Hypo-IMRT using HT is an approach that combines high-precision RT delivery and a hypofractionated regimen. Patients with GBM have a dismal prognosis and a limited life span. The period of best performance is even shorter as clinical deterioration is associated with profound morbidity. Thus, achieving similar clinical outcome while abbreviating treatment course can be of great clinical significance. An additional advantage to hypofractionation is the potential of improving tumour control. Our study is the first to prospectively examine the use of Hypo-IMRT while administering concurrent and adjuvant TMZ for newly diagnosed GBM patients. The efficacy and safety of this regimen were demonstrated by this phase I study. Our data are consistent with the results reported by Stupp et al.  with a median OS of 15.67 months a median PFS of 6.7 months. The study’s inability to show improved outcome, over standard fractionation, could be due to our reluctance to proceed to higher dose levels or merely a reflection of the small sample size tested.
The risk of neurological side effects, particularly radionecrosis of the brain is considered to be the main deterrent of using a hypofractionated scheme. The threshold fraction size above which this risk is clinically significant is difficult to determine. One reason is the considerable variation in the fractionation regimens used  and the major difference in the prognostic characteristics of the patients in whom the regimen was studied. In addition, most of the dose escalation literature is based on conventional methods of radiation delivery. The use of IMRT, HT in particular, is thought to physically allow dose escalation while keeping reasonable dose constraints to at risk normal tissue. Two main studies have prospectively investigated the use of IMRT to deliver a hypofractionated radiation therapy to patients with GBM and each used a different regimen [2, 3]. Floyd et al. , used a dose of 50 Gy at 5 Gy per fraction given to the enhancing primary tumour, residual disease or surgical cavity with a simultaneous dose of 30 Gy at 3 Gy per fraction to the surrounding oedema. Twenty percent of the patients evaluated for late toxicity experienced Grade 4 cerebral necrosis. The latter study by Sultamen et al.  prescribed 60 Gy in 20 fractions to the GTV and 40 Gy in 20 fractions to the PTV in their study. One patient developed blindness 9 months after the completion of radiation treatment but the aetiology of the visual loss was not felt to be attributable to RT. There is a wide range between the fractions sizes used in these studies. Currently, the wide recognition of concurrent and adjuvant TMZ as the standard of care for GBM, diminishes somewhat the applicability of these previous studies to current treatment considerations. The concern of an increased risk of cerebral radionecrosis with the addition of chemotherapy is a valid one based on previous reports . Panet-Raymond et al.  did a retrospective analysis on 35 patients who were treated with post-operative Hypo-IMRT and TMZ according to the Stupp protocol. In this study a total dose of 60 Gy in 20 daily 3-Gy fractions was delivered to the GTV while ensuring that the 95–100% isodose line covered the GTV, and the 65–70% line encompassed the PTV. The median overall survival was 14.4 months, and the median disease-free survival was 7.7 months, both of which were comparable to those reported by the EORTC/NCIC trial . No late toxicity was seen in that cohort of patients but considering their method of prescribing the dose, the PTV could have received a dose per fraction as low as 2 Gy which is not different from conventional fractionation. In our study, we used a dose escalation protocol to establish a safe and tolerable fractionation regimen and no late toxicity was reported with either dose levels. Of interest, a recently published phase I trial of Hypo-IMRT with TMZ used a dose escalation protocol . The study included 16 patients and examined four dose levels. A total dose of 60 Gy in 3 Gy/fraction, 4 Gy/fraction, 5 Gy/fraction, and 6 Gy/fraction, was prescribed in dose levels 1, 2, 3 and 4, respectively. One patient who was treated at level 2 with 60 Gy in 4 Gy/fraction developed Grade 4 visual loss at 7 months following RT. Three patients developed pathologically confirmed extensive necrosis. One was treated at level 1, one at level 2 and the third at level 4. Therefore, most of the up-to-date hypofractionation studies suggest that fraction sizes larger than 3 Gy could be highly associated with detrimental late effects.
With regard to acute toxicity, two patients in our study developed Grade 3–4 myelosuppression which is expected given the toxicity profile reported in the EORTC/NCIC trial . One patient developed severe Pneumocystis jiroveci pneumonia infection requiring ICU admission. This has been previously reported in patients with brain tumours  and because of which prophylaxis against this opportunistic infection was mandatory for all patients receiving TMZ in the EORTC/NCIC trial. All of our patients completed concurrent chemoradiation but only 76% of the patients received adjuvant TMZ. The median number of adjuvant TMZ cycles was 2 and in the EORTC/NCIC trial, the median number was 3. However, due to the small sample size of our study, it is difficult to conclude whether the intensified radiation dose could have reduced the patients’ tolerance to adjuvant TMZ. Two patients had persistent thrombocytopenia precluding the use of adjuvant TMZ. The impact of thrombocytopenia was examined in a retrospective analysis of 52 consecutive patients with newly diagnosed high-grade gliomas treated with the Stupp protocol . The rate of Grade 3–4 thrombocytopenia was found to be 19% with a significant risk of prolonged, possibly irreversible, low platelet count. On multivariate analysis, we found that the administration of 3 or more cycles of adjuvant TMZ was associated with better OS and PFS. Our current guidelines suggest 6 cycles of adjuvant TMZ as per EORTC/NCIC trial . However, the optimal duration of this adjuvant therapy, beyond six months, is currently being prospectively investigated.
