Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Brain metastases in patients with ALK+ non-small cell lung cancer: clinical symptoms, treatment patterns and economic burden

Brain metastases in patients with ALK+ non-small cell lung cancer: clinical symptoms, treatment... Annie Guerin Analysis Group, Inc., Montreal, QC, Canada Objective: Medha Sasane Brain metastases (BM) are highly prevalent among anaplastic lymphoma kinase positive (ALKþ) non-small Jie Zhang cell lung cancer (NSCLC) patients; yet little is known about their real-world treatment patterns and clinical Kenneth W. Culver and economic burdens. This study aimed to describe these patients’ treatment patterns, symptoms, and Novartis Pharmaceuticals Corporation, East Hanover, costs. NJ, USA Research design and methods: Katherine Dea Retrospective study pooling data from three large administrative databases in the US (08/2011–06/2013). Roy Nitulescu ALKþ NSCLC patients with BM and continuous enrollment for 60 days before and30 days after the first Analysis Group, Inc., Montreal, QC, Canada observed BM diagnosis were identified by pharmacy records for crizotinib among patients with lung cancer Eric Qiong Wu and BM diagnostic codes. Analysis Group, Inc., Boston, MA, USA Main outcome measures: Address for correspondence: Treatment patterns, symptoms, healthcare resource utilization, and costs, before and after BM diagnosis. Annie Gue ´ rin, 1000 de la Gauchetie ` re Ouest, Bureau 1200, Montre ´ al, QC, H3B 4W5 Canada. Results: Tel: 514-394-4484; Fax: 514-394-4461; Of the 213 crizotinib patients with BM diagnoses meeting the selection criteria, 23.0% had BM prior to annie.guerin@analysisgroup.com NSCLC diagnosis; 47.4% had BM prior to crizotinib initiation; 19.2% during crizotinib treatment; and 10.3% post-crizotinib treatment. For those diagnosed with BM after NSCLC diagnosis, the median time between Keywords: the NSCLC and BM diagnoses was 88 days. Following the first observed BM diagnosis, 88.7% used Non-small-cell lung cancer – Brain neoplasms – chemotherapy, 63.4% had radiotherapy, and 31.9% had stereotactic radiosurgery. The prevalence Neoplasm metastasis – Costs and cost analysis – of BM-related symptoms substantially increased post-BM-diagnosis: fatigue (from 15% to 39%), Crizotinib – Symptoms headaches (from 5% to 24%), and depression (from 5% to 15%). Monthly costs per patient averaged Accepted: 29 December 2014; published online: 6 February 2015 $5983 before the BM diagnosis and $22,645 after diagnosis. Patients’ resource utilization increased Citation: J Med Econ 2015; 18:312–22 significantly post-BM-diagnosis, with a 3-fold increase in OP visits and a 6-fold increase in IP stays. Post-BM-diagnosis costs were driven by pharmacy (42.0%), inpatient (29.6%), and outpatient costs (26.0%). Limitations: The study sample was limited to crizotinib-treated patients. Conclusions: Post-BM-diagnosis, patients experience high symptom burden. Post-BM-diagnosis, treatment is highly variable and costly: average monthly costs per patient almost quadrupled post-BM-diagnosis. 312 Burden of brain metastasis in ALKþ lung cancer Gue´ rin et al. www.informahealthcare.com/jme 2015 Informa UK Ltd Journal of Medical Economics Volume 18, Number 4 April 2015 Introduction Methods Non-small-cell lung cancer (NSCLC) accounts for more Data sources than 85% of all lung cancers in the US . A large majority This study combined data from three large retrospective of NSCLC patients are diagnosed at an advanced stage of administrative-claims databases (Source Healthcare the disease and, as a result, experience limited success with Analytics’ Source Lx database for June 2011–June 2013, treatment. Only a small fraction of patients (i.e., 4% of IMS LifeLink Health Plan Claims database for January metastatic and 18% of all NSCLC) survive more than 2001–March 2013, and Truven Health Analytics 5 years post-diagnosis . Growing understanding of the MarketScan database for January 2002–September 2012) molecular pathogenesis of lung cancer has led to the devel- to extract patient data. The databases contain enrollment opment of targeted therapies in an effort to improve 3,4 records, medical claims ( physician and facility), and phar- patient survival . In particular, tumors with activating macy claims. All census regions are represented in all three mutations of select oncogenes, such as the fusion of the databases. kinase domain of the anaplastic lymphoma kinase (ALK) The Source Lx database contains linked healthcare to the amino-terminal section of the echinoderm micro- data for more than 156 million annually covered lives, tubule-associated protein-like 4 (EML4) gene product, 5 derived from three basic sources: pharmacy point-of-ser- have been found to be susceptible to targeted therapies . vice; switch/network transactions; and additional direct The EML4-ALK fusion results in the constitutive activa- prescription, medical, and hospital claims data. 6,7 tion of a crucial cell-division-cascade ; thus, these tumors The IMS LifeLink Health Plan Claims database 8,9 tend to have an aggressive outlook with a high chance of contains combined claims for over 100 healthcare plans, early relapse . Patients with ALKþ NSCLC, which rep- representing 42 million annually covered lives. 11–19 resent 2–8% of all patients diagnosed with NSCLC , The Truven Health Analytics MarketScan database tend to have a more advanced stage of the disease and contains combined claims for more than 130 employers multiple metastatic sites at diagnosis compared to other that sponsor private health insurance, representing 40 NSCLC types . million annually covered lives. In addition, the The treatment of ALKþ NSCLC is often complex due MarketScan Medicare Supplemental and Coordination to the frequent occurrence of brain metastases: initial of Benefits (Medicare Supplemental) database profiles occurrence of brain metastasis has been reported in the healthcare experience of retirees and their dependents 20–22 15–35% of cases , and, over the course of first-line with Medicare supplemental insurance paid for by therapy either with crizotinib or chemotherapy the fre- employers. quency of brain metastases can increase to 60% . Brain All data used in this study had been de-identified metastases are associated with many complications: some according to the patient confidentiality requirements of of these are neurocognitive, psychological, physical the Health Insurance Portability and Accountability Act 24,25 impairments , severe comorbidity, and a decreased (HIPAA) and did not require approval from the 26,27 life expectancy . Treatment of brain metastases usually Institutional Review Board or a waiver of authorization involves radiotherapy (either whole-brain radiotherapy for use. [WBRT] or stereotactic radiosurgery [SRS]) and may involve surgical resection (for isolated metastases or Study design cases with a good prognosis) and/or chemotherapy (for 28–30 salvage or experimental cases) . In addition, post- For this retrospective cross-sectional cohort study, patients treatment, rehabilitation therapy may be needed ,as who met the following selection criteria were selected: (1) well as help with daily living tasks for recovering 18 years of age; (2) 1 diagnosis for lung cancer (ICD-9 32,33 patients . Thus, disease management and treatment CM: 162.xx); (3) 1 prescription fill for crizotinib (this planning for patients with brain metastases typically criterion is used as a proxy to identify patients with ALKþ requires a multi-disciplinary approach and highly specia- NSCLC since there is no ICD-9 CM diagnosis code to 34–36 32 lized care , and involves a substantial caregiver and identify ALKþ patients directly and crizotinib was economic burden. restricted to the treatment of this population only); (4) Few studies have characterized the clinical and eco- no medical claims associated with a clinical trial (identi- nomic burden experienced by patients with brain metas- fied using ICD-9 CM code for V70.7); (5) 1 diagnosis of 27,37,38 tases , and none has specifically analyzed the brain metastasis (ICD-9 CM codes 198.3 and 198.4); and (ALKþ) NSCLC population. The objective of this study (6) continuous eligibility2 months before and1 month is to evaluate the impact of developing brain metastases on after the date of the first observed diagnosis of brain metas- patients with ALKþ NSCLC in terms of economic and tasis. Patients selected for this study were referred to as clinical outcomes. ‘crizotinib patients’. 2015 Informa UK Ltd www.informahealthcare.com/jme Burden of brain metastasis in ALKþ lung cancer Gue´ rin et al. 313 Journal of Medical Economics Volume 18, Number 4 April 2015 observed brain metastasis diagnosis, while the post-brain Patient characteristics metastasis-diagnosis period spanned between 30 days The reported patient characteristics included demograph- before the first observed brain metastasis diagnosis ics (age, gender, geographical region), the year of the first and the end of data availability or of continuous eligibility observed brain metastasis diagnosis, the Charlson (Figure 1). The pre-metastasis-diagnosis period was trun- Comorbidity Index (CCI; a higher score indicates a greater cated since some patients had very long observation peri- comorbidity burden) , an indicator of the patient’s overall ods, and truncation was done at 1 year since the median comorbidity profile, and the CCI components reported length of this period was 1 year. The following symptoms individually. Age was measured as of the date of the first that required medical attention and the proportion of observed brain metastasis diagnosis, while the comorbid- patients with these symptoms that are the most commonly ities were measured during the pre-brain-metastasis- reported in the literature were analyzed during these two diagnosis period. The pre-brain-metastasis-diagnosis observation periods: weakness or fatigue, shortness of period was defined as a maximum 1-year period starting breath, nausea or vomiting, headaches, pain or numbness, during the period of continuous eligibility and ending altered mental state, depression, seizure, stroke/transient 30 days before the first observed brain metastasis diagnosis. ischemic attack, anxiety, vision disorder, drowsiness, The timing (measured in days) of the development of loss of appetite, cognitive impairment, speech problems, brain metastases relative to NSCLC diagnosis and crizoti- problems with memory, focal neurologic deficits (motor or nib treatment was captured for all eligible patients. This sensory), and changes in mood or personality . included the proportion of patients who developed brain metastases (a) on or prior to the date of the first observed Treatment patterns related to brain metastases NSCLC diagnosis, (b) after the first observed NSCLC diagnosis but prior to the date of the first observed For the analysis of treatment patterns, the observation prescription for crizotinib, (c) during crizotinib monother- period started on the day of the first observed brain metas- apy-treatment, and (d) after crizotinib monotherapy- tasis diagnosis and ended with the end of data availability treatment. The average time between the first observed or of continuous eligibility (Figure 1). During the observa- diagnosis of NSCLC and the first observed diagnosis of tion period, the following treatments were analyzed: brain metastases recorded in the database were also chemotherapy, crizotinib, other targeted therapies (e.g., reported among patients diagnosed with brain metastases bevacizumab, erlotinib), corticosteroids, radiation therapy after their first observed diagnosis for NSCLC. (including WBRT), SRS, brain surgery, and brachyther- apy. The number of patients with at least one of the above treatments was reported. In addition, the list of all types of Symptoms of brain metastases treatments received after brain metastasis diagnosis and For the analysis of symptoms frequently observed the treatment patterns of the above treatments before, in patients with brain metastasis, the observation during, and after crizotinib treatment, regardless of the period was divided into two periods: the pre- and sequence and the course of the treatment, was summarized. post-brain-metastasis-diagnosis periods. The pre- Since at the time of the study, crizotinib was the only brain-metastasis-diagnosis period was defined as a treatment available for patients with ALKþ NSCLC, maximum 1-year period starting during the period of con- the treatment patterns were expected to be influenced by tinuous eligibility and ending 30 days before the first the timing of the crizotinib course of treatment. Beginning of continuous First observed brain End of continuous eligibility or beginning of metastases eligibility or end of data data availability diagnosis availability 1 year 30 days Pre-brain-metastasis- Post-brain-metastasis- diagnosis period diagnosis period Definition of periods for the analyses of treatments Definition of periods for the symptoms, cost and HRU analyses Figure 1. Study design. 