External Radiation Versus Internal Radiation for Patients With Advanced Gastroenteropancreatic Neuroendocrine Neoplasms: A Retrospective Cohort Study

Abstract

Introduction

While surgical resection is a common treatment for gastroenteropancreatic neuroendocrine neoplasms (GEP-NENs), it is frequently unsuitable for advanced cases. Radiation therapy is emerging as a viable alternative. This retrospective cohort study aimed to compare the long-term outcomes of external versus internal radiation in advanced GEP-NENs.

Methods

Advanced GEP-NENs patients diagnosed from 2000 to 2020 were identified from the Surveillance, Epidemiology, and End Results (SEER) database. Propensity score matching (PSM) was applied to mitigate selection bias. The Kaplan-Meier method and multivariate Cox proportional hazards models were utilized to evaluate overall survival (OS) and cancer-specific survival (CSS).

Results

A total of 758 patients were enrolled in the study, with 597 receiving external radiation and 161 receiving internal radiation. A higher percentage of older patients, with advanced grades, lymph node metastases, and larger tumors, received external radiation. Before PSM, the OS (P < 0.001) and CSS (P < 0.001) in the external radiation group were worse than those in the internal radiation group. Multivariate Cox analysis also showed that radiation therapy was an independent risk factor affecting OS and CSS in patients with GEP-NENs. After PSM, the OS (P < 0.001) and CSS (P < 0.001) in the external radiation group were still worse than those in the internal radiation group. Multivariate Cox analysis similarly indicated that the prognosis for external radiation was worse than that for internal radiation.

Conclusions

Our study suggests a potential survival benefit associated with internal radiation compared to external radiation in advanced GEP-NENs patients.

Keywords

gastroenteropancreatic neuroendocrine neoplasms radiation therapy SEER

Introduction

Gastroenteropancreatic neuroendocrine neoplasms (GEP-NENs) constitute a highly heterogeneous group of tumors originating from stem cells, characterized by their ability to produce bioactive amines and/or polypeptide hormones.^1,2 These neoplasms primarily affect the digestive tract or pancreas and represent the most prevalent subtype of neuroendocrine neoplasms (NENs), accounting for over 70% of all NEN cases.³ According to data from the Surveillance, Epidemiology, and End Results (SEER) database, the incidence of gastrointestinal pancreatic neuroendocrine tumors (GEP-NETs) in the United States is estimated to be 3.56 cases per 100,000 individuals annually.⁴ For patients with resectable GEP-NENs, surgical intervention remains the primary treatment modality.^5,6 However, in advanced stages of the disease, surgical intervention may not significantly improve prognosis, necessitating a shift towards comprehensive treatment strategies, including biotherapy, chemotherapy, targeted therapy, and radiotherapy.^7-10 For patients with metastatic disease who are ineligible for surgical resection or those experiencing hormone-related syndromes, long-acting somatostatin analogs (SSAs) have become the cornerstone of first-line therapy.^11,12

Since the 1990s, radiolabeled somatostatin analogs have demonstrated considerable potential in treating advanced, well-differentiated neuroendocrine tumors, which typically express high levels of somatostatin receptors.^13-15 Peptide receptor radionuclide therapy (PRRT), a type of radioligand therapy, involves the administration of radiolabeled somatostatin analogs, such as (177Lu-DOTATATE), which specifically bind to somatostatin receptors (SSTRs) on tumor cells. This targeted delivery of radiation enables the destruction of tumor cells while minimizing damage to surrounding healthy tissues. PRRT has emerged as a viable second- or third-line treatment option for patients with SSTR-positive, advanced, well-differentiated GEP-NENs.¹⁶ Numerous clinical studies have corroborated the efficacy of PRRT in controlling disease progression, mitigating mortality risks, and enhancing patient quality of life.^17-19

While surgical resection remains a primary treatment for localized GEP-NENs, radiotherapy, including both external and internal approaches, is being increasingly explored as a therapeutic option for advanced or unresectable cases.^20,21 Although its role and significance are still subjects of ongoing research and debate. However, despite the growing use of these radiotherapy modalities in clinical practice, there remains a significant knowledge gap regarding their comparative efficacy in terms of patient prognosis. Specifically, while PRRT has shown promise in targeting tumor cells expressing SSTRs, external radiotherapy may offer benefits in terms of broader tumor coverage or accessibility for certain patients. Given the lack of direct comparative studies between PRRT and external radiotherapy in advanced GEP-NENs, there is an urgent need to evaluate their relative effectiveness in improving long-term outcomes for these patients. To bridge this knowledge gap, we conducted a comprehensive study utilizing data from the SEER database to evaluate the long-term prognosis of patients with advanced GEP-NENs who underwent either external or internal radiation treatment. The objective of this study is to elucidate the preferred radiotherapy regimen for these patients, ultimately aiming to enhance their treatment outcomes and prognosis.

