Perioperative Management of Pediatric Giant Supratentorial Tumors: Challenges and Management Strategies.

BACKGROUND: Intracranial tumors are the most common pediatric solid tumors. Only one-third of these tumors arise from the supratentorial compartment. The abnormal intracranial tumors are unusual but can bleed to an extent causing hemorrhagic shock necessitating blood transfusion in the perioperative period. The perioperative management of these subset of patients poses a unique challenge to both the neurosurgeons and the neuroanesthetic team.
MATERIALS AND METHODS: This study included a case series of 30 patients with giant supratentorial neoplasms who underwent craniotomy and tumor resection from 2014 to 2017 in our Tertiary Care Institute. The clinical data were collected from the patient's records obtained from the Department of Neurosurgery, National Institute of Mental Health and Neurosciences (NIMHANS), Bangalore, Karnataka, India. The aim of this case series was to characterize the perioperative challenges, management strategies, course, and outcome in 30 children who were operated for elective or emergency resection of giant supratentorial lesions in our hospital. We also reviewed the literature available to guide the anesthetic management of pediatric patients with intracranial tumors.
RESULTS: Among the 30 patients, four had significant intraoperative fluid shifts necessitating massive blood transfusion perioperatively. The overall incidence of mortality in our study cohort was 16.67% (5/30).
CONCLUSION: The maintenance of systemic physiological homeostasis by anticipation of complications, vigilant monitoring, and prompt resuscitation is critical to foster favorable outcomes in unison with optimal and safe surgical extirpation of the primary cerebral lesion. Copyright:

Introduction

P ediatric neurosurgery and pediatric anesthesia for large tumors are challenging to even the very experienced. Intracranial tumors are the most common pediatric solid tumors. Only one-third of these tumors arise from the supratentorial compartment. Supratentorial Pediatric tumors are uncommon and the literature is sparse with respect to infratentorial tumors. In India, children with cerebral tumors usually present late, hence having to be considered for urgent craniotomies for definitive treatment. The perioperative management of these subset of patients poses a unique challenge to both the neurosurgical and the neuroanesthetic team. The surgical resection of giant intracranial tumors can cause significant bleeding, which can progress to hemorrhagic shock necessitating perioperative blood transfusion. These perioperative fluid shifts are dicey to approach for stabilization in pediatric patients. The circulating blood volume should be optimized to avoid the hazardous extremes of hypoperfusion-induced cerebral ischemia and fluid overload. The successful management of these patients especially in the perioperative period comes with various inherent challenges to encounter. The anesthetic and surgical techniques should be tailored with appropriate anticipation of complications and adequate preparation to foster favorable outcomes in these subset of patients. Here, we present a case series of the perioperative course (including fluid shifts, hypotension, hypothermia, acid base imbalance, hypoxia, and delayed recovery) of 30 pediatric supratentorial neuro-onco-surgical patients operated over a period of three years, from 2014 to 2017 in National Institute of Mental Health and Neurosciences (NIMHANS), Bangalore, Karnataka, India––a tertiary care neurosciences hospital. The aim of this case series was to characterize the current practices in the perioperative management of pediatric patients with giant supratentorial tumors. We also reviewed the literature available on the anesthetic management in these subset of patients.

Subjects and Methods

The data about pediatric patients diagnosed with giant supratentorial tumors who underwent craniotomy and tumor resection over a period of three years from May 2014 to June 2017 were collected from the patient’s records from the Department of Neurosurgery, NIMHANS. The patient’s records, including the intraoperative anesthetic records of 30 pediatric patients, were reviewed retrospectively to collect data suggesting the perioperative course and the functional/neurological outcome of these patients. We also discussed the possible factors that could have influenced the functional outcome (condition at discharge) in these children.

All the patients were managed by the Departments of Neurosurgery, Neuroanaesthesiology, and Neuro-Critical Care. Additional data from the patient’s case file were used to collect the pre- and postoperative patient information.

