Abstract: Value Pathways powered by NCCN™ provides oncologists with a list of clinically proven treatment options that support the delivery of high-quality, cost-effective cancer care. We evaluated the economic and clinical impact of Value Pathways in patients with metastatic mutation-negative non-small cell lung cancer (NSCLC) who received first-line systemic therapy at US Oncology Network practices participating in the Oncology Care Model. Eligible Medicare patients initiating systemic therapy between July 1, 2016, and June 30, 2018, (N=584) were matched by age and sex (438 in the on-pathway and 146 in the off-pathway cohort). The mean total cost of care was 19% lower for patients treated on-pathway vs off-pathway ($47,287 vs $58,564; P <.0001). Chemotherapy (Part B) drug costs were 40% lower for the on- vs off-pathway cohorts ($24,967 vs $41,806; P <.0001). Median overall survival was not significantly different between the two cohorts (11.9 vs 13.4 months, P=.239). This analysis indicates that treating patients in accordance with evidence-based Value Pathways lowers the total cost of care to patients with mutation-negative metastatic NSCLC while maintaining survival benefit.
Key Words: non-small cell lung cancer, Value pathways, total cost of care, overall survival
Lung cancer, the second most commonly diagnosed cancer in the United States, is the leading cause of cancer-related deaths, with an estimated 228,820 new cases and 135,720 deaths from the disease in 2020.1,2 Non–small cell lung cancer (NSCLC) accounts for approximately 84% of lung cancer cases in the United States. Survival rates for lung cancer vary based on the stage of disease at diagnosis. Approximately 60% of patients present with distant metastases. Until recently, chemotherapy was the only available treatment option for patients with advanced NSCLC whose tumors lacked specific mutations that allowed for treatment with targeted therapies. Although chemotherapy was associated with clinical benefit, durable responses were rarely observed and prognosis for these patients was poor, with a 5-year survival rate of approximately 5%.2 Recent advances in the development of immune checkpoint inhibitors (ICIs) have provided novel options for the treatment of mutation-negative metastatic NSCLC, allowing for durable responses compared with chemotherapy and improvement of 5-year overall survival (OS) rates to almost 15%.3-5
New treatment options have increased the cost of lung cancer care. In 2018, NSCLC was the fifth costliest cancer in the United States, with an estimated 2018 national expenditure of $14.2 billion.6 As novel, more efficacious, and more expensive treatment options become available, oncologists must balance the selection of treatments that deliver better patient outcomes with the societal responsibility to provide care of the highest value without compromising efficacy. The Oncology Care Model (OCM), introduced by the Center for Medicare & Medicaid Innovation (CMMI) in 2016, is a total cost of care model that requires participating practices to maintain or improve the quality of care delivered to patients while simultaneously reducing the total cost of care.7,8 Thus, as the cost of cancer drugs keeps rising, delivering consistent, high-quality, and cost-effective care continues to be a challenge.
In 2006, a team of 13 physicians in The US Oncology Network (USON) developed clinical pathways with the goal of helping their oncologists identify the regimens that provided better value when multiple therapeutic options were available for a given patient.9 Clinical pathways provide physicians with a choice of clinically proven treatment options for each patient while balancing clinical outcomes, treatment toxicities, and cost. These pathways are reviewed and updated frequently to incorporate the most recent clinical evidence based on new literature, as well as newly approved therapies. In 2013, USON entered into a partnership with the National Comprehensive Cancer Network® (NCCN®). The additional clinical expertise provided by the evidence-based NCCN Guidelines® led to the creation of “Value Pathways powered by NCCN™” (Value Pathways), which were integrated into USON’s system-wide electronic health record (EHR), iKnowMed (iKM)SM. This tool presents physicians with recommended pathway choices and also allows physicians the option of choosing an off-pathway regimen when clinically appropriate. Thus, incorporating Value Pathways into routine cancer care can help simplify the integration of evidence-based best practices and improve or maintain the quality of care delivered while reducing costs for both patients and payers.
