosis along with more frequent involvement of the optic chiasm [2, 8, 9]. In contrast, NF1-related OPGs typically have a relatively stable clinical manifestation and slow progression with more frequent involvement of the optic nerve than the optic chiasm [10, 11]. OPGs occur in different locations, including the optic nerve, optic chiasm, optic tract, and optic radiation, and are clinically categorized and named based on the location of
123
358
the primary pathological event. OPGs with primary lesions located in the optic chiasm are called optic chiasmatic gliomas(OCGs). The intervention timing and strategies for treating OPGs, including OCGs, are still controversial [9]. Most researches support that positive intervention is necessary to treat OCGs, especially sporadic chiasmatic gliomas that are not related to NF1 [10–12]. Although the current treatment strategies are more inclined to combine chemotherapy with radiotherapy as the first-line therapy [13], surgical treatment is still an important option for treating OPGs [10, 12, 14]. Magnetic resonance diffusion tensor imaging (MRDTI) was first used in the 1990s to clearly show the orientation of white matter fiber bundles [15–17]. In 2010, Nickerson et al. adopted MRDTI for preoperative examinations of pediatric sellar tumors to characterize optic nerve gliomas and other common tumors in the sellar region [15]. Diffusion tensor tractography (DTT) is based on MRDTI technology. DTT can more directly visualize the complete orientation of a target white matter fiber bundle and its spatial relationship with the pathological changes, offering clinical guidance. The optic nerve, optic chiasm, and optic tract are continuous and are composed of simple white matter fiber bundles that can be visualized using traditional MRI and can be distinguished from brain tissues [15]. Thus, theoretically, DTT technology can be employed to trace these structures. In 2008, Yamada et al. pioneered the use of DTT in imaging and reconstructing the optic pathway in animal models [18]. In 2012, Lober et al. conducted a retrospective study using this technology in pediatric OPGs to investigate the positional relationships between tumors and optic pathway fiber bundles [19]. As a prospective cohort study, this study integrated DTT technology with traditional MRI technology to explore the feasibility of utilizing DTT to guide the surgical treatment of patients with OCGs and to predict their prognosis.
Patients and methods Patient inclusion criteria Patients who were admitted to the Department of Pediatric Neurosurgery at Beijing Tiantan Hospital, affiliated with Capital Medical University, from July to December in 2013 and who were preoperatively diagnosed with optic pathway glioma were enrolled into this clinical study if they exhibited primary tumor location in the optic chiasm and were greater than 3 years old at the time of admission; informed signed consent was obtained from their parents. Patients were excluded from this study if they were not pathologically diagnosed with a neuroepithelial tumor or if
123
J Neurooncol (2015) 122:357–366
they could not cooperate and made obvious head movements in the procedure of DTT. This study protocol was approved by the Medical Ethics Committee of the participating hospital. Clinical assessment The enrolled patients received routine preoperative neuro-ophthalmology, neuroimaging, and neuroendocrine examinations. The ophthalmic examinations included a visual acuity test using a standard logarithmic visual acuity chart, an automated perimetry exam (Octopus 900, Switzerland), and color fundus photography. For ease of comparison, the visual acuity exam results using the logarithmic visual acuity chart were converted into a logarithmic expression of the minimum angle of resolution (logMAR). Based on previous literature, finger counting was defined as 1.6, perception of hand motion as 2, light perception as 2.5, and no light perception as 3 on the logMAR scale [6]. Fundus changes were classified by an independent neuro-ophthalmologist into three categories, papilledema, optic atrophy, and normal, based on the optic disc morphology. The neuroimaging examination included 3.0T head magnetic resonance imaging (plain, enhanced, DTI, and DTT).The neuroendocrine examinations included measuring serum PRL, GH, CRO, TT3, TT4, TSH, FT3, and FT4 levels as well as the serum levels of 3 tumor biomarkers, AFP, CEA, and b-HCG. One week after surgery, the enrolled patients received all the above routine neuro-ophthalmology, neuroimaging, and neuroendocrine examinations, except for the analysis of the tumor biomarkers. DTI examination parameters The enrolled patients underwent a DTI exam using 3.0T magnetic resonance imaging (Siemens Trio Tim, Germany) with an 8-channel head coil and sequence signal acquisition via interleaved echo planar imaging (IEPI) from 30 diffusion gradient directions. The specific parameters includes: FOV 256 mm,TR/TE 11000/94 ms, Slice thickness/gap 2.0/0 mm, EPI factor 128, voxel size 2.0 9 2.0 9 2.0 mm, Band width 1502 Hz/Px, and scan time duration 6 min 16 s. Image processing methods Image processing was conducted using manually outlined regions of interest (ROIs). A scanner workstation equipped with Neuro 3D image processing software (3.0T Siemens Trio Tim, Germany) was used to draw the bilateral optic nerves, optic chiasm, optic tract, and optic radiation.
