Glioblastoma multiforme is the most common malignant primary brain tumor in adults, with an estimated incidence of 4.43 per 100,000 person-years in the United States and a median age at presentation of 64 years.1 Glioblastoma multiforme is characterized by seizures; nausea; vomiting; headaches; and progressive memory, personality, or neurologic deficits, as well as treatment resistance.2 The treatment of glioblastoma multiforme is a challenge, and despite the approval of multiple new therapies in the past decade, survival remains poor.
Based on a national report on the status of cancer published in 2011 in the Journal of the National Cancer Institute, the 5-year relative survival rates for glioblastoma multiforme among adults between 2000 and 2006 was only 21.3% for patients aged 20 to 39 years, 5.3% for those aged 40 to 64 years, and only 1.1% for patients aged ≥65 years in the United States.1 These national 5-year relative survival rates were slightly better when considering all tumors of the neuroepithelial tissue (65.1%, 26.6%, and 4.6% for the same 3 age-groups, respectively).1
The current standard of care for newly diagnosed glioblastoma is derived from a randomized clinical trial published in 2005 and consists of maximal feasible surgical resection followed by radiotherapy with concurrent and adjuvant temozolomide.3 This treatment regimen, known as the Stupp regimen, has resulted in a median survival of 14.6 months in patients receiving temozolomide therapy alone compared with 12.1 months in patients receiving external beam radiation alone.3 The adoption of the Stupp regimen has been credited for improvement in the survival of patients with glioblastoma multiforme from 2005 to 2008 compared with the survival from 2000 to 2003, particularly among younger patients.4,5
The US Food and Drug Administration (FDA) approved the use of temozolomide for the treatment of glioblastoma multiforme in March 2005. The FDA also approved carmustine wafers (initially in 1997) and bevacizumab (in 2009, for glioblastoma multiforme that has progressed after initial treatment) for the treatment of glioblastoma multiforme, but neither of these treatments has demonstrated a significant role in the upfront treatment of this disease.6
The financial costs associated with the addition of temozolomide are significant and have been well documented, particularly in European and Canadian health systems.7 In the United States, several analyses have underscored the overall costs and burden of out-of-pocket (OOP) costs incurred by patients with glioblastoma multiforme for hospital visits, ancillary care, and drug costs.8,9 The total expenditures in this patient population have also been described in 2007 by Kutikova and colleagues for 653 patients with primary malignant brain tumors and were estimated at $6364 per month compared with $277 for controls.9 These costs were mostly associated with inpatient care and likely reflect patient care before the widespread use of temozolomide.9
To our knowledge, no study has comprehensively described the total healthcare costs associated with the treatment of glioblastoma and malignant gliomas in the temozolomide era in the United States. We sought to understand the treatment patterns, survival, and economic burden incurred by patients with glioblastoma in clinical practice in the United States. In this study, we used a large commercial claims database and specifically sought to identify a cohort of patients based on the International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) 191.xx codes that most likely represent newly diagnosed glioblastoma to describe patient survival, comorbidities, treatment duration, and healthcare expenditures in the time period after the FDA’s approval of temozolomide.
Data Source and Study Design
Data for this study were gathered and linked from 2 sources: (1) healthcare claims from the Truven Health Analytics MarketScan Commercial and Medicare Supplemental Databases, and (2) the Social Security Administration (SSA) master death files.
The Truven Health databases include fully integrated real-world patient-level data, including pharmacy and medical claims and associated diagnosis and procedure codes, and enrollment data from approximately 25 million lives covered annually by self-insured employers and private health insurance plans, geographically diversified across the United States. For patients with glioma who receive supplemental Medicare benefits through employer-sponsored health plans, information on the employer-paid portion of Medicare-paid benefits and patients’ OOP expenses for their medical and pharmacy services were also available. The SSA death files were linked to patient enrollment data to identify patients with glioma who died, and the date of death.
The combined data set was used to study the treatment patterns, survival, and healthcare costs of patients with an incident glioma who initiated treatment with brain surgery. A retrospective cohort study design was used with the patients stratified into 1 of 4 treatment cohorts based on the receipt of temozolomide and/or external beam radiation after their initial brain surgery.
