Abstract
Background:
This study explored female and male overall mortality and lung cancer (LC) survival in two LC screening (LCS) populations, focusing on the predictive value of coronary artery calcification (CAC) at baseline low-dose computed tomography (LDCT).
Methods:
This retrospective study analysed data of 6495 heavy smokers enrolled in the MILD and BioMILD LCS trials between 2005 and 2016. The primary objective of the study was to assess sex differences in all-cause mortality and LC survival. CAC scores were automatically calculated on LDCT images by a validated artificial intelligence (AI) software. Sex differences in 12-year cause-specific mortality rates were stratified by age, pack-years and CAC score.
Results:
The study included 2368 females and 4127 males. The 12-year all-cause mortality rates were 4.1 % in females and 7.7 % in males (p < 0.0001), and median CAC score was 8.7 vs. 41 respectively (p < 0.0001). All-cause mortality increased with rising CAC scores (log-rank test, p < 0.0001) for both sexes. Although LC incidence was not different between the two sexes, females had lower rates of 12-year LC mortality (1.0 % vs. 1.9 %, p = 0.0052), and better LC survival from diagnosis (72.3 % vs. 51.7 %; p = 0.0005), with a similar proportion of stage I (58.1 % vs. 51.2 %, p = 0.2782).
Conclusions:
Our findings demonstrate that female LCS participants had lower rates of all-cause mortality at 12 years and better LC survival than their male counterparts, with similar LC incidence rates and stage at diagnosis. The lower CAC burden observed in women at all ages might contribute to explain their lower rates of all-cause mortality and better LC survival.
Keywords: Female lung cancer, Lung cancer screening, Low-dose computed tomography, All-cause mortality, Artificial intelligence, Coronary artery calcifications
1. Introduction
Over the last century, life expectancy has significantly increased and mortality rates steadily decreased in both males and females, but differences remain between sexes and biological aspects (e.g., hormone levels) seem to explain such a diversity only partly [1,2].
Low-dose computed tomography (LDCT)-based lung cancer screening (LCS) has been able to reduce lung cancer (LC) mortality in heavy smokers [3–5], but the magnitude of benefit was greater [3,4] or only evident [6] in females. Previous studies have investigated tumor-related (e.g., histological type and location) and lifestyle-related (e.g., awareness of risk and screening behavior) factors, but the reasons for the sex gap are still not fully understood [7–10]. Moreover, mortality rates for other causes than LC have not been deeply investigated in LCS populations.
Coronary artery calcification (CAC) is a well-known independent risk factor for major cardio-vascular (CV) events and mortality in both sexes [11–13], but more recent studies suggested that CAC can be considered a valuable risk-stratification tool for all-cause and cancer-related mortality [14,15]. LDCT-based LCS allows for the detection and automatic quantification of CAC [16–19], but only few studies have explored the impact of CAC on LCS outcome [20–23].
Understanding sex-specific mortality rates within the high-risk LCS population may improve prevention strategies by allowing the development of targeted interventions for each sex or risk level.
The present study is a retrospective analysis of two LCS trials that aimed to explore the cause-specific mortality rates of females and males, and to assess the predictive value of a CAC score automatically measured on baseline LDCT images.
2. Material and methods
2.1. Study design
This study was a retrospective analysis of two LCS trials conducted at the Fondazione IRCCS Istituto Nazionale dei Tumori of Milan, namely, the Multicentric Italian Lung Detection trial (MILD, ClinicalTrials.gov Identifier: NCT02837809), which randomized 4099 volunteers to LDCT screening (n = 2376) or observation (n = 1723) from September 2005 to January 2011 [24,25], and the BioMILD (clinicaltrials.gov ID: NCT02247453) trial, which enrolled 4119 volunteers to LDCT screening from January 2013 to March 2016 [26]. The original Institutional Review Board approval and written informed consent allowed the use of data for future research, including the present analyses.
2.2. Study participants
The present analysis included the 2376 volunteers randomized to the LDCT screening arm of MILD trial and the 4119 participants of BioMILD trial. MILD eligiblility criteria were: current or former smokers from < 10 years, ≥20 pack-years, aged 49–75 years, with no history of cancer in the previous 5 years. BioMILD eligiblility criteria were: current or former smokers from < 10 years, ≥30 pack-years, aged 50–75 years, with no history of cancer in the previous 5 years. Details of the MILD and BioMILD study design are reported in the supplementary materials (Section 1).
2.3. Demographic and clinical data
Demographic (e.g., age, sex, marital status, ethnicity, education, employment) and clinical data (e.g., smoking history, prior disease, family history of cancer, drugs used, anthropometric data) were collected through dedicated questionnaires and direct interviews with a study investigator at the baseline screening round. At the same time, all participants underwent spirometry with measurement of forced expiratory volume in 1 s (FEV1%) and blood sample collection for quantification of C-reactive protein (CRP) level.
2.4. Imaging acquisition and evaluation of coronary artery calcification
MILD trial baseline LDCT images were acquired on a 16–detector row CT scanner (Somatom Sensation 16; Siemens Medical Solutions, Forchheim, Germany), while BioMILD trial baseline LDCTs were acquired on a second-generation dual-source CT scanner (Somatom Definition Flash, Siemens Medical Solutions, Forchheim, Germany). Details on LDCT acquisition and imaging interpretation have been reported elsewhere [25,26].
LDCT images were transferred to a dedicated graphic station (Alienware Area 51 R6 equipped with Dual NVIDIA GeForce RTX 2080 ◦C graphics), and CAC was automatically evaluated using commercially available AI-based software (AVIEW, Coreline Soft, Seoul, Korea). CAC was assessed based on the Agatston score and stratified as follows: 0, 1–10, 11–100, 101–400, and > 400 [22,23,27,28].
2.5. Follow-up data
Data on each patient’s vital status and date of death were obtained through the Istituto Nazionale di Statistica (ISTAT, SIATEL 2.0 platform). Participants accumulated person-years of follow-up from the date of LDCT baseline until either death or the date of the last follow-up as of August 2022. Data on the causes of death were collected through direct contact with general practitioners, referring hospitals and Italian cancer registries.
2.6. Study objectives
The primary objectives of the present study were to assess sex differences in all-cause mortality and LC survival, as well as the contribution of CAC score in predicting male and female outcomes. Secondary objectives were to evaluate sex differences and predictivity of CAC in non-cancer mortality, CV mortality and non-CV mortality.
