Literature Review Paper

s t t t a c U a

i i o y f o a c m n d

o c t o o n

p a t m

F C 1 © d

2

Brain Tumors in Children: Evaluation and Management

Adam S. Levy, MD

e a w

C T

b t t

I i c t p t r t a p e n i c i j r m a t w a L s s

v

A ccording to the National Cancer Institute’s Surveillance, Epidemiology, and End Results (SEER) data, the incidence of central nervous

ystem (CNS) tumors is 27 per 1 million children less han 19 years old. After leukemias, CNS tumors are he second most common neoplasm in children. Brain umors are the most common solid tumors of children nd account for the greatest mortality from cancer in hildren.1,2 Approximately 11,000 children in the nited States will be diagnosed with cancer each year,

nd of these just over 2000 will be CNS tumors. Over the last 30 years, there appears to be a slow

ncrease in the overall incidence of childhood cancers n the US, including CNS tumors in children. The verall 5-year survival for patients younger than 19 ears with CNS tumors is around 65%.3 The outcome or children with CNS tumors has shown improvement ver the last several decades, but remains guarded for large subset of children with certain brain tumors. In

omparison to the outcome for children with leuke- ias, the outcome for children with brain tumors has

ot had as dramatic improvement over the past three ecades. CNS tumors represent a diverse group of neoplasms f the brain and spine with varying histology. The linical presentation is often dictated by the location of he tumor in terms of the neurological deficit or bstructive hydrocephalus. The onset may be indolent r rapid depending on the histology and aggressive- ess of the tumor. The goal of this article is to discuss the common resentations of pediatric CNS tumors, to detail the ppropriate initial evaluation and management, to review he more common CNS tumors, and to discuss treatment odalities. Long-term follow-up for survivors and the

rom the The Children’s Hospital at Montefiore, Bronx, New York. urr Probl Pediatr Adolesc Health Care 2005;35:230 –245 538-5442/$ - see front matter 2005 Elsevier Inc. All rights reserved.

ioi:10.1016/j.cppeds.2005.04.001

30

ffects on the family of a child with a CNS tumor will be ddressed. Finally, palliative care strategies for children ith CNS tumors will be discussed.

ommon Presentations of CNS umors The presentation of a child with a CNS tumor may e subtle or dramatic and is dependent on a combina- ion of the patient’s age, the tumor location, and the umor histology. In infancy, brain tumors may present insidiously.

ndeed, the only sign of a CNS lesion may be an ncreased head circumference secondary to hydro- ephalus. When the head circumference is dispropor- ionately large compared with the length or weight, the ediatrician must consider an intracranial process as he cause. While it may be more common that the elatively increased head circumference is familial, his conclusion must be reached with great caution, nd a trend of increasing head circumference out of roportion to the length and weight must be further valuated. Unusual fullness or bulging of the fonta- elle may also be a sign of hydrocephalus or increas- ng intracranial pressure. Also, from increased intra- ranial pressure, infants or toddlers may present with rritability or Parinaud’s syndrome (paralysis of con- ugate upward movement of the eyes and poorly eactive pupils). Developmental delay may be the ain symptom at diagnosis, but more typical would be loss of previously attained milestones. Thus, a young

oddler who is not yet walking may not be particularly orrisome, but a toddler who had been walking well

nd then becomes ataxic warrants further evaluation. ikewise, a child who has a primary speech delay hould be approached differently than a child whose peech regresses. Toddlers and school-age children may present in a ariety of ways. New onset seizure (unrelated to fever)

s more likely to be secondary to a seizure disorder,

Curr Probl Pediatr Adolesc Health Care, July 2005

b o s m ( t

i c c a a o p fl m f b n a p

p m e o

w c o c p c p o

m s t s I H e

I S

d

a p c h h t

s a r w p C C a a a

w q t a a s s d t p i a

f u r s c t m fi h t M t t

s m i a

C

ut an intracranial mass must be considered. The level f evaluation and timing of studies for patients pre- enting with new onset seizure is debated, though any would agree that a magnetic resonance imaging

MRI) of the brain is warranted for new onset seizure hat is not associated with fever. Many of the symptoms of a brain tumor result from

ncreased intracranial pressure (ICP) from either hydro- ephalus or the mass effect of the tumor itself. The lassic presentation of increased ICP is morning head- che with vomiting. Frequently, there is no nausea ssociated with the emesis. On lying down overnight, bstructive hydrocephalus may be worsened as the sleep osition does not allow gravity to promote cerebrospinal uid (CSF) flow. The increased ICP causes headache and ay trigger emesis, and the headache is often relieved

ollowing the emesis. It is not unusual for patients with a rain tumor to present with emesis that had been diag- osed as acute gastroenteritis. Isolated emesis without ccompanying diarrhea or dyspepsia should alert the ediatrician to the possibility of increased ICP. In older children and adolescents, brain tumors may resent secondary to endocrine dysfunction. Some tu- ors, such as prolactinomas, are the primary cause of

ndocrine dysfunction. Other times, tumor compression n the pituitary may cause pituitary dysfunction. Depending on the tumor location, children may present ith a focal neurological deficit. Diplopia secondary to

ranial nerve deficits is not an uncommon presenting sign f a brain tumor. For tumors located near the motor ortex, specific losses of physical function may be the resenting symptom. Tumors affecting the brainstem ommonly present with ataxia. Spinal cord tumors may resent with incontinence to urine or stool as well as ther focal deficits. With ICP increasing to a critical level, a patient’s ental status may be affected and the patient’s vital

igns may be altered. Specifically, the triad of hyper- ension, bradycardia, and tachypnea (Cushing’s Triad) hould alert the clinician to the possibility of increased CP. This is usually a late sign of increased ICP. ypertension compensates for the raised ICP in an

ffort to normalize intracranial blood flow.

nitial Evaluation and Management of uspected CNS Tumor

The initial evaluation of a suspected CNS tumor

epends on the presenting signs, symptoms, and avail- g

urr Probl Pediatr Adolesc Health Care, July 2005

ble resources at the health center. For example, a atient presenting with signs and symptoms of in- reased ICP requires specific management including igh-dose corticosteroids, relative fluid restriction, ead elevation, and possibly neurosurgical interven- ion. Urgent brain imaging is also indicated. Obtaining an urgent computerized tomography (CT)

can may be more realistic than obtaining an MRI scan s MRIs are less commonly available and are likely to equire prolonged sedation for pediatric patients. Thus, hile an MRI may be the ideal scan to evaluate a atient for a brain tumor, an initial evaluation with a T scan is more common and still likely to be useful. T scans allow accurate evaluation of ventricle size nd detection of hemorrhage or stroke. CT scans will lso frequently detect brain tumors, especially those ssociated with calcification. In general, the diagnostic test of choice is an MRI ith and without contrast (Gadolinium). FLAIR se- uences are often useful in determining tumor infiltra- ion or surrounding edema.4 Diffusion-weighted im- ging can help differentiate cysts from solid tumors nd identify areas of necrosis.5,6 However, while tandard MRI provides excellent images of anatomical tructure of the brain and spinal cord, standard MRI oes not provide information regarding tissue func- ion, metabolic state, blood supply, and malignant otential.7 Such information would be clinically useful n determining surgical approach, response to therapy, nd monitoring for relapse. Newer imaging techniques, broadly referred to as

unctional imaging, are under development and being sed more frequently in clinical practice. Magnetic esonance spectroscopy (MRS) allows for the mea- urement of biologically important molecules such as holine, creatine, N-acetyl aspartate, glutamate, lac- ate, and lipid. The profile of the concentration of these olecules may differ among tumor, normal brain, or brosis and scarring.8 Some studies suggest that tumor istology can be determined by the signal pattern of he molecules.9 Enhanced tumor localization with

RS may also help determine the ideal location for issue biopsy10 and may predict outcome and response o therapy.11

Positron emission tomography (PET) scan is a sen- itive and specific imaging tool to evaluate the tissue etabolism. 18Fluorodeoxyglucose, a glucose analog,

s taken up by tissues with high-glucose metabolism, nd this uptake can be measured by PET. Again,

lucose metabolism may be markedly distinct in tumor

231

c f c

t d d t d i m t

N T

o l o m r s l b c

C t o t l b r

t t p s m p s n w s d o

c

d r p p

e s c t t v m p

o i s F m o b r a fi

a F T i s w s r p

i b t n t n c o

R P

2

ompared with normal brain or fibrosis and may in the uture become a more standard tool to help direct linical decisions.7

In some circumstances, the radiographic presenta- ion may be diagnostic obviating the need for tissue iagnosis. The most common examples of this are iffuse intrinsic brain stem tumors and optic pathway umors that are usually low-grade gliomas (to be iscussed in detail later). In most other settings, tissue s needed to confirm the tumor histology. As such, the ajority of pediatric patients presenting with a CNS

umor will require a neurosurgical procedure.

