MEDLINE literature search performed through August 2011. The literature on the use of intensity modulated radiation therapy (IMRT) in the central nervous system (CNS) consists of dosimetry planning studies and case series; no comparative studies using IMRT versus other conformal radiation modalities (e.g., 3-dimensional conformal radiation [3D-CRT]) were identified.
High-grade malignant tumors
Amelio and colleagues (2010) conducted a systematic review on the clinical and technical issues of using IMRT in newly diagnosed glioblastoma multiforme (GBM). (1) The articles included in the review were through December 2009 and included 17 studies (9 related to dosimetric data and technical considerations, 7 to clinical results, and 1 to both dosimetric and clinical results) for a total of 204 treated patients and 148 patient datasets used in planning studies. No randomized controlled studies (RCTs) were identified, and a meta-analysis was not performed.
For the 6 papers related to planning studies that compared either 3D-CRT versus IMRT, 1 study showed a noticeable difference between 3D-CRT and IMRT for the planning target volume (PTV) (13% benefit in V95 [volume that received 95% of the prescribed dose] from IMRT, p<0.001) (4); the remaining studies suggested that IMRT and 3D-CRT provide similar PTV coverage, with differences between 0 and 1%. Target dose conformity was found to be improved with IMRT.
The organs at risk (OAR) typically under consideration in the studies were the brainstem, optic chiasm, optic nerves, lens and retina. In general, IMRT allowed better sparing of the OAR than 3D-CRT but with considerable variation from study to study.
The 8 studies that included clinical results included 3 retrospective, 1 prospective Phase I and IV prospective Phase II single institution studies. Of these 8 studies, 2 used conventional total dose and dose per fraction, 2 used a hypofractionated regimen, and in the remaining, a hypofractionated scheme using a simultaneous integrated boost. Chemotherapy was administered in 6 of 8 series, concomitantly with radiation and in the adjuvant phase. Median follow-up ranged from 8.8 and 24 months. Almost all patients (96%) were able to complete the treatment without interruption/discontinuation due to toxicity. Acute toxicity was reported as negligible with grade-3 side effects observed in only 2 studies at rates of 7% and 12%. Grade-4 toxicity was recorded in only 1 series with an absolute rate of 3%. Data for late toxicities were available in 6/8 studies, with 1 study recording grade-4 side effects with an incidence of 20%. One-year and 2-year overall survival (OS) varied between 30% and 81.9% and between 0% and 55.6%, respectively. When OS was reported as a median time, its value ranged from 7 to 24 months. Progression-free survival (PFS) ranged from 0% and 71.4% at 1 year and 0% and 53.6% at 2 years. Median PFS was reported as ranging from 2.5 to 12 months.
The authors also carried out a comprehensive qualitative comparison with data reported in the literature on similar non-IMRT clinical studies and offered the following conclusions. The results of the planning comparisons showed 3D-CRT and IMRT techniques provide similar results in terms of target coverage, IMRT is somewhat better than 3D-CRT in reducing the maximum dose to the OAR, although the extent varied from case to case, IMRT is clearly better than 3D-CRT in terms of dose conformity and sparing of the healthy brain at medium to low doses and that (in general) there were no aspects where IMRT seemed worse than 3D-CRT.
This evidence is limited by a number of factors. There is an absence of comparative studies with clinical outcomes, all of the studies were small in size, from a single institution, a majority of patients (53%) were retrospectively analyzed, and the administration of chemotherapy was variable across studies.
A representative sample of the comparative studies on dose planning and the single-arm studies with clinical outcomes are discussed below.
MacDonald and colleagues (2007) compared the dosimetry of IMRT and 3D-CRT in 20 patients treated for high-grade glioma. (5) Prescription dose and normal-tissue constraints were identical for the 3D-CRT and IMRT treatment plans. The IMRT plan yielded superior target coverage as compared with the 3D-CRT plan. The IMRT plan reduced the percent volume of brainstem receiving a dose greater than 45 Gy by 31% (p=0.004) and the percent volume of brain receiving a dose greater than 18 Gy, 24 Gy, and 45 Gy by 10% (p=0.059), 14% (p=0.015), and 40% (p< or=0.0001), respectively. With IMRT, the percent volume of optic chiasm receiving more than 45 Gy was reduced by 30.4% (p=0.047). As compared with 3D-CRT, IMRT significantly increased the tumor control probability (p< or=0.0005) and lowered the normal-tissue complication probability for brain and brain stem (p<0.033).