314 Burden of brain metastasis in ALKþ lung cancer Gue´ rin et al. www.informahealthcare.com/jme 2015 Informa UK Ltd Journal of Medical Economics Volume 18, Number 4 April 2015 Paents with at least one diagnosis for lung cancer Healthcare resource utilization and costs n = 1,352,449 Healthcare resource utilization and costs were each assessed during the pre- and the post-brain-metastasis- diagnosis periods described above. Specifically, the Adult paents with at least one prescripon fill for crizonib – this criteria was used as a proxy to idenfy paents with ALK+ NSCLC number of inpatient (IP) stays, number of IP days, n = 947 number of emergency room (ER) visits, number of out- patient (OP) visits, and number of other medical services (e.g., laboratory, radiology, patient’s home, hospice, or Paents without any medical claims associated with a clinical trial other ancillary services) were captured for each patient n = 753 and reported as per-patient-per-month (PPPM) values. Healthcare costs, measured from the payer’s perspective (i.e., including all charges reimbursed by payers— Paents with at least one diagnosis for brain metastases excluding the Medicare portion for patients 65 years n = 253 and older from the IMS LifeLink Health Plan Claims data- base), were adjusted for inflation and expressed in 2013 US dollars using the Consumer Price Index medical care Paents with at least 2 months of connuous eligibility before and component. Total healthcare costs were comprised of 1 month of connuous eligibility aer the date of the first observed pharmacy costs and medical service costs. Pharmacy diagnosis of brain metastases n = 213 costs included NSCLC-specific therapy costs (chemother- apy, targeted therapy, corticosteroids, or crizotinib) as well Figure 2. Sample selection. as costs for other therapies used for symptom management, and other pharmacy costs. Medical service costs included monotherapy. The median time from the first observed costs associated with IP stays, ER visits, OP visits, and NSCLC diagnosis to the first observed brain metastasis other medical services. Healthcare costs associated diagnosis among patients diagnosed with brain metastases with radiotherapy were also reported separately for each was 88 days. component of the medical costs. Only descriptive statistics were reported in all analyses. All statistical analyses were conducted using SAS 9.3 (SAS Institute Inc., Cary, NC). Frequent symptoms in patients with brain metastases Overall, the symptom burden was high both before and Results after the diagnosis of brain metastasis in crizotinib patients, Patient characteristics but patients experienced a substantially higher number of symptoms after brain metastasis diagnosis than before: the A total of 753 crizotinib patients were identified from a most frequent symptoms were fatigue (15.0% pre-diagnosis pool of 1,352,449 lung cancer patients; of these, 213 were and 39.0% post-diagnosis), shortness of breath (14.1% and diagnosed with brain metastases and satisfied all other 38.0%, respectively), nausea or vomiting (10.8% and selection criteria (Figure 2). The mean patient age for 33.3%, respectively), and headaches (4.7% and 23.9%, the study sample was 54.8 years old and 55.9% of patients respectively) (Figure 3). were female (Table 1). The study sample had an average CCI score of 2.7 (SD¼ 3.0). The most common comorbid- ities—other than cancer—in the patient sample were Treatment patterns among brain metastases chronic pulmonary disease (16.9%), diabetes (8.0%), patients and mild liver disease (4.2%). In total, 70.4% of patients (n¼ 150) were diagnosed After the diagnosis of brain metastases, 88.7% received with brain metastasis prior to receiving targeted therapy chemotherapy, 84.0% received crizotinib, 63.4% received for ALKþ NSCLC. About 23.0% (n¼ 49) of patients radiation therapy, and 31.9% had SRS at some point were diagnosed with brain metastasis at the time of the during the observation period. In addition to these anti- initial NSCLC diagnosis and 47.4% (n¼ 101) were cancer therapies, 77.5% of the study sample received cor- reported to have their first observed brain metastasis diag- ticosteroids (Table 2). A good portion (69%) of patients nosis following NSCLC diagnosis but prior to crizotinib received multiple treatments after the diagnosis of brain initiation. Similarly, 19.2% (n¼ 41) had their first metastases, regardless if they were already treated with observed brain metastasis diagnosis observed during crizo- crizotinib (Table 3). Although treatment combinations tinib monotherapy, and 10.3% (n¼ 22) after crizotinib varied, the most frequently used treatment combination 2015 Informa UK Ltd www.informahealthcare.com/jme Burden of brain metastasis in ALKþ lung cancer Gue´ rin et al. 315 Journal of Medical Economics Volume 18, Number 4 April 2015 Table 1. Patient characteristics. received radiotherapy only. Detailed treatment patterns are summarized in Table 3. Crizotinib patients with brain metastases (n¼ 213) Healthcare resource utilization and cost Crizotinib patients’ resource utilization increased after Age (years; mean (SD)) 54.8 (11.6) Female, n (%) 119 (55.9%) the diagnosis of brain metastasis, with OP visits increasing Region, n (%) from 1.37 visits PPPM pre-diagnosis to 3.25 visits PPPM Northeast 45 (21.1%) post-diagnosis, IP stays increasing from 0.04 stays PPPM South 81 (38.0%) West 42 (19.7%) pre-diagnosis to 0.25 stays PPPM post-diagnosis, and Midwest 42 (19.7%) patients having more IP days post-diagnosis (increasing Unknown 3 (1.4%) Data Source, n (%) from 0.25 days PPPM pre-diagnosis to 1.17 days PPPM Source Healthcare Analytics’ Source Lx 136 (63.8%) post-diagnosis) (Table 5). Patients also had an almost database 6-fold increase in the utilization of other medical services IMS LifeLink Health Plan Claims database 22 (10.3%) Truven Health Analytics MarketScan database 55 (25.8%) post-diagnosis, increasing from 0.19 PPPM pre-diagnosis Year of first observed brain metastasis diagnosis, n (%) to 1.09 visits PPPM in the post-diagnosis period. 2007 1 (0.5%) Higher resource utilization resulted in higher 2008 5 (2.3%) 2009 2 (0.9%) healthcare costs, as costs rose in the post-brain 2010 8 (3.8%) metastasis-diagnosis period compared to the pre-brain- 2011 94 (44.1%) metastasis-diagnosis period: patients had an average total 2012 95 (44.6%) 2013 8 (3.8%) monthly cost of $5983 in the pre-brain-metastasis-diagno- Charlson Comorbidity Index, Mean (SD) 2.66 (2.97) sis period vs $22,645 in the post-brain-metastasis-diagnosis Charlson Comorbidity Index Components, n (%) period (Figure 4). Furthermore, in the pre-brain-metasta- Metastatic Carcinoma 63 (29.6%) Cancer 46 (21.6%) sis-diagnosis period, 79% of patients had total healthcare Chronic Pulmonary Disease 36 (16.9%) costs between $0–$10,000, compared to 26% of patients Diabetes 17 (8.0%) in the post-brain-metastasis-diagnosis period (Figure 5). Mild Liver Disease 9 (4.2%) Congestive Heart Failure 6 (2.8%) Post-diagnosis, medical costs increased 4-fold and repre- Cerebrovascular Disease 6 (2.8%) sented a higher portion of the total healthcare costs (58%), Peripheral Vascular Disease 4 (1.9%) with the cost of hospitalization and associated procedures Myocardial Infarction 3 (1.4%) Connective Tissue Disease/Rheumatic 2 (0.9%) dominating this cost category (IP costs were 51% of the Disease medical care costs). Pharmacy costs increased more than Renal Disease 2 (0.9%) 3-fold in the post-diagnosis period. In the pre-brain-metas- Dementia 1 (0.5%) Peptic Ulcer Disease 1 (0.5%) tasis-diagnosis period, crizotinib cost accounted for 52.7% Hemiplegia or Paraplegia 1 (0.5%) of the total pharmacy costs ($2924), while in the post- Moderate-to-Severe Liver Disease 1 (0.5%) brain-metastasis-diagnosis period, crizotinib costs AIDS/HIV (Excluding Asymptomatic – HIV Infection) accounted for 68.3% of the total pharmacy costs Treatments before Brain Metastasis ($9520). Radiotherapy was mainly performed on an Diagnosis, n (%) OP basis, and was responsible for 43% of the OP costs. Chemotherapy 46 (21.6%) Targeted Therapy (excluding crizotinib) 21 (9.9%) Pre-diagnosis, 51% of total costs were due to medical Crizotinib 53 (24.9%) care costs (with OP costs representing the majority— Radiotherapy 32 (15.0%) 59%—of these costs), and the remainder was due to the Brain Surgery 0 (0.0%) Brachytherapy 1 (0.5%) cost of pharmacological treatment. consisted of crizotinibþ radiotherapy (18.3%). Among Discussion patients who developed brain metastases before initiating crizotinib (n¼ 150), 16.7% received crizotinib only and The ALKþ translocation confers special clinicopathologic 21.3% received crizotinib and radiotherapy after the diag- features to the NSCLC cancer sub-type: ALKþ NSCLC nosis of brain metastases. Among patients who developed patients tend to be diagnosed with lung cancer at a younger 14,15,19,21,40 brain metastases while on crizotinib (n¼ 41), 31.7% con- age , and have a higher risk of developing brain tinued on crizotinib monotherapy while 17.1% added or (and other) metastases than patients with other NSCLC 20,21,41,42 switched to radiotherapy. Finally, among patients who sub-types . Brain metastases are known to severely developed brain metastases after discontinuing crizotinib impact a patient’s quality-of-life ( physical, cognitive, and 24,31,33 (n¼ 22), 63.6% of them remained untreated and 13.6% functional impairments) , as well as to result in high 316 Burden of brain metastasis in ALKþ lung cancer Gue´ rin et al. www.informahealthcare.com/jme 2015 Informa UK Ltd Journal of Medical Economics Volume 18, Number 4 April 2015 45% Pre-brain-metastasis-diagnosis period 39.0% 40% 38.0% Post-brain-metastasis-diagnosis period 35% 33.3% 30% 23.9% 25% 20% 16.9% 15.0% 14.6% 14.1% 15% 13.1% 12.2% 10.8% 9.9% 10% 8.5% 8.5% 6.1% 5.6% 5.2% 5.2% 4.7% 4.7% 4.7% 3.8% 5% 2.3% 2.3% 1.9% 1.4% 1.4% 1.4% 0.9% 0.9% 0.5% 0.5% 0.5% 0.0% 0.0% 0.0% 0% Figure 3. Symptoms during the pre- and post-brain-metastasis-diagnosis periods. Table 2. Treatments received after brain metastasis diagnosis. burden, resource use, and costs in ALKþ NSCLC patients (identified by the use of crizotinib). Patients with brain Crizotinib patients with metastases experienced a high symptom burden, particu- brain metastases (n¼ 213) larly post-diagnosis of these metastases: more than a Treatments, n (%) n % third suffered from fatigue, shortness of breath, nausea/ vomiting, and almost a quarter suffered from headaches Non-Systemic treatments Radiation Therapy 135 63.4% post-diagnosis. Prior studies also report that neuro- Stereotactic Radiosurgery 68 31.9% 44,45 logical symptoms are frequently encountered (in 36– Brain Surgery 19 8.9% 44–46 Brachytherapy 1 0.5% 82% of cases) in lung patients with brain metastases. Systemic treatments Temporary relief from symptoms can be achieved with Crizotinib 179 84.0% 22,45,47–49 