Sex-specific risk of heart failure and acute myocardial infarction associated with osimertinib in older patients with non-small cell lung cancer

Abstract

Background:

The cardiovascular risk associated with the newer epidermal growth factor receptor tyrosine kinase inhibitor (EGFR-TKI), osimertinib, in non-small cell lung cancer (NSCLC) is being reported. While more than half of patients with NSCLC are aged 65 or older, whether the cardiovascular risk increases in older patients treated with osimertinib compared to traditional EGFR-TKIs has not been clearly identified.

Objectives:

We aimed to identify the risk of major adverse cardiovascular events (MACEs) in older patients treated with osimertinib and traditional EGFR-TKIs, stratifying by sex, age, and race/ethnicity.

Design:

A retrospective cohort study.

Methods:

Using the 2006–2019 Surveillance, Epidemiology, and End Results (SEER)-Medicare, older patients with advanced NSCLC prescribed EGFR-TKIs were included. Hazard ratio (HR) and 95% confidence interval (CI) of incident MACEs were calculated using Cox proportional hazard model.

Results:

Osimertinib had a significantly higher risk of heart failure (HF) [HR, 1.20 (95% CI, 1.01–1.42)] and a lower risk of angina [HR, 0.36 (95% CI, 0.20–0.64)] than first/second-generation EGFR-TKI. The risk of HF and acute myocardial infarction (AMI) was significantly elevated in female [HR, 1.40 (95% CI, 1.13–1.73)] and male [HR, 1.98 (95% CI, 1.13–3.48)] subgroups, respectively, when used osimertinib than first/second-generation EGFR-TKIs. Within osimertinib users, the risk of HF was higher in patients aged ⩾75 (versus aged 65–74) [HR, 1.71 (95% CI, 1.07–2.71)], White patients (versus Asian/Pacific Islanders patients) [HR, 1.80 (95% CI, 1.04–3.12)]. The risk of AMI within osimertinib users was higher in White [HR, 3.23 (95% CI, 1.34–7.80)] and Black [HR, 5.63 (95% CI, 1.23–25.87)] than Asian/Pacific Islanders patients.

Conclusion:

Increased risk of HF was observed in patients who are female, White, and aged ⩾75 while male patients had a higher risk of AMI when treated with osimertinib than first/second-generation EGFR-TKIs. The in-depth examination of osimertinib-induced cardiotoxicity provides evidence for individualized cardiotoxicity surveillance, prevention, and treatment for osimertinib users.

Keywords

cardio-oncology epidermal growth factor receptor tyrosine kinase inhibitors major adverse cardiovascular event non-small cell lung cancer pharmacoepidemiologic study

Introduction

As epidermal growth factor receptor-tyrosine kinase inhibitors (EGFR-TKIs) improved survival and quality of life in patients with advanced non-small cell lung cancer (NSCLC) with EGFR-mutation,^1,2 EGFR-TKIs have been utilized in the treatment of NSCLC since 2003. The most recently approved third-generation EGFR-TKI, osimertinib, has been used since 2015, showing improved overall survival compared to first/second-generation EGFR-TKIs.³

However, there has been a growing concern regarding the cardiovascular risk associated with EGFR-TKIs,^4,5 especially osimertinib.^6,7 In a previous study based on reported cardiovascular adverse events associated with EGFR-TKIs, osimertinib was found to have a higher risk of heart failure (HF) and myocardial infarction compared to first/second-generation EGFR-TKIs.⁸ In patients with lung cancer, concurrent cardiovascular disease (CVD) is associated with significantly worse survival,⁹ lower quality of life,¹⁰ and higher economic burden.¹¹ While over half of patients with lung cancer are diagnosed at 65 years old or older,¹² older patients are at a higher risk of developing CVD.¹³ Patients aged ⩾65 account for more than 50% of hospitalizations due to CVD and about 80% of cardiovascular-related deaths.¹⁴ In addition to increasing age,¹⁵ sex, and race/ethnicity are known to be associated with the risk of CVD with cancer therapies.^16,17 Characterization of the CVD risk associated with new-generation EGFR-TKIs is essential in clinical risk stratification of TKI cardiotoxicity and optimizing prevention, mitigation, and treatment of TKI-associated cardiac dysfunction. However, the evidence on new-generation EGFR-TKI-associated cardiovascular risk is scarce, especially the impact of demographics (age, sex, race/ethnicity) on the risk of cardiotoxicity.

Major adverse cardiovascular events (MACEs) are commonly used to assess the risk of CVD in cancer treatment.^18,19 Three-point MACE was defined to include acute myocardial infarction (AMI), stroke, and cardiac death by the U.S. Food and Drug Administration (FDA) in 2008 and the European Medicines Agency in 2012.²⁰ Thereafter, studies have assessed four-point MACE that additionally included unstable angina to the three-point MACE,²¹ and five-point MACE that added HF to the four-point MACE.²² Since HF is an important CVD included when assessing the cardiovascular risk in cancer treatment,^19,23,24 assessing the risk of five-point MACE would enable a comprehensive assessment of EGFR-TKI-associated cardiotoxicity.

In this study, we compared the risk of MACE between osimertinib and first/second-generation EGFR-TKIs in older patients with advanced NSCLC using a nationwide cancer registry and administrative data.