The following information was recorded from each patient:

Patient demographic characteristics.

Presenting symptoms.

Location and histopathological diagnosis of the supratentorial lesion.

Perioperative fluid shifts.

Perioperative complications.

Patient’s condition at discharge and during the subsequent follow-ups.

Operative details of all procedures performed on these patients were analysed for significant perioperative events.

Results

Thirty children who presented with giant supratentorial tumors underwent craniotomy for surgical tumor excision. The demographic characteristics of the patients are depicted in Table 1. The mean age of the patients at presentation was 2.36 years (range 1–4 years). There were 17 male and 13 female patients. Eleven patients presented with seizures, seven with limb weakness, four with only headache, and six with intracranial hypertension (ICHTN). Gross total excision could be achieved in 21 patients and subtotal in 9 patients. Diagnosis included choroid plexus papilloma (26.6%), anaplastic ependymoma (20%), glioma (16.6%), astrocytoma (13.33%), meningioma (6.6%), ganglioneuroblastoma (6.66%), and 3.33% each of germinoma, teratoid ,and xanthomatous lesion as shown in Table 2. The intraoperative blood loss exceeded the minimum allowable volume in 24 (80%) of 30 patients necessitating blood transfusion in 22 (73.3%) of 30 patients. The intraoperative fluid shifts and complications are depicted in Tables 3 and 4, respectively. Massive blood loss exceeding the patient’s entire blood volume[1] occurred in 4 (13.33%) of 30 patients triggering massive transfusion protocol. Among them, 1 (3.33%) patient developed transfusion-related acute lung injury (TRALI). There was a 3.33% (1/30) incidence each of diabetes insipidus, ventriculitis, meningitis with ipsilateral posterior cerebral artery infarct, brain abscess with bilateral internal carotid artery infarct, and tumor recurrence, which complicated the postoperative course of these 30 children. The child with tumor (ganglioneuroblastoma grade IV) recurrence presented with signs of ICHTN and neurogenic pulmonary edema following one year after the primary surgery and four cycles of radiotherapy. The overall case fatality was 16.66% (5/30). The histopathological diagnosis in these five patients were anaplastic ependymoma III, ganglioneuroblastoma IV, xanthomatous lesion, pilocytic astrocytoma, and choroid plexus papilloma. These five children who succumbed to death had an apparently uneventful perioperative course except for one.

Table 1

The demographic parameters, tumor characteristics, anesthetics administered, and neurological outcome