eurosurgical Approach to CNS umors

The neurosurgical approach to a CNS tumor depends n the clinical status of the patient, the location of the esion, and the suspected histology of the lesion based n the imaging studies. Urgent surgical intervention ay be indicated to manage hydrocephalus, hemor-

hage, or the acute mass effect of a tumor. In these ettings, the primary goal is to manage the acute ife-threatening problem. In such circumstances it may e necessary to perform a second neurosurgical pro- edure to deal more definitively with the tumor. Fortunately, most of the time the presentation of a NS tumor is not this acute, thereby allowing for

houghtful planning. At a minimum, the primary goal f nonurgent neurosurgical intervention is to obtain issue for a histological diagnosis. Depending on the ocation of the mass, this may be accomplished by a iopsy, partial resection, or complete resection (also eferred to as a gross total resection). In general, the outcome for children with brain

umors is associated with the extent of surgical resec- ion.12,13 Thus, there is a tendency for experienced ediatric neurosurgeons to favor a relatively aggres- ive approach. However, the desire to accomplish a aximal resection or tumor debulking must be tem-

ered by the possible neurological insult from aggres- ive surgery. For this reason, it is preferable for eurosurgical interventions to be made by surgeons ith experience with pediatric CNS tumors in the

etting of a tertiary care hospital where the case can be iscussed with a pediatric oncologist and a radiation ncologist with pediatric CNS tumor experience. In a situation where the tumor location precludes

omplete resection, a biopsy alone will likely help b

32

etermine adjuvant therapy (either chemotherapy or adiation therapy). Following adjuvant therapy, a com- lete resection may then be more safely accom- lished.14

Recent advances in neurosurgical techniques have nhanced the ability to biopsy and resect tumors more afely. State-of-the-art neurosurgical techniques in- lude intraoperative navigation systems (neuronaviga- ion), intraoperative MRI, and frameless strereotaxy hat allow intraoperative guidance with the ability to isualize the tumor in multiple planes.15 Cortical apping is frequently needed for tumors in close

roximity to the motor and speech cortex.16

Unfortunately, complications from either the tumor r the surgery are not uncommon. Postoperative swell- ng or edema commonly results in worsening of ymptoms in the immediate postoperative period.17

ocal neurological deficits following surgery are com- on but generally have some or complete resolution

ver time. Hydrocephalus may occur before surgery ut may also develop even after a complete tumor esection either from edema, a hematoma, or an lteration of CSF dynamics that is not clearly de- ned.15

Following the resection of a posterior fossa tumor, pproximately 10% of patients may develop “Posterior ossa Syndrome,” also known as cerebellar mutism. his syndrome is characterized by postoperative mut-

sm, ataxia, 6th and 7th nerve palsies, and hemipare- is,18 and this syndrome is more commonly associated ith malignant tumors as compared with benign le-

ions. The onset is typically within days of surgery and ecovery may take weeks to months with up to 20% of atients having permanent features.15

Other postoperative complications include wound nfection, pseudomeningocele, aseptic meningitis, acterial meningitis, stroke, and seizures. Given the echnical and clinical considerations involved in the eurosurgical management of pediatric CNS tumors, he importance of referral to an experienced pediatric eurosurgeon with state-of-the-art operative tools at a enter with experience treating such patients cannot be veremphasized.

adiation Therapy Approach to ediatric CNS Tumors

For those pediatric brain tumor patients that cannot

e cured by surgery alone, radiation therapy is fre-

Curr Probl Pediatr Adolesc Health Care, July 2005

q T o t c

d n t c d t d f t g e r

o i r i r c r

b r d d h i y r a m c c e t p m

s m r d v v

m n s i t

d p n i b n w t r d r a s d w

r p i l t e s

c i t t

l o r i t t

i e r c o L i

C

uently the definitive adjunctive treatment of choice. he majority of brain tumor patients with high-grade r malignant tumors will receive radiation to either the umor, the resection bed, the whole brain, or the entire raniospinal axis. Radiation therapy takes advantage of the inherent ifferences in radiosensitivity between tumor cells and ormal tissues. In general, normal tissue is better able o tolerate the insult of ionizing radiation than tumor ells. Ionizing radiation causes DNA damage either irectly or indirectly by the formation of free radicals hat chemically interact with DNA.19,20 Since the amaging effects are thought to be the result of ree-radical formation, many clinicians transfuse pa- ients as needed to maintain relatively normal hemo- lobin concentrations in the blood in an effort to nsure adequate oxygen delivery to the tumor during adiation therapy. Just as some tumors are more radiosensitive than thers, normal tissues vary in degree of radiosensitiv- ty. As compared with hematopoietic stem cells, neu- ons are significantly less radiosensitive. This is clin- cally evident in patients receiving craniospinal adiation for whom myelosuppression is more likely to ause an interruption in radiation therapy than acute adiation effects on the CNS. An acceptable therapeutic ratio can also be achieved y targeted delivery of radiation.20 A main focus of adiation therapy advances has been to more precisely eliver radiation to the tumor target. To repeatedly eliver radiation to a limited target volume, methods ave been developed to place a patient in a reproduc- ble position over the days to weeks of therapy.19 For ounger pediatric patients, the positioning needed for adiation administration can be so challenging that nesthesia is frequently necessary. Essentially an im- obilization device such as a plastic facemask is

ustomized for the patient. The mask is placed on the hild and then attached to the treatment table in an ffort to limit movement as much as possible. With echnology that is designed to deliver radiation with recision on the order of millimeters, patient move- ent cannot be allowed. Radiation oncologists utilize MRIs, CT scans, and

ophisticated computer systems in conjunction with edical physicists to define the target volume of the

adiation dose and determine the treatment plan to eliver a given amount of radiation to the target olume. The target volume includes the gross tumor

olume and surrounding tissue at risk for spread of b

urr Probl Pediatr Adolesc Health Care, July 2005

icroscopic disease.19 For tumors with high malig- ant potential to spread throughout the CNS, cranio- pinal radiation with a boost to the tumor may be ndicated. The radiation dose is based on the type of umor and its location in the body. There have been recent advances in the ability to eliver radiation therapy in the full doses needed to rovide tumor response while limiting the exposure of ormal surrounding tissue.21 If one considers deliver- ng a single high-energy beam to a tumor within the rain surrounded by normal tissue, it is clear the ormal tissue between the energy source and the tumor ill receive the highest dose of energy. The normal

issue on the other side of the tumor will receive a high adiation dose as the beam exits the body. Three- imensional conformal radiation therapy allows the adiation therapy team to select beam shapes and ngles that divide the total dose into multiple inter- ecting energy beams. Thus, the treatment high-energy ose can more precisely conform to the target volume hile limiting the dose to normal tissue.21

Intensity-Modulated Radiation Therapy (IMRT) is a elatively new approach that allows even greater recision of radiation delivery by modulating the ntensity of each beam with individual radio-opaque eaves. The individual leaves are moved in and out of he radiation field during treatment such that the nergy beam can be sculpted to conform to irregular- haped tumors within the body.19,21

Radiosurgery utilizes multiple radiation beams all onverging on the same tumor target thereby deliver- ng a single high-dose fraction of radiation to the umor. Radiosurgery is usually limited to use for small reatment volumes.19,21

Proton Beam Therapy is particularly helpful in imiting exposure of normal tissue. Unlike the photons r more standard RT, protons enter the body with a elatively low radiation dose and that energy is depos- ted in the tumor target with very little energy passing hrough the target into surrounding tissue. Normal issues distal to the target are not irradiated.21,22

Toxicity of radiation therapy is of particular concern n pediatrics and can be characterized as acute and late ffects. Common acute effects include local skin eactions, anorexia, nausea, somnolence, focal alope- ia, mucositis (depending on radiation beam path), titis, and myelosuppression (with spinal radiation). ong-term effects are of even greater concern and

nclude radionecrosis of normal tissue, vascular injury,

one growth stunting, endocrine abnormalities, hear-

233

i a

C P

g p p l r m w r d l l

h s a b l W b i p C s t i

a m p m d e O w c c a s a m f p t

p c a c c u t e “ b

i s b O i a b u B b t s

m m C p r e t v a t

t p m p p s e c r

f i m

2

ng loss, secondary malignancies, cognitive deficits, nd neuropsychological effects.17,22,23

hemotherapeutic Approach to ediatric CNS Tumors Over the past several decades, chemotherapy has ained increasing acceptance as an important thera- eutic approach for a variety of brain tumors. For atients with low-grade tumors, complete resection is ikely to be curative. However, when a complete esection cannot be accomplished for low-grade tu- ors, additional therapy must be considered. Like- ise, for high-grade lesions regardless of the extent of

esection adjuvant therapy must be considered as issemination and recurrence after surgery alone are ikely. In general, the younger the patient, the more ikely the late effects of RT will be devastating. The mainstay of brain tumor management in adults as been surgery and radiation with chemotherapy in ome cases. However, because of the effects of radi- tion therapy (RT) on the developing brain, there has een a movement to try and avoid RT in children or at east delay RT administration until the patient is older.

