doi: /j.ijrobp CLINICAL INVESTIGATION

Transcription

1 doi: /j.ijrobp Int. J. Radiation Oncology Biol. Phys., Vol. 67, No. 1, pp , 2007 Copyright 2007 Elsevier Inc. Printed in the USA. All rights reserved /07/$ see front matter CLINICAL INVESTIGATION Brain EFFECT OF PROPHYLACTIC HYPERBARIC OXYGEN TREATMENT FOR RADIATION-INDUCED BRAIN INJURY AFTER STEREOTACTIC RADIOSURGERY OF BRAIN METASTASES TAKAYUKI OHGURI, M.D.,* HAJIME IMADA, M.D.,* KIYOTAKA KOHSHI, M.D., SHINGO KAKEDA, M.D.,* NORIHIRO OHNARI, M.D.,* TOMOAKI MORIOKA, M.D.,* KEITA NAKANO, M.D.,* NOBUHIDE KONDA, M.D., AND YUKUNORI KOROGI, M.D.* Departments of *Radiology, Hyperbaric Medicine, and Neurosurgery, University of Occupational and Environmental Health, Kitakyushu, Japan Purpose: The purpose of the present study was to evaluate the prophylactic effect of hyperbaric oxygen (HBO) therapy for radiation-induced brain injury in patients with brain metastasis treated with stereotactic radiosurgery (SRS). Methods and Materials: The data of 78 patients presenting with 101 brain metastases treated with SRS between October 1994 and September 2003 were retrospectively analyzed. A total of 32 patients with 47 brain metastases were treated with prophylactic HBO (HBO group), which included all 21 patients who underwent subsequent or prior radiotherapy and 11 patients with common predictors of longer survival, such as inactive extracranial tumors and younger age. The other 46 patients with 54 brain metastases did not undergo HBO (non-hbo group). The radiation-induced brain injuries were divided into two categories, white matter injury (WMI) and radiation necrosis (RN), on the basis of imaging findings. Results: The radiation-induced brain injury occurred in 5 lesions (11%) in the HBO group (2 WMIs and 3 RNs) and in 11 (20%) in the non-hbo group (9 WMIs and 2 RNs). The WMI was less frequent for the HBO group than for the non-hbo group (p 0.05), although multivariate analysis by logistic regression showed that WMI was not significantly correlated with HBO (p 0.07). The 1-year actuarial probability of WMI was significantly better for the HBO group (2%) than for the non-hbo group (36%) (p < 0.05). Conclusions: The present study showed a potential value of prophylactic HBO for the radiation-induced WMIs, which justifies further evaluation to confirm its definite benefit Elsevier Inc. Hyperbaric oxygen therapy, Stereotactic radiosurgery, Radiation injury, Brain metastasis, White matter injury. INTRODUCTION Stereotactic radiosurgery (SRS) has become an important therapeutic approach for the treatment of vascular malformations, benign and malignant tumors, because it allows a dose increase in the target volume and sparing of normal tissue. This treatment technique is deemed to improve patient quality of life and to control the disease (1 4). However, high radiation doses bear an increased risk of radiation injury to normal brain tissue. Immediate toxicity after SRS has only rarely been reported, with seizure being the most common complication, and has in general been considered uncommon (5). One of the complications after SRS is radiation necrosis (RN) of the brain as delayed toxicity, which typically occurs 3 months or more after treatment. It was estimated that if all the cases with brain metastasis Reprint requests to: Takayuki Ohguri, M.D., Department of Radiology, University of Occupational and Environmental Health, 1-1 Iseigaoka, Yahatanishi-ku, Kitakyushu , Japan. Tel: ( 81) ; Fax: ( 81) ; ogurieye@ med.uoeh-u.ac.jp Presented at the 38th Annual Scientific Meeting of the Undersea and Hyperbaric Medical Society, Orlando, Florida, June 2006, and treated by SRS could be followed up over 18 months, RN would arise in 10% of these cases (6). Necrosis developing from radiation exposure presents as a coagulative process predominantly affecting the white matter. The events leading up to RN are considered to center around small vessel injury, ultimately resulting in vascular hyalinization and thrombus deposition leading to vascular occlusion (7, 8). RN typically develops as a focal process at or near the site of the brain tumor (9). White matter injury (WMI), which is defined as peritumoral hyperintense areas on T2-weighted images of magnetic resonance imaging or radiation-induced brain edema, is also a complication of SRS, and it is of concern especially for arteriovenous malformation or benign brain tumors as well as brain metastasis (1, 2, 10). Classic treatments for such conditions include steroids, non- the 17th Annual Scientific Meeting of the Japanese Society for Therapeutic Radiation and Oncology, Chiba, Japan, November Conflict-of-interest: none. Received Feb 26, 2006, and in revised form Aug 1, Accepted for publication Aug 4,