Patient Selection and Cohorts
Patients were included if they (1) were diagnosed with malignant cancer of the brain (ICD-9-CM, 191.xx) on or between January 1, 2006, and December 31, 2010; (2) had undergone brain-related surgery (set as the index event) within 90 days (before or after) of the first diagnosis of 191.xx; (3) were aged ≥18 years at the index date; and (4) were continuously enrolled with medical and pharmacy benefits for 6 months before the index date.
Patients were excluded if they (1) had a diagnosis of another primary cancer (ICD-9-CM, 140.xx-195.xx and 200.xx-208.xx) in the 6 months before the index date; (2) had a diagnosis of secondary brain metastases (ICD-9-CM, 198.3) before the index date; (3) received chemotherapy or temozolomide, or had index-eligible brain surgery during the preindex period; or (4) used an off-label or nonstandard-of-care therapy as part of their first line of therapy (ie, in the 90 days after their index brain surgery), including carmustine wafer, bevacizumab, or other chemotherapy, or stereotactic radiosurgery.
The follow-up period varied for each patient and ran from the index date until a patient’s date of death, disenrollment from an eligible health plan, or until March 31, 2011, whichever occurred first. Although there was no minimal postindex continuous enrollment requirement, fully adjudicated data were available through March 31, 2011, providing at least 3 months of potential data availability for all patients. To evaluate mortality as a study outcome, the analysis was limited to the subset of patients in the commercial and Medicare databases that could be linked to the SSA death data, to determine if a patient died during the study period. Patient survival was censored at the end of the follow-up.
Patients were divided into the following 4 mutually exclusive cohorts based on whether they received temozolomide and/or external beam radiation in the 90 days after the index brain surgery. The 4 cohorts included those who (1) received temozolomide only, (2) received external beam radiation only, (3) received both temozolomide plus external beam radiation, and (4) received neither temozolomide nor external beam radiation.
This study is focused on 3 types of outcomes—patterns of treatment, survival, and healthcare costs.
Treatment patterns. Among patients in 1 of the 2 temozolomide cohorts, the total duration and medication possession ratio of the first temozolomide episode is described. The end of the initial episode of temozolomide is defined as either patient death or disenrollment, or the start of a 60-day gap in temozolomide therapy. The proportion of patients restarting temozolomide therapy after a 60-day gap was also calculated.
Survival. The survival time was calculated using the date of death as obtained from the SSA; patients without a date of death were censored at the end of follow-up.
Healthcare costs. Total insurance-covered healthcare costs are reported, including both patient and plan portions of each claim for all services utilized during the study period (including those not specifically listed below). The data source includes only fully adjudicated and paid claims. The costs are reported in 3 categories of expenditures based on the location and type of healthcare resource used: inpatient, outpatient, and pharmacy. The outpatient expenditures are separated into emergency, outpatient hospital, and office visits. The pharmacy costs are classified by antiemetics, cancer therapies, neutropenia-related drugs, and pain-related drugs.
The expenditures were evaluated in 3 time periods relative to the index brain surgery—6 months before the index brain surgery, 6 months after the index brain surgery, and 12 months after the index brain surgery.
Descriptive statistics were used for all 3 outcomes. In addition, the baseline patient clinical and demographic characteristics are described, including age, sex, health plan type, and urbanicity, as well as relevant concomitant medications and comorbid conditions.
We identified 69,495 patients with an ICD-9-CM code of 191.xx in the MarketScan Research databases, of which 17,137 (24.7%) had brain surgery within 90 days of the 191.xx diagnosis. Of those patients, 12,143 (70.9%) were adults at the time of the index brain surgery and had 6 months of continuous medical and pharmacy coverage before the index brain surgery. An additional 4773 patients were excluded for either having another primary cancer, evidence of brain metastases, having had brain surgery, chemotherapy administration, or the use of a carmustine wafer in the 6 months of having the index brain surgery, leaving 7370 evaluable patients with a malignant brain tumor. Data regarding patient survival were available for 2484 of these patients. An additional 212 patients used a nonstandard-of-care therapy as a first-line treatment, resulting in a final study sample of 2272 patients.