Case	Age	Pathology	Presentation	Size and tumor localization	Postoperative complications	Outcome
1	3	Germinoma/pinealoblastoma IV	Ear discharge	Third ventricular lesion	Decline in sensorium (M6–M5)	M5
2	3	Anaplastic ependymoma	Left neck tilt, seizures	Left fronto parietal lesion	–	M6
3	3	Rhabdoid/teratoid IV	Right upper limb weakness	Right capsuloganglionic lesion	–	M6
4	3	Anaplastic ependymoma III	Seizures	Left frontal lesion	Ventriculitis, decline in sensorium (M6–M4)	M4
5	3	Giant cell astrocytoma I	Seizures	Right temperoparietal lesion	–	M6
6	3	High-grade glioma	Seizure, vomiting, headache	Petro clival lesion	–	M6
7	3	GBM	Seizures	Right temperoparietal lesion	–	M6
8	2	Anaplastic ependymoma	Headache, vomiting	Right thalamic lesion	–	M6
9	1	Choroid plexus papilloma	Vomiting, right limb weakness	Left thalamic lesion	–	M6
10	2	Glioma	Complex partial seizures	Right temporal lesion	–	M6
11	1	Anaplastic ependymoma	Headache, fever	Right fronto parietal lesion	Decline in sensorium (M6–M4)	M6–M1, expired
12	3	Ganglio neuroblastoma IV	Giddiness, headache, vomiting	Left parieto occipital lesion	Recurrence, with ICHTN, NPE	Expired after one year
13	2	Choroid plexus papilloma	Headache, vomiting	Intra ventricular lesion	–	M6
14	1	Choroid plexus papilloma	Right-sided weakness	Left lateral ventricular lesion	–	M6
15	2	Pilomyxoid astrocytoma	Seizures	Right tempero parietal lesion	–	M6
16	2	Pilomyxoid astrocytoma	Headache	Suprasellar lesion with hydrocephalus	–	M6
17	2	Meningioma	Left-sided weakness, headache	Right falcotentorial lesion	–	M6
18	2	Recurrent ependymoma	Headache	Left temperoparietal lesion	–	M6
19	3	GBM	Right-hand tremors	Left thalamic lesion	–	M6, new onset right sided weakness
20	2	Xanthomatous lesion	Seizures, altered sensorium	Third ventricular lesion	Meningitis left PCA infarct	Expired after one month
21	2	Pilocytic astrocytoma	Increase in head size, vomiting	Third ventricular lesion	Decline in sensorium (M6–M2), diabetes insipidus	E1VTM2 expired
22	4	Choroid plexus papilloma	Increase in head circumference	Right trigonal lesion	Brain abscess, B/L ICA infarct	M2, expired
23	2	Anaplastic ependymoma	Left-sided weakness, seizures	Right frontal recurrent lesion	–	M6
24	2	Ganglioneuroblastoma IV	Complex partial seizures	Right parieto occipital lesion	–	M6
25	3	Glioma	Seizures	Optochiasmatic lesion	–	M6
26	3	Choroid plexus papilloma	Headache, vomiting	Lateral ventricular	–	M6
27	2	Atypical meningioma	Imbalance while walking	Posterior third ventricular lesion	Decline in sensorium (M6–M3)	M3
28	3	Anaplastic ependymoma IV	Imbalance while walking	Lateral ventricular	TRALI	M6
29	2	Choroid plexus papilloma	Headache, vomiting, left sided weakness	Right lateral ventricular	–	M6
30	2	Choroid plexus carcinoma	Vomiting, right-sided weakness	Left lateral ventricle	–	M6

M = motor component of the Glasgow Coma Scale, ICHTN = intracranial hypertension, NPE = neurogenic pulmonary edema, PCA = posterior cerebral artery, ICA = internal carotid artery, TRALI = transfusion-related acute lung injury

Table 2

The incidence of different tumor types based on histopathological diagnosis

Diagnosis	Number	Percentage	Recurrence	Death
Anaplastic ependymoma	6	20%		1
Glioma	5	16.6%
Ganglioneuroblastoma	2	6.66%	1	1
Xanthomatous lesion	1	3.33%		1
Choroid plexus papilloma	8	26.6%		1
Astrocytoma	4	13.33%		1
Meningioma	2	6.6%
Germinoma	1	3.33%
Teratoid	1	3.33%

Table 3

The anesthetics administered and the intraoperative fluid management

Case	Anesthetic induction	Anesthetic maintenance	Blood transfused	Use of colloids (mL)	Total IVF (mL)	Urine output (mL)	FFP	PLT
1	Thiopentone	Sevoflurane	450		500	350
2	Thiopentone	Sevoflurane	800		1000	700	200	100
3	Propofol	Sevoflurane	100		1100	850
4	Propofol	Sevoflurane	–		600	250
5	Thiopentone	Sevoflurane	200		900	250
6	Thiopentone	Desflurane	100		1000	600
7	Propofol	Sevoflurane	–		450	200
8	Thiopentone	Sevoflurane	–		400	150
9	Propofol	Sevoflurane	900	250	1000	400	100	50
10	Thiopentone	Isoflurane	350		500	200	100
11	Thiopentone	Sevoflurane	100		500	250
12	Propofol	Sevoflurane	1100		1500	1000
13	Propofol	Sevoflurane	800		200	100
14	Thiopentone	Sevoflurane	300	100	900	600
15	Thiopentone	Sevoflurane	450		400	300
16	Thiopentone	Sevoflurane	300		500	300	100
17	Thiopentone	Sevoflurane	–		700	200
18	Thiopentone	Isoflurane	500		1000	150
19	Propofol	Sevoflurane	200		1300	350
20	Thiopentone	Isoflurane	200		400	300
21	Nil	300	100		1500	500
22	Hemorrhagic shock	1000	500		2000	600	50	50
23	Nil	150	–		800	500
24	Yes	400	200		1000	700
25	Nil	150	–		1000	500
26	Nil	300	200		300	400
27	Nil	200	–		1250	700
28	Yes	800	800		1500	600	50
29	Nil	150	150		200	150
30	Nil	200	–		1000	350

