Prediction of distant metastasis and specific survival prediction of small intestine cancer patients with metastasis: a population-based study

doi:10.21203/rs.3.rs-2609887/v1

Background: Small intestine cancer (SIC) is a rarely found gastrointestinal malignancy, however early diagnosis of SIC is difficult as patients often present poor prognoses due to distant metastasis (DM) of the tumor by the time of diagnosis. Although nomograms for SIC have been developed, there are no relevant studies on the diagnosis and prognostic assessment of DM in patients with SIC.

Methods: The data of patients diagnosed with SIC between 2010 and 2015 was extracted from the Surveillance, Epidemiology and End Results (SEER) database. All patients were randomly assigned into the training and validation sets (7:3). Independent risk factors for DM in SIC patients were then determined by univariate and multifactor logistic regression analysis. In addition, independent prognostic factors for overall survival (OS) and cancer-specific survival (CSS) in patients with DM were analyzed by univariate and multifactor Cox regression analysis, respectively. We then constructed the corresponding three nomograms and assessed the clinical efficacy of the nomograms by receiver operating characteristic (ROC) curves, decision curve analysis (DCA), and calibration curves.

Result: The entire cohort consisted of 6773 SIC patients, of whom 1327 had DM at diagnosis. The results of multifactorial logistic regression analysis showed that T-stage, N-stage, tumor collaborative stage (CS) extension and histological type acted as independent risk factors for DM in patients with SIC. The results of multifactorial Cox regression analysis indicated that age, sex, histological type, N stage and tumor CS extension were independent predictors of OS; sex, histological type, N stage and tumor CS extension served as independent predictors of CSS. The results of ROC curves, DCA, calibration curves, and Kaplan-Meier (K-M) survival curves in the training and validation sets further confirmed the excellent accuracy of the three nomograms in predicting DM and prognosis in SIC patients.

Conclusion: This study constructed and validated nomograms for predicting DM in SIC patients as well as OS and CSS in patients with DM, which appear to function, as excellent tools to aid the physicians make more rational and personalized clinical decisions.

Small intestine cancer

distant metastases

overall survival

cancer-specific survival

nomogram

The small intestine constitutes 75% of the length of the gastrointestinal tract (approximately 5–6 meters) and its mucosal area accounts for the vast majority (> 90%) of the mucosal surface area of the gastrointestinal tract¹. However, small intestine cancer (SIC) is extremely rarely diagnosed, and only about 3% of gastrointestinal malignancies arise there². However, the incidence of SIC has increased annually, as shown in Yao et al’s³ study, from 1976 to 2016, the incidence of SIC has significantly increased by 130%, along with a relative increase in mortality of 26%, the authors suggest this could signal overdiagnosis in the United States. Owing to the fact that many SIC patients might remain asymptomatic until the detection of advanced disease and SIC is difficult to identify by imaging examination, there is a delay in diagnosis in many cases, and a late diagnosis can often lead to a poor prognosis⁴. The data from the Surveillance, Epidemiology, and End Results (SEER) in the United States from 2015 to 2019 indicated that approximately 56% of SIC patients involved locally advanced tumor growth or distant metastases (DM) when they were first diagnosed, and the 5-year relative survival rate for SIC patients with DM was only 42.6%⁵. A number of previous studies have suggested that surgical resection, tumor stage, histologic grade, and the presence of DM are determinants of survival in SIC patients^6–8. However, these variables primarily serve as single indicators and are unable to accurately predict the survival rate of SIC patients, much less the survival rate of SIC patients with DM.

At present, the various risk factors associated with the development of DM in SIC have been reported in very few literatures. Moreover, the nomogram for DM of SIC has not been developed and validated, whereas the previous nomograms for prognostic analysis of SIC have only been built for the single SIC histological subtype, so we postulated that a reliable and well-performing clinical model for SIC patients is needed.

The SEER database is the authoritative source of the cancer statistics in the United States (US). In this study, we have modeled DM by using SEER cancer registry data for patients diagnosed with SIC from 2010 to 2015. On this basis, we had also developed an overall survival (OS) prediction model and a cancer-specific survival (CSS) prognosis model for SIC patients with DM.

