Introduction
While I was a rotating surgical intern at St. Luke’s International Hospital in Tokyo in 1970-1972, two pediatric patients struck me the most. One of them a 10-year-old boy with cystic craniopharyngioma with recurring cyst needing frequent Ommaya taps, the only treatment that could be offered to him then. Another was a 7-year-old girl with pontine glioma who progressed rapidly and succumbed to this terrible disease. These young patients started me on my path in pediatric neurosurgery, hoping someday I could discover a “cure”. That was more than 50 years ago.
I received my neurosurgery training with two years under Satoshi Matsumoto at Kobe University, Kobe, Japan and six years under Anthony J Raimondi at Northwestern University in Chicago, Illinois. I started my full-time practice at Children’s Memorial Hospital (CMH) in Chicago on July 1, 1981. After a presence of 130 years at its Lincoln Park location, CMH was moved to the Streeterville neighborhood of downtown Chicago in 2012 and it was renamed Ann & Robert H. Lurie Children’s Hospital of Chicago (Lurie). Following David G. McLone’s retirement, I was appointed as a Division Head in July 2001, and served until the end of 2019, and retired from clinical practice at the end of 2022. I proudly say, I was the last pediatric neurosurgery chief at CMH and the first one at Lurie.
The pediatric neurosurgery service of CMH, which was started by Luis V. Amador in 1952, is one of the oldest programs in the nation and has produced many national and international leaders in the field, including 12 (24%) of 50 past presidents of ISPN trained in various capacities at CMH/Lurie. The service has focused on Clinical Care, Education, Research and Scholarly Inquiry, the four pillars of academic medicine over the years. Many pediatric trainees have become academic pediatric neurosurgeons and taken leadership positions in their own departments.
This manuscript is written at the request from AANS Neurosurgeon to write a personal perspective on the evolutions in pediatric neurosurgery, reflecting on my own experience. As the field of pediatric neurosurgery has grown exponentially in the past few decades, I cover only on hydrocephalus and brain tumor in this communication to share my reflection.
Hydrocephalus
Since 1970, ventriculoperitoneal shunt became the dominant method for the treatment of hydrocephalus. In order to avoid shunt-related complications such as over-or under-drainage, post-shunt subdural hematoma, shunt malfunction, slit ventricles etc., multiple valve systems have been invented. They have included ball valves, slit valve, gravity assisted valve, flow-regulated valve and then came antisiphon devices and adjustable valves. I have used all of them. Although antibiotics impregnated catheter became available, preoperative use of systemic antibiotics and a double glove technique during the surgery have been the key to the reduction of shunt infection rates. There is no question that hydrocephalus shunt revision rates have steadily diminished in the last decade, however, a 100% reliable shunt is yet to be found.
Neuroendoscopy has revolutionized the management of pediatric hydrocephalus. In the early 1980s, prior to the availability of dedicated ventriculoscopy, I used a pediatric cystoscope and was fascinated by the view at navigating through the hydrocephalic ventricle of children. Since 1993, several manufacturers have designed advanced ventriculoscopes with working channels, enabling us to biopsy or resect a tumor and to fenestrate a cyst. A flexible scope became available recently, expanding the endoscopic views.
ETV revolutionized management pediatric obstructive hydrocephalus since 1990. Years of experience taught us the indications and success factors for ETV. Poor responses were noted among young infants and posthemorrhagic/postinfectious hydrocephalus and spina bifida. Endoscopic choroid plexus coagulation +/- ETV has been exercised for neonatal hydrocephalic cases but remains controversial. Controlling hemorrhage during the ventriculoscopic procedure remains difficult and often irrigation, flushing and waiting with a little prayer are the only way to stop it.
Brain Tumors: Diagnosis and Treatment
Prior to CT era in the early 1970s, Diagnostic neuroimaging available were cerebral angiography, pneumoencephalogram and radioisotope nuclear brain scan. Diagnosing brain tumor, its location and extension, needed training and profound knowledge. It became less stressful to both clinicians and children, after CT/MR. Pathological diagnosis was traditionally based on histological features. Additional molecular classifications have been adopted for many CNS tumors by the WHO since 2016.
