Background

Lung cancer is the leading cause of cancer-related mortality in the United States and is projected to account for 127,070 deaths in 2023.¹ Lung cancer is comprised of 2 major histologic subtypes: non–small cell lung cancer (NSCLC) and small cell lung cancer.² The rates of lung cancer by histologic type are influenced by several factors, including sex and race.^3,4 Risk factors for lung cancer include cigarette smoking, radon exposure, secondhand smoke, genetics, and diet, and all play a significant role in the pathogenesis of the disease.⁵ Despite advances in treatment modalities and a decrease in lung cancer incidence, lung cancer remains the leading cause of cancer-related mortality in the United States.⁶ Lung cancer is associated with significant financial burden and economic impairment, which have led to an increasing focus on screening and the development of novel treatments.⁷

Despite the significant burden that lung cancer causes nationally, there are variations in contemporary mortality estimates in the United States. Furthermore, geographic trends in lung cancer mortality are largely understudied, which is especially important in light of recent concerns regarding regional disparities in screening access and compliance.⁸ We analyzed a nationwide database linked to death certificates to describe the trends in lung cancer mortality in the United States from 1999 through 2020.

Patients and Methods

Results

From 1999 through 2020, lung cancer led to 3,380,830 deaths. The AAMR decreased by 42%, from 55.4 in 1999 to 31.8 in 2020, with an associated APC of −2.6% (95% CI, −2.5% to −2.7%). From 1999 through 2011, the AAMR decreased, with an associated APC of −1.55% (95% CI, −1.4% to −1.7%). Following this initial decline, the AAMR decreased from 2011 through 2020 at a rate of −4.2% annually (95% CI, −3.8% to −4.6%) (Figure 1).

View Full Size

Trends in lung cancer AAMRs stratified by sex in the United States from 1999 through 2020.

Overall: 1999–2005 APC, −1.0%^a (95% CI, −0.7% to −1.3%); 2005–2013 APC, −2.3%^a (95% CI, −2.0% to −2.6%); 2013–2020 APC, −4.5%^a (95% CI, −4.1% to −4.8%).

Female: 1999–2006 APC, −0.06% (95% CI, 0.3% to −0.5%); 2006–2014 APC, −2.0%^a (95% CI, −1.7% to −2.2%); 2014–2020 APC, −4.2%^a (95% CI, −3.9% to −4.6%).

Male: 1999–2005 APC, −1.9%^a (95% CI, −1.6% to −2.2%); 2005–2013 APC, −2.9%^a (95% CI, −2.7% to −3.2%); 2013–2020 APC, −5.0%^a (95% CI, −4.8% to −5.2%).

Abbreviations: AAMR, age-adjusted mortality rate; APC, annual percent change.

^aIndicates that the APC is significantly different from 0 at α=0.05.

Citation: Journal of the National Comprehensive Cancer Network 22, 6; 10.6004/jnccn.2024.7004

In 1999, men had an AAMR of 76.8, which was almost twice as high as women, who had an AAMR of 40.2. These differences became less pronounced over time, and in 2020, men had an AAMR of 38.1, whereas women had an AAMR of 26.9. The AAMR for men decreased at a rate of −2.38% (95% CI, −2.2% to −2.7%) annually from 1999 through 2011 compared with women, who experienced a decrease in mortality at a slower APC of −1.0% (95% CI, −1.0% to −1.2%) during the same period. Following the initial decline, a steeper decline in mortality was seen from 2012 through 2020, at a rate of −4.7% (95% CI, −4.5% to −4.9%) annually in men and −3.9% annually (95% CI, −3.9% to −4.3%) in women (Figure 1).

In 1999, non-Hispanic Black or African American individuals experienced the highest AAMR at 65.7, with an annual decrease of −3.0% (95% CI, −2.7% to −3.3%) to 33.4 in 2020, followed by non-Hispanic White individuals, who had an AAMR of 57.2 in 1999, with an annual decrease of −2.5% (95% CI, −2.1% to −2.8%) to 34.9 in 2020—the highest of any subgroup. Non-Hispanic American Indian or Alaska Native individuals experienced an AAMR of 39.9 in 1999, which decreased at an APC of −1.8% (95% CI, −1.2% to −2.4%) to 26.7 in 2020. Hispanic and non-Hispanic Asian or Pacific Islander individuals had similar AAMRs in 1999 of 25.0 and 27.9, respectively, which decreased to 14.1 and 18.5, respectively, in 2020 at rates of −2.7% (95% CI, −2.5% to −3.0%) and −2.0% (95% CI, −1.7% to −2.4%) (Figure 2).