ith advances in supportive care, chemotherapy can e delivered more safely and with greater predictabil- ty of side effects thereby justifying attempts to incor- orate chemotherapy into a broader range of pediatric NS tumor treatment plans. The goal is to improve

urvival and decrease the morbidity of RT. Of course here is the added morbidity of the chemotherapy tself. Though chemotherapy regimens may appear haphaz-

rd, they are generally designed to provide comple- entary insults to the tumor while avoiding overlap-

ing dose-limiting toxicities. For example, the ideal ulti-agent chemotherapy regimen would include

rugs that have different mechanisms of action in an ffort to avoid intrinsic or acquired tumor resistance. n the other hand, the toxicity profile must in some ays be dissimilar so that the chemotherapy agents

an be given contemporaneously. In other pediatric ancers such as acute lymphoblastic leukemia, multi- gent chemotherapy regimens have successfully re- ulted in impressive improvement in survival with cceptable morbidity. For a subset of brain tumors ulti-agent chemotherapy has proven successful, but

or other brain tumors results have been equivocal or roven to be unsuccessful (to be discussed below for

he common pediatric brain tumors). t

34

The range of intensity of chemotherapy regimens for ediatric brain tumors is great. Some treatment plans all for single-agent oral chemotherapy that is gener- lly very well tolerated24; in other settings high-dose hemotherapy with stem cell rescue (autologous stem ell transplant) is indicated.25 Despite the oncologist’s sual comfort in chemotherapy administration, the herapeutic index for chemotherapeutic agents is gen- rally small. Thus, while families are often willing to try anything” in desperate situations, physicians must e reminded to only provide rational therapy. A variety of chemotherapy drugs are commonly used

n the treatment of pediatric brain tumors. In general, ingle-agent Phase I trials are utilized to show feasi- ility and perhaps some benefit in a specific tumor. ver time agents with known benefit have been tried

n combination. The most common chemotherapy gents for pediatric brain tumors will be detailed elow. For a detailed review of chemotherapy agents sed in pediatrics, see Balis and coworkers26 and anerjee and Mathay27 from which the discussion elow is derived (see also Table 1). Combination reatment plans will be reviewed later as they relate to pecific tumor types. Vincristine is a commonly used drug in the manage- ent of low-grade gliomas as well as high-grade and alignant brain tumors despite its limited documented NS penetration.28 Usually given as an intravenous ush in combination with other agents, vincristine is elatively well tolerated with the most common side ffects being constipation, jaw pain with administra- ion, and peripheral neuropathy which is usually re- ersible and related to cumulative doses. Vincristine is plant alkaloid that induces metaphase arrest through

he binding of tubulin.26

Etoposide also binds tubulin and interacts with opoisomerase II resulting in cell cycle arrest.26 Eto- oside is used intravenously in combination with ultiple agents to aggressively treat brain tumor

atients, but can also be prescribed orally in a more alliative setting with potential effect and acceptable ide effects. Common toxicities include nausea, my- losuppresion, and alopecia. Frequently the greatest oncern with etoposide administration is the increased isk of secondary malignancies. Cisplatin and carboplatin are alkylating agents used

or a variety of CNS tumors given as intravenous nfusions. Cisplatin generally requires close manage- ent of fluid and electrolyte balance and is more likely

o cause significant renal toxicity and hearing loss.

Curr Probl Pediatr Adolesc Health Care, July 2005

C t p n n

i g m c c a

( i C n

c i u w c h n

a h

w t

c c o c a s t v a m c s a

a w m p e m l t a T

T

V

E

C

C

N

M

T

C

isplatin is also one of the most emetogenic chemo- herapeutic agents used. The toxicity profile of carbo- latin is similar but usually less severe with respect to ausea, ototoxicity, and renal toxicity.27 Both plati- um compounds are myelosuppressive. Cyclophosphamide is an alkylating agent that inhib-

ts DNA synthesis. Cyclophosphamide is usually iven as an intravenous infusion for high-grade or alignant brain tumors.27 Common side effects in-

lude myelosuppression, nausea, and alopecia. Other oncerning side effects include hemorrhagic cystitis nd impaired fertility. The nitrosureas Lomustine (CCNU) and Carmustine

BCNU) are alkylating agents that are given orally and ntravenously, respectfully.27 These agents have good NS penetration. The most common toxicities are ausea and delayed myelosuppression. Methotrexate is an antifolate drug that causes de-

reased DNA synthesis. It can be given by mouth, ntravenously, or intrathecally. Methotrexate has been sed recently to intensify therapy for pediatric patients ith high-grade or malignant brain tumors.29 Con-

erning side effects are renal toxicity when given in igh doses. Myelosuppression is also possible as are eurological sequelae, stomatitis, and photosensitivity. Temozolomide is a relatively new oral alkylating

gent that has been used for low-grade as well as

ABLE 1. Commonly used chemotherapeutic agents for the treatment of

Chemotherapeutic agent Mechanism of action

incristine Induces metaphase arrest through the binding of tubulin

toposide Binds tubulin and interacts with topoisomerase II resulting in cell-cycle arrest

isplatin, Carboplatin Alkylating agents

yclophosphamide Alkylating agent that inhibits DNA synthesis

itrosureas: Alkylating agents Lomustine (CCNU, given orally), Carmustine (BCNU, given IV) ethotrexate Anti-metabolite; interferes with

folate metabolism

emozolomide Alkylating agent

igh-grade pediatric brain tumors.27 It is generally b

urr Probl Pediatr Adolesc Health Care, July 2005

ell tolerated but is known to cause nausea, constipa- ion, and myelosuppression. Tumor cells may have intrinsic resistance to a given

hemotherapy. That is, the tumor resists chemotherapy ell kill based on some inherent tumor biology. On the ther hand, a tumor may be relatively sensitive to a hemotherapeutic agent initially but resistance may be cquired as tumor cells that are relatively less sensitive urvive. The concept of tumor resistance in brain umors is similar to other pediatric neoplasms. Adju- ant therapies are under development to obviate mech- nisms of resistance. In other pediatric tumors, che- otherapy dose intensification and multi-agent

hemotherapy regimens have often improved overall urvival, though the intensification of chemotherapy is t the expense of added iatrogenic toxicity. The use of chemotherapy in CNS tumors has the

dded obstacle of the blood brain barrier (BBB), hich under normal circumstances limits many large olecules from leaving the peripheral bloodstream to

enetrate the CNS. The BBB is conceptualized as tight ndothelial cell junctions that allow small lipophilic olecules to permeate into the CNS while excluding

arge hydrophilic molecules.27 Thus, there is concern hat systemically administered chemotherapy may not chieve optimal concentrations within the tumor. hus, CNS penetration of chemotherapy agents must

tumors

Tumors treated Side effects

ow-grade gliomas as well as high-grade and malignant brain tumors

Constipation, jaw pain with administration, and peripheral neuropathy, which is usually reversible and related to cumulative doses

alignant brain tumors Nausea, myelosuppression, and alopecia; increased risk of secondary malignancies

ariety of CNS tumors Renal toxicity and hearing loss; highly emetogenic

igh-grade or malignant brain tumors

Myelosuppression, nausea, alopecia, hemorrhagic cystitis, and impaired fertility

ariety of CNS tumors Nausea and delayed myelosuppression

igh grade or malignant brain tumors

Renal toxicity (high doses) Myelosuppression, neurological

sequelae, stomatitis, and photosensitivity

ow-grade as well as high-grade pediatric brain tumors

Nausea, constipation, and myelosuppression

brain

L

M

V

H

V

H

L

e considered when choosing among them. Despite

235

t e a e d D t a

C

t t t e a e S a i t s h a s ( c p t i a

n c s e a h e m d w a a p

P f

t c e p n o a

F

c b u p w d u t G a a F s

E

e h i n s o

s o c t m c fi b c t

C

2

hese theoretical concerns, it is thought that the pres- nce of a CNS tumor partially disrupts the BBB llowing for even large water-soluble molecules to be ffective chemotherapy agents. This is clinically evi- ent in the uptake of gadolinium on MRI as well.27

isruption of the BBB in an effort to enhance chemo- herapy penetration into the CNS remains an area of ctive research interest.

linical Trials

Great advances in the management of pediatric umors have been realized largely from the participa- ion of patients and their care providers in clinical rials. The culture within pediatric oncology is one that xpects enrollment of most patients on a treatment trial nd a study of tumor biology. The majority of patients ligible for an available clinical trial in the United tates are enrolled in one. This is in sharp contrast to dult cancer patients who are not likely to be enrolled n a clinical trial. The relative rarity of pediatric umors makes clinical trial participation through con- ortiums necessary as most single institutions will not ave enough pediatric tumor patients within a reason- ble time period to perform a clinical trial for a pecific disease. The Children’s Oncology Group COG) is an international consortium devoted to linical and laboratory research for patients with ediatric tumors. Most tertiary care centers throughout he US that treat pediatric cancer patients are COG nstitutions and many nontertiary centers are affiliated s well. COG is divided by disease discipline and there are a umber of clinical and biology trials available for hildren with brain tumors. The goal of each COG ubcommittee is to have an available clinical trial for ach specific disease. As such there is presently vailable a distinct clinical trial for medulloblastoma, igh-grade gliomas, recurrent high-grade gliomas, pendymomas, low-grade gliomas, brain stem glio- as, and others. Enrollment in a given study is at the

iscretion of the treating oncologist and must be made ith the full informed consent of a parent and the

ssent of older children. Patients may only be enrolled t COG institutions that have institutional IRB ap- roval for the clinical trial to be offered. According to their website (www.pbtc.org), The ediatric Brain Tumor Consortium (PBTC) was

ormed by the National Cancer Institute (NCI) in 1999 t

36

o improve the treatment of primary brain tumors in hildren. The PBTC comprises 10 institutions with xtensive clinical experience and research focus on ediatric brain tumors. The PBTC’s main focus is on ovel therapeutic approaches. Other goals are devel- ping more accurate prognostic markers and neuroim- ging techniques.

