Fractionated Radiotherapy for Spine Tumors

Published on 26/03/2015 by admin

Last modified 26/03/2015

Print this page

This article have been viewed 1819 times

CHAPTER 252 Fractionated Radiotherapy for Spine Tumors

In the 1920s, after biologic experiments, Coutard and Regaud showed that by dividing the total dose into many smaller treatments (i.e., fractions) delivered over a period of several weeks, tumors could be cured without severe normal tissue damage.¹ Since that time, fractionated radiotherapy had been used for the treatment of tumors (benign or malignant) of the spine. Fractionated radiotherapy takes advantage of the differential radiosensitivity of normal tissues and the target lesion. The time between fractions allows some recovery of normal tissues, which reduces the risk associated with radiation injury. The inherent advantages of fractionated radiotherapy are especially appealing when treating near critical dose-limiting structures such as the spinal cord. Before proceeding with a discussion of the treatment of spine tumors, it is essential to have a solid understanding of the tolerance of the spinal cord to fractionated radiotherapy.

Radiation Tolerance of the Spinal Cord

Because spinal cord injury is disabling and untreatable, understanding the radiation tolerance of the spinal cord and risk factors for injury is of paramount importance for radiation oncologists. Two types of radiation injury have been well documented. The first is a transient, self-limited, reversible myelopathy called Lhermitte’s sign that occurs 2 to 6 months after radiotherapy. Characterized by numbness, tingling, and shock-like sensations radiating to the hands and feet when the neck is flexed, Lhermitte’s sign is a classic finding in patients with transient myelopathy. It is believed that this phenomenon is related to transient demyelination mediated by damage to oligodendrocytes. This syndrome is self-limited and resolves within weeks to months. It is not associated with chronic progressive myelitis when standard radiation doses have been delivered.²

Chronic progressive or delayed myelopathy can occur months to years after radiotherapy but fortunately is quite rare with conventional fractionation and adequate time between radiation treatments.³ Permanent myelopathy is characterized by progressive neurological signs and symptoms, including paresthesias, motor weakness, and loss of pain or temperature sensation. Patients ultimately lose bowel and bladder control and exhibit complete sensory and motor function loss. The latency period of chronic myelopathy is bimodal, with peaks of incidence occurring at 13 and 29 months. It is hypothesized that the early peak is due to white matter injury with subsequent demyelination whereas the latter peak results from radiation injury to the cord microvasculature.⁴ Because of the rarity of radiation myelitis after conventional treatment, spinal cord myelopathy caused by radiation is a diagnosis of exclusion, with recurrence of the tumor being the most common cause. Magnetic resonance imaging (MRI) may assist in the diagnosis, both to confirm that the changes have occurred within the irradiated volume and to substantiate the presence of the classic findings associated with radiation-induced myelopathy. In the early delayed phase, cord edema is seen frequently. Within 8 months of the onset of symptoms, T1-weighted images may show low intensity, whereas T2-weighted images show high intensity. Gadolinium enhancement is also common. Late changes may include spinal cord atrophy.⁵

The occurrence of chronic progressive myelopathy is dependent on the total dose, fraction size, volume irradiated, and use of chemotherapy.^4,^6,⁷ Historically, radiation oncologists have limited the spinal cord dose to 45 to 50 Gy, with conventional fractionation schedules of 1.8 to 2.0 Gy/day. A review of 1112 patients treated with multiple different fractionation schedules and a range of doses but no chemotherapy found only two cases of myelopathy in patients receiving less than 50 Gy. Because there were no identifiable risk factors that set these patients apart from others receiving similar doses, the authors argued that the onset of permanent myelopathy in patients receiving less than 50 Gy was idiosyncratic. The actual incidence of myelopathy with these conventionally fractionated doses is less than 0.2% to 0.5% after 50 Gy and 1% to 5% after 60 Gy.^3,⁴ The dose required to cause a 5% risk for myelopathy at 5 years (TD 5/5) is in the range of 57 to 61 Gy, and the 50% rate of myelopathy at 5 years (TD 50/5) is believed to be 68 to 73 Gy.⁸ Risk for spinal cord injury increases with fraction sizes above 4 to 6 Gy.⁴ Time between fractions of radiotherapy is also important. To minimize risk for injury, time between radiation treatments should be greater than 6 hours, and total dose constraints to the spinal cord should be reduced in treatment regimens administering more than one fraction per day.^4,⁹ Although historically the thoracic spinal cord was thought to be more radiosensitive than other regions of the spinal cord, newer data do not support differential sensitivities based on region of the cord. The total volume of spinal cord irradiated has been associated with risk for radiation injury, with larger volumes of cord treated associated with higher risk for myelopathy.⁴

