- Email: [email protected]
Treatment Patterns and Clinical Outcomes Among Metastatic Non–Small-Cell Lung Cancer Patients Treated in the Community Practice Setting
Accepted Manuscript Treatment Patterns and Clinical Outcomes Among Metastatic Non-Small Cell Lung Cancer Patients Treated in the Community Practice Setting Eric Nadler, Janet L. Espirito, Melissa Pavilack, Marley Boyd, Andrea Vergara-Silva, Ancilla Fernandes PII:
S1525-7304(18)30021-4
DOI:
10.1016/j.cllc.2018.02.002
Reference:
CLLC 745
To appear in:
Clinical Lung Cancer
Received Date: 30 October 2017 Revised Date:
1 February 2018
Accepted Date: 12 February 2018
Please cite this article as: Nadler E, Espirito JL, Pavilack M, Boyd M, Vergara-Silva A, Fernandes A, Treatment Patterns and Clinical Outcomes Among Metastatic Non-Small Cell Lung Cancer Patients Treated in the Community Practice Setting, Clinical Lung Cancer (2018), doi: 10.1016/j.cllc.2018.02.002. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
Treatment Patterns and Clinical Outcomes Among Metastatic NonSmall Cell Lung Cancer Patients Treated in the Community Practice
RI PT
Setting Eric Nadler,1 Janet L. Espirito,2 Melissa Pavilack,3 Marley Boyd,2 Andrea Vergara-Silva,3 Ancilla Fernandes3 1
Texas Oncology, 3410 Worth St #400, Dallas, TX 75246, USA
2
SC
McKesson Specialty Health, 10101 Woodloch Forest, The Woodlands, TX, USA
3
M AN U
AstraZeneca, 950 Wind River Lane, Gaithersburg, MD 20878, USA
Address for correspondence: Eric Nadler, MD Texas Oncology
3410 Worth St. HR081 Dallas, TX 75246
Phone: +1 (214) 370-1000
TE D
Email: [email protected]
Target Journal: Clinical Lung Cancer
Word Count (Body only): 3186 of 8000 Max
AC C
EP
Figures/Tables: 7 of 7 max
1
ACCEPTED MANUSCRIPT
Conflict of Interest/Disclosure Dr Nadler has received grants from AstraZeneca; financial compensation for speaker bureau activities from Genentech, Merck, and Eli Lilly; and financial compensation for consulting roles from Eli Lilly. Dr Espirito is an employee of and owns stocks in McKesson Specialty Health. Dr
RI PT
Pavilack is an employee of AstraZeneca. Dr Boyd has received grants from AstraZeneca and is an employee of McKesson Specialty Health. Drs Vergara-Silva and Fernandes are employees of and own stocks in AstraZeneca. Dr Vergara-Silva has received nonfinancial support (travel
AC C
EP
TE D
M AN U
SC
expenses) from AstraZeneca.
2
ACCEPTED MANUSCRIPT
Micro Abstract (word limit 60 max; current = 57) New targeted and immunotherapies expand the treatment options for mNSCLC. This study assessed contemporary treatment patterns and outcomes using a large EHR database of >10,000 patients. The majority of targeted and immunotherapy use was in the 2L and 3L
AC C
EP
TE D
M AN U
SC
actionable mutations have a survival advantage.
RI PT
treatment setting. Survival differences observed by treatment types suggest that patients with
3
ACCEPTED MANUSCRIPT
Abstract (word limit 250 max; current = 247)
RI PT
Introduction: Multiple therapeutic options now exist for metastatic non-small cell lung cancer (mNSCLC). This study evaluated treatment patterns and outcomes in mNSCLC patients receiving first-line (1L), second-line (2L), and third-line (3L) therapy.
Patients and Methods: A retrospective, observational cohort study was conducted using an
SC
electronic health record database of mNSCLC patients who received initial treatment from
January 2012 through April 2016, with follow-up through June 2016. Patient characteristics and treatment patterns were characterized. Overall survival (OS) was assessed using the Kaplan-
M AN U
Meier method.
Results: 10,689 1L patients were identified. Median age was 68 years, and 54% were male. Most patients (59%) had a performance status of 1, and 77% had nonsquamous histology. 1L treatment was chemotherapy in 93% of patients, and targeted therapy in 6%. Median OS (mOS) for all patients in 1L was 12.3 mo (95% confidence interval [CI], 11.9-12.7), and 24.3 mo in 1L
TE D
patients receiving targeted therapy. Among patients who received 2L therapy (n=4235), 66%, 17%, and 17% received chemotherapy, targeted therapy, and immunotherapy, respectively. mOS from 2L therapy was 9.6 mo (95% CI, 9.1-10.1). In patients receiving 3L therapy (n=1580), 58%, 22%, and 19% received chemotherapy, targeted therapy, and immunotherapy,
EP
respectively. mOS from 3L therapy was 8.2 mo (95% CI, 7.3-8.7).
Conclusion: Targeted therapy and immunotherapy was most frequently used in the 2L and 3L
AC C
setting during the study time frame. Survival differences observed by treatment types are likely due to biologic differences, and suggest that patients with actionable mutations have a survival advantage.
Keywords (5 max): NSCLC, Targeted therapy, Immunotherapy, Chemotherapy, Real-world experience
4
ACCEPTED MANUSCRIPT
Introduction Lung cancer is the leading cause of cancer-related deaths in both men and women in the United States. In 2017, approximately 222,500 new cases and 155,870 deaths due to lung cancer are estimated.1 The majority of patients are diagnosed with metastatic non-
RI PT
small cell lung cancer (mNSCLC), and treatment is largely palliative. The estimated
5-year survival for patients with mNSCLC from 2007 to 2013 was 4.5%.2 The majority (70% to 75%) of NSCLC subtypes are grouped into the nonsquamous histologic
category, with the remaining grouped into the squamous category, which accounts for
SC
approximately 25% to 30%.3
Platinum-based combination chemotherapy has been the standard of care for treatment of mNSCLC for several years, and is associated with improved survival, symptom control, and
M AN U
improved quality of life versus best supportive care.4 However, treatment of mNSCLC has evolved over the past decade with the development of new therapeutic options for metastatic disease.5-8 Several biomarkers such as epidermal growth factor receptor (EGFR), anaplastic lymphoma kinase (ALK), ROS1 receptor tyrosine kinase (ROS1), and programmed death cell ligand-1 (PD-L1) have emerged as predictive for therapeutic efficacy.4 Therapies targeting these molecules were initially approved by the U.S. Food and Drug Administration (FDA) after disease
TE D
progression on at least one prior chemotherapy regimen.9-11 Subsequent studies have demonstrated improved progression-free survival (PFS) in patients with identified mutations who receive targeted therapy compared with standard chemotherapy, and EGFR and ALK inhibitors are now recommended as first-line (1L) therapy in patients with known mutations.4,12,13
EP
However, a majority of patients do not have targetable gene abnormalities, and even among tumors that harbor these mutations, resistance to therapy eventually develops.4 Novel targeted therapies are also now available for patients with known resistance mutations emerging after
AC C
EGFR TKI therapy14 or for patients with ALK-positive disease and progression after crizotinib.15 In 2015, the first immunotherapy for mNSCLC was FDA-approved for use after progression on platinum-based chemotherapy and in patients with EGFR mutations or ALK rearrangements after progression on targeted therapy.16 Multiple therapeutic options now exist for mNSCLC, including chemotherapy, targeted therapy, and immunotherapy, creating questions about how best to integrate these treatments into the course of care.4 To date, no studies have evaluated the best sequence of therapies and, therefore, treatment selection often depends on tumor molecular profile, performance status, and toxicities.4 This study aims to understand how these different treatment options currently 5
ACCEPTED MANUSCRIPT
available are sequenced during the course of therapy in this patient population. The objectives were (1) to describe the demographic and clinical characteristics of patients with mNSCLC initiating treatment with chemotherapy, targeted therapy, or immunotherapy; (2) to evaluate realworld systemic treatment patterns among patients with metastatic disease by line of therapy
RI PT
(LOT); and (3) to evaluate overall survival (OS) with systemic treatment for patients with mNSCLC by LOT and treatment category.
Materials and Methods
This was a retrospective observational cohort study of patients with mNSCLC. Data were
SC
obtained via programmatic queries of The US Oncology Network iKnowMed (iKM) electronic health record (EHR) system to collect the structured, documented information available from the records. McKesson Specialty Health maintains iKM, an integrated web-based database and
M AN U
oncology-specific EHR system that captures outpatient practice encounter histories from network community oncology practices affiliated with over 1000 physicians in more than 25 practices across 400 sites of care in 19 states. Vital status data were supplemented with information obtained from the U.S. Department of Social Security Death Index. Institutional review board approval was obtained.
TE D
The study population included patients with documented mNSCLC, defined as either stage IV at diagnosis or patients diagnosed with earlier stages who later developed evidence of metastatic disease. Patients were required to receive their initial treatment for metastatic disease between January 1, 2012, and April 30, 2016, with follow-up through June 30, 2016, or until date of last record, whichever occurred first, for a minimum potential follow-up of 2 months. The study time
in 2015.
EP
period was selected to capture maximum data on immunotherapy utilization following approval
AC C
Patients were required to have ≥2 visits within The US Oncology Network during the study period. Patients were excluded if they were aged <18 years at diagnosis, had another documented cancer prior to their mNSCLC, or, to avoid confounding effects of investigational therapies, if they were enrolled in a clinical trial before their initial treatment for metastatic disease.
All patients who met eligibility criteria for the treatment of mNSCLC were included in the study. Assignment of patients into treatment cohorts was based on their 1L mNSCLC regimen and followed for up to 3 lines of therapy. Systemic treatments and LOTs, including maintenance therapy (after 1L) during the study period, were programmatically sequenced based on 6
ACCEPTED MANUSCRIPT
treatment dates and categorized as chemotherapy, targeted therapy, or immunotherapy. Regimens received by <10% of patients were categorized as “other.”
Statistical Analysis
RI PT
Descriptive analyses were conducted to assess demographic, clinical, and treatment characteristics among the cohorts. Wilcoxon rank-sum tests were performed for continuous variables, and chi-square tests were used for categorical variables when all cell counts were ≥5. The Fisher test for categorical variables was used when cell counts were <5. Missing data were identified and reported as percentages. The Kaplan-Meier method was used to estimate OS,
SC
stratified by LOT and treatment category. Statistical analyses were performed using SAS 9.4
M AN U
(SAS Institute, Cary, NC).
Results
Demographics and Baseline Characteristics
A total of 10,689 patients who met eligibility criteria and received 1L therapy during the study period were included in the analysis. The 1L treatment cohorts consisted predominantly of
TE D
patients who initiated 1L treatment with chemotherapy (93%), 6% of patients with targeted therapy, and <1% of patients with immunotherapy (Figure 1). The median age at the start of therapy was 68 years (range 27 to ≥90); 79% were white and 54% were male. The majority of patients (64%) were initially diagnosed with stage IV disease. Most patients (59%) had an
EP
Eastern Cooperative Oncology Group (ECOG) performance status of 1, 77% had nonsquamous histology, and 85% were former or current smokers. Among patients receiving initial targeted therapy, most were never smokers (48%) versus those who received initial chemotherapy and
AC C
immunotherapy (<10% for both; P < .0001). Tumor markers were assessed through documented structured data in the EHR. EGFR and ALK status was available in approximately 40% and 35% of patients, respectively. Among those with documented results, EGFR status was positive in 13%; ALK was positive in 3%; PD-L1 status was known in <1%; and ROS1 was not collected in this study (Table 1).
Treatment Patterns The most frequently used chemotherapies in the 1L setting were carboplatin + paclitaxel (37%), carboplatin + pemetrexed (21%), and carboplatin + paclitaxel + bevacizumab (13%). Among 7
ACCEPTED MANUSCRIPT
patients receiving targeted therapy in the 1L, erlotinib (57%) and crizotinib (12%) were most often used (Table 2).
Sixty percent of the 1L patients did not receive second-line (2L) therapy within the study period.
RI PT
Among the 40% of 1L patients (n = 4235) who did receive 2L therapy, 66%, 17%, and 17% received chemotherapy, targeted therapy, and immunotherapy, respectively. Pemetrexed or docetaxel monotherapy was most commonly used for 2L chemotherapy, erlotinib was most used as 2L targeted therapy, and nivolumab was most used for 2L immunotherapy. Among patients who received 1L chemotherapy and continued to 2L, the majority continued with 2L
SC
chemotherapy (27%). Among patients who received 1L targeted therapy and continued to 2L, the majority continued with 2L targeted therapy (33%). Among patients who received
bevacizumab (15%; data not shown).