Materials and methods

Data source

This retrospective observational study used the 2006–2019 Surveillance, Epidemiology, and End Results (SEER)-Medicare linked data. The SEER provides population-based cancer registry data of 48% of the US population,²⁵ and Medicare covers over 97% of patients aged ⩾65 in the United States^26,27 Likewise, the SEER-Medicare enables epidemiological and health services research based on population-level sociodemographic/clinical information (sex, age, race/ethnicity, income, geographic region, urbanicity, plan benefits, marital status, year of diagnosis, stage, histology, tumor size at diagnosis, radiation/surgery) from SEER and enrollment/claims data (diagnosis, healthcare resource utilization, healthcare costs) from Medicare. The direct identifiers are not available, but geographic locations or dates of admission/discharge/service/birth are available through SEER-Medicare data. Therefore, researchers must submit research-specific application that contains a data protection, plan for storage, institutional review board (IRB) determination, and a signed Data Usage Agreement to National Cancer Institute (NCI).²⁶

The IRB of the University of Texas at Austin determined this study is not involving human subjects, therefore does not need IRB review and approval. This study followed the strengthening the reporting of observational studies in epidemiology guideline when reporting (Supplemental Method 1).²⁸

Patient population

We included older patients aged ⩾65 at diagnosis of advanced stage NSCLC and first prescribed EGFR-TKI between 1 January 2007 and 31 December 2017 (i.e. index period) in the analysis. International Classification of Diseases, 10th Revision, Clinical Modification (ICD-10-CM) (C34) was used to identify the diagnosis of lung cancer. ICD-O-3 histology codes for adenocarcinoma, squamous cell carcinoma, and large cell carcinoma were used to identify NSCLC.^29,30 American Joint Committee on Cancer (AJCC) stage (stages IIIB and IV) and SEER summary stage (regional and distant) were used to identify advanced staged NSCLC. Patients were followed up from the date of first prescription of EGFR-TKIs (i.e. index date) until the earliest date among incident MACE, death, or end of study period (31 December 2019) (i.e. post-index period) (Supplemental Figure 1). We excluded patients who were (1) treated with both osimertinib and first/second-generation EGFR-TKIs between 2006 and 2019, (2) included in Medicare due to end-stage renal disease or disability between 2006 and 2019, (3) included in Medicare health maintenance organizations during 1 year prior to the index date (i.e. pre-index period) and the post-index period, (4) not continuously enrolled in Medicare parts A and B in the pre-/post-index period, or (5) not continuously enrolled in Medicare part D in the post-index period.

We created five cohorts with patients who had not been diagnosed with each of the following diseases (i–v) during the pre-index period: (i) MACE [a composite event of the following diseases (ii–v)], (ii) AMI, (iii) angina, (iv) HF, and (v) stroke. The five cohorts (MACE cohort, AMI cohort, angina cohort, HF cohort, and stroke cohort) were used for the analysis separately. For inter/intragroup subgroup analyses, the patients were classified into subgroups based on age (65–74, ⩾75), sex (male, female), and race/ethnicity (Hispanic patients, non-Hispanic Asian/Pacific Islander patients, non-Hispanic Black patients, non-Hispanic White patients, others).

Exposure/outcomes

Patients who used osimertinib and first/second-generation EGFR-TKIs (gefitinib, erlotinib, afatinib) were classified as osimertinib group and first/second-generation EGFR-TKI group, respectively. Medication usage was identified using national drug codes. According to the five-point MACE,²² we assessed the incident MACE, AMI, angina, HF, and stroke as the primary outcomes using ICD-9-CM and ICD-10-CM (Supplemental Table 1).

Covariates

Socioeconomic (e.g. sex, age group, race/ethnicity, census tract median income, geographic region, urbanicity, plan benefits, marital status) and clinical [e.g. year of diagnosis, stage, tumor histology, tumor size at diagnosis, radiation, surgery, other cancers during the pre-index period, other cancers during the post-index period, Charlson comorbidity index (CCI), tobacco use disorder] characteristics were used for propensity score (PS) calculation. The morbidities included in calculating CCI were selected based on the 2021 version of CCI defined by NCI.^31,32 The preexisting MACEs other than the outcome of interest in each cohort were included as covariates in the additional intragroup analysis (Supplemental Method 2).

Statistical analysis

The continuous variables were presented as the mean with standard deviation (SD) or median with interquartile range. The categorical variables were presented with the number of patients and percentage. PS, the probability of being in the osimertinib group, was calculated with logistic regression using the covariates explained earlier. Model discrimination, one of the key elements indicating the performance of logistic regression, was assessed using the concordance statistic (i.e. c-statistic).³³ C-statistics between 0.6 and 0.8 in the fitted logistic regression model were considered to indicate adequate model discrimination.³⁴ After PS estimation, we trimmed 1% of patients who were at the extremes of the PS in each group.³⁵ Using PS, we conducted inverse probability of treatment weight (IPTW) by applying 1/PS as weight in the osimertinib group and 1/(1 − PS) as weight in the first/second-generation EGFR-TKI group. Dependent variable of interest in this study, which is the incidence of MACE, is assumed to occur independently and have a constant probability of occurrence at any time point. Then the number of incident MACE within a time interval follows a Poisson distribution.³⁶ Therefore, the crude incidence rate (IR) and its 95% confidence interval (CI) were calculated with a weighted number of patients and weighted person-years using a generalized linear model with Poisson distribution and log link function. To compare the effect of whether the patients used osimertinib or first/second-generation EGFR-TKIs on the incidence and time to the incident MACE, which is a time-to-event outcome, the Cox proportional hazards model was chosen to obtain the hazard ratio (HR) and 95% CI. To use the Cox proportional hazards model, we used Schoenfeld’s residual test to test the proportional hazard assumptions. Standardized mean difference (SMD) was presented to show the balance of characteristics between the two groups before and after IPTW. Covariates with SMD ⩾0.1 in each cohort were included as covariates in Cox proportional hazard model for HR estimation.

For intergroup subgroup analyses, we calculated PS for each subgroup and conducted 1% trimming and IPTW based on the PS of each subgroup. The weight estimated from IPTW was applied to each subgroup to estimate the HR of MACEs between osimertinib and first/second-generation EGFR-TKI users. For the intragroup subgroup analysis within the osimertinib group, HRs of MACEs between subgroups [age: ⩾75 (reference: 65–74); sex: female (reference: male); race/ethnicity: Hispanic patients, non-Hispanic Black patients, non-Hispanic White patients (reference: non-Hispanic Asian/Pacific Islanders patients)] were estimated without applying weight.

