Christian Schneider, Vijay Ramaswamy, Abhaya V. Kulkarni, James T. Rutka, Marc Remke, Uri Tabori, Cynthia Hawkins, Eric Bouffet, and Michael D. Taylor
While medulloblastoma was initially thought to comprise a single homogeneous entity, it is now accepted that it in fact comprises 4 discrete subgroups, each with its own distinct demographics, clinical presentation, transcriptomics, genetics, and outcome. Hydrocephalus is a common complication of medulloblastoma and not infrequently requires CSF diversion. The authors report the incidence of CSF diversion surgery in each of the subgroups of medulloblastoma (Wnt, Shh, Group 3, and Group 4).
The medical and imaging records for patients who underwent surgery for medulloblastoma at The Hospital for Sick Children were retrospectively reviewed. The primary outcome was the requirement for CSF diversion surgery either before or within 60 days of tumor resection. The modified Canadian Preoperative Prediction Rule for Hydrocephalus (mCPPRH) was compared among subgroups.
Of 143 medulloblastoma patients, treated from 1991 to 2013, sufficient data were available for 130 patients (15 with Wnt, 30 with Shh, 30 with Group 3, and 55 with Group 4 medulloblastomas). Of these, 28 patients (22%) ultimately underwent CSF diversion surgery: 0% with Wnt, 29% with Shh, 29% with Group 3, and 43% with Group 4 tumors. Patients in the Wnt subgroup had a lower incidence of CSF diversion than all other patients combined (p = 0.04). Wnt patients had a lower mCPPRH score (lower risk of CSF diversion, p = 0.045), were older, had smaller ventricles at diagnosis, and had no leptomeningeal metastases.
The overall rate of CSF diversion surgery for Shh, Group 3, and Group 4 medulloblastomas is around 30%, but no patients in the present series with a Wnt medulloblastoma required shunting. The low incidence of hydrocephalus in patients with Wnt medulloblastoma likely reflects both host factors (age) and disease factors (lack of metastases). The absence of hydrocephalus in patients with Wnt medulloblastomas likely contributes to their excellent rate of survival and may also contribute to a higher quality of life than for patients in other subgroups.
Jennifer L. Quon, Michelle Han, Lily H. Kim, Mary Ellen Koran, Leo C. Chen, Edward H. Lee, Jason Wright, Vijay Ramaswamy, Robert M. Lober, Michael D. Taylor, Gerald A. Grant, Samuel H. Cheshier, John R. W. Kestle, Michael S. B. Edwards, and Kristen W. Yeom
Imaging evaluation of the cerebral ventricles is important for clinical decision-making in pediatric hydrocephalus. Although quantitative measurements of ventricular size, over time, can facilitate objective comparison, automated tools for calculating ventricular volume are not structured for clinical use. The authors aimed to develop a fully automated deep learning (DL) model for pediatric cerebral ventricle segmentation and volume calculation for widespread clinical implementation across multiple hospitals.
The study cohort consisted of 200 children with obstructive hydrocephalus from four pediatric hospitals, along with 199 controls. Manual ventricle segmentation and volume calculation values served as “ground truth” data. An encoder-decoder convolutional neural network architecture, in which T2-weighted MR images were used as input, automatically delineated the ventricles and output volumetric measurements. On a held-out test set, segmentation accuracy was assessed using the Dice similarity coefficient (0 to 1) and volume calculation was assessed using linear regression. Model generalizability was evaluated on an external MRI data set from a fifth hospital. The DL model performance was compared against FreeSurfer research segmentation software.
Model segmentation performed with an overall Dice score of 0.901 (0.946 in hydrocephalus, 0.856 in controls). The model generalized to external MR images from a fifth pediatric hospital with a Dice score of 0.926. The model was more accurate than FreeSurfer, with faster operating times (1.48 seconds per scan).
The authors present a DL model for automatic ventricle segmentation and volume calculation that is more accurate and rapid than currently available methods. With near-immediate volumetric output and reliable performance across institutional scanner types, this model can be adapted to the real-time clinical evaluation of hydrocephalus and improve clinician workflow.