View Full Size

Trends in lung cancer AAMRs stratified by race in the United States from 1999 through 2020.

Hispanic: 1999–2005 APC, −1.5%^a (95% CI, −0.6% to −2.4%); 2005–2014 APC, −2.5%^a (95% CI, −2.0% to −3.0%); 2014–2020 APC, −4.4%^a (95% CI, −3.7% to −5.0%).

NH Black or African American: 1999–2004 APC, −1.6%^a (95% CI, −1.1% to −2.2%); 2004–2013 APC, −2.5%^a (95% CI, −2.3% to −2.8%); 2013–2020 APC, −5.0%^a (95% CI, −4.7% to −5.4%).

NH White: 1999–2007 APC, −1.0%^a (95% CI, −0.7% to −1.2%); 2007–2014 APC, −2.4%^a (95% CI, −2.1% to −2.7%); 2014–2020 APC, −4.4%^a (95% CI, −4.1% to −4.7%).

NH Asian or Pacific Islander: 1999–2013 APC, −1.1%^a (95% CI, −0.9% to −1.4%); 2013–2018 APC, −4.3%^a (95% CI, −3.0% to −5.5%); 2018–2020 APC, −2.4% (95% CI, −6.3% to 1.8%).

NH American Indian or Alaska Native: 1999–2004 APC, −2.0% (95% CI, −1.3% to 5.4%); 2004–2012 APC, −1.0% (95% CI, −2.7% to 0.7%); 2012–2020 APC, −4.2%^a (95% CI, −3.0% to −5.5%).

Abbreviations: AAMR, age-adjusted mortality rate; APC, annual percent change; NH, non-Hispanic.

^aIndicates that the APC is significantly different from 0 at α=0.05.

Citation: Journal of the National Comprehensive Cancer Network 22, 6; 10.6004/jnccn.2024.7004

From 1999 through 2020, people aged 25 to 44 years had the lowest AAMR, followed by those aged 45 to 64 years and those aged ≥65 years, respectively. In those aged 25 to 44 years, the AAMR was lowest, decreasing from 3.5 in 1999 to 1.1 in 2020. From 1999 to 2020, the average APC in young adults was the steepest of any subgroup at −6.2% (95% CI, −5.8% to −6.5%). From 1999 through 2002, this trend was largely stagnant, and then in 2002, the AAMR began to decrease at an APC of −6.6% (95% CI, −6.3% to −6.9%). In those aged 45 to 64 years, the AAMR slowly decreased from 67.4 in 1999 to 44.4 in 2014 at an APC of −2.9% (95% CI, −2.8% to −3.0%); subsequently, the AAMR decreased at a greater APC of −5.3% (95% CI, −4.9% to −5.6%) until 2020. Similar trends were seen for the oldest adults, aged ≥65 years. This group had the highest AAMR across all study years, decreasing from 311.8 in 1999 to 192.6 in 2020. There was a slow initial decline from 1999 to 2010 at an APC of −0.9% (95% CI, −0.7% to −1.2%), with a steeper decline from 2010 to 2020 at an APC of −3.9% (95% CI, −3.6% to −4.1%) (Figure 3).

View Full Size

Trends in lung cancer AAMRs stratified by age in the United States from 1999 through 2020.

Age 25–44 years: 1999–2002 APC, 0.1% (95% CI, −4.8% to 5.3%); 2002–2017 APC, −6.8%^a (95% CI, −7.2% to −6.3%); 2017–2020 APC, −4.5% (95% CI, −9.2% to 0.4%).

Age 45–64 years: 1999–2001 APC, −1.8%^a (95% CI, −3.5% to −0.2%); 2001–2014 APC, −2.9%^a (95% CI, −3.0% to −2.8%); 2014–2020 APC, −5.2%^a (95% CI, −5.5% to −4.9%).

Age ≥65 years: 1999–2005 APC, −0.3% (95% CI, −0.7% to 0.1%); 2005–2012 APC, −1.9%^a (95% CI, −2.2% to −1.5%); 2012–2020 APC, −4.1%^a (95% CI, −4.4% to −3.9%).