2.7. Statistical analysis
Categorical variables were reported as numbers and percentages, whereas continuous variables were reported as medians with interquartile ranges (IQRs). Measures of association were evaluated by the chi-square test or Fisher’s exact test for categorical data and by the Mann-Whitney U test for continuous variables. Comparisons were performed to evaluate differences between the sexes at baseline and at the last follow-up. Characteristics including age, pack-years, smoking status, body mass index (BMI), CRP, FEV1%, CAC score, self-reported prior CV disease (CVD; myocardial infarction, stroke, thrombosis, or angina), self-reported chronic obstructive pulmonary diasease (COPD) (i.e., emphysema and chronic bronchitis), self-reported diabetes, and PLCOm2012noRace risk score were evaluated [29].
Follow-up was censored at 12 years. Twelve-year all-cause mortality and LC incidence were plotted in Kaplan-Meier curves, and differences between sexes were evaluated by the log-rank test. Different all-cause mortality curves for females and males stratified by CAC score were reported in the Supplementary Material. Multivariate Cox proportional hazard regression models were applied to estimate the 12-year all-cause mortality hazard ratio (HR) and 95 % confidence interval (CI). In model A, differences between males and females (reference category) and age classes were tested; model B was further adjusted for CAC score categories; model C was further adjusted for pack-years, smoking status at baseline, smoking status at the last follow-up, and BMI; model D also included FEV1% and CRP; and model E included self-reported prior diseases categorized into three classes (only one disease among CVD, COPD and diabetes, two conditions reported, and three conditions reported). As a supplementary analysis, similar Cox proportional models were estimated for females and males separately and for volunteers without prior CVD reported. Twelve-year noncancer-, CV-, non-CV- and LC-specific mortalities were estimated by cumulative incidence functions for competing risk, and comparisons between sexes were evaluated with Gray’s test. In the supplementary material multivariate Fine-Gray models were applied to estimate the 12-year LC mortality hazard ratio (HR) and 95 % confidence interval (CI), as described above. LC survival at 5 years stratified by sex was also estimated using the Kaplan-Meier method. All analyses were carried out using Statistical Analysis System Software (Release SAS: 9.04; SAS Institute, Cary, NC).
3. Results
3.1. Participant characteristics
The current study cohort included 6495 participants: 2368 (36.5 %) were female and 4127 (63.5 %) were male (Table 1). Screened females were significantly younger than males (median age: 58 and 59 years, p < 0.0001), with a greater percentage < 55 years (29 % vs. 25.8 %) and a lower percentage ≥ 65 years (18.8 % vs. 22.7 %). A higher percentage of current smokers was observed among females at baseline (81.4 % vs. 71.8 %, p < 0.0001) as well as at the last follow-up (65.8 % vs. 54.9 %, p < 0.0001). Females had a significantly lower median BMI than males (23.5 vs. 26.3, p < 0.0001); more were underweight (3.7 % vs. 0.4 %), and fewer were overweight or obese (35.4 % vs. 66.1 %). Females also had a lower median level of CRP (1.2 mg/L vs. 1.6 mg/L, p < 0.0001) and a higher median FEV1% (98 vs. 96, p < 0.0001). No significant difference in the PLCOm2012noRace score was observed between females and males (p = 0.3700). Females had lower CAC scores: the median score was 8.7 in females vs. 41 in males (p < 0.0001), and only 4.1 % of females had an CAC score > 400 as compared to 15.7 % of males (p < 0.0001).
Table 1.
Baseline demographic and clinical characteristics of 6495 volunteers stratified by sex.
| TOTAL | FEMALES | MALES | P value | ||
|---|---|---|---|---|---|
|
| |||||
| 6495 | 2368 (36.5 %) | 4127 (63.5 %) | |||
| Age (years) | <55 | 1753 (27.0 %) | 687 (29.0 %) | 1066 (25.8 %) | 0.0003a |
| 55–59 | 1753 (27.0 %) | 665 (28.1 %) | 1088 (26.4 %) | ||
| 60–64 | 1606 (24.7 %) | 571 (24.1 %) | 1035 (25.1 %) | ||
| ≥65 | 1383 (21.3 %) | 445 (18.8 %) | 938 (22.7 %) | ||
| Median (IQR) | 59 (54–64) | 58 (54–63) | 59 (54–64) | <0.001b | |
| Pack-years | <35 | 1836 (28.3 %) | 816 (34.5 %) | 1020 (24.7 %) | <0.001a |
| ≥35 | 4659 (71.7 %) | 1552 (65.5 %) | 3107 (75.3 %) | ||
| Smoking status at baseline | Current smokers | 4892 (75.3 %) | 1928 (81.4 %) | 2964 (71.8 %) | <0.001a |
| Former smokers | 1603 (24.7 %) | 440 (18.6 %) | 1163 (28.2 %) | ||
| Smoking status at the last follow-up | Current smokers | 3822 (58.9 %) | 1558 (65.8 %) | 2264 (54.9 %) | <0.001a |
| Former smokers | 2673 (41.2 %) | 810 (34.2 %) | 1863 (45.1 %) | ||
| BMI c | Underweight | 103 (1.6 %) | 88 (3.7 %) | 15 (0.4 %) | <0.001a |
| Normal weight | 2739 (42.2 %) | 1421 (60.0 %) | 1318 (31.9 %) | ||
| Overweight/obese | 3564 (54.9 %) | 837 (35.3 %) | 2727 (66.1 %) | ||
| Median (IQR) | 25.5 (23.0–28.1) | 23.5 (21.2–26.4) | 26.3 (24.3–28.7) | <0.001b | |
| CRP (mg/L) | Median (IQR) | 1.5 (0.7–2.9) | 1.2 (0.6–2.6) | 1.6 (0.8–3.1) | <0.001b |
| FEV1% | Median (IQR) | 97 (87–107) | 98 (88–109) | 96 (86–106) | <0.001b |
| PLCOm2012noRac | Median (IQR) | 2.5 (1.5–4.4) | 2.5 (1.5–4.3) | 2.5 (1.5–4.5) | 0.37b |
| CAC Score d | 0 | 405 (6.2 %) | 203 (8.6 %) | 202 (5.0 %) | <0.001 a |
| 1–10 | 2041 (31.4 %) | 1021 (43.1 %) | 1020 (24.7 %) | ||
| 11–100 | 2127 (32.7 %) | 792 (33.4 %) | 1335 (32.3 %) | ||
| 101–400 | 1046 (16.1 %) | 218 (9.2 %) | 828 (20.1 %) | ||
| >400 | 745 (11.5 %) | 96 (4.1 %) | 649 (15.7 %) | ||
| Median (IQR) | 22 (4–126) | 8.7 (1.9–39.9) | 41 (6–207) | <0.001b | |
| Prior CVD e | 1608 (24.8 %) | 495 (20.9 %) | 1113 (27.0 %) | <0.001a | |
| Prior COPD f | 1047 (16.1 %) | 418 (17.8 %) | 629 (15.4 %) | 0.0155a | |
| Diabetes | 476 (7.3 %) | 72 (3.1 %) | 404 (9.9 %) | <0.001a | |
BMI, body mass index; CRP, C-reactive protein; CVD, cardiovascular disease; COPD, Chronic Obstructive Pulmonary Disease; FEV1%, percentage of the forced expiratory volume in 1 s; IQR, interquartile range.
