Evaluation of Erlotinib in Advanced Non-Small Cell Lung Cancer: Impact on the Budget of a U.S. Health Insurance Plan

BACKGROUND: Lung cancer is the most common non-skin cancer and the leading cause of cancer death among men and women in North America. More than half of all patients diagnosed with lung cancer are diagnosed with advanced disease. Most cases of lung cancer are non-small cell lung cancer (NSCLC). Erlotinib monotherapy is indicated for the treatment of patients with locally advanced or metastatic NSCLC after failure of at least 1 prior chemotherapy regimen. OBJECTIVES: To assess the budgetary impact, from the health plan perspective, of covering erlotinib for treating patients with NSCLC stage IIIb/IV who have failed at least 1 prior chemotherapy regimen. METHODS: An Excel-based model was developed to evaluate costs for U.S. Food and Drug Administration-approved and National Comprehensive Cancer Network guideline-recommended treatment options for second- and third-line NSCLC from the perspective of a U.S. health insurer. In particular, the model compares a formulary with erlotinib and a formulary without erlotinib, including the costs of treatment, drug administration, and adverse effects. The incidence of advanced NSCLC is based on the Surveillance, Epidemiology, and End Results Cancer Registry and adverse effects related to treatment (all agents) in published results of clinical trials. Drug and treatment costs were obtained from publicly available sources in 2005. RESULTS: The base case considers a health plan of 500,000 enrollees. Assuming that erlotinib comprises 30% of second-line treatments and 90% for third-line, total costs of treating stage IIIb/IV NSCLC patients over 1 year are $382,418 with erlotinib and $380,968 without erlotinib (difference: $1,450; 90% confidence interval, -$61,376 to $29,855), less than $0.01 per member per month (PMPM) in 2005. Erlotinib direct cost is offset by reductions in standard chemotherapy-related infusion costs and adverse events. CONCLUSIONS: Based on the analysis, the inclusion of erlotinib on a formulary appears to have a relatively small impact on the annual health care budget or PMPM expenditures if it is used consistent with the product label indications.

ith more than 173,000 new cases diagnosed and more than 160,000 deaths projected, lung cancer was the most common non-skin cancer and the leading cause of cancer death among men and women in North America in 2005. 1 Most cases of lung cancer are non-small cell type; more than half of all patients diagnosed with lung cancer are diagnosed with advanced disease. In these cases, platinumbased chemotherapy offers symptomatic relief and modest improvement in survival, 2 yet responses are brief and most patients experience disease progression. Traditionally, secondline chemotherapy with docetaxel has been shown to prolong survival after platinum-based therapy for non-small cell lung cancer (NSCLC). 3,4 Recently several new agents have been approved for secondline therapy of advanced NSCLC. On November 18, 2004, erlotinib (Tarceva) was approved as monotherapy for the treatment of patients with locally advanced or metastatic NSCLC after failure of at least 1 prior chemotherapy regimen, i.e., as second-line therapy. Erlotinib is the only epidermal growth factor receptor tyrosine kinase inhibitor (EGFR TKI) therapy shown in a randomized phase III trial to provide a survival benefit to NSCLC patients (hazard ratio [HR] = 0.73). 5 In addition, in a recent Phase III trial of second-line therapy, the novel antifolate pemetrexed showed comparable activity to docetaxel with significantly reduced toxicity. 6 Although newer agents modestly lengthen overall survival in patients with advanced NSCLC, they are costly, and some have raised concerns about costs to patients. 7 Given the high prevalence of lung cancer, health plan managers may also be concerned about the impact of these novel therapies on their health plan budgets.
We developed a decision model to assess the impact of adding erlotinib to the health plan formulary on total health plan expenditures. The model evaluates patients with advanced NSCLC who have shown disease progression after treatment with chemotherapy and thus are eligible for second-or thirdline treatment. The analysis is conducted from the perspective of a private U.S. health insurer.

