glioblastoma continues to be a challenging disease to treat, with most patients succumbing to their disease in the course of a few months to a few years. Since the late 1970s when adjuvant radiation was found to improve survival, very little progress has been made in further improving the survival. In 2005, the EORTC and NCIC conducted a phase III trial, investigating temozolomide during and after radiation [2
]; this regimen yielded a modest but significant improvement in survival for this aggressive tumor, a small impact made almost after 25 years since the report by Walker et al established that 60 Gy as the most effective dose for glioblastoma from a pooled retrospective analysis [5
The inability to control the tumor locally, with distant brain metastases being far less common than local recurrence, led to the hypothesis that by increasing the radiation dose, local control and survival may improve. SRS is one of the methods of dose escalation which had been hypothesized to improve outcome [33
]. The RTOG 93-05 study randomized >200 patients with glioblastoma to SRS boost in addition to 60 Gy fractionated radiation and BCNU chemotherapy; no survival benefit was appreciated with the addition of an SRS boost [16
]. As size was a criterion for eligibility for SRS boost, a subsequent analysis of RTOG data found even by RPA stratification, SRS offers no significant added advantage [35
]. In an attempt to further clarify the role of SRS in malignant glioma, Tsao M et al. reviewed the literature and concluded that there is level I-III evidence that the addition of an upfront SRS boost to fractionated radiation and BCNU offers no benefit to survival, local control or quality of life in malignant glioma, with a greater risk of toxicity [36
]. Similar conclusions were reached in an evidenced based review in this journal, though the authors conclude that "selected patients may benefit but the specific characteristics of this group have yet to be identified" [1
Because a high percentage of patients experience local recurrence following standard therapy, salvage therapy is often challenging. Resection, chemotherapy or radiation are the various options which play an important role in the management of recurrent disease [37
]. Retrospective data, including matched pair analyses suggest that salvage therapy prolongs progression free survival and overall survival, albeit with a selection bias favoring those patients undergoing salvage [39
]. Since these tumors are quite infiltrative, a second surgery is often not feasible. Fractionated re-irradiation can be quite risky because a larger volume of previously treated brain is enclosed within the radiated volume. SRS, by virtue of rigid immobilization, allows for minimal radiation dose exposure to surrounding tissue. SRS is therefore well suited for patients undergoing a second course of radiation, as the normally accepted dose tolerances of normal structures would otherwise likely need to be exceeded to ensure adequate target coverage. In the RTOG 93-05 study, 19% of patients randomized to the arm without upfront SRS received SRS as salvage, 6% received fractionated radiation and 35% underwent salvage surgery. In contrast, 6% of patients randomized to the arm with upfront SRS received SRS as salvage, 7% received fractionated radiation and 33% underwent resection. Arguably, one can conclude that the RTOG 93-05 study does not necessarily show that SRS is not beneficial in patients with glioblastoma, but rather SRS can be delayed as part of salvage therapy without a detriment in survival.
The data on salvage SRS for recurrent glioblastoma is sparse and mostly retrospective making it difficult to interpret. A few retrospective studies have examined SRS with recurrent glioblastoma; these are summarized in the Table . The median survival after SRS is on the order of 8–12 months. Late toxicity other than radiation necrosis is uncommonly reported. Among those patients who undergo a neurosurgical procedure for progression and/or suspected radionecrosis, necrosis is admixed with viable tumor cells in nearly all samples. Generally, radionecrosis is reported in roughly 5–10% of cases,36,37 though most patients develop radiographic evidence of necrosis [40
]. The number of patients who are found to have pathologic necrosis will obviously be impacted by the aggressiveness of the neurosurgeon and the amenability of the tumor and patient to surgical exploration and debulking. As a consequence, the extent of radionecrosis after SRS is difficult to accurately quantify and characterize, but certainly it is not prohibitive to treatment, particularly since tumor progression is the primary means of death and diminished quality of life in most patients.
Summary or studies on the use of salvage stereotactic radiosurgery for recurrent glioblastoma
In our study, the median survival following SRS was 6.7 months which is comparable with the most reported series. The majority of the patients who received SRS as an upfront boost were treated prior to the reported results of RTOG 95-03. The median survival from the time of SRS for the up-front SRS was slightly better (10.3 months) compared to the patients who had SRS at recurrence (5.3 months). However, the median interval between the diagnosis and SRS was 1.3 months versus 12.1 months for patients who had consolidative SRS and SRS at recurrence, respectively. When we looked at the overall survival from the time of diagnosis, there is no difference in these two groups. Interestingly, we found significant improvement in progression-free survival when SRS was added as a consolidative treatment. We did not analyze the quality of life measure to see if delay in progression free survival is associated with any improvement in quality of life.
We have two long term survivors over 6 years in each group indicating it is not only the treatment but also the tumor biology which is probably crucial. As we develop greater understanding of tumor biology, we might be able to identify a subset of patients who would require more aggressive therapy compared to those who will do better without any aggressive therapy [41
]. A recent article by Krex et al. suggested that patients who have hypermethylated MGMT protein are the long-term survivors [42
]. Hegi et al. reported from the EORTC randomized trial that those patients who had methylated MGMT gene benefited significantly better from temozolomide than the patients who did not [43
]. Most of the SRS literature pre-dates the temozolomide era. Currently, whether more aggressive initial local therapy, including SRS, will have more benefit among patients receiving temozolomide, with or without the methylated MGMT gene, is unknown.
In our series, there was no RTOG >grade 2 acute toxicity seen. One patient developed grade 4 late toxicity and underwent craniotomy for this. The pathology largely showed necrosis (80%) with some viable tissue including both fibrosis and tumor cells and minimal Ki-67 activity. Among our 4 long-term survivors who lived over 4 years since the diagnosis of their glioblastoma, three are still alive. One patient died presumably from treatment related toxicity. However, he received aggressive surgery, multiple courses of radiation and prolonged chemotherapy and all these factors might have contributed to the toxicity rather than just radiation. The other patient who is still alive after 6 years has grade 2 late toxicity.
In conclusion, SRS is generally a well tolerated treatment both as a boost and as a salvage therapy. As there was no difference in survival between the two groups, the decision of adding SRS to fractionated treatment should be based on individual patient status and preference. In the temozolomide era, the role of SRS needs to be better defined by future studies.