2 Prophylactic hyperbaric oxygen for radiation brain injury T. OHGURI et al. 249 steroidal anti-inflammatory agents, anticoagulants, and surgical resections. Hyperbaric oxygen (HBO) therapy has been used to assist in the repair of radiation-induced damage (11 15). Besides improving temporarily the oxygenation of tissue, it is thought that high oxygen tension promotes neovascularization in damaged tissues of radiation-treated patients (11). Studies have shown that HBO therapy effectively treats irradiated soft tissue necrosis and has also been used empirically to treat mandibular osteoradionecrosis, radiation cystitis, radiation proctitis, and other radiation side effects (12 15). Some studies have reported the effects of HBO therapy for the RN (16 19). Tandon et al. treated a patient with severe RN, managing its progression by steroid and anticoagulant treatments combined with HBO therapy; this case had a gradual clinical improvement after the initiation of HBO therapy (17). In another case report, a patient treated with repeated SRS had RN; the patient clinically failed steroid treatment alone but showed an improvement after HBO therapy both clinically and radiologically (19). In addition, Feldmeier et al. demonstrated in an animal study a delay in the onset of myelitis with HBO applied as prophylaxis after irradiation (20). To our knowledge, however, there are no clinical reports of prophylactic HBO for radiation-induced brain injury. The purpose of our study was to assess the prophylactic effect of HBO for radiation-induced brain injury in patients with brain metastases treated with SRS. METHODS AND MATERIALS Between October 1994 and September 2003, 78 patients presenting with 101 brain metastases were treated with radiosurgery in our hospital. There were 52 males and 26 females. Median age was 53 years (range, years). All patients had pathologically confirmed extracranial malignancies. The characteristics and treatments of the patients are given in Table 1. A total of 32 patients with 47 brain metastases were treated with prophylactic HBO (HBO group). Another 46 patients with 54 brain metastases did not undergo HBO (non-hbo group). Regarding treatment results of SRS for brain metastasis, inactive extracranial tumors, younger age, and higher Karnofsky performance status are considered the main significant factors associated with longer survival (21, 22). Therefore, we made positive use of prophylactic HBO treatment for the 11 patients with the predictors of longer survival, such as inactive extracranial tumors and younger age. Also, as a candidate of prophylactic HBO, we chose the patients who underwent conventional radiotherapy; all the 21 patients with subsequent or prior radiotherapy received prophylactic HBO to prevent radiation-induced brain injury. A total of 18 patients received SRS as a boost (median marginal dose, Gy), immediately preceding or after whole or limited volume brain fractionated radiotherapy (median dose, Gy; fractions of 2 3 Gy daily, given 5 days/week) within 6 weeks of one another. Three patients had undergone prior whole and/or limited volume brain fractionated radiotherapy (total dose, Gy; fractions of 2 3 Gy daily, given 5 days/week), and were treated at recurrence; intervals from prior radiation to recurrence were 4 16 months. Table 1. Summary of patient parameters between HBO group and non-hbo group HBO Non-HBO p value Patient age Gender: Male/female 22/10 30/ KPS: 80/ 70 14/18 16/ Tumor volume (cc) Primary controlled and 16/16 12/ metastases-brain only: Yes/No RPA*: Class I/II/III 6/12/14 4/22/ Hypertension: Yes/No 2/30 10/ Diabetes: Yes/No 1/31 5/ Marginal dose (Gy) Maximal dose (Gy) MX/MR : 1.5/ 1.5 6/41 10/ Lesion(s) treated: 1/2/ 3 24/5/3 39/6/ Prior radiotherapy: Yes/No 3/29 0/ Subsequent radiotherapy: 18/14 0/ Yes/No Histology 0.61 Non small-cell lung cancer Small-cell lung cancer 0 4 Renal cell cancer 1 1 Melanoma 1 1 Breast cancer 2 2 Others 8 8 Lesion location 0.16 Frontal lobe Temporal lobe 10 3 Parietal lobe 10 9 Occipital lobe 5 9 Cerebellar 7 10 Brainstem 0 2 Abbreviations: HBO hyperbaric oxygen; KPS Karnofsky performance status; RPA recursive partitioning analysis; MR marginal dose; MX maximal dose within treatment volume. * Recursive partitioning analysis (RPA) classes defined by Gaspar et al. (38). Maximal dose within treatment volume (MX), Marginal dose (MR), Ratio of MX/MR (measure of dose heterogenity within target volume). Stereotactic radiosurgery was performed using a 6-MV linear accelerator (Mevatron; Siemens, Munich, Germany); a F.L. Fischer system was used for head fixation during computed tomography planning and treatment. Treatment planning was carried out using a dedicated computed tomography scanner with 2-mm axial slices and intravenously administered contrast medium for optimal tumor visualization. The median number of rotational arcs used during treatment was 5 (range, 3 6); the range of the circular collimator size used for treatment was 5 20 mm. The radiation dose and dose heterogeneity within target volume are listed in Table 1. All patients were treated using a single isocenter, except one in whom two isocenters were selected to conform to a large, irregularly shaped lesion. Corticosteroid medications were prescribed in 30 of 32 patients (94%) in the HBO group and 45 of 47 (96%) in the non-hbo group, before, during, and after SRS as clinically indicated, and they were tapered or discontinued after SRS. Exposure to HBO was initiated less than 1 week after SRS; it was administered in hyperbaric chambers according to the follow-