M AN U
maintenance therapy after 1L, pemetrexed (48%) was most frequently used, followed by
For patients who received 2L chemotherapy (n = 2790) and continued to receive third-line (3L) treatment (n = 1216), the majority continued with 3L chemotherapy (57%). For patients who received 2L targeted therapy (n = 718) and went on to receive 3L treatment (n = 244), more than half (52%) switched to chemotherapy in the 3L, and 34% continued with targeted therapy
TE D
in the 3L. Among all patients who received 3L therapy during the study period (n = 1580), immunotherapy treatment with nivolumab was most frequently used (17%).
EP
Clinical Outcomes
The median follow-up time from 1L treatment initiation was 6.9 months (range 0 to 54 months) for all patients in the study. Kaplan-Meier survival curves for patients from the initiation of each
AC C
line of treatment are shown in Figure 2, Figure 3, and Figure 4. Median OS (mOS) for all patients in 1L was 12.3 months (95% confidence interval [CI], 11.9-12.7). Patients who received targeted therapy in 1L were observed to have the longest survival of 24.3 months. In all patients who received 2L therapy, mOS from start of 2L was 9.6 months (95% CI, 9.1-10.1). mOS was similarly observed to be longest for patients who received targeted therapy in 2L, at 11.2 months. In all patients who received 3L therapy, the mOS from the start of 3L was 8.2 months (95% CI, 7.3-8.7). The mOS appeared longest, at 11.3 months, in patients who received 3L immunotherapy.
8
ACCEPTED MANUSCRIPT
Among the subset of patients receiving only 1L therapy of any type (n = 6396), mOS was 7.2 mo (95% CI, 6.9-7.5). The mOS among patients who received only 1L and 2L therapy of any type (n = 2623) was 14.7 mo (95% CI, 13.8-15.4). The mOS among patients who received 1L,
RI PT
2L, and 3L therapy of any type (n = 1568) was 23.0 mo (95% CI, 21.8-24.8; Figure 5).
Discussion
A large EHR database population was used to examine real-world patient characteristics,
SC
treatment patterns, and clinical outcomes involving the use of chemotherapy, newer targeted therapies, and immunotherapies for mNSCLC. Use of EHR database information provided insight into initial treatment utilization in real-world clinical practice, including at a time when
M AN U
biomarker testing was not universal and the impact of mutational status was not as well known.
In this study, the majority of the patient population received initial platinum-based chemotherapy in the 1L setting (71%), as this has been the standard of care for several years and many of the new drugs that gained FDA approval during the study period required disease progression following platinum-based therapy.4 These results are similar to other smaller published realworld utilization studies of similar time frames reporting a range of 61% to 85% utilization of 1L
TE D
platinum-containing chemotherapy, depending on EGFR or ALK status.17,18 Utilization rate of 1L chemotherapy is lower when there is a known actionable mutation. More variability in treatment modality was observed in this study in the 2L or 3L setting, with higher use of targeted therapy or immunotherapy in 2L or 3L compared with 1L. In this study, the most frequently utilized
EP
regimens in the 2L setting were pemetrexed (16%), docetaxel (16%), nivolumab (16%), and erlotinib (11%). The same drugs at slightly different rates were most frequently used in the 3L
AC C
setting: nivolumab (17%), docetaxel (14%), erlotinib (14%), and pemetrexed (10%). This is also consistent with the aforementioned real-world study reporting 14% use of docetaxel in the 2L setting, and a diverse range of other 2L therapies with over 90 different regimens.17
It is important to note that the use of targeted therapies in a nontargeted population was an established treatment option during the study time period. Based on the study time frame, the majority of patients who received targeted therapy as 1L did so as a result of known gene mutations. However, the large proportion of patients who received targeted therapy in the 2L may have included those patients who did not have mutation testing results known by the time 9
ACCEPTED MANUSCRIPT
chemotherapy was initiated (e.g., poor tissue sample or symptomatic patients necessitating initial chemotherapy treatment) and therefore received targeted therapy in 2L. In addition, some patients may still have received EGFR tyrosine kinase inhibitors (TKIs) as palliative therapy regardless of EGFR status, if their status was unknown or if the patient had poor performance
RI PT
status. Patients who received 3L EGFR TKIs were likely given this therapy based on factors other than mutation status. With increasing data demonstrating the predictive value of
biomarkers (e.g., EGFR mutational status), changes in the labels for EGFR TKIs have occurred over time. In 2004, erlotinib was first FDA-approved for mNSCLC regardless of EGFR status.9 Now, it is indicated only for patients with known EGFR mutations.5 In 2013, approval was
SC
modified to include the 1L treatment of patients whose tumors have EGFR exon 19 deletions or exon 21 (L858R) substitution mutations.19 Effective October 2016, the labelling change also applied to patients receiving maintenance or 2L or greater treatment, requiring patients to have
M AN U
a known EGFR-sensitizing mutation for use.5,20 In 2015, the pathologic classification of lung cancer mandated immunohistochemistry and molecular analysis in routine clinical practice.4,21
Phase 3 studies such as the Iressa Pan-Asia Study (IPASS) and the European Randomised Trial of Tarceva versus Chemotherapy (EURTAC) demonstrated that patients with EGFR mutations receiving 1L chemotherapy followed by an EGFR TKI, or 1L EGFR TKI followed by
TE D
chemotherapy, had the same outcome in terms of OS, but the benefits in tolerability and PFS were much higher with front-line targeted therapies.12,22 In both IPASS and EURTAC, mOS did not differ significantly between the gefitinib arm and erlotinib arm versus the chemotherapy arms.12,20 In IPASS, mOS with 1L targeted therapy with gefitinib was 18.8 months versus 17.4
EP
months with 1L chemotherapy (P = .109).22 In the EURTAC study, mOS was 19.3 months in the 1L erlotinib group compared with 19.5 months in the standard chemotherapy group (P = .87).12
AC C
A large consortium study assessed the frequency and outcomes associated with various oncogenic drivers in patients with stage IV or recurrent adenocarcinoma of the lung, and at least one gene mutation was found in 64% of patients. Among the actionable mutations with targeted therapies, an EGFR-sensitizing mutation was found in 17% of patients, and ALK rearrangement in 8%. It was demonstrated that patients with oncogenic drivers treated with targeted therapies had improved survival over those who did not receive targeted therapies, hence potentially treated with chemotherapy, with a median survival of 3.5 years versus 2.4 years (hazard ratio 0.69; [95% CI, 0.53-0.9], P = .006).23 In our real-world study, we observed improved OS of 24.3 months in patients who received 1L targeted therapy versus 11.7 months for 1L chemotherapy.
10
ACCEPTED MANUSCRIPT
Our data provide additional evidence that patients with actionable mutations derive greater survival benefit from targeted therapies. In our study, 5% and 1% of the entire population of treated patients were known to be EGFR mutation positive and ALK rearrangement positive, while the status was unknown for more than 60% of patients. One cannot be certain if the
RI PT
patients were never tested, or if the results were not readily available. As growing evidence from studies demonstrate improved outcomes in those who receive treatment for known actionable targets, this further reinforces the importance of testing.
It should be noted that survival estimates in this study reflect those of metastatic patients
SC
diagnosed and treated, since eligibility included all patients who initiated 1L treatment, and did not include patients who did not receive therapy. One study reported that only 24% of NSCLC
M AN U
patients receive 1L chemotherapy, and only 31% of these patients receive 2L therapy.24 So OS estimates may be lower if untreated patients were included in this analysis.
OS estimates from the present study were consistent with the ranges reported in the immunotherapy registrational clinical trials.16,25 Among patients who initiated immunotherapy in the 2L and 3L setting in this study, the mOS was 9.7 and 11.3 months, respectively. In comparison, the Checkmate-057 and Keynote-010 studies that evaluated the use of nivolumab
TE D
or pembrolizumab versus docetaxel for ≥2L treatment of mNSCLC reported significantly improved mOS of 12.2 and 12.7 months with nivolumab and pembrolizumab, versus 9.4 and 8.5 months with docetaxel, respectively.16,25 Third-line EGFR TKIs in this real-world study were likely not given to patients based on mutational status, and in these patients the OS was similar
EP
to chemotherapy and of modest benefit with either therapy.
Comparisons of patient outcomes among treatment groups should be interpreted with caution,
AC C
as it is known that the groups are biologically different given the variable rate of known biomarker status among the population during the study time period. The treatment landscape for use of targeted therapies in NSCLC has evolved over the years, yet there continues to be unmet need in those patients without known targets.
The increases in mOS reported among those patients receiving 1L only, 1L and 2L only, and up to 3L also illustrate the increased likelihood of response to subsequent lines of therapy among patients who respond to 1L as compared with those who do not. These observations are likely
11
ACCEPTED MANUSCRIPT
also explained by differences in the underlying biology of disease in these patients, which is not well understood.
The strength of this study lies in the large, multisite population of mNSCLC patients who were
RI PT
treated with multiple lines of therapy in the community oncology setting. Data collected from the EHR database represent actual treatment information documented as part of routine clinical care, and therefore demonstrate real-world clinical treatment patterns and outcomes, not only those of patients selected as part of clinical trials.
SC
Limitations of this study include the retrospective nature of the evaluation, short follow-up time in more recently selected patients, and potential for documentation bias if there were omissions or errors in the EHR database. Missing data cannot confirm the absence of a condition or value in
M AN U
patients’ medical histories, only that it was not documented. Oral therapies are prescribed through iKM, but the fulfillment of those prescriptions is not observable; therefore, an intent-totreat approach was used for oral therapies. This study was limited to clinics that are part of The US Oncology Network and may not be generalizable to practices outside of the network.
Conclusion
TE D
These data demonstrate timely translation of randomized trial data and integration of novel therapies for mNSCLC in the community setting. Use of targeted therapy was frequently observed in the front-line setting for mutation-positive patients, and survival with immunotherapy in the 2L was similar to that reported in the registrational trials. OS for patients who received multiple lines of therapy was better than in patients who did not. Differences in survival by
AC C
EP
treatment types are likely due to biologic differences.
12
ACCEPTED MANUSCRIPT
Clinical Practice Points •
Identification of biomarkers in mNSCLC such as EGFR mutations and ALK rearrangements helps select patients who will derive maximal benefit from TKIs directed
•
RI PT
against these targets.
In 2015, the first immunotherapy for mNSCLC was FDA-approved for use in patients after progression on platinum-based chemotherapy and after targeted therapy in patients with known mutations. Recent and ongoing studies are evaluating different sequencing options for these treatments.
Effective October 2016, the indication for using EGFR TKIs changed to require patients
SC
•
treated in all lines of therapy for metastatic disease to have a known EGFR-sensitizing
by an FDA-approved test). •
M AN U
mutation (EGFR exon 19 deletions or exon 21 L858R substitution mutations as detected
During the study period, platinum-based chemotherapy was the most frequently used treatment in the 1L setting, followed by targeted therapy or immunotherapy in 2L or 3L.
•
These data demonstrate that patients with actionable mutations can derive greater survival benefit from targeted therapies compared with chemotherapy, and provide
chemotherapy.
Acknowledgements
TE D
evidence that incorporation of immunotherapy can improve OS more than
The authors thank The Lockwood Group (Stamford, CT, USA) for editorial support, which was in
EP
accordance with Good Publication Practice (GPP3) guidelines and funded by AstraZeneca
AC C
(Wilmington, DE, USA).
Funding
This work was supported by AstraZeneca.
References 1. Siegel RL, Miller KD, Jemal A. Cancer statistics, 2017. CA Cancer J Clin. 2017;67(1):730.
13
ACCEPTED MANUSCRIPT
2. National Cancer Institute. Surveillance, Epidemiology, and End Results (SEER) Program. Cancer stat facts: lung and bronchus cancer. https://seer.cancer.gov/statfacts/html/lungb.html. Accessed June 20, 2017. 3. American Cancer Society. What is non-small cell lung cancer?
RI PT
www.cancer.org/cancer/non-small-cell-lung-cancer/about/what-is-non-small-cell-lungcancer.html. Accessed June 20, 2017.
4. National Comprehensive Cancer Network® (NCCN®) Clinical Practice Guidelines in Oncology. Non-small cell lung cancer, v9.2017.
www.nccn.org/professionals/physician_gls/pdf/nscl.pdf. Accessed October 2, 2017.
SC
5. Tarceva [package insert]. Northbrook, IL: OSI Pharmaceuticals, LLC; 2016.
6. Iressa [package insert]. Wilmington, DE: AstraZeneca Pharmaceuticals LP; 2015.
M AN U
7. Gilotrif [package insert]. Ridgefield, CT: Boehringer Ingelheim Pharmaceuticals, Inc.; 2016.
8. Xalkori [package insert]. New York, NY: Pfizer Labs; 2017.
9. Cohen MH, Johnson JR, Chen YF, et al. FDA drug approval summary: erlotinib (Tarceva) tablets. Oncologist. 2005;10(7):461-466.