Abbreviations: AAMR, age-adjusted mortality rate; APC, annual percent change.

^aIndicates that the APC is significantly different from 0 at α=0.05.

Citation: Journal of the National Comprehensive Cancer Network 22, 6; 10.6004/jnccn.2024.7004

Overall, rural populations had the greatest AAMR, followed by suburban and urban populations. From 1999 through 2020, rural populations experienced the highest AAMR, at 52.3, and the slowest rate of decrease, at an APC of −1.7% (95% CI, −1.4% to −2.0%). Urban populations experienced the lowest AAMR of 40.6 and fastest APC decrease of −3.1% (95% CI, −2.8% to −3.4%), whereas the AAMR of suburban populations was 45.5, with an associated APC decrease of −2.7% (95% CI, −2.4% to −3.0%) (Figure 4).

View Full Size

Trends in lung cancer AAMRs stratified by geographic density^a in the United States from 1999 through 2020.

Rural: 1999–2002 APC, 0.8% (95% CI, −0.5% to 2.2%); 2002–2012 APC, −1.1%^b (95% CI, −1.4% to −0.9%); 2012–2020 APC, −3.4%^b (95% CI, −3.7% to −3.1%).

Suburban: 1999–2005 APC, −0.8%^b (95% CI, −1.2% to −0.5%); 2005–2014 APC, −2.6%^b (−2.8% to −2.4%); 2014–2020 −4.6%^b (95% CI, −4.9% to −4.3%).

Urban: 1999–2004 APC, −1.6%^b (95% CI, −1.9% to −1.3%); 2004–2013 APC, −2.6%^b (95% CI, −2.7% to −2.5%); 2013–2020 APC, −5.3%^b (95% CI, −5.5% to −5.1%).

Abbreviations: AAMR, age-adjusted mortality rate; APC, annual percent change.

^aBased on the 2013 US census classification.

^bIndicates that the APC is significantly different from 0 at α=0.05.

Citation: Journal of the National Comprehensive Cancer Network 22, 6; 10.6004/jnccn.2024.7004

The AAMR from 1999 through 2020 varied widely by state, from 68.6 in Kentucky to 20.7 in Utah. States in the 90th percentile of mortality included Arkansas and Kentucky, whereas states in the 10th percentile of mortality included Hawaii and Utah. The AAMR from 1999 through 2020 was highest in the Midwest (55.7), followed by the South (54.5), Northeast (51.4), and West (37.0). The West experienced the fastest decrease at −3.1% annually (95% CI, −2.8% to −3.3%), whereas the Midwest experienced the slowest decrease at −2.0% annually (95% CI, −1.7% to −2.3%) (Figure 5).

View Full Size

State‐level lung cancer AAMRs in the United States from 1999 through 2020.

Abbreviation: AAMR, age-adjusted mortality rate.

Citation: Journal of the National Comprehensive Cancer Network 22, 6; 10.6004/jnccn.2024.7004

Discussion

In our nationwide study utilizing CDC death certificate data to describe lung cancer mortality in the United States from 1999 through 2020, we found many notable trends. First, after the initial decline in mortality from 1999 through 2011, there was a sharper decline from 2012 through 2020, at a rate of −4.2% each year. Throughout the study period, men had a higher AAMR than women, with a consistently steeper decline in mortality. Non-Hispanic Black or African American individuals experienced the highest AAMR related to lung cancer in 1999; however, these differences became less significant throughout the study period, and in 2020, non-Hispanic White individuals had a higher AAMR than non-Hispanic Black or African American individuals, with similar annual decreases of −3.0% and −2.5%, respectively. Lung cancer–related AAMR was highest in rural counties throughout the study period. Lung cancer mortality was highest in the Mid west and the South and lowest in the West.