3 250 I. J. Radiation Oncology Biology Physics Volume 67, Number 1, 2007 Fig. 1. (a) A 66-year-old patient with brain metastasis in the frontal robe (an example of white matter injury). Fluid-attenuated inversionrecovery axial magnetic resonance image before stereotactic radiosurgery shows small hyperintense signals around tumor. (b) Fluidattenuated inversion-recovery axial magnetic resonance image 2.3 months after stereotactic radiosurgery demonstrates enlargement of the hyperintense signals without tumor regrowth. Fig. 2. (a) A 69-year-old patient with brain metastasis in the cerebellum (an example of radiation necrosis). T1-weighted axial magnetic resonance image before stereotactic radiosurgery shows an enhancing lesion. (b) T1-weighted axial magnetic resonance image 10.3 months after stereotactic radiosurgery demonstrates an irregular, ring-enhancing lesion at irradiated area. Because the enhancing lesion was unchanged on further follow-up magnetic resonance examinations until 18.6 months after stereotactic radiosurgery, accompanied by neurologic improvement during the follow-up period, we considered that this lesion was radiation necrosis. ing schedule: 15 min of compression with air, 60 min of 100% oxygen inhalation using an oxygen mask at 2.5 atmospheres absolute, and 10 min of decompression with oxygen inhalation. A total of 20 sessions of HBO, 5 times per week, were performed. The endpoint of radiation-induced brain injury was analyzed with follow-up imaging, either magnetic resonance imaging or computed tomography, as well as neurologic examinations in all 78 patients. Magnetic resonance imaging was used in 61 patients (78%), and computed tomography in 21 patients (27%); 4 patients underwent both magnetic resonance imaging and computed tomography. The mean follow-up intervals for images were 2.5 months within 6 months after SRS, and 3.8 months thereafter; the mean of maximal follow-up periods was 11.8 months. Magnetic resonance images were obtained by using 1.5-T units at our institution or at vicinal hospitals, consisting of precontrast and postcontrast T1-weighted spin-echo imaging, T2-weighted fast spinecho imaging, and fluid-attenuated inversion-recovery imaging. The tumor size was determined by the enhanced area on contrast enhancement T1-weighted images or computed tomographic images in comparison with nonenhanced images. All imaging studies were evaluated by two neuroradiologists in consensus, without knowledge of the exposure to HBO. The radiation-induced brain injuries were divided into two categories, WMI and RN, on the basis of imaging findings (Figs. 1 and 2,ID 2). The WMI was defined as the detection of hyperintense signals on T2-weighted or fluid-attenuated inversion-recovery images or low attenuation area on computed tomographic images without tumor regrowth. To quantify the WMI further, a descriptive method was chosen to categorize the tissue changes according to their extent, small perifocal, intermediate perifocal, and large lesions. Because all the patients with tumor regrowth underwent neither surgical resection nor stereotactic biopsy, we determined clinicoradiologically whether the lesions were RN or recurrence, considering temporal characteristics of the lesions with sufficient follow-up period, which was more than 7 months. Lesions were regarded as recurrence if the enhancing lesions increased in size after SRS with continued progression on serial image examinations, and the patient s clinical condition deteriorated progressively during that period. Lesions were classified as RN if the enhancing lesions, which once increased in size after SRS, disappeared, were unchanged, or decreased in size on serial follow-up magnetic resonance or computed tomographic examinations, accompanied by neurologic improvement during the follow-up period. Fisher exact probability test was used for statistical analysis of the differences in gender, Karnofsky performance status, primary control, recursive partitioning analysis class, hypertension, diabetes, number of lesion treated, prior radiotherapy, subsequent radiotherapy, histology, lesion location, dose-diameter data, WMI, RN, and recurrence