The benefit of using pathways was evaluated in an earlier analysis of costs and outcomes in NSCLC conducted in a private payer setting and before the introduction of OCM (2006-2007).10 This analysis showed that implementing pathways maintained the same clinical outcomes, such as OS, but at a lower cost. The retrospective cohort study presented here analyzes data from patients with metastatic NSCLC who were treated at USON practices participating in the OCM to determine the impact of following clinical pathways by comparing total cost of care and patient survival in patients who were treated on-pathway vs off-pathway. This study evaluates the use of Value Pathways in an older population with more comorbidities and in a market landscape that includes substantially more expensive therapies, including ICIs.
Methods
This multicenter, retrospective, observational study evaluated patients with metastatic mutation-negative NSCLC who received systemic therapy in the first-line setting at USON practices that participated in the OCM. The primary objective of the study was to compare the total cost of care for patients receiving treatment for mutation-negative metastatic NSCLC using on-pathway regimens vs off-pathway regimens. The secondary objective was to determine OS for the two cohorts.
The study population included all Medicare patients enrolled in the OCM from the iKM database who received first-line chemotherapy for metastatic NSCLC starting between July 1, 2016, and June 30, 2018. To avoid confounding of costs and survival estimates, patients enrolled into clinical trials; patients with a concurrent diagnosis of a cancer other than NSCLC; patients who had missing treatment, survival, or cost data in either the iKM or the Centers for Medicare & Medicaid Services (CMS) database; and all patients whose tumors expressed genetic mutations in BRAF, ROS1, EGFR, or ALK were excluded from the analysis. Patients covered by commercial payers (including Medicare Advantage plans), Medicaid, those uninsured, or those who did not qualify for participation in OCM were also excluded. To eliminate possible selection bias, as well as differences in baseline characteristics that could affect cost and clinical outcomes, propensity scores were developed using logistic regression and used to match patients on the variables of age and sex.11,12 A 1:3 ratio was obtained for patients matched off-pathway to on-pathway using a greedy algorithm. If an off-pathway patient could not be matched to three on-pathway patients with similar propensity scores (a maximum distance between scores of 0.1), the patient was not included in the matched cohort. Similarly, unmatched on-pathway patients were also excluded from the analysis. Details on treatment selections and outcomes were obtained from iKM.
All included patients were assigned a pathway status depending on concordance with Value Pathways for NSCLC. Patients were classified as treated on-pathway if they received care according to the recommendations of Value Pathways and off-pathway if they were treated with regimens not included in Value Pathways.
Cost of care details from the CMS claims database were available to all practices participating in OCM, and clinical outcomes were extracted from iKM. The total cost of care for patients treated with on-pathway vs off-pathway regimens was determined using the Kaplan-Meier Sample Average approach. All charges incurred after the initiation of chemotherapy were obtained from the CMS database and included in the analysis. Medicare-assigned benchmark prices were obtained from the OCM database and assigned for the time period in which each patient was enrolled and treated. Total cost of care included all costs from the start of first-line metastatic treatment (the first day of the first chemotherapy cycle for metastatic disease) through follow-up and was estimated using the Wilcoxon rank sum test.
OS was estimated by the Kaplan-Meier method and was calculated from the start of first-line metastatic treatment to the date of death.
Results
Patients
A total of 1270 Medicare patients who were enrolled in the OCM and treated at USON practices initiated systemic therapy for first-line treatment of NSCLC between July 1, 2016, and June 30, 2018. After excluding patients who did not meet eligibility criteria, 1242 patients were identified, of whom 1096 were treated on-pathway and 146 were treated off-pathway. Propensity score matching of patients resulted in the identification of 584 patients matched in a 3:1 ratio, with 438 in the on-pathway cohort and 146 in the off-pathway cohort.
Patient characteristics in the on- and off-pathway cohorts are shown in Table 1. Consistent with the Medicare-aged population being studied, approximately 95% of patients were aged 65 years or older, with 39% of patients aged 75 years old or older. Patients in the two groups were similar with respect to stage at diagnosis and comorbidities at baseline. A larger proportion of patients in the off-pathway cohort had unknown Eastern Cooperative Oncology Group performance status, resulting in possible differences in the functional status of patients in the two groups. The majority of patients had non-squamous NSCLC (73%) and 22% had no documented non-cancer comorbidities based on the CMS-HCC Risk Model.