treatment , but not all patients will respond to Targeted Therapy 43 20.2% Chemotherapy 189 88.7% treatment , and resistance to treatment eventually Supportive treatments occurs . Corticosteroids 165 77.5% In this study, we observed that, after being diagnosed No treatment reported 19 8.9% with brain metastases, most patients received several treat- Patients may have received multiple types of treatments. Therefore, counts ments—chemotherapy, crizotinib, corticosteroids, and are not mutually exclusive of one another. radiation therapy—but treatment combinations and pat- Targeted therapies received after brain metastasis diagnosis included bev- acizumab, bortezomib, and erlotinib. terns remained highly variable among the sample studied. According to the NCCN guidelines, radiation therapy for 28–30 brain metastases and crizotinib for advanced ALKþ 37 43 50 healthcare resource utilization and substantial clinical , NSCLC are the de facto treatment options for most 27 32 ALKþ patients with or without brain metastases, but, for economic , and caregiver burden. In this study, we found that the development of brain intolerant or resistant patients, there are several recom- metastases is associated with a substantial treatment mended treatment options, with no clear hierarchical 2015 Informa UK Ltd www.informahealthcare.com/jme Burden of brain metastasis in ALKþ lung cancer Gue´ rin et al. 317 Journal of Medical Economics Volume 18, Number 4 April 2015 Table 3. All treatments received after brain metastasis diagnosis . Crizotinib patients with brain metastases (n¼ 213), n (%) Period after brain metastasis diagnosis (days), mean SD [median] 333.3 298.9 [251] Crizotinib and radiotherapy 39 (18.3%) Crizotinib only 38 (17.8%) No treatment reported 19 (8.9%) Crizotinib, chemotherapy, and radiotherapy 14 (6.6%) Crizotinib, chemotherapy, radiotherapy, and targeted therapy 13 (6.1%) Crizotinib, radiotherapy, and stereotactic radiosurgery 13 (6.1%) Crizotinib and chemotherapy 10 (4.7%) Crizotinib, chemotherapy, radiotherapy, and stereotactic radiosurgery 10 (4.7%) Crizotinib, chemotherapy, radiotherapy, stereotactic radiosurgery, and targeted therapy 8 (3.8%) Crizotinib, chemotherapy, and targeted therapy 7 (3.3%) Radiotherapy only 7 (3.3%) Crizotinib, radiotherapy, and targeted therapy 6 (2.8%) Crizotinib, brain surgery, radiotherapy, and stereotactic radiosurgery 5 (2.3%) Chemotherapy and radiotherapy 3 (1.4%) Crizotinib, brain surgery, chemotherapy, and radiotherapy 3 (1.4%) Crizotinib, brain surgery, chemotherapy, radiotherapy, stereotactic radiosurgery, and targeted therapy 3 (1.4%) Chemotherapy, radiotherapy, and targeted therapy 2 (0.9%) Crizotinib, brain surgery, chemotherapy, radiotherapy, and stereotactic radiosurgery 2 (0.9%) Crizotinib, brain surgery, chemotherapy, radiotherapy, and targeted therapy 2 (0.9%) Crizotinib, brain surgery, and radiotherapy 2 (0.9%) Crizotinib and stereotactic radiosurgery 2 (0.9%) Brain surgery, chemotherapy, and radiotherapy 1 (0.5%) Chemotherapy, radiotherapy, stereotactic radiosurgery, and targeted therapy 1 (0.5%) Crizotinib, brachytherapy, chemotherapy, radiotherapy, and stereotactic radiosurgery 1 (0.5%) Crizotinib, brain surgery, and targeted therapy 1 (0.5%) Stereotactic radiosurgery only 1 (0.5%) The list of treatments does not take into account the sequence of treatment. There may also be several courses of treatment within the same list of treatments. $8,000 Pre-brain-metastasis-diagnosis $9,520 $6,692 Period $7,000 Post-brain-metastasis-diagnosis Period $5,877 $6,000 Radiotherapy-related Costs $5,000 $4,000 $3,000 $2,924 $1,813 $2,000 $1,174 $1,000 $507 $49 $61 $11 $0 Inpaent Costs Outpaent Costs Emergency Room Other Medical Pharmacy Costs Service Costs Figure 4. Healthcare costs during the pre- and post-brain-metastasis-diagnosis periods. path (for all steps subsequent to crizotinib treatment, the not unexpected, and a similar use of systemic therapy NCCN recommendations are based upon lower-level evi- (65%), radiation therapy (47%), and corticosteroids dence; there is NCCN consensus that the intervention is (79%) has been noted previously in a retrospective study 50 27 appropriate ). Thus, the observed treatment variability is of lung cancer patients with brain metastases . 318 Burden of brain metastasis in ALKþ lung cancer Gue´ rin et al. www.informahealthcare.com/jme 2015 Informa UK Ltd Journal of Medical Economics Volume 18, Number 4 April 2015 90% 80% 70% 60% Pre-brain-metastasis-diagnosis Period 50% Post-brain-metastasis-diagnosis Period 40% 30% 20% 10% 0% 0 - 10K 10K - 20K 20K - 30K 30K - 40K 40K - 50K 50K - 60K 60K - 70K 70K - 80K 80K - 90K 90K + Heathcare Costs PPPM Range Figure 5. Distribution of total healthcare costs PPPM. With the increase in symptoms and the variety of treat- distribution is that using targeted agents for the treatment ments post-diagnosis of brain metastases, the current study of (brain) metastatic ALKþ NSCLC increases pharmacy found that the already high economic burden of treatment costs, but reduces the use of inpatient services, possibly pre-diagnosis of brain metastases drastically increased after indicating that the disease has been managed (for a diagnosis. The average monthly costs per patient almost period of time). Another study evaluating the costs of quadrupled after the diagnosis of brain metastases, with treating brain-metastatic cancer patients (of which 55% monthly costs per patient averaging $5983 (interquartile were lung cancer patients) also found that brain metastases are associated with a significant economic burden, con- range: $118–$8667) vs $22,645 (interquartile range: cluding that the total costs per year of life amounted to $9452–$26,693) for the pre- and post-brain metastasis per- $62,130 (where less than 25% of the patients survived iods, respectively. The estimate reported by Ray et al. of the first year of treatment) . post-diagnosis monthly costs was similar to the mean total This study has demonstrated that the development of $17,007 for all-cause healthcare costs (PPPM) reported of brain metastases, common in ALKþ NSCLC, can be for lung cancer patients, of all types, with newly diagnosed costly and can present a substantial clinical burden. brain metastases. However, there are several differences in Further analyses are warranted to explore how treatment how costs are distributed for brain-metastasis diagnosed in patterns and the burden are impacted post-crizotinib, now crizotinib patients vs in lung cancer patients: crizotinib that a new therapy is available for patients who do not patients had much higher pharmacy costs post-diagnosis respond to crizotinib. of brain metastases ($9520 vs $809 PPPM), but lower med- ical costs ($13,125 vs $16,198 PPPM, respectively), with similar IP stays, but fewer ER visits (0.04 visits vs 0.21 visits Limitations PPPM, respectively) and more OP visits (3.25 visits vs 2.8 visits PPPM, respectively). It must be noted that crizotinib The current study was subject to some limitations. First, was granted accelerated regulatory approval in the US in patients were selected based on several proxy indicators of August 2011, and erlotinib was approved for NSCLC the type of the disease, as there was no ICD-9 CM code to treatment in April 2010, hence these agents were not identify ALKþ and/or NSCLC patients. A prescription fill available at the time Ray et al. conducted their study. for crizotinib was used as a proxy to identify patients with In addition, since ALKþ NSCLC has a relatively rare ALKþ NSCLC among patients with a diagnosis of lung rate of occurrence (2–8%) among all NSCLC lung cancer. Thus, the findings of the study may not be repre- 11–19 cancers , the treatment patterns reported by Ray sentative of ALKþ NSCLC patients who did not receive et al. are likely not indicative of this type of cancer. crizotinib. Second, patients were not required to have A probable explanation for the differences in cost more than one diagnosis of brain metastases. Thus, their 2015 Informa UK Ltd www.informahealthcare.com/jme Burden of brain metastasis in ALKþ lung cancer Gue´ rin et al. 319 Percentage of Paents per Cost Category Journal of Medical Economics Volume 18, Number 4 April 2015 Table 4. Treatment patterns after brain metastasis diagnosis relative to crizotinib treatment . Crizotinib patients with brain metastases (N¼ 213) Among patients who developed brain metastases before initiating crizotinib N¼ 150 Patients who received crizotinib first following NSCLC diagnosis (n¼ 30) Crizotinib only 25 (16.7) And radiotherapy 3 (2.0) And chemotherapy 1 (0.7) And chemotherapy, radiotherapy, and SRS 1 (0.7) Patients who received at least one other treatment before crizotinib (n¼ 120) Radiotherapy only 29 (19.3) Radiotherapyþ chemotherapy 11 (7.3) Radiotherapyþ chemotherapyþ targeted therapy 11 (7.3) Radiotherapyþ targeted therapy 6 (4.0) Radiotherapyþ SRS 10 (6.7) Radiotherapyþ SRSþ chemotherapy 5 (3.3) Radiotherapyþ SRSþ chemotherapyþ targeted therapy 6 (4.0) Radiotherapyþ SRSþ chemotherapyþ brachitherapy 1 (0.7) Chemotherapy only 9 (6.0) Chemotherapyþ targeted therapy 7 (4.7) Brain surgeryþ radiotherapy 2 (1.3) Brain surgeryþ radiotherapyþ SRS 4 (2.7) Brain surgeryþ radiotherapyþ SRSþ chemotherapy 2 (1.3) Brain surgeryþ radiotherapyþ SRSþ chemotherapyþ targeted therapy 3 (2.0) Brain surgeryþ radiotherapyþ chemotherapy 2 (1.3) Brain surgeryþ radiotherapyþ chemotherapyþ targeted therapy 2 (1.3) Brain surgeryþ SRS 1 (0.7) Brain surgeryþ targeted therapy 1 (0.7) SRSþ chemotherapy 5 (3.3) SRS only 3 (2.0) Among patients who developed brain metastases while on crizotinib N¼ 41 Patients who continued crizotinib (n¼ 28) Crizotinib monotherapy 13 (31.7) Crizotinibþ radiotherapy/SRS/brain surgery/ 10 (24.4) Added or switched to radiotherapy 7 (17.1) Added SRS 1 (2.4) Added radiotherapyþ SRS 1 (2.4) Added radiotherapyþ brain surgery, then continued with targeted therapyþ chemotherapy 1 (2.4) Crizotinibþ chemotherapyþ radiotherapy 1 (2.4) Switched to radiotherapy, chemotherapy or other targeted therapy 5 (12.2) Switched to chemotherapy 2 (4.9) Switched to other targeted therapy and chemotherapy 2 (4.9) Switched to radiotherapy, þtargeted therapy, þchemotherapy 1 (2.4) Patients who discontinued crizotinib (n¼ 12) Remained untreated 5 (12.2) Received radiotherapy only 4 (9.8) Received radiotherapyþ chemotherapy 1 (2.4) Received radiotherapyþ targeted therapyþ chemotherapy 1 (2.4) Received radiotherapyþ brain surgeryþ targeted therapyþ chemotherapy 1 (2.4) Among patients who developed brain metastases after discontinuing crizotinib N¼ 22 Remained untreated 14 (63.6) Received radiotherapy only 3 (13.6) Received radiotherapyþ chemotherapyþ targeted therapy 2 (9.1) Received radiotherapyþ chemotherapy 1 (4.5) Received radiotherapyþ SRSþ chemotherapyþ targeted therapy 1 (4.5) Received SRS only 1 (4.5) SRS, stereotactic radiosurgery. The list of treatments does not take into account the sequence of treatment. There may also be several courses of treatment within the same list of treatments. brain metastases were not confirmed with at least two estimation of costs experienced by patients covered by independent diagnoses. Third, information needed to Medicare: only the MarketScan database had linked infor- identify ER visits in the Source Healthcare Analytics mation on previous employees now covered by Medicare, Source Lx database was not available, thus the use of thus this study had very limited information on the ER services, as well as associated costs, may be under- Medicare-covered population, and its findings may not estimated in this study. A similar limitation arises in the be immediately generalizable to non-privately-insured 320 Burden of brain metastasis in ALKþ lung cancer Gue´ rin et al. www.informahealthcare.com/jme 2015 Informa UK Ltd Journal of Medical Economics Volume 18, Number 4 April 2015 Table 5. Healthcare resource utilization pre- and post-brain-metastasis-diagnosis period. Healthcare