Materials and Methods

Patient Selection

The retrospective cohort study focused on individuals diagnosed with GEP-NENs between the years 2000 and 2020, utilizing data derived from the Surveillance, Epidemiology, and End Results (SEER) database. Patients were identified using the International Classification of Diseases for Oncology, Third Edition (ICD-O-3) with specific codes, including 8013/3 for large cell neuroendocrine carcinoma, 8150/3 for malignant pancreatic endocrine tumor, 8151/3 for malignant insulinoma, 8152/3 for malignant glucagonoma, 8154/3 for malignant mixed pancreatic endocrine and exocrine tumor, 8155/3 for malignant vipoma, 8156/3 for malignant somatostatinoma, 8240/3 for carcinoid tumor, NOS, 8243/3 for goblet cell carcinoid, 8244/3 for mixed adenoneuroendocrine carcinoma (ICD-O-3 update), 8245/3 for adenocarcinoid tumor, 8246/3 for neuroendocrine carcinoma, NOS, 8249/3 for atypical carcinoid tumor, and 8574/3 for adenocarcinoma with neuroendocrine differentiation. Our study included only patients with advanced stages, specifically those classified as TNM stage III and IV according to the 7th edition of the AJCC Stage. However, we excluded cases based on the following criteria: (1) lack of a definitive pathological diagnosis; (2) missing survival data; (3) absence of radiotherapy, unspecified radiotherapy methods, or multiple combined radiotherapies; (4) classification as TNM stage I, II, or unknown staging. A visual overview of the research methodology is presented in Figure 1. The investigation covered various variables, such as age, race, gender, tumor site, tumor grade, marital status, TNM stage, tumor size, radiation modality, chemotherapy status, overall survival (OS), cancer-specific survival (CSS), and duration of follow-up. According to the SEER Program’s Coding and Staging Guidelines, radiotherapy was categorized as either external or internal radiation. External radiation primarily involved beam radiation therapy. Internal radiation included any form of radioisotope-based therapy, such as PRRT, brachytherapy, or other radioactive implants—coded as “radioisotope' and “radioactive implants' in the SEER database. Critically, the SEER database does not distinguish between these distinct internal radiation modalities, which differ fundamentally in mechanism (systemic receptor-mediated versus local), patient selection (SSTR-positive well-differentiated tumors versus others), and therapeutic intent. Thus, our “internal radiation' category is a pragmatic but clinically heterogeneous grouping. Participants were divided into two age groups: those under 60 years (younger cohort) and those 60 years and older (older cohort). Race was categorized as white, black, or other races, including American Indian/Alaska Native, and Asian/Pacific Islander ethnicities. Marital status was dichotomized as married or unmarried, with the unmarried category encompassing divorced, separated, single, or widowed statuses. The primary objective of the study was to assess OS and CSS in patients with GEP-NENs. Both OS and CSS were calculated from the date of GEP-NENs diagnosis until death, cancer-related mortality, or the end of follow-up, whichever occurred first. Since the SEER database provides de-identified, publicly accessible data without any involvement of human subjects and with all patient information anonymized, neither Institutional Review Board (IRB) approval nor informed consent was needed. This study was conducted in accordance with the Helsinki Declaration of 1975, as revised in 2024. The reporting of this study conforms to the strengthening the reporting of observational studies in epidemiology (STROBE) guidelines.²² According to the predefined histological and temporal criteria, patients were identified from the SEER database, forming a population-based cohort as opposed to a consecutive, random, or selectively recruited sample.

Figure 1.

Flow chart of the study

Statistical Analysis

Categorical data were displayed as frequency counts and proportions, and discrepancies between groups were examined using chi-square tests. For continuous variables that deviated from normality, the median and the interquartile range (IQR) were reported, and comparisons were made using the Mann-Whitney U test. A 1:1 propensity score matching (PSM) method was applied to balance the external and internal radiation groups, considering factors like age, race, gender, tumor site, tumor grade, marital status, TNM stage, T stage, N stage, tumor size, and chemotherapy status, with a caliper set at 0.01. Survival analysis for OS and CSS was conducted using the Kaplan-Meier method, and survival curves were compared via the log-rank test. Both univariate and multivariate analyses were performed using Cox proportional hazards models to produce hazard ratios (HRs) with 95% confidence intervals (CIs). Variables with a P-value below 0.05 in the univariate analysis were included in the multivariate Cox regression. To enhance the statistical robustness of the study, particularly in addressing missing data in the SEER database, a combined strategy utilizing multiple imputation (MI) and the random forest algorithm was employed. All statistical analyses were carried out using R software (version 4.1.0), with statistical significance defined as a P-value less than 0.05.

Results

Patient Characteristics

A total of 758 patients were enrolled in the study and split into two cohorts: 597 assigned to the external radiation cohort and 161 to the internal radiation cohort. Prior to implementing PSM, substantial discrepancies were noticeable in multiple dimensions, including age, tumor site, tumor grade, marital status, TNM stage, N stage, tumor size, and chemotherapy administration, all of which were statistically significant at P < 0.05. It is noteworthy that a higher percentage of older patients, those with advanced grades, those exhibiting lymph node metastasis, and those with larger tumors underwent external radiation. Nevertheless, following a 1:1 PSM, both cohorts were equilibrated to 161 participants each, and no meaningful distinctions were detected in the baseline characteristics (P > 0.05), as delineated in Table 1. Notably, survival outcomes (e.g., median survival months) were not used for matching and therefore remained significantly different between groups after PSM, reflecting the primary study findings. This table furnishes a detailed breakdown of the patients' demographic and clinical attributes. The study also confronted challenges with incomplete data, particularly in variables such as race (0.3% unreported), grade (27.3% unreported), marital status (3.0% unreported), T stage (0.4% unreported), N stage (4.9% unreported), and tumor size (22.8% unreported). To aid comprehension, Supplementary Table 1 presents a rigorous pre-MI comparison of these variables across the two cohorts.

Table 1.

Demographic and Clinical Characteristics of Patients Before and After PSM

Variables	Before PSM				After PSM
Variables	Total (n=758)	External radiation (n=597)	Internal radiation (n=161)	P-value	External radiation (n=161)	Internal radiation (n=161)	P-value
Age, years, n (%)				0.006			0.910
≤60	355 (46.8%)	264 (44.2%)	91 (56.5%)		92 (57.1%)	91 (56.5%)
>60	403 (53.2%)	333 (55.8%)	70 (43.5%)		69 (42.9%)	70 (43.5%)
Race, n (%)				0.608			0.914
White	597 (78.8%)	466 (78.1%)	131 (81.4%)		128 (79.5%)	131 (81.4%)
Black	91 (12.0%)	73 (12.2%)	18 (11.2%)		20 (12.4%)	18 (11.2%)
Others	70 (9.2%)	58 (9.7%)	12 (7.5%)		13 (8.1%)	12 (7.5%)
Sex, n (%)				0.225			0.737
Male	446 (58.8%)	358 (60.0%)	88 (54.7%)		85 (52.8%)	88 (54.7%)
Female	312 (41.2%)	239 (40.0%)	73 (45.3%)		76 (47.2%)	73 (45.3%)
Location, n (%)				<0.001			0.738
Gastrointestine	488 (64.4%)	406 (68.0%)	82 (50.9%)		79 (49.1%)	82 (50.9%)
Pancreas	270 (35.6%)	191 (32.0%)	79 (49.1%)		82 (50.9%)	79 (49.1%)
Grade, n (%)				<0.001			>0.999
Well	205 (27.0%)	122 (20.4%)	83 (51.6%)		83 (51.6%)	83 (51.6%)
Moderately	155 (20.4%)	97 (16.2%)	58 (36.0%)		58 (36.0%)	58 (36.0%)
Poorly	282 (37.2%)	267 (44.7%)	15 (9.3%)		16 (9.9%)	15 (9.3%)
Undifferentiated	116 (15.3%)	111 (18.6%)	5 (3.1%)		4 (2.5%)	5 (3.1%)
Marital status, n (%)				<0.001			0.185
Married	488 (64.4%)	359 (60.1%)	129 (80.1%)		119 (73.9%)	129 (80.1%)
Unmarried	270 (35.6%)	238 (39.9%)	32 (19.9%)		42 (26.1%)	32 (19.9%)
TNM stage, n (%)				<0.001			>0.999
III	171 (22.6%)	168 (28.1%)	3 (1.9%)		2 (1.2%)	3 (1.9%)
IV	587 (77.4%)	429 (71.9%)	158 (98.1%)		159 (98.8%)	158 (98.1%)
T stage, n (%)				0.081			0.765
T0	6 (0.8%)	3 (0.5%)	3 (1.9%)		3 (1.9%)	3 (1.9%)
T1	31 (4.1%)	24 (4.0%)	7 (4.3%)		6 (3.7%)	7 (4.3%)
T2	144 (19.0%)	103 (17.3%)	41 (25.5%)		31 (19.3%)	41 (25.5%)
T3	225 (29.7%)	186 (31.2%)	39 (24.2%)		45 (28.0%)	39 (24.2%)
T4	167 (22.0%)	133 (22.3%)	34 (21.1%)		32 (19.9%)	34 (21.1%)
TX	185 (24.4%)	148 (24.8%)	37 (23.0%)		44 (27.3%)	37 (23.0%)
N stage, n (%)				<0.001			0.654
N0	222 (29.3%)	153 (25.6%)	69 (42.9%)		61 (37.9%)	69 (42.9%)
N1	404 (53.3%)	335 (56.1%)	69 (42.9%)		74 (46.0%)	69 (42.9%)
NX	132 (17.4%)	109 (18.3%)	23 (14.3%)		26 (16.1%)	23 (14.3%)
Chemotherapy, n (%)				<0.001			0.574
No/Unknown	264 (34.8%)	170 (28.5%)	94 (58.4%)		89 (55.3%)	94 (58.4%)
Yes	494 (65.2%)	427 (71.5%)	67 (41.6%)		72 (44.7%)	67 (41.6%)
Tumor size, cm, n (%)				<0.001			0.467
≤2.0	123 (16.2%)	84 (14.1%)	39 (24.2%)		33 (20.5%)	39 (24.2%)
2.1-5.0	344 (45.4%)	250 (41.9%)	94 (58.4%)		92 (57.1%)	94 (58.4%)
>5.0	291 (38.4%)	263 (44.1%)	28 (17.4%)		36 (22.4%)	28 (17.4%)
Survival months, median (IQR)	15.0 (7.0, 36.0)	13.0 (6.0, 28.0)	32.0 (14.0, 64.0)	<0.001	17.0 (7.0, 41.0)	32.0 (14.0, 64.0)	<0.001

PSM: propensity score matching; Others: American Indian, Alaska Native, Asian/Pacific Islander; IQR: interquartile range; bold values indicate P < 0.05.

Comparison Between the External Radiation Group and the Internal Radiation Group on OS and CSS

Over a follow-up period with a median duration of 13.0 months (IQR: 6.0-28.0 months), 432 fatalities were documented in the group receiving external radiation, with 397 of those deaths linked to GEP-NENs. Conversely, in the internal radiation group, there were 73 deaths, 66 of which were attributed to GEP-NENs. Prior to PSM, the external radiation group demonstrated markedly inferior OS rates (HR 2.46, 95% CI 1.91-3.15, P < 0.001) (Figure 2A) and CSS rates (HR 2.47, 95% CI 1.90-3.21, P < 0.001) (Figure 2B) when compared to the internal radiation group. Meanwhile, following PSM, the OS (HR 1.97, 95% CI 1.46-2.65, P < 0.001) (Figure 3A) and CSS (HR 1.97, 95% CI 1.44-2.79, P < 0.001) (Figure 3B) for the external radiation group continued to be lower than those for the internal radiation group.

Figure 2.

Overall survival (OS) and cancer-specific survival (CSS) were compared between the external radiation group and the internal radiation group before propensity score matching A. OS; B CSS

Figure 3.

Overall survival (OS) and cancer-specific survival (CSS) were compared between external radiation group and internal radiation group after propensity score matching A. OS; B CSS

Univariate and Multivariate Cox Regression

Before PSM, an initial univariate Cox regression analysis uncovered a range of factors—encompassing radiation modality, patient age, tumor grade, marital status, TNM stage, chemotherapy regimen, and tumor size—that independently influenced OS in individuals diagnosed with GEP-NENs. This discovery was further corroborated through a multivariate Cox regression analysis, which singled out radiation modality, age, tumor grade, marital status, TNM staging, and chemotherapy status as the primary independent predictors of OS within this patient cohort (Table 2). In parallel, a univariate Cox regression analysis also revealed these same factors to have an independent impact on CSS in GEP-NENs patients. A subsequent multivariate Cox regression analysis reinforced the notion that radiation modality, tumor grade, marital status, TNM stage, chemotherapy status, and tumor size were crucial independent predictors of CSS in this patient population (Table 3).

Table 2.