J Neurooncol (2015) 122:357–366
Specifically, images were extracted from the T1-MPRAGE scanning sequence, and multiple seed regions were selected in both tumor tissue and adjacent normal tissue in the bilateral optic nerves and optic chiasm on the axial images. Fractional anisotropy threshold 0.05, curvature threshold 308. The image processing was performed by an independent neuroimaging technician and a neurosurgeon. Treatment All the enrolled patients received tumor resections by three senior physicians. Approaches were chosen based on their clinical and radiological manifestations, including the preoperative DTT results. The degree of resection was defined as: near-total resection(more than 90 % tumor resection), subtotal resection (50–90 % tumor resection), and partial resection (less than 50 % tumor resection).
Results Clinical features This study enrolled five male and six female patients with an average age of 8.1 years. Two patients had previously received ventricular-peritoneal shunt, and one patient had undergone frontotemporal craniotomy for biopsy. The remaining patients were not treated with any interventions, including radio- or chemotherapy. The preoperative symptoms, signs, and neuroendocrine results are summarized in Table 1. In this study, seven patients underwent tumor resection via the interhemispheric approach with a right frontal coronal incision, whereas the transcallosal interforniceal approach was selected with a right frontal incision for three patients, and with a right frontotemporal approach for one patient. The definitive pathological diagnoses were attained postoperatively. The operation and pathological results are shown in Table 1. Preoperative MRI features This cohort mostly presented with equal T1 and long T2 weighted images in the preoperative MRI with heterogeneous enhancement in the center of the lesions, as shown in Fig. 1a–c. However, one patient exhibited equal T1 and equal T2 weighted images without significant enhancement. Another lesion presented with cystic changes, but no significant enhancement was observed in the cyst wall. Three patients had lesions restricted to the suprasellar region, while 5 lesions protruded upward into the third ventricle. One patient had lesion (Fig. 2a–c) growing toward the anterior skull base and protruding upward into the third ventricle. Two patients had lesions with lateral
359
growth toward the medial temporal lobe through the lateral side of the optic nerve, as demonstrated in Fig. 3a–c. The tumor origin was determined based on preoperative imaging and intraoperative microscopy; none of the patients had tumors that originated solely from the optic nerve. Six patients presented with lesions originating from the optic chiasm, three patients had lesions originating from both the optic chiasm and the optic nerve, and two patients had lesions from both the optic chiasm and the optic tract. Based on the modified Dodge classification criteria [20], Dodge III were most frequently observed in nine patients (81.8 %), and the other two patients presented with Dodge II lesions. Preoperative DTT features Preoperative DTT examinations were utilized to evaluate the positional relationships between the lesions and the visual pathway structures. The optic nerves were visualized using DTT in 10 patients, with a visualization rate of 90.9 %; the optic nerves were frequently pushed to the temporal side of the tumor. The DTT indicated that the optic nerve fibers were predominantly located in the bilateral walls of the tumor (eight