10. Cohen MH, Williams GA, Sridhara R, et al. FDA drug approval summary: gefitinib (ZD1839) (Iressa) tablets. Oncologist. 2003;8(4):303-306.
TE D
11. Kazandjian D, Blumenthal GM, Chen HY, et al. FDA approval summary: crizotinib for the treatment of metastatic non-small cell lung cancer with anaplastic lymphoma kinase rearrangements. Oncologist. 2014;19(10):e5-e11. 12. Rosell R, Carcereny E, Gervais R, et al. Erlotinib versus standard chemotherapy as first-
EP
line treatment for European patients with advanced EGFR mutation-positive non-smallcell lung cancer (EURTAC): a multicentre, open-label, randomised phase 3 trial. Lancet Oncol. 2012;13(3):239-246.
AC C
13. Solomon BJ, Mok T, Kim DW, et al. PROFILE 1014 Investigators. First-line crizotinib versus chemotherapy in ALK-positive lung cancer. N Engl J Med. 2014;371(23):2167-
2177.
14. Tagrisso [package insert]. Wilmington, DE: AstraZeneca Pharmaceuticals LP; 2017. 15. Alecensa [package insert]. South San Francisco, CA: Genentech USA, Inc.; 2016. 16. Borghaei H, Paz-Ares L, Horn L, et al. Nivolumab versus docetaxel in advanced nonsquamous non-small-cell lung cancer. N Engl J Med. 2015;373(17):1627-1639.
14
ACCEPTED MANUSCRIPT
17. McKay C, Burke T, Cao X, Abernethy AP, Carbone DP. Treatment patterns for advanced non-small-cell lung cancer after platinum-containing therapy in U.S. community oncology clinical practice. Clin Lung Cancer. 2016;17(5):449-460.e7. 18. Abernethy AP, Arunachalam A, Burke T, et al. Real-world first-line treatment and overall
RI PT
survival in non-small cell lung cancer without known EGFR mutations or ALK rearrangements in US community oncology setting. PLoS One. 2017;12(6):e0178420. 19. Khozin S, Blumenthal GM, Jiang X, et al. U.S. Food and Drug Administration approval summary: erlotinib for the first-line treatment of metastatic non-small cell lung cancer
mutations. Oncologist. 2014;19(7):774-779.
SC
with epidermal growth factor receptor exon 19 deletions or exon 21 (L858R) substitution
20. U.S. Food and Drug Administration. FDA-approved drugs: erlotinib (Tarceva).
M AN U
www.fda.gov/Drugs/InformationOnDrugs/ApprovedDrugs/ucm525739.htm. Accessed June 20, 2017.
21. Travis WD, Brambilla E, Nicholson AG, et al. The 2015 World Health Organization classification of lung tumors: impact of genetic, clinical and radiologic advances since the 2004 classification. J Thorac Oncol. 2015;10(9):1243-1260. 22. Fukuoka M, Wu YL, Thongprasert S, et al. Biomarker analyses and final overall survival results from a phase III, randomized, open-label, first-line study of gefitinib versus
TE D
carboplatin/paclitaxel in clinically selected patients with advanced non-small-cell lung cancer in Asia (IPASS). J Clin Oncol. 2011;29(21):2866-2874. 23. Kris MG, Johnson BE, Berry LD, et al. Using multiplexed assays of oncogenic drivers in lung cancers to select targeted drugs. JAMA. 2014;311(19):1998-2006.
EP
24. Sacher AG, Le LW, Lau A, Earle CC, Leighl NB. Real-world chemotherapy treatment patterns in metastatic non-small cell lung cancer: Are patients undertreated? Cancer. 2015;121(15):2562-2569.
AC C
25. Herbst RS, Baas P, Kim DW, et al. Pembrolizumab versus docetaxel for previously treated, PD-L1-positive, advanced non-small-cell lung cancer (KEYNOTE-010): a randomised controlled trial. Lancet. 2016;387(10027):1540-1550.
15
ACCEPTED MANUSCRIPT
Tables Table 1. Baseline Patient and Clinical Characteristics
RI PT
Index Treatment Cohorts Nontargeted Targeted Therapy Chemotherapy (n=685) (n=9969)
Overall (N=10,689)
Immunotherapy (n=35)
68 (27, 90+)
76 (48, 87)
68 (28, 90+)
≤64
3845 (36.0%)
4 (11.4%)
3555 (35.7%)
286 (41.8%)
65-69
2131 (19.9%)
4 (11.4%)
2017 (20.2%)
110 (16.1%)
≥70
4713 (44.1%)
27 (77.1%)
4397 (44.1%)
289 (42.2%)
White
8453 (79.1%)
29 (82.9%)
7943 (79.7%)
481 (70.2%)
African American
1011 (9.5%)
2 (5.7%)
958 (9.6%)
51 (7.4%)
Unknown
1225 (11.5%)
4 (11.4%)
1068 (10.7%)
153 (22.3%)
Race
TE D
Gender
M AN U
Median (min, max) Age group at index
67 (27, 90+)
SC
Age at index date
4873 (45.6%)
18 (51.4%)
4448 (44.6%)
407 (59.4%)
Male
5816 (54.4%)
17 (48.6%)
5521 (55.4%)
278 (40.6%)
6610 (61.8%)
27 (77.1%)
6179 (62.0%)
404 (59.0%)
2554 (23.9%)
5 (14.3%)
2381 (23.9%)
168 (24.5%)
1019 (9.5%)
2 (5.7%)
935 (9.4%)
82 (12.0%)
505 (4.7%)
1 (2.9%)
473 (4.7%)
31 (4.5%)
1 (0.0%)
0 (0.0%)
1 (0.0%)
0 (0.0%)
2264 (21.2%)
14 (40.0%)
2125 (21.3%)
125 (18.2%)
IIIB
765 (7.2%)
1 (2.9%)
740 (7.4%)
24 (3.5%)
IV
6865 (64.2%)
15 (42.9%)
6347 (63.7%)
503 (73.4%)
NA
795 (7.4%)
5 (14.3%)
757 (7.6%)
33 (4.8%)
Practice region South West
AC C
Midwest
EP
Female
Northeast
No information
P value*
.0322
.0017
.3938
<.0001
.1316
Stage at diagnosis I-IIIA
ECOG
16
<.0001
747 (7.0%)
2 (5.7%)
669 (6.7%)
76 (11.1%)
1
6337 (59.3%)
23 (65.7%)
5903 (59.2%)
411 (60.0%)
2
1896 (17.7%)
1 (2.9%)
1804 (18.1%)
91 (13.3%)
3
166 (1.6%)
0 (0.0%)
148 (1.5%)
18 (2.6%)
1543 (14.4%)
9 (25.7%)
1445 (14.5%)
89 (13.0%)
Adenocarcinoma
6392 (59.8%)
15 (42.9%)
5802 (58.2%)
575 (83.9%)
<.0001
Adenosquamous
196 (1.8%)
1 (2.9%)
188 (1.9%)
7 (1.0%)
.1027
Large cell
174 (1.6%)
0 (0.0%)
172 (1.7%)
2 (0.3%)
.0015
Bronchioloalveolar
105 (1.0%)
0 (0.0%)
93 (0.9%)
12 (1.8%)
.0358
Squamous cell
2407 (22.5%)
13 (37.1%)
2355 (23.6%)
39 (5.7%)
<.0001
Neuroendocrine
82 (0.8%)
0 (0.0%)
82 (0.8%)
0 (0.0%)
.0172
Other
277 (2.6%)
0 (0.0%)
268 (2.7%)
9 (1.3%)
.0288
1077 (10.1%)
6 (17.1%)
1029 (10.3%)
42 (6.1%)
.0004
Current
2307 (21.6%)
4 (11.4%)
2257 (22.6%)
46 (6.7%)
Former
6709 (62.8%)
28 (80.0%)
6389 (64.1%)
292 (42.6%)
Never
1315 (12.3%)
3 (8.6%)
985 (9.9%)
327 (47.7%)
232 (2.2%)
0 (0.0%)
219 (2.2%)
13 (1.9%)
0 (0.0%)
119 (1.2%)
7 (1.0%)
NA
RI PT
0
EP
ACCEPTED MANUSCRIPT
TE D
Smoking status
Other tobacco use Unknown
M AN U
Unknown
126 (1.2%)
SC
Histology
<.0001
<.0001
AC C
EGFR mutation status Negative
3454 (32.3%)
13 (37.1%)
3317 (33.3%)
124 (18.1%)
Positive
524 (4.9%)
0 (0.0%)
170 (1.7%)
354 (51.7%)
6694 (62.6%)
22 (62.9%)
6478 (65.0%)
194 (28.3%)
0 (0.0%)
4 (0.0%)
13 (1.9%)
Unknown
<.0001
EGFR T790M mutation status T790M
17 (0.2%)
ALK rearrangement status
17
NR
ACCEPTED MANUSCRIPT
Negative
3536 (33.1%)
12 (34.3%)
3219 (32.3%)
305 (44.5%)
Positive
119 (1.1%)
0 (0.0%)
33 (0.3%)
86 (12.6%)
Equivocal
72 (<0.7%)
1 (2.9%)
64 (0.6%)
7 (1.0%)
NR
Unknown
6962 (65.1%)
22 (62.8%)
6653 (66.7%)
287 (41.9%)
NR
Negative
74 (0.7%)
0 (0.0%)
66 (0.7%)
8 (1.2%)
Positive
51 (0.5%)
2 (5.7%)
46 (0.5%)
3 (0.4%)
10,564 (98.8%)
33 (94.3%)
9857 (98.9%)
674 (98.4%)
Unknown
.5234
SC
PD-L1 status
RI PT
<.0001
M AN U
*P values were comparisons between the targeted and chemo groups. Wilcoxon rank-sum tests for continuous variables and chi-square tests were used for categorical variables when all cell counts were ≥5. The Fisher test for categorical variables was used when there were cell counts <5. Abbreviations: ECOG = Eastern Cooperative Oncology Group; EGFR = epidermal growth factor receptor;
AC C
EP
TE D
ALK = anaplastic lymphoma kinase.
18
ACCEPTED MANUSCRIPT
Table 2. Treatment Patterns
Index Treatment Cohorts Overall
Nontargeted Chemotherapy (n=9969)
Targeted Therapy (n=685)
Immunotherapy (n=35)
Carboplatin + paclitaxel
3914 (36.6%)
3914 (39.3%)
0 (0.0%)
0 (0.0%)
Carboplatin + pemetrexed
2231 (20.9%)
2231 (22.4%)
0 (0.0%)
0 (0.0%)
Bevacizumab + carboplatin + paclitaxel
1395 (13.1%)
1395 (14.0%)
0 (0.0%)
0 (0.0%)
392 (3.7%)
Crizotinib
82 (0.8%)
Nivolumab
32 (0.3%)
Other*
M AN U
Erlotinib
2643 (24.7%)
Chemotherapy Immunotherapy
0 (0.0%)
0 (0.0%)
82 (12.0%)
0 (0.0%)
0 (0.0%)
0 (0.0%)
32 (91.4%)
2429 (24.4%)
211 (30.8%)
3 (8.6%)
2695 (27.0%)
89 (13.0%)
6 (17.1%)
727 (6.8%)
715 (7.2%)
11 (1.6%)
1 (2.9%)
718 (6.7%)
491 (4.9%)
227 (33.1%)
0 (0.0%)
6068 (60.9%)
358 (52.3%)
28 (80.0%)
6454 (60.4%)
AC C
None
392 (57.2%)
2790 (26.1%)
EP
Targeted therapy
0 (0.0%)
TE D
LOT2 treatment group
SC
LOT1 regimens
RI PT
(N=10,689)
Overall (N=4235)
Index Treatment Cohorts Nontargeted Chemotherapy (n=2790)
Targeted Immunotherapy Chemotherapy (n=727) (n=718)
LOT2 regimens Pemetrexed
696 (16.4%)
696 (16.4%)
0 (0.0%)
0 (0.0%)
Docetaxel
694 (16.4%)
694 (16.4%)
0 (0.0%)
0 (0.0%)
Nivolumab
674 (15.9%)
0 (0.0%)
0 (0.0%)
674 (92.7%)
19
ACCEPTED MANUSCRIPT
Erlotinib
458 (10.8%)
0 (0.0%)
458 (10.8%)
0 (0.0%)
Other*
1713 (40.4%)
1400 (50.2%)
260 (36.2%)
53 (7.3.3%)
Chemo
921 (21.7%)
697 (25.0%)
127 (17.7%)
97 (13.3%)
Targeted
355 (8.4%)
256 (9.2%)
84 (11.7%)
15 (2.1%)
Immunotherapy
304 (7.2%)
263 (9.4%)
33 (4.6%)
8 (1.1%)
2655 (62.7%)
1574 (56.4%)
474 (66.0%)
607 (83.5%)
Nivolumab
270 (17.1%)
245 (20.1%)
20 (8.2%)
5 (4.2%)
Docetaxel
227 (14.4%)
160 (13.2%)
31 (12.7%)
36 (30.0%)
Erlotinib
224 (14.2%)
208 (17.1%)
5 (2.0%)
11 (9.2%)
Pemetrexed
158 (10.0%)
126 (10.4%)
24 (9.8%)
8 (6.7%)
Gemcitabine
128 (8.1%)
104 (8.6%)
7 (2.9%)
17 (14.2%)
157 (64.3%)
43 (35.8%)
TE D
Other*
M AN U
LOT3 regimens
SC
None
RI PT
LOT3 treatment group
573 (36.3%)
373 (30.7%)
*“Other” treatments individually account for less than 10% of each treatment cohort.