Patients

The dataset of SIC patients in this study was extracted from the SEER database from 2010 to 2015. The inclusion criteria used were as follows: (1) patients with SIC; (2) detailed clinical statistical variables were available, such as age, sex, and race; (3) complete clinicopathological information, including the primary tumor histological type, grade, TNM stage, tumor size, and collaborative stage (CS) extension. In addition, we excluded those patients who were diagnosed by autopsy or based on the death certificate from this study. Ultimately, 6773 patients were included in this study, of which 1327 patients had DM. Since the data from SEER is publicly available and de-identified, this study was exempt from local institutional review board review.

We formed a diagnostic cohort of all the patients to explore the potential risk factors associated with DM and build a predictive nomogram. In addition, we formed a prognostic cohort of 1327 SIC patients with DM (survival time ≥ 1 month) to investigate the various factors associated with OS and CSS in DM patients and develop new prognostic nomograms.

In each individual cohort, patients were randomly divided into a training set (70%) and a validation set (30%) in a ratio of 7:3, with patients in the training set primarily being used to construct the nomogram whereas patients in the validation set being used to validate the nomogram. (TABLE3)

Statistical analysis

In this study, statistical analysis of all the data was performed by using R software (version 4.2.0), and P values < 0.05 on both sides were considered to be statistically significant⁹. We randomly divided all SIC patients into the training and validation sets in the R software and employed the Chi-square test to compare the distribution of variables between the two groups.

In the diagnostic cohort, univariate logistic regression analysis was performed to identify DM-related risk factors. The variables with P < 0.05 in the univariate analysis were screened for further multifactorial logistic regression analysis to identify the different independent risk factors associated with DM in SIC patients. Thereafter, a new diagnostic nomogram was constructed by using the "rms" software package based on the independent risk factors obtained from the multifactorial logistic regression analysis. We then generated ROC curves¹⁰ for the nomogram and all independent variables in both the training as well as the validation sets and calculated the corresponding area under the curve (AUC)¹¹ to assess the diagnostic accuracy and sensitivity. In addition, calibration curves¹² and decision curve analysis (DCA)¹³ were used to test the discriminatory power and predictive ability of the model.

For the determination of the prognostic factors in patients with DM, univariate Cox regression analysis was applied to identify the different factors associated with OS and CSS in patients with DM. Thereafter, the significant variables with P < 0.05 were included in the multifactorial Cox regression analysis to further identify independent prognostic factors. Based on these identified independent prognostic factors, two prognostic nomograms were respectively created to predict OS and CSS in patients with DM, and individual OS and CSS risk score values were respectively calculated using the nomograms. In addition, nomograms of OS and CSS were generated at 12, 36, and 60 months and time-dependent ROC curves for all independent prognostic factors in the training and validation sets. In addition, the corresponding time-dependent AUCs were evaluated to demonstrate accuracy as well as sensitivity. The calibration curves and DCA at 12, 36, and 60 months were also plotted to evaluate the model. Based on the cutoff stratification results of OS and CSS risk score values of patients, all SIC patients with DM were next divided into high- and low-risk groups, and Kaplan-Meier (K-M) survival curves as well as log-rank tests were performed to show the potential differences in OS and CSS status between the two groups, respectively.

Baseline Characteristics of the Study Population

Between 2010 and 2015, there were 6773 SIC patients in the SEER database who met the inclusion criteria for this study (Table 3), among whom 1327 patients developed DM. The patients were randomized into the training set (n = 4741) and the validation set (n = 2032) according to a 7:3 ratio. Table 3 lists the demographic and clinicopathological characteristics of the patients included in the training and validation sets. As shown in Table 3, 65.5% of the patients were older than 60 years old. The most common tumor differentiation grade was grade I (56.3% in the training set and 57.5% in the validation set). In addition, the most common T-stage and N-stage were, respectively, T3 (38.7% in the training set and 39.5% in the validation set) and N0 (48.9% in the training set and 48.1% in the validation set). Among the histological types of SIC patients analyzed, adenoma/adenocarcinoma accounted for 87.8% in the training set and 88.8% in the validation set. Moreover, the Chi-square test demonstrated that the bias between the training and validation sets was completely random and was not statistically significant, because the P-values of the chi-square test results for each variable between the two sets in Table 3 are greater than 0.05. In this study, the grade classification was defined by ICD-O-2 (TABLE2); the histological type was defined by ICD-O-3 whereas the TNM stage was based on the 7th edition of AJCC stage criteria (TABLE1).