FFP = fresh frozen plasma, IVF = intravenous fluid, PLT = platelets transfused, EBL = estimated blood loss

Table 4

The incidence of intraoperative complications

Intraoperative complications	Number	Percentage
Significant blood loss	24	80%
Blood transfusion	22	73.3%
Massive blood loss	4	13.33%
Massive blood transfusion	4	13.33%
Hypotension	8	26.66%
Hypothermia	2	6.66%
Delayed recovery	4	13.33%

One child had preoperative anemia with a hemoglobin (Hb) of 8g/dL and had to be operated for emergent tumor excision because of poor neurological status at presentation. With limited time to correct anemia preoperatively, the perioperative homeostasis was maintained and hemodynamic fluctuations were managed with ongoing blood transfusion.

All surgeries performed in the 30 patients included in this study were conducted under general anesthesia with endotracheal intubation and controlled mechanical ventilation. The intravenous (IV) anesthetic induction technique was administered using thiopentone injection in most cases (25/30) followed by preference propofol injection (5/30). The balanced anesthesia technique was used to maintain the anesthesia. The inhalational anesthetic (sevoflurane in 24/30 vs. isoflurane in 6/30 patients) below one minimal alveolar concentration was used for the maintenance of anesthesia, along with intermittent opioid boluses and neuromuscular blocking agents. None of the patients received total intravenous anesthesia (TIVA).

The haemostatic pharmacotherapy using tranexaemic acid injection was administered during 12 of 30 surgeries, in anticipation of significant blood loss.

Discussion

Intracranial tumors are the second most common pediatric tumors only after leukemia.[2] They represent approximately 35% of all pediatric malignancies and remain a prominent cause of cancer-related childhood deaths.[3]

Of the average of 130 pediatric intracranial tumors that present to our hospital in a year,[4] giant neoplasms located in the supratentorial region were selected because of the additional inherent challenges encountered in the management of these patients.

The choroid plexus papilloma––the most common diagnosis (26.6%) in our study––was seen in almost one-third of our study population (8/30 patients). This was contrary to literature in which astrocytomas (13.3%) were the commonest in a pediatric population.[4] We attribute this difference to the restriction of our study cohort to giant sized lesions. The pediatric glial neoplasms and primitive embryonal neoplasms are other common encounters in the context of Indian pediatric population.

The treatment decision and perioperative management in pediatric giant supratentorial lesions should consider the developing and maturing physiological and neurological status. Hypotension and hypoxia significantly influence the outcome in neurosurgical patients. The cerebral blood flow (CBF) is coupled to the cerebral metabolic rate of oxygen (CMRO2), which is higher in the pediatric population (5.2mL/100g/min) than that in adults (3.5mL/100g/min), rendering them less tolerant to hypoxia. The uncontrolled fluctuations in blood pressure beyond the autoregulatory range predispose to cerebral ischemia and intraventricular hemorrhage.[5] Hence, the optimization of both cerebral perfusion and oxygenation cannot be overemphasized.