Chi-square test.
Mann-Whitney U test.
Underweight BMI < 18.5; Normal weight BMI 18.5–24.9; Overweight BMI 25.0–29.9; and Obesity BMI ≥ 30.0; n = 89 missing value.
Missing CAC score n = 131.
Myocardial infarction, stroke, thrombosis, or angina.
Emphysema and chronic bronchitis.
Finally, fewer women than men reported a history of prior CVD (20.9 % vs. 27 %, p < 0.0001) and diabetes (3.1 % vs. 9.9 %, p < 0.0001). However, more women reported a history of COPD (17,8% vs. 15,4%, p = 0.0155) as compared to their men counterparts.
Although women had lower scores than men at the same age, the CAC score increased with age in both sexes (Fig. 1). A similar trend in the CAC score was found with increasing pack-years in both sexes (Fig. S1). The AI software was able to obtain CAC quantification in 6364/6495 (98 %) baseline LDCTs, with only 131 (2 %) missing values. Of interest, the large majority of volunteers (75.2 %) reported no prior CVD, and among them 15 % had a CAC score between 101 and 400, and 8.6 %% >400 (p < 0.0001, Table S1). A significant association between median value of FEV1% and GOLD criteria classification with COPD was observed (p.value < 0.0001) (Table S2). We also found a significant association between diabetes and BMI (p.value <0.0001) as well as CVD and BMI (p.value < 0.0001), being subjects with an history of diabetes or CVD more likely overweight (Table S3 and S4).
Fig. 1.
Boxplots of the distribution of CAC score by sex and age.
Median follow-up of the whole cohort was 100 months (99 for females and 101 for males, IQR: 85.9–185.5).
3.2. All-cause mortality
Four hundred and seventeen deaths were recorded at the 12-year follow-up, and data on the cause of death were missing for 45 (10.8 %) subjects. The 12-year all-cause mortality was 6.4 % overall (417/6495), 4.1 % (98/2368) among females, and 7.7 % (319/4127) among males (p < 0.0001) (Table 2). Compared with males, females had a significantly lower risk of all-cause mortality (log-rank test p < 0.0001) (Fig. 2A). Significant differences were also observed after stratification by age, confirming higher mortality rates for men in each age group (Fig. S2). Differences in 12-year all-cause mortality were further tested by Cox proportional hazards regression models (Table 3). In multivariate model A adjusted for age, females had a statistically significant lower risk of mortality, with the male-to-female HR (95 % CI) 1.72 (1.36–2.20, p < 0.0001). After adjustment for CAC score (model B), the male-to-female HR was still significant (p = 0.0035), with an increased risk as CAC score categories increased: compared to score 0–10 HRs were 1.47 (p = 0.0108) for score 11–100, 1.79 (p = 0.0004) for score 101–400, and 2.36 (p < 0.0001) for score > 400. This statistical significance was confirmed in multivariate models C,D, and E. In final model E, the male-to-female HR (95 % CI) was 1.47 (1.13–1.91, p = 0.004). Different Cox regression models for females and males were performed separately (Tables S5 and S6, respectively). Kaplan-Meier curves showed an increased risk of all-cause mortality with increasing CAC score (log-rank test, p < 0.0001) in both sexes (Fig. S3). In the sensitivity analysis, 12-year all-cause mortality was evaluated for volunteers without prior CVD using Cox regression models (Table S7), and the data confirmed the models’ results described above (Table 3).
Table 2.
Twelve-year mortality and incidence data stratified by sex.
| TOTAL 6495 | FEMALES 2368 (36.5 %) | MALES 4127 (63.5 %) | Chi-square p-value | |
|---|---|---|---|---|
|
| ||||
| 12-year outcomes All-cause deaths* |
417 (6.4 %) | 98 (4.1 %) | 319 (7.7 %) | <0.001 |
| Lung cancer deaths | 103 (1.6 %) | 24 (1.0 %) | 79 (1.9 %) | 0.0052 |
| Other cancer deaths | 120 (1.9 %) | 32 (1.4 %) | 88 (2.15 %) | 0.0245 |
| CVD deaths | 73 (1.1 %) | 17 (0.7 %) | 56 (1.4 %) | 0.0187 |
| Other causes of deaths | 76 (1.2 %) | 15 (0.6 %) | 61 (1.5 %) | 0.0023 |
| Non-cancer deaths | 149 (2.3 %) |
32 (1.4 %) | 117 (2.8 %) | 0.001 |
| Non-CVD deaths | 299 (4.6 %) |
71 (3.0 %) | 228 (5.5 %) | <0.001 |
| Lung cancer incidence | 299 (4.6 %) |
105 (4.4 %) | 194 (4.7 %) | 0.62 |
| Proportion of stage I | 161 (53.9 %) | 61 (58.1 %) | 100 (51.2 %) | 0.27 |
Missing causes of deaths N = 45.
Fig. 2.
Twelve-year mortality curves stratified by sex: all-cause mortality (A) and noncancer mortality (B).
Table 3.