Univariate and Multivariate Cox Regression for Analyzing the Overall Survival for Patients With GEP-NENs Before PSM

	Univariate			Multivariate
	HR	95%CI	P-value	HR	95%CI	P-value
Radiation therapy
Internal radiation	1	Reference		1	Reference
External radiation	2.46	1.91, 3.15	<0.001	2.17	1.65, 2.86	<0.001
Age, years
≤60	1	Reference		1	Reference
>60	1.44	1.20, 1.72	<0.001	1.26	1.05, 1.51	0.013
Race
White	1	Reference		-	-	-
Black	0.83	0.63, 1.09	0.176	-	-	-
Others	0.82	0.60, 1.12	0.211	-	-	-
Sex
Male	1	Reference		-	-	-
Female	0.86	0.72, 1.03	0.099	-	-	-
Location
Gastrointestine	1	Reference		-	-	-
Pancreas	0.95	0.79, 1.14	0.570	-	-	-
Grade
Well	1	Reference		1	Reference
Moderately	1.66	1.24, 2.22	<0.001	2.04	1.52, 2.74	<0.001
Poorly	3.41	2.65, 4.37	<0.001	4.56	3.39, 6.15	<0.001
Undifferentiated	4.47	3.35, 5.95	<0.001	5.47	3.90, 7.68	<0.001
Marital status
Married	1	Reference		1	Reference
Unmarried	1.42	1.19, 1.70	<0.001	1.27	1.06, 1.53	0.010
TNM stage
III	1	Reference		1	Reference
IV	1.79	1.43, 2.24	<0.001	3.30	2.59, 4.21	<0.001
T stage
T0	1	Reference		-	-	-
T1	1.28	0.38, 4.36	0.689	-	-	-
T2	1.26	0.40, 3.97	0.698	-	-	-
T3	1.12	0.36, 3.51	0.849	-	-	-
T4	1.44	0.46, 4.53	0.535	-	-	-
TX	1.70	0.54, 5.33	0.364	-	-	-
N stage
N0	1	Reference		-	-	-
N1	0.83	0.68, 1.02	0.073	-	-	-
NX	1.10	0.85, 1.43	0.458	-	-	-
Chemotherapy
No/Unknown	1	Reference		1	Reference
Yes	1.32	1.09, 1.60	0.005	0.63	0.50, 0.79	<0.001
Tumor size, cm
≤2.0	1	Reference		1	Reference
2.1-5.0	1.33	1.01, 1.77	0.046	0.95	0.71, 1.26	0.707
>5.0	2.63	1.99, 3.49	<0.001	1.29	0.95, 1.74	0.098

GEP-NENs: gastroenteropancreatic neuroendocrine neoplasms; PSM: propensity score matching; Others: American Indian, Alaska Native, Asian/Pacific Islander; HR: hazard ratios; bold values indicate P < 0.05.

Table 3.

Univariate and Multivariate Cox Regression for Analyzing the Cancer-specific Survival for Patients With GEP-NENs Before PSM

	Univariate			Multivariate
	HR	95%CI	P-value	HR	95%CI	P-value
Radiation therapy
Internal radiation	1	Reference		1	Reference
External radiation	2.47	1.90, 3.21	<0.001	2.11	1.58, 2.81	<0.001
Age, years
≤60	1	Reference		1	Reference
>60	1.35	1.12, 1.62	0.002	1.17	0.97, 1.42	0.099
Race
White	1	Reference		-	-	-
Black	0.84	0.63, 1.12	0.225	-	-	-
Others	0.75	0.53, 1.05	0.093	-	-	-
Sex
Male	1	Reference		-	-	-
Female	0.90	0.74, 1.08	0.248	-	-	-
Location
Gastrointestine	1	Reference		-	-	-
Pancreas	1.01	0.83, 1.22	0.942	-	-	-
Grade
Well	1	Reference		1	Reference
Moderately	1.81	1.33, 2.48	<0.001	2.23	1.62, 3.05	<0.001
Poorly	3.80	2.90, 4.98	<0.001	5.13	3.73, 7.05	<0.001
Undifferentiated	5.07	3.74, 6.87	<0.001	6.28	4.39, 9.00	<0.001
Marital status
Married	1	Reference		1	Reference
Unmarried	1.45	1.20, 1.74	<0.001	1.30	1.08, 1.58	0.007
TNM stage
III	1	Reference		1	Reference
IV	1.85	1.45, 2.34	<0.001	3.38	2.62, 4.37	<0.001
T stage
T0	1	Reference		-	-	-
T1	1.12	0.33, 3.86	0.853	-	-	-
T2	1.13	0.36, 3.58	0.833	-	-	-
T3	1.04	0.33, 3.27	0.943	-	-	-
T4	1.32	0.42, 4.16	0.634	-	-	-
TX	1.48	0.47, 4.65	0.504	-	-	-
N stage
N0	1	Reference		-	-	-
N1	0.86	0.70, 1.05	0.139	-	-	-
NX	1.02	0.78, 1.35	0.870	-	-	-
Chemotherapy
No/Unknown	1	Reference		1	Reference
Yes	1.38	1.12, 1.68	0.002	0.63	0.50, 0.80	<0.001
Tumor size, cm
≤2.0	1	Reference		1	Reference
2.1-5.0	1.47	1.08, 2.00	0.013	1.02	0.75, 1.40	0.882
>5.0	2.98	2.20, 4.04	<0.001	1.40	1.01, 1.93	0.044

After PSM, a fresh univariate Cox regression analysis shed light on a refined set of factors—namely, radiation modality, age, tumor grade, and tumor size—that exerted an independent effect on OS in GEP-NENs patients. This was further validated by a multivariate Cox regression analysis, which pinpointed radiation modality, age, and tumor grade as the key independent predictors of OS in this specific patient group (Supplementary Table 2). Similarly, a univariate Cox regression analysis following PSM indicated that radiation modality and tumor size had an independent influence on CSS in GEP-NENs patients. A subsequent multivariate Cox regression analysis underscored the significance of radiation modality and tumor grade as the paramount independent predictors of CSS in this particular patient cohort (Supplementary Table 3). To assess the robustness of our findings, we performed a complete-case sensitivity analysis by excluding patients with missing data. The results of the sensitivity analysis (Supplementary Table 4 for OS and Supplementary Table 5 for CSS) were consistent with those from the primary multiple imputation analysis, showing that external radiation remained significantly associated with worse OS and CSS compared with internal radiation (both P < 0.001).