Posterior fossa craniotomy instead of craniectomy and laminoplastic laminotomy instead of laminectomy, for posterior fossa tumors and spinal cord tumor respectively, became a standard method for pediatric neurosurgery since the mid-1970s. Surgical laser and ultrasonic aspirator augmented the surgical armaments since the 1980s. Surgical microscopy was at its infancy when I started my training in 1970. Since then, neurosurgical microscopes have advanced to the current level, further enhanced with neuronavigational attachment, blood flow detection, 3-D imaging, Fluorescence-Guided Surgery, robotic guidance, etc.. Frameless neuronavigational system was introduced and has been improved since the mid-1980s. My techniques of tumor resection, however, have not changed much over the decades and micro-dissector, suction and bipolar cautery under microscope have been the main tool.
We are aware of latent radiation side effects to the growing CNS such as cognitive delay, endocrinopathy, cranial nerve neuropathy, vasculopathy and second neoplasms. Particularly children with NF 1 face higher risks of malignant transformation of gliomas and increased risks of vasculopathy following RT. We have made every effort to lower the dose and field of radiation, or ideally eliminate irradiating babies and young children altogether. Instead, chemotherapy has gained a significant role to accomplish these purposes over the past three decades. Some tumors have been treated with target therapy to specific molecules.
Medulloblastomas/Ependymoma
CNS embryonal tumor is one of the most representative and studied childhood brain tumors. In the 2007 WHO classification of CNS tumors, medulloblastoma was included among embryonal tumors together with atypical teratoid rhabdoid tumor (ATRT) and CNS PNET. Staging medulloblastoma for prognosis, average-risk vs. high-risk, has been established based upon the patient age, tumor dissemination, extent of tumor resection and tumor histological and molecular classification.
Children with medulloblastoma have been treated with craniospinal irradiation (CSI) and additional dose of RT to the posterior fossa followed by multiagent chemotherapy. In the 1970s, it was a common finding to see children with short stature, small head and developmental delay among survivors after medulloblastoma therapy because of high dose of RT to the CSI. The CSI dose was lowered to 25 Gy in the early 1980s after finding out high dose “prophylactic” RT was the cause of developmental delay. Adjuvant chemotherapy to surgical resection and RT improved medulloblastoma survival from only 20% with RT alone in the 1970s to 80% after the use of adjuvant chemotherapy.
Medulloblastoma invades the brain stem in more than 30% and invades the cerebellar peduncle additional 40% of the cases in my experience. Posterior fossa syndrome or cerebellar mutism is unfortunately common after radical resection of fourth ventricle medulloblastoma occurring in 20-25% of cases, although its definitive pathomechanism has yet to be defined for these often-devastating surgical complications. We continue maximum cytoreduction in a “safe” margin.
ATRT coined by Lucy Rorke in 1996 is a highly aggressive, often chemo resistant tumor affecting infants and children younger than 3-years-old. Mutations of the INI1 (SMARCB1) gene were found among ATRT in 1999, yet its pathogenesis remains poorly understood. After searching for potential molecular targets, we found an inhibitor of Polo-like Kinase 4 to be a hopeful candidate.
Posterior fossa ependymoma remains a difficult tumor to control. We learned over time that the extent of resection affects the prognosis, thus it has been considered to be a “surgical disease”. However, its propensity of involving the brain stem/cranial nerves, makes total resection often hazardous. Unfortunately, chemotherapy has not been effective, which we speculated due to their expression of multidrug resistance gene. Based on the molecular information, targeted treatments for ependymoma have been in search.
Craniopharyngioma
There are two distinct types among craniopharyngiomas: adamantinomatous occurring childhood and papillary type in adults. Surgical resection of craniopharyngioma is often challenging even to the experienced pediatric neurosurgeon. I have personally experienced visual disturbances, hypothalamic obesity, pan-hypopituitarism, hemiparesis, pseudoaneurysm, etc. following aggressive craniopharyngioma resection in my career. Adamantinomatous type are invariably cystic and calcified. Obstructive hydrocephalus is common among retro-chiasmal third ventricular craniopharyngiomas but is resolved by endoscopic cyst drainage. Pre-chiasmal craniopharyngiomas often present with visual deterioration and require urgent decompression of the optic pathway. Multiple surgical approaches have been practiced; subfrontal/pterional, transcallosal-ventricular or transsphenoidal. There has been a trend toward the endonasal endoscopic transsphenoidal approach because of better visualization of the infra and retro-chiasmal structures. However, a smaller nasal structure and poorly pneumatized sinuses may pose a technical challenge in young children. Controversy as to a limited resection followed by focal RT vs. a radical tumor resection remains unsettled among pediatric neurosurgeons. The results of recurrence free survival after either therapy is comparable, about 20% based on my own series. Functional outcomes of children with limited resection with RT did better based on my experience. Intracystic treatments of bleomycin, alpha-interferon or radioisotope have been scarcely practiced over the years, but their results remain inconclusive.
Recent molecular investigations of craniopharyngioma have confirmed a high incidence of CTNNB1 mutation in adamantinomatous type, whereas high incidence of BRAF V600E mutation was noted among papillary type. The latter has been treated with molecular target therapy, BRAF and MEK inhibitors, resulting in a therapeutic response. On the other hand, WNT/β-catenin therapy is potentially an attractive target for molecular therapy for adamantinomatous craniopharyngiomas.
Diffuse Intrinsic Pontine Glioma
I have treated over the years numerous children with DIPG with multiple treatment protocols including RT with chemotherapy, RT with radiosensitizer, hyper-fractionated irradiation, intra-arterial chemotherapy and convection enhanced interstitial chemotherapy; but unfortunately all failed. Although, previously, biopsy was not recommended for a typical DIPG [9], biopsy for tissue sampling to identify molecular expression has been done to search for effective therapy for DIPG since a decade ago. The 2016 WHO classification of CNS tumors defined diffuse midline glioma (H3K27M-mutant) as an infiltrative midline high-grade glioma with predominantly astrocytic differentiation. A new drug specific for gliomas with H3K27M–mutation is currently under investigation.
Conclusion
In addition to above-stated hydrocephalus and brain tumor managements, I have witnessed rather dynamic progress in the practice of pediatric neurosurgery. They include the reduction of incidence of neural tube defect after the use of folic acid for women of childbearing age since 1992; prenatal repair of myelomeningocele after MOMS study in 2011; minimally invasive endoscopic repair of craniosynostosis; disappearance of halo-vest using spine instrumentation; endovascular intervention for vascular malformation and tumors; selective dorsal rhizotomy for spasticity since 1986 and baclofen pump for spasticity/dystonia since 1993; functional hemispherotomy from anatomical hemispherectomy for intractable seizures since the 1990s; neuromodulation for seizures with VNS in 1997 and RNS since 2011; the 80-hour workweek restriction for medical residents implemented by the ACGME in 2003 and the introduction of the RVU system by the Centers for Medicare and Medicaid Services in 1992.
Unfortunately, management of children with failed craniopharyngioma or DIPG did not make progress even after a half century later, hopefully better days are ahead. Due to medical sophistication and sub-specialization within pediatric neurosurgery, multidisciplinary team approaches are an essential integral part for successful patient care. Subspecialty programs include brain tumor, spina bifida, craniofacial, pediatric spine, pediatric epilepsy and others. It is indispensable that pediatric neurosurgeons take a major role in respective multidisciplinary programs. With these commitments, pediatric neurosurgeons contribute to further advance in patient care of many neurological disorders of childhood.