Results

Patient characteristics

Of the 235,223 patients who were aged ⩾65 and diagnosed with advanced NSCLC between 2007 and 2017, 563 and 1175 patients were included in the osimertinib group and first/second-generation EGFR-TKI group, respectively (Supplemental Figure 2). After excluding those with a previous diagnosis of outcome in each cohort and trimming 1% of patients with extreme PS, 448 and 851 patients were included in osimertinib and first/second-generation EGFR-TKI group, respectively, in the MACE cohort.

In the MACE cohort, before IPTW, the osimertinib group had a larger proportion of females (71.0% versus 65.7%), patients at AJCC stage IV or SEER stage distant (87.3% versus 79.4%), patients with adenocarcinoma (96.9% versus 89.8%) than first/second-generation EGFR-TKI group (Table 1). The mean (SD) age was 75.9 (6.2) and 76.7 (6.9) in osimertinib and first/second-generation EGFR-TKI group, respectively. More than half of the patients were non-Hispanic White, followed by non-Hispanic Asian/Pacific Islanders. After IPTW, the number of patients increased to 1249 and 1282 in the osimertinib and first/second-generation EGFR-TKI group, respectively. The characteristics between the two groups were balanced after IPTW. Baseline characteristics of AMI, angina, HF, and stroke cohorts are described in Supplemental Tables 2–5. The characteristics of each subgroup in the MACE cohort before IPTW are described in Supplemental Tables 6–8.

Table 1.

Socioeconomic/clinical characteristics of patients without previous MACE before and after inverse probability of treatment weighting.

Characteristic	Before IPTW			AFTER IPTW
	Osimertinib	First/second-generation EGFR-TKI	SMD	Osimertinib	First/second-generation EGFR-TKI	SMD
Total, N	448	851		1249	1282
Socioeconomic demographic
Sex			−0.11			−0.06
Male, N (%)	130 (29.0)	292 (34.3)		373 (29.8)	419 (32.7)
Female, N (%)	318 (71.0)	559 (65.7)		876 (70.2)	863 (67.3)
Age at diagnosis, years			−0.12			−0.02
Mean (SD)	75.9 (6.2)	76.7 (6.9)		76.3 (10.9)	76.5 (8.3)
Median (IQR)	75 (71–80)	76 (71–82)		75 (71–81)	76 (71–82)
Age group, years, N (%)			0.19			0.04
65–69	77 (17.2)	148 (17.4)		214 (17.1)	227 (17.7)
70–74	125 (27.9)	219 (25.7)		335 (26.8)	332 (25.9)
75–79	127 (28.4)	192 (22.6)		318 (25.5)	312 (24.3)
80+	119 (26.6)	292 (34.3)		382 (30.6)	411 (32.1)
Race, N (%)^a			0.09			0.06
Hispanic	36 (8.0)	64 (7.5)		90 (7.2)	98 (7.7)
Non-Hispanic Asian/Pacific Islanders	142 (31.7)	274 (32.2)		406 (32.5)	412 (32.1)
Non-Hispanic Black	20 (4.5)	49 (5.8)		66 (5.3)	70 (5.5)
Non-Hispanic White	250 (55.8)	462 (54.3)		686 (55.0)	700 (54.6)
Census tract median income, N (%)^a			0.24			0.10
Q1 (least wealthy)	86 (19.2)	219 (25.7)		279 (22.4)	304 (23.7)
Q2	112 (25.0)	223 (26.2)		327 (26.2)	333 (26.0)
Q3	114 (25.5)	210 (24.7)		300 (24.0)	323 (25.2)
Q4	124 (27.7)	194 (22.8)		319 (25.6)	312 (24.3)
Geographic region, N (%)			0.26			0.06
Northeast	173 (38.6)	271 (31.8)		446 (35.7)	431 (33.6)
Midwest	20 (4.5)	32 (3.8)		49 (3.9)	51 (4.0)
South	26 (5.8)	107 (12.6)		114 (9.1)	134 (10.5)
West	229 (51.1)	441 (51.8)		640 (51.3)	666 (52.0)
Rural/urban, N (%)^a			0.09			0.04
Metropolitan	429 (95.8)	799 (93.9)		1186 (95.0)	1212 (94.5)
Urban	19 (4.2)	51 (6.0)		63 (5.1)	69 (5.4)
Plan benefits, N (%)			0.19			0.06
Medicaid	28 (6.3)	81 (9.5)		92 (7.4)	108 (8.4)
Private insurance	90 (20.1)	216 (25.4)		279 (22.3)	307 (24.0)
Others	330 (73.7)	554 (65.1)		878 (70.3)	867 (67.6)
Marital status, N (%)			0.24			0.07
Married	209 (46.7)	397 (46.7)		586 (47.0)	603 (47.1)
Not married	112 (25.0)	285 (33.5)		357 (28.6)	397 (30.9)
Unknown	127 (28.4)	169 (19.9)		305 (24.5)	282 (22.0)
Clinical characteristics
Year of diagnosis, N (%)			0.39			0.08
2007–2010	21 (4.7)	72 (8.5)		87 (7.0)	92 (7.2)
2011	18 (4.0)	49 (5.8)		66 (5.3)	67 (5.2)
2012	23 (5.1)	91 (10.7)		91 (7.3)	118 (9.2)
2013	64 (14.3)	113 (13.3)		179 (14.3)	178 (13.9)
2014	93 (20.8)	100 (11.8)		197 (15.7)	181 (14.1)
2015	105 (23.4)	160 (18.8)		259 (20.8)	262 (20.5)
2016	72 (16.1)	182 (21.4)		242 (19.4)	253 (19.7)
2017	52 (11.6)	84 (9.9)		128 (10.2)	130 (10.2)
Stage, N (%)			−0.21			−0.03
IIIB (AJCC) or regional (SEER stage)	57 (12.7)	175 (20.6)		16 (17.4)	60 (18.5)
IV (AJCC) or distant (SEER stage)	391 (87.3)	676 (79.4)		1034 (82.6)	1048 (81.5)
Tumor histology, N (%)^a			0.31			0.20
Adenocarcinoma	434 (96.9)	764 (89.8)		1200 (96.1)	1179 (92.0)
Tumor size at diagnosis, cm, N (%)			0.22			0.08
0.1–2.0	18 (4.0)	62 (7.3)		67 (5.3)	81 (6.3)
2.1–3.0	50 (11.2)	123 (14.5)		157 (12.5)	172 (13.4)
3.1–4.0	75 (16.7)	126 (14.8)		194 (15.6)	196 (15.3)
4.1–5.0	44 (9.8)	87 (10.2)		126 (10.1)	132 (10.3)
5.1–6.0	77 (17.2)	164 (19.3)		218 (17.4)	243 (19.0)
Unknown	184 (41.1)	289 (34.0)		488 (39.1)	459 (35.8)
Radiation, N (%)			−0.02			−0.04
Yes	115 (25.7)	227 (26.7)		312 (25.0)	344 (26.8)
No	333 (74.3)	624 (73.3)		937 (75.0)	939 (73.2)
Surgery of primary cancer site, N (%)			−0.08			−0.03
Yes	40 (8.9)	97 (11.4)		125 (10.0)	139 (10.8)
No	408 (91.1)	754 (88.6)		1124 (90.0)	1143 (89.2)
Other cancers during pre-index period, N (%)			0.01			0.00
Yes	125 (27.9)	235 (27.6)		345 (27.6)	353 (27.6)
No	323 (72.1)	616 (72.4)		904 (72.4)	929 (72.4)
Other cancers during post-index period, N (%)			0.04			0.04
Yes	143 (31.9)	255 (30.0)		401 (32.1)	388 (30.2)
No	305 (68.1)	596 (70.0)		848 (67.9)	895 (69.8)
Charlson comorbidity index			−0.11			−0.04
Mean (SD)	1.38 (1.5)	1.55 (1.7)		1.4 (2.5)	1.5 (2.0)
Median (IQR)	1 (0–2)	1 (0–2)		1 (0–2)	1 (0–2)
Charlson comorbidity index, N (%)			0.08			0.05
0	207 (46.2)	399 (46.9)		582 (46.6)	599 (46.7)
1	104 (23.2)	188 (22.1)		286 (22.9)	287 (22.4)
2	70 (15.6)	116 (13.6)		191 (15.3)	182 (14.2)
3+	67 (15.0)	148 (17.4)		190 (15.2)	214 (16.7)
Tobacco use disorder, N (%)			−0.09			−0.02
Yes	25 (5.6)	66 (7.8)		80 (6.4)	88 (6.9)
No	423 (94.4)	785 (92.2)		1168 (93.6)	1195 (93.2)