AC C
EP
Abbreviation: LOT = line of therapy.
20
ACCEPTED MANUSCRIPT
Figure Captions
Figure 1. Consort Diagram. A total of 10,689 patients met eligibility criteria and initiated 1L during the study period.
RI PT
therapy with either nontargeted chemotherapy, targeted therapy, or other (immunotherapy)
patients from the initiation of first line of treatment.
SC
Figure 2. Overall Survival by First-Line. Kaplan-Meier estimated overall survival for
Figure 3. Overall Survival by Second-Line. Kaplan-Meier estimated overall survival for
M AN U
patients from the initiation of second line of treatment.
Figure 4. Overall Survival by Third-Line. Kaplan-Meier estimated overall survival for
AC C
EP
TE D
patients from the initiation of third line of treatment.
21
ACCEPTED MANUSCRIPT
Figures
AC C
EP
TE D
M AN U
SC
RI PT
Figure 1. Consort Diagram
Abbreviations: NSCLC = non-small cell lung cancer; USON = US Oncology Network.
22
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
Figure 2. Overall Survival by First-line
Abbreviations: CI = confidence interval; LOT = line of therapy; OS = overall survival.
23
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
Figure 3. Overall Survival by Second-line
Abbreviations: CI = confidence interval; OS = overall survival. 24
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
Figure 4. Overall Survival by Third-Line
Abbreviations: CI = confidence interval; OS = overall survival. 25
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
Figure 5. Overall Survival by LOTs received
Abbreviations: CI = confidence interval; LOT = line of therapy; OS = overall survival.
26
ACCEPTED MANUSCRIPT
Table 1. Baseline Patient and Clinical Characteristics
Index Treatment Cohorts Nontargeted Targeted Therapy Chemotherapy (n=685) (n=9969)
Immunotherapy (n=35)
68 (27, 90+)
76 (48, 87)
68 (28, 90+)
67 (27, 90+)
≤64
3845 (36.0%)
4 (11.4%)
3555 (35.7%)
286 (41.8%)
65-69
2131 (19.9%)
4 (11.4%)
2017 (20.2%)
110 (16.1%)
≥70
4713 (44.1%)
27 (77.1%)
4397 (44.1%)
289 (42.2%)
White
8453 (79.1%)
29 (82.9%)
7943 (79.7%)
481 (70.2%)
African American
1011 (9.5%)
2 (5.7%)
958 (9.6%)
51 (7.4%)
Unknown
1225 (11.5%)
4 (11.4%)
1068 (10.7%)
153 (22.3%)
Female
4873 (45.6%)
18 (51.4%)
4448 (44.6%)
407 (59.4%)
Male
5816 (54.4%)
17 (48.6%)
5521 (55.4%)
278 (40.6%)
South
6610 (61.8%)
27 (77.1%)
6179 (62.0%)
404 (59.0%)
West
2554 (23.9%)
5 (14.3%)
2381 (23.9%)
168 (24.5%)
1019 (9.5%)
2 (5.7%)
935 (9.4%)
82 (12.0%)
505 (4.7%)
1 (2.9%)
473 (4.7%)
31 (4.5%)
1 (0.0%)
0 (0.0%)
1 (0.0%)
0 (0.0%)
2264 (21.2%)
14 (40.0%)
2125 (21.3%)
125 (18.2%)
765 (7.2%)
1 (2.9%)
740 (7.4%)
24 (3.5%)
IV
6865 (64.2%)
15 (42.9%)
6347 (63.7%)
503 (73.4%)
NA
795 (7.4%)
5 (14.3%)
757 (7.6%)
33 (4.8%)
0
747 (7.0%)
2 (5.7%)
669 (6.7%)
76 (11.1%)
1
6337 (59.3%)
23 (65.7%)
5903 (59.2%)
411 (60.0%)
RI PT
Overall (N=10,689)
Race
Midwest Northeast
AC C
No information
EP
Practice region
TE D
Gender
M AN U
Median (min, max) Age group at index
SC
Age at index date
P value*
.0322
.0017
.3938
<.0001
.1316
Stage at diagnosis I-IIIA IIIB
<.0001
ECOG <.0001
2
1896 (17.7%)
1 (2.9%)
1804 (18.1%)
91 (13.3%)
3
166 (1.6%)
0 (0.0%)
148 (1.5%)
18 (2.6%)
1543 (14.4%)
9 (25.7%)
1445 (14.5%)
89 (13.0%)
Adenocarcinoma
6392 (59.8%)
15 (42.9%)
5802 (58.2%)
575 (83.9%)
Adenosquamous
196 (1.8%)
1 (2.9%)
188 (1.9%)
7 (1.0%)
.1027
Large cell
174 (1.6%)
0 (0.0%)
172 (1.7%)
2 (0.3%)
.0015
Bronchioloalveolar
105 (1.0%)
0 (0.0%)
93 (0.9%)
12 (1.8%)
.0358
Squamous cell
2407 (22.5%)
13 (37.1%)
2355 (23.6%)
39 (5.7%)
<.0001
Neuroendocrine
82 (0.8%)
0 (0.0%)
82 (0.8%)
0 (0.0%)
.0172
Other
277 (2.6%)
0 (0.0%)
268 (2.7%)
9 (1.3%)
.0288
1077 (10.1%)
6 (17.1%)
1029 (10.3%)
42 (6.1%)
.0004
Current
2307 (21.6%)
4 (11.4%)
2257 (22.6%)
46 (6.7%)
Former
6709 (62.8%)
28 (80.0%)
6389 (64.1%)
292 (42.6%)
Never
1315 (12.3%)
3 (8.6%)
985 (9.9%)
327 (47.7%)
Other tobacco use
232 (2.2%)
0 (0.0%)
219 (2.2%)
13 (1.9%)
Unknown
126 (1.2%)
0 (0.0%)
119 (1.2%)
7 (1.0%)
EP
ACCEPTED MANUSCRIPT
13 (37.1%)
3317 (33.3%)
124 (18.1%)
524 (4.9%)
0 (0.0%)
170 (1.7%)
354 (51.7%)
6694 (62.6%)
22 (62.9%)
6478 (65.0%)
194 (28.3%)
0 (0.0%)
4 (0.0%)
13 (1.9%)
NR
12 (34.3%)
3219 (32.3%)
305 (44.5%)
<.0001
NA
TE D
Smoking status
SC
M AN U
Unknown
RI PT
Histology <.0001
<.0001
EGFR mutation status
3454 (32.3%)
AC C
Negative Positive
Unknown
<.0001
EGFR T790M mutation status T790M
17 (0.2%)
ALK rearrangement status Negative
3536 (33.1%)
2
ACCEPTED MANUSCRIPT
Positive
119 (1.1%)
0 (0.0%)
33 (0.3%)
86 (12.6%)
Equivocal
72 (<0.7%)
1 (2.9%)
64 (0.6%)
7 (1.0%)
NR
Unknown
6962 (65.1%)
22 (62.8%)
6653 (66.7%)
287 (41.9%)
NR
Negative
74 (0.7%)
0 (0.0%)
66 (0.7%)
8 (1.2%)
Positive
51 (0.5%)
2 (5.7%)
46 (0.5%)
3 (0.4%)
10,564 (98.8%)
33 (94.3%)
9857 (98.9%)
674 (98.4%)
.5234
SC
Unknown
RI PT
PD-L1 status
*P values were comparisons between the targeted and chemo groups. Wilcoxon rank-sum tests for continuous variables and chi-square tests were used for categorical variables when all cell counts were
M AN U
≥5. The Fisher test for categorical variables was used when there were cell counts <5. Abbreviations: ECOG = Eastern Cooperative Oncology Group; EGFR = epidermal growth factor receptor;
AC C
EP
TE D
ALK = anaplastic lymphoma kinase.
3
ACCEPTED MANUSCRIPT
Table 2. Treatment Patterns
Index Treatment Cohorts Overall
Nontargeted Chemotherapy (n=9969)
Targeted Therapy (n=685)
Immunotherapy (n=35)
Carboplatin + paclitaxel
3914 (36.6%)
3914 (39.3%)
0 (0.0%)
0 (0.0%)
Carboplatin + pemetrexed
2231 (20.9%)
2231 (22.4%)
0 (0.0%)
0 (0.0%)
Bevacizumab + carboplatin + paclitaxel
1395 (13.1%)
1395 (14.0%)
0 (0.0%)
0 (0.0%)
392 (3.7%)
Crizotinib
82 (0.8%)
Nivolumab
32 (0.3%)
Other*
M AN U
Erlotinib
2643 (24.7%)
Chemotherapy Immunotherapy
0 (0.0%)
0 (0.0%)
82 (12.0%)
0 (0.0%)
0 (0.0%)
0 (0.0%)
32 (91.4%)
2429 (24.4%)
211 (30.8%)
3 (8.6%)
2695 (27.0%)
89 (13.0%)
6 (17.1%)
727 (6.8%)
715 (7.2%)
11 (1.6%)
1 (2.9%)
718 (6.7%)
491 (4.9%)
227 (33.1%)
0 (0.0%)
6068 (60.9%)
358 (52.3%)
28 (80.0%)
6454 (60.4%)
AC C
None
392 (57.2%)
2790 (26.1%)
EP
Targeted therapy
0 (0.0%)
TE D
LOT2 treatment group
SC
LOT1 regimens
RI PT
(N=10,689)
Overall (N=4235)
Index Treatment Cohorts Nontargeted Chemotherapy (n=2790)
Targeted Immunotherapy Chemotherapy (n=727) (n=718)
LOT2 regimens Pemetrexed
696 (16.4%)
696 (16.4%)
0 (0.0%)
0 (0.0%)
Docetaxel
694 (16.4%)
694 (16.4%)
0 (0.0%)
0 (0.0%)
Nivolumab
674 (15.9%)
0 (0.0%)
0 (0.0%)
674 (92.7%)
ACCEPTED MANUSCRIPT
Erlotinib
458 (10.8%)
0 (0.0%)
458 (10.8%)
0 (0.0%)
Other*
1713 (40.4%)
1400 (50.2%)
260 (36.2%)
53 (7.3.3%)
Chemo
921 (21.7%)
697 (25.0%)
127 (17.7%)
97 (13.3%)
Targeted
355 (8.4%)
256 (9.2%)
84 (11.7%)
15 (2.1%)
Immunotherapy
304 (7.2%)
263 (9.4%)
33 (4.6%)
8 (1.1%)
2655 (62.7%)
1574 (56.4%)
474 (66.0%)
607 (83.5%)
Nivolumab
270 (17.1%)
245 (20.1%)
20 (8.2%)
5 (4.2%)
Docetaxel
227 (14.4%)
160 (13.2%)
31 (12.7%)
36 (30.0%)
Erlotinib
224 (14.2%)
208 (17.1%)
5 (2.0%)
11 (9.2%)
Pemetrexed
158 (10.0%)
126 (10.4%)
24 (9.8%)
8 (6.7%)
Gemcitabine
128 (8.1%)
104 (8.6%)
7 (2.9%)
17 (14.2%)
157 (64.3%)
43 (35.8%)
TE D
Other*
M AN U
LOT3 regimens
SC
None
RI PT
LOT3 treatment group
573 (36.3%)
373 (30.7%)
*“Other” treatments individually account for less than 10% of each treatment cohort.
AC C
EP
Abbreviation: LOT = line of therapy.
2
ACCEPTED MANUSCRIPT
Figure Captions
Figure 1. Consort Diagram. A total of 10,689 patients met eligibility criteria and initiated 1L during the study period.
RI PT
therapy with either nontargeted chemotherapy, targeted therapy, or other (immunotherapy)
patients from the initiation of first line of treatment.
SC
Figure 2. Overall Survival by First-Line. Kaplan-Meier estimated overall survival for
Figure 3. Overall Survival by Second-Line. Kaplan-Meier estimated overall survival for
M AN U
patients from the initiation of second line of treatment.