Subgroup Analysis

In our initial analysis, we conducted subgroup evaluations to explore the potential differences in OS and CSS between the internal and external radiation cohorts across various patient characteristics. Before PSM, our findings suggested that the internal radiation cohort generally had more favorable OS and CSS compared to the external radiation cohort across most subgroups (P < 0.05). However, we observed notable exceptions in subgroups characterized by black race, moderate, poor, or undifferentiated tumor grade, T0 or T1 stage, and tumor sizes ≤ 2.0 cm, where both radiation groups displayed equivalent OS and CSS (P > 0.05) (Supplementary Figures 1 and 2). Following PSM, we reiterated the subgroup analyses to ensure the robustness of our findings. The results largely supported the initial findings, showing that the internal radiation cohort had advantageous OS and CSS relative to the external radiation cohort in the majority of subgroups (P < 0.05). Nevertheless, we acknowledge the importance of interpreting these subgroup analyses with caution, particularly in subgroups defined by black or other races, undifferentiated grade, unmarried status, T0, T1, or TX stage, and NX stage, where the two radiation groups exhibited comparable OS and CSS (P > 0.05) (Figures 4 and 5).

Figure 4.

Subgroup analysis of OS between external radiation group and internal radiation group after PSM

Figure 5.

Subgroup analysis of CSS between external radiation group and internal radiation group after PSM

Discussion

In this research, we utilized the SEER database to explore the optimal radiation therapy for advanced GEP-NENs by comparing external radiation to internal radiation. Our findings suggested a more favorable prognosis associated with internal radiation compared to external radiation. Before PSM analysis, patients in the external radiation group exhibited more advanced disease features, such as older age, higher grades, lymph node metastases, and larger tumors. Even after PSM, the external radiation group still had significantly worse OS and CSS compared to the internal radiation group. Multivariate Cox analysis similarly indicated that the prognosis for external radiation was worse than that for internal radiation. These findings suggest that internal radiation may be associated with better long-term outcomes in patients with advanced GEP-NENs, although causality cannot be inferred from this observational study.

Surgical intervention is the primary approach for treating resectable GEP-NENs,^5,6 but it may not be feasible for certain patients due to coexisting conditions or advanced disease stage. In such cases, systemic therapy and non-surgical local ablative treatments, such as radiofrequency ablation, internal radiation, or external radiation, can be considered.²³ Both internal and external radiation have been used for the treatment of patients with advanced GEP-NENs, with PRRT with 177Lu-DOTATATE being a frequently employed internal radiation technique, and intensity-modulated radiotherapy (IMRT) and three-dimensional conformal radiotherapy (3D-CRT) being commonly utilized external radiation techniques.^24-26

A large, multicenter cohort study revealed positive outcomes with PRRT for patients with advanced G3 GEP-NENs, including high response rates, effective disease control, prolonged progression-free survival (PFS), enhanced OS, and manageable toxicity.²⁷ Similarly, the research conducted by Ezziddin et al. 2²⁸ demonstrated beneficial responses and long-lasting results with PRRT for individuals with G1 and G2 GEP-NENs. Furthermore, another investigation examined the use of PRRT in advanced GEP-NEN patients with substantial liver tumor loads and found it to be safe and effective, with minimal risk of hepatotoxicity.²⁹ Several review analyses have further emphasized the efficacy of PRRT in managing advanced GEP-NENs.^17,30,31 Our study also demonstrates that the 5-year OS and CSS rates for advanced GEP-NENs patients who received internal radiation therapy can exceed 50%.

Lee et al. 2²⁴ conducted a retrospective analysis of the efficacy of external radiation therapy in nine patients with pancreatic neuroendocrine tumors (PNETs). The findings revealed that external radiation is efficient in halting the progression of advanced PNETs. In another study, the impact of external radiation therapy was assessed in 36 PNET patients, with results indicating a 39% overall response rate (divided into 13% complete response, 26% partial response, 56% stable disease, and 4% progressive disease). It was observed that external radiation significantly reduces patients' symptoms and manages local tumor advancement.³² Strosberg et al. 3³³ investigated the combination of chemotherapy and external radiation in six advanced PNETs patients and found that all patients saw a decrease in tumor size after chemoradiotherapy, with no reported cases of local or metastatic disease progression. A systematic review examined the role of external radiation in patients with GEP-NENs, concluding that external radiation is generally well-tolerated and exhibits promising results in appropriately selected PNETs patients.²¹

Although numerous previous studies have analyzed the efficacy of internal radiation and external radiation for patients with GEP-NENs, no research has yet compared and analyzed the long-term prognosis of these two radiation methods specifically for advanced GEP-NENs. Our study is the first to utilize data from the SEER database to compare the long-term prognosis of advanced GEP-NENs treated with internal radiation versus external radiation. We found that internal radiation was associated with better prognosis than external radiation. Several factors may contribute to this observed association: 1) Internal radiation enables precise dose distribution within the tumor, directly targeting it with high-dose radiation while minimizing damage to surrounding normal tissues. This highly targeted treatment may helps more effectively kill tumor cells and reduce the risk of recurrence. Additionally, continuous low-dose-rate radiation may exert a stronger killing effect on certain tumor cells by interfering with their DNA repair mechanisms, thereby potentially enhancing the overall treatment effect. 2) Internal radiation generally exhibits lower toxicity to other organs and systems compared to external radiation. This reduces systemic adverse reactions caused by the treatment, improves the patient’s quality of life, and may indirectly improve long-term prognosis. However, our study also found that factors such as age and tumor grade, apart from the radiation method, influence the prognosis of patients with advanced GEP-NENs. Therefore, prospective randomized controlled trial (RCT) studies are still needed to compare the impact of these two radiation methods on the prognosis of advanced GEP-NENs.

In the management of advanced GEP-NENs, a diverse array of treatment options beyond surgery and radiotherapy is available, emphasizing the need for a tailored therapeutic strategy based on tumor biology and patient factors. Chemotherapy remains a relevant consideration, particularly for poorly differentiated NENs or in cases with aggressive disease progression. As highlighted by Espinosa-Olarte et al. 3³⁴ chemotherapy continues to play a role, especially in high-grade neoplasms where platinum-based regimens are often employed to achieve tumor control and symptom relief, even amid the rise of newer therapies. Concurrently, the advent of precision medicine has revolutionized the approach to GEP-NENs, incorporating molecular profiling and targeted agents to address specific tumor characteristics. Fazio et al. 3³⁵ underscore that precision-oriented treatments, such as inhibitors targeting mTOR pathways or angiogenesis, as well as immune checkpoint inhibitors, are increasingly pivotal for optimizing outcomes in molecularly defined subgroups, thereby enhancing therapeutic efficacy while mitigating off-target effects. Collectively, these modalities—ranging from cytotoxic drugs to personalized biologic therapies—underscore the heterogeneity of NENs and reinforce the importance of multidisciplinary decision-making to leverage the most appropriate systemic options for individualized patient care.

A critical caveat must be acknowledged when interpreting the observed survival advantage of internal radiation over external radiation. In routine clinical practice, internal radiotherapy, particularly PRRT, is preferentially offered to patients with well-differentiated, SSTR-positive, and generally indolent tumors. In contrast, external beam radiotherapy is more often applied in poorly differentiated, locally aggressive, or symptomatic diseases. The SEER database does not provide information on SSTR status, Ki-67 proliferation index, performance status, or specific histological grading details that would allow us to disentangle treatment effects from underlying tumor biology. Thus, despite the use of propensity score matching to adjust for observable covariates, the possibility of unmeasured confounding—particularly by tumor aggressiveness and differentiation—cannot be excluded. Accordingly, the survival benefit associated with internal radiation observed in this study should be interpreted as an association rather than a causal treatment effect.

Our research encounters several limitations. Firstly, and most fundamentally, the SEER database aggregates mechanistically and clinically distinct internal radiation modalities—including PRRT, brachytherapy, and other radioisotope implants—into a single category. PRRT is a systemic, somatostatin receptor-targeted therapy typically reserved for well-differentiated, SSTR-positive, indolent tumors, whereas brachytherapy is a local modality applied in a different clinical context (e.g., locally advanced but non-metastatic disease). Grouping these together obscures profound differences in mechanism, patient selection, and therapeutic intent. No statistical adjustment, including propensity score matching, can fully compensate for this inherent limitation. Therefore, our comparison is not between PRRT and external radiation, but between two administratively defined, heterogeneous categories. The observed survival advantage associated with the “internal radiation' category should not be interpreted as evidence that PRRT or any specific internal radiation technique is superior to external radiation; rather, it may largely reflect baseline differences in tumor biology and patient selection that are not captured in the SEER database. Secondly, and relatedly, the SEER database lacks critical biological and clinical variables—including somatostatin receptor status, Ki-67 proliferation index, performance status, and precise tumor grade differentiation—that are known to guide radiotherapy selection in advanced GEP-NENs. Consequently, even if internal radiation modalities could be disaggregated, residual confounding by tumor biology and patient selection would remain highly likely. Thirdly, the study spans 2000–2020, and modern PRRT only became widely used after 2017–2018, introducing potential temporal bias. We agree that adjusting for year of diagnosis would be ideal; however, only 62 internal radiation patients were diagnosed in or after 2017, and the matched sample size after PSM was too small for meaningful adjustment or stratified analysis. Therefore, we were unable to include year of diagnosis in the model. The observed survival benefit should be interpreted with caution given this potential era effect. Fourthly, due to inherent limitations of the SEER database, we cannot distinguish between specific external beam modalities (e.g., IMRT, 3D-CRT, SBRT) or specific radioisotopes used in internal radiation (e.g., ¹⁷⁷Lu-DOTATATE). Critical data on radiation dosing, treatment duration, and lines of therapy are also unavailable. These omissions restrict our ability to perform granular comparative analyses. Lastly, the multiple unadjusted subgroup analyses conducted in our study raise concerns about false positives. No formal sample size calculation was performed, and the study may be underpowered for certain subgroup analyses. The limited number of patients in some subgroups should be considered when interpreting the results. Nevertheless, to our knowledge, this research marks the initial comprehensive study comparing the long-term effects of internal versus external radiotherapy in patients with advanced GEP-NENs, and future prospective studies are needed to confirm these associations and to provide a more comprehensive assessment of treatment efficacy.

Conclusion

Our research indicates that internal radiation is associated with better long-term survival than external radiation in patients with advanced GEP-NENs. However, due to the absence of key biological data (e.g., SSTR status, Ki-67 index) in the SEER database, this association may be substantially influenced by unmeasured differences in tumor biology and patient selection. Therefore, internal radiation should not be concluded as a superior therapeutic approach based solely on these findings. Prospective studies incorporating pathological and functional imaging data are needed to determine whether a true treatment effect exists.

Supplemental Material

Supplemental Material - External Radiation Versus Internal Radiation for Patients With Advanced Gastroenteropancreatic Neuroendocrine Neoplasms: A Retrospective Cohort Study

Supplemental Material for External Radiation Versus Internal Radiation for Patients With Advanced Gastroenteropancreatic Neuroendocrine Neoplasms: A Retrospective Cohort Study by Jingxian Shi, Mingyu Yao, Ping Yin, Jie Lu, Jun Yang, Jing Zhao, Zhenguo Qiao, Chengjie Lu in Cancer Control

Supplemental Material

Supplemental Material - External Radiation Versus Internal Radiation for Patients With Advanced Gastroenteropancreatic Neuroendocrine Neoplasms: A Retrospective Cohort Study

Footnotes

ORCID iD

Zhenguo Qiao

Ethical Considerations

This study used de-identified data from the publicly accessible SEER database. Since the research did not involve direct contact with human participants or the use of identifiable personal information, neither ethical approval nor informed consent was necessary. The study was carried out in compliance with the Declaration of Helsinki and its later amendments.

Author Contributions

Conception and design: Jingxian Shi, Mingyu Yao.

Administrative support: Chengjie Lu.

Provision of study materials or patients: Zhenguo Qiao, Jing Zhao.

Collection and assembly of data: Ping Yin.

Data analysis and interpretation: Jun Yang, Jie Lu.

Manuscript writing: All authors.

Final approval of manuscript: All authors.

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 the Science and Technology Development Program of Suzhou (SYWD2024322 and SYWD2024077), the Program for the Talents in Science and Education of Wujiang District, Suzhou, China (WWK202117, WWK202510 and WWK202502), the Scientific Research Project of Suzhou Ninth People’s Hospital (YK202438 and YK202307). There was no additional external funding received for this study.

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Data Availability Statement

Publicly available datasets were analyzed in this study. These data can be found here: https://seer.cancer.gov/. The datasets supporting the conclusions of this article are included within the article.*

Supplemental Material

Supplemental material for this article is available online.

Appendix

References

Pavel

Öberg

Falconi

, et al. Gastroenteropancreatic neuroendocrine neoplasms: ESMO Clinical Practice Guidelines for diagnosis, treatment and follow-up. Ann Oncol. 2020;31(7):844-860. doi:10.1016/j.annonc.2020.03.304. Epub 2020 Apr 6. PMID: 32272208.

Ito

Masui

Komoto

, et al. JNETS clinical practice guidelines for gastroenteropancreatic neuroendocrine neoplasms: diagnosis, treatment, and follow-up: a synopsis. J Gastroenterol. 2021;56(11):1033-1044. doi:10.1007/s00535-021-01827-7. Epub 2021 Sep 29. PMID: 34586495; PMCID: PMC8531106.

Oronsky

Morgensztern

, et al. Nothing But NET: A Review of Neuroendocrine Tumors and Carcinomas. Neoplasia. 2017;19(12):991-1002. doi:10.1016/j.neo.2017.09.002. Epub 2017 Nov 5. PMID: 29091800; PMCID: PMC5678742.

Dasari

Shen

Halperin

, et al. Trends in the Incidence, Prevalence, and Survival Outcomes in Patients With Neuroendocrine Tumors in the United States. JAMA Oncol. 2017;3(10):1335-1342. doi:10.1001/jamaoncol.2017.0589. PMID: 28448665; PMCID: PMC5824320.

Cives

Strosberg

. Gastroenteropancreatic Neuroendocrine Tumors. CA Cancer J Clin. 2018;68(6):471-487. doi:10.3322/caac.21493. Epub 2018 Oct 8. PMID: 30295930.

Wang

Zhang

Fan

Lin

. The impact of surgery and survival prediction in patients with gastroenteropancreatic neuroendocrine tumors: a population-based cohort study. Int J Surg. 2023;109(6):1629-1638. doi:10.1097/JS9.0000000000000336. PMID: 37133986; PMCID: PMC10389215.

Que

Zhang

Bao

Ling

. Role of surgical treatments in high-grade or advanced gastroenteropancreatic neuroendocrine neoplasms. World J Gastrointest Surg. 2022;14(5):397-408. doi:10.4240/wjgs.v14.i5.397. PMID: 35734618; PMCID: PMC9160682.

Singh

Halperin

Myrehaug

, et al. all the NETTER-2 Trial Investigators. [177Lu]Lu-DOTA-TATE plus long-acting octreotide versus high-dose long-acting octreotide for the treatment of newly diagnosed, advanced grade 2-3, well-differentiated, gastroenteropancreatic neuroendocrine tumours (NETTER-2): an open-label, randomised, phase 3 study. Lancet. 2024;403(10446):2807-2817. doi:10.1016/S0140-6736(24)00701-3. Epub 2024 Jun 5. PMID: 38851203.

Sorbye

Hjortland

Vestermark

, et al. NETest in advanced high-grade gastroenteropancreatic neuroendocrine neoplasms. J Neuroendocrinol. 2024;27:e13428. doi:10.1111/jne.13428. Epub ahead of print. PMID: 38937137.

10.

Abdel-Rahman

Fouad

. Bevacizumab-based combination therapy for advanced gastroenteropancreatic neuroendocrine neoplasms (GEP-NENs): a systematic review of the literature. J Cancer Res Clin Oncol. 2015;141(2):295-305. doi:10.1007/s00432-014-1757-5. Epub 2014 Jul 3. PMID: 24990591.

11.

Rinke

Müller

Schade-Brittinger

, et al. Placebo-controlled, double-blind, prospective, randomized study on the effect of octreotide LAR in the control of tumor growth in patients with metastatic neuroendocrine midgut tumors: a report from the PROMID Study Group. J Clin Oncol. 2009;27(28):4656-4663. doi:10.1200/JCO.2009.22.8510. Epub 2009 Aug 24. PMID: 19704057.

12.

Caplin

Pavel

Ćwikła

, et al. Lanreotide in metastatic enteropancreatic neuroendocrine tumors. N Engl J Med. 2014;371(3):224-233. doi:10.1056/NEJMoa1316158. PMID: 25014687.

13.

Krenning

Kooij

Bakker

, et al. Radiotherapy with a radiolabeled somatostatin analogue, [111In-DTPA-D-Phe1]-octreotide. A case history. Ann N Y Acad Sci. 1994;733:496-506. doi:10.1111/j.1749-6632.1994.tb17300.x. PMID: 7978900.

14.

Krenning

de Jong

Kooij

, et al. Radiolabelled somatostatin analogue(s) for peptide receptor scintigraphy and radionuclide therapy. Ann Oncol. 1999;10(Suppl 2):S23-S29. doi:10.1093/annonc/10.suppl_2.s23. PMID: 10399029.

15.

Hörsch

Ezziddin

Haug

, et al. Effectiveness and side-effects of peptide receptor radionuclide therapy for neuroendocrine neoplasms in Germany: A multi-institutional registry study with prospective follow-up. Eur J Cancer. 2016;58:41-51. doi:10.1016/j.ejca.2016.01.009. Epub 2016 Mar 2. PMID: 26943056.

16.

Becx

Minczeles

Brabander

de Herder

Nonnekens

Hofland

. A Clinical Guide to Peptide Receptor Radionuclide Therapy with 177Lu-DOTATATE in Neuroendocrine Tumor Patients. Cancers (Basel). 2022;14(23):5792. doi:10.3390/cancers14235792. PMID: 36497273; PMCID: PMC9737149.

17.

Das

Al-Toubah

El-Haddad

, et al. 177Lu-DOTATATE for the treatment of gastroenteropancreatic neuroendocrine tumors. Expert Rev Gastroenterol Hepatol. 2019;13(11):1023-1031. doi:10.1080/17474124.2019.1685381. Epub 2019 Oct 30. PMID: 31652074; PMCID: PMC7227421.

18.

Hofland

Brabander

Verburg

Feelders

de Herder

. Peptide Receptor Radionuclide Therapy. J Clin Endocrinol Metab. 2022;107(12):3199-3208. doi:10.1210/clinem/dgac574. PMID: 36198028; PMCID: PMC9693835.

19.

Kashyap

Raja

Adusumilli

, et al. Role of neoadjuvant peptide receptor radionuclide therapy in unresectable and metastatic gastro-entero-pancreatic neuroendocrine neoplasms: A scoping review. J Neuroendocrinol. 2024;27:e13425. doi:10.1111/jne.13425. Epub ahead of print. PMID: 38937270.

20.

Giraudet

. Combinaison de la radiothérapie interne vectorisée et de la radiothérapie externe [Combination of internal and external beam radiotherapy]. Cancer Radiother. 2023;27(8):754-758. French. doi: 10.1016/j.canrad.2023.08.005. Epub 2023 Nov 10. PMID: 37953187.

21.

Chan

Thompson

Lam

, et al. External Beam Radiotherapy in the Treatment of Gastroenteropancreatic Neuroendocrine Tumours: A Systematic Review. Clin Oncol (R Coll Radiol). 2018;30(7):400-408. doi:10.1016/j.clon.2018.03.006. Epub 2018 Mar 31. PMID: 29615284.

22.

von

Altman

Egger

, et al. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. J Clin Epidemiol. 2008;61(4):344-349. doi:10.1016/j.jclinepi.2007.11.008. PMID: 18313558.

23.

Yao

Hassan

Phan

, et al. One hundred years after “carcinoid”: epidemiology of and prognostic factors for neuroendocrine tumors in 35,825 cases in the United States. J Clin Oncol. 2008;26(18):3063-3072. doi:10.1200/JCO.2007.15.4377. PMID: 18565894.

24.

Lee

Choi

, et al. Role of radiotherapy for pancreatobiliary neuroendocrine tumors. Radiat Oncol J. 2013;31(3):125-130. doi:10.3857/roj.2013.31.3.125. Epub 2013 Sep 30. PMID: 24137557; PMCID: PMC3797271.

25.

Saif

Ove

Russo

. Radiotherapy in the management of pancreatic neuroendocrine tumors (PNET): experience at three institutions. Anticancer Res. 2013;33(5):2175-2177. PMID: 23645772.

26.

Arvold

Willett

Fernandez-del Castillo

, et al. Pancreatic neuroendocrine tumors with involved surgical margins: prognostic factors and the role of adjuvant radiotherapy. Int J Radiat Oncol Biol Phys. 2012;83(3):e337-e343. doi:10.1016/j.ijrobp.2011.12.068. Epub 2012 Mar 11. PMID: 22414286.

27.

Carlsen

Fazio

Granberg

, et al. Peptide receptor radionuclide therapy in gastroenteropancreatic NEN G3: a multicenter cohort study. Endocr Relat Cancer. 2019;26(2):227-239. doi:10.1530/ERC-18-0424. PMID: 30540557.

28.

Ezziddin

Attassi

Yong-Hing

, et al. Predictors of long-term outcome in patients with well-differentiated gastroenteropancreatic neuroendocrine tumors after peptide receptor radionuclide therapy with 177Lu-octreotate. J Nucl Med. 2014;55(2):183-190. doi:10.2967/jnumed.113.125336. Epub 2014 Jan 16. PMID: 24434296.

29.

Gococo-Benore

Kuhlman

Parent

, et al. Evaluation of Hepatotoxicity from Peptide Receptor Radionuclide Therapy in Patients with Gastroenteropancreatic Neuroendocrine Tumors and a Very High Liver Tumor Burden. J Nucl Med. 2023;64(6):880-884. doi:10.2967/jnumed.122.264533. Epub 2023 Apr 20. PMID: 37080760.

30.

Kolasińska-Ćwikła

Łowczak

Maciejkiewicz

Ćwikła

. Peptide Receptor Radionuclide Therapy for Advanced Gastroenteropancreatic Neuroendocrine Tumors - from oncology perspective. Nucl Med Rev Cent East Eur. 2018;21(2):115-124. doi:10.5603/NMR.2018.0019. PMID: 29741203.

31.

Alsadik

Yusuf

Al-Nahhas

. Peptide Receptor Radionuclide Therapy for Pancreatic Neuroendocrine Tumours. Curr Radiopharm. 2019;12(2):126-134. doi:10.2174/1874471012666190201164132.PMID: 30714538.

32.

Contessa

Griffith

Wolff

, et al. Radiotherapy for pancreatic neuroendocrine tumors. Int J Radiat Oncol Biol Phys. 2009;75(4):1196-1200. doi:10.1016/j.ijrobp.2008.12.044. Epub 2009 Mar 26. PMID: 19327905.

33.

Strosberg

Hoffe

Gardner

Choi

Kvols

. Effective treatment of locally advanced endocrine tumors of the pancreas with chemoradiotherapy. Neuroendocrinology. 2007;85(4):216-220. doi:10.1159/000102969. Epub 2007 May 16. PMID: 17541257.

34.

Espinosa-Olarte

La Salvia

Riesco-Martinez

, et al.

Chemotherapy in NEN: still has a role?

Rev Endocr Metab Disord. 2021;22(3):595-614. doi: 10.1007/s11154-021-09638-0.

35.

Fazio

La Salvia

. Precision medicine in gastroenteropancreatic neuroendocrine neoplasms: Where are we in 2023? Best Pract Res Clin Endocrinol Metab. 2023;37(5):101794. doi:10.1016/j.beem.2023.101794. Epub 2023 Jun 28. PMID: 37414651.

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.

0.95 MB

0.00 MB

1.11 MB