The table is not containing certain categories with small number (<11) of patients.

AJCC, American Joint Committee on Cancer; EGFR-TKI, epidermal growth factor tyrosine kinase inhibitor; IPTW, inverse probability of treatment weighting; IQR, interquartile range; MACE, major adverse cardiovascular event; SD, standard deviation; SEER, surveillance epidemiology and end results; SMD, standardized mean difference.

Incidence rate

The IR of MACE was 28.22 (95% CI, 25.69–31.00) in the osimertinib group, and 28.95 (95% CI, 26.28–31.89) in the first/second-generation EGFR-TKI group (Table 2). The IRs of AMI and HF were higher in osimertinib group [AMI: IR, 6.21 (95% CI, 5.25–7.34); HF: IR, 15.07 (95% CI, 13.43–16.92)] compared to first/second-generation EGFR-TKI group [AMI: IR, 4.96 (95% CI, 4.09–6.03); HF: IR, 12.84 (95% CI, 11.29–14.61)]. Conversely, angina and stroke had higher IR in the first/second-generation EGFR-TKI group [angina: IR, 2.26 (95% CI, 1.68–3.02); stroke: IR, 16.58 (95% CI, 14.75–18.65)] than osimertinib group [angina: IR, 0.75 (95% CI, 0.46–1.22); stroke: IR, 14.65 (95% CI, 13.01–16.50)]. The IRs of MACE, AMI, angina, HF, and stroke in age, sex, and race/ethnicity subgroups are presented in Supplemental Table 9.

Table 2.

Number of events, person-years, and IR of MACE in osimertinib and first/second-generation EGFR-TKIs group.

	Osimertinib			First/second-generation EGFR-TKIs
	No. events (No. total patients)	Person-years	IR^a (95% CI)	No. events (No. total patients)	Person-years	IR^a (95% CI)
MACE	435 (1249)	1542	28.22 (25.69–31.00)	411 (1282)	1421	28.95 (26.28–31.89)
Acute myocardial infarction	137 (1584)	2200	6.21 (5.25–7.34)	101 (1627)	2043	4.96 (4.09–6.03)
Angina	17 (1582)	2205	0.75 (0.46–1.22)	45 (1626)	1991	2.26 (1.68–3.02)
Heart failure	288 (1445)	1913	15.07 (13.43–16.92)	232 (1482)	1809	12.84 (11.29–14.61)
Stroke	272 (1403)	1858	14.65 (13.01–16.50)	279 (1453)	1682	16.58 (14.75–18.65)

The IR was calculated with the weighted number of events and the weighted person-years. The IR presented in the table is crude IR without adjustments for covariates.

The unit of IR is number of events/100 person-years.

CI, confidence interval; EGFR-TKI, epidermal growth factor tyrosine kinase inhibitor; IR, incidence rate; MACE, major adverse cardiovascular disease.

Hazard ratio

The risk of MACE was not significantly different between the osimertinib and first/second-generation EGFR-TKI users of all ages [HR, 0.99, (95% CI, 0.86–1.13)] (Figure 1). However, in the intragroup subgroup analysis with osimertinib users, patients aged ⩾75 had a significantly higher risk of MACE than patients aged 65–74 with an HR of 1.58 (95% CI, 1.11–2.23) (Figure 2).

Figure 1.

The risk of major adverse cardiovascular events in osimertinib group compared to first/second-generation EGFR-TKI group.