■■ Methods
The erlotinib budget impact model was developed to assess the budgetary impact of covering erlotinib for the treatment of patients with locally advanced or metastatic (stage IIIb and IV) NSCLC, who have failed at least 1 prior chemotherapy regimen. Treatment options considered for the analysis were based on U.S. Food and Drug Administration (FDA)-approved indications and practice guidelines recommended by the National Comprehensive Cancer Network (NCCN) for secondand third-line therapy. 8 The model was developed as a Microsoft Excel workbook. 9 Default estimates are provided for all model inputs based on various public and private data sources. The proportion of patients eligible for treatment with erlotinib were estimated using data from the U.S. Census Bureau and the National Cancer Institute' s Surveillance, Epidemiology, and End Results (SEER) Cancer Registry. 10 The model considers 2 scenarios. In the first scenario, erlotinib is not a treatment option in either the second-or the third-line treatment situation. Rather, treatment options for patients include 2 chemotherapy regimens, docetaxel and pemetrexed, that have been approved by the FDA for the treatment of advanced NSCLC and that has shown evidence of progression after first-line treatment. In the second scenario, erlotinib monotherapy is a treatment option and is used in place of competing therapies for a proportion of eligible patients. Comparing the total costs of treating patients in these 2 scenarios in the second-and the third-line treatment situation provides an assessment of the economic impact on a health plan' s budget of adding erlotinib to the formulary.

Estimation of the Patient Population Eligible for Erlotinib
The number of second-and third-line chemotherapy treatmenteligible patients in a hypothetical private health insurance plan with 500,000 covered lives, including a Medicare-eligible group, is based on a stratification of health plan enrollees by age group and age-specific estimates of the incidence of stage IIIb/IV NSCLC for each group ( Table 1).
The age distribution is based on U.S. Census Bureau data. 11 Age-and gender-stratified incidence estimates of advanced lung cancer are based on the incidence for lung cancers listed in the SEER registry as "distant stage." The distant stage designation approximates the American Joint Committee on Cancer (AJCC) stage IIIb/IV designation. The percentage of all metastatic distant stage lung cancer patients with NSCLC pathology is based on the proportion of such cases in the SEER registry. 10

Treatment Distribution
The model considers treatment options that are (1) FDA approved in the particular indication and (2) recommended by the NCCN guidelines for NSCLC. 8 Thus, patients in the secondline and third-line situations are assumed to receive pemetrexed, docetaxel, or erlotinib. Gefitinib is considered to have minimal use due to recent FDA labeling that severely restricts its use. Off-label treatments for stage IIIb/IV NSCLC are not considered in this model. The distribution of patients estimated to receive the various second-and third-line chemotherapy treatment options in the 2 treatment scenarios described above is based on 2004 market research data for erlotinib, docetaxel, and pemetrexed use in patients with lung cancer; it was derived from a commercially available survey. 13 For the scenario without erlotinib, we assume that 50% of patients receive docetaxel and 50% receive pemetrexed. Under the scenario with erlotinib, we assume that 30% of eligible patients receive erlotinib in the second-line situation (with corresponding equal reductions in use of pemetrexed and docetaxel), and 90% of patients receive erlotinib in the third-line situation.

Drug Utilization and Costs
Drug costs for erlotinib, docetaxel, and pemetrexed are estimated based on the medication dose specified in the product label, expected duration of treatment as reported in the pivotal trials for each product, and national reimbursement surveys of the medications 3,14,15 (Table 2). Dose reductions observed in the clinical trials for each agent were accounted for in the analysis. 3,4,6

Drug Administration Costs
Erlotinib is a self-administered oral tablet, hence no administration costs are assumed for this medication. The model assumes  1 outpatient physician office visit for the first prescription, with refill prescriptions for the remaining prescriptions not requiring additional physician office visits beyond scheduled follow-up evaluations. The total cost of administration for pemetrexed and docetaxel is based on the number of infusions per patient per year and the average cost of administering the infusion. One physician office visit is assumed each time the patient received an infusion. For each infusion, there is assumed to be a fixed cost for the first hour. The cost of the infusion is based on 2005 Centers for Medicare & Medicaid Services (CMS) payment rates for Current Procedural Terminology (CPT) code 96410 (chemotherapy, infusion method). 16 The cost of an outpatient physician visit is based on the 2005 CMS payment rates for CPT 99215 (office or other outpatient visit). 17

Costs of Adverse Events
The model includes treatment costs for Grade 3 or Grade 4 adverse events (AEs) with an incidence rate of 5% or greater, as listed in the prescribing information for each of the therapies. AEs requiring hospitalizations were included in the model regard-less of magnitude of incidence if reported in the pivotal trial publications of the respective therapies (see Table 3 for details on costs of AEs). All patients experiencing Grade 3 or 4 AEs (severe AEs) related to erlotinib were assumed to require 1 outpatient visit.
In clinical trials for erlotinib, rash occurred within 2 weeks of initiating the treatment. Based on treatment recommendations by clinicians, patients experiencing a rash were assumed to receive clindamycin gel for the duration of treatment. 18 Patients experiencing diarrhea while taking erlotinib were assumed to receive treatment with loperamide. Costs of hospitalizations due to severe diarrhea are assumed to include costs of hospitalization as well as costs due to inpatient physician visits.
Febrile neutropenia is assumed to require hospitalization and inpatient physician visits. Patients with severe anemia are assumed to have 1 outpatient visit. In addition, a proportion of patients with anemia are assumed to receive treatment with erythropoietin (Epogen) or a red blood cell (RBC) count transfusion. Proportions of patients receiving erythropoietin or RBC transfusions are obtained from the publication of the pivotal trial that compared pemetrexed with docetaxel in the secondline treatment of NSCLC patients. 6 Patients with neutropenia (nonfebrile) were assumed to have 1 additional outpatient physician visit. In addition, a proportion of patients are assumed to receive 1 course of granulocyte colony stimulating factors (G-CSF) as prophylaxis for neutropenia. The proportion of patients receiving G-CSF is based on the pivotal trial publication of pemetrexed compared with docetaxel in the second-line treatment of NSCLC patients. 6 Hospitalization rates due to neutropenic fever and other drugrelated AEs following treatment with pemetrexed or docetaxel are based on published studies. 6 Costs for hospitalization were based on Medicare reimbursement rates for the aforementioned complications of interest. Costs for outpatient visits were based on Medicare fee schedules. Drug costs are based on national reimbursement schedules provided by Price-Chek PC, a database published by Medispan/First DataBank. 19

■■ Analysis
The model was used to estimate the impact of erlotinib on expenditures over 1 year after the drug was made available as a benefit for members of a hypothetical health plan with a total enrollment of 500,000 persons. Analyses of the budget impact of erlotinib focused on comparing costs of treating stage IIIb/IV NSCLC in the second-and third-line treatment situation between scenarios where erlotinib is and is not a treatment option. In addition to aggregate costs to the health plan, an estimate of costs per member per month (PMPM) was computed. The PMPM costs are calculated by dividing the total cost to the plan by the number of members in the plan, then dividing by 12. After analyses were performed using base-case (default) esti-Medication Use* and Costs † of Pemetrexed, Docetaxel, and Erlotinib mates for the parameters, sensitivity analyses were performed on key model parameters to assess the robustness of the model results. Sensitivity analyses were conducted to evaluate the influence of variation in key parameters such as the percentage of patients receiving erlotinib, treatment costs, rates of AEs on the budget impact in the second-line situation only, and in second-and third-line situations together. Parameters were varied across confidence intervals, when available from clinical studies, or by ±25%, when data were unavailable.

Results Based on Default Parameter Estimates
Based on age-adjusted incidence and observed rates of secondand third-line therapy (21% and 11% of stage IIIb/IV NSCLC patients, respectively), the model estimates that in a hypothetical health insurance plan with 500,000 covered lives, 98 patients with advanced NSCLC will be eligible for second-and third-line therapy during the year of analysis (Table 1). Table 4 lists annual health plan costs of treatment of second-and third-line stage IIIb/IV NSCLC, including drug costs, administration costs, and costs related to AEs. The total expected cost of treating secondand third-line stage IIIb/IV NSCLC is estimated to be $380,968 when erlotinib is not a treatment option. With erlotinib as a treatment option, the treatment cost is increased to $382,418. Similarly, in the second-line situation only, total expected cost is $377,165 when erlotinib is not a treatment option and $308,582 when erlotinib is a treatment option. Expressed as changes in PMPM costs for all members in the health plan, the model estimated that the addition of erlotinib changes these costs by approximately $0.01.

Sensitivity Analyses
Sensitivity analyses demonstrate that estimates of the budget impact of adding erlotinib as a treatment option in second-and third-line treatment situations are most sensitive to (in order of importance) the unit cost of erlotinib, the proportion of patients who are switched from pemetrexed to erlotinib, the proportion of patients switched from docetaxel to erlotinib, and the proportion of patients receiving second-and third-line treatment (Table 5). The cost of managing AE rates for all drugs did not substantially influence the outcome. The expected difference in costs between scenarios is a $1,450 savings with erlotinib available. Using multiway sensitivity analysis, the 90% confidence interval ranges from a $61,376 savings to $29,855 higher costs for the erlotinib scenario.

■■ Discussion
The purpose of this erlotinib budget impact model is to estimate the impact on health plan budgets of introducing erlotinib as second-and third-line therapy for patients with advanced   NSCLC. In a health plan of 500,000 enrollees, including erlotinib on the formulary has a modest positive impact on health plan expenditures. The higher direct drug costs of erlotinib are offset by the reduced costs of administration and fewer costs incurred for treatment of severe side effects when compared with pemetrexed and docetaxel. In 1-way sensitivity analysis, the results are relatively insensitive to estimates of the incidence of erlotinib-related AEs. Multiway sensitivity analyses suggest that the difference in costs between the regimens is unlikely to be statistically significant. This decision model suggests that health plans can include erlotinib on their formulary with a relatively low impact on their annual health care budget if erlotinib is used as labeled and in place of rather than in addition to traditional second-line chemotherapy. The budget impact of previous changes in chemotherapy regimens, notably taxanes, has been more substantial than erlotinib. 20 Some chemotherapeutic substitutions may result in cost savings for treating NSCLC patients in certain situations. 21 The budget impact model follows NCCN-recommended guidelines in the third-line situation, where erlotinib is the only recommended treatment alongside best supportive care. If, in practice, other types of chemotherapy are used in place of erlotinib, then substituting erlotinib for these agents may result in savings in this line of treatment. On the other hand, if other chemotherapy agents are typically added alongside erlotinib as third-line treatment, total treatment costs could be higher than predicted.

Limitations
This model does not consider the other labeled indication for erlotinib-locally advanced pancreatic cancer-nor does it consider "off label" uses, such as head and neck cancer. 22 Although these cancers are relatively uncommon in workingage adults, use of erlotinib in these situations will increase PMPM costs for the health plan because, in these cancers, erlotinib is used in addition to rather than in place of other therapies.
Therapy costs are ultimately influenced by patient adherence to treatment schedules. We assume that patients purchase (and health plans incur) the cost of a full erlotinib prescription, whether or not the patient ultimately takes all the medication as prescribed. This is a conservative assumption; that is, it negatively affects erlotinib relative to the alternatives. In a review of the literature, Lucero and colleagues note that adherence may be similar for both injectable and oral therapies. 23 This budget impact model does not consider efficacy beyond its immediate impact on expenditures. A more formal evaluation of the cost-effectiveness of erlotinib, with outcomes expressed as years of life gained or quality-adjusted life-years, would provide a more complete picture of the value of this drug as second-and third-line therapy for patients with non-small cell lung cancer.

■■ Conclusion
Erlotinib, a new epidermal growth factor receptor tyrosine kinase inhibitor that has been shown to modestly improve survival for patients with advanced stage non-small cell lung cancer, is expected to replace standard second-and third-line chemotherapy use in these patients. Because erlotinib is costly, we constructed a decision model to estimate the budget impact of adding erlotinib to the formulary. Although drug expenditures increase when erlotinib replaces existing treatments, the net budget impact to health plans of adopting erlotinib is quite modest due to erlotinib's superior side-effect profile. The estimates only consider use that is consistent with product labeling. Off-label use for lung or other cancers, if common, could have more substantial impacts on net health plan expenditures.

DISCLOSURES
Funding for this study was provided by Genentech, Incorporated, and was obtained by author Scott D. Ramsey. Authors Tripthi V. Kamath and Deborah Lubeck are employed by Genentech, Incorporated. Ramsey and author Lauren Clarke disclose no potential bias or conflict of interest relating to this article. Ramsey served as principal author of the study. Study concept and design were contributed by all authors. Data collection was the work of Ramsey and Clarke; data interpretation was the work of all authors. Writing of the manuscript was primarily the work of Ramsey, with input from Clarke; its revision was the work of Ramsey.