Twelve-year all-cause mortality Cox regression models.
| 12-year all-cause mortality | Multivariate Model Aa HR (95 %CI) P value | Multivariate Model Bb HR (95 %CI) P value | Multivariate Model Cc HR (95 %CI) P value | Multivariate Model Dd HR (95 %CI) P value | Multivariate Model Ee HR (95 %CI) P value | |
|---|---|---|---|---|---|---|
|
| ||||||
| Sex | Males vs. Females | 1.72 (1.36–2.20) |
1.45 (1.13–1.86) |
1.53 (1.17–1.98) |
1.49 (1.15–1.94) |
1.47 (1.13–1.91) |
|
|
|
<0.001 | 0.0035 | 0.0015 | 0.0027 | 0.004 |
| Age (years) | 55–59 vs. < 55 | 1.28 (0.87–1.88) | 1.18 (0.80–1.74) | 1.18 (0.80–1.73) | 1.18 (0.80–1.74) | 1.16 (0.79–1.72) |
| 0.21 | 0.39 | 0.41 | 0.41 | 0.4487 | ||
| 60–64 vs. < 55 | 2.41 (1.70–3.43) | 2.10 (1.47–2.99) | 1.94 (1.36–2.78) | 1.85 (1.29–2.66) | 1.79 (1.25–2.57) | |
| <0.001 | <0.001 | 0.0003 | 0.0008 | 0.0015 | ||
| ≥65 vs. < 55 | 5.47 (3.95–7.58) | 4.37 (3.12–6.12) | 3.67 (2.60–5.18) | 3.31 (2.34–4.68) | 3.13 (2.21–4.44) | |
| <0.001 | <0.001 | <0.001 | <0.001 | <0.001 | ||
| CAC score | 11–100 vs. 0–10 | 1.47 (1.09–1.97) | 1.43 (1.07–1.93) | 1.41 (1.05–1.90) | 1.42 (1.05–1.90) | |
| 0.0108 | 0.0175 | 0.0218 | 0.0215 | |||
| 101–400 vs. 0–10 | 1.79 (1.30–2.46) | 1.76 (1.28–2.43) | 1.70 (1.23–2.34) | 1.67 (1.21–2.30) | ||
| 0.0004 | 0.0005 | 0.0012 | 0.0017 | |||
| >400 vs. 0–10 | 2.36 (1.70–3.26) | 2.23 (1.61–3.09) | 2.17 (1.56–3.00) | 2.02 (1.46–2.81) | ||
| <0.001 | <0.001 | <0.001 | <0.001 | |||
| Pack-years | 35–44 vs. < 35 | 1.38 (0.99–1.93) 0.0607 |
1.32 (0.95–1.85) 0.1 |
1.31 (0.94–1.83) 0.1163 |
||
| ≥45 vs. < 35 | 2.17 (1.60–2.94) <0.001 |
1.94 (1.43–2.64) <0.001 |
1.90 (1.40–2.59) <0.001 |
|||
| Smoking status | Current vs. Former smokers | 0.96 (0.70–1.31) | 0.89 (0.64–1.21) | 0.88 (0.64–1.20) | ||
| at baseline | 0.78 | 0.45 | 0.4145 | |||
| Smoking status | Current vs. Former smokers | 1.58 (1.20–2.08) | 1.64 (1.24–2.17) | 1.64 (1.24–2.16) | ||
| At the last follow-up | 0.0013 | 0.0005 | 0.0005 | |||
| BMI | Underweight vs. normal weight | 2.04 (1.10–3.80) 0.0245 |
2.05 (1.10–3.83) 0.0243 |
1.97 (1.05–3.70) 0.0336 |
||
| Overweight/obese vs. normal weight | 0.89 (0.71–1.10) 0.27 |
0.84 (0.67–1.05) 0.12 |
0.82 (0.66–1.02) 0.0746 |
|||
| CRP (mg/L) | >2 vs. <=2 | 1.50 (1.21 –1.85) 0.0002 |
1.49 (1.21–1.84) 0.0002 |
|||
| FEV1% | 90–70 vs. > 90 <70 vs. > 90 |
1.03 (0.81–1.31) 0.81 2.24 (1.69–2.98) <0.001 |
0.97 (0.76–1.23) 0.7829 2.04 (1.53–2.72) <0.001 |
|||
| Prior diseases | 1 vs 0 2 vs 0 3 vs 0 |
1.18 (0.93–1.49) 0.1742 1.67 (1.22–2.29) 0.0015 2.82 (1.60–4.96) 0.0003 |
||||
BMI, body mass index; CI, confidence interval; CRP, C-reactive protein; FEV1%, percentage of forced expiratory volume in 1 s; HR, hazard ratio.
Model A included sex and age classes.
Model B included Model A variables and CAC score categories.
Model C included Model B variables and pack-years, smoking status at baseline, smoking status at the last follow-up, and BMI.
Model D included Model C variables and FEV1% and CRP.
Model E included Model D variables and prior diseases (cardiovascular disease, chronic obstructive pulmonary disease and diabetes).
3.3. Noncancer, cardiovascular and noncardiovascular mortality
The 12-year noncancer mortality was 2.3 % for the entire study cohort (149/6495), 1.4 % (32/2368) for females only, and 2.8 % (117/4127) for males only (p < 0.0001) (Table 2). Compared with males, females had a significantly lower risk of noncancer mortality (Gray’s test p = 0.0005) (Fig. 2B). The 12-year CV mortality was 1.1 % overall (73/6495), 0.7 % (17/2368) for females only, and 1.4 % (56/4127) for males only (p = 0.0187) (Table 2). Additionally, females had a significantly lower risk of CV mortality (Gray’s test p = 0.0338) (Fig. S4A). The 12-year non-CV mortality was 4.6 % overall (299/6495), 3.0 % (71/2368) for females only, and 5.5 % (228/4127) for males only (p < 0.0001) (Table 2). The difference in non-CV between sexes was confirmed to be significant (Gray’s test p < 0.0001, Fig. S4B).
