DOI: https://doi.org/10.21203/rs.3.rs-2670258/v1
Background：Rectal squamous cell carcinoma（rSCC）is a rare pathological type of rectal malignant tumors. There is no consensus on the treatment paradigm of patients with rSCC. This study aims to provide a paradigm for clinical treatment via analyzing the efficacy of different treatment regimens for patients with different TNM stages.
Methods：Patients diagnosed with rSCC between 2010 and 2019 were extracted from the Surveillance, Epidemiology, and End Results (SEER) database. According to the TNM staging system, Kaplan-Meier（K-M）survival analysis was used to identify the survival benefits of patients with rSCC in different treatment groups. The Cox regression method was used to identify independent prognostic risk factors. Nomograms were evaluated by Harrell's concordance index, calibration curves, DCA and K-M curve.
Results：463 patients with rSCC were extracted from the SEER database. Survival analysis showed that there was no significant difference in cancer specific survival (CSS) among radiotherapy（RT）, chemoradiotherapy（CRT）and surgeryin Stage 1 (P = 0.249). In TNM Stage 2, there was significant difference in CSS among surgery, RT, and CRT (P = 0.003). In TNM Stage 3, there was significant difference in CSS between CRT, no treatment and CRT plus surgery (P < 0.001). In TNM Stage 4, there was significant difference in CSS between CRT, no treatment and CT (P = 0.041). COX regression analysis showed that Age, Marital status, N, M, CEA, PIN, Size, RT, CT, and surgery were the independent risk factors. The 1-, 3-, and 5-year C-index was 0.869,0.777,0.759, respectively. The calibration curve showed that the model had excellent calibration. The DCA curve showed that the model had excellent clinical application value.
Conclusion：RT or surgery is recommended for patients with rSCC at Stage 1, and CRT is recommended for patients with rSCCat Stage 2, Stage 3, and Stage 4. Age, marital status, N, M, PIN, size, RT, CT and surgery are independent risk factors for CSS in patients with rSCC. The prediction model composed of the above independent risk factors has excellent prediction efficiency.
Colorectal cancer (CRC) is a common malignant tumor in the world. The incidence and mortality of patients with CRC rank the third and second, respectively, and the incidence of rectal cancer (RC) accounts for 29% of the incidence of CRC.(1, 2) The main pathological type of RC was adenocarcinoma (AC). In addition, a little proportion of the pathological type is squamous cell carcinoma (SCC), accounting for about 0.1–0.3% of all RC.(3) Unlike rectal AC (rAC), there are significant differences in etiology, epidemiology, pathogenesis, and treatment.(4) The prognosis of rectal SCC (rSCC) is worse than that of rAC.(5, 6) The molecular expression of rSCC is similar to that of anal SCC (aSCC) and is quite different from that of rAC.(7) The pathogenesis of rSCC is not well understood and may be related to smoking, previous radiation exposure, chronic rectal inflammation, intestinal infection due to squamous metaplasia, HIV infection, HPV infection, or malignant transformation of persistent ectopic embryonic nests of ectodermal cells.(3)
Since the first discovery of rSCC in 1933, the number of reported rSCC has been so rare that some researchers have questioned whether this malignancy really exists. The diagnosis of rSCC must meet William's four criteria.(8) First, there is no continuity between the tumor and the anal squamous epithelium or the gynecological tract. Second, Absence of squamous lined fistula in the context of inflammatory bowel disease. Third, Absence of a SCC in another primary site. Fourth, finally histological confirmation.
Up to now, there is no clinical guideline (NCCN, ESMO, ASCO or ASCO) or expert consensus on the treatment regimens of rSCC. Current regimens options have been derived based on rAC and aSCC. Most treatments are like those used for aSCC. In the past, surgery was the standard treatment.(9) In recent years, chemoradiotherapy (CRT) is the preferred treatment.(10) Most studies believe that radiotherapy (RT) can significantly improve the overall survival (OS), local recurrence and distant metastasis of patients with rSCC.(11, 12) In most treatment options, adjuvant chemotherapy (CT) with 5-FU plus carboplatin or mitomycin plus radiotherapy has achieved satisfactory results.(13–18)。
The study of rSCC is limited by its rarity. The Surveillance, Epidemiology, and End Results (SEER) database covers approximately 28% of the United States population, which provides an adequate sample for the study of rare diseases. Evaluating the survival benefits of different treatment options for patients with rSCC through the SEER database to provide a paradigm for clinical treatment. At the same time, the independent risk factors of cancer specific survival (CSS) in patients with rSCC were analyzed to accurately predict the survival prognosis.