amily and Physician Resources There are many local resources and foundations for

hildren with cancer and specifically for children with rain tumors. Family education is vital so that they nderstand the rationale behind their child’s treatment lan and the resources available to families of children ith cancer. Centers with experience caring for chil- ren with cancer should facilitate contact for families nder their care. CureSearch (www.curesearch.org) is he parent organization of the Children’s Oncology roup and the National Childhood Cancer Foundation

nd can provide outstanding information for families nd physicians alike. The Children’s Brain Tumor oundation (www.cbtf.org) is another suggested re- ource for families of children with brain tumors.

tiology of Pediatric Brain Tumors For the vast majority of CNS tumors in children, the

tiology cannot be defined. Known risk factors include igh-dose therapeutic radiation (uncommonly a factor n pediatric patients) and genetic syndromes including eurofibromatosis type 1 and type 2, Li-Fraumeni yndrome, Turcot syndrome, Gorlin syndrome, tuber- us sclerosis, and von Hippel-Lindau syndrome.30-32

Several environmental factors have been extensively tudied but remain inconclusive. A common concern f patients and their families is the risk of exposure to ellular phone use, but at least two large studies failed o show an association between brain tumor develop- ent and cellular phone use.33,34 Another common

oncern is the effect of nonionizing electromagnetic elds that can be emitted from power lines. This has een extensively studied and the data do not support a orrelation between electromagnetic fields and brain umors.32,35,36

ommon Pediatric CNS Tumors The World Health Organization (WHO) classifica-

ion is the most widely accepted classification system

Curr Probl Pediatr Adolesc Health Care, July 2005

a 2 o t e r l m o s w I D c m r

d t i a t m c n

b t t o v i t m f r I fl c w I

d i m a m a t o

a e

L

t p t m s e d m m t e

g F v v g p

T b

T a

S

I

C

nd is categorized by histological appearance37 (Table ). The WHO system classifies CNS tumors as tumors f neuroepithelial tissue, tumors of peripheral nerves, umors of the meninges, lymphomas, and hematopoi- tic neoplasms, germ cell tumors, tumors of the sellar egion, and metastatic tumors. In the pediatric popu- ation tumors of neuroepithelial origin are by far the ost common and include astrocytic tumors, embry-

nal tumors, and ependymal tumors. The WHO clas- ification also includes tumor Grades I, II, III, and IV, hich indicate varying malignant potential with Grade

V tumors having the greatest malignant potential. espite the WHO system, many tumors remain diffi-

ult to classify and neuropathology remains one of the ore challenging subspecialties within pathology with

egard to definitive diagnosis. While cumbersome, classification is important in ictating treatment strategy and prognosis. Thus, fur- her delineation of the more common pediatric tumors s warranted. Astrocytic tumors include low-grade strocytomas, anaplastic astrocytomas, and glioblas- oma multiforme. Ependymal tumors include ependy- omas of various subtypes. Embryonal tumors in-

lude medulloblastomas and supratentorial primitive euroectodermal tumors (PNETs). Many clinicians find it useful to classify CNS lesion y location as well as tumor type38 (Table 3). Supra- entorial tumors may present with symptoms or signs hat reflect their location. Tumors in proximity to the ptic chiasm or hypothalamus often present with isual loss or visual field deficits, hormonal abnormal- ties, and behavioral changes. Tumors in the region of he pineal gland can present with abnormalities of eye ovements. Hemispheric lesions can present with

ocal neurologic signs or induce seizures. Supratento- ial tumors can also present with signs of increased CP secondary to obstruction of flow of cerebrospinal uid. Infratentorial tumors (cerebellar tumors, pilo- ytic astrocytomas, medulloblastomas) can present ith gait abnormalities, ataxia, and signs of increased

CP as a result of obstruction of the fourth ventricle.39

The frequency of certain CNS tumor types in chil- ren is markedly distinct from that observed in adults n whom metastatic lesions and high-grade gliomas are uch more common. Approximately half of all pedi-

tric brain tumors are low-grade cerebellar astrocyto- as (gliomas). PNETs (of which medulloblastomas

re a subset) are the most common malignant CNS umors in pediatric patients and account for over 20%

f all pediatric brain tumors. Other gliomas represent e

urr Probl Pediatr Adolesc Health Care, July 2005

pproximately 20% of pediatric brain tumors, and pendymomas account for approximately 10%.30-32

ow-Grade Gliomas Low-grade gliomas are the most common brain

umors in children accounting for just over half of all ediatric brain tumors. Within “low-grade gliomas” here are variations in histology, but in general treat- ent decisions are made based on the general diagno-

is of low-grade glioma rather than the subtype. For xample, children eligible for the most recent Chil- ren’s Oncology Group Protocol for low-grade glio- as included those with fibrillary, protoplasmic, and ixed variants. Other types include pilocytic astrocy-

omas, pleomorphic xanthroastrocytomas, and sub- pendymal giant cell astrocytomas. Gross total resection is likely to cure a low-grade lioma. Frequently complete resection is not possible. ollowing incomplete resection, the options are obser- ation alone with regularly scheduled MRIs or adju- ant treatment. Some data suggest that low-grade liomas may be stable or even regress following artial resection.40 As such, intervention is only nec-

ABLE 2. Classification strategies for common pediatric brain tumors y histology/tissue type: WHO classification

Neuroepithelial tumors Astrocytic

Low-grade astrocytoma Anaplastic astrocytoma Glioblastoma multiforme

Embryonal Medulloblastomas Supratentorial primitive neuroectodermal tumors (PNETs) Atypical teratoid/rhabdoid tumor

Ependymal Various subtypes of ependymomas

ABLE 3. Classification strategies for brain tumors by location: with ssociated symptoms signs

Location Associated symptoms/signs

upratentorial Signs of increased intracranial pressure Optic chiasm Visual field defects, visual loss, hormonal

abnormalities, behavior changesHypothalamic Pineal Abnormal eye movements Hemispheric Focal neurologic signs, seizures

nfratentorial Cerebellar Gait abnormalities, ataxia Brain stem tumors Signs of increased intracranial pressure/

4th ventricle obstruction/Cranial nerve dysfunction

Medulloblastoma

ssary if the patient is symptomatic or if the tumor is

237

l h i i

i l g m i t m a

r f g y s t c p p n o p t C z

M

C c l t c f t e p t p 2 C b i m p

r h s i t o p d t o r t b e i i

m n n t n o a e t w a

c p c t t e c

e c r n p ( c d p p n

2

ocated where even minimal increase in size could ave profound detrimental effect. Treatment is clearly ndicated if there is tumor progression following ncomplete resection or recurrence. Children with neurofibromatosis Type 1 have an

ncreased risk of developing CNS tumors, especially ow-grade optic pathway gliomas. The natural pro- ression of these tumors is not predictable and treat- ent can be deferred until there is clear progression or

mpending neurological compromise. Given the poten- ial added toxicity and increased risk of secondary alignancies in patients with neurofibromatosis, RT is

voided when possible for these patients. RT has been the standard therapy for progressive or

ecurrent low-grade gliomas. However, given the ef- ects of RT on young children, chemotherapy is enerally the first-line therapy for children under 10 ears of age. A recent Children’s Oncology Group tudy (No. 9952) randomly compared two chemo- herapy plans. Regimen A utilized vincristine and arboplatin, while Regimen B included thioguanine, rocarbazine, CCNU, and vincristine given for ap- roximately 1 year or until progression. The results are ot yet published regarding efficacy and safety. An- ther approach has been oral temozolomide, which rovides ease of administration with promising poten- ial tumor control.24 A new Phase I trial through the OG incorporates vincristine, carboplatin, and temo- olomide (ACNS0126).