3.4. Lung cancer incidence, mortality, and survival
At the 12-year follow-up, 299 (4.6 %) cases of LC were registered, with no significant difference in incidence between women (105/2368, 4.4 %) and men (194/4127, 4.7 %) (p = 0.6216) and with no differences in the proportion of stage I LC diagnoses between women (63/105; 60.0 %) and men (99/194; 51.0 %) (p = 0.2782, Table 2). Analysis of the LC incidence curve confirmed that there was no significant difference between the two sexes (log-rank test p = 7028) (Fig. S5A). Conversely, 12-year LC mortality was significantly lower in females (1.0 %, 24/2368) than in males (1.9 %, 79/4127) (p = 0.0052, Table 2) and this difference was confirmed by Gray’s test (p = 0.0105, Fig. S5B). Twelve-year LC mortality Fine-Gray models were performed (Table S8). Males had a greater risk compared to females (p < 0.05 in Model A, Model C and Model D), while HRs of CAC score categories were not statistically significant. A significant difference between-sex in LC survival was observed, with higher rates in women as compared to their men counterparts (72.3 % vs. 51.7 %; p = 0.0005) (Fig. 3). After stratifying LC cases as stage I or stage II-IV, the difference between sexes remained significant in the early and late stages (Fig. S6). The proportions of LC cases diagnosed by death certificate only were similar between the sexes, at 5.7 % (6/105) among females and 6.7 % (13/194) among males (p = 0.7385). In summary, LC incidence and stage distribution was similar in the two sexes while women had a lower LC mortality.
Fig. 3.
Five-year survival curves from the time of lung cancer diagnosis stratified by sex.
4. Discussion
This retrospective analysis of two LCS trials with a 12-year follow-up period showed that females had significantly lower rates of all-cause, LC and noncancer mortality than males. Moreover, we demonstrated that CAC burden, retrospectively quantified on baseline LDCT images by a fully automated software, was significantly lower in women at all ages and regardless of pack-year exposures. We explored the potential role of CAC in predicting long-term mortality rates in both sexes, observing that, beyond age, CAC burden explained part of the sex-related differences in all-cause mortality. Notably, all-cause mortality was similar between the two sexes within each CAC score stratum.
Numerous studies had revealed that a high CAC score is an independent risk factor for all-cause mortality in both sexes [30–32], and we observed that women have a lower CAC burden, regardless of age, in keeping with prior reports [33,34]. Such sex difference has been attributed to the pathophysiological process of atherosclerotic plaque development. Although robust evidence suggests that high CAC is associated with atheroma development in both sexes, women tend to have fewer calcified lesions and a lower volume of calcifications, explaining why a quantifiable CAC level in women is detected on average 10 years later [13,35,36]. Higher oestrogen levels are well known to protect women against CV events, thus explaining why women aged 40–65 years are less prone to experience major CV events and have a lower CV mortality risk, including those with elevated CAC scores [14]. The observation that 75.2 % of our LCS population (79.1 % of females) had no prior history of CVD, and the lag-time of 1–2 years between baseline LDCT and mortality rise, opens new prospects for CV prevention within LSC programs. Indeed, despite the relative low rate of history of CVD, up to 15 % of participants had a CAC score between 101 and 400, suggesting that LCS program might serve as a secondary prevention strategy in subjects at a greater risk of experiencing CV events, who would benefit from an early referral for CV risk assessment. Notably, the results of such a CV assessment may affect the management of participants diagnosed with lung cancer, with those at a higher risk potentially benefit from a less invasive approach.
Female participants were also found to have significantly lower CRP blood levels and BMI values. CRP, whose levels tend to rapidly rise in response to inflammation, may also increase in patients affected by malignant tumours without concomitant infections [37]. A large body of literature has investigated the role of CRP in LC risk and prognosis [38,39], demonstrating that high levels are associated with poorer survival [40,41]. Our results appear in line with previous evidence, but there are only few studies focusing on the predictive value of CRP in LCS populations [41] and further data are warranted to confirm the significance of high CRP levels in such a setting.
If low BMI is known to be associated with increased mortality in patients with LC [42], there is limited evidence on the association of BMI with mortality risk in LC screenees. Immune-metabolic tumour microenvironment plays a crucial role in the natural history of disease [43] and may improve cancer patients’ stratification and treatment options. In view of this, it can be speculated that immune-metabolic profiling may warrant a more tailored and sex specific LCS strategy.
The ability of LCS to reduce mortality has been demonstrated in previous randomised trials, also showing a more favourable effect in women [3–6]. The present study, by including a longer observation time, a higher proportion of female participants, and assessment of cause-specific mortality, provided a more robust evidence of LCS benefit in women.
Our study also revealed that despite similar LC incidence rates by sex, women had significantly lower LC mortality rates. Of interest, LC stage at diagnosis was not significantly different between women and men, but women had a significantly higher 5-year survival from diagnosis. Many studies have explored the reasons why men have higher cancer mortality rates and have identified both tumour and lifestyle-related factors, including awareness of risk and screening behaviour, that may partially explain this phenomenon, but the explanation of such gap is not fully understood [8–10]. The original data generated by our prolonged LCS experience provide new insight into the reasons for sex disparities in mortality rates, and the magnitude of benefit represents a strong reason to expand LCS programs in females [44,45].
Trends in LC incidence and mortality between sexes have changed over the past 20 years [46], with a dramatic increase in the LC incidence among women [7,47] in all countries except in the USA, and a declining incidence in men [48]. These changes are likely due to shifting trends in smoking over the last decades, resulting in a higher prevalence of smoking among women [49]. In line with these considerations, we observed a significantly higher prevalence of current smokers among women than men at baseline and at the last follow-up.
The study has several strengths, in terms of robustness and reproducibility. First, our analysis, performed through a fully automated software, provides researchers with a new information on the value of CAC burden in LCS. Such data have the potential to improve LCS performance by introducing targeted interventions against CV mortality. Second, our study population includes a high percentage of women. Third, the LCS duration and length of follow-up allows for a comprehensive assessment of long-term outcomes. Fourth, we explored specific non-cancer mortality rates. Last, the 98 % performance shown by this AI software on the relatively old16-slice MILD scanner [23], opens a real possibility of implementing retrospective CAC quantification of large LCS populations on a global level.
The present study has also intrinsic limitations. First, the single-centre design reduced the generalizability of our results. Second, we could not assess racial/ethnic differences since our population included only Caucasian volunteers.
In conclusion, female LC screenees had lower rates of all-cause mortality at 12 years as compared to their male counterparts and CAC burden might play a role in such a discrepancy, suggesting that the implementation of automated CAC score measurement in LCS programs might enforce targeted preventive interventions. Furthermore, women exhibited higher LC survival rates than men, despite similar incidence and stage, and different CRP blood levels and BMI values, suggesting that inflammatory and metabolic profile might be worth investigated to provide further insights for personalized LCS strategies. The analysis of genetic, metabolic, and hormonal features that are likely to influence the observed sex-related mortality difference are currently ongoing in our cohorts to answer some of the open questions. Such integrated approach will aim at reduction all-cause mortality in LCS participants.