Patient data were obtained from SEER∗Stat software (version: 8.4.0.1). Patients with RC diagnosed by pathology from 2010 to 2019 were included. Individual data included age, sex, race, marital status, carcinoembryonic antigen (CEA), AJCC T stage, AJCC N stage, AJCC M stage, differentiation, perineural invasion (PNI), tumor size, surgery, CT, RT, survival status, and survival time. Inclusion criteria: (1) age at diagnosis ≥ 18 years old; (2) patients with primary RC; (3) rSCC identified using the International Classification of Oncology, third Revision, histological coding (8070–8077). Exclusion criteria: (1) age < 18 years or survival time < 1 month; (2) patients with more than one primary cancer; (3) patients with missing or incomplete survival data; (4) Lack of TNM and degree of differentiation. The flow chart is shown in Fig. 1.
The variables included in this study were age, gender, race, marital status, CEA, T, N, M, differentiation, PNI, tumor size, surgery, CT, RT, survival status, and survival time. The age at diagnosis was grouped as༜70 years and ≥ 70 years. Ethnicity included White, Black, and Other (American Indian/AK Native, Asian/Pacific Islander). Marital status included three types: Unknown, Have partner and NO partner. Unmarried or Domestic Partner and Married (including common law) were classified as "Have partner". Divorced, Widowed, Single (never married), Separated is "NO partner." Tumor sizes included༜73 mm, ≥ 73 mm, and Unknow. Age and tumor size were re-segmented using X-tile software, as shown in Fig. 2. The outcome variable was cancer-specific survival (CSS). CSS was defined as the time alive from diagnosis to death from cancer causes.
All statistical analyses were performed by R software (version 4.2.1). In all statistical analyses, p < 0.05 was considered statistically significant. First, according to the TNM staging system, all patients were divided into four subtypes: TNM Stage 1, TNM Stage 2, TNM Stage 3, and TNM Stage 4. The treatment methods were divided into seven subtypes: RT, CT, CRT, surgery, surgery plus RCT, surgery plus RT, surgery plus CT, and no-treatment (NO). In different TNM stage subgroups, Kaplan-Meier method was used to analyze the efficacy of different treatment modalities. Second, univariate Cox regression analysis was used to identify CSS-related risk factors in patients with rSCC. Variables that were statistically significant in the univariate Cox analysis were subsequently included in the multivariate Cox regression analysis to identify independent prognostic factors for CSS. Third, a prognostic model for CSS was developed based on independent prognostic factors. The model was visualized as a nomogram. Fourth, receiver operating characteristic (ROC) curves of CSS at 1-, 3-, and 5-years were plotted, and the corresponding area under the curve (AUC) values were used to evaluate the discrimination of the model. The corresponding calibration curve was drawn to show the calibration degree of the model. Decision curve analysis (DCA) was performed to show the clinical benefit of the model. Furthermore, Subgroup analysis was performed in age༜70 and ≥ 70, have partner and no partner, M0 and M1, PIN negative and positive, tumor size༜73mm and ≥ 73mm, radiation yes and no, chemotherapy yes and no, surgery yes and no, NO, N1 and N2. The Kaplan-Meier survival curves for each subgroup were generated.
