Abstract
Introduction
Precision medicine has resulted in improved outcomes for non small cell lung cancer (NSCLC), whereas biomarker testing is considered critical for guiding treatment decisions for advanced-stage NSCLC, and adoption of testing in routine practice is variable. We studied the utilization of biomarker testing in advanced NSCLC.
Methods
The American Society of Clinical Oncology (ASCO) CancerLinQ Discovery data set was queried to identify patients diagnosed with lung cancer between 2010 and 2018. Data on demographics, tumor stage, histology, and treatments were extracted, and receipt of biomarker testing was investigated as the primary outcome. Univariate association of each clinicopathological variable with the biomarker testing outcome was performed using a chi-square test for categorical variables and an analysis of variance test for numerical variables. A multivariable logistic regression analysis with backward selection at an alpha of 0.05 was reported. All analyses were conducted using SAS 9.4.
Results
A total of 37,925 patients with stage IV NSCLC were analyzed. The patients had a median age of 65 years; meanwhile, 51% of the participants were male individuals, 68% were white, and 33.5% had adenocarcinoma. Approximately 22% of all patients with NSCLC had biomarker testing results. Among the patients with adenocarcinoma, 49% had biomarker testing results available. In the stage IV group, 47% were treated with chemotherapy, 16% with immunotherapy, and 3.5% with targeted therapy. On multivariable analysis, female patients were more likely to have molecular testing compared with male patients (OR = 1.29, 95% confidence interval [CI]: 1.21–1.36, p < 0.001). Compared with white patients, black patients were less likely to have biomarker testing (OR = 0.89, 95% CI: 0.81–0.97, p = 0.009), and Asians were more likely to undergo testing (OR = 2.21, 95% CI: 1.78–2.73, p < 0.001). Hispanic patients were more likely to undergo biomarker testing than non-Hispanic (OR = 1.23, 95% CI: 1.01–1.50, p = 0.03). In addition, treatment with immunotherapy (OR = 1.86, 95% CI: 1.72–2.02, p < 0.001) and targeted therapy (OR = 2.52, 95% CI: 2.21–2.88, p < 0.001) were associated with a significantly higher likelihood of having biomarker testing. These results were also confirmed in a subgroup analysis of patients with adenocarcinoma.
Conclusion
In this analysis of a United States–based real-world data set of patients with stage IV NSCLC, the Asian race and female sex were associated with a higher likelihood of having biomarker testing performed. The overall percentage of patients undergoing testing remained suboptimal.
Keywords: Biomarker testing, NSCLC, Mutation, Targeted therapy
Introduction
An estimated 654,620 people living in the United States have a history of lung cancer, and an additional 238,340 are expected to be diagnosed in 2023.1,2 The five-year survival rates are low for advanced or metastatic NSCLC.3 In recent years, the mortality rate for lung cancer in the United States has decreased by approximately 4% per year for male individuals and 2% per year for female individuals.4 This corresponds to delays in death for approximately 6800 men and 3200 women in the United States between 2014 and 2016. The reduction in mortality is due to multiple factors, including a decline in smoking and the availability of improved treatment options for lung cancer. The discovery of targetable mutations and personalized medicine have dramatically changed the treatment for NSCLC.
Patients with advanced-stage metastatic NSCLC were previously treated with platinum-based systemic therapies5 with median overall survival (OS) being 10 to 12 months and 5-year OS estimated to be 2%.6,7 In the past two decades, the treatment of advanced NSCLC has evolved from cytotoxic therapies to precision medicine. Various subsets of patients are treated with targeted therapies or immunotherapy based on the molecular alteration in the tumor or the expression level of programmed death ligand-1.8 In the Lung Cancer Mutation Consortium study, patients with lung adenocarcinoma treated with targeted therapy on the basis of their oncogenic driver mutations had significantly improved survival.9 Current National Comprehensive Cancer Network guidelines include recommendations for biomarker testing and that patients with advanced NSCLC with EGFR mutation or ALK rearrangement be treated with EGFR or ALK inhibitors, respectively.3,10 In addition, Food and Drug Administration (FDA)–approved and guideline-recommended therapies are available for patients with alterations in RET, ROS1, BRAF V600E, MET exon 14, HER-2, KRAS G12C, NTRK, and EGFR exon 20. Results from numerous studies underscore the importance of biomarker testing before first-line treatment of advanced NSCLC.11 In 2015, nivolumab, an immune checkpoint inhibitor, was approved by the FDA for the treatment of patients who had disease progression on or after platinum-based therapy, marking a critical advancement in the management of advanced-stage NSCLC.12 Other immune checkpoint inhibitors, including pembrolizumab13 and atezolimumab,14 have also reported significant improvement in the OS of patients with advanced NSCLC and now have approval in the first-line setting.
Routine biomarker testing represents a critical component in the management of advanced NSCLC, paving the way for the individualized treatment of specific subgroups of patients. Nevertheless, utilization of biomarker testing continues to be suboptimal; in this work, we studied the utilization of biomarker testing for advanced NSCLC in the United States.
Materials and Methods
Data Source
CancerLinQ, developed in 2014 as a subsidiary of American Society of Clinical Oncology (ASCO), is a nonprofit, big-data, health technology platform in cancer. It has followed an electronic health record agnostic approach to provide a harmonized data repository with structured and unstructured data captured from participating oncology organizations and practices. By March of 2020, it included data from 63 organizations across the United States that utilize nine different electronic health records.15 The CancerLinQ Discovery was created as part of an initiative to make de-identified, real-world data publicly available for research16 and exempted from institutional review board review and approval.
Study Population
The CancerLinQ Discovery data set was queried to identify patients diagnosed with lung cancer between 2010 and 2018. De-identified patient information was abstracted from the data files for this work.
Study Variables
Patient information related to demographic data, including age, sex, race, year of diagnosis, and geographic location, were extracted. Data on tumor stage, histology, and treatments were also mined for this work. We queried the patient IDs in one of the tables having “Smoking cessation advice” and then cross-referenced it against the unique patient IDs in the overall cohort, adding an indicator for whether they had the advice or not. This method allowed for identifying all individuals that had received at least some level of smoking cessation advice in their history. This analysis was limited to patients with stage IV NSCLC as biomarker testing is considered the standard of care for this subset. The treatment variables included in this analysis included targeted therapy, chemotherapy, and immunotherapy. The drugs were searched in the database using their generic and brand names.
Study Outcomes
Receipt of biomarker testing was investigated as the primary outcome. Biomarker testing, which also included programmed death ligand-1 testing for patients, was determined using the laboratory code labels that included individual gene-level tests, fusion/rearrangement testing, and large-scale genomic testing. The code labels that indicated a genomic/genetic result for patients were identified and included in the analysis. A positive result in this analysis was identified on the basis of the corresponding labtestvalue_codelabel field values for the returned results. These labtestvalue_codelabels were a list of values that were manually curated from all possible codelabels in the data set and confirmed to be equivalent to a “positive” result for the given test type.
Statistical Analysis
Statistical analysis was conducted using SAS version 9.4. Descriptive statistics for the stage IV group were reported. Univariate association of each clinicopathological variable with the biomarker testing outcome was performed using a chi-square test for categorical variables and analysis of variance test for numerical variables. A multivariable logistic regression analysis with backward elimination at an alpha of 0.05 was reported. To inform about causal inference, a few selected variables—sex, race, received immunotherapy, received targeted therapy, received chemotherapy, had surgery (for the full cohort), and the age for the adenocarcinoma cohort—were fit in separate multivariable models. A similar analysis was performed on the subgroup of patients with adenocarcinoma. Statistical significance was assessed at the 0.05 level.
Results
Patient characteristics
A total of 37,917 patients with stage IV NSCLC were identified and analyzed in this work. Patient characteristics are outlined in Table 1. The median age was 65 years, with 51% of the patients being male. The white race accounted for 68% of the group. Approximately 22% of all the patients with NSCLC had biomarker testing results available in this data set. About 33.5% of the patients had the recorded histologic subtype of adenocarcinoma. In patients with adenocarcinoma, 49% had biomarker testing results available. The characteristics of this subgroup of patients are provided in Table 2.
Table 1.
Patient Characteristics
| Variable | Level | n (%) (N = 37,917) |
|---|---|---|
| Biomarker testing | No | 29,560 (78.0) |
| Yes | 8357 (22.0) | |
| Positive molecular test result | No | 5006 (59.9) |
| Yes | 3351 (40.1) | |
| Missing | 29,560 | |
| Sex | Male | 19,336 (51.0) |
| Female | 18,578 (49.0) | |
| Missing | 3 | |
| Race | White | 25,707 (67.8) |
| Black or African American | 5618 (14.8) | |
| Asian | 572 (1.5) | |
| American Indian or Alaska Native | 166 (0.4) | |
| Native Hawaiian or Other Pacific | 47 (0.1) | |
| Other race | 2323 (6.1) | |
| Unknown | 3484 (9.2) | |
| Ethnicity | Hispanic or Latino | 869 (2.3) |
| Not Hispanic or Latino | 23,846 (62.9) | |
| Unknown | 13,202 (34.8) | |
| Region | South region | 16,715 (44.1) |
| Midwest region | 11,018 (29.1) | |
| Northeast region | 3334 (8.8) | |
| West region | 3773 (10.0) | |
| Unknown | 3077 (8.1) | |
| Cancer stage | Stage IV | 37,917 (100.0) |
| Received immunotherapy? | No | 31,949 (84.3) |
| Yes | 5968 (15.7) | |
| Received targeted therapy? | No | 36,599 (96.5) |
| Yes | 1318 (3.5) | |
| Received chemotherapy? | No | 20,226 (53.3) |
| Yes | 17,691 (46.7) | |
| Had surgery that is related to or targeted the lung? | No | 34,445 (90.8) |
| Yes | 3472 (9.2) | |
| Receive smoking cessation advice? | No | 31,359 (82.7) |
| Yes | 6558 (17.3) | |
| Histology | Adenocarcinoma | 12,706 (33.5) |
| Squamous cell carcinoma | 3740 (9.9) | |
| NOS | 21,471 (56.6) | |
| Age at diagnosis (y) | Median | 65.00 |
| Minimum | 18.00 | |
| Maximum | 102.00 |
NOS, not otherwise specified.
Table 2.
Patient Characteristics: Stage IV Adenocarcinoma
| Variable | Level | n (%) (N = 12,706) |
|---|---|---|
| Biomarker testing | No | 6467 (50.9) |
| Yes | 6239 (49.1) | |
| Positive molecular test result | No | 3528 (56.5) |
| Yes | 2711 (43.5) | |
| Missing | 6467 | |
| Sex | Male | 6212 (48.9) |
| Female | 6491 (51.1) | |
| Missing | 3 | |
| Race | White | 8208 (64.6) |
| Black or African American | 1944 (15.3) | |
| Asian | 308 (2.4) | |
| American Indian or Alaska Native | 52 (0.4) | |
| Native Hawaiian or Other Pacific | 18 (0.1) | |
| Other race | 836 (6.6) | |
| Unknown | 1340 (10.5) | |
| Ethnicity | Hispanic or Latino | 310 (2.4) |
| Not Hispanic or Latino | 8282 (65.2) | |
| Unknown | 4114 (32.4) | |
| Region | South region | 5326 (41.9) |
| Midwest region | 3983 (31.3) | |
| Northeast region | 851 (6.7) | |
| West region | 1499 (11.8) | |
| Unknown | 1047 (8.2) | |
| Received immunotherapy? | No | 10,857 (85.4) |
| Yes | 1849 (14.6) | |
| Received targeted therapy? | No | 11,881 (93.5) |
| Yes | 825 (6.5) | |
| Received chemotherapy? | No | 6651 (52.3) |
| Yes | 6055 (47.7) | |
| Had surgery that is related to or targeted the lung? | No | 10,587 (83.3) |
| Yes | 2119 (16.7) | |
| Receive smoking cessation advice? | No | 10,477 (82.5) |
| Yes | 2229 (17.5) | |
| Age at diagnosis | Median | 65.00 |
| Minimum | 21.00 | |
| Maximum | 96.00 |
Biomarker Testing Patterns
The univariate association of each clinicopathologic variable with the biomarker testing outcome is shown in Table 3, which shows that 24% of all female patients were tested compared with 20% of males, and 19% of black patients were tested compared with 22% of white and 46% of Asian patients. These results were confirmed in multivariable analysis (Table 4), demonstrating that female patients were more likely to have biomarker testing compared with males (OR = 1.29, 95% confidence interval [CI]: 1.21–1.36, p < 0.001). Compared with white patients, black patients were less likely to have biomarker testing (OR = 0.89, 95% CI: 0.81–0.97, p = 0.009), and Asians were more likely to undergo testing (OR = 2.21, 95% CI: 1.78–2.73, p < 0.001). Hispanic patients were more likely to undergo biomarker testing than non-Hispanics (OR =1.23, 95% CI: 1.01–1.50, p = 0.04). These results were also consistent with the subgroup analysis of patients with adenocarcinoma. The univariate analysis of patients with stage IV adenocarcinoma found that 52% of female patients were tested compared with 46% of male patients, and 44% of Black patients were tested compared with 49% of whites and 74% of Asian patients. These results for patients with adenocarcinoma were further confirmed with multivariable analyses (Table 5). Female patients with stage IV adenocarcinoma were more likely to have a biomarker test performed (OR = 1.31, 95% CI: 1.22–1.42, p < 0.001), and black patients had a lower likelihood of a biomarker test performed (OR = 0.82, 95% CI: 0.73–0.91, p < 0.001). On the basis of multivariate analysis, patients from the Midwest and West Regions of the United States were found to be less likely to undergo biomarker testing compared with those from the South Region. Nevertheless, patients from the Northeast Region are more likely to be tested than those from the South Region (OR = 1.42, 95% CI: 1.27–1.59, p < 0.001; Table 4). The multivariable analysis of selected variables to understand causal inference for all patients with stage IV NSCLC and those with stage IV adenocarcinoma were consistent with these results.
Table 3.
Univariate Association with Biomarker Testing
| Variable | Statistics | Level | Biomarker Testing |
Parametric p Valuea | |
|---|---|---|---|---|---|
| No (n = 29,560) | Yes (n = 8357) | ||||
| Sex | n (row %) | Male | 15,439 (79.85) | 3897 (20.15) | <0.001b |
| n (row %) | Female | 14,118 (75.99) | 4460 (24.01) | ||
| Race | n (row %) | White | 20,142 (78.35) | 5565 (21.65) | <0.001b |
| n (row %) | Black or African American | 4525 (80.54) | 1093 (19.46) | ||
| n (row %) | Asian | 307 (53.67) | 265 (46.33) | ||
| n (row %) | American Indian or Alaska Native | 117 (70.48) | 49 (29.52) | ||
| n (row %) | Native Hawaiian or other Pacific | 34 (72.34) | 13 (27.66) | ||
| n (row %) | Other Race | 1712 (73.7) | 611 (26.3) | ||
| n (row %) | Unknown | 2723 (78.16) | 761 (21.84) | ||
| Ethnicity | n (row %) | Hispanic or Latino | 619 (71.23) | 250 (28.77) | <0.001b |
| n (row %) | Not Hispanic or Latino | 17,687 (74.17) | 6159 (25.83) | ||
| n (row %) | Unknown | 11,254 (85.24) | 1948 (14.76) | ||
| Region | n (row %) | South region | 13,076 (78.23) | 3639 (21.77) | <0.001b |
| n (row %) | Midwest region | 8781 (79.7) | 2237 (20.3) | ||
| n (row %) | Northeast region | 2454 (73.61) | 880 (26.39) | ||
| n (row %) | West region | 3002 (79.57) | 771 (20.43) | ||
| N (Row %) | Unknown | 2247 (73.03) | 830 (26.97) | ||
| Received immunotherapy? | n (row %) | No | 25,408 (79.53) | 6541 (20.47) | <0.001b |
| n (row %) | Yes | 4152 (69.57) | 1816 (30.43) | ||
| Received targeted therapy? | n (row %) | No | 28,926 (79.03) | 7673 (20.97) | <0.001b |
| n (row %) | Yes | 634 (48.1) | 684 (51.9) | ||
| Received chemotherapy? | n (row %) | No | 16,627 (82.21) | 3599 (17.79) | <0.001b |
| n (row %) | Yes | 12,933 (73.1) | 4758 (26.9) | ||
| Had surgery that is related to or targeted the lung? | n (row %) | No | 27,725 (80.49) | 6720 (19.51) | <0.001b |
| n (row %) | Yes | 1835 (52.85) | 1637 (47.15) | ||
| Receive smoking cessation advice? | n (row %) | No | 25,028 (79.81) | 6331 (20.19) | <0.001b |
| n (row %) | Yes | 4532 (69.11) | 2026 (30.89) | ||
| Histology | n (row %) | Adenocarcinoma | 6467 (50.9) | 6239 (49.1) | <0.001b |
| n (row %) | Squamous cell carcinoma | 3199 (85.53) | 541 (14.47) | ||
| n (row %) | NOS | 19,894 (92.66) | 1577 (7.34) | ||
| Age at diagnosis | N | 29,559 | 8357 | 0.343 | |
| Mean | 64.84 | 64.7 | |||
| Median | 65 | 65 | |||
| Min | 18 | 19 | |||
| Max | 102 | 95 | |||
| SD | 11.54 | 11.64 | |||
ANOVA, analysis of variance; NOS, not otherwise specified.
The parametric p value is calculated by ANOVA for numerical covariates and chi-square test for categorical covariates.
Statistically significant.
Table 4.
Multivariable Analysis of Patients of Biomarker Testing
| Covariate | Level | N | Biomarker Testing = Yes |
||
|---|---|---|---|---|---|
| OR (95% CI) | OR p Value | Type 3 p Value | |||
| Sex | Female | 18,578 | 1.29 (1.21–1.36) | <0.001 | <0.001a |
| Male | 19,336 | - | - | ||
| Race | Black or African American | 5618 | 0.89 (0.81–0.97) | 0.009 | <0.001a |
| Asian | 572 | 2.21 (1.78–2.73) | <0.001 | ||
| American Indian or Alaska Native | 166 | 1.66 (1.09–2.52) | 0.018 | ||
| Native Hawaiian or other Pacific | 47 | 1.56 (0.69–3.55) | 0.285 | ||
| Other Race | 2323 | 1.41 (1.24–1.60) | <0.001 | ||
| Unknown | 3483 | 1.14 (1.02–1.27) | 0.017 | ||
| White | 25,705 | - | - | ||
| Ethnicity | Hispanic or Latino | 869 | 1.23 (1.01–1.50) | 0.039 | <0.001a |
| Unknown | 13,200 | 0.53 (0.49–0.57) | <0.001 | ||
| Not Hispanic or Latino | 23,845 | - | - | ||
| Region | Midwest region | 11,018 | 0.79 (0.73–0.85) | <0.001 | <0.001a |
| Northeast region | 3334 | 1.42 (1.27–1.59) | <0.001 | ||
| West region | 3772 | 0.85 (0.76–0.94) | 0.002 | ||
| Unknown | 3075 | 1.38 (1.23–1.55) | <0.001 | ||
| South region | 16,715 | - | - | ||
| Received immunotherapy? | Yes | 5968 | 1.86 (1.72–2.02) | <0.001 | <0.001a |
| No | 31,946 | - | - | ||
| Received targeted therapy? | Yes | 1318 | 2.52 (2.21–2.88) | <0.001 | <0.001a |
| No | 36,596 | - | - | ||
| Received chemotherapy? | Yes | 17,690 | 1.39 (1.31–1.48) | <0.001 | <0.001a |
| No | 20,224 | - | - | ||
| Had surgery that is related to or targeted the lung? | Yes | 3471 | 2.71 (2.48–2.96) | <0.001 | <0.001a |
| No | 34,443 | - | - | ||
| Receive smoking cessation advice? | Yes | 6558 | 1.85 (1.71–2.00) | <0.001 | <0.001a |
| No | 31,356 | - | - | ||
| Histology | Other/unknown | 21,471 | 0.10 (0.09–0.10) | <0.001 | <0.001a |
| Squamous cell carcinoma | 3740 | 0.15 (0.14–0.17) | <0.001 | ||
| Adenocarcinoma | 12,703 | - | - | ||
Note: The number of observations in the original data set was 37,917. The number of observations used was 37,914. Backward selection with an alpha level of removal of 0.05 was used. No variables were removed from the model.
CI, confidence interval.
Statistically significant.
Table 5.
Stage IV Adenocarcinoma Patients: Multivariable Analysis of Biomarker Testing
| Molecular Testing = Yes |
|||||
|---|---|---|---|---|---|
| Covariate | Level | N | OR (95% CI) | OR p Value | Type 3 p Value |
| Sex | Female | 6491 | 1.31 (1.22-1.42) | <0.001 | <0.001a |
| Male | 6212 | - | - | ||
| Race | Black or African American | 1944 | 0.82 (0.73-0.91) | <0.001 | <0.001a |
| Asian | 308 | 2.32 (1.75-3.09) | <0.001 | ||
| American Indian or Alaska Native | 52 | 1.14 (0.63-2.05) | 0.674 | ||
| Native Hawaiian or Other Pacific | 18 | 2.29 (0.80-6.60) | 0.124 | ||
| Other race | 836 | 1.78 (1.51-2.11) | <0.001 | ||
| Unknown | 1339 | 1.06 (0.92-1.21) | 0.422 | ||
| White | 8206 | - | - | ||
| Ethnicity | Hispanic or Latino | 310 | 1.20 (0.92-1.57) | 0.176 | <0.001a |
| Unknown | 4112 | 0.45 (0.41-0.50) | <0.001 | ||
| Not Hispanic or Latino | 8281 | - | - | ||
| Region | Midwest region | 3983 | 0.79 (0.72-0.88) | <0.001 | <0.001a |
| Northeast region | 851 | 0.79 (0.67-0.93) | 0.005 | ||
| West region | 1498 | 0.60 (0.52-0.69) | <0.001 | ||
| Unknown | 1045 | 1.78 (1.53-2.09) | <0.001 | ||
| South region | 5326 | - | - | ||
| Received immunotherapy? | Yes | 1849 | 1.92 (1.71-2.16) | <0.001 | <0.001a |
| No | 10,854 | - | - | ||
| Received targeted therapy? | Yes | 825 | 2.69 (2.28-3.18) | <0.001 | <0.001a |
| No | 11,878 | - | - | ||
| Receive chemotherapy? | Yes | 6054 | 1.68 (1.55-1.82) | <0.001 | <0.001a |
| No | 6649 | - | - | ||
| Had surgery that is related to or targeted the lung? | Yes | 2118 | 2.05 (1.84-2.29) | <0.001 | <0.001a |
| No | 10,585 | - | - | ||
| Receive smoking cessation advice? | Yes | 2229 | 1.64 (1.48-1.83) | <0.001 | <0.001a |
| No | 10,474 | - | - | ||
| Age at diagnosis | 12,703 | 0.99 (0.98-0.99) | <0.001 | <0.001a | |
Note: The number of observations in the original data set was 12,706. The number of observations used was 12,703. Backward selection with an alpha level of removal of .05 was used. No variables were removed from the model.
CI, confidence interval.
Statistically significant.
Treatment Patterns
In the stage IV group, 47% were treated with chemotherapy, 16% with immunotherapy, and 3.5% with targeted therapy (Table 1). The targeted therapy drugs included in the analysis are erlotinib, gefitinib, afatinib, osimertinib, alectinib, ceritinib, crizotinib, and brigatinib. The immunotherapy drugs included in the analysis are pembrolizumab, lambrolizumab, nivolumab, atezolizumab, durvalumab, and ipilimumab. Treatment with immunotherapy (OR = 1.86, 95% CI: 1.72–2.01, p < 0.001), targeted therapy (OR = 2.52, 95% CI: 2.21–2.88, p < 0.001) and chemotherapy (OR = 1.39, 95% CI: 1.31–1.48, p < 0.001) were associated with significantly higher likelihood of having biomarker testing performed (Table 4). These results were consistent with the analysis of treatment patterns in the subgroup analysis of patients with stage IV adenocarcinoma (Table 5).
Discussion
The IPASS study17 was the first randomized trial in advanced NSCLC to reveal the critical importance of biomarker testing to personalize therapy for patients. Multiple studies have subsequently noted improved outcomes for patients with certain driver mutations on receiving targeted therapy.8,9 As the number of known driver mutations with effective therapies has grown, guidelines have been continually updated to include them in the biomarker testing recommendations. In the United States, nine different molecular targets have approved targeted therapies for lung cancer. Driver mutations are seen predominantly in patients with adenocarcinoma histology; for squamous cell histology, very limited treatable driver mutations have been identified to date. Although a number of testing methods are used across centers, increasing utilization of next-generation sequencing has become the most practical approach to testing for multiple driver mutations. Given the critical importance of biomarker testing to the treatment of NSCLC, we studied the uptake of testing in the United States by querying the ASCO CancerLinQ database.
Our analysis notes that testing rates remain low even for patients with lung adenocarcinoma where targeted therapies are applicable to a high proportion of patients; other groups have also reported on lower rates of biomarker testing in the United States.18,19 The reasons behind this are likely to be multifactorial; lack of tumor tissue for testing, lack of awareness, and insurance coverage are all possible reasons for low adoption of biomarker testing. One common mistaken belief may be that only certain clinical characteristics are associated with the prevalence of driver mutation. In our analysis, the female sex and Asian race were most likely to be associated with having undergone biomarker testing. This parallels the initial observations that EGFR mutations were most common in these two patient subsets; therefore, EGFR inhibitors were preferentially used for these subsets of patients in the early days of the adoption of targeted therapy in NSCLC. Subsequent studies have noted that clinical characteristics should not be used to decide which patients undergo molecular testing.20 Nevertheless, our results note that the wide testing approach is yet to be fully adopted in the United States across practice sites.
Another glaring observation is that black patients are less likely to undergo molecular testing; these findings were also noted by Bruno et al.21 Although our study does not have the depth of understanding the specific reasons behind this disparity, prior studies have noted a variety of factors that contribute to disparities in black patients including educational status, socio-economic status and health insurance status.21,22 It is also to be noted that our group has previously shown that black patients who are treated with targeted therapy achieve comparable outcomes to White patients.23 In our analysis, Hispanic patients were more likely to be tested than the non-Hispanic population. Nevertheless, we acknowledge that this result could be potentially biased owing to the high volume of missing data in this population.
We also noticed variation in testing levels across various regions of the United States; the Northeast Region seems to have the highest testing rates in the country, followed by the South Region. Nevertheless, it is important to note that only 9% of the patients in this data set were from the Northeast compared with 44% from the South. In recent years, state-level legislations have been passed to require insurance companies to cover biomarker testing costs, and the impact of this on the uptake of testing remains to be studied. Only 3% of this cohort received targeted therapy, which is much lower than the overall prevalence of various driver mutations in the U.S. population, suggesting lost therapeutic opportunity for many patients.
The strengths of our analysis include the use of a real-world database with several hundred thousand patients included; the ASCO CancerLinQ database is also highly curated for lung cancer patients and has a high capture rate for biomarker testing results. It also provides a high representation of community sites across the United States. We specifically queried for patients diagnosed in 2010 and later because that corresponds to the biomarker testing era for NSCLC. Nevertheless, our work has the known limitations of real-world studies and does not provide light on the underlying reasons that drive testing decisions. The database does not provide information on the specific testing method that was used for biomarker testing. Regardless, these results reveal a clear need to enhance efforts to improve the adoption of biomarker testing in NSCLC. There is a need for focused studies to understand the reasons that contribute to non-testing. Education at the patient and care provider level will contribute to testing. The reduction in mortality related to lung cancer will be even more dramatic if every patient undergoes testing and has access to FDA-approved targeted therapies.
CRediT Authorship Contribution Statement
Madhusmita Behera: Conceptualization, Investigation, Methodology, Supervision, Writing - original draft.
Gregory Joseph: Methodology, Data curation, Formal analysis, Writing -review & editing.
Manali Rupji: Methodology, Statistical analysis, Writing - review & editing.
Zhonglu Huang: Methodology, Data curation, Formal analysis, Writing - review & editing.
Becky Bunn: Project administration, Funding acquisition, Writing - review & editing.
Murry W. Wynes: Project administration, Funding acquisition, Writing - review & editing.
Jeffrey Switchenko: Methodology, Writing - review & editing.
Giorgio Scagliotti: Conceptualization, Writing - review & editing.
Ming Tsao: Conceptualization, Writing - review & editing.
Chandra P. Belani: Conceptualization, Writing - review & editing.
Lecia Sequist: Conceptualization, Writing - review & editing.
Suresh S. Ramalingam: Conceptualization, Investigation, Supervision, Writing - review & editing.
Disclosure
Dr. Behera reports receiving funding or honoraria from LabKey and IntegraConnect. Dr. Scagliotti reports receiving funding or honoraria from BeiGene, AstraZeneca, Verastem, AstraZeneca, Merck Sharp & Dohme, Pfizer, Johnson and Johnson, and Takeda. Dr. Tsao reports receiving funding or honoraria from Bayer, Sanofi, Amgen, AstraZeneca, Daiichi-Sankyo, AbbVie, and Regeneron. Dr. Ramalingam reports receiving funding or honoraria from AstraZeneca, Bristol-Myers Squibb, Merck, GlaxoSmithKline, and Amgen. Dr. Sequist reports receiving grants or contracts from AstraZeneca, Novartis and Delfi Diagnostics. The remaining authors declare no conflict of interest.
Acknowledgments
Research reported in this publication was supported in part by the Data and Technology Shared Resource and Biostatistics Shared Resource of Winship Cancer Institute of Emory University and NIH/NCI under award number P30CA138292, award Number P50CA217691 and by International Association for The Study of Lung Cancer (IASLC). The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. The data used in this study was provided by ASCO CancerLinQ.
Footnotes
Cite this article as: Behera M, Joseph G, Rupji M, et al. Biomarker testing and patterns of treatment in patients with NSCLC: an International Association for The Study of Lung Cancer analysis of Association of Schools and Colleges of Optometry CancerLinQ Discovery data. JTO Clin Res Rep. 2026;7:100816.
References
- 1.Siegel R.L., Miller K.D., Wagle N.S., Jemal A. Cancer statistics, 2023. CA Cancer J Clin. 2023;73:17–48. doi: 10.3322/caac.21763. [DOI] [PubMed] [Google Scholar]
- 2.Miller K.D., Nogueira L., Devasia T., et al. Cancer treatment and survivorship statistics, 2022. CA Cancer J Clin. 2022;72:409–436. doi: 10.3322/caac.21731. [DOI] [PubMed] [Google Scholar]
- 3.Lee D.H., Tsao M.S., Kambartel K.O., et al. Molecular testing and treatment patterns for patients with advanced non-small cell lung cancer: PIvOTAL observational study. PLoS One. 2018;13 doi: 10.1371/journal.pone.0202865. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Howlader N., Forjaz G., Mooradian M.J., et al. The effect of advances in lung-cancer treatment on population mortality. N Engl J Med. 2020;383:640–649. doi: 10.1056/NEJMoa1916623. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Pacheco J.M. Approach to stage IV non-small-cell lung cancer: how to select among first-line therapy options? Curr Opin Pulm Med. 2020;26:311–320. doi: 10.1097/MCP.0000000000000659. [DOI] [PubMed] [Google Scholar]
- 6.Cetin K., Ettinger D.S., Hei Y.J., O’Malley C.D. Survival by histologic subtype in stage IV nonsmall cell lung cancer based on data from the Surveillance, Epidemiology and End Results Program. Clin Epidemiol. 2011;3:139–148. doi: 10.2147/CLEP.S17191. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Pacheco J.M., Gao D., Smith D., et al. Natural history and factors associated with overall survival in stage IV ALK-rearranged non-small cell lung cancer. J Thorac Oncol. 2019;14:691–700. doi: 10.1016/j.jtho.2018.12.014. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Herbst R.S., Morgensztern D., Boshoff C. The biology and management of non-small cell lung cancer. Nature. 2018;553:446–454. doi: 10.1038/nature25183. [DOI] [PubMed] [Google Scholar]
- 9.Kris M.G., Johnson B.E., Berry L.D., et al. Using multiplexed assays of oncogenic drivers in lung cancers to select targeted drugs. JAMA. 2014;311:1998–2006. doi: 10.1001/jama.2014.3741. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Ettinger D.S., Wood D.E., Aisner D.L., et al. Non-small cell lung cancer, version 5.2017, NCCN clinical practice guidelines in oncology. J Natl Compr Canc Netw. 2017;15:504–535. doi: 10.6004/jnccn.2017.0050. [DOI] [PubMed] [Google Scholar]
- 11.Tanner N.T., Pastis N.J., Sherman C., Simon G.R., Lewin D., Silvestri G.A. The role of molecular analyses in the era of personalized therapy for advanced NSCLC. Lung Cancer. 2012;76:131–137. doi: 10.1016/j.lungcan.2011.11.013. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Kazandjian D., Suzman D.L., Blumenthal G., et al. FDA approval summary: nivolumab for the treatment of metastatic non-small cell lung cancer with progression on or after platinum-based chemotherapy. Oncologist. 2016;21:634–642. doi: 10.1634/theoncologist.2015-0507. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Herbst R.S., Baas P., Kim D.W., 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;387:1540–1550. doi: 10.1016/S0140-6736(15)01281-7. [DOI] [PubMed] [Google Scholar]
- 14.Rittmeyer A., Barlesi F., Waterkamp D., et al. Atezolizumab versus docetaxel in patients with previously treated non-small-cell lung cancer (OAK): a phase 3, open-label, multicentre randomised controlled trial. Lancet. 2017;389:255–265. doi: 10.1016/S0140-6736(16)32517-X. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Potter D., Brothers R., Kolacevski A., et al. Development of CancerLinQ, a health information learning platform from multiple electronic health record systems to support improved quality of care. JCO Clin Cancer Inform. 2020;4:929–937. doi: 10.1200/CCI.20.00064. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Potter D.M., Riffon M.F., Manning B., et al. Summary of the 12 most common cancers in the CancerLinQ Discovery (CLQD) database. JCO Clin Cancer Inform. 2021;5:658–667. doi: 10.1200/CCI.21.00011. [DOI] [PubMed] [Google Scholar]
- 17.Mok T.S., Wu Y.L., Thongprasert S., et al. Gefitinib or carboplatin-paclitaxel in pulmonary adenocarcinoma. N Engl J Med. 2009;361:947–957. doi: 10.1056/NEJMoa0810699. [DOI] [PubMed] [Google Scholar]
- 18.Nadler E., Vasudevan A., Wang Y., Ogale S. Real-world patterns of biomarker testing and targeted therapy in de novo metastatic non-small cell lung cancer patients in the US oncology network. Cancer Treat Res Commun. 2022;31 doi: 10.1016/j.ctarc.2022.100522. [DOI] [PubMed] [Google Scholar]
- 19.Gutierrez M.E., Choi K., Lanman R.B., et al. Genomic profiling of advanced non-small cell lung cancer in community settings: gaps and opportunities. Clin Lung Cancer. 2017;18:651–659. doi: 10.1016/j.cllc.2017.04.004. [DOI] [PubMed] [Google Scholar]
- 20.Rosell R., Moran T., Queralt C., et al. Screening for epidermal growth factor receptor mutations in lung cancer. N Engl J Med. 2009;361:958–967. doi: 10.1056/NEJMoa0904554. [DOI] [PubMed] [Google Scholar]
- 21.Bruno D.S., Hess L.M., Li X., Su E.W., Patel M. Disparities in biomarker testing and clinical trial enrollment among patients with lung, breast, or colorectal cancers in the United States. JCO Precis Oncol. 2022;6 doi: 10.1200/PO.21.00427. [DOI] [PubMed] [Google Scholar]
- 22.Harrison S., Judd J., Chin S., Ragin C. Disparities in lung cancer treatment. Curr Oncol Rep. 2022;24:241–248. doi: 10.1007/s11912-022-01193-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Steuer C.E., Behera M., Berry L., et al. Role of race in oncogenic driver prevalence and outcomes in lung adenocarcinoma: results from the Lung Cancer Mutation Consortium. Cancer. 2016;122:766–772. doi: 10.1002/cncr.29812. [DOI] [PMC free article] [PubMed] [Google Scholar]