The baseline characteristics of the whole study cohort were presented in Table 1. In general, a total of 2,705 lung ASC patients were identified in the SEER database. Of these, the median age at diagnosis was 69 years (IQR: 61–76). A larger proportion of patients were aged above 65 years (1,776, 65.7%), male (1,437, 53.1%), married (1,590, 58.8%), and white race (2,248, 83.1%). The majority of tumors were located in the upper lobe (1,651, 61.0%), and on the right side (1,564, 57.8%). Most patients were diagnosed with histological grade III (64.7%), followed by grade II (31.7%), grade IV (2.3%), and grade I (1.3%). The proportion of AJCC stage was 42.2% for stage I, 12.3% for stage II, 23.4% for stage III, and 22.1% for stage IV. A total of 67.2% of patients received surgical treatment.
Table 1
Cumulative incidences of death among patients with ASC.
Variables | Nt (%) | Ne (%) | | LC-SM (%) | | | OCSM (%) | |
3-year (95% CI) | 5-year (95% CI) | P | 3-year (95% CI) | 5-year(95% CI) | P |
Total | 2705 | 1897 | 49.6(47.7–51.5) | 55.8(53.8–57.8) | | 8.2(7.1–9.2) | 11.8(10.5–13.1) | |
Age (years) | | | | | | | | |
median (IQR) | 69(61–76) | 70(62–77) | | | | | | |
< 65 years | 929(34.3) | 608(32.1) | 49.2(45.9–52.4) | 55.0(51.6–58.3) | > 0.05 | 5.1(3.7–6.5) | 8.0(6.2–9.9) | < 0.001 |
| 1776(65.7) | 1289(67.9) | 49.9(47.5–52.2) | 56.3(53.8–58.7) | | 9.8(8.4–11.2) | 13.9(12.2–15.6) | |
Gender | | | | | 0.002 | | | > 0.05 |
Female | 1268(46.9) | 858(45.2) | 45.8(43.0-48.6) | 52.4(49.5–55.3) | | 8.3(6.8–9.9) | 11.9(10.0-13.8) | |
Male | 1437(53.1) | 1039(54.8) | 53.0(50.4–55.6) | 58.8(56.1–61.5) | | 8.0(6.6–9.5) | 11.8(10.0-13.6) | |
Race | | | | | 0.008 | | | 0.03 |
Black | 262(9.7) | 195(10.3) | 56.8(50.7–63.0) | 62.9(56.7–69.1) | | 7.0(3.9–10.2) | 12.4(8.0-16.8) | |
White | 2248(83.1) | 1571(82.8) | 48.4(46.3–50.5) | 54.4(52.2–56.6) | | 8.6(7.4–9.8) | 12.2(10.8–13.7) | |
Other race | 195(7.2) | 131(6.9) | 54.3(46.9–61.7) | 63.1(55.6–70.6) | | 4.3(1.4–7.3) | 6.0(2.3–9.7) | |
Marital status | | | | | 0.02 | | | > 0.05 |
Married | 1590(58.8) | 1091(57.5) | 48.1(45.5–50.6) | 54.2(51.6–56.8) | | 7.9(6.5–9.2) | 11.3(9.7–13.0) | |
Unmarried | 1115(41.2) | 806(42.5) | 51.9(48.9–54.9) | 58.1(55.1–61.2) | | 8.6(6.9–10.3) | 12.6(10.5–14.7) | |
Tumor site | | | | | < 0.001 | | | > 0.05 |
Upper lobe | 1651(61.0) | 1130(59.6) | 47.8(45.3–50.3) | 53.7(51.2–56.3) | | 8(6.7–9.4) | 12(10.3–13.7) | |
Middle lobe | 113(4.2) | 79(4.2) | 42.4(33.1–51.7) | 53.8(43.7–63.8) | | 11.2(5.1–17.2) | 18.6(10.6–26.6) | |
Lower lobe | 838(31.0) | 605(31.9) | 52.3(48.8–55.7) | 58.2(54.7–61.7) | | 8.5(6.5–10.4) | 11.4(9.1–13.7) | |