edulloblastomas

Medulloblastomas are the most common malignant NS tumor in pediatrics. The peak incidence occurs in hildren between 5 and 9 years old.30 Given the ocation of medulloblastomas in the posterior fossa, he typical presenting signs are consistent with hydro- ephalus resulting from compression or invasion of the ourth ventricle. Other typical symptoms include irri- ability, decline in school performance, headache, mesis, truncal ataxia, nystagmus, and cranial nerve alsies.30,41 The time course of symptoms is usually 2 o 6 months before diagnosis.42 Medulloblastomas resent with distant spread of disease in approximately 0% of patients and rarely can spread outside the NS.41 A small percentage of patients with medullo- lastoma have a germline mutation such as one found n Gorlin’s syndrome, Turcot’s syndrome, Li-Frau- enmi syndrome, or ataxia telangiectasia, but for most

atients the etiology is unknown.30,41,42 R

38

Diagnosis of medulloblastoma can be predicted by adiographic appearance and location, but requires istological confirmation. Given the tendency to pread throughout the CNS, an MRI of the spine is ndicated to rule out “drop-metastases.” It is important hat the spine MRI be performed either before surgery r 10 to 14 days after surgery as postoperative blood roducts in the CNS may be confused with metastatic isease.42 Cytological evaluation of CSF is also part of he standard metastatic evaluation. Of note, CSF btained from a lumbar puncture more accurately eflects the presence of CSF seeding than CSF ob- ained from ventricular sampling.43 Since medullo- lastomas only rarely metastasize outside the CNS, xtent of disease evaluation of bone and bone marrow s not standard but must be considered if clinically ndicated. Surgical resection is a cornerstone in management of edulloblastoma; however, a gross total resection has

ot been shown to improve outcome compared with ear total resection. Patients with residual tumor less han 1.5 cm2 have had the same outcome as those with o radiographically detected residual disease.44 Post- perative MRIs obtained more than 48 to 72 hours fter surgery may be difficult to interpret based on dema, gliosis, and residual blood product degrada- ion; as such, postoperative MRI should be obtained ithin 1 or 2 days after surgery to accurately assess the

mount of residual tumor.42

Patients with medulloblastoma can be stratified ac- ording to age, stage, and risk group. Average risk atients are those 3 years or older with less than 1.5 m2 of residual tumor and no evidence of dissemina- ion. High-risk patients are less than 3 years old or hose with greater than 1.5 cm2 of residual tumor or vidence of dissemination either by radiograph or by ytological analysis of CSF. Treatment with surgery alone is not sufficient to

nsure durable remission. Standard approaches for hildren over 3 years old include surgery followed by adiation therapy to the posterior fossa and complete euraxis as well as combination chemotherapy. The resent Children’s Oncology Group clinical trial ACNS0331) for standard risk medulloblastoma in hildren 3 to 7 years old seeks to determine if the total ose of craniospinal RT and the target volume to the rimary tumor site can be reduced. Following surgery, atients are randomized to either standard-dose cra- iospinal RT (23.4 Gy) or reduced-dose craniospinal

T (18 Gy). Patients are then randomized to receive a

Curr Probl Pediatr Adolesc Health Care, July 2005

r l r r c a a c

b m t g t W a i

H

g a t H p C r o w v w g e h t r c T o o w

m t n T o e r g

m o s

E

d E t m f l r m e i c

m m f d W n s a m

t t ( a a e a t e w s i r b h c c

C

o

C

adiation boost to either the entire posterior fossa or a ocal boost to the tumor bed. During RT, patients eceive weekly vincristine. Following RT, patients eceive chemotherapy including vincristine, CCNU, isplatin, and cyclophosphamide. Children 8 years of ge and older receive standard craniospinal RT dosing, s the concern for RT toxicity is outweighed by the oncern of distant recurrence in this older age group. There is no standard approach for high-risk patients, ut a common approach in young children with edulloblastoma is to utilize more intensive chemo-

herapy with myeloablative chemotherapy and autolo- ous stem cell rescue. This approach has been shown o be feasible with an overall 3-year survival of 60%.

hile this survival rate is not acceptable, it is encour- ging that this survival rate was achieved with avoid- ng or delaying RT in many patients.29

igh-Grade Gliomas

Like “low-grade glioma,” “high-grade glioma” is a eneral pathological description that encompasses naplastic astrocytomas (WHO III), glioblastoma mul- iforme (WHO IV), and the less common gliosarcoma. igh-grade gliomas account for approximately 15% of ediatric CNS tumors and may arise throughout the NS.30 As a class of tumors, high-grade gliomas are

elatively refractory to adjuvant therapy. The chance f cure depends on the extent of resection. Patients ho undergo a resection of less than 90% of the tumor olume have significantly worse outcome than those ho undergo a greater resection.45 Even following a ross total resection, recurrence is the rule. With the xception of the youngest patients, children with igh-grade gliomas receive RT even following a gross otal resection. Even those who undergo a gross total esection and receive adjuvant RT with or without hemotherapy have less than 50% 5-year survival rate. he histology within high-grade gliomas is predictive f outcome as described by WHO grade. That is, the utcome for glioblastoma multiforme (WHO IV) is orse than for anaplastic astrocytomas (WHO III).37

The current standard therapeutic approach includes aximal surgical resection followed by conformal RT

o the area of the tumor with concurrent and mainte- ance oral temozolomide (Children’s Oncology Group rial ACNS0126). For children younger than 3 years ld, multi-agent chemotherapy with or without my- loablative chemotherapy and autologous stem cell escue has been used with limited success unless a

ross total resection was obtained.46,47 These regi- c

urr Probl Pediatr Adolesc Health Care, July 2005

ens may allow RT to be deferred until the patient is lder. So, while not curative, many believe the expo- ure to intensive chemotherapy is warranted.

pendymomas

Ependymomas constitute approximately 9% of pe- iatric CNS tumors and can arise throughout the CNS. pendymal epithelium that lines the ventricles is

hought to be the cell of origin for these tumors. The ost common sites for ependymomas to occur are the

ourth, third, and lateral ventricles as well as the umbosacral spinal cord.48 The peak incidence occur- ence is children less than 6 years old and the great ajority are intracranial.49 Approximately 10% of

pendymomas occur in the spine, but there is an ncreased frequency of ependymomas of the spinal ord in patients with neurofibromatosis type II.30

The overall survival for all children with ependy- oma is between 50 and 60%.48 However, ependy- omas include a range of histologic variants ranging

rom benign to anaplastic (WHO III) and prognosis is ependent on histology as well as extent of resection. hile ependymomas are responsive to chemotherapy,

o study has shown that chemotherapy affects overall urvival. However, chemotherapy has been shown to llow for delayed RT in young patients with ependy- oma.50,51

For older children, standard therapy includes resec- ion followed by focal RT.49,52 The current clinical rial available through the Children’s Oncology Group ACNS0121) takes into account extent of resection nd ependymoma histology. Patients with differenti- ted histology (that is, low malignant potential) who xperience a gross total resection will not receive djuvant therapy and will be followed with observa- ion alone. Patients with a gross total resection of an pendymoma with anaplastic (malignant) histology ill receive conformal RT as will patients with micro-

copic disease regardless of histology. Patients follow- ng a subtotal resection regardless of histology will eceive a trial of chemotherapy with vincristine, car- oplatin, cyclophosphamide, and etoposide with the ope that response will make second surgery with a omplete resection possible. Patients then receive onformal RT.

raniopharyngiomas

Craniopharyngiomas account for approximately 5% f all pediatric brain tumors with a peak incidence in

hildren between 5 and 14 years old. While benign in

239

t o v i C p a

c e t l

v t t s i f R s h f

B

s g r o n a a

t b g R c i l

s n a h ( G

E

t p d s o f d

t p e f e c

a t i i y h t i t l e s p b t a

o w h p f l g

v a e e t c

2

hat they do not disseminate, their local effect on the ptic pathway and pituitary can result in significant isual and endocrine effects at presentation or follow- ng therapy.53 Diagnosis is based on MRI findings, but T scan is particularly helpful to distinguish cranio- haryngiomas in children as 80% will have tumor- ssociated calcifications best seen on CT.53,54

Given the endocrine abnormalities associated with raniopharygiomas, it is helpful to involve a pediatric ndocrinologist in the patient’s care soon after presen- ation. Delayed endocrine defects may require pro- onged management. Treatment of craniopharyngiomas remains contro- ersial, with limited evidence-based data from clinical rials to support surgical resection versus radiation herapy. This is a polarizing issue without clear con- ensus.53,55,56 Many believe that for older patients RT s the preferred modality and therefore reserve surgery or younger patients in an effort to avoid toxicity from T. Chemotherapy has not been used with consistent

uccess and in general is reserved for patients that ave multiple recurrences and that are not candidates or further surgery or RT.

rain Stem Tumors

“Brain stem tumors” include diffuse intrinsic brain tem gliomas, exophytic brain stem gliomas, and tectal liomas. Intrinsic brain stem gliomas have a classic adiographic appearance obviating the need for bi- psy. Brain stem gliomas can present with cranial erve abnormalities or upper motor neuron signs, such s contralateral hemiparesis, speech changes, stridor, nd hyperreflexia.39

Surgical excision is not possible because of the umor location. As such, RT and chemotherapy have een utilized with minimal success. The 2-year pro- ression-free survival remains less than 10% despite T intensification, chemotherapy intensification, and ombination therapy. The median time to progression s generally 6 months and the median time to death is ess than 1 year.25,57,58