Supplementary Material
Acknowledgments
The authors thank the MILD and BioMILD staff C. Banfi, P. Suatoni, A. Calanca, and C. Ninni for recruitment and LDCT recalls; C Jacomelli for data management; and the MILD and BioMILD trial participants. The authors also thank the Italian Association for Cancer Research (AIRC), the Italian Ministry of Health and the National Cancer Institute for the funding received.
Funding
The study was supported by grants from the AIRC Investigator grant nos. 23244 and the Italian Ministry of Health (RF-2018-12367824). The MILD trial was supported by grants from the Italian Ministry of Health (RF 2004), the Italian Association for Cancer Research (AIRC 2004 IG 1227 and AIRC 5xmille IG 12162), Fondazione Cariplo (2004-1560), and the National Cancer Institute (EDRN UO1 CA166905).The bioMILD trial was supported by grants from the Italian Association for Cancer Research (AIRC 5xmille IG 12162, IG 11991 and IG 18812), the Italian Ministry of Health (RF 2010-2306232 and 2010-2310201), the National Cancer Institute (EDRN UO1 CA166905) and Gensignia Life Science.
The sponsors had no role in conducting and interpreting the study.
Footnotes
CRediT authorship contribution statement
Margherita Ruggirello: Investigation, Methodology, Writing – original draft, Writing – review & editing. Camilla Valsecchi: Formal analysis, Investigation, Methodology, Data curation, Writing – original draft, Writing – review & editing. Roberta Eufrasia Ledda: Investigation, Methodology, Writing – original draft, Writing – review & editing. Federica Sabia: Formal analysis, Investigation, Methodology, Data curation, Writing – original draft, Writing – review & editing. Raffaella Vigorito: Writing – original draft. Gabriella Sozzi: Methodology, Writing – original draft, Writing – review & editing. Ugo Pastorino: Conceptualization, Supervision, Formal analysis, Investigation, Methodology, Writing – original draft, Writing – review & editing.
Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Appendix A. Supplementary data
Supplementary data to this article can be found online at https://doi.org/10.1016/j.lungcan.2023.107387.
Data availability
The study data are available upon reasonable request to the corresponding author (UP).
References
- [1].https://ourworldindata.org/, Our World in Data. 2023.
- [2].Tokatli MR, Sisti LG, Marziali E, Nachira L, Rossi MF, Amantea C, Moscato U, Malorni W, Hormones and sex-specific medicine in human physiopathology, Biomolecules 12 (3) (2022) 413. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [3].National Lung Screening Trial Research, T., Lung cancer incidence and mortality with extended follow-up in the national lung screening trial. J. Thorac. Oncol, 2019. 14(10): p. 1732–1742. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [4].de Koning HJ, van der Aalst CM, de Jong PA, Scholten ET, Nackaerts K, Heuvelmans MA, Lammers J-W, Weenink C, Yousaf-Khan U, Horeweg N, van ‘t Westeinde S, Prokop M, Mali WP, Mohamed Hoesein FAA, van Ooijen PMA, Aerts JGJV, den Bakker MA, Thunnissen E, Verschakelen J, Vliegenthart R, Walter JE, ten Haaf K, Groen HJM, Oudkerk M, Reduced lung-cancer mortality with volume CT screening in a randomized trial, N. Engl. J. Med. 382 (6) (2020) 503–513. [DOI] [PubMed] [Google Scholar]
- [5].Pastorino U, Silva M, Sestini S, Sabia F, Boeri M, Cantarutti A, Sverzellati N, Sozzi G, Corrao G, Marchianò A, Prolonged lung cancer screening reduced 10-year mortality in the MILD trial: new confirmation of lung cancer screening efficacy, Ann. Oncol. 30 (7) (2019) 1162–1169. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [6].Becker N, Motsch E, Trotter A, Heussel CP, Dienemann H, Schnabel PA, Kauczor H-U, Maldonado SG, Miller AB, Kaaks R, Delorme S, Lung cancer mortality reduction by LDCT screening-Results from the randomized German LUSI trial, Int. J. Cancer 146 (6) (2020) 1503–1513. [DOI] [PubMed] [Google Scholar]
- [7].Siegel RL, Miller KD, Wagle NS, Jemal A, Cancer statistics, 2023, CA Cancer J. Clin. 73 (1) (2023) 17–48. [DOI] [PubMed] [Google Scholar]
- [8].Radkiewicz C, Johansson ALV, Dickman PW, Lambe M, Edgren G, Sex differences in cancer risk and survival: A Swedish cohort study, Eur. J. Cancer 84 (2017) 130–140. [DOI] [PubMed] [Google Scholar]
- [9].Ellison LF, Differences in cancer survival in Canada by sex, Health Rep. 27 (4) (2016) 19–27. [PubMed] [Google Scholar]
- [10].Jung K-W, Park S, Shin A, Oh C-M, Kong H-J, Jun JK, Won Y-J, Gorlova OY, Do female cancer patients display better survival rates compared with males? Analysis of the Korean National Registry data, 2005–2009, PLoS One 7 (12) (2012) e52457. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [11].Bigeh A, Shekar C, Gulati M, Sex Differences in Coronary Artery Calcium and Long-term CV Mortality, Curr. Cardiol. Rep. 22 (4) (2020) 21. [DOI] [PubMed] [Google Scholar]