The time window to preoperatively optimize the hematological parameters of the patients is usually minimized in children who often present late with abnormal tumors, when the compensatory mechanism by macrocephaly is exhausted/overwhelmed causing ICHTN. Further delay in definitive management can cause brain herniation and prove fatal. The children diagnosed with perisellar lesions, most common of which is craniopharyngioma, may have associated preoperative diabetes insipidus because of hypothalamic involvement. Optimizing the intravascular volume is an additional challenge in these patients. None of our patients had a diagnosis of perisellar lesion.

The decision for either gross total excision or partial excision (to be followed by chemotherapy or radiotherapy) depends on the tumors histopathology, size and the tumors grade at presentation. Maximal safe resection of gliomas has been found to have an important prognostic role in patient survival and outcome.[6] The performance of safe maximal resection/near total excision aiming for good outcome comes with an associated potential for massive blood loss triggering massive blood transfusion (MBT) protocol with its attendant complications.

Children presenting with hydrocephalus with or without associated congenital anomalies can present a difficult airway to manage in the intraoperative period. Fortunately, none of our study patient’s airway were difficult to manage. All the 30 children received anesthetic induction using IV agents. The choice of anesthetic drugs used in this study were related to availability and therefore may not have much influence on the study outcome. Hence, this was not analyzed as an outcome predictor. But this can be considered useful for future reference.

A balanced anesthetic technique was used in all patients with preference for inhalational anesthetics rather than TIVA for maintenance. Fentanyl was the most commonly administered opioid for analgesia (30/30) with concomitantly administered morphine in 10/30 patients. Atracurium was preferentially used for neuromuscular blockade in 26/30 followed by rocuronium in the remaining 4/30 children.

Controlled mechanical ventilation was used to titrate end tidal carbon dioxide levels targeting the arterial partial pressures of carbondioxide to acceptable limits. Children on chronic anticonvulsant therapy required more frequent boluses of muscle relaxants and narcotics because of the enzymatic induction of these drugs. Positioning of patients for surgery may also influence the degree of blood loss, and hence reverse Trendelenburg position technique favored for neurosurgeries. All our 30 patients had craniotomy performed in the supine position. Other measures to minimize blood loss, such as autologous predonation, acute normovolemic hemodilution, and hypovolemic hemodilution, have limited applications in pediatric population. Cell salvage, although a feasible option, has its own technical limitations[7] and fear of inadvertent autotransfusion of tumors seedlings.

Such prolonged surgeries may predispose these children to hypothermia, which can augment surgical bleeding by inducing coagulopathy. Hence it is mandatory to monitor the core body temperature and use warming techniques such as warm air blowers and fluid-warming devices. This explains the low incidence of only two of 30 patients becoming hypothermic despite the above measures. Blood loss was significant in one of these two patients who were hypothermic.

Intracranial surgery may be associated with hemodynamic changes because of sudden blood loss, venous air embolism (VAE), and manipulation of cranial nerves. Transducing of intra-arterial pressures to capture beat-to-beat blood pressure changes is hence indicated and was monitored in most (26/30) of these patients as determined by the size and location of their tumors. Arterial cannula placement also aided in the frequent analysis of arterial blood gas, electrolytes, blood glucose, and Hb values.

Massive fluid shifts can occur during surgical resection of abnormal intracranial tumors, contributing to significant hemodynamic instability. Extensive bleeding can occur during scalp dissection, periosteal elevation, and tumors resection. Massive blood loss is one cardinal factor among the various contributors to significant morbidity in the perioperative period. Intraoperative hemodynamic instability may dictate or permit only partial tumor excision leaving residue. In addition to this, some tumors post-resection have high recurrence rates and present for a reexploration/redo surgery. The incidence of tumor recurrence in our study cohort was 1/30 (3.33%) with a histopathological diagnosis of ganglioneuroblastoma IV.

