The Prognosis of Patients With Large Cell Carcinoma of the Cervix: A Retrospective Study of the SEER Database and a Chinese Multi-Center Registry

Abstract

Introduction

Large cell neuroendocrine carcinoma of the cervix (LCNEC), a rare and aggressive malignancy, has poor survival. Due to the scarcity of specific treatment guidelines and the limitations of small retrospective studies, this study aims to assess prognostic factors and treatment outcomes in cervical LCNEC using data from both the SEER database and a Chinese multicenter registry.

Methods

This retrospective multicenter cohort study included 130 patients diagnosed with cervical LCNEC between 2004 and 2021, including 87 from the SEER database and 43 from the Chinese registry. We compared clinicopathological variables, treatment approaches such as surgery, chemotherapy, radiotherapy and survival outcomes. Kaplan-Meier and Cox regression analyses were employed to identify prognostic factors.

Results

The Chinese cohort exhibited a lower proportion of metastatic cases (13.95% compared to 49.43%, p<0.001), earlier disease stages (AJCC IV: 13.95% vs. 49.43%, p<0.001), and higher rates of surgery (76.74% vs. 56.32%, p=0.023) and chemotherapy (93.02% vs. 83.91%, p=0.147). The median overall survival was 22.0 months (95% CI: 17.0–29.5), with the Chinese cohort demonstrating superior 5-year survival (44.0% vs. 27.3%, p<0.001). Surgery was found to reduce the risk of mortality in both cohorts, particularly for locally advanced disease. Chemotherapy improved survival in the SEER cohort, while radiotherapy showed no significant benefit. Advanced T-stage and nodal involvement were identified as predictors of poor prognosis.

Conclusion

Surgery and chemotherapy remain cornerstone interventions for survival improvement, with surgical resection showing particularly pronounced benefits in locally advanced cases. Prospective studies are needed to optimize guidelines.

Keywords

large cell neuroendocrine carcinoma of the cervix (LCNEC)prognosis SEER database Chinese multi-center registry surgery

Introduction

Cervical cancer remains a considerable global health concern, with an estimated 661,021 new cases and 348,189 deaths in 2022, even with advances in HPV vaccination and screening.¹ Among its histological varieties, neuroendocrine carcinomas (NECs) of the cervix are exceptionally uncommon and aggressive, comprising less than 2% of invasive cervical cancers.^2-4 Large cell neuroendocrine carcinoma of the cervix (LCNEC), a distinct NEC subtype, is notable for its high mitotic rate, rapid progression, and early lymphovascular invasion and distant metastasis.^5,6 Compared to more prevalent squamous cell carcinoma or adenocarcinoma, LCNEC has a significantly poorer prognosis, with reported 5-year survival rates below 25% even in early stages.^7,8

Current treatment approaches for LCNEC are often based on guidelines for neuroendocrine carcinoma or other aggressive cervical cancers. The NCCN guidelines recognize the distinctiveness of neuroendocrine tumors but offer limited stage-specific therapy guidance for LCNEC.⁹ Currently, the prevailing treatment approaches include surgery, chemotherapy, and radiotherapy.^4,10 Radical surgery combined with chemotherapy stands as the prevalent choice, particularly for patients with resectable tumors in the early stage.^10,11 However, their precise impact on prognosis remains unclear, with conflicting findings reported in current studies.^12-15

The scarcity of LCNEC has impeded the conduct of prospective clinical trials, resulting in significant evidence-based management gaps. Existing studies are often single-center, small-scale, or variably designed, limiting their broader applicability.^8,11 Additionally, LCNEC’s biological behavior and treatment response may differ from other neuroendocrine subtypes, necessitating targeted research.¹⁶ Large-scale, multi-institutional studies are urgently required to clarify prognostic factors, refine treatment algorithms, and validate stage-specific interventions.

In this study, we undertook a retrospective analysis, integrating data from the SEER database and a Chinese multi-center registry. Our aim was to assess the influence of different factors on survival outcomes in cervical LCNEC. By combining two geographically distinct patient groups, we hoped to identify tailored treatment strategies and contribute evidence to enhance clinical guidelines for this underserved malignancy.

Methods

Patient Selection

In this retrospective study, we utilized data from a multi-center registry in China and the Surveillance, Epidemiology, and End Results (SEER) database. The Chinese registry comprised data on patients with LCNEC diagnosed between January 1, 2004, and December 31, 2021, at Fujian Cancer Hospital and Shandong Cancer Hospital. Inclusion criteria were a pathological diagnosis of primary LCNEC (based on the WHO Classification of Tumors of Female Reproductive Organs, with ICD-O-3 code 8013/3). Exclusion criteria were non-primary tumors (such as metastatic carcinomas) and cases without complete survival information.

Using the SEER database of the National Cancer Institute (NCI) (SEER*Stat Database: Incidence - SEER Research Data, 22 Registries, Nov 2023 Sub [2000-2021] - Linked To County Attributes - Time Dependent [1990-2022] Income/Rurality, 1969-2022 Counties, National Cancer Institute, DCCPS, Surveillance Research Program, released April 2024, based on the November 2023 submission), we selected patients diagnosed with cervical LCNEC between January 1, 2004, and December 31, 2021, to form the SEER cohort. Cases were limited to those with a primary site of the cervix (ICD-O-3 codes: C53.0, C53.1, C53.8, C53.9) and a histologic code (ICD-O-3 code) of 8013/3 (large cell neuroendocrine carcinoma). All patients required pathological confirmation and complete survival data. Patients without primary tumors were excluded.

