Neha Vapiwala, MD and John P. Plastaras, MD, PhD
Updated by J. Taylor Whaley, MD
The Abramson Cancer Center of the University of Pennsylvania
Last Modified: August 12, 2011
Almost 20% of all childhood cancers start in the central nervous system (CNS), which consists of the brain, the spinal cord, and the surrounding fluid (cerebrospinal fluid, or CSF), lining tissues (meninges) and bone (cranium and vertebrae).
In the past several years, the incidence of pediatric CNS tumors has been increasing. This increase is partially explained by medical advances that have led to earlier detection and diagnosis of brain tumors.
There are many different types of pediatric CNS cancers, all of which have long, rather complicated names. The tumor type is generally determined by two things: 1) where in the CNS the tumor starts, or the location, and 2) how the tumor looks under a microscope, also known as the histology. Primary brain tumors are tumors that arise in the brain, while primary spinal tumors grow in the spinal cord. However, some brain tumors can spread to involve parts of the spine, and vice versa. On rare occasion, certain types of tumors can even spread to areas outside of the CNS, such as distant bones or bone marrow.
Ultimately, every cancer treatment plan is individualized for that patient and takes into account not only the stage and the clinical data, but also the goals and desires of the patient and his or her family.
Before proceeding any further, it is worthwhile to properly define and clarify some important concepts that are relevant to the subject of brain tumors.
The word tumor itself refers to an abnormal growth or mass of cells, also referred to in general terms as a "cancer".
Tumors (cancers) can be either "benign" or "malignant", depending on how "bad" the tumor cells look under a microscope. In simple terms, "benign" tumor cells have a sort of "innocent" appearance under the microscope; they usually do not have features associated with invasion or spread (metastasis) to other locations. In contrast, "malignant" tumors have cells that can, and are likely to, invade and spread, especially if left untreated.
However, it is very important to realize that a benign microscopic tumor is NOT necessarily clinically benign or innocent. Remember, a benign tumor is still a tumor, and it represents uncontrolled, inappropriate growth. Although benign tumors may not spread far beyond where they started, they can continue to grow and thus can still cause trouble from local expansion, particularly in an otherwise small, enclosed space like the brain. Pressure and obstruction from a growing mass can lead to severe neurological deficits and, worst case scenario, even death.
Craniopharyngioma is a benign brain tumor that accounts for about 3-9% of all pediatric CNS cancers. This tumor results from abnormal overgrowth of a part of the brain called Rathke's pouch, located near the pituitary gland. It typically consists of a combination of calcium deposits mixed with cysts, or pockets, of fluid. The fluid characteristically contains protein, lipid (fat) and cholesterol pieces, giving it a so-called "crankcase oil" texture.
Craniopharyngiomas can occur in both children and adults, with a peak in incidence at 9 to 14 years of age. There are approximately 120 cases diagnosed each year in the United States in patients under the age of 19 years old. In fact, more than 50% of all patients with craniopharyngioma are under the age of 18 years. There is no clear association of the tumor with a particular gender or race. It is not really known what causes craniopharyngiomas, but they do not appear to "run in families" or to be directly inherited from the parents.
The most common symptoms of this tumor are visual changes, most often loss of peripheral vision in one eye (visual field cuts) or blurriness. This is because of the proximity of craniopharyngiomas to the visual nerve pathways in the brain. Similarly, craniopharyngiomas are also close to the pituitary gland, and thus children can present with hormonal, or endocrine, problems. This is because the pituitary gland controls various important hormone systems, including those vital for body growth, sexual organ maturation, and thyroid gland function. Up to half of children can even develop notable personality changes or cognitive deficits. Finally, general signs like nausea, vomiting and headache may occur due to increased and excessive pressure in the brain from the tumor.
Please note that the symptoms mentioned here do not necessarily or automatically mean a child has a brain tumor, but further medical evaluation is required to rule out the possibility of a brain tumor such as craniopharyngioma.
Typically, the diagnosis is made based on a radiology study such as magnetic resonance imaging (MRI) scan of the brain with gadolinium contrast-enhancement. A craniopharyngioma has a very characteristic appearance on MRI scan: a well-defined mass with a combination of solid and cystic parts. Importantly, this is a tumor that occurs in the suprasellar region, a part of the brain also containing the optic chiasm (contains nerves for vision) and hypothalamus (regulates release of growth, thyroid, and stress hormones, among others).
As mentioned earlier, craniopharyngioma is a benign tumor and thus is not expected to spread from the brain to the spine, cerebrospinal fluid (CSF), or other sites outside the CNS. Thus, no additional scans are required unless indicated by symptoms or physical exam findings.
Ironically, although craniopharyngiomas are benign tumors that should have relatively good long-term outcome, the optimal management of them is a very controversial issue in pediatric neurooncology. One major explanation for this is the inherent conflict between wanting to "cure" the patient and wanting to minimize the long-term morbidity, or negative side effects. Physicians are understandably hesitant to recommend and pursue the most aggressive treatment, if it risks leaving the child with worse neurologic problems than the underlying tumor itself. Recent publications from St. Jude’s Children’s Hospital highlight these concerns. Of 55 patients treated, 93% remained alive at the time of analysis; however, long term side effects included hormone deficiencies, shunt dependence, and seizures. Due to the potential for long term problems, it is very beneficial for children with craniopharyngiomas to be followed with multidisciplinary care to minimize the late complications.
So although no clear consensus has been reached in the medical community on the best therapeutic regimen for craniopharyngioma, presented below are some general management principles. It is important to remember, however, that each patient's case should be approached individually, and the therapy plan designed accordingly.
Surgery is typically the initial and major component of therapy, although the "right" or "best" extent of surgery is a matter of much debate. Some believe that the main goal should be gross total resection of the tumor, or in other words, complete surgical removal of all visible tumor. It is technically possible to perform a total resection in about 70-90% of cases. The success rate of total resection alone also ranges from 70-90%, (meaning the tumor comes back in about 10-30% of patients), which is quite good. So if total tumor resection is achieved based on both what the surgeon describes and what the post-surgical MRI scan shows, then careful observation after surgery is reasonable. Any additional treatment could be used only if and when indicated (ie: if the tumor comes back in future).
A complete surgery with no other required treatment sounds great, but bear in mind that the rates of serious surgical complications following complete tumor resection are high: 50-90% risk of damage to the hypothalamus, 1-12% risk of brain hemorrhage during the surgery, and even a small risk of death. Not surprisingly, the skill and experience of the surgeon plays an important role in the outcome, both in terms of better tumor control and fewer surgical complications.
However, even in the best of hands, there are many cases in which the craniopharyngioma may press against or grow rather close to various critical brain structures (optic chiasm, hypothalamus, blood vessels), thereby making total resection essentially impossible. Instead, the surgeon removes as much as tumor as is safely possible. Simply observing after such a surgery would seem like a bad idea, since there is known tumor still left behind and untreated.
This has led to several important questions: Can you perform a partial* resection in these difficult cases, but then follow it with radiation therapy to eradicate, or "mop up", the remaining tumor cells? Could this approach achieve equally good results as a total resection (if it had been possible), while decreasing the possible long-term side effects of a big surgery? In fact, knowing that the naked eye at the time of surgery cannot possibly see individual, microscopic tumor cells that might be left behind, should gross total resection alone EVER be done? Instead, should partial surgery and postoperative radiation for residual tumor be the goal for all patients, thus minimizing surgical risk and maximizing microscopic tumor control.
*(Please note: the term "partial" will be used synonymously with "incomplete" or "subtotal", all of which describe any surgery that is less than a total resection.)
Studies have tried to answer these questions, and found that craniopharyngioma patients treated with partial resection and radiation have 10-year and 20-year progression-free-survival rates as high as 70-90%. In other words, these patients are living without signs of growing or returning tumor as long as 10 and 20 years after treatment. Notice that this is comparable to the rates mentioned earlier for complete surgery alone, but with much lower rates of problems like hypothalamus damage or brain hemorrhage.
Regarding post-surgical radiation therapy: Firstly, a post-operative MRI scan of the brain is usually obtained to look for any leftover tumor. Residual craniopharyngioma can appear as anything from calcified little "flecks", to a more standard-looking cystic-solid mass. Based on what the MRI shows, together with what the surgeon reports he or she was able to do in the operating room, the surgery is classified as a "total" versus "less-than-total" (aka "partial", "subtotal", or "incomplete") resection.
If it is a partial resection, the patient should receive radiation therapy to the tumor site, which consists of both solid and cystic component of the tumor, to reduce the risk of local tumor recurrence. If the radiation is delayed in such cases and the patient is observed, the tumor will undoubtedly return within 3 years (70-100% of these patients experience relapse). On top of that, treatment at the time of relapse is often more complicated and dangerous. Thus it is generally recommended that patients who get partial resections should receive the radiation therapy upfront, rather than delaying it until the tumor comes back. In the postoperative setting, radiation is usually delivered only to the region of the original tumor, including the surgical area and all remaining tumor. Surrounding brain tissue and other critical structures are excluded from the treatment field as much as possible.
Much attention has understandably been paid to the possible long-term complications of radiation therapy to the brain and spine of a growing child. These can include deficits in memory, learning, and social/emotional adjustment, among other things. The development of such side effects depends on many factors, including extent of pre-radiation surgery, amount and location of brain that is treated, age of the child at the time of radiation, and how much radiation dose is given, among others.
However, modern radiotherapy techniques and proper attention to minimizing radiation dose to important brain structures whenever feasible can allow for safe and effective treatment, even in younger children. There have been considerable efforts within the field to decrease the dose of radiation and volume of brain that receives radiation to minimize long term toxicity. While no therapy is without its side effects, radiation therapy can be planned and delivered in such a way as to minimize any potential long-term side-effects. Newer methods of delivering radiation therapy, such as stereotactic radiosurgery or proton radiotherapy (see below), may be promising options to limits radiation to normal brain tissues. Radiotherapy for craniopharyngioma is best accomplished at a major radiation oncology center where physicians and staff are familiar with pediatric patients and technologically capable of treating childhood cancers.
Proton radiotherapy is a special kind of radiotherapy that uses protons instead of the usual x-rays. The advantage of proton radiotherapy is that it can be shaped to a patient's tumor better than x-ray therapy, avoiding treatment of normal structures, and thus decreasing radiation side effects.
At Loma Linda, 16 patients with craniopharyngioma (age 7-34 yr) were treated with protons after at least one resection (Cancer, 2006). They were treated with 50.4-59.4 CGE (Cobalt Gray Equivalents, the accepted unit of proton radiation that equates to 1 Gray of photon radiation), and 14/15 patients had not recurred after 25 months of follow-up. There were few acute side effects from treatment, but there were late effects observed in four patients, namely panhypopituitarism, one stroke, and one meningioma that occurred out of the proton-field in the patient that had previous photon radiation.
In Boston, 24 children with craniopharyngioma were treated with protons between January 2001 to August 2007 (IJROBP, 2009). Eight had recurrent tumors. The median dose was 52.8 GCE treated with a median of 4 proton fields. Due to intermittent cyst growth during treatment, MRI’s were obtained to monitor cysts to ensure that the tumor was covered completely with radiation throughout treatment. Interestingly, when patients were rescanned during treatment, six were found to have cyst growth during radiation, four required re-planning. After median follow up of 40 months of follow-up, there were no local failures seen.
In Orsay, France, proton radiotherapy was used to increase radiation doses delivered to craniopharyngiomas to 59 CGE. They were able to limit radiation dose to important structures, such as the hearing apparatus, optic nervous structures, and brainstem.
At St. Jude Children’s Hospital from 2004 to 2007, 14 children with craniophayngioma underwent weekly MRI to compare various treatment techniques with photon and proton radiation. Although IMRT (intensity modulate radiation therapy) had lower dose to visual structures, proton therapy had lower dose to hearing structures, normal brain tissue, and total body doses. Their conclusions of the study were that protons have the potential to decrease low dose radiation in pediatric patients with craniopharyngiomas. However, because of the risk of cyst growth, MRI’s should be performed at intervals during treatment to ensure the tumor remains stable in size through treatment.
At the 2006 and 2010 PTCOG meetings (proton therapy professional society meetings), Indiana University presented their preliminary results for proton therapy used in pediatric craniopharyngiomas. From 2004 to 2010, 21 patients (age 5-21 years old) were treated to median of 54 CGE (range- 50.4 to 57.6 CGE) with 3 fields. At a median follow up of 32 months, all patients remained alive with one local failure in the radiation field and one failure outside of the radiation field (Event Free Suvival was 90.5%). 2 patients required cyst drainage. Toxicities during treatment included intermittent headaches, alopecia, and fatigue. No hormonal deficits developed after radiation, although these were present in many prior to radiation treatments. No visual deficits have developed and one patient had neurocognitive decline requiring special education. Although this was a small subset of patients, proton beam radiotherapy for treatment of craniopharyngiomas appears to be feasible and safe in the acute setting.
At this time, proton therapy for treatment of craniopharyngiomas remains experimental, but it offers promise to deliver radiation doses high enough to prevent regrowth with the added benefit of potentially fewer side effects.
There is currently no role for chemotherapy in the treatment of craniopharyngioma. As it is a benign tumor, disease spread to distant locations is not typically expected, and thus systemic (ie: "body-wide") medication like chemotherapy has not been extensively used. Some have tried to inject chemotherapy directly into craniopharyngioma cysts in order to delay the need for surgery/radiation, but this approach is experimental.
Beltran C, Roca M, Merchant TE. On the Benefits and Risks of Proton Therapy in Pediatric Craniopharyngioma. Int J Radiat Oncol Biol Phys. 2011 May 11.
Crom DB, Smith D, Xiong Z, Onar A, Hudson MM, Merchant TE, Morris EB. Health status in long-term survivors of pediatric craniopharyngiomas. J Neurosci Nurs. 2010 Dec;42(6):323-8; quiz 329-30.
Halperin EC, Constine LS, Tarbell NJ, Kun LE. Pediatric Radiation Oncology. 4th Edition. Lippincott Williams & Wilkins (2205).
Luu QT, Loredo LN, Archambeau JO, Yonemoto LT, Slater JM, Slater JD. Fractionated proton radiation treatment for pediatric craniopharyngioma: preliminary report. Cancer J. 2006 Mar-Apr;12(2):155-9.
Winkfield KM, Linsenmeier C, Yock TI, Grant PE, Yeap BY, Butler WE, Tarbell NJ.. Surveillance of craniopharyngioma cyst growth in children treated with proton radiotherapy.
Int J Radiat Oncol Biol Phys. 2009 Mar 1;73(3):716-21.