between the two groups. A Mann-Whitney U test was applied to compare patient age, tumor volume, marginal dose, and maximal dose between the two groups. To identify predictors for radiation-induced brain injury, univariate analysis by the Kaplan-Meier approach with log rank testing was performed using patient age, tumor volume, marginal dose, maximal dose, dose heterogeneity, and HBO. Multivariate analyses by logistic regression also were used to compare radiation-induced brain injury with patient age, tumor volume, marginal dose, maximal dose, dose heterogeneity, and HBO. RESULTS Mean follow-up periods were 13.7 months (range, months) for the HBO group and 10.5 months (range, months) for the non-hbo group. Incidence and time to development of WMI, RN, and recurrence in the HBO group and the non-hbo group are summarized in Table 2. The radiationinduced brain injury occurred in 5 lesions (11%) for the HBO group (2 WMIs and 3 RNs) and in 11 (20%) for the non-hbo group (9 WMIs and 2 RNs). The WMI was less frequent in the HBO group than in the non-hbo group (p 0.05). Tables 3 and 4 list the patients with radiation-induced brain injury, associated with the clinical characteristics and treatment parameters. As for extent of the WMI, although all cases of the HBO group showed small lesions, 6 (67%) of the 9 lesions of the non-hbo group were large. WMI associated with neurologic symptoms was not seen in the HBO group, but was recognized in 3 of 9 in the non-hbo group. Two of 3 RNs in the HBO group and 1 of 2 in the non-hbo group were accompanied by neurologic symp-

4 Prophylactic hyperbaric oxygen for radiation brain injury T. OHGURI et al. 251 Table 2. Incidence and time to development of WMI, RN, and recurrence between HBO group and non-hbo group Mean time to development (mo) HBO (n 47) (%) Non-HBO (n 54) (%) p value* HBO (range) Non-HBO (range) WMI 2 (4) 9 (17) ( ) 9.3 ( ) RN 3 (6) 2 (4) ( ) 3.7 ( ) Recurrence 6 (13) 5 (9) ( ) 11.1 ( ) Abbreviations: WMI white matter injury; RN radiation necrosis; HBO hyperbaric oxygen. * Fisher exact test. toms. All the patients with neurologic symptoms needed to increase or resume steroid use. Table 5 presents the results of multivariate analysis by logistic regression to evaluate effects of certain factors on WMI or WMI plus RN. None of the factors showed a statistically significant association, although trends toward significance were seen for HBO on WMI (p 0.07). Table 6 summarizes the univariate analysis by the Kaplan-Meier approach with log rank testing for evaluation of certain factors on WMI or WMI plus RN. The 1-year actuarial probability of WMI was significantly better for the HBO group (2%) than for the non-hbo group (36%) (p 0.05) (Fig. 3). The 1-year actuarial probability of WMI plus RN was also better for the HBO group (13%) than for the non-hbo group (40%), although the difference was not significant (p 0.07). The daily HBO treatment was well tolerated. Although a small number of patients experienced auditory problems ranging from hearing difficulties to ear pain during or shortly after HBO, serious or life-threatening complications with HBO were not observed. DISCUSSION Some authors have reported the usefulness of HBO therapy for the prevention of radiation-induced mandibular injuries; Marx et al. performed a randomized controlled trial showing the successful use of HBO in preventing mandibular radiation necrosis by giving HBO before and after dental extractions (23, 24). Although some studies have reported the effects of HBO therapy for RN (16 19), there have been no reports regarding HBO therapy for the prevention of delayed radiation injuries. The present study is the first investigation attempting to assess the prophylactic effect of HBO for radiation-induced brain injury in patients treated with SRS, which demonstrated that radiation-induced WMI was less frequent for the HBO group than for the non-hbo group, although multivariate analysis indicated that the WMI was not significantly correlated with HBO. Because of the slow progression of radiation-induced brain injury, the effectiveness of HBO treatment might be very much dependent on timing. Some animal studies were done to explore the temporal characteristics of radiosurgical lesions (7, 8). The lesions at 3.5 weeks demonstrated edema, demyelination, axonal loss, and neuronal death; in addition, coagulative necrosis without cavitation was also seen (7). Further necrosis associated with perilesional edema was observed at 6 weeks. Resorption of necrotic debris, as evidenced by progressive cavitation and phagocytic response, and increased perilesional vascularity occurred dur- Table 3. Clinical characteristics in 10 patients with 11 brain metastases that caused WMI Case Age Lesion location TV (cc) MD (Gy) Primary lesion Prior or subsequent RT Time* (mo) Extent of WMI Neurologic symptoms HBO group 1 47 Temporal lobe NSCLC None 16.3 Small None 2 66 Frontal lobe NSCLC None 2.3 Small None Non-HBO group 3 66 Frontal lobe NSCLC None 8.3 Large None 4 70 Temporal lobe NSCLC None 24.4 Large Quadriplegia 5 63 Parietal lobe NSCLC None 11.6 Large Defect in visual field, Weakness of a grip Frontal lobe NSCLC None 9.2 Large None 6 79 Frontal lobe NSCLC None 3.1 Large Quadriplegia 7 78 Cerebellum NSCLC None 7.2 Small None 8 58 Cerebellum Rectal ca. None 6.2 Small None 9 72 Frontal lobe Ovarian ca. None 10.1 Intermediate None Parietal lobe NSCLC None 3.6 Large None Abbreviations: WMI white matter injury; TV tumor volume; MD marginal dose; HBO hyperbaric oxygen; NSCLC non small-cell lung cancer; RT radiotherapy. * Time to development.

5 252 I. J. Radiation Oncology Biology Physics Volume 67, Number 1, 2007 Table 4. Clinical characteristics in 5 patients with brain metastases that caused RN Case Age Lesion location TV (cc) MD (Gy) Primary lesion Prior or subsequent RT (Gy) Time* (mo) Neurologic symptoms HBO group 1 71 Frontal lobe NSCLC 36 (subsequent) 10.7 Stagger, Weakness of legs 2 69 Cerebellum NSCLC None 10.3 Vomit 3 76 Frontal lobe NSCLC None 14.4 None Non-HBO group 4 70 Frontal lobe Rectal ca. None 2.8 Vomit 5 63 Frontal lobe Melanoma None 4.6 None Abbreviations: RN radiation necrosis; TV tumor volume; MD marginal dose; HBO hyperbaric oxygen; NSCLC non small cell lung cancer; RT radiotherapy. * Time to development. ing the intermediate stage (12 to 29 weeks) (7). The late phase was characterized by a glial scar without hypervascularity. In our study, exposure to HBO was initiated less than 1 week after SRS to prevent injury of normal brain and to promote neovascularization, because necrosis already was recognized at 3.5 weeks in the above experimental studies. We supposed that our results for the WMI were strongly associated with the timing of prophylactic HBO treatment to control the early necrotic change. Both endothelial cells and oligodendrocytes are considered to be the target cells in radiation-induced brain injury (25). After irradiation of the rat brain, it was demonstrated that the reduction of vascular endothelial cells occurred concomitantly with the start of regeneration characterized by the repopulation of endothelial cells and glial progenitor cells (25, 26). It seems that the duration of the latent period and severity of the tissue reaction are determined by the balance between the progression and recovery from radiation injury. Using the animal experiment, Feldmeier et al. investigated the prophylactic effect of HBO for radiation myelitis (20). All animals received identical spinal cord radiation dose of 69 Gy in 10 daily fractions. The prophylactic HBO consisted of 90 min oxygen at 2.4 atmospheres absolute for 20 daily treatments, initiated 6 weeks after completion of irradiation. In their results, the group with the prophylactic HBO exhibited significantly milder myelitis than the control group. Although our results also indicated a potential value of prophylactic HBO in WMI, further evaluations for HBO treatment protocol such as number of times, timing, and atmosphere, using clinical trials and experimental analysis, are needed to confirm its definite benefit. Flickinger et al. reported that, in 116 patients with brain metastases who underwent SRS, peritumoral hyperintense areas on T2-weighted magnetic resonance images, i.e., the WMI, along with neurologic symptoms developed in 4 patients (4%) 2 to 9 months after SRS (3). Mehta et al. also described delayed reappearance or progression of the WMI in 4 of 40 (10%) patients with brain metastases treated by SRS (27). WMI after SRS is recognized more frequently for arteriovenous malformations than for brain metastases; the incidence for arteriovenous malformations has been reported as 28% to 50% (1, 10). For benign glial neoplasms, the WMI as complications with peritumoral edema occurred in 4 of 13 (31%) patients treated by SRS (2). We believe that prophylactic use of HBO for prevention of the WMI might be more effective for arteriovenous malformations Table 5. Multivariate analyses by logistic regression to evaluate effects of certain factors on WMI or WMI plus RN in patients with brain metastasis WMI WMI plus RN Variable OR (95% CI) p value OR (95% CI) p value Patient age: 66/ ( ) ( ) 0.52 Tumor volume (cc): 5/ ( ) ( ) 0.47 Marginal dose (Gy): 20/ ( ) ( ) 0.86 Maximal dose (Gy): 25/ ( ) ( ) 0.84 MX/MR*: 1.5/ ( ) ( ) 0.91 HBO: No/Yes 4.34 ( ) ( ) 0.17 Abbreviations: WMI white matter injury; RN radiation necrosis; OR odds ratio; CI confidence interval; HBO hyperbaric oxygen; MX maximal dose within treatment volume; MR marginal dose. * Maximal dose within treatment volume (MX), Marginal dose (MR), Ratio of MX/MR (measure of dose heterogeneity within target volume).

