Presenter: T. E. Merchant Presenter's Affiliation: St. Jude Children's Research Hospital, Memphis, TN Type of Session: Scientific
Central nervous system (CNS) tumors represent the most common solid tumor of childhood, and are the most common cause of cancer-related death in this population.
Low-grade gliomas are a subset of intracranial tumors that are generally treated with chemotherapy in combination with radiotherapy and surgical resection when possible. Conventionally, radiotherapy for low-grade gliomas has been delivered with clinical target volume (CTV) margins of 1.5 – 2 cm because these tumors have been observed to enlarge due to their frequently cystic nature during the course of radiotherapy [Jahraus, Ped Blood Cancer, 2006].
Radiation late effects are particularly important in the pediatric population, particularly those encountered by patients requiring central nervous system radiation.
Pediatric patients who receive radiation treatment to the brain are often left with severe functional deficits in cognition. Visual, auditory, and endocrine deficits are also frequently observed. These effects have been demonstrated to be related to both dose of radiation delivered and volume of tissue radiated.
Recent research in pediatric radiation oncology has focused on eliminating or reducing the need for radiotherapy in this vulnerable population while still maintaining optimal tumor sterilization.
Treatment of many pediatric malignancies, including low-grade gliomas, remains controversial. This is particularly true for young patients, as deficits tend to be most severe when patients receive radiation at very young ages.
This phase II study was performed to investigate the clinical disease outcomes and functional outcomes for children undergoing radiotherapy for treatment of low-grade glioma using a 10 mm CTV margin.
Materials and Methods
78 pediatric patients who were treated for low-grade glioma between 8/1997 and 8/2006 were enrolled on this study.
Median patient age was 8.9 years, with range +/- 4.3 years.
Tumor locations were diencephalic (n = 58), cerebral hemispheric (n = 3), and posterior fossa (n = 17). Tumors were World Health Organization (WHO) grade I in 68 patients, and grade II in 10 patients. All patients received chemotherapy, and 25 had failed previous treatment with chemotherapy with or without surgery.
Thirty-one patients had hydrocephalus at the time of diagnosis, 29 required ventriculoperitoneal shunting, and 13 had history of neurofibromatosis type I.
Treatment planning was accomplished with CT-MRI image registration, using T2-weighted MRI images fused with CT scans obtained for simulation. Target and normal tissues were defined using these registered images.
Gross tumor volume (GTV) was defined as the tumor for patients who did not undergo surgical resection, areas of residual disease + tumor bed for those who underwent subtotal resection, and tumor bed for those who underwent gross total resection. Clinical target volume was expanded 10 mm from GTV, and planning target volume was expanded 3-5 mm from CTV to account for set-up error and motion.
Patients were prescribed to receive 54 Gray (Gy) delivered to the CTV.
Patients underwent MRI at weeks 3 and 5 during the radiotherapy course. Following completion of radiation, they were followed with MRI every 3 months during the first 2 years after treatment, every 6 months for years 2 – 5, and annually thereafter.
Disease end-points evaluated included local control (LC), event-free survival (EFS), and overall survival (OS). Functional endpoints included the cumulative incidence of endocrinopathy, need for hormone-replacement therapy, multifrequency hearing loss, estimated IQ loss, and measures of deficits in memory, academic achievement, and adaptive behavior.
Median follow-up was 55 months. Five-year rates of LC, EFS, and OS were 91%, 87%, and 98%, respectively. Nine patients developed recurrent disease: Of these, one had a marginal failure near the optic chiasm, 5 experienced local failure, and 3 developed distant metastatic disease.
Five-year EFS was 89% for patients with WHO I tumors, and 64% for patients with WHO II tumors (p = 0.17).
Change in planned target volume was necessary in 15% of patients based on MRI obtained during the course of radiotherapy demonstrating cystic tumor enlargement. Persistent nausea, anorexia, and fatigue was experienced by 10% of patients due to cystic expansion.
Growth hormone supplementation was required 5 years after therapy in 40% of patients, and was positively correlated with mean radiotherapy dose to the pituitary region. Thyroid hormone supplementation was required 5 years after therapy in 52% (7% had required thyroid supplementation before radiotherapy). Sixteen percent of patients required adrenocorticotropic hormone supplementation 5 years after therapy as compared to 3% before therapy.
Precocious puberty was present in 28% of patients after therapy as compared to 12% before therapy.
IQ estimated before therapy and 5 years after therapy was not significantly altered. A decline in reading and spelling skills was observed between the period before radiotherapy and 5 years thereafter (p = 0.0038). This decline was positively associated with larger PTV size, increased mean radiation dose, and extent of surgery. The presence of hydrocephalus also increased the risk of significant decline in reading and spelling skills. Finally, reading deficits were correlated with tumor location.
Memory appeared to be preserved during the period before and after treatment. Both behavioral scores and visual/ auditory learning scores improved between the period before treatment and 5 years after.
Adaptive behavior appeared to be preserved during this period, although mild decline in communication skills was observed between the period before treatment and 5 years after.
Hearing loss was very uncommon when cochlear dose was maintained below 35 Gy.
Symptomatic vasculopathy developed in 10% of patients, and was present in 5% before treatment.
The authors conclude that excellent disease control can be achieved with CTV volume limited to 10 mm.
They note that the side effect profile from this treatment is limited and predictable, and that these observations are very important contributions to our current understanding of radiation-induced side effects.
The authors recognize that this information will be important in future investigations, one goal of which will be to continue to reduce the volume of irradiated tissue.
This study is a large, systematic, prospective trial, the like of which is not always available in the pediatric body of literature due to the rare nature of many childhood malignancies.
The authors demonstrate excellent overall survival and local control rates with reduced CTV volume, with only one patient having experienced a marginal failure. They demonstrate that use of MRI during treatment may allow volume reduction as long as treatment volumes can be efficiently adjusted should cystic enlargement be observed.
The ability to reduce the volume of tissue irradiated is particularly important in the setting of pediatric CNS malignancies. Previous literature has demonstrated that no “safe” or threshold radiation dose to the pediatric brain exists, and that cognitive and other deficits are related to integral tissue dose [Merchant T, IJROBP, 2006]. As such, reducing the volume of tissue exposed to radiation is expected to reduce normal tissue damage.
This study demonstrates relatively minimal cognitive deficits as a result of radiation treatment, with predictable deficits in reading and spelling observed related to tumor and disease characteristics. The lack of significant change in IQ and memory, as well as improvements in behavior and visual and auditory learning in the interim after treatment, are reassuring with regards to the ability of these patients to function well as adults.
As technologies continue to improve and become available, this data will serve as an excellent starting point for future research. Perhaps most notably, as proton beam therapy becomes available at increasing numbers of centers, comparison of disease and functional outcomes between x-ray and proton therapy will be of paramount interest. This data provides a comprehensive assessment of our current ability to treat pediatric patients with low-grade glioma, and will likely serve as an important source of data as further treatment options emerge, and ability to reduce dose and volume continues to progress.
Partially funded by an unrestricted educational grant from Bristol-Myers Squibb.
Feb 24, 2012 - Long-term survivors of childhood high-grade glioma have intellectual functioning within low-average ranges and low neuropsychological functioning, but the majority of patients report within or above normal quality of life, according to a study published online Feb. 21 in the Journal of Clinical Oncology.