Incidence and Risk Factors of Distant Metastasis in Small Intestinal Tumor Patients

A total of 1327 SIC patients (19.6%) were identified with DM during the first diagnosis. As shown in Table 4, nine potential factors (age, sex, race, histological type, grade, T-stage, N-stage, tumor size, and tumor CS extension) were analyzed by applying univariate logistic analysis, resulting in the identification of the six distinct variables associated with DM, including grade, histological type, T-stage, N-stage, tumor size, and CS extension. Moreover, the findings of the multifactorial logistic analysis indicated that higher N stage, higher CS extension values, larger tumor size, and different tumor histologic types served as independent risk predictors of DM in SIC patients (TABLE4).

Diagnostic Nomogram Development and Validation

We built a novel nomogram for predicting DM risk in SIC patients (FIGURE1, A) based on four independent predictors (T-stage, N-stage, tumor CS extension, and histological type). In addition, we generated ROC curves for the training and validation sets, and the AUC of the nomogram was found to be 0.714 and 0.712 in the training and validation sets, respectively (FIGURE1, B, E). Moreover, ROC curves for all independent predictors were also constructed (FIGURE2), and the results indicated that the newly generated nomogram demonstrated better accuracy than other individual factors in both the training and validation sets. Interestingly, the calibration curves of the nomogram illustrate excellent agreement between the observed and predicted results (FIGURE1, C, F). As shown in Fig. 1, the DCA curves indicated (FIGURE1, D, G) that the diagnostic nomogram constructed in this study can be potentially used as an accurate tool for DM assessment in SIC patients.

Prognostic Factors for Small Intestinal Tumor Patients with DM

A total of 1327 SIC patients with DM were included in the dataset of the present study for analysis. 1327 eligible SIC patients with DM in the study were separately used to explore the various prognostic factors (TABLE5). As shown in Table 5, the most common age group of patients was 60–79 years (52.2%), the most common grade of tumor differentiation was grade I (49.1%), and the most common N stage was N1 (62.5%). In addition, the Chi-square test showed that the differences in all the variables between the training and validation sets were not statistically significant. Subsequently, the independent prognostic factors for OS and CSS were obtained using univariate and multifactorial Cox regression analyses successively. Finally, the forest plots were constructed to depict the independent prognostic factors and their P values for OS (FIGURE3) and CSS (FIGURE9), respectively. Multifactorial Cox regression models indicated that OS was associated with age, sex, histological type, tumor CS extension and N stage; CSS was associated with sex, histological type, tumor CS extension and N stage.

Prognostic Nomogram Development and Validation

Based on independent prognostic factors derived from the multifactorial Cox regression analysis (OS: age, sex, histological type, N-stage and tumor CS extension; CSS: sex, histological type, N-stage and tumor CS extension), we constructed nomograms to predict OS (FIGURE4, A) and CSS (FIGURE10, A) for SIC patients with DM, respectively. The generated nomograms calibration curves for OS (FIGURE4, B-D; FIGURE5, A-C) and CSS (FIGURE10, B-D; FIGURE11, A-C) at 12 months, 36 months, and 60 months respectively showed a significant higher agreement between the nomograms predictions and the actual results in the training and validation sets. In addition, the DCA curves also confirmed that the predicted nomograms of OS (FIGURE4, E-G; FIGURE5, D-F) and CSS (FIGURE10, E-G; FIGURE11, D-F) in this study possessed good performance in clinical practice. In addition, the ROC curves of the established patient OS prediction models showed that the AUCs of the nomogram reached 0.762, 0.751, and 0.754 for 12, 36, and 60 months in the training set (FIGURE6A), and 0.756, 0.762, and 0.770 in the validation set (FIGURE6B), respectively. Moreover, the AUCs of the patient CSS prediction model ROC curves indicated that the AUCs of the nomogram reached 0.743, 0.737, and 0.715 in the training set at 12, 36, and 60 months (FIGURE12, A), and 0.755, 0.740, and 0.749 in the validation set (FIGURE12, B), respectively, which demonstrated the ability of these two distinct nomograms in predicting the OS and CSS of the SIC patients with DM. The K-M curves indicated that the OS (FIGURE8) and CSS (FIGURE14) of patients in the high-risk group were significantly lower compared those in the low-risk group. In addition, AUC between the nomograms of OS and CSS were compared with each of their respective independent prognostic factors separately, and the results indicated that the discriminatory power of the nomograms of OS (FIGURE7) and CSS (FIGURE13) was significantly better than that of all independent prognostic factors at 12, 36, and 60 months.

