paediatrics Brussels 17

Childs Nerv Syst

Hypothalamic dose volume effects have been modeled for patients with ependymoma suggesting that the risk of endocrinopathy is low for most patients and that while even low doses to the hypothalamus result in a risk for growth hormone deficiency, other endocrinopathies are even less common if baseline assessments prove to be normal. Preexisting endocrine deficiencies in these children corre- late with ventricle size (hydrocephalus) at diagnosis [ 28 , 29 ]. The peer-reviewed scientific literature contains numerous references to highly focused focal irradiation for intracra- nial ependymoma in children; however, there is only one perspective conformal series. The primary measure of success for conformal radiation therapy is local tumor control corrected for extent of resection which is the most important prognostic factor. There are several contemporary series that utilized fully or to a large extent conformal radiation therapy. Local tumor control has been estimated at 68 – 89% when measured at 3 – 5 years (Table 1 ) [ 18 , 30 – 32 ]. The rates of local control in the modern series are considerably higher than those inferred from historic series where event-free survival and not local control rates have been reported (Table 2 ) [ 2 – 10 ]. The use of IMRT in very young children has raised concern about extraneous dose to normal tissues. Mansur et al. [ 33 ] found that IMRT lowered peripheral doses near the target. This was attributed to reduced internal scatter due to smaller effective field sizes. The thyroid was given as an example of a critical peripheral organ near to the targeted volume. The peripheral dose was similar for both IMRT and three-dimensional CRT indicating that peripheral dose was difficult to predict by monitor units which are often significantly greater for IMRT. Result IMRT

The optic chiasm dose should be managed in a very similar fashion to the spinal cord and should be defined on CT or MR and appearing on at least two successive images. If the cumulative treatment dose may exceed 54 Gy, the chiasm should be excluded from the treatment after 54 Gy and receive no more than 1.25 Gy per fraction at any point. These guidelines also allow for the coverage of the target volumes to be compromised after 54 Gy in selected cases. Each cochlea should be contoured separately on the CT data as a circular structure within the petrous portion of the temporal bone. The size and position of the contoured cochlea should be confirmed by viewing the structures in three dimensions using the treatment planning system and on two successive CT images. The mean dose to the cochleae should be limited to 35 Gy. At these levels, the risk of hearing loss is less than 5% [ 25 ]. The brainstem is central to the irradiated volume in patients with posterior fossa tumors, and while major side effects from radiation therapy have not been widely reported, investigators remain very concerned about the long-term effects of irradiation especially for children who suffer neurological effects from tumor and surgery. Recent data suggest that factors impeding neurologic recovery in children with ependymoma treated with high-dose postop- erative radiation therapy do not include radiation dose, rather, the volume of tumor and clinical and treatment factors related to tumor and surgery. Given the safety profile of radiation therapy as administered in recent trials and plans to further reduce the target volume for radiation therapy, the risk of side effects involving the brainstem should be further diminished [ 26 ]. Temporal lobe and whole brain doses of radiation therapy are correlated with cognitive outcome corrected for the age of the patient at the time of irradiation. This important knowledge has driven investigators to find new ways to reduce dose to normal tissues (shrinking target volume margins) or this high-dose volume of irradiation using conformal methods. Evaluating patients with ependy- moma and considering radiation effects should not be absent the potential effects of hydrocephalus [ 18 , 27 ].

Table 1 Local tumor control estimates from contemporary reports using postoperative irradiation

Series

Time period

Patients

GTR (%)

Local control

MacDonald-PBRT Schroeder-IMRT Massimino-HFRT Merchant-CRT/IMRT

2000 – 2006 1994 – 2005 1993 – 2001 1997 – 2007

17 22 46

76 77 74 85

86% at 2 years 68% at 3 years 70% at 4 years 89% at 5 years

153

PBRT proton beam radiation therapy, IMRT intensity-modulated radiation therapy, HFRT hyperfractionated radiation therapy, CRT conformal radiation therapy, GTR gross-total resection