The most commonly used on-pathway treatment regimens were carboplatin/pemetrexed, carboplatin/paclitaxel, and single-agent pembrolizumab, accounting for 37%, 36%, and 14% of on-pathway regimens, respectively. The most common treatment regimens used off-pathway were pembrolizumab/carboplatin/pemetrexed, ingle-agent nivolumab, single-agent pembrolizumab, and carboplatin/albumin-bound paclitaxel, accounting for 22%, 12%, 12%, and 10% of off-pathway regimens, respectively. It should be noted that single-agent pembrolizumab could be used either on- or off-pathway based on PD-L1 levels. Based on the strength of data available during the study period, pembrolizumab + chemotherapy was not included as an on-pathway option. White blood cell growth factors were not used for patients in either cohort.
Cost Analysis
Patients who were treated on-pathway had a 19% lower mean total cost of care than those who were treated off-pathway ($47,287 ± 25,193 vs $58,564 ± 29,975;
P <.0001; Table 2). The mean Medicare-assigned benchmark price, which was determined using the OCM database for the time period when the patient was enrolled and treated, did not differ significantly between the two groups. However, the absolute difference between the benchmark price and the total cost of care was significantly lower for the on-pathway group when compared with the off-pathway group ($4462 ± 26,902 vs $17,846 ± 31,581; P <.0001).
An analysis of the different contributing costs showed that the costs for chemotherapy (Part B) drugs were 40% lower in patients who were treated on-pathway compared with those who received off-pathway therapy ($24,967 ± 22,437 vs $41,806 ± 33,836; P <.0001). There were no significant differences in the costs associated with inpatient admissions, emergency room visits, radiation oncology, physician services, outpatient care, and hospice care between the two cohorts. Thus, the costs for chemotherapy drugs were the major contributors to the total cost of care.
Overall Survival
At median follow-ups of 9 months (range, 0.2, 27.7) in the on-pathway cohort and 7.6 months (range, 0.5, 28.6) in the off-pathway cohort, median OS was not significantly different between the two groups (11.9 vs 13.4 months,
P = .239), indicating that following pathways had no adverse effect on clinical outcomes. At 12 months, OS rates for the on-pathway and off-pathway populations were 50% and 54%, respectively (Figure 1).
Discussion
Significant changes in the overall management of metastatic NSCLC have occurred in the last decade with the introduction of new effective therapies, such as ICIs, that have prolonged progression-free survival and OS.3-5 Newer therapies have resulted not only in substantial improvements in patient outcomes but also in an increasing financial burden on the health care system. Over the past decade, several studies have evaluated the impact of incorporating evidence-based clinical cancer pathways into routine clinical practice with the goal of reducing drug spend for cancer care while maintaining clinical outcomes.10,13-15 This retrospective analysis, based on recent real-world data in Medicare patients with metastatic NSCLC, shows that following alue Pathways enabled a significant 19% reduction in the total cost of care while OS rates remained comparable. The minimal impact on survival benefit in the off-pathway group indicates that the added cost of off-pathway therapies does not translate into improved outcomes.
It is important to point out that the inclusion of therapies in Value Pathways is based on a careful consideration of the available clinical evidence.9 Expensive drugs, such as pembrolizumab, nivolumab, and bevacizumab, are included as options in Value Pathways for the appropriate patients when supported by robust clinical evidence. These results reinforce our previous study10 demonstrating that treating patients with therapies that do not meet pathway inclusion criteria does not improve outcomes. Furthermore, these findings hold true even in an older population, which includes frailer patients with multiple comorbidities.
As expected, the lower total cost of care observed with the on-pathway group resulted, to a large extent, from the significantly lower cost of chemotherapy Part B drugs. The use of non-chemotherapy drugs, such as supportive care agents, were not included in this analysis, but did not contribute to changes in total cost of care. There were no significant differences in other costs that contributed to the total cost of care, such as physician services, outpatient care costs, emergency room visits, or hospitalization costs.
Limitations of this analysis include its retrospective and observational nature, as well as the relatively small patient populations. Although patients were matched in the on- and off-pathway cohorts using propensity scoring, these scores may not have accounted for all factors associated with total cost of care or survival, such as performance status, which could affect treatment decisions. It should also be noted that the performance status of approximately 40% of patients was unknown in the off-pathway cohort. In addition, patients with unknown biomarker status according to iKM were assumed to be mutation-negative, although it is likely that the treating physician was aware of the mutation-negative status of the patient when selecting an appropriate treatment option. Nonetheless, the strength of this study lies in the wealth of clinical data obtained from the iKM EHR database. These clinical treatment and survival data, when used in conjunction with the CMS claims data, allow us to study the impact of Value Pathways on costs and outcomes.