Figure 2.

The risk of major adverse cardiovascular events between age, sex, race/ethnicity subgroups within osimertinib users.

The risk of AMI was elevated in osimertinib group compared to the first/second-generation EGFR-TKI group, but not significantly with an HR of 1.21 (95% CI, 0.94–1.57). In the male subgroup, osimertinib users had a significantly higher risk of AMI than the first/second-generation EGFR-TKI group with an HR of 1.98 (95% CI, 1.13–3.48) (Figure 1). Within the osimertinib group, non-Hispanic White and non-Hispanic Black had 3.23 [HR, 3.23 (95% CI, 1.34–7.80)] and 5.63 [HR, 5.63 (95% CI, 1.23–25.87)] times higher risk of AMI than non-Hispanic Asian/Pacific Islanders (Figure 2).

The risk of angina was significantly lower in the osimertinib group compared to the first/second-generation EGFR-TKI group [HR, 0.36 (95% CI, 0.20–0.64)]. In subgroup analyses on patients aged 65–74 [HR, 0.18 (95% CI, 0.03–0.92)], female [HR, 0.30 (95% CI, 0.14–0.64)], and non-Hispanic Asian/Pacific Islanders [HR, 0.11 (95% CI, 0.03–0.38)], osimertinib was associated with even lower risk of angina than first/second-generation EGFR-TKIs (Figure 1). The risk of angina between subgroups within osimertinib users did not differ significantly (Figure 2).

The risk of HF was significantly higher in the osimertinib group than in the first/second-generation EGFR-TKI group [HR, 1.20 (95% CI, 1.01–1.42)]. The risk of HF in osimertinib users compared to first/second-generation EGFR-TKI users was even higher in the subgroup of patients aged ⩾75 [HR, 1.39 (95% CI, 1.12–1.72)], female [HR, 1.40 (95% CI, 1.13–1.73)], and non-Hispanic White [HR, 1.44 (95% CI, 1.15–1.82)] (Figure 1). Within osimertinib users, patients aged ⩾75 (versus 65–74) had a significantly higher risk of HF [HR, 1.71 (95% CI, 1.07–2.71)] and non-Hispanic White (versus non-Hispanic Asian/Pacific Islanders) had 1.8 times higher risk of HF [HR, 1.80 (95% CI, 1.04–3.12)] (Figure 2).

The risk of stroke was not significantly different between osimertinib users and first/second-generation EGFR-TKI users. Intragroup subgroup analysis with osimertinib users also showed no significant difference in stroke risk between subgroups (Figures 1 and 2).

In the additional intragroup analysis, preexisting AMI and angina had an association with the increased risk of HF [AMI: HR, 2.13 (95% CI, 1.27–3.60); Angina: HR, 1.55 (95% CI, 1.02–2.31)] and stroke [AMI: HR, 1.69 (95% CI, 1.05–2.72); Angina: HR, 1.69 (95% CI, 1.05–2.72)] in the first/second-generation EGFR-TKI users (Supplemental Table 10).

Discussion

In a previous study based on FDA Adverse Events Reporting System data, osimertinib had a higher but statistically insignificant reported odds ratio of myocardial infarction compared to first/second-generation EGFR-TKIs [reported odds ratio, 1.2 (95% CI, 0.8–3.3)].⁸ Similarly, in our study, osimertinib was associated with a higher risk of AMI but without statistical significance, with an HR of 1.21 (95% CI, 0.94–1.57) than first/second-generation EGFR-TKIs in the overall population and the elevated risk of AMI was statistically significant in the male subgroup. Moreover, consistent with significantly higher reported odds ratio of HF in the previous study [reported odds ratio, 2.2 (95% CI, 1.5–3.2)],⁸ our study showed a 1.2 times higher risk of HF with osimertinib compared to first/second-generation EGFR-TKIs, with even more elevated risk in the female subgroup. Higher risk of osimertinib-associated AMI in male and HF in female patients shown in our study is consistent with a higher risk of AMI in male³⁷ and HF in female^38,39 of the general population. In contrast to AMI and HF, the risk of angina was significantly lower in osimertinib when compared to older-generation EGFR-TKIs. Future studies would be needed to address the discrepancy between the osimertinib-associated risk of AMI and angina in our study; however, our results are highly suggestive of an increased risk of acute plaque rupture instead of progressive coronary disease when using osimertinib.⁴⁰

Possible mechanisms of TKI-induced cardiotoxicity include TKI’s negative effects on vascular endothelial cells and cardiac postmitotic cells such as cardiomyocytes.^41,42 Cardiomyocytes have downregulated oxidative phosphorylation and subsequent mitochondrial energetic defects after TKI use.⁴³ The mechanism underlying the differential cardiotoxicity profiles of TKIs between male and female patients has not been clearly studied, but a higher level of cardiac apoptosis, worsened myofiber structure, and lower cardiac actin level after TKI use in males compared to females could be related to the observed differential risk by sex.⁴⁴ Future studies focusing on the mechanisms of EGFR-TKI or osimertinib-associated cardiotoxicity are needed to understand the increased cardiovascular risk in osimertinib, particularly in specific subgroups.

As advanced age is a known risk factor of HF and MACE in the general population,^45,46 our study also showed the significantly elevated HF risk of osimertinib compared to other EGFR-TKIs in patients aged ⩾75. The significantly higher risk of MACE and HF in osimertinib users aged ⩾75 than osimertinib users aged 65–74 in our intragroup analysis also show that advanced age could also be associated with increased risk of MACE in patients with NSCLC who are treated with osimertinib. Racial/ethnic differences are known to exist in cardiovascular risk.⁴⁷ Consistent with that non-Hispanic Asian/Pacific Islanders have a lower risk of AMI in the general population,⁴⁸ non-Hispanic Asian/Pacific Islanders showed significantly lower risk of AMI than non-Hispanic White/Black among osimertinib users in our study. Considering other ethnicity/race groups not only have an increased risk of AMI but also have a higher mortality rate than Asian when having AMI,⁴⁹ careful monitoring of the management of AMI, particularly in race/ethnicity groups with higher risk of AMI will be needed when prescribing osimertinib.