Patients that achieve a good initial response to teroids and RT may enjoy a window of neurological ormalization, but almost all progress. New strategies re needed. The Children’s Oncology Group presently as a trial incorporating oral Temozolomide with RT ACNS0126) and a Phase I trial utilizing Motexafin-

adolinium as a radiosensitizer for involved field RT. e

40

nd of Therapy

Even when treatment has been successful and well- olerated, the end of therapy is frequently anxiety- rovoking for the patient and family. Many have ifficulty transferring from the active role of “doing omething” to treat the tumor to the more passive role f observation. Patients must also adjust to the less requent reinforcement and support that they received uring the more intense treatment period. Just as anticipatory guidance is an important role for

he general pediatrician, anticipatory guidance from ediatric oncologists can help allay many fears and nsure better follow-up. An open discussion about the amily’s feelings about the end of therapy is an ssential part of comprehensive and compassionate are. In general, the tumor follow-up and surveillance

fter completion of therapy is straightforward. Most of he time patients require follow-up MRIs at 3-month ntervals for the first years with a gradual decrease in maging frequency over the next few years until a early follow-up MRI is adequate. It is important to ave the scans performed in a consistent manner so hat slight alterations in technique do not confound nterpretation. Many neuroradiologists prefer the scans o be performed at their center under specific guide- ines. For patients with residual MRI abnormalities or nhancement, it is important to compare each new can to a true baseline scan such as the first scan erformed at the end of therapy. Subtle changes etween two subsequent scans may be more difficult o appreciate than differences between each new scan nd the true baseline scan. When a patient can be considered “cured” depends n the type of tumor. In general, the longer off therapy ithout recurrence, the less likely recurrence will appen. Nonetheless, a small chance of recurrence ersists for several years after completion of therapy or most patients. Other important medical surveil- ance includes serial endocrine surveillance, audio- rams, and neuropsychological evaluation. For the patient, the return to routine activities can be ery challenging. Most patients have prolonged school bsences. The return to school can be very difficult specially when the patient has suffered physical, sthetic, cognitive, or emotional effects from the umor or therapy. Some centers have child-life spe- ialists that visit the school to meet with teachers or

ven the patient’s classmates to help answer questions

Curr Probl Pediatr Adolesc Health Care, July 2005

a c n p p k

f p p s m p t

c t s m T t a

L

p n c p l c r n b t

c g i m p a s t l I l v

a h

i A R f o

b o t b t d a b p

t p m r t a

f a o f e h c h s l

p d d m t f

m v t ( t

C

bout the student’s absence and return. Other physi- ians prefer to contact the school principal, teacher, or urse to discuss particular concerns regarding the atient. Of course such contact must be made with the arent’s permission and preferably with the patient’s nowledge. Another transition is regarding medical care. For

amily convenience as well as medical concerns, ediatric oncologists sometimes serve as the general ediatrician for patients receiving therapy. After inten- ive therapy, many patients’ families are unsure what edical issues should be referred to their general

ediatrician and what issues require involvement of he oncologist. Following up with immunizations is another con-

ern. During chemotherapy, few oncologists give rou- ine vaccinations. The recommendation is to provide easonal influenza vaccination. Immediately after che- otherapy, the efficacy of vaccination is variable. hree to 6 months after immune-suppressive chemo-

herapy, patients should receive routine vaccinations gain and vaccinations to catch up for those missed.

ong-Term Follow-up

With the increasing success for pediatric oncology atients in general, there has been an enhanced aware- ess regarding issues of long-term follow-up. Some enters have clinic sessions devoted strictly to these atients. Most centers do not have dedicated clinics for ong-term follow-up and rely on the oncologist to oordinate and oversee this aspect of care. While ecurrence is the most common cause of late death, eurological, neurocognitive, and endocrine distur- ances are prevalent disabilities among pediatric brain umor long-term survivors.59

Deficits may result from the tumor itself or as a onsequence of surgery, RT, or chemotherapy. In eneral most consider RT the most detrimental factor n later cognitive development. Cognitive deficits are ost completely described among children surviving

osterior fossa tumors (specifically medulloblastomas nd ependymomas). These patients tend to suffer ignificant losses in IQ scores likely from an inability o keep learning at a normal pace rather than from a oss of knowledge. Risk factors for a greater degree of Q loss include younger age at time of treatment, onger time since treatment, female sex, and clinical

ariables such as exposure to radiotherapy, radiother- s

urr Probl Pediatr Adolesc Health Care, July 2005

py dose, and treatment volume, and presence of ydrocephalus.60

There is a consensus that radiation to young children s more likely to result in significant cognitive delay. s discussed, studies have been directed at avoiding T in young patients or at least delaying RT by up

ront chemotherapy in an effort to preserve intellectual utcome.61

A recent study illustrated that survivors of medullo- lastoma demonstrated significantly less development f normal-appearing white matter (NAWM) volume han healthy controls. As NAWM volume is known to e associated with neurocognitive test performance, he authors suggest volumetric monitoring of brain evelopment to be used to guide the care of survivors nd assess clinical trial toxicity.62 Pediatric medullo- lastoma survivors also demonstrate abnormal hip- ocampal development following treatment.63

Given the known neurocognitive effects of brain umors and their management, comprehensive neuro- sychological testing is indicated for years after treat- ent. Many clinical trials provide clear guidelines

egarding the precise testing required. Multiple testing ools are often needed to provide a comprehensive nalysis from specialized psychologists.64

Endocrine complications are another major concern or pediatric brain tumor survivors. Tumor invasion nd surgical causes of endocrine dysfunction can ccur but studies suggest the risk of endocrine dys- unction is increased by RT.65 The most frequent ndocrine abnormality secondary to lesions of the ypothalamic-pituitary area is growth hormone defi- iency.66 Growth hormone replacement in growth ormone deficient brain tumor survivors has been hown to result in an overall greater final height in at east one study.67

Pediatric brain tumor patients who receive RT for osterior fossa tumors have been shown to have iminished total body and lumbar spine bone mineral ensity compared with the general population despite anagement of growth hormone deficiency and hypo-

hyroidism.68 Hearing loss is not uncommon either rom cochlear exposure to RT or from cisplatin.42

As with other pediatric cancer patients, second alignancies remain a concern for brain tumor survi-

ors. A recent study showed that the 15-year cumula- ive incidence rate for malignant second neoplasms ie, not primary tumor recurrence) was 4%. While the otal number of patients with second malignancies was

mall, there was no clear difference in estimated

241

i r

P

t c b

t U w a p h r s s H t

t t n s s

p T c m m t T t t

a p t a

n t i g a f

c t w l

S m a r r p o e a m a m d a h

t t l c d f

d w h h f c h h t c

i r s c h p w t c h

2

ncidence of second malignancy among patients that eceived different types of therapy.69

alliative Care

For many brain tumor patients, palliation becomes he main goal of therapy. Often this transition from ure-oriented care to palliative care is clear, but it may e quite subtle. The pediatric oncologist must consider the effect of

his transition on the wide range of people involved. sually the decision to pursue palliative care is made ith the patient’s parents. Parents may be resistant to

cknowledge that there is little or no hope for cure, but arents are usually thankful that they and their child ad the opportunity to make decisions based on the eality of the disease. Parents who pursue very aggres- ive therapy despite a very poor prognosis often regret ubjecting their child to the difficulties of therapy. owever, many consider a worse regret would be to

hink they had not “done everything.” Physicians need to balance the hope of new drug

rials and chemotherapy protocols with realistic expec- ations. Even for physicians and families that seek ovel or aggressive therapy, a conscious decision to witch the goal of therapy is likely and appropriate at ome point. The goal of palliative care to most oncologists is to rovide the patient the best quality of life possible. his concept is not mutually exclusive with providing hemotherapy depending on the chemotherapy regi- en and expected toxicity. Some chemotherapy regi- ens are well-tolerated, easy to take, and may prolong

he period of time that the patient has good function. he intent, however, of palliative chemotherapy is not

o cause tumor eradication, but to enhance the pa- ient’s quality of life. Eventually, palliative chemotherapy becomes less

ppropriate in the view of most oncologists as the atient’s physical and cognitive condition worsens. At his point, the goal of care is to provide comfort care nd support to the patient and family. Except for the youngest patients, parents usually eed guidance regarding how to discuss the changes in reatment with the patient. Most oncologists learn that ncorporating the patient in the discussion of treatment oals is appropriate. Of course this needs to be done in developmentally sensitive manner. While parents

requently request withholding information from their l

42

hild, it is important that patients feel they can trust heir doctor. It is likely that patients will conclude hat is going on or overhear conversations and then

earn that others are being deceptive. Physicians have different approaches to these issues. ome feel very strongly that patients of a certain age ust be told their prognosis. Other physicians will

ctively withhold or misrepresent information out of espect for the parents’ wishes. Other physicians espect parents’ wishes by agreeing not to proactively rovide details to the patient but allow the patient the pportunity to ask questions alone and with the par- nts. Questions are answered truthfully in a way that is ppropriate for the child’s understanding and develop- ent. This compromise acknowledges the parent’s

uthority and respects the patient’s autonomy. While any in health care feel compelled to disclose every

etail to a dying patient, others feel that patients will sk what they want to know when they are ready to ear the likely answer. In addition to the patient and the patient’s parents,

he oncologist must consider the needs of the rest of he patient’s family. Grandparents must cope with the oss of their grandchild and the suffering of their hildren. After experiencing major changes in family ynamics, siblings must deal with their loss and requently experience survivor guilt. In most cases, there is time to have thoughtful iscussions with parents and patients about where they ant the patient to die. Some prefer the concept of aving their child at home with coordinated home- ospice care. It is not uncommon even for these amilies to request hospitalization as death becomes learly more imminent. There are limited inpatient ospice facilities for children. Many parents prefer to ave their child die in the hospital at which s/he was reated where the staff is familiar and the parents have onfidence that the patient will be well cared for. Whether the patient is at home or in the hospital, it

s important to discuss the parent’s and patient’s desire egarding resuscitation. While a formal “Do Not Re- uscitate” consent is not strictly required when resus- itation would be medically futile, most clinicians and ospital staff prefer an open discussion with the arents and patient when appropriate to clarify their ishes. This is particularly important for patients in

eaching hospitals so that house staff and nurses have lear guidelines, and it is crucial for patients receiving ospice at home so that when the patient dies ambu-

ance staff has clear direction.

