Malignant brain tumours (MBTs) comprise primary malignant tumours (PMTs) and metastatic tumours (METs). MBTs are responsible for 3 % of all cancer deaths worldwide and are the second most common cause of cancer death in young people. They carry a very dismal prognosis, typically representing a death sentence that may be deferred by up to merely 36 weeks inspite of treatment (Obwegeser et al. 1995; Eljamel 2004). PMTs represent 40 % of all primary brain tumours, and glioblastoma multiforme (GBM) is the commonest type (Stupp 2007). Although progress has been made in the pathophysiology of GBMs, overall survival remains very poor and individual prediction of clinical outcome remains an elusive goal. Despite extensive clinical trials, the median survival of patients with GBM is still 12–14 months, with fewer than 26 % surviving for 2 years. In a series of 279 patients, only five (1.8 %) survived for 3 years (Lamborn et al. 2004). Multimodality therapeutic approaches would be necessary to improve this dismal outcome.

The joint tumour section of the American Association of Neurological Surgeons/Congress of Neurological Surgeons produced guidelines for the management of GBMs (Scott et al. 1998). The guidelines recommend maximum safe surgical resection, followed by 60 Gy of radiotherapy to the enhancing lesion, encompassing a 2-cm cuff around the lesion. The guidelines also recommend concurrent chemoradiation using Temozolomide and BCNU (Gliadel) in those who undergo craniotomy. Temozolomide offers a 2-year survival rate of merely 26 %, with median survival of 14.6 months and progression-free survival of 7.2 months (Olsen and Ryken 2008), while Gliadel offers survival of 13.9 months (Olsen and Ryken 2008). The poor outcome of these tumours is due to local invasion and local relapse. Eighty percent recur locally within 2 cm of the resection margin and patients often succumb to local recurrence, emphasising that a more aggressive local therapy is still awaited. However, complete radical surgical excision is hindered by the significant amount of tumour that is invisible to the surgeon even with the aid of the surgical microscope; indeed, its complete identification at surgery is an impossible task. Most of these tumours have invaded the brain widely by the time they manifest clinically, making a wider excision margin out of the question.

On the other side of the coin, METs are much more common (Vecht et al. 1990). They have an incidence of 12/100,000 per year and 15–40 % of systemic cancers metastasise to brain. The most frequent sources are lung tumours (40–50 %), breast cancers (15–25 %) and melanoma (5–20 %) (Vecht et al. 1990). The frequency of METs appears to be rising as a result of improved brain imaging and more effective treatment of primary cancers (Klos and O’Neill 2004; Chang et al. 2007). Common clinical features of METs are similar to those of PMTs, i.e. headache, neurological deficit and seizures.

If METs remain untreated, they are rapidly fatal. Surgery is indicated for solitary metastases in patients with a good performance status, and when the primary cancer is under control (Mehta et al. 2003). However, eradication of METs is not always possible owing to inability to resect the tumour en masse with a cuff of normal brain tissue, and adjuvant therapies are often used, e.g. whole brain radiotherapy, stereotactic radiosurgery, interstitial radiotherapy and chemotherapy.

Overall, therefore, aggressive local multimodality therapy holds the key in the successful treatment of MBTs, and the Photon Radiosurgery System (PRS) provides a means to deliver local interstitial radiotherapy at the same time as surgery.