Statistical Analysis

Descriptive statistical analysis was used to analyze the data. The baseline categorical characteristics were compared between different groups (SEER database group and Chinese registry group, surgery group and non-surgery group) using the χ2 test or Fisher’s exact test. The overall survival (OS) from diagnosis to death from any cause or last follow-up was evaluated. The final follow-up date for survival analyses was April 10, 2025. Kaplan-Meier method was used to construct survival curves, and the log-rank trend test was employed to determine the statistical significance of differences in Kaplan-Meier survival.

The Cox proportional hazards model was utilized for both univariate and multivariate survival analyses. All variables (Year of diagnosis, Age, Marital status, Tumor size, AJCC 8^th stage, FIGO 2018^th stage, Surgery, Radiation, Chemotherapy) were included in the univariate Cox regression analyses. Variables that significant in the univariate analyses or clinically considered to affect patient prognosis were incorporated into the multivariate Cox regression analyses. Hazard ratios (HR) were reported with 95% confidence intervals (CI). All tests were two-sided, and a p-value < 0.05 was considered statistically significant. All statistical analyses were performed using R (version 4.3.2).

Ethics Statement

This retrospective multicenter cohort study was conducted in accordance with the Declaration of Helsinki as revised in 2024. The reporting of this study conforms to the STROBE guidelines.¹⁷ Ethical approval was obtained from the Institutional Review Board at Fujian Provincial Cancer Hospital (approval No. K2024-206-01; approval date: June 12, 2024). Written informed consent was obtained from all patients in the Chinese multi-center registry. The SEER database contains de-identified publicly available data and therefore did not require additional institutional ethics approval or informed consent for secondary analysis. All patient details were de-identified before analysis. Patients from the Chinese multi-center registry were included consecutively during the study period.

Results

Clinicopathological Features

A total of 130 patients with LCNEC were included in this study: 87 from the SEER database and 43 from the Chinese registry. Baseline characteristics are summarized in Table 1. The Chinese cohort had a higher proportion of patients diagnosed between 2018 and 2021 (88.37% vs. 25.29% in SEER, p<0.001) and were younger (mean age 45±10 vs. 49±15 years, p=0.065). The SEER cohort had a higher proportion of tumors >40mm (36.78% vs. 13.95%, p=0.003). Notably, the Chinese cohort had significantly fewer metastatic cases (13.95% vs. 49.43%, p<0.001) and earlier AJCC 8^th stages (Stage IV: 13.95% vs. 49.43%, p<0.001). Additionally, surgery (76.74% vs. 56.32%, p=0.023) and chemotherapy (93.02% vs. 83.91%, p=0.147) were more common in Chinese cohort.

Table 1.

Patient Demographics and Baseline Characteristics From the SEER Database and Chinese Registry

Characteristic	Group		p-value
Characteristic	SEER database,N = 87¹	Chinese registry,N = 43¹	p-value
Year of diagnosis			<0.001²
2004-2012	27 (31.03%)	2 (4.65%)
2013-2017	38 (43.68%)	3 (6.98%)
2018-2021	22 (25.29%)	38 (88.37%)
Age	49 ± 15	45 ± 10	0.065³
Marital status			<0.001²
Married	43 (49.43%)	36 (83.72%)
Others	44 (50.57%)	7 (16.28%)
Tumor Size			0.003²
≤20mm	15 (17.24%)	6 (13.95%)
20-40mm	18 (20.69%)	22 (51.16%)
>40mm	32 (36.78%)	6 (13.95%)
Unknown	22 (25.29%)	9 (20.93%)
T			<0.001⁴
T1	41 (47.13%)	12 (27.91%)
T2	16 (18.39%)	16 (37.21%)
T3	19 (21.84%)	5 (11.63%)
T4	6 (6.90%)	0 (0.00%)
Unknown	5 (5.75%)	10 (23.26%)
N			0.004⁴
N0	41 (47.13%)	29 (67.44%)
N1	46 (52.87%)	12 (27.91%)
Unknown	0 (0.00%)	2 (4.65%)
M			<0.001²
M0	44 (50.57%)	37 (86.05%)
M1	43 (49.43%)	6 (13.95%)
AJCC 8 ^th stage			<0.001⁴
Stage I	27 (31.03%)	13 (30.23%)
Stage II	6 (6.90%)	11 (25.58%)
Stage III	11 (12.64%)	5 (11.63%)
Stage IV	43 (49.43%)	6 (13.95%)
Unknown	0 (0.00%)	8 (18.60%)
FIGO 2018 ^th stage			<0.001⁴
Stage I	27 (31.03%)	8 (18.60%)
Stage II	4 (4.60%)	9 (20.93%)
Stage III	13 (14.94%)	13 (30.23%)
Stage IV	43 (49.43%)	4 (9.30%)
Unknown	0 (0.00%)	9 (20.93%)
Surgery			0.023²
No	38 (43.68%)	10 (23.26%)
Yes	49 (56.32%)	33 (76.74%)
Radiation			0.188²
No	43 (49.43%)	16 (37.21%)
Yes	44 (50.57%)	27 (62.79%)
Chemotherapy			0.147²
No	14 (16.09%)	3 (6.98%)
Yes	73 (83.91%)	40 (93.02%)

¹n (%); Mean ± SD; Median (IQR).

²Pearson’s Chi-squared test.

³Welch Two Sample t-test.