Because population-based screening exhibits minimal benefit for asymptomatic SIC patients¹⁴ and a single diagnostic test is often insufficient to achieve a definitive diagnosis¹⁵, most SIC patients suffer from delayed diagnosis and have a very poor prognosis¹⁶, usually presenting with advanced subocclusive crisis¹⁷. A number of previous studies⁷ have shown that the presence of DM is an important prognostic influence for SIC patients, and the primary reason for this may be related to the fact that DM patients are usually not recommended for the primary tumor resection¹⁸. However, there is still a substantial gap in research related to the risk factors for the development of distant metastases in SIC patients and the prognosis of SIC patients with DM. Therefore, in the present study, we have constructed a diagnostic nomogram of the distant metastases in SIC patients and a prognostic nomogram of DM patients. By obtaining the key clinical features on the nomogram, we were able to calculate the diagnosis-related and prognosis-related scores of patients and guide the subsequent clinical intervention programs.

In this study, we have used a large sample of data with comprehensive and detailed clinical information from the SEER database and observed that the probability of DM in SIC patients was 19.6%. We have also identified four key predictors of DM in SIC patients, namely N stage, tumor size, histological type, and tumor CS extension. Among them, the CS extension score encodes the specific depth and extent of the tumor infiltration. In our study, a high score of CS extension was positively correlated with the incidence of DM, and the various previous studies have confirmed that the cell migration and invasion are important processes associated with the tumor development and metastasis which accounts for around 90% mortality in SIC patients¹⁹. Surprisingly, patients with tumor diameters above 40 mm displayed a lower risk of metastasis than those with the tumor diameters between 20 and 39 mm. We speculate that this could be attributed to the greater heterogeneity of the tumor diameters in the "≥40 mm" group of patients. In our study cohort, the tumor diameters of SIC patients ranged from 1 mm to 500 mm, and the "≥40 mm" group covered a wide range of the tumor diameters, thereby resulting in greater heterogeneity among patients. As the prognosis of SIC patients with DM is extremely poor, early detection of DM is extremely crucial for patients to receive appropriate surgical resection, chemotherapy and radiotherapy²⁰. To our knowledge, this nomogram is the first clinical model specifically designed to predict the potential risk of DM in SIC patients, thus filling an important gap in the field and demonstrating its excellent clinical utility through calibration curves, ROC curves and DCA, which might provide additional reference for the selection of clinical treatment options.

We have also identified the prognostic factors of small intestine cancer patients with DM (OS: age, sex, histological type, N stage and tumor CS extension; CSS: sex, histological type, N stage and tumor CS extension) and constructed the corresponding prognostic nomogram. In both the prognostic nomograms, the score of N1 was higher than N0, which was opposite to our general perception. The AJCC 7th edition staging was used in the present study, where N0 represents no regional lymph node metastases and N1 represents 1–3 regional lymph node metastases. Interestingly, Mohammad et al²¹ showed that for small intestine neuroendocrine tumors patients, there was no significant difference in recurrence-free survival (RFS) between patients with 0, 1, 1 to 2, 2, or 3 positive lymph nodes. However, once the threshold of 4 positive lymph nodes was reached, there was a substantial decrease noted in RFS, which might explain the difference in N-stage scores in the nomograms to some extent. It is worth noting that some prior studies have suggested that age and N stage are not regarded as prognostic factors for SIC patients^6,7, which differs from the findings of the present study. Of course, there are also few studies indicating that both age and N stage can prognostic factors for small intestine adenocarcinoma and small intestine carcinoid tumors^22,23. We speculate that the reasons for these observed differences may be the following: the small sample size included in previous studies, the unreasonable age grouping of the patients, and the fact that the subjects of this study were SIC patients with DM. However, when we explored the CSS of SIC patients with DM, we also observed that age was not an independent predictor of prognosis, which needs to be further investigated in detail. We also noted that the prognosis of female patients was better than that of male patients, and some prior reports have shown that the age-adjusted average incidence rate of small intestine cancer per 100,000 people was 1.9 for males and 1.4 for females, which might have influenced the different prognosis of patients of different genders²⁴. In addition, we have also demonstrated the excellent performance of the two prognostic nomograms by calibration curves, ROC curves, and DCA, but the predictive power of the nomograms was also confirmed to be significantly superior to any of the independent predictors.