- [12].Wong ND, Cordola Hsu AR, Rozanski A, Shaw LJ, Whelton SP, Budoff MJ, Nasir K, Miedema MD, Rumberger J, Blaha MJ, Berman DS, Sex Differences in Coronary Artery Calcium and Mortality From Coronary Heart Disease, Cardiovascular Disease, and All Causes in Adults With Diabetes: The Coronary Calcium Consortium, Diabetes Care 43 (10) (2020) 2597–2606. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [13].Shaw LJ, Min JK, Nasir K, Xie JX, Berman DS, Miedema MD, Whelton SP, Dardari ZA, Rozanski A, Rumberger J, Bairey Merz CN, Al-Mallah MH, Budoff MJ, Blaha MJ, Sex differences in calcified plaque and long-term cardiovascular mortality: observations from the CAC Consortium, Eur. Heart J. 39 (41) (2018) 3727–3735. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [14].Dzaye O, Al Rifai M, Dardari Z, Shaw LJ, Al-Mallah MH, Handy Marshall C, Rozanski A, Mortensen MB, Duebgen M, Matsushita K, Rumberger JA, Berman DS, Budoff MJ, Miedema MD, Nasir K, Blaha MJ, Whelton SP, Coronary Artery Calcium as a Synergistic Tool for the Age- and Sex-Specific Risk of Cardiovascular and Cancer Mortality: The Coronary Artery Calcium Consortium, J. Am. Heart Assoc. 9 (8) (2020) e015306. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [15].Tota-Maharaj R, Joshi PH, Budoff MJ, Whelton S, Zeb I, Rumberger J, Al-Mallah M, Blumenthal RS, Nasir K, Blaha MJ, Usefulness of regional distribution of coronary artery calcium to improve the prediction of all-cause mortality, Am. J. Cardiol. 115 (9) (2015) 1229–1234. [DOI] [PubMed] [Google Scholar]
- [16].Yip R, Jirapatnakul A, Hu M, Chen X, Han D, Ma T, Zhu Y, Salvatore MM, Margolies LR, Yankelevitz DF, Henschke CI, Added benefits of early detection of other diseases on low-dose CT screening, Transl. Lung Cancer Res. 10 (2) (2021) 1141–1153. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [17].Schreuder A, Jacobs C, Lessmann N, Broeders MJM, Silva M, Išgum I, de Jong PA, Sverzellati N, Prokop M, Pastorino U, Schaefer-Prokop CM, van Ginneken B, Combining pulmonary and cardiac computed tomography biomarkers for disease-specific risk modelling in lung cancer screening, Eur. Respir. J. 58 (3) (2021) 2003386. [DOI] [PubMed] [Google Scholar]
- [18].Takx RAP, Išgum I, Willemink MJ, van der Graaf Y, de Koning HJ, Vliegenthart R, Oudkerk M, Leiner T, de Jong PA, Quantification of coronary artery calcium in nongated CT to predict cardiovascular events in male lung cancer screening participants: results of the NELSON study, J. Cardiovasc. Comput. Tomogr. 9 (1) (2015) 50–57. [DOI] [PubMed] [Google Scholar]
- [19].Vonder M, Zheng S, Dorrius MD, van der Aalst CM, de Koning HJ, Yi J, Yu D, Gratama JWC, Kuijpers D, Oudkerk M, Deep Learning for Automatic Calcium Scoring in Population-Based Cardiovascular Screening, J. Am. Coll. Cardiol. Img. 15 (2) (2022) 366–367. [DOI] [PubMed] [Google Scholar]
- [20].Jacobs PC, Gondrie MJA, van der Graaf Y, de Koning HJ, Isgum I, van Ginneken B, Mali WPTM, Coronary artery calcium can predict all-cause mortality and cardiovascular events on low-dose CT screening for lung cancer, AJR Am. J. Roentgenol. 198 (3) (2012) 505–511. [DOI] [PubMed] [Google Scholar]
- [21].Watts JR, Sonavane SK, Snell-Bergeon J, Nath H, Visual scoring of coronary artery calcification in lung cancer screening computed tomography: association with all-cause and cardiovascular mortality risk, Coron. Artery Dis. 26 (2) (2015) 157–162. [DOI] [PubMed] [Google Scholar]
- [22].Balbi M, Sabia F, Ledda RE, Milanese G, Ruggirello M, Silva M, Marchianò AV, Sverzellati N, Pastorino U, Automated Coronary Artery Calcium and Quantitative Emphysema in Lung Cancer Screening Association With Mortality, Lung Cancer Incidence, and Airflow Obstruction, J. Thorac. Imaging 38 (4) (2023) W52–W63. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [23].Sabia F, Balbi M, Ledda RE, Milanese G, Ruggirello M, Valsecchi C, Marchianò A, Sverzellati N, Pastorino U, Ceruti S, Fully automated calcium scoring predicts all-cause mortality at 12 years in the MILD lung cancer screening trial, PLoS One 18 (5) (2023) e0285593. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [24].Pastorino U, Rossi M, Rosato V, Marchianò A, Sverzellati N, Morosi C, Fabbri A, Galeone C, Negri E, Sozzi G, Pelosi G, La Vecchia C, Annual or biennial CT screening versus observation in heavy smokers: 5-year results of the MILD trial, Eur. J. Cancer Prev. 21 (3) (2012) 308–315. [DOI] [PubMed] [Google Scholar]
- [25].Pastorino U, Sverzellati N, Sestini S, Silva M, Sabia F, Boeri M, Cantarutti A, Sozzi G, Corrao G, Marchianò A, Ten-year results of the Multicentric Italian Lung Detection trial demonstrate the safety and efficacy of biennial lung cancer screening, Eur. J. Cancer 118 (2019) 142–148. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [26].Pastorino U, Boeri M, Sestini S, Sabia F, Milanese G, Silva M, Suatoni P, Verri C, Cantarutti A, Sverzellati N, Corrao G, Marchianò A, Sozzi G, Baseline computed tomography screening and blood microRNA predict lung cancer risk and define adequate intervals in the BioMILD trial, Ann. Oncol. 33 (4) (2022) 395–405. [DOI] [PubMed] [Google Scholar]
- [27].Shao L, Yan AT, Lebovic G, Wong HH, Kirpalani A, Deva DP, Prognostic value of visually detected coronary artery calcification on unenhanced non-gated thoracic computed tomography for prediction of non-fatal myocardial infarction and all-cause mortality, J. Cardiovasc. Comput. Tomogr. 11 (3) (2017) 196–202. [DOI] [PubMed] [Google Scholar]
- [28].Takx RAP, de Jong PA, Leiner T, Oudkerk M, de Koning HJ, Mol CP, Viergever MA, Išgum I, Muñoz-Barrutia A, Automated coronary artery calcification scoring in non-gated chest CT: agreement and reliability, PLoS One 9 (3) (2014) e91239. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [29].Pasquinelli MM, Tammemagi MC, Kovitz KL, Durham ML, Deliu Z,¨ Rygalski K, Liu L.i., Koshy M, Finn P, Feldman LE, Risk Prediction Model Versus United States Preventive Services Task Force Lung Cancer Screening Eligibility Criteria: Reducing Race Disparities, J. Thorac. Oncol. 15 (11) (2020) 1738–1747. [DOI] [PubMed] [Google Scholar]