Main bronchus | 58(2.1) | 53(2.8) | 82(71.6–92.5) | 86.8(76.9–96.7) | | 5.2(0–11) | 5.2(0–11.0) | |
Overlapping lesion | 45(1.7) | 30(1.6) | 42.8(28-57.6) | 52.8(37.5–68.0) | | 4.6(0–11) | 7.1(0–15.0) | |
Tumor laterality | | | | | > 0.05 | | | > 0.05 |
Right | 1564(57.8) | 1105(58.2) | 48.7(46.2–51.2) | 55.2(52.6–57.8) | | 8.6(7.2–10) | 12.9(11.1–14.7) | |
Left | 1135(42.0) | 787(41.5) | 50.8(47.9–53.8) | 56.4(53.4–59.4) | | 7.6(6-9.2) | 10.4(8.5–12.3) | |
Bilateral | 6(0.2) | 5(0.3) | 66.7(22.2-111.2) | 83.3(43.8–100) | | 0(0–0) | 0(0–0) | |
Grade | | | | | < 0.001 | | | < 0.001 |
Grade I | 35(1.3) | 21(1.1) | 21.1(6.8–35.3) | 21.1(6.8–35.3) | | 21.5(7–36) | 25.6(9.6–41.5) | |
Grade II | 857(31.7) | 529(27.9) | 38.2(34.8–41.5) | 45.9(42.4–49.5) | | 7.5(5.7–9.3) | 12.8(10.4–15.3) | |
Grade III | 1751(64.7) | 1294(68.2) | 55.4(53.1–57.8) | 61.1(58.7–63.5) | | 8.3(6.9–9.6) | 11.1(9.5–12.6) | |
Grade IV | 62(2.3) | 53(2.8) | 60.1(47.7–72.6) | 62.1(49.6–74.6) | | 8.2(1.2–15.2) | 12.2(3.5–20.8) | |
T stage | | | | | < 0.001 | | | < 0.001 |
T1 | 736(27.2) | 402(21.2) | 26.8(23.5–30.1) | 32.8(29.2–36.4) | | 9.4(7.2–11.6) | 16.3(13.3–19.2) | |
T2 | 1213(44.8) | 839(44.2) | 47.2(44.4–50.1) | 54.2(51.3–57.2) | | 9(7.4–10.7) | 12.1(10.2–14.0) | |
T3 | 211(7.8) | 174(9.2) | 68.9(62.5–75.3) | 74.6(68.4–80.9) | | 6.6(3.1–10.1) | 9.9(5.5–14.4) | |
T4 | 545(20.1) | 482(25.4) | 78.2(74.6–81.7) | 83.1(79.7–86.5) | | 5.2(3.3–7.1) | 6.1(4.0-8.2) | |
N stage | | | | | < 0.001 | | | < 0.001 |
N0 | 1475(54.5) | 892(47.0) | 33.8(31.4–36.3) | 40.6(38.0-43.2) | | 9.6(8-11.1) | 14.9(12.9–16.9) | |
N1 | 362(13.4) | 245(12.9) | 51.5(46.2–56.7) | 58.6(53.2–64.0) | | 5.6(3.1-8) | 8.3(5.2–11.4) | |
N2 | 711(26.3) | 617(32.5) | 73.8(70.5–77.1) | 79.2(76.0-82.3) | | 7(5.1–8.9) | 8.3(6.2–10.4) | |
N3 | 157(5.8) | 143(7.5) | 85.1(79.2–91.0) | 87.3(81.5–93.2) | | 6.6(2.6–10.5) | 6.6(2.6–10.5) | |
M stage | | | | | < 0.001 | | | < 0.001 |
M0 | 2108(77.9) | 1344(70.8) | 39.8(37.7–42) | 46.9(44.7–49.2) | | 8.9(7.6–10.1) | 13.3(11.8–14.9) | |
M1 | 597(22.1) | 553(29.2) | 84.2(81.2–87.1) | 87.1(84.2–89.9) | | 5.8(3.9–7.7) | 6.5(4.5–8.6) | |
AJCC Stage | | | | | < 0.001 | | | < 0.001 |
I | 1142 (42.2) | 624(32.9) | 23.9(21.4–26.4) | 31.1(28.3–34.0) | | 11.2(9.3–13.1) | 17.2(14.8–19.6) | |
II | 332(12.3) | 213(11.2) | 47.1(41.6–52.6) | 52.7(47.1–58.4) | | 5.7(3.1–8.3) | 9.4(6.0-12.8) | |
III | 634(23.4) | 507(26.7) | 64.9(61.1–68.8) | 72.8(69.1–76.5) | | 6.3(4.3–8.2) | 8.4(6.0-10.7) | |
IV | 597(22.1) | 553(29.2) | 84.2(81.2–87.1) | 87.1(84.2–89.9) | | 5.8(3.9–7.7) | 6.5(4.5–8.6) | |
Surgery | | | | | < 0.001 | | | < 0.001 |