Michal Zapotocky, Daddy Mata-Mbemba, David Sumerauer, Petr Liby, Alvaro Lassaletta, Josef Zamecnik, Lenka Krskova, Martin Kyncl, Jan Stary, Suzanne Laughlin, Anthony Arnoldo, Cynthia Hawkins, Uri Tabori, Michael D. Taylor, Eric Bouffet, Charles Raybaud, and Vijay Ramaswamy
Metastatic dissemination is a major treatment challenge and cause of death in patients with medulloblastoma. However, the influence of molecular biology on the pattern of metastatic dissemination at diagnosis is not known. In this study, the authors sought to define the location, pattern, and imaging characteristics of medulloblastoma metastases across subgroups at diagnosis.
A consecutive cohort of patients with metastatic medulloblastoma at The Hospital for Sick Children and the University Hospital Motol, who underwent up-front MRI of the craniospinal axis, was assembled and allocated to subgroups using NanoString limited gene–expression profiling. Radiological characteristics (including location, morphology, size, diffusion restriction, and contrast enhancement) were discerned through a retrospective review.
Forty metastatic medulloblastomas were identified with up-front neuroimaging of the craniospinal axis: 5 sonic hedgehog (SHH), 16 Group 3, and 19 Group 4 metastases. Significant subgroup-specific differences were observed, particularly with respect to tumor location, size, and morphology. Group 3 metastases were most frequently laminar compared with a more nodular pattern in Group 4 (14 of 16 in Group 3 vs 8 of 19 in Group 4; p = 0.0004). Laminar metastases were not observed in patients with SHH medulloblastoma. Suprasellar metastases are highly specific to Group 4 (p = 0.016). Two of the 5 SHH cases had multifocal lesions in the cerebellum, raising the possibility that these were in fact synchronous primary tumors and not true metastases. A minority of patients with Group 4 metastases harbored metastatic deposits that did not enhance on MRI after contrast administration, often in patients whose primary tumor did not enhance.
The location, morphology, and imaging characteristics of metastatic medulloblastoma differ across molecular subgroups, with implications for diagnosis and management. This suggests that the biology of leptomeningeal dissemination differs among medulloblastoma subgroups.
Christian Schneider, Ian Kamaly-Asl, Vijay Ramaswamy, Lucie Lafay-Cousin, Abhaya V. Kulkarni, James T. Rutka, Marc Remke, Daniel Coluccia, Uri Tabori, Cynthia Hawkins, Eric Bouffet, and Michael D. Taylor
Choroid plexus carcinomas (CPCs) are rare brain tumors originating from the ventricular choroid plexus. They account for 2%–4% of all pediatric brain tumors and are most frequently seen in very young children. This pediatric proclivity, in combination with a marked vascularity, renders an aggressive resection a difficult and often dangerous endeavor. Blood losses of several total blood volumes in small children are not uncommon, sometimes forcing the neurosurgeon to abort the procedure, often leaving residual tumor. Great extent of tumor resection is an accepted beneficial factor for overall survival. Therefore, a second resection usually follows the administration of adjuvant chemotherapy. Second-look surgery appears to be associated with markedly decreased blood loss. Histological examination of specimens obtained at a second intervention shows decreased vascularity and fibrotic changes in tumor tissue. At the Hospital for Sick Children in Toronto, this empirical finding led to the strategy of neoadjuvant chemotherapy to minimize blood loss and maximize cytoreduction. The authors undertook this study to assess the potentially beneficial effect of neoadjuvant chemotherapy on blood loss during surgery for CPCs.
In this retrospective cohort review, the demographic, clinical, and treatment parameters of 22 consecutive patients diagnosed with CPC are presented. All underwent surgical treatment at the Hospital for Sick Children from 1982 to 2013. Special attention was given to the impact of neoadjuvant chemotherapy on extent of resection and intraoperative blood loss. Extent of resection was calculated based on perioperative neuroimaging, and amount of blood loss was estimated based on transfusion parameters and perioperative changes in hematocrit.
Ten patients did not receive neoadjuvant chemotherapy, and 12 were treated with 2–5 cycles of ICE (ifosfamide, carboplatin, etoposide) chemotherapy in a neoadjuvant fashion. The 22 patients included in the study underwent a total of 37 tumor resection surgeries. In all of the cases in which neoadjuvant chemotherapy was used, at least a near-total resection (> 95% of tumor volume) was achieved. Patients who underwent gross-total resection had prolonged overall survival. Of the 37 resections, 18 were performed after chemotherapy. Mean blood loss in the neoadjuvant chemotherapy group was 22% of total estimated blood volume as opposed to 96% in patients without preoperative chemotherapy.
In children with CPC, the administration of neoadjuvant chemotherapy decreases intraoperative blood loss and increases extent of resection with a significant positive effect on overall survival.