cases, 72.7 %). Among these eight patients, two exhibited an interruption in the continuity of the unilateral optic nerve fiber bundles in the lesioned area, as shown in Fig. 1d and 4b. In the two patients with tumors showing lateral growth, the DTT images revealed that one side of the optic nerve fiber bundle traveled through the tumor rather than being pushed to the lateral wall of the tumor. The optic chiasm was the predominant origin of the primary lesions in the patients in this study and therefore exhibited a closer positional relationship with the tumors, which resulted in more complicated DTT imaging. The optic chiasm fiber bundles visualized by DTT exhibited different positional relationships with the tumors, which demonstrates their different growth patterns. Based on the DTT images, the patients were divided into the following categories, which has been summarized in Table 1. Type I comprised the infiltrative endophytic tumors, and there was only 1 patient in this category, which was just the NF-1 related one. In this type, the optic chiasm was fused to the tumor, which could not be distinguished from normal structures using either traditional MRI or DTT images. Patients with expanded growth were categorized into Type II, the inflated type. 10 patients were in this category. These lesions all can be distinguished from the remaining optic chiasm by DTT. DTT of five patients showed the residual optic chiasm fiber bundles diverged around the lesions, as shown in Fig. 1d, e. The visualized optic chiasm fiber bundles in the other five patients exhibited trajection in the tumors and interrupted continuity in certain fiber bundles, as demonstrated in Figs. 2d, 3d, and 4b.
123
123
9.0, M
15.0, F
9.0, F
6.0, M
3.6, M
13.0, M
4.0, F
11.0, F
5.5, F
6.0, F
7.0, M
1
2
3
4
5
6
7
8
9
10
11
Binocular visual deterioration & precocious puberty for 1 year
Binocular visual deterioration for 2 years
Visual deterioration & limb weakness for 12 ms
Right visual deterioration for 2 years & left visual deterioration for 2 ms
Headache & nausea for 2 ms
Headache & polyuria for 2ws
Accidental finding for 3 ms
Binocular visual deterioration for 2 ms
Visual deterioration for 1 y, headache for 1 m
Headache & right visual deterioration for 3 ws
Headache & left visual deterioration for 2 ms
Symptoms
no
yes
no
no
no
no
no
no
no
no
no
NF1
Interhemispheric
Interhemispheric
Frontotemporal
Interhemispheric
Transcallosal
Interhemispheric
Transcallosal
Interhemispheric
Transcallosal
Interhemispheric
Interhemispheric
Surgical approach
Subtotal
Partial
Subtotal
Subtotal
Subtotal
Subtotal
Subtotal
Subtotal
Subtotal
Neartotal
Subtotal
Resection degree
Astrocytoma
PMA
Astrocytoma
PMA
PMA
PMA
Astrocytoma
Astrocytoma
PMA
PMA
PA
Pathological result
IIb
I
IIb
IIb
IIb
IIa
IIb
IIa
IIb
IIb
IIb
DTT type
TT3, TSH;/ FT3, TT3, FT4, TT4;
OD 0.9/1.2,
N/N
FT3, TT3:, TSH;/N
N/N
N/N
T, E2:/N
FT3, TT3:/N
N/N
N/N
N/N
PRL;/N
Pre-/postoperative neuroendocrine results
OS0.8/1.2,
OS0.7/0.7,
OD 0.7/0.7,
OS1.0/1.0
OD 1.0/1.0,
OD 3/3, OS 0.2/0.3
Do not cooperate
OS 0.6/2
OD -0.1/3,
Do not cooperate
OS 0.6/3,
OD 0.1/0.2,
OS 0.9/0.9,
OD 0.7/0.7,
OS -0.1/-0.12
OD 0.7/0.4,
OS -0.2/-0.2
OD 0.4/0.4,
Pre-/postoperative ophthalmic findings
PA pilocytic astrocytoma, PMA pilocytic mucinous astrocytoma, m month, y year, w week, N normal, PRL serum prolactin, FT3 free triiodothyronine, TT3 total triiodothyronine, T testosterone, E2 estradiol, TSH thyroid stimulating hormone, : higher than normal level, ; lower than normal level
Age(years), sex
Case no.