The decline in mortality noted in our study may be explained by a decrease in lung cancer incidence.¹⁴ Several studies of national data have demonstrated trends by sex and geography, with varying results.^6,15 A CDC incidence database study exploring trends in lung cancer incidence in metropolitan and nonmetropolitan counties from 2007 through 2016⁶ showed that incidence rates declined in both county types during the study period, but the specific trends differed by sex and geography: men and those residing in nonmetropolitan counties had higher incidence rates. Furthermore, one study examined lung cancer incidence rates and trends by histologic subtype and geographic residence using the United States Cancer Statistics dataset, covering 100% of the US population, from 2004 through 2013.¹⁵ They found that incidence of squamous cell, small cell, and large cell carcinoma had decreased in both metropolitan and nonmetropolitan counties; however, adenocarcinoma incidence has been increasing in both county types. Additionally, incidence was higher in non-Hispanic Black individuals in rural counties than any other subgroup. Thus, the decrease in mortality demonstrated in our study may be explained in part due to a decrease in the incidence of lung cancer. These decreases in incidence may also explain why men experienced a steeper decline than women in lung cancer mortality from 2012 through 2020—men experienced greater decreases than women in incidence during this time period, which may reflect changes in smoking habits.¹⁶ Although clinicians are improving in their ability to detect lung cancer, these decreasing trends and incidence may be driven by a decrease in smoking rates during this time period.^17,18

Additionally, the decrease in lung cancer mortality seen in our study is also explained by advances in treatment and survival. One study used the SEER database to assess lung cancer 2-year survival.¹⁹ It found that 2-year relative survival was higher among women than men and increased in each racial subgroup from 2001 through 2014. These changes may be driven by an evolving treatment paradigm, especially for NSCLC, which comprises roughly 75% of all lung cancer cases. The recent advancement of immune-based therapies targeting PD-1 and PD-L1 have led to an improvement in outcomes. Additionally, the identification of targetable oncogenes, such as EGFR and ALK, has led to the development of targeted therapies that have improved survival. Further studies are ongoing to address the cost-effectiveness of these targeted therapies as well as to determine strategies for early identification of patients who may derive benefit.²⁰

Rural and urban differences in lung cancer incidence and mortality may also be explained by behavioral or environmental factors. A recent analysis demonstrated that mortality due to smoking has declined the fastest in large and coastal cities and the slowest in nonmetropolitan areas of the United States.¹⁸ Furthermore, another analysis of the Behavioral Risk Factor Surveillance System showed that rural areas had the highest smoking rates of any subgroup, with increases in smoking noted in many states in the South.¹⁷ Another well-known risk factor for lung cancer is exposure to environmental factors, such as asbestos or radon. Previous studies have shown that rural areas are more likely to rely on well water or agricultural or mining industries, placing them at a higher risk of exposure to these environmental toxins.²¹ These same groups are also less likely to undergo home testing for radon, exposure to which would increase their lifetime risk of lung cancer.²² These environmental and behavioral factors may help to explain why the AAMR was highest in rural counties, especially those in the South. Additionally, baseline population health may explain the increased AAMR in the Midwest when compared with the West. Smoking prevalence is highest in the Midwest and South and lowest in the Northeast and West, which could explain differences in US census region mortality. Increasing formation of networks may serve to elevate screening access and participation among these at-risk populations.

In our study, we noted higher lung cancer–related mortality in non-Hispanic Black or African American and non-Hispanic White individuals when compared with Hispanic or non-Hispanic Asian or Pacific Islander populations. As mentioned previously, these differences may also be due to incidence: Hispanic populations had the lowest incidence of lung cancer in one study.¹⁵ Furthermore, likelihood of engaging in behavioral risk factors may differ between subgroups. For example, non-Hispanic Black or African American and non-Hispanic White individuals had the highest smoking rates, at 16.8% and 16.6%, respectively, whereas Hispanic and non-Hispanic Asian or Pacific Islander individuals had the lowest smoking rates, at 10.1% and 7.0%, respectively.²³ Additionally, differences in screening rates may explain the increased risk that could drive a higher mortality in certain subgroups. There are currently no large, published studies assessing lung cancer screening rates in different racial groups. However, low quality of provider information regarding lung cancer screening has been proposed as a barrier to undergoing screening in non-Hispanic Black or African American individuals.²⁴ Access to care is another barrier that may explain the higher mortality seen in non-Hispanic Black or African American persons: rural Black patients have reported more difficulty with access to cancer care when compared with White patients.²⁵ In addition, several studies have shown that non-Hispanic Black or African American populations are less likely to receive guideline-concordant care, including slower times to treatment and a lower likelihood of being offered surgical resection.^26–28 These factors may contribute to the higher AAMR seen in non-Hispanic Black or African American individuals. However, the results of our study demonstrate that these disparities may be decreasing over time, with non-Hispanic Black or African American populations experiencing the highest rate of decline at −3.0% annually, which should encourage further positive change.