- [30].Chen LQ, Weber J, Christian T, Guerci AD, Cao JJ, Long-term all-cause mortality among asymptomatic individuals with 80th percentile of coronary calcium score based on age and gender in the St. Francis Heart Study, Coron. Artery Dis. 32 (7) (2021) 639–643. [DOI] [PubMed] [Google Scholar]
- [31].Peng AW, et al. , Long-Term All-Cause and Cause-Specific Mortality in Asymptomatic Patients With CAC >/=1,000: Results From the CAC Consortium, J. Am. Coll. Cardiol. Img. 13 (1 Pt 1) (2020) 83–93. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [32].Ota K, Nakanishi R, Hashimoto H, Okamura Y, Watanabe I, Yabe T, Okubo R, Ikeda T, Association between coronary artery calcium score on non-contrast chest computed tomography and all-cause mortality among patients with congestive heart failure, Heart Vessels 37 (2) (2022) 262–272. [DOI] [PubMed] [Google Scholar]
- [33].McClelland RL, Chung H, Detrano R, Post W, Kronmal RA, Distribution of coronary artery calcium by race, gender, and age: results from the Multi-Ethnic Study of Atherosclerosis (MESA), Circulation 113 (1) (2006) 30–37. [DOI] [PubMed] [Google Scholar]
- [34].Lessmann N, de Jong PA, Celeng C, Takx RAP, Viergever MA, van Ginneken B, Išgum I, Sex Differences in Coronary Artery and Thoracic Aorta Calcification and Their Association With Cardiovascular Mortality in Heavy Smokers, J. Am. Coll. Cardiol. Img. 12 (9) (2019) 1808–1817. [DOI] [PubMed] [Google Scholar]
- [35].Leifheit-Limson EC, et al. , Sex Differences in Cardiac Risk Factors, Perceived Risk, and Health Care Provider Discussion of Risk and Risk Modification Among Young Patients With Acute Myocardial Infarction: The VIRGO Study, J. Am. Coll. Cardiol. 66 (18) (2015) 1949–1957. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [36].Ceponiene I, et al. , Coronary Artery Calcium Progression Is Associated With Coronary Plaque Volume Progression: Results From a Quantitative Semiautomated Coronary Artery Plaque Analysis, J. Am. Coll. Cardiol. Img. 11 (12) (2018) 1785–1794. [DOI] [PubMed] [Google Scholar]
- [37].van der Meer V, et al. , Diagnostic value of C reactive protein in infections of the lower respiratory tract: systematic review. BMJ, 2005. 331(7507): p. 26. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [38].Zhu M, Ma Z, Zhang X.u., Hang D, Yin R, Feng J, Xu L, Shen H, C-reactive protein and cancer risk: a pan-cancer study of prospective cohort and Mendelian randomization analysis, BMC Med. 20 (1) (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
- [39].Klümper N, Saal J, Berner F, Lichtensteiger C, Wyss N, Heine A, Bauernfeind FG, Ellinger J, Brossart P, Diem S, Schmid S, Joerger M, Frueh M, Ritter M, Hölzel M, Flatz L, Bald T, C reactive protein flare predicts response to checkpoint inhibitor treatment in non-small cell lung cancer, J. Immunother. Cancer 10 (3) (2022) e004024. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [40].Nagata M, Ito H, Matsuzaki T, Furumoto H, Isaka T, Nishii T, Yokose T, Nakayama H, Body mass index, C-reactive protein and survival in smokers undergoing lobectomy for lung cancerdagger, Eur. J. Cardiothorac. Surg. 51 (6) (2017) 1164–1170. [DOI] [PubMed] [Google Scholar]
- [41].Pastorino U, Morelli D, Marchianò A, Sestini S, Suatoni P, Taverna F, Boeri M, Sozzi G, Cantarutti A, Corrao G, Inflammatory status and lung function predict mortality in lung cancer screening participants, Eur. J. Cancer Prev. 27 (4) (2018) 289–295. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [42].Wang J, Xu H, Zhou S, Wang D, Zhu L, Hou J, Tang J, Zhao J, Zhong S, Body mass index and mortality in lung cancer patients: a systematic review and meta-analysis, Eur. J. Clin. Nutr. 72 (1) (2018) 4–17. [DOI] [PubMed] [Google Scholar]
- [43].Vavalà T, Catino A, Pizzutilo P, Longo V, Galetta D, Gender Differences and Immunotherapy Outcome in Advanced Lung Cancer, Int. J. Mol. Sci. 22 (21) (2021. Nov 4) 11942, 10.3390/ijms222111942. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [44].van der Toorn JE, Rueda-Ochoa OL, van der Schaft N, Vernooij MW, Ikram MA, Bos D, Kavousi M, Arterial calcification at multiple sites: sex-specific cardiovascular risk profiles and mortality risk-the Rotterdam Study, BMC Med. 18 (1) (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
- [45].Gu H, Gao Y, Wang H, Hou Z, Han L, Wang X, Lu B, Sex differences in coronary atherosclerosis progression and major adverse cardiac events in patients with suspected coronary artery disease, J. Cardiovasc. Comput. Tomogr. 11 (5) (2017) 367–372. [DOI] [PubMed] [Google Scholar]
- [46].Our World in Data. 2023. https://ourworldindata.org/.
- [47].Huang J, Deng Y, Tin MS, Lok V, Ngai CH, Zhang L, Lucero-Prisno DE 3rd, Xu W, Zheng ZJ, Elcarte E, Withers M, Wong MCS, Distribution, Risk Factors, and Temporal Trends for Lung Cancer Incidence and Mortality: A Global Analysis, Chest 161 (4) (2022) 1101–1111, 10.1016/j.chest.2021.12.655. [DOI] [PubMed] [Google Scholar]
- [48].Santucci C, Carioli G, Bertuccio P, Malvezzi M, Pastorino U, Boffetta P, Negri E, Bosetti C, La Vecchia C, Progress in cancer mortality, incidence, and survival: a global overview, Eur. J. Cancer Prev. 29 (5) (2020) 367–381. [DOI] [PubMed] [Google Scholar]
- [49].Dai X, Gakidou E, Lopez AD, Evolution of the global smoking epidemic over the past half century: strengthening the evidence base for policy action, Tob. Control 31 (2) (2022) 129–137. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The study data are available upon reasonable request to the corresponding author (UP).