6 Prophylactic hyperbaric oxygen for radiation brain injury T. OHGURI et al. 253 Table 6. Univariate analysis by the Kaplan-Meier approach with log rank testing for evaluation of certain factors on WMI or WMI plus RN WMI WMI plus RN Variable No. of brain metastases Probability (%) (1 year) p value Probability (%) (1 year) p value Patient age Tumor volume (cc) Marginal dose (Gy) Maximal dose (Gy) MX/MR* HBO Yes No Abbreviations: WMI white matter injury; RN radiation necrosis; HBO hyperbaric oxygen. * Maximal dose within treatment volume (MX), Marginal dose (MR), Ratio of MX/MR (measure of dose heterogeneity within target volume). and benign brain tumors than for brain metastases, because the WMI is more frequent in those benign lesions. In previous reports, marginal doses were usually set smaller for SRS with subsequent whole brain radiotherapy than for SRS alone to prevent the development of radiationinduced brain injury (28, 29). For example, in randomized clinical trial comparing SRS alone and whole brain radiotherapy plus SRS, the marginal dose was 30 Gy for SRS alone and 20 Gy for the whole brain radiotherapy plus SRS Fig. 3. Actuarial risk of developing white matter injury for the hyperbaric oxygen (HBO) group and the non-hbo group. The 1-year actuarial probability was 2% and 36% between the HBO group and the non-hbo group, respectively (p 0.02). *The number of lesions in patients who are alive at a given time point. **The number of lesions that are scanned for evaluation at a given time point. (29). In the present study, marginal doses of SRS were nearly equal between the groups with or without conventional radiation therapy. Because all cases with subsequent radiotherapy underwent HBO, and the WMI was less frequent in the HBO group, we believe that prophylactic HBO could prevent the radiation-induced WMI effectively. Radiation necrosis is also a serious complication after SRS and develops at the site of the initial tumor. It is usually difficult to distinguish RN with recurrence using conventional computed tomography and magnetic resonance imaging (30); in both pathologic abnormalities, contrast enhancements of the lesions are frequently seen, which only indicates breakdown of the blood brain barrier. More recently, usefulness of magnetic resonance spectroscopy, functional magnetic resonance imaging, or positron emission tomography for discrimination between RN and recurrence has been reported (31 33). In our study, the prophylactic use of HBO did not prevent the RN. Because our patients underwent neither surgical resection nor stereotactic biopsy, our diagnosis on recurrence or RN might be incorrect, and results for RN in our study may not allow final conclusions on the additional value of prophylactic HBO. Therefore, appropriately designed further studies using more convincing investigation to discriminate between RN and recurrence will be needed. A number of risk factors have been found to be associated with the development of radiation-induced brain injury after SRS. Nedzi et al. found the following factors to be associated with increased risk: tumor volume, the use of multiple isocenters, tumor homogeneity, and marginal and maximal

7 254 I. J. Radiation Oncology Biology Physics Volume 67, Number 1, 2007 doses (34). In this study, the maximal dose of SRS in the non-hbo group was higher than that in the HBO group, and the difference tended to be significant (p 0.06). Although the maximal dose did not yield a significant association on multivariate analysis, this might still be an important distinction between the two groups. Our non-hbo group was approximately 5 years older than the HBO group on average. The so-called diffuse white matter hyperintensity on magnetic resonance imaging, which may resemble the radiation-induced WMI, is more commonly seen in the elderly; one of the risk factors for the diffuse white matter hyperintensity is considered to be aging. The white matter may also be more vulnerable to radiation in the elderly than in younger patients. In some previous reports of conventional radiotherapy for adults and pediatric patients with brain tumors, higher age as well as total radiation dose was the significant predictor of the incidence of the WMI (35, 36). Constine et al. reported 56% of adults and 20% of children experienced severe white matter injury (36). Although the patient s age was not a predictor for the WMI in our multivariate analysis by logistic regression, the contribution of age, especially in the elderly, could not be neglected. According to Niibe et al., survival benefits of SRS for metastatic brain tumors were observed in patients with controlled primary lesions and no other distant metastases, whose 1-year and 3-year overall survival rates were 88.9% and 51.9%, respectively (37). Therefore, we applied prophylactic HBO to the patients with such predictors of longer survival. It may be necessary to assess the effect of prophylactic HBO on patients with limited prognoses. In summary, this is the first report attempting to assess the prophylactic effect of HBO for radiation-induced brain injury in patients treated with SRS. It demonstrated a potential value of prophylactic HBO in the WMI, and at least, the results justify further evaluation using both clinical and experimental analysis to confirm its definite benefit. REFERENCES 1. Flickinger JC, Lunsford LD, Kondziolka D, et al. Radiosurgery and brain tolerance: An analysis of neurodiagnostic imaging changes after gamma knife radiosurgery for arteriovenous malformations. Int J Radiat Oncol Biol Phys 1992;23: Grabb PA, Lunsford LD, Albright AL, et al. Stereotactic radiosurgery for glial neoplasms of childhood. Neurosurgery 1996;38: Flickinger JC, Kondziolka D, Lunsford LD, et al. A multiinstitutional experience with stereotactic radiosurgery for solitary brain metastasis. Int J Radiat Oncol Biol Phys 1994;28: DiBiase SJ, Chin LS, Ma L. Influence of gamma knife radiosurgery on the quality of life in patients with brain metastases. Am J Clin Oncol 2002;25: Chin LS, Lazio BE, Biggins T, et al. Acute complications following gamma knife radiosurgery are rare. Surg Neurol 2000;53: Loeffler JS, Alexander E. Radiosurgery for the treatment of intracranial metastases. In: Alexander E, Loeffler JS, Lunsford LD, editors. Stereotactic radiosurgery. New York: McGraw- Hill; p Blatt DR, Friedman WA, Bova FJ, et al. Temporal characteristics of radiosurgical lesions in an animal model. J Neurosurg 1994;80: Kondziolka D, Lunsford LD, Claassen D, et al. Radiobiology of radiosurgery: Part I. The normal rat brain model. Neurosurgery 1992;31: Fitzek MM, Thornton AF, Rabinov JD, et al. Accelerated fractionated proton/photon irradiation to 90 cobalt gray equivalent for glioblastoma multiforme: Results of a phase II prospective trial. J Neurosurg 1999;91: Guo WY, Nordell B, Karlsson B, et al. Target delineation in radiosurgery for cerebral arteriovenous malformations. Assessment of the value of stereotaxic MR imaging and MR angiography. Acta Radiol 1993;34: Marx RE, Ehler WJ, Tayapongsak P, et al. Relationship of oxygen dose to angiogenesis induction in irradiated tissue. Am J Surg 1990;160: Hart GB, Mainous EG. The treatment of radiation necrosis with hyperbaric oxygen (OHP). Cancer 1976;37: Maier A, Gaggl A, Klemen H, et al. Review of severe osteoradionecrosis treated by surgery alone or surgery with postoperative hyperbaric oxygenation. Br J Oral Maxillofac Surg 2000;38: Nakada T, Yamaguchi T, Sasagawa I, et al. Successful hyperbaric oxygenation for radiation cystitis due to excessive irradiation to uterus cancer. Eur Urol 1992;22: Bui QC, Lieber M, Withers HR, et al. The efficacy of hyperbaric oxygen therapy in the treatment of radiationinduced late side effects. Int J Radiat Oncol Biol Phys 2004;60: Chuba PJ, Aronin P, Bhambhani K. Hyperbaric oxygen therapy for radiation induced brain injury in children. Cancer 1997;80: Tandon N, Vollmer DG, New