Although some nomograms have been created in the previous studies on SIC patients, we believe that our study significantly improves and expands on previous findings. Compared with the existing nomograms for SIC patients^23,25−27, our study possesses the following advantages. First, our study population differs from those analyzed by Wang et al²⁵ who only studied patients with small intestine adenocarcinoma as well as Modlin et al²³ who chose to include patients with small intestine carcinoid tumors. However, we did not limit ourselves to a single histological subtype and selected patients with DM who lacked effective treatment and displayed a poor prognosis, which has never been studied and is more valuable for clinical guidance. Second, our study included the construction of an innovative predictive model for the development of DM in SIC patients and the prediction of OS and CSS in patients with DM. Thus, in contrast, our study is more comprehensive and opens up novel avenue for predicting DM that has not yet been studied. Third, our study has used more common and easily accessible clinical variables, which has further improved the utility and feasibility of the prediction model and provided better AUC values and DCA.

However, we must admit that our study still has some major shortcomings. First, the limited sample size we included (N = 6773) and the further shrinking number of SIC patients with DM (N = 1327) might produce some errors. Second, we have constructed the prediction model in the training set and validated it in the validation set to confirm its excellent confidence and performance, but the nomograms still lacked sufficient external data for complete validation, which could lead to the existence of internal bias. Third, the relevant information in the SEER database we included was collected only at the time of patient diagnosis, and thus our study did not include patients who developed DM at a later stage. Fourth, although we found in our study that race did not acted as an independent predictor of the development of DM in SIC patients and the prognosis of patients with DM, our study population was predominantly white, which has our prediction model not applicable to other populations. In the future, we hope to further refine our model by examining data from other populations. In addition, the predictors in this study only encompassed the common clinical variables, as several important variables such as CEA and CA-199 were not recorded in the SEER database. Finally, this is only a retrospective study, and we still need to validate the nomograms designed in this study with relevant prospective studies in the future.

This study developed a novel predictive model for DM in SIC patients and a predictive model for OS and CSS in patients with DM. The three prediction models are clinically important and can further assist the clinicians in the development of optimal treatment plans for SIC patients in the future, thus helping the patients to achieve more efficient treatment benefits.

Ethics approval and consent to participate:

Since the data from SEER is publicly available and de-identified, this study was exempt from local institutional review board review.

Consent for publication:

Not applicable

Availability of data and materials:

The datasets generated during the current study are available in the SEER repository, seer.cancer.gov

Competing interests:

The authors declare that they have no competing interests

Funding:

This study was supported by National Natural Science Foundation of China (Grant No: 81960437); Science and Technology Project of Jiangxi Province (Grant No: 20192BBH80025)

Authors' contributions:

BL, JC and SM conceived the idea; JX conducted the analyses; GL, OX and ZY provided the data; JX and ZY contributed to the writing and revisions.

Acknowledgements:

Not applicable

North JH, Pack MS. Malignant tumors of the small intestine: a review of 144 cases. Am Surg. 2000;66:46–51.
Jasti R, Carucci LR. Small Bowel Neoplasms: A Pictorial Review. Radiographics. 2020;40:1020–38. 10.1148/rg.2020200011.
Fedewa SA. Why Is Cancer of the Small Intestine Increasing? Gastroenterology. 2022;162:1593–4. 10.1053/j.gastro.2022.03.005.
Anzidei M, et al. Malignant tumours of the small intestine: a review of histopathology, multidetector CT and MRI aspects. Br J Radiol. 2011;84:677–90. 10.1259/bjr/20673379.
US National Cancer Institute. Surveillance Epidemiology and End Results (SEER) data base, https://seer.cancer.gov (
Talamonti MS, Goetz LH, Rao S, Joehl RJ. Primary cancers of the small bowel: analysis of prognostic factors and results of surgical management. Arch Surg. 2002;137:564–70. 10.1001/archsurg.137.5.564. discussion 570 – 561.
Brücher BL, et al. Prognostic factors in resected primary small bowel tumors. Dig Surg. 1998;15:42–51. 10.1159/000018585.
Taghipour Zahir S, Heidarymeybodi Z, AleSaeidi S. Prognostic Factors and Survival Time in Patients with Small Bowel Tumors: A Retrospective Observational Study. Int J Surg Oncol 2019, 2912361, doi:10.1155/2019/2912361 (2019).
Zhu Y, Fang X, Wang L, Zhang T, Yu D. A Predictive Nomogram for Early Death of Metastatic Gastric Cancer: A Retrospective Study in the SEER Database and China. J Cancer. 2020;11:5527–35. 10.7150/jca.46563.
Streiner DL, Cairney J. What's under the ROC? An introduction to receiver operating characteristics curves. Can J Psychiatry. 2007;52:121–8. 10.1177/070674370705200210.
Wu J, et al. A nomogram for predicting overall survival in patients with low-grade endometrial stromal sarcoma: A population-based analysis. Cancer Commun (Lond). 2020;40:301–12. 10.1002/cac2.12067.
Gerds TA, Andersen PK, Kattan MW. Calibration plots for risk prediction models in the presence of competing risks. Stat Med. 2014;33:3191–203. 10.1002/sim.6152.
Kerr KF, Brown MD, Zhu K, Janes H. Assessing the Clinical Impact of Risk Prediction Models With Decision Curves: Guidance for Correct Interpretation and Appropriate Use. J Clin Oncol. 2016;34:2534–40. 10.1200/jco.2015.65.5654.
Kim JS, Park SH, Hansel S, Fletcher JG. Imaging and Screening of Cancer of the Small Bowel. Radiol Clin North Am. 2017;55:1273–91. 10.1016/j.rcl.2017.06.008.
Rondonotti E, Koulaouzidis A, Georgiou J, Pennazio M. Small bowel tumours: update in diagnosis and management. Curr Opin Gastroenterol. 2018;34:159–64. 10.1097/mog.0000000000000428.
Maglinte DD, O'Connor K, Bessette J, Chernish SM, Kelvin FM. The role of the physician in the late diagnosis of primary malignant tumors of the small intestine. Am J Gastroenterol. 1991;86:304–8.
Buckley JA, Fishman EK. CT evaluation of small bowel neoplasms: spectrum of disease. Radiographics. 1998;18:379–92. 10.1148/radiographics.18.2.9536485.
Puccini A, Battaglin F, Lenz HJ. Management of Advanced Small Bowel Cancer. Curr Treat Options Oncol. 2018;19:69. 10.1007/s11864-018-0592-3.
Peng HY, et al. Knockdown of ELMO3 Suppresses Growth, Invasion and Metastasis of Colorectal Cancer. Int J Mol Sci. 2016;17. 10.3390/ijms17122119.
Zhou Z, Ge H, Li Y, Wang D, Gungor C. Survival effects of primary and metastatic surgical treatment in metastatic small intestinal tumors: A propensity score-matching study. PLoS ONE. 2022;17:e0270608. 10.1371/journal.pone.0270608.
Zaidi MY, et al. Prognostic Role of Lymph Node Positivity and Number of Lymph Nodes Needed for Accurately Staging Small-Bowel Neuroendocrine Tumors. JAMA Surg. 2019;154:134–40. 10.1001/jamasurg.2018.3865.
Zheng Z, et al. Nomograms predict survival of patients with small bowel adenocarcinoma: a SEER-based study. Int J Clin Oncol. 2021;26:387–98. 10.1007/s10147-020-01813-8.
Modlin IM, et al. A nomogram to assess small-intestinal neuroendocrine tumor ('carcinoid') survival. Neuroendocrinology. 2010;92:143–57. 10.1159/000319784.
Schottenfeld D, Beebe-Dimmer JL, Vigneau FD. The epidemiology and pathogenesis of neoplasia in the small intestine. Ann Epidemiol. 2009;19:58–69. 10.1016/j.annepidem.2008.10.004.
Wang D, et al. Specific survival nomograms based on SEER database for small intestine adenocarcinoma. Ann Palliat Med. 2021;10:7440–57. 10.21037/apm-21-600.
Wang N, et al. A convenient clinical nomogram for predicting the cancer-specific survival of individual patients with small-intestine adenocarcinoma. BMC Cancer. 2020;20:505. 10.1186/s12885-020-06971-6.
Gu Y, Deng H, Wang D, Li Y. Metastasis Pattern and Survival Analysis in Primary Small Bowel Adenocarcinoma: A SEER-Based Study. Front Surg. 2021;8:759162. 10.3389/fsurg.2021.759162.

Table 1 丨TNM classification of small intestine cancer (The Seventh Edition AJCC Cancer Staging)
Primary tumor (T)
Tx	Primary tumor cannot be assessed
T0	No evidence of primary tumor
Tis	Carcinoma in situ
T1a	Tumor invades lamina propria
T1b	Tumor invades submucosa
T2	Tumor invades muscularis propria
T3	Tumor invades through the muscularis propria into the subserosa or into the nonperitonealized perimuscular tissue (mesentery or retroperitoneum) with extension 2 cm or less*
T4	Tumor perforates the visceral peritoneum or directly invades other organs or structures (includes other loops of small intestine, mesentery, or retroperitoneum more than 2 cm, and abdominal wall by way of serosa; for duodenum only, invasion of pancreas)
Regional lymph nodes (N)
Nx	Regional lymph nodes cannot be assessed
N0	No regional lymph node metastasis
N1	1-3 regional lymph node metastases
N2	4 or more regional lymph node metastases
Distant metastasis
M0	No distant metastasis
M1	Distant metastasis
The nonperitonealized perimuscular tissue is, for jejunum and ileum, part of the mesentery and, for duodenum in areas where serosa is lacking, part of the retroperitoneum

Table 2丨 Grading and differentiation codes are defined in ICD-O-2
Grade
I	well differentiated; differentiated, NOS
II	moderately differentiated; moderately differentiated; intermediate differentiation
III	poorly differentiated; differentiated
IV	undifferentiated; anaplastic

TABLE3丨Demographic and clinicopathological characteristics of patients with small intestine cancer
	All subjects (N=6773)	Training cohort (N=4741)	Validation cohort (N=2032)	x²	P
Age				7.564	0.182
<40	264	191	73
40-49	689	497	192
50-59	1506	1069	437
60-69	2047	1391	656
70-79	1495	1058	437
≥80	772	535	237
Sex				1.067	0.302
Female	3183	2248	935
Male	3590	2493	1097
Race				2.603	0.272
White	5430	3787	1643
Black	1031	723	308
Other	312	231	81
Histology				1.659	0.646
8010-8049: epithelial neoplasms, NOS	53	39	14
8140-8389: adenomas and adenocarcinomas	5966	4162	1804
8440-8499: cystic, mucinous and serous neoplasms	239	174	65
8930-8999: complex mixed and stromal neoplasms	515	366	149
Grade				2.080	0.556
I: Well differentiated	3838	2670	1168
II: Moderately differentiated	1987	1415	572
III: Poorly differentiated	809	561	248
IV: Undifferentiated; anaplastic;	139	95	44
T				2.021	0.568
T1	1040	715	325
T2	1194	849	345
T3	2635	1833	802
T4	1904	1344	560
N				0.498	0.780
N0	3297	2320	977
N1	3109	2163	946
N2	367	258	109
M				0.086	0.770
M0	5446	3817	1629
M1	1327	924	403
Tumor size, mm				2.065	0.724
<10	1062	756	306
10-19	1784	1231	553
20-29	1394	987	407
30-39	807	564	243
≥40	1726	1203	523
Extension				0.274	0.872
<300	2143	1504	639
300-600	4143	2892	1251
>600	487	345	142

