» This article has been updated from its originally published version to correct units of measure. See the corresponding erratum notice, DOI: 10.3171/2017.8.JNS08268a. «J Neurosurg 108:000 000, 109:268 272, 2008 Bevacizumab and irinotecan therapy in glioblastoma multiforme: a series of 13 cases Sheikh A. Ali, M.D., 1 Wassim M. McHayleh, M.D., 2 Asif Ahmad, M.D., 3 Rajesh Sehgal, M.D., 2 Molly Braffet, P.A.-C., 3 Mohsin Rahman, M.D., 3 Ghassan Bejjani, M.D., 4 and David M. Friedland, M.D. 2 1 Department of Medicine, Division of Hematology Oncology, Temple University, Philadelphia; and Departments of 2 Hematology/Oncology, 3 Radiology, and 4 Neurosurgery, University of Pittsburgh Cancer Institute, Pittsburgh, Pennsylvania Object. Endothelial proliferation has been recognized as a marker of high-grade or aggressive glioma. Bevacizumab is a humanized immunoglobulin G1 monoclonal antibody to vascular endothelial growth factor that has been shown to have activity in malignant gliomas when combined with irinotecan. The authors report on a case series of 13 patients with recurrent heavily pretreated malignant glioma that was treated with the combination of bevacizumab and irinotecan. Methods. Standard therapy with primary resection followed by adjuvant chemotherapy and radiation had failed in all patients. The median number of therapies applied, including initial surgery, was 5 (range 3 7 therapies). Nine patients were started on bevacizumab at a dose of 5 mg/kg every 2 weeks. Four patients received bevacizumab at a dose of 10 mg/kg; irinotecan was given at a dose of 125 mg/kg every week for 3 weeks. Results. Of the 13 treated patients, 10 (77%) had a radiologically demonstrated partial response and 3 (23%) had stable disease. Six patients (46%) had a clinical response. The median time to disease progression while on treatment was 24 weeks. The median overall survival was 27 weeks. The disease progressed in 8 patients, despite an initial response. Five patients are still responding to therapy. Six of the 8 patients whose disease progressed have died. Bevacizumab was discontinued in 2 patients because of nonfatal intracranial bleeding. Conclusions. The combination of bevacizumab and irinotecan is safe and has excellent activity even in this relapsed, heavily pretreated population of patients with high-grade malignant glioma, most of whom would not be candidates for clinical trials. (DOI: 10.3171/JNS/2008/109/8/0268) Key Words bevacizumab glioblastoma multiforme irinotecan vascular endothelial growth factor Glioblastoma multiforme has been described as an anaplastic, highly cellular tumor with poorly dif ferentiated round or pleomorphic cells, nuclear aty pia, endothelial cell proliferation, neovascularization, and psuedopalisading necrosis. 9 Of the 18,000 primary brain tumors diagnosed annually in the US, more than half are GBMs. 5 They occur most often in the elderly, and they are more common in whites and in males. 5 The median duration of survival in patients with GBMs is 20 weeks with resection, 36 weeks with surgery plus radiation therapy, and 40 50 weeks with the addition of adjuvant chemotherapy. 12,16 Abbreviations used in this paper: ECOG = Eastern Cooperative Oncology Group; GBM = glioblastoma multiforme. Despite the modest survival benefit provided by adju vant radiation therapy and chemotherapy, the major - ity of pa tients experience a relapse following initial therapy. The prognosis for recurrent malignant gliomas is poor, with a reported median survival < 12 months, median pro gression-free survival < 12 weeks, and response rates < 20%. 15 The optimal treatment for patients with recurrent or progressive GBM is unclear, and there is no widely accepted standard of care. Among the therapeutic options are repeated surgery, repeated radiation therapy, conventional chemotherapy, and experimental treatments. The disappointing results of most therapeutic strategies in the setting of relapse 13,14 have led to efforts to find more effective and better tolerated treatment options, the most promising of which appears to be the combination of the newer agent irinotecan and the angiogenesis inhibitor bevacizumab. 7,11,15 268 J. Neurosurg. / Volume 109 / August 2008