Table 1 Clinical characteristics of 11 patients with OCG
360 J Neurooncol (2015) 122:357–366
J Neurooncol (2015) 122:357–366
361
Fig. 1 a–c Preoperative MR imagings and f–h postoperative MR imagings of Case NO.1. DTT imagings d–e show left optic nerve fiber was interrupted and optic chiasm fiber diverged around the lesion
For the cases of inflated type (Type II), the relative positional relationships between the tumors and the remaining optic chiasm fiber bundles on these DTT images were generally divided into two subgroups, Type IIa, inferior to the optic chiasm(IOC), and Type IIb, superior to the optic chiasm(SOC). Depending on whether the tumor was below or above the visualized optic chiasm fiber bundles, as shown in Table 1. Among the patients in this study, two had tumors that were inferior to the optic chiasm (Fig. 4b) and eight had tumors that were superior to the optic chiasm (Fig. 3d). The optic tract and optic radiation fiber bundles were not visualized using DTT imaging in four and five patients,
respectively, with lower visualization rates (63.6 and 54.5 %, respectively) than those for the optic nerve and the optic chiasm. One patient had lesions originating from the optic chiasm and the right optic tract, and the right optic tract fibers were located lateral to the inner side of the tumor on the DTT images. The optic tract fibers were all pushed to the side of the tumors in the remaining six patients. In the six patients with visualized optic radiation fibers, the optic radiation fibers were of significantly different thickness on the two sides in three patients (Fig. 2d). The positional relationships between the visual pathway structures and the tumors in the preoperative DTT images
123
362
J Neurooncol (2015) 122:357–366
Fig. 2 a–c Preoperative MR imagings and f–h postoperative MR imagings of Case NO. 2. DTT imaging d shows the lesion was superior to the residual optic chiasm fiber, and left visual radiation
fibers were more thick than right ones. The postoperative DTT imaging e was quite different from the preoperative ones
were confirmed by the intraoperative microscope and neuronavigation findings in eight patients, with a coincidence rate of 100 % for the clinical and imaging findings, as shown in Figs. 3e, f and 4c, d. The DTT results were not intraoperatively confirmed for the 3 patients who underwent the transcallosal interforniceal procedure with a frontal incision due to insufficient exposure.
thinner in most of the patients, with poor imaging of the visual pathway in certain patients. The postoperative DTT images clearly illustrated the optic nerve and optic radiation in particular patients, but the optic chiasm was poorly visualized (Fig. 2e).
Postoperative DTT features
Within the patients in this study, three patients had worse visual acuity, and one exhibited improved visual acuity after surgery, as listed in Table 1. The postoperative fundus examination showed that there were no notable changes by comparison with their preoperative results.
Morphological changes in the visual pathway were observed by postoperative DTT. The posterior visual pathway (the optic tract and optic radiation) became
123
Variations of clinical results pre-& postoperatively
J Neurooncol (2015) 122:357–366
363
Fig. 3 a–c Preoperative MR imagings and g–j postoperative MR imagings of Case NO. 8. DTT imaging d shows the residual optic chiasm fiber transversed through the lesion. The intraoperative
imaging e verified the superior chiasmatic subtype (Type IIb), and f after the tumor resection, the residual optic pathway structures was demonstrated
Among the patients in this study, four patients experienced a postoperative onset of non-ophthalmologic complications, with an incidence of 36.4 %. One patient had abnormal postoperative hormone levels, specifically reduced thyroid hormone levels. In the two patients with preoperative polydipsia and polyuria, one had a significant improvement postoperatively, whereas the symptom remained in the other patient. One patient experienced transient diabetes insipidus symptoms after surgery, and two patients had transient electrolyte disturbances postoperatively(elevated serum sodium levels), which were all corrected by the time of discharge. None of the patients experienced a postoperative fever, coma, hydrocephalus, cerebral infarction, or death. The results can be seen in Table 1.
Discussion The strategy for treating OPGs, especially OCGs, has gradually, yet controversially, evolved. Surgery combined with postoperative radiotherapy used to be the primary treatment strategy [10, 14, 21]. However, with the progress of European clinical trials of chemotherapy for OPGs, most researchers use chemotherapy and/or radiotherapy as the first-line treatment [13, 22], and surgery is only utilized for the unilateral optic nerve tumors that cause severe monocular vision loss, or the exogrowth and cystic lesions [4]. Nevertheless, Goodden et al. [23] recently claimed that surgery has a clear role for diagnosis, tumor control, and relief of mass effect in children with OCGs. By comparing the treatment effects in 53 children with OPGs, our project
123
364
J Neurooncol (2015) 122:357–366
Fig. 4 a Preoperative MR imaging and e postoperative MR imaging of Case NO. 4. DTT imaging b shows left optic nerve fiber was interrupted and the lesion was inferior to the optic chiasm fiber. The intraoperative imaging c, d verified the inferior chiasmatic subtype
(Type IIa), and his left optic nerve was thinner than right one. The fact that the patient’ left visual outcome turned to worse were consistent with his preoperative DTT imagings and intraoperative findings
team found that surgery combined with stereotactic radiotherapy significantly prolonged overall survival [24]. Therefore, for OCGs with an onset after 4 years old, our project team employs surgical resection as the first-line treatment. OPGs can be divided into the Dodge [20] and McCullough [21] subtypes, and most OCGs are types II–III within the Dodge subtype. In 1974, Miller et al. [25] divided OCGs into pre- and post-chiasmatic types according to the positional relationship between the lesion and the optic chiasm. In 1990, Wisoff et al. [21] divided OCGs into ingrowth and outgrowth types based on whether the tumor broke through the perineurium into the subarachnoid space. In this study, the application of DTT to the perioperative assessment of OCGs provides a more adequate and intuitive basis for classifying these tumors. The high rebuilding imaging rate for the optic nerve and optic chiasm structures in this study provides a premise for the appropriate clinical classification. Moreover, with the knowledge of previous classifications, certain lesions were classified differently when imaged from a different angle. Using DTT images, we revised the typing in Wisoff et al. [21] as infiltrative
endophytic type (Type I) and the inflated type (Type II), and we revised the typing in Miller et al. [25] as inferior chiasmatic ones(Type IIa) versus suprachiasmatic ones(Type IIb). There are two significant reasons for classifying OCGs based on DTT imagings. (1) DTT classification provides a deeper understanding of OCGs. There were significantly more cases of the suprachiasmatic type than the inferior chiasmatic type in this study because most OCGs gradually push the remaining optic chiasm structure downward as the tumor grows. In addition, residual optic chiasm fiber bundles have two arrangements with the relationship of the lesion, either diverge around it or transversed through it. All of these informations can only be able to be captured by DTT imagings. (2) DTT contributes to improving surgical strategy. To retain vision and to reduce the complications of OCGs, it is important to maintain the integrity of the normal optic chiasm structure [26, 27]. We can clearly distinguish the optic nerve, the optic chiasm, and the main tumor body on a preoperative DTT image. Most tumors pushed the bilateral optic nerves to each side; therefore we chose the midline approach (including the interhemispheric or transcallosal
123
J Neurooncol (2015) 122:357–366
365
Table 2 Comparision of clinical results between two series of OPGs from the authors’ group Period range
Num
Sex ratio (M:F)
Mean age (years)
NF1 ratio