Contrary to the increased risk of hospitalization and death due to HF in Black patients of the general population,⁴⁵ our study showed among osimertinib users, non-Hispanic Black patients have a nonsignificant but lower risk of HF than non-Hispanic Asian/Pacific Islander patients, and non-Hispanic White patients have a significantly higher risk of HF than non-Hispanic Asian/Pacific Islander patients. The inconsistency in racial/ethnic differences in HF risk between the general population and osimertinib users might have resulted from the racial/ethnic distribution of osimertinib users. Unlike the general NSCLC population in the U.S., which consists mainly of White (75.2%), followed by Black (12.1%), Asian/Pacific Islanders (6.3%), and Hispanic (5.8%) patients,⁵⁰ our results show osimertinib users are mainly made up of non-Hispanic White (55.8%) and non-Hispanic Asian/Pacific Islanders (31.7%), followed by Hispanic (8.0%) and non-Hispanic Black (4.5%) patients. The racial/ethnic differences in genetic mutation in osimertinib users (EGFR, T790m), characterized by higher prevalence in Asian,⁵¹ might have resulted in more observation of HF in non-Hispanic Asian/Pacific Islanders patients in our study.

To the best of our knowledge, this is the first pharmacoepidemiologic study to use nationwide administrative data to retrospectively compare the risk of MACE after using osimertinib compared to first/second-generation EGFR-TKIs in older patients with advanced NSCLC. Aligning with current management and monitoring practices for TKI-induced cardiotoxicity, such as recommending caution when concomitantly using cardiotoxic drugs or in patients with a history of CVD, and conducting regular monitoring of cardiac function during treatment as stated in the FDA labeling,⁵² the need for considering cardiotoxicity in the osimertinib use is being raised.⁵³ The assessment of each MACE risk and the stratified results by sex, age, and race/ethnicity in our study could assist in planning closer monitoring of MACE for those at a higher risk when using osimertinib. As monitoring of CVD during cancer treatment can help manage the adverse cardiac impact, it is recommended to manage MACE early in EGFR-TKI treatment through collaboration between cardiologists and oncologists.⁵⁴

There are several limitations to our study. First, our study used the disease code of angina and not that of unstable angina due to a small number of events of unstable angina. However, because the disease code of angina includes unstable angina, our study results include the occurrence of unstable angina. Second, due to the small sample size, we were not able to examine the HRs of several race/ethnicity subgroups. Future studies with larger sample sizes of those race/ethnicity groups might be able to compare the risk differences by race/ethnicity with estimated HR and 95% CIs. Third, older patients are more likely to be prescribed multiple medications⁵⁵ that can induce CVD,⁵⁶ but we did not include medication history as a covariate in the analysis. However, we balanced the two groups with their clinical characteristics, including baseline comorbidity; therefore, the clinical status of the patients could have been balanced with the IPTW. Fourth, it should be taken into account when interpreting our results that tobacco use disorder that was included as a covariate when estimating PS has low sensitivity caused by incompleteness in measuring tobacco use disorder in SEER.⁵⁷ Fifth, as cardiovascular adverse event of osimertinib has received attention,^6,7 the ascertainment bias could possibly exist in our study. Potential association between the risk of MACE and osimertinib suggested in our study could be studied for causation in future studies. Lastly, our study results are limited to patients aged 65 and older who are enrolled in Medicare. However, considering that similar results regarding the risk of AMI and HF have been found between the previous reports from the FDA Adverse Events Reporting System that does not restrict patients’ ages and our study, it could be inferred that our study results might be applicable to patients younger than 65 years.

Conclusion

Osimertinib is associated with a significantly higher risk of HF, but a lower risk of angina than first/second-generation EGFR-TKIs. Patients who are female, aged ⩾75, and non-Hispanic White had an increased risk of HF while male patients had a higher risk of AMI when treated with osimertinib compared to first/second-generation EGFR-TKIs. It would be recommended to consider the increased risk of cardiovascular adverse events when prescribing osimertinib particularly in high-risk patients. By investigating an in-depth examination of osimertinib-induced cardiotoxicity, this study provides evidence for individualized cardiotoxicity surveillance, prevention, and treatment for patients being treated with osimertinib. Further investigation into the mechanistic pathways underlying osimertinib-induced cardiotoxicity and its potential modulation by patient demographics and comorbidities could elucidate tailored strategies for managing cardiovascular risk in patients receiving osimertinib.

Supplemental Material

sj-docx-1-tam-10.1177_17588359241264721 – Supplemental material for Sex-specific risk of heart failure and acute myocardial infarction associated with osimertinib in older patients with non-small cell lung cancer

Supplemental material, sj-docx-1-tam-10.1177_17588359241264721 for Sex-specific risk of heart failure and acute myocardial infarction associated with osimertinib in older patients with non-small cell lung cancer by Joo-Young Byun, Sola Han, Yan Liu and Chanhyun Park in Therapeutic Advances in Medical Oncology

Footnotes

Acknowledgements

The funders had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit manuscript for publication. This study used the linked SEER-Medicare database. The interpretation and reporting of these data are the sole responsibility of the authors. The authors acknowledge the efforts of the National Cancer Institute; Information Management Services (IMS), Inc.; and the SEER Program tumor registries in the creation of the SEER-Medicare database. The collection of cancer incidence data used in this study was supported by the California Department of Public Health pursuant to California Health and Safety Code Section 103885; Centers for Disease Control and Prevention’s (CDC) National Program of Cancer Registries, under cooperative agreement 1NU58DP007156; the National Cancer Institute’s SEER Program under contract HHSN261201800032I awarded to the University of California, San Francisco, contract HHSN261201800015I awarded to the University of Southern California, and contract HHSN261201800009I awarded to the Public Health Institute. The ideas and opinions expressed herein are those of the author(s) and do not necessarily reflect the opinions of the State of California, Department of Public Health, the National Cancer Institute, and the Centers for Disease Control and Prevention or their Contractors and Subcontractors.