Early prediction of candidates requiring perioperative MBT is critical as it helps optimize resource use with adequate preparedness while minimizing the waiting period to avail blood or blood products in the event of an emergent indication. MBT exceeding 50% of total blood volume within 3h was triggered in four of our patients.[8]

The estimated blood loss exceeding the allowable maximal limit (calculated based on the estimated blood volume and deficit in hematocrit from baseline), patient’s hemodynamic status, extent and rate of ongoing blood loss, and the stage of tumors resection should collectively be considered to decide on the need, type, and volume of transfusion, aimed at restoring the circulating blood volume and oxygen-carrying capacity. Our institutional practice entails goal-directed transfusion with close monitoring of bleeding, hematological parameters, and point of care coagulation testing rather than a fixed-ratio-driven massive transfusion of blood and blood components. Among 22 of 30 (73.33%) patients who received blood transfusion, four patients received only whole blood. Only packed cells (PRBCs) were transfused in 12 patients. A combination of PRBC, platelets, and plasma were transfused in the remaining six patients in the ratio approximately 1.5–2:1:1.

MBT can be associated with untoward complications in the form of hypothermia, coagulopathy, electrolyte abnormalities (hypocalcemia, hypomagnesemia, hypokalemia, and hyperkalemia), acid/base derangements, citrate toxicity, and transfusion-associated acute lung injury.[9] One of the four patients who received MBT developed TRALI and recovered following two weeks of intensive care. Of the two patients who became hypothermic, one patient received MBT, which did not precede the low core body temperatures recorded in this patient.

A close communication between the neurosurgeon and the neuroanesthetists plays a crucial role in the anticipation and management of untoward significant intraoperative events. Other complications such as seizures, VAE, and arrhythmias were reported in none of the patients.

Normal saline is the most commonly administered crystalloid during pediatric neurosurgical procedures as it is mildly hyperosmolar and hence prevents cerebral edema. This was the most commonly used IV fluid in the patients of our study followed by Ringer’s lactate solution. Only two patients received colloid during surgery.[10]

Rapid awakening for early assessment of neurological status is the goal at the end of neurosurgical procedures. However, four (13.33%) of 30 patients had a delay in recovery from anesthesia. The anesthetic agent used during the perioperative period can partly influence the course of emergence from anesthesia. Due to the fact that sevoflurane offers a better recovery profile than isoflurane in pediatric cohorts,[11] of the four patients who had delayed recovery, only one patient received the former and three received the latter inhalational agent.

Decision to keep the trachea intubated postoperatively was individualized, based on the preoperative neurological status, perioperative adverse events, and deterioration of neurological status in the postoperative period. Nine of 30 patients had their trachea remained intubated in the immediate postsurgical period. Although five of nine patients were extubated on postoperative days (PODs) 1 and 2, the remaining four patients were never extubated (due to deterioration of neurological status) and were among the five of30 patients who succumbed to death despite intensive treatment. One of the deceased patent was extubated in the operating room and later re-intubated on POD 1 due to neurological deterioration associated with seizures.

The results were found better in those with seizures as the presenting symptom as compared with those who presented with raised intracranial pressure (ICP) similar to the findings seen in a case series on parturients who underwent craniotomy for cerebral tumors resections.[12]

Study strengths

We attempt to compile all the data that may be contributive to the outcome in children managed for giant supratentorial lesions. The restriction of analysis to only supratentorial pediatric population serves to provide a definitive solution to a specific research question.

Study limitations

The primary limitation of our study was the variation in time duration of the data collected. In spite of this limitation, this study serves a vital purpose as an initial attempt to define the challenges and possible predictors of outcome in these small subset of patients.

This can have significant connotations in the field of pediatric neurosurgery, to found a larger prospective study in future, attempting to establish a definitive causative role of the possible predictors described in this study, with an intention to promote better functional outcomes.

Conclusion

In this article, we have reviewed our experience with pediatric neuroanesthesia for children with giant supratentorial tumors presenting for definitive neurosurgery which should be tailored to optimize systemic and cerebral homeostasis. The maintenance of systemic physiological homeostasis by anticipation of complications, vigilant monitoring, and prompt resuscitation is crucial to support favorable outcomes in unison with optimal and safe surgical extirpation of the primary cerebral lesion.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.