⁴Fisher’s exact test.

Survival and Prognostic Factor Analyses

The median OS for the entire cohort was 22.00 months (95% CI, 17.00–29.50) (Table S1) and the 1-, 3-, and 5-year OS rates were 67.0% (95% CI, 59.3%, 75.7%), 35.9% (95% CI, 27.9%, 46.1%) and 33.2% (95% CI, 25.2%, 43.6%), respectively (Table 2). The Chinese registry demonstrated superior survival at 12 months (90.5% [95% CI, 82.1–99.8%] vs. 55.2% [95% CI, 45.5–66.9%], p<0.001), 36 months (48.0% [95% CI, 33.5–68.7%] vs. 29.3% [95% CI, 20.6–41.5%]), and 60 months (44.0% [95% CI, 29.6–65.5%] vs. 27.3% [95% CI, 18.8–39.7%]) (Table 2).

Table 2.

The Median Survival Rate of Patients From the SEER Database and Chinese Registry

Characteristic	N	Event N	12-months	36-months	60-months	p-value¹
Overall	130	79	67.0% (59.3%, 75.7%)	35.9% (27.9%, 46.1%)	33.2% (25.2%, 43.6%)
Group	130	79				<0.001
SEER database			55.2% (45.5%, 66.9%)	29.3% (20.6%, 41.5%)	27.3% (18.8%, 39.7%)
Chinese registry			90.5% (82.1%, 99.8%)	48.0% (33.5%, 68.7%)	44.0% (29.6%, 65.5%)

¹Log-rank test.

In the univariate Cox regression analyses, tumors exceeding 40 mm in size were tied to a heightened mortality risk in the SEER cohort (HR 4.30; 95% CI 1.64–11.29; p=0.003), though not in the Chinese cohort. Advanced T stages (T2, T3, T4) indicated a poorer prognosis in both groups. In the SEER cohort, nodal involvement (N1 vs N0) was associated with a higher mortality rate (HR, 3.43; 95% CI, 1.98–5.94; p<0.001). FIGO 2018^th stages II, III, and IV demonstrated notable or marginally significant correlations with mortality compared to stage I in the SEER cohort. Surgery proved to be a significant protective factor in both cohorts, reducing mortality risk. Chemotherapy showed a marginally beneficial effect in the China cohort and a significant one in the SEER cohort. Radiation therapy, however, had no significant impact on mortality in either group.

In the multivariate analyses, after adjusting for confounders, no factors were significantly associated with mortality in the Chinese registry. However, in the multivariate analyses of SEER database, patients with higher T stage exhibited substantially increased mortality risk, with T3 (HR, 6.18; 95% CI, 2.70-14.15, p<0.001), T4 (HR, 13.30; 95% CI, 4.03-43.93, p<0.001) showing progressively greater hazards compared to T1. Surgical intervention significantly reduced mortality risk (HR, 0.30; 95% CI, 0.16-0.57, p<0.001), as did chemotherapy receipt (HR, 0.30; 95% CI, 0.13-0.68, p=0.004). The forest plots of the results of the multivariate analyses are shown in Figures 1 and 2. The results of univariate Cox regression analyses are shown in Table 3 and Table 4.

Figure 1.

Forest plot of multivariate cox regression analyses in the SEER database

Figure 2.

Forest plot of multivariate cox regression analyses in the Chinese registry

Table 3.

Univariate and Multivariate Analysis of Patients From SEER Database

Characteristic	Univariate					Multivariate
Characteristic	N	Event N	HR¹	95% CI¹	p-value²	N	Event N	HR¹	95% CI¹	p-value²
Age	87	60	1.03	1.01, 1.05	<0.001***
Year of diagnosis
2004-2012	27	21	—	—		27	21	—	—
2013-2017	38	28	1.17	0.66, 2.06	0.590	38	28	1.79	0.92, 3.45	0.084
2018-2021	22	11	0.87	0.42, 1.83	0.721	22	11	0.72	0.31, 1.68	0.450
Marital status
Married	43	29	—	—
Other	44	31	1.12	0.67, 1.86	0.660
Tumor size
≤20mm	15	5	—	—
20-40mm	18	11	2.73	0.95, 7.89	0.063
>40mm	32	25	4.30	1.64, 11.29	0.003**
Unknown	22	19	7.47	2.75, 20.33	<0.001***
T
T1	41	21	—	—		41	21	—	—
T2	16	10	2.04	0.95, 4.38	0.067	16	10	2.11	0.86, 5.19	0.105
T3	19	18	9.05	4.50, 18.19	<0.001***	19	18	6.18	2.70, 14.15	<0.001***
T4	6	6	12.96	4.77, 35.18	<0.001***	6	6	13.30	4.03, 43.93	<0.001***
Unknown	5	5	15.77	5.25, 47.37	<0.001***	5	5	6.18	1.66, 23.01	0.007**
N
N0	41	20	—	—		41	20	—	—
N1	46	40	3.43	1.98, 5.94	<0.001***	46	40	1.87	1.03, 3.40	0.041*
M
M0	44	20	—	—
M1	43	40	4.23	2.43, 7.39	<0.001***
AJCC 8 ^th stage
Stage I	27	10	—	—
Stage II	6	2	0.95	0.21, 4.35	0.950
Stage III	11	8	4.35	1.71, 11.10	0.002**
Stage IV	43	40	6.19	3.04, 12.61	<0.001***
FIGO 2018 ^th stage
Stage IV	43	40	—	—
Stage I	27	10	0.16	0.08, 0.33	<0.001***
Stage II	4	2	0.22	0.05, 0.92	0.038*
Stage III	13	8	0.53	0.25, 1.13	0.101
Surgery
No	38	34	—	—		38	34	—	—
Yes	49	26	0.24	0.14, 0.40	<0.001***	49	26	0.30	0.16, 0.57	<0.001***
Radiation
No	43	31	—	—
Yes	44	29	0.76	0.46, 1.27	0.299
Chemotherapy
No	14	12	—	—		14	12	—	—
Yes	73	48	0.34	0.18, 0.65	0.001**	73	48	0.30	0.13, 0.68	0.004**

¹HR = Hazard Ratio, CI = Confidence Interval.

²*p<0.05; **p<0.01; ***p<0.001.

n = 87.0; N events = 60.0; statistic.log = 83.1; p.value.log = 0.000; statistic.sc = 94.5; p.value.sc = 0.000; statistic.wald = 67.5; p.value.wald = 0.000; statistic.robust = NA; p.value.robust = NA; R² = 0.615; r.squared.max = 0.995; c-index = 0.824; c-index SE = 0.022; Log-likelihood = -190; AIC = 402; BIC = 425; No. Obs. = 87.

Table 4.

Univariate and Multivariate Analysis of Patients From Chinese Registry

Characteristic	Univariate					Multivariate
Characteristic	N	Event N	HR¹	95% CI¹	p-value²	N	Event N	HR¹	95% CI¹	p-value²
Age	43	19	1.05	1.00, 1.10	0.044*	43	19	1.04	0.95, 1.14	0.449
Year of diagnosis
2004-2012	2	2	—	—
2013-2017	3	1	0.36	0.03, 4.04	0.409
2018-2021	38	16	0.69	0.16, 3.03	0.622
Marital status
Married	36	17	—	—
Others	7	2	0.68	0.16, 2.97	0.613
Tumor size
≤20mm	6	3	—	—		6	3	—	—
20-40mm	22	8	1.08	0.28, 4.07	0.914	22	8	0.91	0.13, 6.22	0.922
>40mm	6	3	2.18	0.43, 10.99	0.344	6	3	2.79	0.24, 32.70	0.415
Unknown	9	5	1.29	0.31, 5.41	0.729	9	5	0.14	0.01, 2.09	0.155
T
T1	12	2	—	—
T2	16	8	5.82	1.22, 27.81	0.027*
T3	5	3	9.93	1.58, 62.52	0.014*
Unknown	10	6	4.81	0.97, 23.90	0.055
N
N0	29	11	—	—
N1	12	6	1.60	0.59, 4.33	0.359
Unknown	2	2	3.64	0.79, 16.71	0.097
M
M0	37	14	—	—
M1	6	5	2.06	0.73, 5.81	0.173
AJCC 8 ^th stage
Stage I	13	3	—	—
Stage II	11	4	2.71	0.60, 12.21	0.193
Stage III	5	3	6.15	1.18, 32.05	0.031*
Stage IV	6	5	4.41	1.03, 18.80	0.045*
Unknown	8	4	2.36	0.53, 10.56	0.262
FIGO 2018 ^th stage
Stage IV	4	4	—	—		4	4	—	—
Stage I	8	0	0.00	0.00, Inf	0.998	8	0	0.00	0.00, Inf	0.998
Stage II	9	4	0.52	0.13, 2.12	0.363	9	4	0.23	0.03, 1.95	0.176
Stage III	13	5	0.51	0.14, 1.94	0.327	13	5	0.19	0.02, 1.55	0.122
Unknown	9	6	0.97	0.27, 3.54	0.965	9	6	2.40	0.46, 12.52	0.298
Surgery
No	10	8	—	—		10	8	—	—
Yes	33	11	0.29	0.12, 0.73	0.008**	33	11	0.33	0.05, 2.31	0.262
Radiation
No	16	6	—	—		16	6	—	—
Yes	27	13	1.15	0.44, 3.04	0.773	27	13	0.91	0.25, 3.39	0.893
Chemotherapy
No	3	2	—	—		3	2	—	—
Yes	40	17	0.59	0.14, 2.55	0.477	40	17	0.09	0.01, 1.00	0.050

¹HR = Hazard Ratio, CI = Confidence Interval.

²*p<0.05; **p<0.01; ***p<0.001.

n = 43.0; N events = 19.0; statistic.log = 30.7; p.value.log = 0.001; statistic.sc = 32.8; p.value.sc = 0.001; statistic.wald = 16.1; p.value.wald = 0.138; statistic.robust = NA; p.value.robust = NA; R² = 0.510; r.squared.max = 0.943; c-index = 0.842; c-index SE = 0.044; Log-likelihood = -46.1; AIC = 114; BIC = 125; No. Obs. = 43.

Surgical Intervention on Survival

Baseline characteristics of patients stratified by surgery status are presented in Table 5. We further investigated the prognostic impact of surgery on locally advanced LCNEC. Patients who underwent surgery had significantly better survival outcomes compared to those who did not both in SEER cohort and Chinese cohort(P=0.002 and P=0.170, respectively). The survival probability was consistently higher in the surgery group throughout the follow-up period. The details are shown in Figure 3.

Table 5.

Baseline Characteristics of Patients Stratified by Surgery Status

Characteristic	Seer database			Chinese registry
Characteristic	No surgery (N = 38)¹	Surgery(N = 49)¹	p-value²	No surgery(N = 10)¹	Surgery(N = 33)¹	p-value³
Year of diagnosis			0.651			0.498
2004-2012	10 (26.32%)	17 (34.69%)		1 (10.00%)	1 (3.03%)
2013-2017	17 (44.74%)	21 (42.86%)		0 (0.00%)	3 (9.09%)
2018-2021	11 (28.95%)	11 (22.45%)		9 (90.00%)	29 (87.88%)
Age	57 ± 15	44 ± 13	<0.001	52 ± 11	43 ± 9	0.047
Marital status			0.338			0.656
Married	21 (55.26%)	22 (44.90%)		8 (80.00%)	28 (84.85%)
Others	17 (44.74%)	27 (55.10%)		2 (20.00%)	5 (15.15%)
Tumor size			<0.001			0.027
≤20mm	0 (0.00%)	15 (30.61%)		0 (0.00%)	6 (18.18%)
20-40m	3 (7.89%)	15 (30.61%)		3 (30.00%)	19 (57.58%)
>40mm	22 (57.89%)	10 (20.41%)		2 (20.00%)	4 (12.12%)
Unknow	13 (34.21%)	9 (18.37%)		5 (50.00%)	4 (12.12%)
T			<0.001			0.003
T1	8 (21.05%)	33 (67.35%)		0 (0.00%)	12 (36.36%)
T2	6 (15.79%)	10 (20.41%)		2 (20.00%)	14 (42.42%)
T3	14 (36.84%)	5 (10.20%)		2 (20.00%)	3 (9.09%)
T4	5 (13.16%)	1 (2.04%)		0 (0.00%)	0 (0.00%)
Unknown	5 (13.16%)	0 (0.00%)		6 (60.00%)	4 (12.12%)
N			0.017			0.004
N0	12 (31.58%)	29 (59.18%)		3 (30.00%)	26 (78.79%)
N1	26 (68.42%)	20 (40.82%)		5 (50.00%)	7 (21.21%)
Unknown	0 (0.00%)	0 (0.00%)		2 (20.00%)	0 (0.00%)
M			<0.001			0.611
M0	8 (21.05%)	36 (73.47%)		8 (80.00%)	29 (87.88%)
M1	30 (78.95%)	13 (26.53%)		2 (20.00%)	4 (12.12%)
AJCC 8 ^th stage			<0.001			0.032
Stage I	2 (5.26%)	25 (51.02%)		0 (0.00%)	13 (39.39%)
Stage II	2 (5.26%)	4 (8.16%)		2 (20.00%)	9 (27.27%)
Stage III	4 (10.53%)	7 (14.29%)		2 (20.00%)	3 (9.09%)
Stage IV	30 (78.95%)	13 (26.53%)		2 (20.00%)	4 (12.12%)
Unknown	0 (0.00%)	0 (0.00%)		4 (40.00%)	4 (12.12%)
FIGO 2018 ^th stage			<0.001			0.147
Stage I	2 (5.26%)	25 (51.02%)		0 (0.00%)	8 (24.24%)
Stage II	2 (5.26%)	2 (4.08%)		1 (10.00%)	8 (24.24%)
Stage III	4 (10.53%)	9 (18.37%)		5 (50.00%)	8 (24.24%)
Stage IV	30 (78.95%)	13 (26.53%)		2 (20.00%)	2 (6.06%)
Unknown	0 (0.00%)	0 (0.00%)		2 (20.00%)	7 (21.21%)
Radiation			0.229			>0.999
No	16 (42.11%)	27 (55.10%)		4 (40.00%)	12 (36.36%)
Yes	22 (57.89%)	22 (44.90%)		6 (60.00%)	21 (63.64%)
Chemotherapy			0.267			0.130
No	8 (21.05%)	6 (12.24%)		2 (20.00%)	1 (3.03%)
Yes	30 (78.95%)	43 (87.76%)		8 (80.00%)	32 (96.97%)

¹n (%); Mean ± SD; Median (IQR).

²Pearson’s Chi-squared test; Welch Two Sample t-test; Fisher’s exact test; Wilcoxon rank sum test.

³Fisher’s exact test; Welch Two Sample t-test; Wilcoxon rank sum test.

Figure 3.

(A) Kaplan-Meier Survival Curves Stratified by Surgery Status in Locally Advanced patients of SEER Database; (B) Kaplan-Meier Survival Curves Stratified by Surgery Status in Locally Advanced patients of Chinese Registry

To further explore the prognostic value of surgery across disease stages, Kaplan–Meier survival analyses were performed stratified by FIGO 2018 stage (I–II vs. III–IV) in both the SEER and Chinese multi-center cohorts (Figures 4 and 5).

Figure 4.

Kaplan–Meier overall survival curves comparing surgical vs. non-surgical treatment in SEER patients stratified by FIGO 2018^th stage: (A) Stage I–II; (B) Stage III–IV

Figure 5.

Kaplan–Meier overall survival curves comparing surgical vs. non-surgical treatment in the Chinese registry stratified by FIGO 2018^th stage: (A) Stage I–II; (B) Stage III–IV

In the SEER cohort, no significant survival difference was observed between surgical and non-surgical patients with stage I–II disease (P=0.350). However, in stage III–IV patients, surgical treatment was associated with a markedly improved overall survival (P<0.001). A similar pattern was observed in the Chinese cohort: surgery did not confer a significant survival benefit in early-stage patients (P=0.560), while a significant improvement in overall survival was noted among stage III–IV patients (P=0.035).

Discussion

SCNEC accounts for only a very small proportion of cervical cancers, whereas SCC remains the predominant histologic subtype. Consistently, survival outcomes of cervical neuroendocrine carcinomas, including SCNEC, are substantially worse than those of SCC in database-based analyses.⁸ Current evidence for cervical LCNEC is limited, its clinical management is largely derived from the broader experience with cervical neuroendocrine carcinoma, especially small cell neuroendocrine carcinoma. However, LCNEC is not completely interchangeable with SCNEC and may differ in clinicopathological features, biological behavior, and therapeutic responsiveness. This highlights the importance of LCNEC-specific studies.^8,18 This study represented the systematic evaluation of prognostic factors and therapeutic strategies for LCNEC through integration of the SEER database with multi-center registry data from China. The results demonstrated that surgery and chemotherapy served as pivotal interventions for survival improvement, with surgery demonstrating particularly significant benefits in locally advanced patients.

The Chinese cohort demonstrated significantly superior survival outcomes compared to the SEER cohort. This disparity may stem from multiple factors. The predominant determinant was likely the higher proportion of metastatic cases in the SEER cohort. This between-cohort difference should be interpreted cautiously because the baseline stage distribution differed markedly between the two datasets. This observation aligned with findings from multinational comparative studies on small cell cervical carcinoma, further underscoring the clinical imperative for early detection and intervention.¹⁹

Our study demonstrated that surgery significantly reduced the mortality risk in patients with LCNEC. This finding was consistent with previous studies on cervical neuroendocrine carcinoma.^20,21 In the SEER cohort, locally advanced patients accepted surgery exhibited markedly prolonged median survival, supporting the pivotal role of surgery in disease control. Notably, although NCCN guidelines favored non-surgical approaches for locally advanced cervical cancer, our findings suggested that LCNEC may possess distinct biological behaviors, and its aggressive nature might be effectively controlled through surgery.^9,18 Our stage-stratified analyses indicate that the survival benefit of surgery is concentrated in patients with FIGO stage III–IV LCNEC. This pattern reflects a well-established principle from advanced epithelial ovarian cancer, in which reducing macroscopic tumor burden improves outcomes when combined with systemic therapy. Randomized trials (EORTC 55971 and CHORUS) demonstrated that neoadjuvant chemotherapy followed by interval debulking is non-inferior to primary debulking followed by chemotherapy, underscoring that the extent of residual disease—rather than the sequence of surgery—determines prognosis; the primary objective of cytoreduction is to achieve no macroscopic residual disease.^22,23 By analogy, LCNEC is a highly proliferative tumor with early metastatic potential, and debulking surgery may augment systemic therapy by reducing total tumor burden and chemo-resistant subclones, removing poorly perfused or sanctuary sites less accessible to chemotherapy or radiation, and improving performance status and symptom control to facilitate timely administration of full-dose multi-agent chemotherapy.²⁴ Contemporary retrospective evidence in cervical neuroendocrine carcinoma, including LCNEC, supports multimodal therapy and suggests an association between surgery—particularly when combined with platinum/etoposide—and improved survival, although effect sizes vary across series and stages. Given the rarity of LCNEC and the absence of randomized trials, the current evidence derives from retrospective cohorts, registries, and multicenter series. We therefore propose a conditional recommendation to consider aggressive cytoreductive surgery in carefully selected stage III–IV LCNEC patients as part of multimodal therapy, typically with platinum-based chemotherapy.^25,26

The survival advantage of chemotherapy was validated in the multivariate analyses, which aligned with the findings reported by Embry et al 2²⁷ Platinum-based agents suppressed tumor proliferation by inducing DNA crosslinking, while etoposide enhanced their cytotoxicity. The synergistic effect of these agents may improve distant metastasis control.²⁸ In the Chinese cohort, the majority of patients were subjected to chemotherapy, which may partially account for the observed survival benefit. Notably, radiotherapy did not demonstrate significant survival benefits in this study, potentially due to insufficient sample size or suboptimal timing of radiotherapy. Further research is warranted to optimize the role of radiotherapy in multimodal treatment strategies.

We found that T stage and lymph node metastasis were identified as independent prognostic factors, which was consistent with the highly aggressive nature of neuroendocrine carcinoma, characterized by its propensity for early invasion into parametrial tissues and the vascular system.²⁹ Notably, even in early-stage patients, the median survival time was only 19 months, significantly lower than that of conventional cervical cancer, highlighting the distinct malignant biological behavior of LCNEC. The associations of tumor size and stage with survival appeared less stable in the Chinese cohort than in the SEER cohort. This discrepancy should be interpreted cautiously and is likely related to the relatively small sample size and limited number of outcome events in the Chinese registry, which may have reduced statistical power and widened confidence intervals. Therefore, the absence of consistent significance in the Chinese cohort should not be interpreted as evidence of no prognostic effect, but rather as an indication that these estimates require validation in larger multicenter datasets. Further exploration of molecular biomarkers to guide risk stratification is warranted. Bellone et al 3³⁰ revealed through integrated genomic analyses that high-frequency alterations in high-grade neuroendocrine carcinoma of the cervix (HGNECC) include activation of the PI3K/AKT/mTOR pathway, EGFR amplification, and mutations in ATR/ATM pathway-related genes. Preclinical models demonstrated that copanlisib (a PI3K inhibitor) combined with afatinib (an EGFR inhibitor) or elimusertib (an ATR inhibitor) significantly suppressed tumor growth, likely through synergistic blockade of DNA damage repair and pro-proliferative signaling pathways. This finding suggested that molecular subtype-based drug combination strategies may overcome drug resistance associated with conventional chemotherapy.

The synergistic effects between PARP inhibitors and immunotherapy are equally noteworthy. Carroll et al 3³¹ demonstrated that the expression rates of PARP1/2 reached 89% in cervical neuroendocrine carcinoma, whereas the PD-L1 positivity rate was merely 21%. In PD-L1-negative patients, PARP inhibitors could enhance tumor cell sensitivity to DNA damage by inducing the “synthetic lethality” effect while simultaneously promoting tumor antigen release. The combination with PD-1/PD-L1 inhibitors may elicit an “abscopal effect.” Animal studies revealed that the combination of olaparib and pembrolizumab resulted in a reduction in tumor volume.³² Additionally, the “molecular signature-driven therapy” strategy proposed by Manchana et al 3³³ emphasized that targeting specific molecular features such as MYC amplification and RB1 deletion with CDK4/6 inhibitors or EZH2 inhibitors may provide precise therapeutic options for rare subtypes.

Despite utilizing the largest sample size to date for LCNEC, this study had several limitations. The retrospective nature of the study introduced inherent selection biases, as evidenced by the surgical cohort comprising younger patients with smaller tumors, potentially leading to overestimation of surgical efficacy. Furthermore, the SEER database lacked detailed chemotherapy regimen information (e.g., platinum dosage, number of cycles), which limited the assessment of chemotherapy intensity. The relatively short follow-up duration in the Chinese cohort may result in underestimation of long-term recurrence risk.

Conclusion

In summary, this study suggests that surgery and chemotherapy remain important components of treatment for cervical LCNEC, while surgical intervention may provide survival benefit in selected patients within a multimodal treatment framework. However, these findings should be interpreted cautiously given the retrospective design, the marked between-cohort differences in stage distribution, and the limited number of advanced-stage cases in the Chinese cohort. Further large-scale prospective studies are warranted to validate these findings and refine treatment strategies.

Supplemental Material

Supplemental material - The Prognosis of Patients With Large Cell Carcinoma of the Cervix: A Retrospective Study of the SEER Database and a Chinese Multi-Center Registry

Supplemental material for The Prognosis of Patients With Large Cell Carcinoma of the Cervix: A Retrospective Study of the SEER Database and a Chinese Multi-Center Registry by Yaxin Kang, Liangwu Zhuang, Lele Chang, Jing Liu, Cong Wang and Qin Xu in Cancer Control.

Footnotes

ORCID iD

Qin Xu

Ethical Considerations

This study was conducted in accordance with the Declaration of Helsinki as revised in 2024. Ethical approval was obtained from the Institutional Review Board at Fujian Provincial Cancer Hospital (approval No. K2024-206-01; approval date: June 12, 2024).

Consent to Participate

Written informed consent was obtained from all patients in the Chinese multi-center registry. For the SEER cohort, informed consent was not required because all data were de-identified and publicly available.

Consent for Publication

All the authors agreed to publish this work.

Author Contributions

YK and LZ proposed the main idea and drafted the initial version of the manuscript. LC and YK conducted the data analysis and data interpretation, while LZ and JL were responsible for data collection and validation. JL and CW conducted the data visualization and validation. QX and CW provided guidance on the methodology and structure of the article. QX and JL contributed to funding acquisition. QX and CW has carefully polished and revised the overall article.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by Fujian Cancer Hospital High-Level Talent Training Program (Grant number: 2023YNG09), Joint Funds for the innovation of science and Technology, Fujian province (Grant number: 2023Y9449), Natural Science Foundation of Fujian Province (Grant number: 2024J011099 and 2023J011273).

Declaration of Conflicting Interests

The authors declare that there were no potential conflicts of interest in this research.

Data Availability Statement

The data supporting this study’s findings are available from the corresponding author upon reasonable request.*

Supplemental Material

Supplemental material for this article is available online.

Appendix

References

Bray

FA-O

Laversanne

Sung

HA-O

, et al. Global cancer statistics 2022: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries:1542-4863. (Electronic)).

Verma

Dubey

Gupta

Ranjan

Goel

Sharma

. Neuroendocrine carcinoma of cervix and review literature:2210-2612. (Print)).

Chen

Huang

Sun

Yang

. Cervical neuroendocrine carcinoma in the third trimester: a rare case report and literature review:1471-2393. (Electronic)).

Tempfer

Tischoff

Dogan

, et al. Neuroendocrine carcinoma of the cervix: a systematic review of the literature:1471-2407. (Electronic)).

Burkeen

Chauhan

AA-O

Agrawal

, et al. Gynecologic large cell neuroendocrine carcinoma: A review; 2036-3605. (Print)).

Prodromidou

Phelps

Pergialiotis

, et al. Clinicopathological characteristics and survival outcomes of patients with large cell neuroendocrine carcinoma of the uterine cervix: A systematic review and meta-analysis:1872-7654. (Electronic)).

Siu

WYS

Cheng

C-H

Ding

D-C

. Large-Cell Neuroendocrine Carcinoma of the Cervix: Case Report and Literature Review. Diagnostics. 2025;15(6):775.

Prodromidou

Phelps

Pergialiotis

NCCN . The NCCN clinical practice guidelines in oncology—cervical cancer (version 4.2025). 2025. https://www.nccn.org/guidelines/guidelines-detail?category=1&id=1426. (accessed 10 May 2025).

10.

Mandic

Maksimovic

Nadj

Maricic

Celebic

. Neuroendocrine cervical cancer: Have we made any steps forward in its management?; (1532-2157 (Electronic)).

11.

Salvo

GA-O

Ramalingam

Flores Legarreta

, et al. Role of radical hysterectomy in patients with early-stage high-grade neuroendocrine cervical carcinoma: a NeCTuR study:1525-1538. (Electronic)).

12.

Dikmen

Kazandi

Fau - Zekioglu

, et al. Large cell neuroendocrine carcinoma of the uterine cervix: a report of a case and review of the literature:0932-0967. (Print)).

13.

Garzon

Uccella

Zorzato

, et al. Fertility-sparing management for endometrial cancer: review of the literature:1827-1869. (Electronic)).

14.

Caruso

Palaia

DID

, et al. Radical Surgery After Neoadjuvant Chemotherapy for Locally Advanced Neuroendocrine Cancer of the Cervix:1791-7530. (Electronic)).

15.

Salvo

GA-O

Gonzalez Martin

Gonzales

Frumovitz

. Updates and management algorithm for neuroendocrine tumors of the uterine cervix:1525-1538. (Electronic)).

16.

Lee

Nikas

, et al. Clinical Significance of Different Histology and High-Risk HPV Types in Neuroendocrine Carcinomas of the Cervix:1532-1979. LID - 10.1097/PAS.0000000000002391 [doi] (Electronic).

17.

Cuschieri

. The STROBE guidelines. Saudi J Anaesth. 2019;13(Suppl 1):S31-S34. doi:10.4103/sja.SJA_543_18.

18.

Fattorini

Rocca

Giusti

, et al. Mixed mucinous adenocarcinoma/large cell neuroendocrine carcinoma of the uterine cervix: case report and molecular characterization of a rare entity:1432-2307. (Electronic)).

19.

Chu

Meng

, et al. The prognosis of patients with small cell carcinoma of the cervix: a retrospective study of the SEER database and a Chinese multicentre registry:1474-5488. (Electronic)).

20.

Kim

Shim

Nam

, et al. Prognostic Factors and Impact of Minimally Invasive Surgery in Early-stage Neuroendocrine Carcinoma of the Cervix:1553-4669. (Electronic)).

21.

Zhang

Lou

HA-O

. The clinicopathological features and treatment modalities associated with survival of neuroendocrine cervical carcinoma in a Chinese population:1471-2407. (Electronic)).

22.

Vergote

Tropé Cg Fau - Amant

Amant

Fau - Kristensen

, et al. Neoadjuvant chemotherapy or primary surgery in stage IIIC or IV ovarian cancer:1533-4406. (Electronic)).

23.

Kehoe

Hook

Nankivell

, et al. Primary chemotherapy versus primary surgery for newly diagnosed advanced ovarian cancer (CHORUS): an open-label, randomised, controlled, non-inferiority trial:1474-1547X. (Electronic)).

24.

Pang

Chen

Chang

. Large-cell neuroendocrine carcinoma of the gynecologic tract: Prevalence, survival outcomes, and associated factors:2234-2943X. (Print)).

25.

Salvo

GA-O

Flores Legarreta

AA-O

Ramalingam

, et al. Clinicopathologic characteristics, oncologic outcomes, and prognostic factors in neuroendocrine cervical carcinoma: a Neuroendocrine Cervical Tumor Registry study:1525-1538. (Electronic)).

26.

CA-O

Zhang

, et al. Prognostic model for survival in patients with neuroendocrine carcinomas of the cervix: SEER database analysis and a single-center retrospective study:1932-6203. (Electronic)).

27.

Embry

Kelly Mg Fau - Post

Post

Fau - Spillman

Spillman

. Large cell neuroendocrine carcinoma of the cervix: prognostic factors and survival advantage with platinum chemotherapy:1095-6859. (Electronic)).

28.

Zhang

Huang

Luo

, et al. Therapeutic strategy analysis of patients with advanced stage high-grade neuroendocrine cervical cancer: A real-world multicenter study:1879-3479. (Electronic)).

29.

Inzani

Santoro

Angelico

, et al. Neuroendocrine Carcinoma of the Uterine Cervix: A Clinicopathologic and Immunohistochemical Study with Focus on Novel Markers (Sst2-Sst5). LID - 10.3390/cancers12051211 [doi] LID - 1211. (2072-6694 (Print)).

30.

Bellone

Jeong

Halle

MA-O

, et al. Integrated mutational landscape analysis of poorly differentiated high-grade neuroendocrine carcinoma of the uterine cervix:1091-6490. (Electronic)).

31.

Carroll

MA-O

Ramalingam

Salvo

GA-O

, et al. Evaluation of PARP and PDL-1 as potential therapeutic targets for women with high-grade neuroendocrine carcinomas of the cervix:1525-1538. (Electronic)).

32.

Liu

. Positive clinical outcomes following therapy with programmed cell death protein 1/programmed cell death ligand 1 inhibitors in neuroendocrine carcinoma of the cervix:1663-9812. (Print)).

33.

Manchana

Ittiwut

Mutirangura

Kavanagh

. Targeted therapies for rare gynaecological cancers. The Lancet Oncology. 2010;11(7):685-693.

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

0.07 MB