The overall incidence of maligant brain tumors was 0.0056% between the years 2006 and 2010, with a range in individual years from 0.0053% to 0.0060% among patients in the MarketScan Research databases. The remaining 2272 patients were divided into 4 treatment cohorts based on their receipt of temozolomide and/or external beam radiation in the 90 days after their index brain surgery. The reason for this was to uncover real-world treatment patterns and to further identify a specific cohort that most likely represents a group of patients with malignant glioma receiving standard-of-care therapy.
The largest group included 1029 patients (45.3%) who did not receive temozolomide or external beam radiation in the 90 days after their index brain surgery. The second group of 841 (37%) patients received both temozolomide and radiation therapy, and is the cohort that most clearly represents the current standard-of-care therapy for patients with malignant gliomas in the United States. Smaller percentages of patients received radiation alone (13.8%) and temozolomide alone (3.9%).
Demographic and clinical characteristics are displayed in Table 1. These groups were generally well balanced in terms of age, sex, geographic region, type of health plan, and the length of follow-up. The median age of the final study sample was 58 years (mean, 56.9; standard deviation, 14.5); a total of 63.1% of the patients were male, and 29.9% were aged >65 years. Corticosteroids (40.2%), anticonvulsants (34.5%), narcotic analgesics (26.1%), and anxiety medications (18.2%) were frequently used in the 6 months before the index brain surgery. Similarly, during the 6 months before the index, nearly 22.1% of patients had a claim with a diagnosis of headache, 20.2% had a diagnosis of cerebrovascular disease or stroke, and 18.9% had a diagnosis of seizures.
The median survival time was 456 days across all 2272 patients, ranging from a median of 331 days among patients who received only temozolomide to 529 days among patients who received neither temozolomide nor radiation therapy. In the cohort of patients receiving both radiation and temozolomide, the median survival was 426 days (14.2 months). Death was observed in nearly half (48.2%) of the patients, ranging from 28.4% among those who received neither temozolomide nor radiation therapy to 68% among patients who received both temozolomide and radiation.
As shown in Table 2, 39.6% of patients used temozolomide in the 90 days after their index surgery. The duration of this therapy was, on average, 180 days for patients who received both temozolomide and radiation therapy and 155 days for patients who did not receive radiation therapy. Medication possession ratios were similar for patients who did and did not receive radiation therapy (72.6% and 73.4%, respectively). The majority of patients received temozolomide for at least 120 days (59.9%), and nearly half (48.1%) of the patients received temozolomide for at least 180 days. A total of 30.9% of the patients received temozolomide for at least 270 days.
Table 3 lists the healthcare costs incurred by patients before and after their index brain surgery. The mean total cost over the 6 months before the index brain surgery was $11,949, ranging from $11,564 in the radiation-only cohort to $12,850 in the temozolomide-only cohort. The average total cost in the 6 months after the index surgery was markedly higher, averaging $106,896, ranging from $79,099 among patients who did not receive temozolomide or radiation to $138,767 among patients who received both temozolomide and radiation.
The total costs in the 12 months after the index surgery were higher than the costs after only 6 months, with an average cost of $131,815, and ranging from $88,827 among patients who did not receive temozolomide or radiation to $184,107 among patients who received both temozolomide and radiation. All 3 components of the total cost (ie, inpatient, oupatient, and pharmacy) showed a pattern similar to the total costs in the periods before and after the surgery.
Although the Stupp regimen—which consists of radiotherapy and temozolomide being administered concomitantly, and then temozolomide is used after radiotherapy—has become the standard of care for newly diagnosed patients with glioblastoma, there are limited data on this patient population since the introduction of this regimen in a clinical practice setting. Using administrative claims and mortality data, this study provides data that are useful for understanding treatment patterns, survival, and the healthcare costs associated with glioblastoma as observed in a large, representative commercially insured US population.