Bevacizumab and irinotecan therapy in glioblastoma multiforme Endothelial proliferation has been recognized as a marker of high-grade or aggressive glioma in several grading classifications, and it has been demonstrated that the degree of microvascularity as assessed by the endothelial cell/capillary density correlates well with the biological aggressiveness of astrocytic tumors. 1 It has also been demonstrated that malignant glioma cells overexpress nu merous growth factors including vascular endo thelial growth factor 6,10 and epidermal growth factor receptor. 15 Based on this information, angiogenesis inhibition could play a role in the treatment of these highly vascular and aggressive tumors, a hypothesis that has been tested with varying success in both the laboratory 4 and clinical setting. 11 Historically, the results of using chemotherapy for recurrent malignant astrocytomas, particularly GBM, have been mostly disappointing; however, more recent studies have shown that the combination of temozolomide and irinotecan has activity and is of benefit in recurrent malignant gliomas. 8 Even in the past, irinotecan has been reported to have response rates as high as 15% in malignant glioma 3 and is an agent known to have modest activity in this disease, especially when used as combination therapy. It would thus be appropriate to the basic biology of the disease to consider combining an angiogenesis inhibitor such as bevacizumab with a chemotherapeutic agent with demonstrated activity in relapsed GBM such as irinotecan. The first clinical evidence to this effect came from data reported by Stark-Vance, 11 who used the combination of bevacizumab and irinotecan for the treatment of recurrent malignant glioma. In her report Stark- Vance described 1 patient with a complete response, 8 patients with partial responses, and 11 with stable disease among 21 patients treated with a combination of bevacizumab and irinotecan. Her report was bolstered by the results of a Phase II study by Vredenburgh et al., 15 which also showed the activity of this regimen in relapsed pretreated gliomas. These authors reported a response rate of 63% (19 partial responses and 1 complete response) with a median progression-free survival of 23 weeks and an overall 6-month survival probability of 72% in the 32 patients who were enrolled. Recently, Pope et al. 7 described an imaging response rate of 50% in patients treated with the same combination of irinotecan and bevacizumab. In this paper we describe a series of 13 cases in which this novel approach of using a combination of angiogenesis inhibition and chemotherapy was tried with success in a population of heavily pretreated patients with recurrent GBM, including several with marginal (ECOG Grades 2 3) performance status. Methods Patient Population After standard therapy with primary resection followed by adjuvant chemotherapy and radiation had failed, all patients were considered for therapy with bevacizumab and irinotecan. Patients were required to have a histologically proven diagnosis of a high-grade glioma (World Health Organization grade III) and failure of standard J. Neurosurg. / Volume 109 / August 2008 therapy including surgery followed by adjuvant chemotherapy and radiation. They had to be at least 6 weeks out from their most recent brain surgery, and they needed neurosurgical clearance before starting the bevacizumab therapy. Treatment Regimen Nine patients were started on bevacizumab at a dose of 5 mg/kg every 2 weeks and irinotecan at a dose of 125 mg/kg every week for 3 weeks with 1 week off medication. Four patients received bevacizumab at a dose of 10 mg/kg together with irinotecan at 125 250 mg/kg every 2 weeks. Dose modifications to the irinotecan were made in case of side effects including diarrhea, cytopenias, and poor tolerance. Bevacizumab was to be stopped in case of significant bleeding at any site. Therapy was changed or discontinued if patients had progressive disease as defined by our radiological response criteria. Response Measurement and Criteria Radiographic Response. Radiographic response was measured by comparing each patient s baseline MR imaging performed before the initiation of therapy and serial MR imaging performed every 4 8 weeks after starting therapy. All images were obtained on a GE Signa 1.5-T MR imaging unit. All postcontrast MR images were obtained after the intravenous administration of 20 ml Multihance (gadobenate dimeglumine). Axial T2-weighted fast spin echo, axial T2 FLAIR, axial postcontrast, and coronal postcontrast images were reviewed by a neurora diology staff member (M.R.) and a neuroradiology fellow (A.A.), both of whom were blinded to the clinical outcome. Imaging response criteria were defined as follows: partial response, a > 50% reduction in the overall degree of enhancement, a > 50% reduction in the degree of vasogenic edema, decreased mass effect, and a reduction by at least 25% in maximal tumor diameter; complete response, a complete resolution of vasogenic edema, enhancement, and a nonenhancing tumor; stable disease, a < 50% reduction in the overall degree of enhancement and vasogenic edema, and a < 25% reduction in maximal tumor diameter; and progressive disease, an increase in the degree of enhancement and vasogenic edema by at least 25% and/or an increase in maximal tumor diameter by at least 20%. Clinical Response. Any subjective improvement in neurological function, cognition, speech, mobility, or ability to perform activities of daily living was considered a clinical response. Results Thirteen patients, including 7 men and 6 women, were treated with this regimen (Table 1). The median age was 53 years (range 32 76 years). All patients had at least 1 recurrence after the initial definitive treatment and before being treated with bevacizumab and irinotecan, for a median number of 5 therapies including the initial surgery (range 3 7 therapies). The median ECOG performance status in our patient population was Grade 2. Of the 13 269
S. A. Ali et al. TABLE 1 Summary of characteristics in 13 patients with GBM Age at Diagnosis Type of GBM ECOG Performance (yrs), Sex Diagnosed Status Grade 37, F primary 1 76, F primary 1 65, M primary 2 56, F secondary 2 61, F primary 1 62, M secondary 3 40, M primary 1 53, M primary 2 32, F primary 1 46, M primary 1 50, M primary 2 47, M primary 3 57, F primary 2 treated patients, 10 (77%) had a partial response and 3 (23%) had stable disease according to MR imaging (Table 2). Radiographically, the patients showing a response had dramatic reductions in the size of the tumor masses and associated edema (Figs. 1 and 2). Six patients (46%) had a clinical response described as an improvement in neurological deficits or symptoms after starting therapy. The median time to disease progression on treatment was 24 weeks. The median overall survival after treatment had been started was 27 weeks. Of the 13 patients who were treated, 8 have had disease progression despite having an initial response and 5 are still responding to therapy. Six of the 8 patients who demonstrated disease progression have died. Bevacizumab was discontinued in 2 patients because of nonfatal intracranial bleeding. Patients whose disease progressed while on bevacizumab were found to have enhancing tumors. Discussion Approximately 35 years ago, Folkman 2 proposed the hypothesis that solid tumor growth was dependant on the development of tumor-associated blood vessels, a process called angiogenesis. Many studies on experimental and human tumors since then have confirmed the central role played by angiogenesis in various tumors including GBM. 6 Glioblastoma multiforme is a highly malignant and refractory disease that results in a high disease-specific mortality rate. Prior experience has demonstrated long-term survival in only anecdotal cases. Recent reports of authors using therapeutic combinations, including irinotecan and the anti vascular endothelial growth factor monoclonal antibody bevacizumab, have revealed high response rates. In the present series, after using the combination of irinotecan and bevacizumab, we report a partial response rate of 77% together with a progression-free interval of 24 weeks. Many of the patients had symptomatic, palliative benefits from the treatment. What makes these data even more compelling is the fact that this treatment led to substantial responses although the patients had been very TABLE 2 Treatment response and complications in 13 patients with GBM* Radiologically Bevacizu- Time to Dis- Case Initial Clinical Demonstrated mab Dose Survival ease Progres- No. Treatment Resection Response Response Complications (mg/kg) (wks) sion (wks) 1 surgery/xrt, temozolomide complete no PR none 5 16 16 2 surgery/xrt, BCNU wafer, complete no PR intracranial 10 13 13 temozolomide bleeding 3 surgery/xrt, temozolomide, complete no PR none 5 120 71 BCNU, irinotecan 4 surgery/xrt, temozolomide complete yes stable none 5 53 53 vaccine, GKS 5 surgery/xrt, temozolomide complete yes PR none 5 21 21 6 surgery/xrt, temozolomide, incomplete yes PR none 5 42 42 carboplatin, etoposide 7 surgery/xrt, temozolomide, complete no PR none 5 56 28 CCNU, irinotecan, erlotonib 8 surgery/xrt, temozolomide, complete yes PR DVT 5 65 48 thalidomide, etoposide, BCNU wafer, vincristine 9 surgery/xrt, temozolomide, incomplete no stable intracranial 5 48 43 BCNU, irinotecan bleeding 10 surgery/xrt, temozolomide, complete no PR none 5 22 13 BCNU, paclitaxel 11 surgery/xrt, temozolomide, incomplete no PR none 10 22 20 irinotecan, GKS 12 surgery/xrt, temozolomide, incomplete no PR none 5 27 19 BCNU, clinical trial 13 surgery/xrt, temozolomide incomplete yes PR none 10 24 24 * BCNU = carmustine; CCNU = lomustine; DVT = deep venous thrombosis; GKS = Gamma Knife surgery; PR = partial response; XRT = radiation treatment. No means no response to treatment and no disease progression. Since treatment with bevacizumab and/or irinotecan. 270 J. Neurosurg. / Volume 109 / August 2008
Bevacizumab and irinotecan therapy in glioblastoma multiforme Fig. 1. Axial radiographs obtained in a 47-year-old man before (A) and after treatment with irinotecan plus bevacizumab for ~ 19 weeks, showing an initial dramatic reduction in the size of the mass and associated edema (B), followed by progression with an enhancing, more aggressive pattern while on the same treatment (C). heavily pretreated and numerous surgical and medical ther apies had failed before the regimen was started. Additionally, many of these patients had a poor performance status (median ECOG Grade 2), which has historically been associated with a poor outcome 5,8 and would have made them ineligible for clinical trials. These findings sug gest that this angiogenesis-based therapy for GBM is tolerable and consistently active. Although toxicity is gen erally moderate, the antiangiogenesis component of the treatment can produce serious complications, including a risk of intracranial bleeding. When tumor progress - ion occurred, the radiographic features were consistent with the original presentation, showing infiltrative enhanc ing lesions. Although there was no evidence of extra cranial dissemination in anyone in this series, several pa tients had aggressive patterns of disease progression, in cluding gliomatosis cerebri in at least 2 cases. One patient who had undergone salvage surgery after the experimental therapy demonstrated a pathological conversion to glio sarcoma. Conclusions From the limited experience with antiangiogene sis therapy in GBM, it is not yet clear whether disease pro gression during or after this type of therapy will show different clinical or pathological features. Upcoming studies like the current Radiation Therapy Oncol - ogy Group protocol in which investigators will evaluate bevacizumab in combination with irinotecan or temozolomide in GBM at the first recurrence will shed further light on this approach in less heavily pretreated patients and will, we hope, lead to better survival outcomes in this disease. Ultimately, if this regimen proves to be superior Fig. 2. Axial radiographs obtained in a 37-year-old woman before (left) and after (right) treatment with bevacizumab and irinotecan for ~ 8 weeks, showing significant reduction in the mass and mass effect. J. Neurosurg. / Volume 109 / August 2008 271
S. A. Ali et al. to standard therapy, a randomized trial in the immediate perioperative setting will be warranted. References 1. Brem S, Cotran R, Folman J: Tumor angiogenesis: A quantita tive method for histologic grading. J Natl Cancer Inst 48: 347 356, 1972 2. Folkman J: Tumor angiogenesis: therapeutic implications. N Engl J Med 285:1182 1186, 1971 3. Friedman HS, Petros WP, Friedman AH, Schaaf LJ, Kerby T, Lawyer J, et al: Irinotecan therpy in adults with recur rent or progressive malignant glioma. J Clin Oncol 17:1516 1525, 1999 4. Goldbrunner RH, Bendszus M, Wood J, Kiderlen M, Sasaki M, Tonn JC: PTK787/ZK222584, an inhibitor of vascu lar endothelial growth factor receptor tyrosine kinases, de creas es glioma growth and vascularization. Neurosurgery 55:426 432, 2004 5. Grossman SA, Batara JF: Current management of glioblastoma multiforme. Semin Oncol 31:635 644, 2004 6. Plate KH, Brier KH, Brier G, Weich HA: Vascular endothelial growth factor is a potential tumor angiogenesis factor in human gliomas in vivo. Nature 359:845 848, 1992 7. Pope WB, Lai A, Nghiemphu P, Mischel P, Cloughesy TF: MRI in patients with high-grade gliomas treated with bevacizumab and chemotherapy. Neurology 66:1258 1260, 2006 8. Reardon DA, Quinn JA, Rich JN, Desjardins A, Vredenburgh J, Gururangan S, et al: Phase 1 trial of irinotecan plus temozolomide in adults with recurrent malignant glioma. Cancer 104:1478 1486, 2005 9. Reardon DA, Wen PY: Therapeutic advances in the treat- ment of glioblastoma: rationale and potential role of targeted agents. Oncologist 11:152 164, 2006 10. Shweiki D, Itin A, Soffer D, Keshet E: Vascular endothelial growth factor induced by hypoxia may mediate hypoxia-initiated angiogenesis. Nature 359:843 845, 1992 11. Stark-Vance V: Bevacizumab and CPT-11 in the treatment of relapsed malignant glioma. Neuro-oncol 7:369, 2005 12. Stupp R, Mason WP, van den Bent MJ, Weller M, Fisher B, Taphoorn MJ, et al: Radiotherapy plus concomitant and adjuvant temozolomide for glioblastoma. N Engl J Med 352: 987 996, 2005 13. Tatter SB: Recurrent malignant glioma in adults. Curr Treat Options Oncol 3:509 524, 2002 14. Voynov G, Kaufman S, Hong T, Pinkerton A, Simon R, Dowsett R: Treatment of recurrent malignant gliomas with stereotactic intensity modulated radiation therapy. Am J Clin Oncol 25:606 611, 2002 15. Vredenburgh JJ, Desjardins A, Herndon JE, Dowell JM, Reardon DA, Quinn JA, et al: Phase II Trial of bevacizumab and irinotecan in recurrent malignant glioma. Clin Cancer Res 13:1253 1259, 2007 16. Walker MD, Alexander E Jr, Hunt WE, MacCarty CS, Mahaley MS Jr, Mealy J Jr, et al: Evaluation of BCNU and/or radiotherapy in the treatment of anaplastic gliomas. A cooperative clinical trial. J Neurosurg 49:333 343, 1978 Manuscript submitted May 30, 2007. Accepted October 18, 2007. Sources of support: none reported. Address correspondence to: Wassim McHayleh, M.D., 5150 Centre Avenue, Pittsburgh, Pennsylvania 15232. email: Mchaylehw@ upmc.edu. 272 J. Neurosurg. / Volume 109 / August 2008