Curr Probl Pediatr Adolesc Health Care, July 2005

t d d p a c a

b f b f t d m w d

p s c r g m h p c d a

S

c i t s m w g G l c

1

1

1

1

1

1

1

1

1

C

In general, malignancy is considered a contraindica- ion for organ donation.70 In some circumstances, eye onation may be possible. While most clinicians have ifficulty raising this prospect with families, many arents will be thankful for the opportunity. It is ppropriate to check with the organ procurement enter so that parents are given accurate information nd the topic is raised only when donation is possible. After a child’s death, health care providers must alance the need to continue to provide support to the amily with the family’s need for privacy. The correct alance can only be defined for each family. Many amilies are grateful when members of the health care eam attend funerals or make contact after the child’s eath. Other families prefer more distance. At a inimum it is important that families know they are elcome to call for help and support. The relationship oes not end with the death of their child. Finally, it is important to note the effect of a atient’s death on colleagues and staff. While sea- oned oncologists may have developed a successful oping mechanism regarding patient death, the refer- ing pediatrician, junior colleagues, and staff may need uidance and support. Many centers have regular emorial services for their patients. Some centers

ave structured discussions to provide an outlet for hysicians, nurses, and other staff. Just as the health are team looked to the lead physician for treatment ecisions, members of the team may look to the ttending physician at the time of a patient’s death.

ummary Brain and spinal cord tumors represent the most

ommon solid tumors in children. Recent advances in maging techniques, neurosurgical techniques, chemo- herapy approaches, and radiation oncology have re- ulted in some improvement in overall survival and orbidity. However, the prognosis for many children ith high-grade and malignant brain tumors remains uarded in terms of mortality and long-term sequelae. iven the rarity of tumors in children, patients are

ikely best served at tertiary centers that can offer a omprehensive multidiscipline approach.

References 1. Gurney JG, Smith MA, Bunin GR. CNS and miscellaneous

intracranial and intraspinal neoplasms. In: Reis LAG, Sith MA, Gurney JG, et al, editors. Cancer Incidence and Survival

among Children and Adolescents: United States SEER Pro- 1

urr Probl Pediatr Adolesc Health Care, July 2005

gram 1975-1995. Bethesda, MD: National Cancer Institute, SEER Program; NIH Publication No. 99-4649, 1999.

2. Reddy AT. Advances in biology and treatment of childhood brain tumors. Curr Neurol Neurosci Rep 2001;1(2):137-43.

3. Gurney, JG, Kadan-Lottick, N. Brain and other central ner- vous system tumors: rates, trends, and epidemiology. Curr Opin Oncol 2001;13(3):160-6.

4. Husstedt HW, Sickert M, Kostler H, et al. Diagnostic value of the fast-FLAIR sequence in MR imaging of intracranial tumors. Eur Radiol 2000;10(5):745-52.

5. Bryant SO, Cha S, Barkovich AJ. Modern Neuroimaging of Pediatric CNS Tumors. In: Gupta N, Banerjee A, Haas-Kogan D, editors. Pediatric CNS Tumors. Berlin Heidelberg: Springer-Verlag, 2004.

6. Chang YW, Yoon HK, Shin HJ, et al. MR imaging of glioblastoma in children: usefulness of diffusion/perfusion- weighted MRI and MR spectroscopy. Pediatr Radiol 2003 Dec;33(12):836-42. Epub 2003 Oct 17.

7. Peet AC, Leach MO, Pinkerton CR, et al. The development of functional imaging in the diagnosis, management and under- standing of childhood brain tumours. Pediatr Blood Cancer 2005;44(2):103-13.

8. Hunter JV, Wang ZJ. MR spectroscopy in pediatric neurora- diology. Magn Reson Imaging Clin North Am 2001;9:165-89.

9. Preul MC, Leblanc R, Caramanos Z, et al. Magnetic resonance spectroscopy guided brain tumor resection: dif- ferentiation between recurrent glioma and radiation change in two diagnostically difficult cases. Can J Neurol Sci 1998 Feb;25(1):13-22.

0. Burtscher IM, Skagerberg G, Geijer B, et al. Proton MR spectroscopy and preoperative diagnostic accuracy: an evalu- ation of intracranial mass lesions characterized by stereotactic biopsy findings. AJNR Am J Neuroradiol 2000;21(1):84-93.

1. Girard N, Wang ZJ, Erbetta A, et al. Prognostic value of protom MR spectroscopy of cerebral hemisphere tumors in children. Neuroradiology 1998;40:121-5.

2. Pollack IF. The role of surgery in pediatric gliomas. J Neu- rooncol 1999;42(3):271-88.

3. Shaw EG, Wisoff JH. Prospective clinical trials of intracranial low-grade glioma in adults and children. Neurooncology 2003;5(3):153-60.

4. Khan RB, Sanford RA, Kun LE, et al. Morbidity of second- look surgery in pediatric central nervous system tumors. Pediatr Neurosurg 2001;35(5):225-9.

5. Gupta N, Perry V, Berger MS. Current surgical management. In: Gupta N, Banerjee A, Haas-Kogan D, editors. Pediatric CNS Tumors. Berlin Heidelberg: Springer-Verlag; 2004.

6. Keles GE, Berger MS. Advances in neurosurgical technique in the current management of brain tumors. Semin Oncol 2004;31(5):659-65.

7. Shiminski-Maher T, Cullen P, Sansalone M. Childhood Brain and Spinal Cord Tumors: A Guide for Families, Friends and Caregivers. Sebastopol, CA: O’Reilly & Associates, Inc.; 2002.

8. Doxey D, Bruce D, Sklar F, et al. Posterior fossa syndrome: identifiable risk factors and irreversible complications. Pediatr Neurosurg 1999;31(3):131-6.

9. Haas-Kogan D, Frisch BM. Advances in radiation therapy. In:

243

2

2

2

2

2

2

2

2

2

2

3

3

3

3

3

3

3

3

3

3

4

4

4

4

4

4

4

4

4

4

5

5

5

2

Gupta N, Banerjee A, Haas-Kogan D, eds. Pediatric CNS Tumors. Berlin Heidelberg: Springer-Verlag; 2004.

0. Kun LE. General principles of radiation therapy. In: Pizzo PA, Poplack DG, eds. Principles and Practice of Pediatric Oncol- ogy. Philadelphia: Lippincott-Raven Publishers; 1997.

1. Kirsch DG, Tarbell NJ. Conformal radiation therapy for childhood CNS tumors. Oncologist 2004;9(4):442-50.

2. Duffner PK. Long-term effects of radiation therapy on cogni- tive and endocrine function in children with leukemia and brain tumors. Neurologist 2004;10(6):293-310.

3. Donahue B. Short- and long-term complications of radiation therapy for pediatric brain tumors. Pediatr Neurosurg 1992;18:207-17.

4. Kuo JK, Weiner HL, Wisoff J, et al. Temozolomide is active in childhood, progressive, unresectable, low-grade gliomas. J Pediatr Hematol Oncol May 2003;25(5):372-8.

5. Wolff JE, Finlay JL. High-dose chemotherapy in childhood brain tumors. Onkologie 2004;27(3):239-45.

6. Balis FM, Holcenberg JS, Blaney S. General principles of chemotherapy. In: Pizzo PA, Poplack DG, eds. Principles and Practice of Pediatric Oncology. Philadelphia: Lippincott, Wil- liams and Wilkins; 2000.

7. Banerjee A, Mathay KK. Chemotherapy. In: Gupta N, Ban- erjee A, Haas-Kogan D, eds. Pediatric CNS Tumors. Berlin Heidelberg: Springer-Verlag; 2004.

8. Kellie SJ, Barbaric D, Koopmans P, et al. Cerebrospinal fluid concentrations of vincristine after bolus intravenous dosing: a surrogate marker of brain penetration. Cancer 2002;94:1815-20.