Figure 4. Overall Survival by Third-Line. Kaplan-Meier estimated overall survival for
AC C
EP
TE D
patients from the initiation of third line of treatment.
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
S1525-7304(18)30021-4
DOI:
10.1016/j.cllc.2018.02.002
Reference:
CLLC 745
To appear in:
Clinical Lung Cancer
Received Date: 30 October 2017 Revised Date:
1 February 2018
Accepted Date: 12 February 2018
Please cite this article as: Nadler E, Espirito JL, Pavilack M, Boyd M, Vergara-Silva A, Fernandes A, Treatment Patterns and Clinical Outcomes Among Metastatic Non-Small Cell Lung Cancer Patients Treated in the Community Practice Setting, Clinical Lung Cancer (2018), doi: 10.1016/j.cllc.2018.02.002. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
Treatment Patterns and Clinical Outcomes Among Metastatic NonSmall Cell Lung Cancer Patients Treated in the Community Practice
RI PT
Setting Eric Nadler,1 Janet L. Espirito,2 Melissa Pavilack,3 Marley Boyd,2 Andrea Vergara-Silva,3 Ancilla Fernandes3 1
Texas Oncology, 3410 Worth St #400, Dallas, TX 75246, USA
2
SC
McKesson Specialty Health, 10101 Woodloch Forest, The Woodlands, TX, USA
3
M AN U
AstraZeneca, 950 Wind River Lane, Gaithersburg, MD 20878, USA
Address for correspondence: Eric Nadler, MD Texas Oncology
3410 Worth St. HR081 Dallas, TX 75246
Phone: +1 (214) 370-1000
TE D
Email: [email protected]
Target Journal: Clinical Lung Cancer
Word Count (Body only): 3186 of 8000 Max
AC C
EP
Figures/Tables: 7 of 7 max
1
ACCEPTED MANUSCRIPT
Conflict of Interest/Disclosure Dr Nadler has received grants from AstraZeneca; financial compensation for speaker bureau activities from Genentech, Merck, and Eli Lilly; and financial compensation for consulting roles from Eli Lilly. Dr Espirito is an employee of and owns stocks in McKesson Specialty Health. Dr
RI PT
Pavilack is an employee of AstraZeneca. Dr Boyd has received grants from AstraZeneca and is an employee of McKesson Specialty Health. Drs Vergara-Silva and Fernandes are employees of and own stocks in AstraZeneca. Dr Vergara-Silva has received nonfinancial support (travel
AC C
EP
TE D
M AN U
SC
expenses) from AstraZeneca.
2
ACCEPTED MANUSCRIPT
Micro Abstract (word limit 60 max; current = 57) New targeted and immunotherapies expand the treatment options for mNSCLC. This study assessed contemporary treatment patterns and outcomes using a large EHR database of >10,000 patients. The majority of targeted and immunotherapy use was in the 2L and 3L
AC C
EP
TE D
M AN U
SC
actionable mutations have a survival advantage.
RI PT
treatment setting. Survival differences observed by treatment types suggest that patients with
3
ACCEPTED MANUSCRIPT
Abstract (word limit 250 max; current = 247)
RI PT
Introduction: Multiple therapeutic options now exist for metastatic non-small cell lung cancer (mNSCLC). This study evaluated treatment patterns and outcomes in mNSCLC patients receiving first-line (1L), second-line (2L), and third-line (3L) therapy.
Patients and Methods: A retrospective, observational cohort study was conducted using an
SC
electronic health record database of mNSCLC patients who received initial treatment from
January 2012 through April 2016, with follow-up through June 2016. Patient characteristics and treatment patterns were characterized. Overall survival (OS) was assessed using the Kaplan-
M AN U
Meier method.
Results: 10,689 1L patients were identified. Median age was 68 years, and 54% were male. Most patients (59%) had a performance status of 1, and 77% had nonsquamous histology. 1L treatment was chemotherapy in 93% of patients, and targeted therapy in 6%. Median OS (mOS) for all patients in 1L was 12.3 mo (95% confidence interval [CI], 11.9-12.7), and 24.3 mo in 1L
TE D
patients receiving targeted therapy. Among patients who received 2L therapy (n=4235), 66%, 17%, and 17% received chemotherapy, targeted therapy, and immunotherapy, respectively. mOS from 2L therapy was 9.6 mo (95% CI, 9.1-10.1). In patients receiving 3L therapy (n=1580), 58%, 22%, and 19% received chemotherapy, targeted therapy, and immunotherapy,
EP
respectively. mOS from 3L therapy was 8.2 mo (95% CI, 7.3-8.7).
Conclusion: Targeted therapy and immunotherapy was most frequently used in the 2L and 3L
AC C
setting during the study time frame. Survival differences observed by treatment types are likely due to biologic differences, and suggest that patients with actionable mutations have a survival advantage.
Keywords (5 max): NSCLC, Targeted therapy, Immunotherapy, Chemotherapy, Real-world experience
4
ACCEPTED MANUSCRIPT
Introduction Lung cancer is the leading cause of cancer-related deaths in both men and women in the United States. In 2017, approximately 222,500 new cases and 155,870 deaths due to lung cancer are estimated.1 The majority of patients are diagnosed with metastatic non-
RI PT
small cell lung cancer (mNSCLC), and treatment is largely palliative. The estimated
5-year survival for patients with mNSCLC from 2007 to 2013 was 4.5%.2 The majority (70% to 75%) of NSCLC subtypes are grouped into the nonsquamous histologic
category, with the remaining grouped into the squamous category, which accounts for
SC
approximately 25% to 30%.3
Platinum-based combination chemotherapy has been the standard of care for treatment of mNSCLC for several years, and is associated with improved survival, symptom control, and
M AN U
improved quality of life versus best supportive care.4 However, treatment of mNSCLC has evolved over the past decade with the development of new therapeutic options for metastatic disease.5-8 Several biomarkers such as epidermal growth factor receptor (EGFR), anaplastic lymphoma kinase (ALK), ROS1 receptor tyrosine kinase (ROS1), and programmed death cell ligand-1 (PD-L1) have emerged as predictive for therapeutic efficacy.4 Therapies targeting these molecules were initially approved by the U.S. Food and Drug Administration (FDA) after disease
TE D
progression on at least one prior chemotherapy regimen.9-11 Subsequent studies have demonstrated improved progression-free survival (PFS) in patients with identified mutations who receive targeted therapy compared with standard chemotherapy, and EGFR and ALK inhibitors are now recommended as first-line (1L) therapy in patients with known mutations.4,12,13
EP
However, a majority of patients do not have targetable gene abnormalities, and even among tumors that harbor these mutations, resistance to therapy eventually develops.4 Novel targeted therapies are also now available for patients with known resistance mutations emerging after
AC C
EGFR TKI therapy14 or for patients with ALK-positive disease and progression after crizotinib.15 In 2015, the first immunotherapy for mNSCLC was FDA-approved for use after progression on platinum-based chemotherapy and in patients with EGFR mutations or ALK rearrangements after progression on targeted therapy.16 Multiple therapeutic options now exist for mNSCLC, including chemotherapy, targeted therapy, and immunotherapy, creating questions about how best to integrate these treatments into the course of care.4 To date, no studies have evaluated the best sequence of therapies and, therefore, treatment selection often depends on tumor molecular profile, performance status, and toxicities.4 This study aims to understand how these different treatment options currently 5
ACCEPTED MANUSCRIPT
available are sequenced during the course of therapy in this patient population. The objectives were (1) to describe the demographic and clinical characteristics of patients with mNSCLC initiating treatment with chemotherapy, targeted therapy, or immunotherapy; (2) to evaluate realworld systemic treatment patterns among patients with metastatic disease by line of therapy
RI PT
(LOT); and (3) to evaluate overall survival (OS) with systemic treatment for patients with mNSCLC by LOT and treatment category.
Materials and Methods
This was a retrospective observational cohort study of patients with mNSCLC. Data were
SC
obtained via programmatic queries of The US Oncology Network iKnowMed (iKM) electronic health record (EHR) system to collect the structured, documented information available from the records. McKesson Specialty Health maintains iKM, an integrated web-based database and
M AN U
oncology-specific EHR system that captures outpatient practice encounter histories from network community oncology practices affiliated with over 1000 physicians in more than 25 practices across 400 sites of care in 19 states. Vital status data were supplemented with information obtained from the U.S. Department of Social Security Death Index. Institutional review board approval was obtained.
TE D
The study population included patients with documented mNSCLC, defined as either stage IV at diagnosis or patients diagnosed with earlier stages who later developed evidence of metastatic disease. Patients were required to receive their initial treatment for metastatic disease between January 1, 2012, and April 30, 2016, with follow-up through June 30, 2016, or until date of last record, whichever occurred first, for a minimum potential follow-up of 2 months. The study time
in 2015.
EP
period was selected to capture maximum data on immunotherapy utilization following approval
AC C
Patients were required to have ≥2 visits within The US Oncology Network during the study period. Patients were excluded if they were aged <18 years at diagnosis, had another documented cancer prior to their mNSCLC, or, to avoid confounding effects of investigational therapies, if they were enrolled in a clinical trial before their initial treatment for metastatic disease.
All patients who met eligibility criteria for the treatment of mNSCLC were included in the study. Assignment of patients into treatment cohorts was based on their 1L mNSCLC regimen and followed for up to 3 lines of therapy. Systemic treatments and LOTs, including maintenance therapy (after 1L) during the study period, were programmatically sequenced based on 6
ACCEPTED MANUSCRIPT
treatment dates and categorized as chemotherapy, targeted therapy, or immunotherapy. Regimens received by <10% of patients were categorized as “other.”
Statistical Analysis
RI PT
Descriptive analyses were conducted to assess demographic, clinical, and treatment characteristics among the cohorts. Wilcoxon rank-sum tests were performed for continuous variables, and chi-square tests were used for categorical variables when all cell counts were ≥5. The Fisher test for categorical variables was used when cell counts were <5. Missing data were identified and reported as percentages. The Kaplan-Meier method was used to estimate OS,
SC
stratified by LOT and treatment category. Statistical analyses were performed using SAS 9.4
M AN U
(SAS Institute, Cary, NC).
Results
Demographics and Baseline Characteristics
A total of 10,689 patients who met eligibility criteria and received 1L therapy during the study period were included in the analysis. The 1L treatment cohorts consisted predominantly of
TE D
patients who initiated 1L treatment with chemotherapy (93%), 6% of patients with targeted therapy, and <1% of patients with immunotherapy (Figure 1). The median age at the start of therapy was 68 years (range 27 to ≥90); 79% were white and 54% were male. The majority of patients (64%) were initially diagnosed with stage IV disease. Most patients (59%) had an
EP
Eastern Cooperative Oncology Group (ECOG) performance status of 1, 77% had nonsquamous histology, and 85% were former or current smokers. Among patients receiving initial targeted therapy, most were never smokers (48%) versus those who received initial chemotherapy and
AC C
immunotherapy (<10% for both; P < .0001). Tumor markers were assessed through documented structured data in the EHR. EGFR and ALK status was available in approximately 40% and 35% of patients, respectively. Among those with documented results, EGFR status was positive in 13%; ALK was positive in 3%; PD-L1 status was known in <1%; and ROS1 was not collected in this study (Table 1).
Treatment Patterns The most frequently used chemotherapies in the 1L setting were carboplatin + paclitaxel (37%), carboplatin + pemetrexed (21%), and carboplatin + paclitaxel + bevacizumab (13%). Among 7
ACCEPTED MANUSCRIPT
patients receiving targeted therapy in the 1L, erlotinib (57%) and crizotinib (12%) were most often used (Table 2).
Sixty percent of the 1L patients did not receive second-line (2L) therapy within the study period.
RI PT
Among the 40% of 1L patients (n = 4235) who did receive 2L therapy, 66%, 17%, and 17% received chemotherapy, targeted therapy, and immunotherapy, respectively. Pemetrexed or docetaxel monotherapy was most commonly used for 2L chemotherapy, erlotinib was most used as 2L targeted therapy, and nivolumab was most used for 2L immunotherapy. Among patients who received 1L chemotherapy and continued to 2L, the majority continued with 2L
SC
chemotherapy (27%). Among patients who received 1L targeted therapy and continued to 2L, the majority continued with 2L targeted therapy (33%). Among patients who received
bevacizumab (15%; data not shown).
M AN U
maintenance therapy after 1L, pemetrexed (48%) was most frequently used, followed by
For patients who received 2L chemotherapy (n = 2790) and continued to receive third-line (3L) treatment (n = 1216), the majority continued with 3L chemotherapy (57%). For patients who received 2L targeted therapy (n = 718) and went on to receive 3L treatment (n = 244), more than half (52%) switched to chemotherapy in the 3L, and 34% continued with targeted therapy
TE D
in the 3L. Among all patients who received 3L therapy during the study period (n = 1580), immunotherapy treatment with nivolumab was most frequently used (17%).