Despite the use of conformal RT, disease recurrence occurs in the treatment volume in the majority of patients . The rate of recurrence outside the tumour bed is low and mostly occurs with or after recurrence of the original disease . Brandes et al.  has recently reported the pattern of failure in 95 patients with newly diagnosed GBM treated with radiotherapy plus concomitant and adjuvant TMZ. A shift in the pattern of failure from local or marginal locations to distant locations was observed. Recurrences outside the treatment field were observed in 20% of the patients. Furthermore, there was a significant higher survival rate in patients with recurrence outside the RT fields (median survival of 26.1 months vs. 17.3 months). This change in the recurrence pattern was not mirrored in our study which could be due to the small sample size of our cohort. The same study by Brandes et al.  also reported strong correlation between the recurrence pattern and the MGMT methylation status. The rate of infield recurrence correlated significantly with MGMT methylation status. The rate was higher in patients with unmethylated MGMT versus those with methylated MGMT (85% vs. 57.9%). This factor could not be assessed in our study as MGMT methylation was unfortunately not tested in our patients.
The incidence of pseudoprogression has been variably reported in the literature. Brandes et al.  reported the incidence of pseudoprogression among 103 patients treated with TMZ concurrent with and adjuvant to RT. Out of 50 patients who developed radiological progression as assessed by MRI done 4 weeks after treatment completion, 32 (64%) were identified to have pseudoprogression. On the other hand, the data published by Sanghera et al.  on a series of 104 patients, only 7 patients (32%) were found to have pseudoprogression out of 22 classified to have early disease progression as indicated by an MRI 8 weeks post-RT. During the 3 month period after the concurrent treatment, radiation-induced brain injury could be associated with an increase of non-enhancing and enhancing tumour component on MRI. In our study, unless otherwise clinically indicated, an MRI was first routinely performed after three months of the end of RT. Accordingly, only nine cases were classified as early progression and all of which were true disease progression. This could reflect the fact that neuroradiological imaging was avoided during the early period following RT during which radiation induced CNS injury, manifested as an increase in contrast enhancement and/or non-enhancing tumour components, is at its peak.
The RPA class is one of the most significant prognostic factors that determine survival in patients with GBM. This has also been shown to hold true in the era of concurrent RT and TMZ (18). Unexpectedly, the RPA in our cohort of patients did not correlate with survival. The most likely explanation is that only patients with a KPS ≥70 were included in our study. This resulted in clustering of the majority of patients within one RPA class and due to our small sample size; the effect of the RPA class could not be statistically appreciated. Blumenthal et al. reported a large analysis on the effect of the time to initiate RT following surgery for GBM patients (19). Approximately, three thousand patients, from the Radiation Therapy Oncology Group (RTOG) database, were examined. The median survival was found to be significantly greater in patients in whom RT was started more than 4 weeks from surgery. All of the patients started RT within 6 weeks as mandated by all the study protocols according to these patients were treated. The authors of the study concluded that short delays in initiating RT, not exceeding 6 weeks, may not significantly affect survival but cautioned that the reported observation could reflect physician’s tendency to expedite treatment in patients with worse outlooks resulting in apparently worse outcome when RT was started earlier. In our study, delays greater than 6 weeks in initiating RT adversely affected outcome. The influence of delaying RT could have been potentiated by the presence of considerable proportion (36%) of patients who only had a biopsy for a tumour notorious for its short doubling time.
In summary, the results of this phase I dose-escalation study have shown that Hypo-IMRT, using HT, with concurrent TMZ is safe and feasible. Our data are consistent with results reported EORTC/NCIC trial, with a median OS of 15.67 months a median PFS of 6.7 months. The Hypo-IMRT regimen is shorter than standard RT schedules. Abbreviating the RT regimen can be clinically meaningful considering the life expectancy of patients with GBM. It is hoped that these advanced technologies can provide better quality of life, more convenient treatment options, and optimal utilization of the resources by health care providers. Our findings warrant further validation of the results by conducting a phase II randomized controlled trial comparing our treatment regimen to conventional fractionation. Careful progression through higher dose levels, in the setting of a clinical trial seems possible.