The Safety and Efficacy of Bevacizumab for Radiosurgery - Induced Steroid - Resistant Brain Edema; Not the Last Part in the Ship of Theseus
Correspondence Address: Source of Support: None, Conflict of Interest: None DOI: 10.4103/0028-3886.271242
Source of Support: None, Conflict of Interest: None
Keywords: Arteriovenous malformation, Bevacizumab, edema, gamma knife, radiation, radiosurgery, tumor, meningioma, vestibular schwannoma, psychosurgeryKey Message: This article provides a preliminary proof that Bevacizumab is an effective and safe treatment option for radiosurgery induced steroid resistant brain edema.
Radiation-induced brain edema (RIBE) is a dreaded complication of cranial irradiation. Gamma knife radiosurgery (GKRS) is a suitable treatment option for several benign and malignant intracranial pathologies. The inherent properties of GKRS, i.e. conformity and precision, offer a very limited radiation exposure to the surrounding normal brain parenchyma. Still, a few patients develop significant RIBE and radiation necrosis (RN). High-dose corticosteroid is the salvage treatment in such a situation, however, long-term continuation of steroids carry a myriad of debilitating side-effects and some patients do not respond favorably in the first place.,,,,,,
The pathophysiology of RIBE is still ill defined. However, the most commonly proposed hypothesis points to a primary endothelial cell dysfunction with release of vascular endothelial growth factor (VEGF) through leaky capillaries. VEGF, also known as “vascular permeability factor” is a substance capable of blood–brain barrier (BBB) disruption and capillary endothelial damage in cranial vasculature.,,, In human cancers, increased expression of VEGF is associated with increased microvascular density, tumor growth, metastasis, and poor prognosis. The role of Bevacizumab is critical in neovascularization and angiogenesis. Bevacizumab is a humanized monoclonal antibody against VEGF. Bevacizumab binds to VEGF, VEGF1 and VEGF2 receptors present on the surface of endothelial cells. There is recent class 1 evidence for efficacy of Bevacizumab in the management of progressive RN through the blockage of VEGF.,,, In this article, we share our experience with Bevacizumab in the management of RIBE, which should be considered a separate clinical entity than RN.
A retrospective search of departmental records of patients in our institute who showed clinic-radiological attributes of RIBE managed with GKRS was performed. Inclusion criteria for treatment with Bevacizumab were “failure to respond” to 4 mg of dexamethasone four times a day for at least a three-week duration and were considered “steroid-resistant edema” [Figure 1]. The clinical, radiological, and pathological details of the patients were evaluated. The study was approved by the Institute Ethics committee (IEC). The present study was conducted according to the declaration of Helsinki and anonymity of the patients was maintained.
All patients received single fraction or fractionated GKRS for various intracranial pathologies [Figure 1]. Dose-fractionated GKRS was administered on a daily or alternate day schedule for consecutively for 2-3 days for large-volume lesions (such as arteriovenous malformation, glomus tumor, pituitary adenoma, and meningioma) which could not be treated in a single fraction [Table 1]. The diagnosis of RIBE was established on clinic-radiological evidence.,,, Patients who developed brain edema adjoining the irradiated target as visible on MRI scans were included. RIBE was considered with hyperintense signal on T2-weighted images and fluid-attenuated inversion recovery (FLAIR) sequences surrounding the lesion, which indicated vasogenic edema. A comprehensive follow-up MRI was performed including MR perfusion and MR spectroscopy. Only those cases were eventually included which fulfilled the criteria for RIBE and RN (peripheral ring enhancement, no hyperperfusion with insignificant elevation in choline) after exclusion of all cases with tumor recurrence [Figure 2].,,, Only patients with clinic-radiological evidence of RIBE and steroid-resistant edema were considered for Bevacizumab [Figure 1] and [Table 1]. The patients were informed about the experimental nature of the treatment in detail and consent was obtained after explaining the uncertain efficacy and chances of potential complications of the treatment. After establishing the clinico-radiological diagnosis of RIBE, the patients were thoroughly evaluated for hypertension, cardiomyopathy (echocardiogram), liver and renal function (complete test panel) and proteinuria. Hemograms and peripheral blood smears were conducted as well. At each follow-up, the patients were regularly evaluated on clinical and biochemical parameters (hematological and renal profile) and any change was regularly monitored.