Increasingly, combined-modality therapy that includes chemotherapy and radiation therapy is used for the optimal treatment of many tumors. Some chemotherapeutic agents are radiosensitizing in addition to their antitumor effects. Unfortunately, this radiosensitization is often not selective and may sensitize the spinal cord as well as the tumor to the effects of radiotherapy. Reports of permanent myelopathy after doses as low as 37 Gy have been noted in patients with Hodgkin’s lymphoma when combined with intensive chemotherapy and bone marrow transplantation.¹⁰ Nitrosoureas, cytosine arabinose, 5-fluorouracil, and vincristine have all been associated with myelopathy when used in combination with radiotherapy.⁶ Although reports are still rare, care should be taken with patients who have been heavily treated with radiosensitizing or neurotoxic chemotherapy until further data are available.

If left untreated, most spinal tumors, especially intramedullary tumors, would be expected to cause severe neurological dysfunction and, eventually, myelopathy. When treating a patient with a spinal cord tumor, the radiation oncologist must weigh the risk of causing myelopathy against the risk of tumor progression and subsequent severe neurological dysfunction. Radiation-induced myelopathy remains rare, in part because of the unwillingness of radiation oncologists to cause spinal cord complications as a result of treatment; this extremely low risk for permanent spinal cord damage after radiotherapy is achieved at the cost of a decrease in the probability of tumor control. Therefore, tumor progression is the most common reason for myelopathy and neurological morbidity, partially because of this caution.⁴

Intramedullary Tumors

Primary spinal tumors can be classified by their anatomic location of origin into extradural or intradural. Extradural tumors are most commonly metastatic. Intradural tumors are further divided into extramedullary and intramedullary locations. Intradural extramedullary tumors are most commonly meningiomas and nerve sheath tumors. Most intramedullary tumors of the spinal cord are glial in origin, with astrocytomas and ependymomas accounting for the majority.

Astrocytoma

Astrocytoma is the most common intramedullary spinal cord tumor and accounts for 40% to 45% of all reported cases. In children, 75% to 90% of intramedullary spinal cord tumors are astrocytomas, and about 85% of them are low-grade fibrillary or juvenile pilocytic astrocytomas.^11,¹² Long-term survival in patients with spinal cord astrocytomas is most strongly correlated with histology: patients with high-grade astrocytomas have a median survival of 8 to 10 months, whereas those with low-grade histology have a median survival greater than 15 years.¹³^–¹⁵ Low-grade tumors often have a low infiltrative nature, and microsurgical resection is the treatment of choice.¹² In contrast, radical surgery in patients with high-grade astrocytomas has no impact on outcome and may contribute to poor postoperative neurological function. For malignant, infiltrative tumors, maximal safe resection is advised. Malignant spinal cord tumors appear to have a higher risk for leptomeningeal dissemination, thus confirming the limited role of aggressive surgery.¹³ Adjuvant radiotherapy is recommended for patients with high-grade or infiltrative tumors associated with neurological dysfunction in whom gross total resection is not possible.^12,^13,¹⁶

The role of adjuvant radiotherapy for subtotally resected low-grade astrocytic tumors of the spinal cord is controversial, with treatment decisions primarily being dependent on tumor grade and age of the patient. In young children, because the late toxicity of radiotherapy is more pronounced, radiotherapy can be reserved until after a second operation for clinical recurrence.¹² Young children in whom astrocytic tumors are diagnosed before their pubertal growth spurt are at significant risk for the development of radiation-induced delay in bone growth with kyphoscoliosis or short stature, especially affecting their sitting height. This radiation-induced deformity is most severe in patients who have extensive tumors or holocord involvement of the spine.^17,¹⁸ In these young children, if their neurological function is good or improved after subtotal resection, close follow-up without adjuvant therapy is a reasonable course of action. Delaying radiotherapy until recurrence or early tumor progression may allow the child to grow at a normal rate for several years before receiving radiotherapy. Careful radiographic and neurological follow-up after surgery is required to ensure that irreversible neurological injury does not develop during these periods of “observation.” Radiotherapy delays the progression of neurological symptoms, but it is unlikely to reverse long-standing neurological deficits. To maximize long-term neurological function, radiotherapy should be initiated after radiologic progression but before clinical neurological deterioration for patients in whom no further surgery is possible or postoperatively for patients with multiply recurrent tumors in whom further surgery is likely to cause significant morbidity. Collaboration between the radiation oncologist and neurosurgeon is critical to determine the most appropriate timing for adjuvant radiotherapy in this group of patients.