No | 886(32.8) | 806(42.5) | 82.1(79.5–84.7) | 85.9(83.4–88.5) | | 7.9(6.1–9.7) | 8.8(6.8–10.8) | |
Yes | 1819(67.2) | 1091(57.5) | 34.1(31.8–36.3) | 41.7(39.3–44.0) | | 8.3(7-9.6) | 13.3(11.6–15.0) | |
Abbreviations: Nt: total number; Ne: number of death events; ASC: adenosquamous carcinoma; CI: confidence interval; LC-SM: lung cancer-specific mortality; OCSM: other cause-specific morality; AJCC: American Joint Committee on Cancer; IQR: interquartile range. |
The median follow-up of the whole study cohort was 21 months (IQR: 8–52). In total, 1,895 (70.1%) patients died throughout the whole follow-up period, of whom 1,535 (81.0%) died owing to lung cancer and 362 (19.0%) died due to non-lung cancer causes. The 3-year and 5-year cumulative incidences of LC-SM and OCSM by different clinicopathological characteristics were displayed in Table 1, and the corresponding CIF curves were presented in Fig. 2. Overall, the 3-year and 5-year LC-SM were 49.6% (CI: 47.7%-51.5%) and 55.8% (CI: 53.8%-57.8%), respectively, while the 3-year and 5-year OCSM were 8.2% (CI: 7.1%-9.2%) and 11.8% (CI: 10.5%-13.1%), respectively. In univariate analysis, male, unmarried status, black race, main bronchus, advanced TNM stage, advanced histological grade, and surgical treatment were related to significantly higher incidences of LC-SM, whereas there were no significant differences for age and tumor laterality. In the meantime, there were significantly higher incidences of OCSM for patients aged ≥ 65 years, diagnosed with histological grade I or earlier TNM stage, and receiving surgery.
In multivariate analysis, age, gender, surgery, T stage, N stage, and M stage were independent predictive factors for LC-SM, in contrast with merely age and surgery acting as independent predictors for OCSM (Table 2). Furthermore, increasing age was associated with an increased probability of LCSM. Male was related to a significantly higher likelihood of LCSM (1.26, CI: 1.10–1.43), while surgery was related to a significantly lower likelihood of LC-SM (0.45, CI: 0.37–0.53). When compared with patients with T1, advanced T-stage patients had a higher probability to face LCSM, with SHR of 1.44 (1.21–1.72), 2.24 (1.72–2.92), and 1.99 (1.59–2.49) for T2, T3, and T4, respectively. Similar phenomenon was observed among advanced N-stage patients, with SHR of 1.52(1.26–1.84), 1.57(1.32–1.87), and 1.51(1.12–2.03) for N1, N2, and N3, respectively, in contrast with N0. For OCSM, patients with older age and receiving surgery were at a higher risk.
Table 2
Multivariate competing risk model for LC-SM and OCSM in patients with lung ASC.