resource utilization (PPPM), mean SD Pre-brain-metastasis-diagnosis period (n¼ 213) Post-brain-metastasis-diagnosis period (n¼ 213) IP stays 0.04 0.09 0.25 0.34 IP days 0.25 1.06 1.17 1.92 OP 1.37 1.73 3.25 2.47 ER 0.02 0.07 0.04 0.13 Other medical services 0.19 0.54 1.09 2.00 PPPM, per-patient-per-month; IP, inpatient; OP, outpatient; ER, emergency room. A. Gue ´rin, K. Dea, R. Nitulescu, and E. Q. Wu are employees of populations. Moreover, since information on different cost Analysis Group Inc., which has received consultancy fees from components (e.g., co-ordination of benefits, out of pocket Novartis. costs) was not available in the Source Lx database, a common measure of costs was used for all three databases. Acknowledgments Fourth, only symptoms that required medical attention The authors thank Ana Bozas, PhD, an employee of Analysis were captured in the study. Fifth, this study examined Group, for writing and editorial support. A synopsis of this only direct medical costs, evaluated from the payers’ per- study’s results has been recently accepted as a poster presentation spective. Information to determine indirect costs, such as at the 19th annual meeting of the Society of Neuro-Oncologists lost productivity and burden to caregivers, was not avail- (SNO) in Miami, Florida from November 13th to 16th (Poster able. Given the high comorbidity burden of the study’s BM-19). patients (general poor health) and frequent hospitaliza- tions, we expect these undetermined indirect costs to be substantial. Further studies are warranted to investigate References this aspect of the economic burden of the ALKþ 1. American Cancer Society. Lung Cancer (Non-Small Cell). Atlanta, GA: NSCLC. Sixth, the treatment choices analyzed in this American Cancer Society; 2014, 77. http://www.cancer.org/acs/groups/con- study do not include ceritinib, the newly-approved tent/@research/documents/document/acspc-041770.pdf. Accessed August second generation tyrosine kinase inhibitor for the treat- 9, 2013 ment of ALKþ NSCLC after progression on crizotinib, as 2. Howlader N, Noone A, Krapcho M, et al., eds. SEER Cancer Statistics Review 1975-2010. Bethesda, MD: National Cancer Institute; 2010. http://seer.can- at the time the study was conducted ceritinib was not yet cer.gov/csr/1975_2010/. Accessed June 6, 2013 available. Finally, further studies are warranted to study 3. Larsen JE, Cascone T, Gerber DE, et al. Targeted therapies for lung cancer: the healthcare resource utilization and costs associated clinical experience and novel agents. Cancer J 2011;17:512-27 with the development of brain metastases in ALKþ 4. Custodio A, Mendez M, Provencio M. Targeted therapies for advanced non- NSCLC patients using more recent data, other data small-cell lung cancer: current status and future implications. Cancer Treat Rev 2012;38:36-53 sources, or different study designs. 5. Bang Y-J. Treatment of ALK-positive non-small cell lung cancer. Arch Pathol Lab Med 2012;136:1201-4 6. Li Y, Ye X, Liu J, et al. Evaluation of EML4-ALK fusion proteins in non-small Conclusions cell lung cancer using small molecule inhibitors. Neoplasia 2011;13:1-11 7. Shaw AT, Solomon B. Targeting anaplastic lymphoma kinase in lung cancer. In this study we found that brain metastases present a sub- Clin Cancer Res 2011;17:2081-6 stantial burden in the ALKþ NSCLC patient population 8. Kim T-J, Park CK, Yeo CD, et al. Simultaneous diagnostic platform of geno- (crizotinib patients) and significant variability in the treat- typing EGFR, KRAS, and ALK in 510 Korean patients with non-small-cell lung ment options was observed after the patient was diagnosed cancer highlights significantly higher ALK rearrangement rate in advanced with brain metastasis. Healthcare utilization and costs stage. J Surg Oncol 2014;110(3):245-51. published online May 29, 2014, doi: 10.1002/jso.23646 after brain metastasis diagnosis were substantially greater 9. Park J, Yamaura H, Yatabe Y, et al. Anaplastic lymphoma kinase gene than before diagnosis. Given the large percentage of rearrangements in patients with advanced-stage non-small-cell lung ALKþ NSCLC patients who will eventually develop cancer: CT characteristics and response to chemotherapy. Cancer Med brain metastases, this study may highlight an important 2013;3:118-23 unmet need. 10. Chun SG, Choe KS, Iyengar P, et al. Isolated central nervous system progres- sion on Crizotinib: an Achilles heel of non-small cell lung cancer with EML4- ALK translocation? Cancer Biol Ther 2012;13:1376-83 11. Soda M, Choi YL, Enomoto M, et al. Identification of the transforming EML4- Transparency ALK fusion gene in non-small-cell lung cancer. Nature 2007;448:561-6 Declaration of funding 12. Koivunen JP, Mermel C, Zejnullahu K, et al. EML4-ALK fusion gene and This study was funded by Novartis. M. Sasane, J. Zhang, and K. efficacy of an ALK kinase inhibitor in lung cancer. Clin Cancer Res Culver are employees of Novartis and own stock/stock options. 2008;14:4275-83 2015 Informa UK Ltd www.informahealthcare.com/jme Burden of brain metastasis in ALKþ lung cancer Gue´ rin et al. 321 Journal of Medical Economics Volume 18, Number 4 April 2015 13. Boland JM, Erdogan S, Vasmatzis G, et al. Anaplastic lymphoma 32. Schubart JR, Kinzie MB, Farace E. Caring for the brain tumor patient: family kinase immunoreactivity correlates with ALK gene rearrangement and tran- caregiver burden and unmet needs. Neuro Oncol 2008;10:61-72 scriptional up-regulation in non-small cell lung carcinomas. Hum Pathol 33. Welzel G, Fleckenstein K, Schaefer J, et al. Memory function before and 2009;40:1152-8 after whole brain radiotherapy in patients with and without brain metastases. 14. Shaw AT, Yeap BY, Mino-Kenudson M, et al. Clinical features and outcome Int J Radiat Oncol Biol Phys 2008;72:1311-18 34. Soffietti R, Ruda R, Trevisan E. Brain metastases: current management and of patients with non-small-cell lung cancer who harbor EML4-ALK. J Clin new developments. Curr Opin Oncol 2008;20:676-84 Oncol 2009;27:4247-53 35. Ricciardi S, de Marinis F. Multimodality management of non-small cell lung 15. Wong DW-S, Leung EL-H, So KK-T, et al. The EML4-ALK fusion gene cancer patients with brain metastases. Curr Opin Oncol 2010;22:86-93 is involved in various histologic types of lung cancers from nonsmokers 36. Narita Y, Shibui S. Strategy of surgery and radiation therapy for brain metas- with wild-type EGFR and KRAS. Cancer 2009;115:1723-33 tases. Int J Clin Oncol 2009;14:275-80 16. Inamura K, Takeuchi K, Togashi Y, et al. EML4-ALK fusion is linked to 37. Nieder C, Norum J, Stemland JG, et al. Resource utilization in patients with histological characteristics in a subset of lung cancers. J Thorac Oncol brain metastases managed with best supportive care, radiotherapy and/or 2008;3:13-17 surgical resection: a Markov analysis. Oncology 2010;78:348-55 17. Takeuchi K, Choi YL, Soda M, et al. Multiplex reverse transcription-PCR 38. Takura T, Hayashi M, Muragaki Y, et al. [Study on medical economic evalu- screening for EML4-ALK fusion transcripts. Clin Cancer Res ation methods for metastatic brain tumors therapy]. No Shinkei Geka 2008;14:6618-24 2010;38:629-37 18. Perner S, Wagner PL, Demichelis F, et al. EML4-ALK fusion lung cancer: a 39. Quan H, Sundararajan V, Halfon P, et al. Coding algorithms and for rare acquired event. Neoplasia 2008;10:298-302 defining comorbidities data in ICD-9-CM administrative. Med Care 19. Rodig SJ, Mino-Kenudson M, Dacic S, et al. Unique clinicopathologic features 2005;43:1130-9 characterize ALK-rearranged lung adenocarcinoma in the western population. 40. Zhang N-N, Liu Y-T, Ma L, et al. The molecular detection and clinical signifi- Clin Cancer Res 2009;15:5216-23 cance of ALK rearrangement in selected advanced non-small cell lung cancer: 20. Doebele RC, Lu X, Sumey C, et al. Oncogene status predicts patterns of meta- ALK expression provides insights into ALK targeted therapy. PLoS One static spread in treatment-naive nonsmall cell lung cancer. Cancer 2014;9:e84501 2012;118:4502-11 41. Paik JH, Choi C-M, Kim H, et al. Clinicopathologic implication of ALK 21. Kang HJ, Lim H-J, Park JS, et al. Comparison of clinical characteristics rearrangement in surgically resected lung cancer: a proposal of diagnostic between patients with ALK-positive and EGFR-positive lung adenocarcinoma. algorithm for ALK-rearranged adenocarcinoma. Lung Cancer 2012;76:403-9 Respir Med 2014;108:388-94 42. Fallet V, Cadranel J, Doubre H, et al. Prospective screening for ALK: 22. Shaw AT, Kim D-W, Nakagawa K, et al. Crizotinib versus chemotherapy in clinical features and outcome according to ALK status. Eur J Cancer advanced ALK-positive lung cancer. N Engl J Med 2013;368:2385-94 2014;50:1239-46 23. Kim D-W, Mehra R, Tan DS-W, et al. Ceritinib in Advanced Anaplastic 43. Sørensen JB, Hansen HH, Hansen M, Dombernowsky P. Brain metastases in Lymphoma Kinase (ALK)-rearranged (ALKþ) Non-small Cell Lung Cancer adenocarcinoma of the lung: frequency, risk groups, and prognosis. J Clin (NSCLC) – Results of the ASCEND-1 Trial (#8003). J Clin Oncol Oncol 1988;6:1474-80 2014;32:5s:abstr 8003 44. Jena A, Taneja S, Talwar V, et al. Magnetic resonance (MR) patterns of brain 24. Wen PY, Loeffler JS. Overview of the clinical manifestations, diagnosis, and metastasis in lung cancer patients: correlation of imaging findings with symp- management of patients with brain metastases. Wolter Kluwer Health; 2013, tom. J Thorac Oncol 2008;3:140-4 p 1–15. UpToDate, Inc., 95 Sawyer Rd., Waltham, MA 02453 45. Komatsu T, Kunieda E, Oizumi Y, et al. Clinical characteristics of brain 25. Gofton TE, Graber J, Carver A. Identifying the palliative care needs of patients metastases from lung cancer according to histological type: pretreatment living with cerebral tumors and metastases: a retrospective analysis. evaluation and survival following whole-brain radiotherapy. Mol Clin Oncol J Neurooncol 2012;108:527-34 2013;1:692-8 26. Bauml J, Mick R, Zhang Y, et al. Determinants of survival in advanced non– 46. Ferrigno D, Buccheri G. Cranial computed tomography as a part of the small-cell lung cancer in the era of targeted therapies. Clin Lung Cancer initial staging procedures for patients with non-small-cell lung cancer. 2013;14:581-91 Chest 1994;106:1025-9 27. Ray S, Dacosta-Byfield S, Ganguli A, et al. Comparative analysis of survival, 47. Haura EB. Treatment of advanced non-small-cell lung cancer: a review of treatment, cost and resource use among patients newly diagnosed with brain current randomized clinical trials and an examination of emerging therapies. metastasis by initial primary cancer. J Neurooncol 2013;114:117-25 Cancer Control 2001;8:326-36 28. Patchell RA. The management of brain metastases. Cancer Treat Rev 48. Frampton JE. Crizotinib: a review of its use in the treatment of anaplastic 2003;29:533-40 lymphoma kinase-positive, advanced non-small cell lung cancer. Drugs 29. Taimur S, Edelman MJ. Treatment options for brain metastases in patients 2013;73:2031-51 with non-small-cell lung cancer. Curr Oncol Rep 2003;5:342-6 49. Wu C, Li YL, Wang ZM, et al. Gefitinib as palliative therapy for 30. Lu-Emerson C, Eichler AF. Brain metastases. Continuum (Minneap Minn) lung adenocarcinoma metastatic to the brain. Lung Cancer 2007;57: 2012;18:295-311 359-64 31. Tang V, Rathbone M, Park Dorsay J, et al. Rehabilitation in primary 50. National Comprehensive Cancer Network. Non-small cell lung cancer, ver- and metastatic brain tumours: impact of functional outcomes on survival. sion 4. 2014. Fort Washington, PA: National Comprehensive Cancer Network, J Neurol 2008;255:820-7 2014. p 1-147 322 Burden of brain metastasis in ALKþ lung cancer Gue´ rin et al. www.informahealthcare.com/jme 2015 Informa UK Ltd http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Journal of Medical Economics Taylor & Francis

Brain metastases in patients with ALK+ non-small cell lung cancer: clinical symptoms, treatment patterns and economic burden

Loading next page...