PZ, et al. Fulminant radiationinduced necrosis after stereotactic radiation therapy to the posterior fossa: Case report and review of the literature. J Neurosurg 2001;95: Leber KA, Eder HG, Kovac H, et al. Treatment of cerebral radionecrosis by hyperbaric oxygen therapy. Stereotact Funct Neurosurg 1998;70: Kohshi K, Imada H, Nomoto S, et al. Successful treatment of radiation-induced brain necrosis by hyperbaric oxygen therapy. J Neurol Sci 2003;209: Feldmeier JJ, Lange JD, Cox SD, et al. Hyperbaric oxygen as prophylaxis or treatment for radiation myelitis. Undersea Hyperb Med 1993;20: Alexander E, Moriarty TM, Davis RB, et al. Stereotactic radiosurgery for the definitive, noninvasive treatment of brain metastases. J Natl Cancer Inst 1995;87: Auchter RM, Lamond JP, Alexander E, et al. A multiinstitutional outcome and prognostic factor analysis of radiosurgery for resectable single brain metastasis. Int J Radiat Oncol Biol Phys 1996;35: Marx RE, Johnson RP, Kline SN. Prevention of osteoradionecrosis: A randomized prospective clinical trial of hyperbaric oxygen versus penicillin. J Am Dent Assoc 1985; 111: Feldmeier JJ, Hampson NB. A systematic review of the literature reporting the application of hyperbaric oxygen preven-

8 Prophylactic hyperbaric oxygen for radiation brain injury T. OHGURI et al. 255 tion and treatment of delayed radiation injuries: An evidence based approach. Undersea Hyperb Med 2002;29: Schultheiss TE, Kun LE, Ang KK, et al. Radiation response of the central nervous system. Int J Radiat Oncol Biol Phys 1995;31: Sminia P, van der Kleij AJ, Carl UM, et al. Prophylactic hyperbaric oxygen treatment and rat spinal cord re-irradiation. Cancer Lett 2003;191: Mehta MP, Rozental JM, Levin AB, et al. Defining the role of radiosurgery in the management of brain metastases. Int J Radiat Oncol Biol Phys 1992;24: Kondziolka D, Patel A, Lunsford LD, et al. Stereotactic radiosurgery plus whole brain radiotherapy versus radiotherapy alone for patients with multiple brain metastases. Int J Radiat Oncol Biol Phys 1999;45: Chougule PB, Burton-Williams M, Saris S, et al. Randomized treatment of brain metastasis with gamma knife radiosurgery, whole brain radiotherapy or both [Abstract]. Int J Radiat Oncol Biol Phys 2000;48(Suppl.): Kumar AJ, Leeds NE, Fuller GN, et al. Malignant gliomas: MR imaging spectrum of radiation therapy- and chemotherapy-induced necrosis of the brain after treatment. Radiology 2000;21: Schlemmer HP, Bachert P, Herfarth KK, et al. Proton MR spectroscopic evaluation of suspicious brain lesions after stereotactic radiotherapy. Am J Neuroradiol 2001;22: Asao C, Korogi Y, Kitajima M, et al. Diffusion-weighted imaging of radiation-induced brain injury for differentiation from tumor recurrence. Am J Neuroradiol 2005;26: Chao ST, Suh JH, Raja S, et al. The sensitivity and specificity of FDG PET in distinguishing recurrent brain tumor from radionecrosis in patients treated with stereotactic radiosurgery. Int J Cancer 2001;96: Nedzi LA, Kooy H, Alexander E 3rd, et al. Variables associated with the development of complications from radiosurgery of intracranial tumors. Int J Radiat Oncol Biol Phys 1991;21: Corn BW, Yousem DM, Scott CB, et al. White matter changes are correlated significantly with radiation dose: Observations from a randomized dose-escalation trial for malignant glioma (Radiation Therapy Oncology Group 83-02). Cancer 1994;74: Constine LS, Konski A, Ekholm S, et al. Adverse effects of brain irradiation correlated with MR and CT imaging. Int J Radiat Oncol Biol Phys 1988;15: Niibe Y, Karasawa K, Nakamura O, et al. Survival benefit of stereotactic radiosurgery for metastatic brain tumors in patients with controlled primary lesions and no other distant metastases. Anticancer Res 2003;23: Gaspar L, Scott C, Rotman M, et al. Recursive partitioning analysis (RPA) of prognostic factors in three Radiation Therapy Oncology Group (RTOG) brain metastases trials. Int J Radiat Oncol Biol Phys 1997;37: