Sage Journals: Discover world-class research

Abstract

Objective

Combined small cell lung cancer (CSCLC) with distant metastasis (DM) is an aggressive disease with a poor prognosis. Effective nomograms are needed to predict DM and early death in patients with CSCLC and DM.

Methods

This retrospective study included patients with CSCLC from the Surveillance, Epidemiology, and End Results database between 2004 and 2015. Risk factors for DM and early death were analyzed by univariate and multivariate logistic regression. Nomograms were constructed based on the results in a training cohort and confirmed in a validation cohort, and their performances were assessed by concordance index (C-index), receiver operating characteristic curve (ROC), calibration curve, and decision curve analysis (DCA).

Results

A total of 788 patients with CSCLC were selected, including 364 patients with metastatic CSCLC. Sex, tumor site, T stage, and N stage were independent risk factors for DM, while age, surgery, chemotherapy, and liver metastasis were independent risk factors for early death. C-index, ROC, calibration, and DCA curve analyses all showed good predictive performances for both nomograms.

Conclusions

These nomograms could reliably predict DM risk in CSCLC patients and early death in CSCLC patients with DM, and may thus help clinicians to assess these risks and implement individualized therapies.

Keywords

Combined small cell lung cancer distant metastasis risk factor early death nomogram prediction

Introduction

According to the Global Cancer Statistics 2020, lung cancer was the leading cause of cancer mortality worldwide, with an estimated 2.2 million new cases and 1.8 million new deaths in 2020.¹ Small cell lung cancer (SCLC) is characterized as a highly lethal and aggressive type, accounting for around 15% of lung cancer cases, while non-small cell lung cancer (NSCLC) accounts for nearly 80%. The World Health Organization Histology Criteria 2015 defined CSCLC as a subset of SCLC diagnosed by pathological biopsy including components of both SCLC and any other subtype of NSCLC, such as adenocarcinoma, squamous cell carcinoma, or large cell carcinoma.² The incidence of CSCLC is variable, accounting for approximately 5% to 20% of all SCLC cases.³ SCLC is a recalcitrant carcinoma with high propensities for metastasis and recurrence. Approximately one third of patients with SCLC are initially diagnosed with early-stage disease, while around 70% of cases are diagnosed at an advanced stage.^4,5 In a retrospective study of 181 patients with CSCLC, 99 patients (54.7%) had recurrence after surgery, leading to poor survival.⁶ Although standard chemotherapy and radiation plus immune checkpoint inhibitors have been recommended, survival is still limited, especially in patients with SCLC and distant metastasis (DM). It is therefore imperative to identify patients with CSCLC at high risk of metastasis.

In a study of 43,156 patients with T1-2 NSCLC, 6944 patients (16.09%) developed DM, and race, sex, histology, T and N stage, marital status, tumor size, grade, and laterality were all identified as independent risk factors for metastasis.⁷ In another series of 18,187 patients with SCLC, 22.07% of patients were diagnosed with bone metastasis, with a median survival of 6 months, with age, sex, N stage, and tumor size all related to bone metastasis.⁸ Zhao et al. reported that low expression levels of Cullin5 and SOCS3 promoted metastasis, drug-resistance, and a poor prognosis in SCLC [13]. Other research also revealed that high neuron-specific enolase, neutrophil-to-lymphocyte, Friend leukemia virus integration 1, and the long noncoding RNA CCAT2 were associated with SCLC metastasis.^9–11 Risk factors for metastasis based on laboratory markers have been identified in patients with NSCLC and SCLC; however, the risk factors based on clinicopathological parameters correlated with metastatic CSCLC remain unclear.

The treatment response is currently transient in most patients with SCLC and DM, resulting in a median survival of 1 year. In a comprehensive analysis of 358 patients with extensive-stage disease, 43.8% of patients died within 12 months,¹¹ while another study in China also reported a median survival of just 11 months, and 58.4% of patients with extensive-stage SCLC died within 1 year.¹² Patients with CSCLC and DM are prone to early death, defined as an overall survival time of less than 3 months after diagnosis. Identification of the factors contributing to early death may help clinicians to develop individualized treatment strategies, which might in turn improve overall survival and alleviate the suffering caused by the disease. There are currently no published in-depth studies on mortality within 3 months of metastatic CSCLC. There is thus a need for a simple-to-use, predictive model to distinguish the potential risk factors for early death.

Nomograms are widely used to predict DM and early death in cancer patients; however, most existing nomograms are derived from patients with metastatic SCLC or CSCLC without DM, and there are no nomograms for diagnosis and early death in patients with CSCLC and DM.

We therefore aimed to construct novel nomograms to predict DM in patients with CSCLC and early death in patients with CSCLC and DM, based on demographic and clinicopathologic variables from the Surveillance, Epidemiology, and End Results (SEER) database. The resulting nomograms may facilitate individualized patient care and medical therapy.

Materials and methods

Patients and data collection

This retrospective study used data extracted from the SEER database between 2004 and 2015. The inclusion criteria were as follows: (a) malignant tumor located in the main bronchus and lung (site code: C340-C349); (b) patients diagnosed with primary CSCLC (histology code: 8045); (c) diagnostic confirmation based on positive histology, laboratory test/marker, or microscopic confirmation; and (d) complete T stage, N stage, and M stage according to the 6th edition American Joint Committee on Cancer staging system. We excluded patients with unknown overall survival time. Data on clinical variables including age at diagnosis, sex, race, grade, tumor site, surgery, chemotherapy, radiotherapy, lung metastasis, brain metastasis, bone metastasis, and liver metastasis were also obtained. We also excluded patients with missing data on race. Age at diagnosis was divided into under 60 years old, 60 to 70 years old, and over 70 years old. Tumor grade was divided into grade I to II, grade III to IV, and unknown.

Patients were divided randomly into training and validation cohorts at a ratio of 7:3, using R software (www.r-project.org). The selection processes for the two study cohorts are shown in Figure 1.

Figure 1.

Study selection process. CSCLC, combined small cell lung cancer; DM, distant metastasis; SEER, Surveillance, Epidemiology, and End Results database.

This retrospective study was approved by the Institutional Review Board of The First Affiliated Hospital Shaoyang University and was exempt from the need for informed consent. All information from the SEER database has been de-identified and no personal identifying information was used in our analysis, and informed consent was therefore not required. The reporting of this study conforms to STROBE guidelines.¹³

Statistical analysis

All statistical analyses were carried out using R software (version 4.0.3) and a P-value <0.05 (two-sided) was considered statistically significant. Numbers and percentages were used to summarize categorical variables. The training and validation cohorts were extracted randomly using R software at a ratio of 7:3, and the distribution and difference between the two cohorts were examined by χ² or Fisher’s exact tests. Independent risk factors for DM in patients with CSCLC and for early death in patients with CSCLC and DM were analyzed by univariate and multivariate logistic regression analyses. Risk factors with a P-value <0.05 in univariate analysis were included in the multivariate analysis. A predictive nomogram was constructed based on the results of multivariate analysis in the training cohort and was validated in the validation cohort using the “rms” package. The discrimination abilities of the nomograms in both the training and validation cohorts were determined using receiver operating characteristic (ROC) curves and the corresponding area under the curve (AUC), and the C-index. Calibration curves were plotted to determine the consistency between the predicted and actual probabilities. Decision curve analysis (DCA) curves were generated to evaluate the clinical benefits and improved performance of the nomogram.

Results

Baseline characteristics

A total of 788 patients diagnosed with primary CSCLC met the inclusion criteria and were enrolled in the present study between 2004 and 2015, including 364 patients (46.2%) diagnosed with DM. The patients’ baseline characteristics are presented in Table 1. The DM rates were similar in the training and validation cohorts (46.5% and 45.6%, respectively).

Table 1.

Baseline characteristics of patients with combined small cell lung cancer.

	Overall	Training cohort	Validation cohort	P
	(n = 788)	(n = 551)	(n = 237)	P
Age, years
<60	179 (22.7%)	128 (23.2%)	51 (21.5%)	0.834
60–70	302 (38.3%)	208 (37.7%)	94 (39.7%)
>70	307 (39.0%)	215 (39.0%)	92 (38.8%)
Sex
Female	349 (44.3%)	236 (42.8%)	113 (47.7%)	0.212
Male	439 (55.7%)	315 (57.2%)	124 (52.3%)
Race
Black	100 (12.7%)	71 (12.9%)	29 (12.2%)	0.944
Other	32 (4.1%)	22 (4.0%)	10 (4.2%)
White	656 (83.2%)	458 (83.1%)	198 (83.5%)
Grade
I–II	48 (6.1%)	39 (7.1%)	9 (3.8%)	0.0525
III–IV	407 (51.6%)	292 (53.0%)	115 (48.5%)
Unknown	333 (42.3%)	220 (39.9%)	113 (47.7%)
Site
Left upper lobe	202 (25.6%)	146 (26.5%)	56 (23.6%)	0.905
Left lower lobe	90 (11.4%)	62 (11.3%)	28 (11.8%)
Right upper lobe	236 (29.9%)	164 (29.8%)	72 (30.4%)
Right middle lobe	33 (4.2%)	23 (4.2%)	10 (4.2%)
Right lower lobe	93 (11.8%)	66 (12.0%)	27 (11.4%)
Main bronchus	59 (7.5%)	37 (6.7%)	22 (9.3%)
Unspecific	75 (9.5%)	53 (9.6%)	22 (9.3%)
T
T1	175 (22.2%)	119 (21.6%)	56 (23.6%)	0.819
T2	265 (33.6%)	188 (34.1%)	77 (32.5%)
T3	33 (4.2%)	25 (4.5%)	8 (3.4%)
T4	315 (40.0%)	219 (39.7%)	96 (40.5%)
N
N0	260 (33.0%)	181 (32.8%)	79 (33.3%)	0.623
N1	79 (10.0%)	51 (9.3%)	28 (11.8%)
N2	336 (42.6%)	236 (42.8%)	100 (42.2%)
N3	113 (14.3%)	83 (15.1%)	30 (12.7%)
M
M0	424 (53.8%)	295 (53.5%)	129 (54.4%)	0.876
M1	364 (46.2%)	256 (46.5%)	108 (45.6%)

DM-related risk factors

We explored potential risk factors for DM by univariate logistic regression analysis. Sex, tumor site, T stage, and N stage were all correlated with DM (P < 0.05) and were further included in multivariate analysis, which identified male sex, main bronchus, T4 stage, and higher N stage as independent risk factors for DM in patients with CSCLC (Table 2).

Table 2.

Univariate and multivariate logistic regression analyses of risk factors associated with distant metastasis in patients with combined small cell lung cancer.

	Univariate analysis			Multivariate analysis
	HR	95%CI	P	HR	95%CI	P
Age, years
<60	Reference
60–70	0.796	0.512–1.239	0.312	1.989	0.831–4.759	0.122
>70	1.016	0.656–1.574	0.943	3.224	1.345–7.724	0.009
Sex
Female	Reference
Male	1.700	1.207–2.394	0.002	1.899	1.291–2.792	0.001
Race
Black	Reference
Other	2.204	0.802–6.055	0.125
White	0.841	0.510–1.387	0.497
Grade
I–II	Reference
III–IV	0.726	0.371–1.422	0.350
Unknown	1.564	0.789–3.099	0.200
Tumor site
Left upper lobe	Reference
Left lower lobe	1.168	0.644–2.119	0.609	1.118	0.582–2.146	0.737
Right upper lobe	0.719	0.456–1.134	0.156	0.665	0.405–1.092	0.107
Right middle lobe	1.938	0.789–4.762	0.149	2.205	0.846–5.749	0.106
Right lower lobe	0.712	0.391–1.295	0.266	0.670	0.349–1.284	0.228
Main bronchus	3.365	1.518–7.455	0.003	2.391	1.016–5.626	0.046
Unspecific	2.639	1.360–5.120	0.004	1.377	0.674–2.813	0.380
T
T1	Reference
T2	1.646	0.993–2.729	0.053	1.291	0.757–2.203	0.3489
T3	2.620	1.081–6.349	0.033	1.781	0.697–4.55	0.2281
T4	5.450	3.321–8.945	<0.001	3.536	2.067–6.048	<0.001
N
N0	Reference
N1	1.600	0.837–3.059	0.155	1.397	0.693–2.819	0.350
N2	3.042	2.016–4.592	<0.001	2.242	1.428–3.519	<0.001
N3	4.619	2.654–8.041	<0.001	2.985	1.640–5.434	<0.001

HR, hazard ratio; CI, confidence interval.

Nomogram construction and validation

We constructed a diagnostic risk nomogram based on the four independent risk factors in multivariate analysis (Figure 2). The C-indexes in the training and validation cohorts were 0.769 (95% confidence interval: 0.730–0.808) and 0.764 (0.704–0.823), respectively, and the AUCs of the ROC curves were 0.769 (0.729–0.808) and 0.769 (0.704–0.824), respectively, indicating that both these factors showed good discrimination (Figure 3a, b). Calibration curves showed good consistency for predicting DM between the predicted and actual results (Figure 3c, d). In addition, DCA showed that the nomogram provided excellent benefits. A larger net benefit was associated with a better predictive performance of the diagnostic risk model, implying that the nomogram could provide an effective and precise tool for DM prediction (Figure 3e, f).

Figure 2.

Diagnostic nomogram for predicting distant metastasis in patients with combined small cell lung cancer. Tumor site code: 1: left upper lobe; 2: left lower lobe; 3: right upper lobe; 4: right middle lobe; 5: right lower lobe; 6: main bronchus; 7: unspecific. DM, distant metastasis.

Figure 3.

Validation of the diagnostic nomogram. Receiver operating characteristic curve (a), calibration curve (c), and decision curve analysis (DCA) (e) in the training cohort, and receiver operating characteristic curve (b), calibration curve (d), and DCA (f) in the validation cohort. DM, distant metastasis; AUC, area under the curve.

Factors associated with early death

Patients lacking data on surgery, radiotherapy, and chemotherapy were excluded from the study on early death, and this cohort therefore included 363 patients with metastatic CSCLC, including 150 patients (41.3%) who experienced early death. The demographic characteristics of all patients with DM are summarized in Table 3. There were no significant differences between the training and validation cohorts. Univariate and multivariate logistic regression analyses identified four independent risk factors associated with early death in patients with DM, namely age, surgery, chemotherapy, and liver metastasis (Table 4).

Table 3.

Baseline characteristics of patients with combined small cell lung cancer and distant metastasis.

	Overall	Training cohort	Validation cohort	P
	(n = 363)	(n = 254)	(n = 109)	P
Age
<60	88 (24.2%)	62 (24.4%)	26 (23.9%)	0.239
60–70	131 (36.1%)	98 (38.6%)	33 (30.3%)
>70	144 (39.7%)	94 (37.0%)	50 (45.9%)
Sex
Female	135 (37.2%)	92 (36.2%)	43 (39.4%)	0.556
Male	228 (62.8%)	162 (63.8%)	66 (60.6%)
Race
Black	50 (13.8%)	31 (12.2%)	19 (17.4%)	0.080
Other	19 (5.2%)	17 (6.7%)	2 (1.8%)
White	294 (81.0%)	206 (81.1%)	88 (80.7%)
Grade
I–II	20 (5.5%)	13 (5.1%)	7 (6.4%)	0.605
III–IV	153 (42.1%)	104 (40.9%)	49 (45.0%)
Unknown	190 (52.3%)	137 (53.9%)	53 (48.6%)
T
T1	39 (10.7%)	28 (11.0%)	11 (10.1%)	0.215
T2	96 (26.4%)	60 (23.6%)	36 (33.0%)
T3	14 (3.9%)	12 (4.7%)	2 (1.8%)
T4	214 (59.0%)	154 (60.6%)	60 (55.0%)
N
N0	71 (19.6%)	42 (16.5%)	29 (26.6%)	0.119
N1	29 (8.0%)	22 (8.7%)	7 (6.4%)
N2	190 (52.3%)	134 (52.8%)	56 (51.4%)
N3	73 (20.1%)	56 (22.0%)	17 (15.6%)
Tumor site
Left upper lobe	82 (22.6%)	59 (23.2%)	23 (21.1%)	0.514
Left lower lobe	38 (10.5%)	26 (10.2%)	12 (11.0%)
Right upper lobe	98 (27.0%)	67 (26.4%)	31 (28.4%)
Right middle lobe	20 (5.5%)	14 (5.5%)	6 (5.5%)
Right lower lobe	32 (8.8%)	24 (9.4%)	8 (7.3%)
Main bronchus	44 (12.1%)	35 (13.8%)	9 (8.3%)
Unspecific	49 (13.5%)	29 (11.4%)	20 (18.3%)
Surgery
No	333 (91.7%)	235 (92.5%)	98 (89.9%)	0.411
Yes	30 (8.3%)	19 (7.5%)	11 (10.1%)
Radiation
No	197 (54.3%)	135 (53.1%)	62 (56.9%)	0.566
Yes	166 (45.7%)	119 (46.9%)	47 (43.1%)
Chemotherapy
No	143 (39.4%)	99 (39.0%)	44 (40.4%)	0.816
Yes	220 (60.6%)	155 (61.0%)	65 (59.6%)
Lung metastasis
No	124 (34.2%)	91 (35.8%)	33 (30.3%)	0.417
Yes	46 (12.7%)	29 (11.4%)	17 (15.6%)
Unknown	193 (53.2%)	134 (52.8%)	59 (54.1%)
Brain metastasis
No	129 (35.5%)	93 (36.6%)	36 (33.0%)	0.668
Yes	43 (11.8%)	28 (11.0%)	15 (13.8%)
Unknown	191 (52.6%)	133 (52.4%)	58 (53.2%)
Bone metastasis
No	117 (32.2%)	80 (31.5%)	37 (33.9%)	0.825
Yes	56 (15.4%)	41 (16.1%)	15 (13.8%)
Unknown	190 (52.3%)	133 (52.4%)	57 (52.3%)
Liver metastasis
No	115 (31.7%)	79 (31.1%)	36 (33.0%)	0.784
Yes	57 (15.7%)	42 (16.5%)	15 (13.8%)
Unknown	191 (52.6%)	133 (52.4%)	58 (53.2%)
Early death
No	213 (81.0%)	149 (58.7%)	64 (58.8%)	0.999
Yes	150 (19.0%)	105 (41.3%)	45 (41.2%)

Table 4.

Univariate and multivariate logistic regression analyses of risk factors associated with early death in patients with combined small cell lung cancer and distant metastasis.

	Univariate analysis			Multivariate analysis
	HR	95%CI	P	HR	95%CI	P
Age, years
<60	Reference
60–70	1.750	0.878–3.486	0.112	1.989	0.831–4.759	0.122
>70	2.882	1.447–5.742	0.003	3.224	1.345–7.724	0.009
Sex
Female	Reference
Male	1.655	0.973–2.815	0.063
Race
Black	Reference
Other	3.333	0.967–11.487	0.056
White	1.227	0.559–2.694	0.610
Grade
I–II	Reference
III–IV	1.619	0.418–6.269	0.485
Unknown	3.285	0.866–12.458	0.080
Tumor site
Left upper lobe	Reference
Left lower lobe	1.950	0.763–4.987	0.163	1.337	0.418–4.274	0.625
Right upper lobe	0.890	0.422–1.878	0.760	0.867	0.344–2.183	0.762
Right middle lobe	1.950	0.600–6.335	0.267	0.902	0.192–4.233	0.896
Right lower lobe	2.305	0.876–6.062	0.091	0.981	0.284–3.393	0.976
Main bronchus	1.300	0.547–3.087	0.552	1.062	0.363–3.109	0.912
Unspecific	2.762	1.107–6.895	0.029	2.696	0.872–8.339	0.085
T
T1	Reference
T2	1.346	0.507–3.575	0.551
T3	0.227	0.025–2.060	0.188
T4	2.373	0.986–5.715	0.054
N
N0	Reference
N1	1.275	0.430–3.782	0.662
N2	1.809	0.865–3.782	0.115
N3	1.673	0.721–3.881	0.231
Surgery
No	Reference
Yes	0.070	0.009–0.533	0.010	0.060	0.006–0.563	0.014
Radiotherapy
No	Reference
Yes	0.415	0.247–0.695	0.001	1.130	0.568–2.25	0.728
Chemotherapy
No	Reference
Yes	0.084	0.046–0.154	<0.001	0.070	0.034–0.146	<0.001
Lung
No	Reference
Yes	1.183	0.505–2.775	0.699
Unknown	1.319	0.765–2.274	0.319
Brain metastasis
No	Reference
Yes	1.504	0.640–3.534	0.349
Unknown	1.342	0.779–2.311	0.289
Bone metastasis
No	Reference
Yes	1.012	0.467–2.189	0.977
Unknown	1.222	0.694–2.152	0.486
Liver metastasis
No	Reference
Yes	2.242	1.042–4.824	0.039	3.585	1.336–9.620	0.011
Unknown	1.529	0.855–2.735	0.152	1.495	0.701–3.192	0.298

HR, hazard ratio; CI, confidence interval.

Nomogram construction and validation

We constructed a nomogram to predict early death in patients with CSCLC and DM based on the independent risk factors screened by multivariate analysis (Figure 4). The C-indexes in the training and validation cohorts were 0.840 (0.792–0.889) and 0.898 (0.833–0.963), respectively, and the AUCs were 0.840 (0.791–0.89) and 0.898 (0.833–0.964), respectively (Figure 5a, b). The calibration curves were good predictors and observers in terms of early death risk (Figure 5c, d). In addition, DCA showed a significant improvement in the net benefit of the nomogram with a wide range of threshold probabilities in both the training and validation sets, implying good clinical application value of this nomogram (Figure 5e, f).

Figure 4.

Risk nomogram for predicting early death in patients with combined small cell lung cancer and distant metastasis. Tumor site code: 1: left upper lobe; 2: left lower lobe; 3: right upper lobe; 4: right middle lobe; 5: right lower lobe; 6: main bronchus; 7: unspecific.

Figure 5.

Validation of the risk nomogram. Receiver operating characteristic curve (a), calibration curve (c), and decision curve analysis (DCA) (e) in the training cohort, and receiver operating characteristic curve (b), calibration curve (d), and DCA (f) in the validation cohort. EM, early death; AUC, area under the curve.

Discussion

To the best of our knowledge, this was the first study to identify independent risk factors for DM and early death in patients with CSCLC based on a large-scale, population-based, national cancer database. We also established two risk nomograms with reliable accuracy and discriminative ability and validated these by ROC, calibration, and DCA curves. The results suggested that these two nomograms may provide a practical tool to help clinicians identify patients at high risks of DM and early death, as well as to determine the optimal clinical interventions for patients diagnosed with metastatic CSCLC.

The incidence of DM at the time of initial diagnosis of SCLC is generally more than 60%, and patients with metastatic SCLC have poorer overall and median survival,¹⁴ whereas the incidence of DM in the current cohort was only 46.3%. Accurately identifying patients at high risk of metastasis may facilitate the implementation of individualized therapeutic strategies; however, despite extensive research on extensive-stage SCLC, studies focusing on metastatic CSCLC are scarce. Cai et al. indicated that the liver was the most common metastatic site and the lung was the least common,¹⁵ and Lu et al. established a diagnostic nomogram based on age, sex, N stage, and tumor size to predict liver metastasis with good accuracy in SCLC.¹⁶ Other researchers exploring metastasis-related factors in SCLC validated some laboratory markers, such as interferon-induced transmembrane protein 1, KMT2C, and DLL3, as useful and efficient biomarkers to estimate the risk of metastasis in SCLC^17–19; however, these are expensive, time-consuming, and invasive, thus limiting their clinical use, and their accuracy and efficiency have not been confirmed in large cohorts. Nomograms based on clinical variables have been used to predict DM efficiently in other tumors. Chao et al. combined age, sex, T stage, N stage, histology, tumor location, and pathological grade to predict metastasis in esophageal cancer,²⁰ while a nomogram based on race, age, primary site, tumor depth, size, and grade was established in patients with superficial gastric cancer without lymph node metastasis.²¹ Nevertheless, no studies have concentrated on metastasis in CSCLC based on clinical characteristics.

In the current study, we identified four risk factors for DM in CSCLC using logistic regression analyses, namely sex, tumor site, T stage, and N stage. As for other tumors, T stage and N stage were confirmed to be associated with DM in CSCLC.^19,22 Patients with centrally located primary tumors have visibly worse survival outcomes compared with patients with peripheral SCLC.²³ This may support the current finding that DM was more frequently associated with CSCLC located in the main bronchus, but further investigations are required to clarify this phenomenon. Males are predominant in both SCLC and CSCLC,^12,24,25 and male sex was associated with a high probability of DM, as reported in previous studies.^8,14,22

Similarly, a novel tool was established to predict early death in patients with uterine sarcoma, which provided a better prediction than the FIGO stage system.²⁶ Shi et al. also built a model to predict early death in patents with stage IV esophageal cancer based on histological grade, chemotherapy, liver metastasis, and bone metastasis.²⁷ These studies demonstrate the feasibility and significance of nomograms for predicting early mortality in patients with cancer. Notably however, recent studies have focused on the prognoses of patients with NSCLC and SCLC without metastases or with brain metastasis, and early death in patients with CSCLC and DM has not been explored. Shen et al. revealed that age and liver metastasis were risk factors for early death in patients with lung cancer with synchronous brain metastasis,²⁸ while another study demonstrated that age, surgery, and chemotherapy were independent risk factors for early death in patients with advanced epithelial ovarian cancer (both FIGO stage III–IV).²⁹ In line with these studies, we found that age was significantly correlated with early death in patients with metastatic CSCLC. Wang et al. reported that extensive disease, no surgery, elevated neuron-specific enolase, and elevated neutrophil/lymphocyte ratio were associated with worse prognosis in patients with CSCLC,²⁵ while surgical resection significantly improved overall and lung cancer-specific survival in a propensity score-matched analysis of stage III SCLC patients from America and China.³⁰ In the current study, surgery improved the risk of early death in patients with metastatic CSCLC; notably however, the limited number of patients with metastatic CSCLC undergoing surgery means that this result must be treated with caution, and more large prospective clinical trials are needed to clarify the benefits of surgery in patients with CSCLC and DM. In a study of 207 patients with metastatic SCLC, response to chemotherapy was the most important factor related to a better prognosis.³¹ Although chemotherapy plus immune checkpoint inhibitors are recommended for patients with extensive-stage SCLC, platinum plus etoposide has also been recommended for many years and is a common treatment for advanced SCLC in clinical practice.^32,33 Chemotherapy reduced the risk of early death in patients with metastatic CSCLC in our study, consistent with previous reports.^34,35 The benefits of radiotherapy for metastatic CSCLC thus remain unknown and require further investigation.

Liver metastasis was usually an early event whereas bone and brain metastases tended to be late occurrences in a series of 1009 patients with SCLC.²³ In addition, patients with extensive-stage with liver metastasis alone or in combination with other organs have worse outcomes. The liver is the most common site of DM in solid tumors because of its anatomical location and vascular features.³⁶ Moreover, Wu et al. revealed that liver metastasis was an independent prognostic factor leading to shorter progression-free and overall survival in patients with stage IV lung adenocarcinoma (median progression-free survival, 6.7 vs 11.2 months, P < 0.001; median overall survival, 9.2 vs 17.5 months, P < 0.001).³⁷ As expected and in line with published data, our study showed that the liver was the most metastatic organ, associated with a significantly increased risk of early death.

This study had some limitations. First, this was a retrospective study with potential selection bias, and prospective randomized controlled studies are therefore required to confirm our results. Second, inherent bias could not be avoided because of the lack of external validation. Third, no detailed information on the CSCLC components was included because the data are not available in the SEER database. Finally, the data on tumor grade and metastatic organs were incomplete, and their impact on metastasis and early death may thus have been underestimated.

Conclusion

We comprehensively demonstrated that sex, tumor site, T stage, and N stage were independent risk factors for DM in patients with CSCLC, while age, liver metastasis, surgery, and chemotherapy were independent risk factors associated with early death in patients with CSCLC and DM. These simple-to-use nomograms will facilitate precise clinical decision-making and improve the prognosis for patients with metastatic CSCLC.

Supplemental Material

sj-pdf-1-imr-10.1177_03000605241238689 - Supplemental material for Risk factors and nomograms for diagnosis and early death in patients with combined small cell lung cancer with distant metastasis: a population-based study

Supplemental material, sj-pdf-1-imr-10.1177_03000605241238689 for Risk factors and nomograms for diagnosis and early death in patients with combined small cell lung cancer with distant metastasis: a population-based study by Hui Yin, Zhi Hu and Jie Yang in Journal of International Medical Research

Footnotes

Acknowledgements

We thank the developers of the R package who shared their code freely. We also acknowledge the contribution of the SEER database for access to the data.

Author contributions

The study was conceptualized and designed by HY and ZH. The figures and tables were generated by HY. Statistical results were completed by HY. The initial manuscript was written by HY and ZH. This study was scrutinized in every aspect by JY who critically reviewed the original article.

Declaration of conflicting interests

None.

Funding

None.

ORCID iD

Jie Yang

References

Sung

Ferlay

Siegel

, et al. Global Cancer Statistics 2020: GLOBOCAN Estimates of Incidence and Mortality Worldwide for 36 Cancers in 185 Countries. CA Cancer J Clin 2021; 71: 209–249. 2021/02/05. DOI: 10.3322/caac.21660.

Travis

Brambilla

Nicholson

, et al. The 2015 World Health Organization Classification of Lung Tumors: Impact of genetic, clinical and radiologic advances since the 2004 Classification. J Thorac Oncol 2015; 10: 1243–1260. 2015/08/21. DOI: 10.1097/JTO.0000000000000630.

Qin

Combined small-cell lung carcinoma. Onco Targets Ther 2018; 11: 3505–3511. 2018/06/29. DOI: 10.2147/OTT.S159057.

Rudin

Brambilla

Faivre-Finn

, et al. Small-cell lung cancer. Nat Rev Dis Primers 2021; 7: 3. 2021/01/16. DOI: 10.1038/s41572-020-00235-0.

Winslow

Sage

Mechanisms of small cell lung cancer metastasis. EMBO Mol Med 2021; 13: e13122. 2020/12/10. DOI: 10.15252/emmm.202013122.

Lei

Feng

Qiang

, et al. Clinical characteristics and prognostic factors of surgically resected combined small cell lung cancer: a retrospective study. Lung Cancer 2020; 146: 244–251. 2020/06/28. DOI: 10.1016/j.lungcan.2020.06.021.

Tao

, et al. Development of nomograms for predicting lymph node metastasis and distant metastasis in newly diagnosed T1-2 non-small cell lung cancer: a population-based analysis. Front Oncol 2021; 11: 683282. 2021/09/28. DOI: 10.3389/fonc.2021.683282.

Liu

Jia

Diagnostic and prognostic nomograms for bone metastasis in small cell lung cancer. J Int Med Res 2021; 49: 3000605211050735. 2021/10/26. DOI: 10.1177/03000605211050735.

Xia

, et al. Predictive value of neuron-specific enolase, neutrophil-to-lymphocyte-ratio and lymph node metastasis for distant metastasis in small cell lung cancer. Clin Respir J 2020; 14: 1060–1066. 2020/08/05. DOI: 10.1111/crj.13242.

10.

Chen

, et al. FLI1 exonic circular RNAs as a novel oncogenic driver to promote tumor metastasis in small cell lung cancer. Clin Cancer Res 2019; 25: 1302–1317. 2018/11/16. DOI: 10.1158/1078-0432.Ccr-18-1447.

11.

Chen

, et al. LncRNA CCAT2 predicts poor prognosis and regulates growth and metastasis in small cell lung cancer. Biomed Pharmacother 2016; 82: 583–588. 2016/07/30. DOI: 10.1016/j.biopha.2016.05.017.

12.

Zhang

Chen

, et al. Prognostic factor analysis of patients with small cell lung cancer: Real-world data from 988 patients. Thorac Cancer 2021; 12: 1841–1850. 2021/05/07. DOI: 10.1111/1759-7714.13846.

13.

Von Elm

Altman

Egger

, et al. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. Ann Intern Med 2007; 147: 573–577. 2007/10/17. DOI: 10.7326/0003-4819-147-8-200710160-00010.

14.

Riihimaki

Hemminki

Fallah

, et al. Metastatic sites and survival in lung cancer. Lung Cancer 2014; 86: 78–84. 2014/08/19. DOI: 10.1016/j.lungcan.2014.07.020.

15.

Cai

Wang

, et al. The prognostic analysis of different metastatic patterns in extensive-stage small-cell lung cancer patients: a large population-based study. Future Oncol 2018; 14: 1397–1407. 2018/01/24. DOI: 10.2217/fon-2017-0706.

16.

Yang

Yuan

, et al. A novel nomogram based on SEER database for the prediction of liver metastasis in patients with small-cell lung cancer. Ann Palliat Med 2020; 9: 3123–3137. 2020/08/14. DOI: 10.21037/apm-20-886.

17.

Sakamoto

Inoue

Kohda

, et al. Interferon-induced transmembrane protein 1 (IFITM1) promotes distant metastasis of small cell lung cancer. Int J Mol Sci 2020; 21: 4934. DOI: 10.3390/ijms21144934.

18.

Sharma

Pourat

Abdel-Atti

, et al. Noninvasive interrogation of DLL3 expression in metastatic small cell lung cancer. Cancer Res 2017; 77: 3931–3941. DOI: 10.1158/0008-5472.Can-17-0299.

19.

Pan

Chen

, et al. KMT2C deficiency promotes small cell lung cancer metastasis through DNMT3A-mediated epigenetic reprogramming. Nat Cancer 2022; 3: 753–767. 2022/04/23. DOI: 10.1038/s43018-022-00361-6.

20.

Zhu

You

Liu

, et al. A nomogram to predict distant metastasis for patients with esophageal cancer. Oncol Res Treat 2020; 43: 2–9. 2019/11/13. DOI: 10.1159/000503613.

21.

Chen

Zhang

, et al. A clinical model to predict distant metastasis in patients with superficial gastric cancer with negative lymph node metastasis and a survival analysis for patients with metastasis. Cancer Med 2021; 10: 944–955. 2020/12/23. DOI: 10.1002/cam4.3680.

22.

Chu

Song

Brain metastasis in patients with small cell lung cancer. Int J Gen Med 2021; 14: 10131–10139. 2022/01/08. DOI: 10.2147/ijgm.S342009.

23.

Megyesfalvi

Tallosy

Pipek

, et al. The landscape of small cell lung cancer metastases: Organ specificity and timing. Thorac Cancer 2021; 12: 914–923. 2021/02/04. DOI: 10.1111/1759-7714.13854.

24.

Men

Hui

Liang

, et al. Further understanding of an uncommon disease of combined small cell lung cancer: clinical features and prognostic factors of 114 cases. Chin J Cancer Res 2016; 28: 486–494. DOI: 10.21147/j.issn.1000-9604.2016.05.03.

25.

Wang

Jiang

Prognostic significance of pretreatment laboratory parameters in combined small-cell lung cancer. Cell Biochem Biophys 2014; 69: 633–640. DOI: 10.1007/s12013-014-9845-3.

26.

Song

Wang

Zhang

, et al. A novel tool to predict early death in uterine sarcoma patients: A Surveillance, Epidemiology, and End Results-Based Study. Front Oncol 2020; 10: 608548. 2020/12/17. DOI: 10.3389/fonc.2020.608548.

27.

Shi

Zhai

GQ.

Models for predicting early death in patients with stage IV esophageal cancer: A Surveillance, Epidemiology, and End Results-Based Cohort Study. Cancer Control 2022; 29: 10732748211072976. 2022/01/18. DOI: 10.1177/10732748211072976.

28.

Shen

Deng

Chen

, et al. The incidence, risk factors and predictive nomograms for early death of lung cancer with synchronous brain metastasis: a retrospective study in the SEER database. BMC Cancer 2021; 21: 825. 2021/07/18. DOI: 10.1186/s12885-021-08490-4.

29.

Song

Zhou

Bai

, et al. A practical nomogram to predict early death in advanced epithelial ovarian cancer. Front Oncol 2021; 11: 655826. 2021/04/06. DOI: 10.3389/fonc.2021.655826.

30.

Gao

Shen

Wang

, et al. Propensity score matched analysis for the role of surgery in stage III small cell lung cancer based on the eighth edition of the TNM classification: a population study of the US SEER database and a Chinese hospital. Lung Cancer 2021; 162: 54–60. DOI: 10.1016/j.lungcan.2021.10.009.

31.

Tas

Aydiner

Topuz

, et al. Factors influencing the distribution of metastases and survival in extensive disease small cell lung cancer. Acta Oncol 1999; 38: 1011–1015. 2000/02/09. DOI: 10.1080/028418699432275.

32.

Ganti

AKP

Loo

Bassetti

, et al. Small Cell Lung Cancer, Version 2.2022, NCCN Clinical Practice Guidelines in Oncology. J Natl Compr Canc Netw 2021; 19: 1441–1464. 2021/12/14. DOI: 10.6004/jnccn.2021.0058.

33.

Dingemans

Früh

Ardizzoni

, et al. Small-cell lung cancer: ESMO Clinical Practice Guidelines for diagnosis, treatment and follow-up. Ann Onco 2021; 32: 839–853. DOI: 10.1016/j.annonc.2021.03.207.

34.

Wang

Yang

, et al. Development and validation of a nomogram prognostic model for SCLC patients. J Thorac Oncol 2018; 13: 1338–1348. 2018/06/15. DOI: 10.1016/j.jtho.2018.05.037.

35.

Zhang

Wang

, et al. Nomogram predicts risk and prognostic factors for bone metastasis of pancreatic cancer: a population-based analysis. Front Endocrinol (Lausanne) 2021; 12: 752176. 2022/04/01. DOI: 10.3389/fendo.2021.752176.

36.

Obenauf

Massagué

Surviving at a distance: organ-specific metastasis. Trends in Cancer 2015; 1: 76–91. 2015/09/01. DOI: 10.1016/j.trecan.2015.07.009.

37.

Tsai

Yang

, et al. Liver metastasis predicts poorer prognosis in stage IV lung adenocarcinoma patients receiving first-line gefitinib. Lung Cancer 2015; 88: 187–194. 2015/03/10. DOI: 10.1016/j.lungcan.2015.02.012.

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

0.16 MB