Predictive Factors for Post-Operative Tracheostomy Requirement in Children Undergoing Surgical Resection of Medulloblastoma

Purpose: Medulloblastoma is the most common pediatric malignancy. Postoperative respiratory failure is an important complication in post-operative recovery of patients undergoing medulloblastoma resection. We aimed to identify factors predicting tracheostomy requirement in children post-operatively. Methods: Retrospective chart review of all patients under 18 undergoing medulloblastoma resection from 2012 to 2020 at Namazi hospital was conducted. Results: 45 patients (26%) needed tracheostomy after the operation. The most common indications were brainstem compression and absence of gag reex prior to operation. Patients who had brainstem compression and inltration by medulloblastoma, bilateral absence of gag reex prior to operation, subtotal resection of tumor, and post-operative brainstem contusion were more likely to require tracheostomy. No statistically signicant difference was observed between males and females and different ages. Conclusion: The results show that if we prevent the invasion of the brainstem by the tumor and resect the tumors totally and accurately, tracheostomy, a highly costly and stressful post-operative complication can be prevented.


Introduction
Brain tumors are the leading cause of cancer-related mortality in children [1]. Medulloblastoma (MB) accounts for approximately 20% of all pediatric central nervous system (CNS) malignancies; making MB the most common malignant childhood brain tumor [2]. The current convention for treating Children With MB comprises a combination of tumor resection, craniofacial radio therapy, and chemotherapy based on the patient's risk category, age, etc. [3]. Recently, a study was published on postoperative tracheostomy requirement in children undergoing surgery for posterior fossa tumors. Apart from this study, there is limited information on post-operative tracheostomy in children with MB undergoing surgical resection.
Moreover, this study was limited by the number of MB cases. Only 65 MB patients were included in this study; solely one of them requiring tracheostomy.
The aim of the present study was to identify pre-and post-operative clinical and radiological factors associated with tracheostomy requirement in children with MB who underwent surgical resection. In the study by Goethe et al., having postoperative dysphagia, ependymoma or astrocytoma (as opposed to MB and atypical teratoid/rhabdoid tumor), and being younger were found to be linked to tracheostomy requirement in children undergoing surgery for posterior fossa tumors [4].

Methods And Materials
A retrospective chart review was conducted of all of the children who underwent initial MB resection between April 2012 and September 2020 at Namazi hospital, an academic tertiary referral center. This study investigates patients under 18-years of age with posterior fossa tumors diagnosed as MB by pathology. Patients with incomplete medical records, equivocal tumor pathology, prior tumor resections, or age over 18 years were excluded from the study. Additionally, patients with any abnormality in chest Xray obtained prior to the operation or those who were already tracheostomy-dependent were excluded from this study. The nal extracted raw data was cross-examined by repeating the data collection for 20% of the patients randomly.
Demographic, clinical, pathologic, and outcome data were examined retrospectively. Variable investigated in this study for correlation with Tracheostomy tube insertion include the presence of hydrocephalus, application of ventriculoperitoneal (VP) shunt, brainstem compression and invasion (in ltration), whether MB tumors were located laterally in the cerebellopontine angle (CPA), preoperative bilateral absence of Gag Re ex (GR), presence of cystic degeneration in the tumor, whether the tumor was totally resected, presence of calci cation in the tumor, postoperative brainstem contusion, age, and sex. This study was carried out between September and December 2020.
Data analysis was carried out using IBM SPSS Statistics version 16 (IBM Corp., Armonk, N.Y., USA).
Comparisons were made using Chi-square tests, Fisher's exact tests, or independent-samples t tests as appropriate. Comparisons were considered signi cant at the p < 0.05 level. Ethical approval was waived in view of the retrospective nature of the study and all the procedures being performed were part of the routine care.

Demographic Findings
There was no statistically signi cant difference between the age and sex of the patients and tracheostomy outcome in these patients.

Clinical and Radiologic Findings
Patients who had brainstem compression and in ltration by MB, bilateral absence of GR prior to operation, subtotal resection of the tumor, and post-operative brainstem contusion were more likely to require tracheostomy. The presence of cystic degeneration, calci cation in the tumor, cervical invasion, and whether the tumor was located at CPA and the tracheostomy outcomes in these patients were not associated with tracheostomy requirement post-operatively.

Discussion
Our results show a remarkably high frequency of tracheostomies in patients whose GR was absent or brainstem was invaded, either compressed or in ltrated by the MB tumors. We found the incidence of tracheostomy requirement in our center for children undergoing MB tumor resection to be 26%, which is relatively high compared to previous studies (1.5%, 16%, and 6.4%) [4][5][6].
Tracheostomy placement may improve respiratory compromise caused by compression of the brainstem by MB tumor or resection operation. However, it is burdensome for the patient and family, challenges the clinical course and quality of life of the patients. A report suggests that it may adversely affect language development [7]. Moreover, tracheostomy placement can cause infection, tracheoinnominate stula, stress on caregivers, place patients at risk for longer hospital stays and death, and impose a great nancial burden on patients, families, and the healthcare system [7][8][9][10].
An important consideration in interpreting the results is that in our center the standard of care for MB resection involves VP shunt insertion two weeks prior to nal operation. Although not everyone will be shunt-dependent for the rest of their lives, VP shunt insertion has several advantages. It provides a period of time for healing of the shunt insertions site. Moreover, there is a risk of upward herniation associated with ETV endoscopic third ventriculostomy in posterior fossa tumor. To be brief, VP shunt application is more favorable in our setting unless proven otherwise. The dominant approach in resecting MB tumors in our center is the midline transvermian posterior fossa approach is the approach practiced by pediatric neurosurgeons in resecting MB tumors in our center.
Postoperatively, we tend to keep the patients intubated for approximately two weeks. Although early tracheostomy is supported, we take a more conservative stance toward tracheostomy. It was noted before that the tracheostomy requirement in patients with brainstem tumors re ects the interruption of respiratory pathways located there [11]. After tumor resection, pressure on the nerves is reduced which might improve the lower cranial neuropathy and help the ventilation function to recover, particularly in pediatric patients. Besides, tracheostomy care is highly costly for the patients and not widely available. It is shown in the results that although the absence of GR is among the most common indications for tracheostomy. However, one-fourth of the patients with absent GR pre-operatively did not require tracheostomy after the operation. It is worth considering that in previous studies, 25% or more of the patients were decannulated one year after the tracheostomy [4,5]. Moreover, once the course of tracheostomy is nished, a second tracheostomy is rarely needed. Recannulation rate in the pediatric setting was measured at 6.5 % [12].
It should be considered that all of the tumors resected in our center were greater than 5 cm in diameter. It points to the ineffective system for screening brain tumors in Iran, resource scarcity, and the long waiting lists.
The lack of an association between the extension of the tumors to CPA and cervical invasion and tracheostomy requirement likely represents the limitation of small sample size and outcome of interest. In previous studies, it was suggested that younger patients were more likely to require tracheostomy. However, the statistical difference between the younger and older patients requiring tracheostomy was not statistically signi cant.
The greatest limitation of the present is the retrospective design of this investigation. This study does not follow the patients long after the operation. The nal outcome of the tracheostomies, whether the patients were expired or decannulated, is not reported. However, the sample size of this study is exceptionally large. Although the patients were not followed long after the operations, valuable factors predicting the requirement of post-operative tracheostomy were included and analyzed. This study can give insight to practitioners and help them to make informed predictions about their needs, patients who may bene t from early tracheostomy, the outcome of the operations, and reduce intubation-associated