The current ICD-9-CM classification system does not clearly discriminate patients by histopathology, and patients with a diagnosis code of 191.xx can represent a heterogeneous group of tumor types. Although this is a limitation of this current analysis, malignant gliomas make up the overwhelming majority of malignant brain tumors in adults, and the contribution of other rare tumors is likely inconsequential.10 One group that may be coded as 191.xx may include lower-grade gliomas, which eventually progress into higher-grade tumors.
To purify our sample, we determined the treatments that patients received to better understand their potential tumor makeup. We created 4 mutually exclusive categories based on the patients’ exposure to radiation and/or temozolomide therapy. Remarkably, more than 40% of patients received neither of these therapies. The relatively high median survival in this group suggests that many of these patients might have had lower-grade histologies, for whom upfront radiation and/or temozolomide are not clearly standard-of-care therapies.11
It is possible that this cohort includes elderly or poorly performing patients in whom a decision was made not to pursue active treatment; however, these patients did not appear substantively different from the other treatment cohorts, particularly in terms of the distribution of age-groups. Therefore, it is more likely that, in this analysis, less-aggressive treatment is an indicator of lower-grade gliomas, where such less-aggressive treatment is clinically reasonable.
The preindex healthcare utilization was similar across the 4 cohorts; therefore, the difference in treatment cohort costs does not appear to be a result of the background comorbid burden, or the patient’s clinical profile. Furthermore, we sought to characterize a cohort of patients that most likely represent patients with newly diagnosed glioblastoma based on the type of treatments that they received. For that purpose, we defined the standard of care based on the Stupp regimen,3 and after applying exhaustive inclusion and exclusion criteria, we were able to identify 841 patients who received radiation plus temozolomide therapy. Although a small portion of these patients could have anaplastic gliomas, given the emerging treatment patterns in tertiary care centers,12 we believe that our cohort of 841 patients receiving temozolomide plus radiation are patients with glioblastoma. This is because anaplastic gliomas are rare compared with glioblastoma, and the median survival in this cohort (of 426 days, or 14.2 months) fits closely with the findings reported with the Stupp regimen for glioblastoma (14.6 months).3
The role of temozolomide for anaplastic gliomas in the upfront setting, however, remains controversial. Anaplastic gliomas, which are World Health Organization grade 3 tumors, make up less than 50% of all malignant gliomas; they are a heterogenous group of tumors in terms of their histology, molecular markers, treatment, and survival.13 Because of the rarity of these tumors, few large phase 3 clinical trials have been conducted to inform optimal therapies.
In the United States, the treatment of these tumors is increasingly extrapolated from studies done on patients with glioblastoma. A recently published large phase 3 trial has confirmed a role of chemotherapy for 1p/19q codeleted anaplastic oligodendrogliomas, but because this study was conceived more than 20 years ago, it investigated procarbazine, lomustine, and vincristine therapy rather than temozolomide therapy.14 An ongoing large phase 3 trial from the Radiation Therapy Oncology Group (RTOG) is addressing the role of temozolomide in anaplastic gliomas (clinicaltrials.gov identifiers NCT01847235 and NCT00033280); but as of now, there are no prospective data that demonstrate efficacy of temozolomide in this population. Given the costs and the lack of evidence, one may consider enrolling such patients in active RTOG trials that are aiming to properly and systematically address the question of temozolomide’s efficacy in anaplastic gliomas.
Not surprising, our study shows that healthcare utilization and costs increase after patients with malignant glioma have surgery. In our analysis, in the 1 year after surgery in the 841 patients with presumed glioblastoma who received radiation and temozolomide therapy, the total healthcare expenditure was $184,107. This cost is substantially higher than in the study by Kutikova and colleagues, which did not report any inclusion of temozolomide data, but that estimated the 1-year healthcare costs to be less than $80,000.9
In this current analysis, the total costs for patients who received both temozolomide and external beam radiation were 1.8 times greater over 6 months than for patients who received neither, and 2.1 times greater over 12 months. After 6 months, the cost for patients who did not receive temozolomide or external beam radiation returned to their presurgery levels; the average 6-month preindex cost was $11,868; the average cost for months 7 to 12 postindex were $9728.
Our descriptive analysis did not identify predictors of healthcare costs, treatment patterns, or patient survival. There are several known predictors of improved survival, such as age, performance status, and tumor O6-methylguanine-DNA methyltransferase (MGMT) methylation.15 Patients managed with methylation of the MGMT promoter have shown improved survival, but broad implementation of testing remains impractical,16 warranting further research and potential promise for treatment prognosis, at least among some subgroups of patients.17
As the US population ages, we expect a rise in the incidence of glioblastoma, and the burden on third-party payers may change substantially over the next few decades.18 The Stupp regimen excluded patients aged >70 years, and there is controversy regarding the role of adjuvant chemotherapy in older patients. A recent phase 3 study showed that temozolomide therapy alone in highly functioning elderly patients is tolerable and is noninferior to radiation therapy alone.19 Another study comparing radiation with temozolomide to radiation alone demonstrated a marginal improvement in survival with chemotherapy,20 and even poorly functioning elderly patients seem to derive some benefit from temozolomide therapy.21
This study had several limitations. In addition to the previously noted limitation regarding the specificity of ICD-9-CM coding, there are additional limitations inherent in the data source used in this analysis. Administrative claims data lack information on disease severity or staging.
Similarly, there is limited information on patient or provider characteristics that may influence medical decision-making. This also limits our ability to differentiate the treatment cohorts or to potentially predict or explain the reasons for administering temozolomide and/or external beam radiation therapy.
Although it is necessary to describe the baseline characteristics and to establish the incident event, the preindex continuous enrollment requirement may bias the study sample toward patients with stable health insurance who may be healthier than patients with intermittent health insurance, or the uninsured.
The 4 cohorts were defined based on treatments received in the 90 days after the index surgery event, because this corresponded with a clinically reasonable time period to initiate either temozolomide or external beam radiation therapy. Patients in any of these cohorts could potentially receive (or initiate) temozolomide therapy outside of this 90-day time period.
Based on our analysis, the addition of temozolomide to the treatment regimens for glioblastoma increases the cost of care, and the use of temozolomide potentially indicates greater disease severity. Although survival in this clinical practice setting–based analysis is similar to the survival reported in clinical trials, further cost-effectiveness and quality-of-life analyses will be critical to better understanding the value of temozolomide therapy in treating this patient population, particularly because of the availability of generic temozolomide.
This study was funded by Abbott Laboratories.
Author Disclosure Statement
Dr Ray is an employee of and has stocks in AbbVie; Dr Bonafede reported no conflicts of interest; Dr Mohile is a consultant to Truven Health Analytics and to Abbott Laboratories, and is on the speaker’s bureau of NovoTTF.
Dr Ray is Director, Oncology, AbbVie (formerly Abbott Laboratories), Abbott Park, IL; Dr Bonafede is Director, Outcomes Research, Truven Health Analytics, Cambridge, MA; Dr Mohile is Assistant Professor of Oncology,
Neuro-Oncology, University of Rochester Medical Center, Cambridge, Rochester, NY.
- Kohler BA, Ward E, McCarthy BJ, et al. Annual report to the nation on the status of cancer, 1975-2007, featuring tumors of the brain and other nervous system. J Natl Cancer Inst. 2011;103:714-736.
- Central Brain Tumor Registry of the United States. CBTRUS statistical report: primary brain and central nervous system tumors diagnosed in the United States in 2004-2007. February 2011. www.cbtrus.org/2011-NPCR-SEER/WEB-0407-Report-3-3-2011.pdf. Accessed March 15, 2014.
- Stupp R, Mason WP, van den Bent MJ, et al; for the European Organisation for Research and Treatment of Cancer Brain Tumor and Radiotherapy Groups; National Cancer Institute of Canada Clinical Trials Group. Radiotherapy plus concomitant and adjuvant temozolomide for glioblastoma. N Engl J Med. 2005;352:987-996.
- Koshy M, Vilano JL, Dolecek TA, et al. Improved survival time trends for glioblastoma using the SEER 17 population-based registries. J Neurooncol. 2012;107:207-212.