Conclusion
In the current environment, in which a plethora of new and expensive therapies are becoming available, the use of an evidence-based approach with platform integration, such as Value Pathways, can help oncologists safely and effectively integrate new treatments into routine practice, while reducing the cost of cancer care.
References
1. Siegel RL, Miller KD, Jemal A. Cancer statistics, 2020. CA Cancer J Clin. 2020;70(1):7-30. doi:10.3322/caac.21590
2. Howlader N, Noone AM, Krapcho M, et al. SEER Cancer Statistics Review, 1975-2017. National Cancer Institute, based on November 2019 SEER data submission. Published April 2020. Accessed October 6, 2020. https://seer.cancer.gov/csr/1975_2017/
3. Brahmer JR, Govindan R, Anders RA, et al. The Society for Immunotherapy of Cancer consensus statement on immunotherapy for the treatment of non-small cell lung cancer (NSCLC). J Immunother Cancer. 2018;6(1):75. doi:10.1186/s40425-018-0382-2
4. Reck M, Rodriguez-Abreu D, Robinson AG, et al. Updated analysis of KEYNOTE-024: Pembrolizumab versus platinum-based chemotherapy for advanced non-small-cell lung cancer with PD-L1 tumor proportion score of 50% or greater. J Clin Oncol. 2019;37(7):537-546. doi:10.1200/JCO.18.00149
5. Topalian SL, Hodi FS, Brahmer JR, et al. Five-year survival and correlates among patients with advanced melanoma, renal cell carcinoma, or non-small cell lung cancer treated with nivolumab. JAMA Oncol. 2019;5(10):1411-1420. doi:10.1001/jamaoncol.2019.2187
6. National Cancer Institute. Cancer Trends Progress Report National Cancer Institute. Published March 2020. Accessed October 6, 2020. https://progressreport.cancer.gov
7. Hoverman JR FJ, Staggs S, et al. From angst to analytics: lessons learned from an Oncology Care Model internal pilot. J Commun Supp Oncol. 2017;15(6):e297-e301.
8. Kline RM, Bazell C, Smith E, Schumacher H, Rajkumar R, Conway PH. Centers for medicare and medicaid services: using an episode-based payment model to improve oncology care. J Oncol Pract. 2015;11(2):114-116. doi:10.1200/JOP.2014.002337
9. Neubauer M. Clinical pathways: reducing costs and improving quality across a network. Am J Manag Care. 2020;26(2):SP60-SP61. doi:10.37765/ajmc.2020.42550
10. Neubauer MA, Hoverman JR, Kolodziej M, et al. Cost effectiveness of evidence-based treatment guidelines for the treatment of non-small-cell lung cancer in the community setting. J Oncol Pract. 2010;6(1):12-18. doi:10.1200/JOP.091058
11. Austin PC. An introduction to propensity score methods for reducing the effects of cconfounding in observational studies. Multivariate Behav Res. 2011;46(3):399-424. doi:10.1080/00273171.2011.568786
12. Haukoos JS, Lewis RJ. The propensity score. JAMA. 2015;314(15):1637-1638. doi:10.1001/jama.2015.13480
13. Jackman DM, Hamilton J, Foster E, et al. Lessons from the front: designing and implementing clinical pathways by and for clinicians. Am J Manag Care. 2020;26(spec 2):SP57-SP59. doi:10.37765/ajmc.2020.42549
14. Jackman DM, Zhang Y, Dalby C, et al. Cost and survival analysis before and after implementation of Dana-Farber clinical pathways for patients with stage IV non-small-cell lung cancer. J Oncol Pract. 2017;13(4):e346-e352. doi:10.1200/JOP.2017.021741
15. Hertler A, Chau S, Khetarpal R, et al. Utilization of clinical pathways can reduce drug spend within the Oncology Care Model. J Oncol Pract. 2020;16(5):e456-e463. doi:10.1200/JOP.19.00753