Narayana and colleagues (2006) reported the outcomes of 58 consecutive patients with high-grade gliomas treated with IMRT. (6) GBM accounted for 70% of cases and anaplastic gliomas for the remainder. Surgery consisted of biopsy alone in 26% of patients and of those that underwent resection, 63% had total or near total resection and 37% had partial resection. Eighty percent of patients received adjuvant chemotherapy. Median follow-up was 24 months. Acute neurotoxicities were grade 1/2 in 36% of patients, grade 3 in 7%, and grade 4 in 3%. Late toxicities were grade 1/2 in 10%, grade 3 in 7%, and no grade 4 or 5. Freedom from late neurotoxicity at 24 months was 85%. Median OS for the anaplastic astrocytomas was 36 months and 9 months for the GBM group. From these data, the authors concluded that the use of IMRT in high-grade gliomas does not appear to improve survival
Narayana et al. (6) also performed a comparison of the IMRT treatment plans with 3D plans performed in 20 patients out of 58 total in that case series. Regardless of tumor location, IMRT did not improve PTV target coverage compared to 3D planning. IMRT decreased the maximum dose to the spinal cord, optic nerves, and eye by 16%, 7%, and 15%, respectively. These data indicate that IMRT may result in decreased late toxicities.
Huang and colleagues (2002) compared ototoxicity with use of conventional (2D) radiotherapy (n=11) versus IMRT (n=15) in 26 pediatric patients with medulloblastoma. (7) All of the patients also received chemotherapy. When compared to conventional radiotherapy, IMRT delivered 68% of the radiation dose to the auditory apparatus, but full doses to the desired target volume. Median follow-up for audiometric evaluation was 51 months (9-107 months) for the conventional radiotherapy group and 18 months (8-37 months) for the group that received IMRT. Thirteen percent of the IMRT group had grade-3 or -4 hearing loss, compared to 64% of the conventional radiotherapy group (p<0.014).
Milker-Zabel and colleagues (2007) reported the results of the treatment of complex-shaped meningiomas of the skull base with IMRT in 94 patients. (8) Patients received radiotherapy as primary treatment (n=26) postoperatively for residual disease (n=14) or after local recurrence (n=54). Tumor histology was World Health Organization grade 1 in 54.3%, grade 2 in 9.6%, and grade 3 in 4.2%. Median follow-up was 4.4%. Overall local tumor control was 93.6%. Sixty-nine patients had stable disease (by computed tomography [CT]/magnetic resonance imaging [MRI]), and 19 had a tumor volume reduction after IMRT. Six patients had local tumor progression on MRI a median of 22.3 months after IMRT. In 39.8% of patients, preexisting neurologic deficits improved. Treatment-induced loss of vision was seen in 1 of 53 re-irradiated patients with a grade-3 meningioma 9 months after retreatment with IMRT.
Mackley and colleagues (2007) reported outcomes of treating pituitary adenomas with IMRT. (9) A retrospective chart review was conducted on 34 patients treated between 1998 and 2003 at the Cleveland Clinic. Median follow-up was 42.5 months. Radiographic local control was 89%, and among patients with secretory tumors, 100% had a biochemical response. One patient required salvage surgery for progressive disease, resulting in a clinical PFS of 97%. One patient who received more than 46 Gy experienced optic neuropathy 8 months after radiation.
Sajja and colleagues (2005) reported the outcomes of 35 patients with 37 meningiomas treated with IMRT. (10) Tumor histology was benign in 35 and atypical in 2 tumors. The median CT/MRI follow-up was 19.1 months (range 6.4-62.4 months). Fifty-four percent of the meningiomas had been previously treated with surgery/radiosurgery prior to IMRT, and 46% were treated with IMRT, primarily after a diagnosis was established by CT/MRI. Three patients had local failure after treatment. No long-term complications from IMRT were documented among the 35 patients.
Uy and colleagues (2002) assessed the safety and efficacy of IMRT in the treatment of intracranial meningioma in 40 patients treated between 1994 and 1999. (11) Twenty-five patients received IMRT after surgery either as adjuvant therapy for incomplete resection or for recurrence, and 15 patients received definitive IMRT after a presumptive diagnosis of meningioma on imaging. Thirty-two patients had skull base lesions and 8 had nonskull base lesions. Follow-up ranged from 6 to 71 months (median 30 months). Defined normal structures generally received a significantly lower dose than the target. The most common acute CNS toxicity was mild headache, usually relieved with steroids. One patient experienced Radiation Therapy Oncology Group (RTOG) Grade-3 acute CNS toxicity, and 2 experienced Grade 3 or higher late CNS toxicity, with one possible treatment-related death. No toxicity was observed with mean doses to the optic nerve/chiasm up to 47 Gy and maximum doses up to 55 Gy. Cumulative 5-year local control, PFS, and OS were 93%, 88%, and 89%, respectively.
Edwards and colleagues (2010) reported outcomes on the use of whole brain radiotherapy (WBRT) with an IMRT boost in 11 patients with metastatic disease to the brain ranging from 25-80 mm in maximum diameter. (3) Patients were excluded if they had more than 4 metastases. Histologies of the metastases included primary lung (n=5), breast (n=4), colon (n=1), and kidney (n=1). There were no acute or subacute complications. All tumors showed response on a 1-month post-radiotherapy scan. Median follow-up was 4 months. Four of the 11 patients died of systemic disease 6-9 months after radiotherapy. The remaining patients were alive with no evidence of progression of the treated brain disease or local recurrence at 2-9 months after radiotherapy. No brain complications occurred to date.
Physician Specialty Society and Academic Medical Center Input
In response to requests, input was received related to the use of IMRT to treat CNS tumors from 3 academic medical centers and 3 specialty medical societies (8 reviewers), for a total of 11 reviewers. While the various physician specialty societies and academic medical centers may collaborate with and make recommendations during this process, through the provision of appropriate reviewers, input received does not represent an endorsement or position statement by the physician specialty societies or academic medical centers, unless otherwise noted.
There was near uniform consensus that IMRT to treat tumors of the CNS should be considered medically necessary, particularly tumors in close proximity to critical structures. Reviewers generally felt that there is sufficient evidence for IMRT being at least as effective as 3D-conformal radiation therapy and that given the possible adverse events that could result if nearby critical structures receive toxic radiation doses (e.g., blindness) that IMRT dosimetric improvements should be accepted as meaningful evidence for its benefit.
A search of online site Clinicaltrials.gov returned no Phase III trials comparing IMRT to other radiation modalities for the treatment of CNS tumors.
The body of evidence available to evaluate IMRT in the treatment of CNS tumors consists of dose planning studies and case series. The case series are limited by small numbers, heterogeneous patient populations, and different types of tumors. No randomized trials have been reported that compare results using IMRT to other conformal radiation therapy modalities, nor do any of the reported case series using IMRT include concurrently treated control groups.
In general, the limited evidence suggests that IMRT provides tumor control and survival outcomes comparable to existing radiotherapy techniques. The evidence from treatment planning studies has shown that the use of IMRT decreases radiation doses delivered to critical CNS structures (e.g., optic chiasm, brainstem) and normal tissue adjacent to the tumor. This potentially lowers the risk of adverse events (acute and late effects of radiation toxicity), although the clinical benefit of reducing the radiation dose to critical structures and surrounding normal tissue using IMRT is theoretical. Determination of whether adverse event rates are reduced with IMRT is further complicated by a lack of high-quality literature defining the adverse effects using 3D conformal radiation therapy for the CNS, the main comparator to IMRT. The single arm case series are of limited usefulness in determining the benefits of IMRT over other conformal radiation modalities.
Due to the limitations in this evidence, this policy underwent clinical vetting. There was near-uniform consensus that the use of IMRT in the CNS is at least as effective as 3D-conformal radiation therapy, and that given the possible adverse events that could result if nearby critical structures receive toxic radiation doses that IMRT dosimetric improvements should be accepted as meaningful evidence for its benefit. The results of the vetting, together with a strong indirect chain of evidence and the potential to reduce harms, led to the decision that IMRT may be considered medically necessary for the treatment of tumors of the central nervous system that are in close proximity to organs at risk.
Practice Guidelines and Position Statements
National Comprehensive Cancer Network Guidelines
The 2011 (v2.2011) National Comprehensive Cancer Network (NCCN) guidelines state that: when radiation is given to patients with low grade gliomas, it is administered with restricted margins. Every attempt should be made to decrease the radiation dose outside the target volume. This can be achieved with 3-dimensional planning or IMRT. (12)
NCCN guidelines do not address the use of IMRT in high-grade tumors or metastases of the CNS. (12)
Medicare National Coverage
There is no national coverage determination.