9. Chi SN, Gardner SL, Levy AS, et al. Feasibility and response to induction chemotherapy intensified with high-dose metho- trexate for young children with newly diagnosed high-risk disseminated medulloblastoma. J Clin Oncol 2004;22 (24):4881-7.

0. Gururangan S, Friedman HS. Recent advances in the treatment of pediatric brain tumors. Oncology Nov 2004;18(13): 1649-61.

1. Strother DR, Pollack IF, Fisher PG, et al. Tumors of the central nervous system: sequelae of treatment. In: Pizzo PA, Polack DG, eds. Principles and Practice of Pediatric Oncol- ogy, 4th ed. Philadelphia: Lippincott, Williams and Wilkins; 2002. p. 805-8.

2. Schuz J, Kaletsch U, Kaatsch P, et al. Risk factors for pediatric tumors of the central nervous system: results from a German population-based case-control study. Med Pediatr Oncol 2001;36(2):274-82.

3. Morgan RW, Kelsh MA, Zhao K, et al. Radiofrequency exposure and mortality from cancer of the brain and lymphat- ic/hematopoietic systems. Epidemiology 2000;11:118-27.

4. Inskip PD, Tarone RE, Hatch EE, et al. Cellular telephone use and brain tumors. N Engl J Med 2001;344:79-86.

5. Gurney JG, van Wijngaarden E. Extremely low frequency electromagnetic fields (EMF) and brain cancer in adults and children: review and comment. Neurooncology 1999 Jul;1(3):212-20.

6. Portier CJ, Wolfe MS. Assessment of health effects from exposure to power-line frequency electric and magnetic fields: NIEHS Working Group report. National Institute of Environ-

mental Health Sciences; 1998; NIH Publication no. 98-3981.

44

7. Gupta N, Haas-Kogan D, Banergee A. Introduction. In: Gupta N, Banerjee A, Haas-Kogan D, editors. Pediatric CNS Tu- mors. Berlin Heidelberg: Springer-Verlag; 2004.

8. Pollack IF. Brain tumors in children. N Engl J Med 1994;331(22):1500-7.

9. Ullrich NJ, Pomeroy SL: Pediatric brain tumors. Neurol Clin 2003 Nov;21(4)897-913. Review.

0. Wisoff J, Kernan JC, Horgan, MA, et al. Spontaneous invo- lution of a diencephalic astrocytoma. Pediatr Neurosurg 1998; 29:149-53.

1. Packer RJ, Cogen P, Vezina G, et al. Medulloblastoma: clinical and biologic aspects. Neurooncology 1999;1:232-50.

2. Fisher PG. Embryonal tumors. In: Gupta N, Banerjee A, Haas-Kogan D, editors. Pediatric CNS Tumors. Berlin Heidelberg: Springer-Verlag; 2004.

3. Gajjar AJ, Fouladi M, Walter A, et al. Comparison of lumbar and shunt cerebrospinal fluid specimens for cytologic detec- tion of leptomeningeal disease in pediatric patients with brain tumors. J Clin Oncol 1999;17:1825-8.

4. Zeltzer PM, Boyett JM, Finlay JL, et al. Metastasis stage, adjuvant treatment, and residual tumor are prognostic factors for medulloblastoma in children: conclusions from the Chil- dren’s Cancer Group 921 randomized phase II study. J Clin Oncol 1999;17:832-45.

5. Wisoff JH, Boyett JM, Berger MS, et al. Current neurosurgical management and the impact of the extent of resection in the treatment of malignant gliomas of childhood: a report of the Children’s Oncology Group trial no. CCG-945. J Neurosurg 1998;89:52-9.

6. Finlay JL, Goldman S, Wong MD, et al. Pilot study of high dose thiotepa and etoposide with autologous bone marrow rescue in children and young adults with recurrent CNS tumors. The Children’s Oncology Group. J Clin Oncol 1996;14:2495-503.

7. Grovas AC, Boyett JM, Lindsley K, et al. Regimen-related toxicity of myeloablative chemotherapy with BCNU, thiotepa, and etoposide followed by autologous stem cell rescue for children with newly diagnosed glioblastoma multiforme: re- port from the Children’s Cancer Group. Med Pediatr Oncol 1999;33:83-7.

8. Horn BN, Smyth M. Ependyma, in Gupta N, Banerjee A, Haas-Kogan D, editors. Pediatric CNS Tumors. Berlin Heidelberg: Springer-Verlag; 2004.

9. Allen JC, Siffert J, Hukin J. Clinical manifestations of childhood ependymoma: a multitude of syndromes. Pediatr Neurosurg 1998;28:49-55.

0. Levy AS, Gardner S, Brady K, et al. Outcome for young children with ependymoma treated with intensive induction chemotherapy followed by myeloablative consolidative che- motherapy with autologous stem cell rescue: the Head Start regimens (abstract 3232). Proc Am Soc Clin Oncol 2003;22:804.

1. Needle MN, Goldwein JW, Grass J, et al. Adjuvant chemo- therapy for the treatment of intracranial ependymoma of childhood. Cancer 1997;80(2):341-7.

2. Robertson PL, Zeltzer PM, Boyett JM, et al. Survival and prognostic factors following radiation therapy and chemother- apy for ependymomas in children: a report of the Children’s

Cancer Group. J Neurosurg 88:695-703, 1998

Curr Probl Pediatr Adolesc Health Care, July 2005

5

5

5

5

5

5

5

6

6

6

6

6

6

6

6

6

6

7

C

3. Du R, Lustig RH, Fisch B, et al. Craniopharyngioma. In: Gupta N, Banerjee A, Haas-Kogan D, editors. Pediatric CNS Tumors. Berlin Heidelberg: Springer-Verlag; 2004.

4. Moore K, Couldwell WT. Craniopharyngioma. In: Bernstein M, Berger MS, editors. Neuro-Oncology: The Essentials. New York: Thieme; 2000.

5. Wisoff J. Surgical management of recurrent craniopharyngi- omas. Pediatr Neurosurg 1994;21(Suppl 1):108-13.

6. Tarbell NJ, Barnes P, Scott RM, et al. Advances in radiation therapy for craniopharyngiomas. Pediatr Neurosurg 1994; 21(Suppl 1):101-7.

7. Packer RJ, Allen JC, Goldwein JL, et al. Hyperfractionated radiotherapy for children with brainstem gliomas: a pilot study using 7,200 cGy. Ann Neurol 1990;27:167-73.

8. Mandell LR, Kadota R, Freeman C, et al. There is no role for hyperfractionated radiotherapy in the management of children with newly diagnosed diffuse intrinsic brainstem tumors: results of a Pediatric Oncology Group phase III trial compar- ing conventional vs. hyperfractionated radiotherapy. Int J Radiat Oncol Biol Phys 1999;43:959-64.

9. Sklar CA. Childhood brain tumors. J Pediatr Endocrinol Metab 2002 May;15(Suppl 2):669-73.

0. Mulhern RK, Merchant TE, Gajjar A, et al. Late neurocogni- tive sequelae in survivors of brain tumours in childhood. Lancet Oncol 2004;5(7):399-408.

1. Lacaze E, Kieffer V, Streri A, et al. Neuropsychological outcome in children with optic pathway tumours when first- line treatment is chemotherapy. Br J Cancer 2003;89(11): 2038-44.

2. Reddick WE, Glass JO, Palmer SL, et al. Atypical white

urr Probl Pediatr Adolesc Health Care, July 2005

matter volume development in children following craniospinal irradiation. Neurooncology 2005;7(1):12-9.

3. Nagel BJ, Palmer SL, Reddick WE, et al. Abnormal hip- pocampal development in children with medulloblastoma treated with risk-adapted irradiation. AJNR Am J Neuroradiol 2004;25(9):1575-82.

4. Grill J, Kieffer V, Kalifa C. Measuring the neuro-cognitive side-effects of irradiation in children with brain tumors. Pediatr Blood Cancer 2004;42(5):452-6.

5. Anderson NE. Late complications in childhood central ner- vous system tumour survivors. Curr Opin Neurol 2003; 16(6):677-83.

6. Brauner R, Adan L, Souberbielle JC. Hypothalamic-pituitary function and growth in children with intracranial lesions. Childs Nerv Syst 1999;15(11-12):662-9.

7. Gleeson HK, Stoeter R, Ogilvy-Stuart AL, et al. Improve- ments in final height over 25 years in growth hormone (GH)-deficient childhood survivors of brain tumors receiving GH replacement. J Clin Endocrinol Metab 2003;88(8):3682-9.

8. Krishnamoorthy P, Freeman C, Bernstein ML, et al. Osteope- nia in children who have undergone posterior fossa or cranio- spinal irradiation for brain tumors. Arch Pediatr Adolesc Med 2004;158(5):491-6.

9. Broniscer A, Ke W, Fuller CE, et al. Second neoplasms in pediatric patients with primary central nervous system tumors: the St. Jude Children’s Research Hospital experience. Cancer 2004;100(10):2246-52.

0. Punnett AS, McMarthy LJ, Dirks PB, et al. Patients with primary brain tumors as organ donors: a case report and review of the literature. Pediatr Blood Cancer 2004;

43(1):73-7.

245