Declarations

ORCID iD

Chanhyun Park

Supplemental material

Supplemental material for this article is available online.

References

Chen

Feng

Zhou

, et al. Quality of life (QoL) analyses from OPTIMAL (CTONG-0802), a phase III, randomised, open-label study of first-line erlotinib versus chemotherapy in patients with advanced EGFR mutation-positive non-small-cell lung cancer (NSCLC). Ann Oncol 2013; 24: 1615–1622.

Maemondo

Inoue

Kobayashi

, et al. Gefitinib or chemotherapy for non-small-cell lung cancer with mutated EGFR. N Engl J Med 2010; 362: 2380–2388.

Ramalingam

Vansteenkiste

Planchard

, et al. Overall survival with osimertinib in untreated, EGFR-mutated advanced NSCLC. N Engl J Med 2020; 382: 41–50.

Zaborowska-Szmit

Krzakowski

Kowalski

, et al. Cardiovascular complications of systemic therapy in non-small-cell lung cancer. J Clin Med 2020; 9: 1268.

Chitturi

Burns

Muhsen

, et al. Cardiovascular risks with epidermal growth factor receptor (EGFR) tyrosine kinase inhibitors and monoclonal antibody therapy. Curr Oncol Rep 2022; 24: 475–491.

Kunimasa

Kamada

Oka

, et al. Cardiac adverse events in EGFR-mutated non-small cell lung cancer treated with osimertinib. JACC CardioOncol 2020; 2: 1–10.

Ewer

Tekumalla

Walding

, et al. Cardiac safety of osimertinib: a review of data. J Clin Oncol 2021; 39: 328–337.

Anand

Ensor

Trachtenberg

, et al. Osimertinib-induced cardiotoxicity: a retrospective review of the FDA Adverse Events Reporting System (FAERS). JACC CardioOncol 2019; 1: 172–178.

Kravchenko

Berry

Arbeev

, et al. Cardiovascular comorbidities and survival of lung cancer patients: medicare data based analysis. Lung Cancer 2015; 88: 85–93.

10.

Maggard

Livingston

EH.

Evaluating health utility in patients with melanoma, breast cancer, colon cancer, and lung cancer: a nationwide, population-based assessment. J Surg Res 2003; 114: 1–5.

11.

Korsnes

Davis

Ariely

, et al. Health care resource utilization and costs associated with nonfatal major adverse cardiovascular events. J Manag Care Spec Pharm 2015; 21: 443–450.

12.

Presley

Reynolds

Langer

CJ.

Caring for the older population with advanced lung cancer. Am Soc Clin Oncol Educ Book 2017; 37: 587–596.

13.

Damluji

Chung

Xue

, et al. Frailty and cardiovascular outcomes in the National Health and Aging Trends Study. Eur Heart J 2021; 42: 3856–3865.

14.

Rich

Chyun

Skolnick

, et al. Knowledge gaps in cardiovascular care of the older adult population: a scientific statement from the American Heart Association, American College of Cardiology, and American Geriatrics Society. Circulation 2016; 133: 2103–2122.

15.

Reddy

Shenoy

Blaes

AH.

Cardio-oncology in the older adult. J Geriatr Oncol 2017; 8: 308–314.

16.

Wilcox

Rotz

Mullen

, et al. Sex-specific cardiovascular risks of cancer and its therapies. Circ Res 2022; 130: 632–651.

17.

Zhu

Shi

Jiang

, et al. Racial and ethnic disparities in all-cause and cardiovascular mortality among cancer patients in the U.S. JACC CardioOncol 2023; 5: 55–66.

18.

Atkins

Chaunzwa

Lamba

, et al. Association of left anterior descending coronary artery radiation dose with major adverse cardiac events and mortality in patients with non-small cell lung cancer. JAMA Oncol 2021; 7: 206–219.

19.

Chen

Liu

Tseng

, et al. Major adverse cardiovascular events in patients with renal cell carcinoma treated with targeted therapies. JACC CardioOncol 2022; 4: 223–234.

20.

Sharma

Pagidipati

Califf

, et al. Impact of regulatory guidance on evaluating cardiovascular risk of new glucose-lowering therapies to treat type 2 diabetes mellitus: lessons learned and future directions. Circulation 2020; 141: 843–862.

21.

Marx

McGuire

Perkovic

, et al. Composite primary end points in cardiovascular outcomes trials involving type 2 diabetes patients: should unstable angina be included in the primary end point? Diabetes Care 2017; 40: 1144–1151.

22.

Bosco

Hsueh

McConeghy

, et al. Major adverse cardiovascular event definitions used in observational analysis of administrative databases: a systematic review. BMC Med Res Methodol 2021; 21: 241.

23.

Herbach

O’Rorke

Carnahan

, et al. Cardiac adverse events associated with chemo-radiation versus chemotherapy for resectable stage III non-small-cell lung cancer: a surveillance, epidemiology and end results-medicare study. J Am Heart Assoc 2022; 11: e027288.

24.

Jacobs

JEJ

L’Hoyes

Lauwens

, et al. Mortality and major adverse cardiac events in patients with breast cancer receiving radiotherapy: the first decade. J Am Heart Assoc 2023; 12: e027855.

25.

National Cancer Institute. Surveillance, epidemiology, and end results program, https://seer.cancer.gov/about/overview.html#:~:text=SEER%20currently%20collects%20and%20publishes,percent%20of%20the%20U.S.%20population (accessed 22 February 2023).

26.

National Cancer Institute. SEER-Medicare: Brief description of the SEER-Medicare database, https://healthcaredelivery.cancer.gov/seermedicare/overview/ (accessed 22 February 2023).

27.

Warren

Klabunde

Schrag

, et al. Overview of the SEER-Medicare data: content, research applications, and generalizability to the United States elderly population. Med Care 2002; 40: IV-3–18.

