EGFR testing and erlotinib use in non-small cell lung cancer patients in Kentucky.

This study determined the frequency and factors associated with EGFR testing rates and erlotinib treatment as well as associated survival outcomes in patients with non small cell lung cancer in Kentucky. Data from the Kentucky Cancer Registry (KCR) linked with health claims from Medicaid, Medicare and private insurance groups were evaluated. EGFR testing and erlotinib prescribing were identified using ICD-9 procedure codes and national drug codes in claims, respectively. Logistic regression analysis was performed to determine factors associated with EGFR testing and erlotinib prescribing. Cox-regression analysis was performed to determine factors associated with survival. EGFR mutation testing rates rose from 0.1% to 10.6% over the evaluated period while erlotinib use ranged from 3.4% to 5.4%. Factors associated with no EGFR testing were older age, male gender, enrollment in Medicaid or Medicare, smoking, and geographic region. Factors associated with not receiving erlotinib included older age, male gender, enrollment in Medicare or Medicaid, and living in moderate to high poverty. Survival analysis demonstrated EGFR testing or erlotinib use was associated with a higher likelihood of survival. EGFR testing and erlotinib prescribing were slow to be implemented in our predominantly rural state. While population-level factors likely contributed, patient factors, including geographic location (areas with high poverty rates and rural regions) and insurance type, were associated with lack of use, highlighting rural disparities in the implementation of cancer precision medicine.

Introduction

Lung cancer is the leading cause of cancer death in the United State [1], and Kentucky leads the nation in both the rate of new cases and deaths due to cancer, with the Appalachian region carrying the highest cancer burden [2-4]. The high incidence and death rates in Kentucky demonstrate a clear need for more effective interventions in lung cancer patients.

Clinical studies associating EGFR mutations with better response to tyrosine kinase inhibitors were reported in 2004 [5-Proc Natl Acad Sci USA. 2004 ">7]. Ongoing clinical trials at that time did not require the presence of an EGFR mutation as an inclusion criteria, and erlotinib was initially approved in late 2004 as a monotherapy for the treatment of patients with locally advanced or metastatic non-small cell lung cancer (NSCLC) after failure of at least one prior chemotherapy regimen. The approval was based on the BR-21 trial, which compared erlotinib to placebo and demonstrated survival was significantly longer for patients treated with erlotinib. Multivariate analyses showed improved survival with erlotinib in the EGFR-positive group by immunohistochemistry, however since the multivariate analyses failed to rule out a small erlotinib survival effect in patients who were EGFR-negative, erlotinib was approved regardless of EGFR status [8].

The first EGFR mutation test was commercialized in 2005, however EGFR testing recommendations were not included in the American Society of Clinical Oncology (ASCO) and National Comprehensive Cancer Network (NCCN) guidelines until 2010 [9, 10]. The OPTIMAL and EUROTAC trials, which compared erlotinib to standard doublet chemotherapy in patients with EGFR mutations, demonstrated both improved progression-free survival and reduced adverse effects in the erlotinib arms. These were published in 2011 and 2012 and supported a new erlotinib indication in the front-line setting for EGFR mutant locally advanced or metastatic NSCLC [11, 12]. Erlotinib indications were updated again in 2016 with the publication of the IUNO trial, which demonstrated no survival benefit in EGFR wild-type individuals, and currently erlotinib is only approved in NSCLC for patients with an EGFR mutation [13]. Current guidelines published by ASCO recommend that all patients with advanced nonsquamous NSCLC, regardless of clinical characteristics such as age, race, or smoking status should undergo testing for EGFR and other actionable mutations. [14].

Despite the availability of an EGFR mutation test as early as 2005 and recommendations for routine EGFR mutation analysis as a part of standard care, not all patients are tested. A 2010 NCCN survey found that less than 50% of oncologists tested their patients for EGFR mutations, and that less than 50% of patients who received erlotinib had EGFR testing done. The same study found that age, location, comorbidity scores, and treatment history of radiation therapy affected whether or not patients received the testing [15]. A later survey found that lack of test availability, unfamiliarity with testing benefits, inadequate tissue for testing, patient refusal, or a lack of access to targeted clinical trials resulted in low mutation testing rates [16].

The purpose of this study is to evaluate EGFR testing and erlotinib use in patients with NSCLC in Kentucky and identify factors associated with lack of testing or erlotinib treatment and associated survival.

Materials and methods

Setting

The Kentucky Cancer Registry (KCR) is a population-based central cancer registry for the Commonwealth of Kentucky. All healthcare facilities that diagnose or treat cancer patients, including all acute care hospitals and associated outpatient facilities, freestanding treatment centers, private pathology laboratories, and physician offices, are required to report each case of cancer to the KCR. The KCR has been part of the Centers for Disease Control and Prevention (CDC) National Program of Cancer Registries since 1994 and the National Cancer Institute’s (NCI) Surveillance and Epidemiology and End Results (SEER) program since 2000. KCR has received the highest level of certification from the North American Association of Central Cancer Registries (NAACCR) indicating its commitment to accuracy, completeness, and quality [17].

KCR performed a probabilistic data linkage to identify matches between KCR and claims from Medicaid, state employee insurance and private insurance groups for cancer cases diagnosed in 2000–2012. Medicare claims were also acquired from the SEER Medicare database. The final data set consolidated the linked claims data, including cancer cases diagnosed in 2000–2011, and claims up to 2015 from sources mentioned above [18].

Study population

The cohort was selected from KCR with claims for cases diagnosed in 2007–2011. Patients must have presented with invasive NSCLC (Stage IIIB–Stage IV), have had continuous healthcare coverage one month prior to the date of diagnosis and one year after, and must have linked claims data. Over this time period, 5.3% of diagnosed cases occurred in uninsured individuals who were excluded from the analysis. Genetic test claims were captured within one month prior to diagnosis and three months after. Drug claims were captured within one year of diagnosis and could have been any line treatment (Table 1). The final cohort included 4957 individuals.

Table 1

Codes used to identify EGFR testing and erlotinib.