Resection degree ratio (subtotal:partial)
Ophthalmologic complications
Non-ophthalmologic complications
Death rate within 1 m
2003–2011 [24]
53
1.5:1
7.6
2/53 (3.8 %)
0:53
–
23/53 (43.4 %)
2/53 (3.8 %)
2013
11
0.8:1
8.1
1/11 (9.1 %)
10:1
3/11 (27.3 %)
4/11 (36.4 %)
0
Num number, M male, F Female, – not mentioned, m month
interforniceal approaches) rather than the lateral approach (such as the frontotemporal approach) to better protect the displaced bilateral optic nerves, which was consistent with Goodden et al. [23]. In addition, for inferior chiasmatic tumors, the transcallosal interforniceal approach and the incision in the perineurium should both be carefully selected to minimize the interference with the remaining optic chiasm. Compared with the previous series from the same group [24], most cases in this series gained more aggressive tumor resection, and better postoperative outcomes with the help of DTT images, as shown in Table 2. It should be noted that DTT imaging and the subsequent classification depend on intraoperative validation and verification. In this study, the intraoperative microscope findings were consistent with the preoperative DTT-determined positional relationships between the tumors and the visual pathway for most patients, which proved the technology in this study was feasible and reliable. The only one previous study of utilizing DTT with OPGs reported that there are no correlation between vision changes and the ability to image the optic fiber bundles [19]. In this series, the pre- and postoperative DTT images of the same patients most had changed substantially. These amplitude of variations are more obvious than their visual changes before and after the operations. These phenomenons may be related to the hemosiderosis and edema in the neighboring brain tissue postoperatively. It remains undetermined whether the sensitivity and specificity of DTT are sufficient to detect changes in vision, which need to be confirmed in the future. There is no doubt that this study has its limitations. Firstly, the DTT technology itself relies greatly on the subjective thinking of examiners, as the choice of ROIs depends on the operators’ understanding of brain anatomy and this disease [15–17, 19]. Secondly, long-term followup results including their DTT images and visual outcomes have not been able to be gathered right now, which needs more continuity of this cohort research.
Conclusion It is feasible and will be beneficial to utilize DTT to diagnose and treat children with OCGs. The application of
DTT to the peri-operative care of children with OCGs is valuable for increasing our understanding of the disease, developing reasonable surgical strategies, and helping determine the overall and visual prognosis of a patient. But, the exact relationship between DTT features and visual outcomes are still uncertain.
Funding This work was supported by the Beijing Nova Program (No. 2011086), which is a chinese official foundation. Conflict of interest The authors have declared no conflicts of interest. No financial interests or affiliations with institutions, organization, or companies are mentioned or have impacts on the views expressed in the article.
References 1. Shapey J, Danesh-Meyer HV, Kaye AH (2011) Diagnosis and management of optic nerve glioma. J Clin Neurosci 18:1585–1591 2. Kornreich L, Blaser S, Schwarz M et al (2001) Optic pathway glioma: correlation of imaging findings with the presence of neurofibromatosis. Am J Neuroradiol 22:1963–1969 3. Shofty B, Constantini S, Bokstein F, Ram Z, Ben-Sira L, Freedman S, Vainer G, Kesler A (2014) Optic pathway gliomas in adults. Neurosurgery 74:273–280 4. Binning MJ (2007) Optic pathway gliomas: a review. Neurosurg Focus 23(5):E2 5. Yeom KW, Lober RM, Andre JB, Fisher PG, Barnes PD, Edwards MS, Partap S (2013) Prognostic role for diffusionweighted imaging of pediatric optic pathway glioma. J