Dosage criteria and treatment endpoints
All patients were initially started on 5 mg/kg of intravenous Bevacizumab followed by 7.5 mg/kg and subsequently 10 mg/kg on every two-week interval. Clinical response was evaluated at two, four and eight-week intervals. The patients were continued on Bevacizumab till the clinical response plateaued or any side affects appeared. This should be in the first paragraph of the Results section. Please move it there with the other patient details. The endpoints of the treatment were complete recovery, partial response with plateauing for the last three preceding cycles, or no response after six cycles [Figure 1]. There is no standard protocol till date for the use of Bevacizumab in RIBE and the pre-decided endpoints were arbitrarily chosen and approved by the IEC in view of the experimental nature of the study.
Patients with RIBE were evaluated with a multiparametric MRI brain scan including the following sequences: T1, T2, T1 contrast, FLAIR, diffusion-weighted (DW), susceptibility-weighted (SW), and perfusion-weighted (dynamic susceptibility contrast imaging) sequence, wherever indicated. The diagnostic MRI of RIBE was considered baseline investigation before starting Bevacizumab. RIBE was labeled in a setting of increasing peritumoral edema with no significant change in the lesion morphology. Care was taken to differentiate it from edema secondary to lesion growth by examination of MR parameters. The follow-up MRI was obtained on an outpatient basis as per the clinical symptomatology. The edema was evaluated and compared on T2-weighted and FLAIR images. Contrast-enhanced scans were evaluated to rule out any ring enhancing lesions. The diagnosis of possible RN/RIBE was established on the basis of previously published guidelines. The follow-up images were taken after third and sixth cycle of Bevacizumab. All patients were followed up with 3 months, 6 months, and yearly after MRI scans once Bevacizumab was stopped.
Every adverse event was evaluated as per the criteria given by the Common Terminology Criteria for Adverse Events (CTCAE). The symptomatology attributable to RIBE was recorded and patients were started on Bevacizumab as per their body weight. Patients were regularly followed for any change in their symptomatology or any adverse event during the treatment. Because of variable pathologies and multifactorial clinical symptomatology, specified clinical criteria for monitoring response could not be defined. The symptomatic response was considered as a decrease in severity of symptoms secondary to RIBE only. Drug-related side-effects such as hypertension, proteinuria, and delayed wound healing were monitored at every clinical visit.
The literature and PubMed was searched with radiation injury, radiation necrosis, edema, Bevacizumab, steroid, radiotherapy, and gamma knife radiosurgery as keywords. Only case series and review articles mentioning the role of Bevacizumab in the management of radiation injury were selected for comparative evaluation. After thorough search, 68 articles were found and 17 were selected. The rest were not selected due to their poorly defined study protocols.
Between 2012 and 2017, 1241 patients received GKRS for various intracranial pathologies. 189 out of the 1241 patients required a short course of steroids for the management of post-radiation brain edem. 17 out of 189 (8.9%) of patients (aged 19–61 years, median 42) received Bevacizumab for RIBE [Figure 1]. None of the patients received any adjuvant, neoadjuvant, or concurrent chemotherapy. Among these 17, eight were males [Table 1]. We identified 19 lesions in 15 patients with arteriovenous malformations (AVM; n = 6), meningioma (n = 9), vestibular schwannoma (n = 4), nonfunctional pituitary adenoma (n = 1), and cavernoma (n = 1). Two patients received ablative GKRS for brachial dystonia and obsessive compulsive disorder (OCD). Patients of brachial dystonia and OCD refused stimulation surgery in view of associated medical comorbidities and the high cost involved with the treatment. One patient had to be considered for Bevacizumab as he developed uncontrolled diabetes with high-dose corticosteroid. None of these patients received whole brain radiation (WBRT) either prior or subsequent to stereotactic radiosurgery (SRS). Two patients [Figure 3] and [Figure 4] received three-session fractionated GKRS for large-volume AVMs. The mean time from GKRS to diagnosis of RIBE was 4.8 months and from diagnosis to initiation of Bevacizumab was 5.6 (range 3–72) months. None of these patients had any histopathological proof of RIBE. All patients received 5 mg/kg of Bevacizumab in the first cycle. The second dose was increased to 7.5 mg/kg, and then the third cycle was 10 mg/kg every two weeks. The mean duration of Bevacizumab administration was seven cycles (range 2–27) before best response. The difference in dosing schedule represents the unpredictable time to response and the lack of standard treatment protocols.,
All patients received high-dose corticosteroids for control of RIBE-associated symptoms before starting treatment with Bevacizumab. The patients were simultaneously continued on Prednisolone and Bevacizumab. Steroid therapy was tapered in most patients by the first follow-up (3–10 weeks). Four (36.4%) patients were weaned off steroids completely after 8 weeks. There was an average reduction in the daily dexamethasone requirement of 2.4 mg. 12 out of 17 (70.6%) patients showed either stable or improved symptomatology with Bevacizumab. One patient needed surgical decompression after 14 cycles as edema was nonresponsive to Bevacizumab and the patient developed clinico-radiological deterioration [Figure 5]. One patient of tubercular sella meningioma showed remarkable visual recovery after the first cycle of Bevacizumab. Four out of six patients of AVM showed clinico-radiological improvement with Bevacizumab. All three patients of vestibular schwannoma showed recovery of facial nerve function with Bevacizumab; however, one patient did not gain any improvement in hearing deficit. One female patient with OCD [Table 1] stopped Bevacizumab after gaining symptomatic relief with two cycles but needed surgical decompression for mass effect and clinical deterioration after 3 months. One patient developed piles after the seventh cycle of Bevacizumab, which could not be explained with drug exposure. The patient temporarily discontinued it for 18 months. Another young patient (19-year-old female) suffered psychiatric complaints and needed antidepressants for a short duration. Contrary to the published literature, no patient developed hypertension or proteinuria.,
The pre-Bevacizumab MRI was obtained at a mean of 92 days (21–482 days) before starting treatment. The average tumor volume was 14.1 mL (range 0.8–39.34). The first posttreatment MRI was obtained at a mean of 68 days and subsequent MRI at intervals of 3–6 months. [Figure 3] and [Figure 4] show representative images obtained before and after Bevacizumab in patients with large-volume AVM treated with fractionated GKRS. Clinical improvement preceded radiological improvement by an average of 8 weeks. The majority of patients showed radiological improvement only after the third cycle of Bevacizumab and improvement plateaued after the eighth and ninth cycle [Table 1].
With the availability of sophisticated treatment modalities such as gamma knife, cyber knife, and intensity modulated radiotherapy (IMRT), a significant number of patients are now treated with upfront or salvage radiosurgery/radiotherapy. RN and RIBE are devastating side-effects of radiation treatment in already compromised brain parenchyma resulting in significant clinical and radiological changes. RN to the brain is a serious side-effect of brain tissue irradiation. It can be classified as acute (during radiation), early delayed (up to 12 weeks post irradiation), and late (months to years post-irradiation). In comparison to RN, RIBE has a latent interval of a few months only following GKRS. Usually the diagnosis of RIBE is clinico-radiological with no definite histopathological evidence. Common modes of presentation are worsening or change in character of headache, seizure, psychomotor slowing, irritability, and location-dependent neurological deficits. The incidence of post SRS-RN has been reported to be 10–12% in earlier case series with no definitive reference for RIBE. The options for management of RIBE are limited, viz. corticosteroids, antiplatelet drugs, anticoagulants, hyperbaric oxygen therapy, multivitamins, and Bevacizumab. There is lack of level one evidence supporting the superiority of one treatment modality over another. At present, steroids are the treatment of choice for RN/RIBE despite their adverse effect profile and limited clinical efficacy. Bevacizumab is considered in cases of steroid-refractory RIBE, steroid intolerance, or as a short-term therapy in cases of intractable edema.
The pathophysiology of RIBE is multifactorial dependent on dose, fractionation, volume, underlying pathology, concurrent treatment (such as chemotherapy), anatomic location, and factors inherent to patients.,,, The clinical mimicry with tumor recurrence often delays correct diagnosis. The pathogenesis of RN centers around endothelial cell dysfunction, overexpression of VEGF, creation of leaky capillary endothelium, breakdown of BBB, etc. It is necessary to differentiate between tumor progression from RIBE and RN. In our study, we found that Bevacizumab has a significant impact in the management of RIBE as evidenced by clinical and radiological improvement in the majority of the patients.
Most available literature approves the role of Bevacizumab in the management of RN for malignant intracranial pathologies such as metastasis and glioma. In 2007, Gonzalez et al. described their experience of the high efficacy of Bevacizumab in the treatment of RN in cases of malignant gliomas. In our study, we have evaluated the role of Bevacizumab in the management of post GKRS RIBE for various benign intracranial pathologies. The significant improvement obtained in our patients might be a result of relative normalization of BBB by a decrease in VEGF levels using Bevacizumab.
Definition of radiation injury
It is very important to differentiate local recurrence (LR) from RN, and a pathology combining these two entities for effective management. Both tumor recurrence and RN present as ring enhancing lesions with significant perilesional edema which is predominantly mediated by VEGF. RN and RIBE are rare injuries, which present with dramatic clinico-radiologicalchanges. The most common clinical symptoms are headache, dizziness, psychomotor slowing, seizures, sensorimotor deficits, and language impairments. Radiological deterioration may present in the absence of clinical findings. Clinical and radiological RN has been estimated to occur in up to 14% and 42% of the patients, respectively. At times, even the radiological imaging may fail to differentiate recurrent lesions from RN. Advanced radiological tools such as perfusion scans, MR spectroscopy, SPECT, and PET may help to differentiate between RN and tumor recurrence. A small subset still remains elusive. Thus histological confirmation remains the definitive diagnostic modality, and surgical management remains the final treatment option for such cases. Surgery provides immediate decompression with relieved mass effect and histopathological evidence. However, it carries a significant morbidity in patients already suffering with neurologic ailments. Surgery fails to address the underlying pathophysiology of RIBE and only provides palliative care, as it does not reverse the process in the majority of patients. One reason for this in the literature are the variations in microscopic analysis of such cases which present a mixed picture of RN, reactive gliosis, and evidence of viable tumor in the surrounding brain parenchyma.,,,,,,,,,
Though there is no definite time limit, RN and RIBE usually appear after three or more months post-radiation therapy. In our patients, RIBE was identified after 4.8 months after GKRS and Bevacizumab could be started only after 5.6 months. The reasons for the delay in treatment were two-fold; delay in the decision by the patient in view of experimental nature of the treatment and the cost of the treatment. Sometimes, patients were noncompliant because of financial reasons and discontinued the treatment despite showing improvement.
Management of radiation injury
In pre-Bevacizumab era, RN was uniformly treated by corticosteroids for an indefinite period. It was frequently complicated by relapse after steroid discontinuation and with the side-effects of long-term steroid administration. The other treatment options are pentoxyphylline, vitamin E, antiplatelet drugs, anticoagulants, hyperbaric oxygen therapy, therapeutic anticoagulation, or surgical decompression. Bevacizumab has significant improvement on quality of life (QOL), steroid/mannitol dose requirement, and overall survival. Improved cranial nerve function in patients of vestibular schwannoma is very important for functional clinical status of these patients.
We like to stress at this point that Bevacizumab should be considered as an early indication in cases of significant symptomatic edema nonresponsive to a short course of dexamethasone (4 mg every 6 hours for 1 week). It can also be considered an option in patients with comorbidities for which steroids cannot be administered such as uncontrolled diabetes, myopathy, and neuropathy.
The basic purpose of this study is to determine indications for starting Bevacizumab and to determine the endpoints of treatment. In our series, we strictly limited the use of Bevacizumab in patients who developed symptomatic edema after GKRS, which was refractory to corticosteroid therapy. Surgical decompression was only considered when the edema was nonresponsive to Bevacizumab and patient was still symptomatic. In our series, two patients did not respond to Bevacizumab, and needed surgical decompression because of significant mass effect. The key questions to answer are the time frame and maximum doses of Bevacizumab when surgical intervention should be considered in individual cases. Yomo et al. in their experience with symptomatic brain metastases have stressed that the functional status of the patient is a very important predictor of the outcome after starting Bevacizumab. Once there is a direct injury, it is very difficult to achieve good functional recovery even with a good radiological response secondary to Bevacizumab administration. Similarly, we observed that pre-Bevacizumab functional status is an important prognostic marker for functional outcome.
Radiological response to Bevacizumab
This study is unique in showing that patients with edem the presentation of patients with edema a after GKRS for different benign tumors and vascular malformations. Given the mixed pathologies and relatively longer follow-up term in the present study, the data provides proof of favorable effect of Bevacizumab on RIBE.
Duration of treatment
The treatment duration is based on patients' clinico-radiological responses to Bevacizumab. In our patients, we continued Bevacizumab for mean of seven times. It is interesting to note that Bevacizumab may provide radiological and clinical improvement in the short term; however, there are chances of disease progression without increased enhancement, the so-called “pseudo-response.”,, Pseudo-response may lead to delayed diagnosis of local failure in the absence of radiologic worsening. The present article provides evidence for the timing of Bevacizumab therapy. This study highlights the issue of relapse of RIBE after discontinuation of Bevacizumab prematurely and demonstrates the value of second-line Bevacizumab in some patients. The relapse can be explained by the decreasing concentration of the drug, with the half-life of Bevacizumab being 20 days (range= 11–50 days).
Complications with Bevacizumab
There are several shortcomings with high doses of Bevacizumab administration including bleeding,, increased chances of cerebral infarction,,, proteinuria, hypertension, and deterioration/rebound phenomenon after initial positive response with or without continued therapy.,, Jeyarethna et al. have reported delayed clinico-radiological deterioration in a previously responsive patient of Bevacizumab attributing the side-effect to overpruning of blood vessels. Bevacizumab also retards wound healing by impairing neovascularization., On the other hand, Levin et al. reported significant toxicity with deep venous thrombosis and superior sagittal sinus thrombosis in >50% patients treated with Bevacizumab. In our series, we witnessed some previously unreported complications, which might be just an incidental finding as well. One patient developed piles after the seventh cycle of Bevacizumab, which improved after stopping the treatment. Another patient developed papular urticaria after the second cycle of Bevacizumab, which did not require discontinuation of therapy. One young female with multiple intracranial lesions developed suicidal thoughts and depression for which Clonzepam and Escitalopram were needed for a short duration. Two patients showed symptomatic deterioration after stopping Bevacizumab after initial good response. None of our patients developed any bleeding manifestation [Table 1].
Shortcomings of the study
Though we have found a definitive benefit of Bevacizumab in the management of RIBE, the limited number of patients and retrospective design of the study warrant further evaluation in multi-institutional randomized control trials with statistically significant sample sizes. There are various unanswered questions regarding Bevacizumab treatment which demand better study designs such as endpoints of treatment, optimum number of treatment cycles, interval between two dose schedules, duration of treatment, clinical and radiological parameters for follow-up, and the role of concurrent medications such as anticoagulatns and antiepileptics. Another confounding variable is concurrent tapering of steroids for variable durations making it difficult to ascertain if the response was secondary to steroids or Bevacizumab treatment. The clinico-radiological diagnosis rather than histological diagnosis of RIBE is another limitation. This means that some of the lesions might be active disease or a mix of local tumor with RIBE. Concurrent corticosteroids might also cause clinico-radiological improvement to some extent, which cannot be quantified separately. Natural history of RN/RIBE remains ill-defined and may lead to a relapsing remitting course. In addition, the cost of Bevacizumab is a major constraint in a resource-stricken country such as India, hampering its usage in a majority of the population. Though the current literature seems promising for the use of Bevacizumab in RIBE, long-term results are needed to count the complications and maintenance of QOL.
Bevacizumab has a definite role in the management of RIBE after SRS. Well-designed randomized controlled prospective trials are needed to establish a management protocol to treat this dreaded complication. It is necessary to distinguish RIBE from RN as RIBE is a relatively benign condition. This study validates the benefits of Bevacizumab with protocol-based dose schedule with a long-term follow-up.
Declaration of patient consent
The authors certify that they have obtained all appropriate patient consent forms. In the form the patient(s) has/have given his/her/their consent for his/her/their images and other clinical information to be reported in the journal. The patients understand that their names and initials will not be published and due efforts will be made to conceal their identity, but anonymity cannot be guaranteed.
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.
[Figure 1], [Figure 2], [Figure 3], [Figure 4], [Figure 5]