In teenagers and adults, observation is reasonable for pilocytic astrocytomas, but adjuvant radiotherapy should be considered for all patients with World Health Organization grade II, III, and IV astrocytomas.¹⁹ The largest experience of spinal cord astrocytomas comes from the Mayo Clinic: 69 pilocytic and 67 infiltrative (nonpilocytic) astrocytomas. The vast majority of the patients in this series were adults (mean age, 35 years). Tumor grade was a strong predictor of prognosis (median survival for grade I not reached; grade II, 4.1 years; grade III, 1.6 years; and grade IV, 0.8 years). Greater extent of resection did not predict a better outcome, but adjuvant radiotherapy did for infiltrative astrocytomas, and this held true on multivariate analysis. Therefore, based on these results and the work of others, adjuvant radiotherapy should be strongly considered for all teenagers and adults with infiltrative (nonpilocytic) astrocytomas.²⁰

Radiotherapy Dose and Techniques

Dose-response data from randomized trials in the cerebrum suggest that there is no difference in outcomes for doses between 45 and 69.4 Gy.^21,²² Retrospective dose-response data from spinal cord tumors suggest an improvement in time to progression for patients receiving greater than 40 Gy versus those who received lower doses.²³ The most commonly used doses for the treatment of both low- and high-grade spinal cord astrocytomas range from 45 to 55 Gy,⁸ depending on the volume of spinal cord being irradiated and the patient’s neurological function. Unfortunately, even with these doses, local recurrence is the predominant pattern of treatment failure, especially for low-grade tumors.^14,²⁴ In rare cases in which the tumor is so advanced that no meaningful recovery of function is expected, dose escalation resulting in anticipated “radiocordectomy” has been shown to control disease, albeit with permanent disability.²⁵

Patient age, overall prognosis, and location and size of the tumor are all factors that influence the radiation technique. Historically, primary tumors of the spinal canal were treated with a direct posterior field. Advantages of this technique include avoidance of delivery of radiation to structures lateral to the spine, including the kidneys and lungs. The primary disadvantages of a single posterior field are the high dose to superficial subcutaneous tissues and a relatively higher dose to immediately anterior structures (such as the larynx in a cervical spine tumor) when compared with a multifield technique. The primary disadvantage of a multifield technique is the larger volume receiving low-dose radiation, which may put the patient at a relatively higher risk for a second malignancy. Thus, knowledge of patient’s prognosis and risk for late effects is important in design of the radiotherapy plan.

For the treatment of adults with radiotherapy, the chance of damage to surrounding soft tissues is exceedingly small; nonetheless, the benefits gained from multifield treatment in reducing acute toxicity outweigh the remote risk for second malignancy, especially since the prognosis is not as favorable as in children. Although a second malignancy is a greater concern in children, higher doses of radiation are associated with reduced vertebral body bone growth ¹⁷ and increased risk for muscle and soft tissue hypotrophy, which can result in reduced paraspinal support in an already unstable spine.¹⁸ Multifield techniques should help reduce the risk for significant hypotrophy in children, although no long-term data exist as yet to support this hypothesis.^24,²⁶

Modern radiotherapy uses imaging information from preoperative and postoperative MRI, as well as treatment-planning computed tomography (CT), to design a conformal radiotherapy plan. With more precise dose administration, it is becoming increasingly important to accurately delineate the radiotherapy target volumes. The gross tumor volume (GTV) is generally the residual tumor visible on imaging. A margin for subclinical spread, generally 1 cm for low-grade tumors and 1 to 2 cm for higher grade tumors, is anatomically confined to the spinal canal to create a clinical target volume (CTV). These tumors appear to spread craniocaudally, so a larger margin superiorly and inferiorly may be advisable for infiltrative, high-grade lesions.²⁷ Careful review of imaging and communication with the neurosurgeon and pathologist to help determine the extent of spread along nerve roots are also important. An additional 0.3- to 0.5-cm margin is added to account for setup error, depending on the treating institution’s immobilization technique, to create the planning target volume (PTV). The dose is typically prescribed so that 95% of the prescription covers the PTV. Three-dimensional conformal techniques use CT guidance to design radiotherapy target volumes, design beams and blocks, and model a radiation plan. Another treatment technique is intensity-modulated radiotherapy (IMRT); it uses sophisticated computer software to control the blocking leaves on the linear accelerator, which then move throughout the treatment to generate a more complex and conformal (around the radiotherapy target) plan.

In special cases, more simple radiotherapy techniques may provide adequate dose coverage while minimizing risk to critical normal organs. Tumors exclusively involving the cervical spine may be treated with opposed lateral fields to avoid incidental irradiation of the thyroid, hypopharynx, and oral cavity. In female patients requiring treatment of the lumbosacral spine for cauda equina tumors, a lateral beam technique to avoid exit irradiation to the pelvis will generally result in the lowest dose to the radiosensitive ovaries. Wedges may be required on these lateral lumbosacral fields to provide a homogeneous dose distribution. If the tumor volume extends superiorly to the L1-3 region, a split-beam technique (lateral fields below the isocenter, anteroposterior fields above the isocenter) to allow anteroposterior arrangement of the beams superiorly for avoidance of the kidneys will eliminate beam divergence between the fields.

Craniospinal or spinal axis irradiation is not generally indicated for the treatment of spinal cord tumors unless neuraxis dissemination is present at the time of diagnosis. In adults with neuraxis dissemination, craniospinal axis irradiation is rarely indicated because of the poor tolerance of craniospinal irradiation in adults and the poor prognosis; for adults, treatment is usually limited to bulky or symptomatic sites (or both). For younger patients with neuraxis dissemination, doses of 39.6 to 50.4 Gy in 1.8-Gy fractions have been used in combination with a boost dose given to bulky areas of disease, especially for higher grade tumors.²⁴ If more than half the spinal cord is irradiated, the total dose should probably not exceed 45 Gy, but small segments may safely tolerate 55 Gy. In young children, the dose to the spinal cord should be limited to between 40 and 45 Gy in 1.5- to 1.75-Gy daily fractions because the developing cord may have a lower tolerance for radiotherapy.

Ependymoma

Ependymomas are usually histologically benign, and they may have a long and often indolent course. Poorly differentiated ependymomas in the spine are rare. Rostral tumors are more frequently cellular variants, whereas distal tumors (including those of the cauda equina) are more commonly a myxopapillary variant. Approximately two thirds of ependymomas occur in the region of the cauda equina; they are not truly intramedullary and are reviewed in the section “Extramedullary Tumors.” Spinal ependymomas are more common in adults than in children.^28,²⁹

Complete resection is more frequently achieved for ependymomas than for astrocytomas, and intraoperative assessment of tumor resection is more reliable with ependymomas than with astrocytomas. Advances in microsurgical techniques have contributed to improved postoperative outcomes for patients with intramedullary spinal cord tumors. The frequency of complete excision varies from 33% to 94%.^30,³¹ For patients undergoing complete en bloc gross excision, the prognosis is excellent, with cure rates exceeding 90% in some series.³² Adjuvant radiotherapy is not recommended in these cases.

Substantial evidence supports the use of postoperative radiotherapy in patients with ependymoma after incomplete or piecemeal excision.^23,³³^–³⁵ Modern series with long-term follow-up confirm these results. Wahab and colleagues reported an 80% progression-free survival rate at 15 years in patients receiving adjuvant radiotherapy after incomplete or piecemeal resection.³⁶ Dose-response data also indirectly suggest a role for postoperative therapy. Some radiotherapy series have demonstrated that increasing doses of radiation are associated with better tumor control in patients with ependymomas. Shaw and colleagues reported that the local failure rate was 35% in patients receiving 50 Gy or less as opposed to just 20% in patients receiving more than 50 Gy.³⁴ Local control and overall survival in patients who receive adjuvant radiotherapy after incomplete resection for low-grade ependymomas are excellent and similar to rates in patients who undergo gross total resection.³⁵^–³⁸

The radiotherapy dose and techniques for intramedullary ependymomas are similar to those for astrocytomas (see earlier). CTV is generally defined as a 1- to 2-cm margin on residual tumor on T1-weighted contrast-enhanced MRI plus any edema seen on T2-weighted or fluid-attenuated inversion recovery (FLAIR) sequences. As with astrocytomas, a more generous CTV is recommended for malignant and infiltrative tumors. If no residual tumor is visible, the postoperative cavity (which preoperative imaging can help delineate) is considered the GTV. A total dose of 45 to 50.4 Gy is typically given in 1.8- to 2-Gy daily fractions to the CTV, along with a boost for a total dose of 54 to 55.8 Gy to areas of bulky residual disease (GTV) (Fig. 252-1).