- Johnson DR, O’Neill BP. Glioblastoma survival in the United States before and during the temozolomide era. J Neurooncol. 2012;107:359-364.
- Stupp R, Tonn JC, Brada M, Pentheroudakis G; for the ESMO Guidelines Working Group. High-grade malignant glioma: ESMO Clinical Practice Guidelines for diagnosis, treatment and follow-up. Ann Oncol. 2010;21(suppl 5):v190-v193.
- Lamers LM, Stupp R, van den Bent MJ, et al; for the EORTC 26981/22981 NCI-C CE3 Intergroup Study. Cost-effectiveness of temozolomide for the treatment of newly diagnosed glioblastoma multiforme: a report from the EORTC 26981/22981 NCI-C CE3 Intergroup Study. Cancer. 2008;112:1337-1344.
- Raborn ML, Pelletier EM, Smith DB, Reyes CM. Patient out-of-pocket payments for oral oncolytics: results from a 2009 US claims data analysis. J Oncol Pract. 2012;8 (3 suppl):9s-15s.
- Kutikova L, Bowman L, Chang S, et al. Utilization and cost of health care services associated with primary malignant brain tumors in the United States. J Neurooncol. 2007;81:61-65.
- Porter KR, McCarthy BJ, Freels S, et al. Prevalence estimates for primary brain tumors in the United States by age, gender, behavior, and histology. Neuro Oncol. 2010;12:520-527.
- Papagikos MA, Shaw EG, Stieber VW. Lessons learned from randomised clinical trials in adult low grade glioma. Lancet Oncol. 2005;6:240-244.
- Abrey LE, Louis DN, Paleologos N, et al; for the Oligodendroglioma Study Group. Survey of treatment recommendations for anaplastic oligodendroglioma. Neuro Oncol. 2007;9:314-318.
- Wen PY, Kesari S. Malignant gliomas in adults. N Engl J Med. 2008;359:492-507. Erratum in: N Engl J Med. 2008;359:877.
- van den Bent MJ, Brandes AA, Taphoorn MJ, et al. Adjuvant procarbazine, lomustine, and vincristine chemotherapy in newly diagnosed anaplastic oligodendroglioma: long-term follow-up of EORTC brain tumor group study 26951. J Clin Oncol. 2013;31:344-350.
- Hegi ME, Diserens AC, Gorlia T, et al. MGMT gene silencing and benefit from temozolomide in glioblastoma. N Engl J Med. 2005;352:997-1003.
- DeAngelis LM. Chemotherapy for brain tumors—a new beginning. N Engl J Med. 2005;352:1036-1038.
- Fietkau R, Putz F, Lahmer G, et al. Can MGMT promoter methylation status be used as a prognostic and predictive marker for glioblastoma multiforme at the present time? A word of caution. Strahlenther Onkol. 2013;189:993-995.
- Werner MH, Phuphanich S, Lyman GH. The increasing incidence of malignant gliomas and primary central nervous system lymphoma in the elderly. Cancer. 1995; 76:1634-1642.
- Wick W, Platten M, Meisner C, et al; for the NOA-08 Study Group of Neuro-oncology Working Group (NOA) of German Cancer Society. Temozolomide chemotherapy alone versus radiotherapy alone for malignant astrocytoma in the elderly: the NOA-08 randomised, phase 3 trial. Lancet Oncol. 2012;13:707-715.
- Malmström A, Grønberg BH, Marosi C, et al; for the Nordic Clinical Brain Tumour Study Group (NCBTSG). Temozolomide versus standard 6-week radiotherapy versus hypofractionated radiotherapy in patients older than 60 years with glioblastoma: the Nordic randomised, phase 3 trial. Lancet Oncol. 2012;13:916-926.
- Gállego Pérez-Larraya J, Ducray F, Chinot O, et al. Temozolomide in elderly patients with newly diagnosed glioblastoma and poor performance status: an ANOCEF phase II trial. J Clin Oncol. 2011;29:3050-3055.