Neurooncol 113:479–483 6. de Blank PM, Berman JI, Liu GT, Roberts TP, Fisher MJ (2013) Fractional anisotropy of the optic radiations is associated with visual acuity loss in optic pathway gliomas of neurofibromatosis type 1. Neuro Oncol 15:1088–1095 7. Lama G, Esposito SM, Grassia C, Calabrese E, Grassia MG, Bismuto R, Melone MA, Russo S, Scuotto A (2007) Neurofibromatosis type 1 and optic pathway glioma. A long-term followup. Minerva Pediatr 59:13–21 8. Avery RA, Liu GT, Fisher MJ, Quinn GE, Belasco JB, Phillips PC, Maguire MG, Balcer LJ (2011) Retinal nerve fiber layer thickness in children with optic pathway gliomas. Am J Ophthalmol 151:542–549 9. Andrew GL (2007) Neuroophthalmological management of optic pathway gliomas. Neurosurg Focus 23:E1 10. Wong JYC, Uhlv Wara WM, Sheline GE (1987) Optic gliomas. Cancer 60:1847–1855
123
366 11. Hiyama H, Kubo O, Nakajima H, Hayashi M, Kagawa M (1992) A case of cystic optic glioma involving chiasma and bilateral posterior optic pathway. No Shinkei Geka 20:1199–1204 12. Ahn Y, Cho BK, Kim SK, Chung YN, Lee CS, Kim IH, Yang SW, Kim HS, Kim HJ, Jung HW, Wang KC (2006) Optic pathway glioma: outcome and prognostic factors in a surgical series. Childs Nerv Syst 22:1136–1142 13. Dalla Via P, Opocher E, Pinello ML, Calderone M, Viscardi E, Clementi M, Battistella PA, Laverda AM, Da Dalt L, Perilongo G (2007) Visual outcome of a cohort of children with neurofibromatosis type 1 and optic pathway glioma followed by a pediatric neuro-oncology program. Neuro Oncol 9:430–437 14. Kanamori M, Shibuya M, Yoshida J, Takayasu M, Kageyama N (1985) Long-term follow-up of patients with optic glioma. Child’s Nerv Syst 1:272–278 15. Nickerson JP, Salmela MB, Koski CJ, Andrews T, Filippi CG (2010) Diffusion tensor imaging of the pediatric optic nerve: intrinsic and extrinsic pathology compared to normal controls. J Magn Reson Imaging 32:76–81 16. Nucifora PP, Verma R, Lee SK, Melhem ER (2007) Diffusiontensor MR imaging and tractography: exploring brain microstructure and connectivity. Radiology 245:367–384 17. Mukherjee P, Berman JI, Chung SW, Hess CP, Henry RG (2008) Diffusion tensor MR imaging and fiber. Am J Neuroradiol 29:632–641 18. Yamada M, Momoshima S, Masutani Y, Fujiyoshi K, Abe O, Nakamura M, Aoki S, Tamaoki N, Okano H (2008) Diffusiontensor neuronal fiber tractography and manganese-enhanced MR
123
J Neurooncol (2015) 122:357–366
19.
20.
21.
22.
23.
24.
25. 26. 27.
imaging of primate visual pathway in the common marmoset: preliminary results. Radiology 249:855–864 Lober RM, Guzman R, Cheshier SH, Fredrick DR, Edwards MS, Yeom KW (2012) Application of diffusion tensor tractography in pediatric optic pathway glioma. J Neurosurg Pediatr 10:273–280 Dodge HW, Love JG, Craig WM, Dockerty MB, Kearns TP, Holman CB, Hayles AB (1958) Gliomas of the optic nerves. AMA Arch Neurol Psychiatry 79:607–621 Wisoff JH, Abbott R, Epstein F (1990) Surgical management of exophytic chiasmatic-hypothalamic tumors of childhood. J Neurosurg 73:661–667 Kell JP, Leary S, Khanna P, Weiss AH (2012) Longitudinal measures of visual function, tumor volume, and prediction of visual outcomes after treatment of optic pathway gliomas. Ophthalmology 119:1231–1237 Goodden J, Pizer B, Pettorini B, Williams D, Blair J, Didi M, Thorp N, Mallucci C (2014) The role of surgery in optic pathway/hypothalamic gliomas in children. J Neurosurg Pediatrics 13:1–12 Tian YJ, Li DL, Jia G (2012) Clinical characteristics and treatment of pediatric optic pathway gliomas. Chin J Neurosurg 28:1136–1140 Miller NR, Iliff WJ, Green WR (1974) Evaluation and management of gliomas of the anterior visual pathways. Brain 97:743–754 Astrup J (2003) Natural history and clinical management of optic pathway glioma. Br J Neurosurg 17:327–335 Kelly JP, Weiss AH (2013) Detection of tumor progression in optic pathway glioma with and without neurofibromatosis type 1. Neuro Oncol 15:1560–1567