A number of strategies have been proposed to improve the disparities seen in lung cancer mortality in the United States. Increased use of lung cancer screening has been identified as one method to improve these disparities. However, the existing screening guidelines do not consider racial, ethnic, socioeconomic, and sex-based differences in smoking behaviors or lung cancer risk. This prompted the American Thoracic Society to release a statement describing strategies to improve screening rates in vulnerable populations, including the recommendation that specific health care institutions and organizations should propose quality metrics to address barriers and reduce disparities.²⁹ Such metrics may lead to an increase in screening rates among vulnerable subgroups; however, an increased focus is needed in groups with increased mortality, such as men and rural populations. Fewer than 15% of eligible individuals actually undergo screening, which represents further potential for change. Furthermore, the American Thoracic Society promotes expanded research examining the use of mobile screening to improve access for rural or other at-risk populations to confront geographic barriers. Another strategy to improve these disparities may be to increase focus on patient navigation in order to provide assistance in overcoming barriers to care across the treatment continuum. One study examining the use of a multidisciplinary lung cancer care coordination program demonstrated reduced time to diagnosis and treatment.³⁰ Finally, expansion of telehealth services has been proposed as another strategy to improve access to care among vulnerable groups. Although there are limited data regarding telehealth utilization in lung cancer, one study assessing telehealth for head and neck cancers showed these models of care to be statistically significantly less expensive and more cost-efficient than the standard model of care, with $642.30 in savings per patient.³¹ Thus, there is some feasibility that telehealth may have similar benefits for patients with lung cancer. Following the COVID-19 pandemic, telehealth platforms have expanded and have been found to be useful and effective across different populations.³² Of particular importance is their impact on disadvantaged patient populations, especially rural patients with geographic barriers to care.^32,33

With regard to socioeconomic disparities, our findings tell a positive story in parts of the data. Since 2014, lung cancer mortality rates in non-Hispanic Black or African American and non-Hispanic White populations have been essentially the same, narrowing the mortality gap seen in 1999. These results may be explained by the intersection of a variety of social determinants of health—it is too simple to suggest that race alone is responsible for differences in mortality. One meta-analysis including 131,378 non-Hispanic Black patients found no significant difference in survival when comparing Black and White patients.³⁴ Several studies have been published accounting for other socioeconomic differences between these 2 groups. One study found a decreased risk of mortality after surgery in Black patients,³⁵ whereas another found increased overall survival in Black patients with stage III NSCLC,³⁶ providing potential explanations for the narrowing mortality gap seen over time between these 2 groups. However, the interaction between social determinants of health is complex and multifaceted. This is supported by the fact that the degree of disparity of lung cancer incidence increases with increasing rurality of residence in all racial subgroups, suggesting that disparities in mortality may be similarly driven.¹⁵ Furthermore, living in rural areas increased the risk of death in Black patients by 54%.³⁷ Thus, heterogeneity related to geographic and residential environment may be vitally important drivers of lung cancer mortality.

Our study is not without limitations. One key limitation is that the CDC WONDER database uses data collected from death certificates, which may inflate the nationwide mortality estimates.¹⁹ This is because occult lung cancer cases identified upon death will be counted in the mortality estimates. Despite this, we opted to query the CDC database because it contains geographic data not accounted for in any other national database. Another limitation of this study is that the CDC database strictly reports data from death certificates and does not contain data regarding histologic subtype, time to treatment, treatments received, or other important predictors of mortality, such as comorbid conditions or disease stage. Additionally, because the database reports the state in which death occurred, it is possible that deaths may be misclassified from the state that the individual resided in, which may impact geographic trends. Lastly, it is possible that the database may have misclassified the cause of death on the death certificates reported. However, the CDC WONDER database is a comprehensive source of data that has been widely used in studies assessing cancer and other causes of mortality and undergoes internal validation and quality assurance measures.^9,10,38

Conclusions

Our study demonstrates a significant improvement in lung cancer mortality in the United States from 1999 through 2020. However, although the mortality rate of lung cancer has been decreasing since 1999, not all demographic groups have experienced the same rates of decline, and disparities in outcomes remain prevalent. These disparities are particularly prevalent among non-Hispanic Black or African American individuals and those residing in rural counties. Vulnerable subgroups may benefit from targeted interventions, such as increased utilization of telehealth platforms, academic center outreach and affiliation with community care centers, screening CT scans, and patient navigation while undergoing cancer care.