TABLE4丨Univariate and multifactorial logistic analysis of distant metastasis in patients with small intestine cancer
	Univariate analysis			Multivariate analysis
	OR	95%CI	P	OR	95%CI	P
Age
≤40	Reference
40-49			0.8432
50-59			0.3444
60-69			0.7363
70-79			0.7345
≥80			0.0921
Sex
Female	Reference
Male			0.833
Race
Black	Reference
White			0.291
Other			0.354
Grade
II	Reference
I	0.789	0.688-0.906	0.005
III	1.548	1.281-1.868	<0.001
IV	1.719	1.163-2.502	0.020
Histology
8010-8049	Reference
8140-8389	0.346	0.203-0.600	0.0012	0.383	0.219-0.680	0.005
8440-8499	0.532	0.292-0.982	0.0860	0.440	0.235-0.829	0.031
8930-8999	0.249	0.139-0.452	<0.001	0.438	0.238-0.818	0.027
T
T1	Reference
T2	2.468	1.752-3.537	<0.001
T3	5.361	3.967-7.423	<0.001
T4	10.249	7.584-14.191	<0.001
N
N0	Reference
N1	2.672	2.341-3.053	<0.001	1.985	1.710-2.308	<0.001
N2	3.356	2.619-4.280	<0.001	1.893	1.448-2.466	<0.001
Tumor size, mm
<10	Reference
10-19	6.564	4.636-9.595	<0.001	3.527	2.451-5.223	<0.001
20-29	10.851	7.675-15.845	<0.001	4.654	3.214-6.934	<0.001
30-39	12.291	8.574-18.154	<0.001	5.101	3.467-7.703	<0.001
≥40	8.451	5.985-12.326	<0.001	3.668	2.512-5.503	<0.001
Extension
<300	Reference
300-600	3.733	3.134-4.473	<0.001	2.139	1.762-2.611	<0.001
>600	9.653	7.593-12.306	<0.001	5.514	4.249-7.172	<0.001

TABLE5丨Demographic and clinicopathological characteristics of SIC patients with small intestine cancer with DM
	All subjects (N=1327)	Training cohort (N=928)	Validation cohort (N=399)	x²	P
Age				1.473	0.689
<40	53	41	12
40-59	461	322	139
60-79	693	482	211
≥80	120	83	37
Sex				1.943	0.163
Female	639	459	180
Male	688	469	219
Race				0.217	0.897
White	1103	772	331
Black	177	122	55
Other	47	34	13
Histology				4.316	0.229
8010-8049: epithelial neoplasms, NOS	22	14	8
8140-8389: adenomas and adenocarcinomas	1170	818	352
8440-8499: cystic, mucinous and serous neoplasms	55	34	21
8930-8999: complex mixed and stromal neoplasms	80	62	18
Grade				3.838	0.280
I	652	444	208
II	398	287	111
III	231	168	63
IV	46	29	17
T				2.704	0.440
T1	50	33	17
T2	116	85	31
T3	532	361	171
T4	629	449	180
N				0.003	0.998
N0	390	273	117
N1	830	580	250
N2	107	75	32
Tumor size, mm				2.282	0.684
<10	37	27	10
10-19	306	208	98
20-29	361	249	112
30-39	236	163	73
≥40	387	281	106
Extension				1.206	0.547
<300	155	114	41
300-600	969	671	298
>600	203	143	60

No competing interests reported.

Prediction of distant metastasis and specific survival prediction of small intestine cancer patients with metastasis: a population-based study

Status:

Version 1

Abstract

Figures

Introduction

Patients And Methods

Patients

Statistical analysis

Results

Baseline Characteristics of the Study Population

Incidence and Risk Factors of Distant Metastasis in Small Intestinal Tumor Patients

Diagnostic Nomogram Development and Validation

Prognostic Factors for Small Intestinal Tumor Patients with DM

Prognostic Nomogram Development and Validation

Discussion

Conclusions

Declarations

References

Tables

Additional Declarations

Status:

Version 1