Characteristics | LC-SM | | OCSM |
Coefficient | SHR (95% CI) | P | | Coefficient | SHR (95% CI) | P |
Age | 0.009 | 1.01(1.00-1.02) | 0.008 | | 0.035 | 1.04(1.02–1.05) | < 0.001 |
Sex | | | | | | | |
Female | Reference | Reference | | | | | |
Male | 0.228 | 1.26(1.10–1.43) | < 0.001 | | | | |
Surgery | | | | | | | |
No | Reference | Reference | | | Reference | Reference | |
Yes | -0.810 | 0.45(0.37–0.53) | < 0.001 | | 0.695 | 2.00(1.46–2.75) | < 0.001 |
T stage | | | | | | | |
T1 | Reference | Reference | | | | | |
T2 | 0.363 | 1.44(1.21–1.72) | < 0.001 | | | | |
T3 | 0.805 | 2.24(1.72–2.92) | < 0.001 | | | | |
T4 | 0.688 | 1.99(1.59–2.49) | < 0.001 | | | | |
N stage | | | | | | | |
N0 | Reference | Reference | | | | | |
N1 | 0.419 | 1.52(1.26–1.84) | < 0.001 | | | | |
N2 | 0.454 | 1.57(1.32–1.87) | < 0.001 | | | | |
N3 | 0.409 | 1.51(1.12–2.03) | 0.007 | | | | |
M stage | | | | | | | |
M0 | Reference | Reference | | | | | |
M1 | 0.506 | 1.66(1.38-2.00) | < 0.001 | | | | |
Abbreviations: ASC: adenosquamous carcinoma; CI: confidence interval; LC-SM: lung cancer-specific mortality; OCSM: other cause-specific morality; SHR: Subdistribution hazard ratio. |
The nomograms on the basis of the competing risk models were developed to calculate the 3-year and 5-year cumulative death probabilities (Fig. 3). For each patient, locate the values of different variables on the corresponding variable rows and draw vertical lines pointing to the “Points” row to get corresponding scores. For instance, for a male patient, by drawing a vertical line straight up to the “Point” row, we would obtain approximately 28 points. Similar processes were performed to other variables. By adding up these scores, a total score was obtained, and can be located on the “Total Points” row. Subsequently, a vertical line was drawn straight down to acquire the 3-year or 5-year cumulative death probabilities. For example, if the total score was 100, the corresponding 3-year and 5-year probabilities of LC-SM would be approximately 30% and 36%, respectively.
The calibration curves accompanied with C-indexes were displayed in Fig. 4. As shown in Fig. 4, if the calibration curves were close to the 45-degree diagonal line, it denoted the developed nomograms were well calibrated (a good agreement between the observed mortality probability and predicted mortality probability). Additionally, the 3-year and 5-year C-indexes for the nomograms predicting probabilities of LC-SM were 0.83 (CI, 0.78–0.87) and 0.82 (CI, 0.73–0.90) for training cohort, and 0.79 (CI, 0.75–0.84) and 0.79 (CI, 0.71–0.88) for validation cohort, respectively, which indicated a superb model discrimination. Besides, when it comes to the nomograms predicting probabilities of OCSM, the 3-year and 5-year C-indexes were 0.60 (CI, 0.55–0.65) and 0.62 (CI, 0.57–0.67) for training cohort, as well as 0.58 (CI, 0.50–0.66) and 0.63 (CI, 0.56–0.70) for validation cohort, respectively. The 10-fold cross validation C-indexes were shown in Table 3. The adjusted 3-year and 5-year C-indexes were 0.81 (CI, 0.80–0.83) and 0.81 (CI, 0.80–0.83) for predicting probabilities of LC-SM, and 0.60 (CI, 0.56–0.63) and 0.63 (CI, 0.58–0.65) for predicting probabilities of OCSM.
Table 3
C-indexes of the predictive model for patients with ASC.
Cohort | | C-indexes | | Adjusted C-indexes |
| | 3-year | 5-year | | 3-year | 5-year |
Training cohort | | | | | | |
LC-SM | | 0.83 (CI, 0.78–0.87) | 0.82 (CI, 0.73–0.90) | | | |
OCSM | | 0.60 (CI, 0.55–0.65) | 0.62 (CI, 0.57–0.67) | | | |
Validation cohort | | | | | | |
LC-SM | | 0.79 (CI, 0.75–0.84) | 0.79 (CI, 0.71–0.88) | | | |
OCSM | | 0.58 (CI, 0.50–0.66) | 0.63 (CI, 0.56–0.70) | | | |
Overall cohort | | | | | | |
LC-SM | | | | | 0.81 (CI, 0.80–0.83) | 0.81 (CI, 0.80–0.83) |
OCSM | | | | | 0.60 (CI, 0.56–0.63) | 0.63 (CI, 0.58–0.65) |