EP
Clinical Outcomes
The median follow-up time from 1L treatment initiation was 6.9 months (range 0 to 54 months) for all patients in the study. Kaplan-Meier survival curves for patients from the initiation of each
AC C
line of treatment are shown in Figure 2, Figure 3, and Figure 4. Median OS (mOS) for all patients in 1L was 12.3 months (95% confidence interval [CI], 11.9-12.7). Patients who received targeted therapy in 1L were observed to have the longest survival of 24.3 months. In all patients who received 2L therapy, mOS from start of 2L was 9.6 months (95% CI, 9.1-10.1). mOS was similarly observed to be longest for patients who received targeted therapy in 2L, at 11.2 months. In all patients who received 3L therapy, the mOS from the start of 3L was 8.2 months (95% CI, 7.3-8.7). The mOS appeared longest, at 11.3 months, in patients who received 3L immunotherapy.
8
ACCEPTED MANUSCRIPT
Among the subset of patients receiving only 1L therapy of any type (n = 6396), mOS was 7.2 mo (95% CI, 6.9-7.5). The mOS among patients who received only 1L and 2L therapy of any type (n = 2623) was 14.7 mo (95% CI, 13.8-15.4). The mOS among patients who received 1L,
RI PT
2L, and 3L therapy of any type (n = 1568) was 23.0 mo (95% CI, 21.8-24.8; Figure 5).
Discussion
A large EHR database population was used to examine real-world patient characteristics,
SC
treatment patterns, and clinical outcomes involving the use of chemotherapy, newer targeted therapies, and immunotherapies for mNSCLC. Use of EHR database information provided insight into initial treatment utilization in real-world clinical practice, including at a time when
M AN U
biomarker testing was not universal and the impact of mutational status was not as well known.
In this study, the majority of the patient population received initial platinum-based chemotherapy in the 1L setting (71%), as this has been the standard of care for several years and many of the new drugs that gained FDA approval during the study period required disease progression following platinum-based therapy.4 These results are similar to other smaller published realworld utilization studies of similar time frames reporting a range of 61% to 85% utilization of 1L
TE D
platinum-containing chemotherapy, depending on EGFR or ALK status.17,18 Utilization rate of 1L chemotherapy is lower when there is a known actionable mutation. More variability in treatment modality was observed in this study in the 2L or 3L setting, with higher use of targeted therapy or immunotherapy in 2L or 3L compared with 1L. In this study, the most frequently utilized
EP
regimens in the 2L setting were pemetrexed (16%), docetaxel (16%), nivolumab (16%), and erlotinib (11%). The same drugs at slightly different rates were most frequently used in the 3L
AC C
setting: nivolumab (17%), docetaxel (14%), erlotinib (14%), and pemetrexed (10%). This is also consistent with the aforementioned real-world study reporting 14% use of docetaxel in the 2L setting, and a diverse range of other 2L therapies with over 90 different regimens.17
It is important to note that the use of targeted therapies in a nontargeted population was an established treatment option during the study time period. Based on the study time frame, the majority of patients who received targeted therapy as 1L did so as a result of known gene mutations. However, the large proportion of patients who received targeted therapy in the 2L may have included those patients who did not have mutation testing results known by the time 9
ACCEPTED MANUSCRIPT
chemotherapy was initiated (e.g., poor tissue sample or symptomatic patients necessitating initial chemotherapy treatment) and therefore received targeted therapy in 2L. In addition, some patients may still have received EGFR tyrosine kinase inhibitors (TKIs) as palliative therapy regardless of EGFR status, if their status was unknown or if the patient had poor performance
RI PT
status. Patients who received 3L EGFR TKIs were likely given this therapy based on factors other than mutation status. With increasing data demonstrating the predictive value of
biomarkers (e.g., EGFR mutational status), changes in the labels for EGFR TKIs have occurred over time. In 2004, erlotinib was first FDA-approved for mNSCLC regardless of EGFR status.9 Now, it is indicated only for patients with known EGFR mutations.5 In 2013, approval was
SC
modified to include the 1L treatment of patients whose tumors have EGFR exon 19 deletions or exon 21 (L858R) substitution mutations.19 Effective October 2016, the labelling change also applied to patients receiving maintenance or 2L or greater treatment, requiring patients to have
M AN U
a known EGFR-sensitizing mutation for use.5,20 In 2015, the pathologic classification of lung cancer mandated immunohistochemistry and molecular analysis in routine clinical practice.4,21
Phase 3 studies such as the Iressa Pan-Asia Study (IPASS) and the European Randomised Trial of Tarceva versus Chemotherapy (EURTAC) demonstrated that patients with EGFR mutations receiving 1L chemotherapy followed by an EGFR TKI, or 1L EGFR TKI followed by
TE D
chemotherapy, had the same outcome in terms of OS, but the benefits in tolerability and PFS were much higher with front-line targeted therapies.12,22 In both IPASS and EURTAC, mOS did not differ significantly between the gefitinib arm and erlotinib arm versus the chemotherapy arms.12,20 In IPASS, mOS with 1L targeted therapy with gefitinib was 18.8 months versus 17.4
EP
months with 1L chemotherapy (P = .109).22 In the EURTAC study, mOS was 19.3 months in the 1L erlotinib group compared with 19.5 months in the standard chemotherapy group (P = .87).12
AC C
A large consortium study assessed the frequency and outcomes associated with various oncogenic drivers in patients with stage IV or recurrent adenocarcinoma of the lung, and at least one gene mutation was found in 64% of patients. Among the actionable mutations with targeted therapies, an EGFR-sensitizing mutation was found in 17% of patients, and ALK rearrangement in 8%. It was demonstrated that patients with oncogenic drivers treated with targeted therapies had improved survival over those who did not receive targeted therapies, hence potentially treated with chemotherapy, with a median survival of 3.5 years versus 2.4 years (hazard ratio 0.69; [95% CI, 0.53-0.9], P = .006).23 In our real-world study, we observed improved OS of 24.3 months in patients who received 1L targeted therapy versus 11.7 months for 1L chemotherapy.
10
ACCEPTED MANUSCRIPT
Our data provide additional evidence that patients with actionable mutations derive greater survival benefit from targeted therapies. In our study, 5% and 1% of the entire population of treated patients were known to be EGFR mutation positive and ALK rearrangement positive, while the status was unknown for more than 60% of patients. One cannot be certain if the
RI PT
patients were never tested, or if the results were not readily available. As growing evidence from studies demonstrate improved outcomes in those who receive treatment for known actionable targets, this further reinforces the importance of testing.
It should be noted that survival estimates in this study reflect those of metastatic patients
SC
diagnosed and treated, since eligibility included all patients who initiated 1L treatment, and did not include patients who did not receive therapy. One study reported that only 24% of NSCLC
M AN U
patients receive 1L chemotherapy, and only 31% of these patients receive 2L therapy.24 So OS estimates may be lower if untreated patients were included in this analysis.
OS estimates from the present study were consistent with the ranges reported in the immunotherapy registrational clinical trials.16,25 Among patients who initiated immunotherapy in the 2L and 3L setting in this study, the mOS was 9.7 and 11.3 months, respectively. In comparison, the Checkmate-057 and Keynote-010 studies that evaluated the use of nivolumab
TE D
or pembrolizumab versus docetaxel for ≥2L treatment of mNSCLC reported significantly improved mOS of 12.2 and 12.7 months with nivolumab and pembrolizumab, versus 9.4 and 8.5 months with docetaxel, respectively.16,25 Third-line EGFR TKIs in this real-world study were likely not given to patients based on mutational status, and in these patients the OS was similar
EP
to chemotherapy and of modest benefit with either therapy.
Comparisons of patient outcomes among treatment groups should be interpreted with caution,
AC C
as it is known that the groups are biologically different given the variable rate of known biomarker status among the population during the study time period. The treatment landscape for use of targeted therapies in NSCLC has evolved over the years, yet there continues to be unmet need in those patients without known targets.
The increases in mOS reported among those patients receiving 1L only, 1L and 2L only, and up to 3L also illustrate the increased likelihood of response to subsequent lines of therapy among patients who respond to 1L as compared with those who do not. These observations are likely
11
ACCEPTED MANUSCRIPT
also explained by differences in the underlying biology of disease in these patients, which is not well understood.
The strength of this study lies in the large, multisite population of mNSCLC patients who were
RI PT
treated with multiple lines of therapy in the community oncology setting. Data collected from the EHR database represent actual treatment information documented as part of routine clinical care, and therefore demonstrate real-world clinical treatment patterns and outcomes, not only those of patients selected as part of clinical trials.
SC
Limitations of this study include the retrospective nature of the evaluation, short follow-up time in more recently selected patients, and potential for documentation bias if there were omissions or errors in the EHR database. Missing data cannot confirm the absence of a condition or value in
M AN U
patients’ medical histories, only that it was not documented. Oral therapies are prescribed through iKM, but the fulfillment of those prescriptions is not observable; therefore, an intent-totreat approach was used for oral therapies. This study was limited to clinics that are part of The US Oncology Network and may not be generalizable to practices outside of the network.
Conclusion
TE D
These data demonstrate timely translation of randomized trial data and integration of novel therapies for mNSCLC in the community setting. Use of targeted therapy was frequently observed in the front-line setting for mutation-positive patients, and survival with immunotherapy in the 2L was similar to that reported in the registrational trials. OS for patients who received multiple lines of therapy was better than in patients who did not. Differences in survival by
AC C
EP
treatment types are likely due to biologic differences.
12
ACCEPTED MANUSCRIPT
Clinical Practice Points •
Identification of biomarkers in mNSCLC such as EGFR mutations and ALK rearrangements helps select patients who will derive maximal benefit from TKIs directed
•
RI PT
against these targets.
In 2015, the first immunotherapy for mNSCLC was FDA-approved for use in patients after progression on platinum-based chemotherapy and after targeted therapy in patients with known mutations. Recent and ongoing studies are evaluating different sequencing options for these treatments.
Effective October 2016, the indication for using EGFR TKIs changed to require patients
SC
•
treated in all lines of therapy for metastatic disease to have a known EGFR-sensitizing
by an FDA-approved test). •
M AN U
mutation (EGFR exon 19 deletions or exon 21 L858R substitution mutations as detected
During the study period, platinum-based chemotherapy was the most frequently used treatment in the 1L setting, followed by targeted therapy or immunotherapy in 2L or 3L.
•
These data demonstrate that patients with actionable mutations can derive greater survival benefit from targeted therapies compared with chemotherapy, and provide
chemotherapy.
Acknowledgements
TE D
evidence that incorporation of immunotherapy can improve OS more than
The authors thank The Lockwood Group (Stamford, CT, USA) for editorial support, which was in
EP
accordance with Good Publication Practice (GPP3) guidelines and funded by AstraZeneca
AC C
(Wilmington, DE, USA).
Funding
This work was supported by AstraZeneca.
References 1. Siegel RL, Miller KD, Jemal A. Cancer statistics, 2017. CA Cancer J Clin. 2017;67(1):730.
13
ACCEPTED MANUSCRIPT
2. National Cancer Institute. Surveillance, Epidemiology, and End Results (SEER) Program. Cancer stat facts: lung and bronchus cancer. https://seer.cancer.gov/statfacts/html/lungb.html. Accessed June 20, 2017. 3. American Cancer Society. What is non-small cell lung cancer?
RI PT
www.cancer.org/cancer/non-small-cell-lung-cancer/about/what-is-non-small-cell-lungcancer.html. Accessed June 20, 2017.
4. National Comprehensive Cancer Network® (NCCN®) Clinical Practice Guidelines in Oncology. Non-small cell lung cancer, v9.2017.
www.nccn.org/professionals/physician_gls/pdf/nscl.pdf. Accessed October 2, 2017.
SC
5. Tarceva [package insert]. Northbrook, IL: OSI Pharmaceuticals, LLC; 2016.
6. Iressa [package insert]. Wilmington, DE: AstraZeneca Pharmaceuticals LP; 2015.
M AN U
7. Gilotrif [package insert]. Ridgefield, CT: Boehringer Ingelheim Pharmaceuticals, Inc.; 2016.
8. Xalkori [package insert]. New York, NY: Pfizer Labs; 2017.
9. Cohen MH, Johnson JR, Chen YF, et al. FDA drug approval summary: erlotinib (Tarceva) tablets. Oncologist. 2005;10(7):461-466.
10. Cohen MH, Williams GA, Sridhara R, et al. FDA drug approval summary: gefitinib (ZD1839) (Iressa) tablets. Oncologist. 2003;8(4):303-306.