	Code Type	Codes Used
Erlotinib	NDC	69189–0063, 50242–062, 50242–063, 50242–064, 54868–5290, 54868–5447, 54868–5474, 54569–5848, 54569–5847
EGFR	CPT	81235,83891, 83894, 83896, 83898, 83903, 83904, 83907, 83909, 83912, 83890, 81401,83969

Demographics variables were extracted from the linked KCR data, including age at diagnosis, race, sex, smoking status, education, poverty status, metropolitan status, Appalachian status, insurance type, comorbidity, hospital type and distance to a hospital. Education level and poverty status were determined by percentage of high school completion rate and percentage of population below poverty range based on the 2000 US Census county estimates, then categorized into four levels based on the quartiles of their corresponding distributions. Metropolitan status was defined based on the 2013 Rural-Urban County Continuum Codes with values 1–3 as Metro and 4–9 as Non-Metro (https://www.ers.usda.gov/data-products/rural-urban-continuum-codes.aspx). Appalachian status was determined by the Appalachia Regional Commission (https://www.arc.gov/appalachian_region/CountiesinAppalachia.asp). A variable with the combination of metro and Appalachian status was also created. The reporting hospitals were categorized into two types: tertiary academic hospital (University of Kentucky and University of Louisville) or not. Carlson comorbidity index was calculated from the linked claims data. Using a Great Circle Distance approach, distance between patient residence and their corresponding hospital was calculated based the geocodes of the locations. Current Procedural Terminology (CPT) codes and National Drug Codes (NDC) were extracted from claims to identify the EGFR mutation test and erlotinib prescription.

Statistical analysis

Descriptive analysis of the demographic and clinical factors was performed. χ2 tests were used to examine associations between demographic/clinical factors and EGFR test/erlotinib prescription. Two logistic regressions were fitted separately to identify significant factors associated with EGFR test or erlotinib prescription while controlling for other covariates. Kaplan-Meier plots and Cox regression survival analysis were also performed to examine how EGFR testing and erlotinib affect overall survival. The final models only kept variables with p-value < 0.1. All analyses were done using SAS Statistical software version 9.4 (SAS Institute, Inc., Cary, North Carolina, USA). All statistical tests were two sided with a P-value ≤ 0.05 used to identify statistical significance.

Ethical considerations

This study was approved by the University of Kentucky IRB #51483. Informed consent was waived as all data was de-identified before analysis All data was treated highly as confidential and was only accessible in password-protected files for authorized study staff.

Results

From 2007 to 2011 the percentage of patients presenting with locally advanced or advanced stage disease that were tested for EGFR mutations increased from 0.1% to 10.6% (Table 2), while erlotinib use ranged from 3.4% to 5.4% with no trend over time. Demographics, including younger age, female gender, non-smokers and not being white or black were associated with EGFR testing and erlotinib prescribing. Individuals living in areas with high poverty, low high school attainment, and with Medicare or Medicaid insurance were significantly less likely to have EGFR testing or an erlotinib prescription. Geographic factors, both distance to an academic medical center and rural Appalachia, were significantly associated with EGFR testing, but not erlotinib prescribing.

Table 2

Bivariate analysis for EGFR testing and erlotinib receipt among NSCLC Stage III and IV patients.

	Had EGFR Testing					Received Erlotinib
	No	%	Yes	%	P	No	%	Yes	%	P
Total	4748	95.8%	209	4.2%		4744	95.7%	213	4.3%
Age					0.0072					0.0058
20–49	162	91.0%	16	9.0%		167	93.8%	11	6.2%
50–64	999	95.4%	48	4.6%		988	94.4%	59	5.6%
65–74	1976	95.9%	85	4.1%		1969	95.5%	92	4.5%
75+	1611	96.4%	60	3.6%		1620	96.9%	51	3.1%
Gender					<0.0001					0.0046
Male	2811	96.7%	96	3.3%		2802	96.4%	105	3.6%
Female	1937	94.5%	113	5.5%		1942	94.7%	108	5.3%
Race					0.0924					<0.0001
White	4438	95.7%	197	4.3%		4441	95.8%	194	4.2%
Black	299	96.8%	10	3.2%		294	95.1%	15	4.9%
Other	11	84.6%	2	15.4%		9	69.2%	4	30.8%
Stage					0.1729					0.2765
Stage IIIb and effusion	278	94.2%	17	5.8%		286	96.9%	9	3.1%
Stage IV	4470	95.9%	192	4.1%		4458	95.6%	204	4.4%
Metro Status					0.0001					0.5738
Metro	2291	94.7%	129	5.3%		2312	95.5%	108	4.5%
Non-Metro	2457	96.8%	80	3.2%		2432	95.9%	105	4.1%
Appalachia Status					0.0053					0.1029
Appalachia	1624	96.9%	52	3.1%		1615	96.4%	61	3.6%
Non-Appalachia	3124	95.2%	157	4.8%		3129	95.4%	152	4.6%
Appalachia and Metro Status					0.0010					0.0629
Appalachia Metro	166	96.5%	6	3.5%		171	99.4%	1	0.6%
Appalachia Non-Metro	1458	96.9%	46	3.1%		1444	96.0%	60	4.0%
Non-Appalachia Metro	2125	94.5%	123	5.5%		2141	95.2%	107	4.8%
Non-Appalachia Non-Metro	999	96.7%	34	3.3%		988	95.6%	45	4.4%
Year of Diagnosis					<0.0001					0.2454
2007	858	99.9%	1	0.1%		823	95.8%	36	4.2%
2008	944	99.6%	4	0.4%		910	96.0%	38	4.0%
2009	914	97.5%	23	2.5%		886	94.6%	51	5.4%
2010	1092	94.1%	69	5.9%		1121	96.6%	40	3.4%
2011	940	89.4%	112	10.6%		1004	95.4%	48	4.6%
Insurance Type					<0.0001					<0.0001
Private	966	93.0%	73	7.0%		972	93.6%	67	6.4%
Medicaid	502	98.2%	9	1.8%		490	95.9%	21	4.1%
Medicare	3280	96.3%	127	3.7%		3282	96.3%	125	3.7%
High School Attainment					<0.0001					0.0176
Very Low	1191	96.4%	44	3.6%		1193	96.6%	42	3.4%
Low	1203	97.3%	33	2.7%		1174	95.0%	62	5.0%
Moderate	1171	96.0%	49	4.0%		1179	96.6%	41	3.4%
High	1183	93.4%	83	6.6%		1198	94.6%	68	5.4%
Poverty					0.0122					0.0032
Low	1179	96.0%	49	4.0%		1193	97.1%	35	2.9%
Moderate	1062	94.1%	66	5.9%		1066	94.5%	62	5.5%
High	1301	96.0%	54	4.0%		1285	94.8%	70	5.2%
Very High	1206	96.8%	40	3.2%		1200	96.3%	46	3.7%
Charlson Comorbidity Index					0.3214					0.1085
0	2074	95.2%	104	4.8%		2071	95.1%	107	4.9%
1	1328	96.0%	56	4.0%		1327	95.9%	57	4.1%
2	682	96.5%	25	3.5%		677	95.8%	30	4.2%
3+	664	96.5%	24	3.5%		669	97.2%	19	2.8%
Smoking					0.0393					0.0088
No	258	92.8%	20	7.2%		256	92.1%	22	7.9%
Yes	4052	96.0%	170	4.0%		4051	95.9%	171	4.1%
Unknown	433	95.8%	19	4.2%		432	95.6%	20	4.4%
Distance to Academic Hospital					0.0001					0.1477
Less than 20 Miles	1111	93.5%	77	6.5%		1123	94.5%	65	5.5%
20–50 Miles	754	97.2%	22	2.8%		745	96.1%	31	4.0%
50–100 Miles	1707	96.3%	65	3.7%		1701	96.0%	71	4.0%
100+ Miles	1176	96.3%	45	3.7%		1175	96.2%	46	3.8%

Factors associated with EGFR testing were assessed through multi-variate logistic regression analysis (Table 3). Clinical variables, including age, gender and smoking status were associated with EGFR testing with younger, female, non-smokers more likely to be tested. Additionally, with the exception of 2008, the testing likelihood increased significantly for each year, 2009 (OR = 22.30, CI = 3.00 to 165.41), 2010 (OR = 58.56, CI = 8.12 to 422.26), and 2011 (OR = 113.47, CI = 15.81 to 814.21) compared to 2007 (P = <0.0001) despite overall rates remaining low. The variables measuring disparities were also significantly associated with a decreased likelihood of receiving testing. Patients enrolled in Medicaid (OR = 0.19, CI = 0.09 to 0.40) or Medicare (OR = 0.61, CI = 0.44 to 0.84) compared to those with private insurance (P = <0.0001) were less likely to receive testing. Those patients living in non-metropolitan areas, whether in Appalachian (OR = 0.51, CI = 0.36 to 0.73) or non-Appalachian regions (OR = 0.60, CI = 0.40 to 0.89), were also significantly less likely to receive testing (P = 0.0011).

Table 3

Factors associated with having EGFR somatic mutation testing in Stage IIIb–Stage IV NSCLC patients.

Modeling Had EGFR Testing
Variable	OR (95% CI)	P-Value
Age (ref = 75+)		0.0001
20–49	4.15 (2.17–7.91)
50–64	1.76 (1.16–2.67)
65–74	1.39 (0.98–1.98)
Sex (ref = Male)		0.0142
Female	1.44 (1.08–1.93)
Appalachian Status (ref = Non-Appalachia/Metro)		0.0011
Appalachian/Metro	0.67 (0.28–1.59)
Appalachian/Non-Metro	0.51 (0.36–0.73)
Non-Appalachian/Non-Metro	0.60 (0.40–0.89)
Year of Diagnosis (ref = 2007)		<0.0001
2008	3.81 (0.43–34.68)
2009	22.30 (3.00–165.41)
2010	58.56 (8.12–422.26)
2011	113.47 (15.81–814.21)
Insurance (ref = Private)		<0.0001
Medicaid	0.19 (0.09–0.40)
Medicare	0.61 (0.44–0.84)
Smoking (ref = No)		0.0266
Yes	0.54 (0.32–0.91)
Unknown	0.83 (0.42–1.66)

OR = odds ratio; CI = confidence interval; (ref) = reference variable

To determine factors associated with erlotinib prescribing, the same variables were examined through multivariate logistic regression analysis (Table 4). Similarly, younger patients and female patients were more likely to receive the drug. In addition, those patients enrolled in Medicaid (OR = 0.55, CI = 0.33 to 0.93) and Medicare (OR = 0.63, CI = 0.46 to 0.87) were significantly less likely to receive the drug compared to those enrolled in private insurance (P = 0.0074). Those patients living in areas with moderate (OR = 1.90, CI = 1.24 to 2.91) and high poverty (OR = 1.84, CI = 1.22 to 2.79) were also significantly less likely to receive the drug compared to those living in low poverty (P = 0.0081).

Table 4

Factors associated with the receiving erlotinib in Stage IIIb- Stage IV NSCLC patients.

Modeling Receive Erlotinib
Variable	OR (95% CI)	P-Value
Age (ref = 75+)		0.0077
20–49	2.05 (1.02–4.14)
50–64	1.97 (1.31–2.95)
65–74	1.56 (1.10–2.21)
Sex (ref = Male)		0.0045
Female	1.49 (1.13–1.97)
Insurance (ref = Private)		0.0074
Medicaid	0.55 (0.33–0.93)
Medicare	0.63 (0.46–0.87)
Poverty (ref = Low)		0.0081
Moderate	1.90 (1.24–2.91)
High	1.84 (1.22–2.79)
Very High	1.33 (0.85–2.09)

OR = odds ratio; CI = confidence interval; (ref) = reference variable

Cox-regression survival analysis was performed to determine factors associated with likelihood of survival in patients with Stage IIIb–IV NSCLC (Table 5). The clinical characteristics associated with improved survival include younger age, female gender and a low co-morbidity score. Several other variables predicted survival. When comparing patients living in metropolitan Appalachia (HR = 1.09, CI = 0.93 to 1.28), rural Appalachia (HR = 1.10, CI = 0.97 to 1.25), and rural non-Appalachian Kentucky (HR = 1.13, CI = 1.04 to 1.23), patients living in rural, non-Appalachian regions had a significantly decreased likelihood of survival compared to those living in a metropolitan region (P = 0.0372). Furthermore, patients enrolled in Medicaid (HR = 1.17, CI = 1.05 to 1.31) and Medicare (HR = 1.11, CI = 1.03 to 1.19) had a significantly lower likelihood survival of compared to those with private insurance survival (P = 0.0053). Finally, those patients receiving the EGFR test had a significantly increased likelihood of survival compared to those who had not received the test (HR = 0.77, CI = 0.67 to 0.89, P = 0.0030). Similarly, those patients that received erlotinib had an increased likelihood of survival compared to those who did not receive the drug (HR = 0.62, CI = 0.54 to 0.71, P = <0.0001).

Table 5

Cox-regression survival analysis for Stage IIIb-IV NSCLC patients.

Variable	HR (95% CI)	P-Value
Age (ref = 75+)		<0.0001
20–49	0.65 (0.55–0.77)
50–64	0.76 (0.70–0.83)
65–74	0.79 (0.74–0.85)
Sex (ref = Male)		<0.0001
Female	0.88 (0.83–0.93)
Appalachian Status (ref = Non-Appalachia/Metro)		0.0372
Appalachian/Metro	1.09 (0.93–1.28)
Appalachian/Non-Metro	1.10 (0.97–1.25)
Non-Appalachian/Non-Metro	1.13 (1.04–1.23)
Insurance (ref = Private)		0.0053
Medicaid	1.17 (1.05–1.31)
Medicare	1.11 (1.03–1.19)
Poverty (ref = Low Poverty)		0.0516
Moderate	1.10 (1.02–1.20)
High	0.98 (0.90–1.07)
Very High	1.01 (0.88–1.16)
Stage (ref = Stage IV)		0.0320
Stage IIIb and effusion	0.88 (0.78–0.99)
Charlson Comorbidity Index (ref = 3+)		<0.0001
0	0.76 (0.70–0.83)
1	0.82 (0.75–0.90)
2	0.85 (0.77–0.95)
EGFR Test (ref = No Test)		0.0003
Received Test	0.77 (0.67–0.89)
Erlotinib Drug (ref = No Drug)		<0.0001
Received Drug	0.62 (0.54–0.71)

HR = hazard ratio; CI = confidence interval; (ref) = reference variable

Kaplan-Meier survival estimates indicate that those patients receiving EGFR testing had an increased survival probability compared to those that did not receive EGFR testing (Fig 1a). Those that received erlotinib also had an increased survival probability compared to those patients not receiving the drug, especially during the 0 to 20 month time period (Fig 1b).

Fig 1

a. Kaplan-Meier survival curves for NSCLC patients by EGFR testing status b. Kaplan-Meier survival curves for NSCLC patients by erlotinib status.

Discussion

The original publications outlining the sensitivity of EGFR-positive NSCLC tumors to tyrosine kinase inhibitors (TKI) were published in 2004, and the first EGFR assay was commercialized in 2005 [5-Proc Natl Acad Sci USA. 2004 ">7]. Despite this, our analysis found that during the years 2007–2011, EGFR testing rates remained low. Erlotinib was approved as a second-line therapy in 2004 for metastatic NSCLC regardless of EGFR status, and its rate of use was also minimal in the years examined [8].

While EGFR testing rates have increased over time, still not all eligible patients receive testing. A study evaluating NSCLC patients seen in community medical oncology practices in New Jersey and Maryland showed between 2013 to 2015, 59% of eligible patients were tested for EGFR mutations, while a second study using data from a national, private health insurance company found testing rates to be around 61% between the years of 2010 to 2012 [19, 20]. In comparison, testing rates in Kentucky were substantially lower during this same time period, with 7% of eligible patients tested in 2010 and 12% tested in 2011, highlighting disparities between urban, privately insured individuals and rural, Medicare recipients. The time lag between the first publications in 2004 and the uptake of the EGFR test and erlotinib use could be due to a number of causes, both at a population level and due to individual patient characteristics. On a population level, the Centers for Medicare and Medicaid Services (CMS) did not approve reimbursement of the EGFR test until 2008, and ASCO and NCCN did not update their guidelines until 2010 [9, 10, 21]. Additionally, FDA-approved indications for erlotinib have changed multiple times since its approval in 2004, with 2013 being the first time it was indicated specifically for those patients with EGFR mutations. Finally, as each year passed, patients were more likely to receive the test compared to 2007, the first year of our analysis, suggesting wider implementation of testing over time.

Our analysis found patient level factors that further influenced testing rates and erlotinib prescribing. Younger patients and female patients were more likely to be tested for EGFR mutations and to receive erlotinib. This is possibly due to EGFR mutations occurring more frequently in younger NSCLC patients as well as in women [22, 23]. Factors that contributed to patients being less likely to receive the EGFR test were enrollment in Medicare or Medicaid and living in a rural area regardless of Appalachian status. Patients enrolled in Medicare or Medicaid and those living in high poverty areas were also significantly less likely to receive the drug.

While population factors, including delays in reimbursement, development of guidelines, and evolving FDA indications likely influenced uptake in Kentucky, we anticipate that patient characteristics associated with decreased testing are over-represented in our population and contribute to the lower than national average testing rates over the same time period. Specifically, our population contained a higher number of Medicare/Medicaid patients compared to the studies described above. In addition, Kentucky’s poverty rate is significantly higher than the national average (KY = 18.3%, national = 14.6%), with several counties in Appalachia reaching 35–40% [24]. This could result in significant health disparities compared with national or less rural populations.

Patients that received EGFR testing had increased survival compared to those who did not. As expected, younger age, female gender, lower stage, and less comorbidities were associated with improved survival. Other factors associated with better survival included having private insurance and living in a non-Appalachia, metropolitan area. Since testing itself should not impact survival, this is likely due to those patients receiving better overall healthcare, related to better access to care or better insurance coverage. Nationally, patients with cancer living in rural areas have worse outcomes when compared to those living in urban areas, related to income and access inequalities, and highlighting these disparities in our population and suggesting better overall healthcare in these patients as a proxy for increasing their chances for survival.[25, 26] Those patients that received erlotinib also had a significantly better chance of survival. This could be an effect of the drug or that patients with EGFR mutant positive NSCLC have an overall better prognosis than those who do not [27].

To our knowledge, this is the largest description of the use of precision medicine in a predominantly rural population and the first to show the impact of precision medicine implementation on patient outcomes. It is also the first to look at precision medicine in Appalachia, a predominantly impoverished and disparate population. Importantly, we demonstrate the uptake of precision medicine in a rural population and suggest that new testing and treatment strategies would similarly lag behind urban and academic medical centers. While the management of NSCLC has changed over the intervening years, this analysis has several advantages, including mature survival data and a comprehensive assessment of implementation over an extended time-period. In addition, data was collected longitudinally using a registry-based cohort, which allows for a large sample size and minimizes selection bias. Lastly, at the time that the data was collected, erlotinib was the only EGFR inhibitor available and NGS panel testing was not performed in Kentucky, which provides the opportunity to observe the implementation of a single precision medicine test and treatment in a population without competing interventions.

This study is not without its limitations. The EGFR status or the prior treatment history of the tested individuals is unknown and so we cannot assess the appropriateness of erlotinib prescribing. Only EGFR testing within three months, and erlotinib prescribing within one year of diagnosis were assessed. It is possible that patients received the testing or the drug outside of this time window, but the median survival of late stage lung cancer at the time of data collection was only twelve months, and we anticipate few, if any patients were missed. In addition, while the number of cases of lung cancer diagnosed in Kentucky were drawn from a population-based cancer registry, the analysis of erlotinib prescribing and EGFR testing was conducted with a linked insurance claims database. Therefore, uninsured patients were not included in the analysis, which represents a selection bias against the poorest end of the spectrum. We anticipate that the 5% of uninsured Kentucky patients with lung cancer were even less likely to receive testing or erlotinib therapy and to have poorer outcomes [28]. Linked claims were only available for the time period reported, so while these results do not reflect current practice patterns, the study presented the opportunity to study the implementation of a single precision medicine intervention without competing interventions. We hypothesize that precision medicine interventions continue to lag in rural communities and this highlights the need for further study. Lastly, we could not measure physician-related factors such as available resources and education.

In conclusion, EGFR testing and prescribing of erlotinib occurred at a low rate in in Kentucky. While population factors likely contributed, patient level factors including residing in rural areas and type of insurance were associated with decreased use and reduced survival, highlighting rural disparities in the implementation of cancer precision medicine.

23 Apr 2020

PONE-D-20-04879

EGFR testing and erlotinib use in non-small cell lung cancer patients in Kentucky

PLOS ONE

Dear Dr Kolesar,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

We would appreciate receiving your revised manuscript by Jun 07 2020 11:59PM. When you are ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter.

To enhance the reproducibility of your results, we recommend that if applicable you deposit your laboratory protocols in protocols.io, where a protocol can be assigned its own identifier (DOI) such that it can be cited independently in the future. For instructions see: http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). This letter should be uploaded as separate file and labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. This file should be uploaded as separate file and labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. This file should be uploaded as separate file and labeled 'Manuscript'.

Please note while forming your response, if your article is accepted, you may have the opportunity to make the peer review history publicly available. The record will include editor decision letters (with reviews) and your responses to reviewer comments. If eligible, we will contact you to opt in or out.

We look forward to receiving your revised manuscript.

Kind regards,

Randall J. Kimple

Academic Editor

PLOS ONE

Journal requirements:

When submitting your revision, we need you to address these additional requirements:

1.    Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at http://www.plosone.org/attachments/PLOSOne_formatting_sample_main_body.pdf and http://www.plosone.org/attachments/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2. In the ethics statement in the manuscript and in the online submission form, please provide additional information about the patient records used in your retrospective study. Specifically, please ensure that you have discussed whether all data were fully anonymized before you accessed them and/or whether the IRB or ethics committee waived the requirement for informed consent. If patients provided informed written consent to have data from their medical records used in research, please include this information.

3. Thank you for including your ethics statement:  "University of Kentucky IRB, Written consent, IRB #51483".

Please amend your current ethics statement to confirm that your named institutional review board or ethics committee specifically approved this study.

Once you have amended this/these statement(s) in the Methods section of the manuscript, please add the same text to the “Ethics Statement” field of the submission form (via “Edit Submission”).

For additional information about PLOS ONE ethical requirements for human subjects research, please refer to http://journals.plos.org/plosone/s/submission-guidelines#loc-human-subjects-research.

4. We note that you have indicated that data from this study are available upon request. PLOS only allows data to be available upon request if there are legal or ethical restrictions on sharing data publicly. For information on unacceptable data access restrictions, please see http://journals.plos.org/plosone/s/data-availability#loc-unacceptable-data-access-restrictions.

In your revised cover letter, please address the following prompts:

a) If there are ethical or legal restrictions on sharing a de-identified data set, please explain them in detail (e.g., data contain potentially identifying or sensitive patient information) and who has imposed them (e.g., an ethics committee). Please also provide contact information for a data access committee, ethics committee, or other institutional body to which data requests may be sent.

b) If there are no restrictions, please upload the minimal anonymized data set necessary to replicate your study findings as either Supporting Information files or to a stable, public repository and provide us with the relevant URLs, DOIs, or accession numbers. Please see http://www.bmj.com/content/340/bmj.c181.long for guidelines on how to de-identify and prepare clinical data for publication. For a list of acceptable repositories, please see http://journals.plos.org/plosone/s/data-availability#loc-recommended-repositories.

We will update your Data Availability statement on your behalf to reflect the information you provide.

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

Reviewer #2: Yes

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: Yes

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: Yes

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: This paper examined the prevalence of EGFR testing and erlotinib use in Kentucky using registry and insurance claim data. The manuscript is clear and well-written with appropriate statistical design. This work may raise community awareness and lead to improved adoption of guidelines for the management of NSCLC; both are strengths.

Several concerns are included below regarding the design and applicability of this work:

1. The cohort examined in this study was from 2007-2011. Since then, management for NSCLC, especially testing and treatment for EGFR-driven disease, has undergone major transformations. With the publication of the FLAURA trial in 2018 and subsequent FDA approval of osimertinib for treatment-naïve EGFR mutated NSCLC, the standard of care for front-line therapy has changed and erlotinib is no longer the recommendated therapy for advanced EGFR-driven NSCLC. The relevance and applicability of this study are questionable given this evolving practice pattern. If the authors could update their cohort to reflect the change in practice pattern, it would significantly strengthen the work.

2. Along these lines, if the authors are able to include more recent data, i.e. post 2016, it would be informative and enhance the story to look at testing pattern and drug usage for other mutation-driven NSCLC such as ALK. Crizotinib was approved in 2016 so the authors would need to have access to registry and insurance claim data post-2016.

3. Similar studies looking at this question have been published previously with more recent data (after 2011), with similar conclusions. The authors even cited one such study in their references. Thus, it is hard to differentiate the novelty of the current work from its predecessors.

4. It would be more informative if the authors were able determine that among the patients who received EGFR testing, how many tested positive? Among those who tested positive, how many received erlotinib.

5. Confounders and biases, some the authors have addressed in the conclusion:

a. Study did not address why utilization of EGFR testing and erlotinib is so low. Was this due to physician education, patient understanding, availability of testing, lack of insurance coverage? The design of the study examined mostly patient-specific and possibly insurance factors but did not address availability of resources or physician-related factors.

b. All patients in this study had insurance coverage. So those who did not were excluded. Unclear how this reflects the broader population of Kentucky.

c. The % of EGFR testing and erlotinib prescription may be falsely low compared to other similar studies or even the national average because the design of the study captured all patients rather than patients who fit the demographics of EGFR-driven disease. Since the general NSCLC patient population in Kentucky probably has lower EGFR prevalence compared to some areas in the US, i.e. West Coast, if looking at all-comers, both testing and treatment may be lower because the prevalence of EGFR mutation is lower.

d. Table 5 Cox-regression survival analysis I do not think that EGFR testing and erlotinib being statistically significant in this model are meaningful because, as the authors pointed out, testing is a likely surrogate for receiving guideline-appropriate care, and EGFR-driven disease is relatively more indolent with better prognosis. Both of these have favorable impact on survival.

6. Minor: please state explicitly how many patients were actually included in the cohort.

Reviewer #2: This study analyzed factors associated with EGFR testing and erlotinib prescribing in Kentucky from 2007 to 2011. The analysis used the Kentucky Cancer Registry linked with health claims from Medicaid, Medicare and private insurance groups. The study concludes that EGFR testing and prescribing of erlotinib occurred at a low rate in Kentucky and factors including residing in rural areas and type of insurance were associated with decreased use and reduced survival. While the methodology appears, appropriate, my main concern is the overall relevance of this study in 2020.

This paper looks at EGFR testing and erlotinb use during a time when EGFR inhibitors were still under clinical investigation and not FDA approved as front line therapy. Practice patterns of oncologists were still adjusting as new data came out. Multiple previous papers, which the authors have cited, have already been published on this topic showing the slow rate of testing and the obstacles of implementing EGFR testing. The analysis has multiple limitations as outlined in the second from last paragraph. It is therefore hard to draw firm conclusions from the data. Not sure, how this data is relevant in 2020 or how it would be used to advanced patient care or current public health policy in Kentucky. The conclusion that rural areas and poverty are barriers to providing health care have already been well documented. An analysis of current data and/or an analysis that examines specific barriers to EGFR testing (such as state policy on testing or laboratory specific barriers or educational programs for oncologists) would have been more impactful.

Comments to be addressed:

• It is unclear what this this study contributes to the field. The analysis appears to be 10 years too late to impact public health policy on precision medicine. Can you explain why this analysis is relevant in 2020? What is the status of EGFR or broad genomic testing for NSCLC in Kentucky in 2020?

• Can the authors speak more about access to EGFR testing in Kentucky? Was testing being done in Kentucky or was it being sent out of state? How many centers in Kentucky were doing EGFR testing during this time? Was there a state effort during this time to assist in EGFR testing?

• Can the authors discuss the barriers that individual oncologists faced with EGFR testing? Was there slow dissemination of knowledge among the healthcare team? What is the distribution of oncologists in regards to rural and metro locations?

• It is unclear if you are analyzing front line use of erlotinib or second-line use or erlotinib. Please clarify in the methods.

• Why was gefitinib not included in the analysis? It was FDA approved in 2003 as second line therapy in metastatic NSCLC.

• Discussion paragraph #5: The survival difference seen in those that had EGFR testing can also be attributed to younger age and because a higher proportion were most likely EGFR mutated and actually derived benefit from erlotinib.

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files to be viewed.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email us at figures@plos.org. Please note that Supporting Information files do not need this step.

10 Jun 2020

Author response to reviewer comments

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at http://www.plosone.org/attachments/PLOSOne_formatting_sample_main_body.pdf and http://www.plosone.org/attachments/PLOSOne_formatting_sample_title_authors_affiliations.pdf

Author response: Revised as suggested

2. In the ethics statement in the manuscript and in the online submission form, please provide additional information about the patient records used in your retrospective study. Specifically, please ensure that you have discussed whether all data were fully anonymized before you accessed them and/or whether the IRB or ethics committee waived the requirement for informed consent. If patients provided informed written consent to have data from their medical records used in research, please include this information.

Author response: revised as suggested

3. Thank you for including your ethics statement: "University of Kentucky IRB, Written consent, IRB #51483".

Please amend your current ethics statement to confirm that your named institutional review board or ethics committee specifically approved this study.

Once you have amended this/these statement(s) in the Methods section of the manuscript, please add the same text to the “Ethics Statement” field of the submission form (via “Edit Submission”).

For additional information about PLOS ONE ethical requirements for human subjects research, please refer to http://journals.plos.org/plosone/s/submission-guidelines#loc-human-subjects-research.

Author response: revised as suggested

4. We note that you have indicated that data from this study are available upon request. PLOS only allows data to be available upon request if there are legal or ethical restrictions on sharing data publicly. For information on unacceptable data access restrictions, please see http://journals.plos.org/plosone/s/data-availability#loc-unacceptable-data-access-restrictions.

In your revised cover letter, please address the following prompts:

a) If there are ethical or legal restrictions on sharing a de-identified data set, please explain them in detail (e.g., data contain potentially identifying or sensitive patient information) and who has imposed them (e.g., an ethics committee). Please also provide contact information for a data access committee, ethics committee, or other institutional body to which data requests may be sent.

Author Response. Raw data cannot be shared because they are both potentially identifying and contain sensitive patient data, including geographic location, dates of diagnosis and dates of testing and receiving a medication. In addition, there are contractual agreements between the University of Kentucky and the Kentucky Cancer Registry precluding data sharing. Any requests for data must be submitted to:

Jacyln K. McDowell, Epidemiologist, Kentucky Cancer Registry

2365 Harrodsburg Rd, Suite A230

Lexington, KY 40504

859-218-2228

b) If there are no restrictions, please upload the minimal anonymized data set necessary to replicate your study findings as either Supporting Information files or to a stable, public repository and provide us with the relevant URLs, DOIs, or accession numbers. Please see http://www.bmj.com/content/340/bmj.c181.long for guidelines on how to de-identify and prepare clinical data for publication. For a list of acceptable repositories, please see http://journals.plos.org/plosone/s/data-availability#loc-recommended-repositories.

We will update your Data Availability statement on your behalf to reflect the information you provide.

Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: This paper examined the prevalence of EGFR testing and erlotinib use in Kentucky using registry and insurance claim data. The manuscript is clear and well-written with appropriate statistical design. This work may raise community awareness and lead to improved adoption of guidelines for the management of NSCLC; both are strengths.

Several concerns are included below regarding the design and applicability of this work:

1. The cohort examined in this study was from 2007-2011. Since then, management for NSCLC, especially testing and treatment for EGFR-driven disease, has undergone major transformations. With the publication of the FLAURA trial in 2018 and subsequent FDA approval of osimertinib for treatment-naïve EGFR mutated NSCLC, the standard of care for front-line therapy has changed and erlotinib is no longer the recommendated therapy for advanced EGFR-driven NSCLC. The relevance and applicability of this study are questionable given this evolving practice pattern. If the authors could update their cohort to reflect the change in practice pattern, it would significantly strengthen the work.

Author response: The discussion was revised as followd: “To our knowledge, this is the largest description of the use of precision medicine in a predominantly rural population and the first to show the impact of precision medicine implementation on patient outcomes. It is also the first to look at precision medicine in Appalachia, a predominantly impoverished and disparate population. Importantly, we demonstrate the uptake of precision medicine in a rural population and suggest that new testing and treatment strategies would similarily lag behind urban and academic medical centers. While the management of NSCLC has changed over the intervening years, this analysis has several advantages, including mature survival data and a comprehensive assessment of implementation over an extended time-period. In addition, data was collected longitudinally using a registry-based cohort, which allows for a large sample size and minimizes selection bias. Lastly, at the time that the data was collected, erlotinib was the only EGFR inhibitor available and NGS panel testing was not performed in Kentucky, which provides the opportunity to observe the implementation of a single precision medicine test and treatment in a population without competing interventions.

2. Along these lines, if the authors are able to include more recent data, i.e. post 2016, it would be informative and enhance the story to look at testing pattern and drug usage for other mutation-driven NSCLC such as ALK. Crizotinib was approved in 2016 so the authors would need to have access to registry and insurance claim data post-2016.

Author response: Please see response to comment 1, above.

3. Similar studies looking at this question have been published previously with more recent data (after 2011), with similar conclusions. The authors even cited one such study in their references. Thus, it is hard to differentiate the novelty of the current work from its predecessors.

Author response: We suggest that the novelty of the finding is that testing was lower in a rural region with known health disparities over the same time period and have revised the discussion as follows (bolded=new):

In comparison, we found that the testing rates in Kentucky were substantially lower during this same time period, with 7% of eligible patients tested in 2010 and 12% tested in 2011, highlighting disparities between urban, privately insured individuals and rural, Medicare recipients.

4. It would be more informative if the authors were able determine that among the patients who received EGFR testing, how many tested positive? Among those who tested positive, how many received erlotinib.

Author response: We agree, and since this data is not available we included it as a limitation

5. Confounders and biases, some the authors have addressed in the conclusion:

a. Study did not address why utilization of EGFR testing and erlotinib is so low. Was this due to physician education, patient understanding, availability of testing, lack of insurance coverage? The design of the study examined mostly patient-specific and possibly insurance factors but did not address availability of resources or physician-related factors.

Author response: We agree, and since this data is not available we revised the limitation to include the following.

Lastly, we could not measure physician-related factors such as available resources and education.

b. All patients in this study had insurance coverage. So those who did not were excluded. Unclear how this reflects the broader population of Kentucky.

Author response: The methods were revised as follows: Over this time period, 5.3% of diagnosed cases occurred in uninsured individuals who were excluded from the analysis.

c. The % of EGFR testing and erlotinib prescription may be falsely low compared to other similar studies or even the national average because the design of the study captured all patients rather than patients who fit the demographics of EGFR-driven disease. Since the general NSCLC patient population in Kentucky probably has lower EGFR prevalence compared to some areas in the US, i.e. West Coast, if looking at all-comers, both testing and treatment may be lower because the prevalence of EGFR mutation is lower.

Author response: Thank-you for your comment. While we know today that EGFR mutation frequency is lower in Kentucky than in other populations, this was not known during the study time period and national guidelines has not recommended testing sub-sets for EGFR mutations and the referenced studies we are comparing to did not exclude patients based on clinical characteristics.

d. Table 5 Cox-regression survival analysis I do not think that EGFR testing and erlotinib being statistically significant in this model are meaningful because, as the authors pointed out, testing is a likely surrogate for receiving guideline-appropriate care, and EGFR-driven disease is relatively more indolent with better prognosis. Both of these have favorable impact on survival.

Author response: We respectfully disagree. Disparities in survival between rural and urban cancer patients is well known and likely related to access and income inequalities. Lack of EGFR testing (and other standard of care treatments) in our population is likely contributing to the survival disadvantage.

The following was added to the discussion: Nationally, patients with cancer living in rural areas have worse outcomes when compared to those living in urban areas, related to income and access inequalities, and highlighting these disparities in our population.

The following references were cited:

Henley SJ, Anderson RN, Thomas CC, Massetti GM, Peaker B, Richardson LC. Invasive cancer incidence, 2004-2013, and deaths, 2006-2015, in nonmetropolitan and metropolitan counties—United States. MMWR Surveill Summ. 2017;66(14):-. doi:10.15585/mmwr.ss6614a1

Iglehart JK. The challenging quest to improve rural health care. N Engl J Med. 2018;378(5):473-479. doi:10.1056/NEJMhpr1707176

6. Minor: please state explicitly how many patients were actually included in the cohort.

Author response: The methods were revised as follows: The final cohort included 4957 individuals.

Reviewer #2: This study analyzed factors associated with EGFR testing and erlotinib prescribing in Kentucky from 2007 to 2011. The analysis used the Kentucky Cancer Registry linked with health claims from Medicaid, Medicare and private insurance groups. The study concludes that EGFR testing and prescribing of erlotinib occurred at a low rate in Kentucky and factors including residing in rural areas and type of insurance were associated with decreased use and reduced survival. While the methodology appears, appropriate, my main concern is the overall relevance of this study in 2020.

This paper looks at EGFR testing and erlotinb use during a time when EGFR inhibitors were still under clinical investigation and not FDA approved as front line therapy. Practice patterns of oncologists were still adjusting as new data came out. Multiple previous papers, which the authors have cited, have already been published on this topic showing the slow rate of testing and the obstacles of implementing EGFR testing. The analysis has multiple limitations as outlined in the second from last paragraph. It is therefore hard to draw firm conclusions from the data. Not sure, how this data is relevant in 2020 or how it would be used to advanced patient care or current public health policy in Kentucky. The conclusion that rural areas and poverty are barriers to providing health care have already been well documented. An analysis of current data and/or an analysis that examines specific barriers to EGFR testing (such as state policy on testing or laboratory specific barriers or educational programs for oncologists) would have been more impactful.

Comments to be addressed:

• It is unclear what this this study contributes to the field. The analysis appears to be 10 years too late to impact public health policy on precision medicine. Can you explain why this analysis is relevant in 2020? What is the status of EGFR or broad genomic testing for NSCLC in Kentucky in 2020?

Author response: Please see response to comment 1, reviewer 1, above.

• Can the authors speak more about access to EGFR testing in Kentucky? Was testing being done in Kentucky or was it being sent out of state? How many centers in Kentucky were doing EGFR testing during this time? Was there a state effort during this time to assist in EGFR testing?

Author response: Thank-you for your comment, but physician related factors this outside the scope of the manuscript. We have revised the discussion to include this as a limitation.

• Can the authors discuss the barriers that individual oncologists faced with EGFR testing? Was there slow dissemination of knowledge among the healthcare team? What is the distribution of oncologists in regards to rural and metro locations?

Author response: Thank-you for your comment, but physician related factors this outside the scope of the manuscript. We have revised the discussion to include this as a limitation.

• It is unclear if you are analyzing front line use of erlotinib or second-line use or erlotinib. Please clarify in the methods.

Author response: The methods were revised as follows (new text bold): Drug claims were captured within one year of diagnosis and could have been any line treatment.

• Why was gefitinib not included in the analysis? It was FDA approved in 2003 as second line therapy in metastatic NSCLC.

Author response: We did assess gefitinib and crizotinib use in the claims cohort, but since no gefitinib was prescribed and less than 5 crizotinib claims were identified we did not include them in the analysis.

• Discussion paragraph #5: The survival difference seen in those that had EGFR testing can also be attributed to younger age and because a higher proportion were most likely EGFR mutated and actually derived benefit from erlotinib.

Author response: We respectfully disagree, the survival analysis (Cox-regression model) controls for other factors associated with survival and having EGFR testing in the model is an independent predictor of survival.

23 Jun 2020

PONE-D-20-04879R1

EGFR testing and erlotinib use in non-small cell lung cancer patients in Kentucky

PLOS ONE

Dear Dr. Kolesar,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

Please submit your revised manuscript by Aug 07 2020 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols

We look forward to receiving your revised manuscript.

Kind regards,

Randall J. Kimple

Academic Editor

PLOS ONE

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #1: (No Response)

Reviewer #2: (No Response)

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

Reviewer #2: Partly

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: Yes

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: Yes

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

**********

6. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: Since both reviewers raised the concern of novelty and relevance of the study, please ask the authors to explain why they did not choose to expand the analysis using the Kentucky Cancer Registry to years beyond 2011 to reflect the changing practice patterns in EGFR-driven NSCLC, rather than simply rephrasing their discussion. If this is feasible, it should done.

Reviewer #2: Table 4: Since EGFR screening is associated with younger non-smoking women (most likely a surrogate for selecting appropriate patients for testing), is it possible that younger non-smoking women were more likely to have private insurance and lower poverty? Have you performed any statistical tests to see if these are associated? If so, it is possible that being a young non-smoker woman is the most important variable associated with testing and not necessary access to testing.

Discussion paragraph 5: The conclusion that that the improved survival with EGFR testing and erlotinib "is likely due to those patients receiving better overall healthcare, relating to care access, insurance coverage or poverty status" is flawed and not supported by your data. Overall survival is most likely better because the patients that received EGFR testing were more likely to have an EGFR mutation and thus obtained benefit from erlotinib. You state that this is not the case because only 10-15% would be positive for the mutation. This 10-15% is for the general public. It is most likely that those receiving EGFR testing in your population have a high chance of having an EGFR mutation given they they are predominately younger non-smoking women. Also, a population enriched with 10-15% of patients with EGFR mutations receiving erlotinib can still be responsible for the overall survival benefit. If you want to make a conclusion that access to care leads to better survival that is fine. But use the metropolitan vs rural and medicaid/medicare vs private insurance analysis instead of the EGFR/erlotinib analysis to support this conclusion.

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

17 Jul 2020

Reviewer #1: Since both reviewers raised the concern of novelty and relevance of the study, please ask the authors to explain why they did not choose to expand the analysis using the Kentucky Cancer Registry to years beyond 2011 to reflect the changing practice patterns in EGFR-driven NSCLC, rather than simply rephrasing their discussion. If this is feasible, it should done.

Author response: The linked claims database was created with a funded research collaboration that covered the time period reported, therefore it is not feasible to update the analysis.

The following was added to the limitations paragraph: “Linked claims were only available for the time period reported, so while these results do not reflect current practice patterns, the study presented the opportunity to study the implementation of a single precision medicine intervention without competing interventions. We hypothesize that precision medicine interventions continue to lag in rural communities and this highlights the need for further study.”

Reviewer #2: Table 4: Since EGFR screening is associated with younger non-smoking women (most likely a surrogate for selecting appropriate patients for testing), is it possible that younger non-smoking women were more likely to have private insurance and lower poverty? Have you performed any statistical tests to see if these are associated? If so, it is possible that being a young non-smoker woman is the most important variable associated with testing and not necessary access to testing.

Author response: Table 4 shows the primary effect of significant factors associated with receiving erlotinib in a logistic regression analysis. The model identified independent association of each factor for the outcome variable. Hence the effects of insurance and poverty are significant regardless the status of age, gender or smoking. No changes were made to the manuscript.

Discussion paragraph 5: The conclusion that that the improved survival with EGFR testing and erlotinib "is likely due to those patients receiving better overall healthcare, relating to care access, insurance coverage or poverty status" is flawed and not supported by your data. Overall survival is most likely better because the patients that received EGFR testing were more likely to have an EGFR mutation and thus obtained benefit from erlotinib. You state that this is not the case because only 10-15% would be positive for the mutation. This 10-15% is for the general public. It is most likely that those receiving EGFR testing in your population have a high chance of having an EGFR mutation given they they are predominately younger non-smoking women. Also, a population enriched with 10-15% of patients with EGFR mutations receiving erlotinib can still be responsible for the overall survival benefit. If you want to make a conclusion that access to care leads to better survival that is fine. But use the metropolitan vs rural and medicaid/medicare vs private insurance analysis instead of the EGFR/erlotinib analysis to support this conclusion.

Author response: The 5th paragraph of the discussion was revised as follows:

As expected, younger age, female gender, lower stage, and less comorbidities were associated with improved survival. Other factors associated with better survival included having private insurance and living in a non-Appalachia, metropolitan area. Since testing itself should not impact survival, this is likely due to those patients receiving better overall healthcare, related to better care access to care or better insurance coverage.

Submitted filename: Response to reviewers.doc.docx

Click here for additional data file.

4 Aug 2020

EGFR testing and erlotinib use in non-small cell lung cancer patients in Kentucky

PONE-D-20-04879R2

Dear Dr. Kolesar,

We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice for payment will follow shortly after the formal acceptance. To ensure an efficient process, please log into Editorial Manager at http://www.editorialmanager.com/pone/, click the 'Update My Information' link at the top of the page, and double check that your user information is up-to-date. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind regards,

Randall J. Kimple

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #1: All comments have been addressed

Reviewer #2: All comments have been addressed

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

Reviewer #2: Yes

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: Yes

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: Yes

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

**********

6. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: While having more recent Kentucky Cancer Registry data would be better, all concerns were addressed by the authors.

Reviewer #2: The authors have addressed all of my concerns and have appropriately revised the manuscript. There are no new concerns.

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

7 Aug 2020

PONE-D-20-04879R2

EGFR testing and erlotinib use in non-small cell lung cancer patients in Kentucky

Dear Dr. Kolesar:

I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

If we can help with anything else, please email us at plosone@plos.org.

Thank you for submitting your work to PLOS ONE and supporting open access.

Kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Dr. Randall J. Kimple

Academic Editor

PLOS ONE