 
/lp/taylor-francis/brain-metastases-in-patients-with-alk-non-small-cell-lung-cancer-hpVvZlKhFx

References (52)

Publisher
Taylor & Francis
Copyright
© 2015 All rights reserved: reproduction in whole or part not permitted
ISSN
1941-837X
eISSN
1369-6998
DOI
10.3111/13696998.2014.1003644
pmid
25565443
Publisher site
See Article on Publisher Site

Abstract

Annie Guerin Analysis Group, Inc., Montreal, QC, Canada Objective: Medha Sasane Brain metastases (BM) are highly prevalent among anaplastic lymphoma kinase positive (ALKþ) non-small Jie Zhang cell lung cancer (NSCLC) patients; yet little is known about their real-world treatment patterns and clinical Kenneth W. Culver and economic burdens. This study aimed to describe these patients’ treatment patterns, symptoms, and Novartis Pharmaceuticals Corporation, East Hanover, costs. NJ, USA Research design and methods: Katherine Dea Retrospective study pooling data from three large administrative databases in the US (08/2011–06/2013). Roy Nitulescu ALKþ NSCLC patients with BM and continuous enrollment for 60 days before and30 days after the first Analysis Group, Inc., Montreal, QC, Canada observed BM diagnosis were identified by pharmacy records for crizotinib among patients with lung cancer Eric Qiong Wu and BM diagnostic codes. Analysis Group, Inc., Boston, MA, USA Main outcome measures: Address for correspondence: Treatment patterns, symptoms, healthcare resource utilization, and costs, before and after BM diagnosis. Annie Gue ´ rin, 1000 de la Gauchetie ` re Ouest, Bureau 1200, Montre ´ al, QC, H3B 4W5 Canada. Results: Tel: 514-394-4484; Fax: 514-394-4461; Of the 213 crizotinib patients with BM diagnoses meeting the selection criteria, 23.0% had BM prior to annie.guerin@analysisgroup.com NSCLC diagnosis; 47.4% had BM prior to crizotinib initiation; 19.2% during crizotinib treatment; and 10.3% post-crizotinib treatment. For those diagnosed with BM after NSCLC diagnosis, the median time between Keywords: the NSCLC and BM diagnoses was 88 days. Following the first observed BM diagnosis, 88.7% used Non-small-cell lung cancer – Brain neoplasms – chemotherapy, 63.4% had radiotherapy, and 31.9% had stereotactic radiosurgery. The prevalence Neoplasm metastasis – Costs and cost analysis – of BM-related symptoms substantially increased post-BM-diagnosis: fatigue (from 15% to 39%), Crizotinib – Symptoms headaches (from 5% to 24%), and depression (from 5% to 15%). Monthly costs per patient averaged Accepted: 29 December 2014; published online: 6 February 2015 $5983 before the BM diagnosis and $22,645 after diagnosis. Patients’ resource utilization increased Citation: J Med Econ 2015; 18:312–22 significantly post-BM-diagnosis, with a 3-fold increase in OP visits and a 6-fold increase in IP stays. Post-BM-diagnosis costs were driven by pharmacy (42.0%), inpatient (29.6%), and outpatient costs (26.0%). Limitations: The study sample was limited to crizotinib-treated patients. Conclusions: Post-BM-diagnosis, patients experience high symptom burden. Post-BM-diagnosis, treatment is highly variable and costly: average monthly costs per patient almost quadrupled post-BM-diagnosis. 312 Burden of brain metastasis in ALKþ lung cancer Gue´ rin et al. www.informahealthcare.com/jme 2015 Informa UK Ltd Journal of Medical Economics Volume 18, Number 4 April 2015 Introduction Methods Non-small-cell lung cancer (NSCLC) accounts for more Data sources than 85% of all lung cancers in the US . A large majority This study combined data from three large retrospective of NSCLC patients are diagnosed at an advanced stage of administrative-claims databases (Source Healthcare the disease and, as a result, experience limited success with Analytics’ Source Lx database for June 2011–June 2013, treatment. Only a small fraction of patients (i.e., 4% of IMS LifeLink Health Plan Claims database for January metastatic and 18% of all NSCLC) survive more than 2001–March 2013, and Truven Health Analytics 5 years post-diagnosis . Growing understanding of the MarketScan database for January 2002–September 2012) molecular pathogenesis of lung cancer has led to the devel- to extract patient data. The databases contain enrollment opment of targeted therapies in an effort to improve 3,4 records, medical claims ( physician and facility), and phar- patient survival . In particular, tumors with activating macy claims. All census regions are represented in all three mutations of select oncogenes, such as the fusion of the databases. kinase domain of the anaplastic lymphoma kinase (ALK) The Source Lx database contains linked healthcare to the amino-terminal section of the echinoderm micro- data for more than 156 million annually covered lives, tubule-associated protein-like 4 (EML4) gene product, 5 derived from three basic sources: pharmacy point-of-ser- have been found to be susceptible to targeted therapies . vice; switch/network transactions; and additional direct The EML4-ALK fusion results in the constitutive activa- prescription, medical, and hospital claims data. 6,7 tion of a crucial cell-division-cascade ; thus, these tumors The IMS LifeLink Health Plan Claims database 8,9 tend to have an aggressive outlook with a high chance of contains combined claims for over 100 healthcare plans, early relapse . Patients with ALKþ NSCLC, which rep- representing 42 million annually covered lives. 11–19 resent 2–8% of all patients diagnosed with NSCLC , The Truven Health Analytics MarketScan database tend to have a more advanced stage of the disease and contains combined claims for more than 130 employers multiple metastatic sites at diagnosis compared to other that sponsor private health insurance, representing 40 NSCLC types . million annually covered lives. In addition, the The treatment of ALKþ NSCLC is often complex due MarketScan Medicare Supplemental and Coordination to the frequent occurrence of brain metastases: initial of Benefits (Medicare Supplemental) database profiles occurrence of brain metastasis has been reported in the healthcare experience of retirees and their dependents 20–22 15–35% of cases , and, over the course of first-line with Medicare supplemental insurance paid for by therapy either with crizotinib or chemotherapy the fre- employers. quency of brain metastases can increase to 60% . Brain All data used in this study had been de-identified metastases are associated with many complications: some according to the patient confidentiality requirements of of these are neurocognitive, psychological, physical the Health Insurance Portability and Accountability Act 24,25 impairments , severe comorbidity, and a decreased (HIPAA) and did not require approval from the 26,27 life expectancy . Treatment of brain metastases usually Institutional Review Board or a waiver of authorization involves radiotherapy (either whole-brain radiotherapy for use. [WBRT] or stereotactic radiosurgery [SRS]) and may involve surgical resection (for isolated metastases or Study design cases with a good prognosis) and/or chemotherapy (for 28–30 salvage or experimental cases) . In addition, post- For this retrospective cross-sectional cohort study, patients treatment, rehabilitation therapy may be needed ,as who met the following selection criteria were selected: (1) well as help with daily living tasks for recovering 18 years of age; (2) 1 diagnosis for lung cancer (ICD-9 32,33 patients . Thus, disease management and treatment CM: 162.xx); (3) 1 prescription fill for crizotinib (this planning for patients with brain metastases typically criterion is used as a proxy to identify patients with ALKþ requires a multi-disciplinary approach and highly specia- NSCLC since there is no ICD-9 CM diagnosis code to 34–36 32 lized care , and involves a substantial caregiver and identify ALKþ patients directly and crizotinib was economic burden. restricted to the treatment of this population only); (4) Few studies have characterized the clinical and eco- no medical claims associated with a clinical trial (identi- nomic burden experienced by patients with brain metas- fied using ICD-9 CM code for V70.7); (5) 1 diagnosis of 27,37,38 tases , and none has specifically analyzed the brain metastasis (ICD-9 CM codes 198.3 and 198.4); and (ALKþ) NSCLC population. The objective of this study (6) continuous eligibility2 months before and1 month is to evaluate the impact of developing brain metastases on after the date of the first observed diagnosis of brain metas- patients with ALKþ NSCLC in terms of economic and tasis. Patients selected for this study were referred to as clinical outcomes. ‘crizotinib patients’. 2015 Informa UK Ltd www.informahealthcare.com/jme Burden of brain metastasis in ALKþ lung cancer Gue´ rin et al. 313 Journal of Medical Economics Volume 18, Number 4 April 2015 observed brain metastasis diagnosis, while the post-brain Patient characteristics metastasis-diagnosis period spanned between 30 days The reported patient characteristics included demograph- before the first observed brain metastasis diagnosis ics (age, gender, geographical region), the year of the first and the end of data availability or of continuous eligibility observed brain metastasis diagnosis, the Charlson (Figure 1). The pre-metastasis-diagnosis period was trun- Comorbidity Index (CCI; a higher score indicates a greater cated since some patients had very long observation peri- comorbidity burden) , an indicator of the patient’s overall ods, and truncation was done at 1 year since the median comorbidity profile, and the CCI components reported length of this period was 1 year. The following symptoms individually. Age was measured as of the date of the first that required medical attention and the proportion of observed brain metastasis diagnosis, while the comorbid- patients with these symptoms that are the most commonly ities were measured during the pre-brain-metastasis- reported in the literature were analyzed during these two diagnosis period. The pre-brain-metastasis-diagnosis observation periods: weakness or fatigue, shortness of period was defined as a maximum 1-year period starting breath, nausea or vomiting, headaches, pain or numbness, during the period of continuous eligibility and ending altered mental state, depression, seizure, stroke/transient 30 days before the first observed brain metastasis diagnosis. ischemic attack, anxiety, vision disorder, drowsiness, The timing (measured in days) of the development of loss of appetite, cognitive impairment, speech problems, brain metastases relative to NSCLC diagnosis and crizoti- problems with memory, focal neurologic deficits (motor or nib treatment was captured for all eligible patients. This sensory), and changes in mood or personality . included the proportion of patients who developed brain metastases (a) on or prior to the date of the first observed Treatment patterns related to brain metastases NSCLC diagnosis, (b) after the first observed NSCLC diagnosis but prior to the date of the first observed For the analysis of treatment patterns, the observation prescription for crizotinib, (c) during crizotinib monother- period started on the day of the first observed brain metas- apy-treatment, and (d) after crizotinib monotherapy- tasis diagnosis and ended with the end of data availability treatment. The average time between the first observed or of continuous eligibility (Figure 1). During the observa- diagnosis of NSCLC and the first observed diagnosis of tion period, the following treatments were analyzed: brain metastases recorded in the database were also chemotherapy, crizotinib, other targeted therapies (e.g., reported among patients diagnosed with brain metastases bevacizumab, erlotinib), corticosteroids, radiation therapy after their first observed diagnosis for NSCLC. (including WBRT), SRS, brain surgery, and brachyther- apy. The number of patients with at least one of the above treatments was reported. In addition, the list of all types of Symptoms of brain metastases treatments received after brain metastasis diagnosis and For the analysis of symptoms frequently observed the treatment patterns of the above treatments before, in patients with brain metastasis, the observation during, and after crizotinib treatment, regardless of the period was divided into two periods: the pre- and sequence and the course of the treatment, was summarized. post-brain-metastasis-diagnosis periods. The pre- Since at the time of the study, crizotinib was the only brain-metastasis-diagnosis period was defined as a treatment available for patients with ALKþ NSCLC, maximum 1-year period starting during the period of con- the treatment patterns were expected to be influenced by tinuous eligibility and ending 30 days before the first the timing of the crizotinib course of treatment. Beginning of continuous First observed brain End of continuous eligibility or beginning of metastases eligibility or end of data data availability diagnosis availability 1 year 30 days Pre-brain-metastasis- Post-brain-metastasis- diagnosis period diagnosis period Definition of periods for the analyses of treatments Definition of periods for the symptoms, cost and HRU analyses Figure 1. Study design. 314 Burden of brain metastasis in ALKþ lung cancer Gue´ rin et al. www.informahealthcare.com/jme 2015 Informa UK Ltd Journal of Medical Economics Volume 18, Number 4 April 2015 Paents with at least one diagnosis for lung cancer Healthcare resource utilization and costs n = 1,352,449 Healthcare resource utilization and costs were each assessed during the pre- and the post-brain-metastasis- diagnosis periods described above. Specifically, the Adult paents with at least one prescripon fill for crizonib – this criteria was used as a proxy to idenfy paents with ALK+ NSCLC number of inpatient (IP) stays, number of IP days, n = 947 number of emergency room (ER) visits, number of out- patient (OP) visits, and number of other medical services (e.g., laboratory, radiology, patient’s home, hospice, or Paents without any medical claims associated with a clinical trial other ancillary services) were captured for each patient n = 753 and reported as per-patient-per-month (PPPM) values. Healthcare costs, measured from the payer’s perspective (i.e., including all charges reimbursed by payers— Paents with at least one diagnosis for brain metastases excluding the Medicare portion for patients 65 years n = 253 and older from the IMS LifeLink Health Plan Claims data- base), were adjusted for inflation and expressed in 2013 US dollars using the Consumer Price Index medical care Paents with at least 2 months of connuous eligibility before and component. Total healthcare costs were comprised of 1 month of connuous eligibility aer the date of the first observed pharmacy costs and medical service costs. Pharmacy diagnosis of brain metastases n = 213 costs included NSCLC-specific therapy costs (chemother- apy, targeted therapy, corticosteroids, or crizotinib) as well Figure 2. Sample selection. as costs for other therapies used for symptom management, and other pharmacy costs. Medical service costs included monotherapy. The median time from the first observed costs associated with IP stays, ER visits, OP visits, and NSCLC diagnosis to the first observed brain metastasis other medical services. Healthcare costs associated diagnosis among patients diagnosed with brain metastases with radiotherapy were also reported separately for each was 88 days. component of the medical costs. Only descriptive statistics were reported in all analyses. All statistical analyses were conducted using SAS 9.3 (SAS Institute Inc., Cary, NC). Frequent symptoms in patients with brain metastases Overall, the symptom burden was high both before and Results after the diagnosis of brain metastasis in crizotinib patients, Patient characteristics but patients experienced a substantially higher number of symptoms after brain metastasis diagnosis than before: the A total of 753 crizotinib patients were identified from a most frequent symptoms were fatigue (15.0% pre-diagnosis pool of 1,352,449 lung cancer patients; of these, 213 were and 39.0% post-diagnosis), shortness of breath (14.1% and diagnosed with brain metastases and satisfied all other 38.0%, respectively), nausea or vomiting (10.8% and selection criteria (Figure 2). The mean patient age for 33.3%, respectively), and headaches (4.7% and 23.9%, the study sample was 54.8 years old and 55.9% of patients respectively) (Figure 3). were female (Table 1). The study sample had an average CCI score of 2.7 (SD¼ 3.0). The most common comorbid- ities—other than cancer—in the patient sample were Treatment patterns among brain metastases chronic pulmonary disease (16.9%), diabetes (8.0%), patients and mild liver disease (4.2%). In total, 70.4% of patients (n¼ 150) were diagnosed After the diagnosis of brain metastases, 88.7% received with brain metastasis prior to receiving targeted therapy chemotherapy, 84.0% received crizotinib, 63.4% received for ALKþ NSCLC. About 23.0% (n¼ 49) of patients radiation therapy, and 31.9% had SRS at some point were diagnosed with brain metastasis at the time of the during the observation period. In addition to these anti- initial NSCLC diagnosis and 47.4% (n¼ 101) were cancer therapies, 77.5% of the study sample received cor- reported to have their first observed brain metastasis diag- ticosteroids (Table 2). A good portion (69%) of patients nosis following NSCLC diagnosis but prior to crizotinib received multiple treatments after the diagnosis of brain initiation. Similarly, 19.2% (n¼ 41) had their first metastases, regardless if they were already treated with observed brain metastasis diagnosis observed during crizo- crizotinib (Table 3). Although treatment combinations tinib monotherapy, and 10.3% (n¼ 22) after crizotinib varied, the most frequently used treatment combination 2015 Informa UK Ltd www.informahealthcare.com/jme Burden of brain metastasis in ALKþ lung cancer Gue´ rin et al. 315 Journal of Medical Economics Volume 18, Number 4 April 2015 Table 1. Patient characteristics. received radiotherapy only. Detailed treatment patterns are summarized in Table 3. Crizotinib patients with brain metastases (n¼ 213) Healthcare resource utilization and cost Crizotinib patients’ resource utilization increased after Age (years; mean (SD)) 54.8 (11.6) Female, n (%) 119 (55.9%) the diagnosis of brain metastasis, with OP visits increasing Region, n (%) from 1.37 visits PPPM pre-diagnosis to 3.25 visits PPPM Northeast 45 (21.1%) post-diagnosis, IP stays increasing from 0.04 stays PPPM South 81 (38.0%) West 42 (19.7%) pre-diagnosis to 0.25 stays PPPM post-diagnosis, and Midwest 42 (19.7%) patients having more IP days post-diagnosis (increasing Unknown 3 (1.4%) Data Source, n (%) from 0.25 days PPPM pre-diagnosis to 1.17 days PPPM Source Healthcare Analytics’ Source Lx 136 (63.8%) post-diagnosis) (Table 5). Patients also had an almost database 6-fold increase in the utilization of other medical services IMS LifeLink Health Plan Claims database 22 (10.3%) Truven Health Analytics MarketScan database 55 (25.8%) post-diagnosis, increasing from 0.19 PPPM pre-diagnosis Year of first observed brain metastasis diagnosis, n (%) to 1.09 visits PPPM in the post-diagnosis period. 2007 1 (0.5%) Higher resource utilization resulted in higher 2008 5 (2.3%) 2009 2 (0.9%) healthcare costs, as costs rose in the post-brain 2010 8 (3.8%) metastasis-diagnosis period compared to the pre-brain- 2011 94 (44.1%) metastasis-diagnosis period: patients had an average total 2012 95 (44.6%) 2013 8 (3.8%) monthly cost of $5983 in the pre-brain-metastasis-diagno- Charlson Comorbidity Index, Mean (SD) 2.66 (2.97) sis period vs $22,645 in the post-brain-metastasis-diagnosis Charlson Comorbidity Index Components, n (%) period (Figure 4). Furthermore, in the pre-brain-metasta- Metastatic Carcinoma 63 (29.6%) Cancer 46 (21.6%) sis-diagnosis period, 79% of patients had total healthcare Chronic Pulmonary Disease 36 (16.9%) costs between $0–$10,000, compared to 26% of patients Diabetes 17 (8.0%) in the post-brain-metastasis-diagnosis period (Figure 5). Mild Liver Disease 9 (4.2%) Congestive Heart Failure 6 (2.8%) Post-diagnosis, medical costs increased 4-fold and repre- Cerebrovascular Disease 6 (2.8%) sented a higher portion of the total healthcare costs (58%), Peripheral Vascular Disease 4 (1.9%) with the cost of hospitalization and associated procedures Myocardial Infarction 3 (1.4%) Connective Tissue Disease/Rheumatic 2 (0.9%) dominating this cost category (IP costs were 51% of the Disease medical care costs). Pharmacy costs increased more than Renal Disease 2 (0.9%) 3-fold in the post-diagnosis period. In the pre-brain-metas- Dementia 1 (0.5%) Peptic Ulcer Disease 1 (0.5%) tasis-diagnosis period, crizotinib cost accounted for 52.7% Hemiplegia or Paraplegia 1 (0.5%) of the total pharmacy costs ($2924), while in the post- Moderate-to-Severe Liver Disease 1 (0.5%) brain-metastasis-diagnosis period, crizotinib costs AIDS/HIV (Excluding Asymptomatic – HIV Infection) accounted for 68.3% of the total pharmacy costs Treatments before Brain Metastasis ($9520). Radiotherapy was mainly performed on an Diagnosis, n (%) OP basis, and was responsible for 43% of the OP costs. Chemotherapy 46 (21.6%) Targeted Therapy (excluding crizotinib) 21 (9.9%) Pre-diagnosis, 51% of total costs were due to medical Crizotinib 53 (24.9%) care costs (with OP costs representing the majority— Radiotherapy 32 (15.0%) 59%—of these costs), and the remainder was due to the Brain Surgery 0 (0.0%) Brachytherapy 1 (0.5%) cost of pharmacological treatment. consisted of crizotinibþ radiotherapy (18.3%). Among Discussion patients who developed brain metastases before initiating crizotinib (n¼ 150), 16.7% received crizotinib only and The ALKþ translocation confers special clinicopathologic 21.3% received crizotinib and radiotherapy after the diag- features to the NSCLC cancer sub-type: ALKþ NSCLC nosis of brain metastases. Among patients who developed patients tend to be diagnosed with lung cancer at a younger 14,15,19,21,40 brain metastases while on crizotinib (n¼ 41), 31.7% con- age , and have a higher risk of developing brain tinued on crizotinib monotherapy while 17.1% added or (and other) metastases than patients with other NSCLC 20,21,41,42 switched to radiotherapy. Finally, among patients who sub-types . Brain metastases are known to severely developed brain metastases after discontinuing crizotinib impact a patient’s quality-of-life ( physical, cognitive, and 24,31,33 (n¼ 22), 63.6% of them remained untreated and 13.6% functional impairments) , as well as to result in high 316 Burden of brain metastasis in ALKþ lung cancer Gue´ rin et al. www.informahealthcare.com/jme 2015 Informa UK Ltd Journal of Medical Economics Volume 18, Number 4 April 2015 45% Pre-brain-metastasis-diagnosis period 39.0% 40% 38.0% Post-brain-metastasis-diagnosis period 35% 33.3% 30% 23.9% 25% 20% 16.9% 15.0% 14.6% 14.1% 15% 13.1% 12.2% 10.8% 9.9% 10% 8.5% 8.5% 6.1% 5.6% 5.2% 5.2% 4.7% 4.7% 4.7% 3.8% 5% 2.3% 2.3% 1.9% 1.4% 1.4% 1.4% 0.9% 0.9% 0.5% 0.5% 0.5% 0.0% 0.0% 0.0% 0% Figure 3. Symptoms during the pre- and post-brain-metastasis-diagnosis periods. Table 2. Treatments received after brain metastasis diagnosis. burden, resource use, and costs in ALKþ NSCLC patients (identified by the use of crizotinib). Patients with brain Crizotinib patients with metastases experienced a high symptom burden, particu- brain metastases (n¼ 213) larly post-diagnosis of these metastases: more than a Treatments, n (%) n % third suffered from fatigue, shortness of breath, nausea/ vomiting, and almost a quarter suffered from headaches Non-Systemic treatments Radiation Therapy 135 63.4% post-diagnosis. Prior studies also report that neuro- Stereotactic Radiosurgery 68 31.9% 44,45 logical symptoms are frequently encountered (in 36– Brain Surgery 19 8.9% 44–46 Brachytherapy 1 0.5% 82% of cases) in lung patients with brain metastases. Systemic treatments Temporary relief from symptoms can be achieved with Crizotinib 179 84.0% 22,45,47–49 treatment , but not all patients will respond to Targeted Therapy 43 20.2% Chemotherapy 189 88.7% treatment , and resistance to treatment eventually Supportive treatments occurs . Corticosteroids 165 77.5% In this study, we observed that, after being diagnosed No treatment reported 19 8.9% with brain metastases, most patients received several treat- Patients may have received multiple types of treatments. Therefore, counts ments—chemotherapy, crizotinib, corticosteroids, and are not mutually exclusive of one another. radiation therapy—but treatment combinations and pat- Targeted therapies received after brain metastasis diagnosis included bev- acizumab, bortezomib, and erlotinib. terns remained highly variable among the sample studied. According to the NCCN guidelines, radiation therapy for 28–30 brain metastases and crizotinib for advanced ALKþ 37 43 50 healthcare resource utilization and substantial clinical , NSCLC are the de facto treatment options for most 27 32 ALKþ patients with or without brain metastases, but, for economic , and caregiver burden. In this study, we found that the development of brain intolerant or resistant patients, there are several recom- metastases is associated with a substantial treatment mended treatment options, with no clear hierarchical 2015 Informa UK Ltd www.informahealthcare.com/jme Burden of brain metastasis in ALKþ lung cancer Gue´ rin et al. 317 Journal of Medical Economics Volume 18, Number 4 April 2015 Table 3. All treatments received after brain metastasis diagnosis . Crizotinib patients with brain metastases (n¼ 213), n (%) Period after brain metastasis diagnosis (days), mean SD [median] 333.3 298.9 [251] Crizotinib and radiotherapy 39 (18.3%) Crizotinib only 38 (17.8%) No treatment reported 19 (8.9%) Crizotinib, chemotherapy, and radiotherapy 14 (6.6%) Crizotinib, chemotherapy, radiotherapy, and targeted therapy 13 (6.1%) Crizotinib, radiotherapy, and stereotactic radiosurgery 13 (6.1%) Crizotinib and chemotherapy 10 (4.7%) Crizotinib, chemotherapy, radiotherapy, and stereotactic radiosurgery 10 (4.7%) Crizotinib, chemotherapy, radiotherapy, stereotactic radiosurgery, and targeted therapy 8 (3.8%) Crizotinib, chemotherapy, and targeted therapy 7 (3.3%) Radiotherapy only 7 (3.3%) Crizotinib, radiotherapy, and targeted therapy 6 (2.8%) Crizotinib, brain surgery, radiotherapy, and stereotactic radiosurgery 5 (2.3%) Chemotherapy and radiotherapy 3 (1.4%) Crizotinib, brain surgery, chemotherapy, and radiotherapy 3 (1.4%) Crizotinib, brain surgery, chemotherapy, radiotherapy, stereotactic radiosurgery, and targeted therapy 3 (1.4%) Chemotherapy, radiotherapy, and targeted therapy 2 (0.9%) Crizotinib, brain surgery, chemotherapy, radiotherapy, and stereotactic radiosurgery 2 (0.9%) Crizotinib, brain surgery, chemotherapy, radiotherapy, and targeted therapy 2 (0.9%) Crizotinib, brain surgery, and radiotherapy 2 (0.9%) Crizotinib and stereotactic radiosurgery 2 (0.9%) Brain surgery, chemotherapy, and radiotherapy 1 (0.5%) Chemotherapy, radiotherapy, stereotactic radiosurgery, and targeted therapy 1 (0.5%) Crizotinib, brachytherapy, chemotherapy, radiotherapy, and stereotactic radiosurgery 1 (0.5%) Crizotinib, brain surgery, and targeted therapy 1 (0.5%) Stereotactic radiosurgery only 1 (0.5%) The list of treatments does not take into account the sequence of treatment. There may also be several courses of treatment within the same list of treatments. $8,000 Pre-brain-metastasis-diagnosis $9,520 $6,692 Period $7,000 Post-brain-metastasis-diagnosis Period $5,877 $6,000 Radiotherapy-related Costs $5,000 $4,000 $3,000 $2,924 $1,813 $2,000 $1,174 $1,000 $507 $49 $61 $11 $0 Inpaent Costs Outpaent Costs Emergency Room Other Medical Pharmacy Costs Service Costs Figure 4. Healthcare costs during the pre- and post-brain-metastasis-diagnosis periods. path (for all steps subsequent to crizotinib treatment, the not unexpected, and a similar use of systemic therapy NCCN recommendations are based upon lower-level evi- (65%), radiation therapy (47%), and corticosteroids dence; there is NCCN consensus that the intervention is (79%) has been noted previously in a retrospective study 50 27 appropriate ). Thus, the observed treatment variability is of lung cancer patients with brain metastases . 318 Burden of brain metastasis in ALKþ lung cancer Gue´ rin et al. www.informahealthcare.com/jme 2015 Informa UK Ltd Journal of Medical Economics Volume 18, Number 4 April 2015 90% 80% 70% 60% Pre-brain-metastasis-diagnosis Period 50% Post-brain-metastasis-diagnosis Period 40% 30% 20% 10% 0% 0 - 10K 10K - 20K 20K - 30K 30K - 40K 40K - 50K 50K - 60K 60K - 70K 70K - 80K 80K - 90K 90K + Heathcare Costs PPPM Range Figure 5. Distribution of total healthcare costs PPPM. With the increase in symptoms and the variety of treat- distribution is that using targeted agents for the treatment ments post-diagnosis of brain metastases, the current study of (brain) metastatic ALKþ NSCLC increases pharmacy found that the already high economic burden of treatment costs, but reduces the use of inpatient services, possibly pre-diagnosis of brain metastases drastically increased after indicating that the disease has been managed (for a diagnosis. The average monthly costs per patient almost period of time). Another study evaluating the costs of quadrupled after the diagnosis of brain metastases, with treating brain-metastatic cancer patients (of which 55% monthly costs per patient averaging $5983 (interquartile were lung cancer patients) also found that brain metastases are associated with a significant economic burden, con- range: $118–$8667) vs $22,645 (interquartile range: cluding that the total costs per year of life amounted to $9452–$26,693) for the pre- and post-brain metastasis per- $62,130 (where less than 25% of the patients survived iods, respectively. The estimate reported by Ray et al. of the first year of treatment) . post-diagnosis monthly costs was similar to the mean total This study has demonstrated that the development of $17,007 for all-cause healthcare costs (PPPM) reported of brain metastases, common in ALKþ NSCLC, can be for lung cancer patients, of all types, with newly diagnosed costly and can present a substantial clinical burden. brain metastases. However, there are several differences in Further analyses are warranted to explore how treatment how costs are distributed for brain-metastasis diagnosed in patterns and the burden are impacted post-crizotinib, now crizotinib patients vs in lung cancer patients: crizotinib that a new therapy is available for patients who do not patients had much higher pharmacy costs post-diagnosis respond to crizotinib. of brain metastases ($9520 vs $809 PPPM), but lower med- ical costs ($13,125 vs $16,198 PPPM, respectively), with similar IP stays, but fewer ER visits (0.04 visits vs 0.21 visits Limitations PPPM, respectively) and more OP visits (3.25 visits vs 2.8 visits PPPM, respectively). It must be noted that crizotinib The current study was subject to some limitations. First, was granted accelerated regulatory approval in the US in patients were selected based on several proxy indicators of August 2011, and erlotinib was approved for NSCLC the type of the disease, as there was no ICD-9 CM code to treatment in April 2010, hence these agents were not identify ALKþ and/or NSCLC patients. A prescription fill available at the time Ray et al. conducted their study. for crizotinib was used as a proxy to identify patients with In addition, since ALKþ NSCLC has a relatively rare ALKþ NSCLC among patients with a diagnosis of lung rate of occurrence (2–8%) among all NSCLC lung cancer. Thus, the findings of the study may not be repre- 11–19 cancers , the treatment patterns reported by Ray sentative of ALKþ NSCLC patients who did not receive et al. are likely not indicative of this type of cancer. crizotinib. Second, patients were not required to have A probable explanation for the differences in cost more than one diagnosis of brain metastases. Thus, their 2015 Informa UK Ltd www.informahealthcare.com/jme Burden of brain metastasis in ALKþ lung cancer Gue´ rin et al. 319 Percentage of Paents per Cost Category Journal of Medical Economics Volume 18, Number 4 April 2015 Table 4. Treatment patterns after brain metastasis diagnosis relative to crizotinib treatment . Crizotinib patients with brain metastases (N¼ 213) Among patients who developed brain metastases before initiating crizotinib N¼ 150 Patients who received crizotinib first following NSCLC diagnosis (n¼ 30) Crizotinib only 25 (16.7) And radiotherapy 3 (2.0) And chemotherapy 1 (0.7) And chemotherapy, radiotherapy, and SRS 1 (0.7) Patients who received at least one other treatment before crizotinib (n¼ 120) Radiotherapy only 29 (19.3) Radiotherapyþ chemotherapy 11 (7.3) Radiotherapyþ chemotherapyþ targeted therapy 11 (7.3) Radiotherapyþ targeted therapy 6 (4.0) Radiotherapyþ SRS 10 (6.7) Radiotherapyþ SRSþ chemotherapy 5 (3.3) Radiotherapyþ SRSþ chemotherapyþ targeted therapy 6 (4.0) Radiotherapyþ SRSþ chemotherapyþ brachitherapy 1 (0.7) Chemotherapy only 9 (6.0) Chemotherapyþ targeted therapy 7 (4.7) Brain surgeryþ radiotherapy 2 (1.3) Brain surgeryþ radiotherapyþ SRS 4 (2.7) Brain surgeryþ radiotherapyþ SRSþ chemotherapy 2 (1.3) Brain surgeryþ radiotherapyþ SRSþ chemotherapyþ targeted therapy 3 (2.0) Brain surgeryþ radiotherapyþ chemotherapy 2 (1.3) Brain surgeryþ radiotherapyþ chemotherapyþ targeted therapy 2 (1.3) Brain surgeryþ SRS 1 (0.7) Brain surgeryþ targeted therapy 1 (0.7) SRSþ chemotherapy 5 (3.3) SRS only 3 (2.0) Among patients who developed brain metastases while on crizotinib N¼ 41 Patients who continued crizotinib (n¼ 28) Crizotinib monotherapy 13 (31.7) Crizotinibþ radiotherapy/SRS/brain surgery/ 10 (24.4) Added or switched to radiotherapy 7 (17.1) Added SRS 1 (2.4) Added radiotherapyþ SRS 1 (2.4) Added radiotherapyþ brain surgery, then continued with targeted therapyþ chemotherapy 1 (2.4) Crizotinibþ chemotherapyþ radiotherapy 1 (2.4) Switched to radiotherapy, chemotherapy or other targeted therapy 5 (12.2) Switched to chemotherapy 2 (4.9) Switched to other targeted therapy and chemotherapy 2 (4.9) Switched to radiotherapy, þtargeted therapy, þchemotherapy 1 (2.4) Patients who discontinued crizotinib (n¼ 12) Remained untreated 5 (12.2) Received radiotherapy only 4 (9.8) Received radiotherapyþ chemotherapy 1 (2.4) Received radiotherapyþ targeted therapyþ chemotherapy 1 (2.4) Received radiotherapyþ brain surgeryþ targeted therapyþ chemotherapy 1 (2.4) Among patients who developed brain metastases after discontinuing crizotinib N¼ 22 Remained untreated 14 (63.6) Received radiotherapy only 3 (13.6) Received radiotherapyþ chemotherapyþ targeted therapy 2 (9.1) Received radiotherapyþ chemotherapy 1 (4.5) Received radiotherapyþ SRSþ chemotherapyþ targeted therapy 1 (4.5) Received SRS only 1 (4.5) SRS, stereotactic radiosurgery. The list of treatments does not take into account the sequence of treatment. There may also be several courses of treatment within the same list of treatments. brain metastases were not confirmed with at least two estimation of costs experienced by patients covered by independent diagnoses. Third, information needed to Medicare: only the MarketScan database had linked infor- identify ER visits in the Source Healthcare Analytics mation on previous employees now covered by Medicare, Source Lx database was not available, thus the use of thus this study had very limited information on the ER services, as well as associated costs, may be under- Medicare-covered population, and its findings may not estimated in this study. A similar limitation arises in the be immediately generalizable to non-privately-insured 320 Burden of brain metastasis in ALKþ lung cancer Gue´ rin et al. www.informahealthcare.com/jme 2015 Informa UK Ltd Journal of Medical Economics Volume 18, Number 4 April 2015 Table 5. Healthcare resource utilization pre- and post-brain-metastasis-diagnosis period. Healthcare resource utilization (PPPM), mean SD Pre-brain-metastasis-diagnosis period (n¼ 213) Post-brain-metastasis-diagnosis period (n¼ 213) IP stays 0.04 0.09 0.25 0.34 IP days 0.25 1.06 1.17 1.92 OP 1.37 1.73 3.25 2.47 ER 0.02 0.07 0.04 0.13 Other medical services 0.19 0.54 1.09 2.00 PPPM, per-patient-per-month; IP, inpatient; OP, outpatient; ER, emergency room. A. Gue ´rin, K. Dea, R. Nitulescu, and E. Q. Wu are employees of populations. Moreover, since information on different cost Analysis Group Inc., which has received consultancy fees from components (e.g., co-ordination of benefits, out of pocket Novartis. costs) was not available in the Source Lx database, a common measure of costs was used for all three databases. Acknowledgments Fourth, only symptoms that required medical attention The authors thank Ana Bozas, PhD, an employee of Analysis were captured in the study. Fifth, this study examined Group, for writing and editorial support. A synopsis of this only direct medical costs, evaluated from the payers’ per- study’s results has been recently accepted as a poster presentation spective. Information to determine indirect costs, such as at the 19th annual meeting of the Society of Neuro-Oncologists lost productivity and burden to caregivers, was not avail- (SNO) in Miami, Florida from November 13th to 16th (Poster able. Given the high comorbidity burden of the study’s BM-19). patients (general poor health) and frequent hospitaliza- tions, we expect these undetermined indirect costs to be substantial. Further studies are warranted to investigate References this aspect of the economic burden of the ALKþ 1. American Cancer Society. Lung Cancer (Non-Small Cell). Atlanta, GA: NSCLC. Sixth, the treatment choices analyzed in this American Cancer Society; 2014, 77. http://www.cancer.org/acs/groups/con- study do not include ceritinib, the newly-approved tent/@research/documents/document/acspc-041770.pdf. Accessed August second generation tyrosine kinase inhibitor for the treat- 9, 2013 ment of ALKþ NSCLC after progression on crizotinib, as 2. Howlader N, Noone A, Krapcho M, et al., eds. SEER Cancer Statistics Review 1975-2010. Bethesda, MD: National Cancer Institute; 2010. http://seer.can- at the time the study was conducted ceritinib was not yet cer.gov/csr/1975_2010/. Accessed June 6, 2013 available. Finally, further studies are warranted to study 3. Larsen JE, Cascone T, Gerber DE, et al. Targeted therapies for lung cancer: the healthcare resource utilization and costs associated clinical experience and novel agents. Cancer J 2011;17:512-27 with the development of brain metastases in ALKþ 4. Custodio A, Mendez M, Provencio M. Targeted therapies for advanced non- NSCLC patients using more recent data, other data small-cell lung cancer: current status and future implications. Cancer Treat Rev 2012;38:36-53 sources, or different study designs. 5. Bang Y-J. Treatment of ALK-positive non-small cell lung cancer. Arch Pathol Lab Med 2012;136:1201-4 6. Li Y, Ye X, Liu J, et al. Evaluation of EML4-ALK fusion proteins in non-small Conclusions cell lung cancer using small molecule inhibitors. Neoplasia 2011;13:1-11 7. Shaw AT, Solomon B. Targeting anaplastic lymphoma kinase in lung cancer. In this study we found that brain metastases present a sub- Clin Cancer Res 2011;17:2081-6 stantial burden in the ALKþ NSCLC patient population 8. Kim T-J, Park CK, Yeo CD, et al. Simultaneous diagnostic platform of geno- (crizotinib patients) and significant variability in the treat- typing EGFR, KRAS, and ALK in 510 Korean patients with non-small-cell lung ment options was observed after the patient was diagnosed cancer highlights significantly higher ALK rearrangement rate in advanced with brain metastasis. Healthcare utilization and costs stage. J Surg Oncol 2014;110(3):245-51. published online May 29, 2014, doi: 10.1002/jso.23646 after brain metastasis diagnosis were substantially greater 9. Park J, Yamaura H, Yatabe Y, et al. Anaplastic lymphoma kinase gene than before diagnosis. Given the large percentage of rearrangements in patients with advanced-stage non-small-cell lung ALKþ NSCLC patients who will eventually develop cancer: CT characteristics and response to chemotherapy. Cancer Med brain metastases, this study may highlight an important 2013;3:118-23 unmet need. 10. Chun SG, Choe KS, Iyengar P, et al. Isolated central nervous system progres- sion on Crizotinib: an Achilles heel of non-small cell lung cancer with EML4- ALK translocation? Cancer Biol Ther 2012;13:1376-83 11. Soda M, Choi YL, Enomoto M, et al. Identification of the transforming EML4- Transparency ALK fusion gene in non-small-cell lung cancer. Nature 2007;448:561-6 Declaration of funding 12. Koivunen JP, Mermel C, Zejnullahu K, et al. EML4-ALK fusion gene and This study was funded by Novartis. M. Sasane, J. Zhang, and K. efficacy of an ALK kinase inhibitor in lung cancer. Clin Cancer Res Culver are employees of Novartis and own stock/stock options. 2008;14:4275-83 2015 Informa UK Ltd www.informahealthcare.com/jme Burden of brain metastasis in ALKþ lung cancer Gue´ rin et al. 321 Journal of Medical Economics Volume 18, Number 4 April 2015 13. Boland JM, Erdogan S, Vasmatzis G, et al. Anaplastic lymphoma 32. Schubart JR, Kinzie MB, Farace E. Caring for the brain tumor patient: family kinase immunoreactivity correlates with ALK gene rearrangement and tran- caregiver burden and unmet needs. Neuro Oncol 2008;10:61-72 scriptional up-regulation in non-small cell lung carcinomas. Hum Pathol 33. Welzel G, Fleckenstein K, Schaefer J, et al. Memory function before and 2009;40:1152-8 after whole brain radiotherapy in patients with and without brain metastases. 14. Shaw AT, Yeap BY, Mino-Kenudson M, et al. Clinical features and outcome Int J Radiat Oncol Biol Phys 2008;72:1311-18 34. Soffietti R, Ruda R, Trevisan E. Brain metastases: current management and of patients with non-small-cell lung cancer who harbor EML4-ALK. J Clin new developments. Curr Opin Oncol 2008;20:676-84 Oncol 2009;27:4247-53 35. Ricciardi S, de Marinis F. Multimodality management of non-small cell lung 15. Wong DW-S, Leung EL-H, So KK-T, et al. The EML4-ALK fusion gene cancer patients with brain metastases. Curr Opin Oncol 2010;22:86-93 is involved in various histologic types of lung cancers from nonsmokers 36. Narita Y, Shibui S. Strategy of surgery and radiation therapy for brain metas- with wild-type EGFR and KRAS. Cancer 2009;115:1723-33 tases. Int J Clin Oncol 2009;14:275-80 16. Inamura K, Takeuchi K, Togashi Y, et al. EML4-ALK fusion is linked to 37. Nieder C, Norum J, Stemland JG, et al. Resource utilization in patients with histological characteristics in a subset of lung cancers. J Thorac Oncol brain metastases managed with best supportive care, radiotherapy and/or 2008;3:13-17 surgical resection: a Markov analysis. Oncology 2010;78:348-55 17. Takeuchi K, Choi YL, Soda M, et al. Multiplex reverse transcription-PCR 38. Takura T, Hayashi M, Muragaki Y, et al. [Study on medical economic evalu- screening for EML4-ALK fusion transcripts. Clin Cancer Res ation methods for metastatic brain tumors therapy]. No Shinkei Geka 2008;14:6618-24 2010;38:629-37 18. Perner S, Wagner PL, Demichelis F, et al. EML4-ALK fusion lung cancer: a 39. Quan H, Sundararajan V, Halfon P, et al. Coding algorithms and for rare acquired event. Neoplasia 2008;10:298-302 defining comorbidities data in ICD-9-CM administrative. Med Care 19. Rodig SJ, Mino-Kenudson M, Dacic S, et al. Unique clinicopathologic features 2005;43:1130-9 characterize ALK-rearranged lung adenocarcinoma in the western population. 40. Zhang N-N, Liu Y-T, Ma L, et al. The molecular detection and clinical signifi- Clin Cancer Res 2009;15:5216-23 cance of ALK rearrangement in selected advanced non-small cell lung cancer: 20. Doebele RC, Lu X, Sumey C, et al. Oncogene status predicts patterns of meta- ALK expression provides insights into ALK targeted therapy. PLoS One static spread in treatment-naive nonsmall cell lung cancer. Cancer 2014;9:e84501 2012;118:4502-11 41. Paik JH, Choi C-M, Kim H, et al. Clinicopathologic implication of ALK 21. Kang HJ, Lim H-J, Park JS, et al. Comparison of clinical characteristics rearrangement in surgically resected lung cancer: a proposal of diagnostic between patients with ALK-positive and EGFR-positive lung adenocarcinoma. algorithm for ALK-rearranged adenocarcinoma. Lung Cancer 2012;76:403-9 Respir Med 2014;108:388-94 42. Fallet V, Cadranel J, Doubre H, et al. Prospective screening for ALK: 22. Shaw AT, Kim D-W, Nakagawa K, et al. Crizotinib versus chemotherapy in clinical features and outcome according to ALK status. Eur J Cancer advanced ALK-positive lung cancer. N Engl J Med 2013;368:2385-94 2014;50:1239-46 23. Kim D-W, Mehra R, Tan DS-W, et al. Ceritinib in Advanced Anaplastic 43. Sørensen JB, Hansen HH, Hansen M, Dombernowsky P. Brain metastases in Lymphoma Kinase (ALK)-rearranged (ALKþ) Non-small Cell Lung Cancer adenocarcinoma of the lung: frequency, risk groups, and prognosis. J Clin (NSCLC) – Results of the ASCEND-1 Trial (#8003). J Clin Oncol Oncol 1988;6:1474-80 2014;32:5s:abstr 8003 44. Jena A, Taneja S, Talwar V, et al. Magnetic resonance (MR) patterns of brain 24. Wen PY, Loeffler JS. Overview of the clinical manifestations, diagnosis, and metastasis in lung cancer patients: correlation of imaging findings with symp- management of patients with brain metastases. Wolter Kluwer Health; 2013, tom. J Thorac Oncol 2008;3:140-4 p 1–15. UpToDate, Inc., 95 Sawyer Rd., Waltham, MA 02453 45. Komatsu T, Kunieda E, Oizumi Y, et al. Clinical characteristics of brain 25. Gofton TE, Graber J, Carver A. Identifying the palliative care needs of patients metastases from lung cancer according to histological type: pretreatment living with cerebral tumors and metastases: a retrospective analysis. evaluation and survival following whole-brain radiotherapy. Mol Clin Oncol J Neurooncol 2012;108:527-34 2013;1:692-8 26. Bauml J, Mick R, Zhang Y, et al. Determinants of survival in advanced non– 46. Ferrigno D, Buccheri G. Cranial computed tomography as a part of the small-cell lung cancer in the era of targeted therapies. Clin Lung Cancer initial staging procedures for patients with non-small-cell lung cancer. 2013;14:581-91 Chest 1994;106:1025-9 27. Ray S, Dacosta-Byfield S, Ganguli A, et al. Comparative analysis of survival, 47. Haura EB. Treatment of advanced non-small-cell lung cancer: a review of treatment, cost and resource use among patients newly diagnosed with brain current randomized clinical trials and an examination of emerging therapies. metastasis by initial primary cancer. J Neurooncol 2013;114:117-25 Cancer Control 2001;8:326-36 28. Patchell RA. The management of brain metastases. Cancer Treat Rev 48. Frampton JE. Crizotinib: a review of its use in the treatment of anaplastic 2003;29:533-40 lymphoma kinase-positive, advanced non-small cell lung cancer. Drugs 29. Taimur S, Edelman MJ. Treatment options for brain metastases in patients 2013;73:2031-51 with non-small-cell lung cancer. Curr Oncol Rep 2003;5:342-6 49. Wu C, Li YL, Wang ZM, et al. Gefitinib as palliative therapy for 30. Lu-Emerson C, Eichler AF. Brain metastases. Continuum (Minneap Minn) lung adenocarcinoma metastatic to the brain. Lung Cancer 2007;57: 2012;18:295-311 359-64 31. Tang V, Rathbone M, Park Dorsay J, et al. Rehabilitation in primary 50. National Comprehensive Cancer Network. Non-small cell lung cancer, ver- and metastatic brain tumours: impact of functional outcomes on survival. sion 4. 2014. Fort Washington, PA: National Comprehensive Cancer Network, J Neurol 2008;255:820-7 2014. p 1-147 322 Burden of brain metastasis in ALKþ lung cancer Gue´ rin et al. www.informahealthcare.com/jme 2015 Informa UK Ltd

Journal

Journal of Medical EconomicsTaylor & Francis

Published: Apr 3, 2015

Keywords: Non-small-cell lung cancer; Brain neoplasms; Neoplasm metastasis; Costs and cost analysis; Crizotinib; Symptoms

There are no references for this article.