TE D
11. Kazandjian D, Blumenthal GM, Chen HY, et al. FDA approval summary: crizotinib for the treatment of metastatic non-small cell lung cancer with anaplastic lymphoma kinase rearrangements. Oncologist. 2014;19(10):e5-e11. 12. Rosell R, Carcereny E, Gervais R, et al. Erlotinib versus standard chemotherapy as first-
EP
line treatment for European patients with advanced EGFR mutation-positive non-smallcell lung cancer (EURTAC): a multicentre, open-label, randomised phase 3 trial. Lancet Oncol. 2012;13(3):239-246.
AC C
13. Solomon BJ, Mok T, Kim DW, et al. PROFILE 1014 Investigators. First-line crizotinib versus chemotherapy in ALK-positive lung cancer. N Engl J Med. 2014;371(23):2167-
2177.
14. Tagrisso [package insert]. Wilmington, DE: AstraZeneca Pharmaceuticals LP; 2017. 15. Alecensa [package insert]. South San Francisco, CA: Genentech USA, Inc.; 2016. 16. Borghaei H, Paz-Ares L, Horn L, et al. Nivolumab versus docetaxel in advanced nonsquamous non-small-cell lung cancer. N Engl J Med. 2015;373(17):1627-1639.
14
ACCEPTED MANUSCRIPT
17. McKay C, Burke T, Cao X, Abernethy AP, Carbone DP. Treatment patterns for advanced non-small-cell lung cancer after platinum-containing therapy in U.S. community oncology clinical practice. Clin Lung Cancer. 2016;17(5):449-460.e7. 18. Abernethy AP, Arunachalam A, Burke T, et al. Real-world first-line treatment and overall
RI PT
survival in non-small cell lung cancer without known EGFR mutations or ALK rearrangements in US community oncology setting. PLoS One. 2017;12(6):e0178420. 19. Khozin S, Blumenthal GM, Jiang X, et al. U.S. Food and Drug Administration approval summary: erlotinib for the first-line treatment of metastatic non-small cell lung cancer
mutations. Oncologist. 2014;19(7):774-779.
SC
with epidermal growth factor receptor exon 19 deletions or exon 21 (L858R) substitution
20. U.S. Food and Drug Administration. FDA-approved drugs: erlotinib (Tarceva).
M AN U
www.fda.gov/Drugs/InformationOnDrugs/ApprovedDrugs/ucm525739.htm. Accessed June 20, 2017.
21. Travis WD, Brambilla E, Nicholson AG, et al. The 2015 World Health Organization classification of lung tumors: impact of genetic, clinical and radiologic advances since the 2004 classification. J Thorac Oncol. 2015;10(9):1243-1260. 22. Fukuoka M, Wu YL, Thongprasert S, et al. Biomarker analyses and final overall survival results from a phase III, randomized, open-label, first-line study of gefitinib versus
TE D
carboplatin/paclitaxel in clinically selected patients with advanced non-small-cell lung cancer in Asia (IPASS). J Clin Oncol. 2011;29(21):2866-2874. 23. Kris MG, Johnson BE, Berry LD, et al. Using multiplexed assays of oncogenic drivers in lung cancers to select targeted drugs. JAMA. 2014;311(19):1998-2006.
EP
24. Sacher AG, Le LW, Lau A, Earle CC, Leighl NB. Real-world chemotherapy treatment patterns in metastatic non-small cell lung cancer: Are patients undertreated? Cancer. 2015;121(15):2562-2569.
AC C
25. Herbst RS, Baas P, Kim DW, et al. Pembrolizumab versus docetaxel for previously treated, PD-L1-positive, advanced non-small-cell lung cancer (KEYNOTE-010): a randomised controlled trial. Lancet. 2016;387(10027):1540-1550.
15
ACCEPTED MANUSCRIPT
Tables Table 1. Baseline Patient and Clinical Characteristics
RI PT
Index Treatment Cohorts Nontargeted Targeted Therapy Chemotherapy (n=685) (n=9969)
Overall (N=10,689)
Immunotherapy (n=35)
68 (27, 90+)
76 (48, 87)
68 (28, 90+)
≤64
3845 (36.0%)
4 (11.4%)
3555 (35.7%)
286 (41.8%)
65-69
2131 (19.9%)
4 (11.4%)
2017 (20.2%)
110 (16.1%)
≥70
4713 (44.1%)
27 (77.1%)
4397 (44.1%)
289 (42.2%)
White
8453 (79.1%)
29 (82.9%)
7943 (79.7%)
481 (70.2%)
African American
1011 (9.5%)
2 (5.7%)
958 (9.6%)
51 (7.4%)
Unknown
1225 (11.5%)
4 (11.4%)
1068 (10.7%)
153 (22.3%)
Race
TE D
Gender
M AN U
Median (min, max) Age group at index
67 (27, 90+)
SC
Age at index date
4873 (45.6%)
18 (51.4%)
4448 (44.6%)
407 (59.4%)
Male
5816 (54.4%)
17 (48.6%)
5521 (55.4%)
278 (40.6%)
6610 (61.8%)
27 (77.1%)
6179 (62.0%)
404 (59.0%)
2554 (23.9%)
5 (14.3%)
2381 (23.9%)
168 (24.5%)
1019 (9.5%)
2 (5.7%)
935 (9.4%)
82 (12.0%)
505 (4.7%)
1 (2.9%)
473 (4.7%)
31 (4.5%)
1 (0.0%)
0 (0.0%)
1 (0.0%)
0 (0.0%)
2264 (21.2%)
14 (40.0%)
2125 (21.3%)
125 (18.2%)
IIIB
765 (7.2%)
1 (2.9%)
740 (7.4%)
24 (3.5%)
IV
6865 (64.2%)
15 (42.9%)
6347 (63.7%)
503 (73.4%)
NA
795 (7.4%)
5 (14.3%)
757 (7.6%)
33 (4.8%)
Practice region South West
AC C
Midwest
EP
Female
Northeast
No information
P value*
.0322
.0017
.3938
<.0001
.1316
Stage at diagnosis I-IIIA
ECOG
16
<.0001
747 (7.0%)
2 (5.7%)
669 (6.7%)
76 (11.1%)
1
6337 (59.3%)
23 (65.7%)
5903 (59.2%)
411 (60.0%)
2
1896 (17.7%)
1 (2.9%)
1804 (18.1%)
91 (13.3%)
3
166 (1.6%)
0 (0.0%)
148 (1.5%)
18 (2.6%)
1543 (14.4%)
9 (25.7%)
1445 (14.5%)
89 (13.0%)
Adenocarcinoma
6392 (59.8%)
15 (42.9%)
5802 (58.2%)
575 (83.9%)
<.0001
Adenosquamous
196 (1.8%)
1 (2.9%)
188 (1.9%)
7 (1.0%)
.1027
Large cell
174 (1.6%)
0 (0.0%)
172 (1.7%)
2 (0.3%)
.0015
Bronchioloalveolar
105 (1.0%)
0 (0.0%)
93 (0.9%)
12 (1.8%)
.0358
Squamous cell
2407 (22.5%)
13 (37.1%)
2355 (23.6%)
39 (5.7%)
<.0001
Neuroendocrine
82 (0.8%)
0 (0.0%)
82 (0.8%)
0 (0.0%)
.0172
Other
277 (2.6%)
0 (0.0%)
268 (2.7%)
9 (1.3%)
.0288
1077 (10.1%)
6 (17.1%)
1029 (10.3%)
42 (6.1%)
.0004
Current
2307 (21.6%)
4 (11.4%)
2257 (22.6%)
46 (6.7%)
Former
6709 (62.8%)
28 (80.0%)
6389 (64.1%)
292 (42.6%)
Never
1315 (12.3%)
3 (8.6%)
985 (9.9%)
327 (47.7%)
232 (2.2%)
0 (0.0%)
219 (2.2%)
13 (1.9%)
0 (0.0%)
119 (1.2%)
7 (1.0%)
NA
RI PT
0
EP
ACCEPTED MANUSCRIPT
TE D
Smoking status
Other tobacco use Unknown
M AN U
Unknown
126 (1.2%)
SC
Histology
<.0001
<.0001
AC C
EGFR mutation status Negative
3454 (32.3%)
13 (37.1%)
3317 (33.3%)
124 (18.1%)
Positive
524 (4.9%)
0 (0.0%)
170 (1.7%)
354 (51.7%)
6694 (62.6%)
22 (62.9%)
6478 (65.0%)
194 (28.3%)
0 (0.0%)
4 (0.0%)
13 (1.9%)
Unknown
<.0001
EGFR T790M mutation status T790M
17 (0.2%)
ALK rearrangement status
17
NR
ACCEPTED MANUSCRIPT
Negative
3536 (33.1%)
12 (34.3%)
3219 (32.3%)
305 (44.5%)
Positive
119 (1.1%)
0 (0.0%)
33 (0.3%)
86 (12.6%)
Equivocal
72 (<0.7%)
1 (2.9%)
64 (0.6%)
7 (1.0%)
NR
Unknown
6962 (65.1%)
22 (62.8%)
6653 (66.7%)
287 (41.9%)
NR
Negative
74 (0.7%)
0 (0.0%)
66 (0.7%)
8 (1.2%)
Positive
51 (0.5%)
2 (5.7%)
46 (0.5%)
3 (0.4%)
10,564 (98.8%)
33 (94.3%)
9857 (98.9%)
674 (98.4%)
Unknown
.5234
SC
PD-L1 status
RI PT
<.0001
M AN U
*P values were comparisons between the targeted and chemo groups. Wilcoxon rank-sum tests for continuous variables and chi-square tests were used for categorical variables when all cell counts were ≥5. The Fisher test for categorical variables was used when there were cell counts <5. Abbreviations: ECOG = Eastern Cooperative Oncology Group; EGFR = epidermal growth factor receptor;
AC C
EP
TE D
ALK = anaplastic lymphoma kinase.
18
ACCEPTED MANUSCRIPT
Table 2. Treatment Patterns
Index Treatment Cohorts Overall
Nontargeted Chemotherapy (n=9969)
Targeted Therapy (n=685)
Immunotherapy (n=35)
Carboplatin + paclitaxel
3914 (36.6%)
3914 (39.3%)
0 (0.0%)
0 (0.0%)
Carboplatin + pemetrexed
2231 (20.9%)
2231 (22.4%)
0 (0.0%)
0 (0.0%)
Bevacizumab + carboplatin + paclitaxel
1395 (13.1%)
1395 (14.0%)
0 (0.0%)
0 (0.0%)
392 (3.7%)
Crizotinib
82 (0.8%)
Nivolumab
32 (0.3%)
Other*
M AN U
Erlotinib
2643 (24.7%)
Chemotherapy Immunotherapy
0 (0.0%)
0 (0.0%)
82 (12.0%)
0 (0.0%)
0 (0.0%)
0 (0.0%)
32 (91.4%)
2429 (24.4%)
211 (30.8%)
3 (8.6%)
2695 (27.0%)
89 (13.0%)
6 (17.1%)
727 (6.8%)
715 (7.2%)
11 (1.6%)
1 (2.9%)
718 (6.7%)
491 (4.9%)
227 (33.1%)
0 (0.0%)
6068 (60.9%)
358 (52.3%)
28 (80.0%)
6454 (60.4%)
AC C
None
392 (57.2%)
2790 (26.1%)
EP
Targeted therapy
0 (0.0%)
TE D
LOT2 treatment group
SC
LOT1 regimens
RI PT
(N=10,689)
Overall (N=4235)
Index Treatment Cohorts Nontargeted Chemotherapy (n=2790)
Targeted Immunotherapy Chemotherapy (n=727) (n=718)
LOT2 regimens Pemetrexed
696 (16.4%)
696 (16.4%)
0 (0.0%)
0 (0.0%)
Docetaxel
694 (16.4%)
694 (16.4%)
0 (0.0%)
0 (0.0%)
Nivolumab
674 (15.9%)
0 (0.0%)
0 (0.0%)
674 (92.7%)
19
ACCEPTED MANUSCRIPT
Erlotinib
458 (10.8%)
0 (0.0%)
458 (10.8%)
0 (0.0%)
Other*
1713 (40.4%)
1400 (50.2%)
260 (36.2%)
53 (7.3.3%)
Chemo
921 (21.7%)
697 (25.0%)
127 (17.7%)
97 (13.3%)
Targeted
355 (8.4%)
256 (9.2%)
84 (11.7%)
15 (2.1%)
Immunotherapy
304 (7.2%)
263 (9.4%)
33 (4.6%)
8 (1.1%)
2655 (62.7%)
1574 (56.4%)
474 (66.0%)
607 (83.5%)
Nivolumab
270 (17.1%)
245 (20.1%)
20 (8.2%)
5 (4.2%)
Docetaxel
227 (14.4%)
160 (13.2%)
31 (12.7%)
36 (30.0%)
Erlotinib
224 (14.2%)
208 (17.1%)
5 (2.0%)
11 (9.2%)
Pemetrexed
158 (10.0%)
126 (10.4%)
24 (9.8%)
8 (6.7%)
Gemcitabine
128 (8.1%)
104 (8.6%)
7 (2.9%)
17 (14.2%)
157 (64.3%)
43 (35.8%)
TE D
Other*
M AN U
LOT3 regimens
SC
None
RI PT
LOT3 treatment group
573 (36.3%)
373 (30.7%)
*“Other” treatments individually account for less than 10% of each treatment cohort.
AC C
EP
Abbreviation: LOT = line of therapy.
20
ACCEPTED MANUSCRIPT
Figure Captions
Figure 1. Consort Diagram. A total of 10,689 patients met eligibility criteria and initiated 1L during the study period.
RI PT
therapy with either nontargeted chemotherapy, targeted therapy, or other (immunotherapy)
patients from the initiation of first line of treatment.
SC
Figure 2. Overall Survival by First-Line. Kaplan-Meier estimated overall survival for
Figure 3. Overall Survival by Second-Line. Kaplan-Meier estimated overall survival for
M AN U
patients from the initiation of second line of treatment.
Figure 4. Overall Survival by Third-Line. Kaplan-Meier estimated overall survival for
AC C
EP
TE D
patients from the initiation of third line of treatment.
21
ACCEPTED MANUSCRIPT
Figures
AC C
EP
TE D
M AN U
SC
RI PT
Figure 1. Consort Diagram
Abbreviations: NSCLC = non-small cell lung cancer; USON = US Oncology Network.
22
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
Figure 2. Overall Survival by First-line
Abbreviations: CI = confidence interval; LOT = line of therapy; OS = overall survival.
23
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
Figure 3. Overall Survival by Second-line
Abbreviations: CI = confidence interval; OS = overall survival. 24
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
Figure 4. Overall Survival by Third-Line
Abbreviations: CI = confidence interval; OS = overall survival. 25
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
Figure 5. Overall Survival by LOTs received
Abbreviations: CI = confidence interval; LOT = line of therapy; OS = overall survival.
26
ACCEPTED MANUSCRIPT
Table 1. Baseline Patient and Clinical Characteristics
Index Treatment Cohorts Nontargeted Targeted Therapy Chemotherapy (n=685) (n=9969)
Immunotherapy (n=35)
68 (27, 90+)
76 (48, 87)
68 (28, 90+)
67 (27, 90+)
≤64
3845 (36.0%)
4 (11.4%)
3555 (35.7%)
286 (41.8%)
65-69
2131 (19.9%)
4 (11.4%)
2017 (20.2%)
110 (16.1%)
≥70
4713 (44.1%)
27 (77.1%)
4397 (44.1%)
289 (42.2%)
White
8453 (79.1%)
29 (82.9%)
7943 (79.7%)
481 (70.2%)
African American
1011 (9.5%)
2 (5.7%)
958 (9.6%)
51 (7.4%)
Unknown
1225 (11.5%)
4 (11.4%)
1068 (10.7%)
153 (22.3%)
Female
4873 (45.6%)
18 (51.4%)
4448 (44.6%)
407 (59.4%)
Male
5816 (54.4%)
17 (48.6%)
5521 (55.4%)
278 (40.6%)
South
6610 (61.8%)
27 (77.1%)
6179 (62.0%)
404 (59.0%)
West
2554 (23.9%)
5 (14.3%)
2381 (23.9%)
168 (24.5%)
1019 (9.5%)
2 (5.7%)
935 (9.4%)
82 (12.0%)
505 (4.7%)
1 (2.9%)
473 (4.7%)
31 (4.5%)
1 (0.0%)
0 (0.0%)
1 (0.0%)
0 (0.0%)
2264 (21.2%)
14 (40.0%)
2125 (21.3%)
125 (18.2%)
765 (7.2%)
1 (2.9%)
740 (7.4%)
24 (3.5%)
IV
6865 (64.2%)
15 (42.9%)
6347 (63.7%)
503 (73.4%)
NA
795 (7.4%)
5 (14.3%)
757 (7.6%)
33 (4.8%)
0
747 (7.0%)
2 (5.7%)
669 (6.7%)
76 (11.1%)
1
6337 (59.3%)
23 (65.7%)
5903 (59.2%)
411 (60.0%)
RI PT
Overall (N=10,689)
Race
Midwest Northeast
AC C
No information
EP
Practice region
TE D
Gender
M AN U
Median (min, max) Age group at index
SC
Age at index date
P value*
.0322
.0017
.3938
<.0001
.1316
Stage at diagnosis I-IIIA IIIB
<.0001
ECOG <.0001
2
1896 (17.7%)
1 (2.9%)
1804 (18.1%)
91 (13.3%)
3
166 (1.6%)
0 (0.0%)
148 (1.5%)
18 (2.6%)
1543 (14.4%)
9 (25.7%)
1445 (14.5%)
89 (13.0%)
Adenocarcinoma
6392 (59.8%)
15 (42.9%)
5802 (58.2%)
575 (83.9%)
Adenosquamous
196 (1.8%)
1 (2.9%)
188 (1.9%)
7 (1.0%)
.1027
Large cell
174 (1.6%)
0 (0.0%)
172 (1.7%)
2 (0.3%)
.0015
Bronchioloalveolar
105 (1.0%)
0 (0.0%)
93 (0.9%)
12 (1.8%)
.0358
Squamous cell
2407 (22.5%)
13 (37.1%)
2355 (23.6%)
39 (5.7%)
<.0001
Neuroendocrine
82 (0.8%)
0 (0.0%)
82 (0.8%)
0 (0.0%)
.0172
Other
277 (2.6%)
0 (0.0%)
268 (2.7%)
9 (1.3%)
.0288
1077 (10.1%)
6 (17.1%)
1029 (10.3%)
42 (6.1%)
.0004
Current
2307 (21.6%)
4 (11.4%)
2257 (22.6%)
46 (6.7%)
Former
6709 (62.8%)
28 (80.0%)
6389 (64.1%)
292 (42.6%)
Never
1315 (12.3%)
3 (8.6%)
985 (9.9%)
327 (47.7%)
Other tobacco use
232 (2.2%)
0 (0.0%)
219 (2.2%)
13 (1.9%)
Unknown
126 (1.2%)
0 (0.0%)
119 (1.2%)
7 (1.0%)
EP
ACCEPTED MANUSCRIPT
13 (37.1%)
3317 (33.3%)
124 (18.1%)
524 (4.9%)
0 (0.0%)
170 (1.7%)
354 (51.7%)
6694 (62.6%)
22 (62.9%)
6478 (65.0%)
194 (28.3%)
0 (0.0%)
4 (0.0%)
13 (1.9%)
NR
12 (34.3%)
3219 (32.3%)
305 (44.5%)
<.0001
NA
TE D
Smoking status
SC
M AN U
Unknown
RI PT
Histology <.0001
<.0001
EGFR mutation status
3454 (32.3%)
AC C
Negative Positive
Unknown
<.0001
EGFR T790M mutation status T790M
17 (0.2%)
ALK rearrangement status Negative
3536 (33.1%)
2
ACCEPTED MANUSCRIPT
Positive
119 (1.1%)
0 (0.0%)
33 (0.3%)
86 (12.6%)
Equivocal
72 (<0.7%)
1 (2.9%)
64 (0.6%)
7 (1.0%)
NR
Unknown
6962 (65.1%)
22 (62.8%)
6653 (66.7%)
287 (41.9%)
NR
Negative
74 (0.7%)
0 (0.0%)
66 (0.7%)
8 (1.2%)
Positive
51 (0.5%)
2 (5.7%)
46 (0.5%)
3 (0.4%)
10,564 (98.8%)
33 (94.3%)
9857 (98.9%)
674 (98.4%)
.5234
SC
Unknown
RI PT
PD-L1 status
*P values were comparisons between the targeted and chemo groups. Wilcoxon rank-sum tests for continuous variables and chi-square tests were used for categorical variables when all cell counts were
M AN U
≥5. The Fisher test for categorical variables was used when there were cell counts <5. Abbreviations: ECOG = Eastern Cooperative Oncology Group; EGFR = epidermal growth factor receptor;
AC C
EP
TE D
ALK = anaplastic lymphoma kinase.
3
ACCEPTED MANUSCRIPT
Table 2. Treatment Patterns
Index Treatment Cohorts Overall
Nontargeted Chemotherapy (n=9969)
Targeted Therapy (n=685)
Immunotherapy (n=35)
Carboplatin + paclitaxel
3914 (36.6%)
3914 (39.3%)
0 (0.0%)
0 (0.0%)
Carboplatin + pemetrexed
2231 (20.9%)
2231 (22.4%)
0 (0.0%)
0 (0.0%)
Bevacizumab + carboplatin + paclitaxel
1395 (13.1%)
1395 (14.0%)
0 (0.0%)
0 (0.0%)
392 (3.7%)
Crizotinib
82 (0.8%)
Nivolumab
32 (0.3%)
Other*
M AN U
Erlotinib
2643 (24.7%)
Chemotherapy Immunotherapy
0 (0.0%)
0 (0.0%)
82 (12.0%)
0 (0.0%)
0 (0.0%)
0 (0.0%)
32 (91.4%)
2429 (24.4%)
211 (30.8%)
3 (8.6%)
2695 (27.0%)
89 (13.0%)
6 (17.1%)
727 (6.8%)
715 (7.2%)
11 (1.6%)
1 (2.9%)
718 (6.7%)
491 (4.9%)
227 (33.1%)
0 (0.0%)
6068 (60.9%)
358 (52.3%)
28 (80.0%)
6454 (60.4%)
AC C
None
392 (57.2%)
2790 (26.1%)
EP
Targeted therapy
0 (0.0%)
TE D
LOT2 treatment group
SC
LOT1 regimens
RI PT
(N=10,689)
Overall (N=4235)
Index Treatment Cohorts Nontargeted Chemotherapy (n=2790)
Targeted Immunotherapy Chemotherapy (n=727) (n=718)
LOT2 regimens Pemetrexed
696 (16.4%)
696 (16.4%)
0 (0.0%)
0 (0.0%)
Docetaxel
694 (16.4%)
694 (16.4%)
0 (0.0%)
0 (0.0%)
Nivolumab
674 (15.9%)
0 (0.0%)
0 (0.0%)
674 (92.7%)
ACCEPTED MANUSCRIPT
Erlotinib
458 (10.8%)
0 (0.0%)
458 (10.8%)
0 (0.0%)
Other*
1713 (40.4%)
1400 (50.2%)
260 (36.2%)
53 (7.3.3%)
Chemo
921 (21.7%)
697 (25.0%)
127 (17.7%)
97 (13.3%)
Targeted
355 (8.4%)
256 (9.2%)
84 (11.7%)
15 (2.1%)
Immunotherapy
304 (7.2%)
263 (9.4%)
33 (4.6%)
8 (1.1%)
2655 (62.7%)
1574 (56.4%)
474 (66.0%)
607 (83.5%)
Nivolumab
270 (17.1%)
245 (20.1%)
20 (8.2%)
5 (4.2%)
Docetaxel
227 (14.4%)
160 (13.2%)
31 (12.7%)
36 (30.0%)
Erlotinib
224 (14.2%)
208 (17.1%)
5 (2.0%)
11 (9.2%)
Pemetrexed
158 (10.0%)
126 (10.4%)
24 (9.8%)
8 (6.7%)
Gemcitabine
128 (8.1%)
104 (8.6%)
7 (2.9%)
17 (14.2%)
157 (64.3%)
43 (35.8%)
TE D
Other*
M AN U
LOT3 regimens
SC
None
RI PT
LOT3 treatment group
573 (36.3%)
373 (30.7%)
*“Other” treatments individually account for less than 10% of each treatment cohort.
AC C
EP
Abbreviation: LOT = line of therapy.
2
ACCEPTED MANUSCRIPT
Figure Captions
Figure 1. Consort Diagram. A total of 10,689 patients met eligibility criteria and initiated 1L during the study period.
RI PT
therapy with either nontargeted chemotherapy, targeted therapy, or other (immunotherapy)
patients from the initiation of first line of treatment.
SC
Figure 2. Overall Survival by First-Line. Kaplan-Meier estimated overall survival for
Figure 3. Overall Survival by Second-Line. Kaplan-Meier estimated overall survival for
M AN U
patients from the initiation of second line of treatment.
Figure 4. Overall Survival by Third-Line. Kaplan-Meier estimated overall survival for
AC C
EP
TE D
patients from the initiation of third line of treatment.
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT