Associations Among Optimal Lung Cancer Treatment, Clinical Outcomes, and Health Care Utilization in Patients Who Underwent Comprehensive Genomic Profiling

Authors:
Adam C. Powell Payer+Provider Syndicate, Newton, Massachusetts

Search for other papers by Adam C. Powell in
Current site
Google Scholar
PubMed
Close
 PhD
,
Elifnur Yay Donderici Guardant Health Inc., Palo Alto, California

Search for other papers by Elifnur Yay Donderici in
Current site
Google Scholar
PubMed
Close
 PhD
,
Nicole J. Zhang Guardant Health Inc., Palo Alto, California

Search for other papers by Nicole J. Zhang in
Current site
Google Scholar
PubMed
Close
 MPH
,
Shaun P. Forbes Guardant Health Inc., Palo Alto, California

Search for other papers by Shaun P. Forbes in
Current site
Google Scholar
PubMed
Close
 PhD
,
Julie Wiedower Guardant Health Inc., Palo Alto, California

Search for other papers by Julie Wiedower in
Current site
Google Scholar
PubMed
Close
 MS, CGC
,
Amy C. McNeal Guardant Health Inc., Palo Alto, California

Search for other papers by Amy C. McNeal in
Current site
Google Scholar
PubMed
Close
 MBA
, and
Mark D. Hiatt Guardant Health Inc., Palo Alto, California

Search for other papers by Mark D. Hiatt in
Current site
Google Scholar
PubMed
Close
 MD, MBA, MS
Full access

Background: Although immune checkpoint inhibitor immunotherapies are contraindicated as first-line treatment of advanced non–small cell lung cancer (NSCLC) in patients with ALK rearrangement and EGFR mutation, many receive them. The purpose of this study was to examine the association between optimal first-line treatment in this population and clinical outcomes. Methods: Claims and genomic data from patients with advanced or metastatic NSCLC were extracted from a nationally representative GuardantINFORM dataset. Patients who had their first claim mentioning advanced or metastatic NSCLC between March 2019 and February 2020 and had ALK rearrangement or EGFR mutation detected by comprehensive genomic profiling were included in this study. Patients were classified as having received optimal or suboptimal first-line treatment. Claims were reviewed to determine real-world time to next treatment, real-world time to discontinuation, and health services utilization (emergency department, inpatient, and outpatient) in the 12 months following first-line treatment initiation. Survival analyses were conducted using Kaplan-Meier plots and Cox proportional hazard models. Health services utilization was compared between the groups using t tests and negative binomial models. Results: Of the 359 patients included, 280 (78.0%) received optimal first-line treatment. Optimally treated patients had longer median real-world time to next treatment (11.2 vs 4.4 months; P<.01) and real-world time to discontinuation (10.4 vs 1.9 months; P<.01). The optimal group had significantly fewer emergency department presentations (0.76 vs 1.27; P<.01) and outpatient visits (22.9 vs 42.7; P<.01) than the suboptimal group but did not significantly differ in inpatient utilization. Adjusted utilization analysis yielded similar findings. Conclusions: Patients with NSCLC who received optimal treatment, as determined by comprehensive genomic profiling using next-generation sequencing–based circulating tumor DNA testing (Guardant360), had significantly superior clinical and utilization outcomes, reinforcing existing guidelines recommending profiling at the onset of treatment.

Background

The optimal first-line approach to treating non–small cell lung cancer (NSCLC) varies in accordance with a patient’s biomarkers. Comprehensive genomic profiling (CGP) tests can be used to determine the course of first-line treatment best attuned to a patient’s biomarkers, because they concurrently detect single-nucleotide variants, genetic insertions, genetic deletions, copy number amplifications, and fusions. Immune checkpoint inhibitor (ICI) immunotherapies for the treatment of NSCLC require patients to have a specific biomarker profile detectable via CGP—negativity for ALK rearrangement (fusions) and EGFR mutations (alterations L858R/EXON19 deletions)—or to have had prior toxicity or progression on oral therapies targeting mutations in these genes.1 Although CGP results can guide care for multiple types of cancer, empirical data suggest that CGP has the greatest impact on the course of treatment for patients with NSCLC.2

ASCO and NCCN have recommended that biomarker testing for ALK and EGFR at the time of diagnosis of advanced NSCLC, and before the administration of ICIs, be considered the standard of care.3,4 Nonetheless, testing continues to be underutilized, underdiagnosis occurs, and evidence shows that patients receiving contraindicated therapy may experience hyperprogression.5,6 One study found that 64.4% of potentially eligible patients with advanced NSCLC were not benefiting from precision oncology care appropriate for their disease.7 A study examining a population of patients with NSCLC who tested positive for EGFR mutation found that 16.0% initiated treatment before EGFR testing results were available.8 Although the evidence supporting the use of genomically targeted treatments is substantial, the real-world evidence is generally sparse on how using CGP in guiding treatment impacts outcomes and utilization.9

CGP enables physicians to rapidly obtain a comprehensive view of a patient’s biomarkers. In contrast, sequential testing introduces delays into the care process, because the results of initial tests are used to make subsequent testing decisions. A model of genomic testing for patients with metastatic NSCLC found that panel and next-generation sequencing approaches to testing yielded results nearly 3 weeks faster than exclusionary and sequential testing.10

Previous researchers have characterized first-line treatment patterns found in the SEER database pertaining to patients with traditional Medicare health plans, aged ≥65 years, with advanced NSCLC, initially diagnosed between 2007 and 2011.11 Since then, numerous new ICIs have been brought to market, including pembrolizumab (in 2014), nivolumab (in 2014), and durvalumab (in 2017), which are not suitable first-line treatments for patients with ALK rearrangement and EGFR mutation.12 As a consequence, research is needed to explore treatment patterns in an environment in which more options are available.

Previous research has shown that receipt of genomic testing and targeted therapy is associated with significantly longer progression-free survival in a matched cohort of patients with metastatic cancer.5 When EGFR testing for patients with nonsquamous metastatic NSCLC was introduced in Alberta, Canada, there was a significant improvement in overall survival.13 Likewise, a study of patients with diverse refractory cancers who underwent CGP and then were treated in a manner that was either matched or unmatched to their genomic profile concluded that patients receiving matched therapy had longer time-to-treatment failure and observed overall survival.14

The purpose of this study was to assess the consequences of receipt of suboptimal first-line treatment in a population of patients who were determined at some point during their treatment to have ALK rearrangement or EGFR mutation. Although all patients included in the study ultimately received CGP, had they all received it more promptly or the results been uniformly heeded, it is possible that the patients would have experienced better outcomes. To highlight the consequences of optimal versus suboptimal first-line treatment, this study examines the association between the first-line treatment selected and downstream clinical outcomes, utilization, and adverse events.

Methods

Study Design

This study was reviewed by the Advarra Institutional Review Board (Pro00061858) and received an exemption from oversight on March 9, 2022, in accordance with the regulations found at 45 CFR 46.104(d)(4). The study was conducted in accordance with the Declaration of Helsinki.

Data Source and Sample Population

This retrospective, observational study was conducted using the nationally representative GuardantINFORM dataset, containing anonymized health care claims and next-generation sequencing–based circulating tumor DNA genomic data pertaining to >200,000 United States–based patients with advanced solid-stage tumors who had at least 1 CGP test performed using a Guardant360 comprehensive liquid biopsy between June 1, 2014, and March 31, 2022. Patients met inclusion criteria if their first claim mentioning advanced or metastatic cancer (ICD-10 diagnosis codes C77.*, C78.*, and C79.*) occurred between March 1, 2019, and February 29, 2020. These dates capture a year of data preceding the COVID-19 pandemic and largely following the FDA’s expansion of the indications for pembrolizumab for the first-line treatment of NSCLC in April 2019.15 The date of the first claim mentioning advanced or metastatic NSCLC served as the index event. Patients were excluded if they did not have ≥1 claims listing an NSCLC diagnosis code in the year before the index date, if they did not have an EGFR mutation or ALK rearrangement detected by CGP, if they did not have a minimum of 1 year of follow-up data after the index date (unless they died during the period), if they did not have ≥2 claims (medical or pharmacy) in the year prior to the index event (a proxy for enrollment), if they participated in clinical trials between 1 year before and 1 year after the index date (as indicated by having claims listing the ICD-10 diagnosis code Z00.6), if they did not start first-line therapy within 60 days after the index date, or if their home state was missing. The patient’s home state was used to construct several control variables and thus played a role in the analysis.

Measurement

A physician and a genetic counselor created definitions for optimal versus suboptimal therapy (supplemental eTable 1, available with this article at JNCCN.org) and diagnosis of lung cancer (supplemental eTable 2) considering guidelines from ASCO and NCCN.3,4 Patients were classified as having received optimal or suboptimal therapy (the independent variable) by reviewing their medication claims in the 60 days following the index event, which was defined as the patient’s first claim mentioning a diagnosis of metastatic lung cancer.

First-line therapy was defined as any treatment in concordance with NCCN Clinical Practice Guidelines in Oncology (NCCN Guidelines) for NSCLC that was commenced within 90 days after metastatic diagnosis date.3 The end of first-line therapy was defined when a 42-day period with no treatment was observed or when a second drug or regimen started on or after day 30. Continuation maintenance therapy (ie, a first-line drug continued as maintenance therapy after a gap of 21–41 days) would be captured as first-line therapy, whereas any switch in therapy initiated after 30 days would be captured as second-line therapy.

Several dependent variables were generated. Real-world time to next treatment (rwTTNT) was defined as the length of time from the start of the first line of treatment to the start of the next line of treatment. A treatment was considered to be second line only if it was initiated at least 21 days after first-line treatment. Death was considered an event if the patient died before the end of therapy or within 90 days of the end of therapy, or if the patient was lost to follow-up and died before the end of therapy. Patients were censored without progression if their last activity date fell before the start of next treatment.

An additional dependent variable was created to measure real-world time to discontinuation (rwTTD), which was defined as the time from the start of the line of treatment to the date the patient discontinued treatment. Death was considered an event when patients died before the end of therapy or within 90 days of the end of therapy, or if the patient was lost to follow-up and died before the end of therapy. Patients were censored if they reached the end of follow-up before the derived end of treatment. Additional dependent variables pertaining to utilization were created to capture the number of emergency department (ED) presentations, inpatient admissions, and outpatient visits that each patient experienced in the 3, 6, and 12 months following the date of first-line treatment initiation.

Adjusted analyses controlled for demographic and social factors with the potential to impact treatment discontinuation and health service utilization, which were the patient’s age on the index date, sex as reported on claims, history of ever using tobacco, Elixhauser comorbidity index16 (defined using the codes provided in supplemental eTable 3), and the percentage of households in rural areas in the patient’s state.17 Additional control variables were considered for inclusion but ultimately eliminated due to collinearity. The control variables omitted were the percentage of adults in the patient’s state who had ever been told that they had asthma as of 2020,18 the percentage of adults in the patient’s state who were high school graduates or higher,19 the percentage of households in the patient’s state earning <$25,000 per year and the percentage of households earning >$200,000 per year,20 the percentage of the population in the patient’s state that reported being White alone,21 and the lung cancer incidence in the patient’s state between 2014 and 2018.22

Analysis

Descriptive statistics were reported for the overall population as well as for the groups receiving optimal and suboptimal treatment. Comparisons were made between the groups using chi-square tests when outcomes were binary and Student’s t tests when outcomes were continuous.

The univariate relationship between optimal versus suboptimal therapy, rwTTNT, and rwTTD was assessed using Kaplan-Meier plots. The plots considered events occurring in the 12 months after the index date. In addition to depicting the relationship between type of therapy and outcomes, the analysis reported the censored log-rank to determine whether the difference in survival between the 2 groups was statistically significant.

To assess the relationship between optimal versus suboptimal treatment, rwTTNT, and rwTTD, after accounting for other factors that may influence care, 2 Cox proportional hazards models were run, considering outcomes observed in the 12 months after first-line initiation. The models each adjusted for the aforementioned control variables. Outcomes from the models were reported as hazard ratios (HRs) with 95% confidence intervals.

Unadjusted health care utilization in the ED, inpatient, and outpatient settings by the optimal and suboptimal groups was compared at 3 separate time intervals: 3, 6, and 12 months after first-line initiation. Student’s t tests were used to assess whether differences in utilization between the groups were significant. Adjusted multivariable analyses were conducted to assess the impact of optimal versus suboptimal treatment on ED, inpatient, and outpatient utilization in the first 12 months following first-line initiation. Because it is common for patients to have no ED or inpatient visits but rare for them to have no outpatient visits, zero-inflated negative binomial models were used to assess which factors were associated with ED and inpatient utilization, whereas a regular negative binomial model was used to assess which factors were associated with outpatient utilization.23

Given that patients receiving treatment for NSCLC often experience adverse events, claims in the 12 months following first-line treatment initiation were reviewed from chemotherapy-related and ICI-related adverse events (defined in supplemental eTable 4). Patients were classified as having had an adverse event if they had ≥1 claim mentioning it. Chi-square tests were used to compare rates of adverse events between the optimal and suboptimal groups.

Results

Population

A total of 59,278 patients were assessed for eligibility (Figure 1). After exclusion criteria were applied, 359 patients remained, of which 280 (78.0%) fell into the optimal category and 79 (22.0%) into the suboptimal category (Table 1). A comparison of the descriptive statistics of the optimal and suboptimal populations revealed that the suboptimal population had a significantly higher average Elixhauser comorbidity index than the optimal population (5.65 vs 4.55). None of the other factors significantly differed between the 2 populations.

Figure 1.
Figure 1.

CONSORT diagram.

Abbreviation: NSCLC, non–small cell lung cancer.

Citation: Journal of the National Comprehensive Cancer Network 22, 1D; 10.6004/jnccn.2023.7073

Table 1.

Descriptive Statistics

Table 1.

As a consequence of the sample selection procedure, all patients had CGP at least once. CGP can be performed before first-line initiation, after first-line initiation, or both. There was a significant difference in the timing of CGP testing between the optimal and suboptimal groups. In the optimal group, 53.6% had CGP testing before first-line initiation versus 20.3% in the suboptimal group (P<.001). Likewise, in the optimal group, 68.2% had CGP testing after first-line initiation versus 87.3% in the suboptimal group (P=.0008). These rates sum to >100% because some patients had testing both before and after first-line initiation: 21.8% in the optimal group versus 7.6% in the suboptimal group (P=.0043). The gap between CGP and first-line therapy was a median of 27 days (minimum, 1 day; maximum, 720 days; IQR, 9–37 days).

Time Series Analyses

Kaplan-Meier plots (Figure 2) revealed significant (P<.0001) differences in rwTTNT, with a median time of 11.2 months in the optimal group and 4.4 months in the suboptimal group. A total of 195 patients initiated second-line therapy: 132 patients in the optimal group and 63 patients in the suboptimal group. The plot for rwTTD revealed similar findings, with significant (P<.0001) differences between the 2 groups. The optimal group had a median rwTTD of 10.4 months, whereas the suboptimal group had a median rwTTD of 1.9 months. The difference in overall survival between the optimal and suboptimal groups was not statistically significant, and therefore these plots are not presented.

Figure 2.
Figure 2.

Kaplan-Meier plots of adjusted (A) real-world time to next treatment and (B) real-world time to discontinuation.

Abbreviation: NR, not reached.

Citation: Journal of the National Comprehensive Cancer Network 22, 1D; 10.6004/jnccn.2023.7073

A series of 3 post hoc sensitivity analyses were conducted, examining rwTTNT and rwTTD for 3 subpopulations: patients who received CGP before first-line initiation (optimal, n=150; suboptimal, n=16), patients who received CGP after first-line initiation (optimal, n=191; suboptimal, n=69), and patients who received CGP both before and after first-line initiation (optimal, n=61; suboptimal, n=6). Within the subpopulation who received CGP before first-line initiation, the difference in rwTTNT between the optimal and suboptimal groups (median, 11.3 vs 6.1 months, respectively; P=.06) did not meet the P<.05 threshold for significance. However, there was a significant difference in rwTTD between the groups (median, 10.3 vs 3.5 months, respectively; P=.02). Among patients who received CGP after first-line initiation, the optimal group had significantly better rwTTD (10.8 vs 2.1 months, respectively; P<.0001;) and rwTTNT (11.4 vs 4.3 months; P<.0001). Patients who received CGP both before and after first-line initiation did not significantly differ in their rwTTD (P=.83) and rwTTNT (P=.54).

Adjusted time-series analysis using Cox proportional hazards models likewise found an association between optimal treatment and better outcomes (Table 2). Patients in the optimal group had significantly longer rwTTNT (HR, 0.31; 95% CI, 0.23–0.42), and patients in a state with a higher percentage of its households living in rural areas had significantly shorter rwTTNT (HR, 1.03; 95% CI, 1.01–1.04). Patients in the optimal group also had significantly longer rwTTD (HR, 0.31; 95% CI, 0.23–0.42). Meanwhile, patients aged ≥65 years (HR, 1.37; 95% CI, 1.02–1.83) and residing in a state with a higher percentage of its households living in rural areas (HR, 1.03; 95% CI, 1.02–1.04) had significantly shorter rwTTD.

Table 2.

Parameter Estimates of Cox Proportional Hazards Models Consider the 12 Months After First-Line Initiation

Table 2.

Utilization

Utilization of health services by the optimal and suboptimal populations was compared at 3, 6, and 12 months following first-line therapy initiation (Table 3). In each period, patients receiving optimal treatment had significantly fewer ED presentations and outpatient visits than those receiving suboptimal treatment. Twelve months after first-line therapy initiation, patients in the optimal group had an average of 0.76 ED presentations, whereas those in the suboptimal group had an average of 1.27 presentations (P<.01). Likewise, patients in the optimal group had an average of 22.93 outpatient visits, whereas those in the suboptimal group had an average of 42.70 (P<.01).

Table 3.

Comparison of Health Care Utilization at 3, 6, and 12 Months Between Optimal and Suboptimal Groups

Table 3.

An adjusted analysis considering health care utilization in the first 12 months concurred (Table 4). Optimal treatment was associated with significantly lower ED utilization (estimate: −0.53; 95% CI, −0.95 to −0.10) and significantly lower outpatient utilization (estimate: −0.62; 95% CI, −0.83 to −0.42). The adjusted analysis additionally found that a higher Elixhauser comorbidity index was significantly associated with greater ED utilization (estimate: 0.08; 95% CI, 0.01 to 0.15) and outpatient utilization (estimate: 0.05; 95% CI, 0.01 to 0.08) and that age ≥65 years was significantly associated with greater inpatient utilization (estimate: 0.69; 95% CI, 0.21 to 1.17) and lower outpatient utilization (estimate: −0.41; 95% CI, −0.60 to −0.23).

Table 4.

Adjusted Analysis Examining Health Care Utilization at 12 Months Following First-Line Initiation

Table 4.

Adverse Events

In the 12 months before first-line initiation, the optimal and suboptimal groups did not differ significantly in the rates at which their members experienced the adverse events considered. In the 12 months following first-line initiation, patients in the suboptimal group were significantly more likely to experience neutropenia (11.4% vs 2.5%; P<.001), tinnitus (36.7% vs 18.2%; P<.001), and pneumonitis (13.4% vs 2.5%; P<.0001). After first-line initiation, the optimal and suboptimal groups did not significantly differ in their rates of acute kidney injury, diabetes mellitus, paraneoplastic neuromyopathy and neuropathy, or transaminitis. The optimal group did not have significantly higher rates of adverse events for any of the categories examined following first-line initiation.

Discussion

This study found that patients who received suboptimal first-line treatment had significantly worse rwTTNT and rwTTD regardless of whether adjustments were made for potential confounders. Furthermore, patients receiving optimal first-line treatment had significantly lower ED and outpatient utilization at 3, 6, and 12 months and fewer adverse events. Collectively, these findings suggest that ensuring that patients are optimally treated leads to improved clinical outcomes and reduced utilization. In doing so, these findings build upon the literature suggesting that patients should receive CGP before initiating first-line treatment.3,24

Prior analysis has suggested that the use of CGP to guide the treatment of NSCLC is associated with both better patient outcomes and a modest financial impact driven by increased survival.1012,2527 Furthermore, patients with EGFR mutations have been shown to have significantly longer time to next treatment or death if they receive an EGFR tyrosine kinase inhibitor as first-line treatment, rather than immunotherapy or chemotherapy.8 The findings of this study build upon the literature by highlighting the inferior outcomes that patients may have if they receive suboptimal first-line treatment. Despite the evidence supporting CGP, patients struggle to access it, particularly in developing countries.27

Patients from states with higher percentages of households living in rural areas were found to have had significantly shorter rwTTNT and rwTTD. Although the urbanicity of the patients in this study is not known, patients from states with proportionally larger rural populations are more likely to come from rural backgrounds themselves. Prior research has found that rural patients have worse access to oncologists and inferior cancer outcomes compared with urban patients.28,29

A few limitations must be considered when interpreting the findings of this study. Patient medical records were not available, and patients were not categorized by stage or subtype of lung cancer in the analysis. Although diagnosis codes listed on claims were used to identify patients with advanced or metastatic lung cancer, it is possible that some patients had claims that were inappropriately coded. Because clinical data were not available, it is unknown why some patients were given suboptimal first-line treatments. It is possible that the patients initiated them while awaiting findings from CGP and then pursued optimal therapy after a washout period. Patients receiving optimal treatment had a significantly lower average Elixhauser comorbidity index (4.55 vs 5.65; P=.001), suggesting that this population was in somewhat better health than the population receiving suboptimal treatment (Table 1). There was likewise a higher proportion of female patients in the optimal group (66.07% vs 54.43%; P=.052), although the difference was not statistically significant using P=.05 as a cutoff. Although the reasons that patients received suboptimal first-line treatment after CGP testing are not documented in the data, a prior study with a similar observation suggested that this may have been due to discrepancies in notifying physicians about test results.8

It is also unclear why no significant differences in inpatient care utilization were observed between the optimal and suboptimal groups. Only procedure codes associated with inpatient care were available, not the diagnosis-related groups into which they are customarily aggregated for billing purposes.30 This omission may have reduced the fidelity of the inpatient care utilization data.

Finally, although the study considered a national sample, the patients were not distributed uniformly across the country. The patients who were included in this study may not have been perfectly representative of American patients with NSCLC, because all included patients had access to CGP and received it. Likewise, all the patients included received a specific CGP test; a different CGP test may have yielded different findings.

Conclusions

In the population examined, when patients with ALK rearrangement or EGFR mutation were provided with optimal NSCLC treatment, they had significantly longer rwTTNT and rwTTD, lower ED and outpatient utilization in the first 12 months, and fewer adverse events. These findings suggest that treatment concordant with CGP findings is associated with better health outcomes. These findings provide further support for the ASCO and NCCN recommendations that treatment of NSCLC be tailored to a patient’s biomarkers as ascertained through CGP.3,4

References

  • 1.

    Johnson DB, Rioth MJ, Horn L. Immune checkpoint inhibitors in NSCLC. Curr Treat Options Oncol 2014;15:658669.

  • 2.

    Singh AP, Shum E, Rajdev L, et al. Impact and diagnostic gaps of comprehensive genomic profiling in real-world clinical practice. Cancers (Basel) 2020;12:1156.

  • 3.

    Ettinger DS, Wood DE, Aisner DL, et al. NCCN Clinical Practice Guidelines in Oncology: Non-Small Cell Lung Cancer. Version 3.2023. Accessed May 1, 2023. To view the most recent version, visit https://www.nccn.org

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 4.

    Leighl NB, Rekhtman N, Biermann WA, et al. Molecular testing for selection of patients with lung cancer for epidermal growth factor receptor and anaplastic lymphoma kinase tyrosine kinase inhibitors: American Society of Clinical Oncology endorsement of the College of American Pathologists/International Association for the Study of Lung Cancer/Association for Molecular Pathology guideline. J Clin Oncol 2014;32:36733679.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 5.

    Bodor JN, Boumber Y, Borghaei H. Biomarkers for immune checkpoint inhibition in non-small cell lung cancer (NSCLC). Cancer 2020;126:260270.

  • 6.

    Viteri S, Minchom A, Bazhenova L, et al. Frequency, underdiagnosis, and heterogeneity of epidermal growth factor receptor exon 20 insertion mutations using real-world genomic datasets. Mol Oncol 2023;17:230237.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 7.

    Sadik H, Pritchard D, Keeling DM, et al. Impact of clinical practice gaps on the implementation of personalized medicine in advanced non-small-cell lung cancer. JCO Precis Oncol 2022;6:e2200246.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 8.

    Apple J, Shenolikar R, De Silva K, et al. Real-world outcomes among patients with EGFR-mutated non-small cell lung cancer treated with EGFR tyrosine kinase inhibitors versus immunotherapy or chemotherapy in the first-line setting. Cancer Med 2023;12:1341513425.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 9.

    Deverka PA, Douglas MP, Phillips KA. Use of real-world evidence in US payer coverage decision-making for next-generation sequencing–based tests: challenges, opportunities, and potential solutions. Value Health 2020;23:540550.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 10

    Pennell NA, Mutebi A, Zhou ZY, et al. Economic impact of next-generation sequencing versus single-gene testing to detect genomic alterations in metastatic non-small-cell lung cancer using a decision analytic model. JCO Precis Oncol. Published online May 16, 2019. doi:10.1200/PO.18.00356

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 11.

    Bittoni MA, Arunachalam A, Li H, et al. Real-world treatment patterns, overall survival, and occurrence and costs of adverse events associated with first-line therapies for Medicare patients 65 years and older with advanced non–small-cell lung cancer: a retrospective study. Clin Lung Cancer 2018;19:e629645.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 12.

    Mazieres J, Drilon A, Lusque A, et al. Immune checkpoint inhibitors for patients with advanced lung cancer and oncogenic driver alterations: results from the IMMUNOTARGET registry. Ann Oncol 2019;30:13211328.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 13.

    Brenner DR, O’Sullivan DE, Jarada TN, et al. The impact of population-based EGFR testing in non-squamous metastatic non-small cell lung cancer in Alberta, Canada. Lung Cancer 2023;175:6067.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 14.

    Chawla A, Janku F, Wheler JJ, et al. Estimated cost of anticancer therapy directed by comprehensive genomic profiling in a single-center study. JCO Precis Oncol 2018;2:PO.18.00074.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 15.

    U.S. Food and Drug Administration. FDA expands pembrolizumab indication for first-line treatment of NSCLC (TPS ≥1%). Accessed November 1, 2022. Available at: https://www.fda.gov/drugs/fda-expands-pembrolizumab-indication-first-line-treatment-nsclc-tps-1

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 16.

    Elixhauser A, Steiner C, Harris DR, et al. Comorbidity measures for use with administrative data. Med Care 1998;36:827.

  • 17.

    U.S. Census Bureau. 2010 census urban and rural classification and urban area criteria. Accessed August 8, 2022. Available at: https://www.census.gov/programs-surveys/geography/guidance/geo-areas/urban-rural/2010-urban-rural.html

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 18.

    Centers for Disease Control and Prevention. BRFSS: table of chronic health indicators. Accessed August 8, 2022. Available at: https://chronicdata.cdc.gov/Behavioral-Risk-Factors/BRFSS-Table-of-Chronic-Health-Indicators/u7k3-tu8b

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 19.

    U.S. Census Bureau. S1501: educational attainment. Accessed August 8, 2022. Available at: https://data.census.gov/cedsci/table?q=S1501&tid=ACSST1Y2019.S1501

  • 20.

    U.S. Census Bureau. S1901: income in the past 12 months (in 2019 inflation-adjusted dollars). Accessed August 8, 2022. Available at: https://data.census.gov/cedsci/table?q=S1501&tid=ACSST1Y2019.S1901

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 21.

    U.S. Census Bureau. Decennial census P.L. 94-171 redistricting data. Accessed January 30, 2022. Available at: https://www.census.gov/programs-surveys/decennial-census/about/rdo/summary-files.html

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 22.

    American Cancer Society. Cancer Statistics Center: incidence rates, 2015–2019. Accessed August 8, 2022. Available at: https://cancerstatisticscenter.cancer.org/#!/data-analysis/IncRate

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 23

    Yau KK, Wang K, Lee AH. Zero‐inflated negative binomial mixed regression modeling of over‐dispersed count data with extra zeros. Biom J 2003;45:437452.

  • 24.

    Cui W, Milner-Watts C, O’Sullivan H, et al. Up-front cell-free DNA next generation sequencing improves target identification in UK first line advanced non-small cell lung cancer (NSCLC) patients. Eur J Cancer 2022;171:4454.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 25.

    Signorovitch J, Zhou Z, Ryan J, et al. Budget impact analysis of comprehensive genomic profiling in patients with advanced non-small cell lung cancer. J Med Econ 2019;22:140150.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 26.

    Harvey MJ, Cunningham R, Sawchyn B, et al. Budget impact analysis of comprehensive genomic profiling in patients with advanced non-small-cell lung cancer. JCO Precis Oncol 2021;5:16111624.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 27.

    Mateo J, Steuten L, Aftimos P, et al. Delivering precision oncology to patients with cancer. Nat Med 2022;28:658665.

  • 28.

    Muluk S, Sabik L, Chen Q, et al. Disparities in geographic access to medical oncologists. Health Serv Res 2022;57:10351044.

  • 29.

    Bhatia S, Landier W, Paskett ED, et al. Rural–urban disparities in cancer outcomes: opportunities for future research. J Natl Cancer Inst 2022;114:940952.

  • 30.

    Centers for Medicare & Medicaid Services. MS-DRG classifications and software. Accessed August 18, 2022. Available at: https://www.cms.gov/Medicare/Medicare-Fee-for-Service-Payment/AcuteInpatientPPS/MS-DRG-Classifications-and-Software

    • PubMed
    • Search Google Scholar
    • Export Citation

Submitted March 9, 2023; final revision received July 21, 2023; accepted for publication August 17, 2023. Published online January 8, 2024.

Author contributions: Conceptualization: Powell, Yay Donderici, Zhang, Wiedower, McNeal, Hiatt. Data curation: Yay Donderici, Zhang. Formal analysis: Yay Donderici, Zhang. Funding acquisition: McNeal. Investigation: Powell. Methodology: All authors. Project administration: McNeal. Resources: Powell. Supervision: Zhang, McNeal, Hiatt. Validation: All authors. Visualization: Yay Donderici, Zhang. Writing—original draft: Powell. Writing—review and editing: Yay Donderici, Zhang, Forbes, Wiedower, McNeal, Hiatt.

Disclosures: Dr. Powell has disclosed being employed by and owning Payer+Provider Syndicate. All remaining authors have disclosed being employed by and owning stock in Guardant Health.

Correspondence: Adam C. Powell, PhD, Payer+Provider Syndicate, 20 Oakland Avenue, Newton, MA 02466. Email: powell@payerprovider.com

Supplementary Materials

  • Collapse
  • Expand
  • Figure 1.

    CONSORT diagram.

    Abbreviation: NSCLC, non–small cell lung cancer.

  • Figure 2.

    Kaplan-Meier plots of adjusted (A) real-world time to next treatment and (B) real-world time to discontinuation.

    Abbreviation: NR, not reached.

  • 1.

    Johnson DB, Rioth MJ, Horn L. Immune checkpoint inhibitors in NSCLC. Curr Treat Options Oncol 2014;15:658669.

  • 2.

    Singh AP, Shum E, Rajdev L, et al. Impact and diagnostic gaps of comprehensive genomic profiling in real-world clinical practice. Cancers (Basel) 2020;12:1156.

  • 3.

    Ettinger DS, Wood DE, Aisner DL, et al. NCCN Clinical Practice Guidelines in Oncology: Non-Small Cell Lung Cancer. Version 3.2023. Accessed May 1, 2023. To view the most recent version, visit https://www.nccn.org

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 4.

    Leighl NB, Rekhtman N, Biermann WA, et al. Molecular testing for selection of patients with lung cancer for epidermal growth factor receptor and anaplastic lymphoma kinase tyrosine kinase inhibitors: American Society of Clinical Oncology endorsement of the College of American Pathologists/International Association for the Study of Lung Cancer/Association for Molecular Pathology guideline. J Clin Oncol 2014;32:36733679.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 5.

    Bodor JN, Boumber Y, Borghaei H. Biomarkers for immune checkpoint inhibition in non-small cell lung cancer (NSCLC). Cancer 2020;126:260270.

  • 6.

    Viteri S, Minchom A, Bazhenova L, et al. Frequency, underdiagnosis, and heterogeneity of epidermal growth factor receptor exon 20 insertion mutations using real-world genomic datasets. Mol Oncol 2023;17:230237.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 7.

    Sadik H, Pritchard D, Keeling DM, et al. Impact of clinical practice gaps on the implementation of personalized medicine in advanced non-small-cell lung cancer. JCO Precis Oncol 2022;6:e2200246.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 8.

    Apple J, Shenolikar R, De Silva K, et al. Real-world outcomes among patients with EGFR-mutated non-small cell lung cancer treated with EGFR tyrosine kinase inhibitors versus immunotherapy or chemotherapy in the first-line setting. Cancer Med 2023;12:1341513425.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 9.

    Deverka PA, Douglas MP, Phillips KA. Use of real-world evidence in US payer coverage decision-making for next-generation sequencing–based tests: challenges, opportunities, and potential solutions. Value Health 2020;23:540550.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 10

    Pennell NA, Mutebi A, Zhou ZY, et al. Economic impact of next-generation sequencing versus single-gene testing to detect genomic alterations in metastatic non-small-cell lung cancer using a decision analytic model. JCO Precis Oncol. Published online May 16, 2019. doi:10.1200/PO.18.00356

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 11.

    Bittoni MA, Arunachalam A, Li H, et al. Real-world treatment patterns, overall survival, and occurrence and costs of adverse events associated with first-line therapies for Medicare patients 65 years and older with advanced non–small-cell lung cancer: a retrospective study. Clin Lung Cancer 2018;19:e629645.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 12.

    Mazieres J, Drilon A, Lusque A, et al. Immune checkpoint inhibitors for patients with advanced lung cancer and oncogenic driver alterations: results from the IMMUNOTARGET registry. Ann Oncol 2019;30:13211328.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 13.

    Brenner DR, O’Sullivan DE, Jarada TN, et al. The impact of population-based EGFR testing in non-squamous metastatic non-small cell lung cancer in Alberta, Canada. Lung Cancer 2023;175:6067.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 14.

    Chawla A, Janku F, Wheler JJ, et al. Estimated cost of anticancer therapy directed by comprehensive genomic profiling in a single-center study. JCO Precis Oncol 2018;2:PO.18.00074.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 15.

    U.S. Food and Drug Administration. FDA expands pembrolizumab indication for first-line treatment of NSCLC (TPS ≥1%). Accessed November 1, 2022. Available at: https://www.fda.gov/drugs/fda-expands-pembrolizumab-indication-first-line-treatment-nsclc-tps-1

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 16.

    Elixhauser A, Steiner C, Harris DR, et al. Comorbidity measures for use with administrative data. Med Care 1998;36:827.

  • 17.

    U.S. Census Bureau. 2010 census urban and rural classification and urban area criteria. Accessed August 8, 2022. Available at: https://www.census.gov/programs-surveys/geography/guidance/geo-areas/urban-rural/2010-urban-rural.html

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 18.

    Centers for Disease Control and Prevention. BRFSS: table of chronic health indicators. Accessed August 8, 2022. Available at: https://chronicdata.cdc.gov/Behavioral-Risk-Factors/BRFSS-Table-of-Chronic-Health-Indicators/u7k3-tu8b

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 19.

    U.S. Census Bureau. S1501: educational attainment. Accessed August 8, 2022. Available at: https://data.census.gov/cedsci/table?q=S1501&tid=ACSST1Y2019.S1501

  • 20.

    U.S. Census Bureau. S1901: income in the past 12 months (in 2019 inflation-adjusted dollars). Accessed August 8, 2022. Available at: https://data.census.gov/cedsci/table?q=S1501&tid=ACSST1Y2019.S1901

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 21.

    U.S. Census Bureau. Decennial census P.L. 94-171 redistricting data. Accessed January 30, 2022. Available at: https://www.census.gov/programs-surveys/decennial-census/about/rdo/summary-files.html

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 22.

    American Cancer Society. Cancer Statistics Center: incidence rates, 2015–2019. Accessed August 8, 2022. Available at: https://cancerstatisticscenter.cancer.org/#!/data-analysis/IncRate

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 23

    Yau KK, Wang K, Lee AH. Zero‐inflated negative binomial mixed regression modeling of over‐dispersed count data with extra zeros. Biom J 2003;45:437452.

  • 24.

    Cui W, Milner-Watts C, O’Sullivan H, et al. Up-front cell-free DNA next generation sequencing improves target identification in UK first line advanced non-small cell lung cancer (NSCLC) patients. Eur J Cancer 2022;171:4454.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 25.

    Signorovitch J, Zhou Z, Ryan J, et al. Budget impact analysis of comprehensive genomic profiling in patients with advanced non-small cell lung cancer. J Med Econ 2019;22:140150.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 26.

    Harvey MJ, Cunningham R, Sawchyn B, et al. Budget impact analysis of comprehensive genomic profiling in patients with advanced non-small-cell lung cancer. JCO Precis Oncol 2021;5:16111624.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 27.

    Mateo J, Steuten L, Aftimos P, et al. Delivering precision oncology to patients with cancer. Nat Med 2022;28:658665.

  • 28.

    Muluk S, Sabik L, Chen Q, et al. Disparities in geographic access to medical oncologists. Health Serv Res 2022;57:10351044.

  • 29.

    Bhatia S, Landier W, Paskett ED, et al. Rural–urban disparities in cancer outcomes: opportunities for future research. J Natl Cancer Inst 2022;114:940952.

  • 30.

    Centers for Medicare & Medicaid Services. MS-DRG classifications and software. Accessed August 18, 2022. Available at: https://www.cms.gov/Medicare/Medicare-Fee-for-Service-Payment/AcuteInpatientPPS/MS-DRG-Classifications-and-Software

    • PubMed
    • Search Google Scholar
    • Export Citation
All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 3229 3229 2355
PDF Downloads 1826 1826 1255
EPUB Downloads 0 0 0