28.

von Elm

Altman

Egger

, et al. STROBE Initiative. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) Statement: guidelines for reporting observational studies, https://www.equator-network.org/reporting-guidelines/strobe/ (accessed 27 February 2024).

29.

Lewis

Check

Caporaso

, et al. US lung cancer trends by histologic type. Cancer 2014; 120: 2883–2892.

30.

Linnoila

Pathology of non-small cell lung cancer. New diagnostic approaches. Hematol Oncol Clin North Am 1990; 4: 1027–1051.

31.

National Cancer Institute, Division of Cancer Control and Population Sciences. Comorbidity SAS macro (2021 version), https://healthcaredelivery.cancer.gov/seermedicare/considerations/macro-2021.html (accessed 7 April 2023).

32.

Stedman

Doria-Rose

Warren

, et al. The impact of different SEER-Medicare claims-based comorbidity indexes on predicting non-cancer mortality for cancer patients. Comorbidity Technical Report. National Cancer Institute, https://healthcaredelivery.cancer.gov/seermedicare/considerations/comorbidity-report.pdf (accessed July 1 2024).

33.

Austin

Steyerberg

EW.

Interpreting the concordance statistic of a logistic regression model: relation to the variance and odds ratio of a continuous explanatory variable. BMC Med Res Methodol 2012; 12: 82.

34.

Westreich

Cole

Funk

, et al. The role of the c-statistic in variable selection for propensity score models. Pharmacoepidemiol Drug Saf 2011; 20: 317–320.

35.

Chesnaye

Stel

Tripepi

, et al. An introduction to inverse probability of treatment weighting in observational research. Clin Kidney J 2022; 15: 14–20.

36.

Everitt

Sophia

R-H.

Analyzing medical data using S-PLUS. New York, NY: Springer, 2013.

37.

Millett

ERC

Peters

SAE

Woodward

Sex differences in risk factors for myocardial infarction: cohort study of UK Biobank participants. BMJ 2018; 363: k4247.

38.

Gao

Chen

Sun

, et al. Gender differences in cardiovascular disease. Med Nov Technol Dev 2019; 4: 100025.

39.

Beale

Meyer

Marwick

, et al. Sex differences in cardiovascular pathophysiology: why women are overrepresented in heart failure with preserved ejection fraction. Circulation 2018; 138: 198–205.

40.

Hong

Mintz

Lee

, et al. Comparison of coronary plaque rupture between stable angina and acute myocardial infarction: a three-vessel intravascular ultrasound study in 235 patients. Circulation 2004; 110: 928–933.

41.

Shyam Sunder

Sharma

Pokharel

. Adverse effects of tyrosine kinase inhibitors in cancer therapy: pathophysiology, mechanisms and clinical management. Signal Transduct Target Ther 2023; 8: 262.

42.

Gomez

JA.

Vascular endothelial growth factor-tyrosine kinase inhibitors: novel mechanisms, predictors of hypertension and management strategies. Am Heart J Plus 2022; 17: 100144.

43.

Wang

Sheehan

Palmer

, et al. Adaptation of human iPSC-derived cardiomyocytes to tyrosine kinase inhibitors reduces acute cardiotoxicity via metabolic reprogramming. Cell Syst 2019; 8: 412–426.e7.

44.

Madonna

Pieragostino

Cufaro

, et al. Sex-related differential susceptibility to ponatinib cardiotoxicity and differential modulation of the Notch1 signalling pathway in a murine model. J Cell Mol Med 2022; 26: 1380–1391.

45.

Ziaeian

Kominski

Ong

, et al. National differences in trends for heart failure hospitalizations by sex and race/ethnicity. Circ Cardiovasc Qual Outcomes 2017; 10: e003552.

46.

Helber

Alves

CMR

Grespan

, et al. The impact of advanced age on major cardiovascular events and mortality in patients with ST-elevation myocardial infarction undergoing a pharmaco-invasive strategy. Clin Interv Aging 2020; 15: 715–722.

47.

Gray

Welsh

, et al. Ethnic differences in cardiovascular risk: examining differential exposure and susceptibility to risk factors. BMC Med 2022; 20: 149.

48.

Chi

Kanter

, et al. Trends in acute myocardial infarction by race and ethnicity. J Am Heart Assoc 2020; 9: e013542.

49.

Hammershaimb

Goitia

Gyurjian

, et al. Racial and ethnic differences in risk factors and outcomes in adults with acute myocardial infarction. Perm J 2023; 27: 113–121.

50.

Primm

Zhao

Hernandez

, et al. Racial and ethnic trends and disparities in NSCLC. JTO Clin Res Rep 2022; 3: 100374.

51.

Zhang

Yuan

Wang

, et al. The prevalence of EGFR mutation in patients with non-small cell lung cancer: a systematic review and meta-analysis. Oncotarget 2016; 7: 78985–78993.

52.

Chaar

Kamta

Ait-Oudhia

Mechanisms, monitoring, and management of tyrosine kinase inhibitors-associated cardiovascular toxicities. Onco Targets Ther 2018; 11: 6227–6237.

53.

Piper-Vallillo

Sequist

LV.

Cardiac risk-informed treatment of EGFR-mutant lung cancer with osimertinib. JACC CardioOncol 2019; 1: 179–181.

54.

Curigliano

Lenihan

Fradley

, et al. Management of cardiac disease in cancer patients throughout oncological treatment: ESMO consensus recommendations. Ann Oncol 2020; 31: 171–190.

55.

Davies

Spiers

Kingston

, et al. Adverse outcomes of polypharmacy in older people: systematic review of reviews. J Am Med Dir Assoc 2020; 21: 181–187.

56.

Murphy

Dargie

HJ.

Drug-induced cardiovascular disorders. Drug Saf 2007; 30: 783–804.

57.

National Cancer Institute. SEER-Medicare. Measures that are limited or not available in the data, https://healthcaredelivery.cancer.gov/seermedicare/considerations/measures.html#2 (accessed 26 April 2023).

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.33 MB