According to the inclusion and exclusion criteria, 463 patients in the SEER database were finally included in this study. 331 female patients and 132 male patients were included. The median CSS was 59 months for women and 54 months for men. The female mortality rate was 26.6 percent and the male mortality rate 36.4 percent. Among females, 6 percent were Well differentiated, 44.4 percent Moderately differentiated, 47.1 percent Poorly differentiated, and 2.4 percent Undifferentiated. Among males, 9.1 percent were Well differentiated, 40.9 percent Moderately differentiated, 46.2 percent Poorly differentiated, and 3.8 percent Undifferentiated. Among females, 33.8 percent were T1, 19.3 percent were T2, 31.7 percent were T3, and 15.1 percent were T4. Among males, 35.6 percent were T1, 15.9 percent were T2, 34.8 percent were T3, and 13.6 percent were T4. Among females, 66.2 percent were N0, 29.9 percent were N1, and 3.9 percent were N2. In males, 57.6% were N0, 33.3% were N1 and 9.1% were N2 (P = 0.047). Among females, M0 accounted for 93.1% and M1 accounted for 6.9%. Among males, M0 and M1 accounted for 88.6% and 11.4%, respectively (P = 0.118). Other information on clinicopathological features is shown in Table 1.
The treatment options in this study for patients with rSCC at different TNM stages are shown in Table 2. In TNM Stage 1, there was no significant difference in CSS among surgery, surgery plus RT, surgery plus CT, surgery plus CRT (Figs. 3A, P = 0.631), and was significant difference among CT, RT, CRT and NO (Figs. 3B, P = 0.002), and was no significant difference among surgery, RT, and CRT (Figs. 3C, P = 0.249).In TNM Stage 2, there was no significant difference among surgery plus CRT (Figs. 3D, P = 0.099), and was significant difference among RT, CRT and NO (Figs. 3E, P < 0.001), and was significant difference among surgery, RT, and CRT (Figs. 3F, P = 0.003). In TNM Stage 3, there was no significant difference among CT, RT and NO (Figs. 3G, P = 0.508), and was significant difference among CRT, surgery plus CRT and NO (Figs. 3H, P < 0.001). In TNM Stage 4, there was significant difference between CT, CRT and NO (Figs. 3I, P = 0.041).
To identify independent risk factors related to CSS, covariates such as age, sex, race, marital status, CEA, T, N, M, differentiation, PNI, tumor size, surgery, CT, and RT were included in univariate Cox analysis. Then the variables with statistically significant differences in univariate COX regression analysis were included in multivariate COX regression analysis. The results of multivariate COX regression analysis showed that age, marital status, N, M, PIN, size, RT, CT, and surgery were independent prognostic factors for CSS. The results of univariate and multivariate COX regression analysis are shown in Table 3.
As shown in Fig. 4, to predict CSS in patients with rSCC, all independent prognostic factors were developed a prognostic model, which is visualized by a nomogram. As shown in Fig. 5A-5C, the ROC curve was drawn, and the results showed that the AUC values of CSS at 1-, 3-, and 5-years were 0.869、0.777、0.759, respectively, which indicating that the model had good discrimination. As shown in Fig. 5D-5F, the 1-, 3-, and 5-year calibration curves for CSS in patients with rSCC indicate a strong calibration of the model. As shown in Figs. 5G-5I, DCA at 1-, 3-, and 5-years for CSS of patients with rSCC shows that the model has high clinical benefit.
According to the prognostic model established in this study, rSCC patients can be divided into two groups: low risk and high risk. As shown in Fig. 6A-6F, Fig. 6H, Fig. 6J and Fig. 6L-6S, the results of Kaplan-Meier survival curves suggested that there were significant differences in survival patterns among the subgroups. In the four subgroups of M1, PIN Positive, Size ≥ 73mm and N2, P was greater than 0.05, which may be related to the small sample size of the subgroup. With the increase of risk score, the prognosis of patients is worse. The above results suggest that the model can classify rectal SCC patients into two groups with significantly different prognosis.
rSCC was first discovered in 1933.(19) rSCC is a rare pathological type. The occurrence of rSCC may be related to several reasons. The pluripotent stem cells of mucosal endodermal origin have the ability to differentiate in multiple directions, which may lead to the development of squamous epithelium and malignant transformation.(20) Damage to the mucosa may result in the proliferation of basal cells into squamous cells.(21) Human papillomavirus (HPV) can induce rSCC by disrupting local cell proliferation.(22) rSCC is different from AC in terms of treatment and prognosis.(3, 23) rSCC tend to occur in the elderly, women.(24)
In clinical guidelines, there is a lack of consensus on the treatment of patients with rSCC. The principles of treatment for different stages are discussed in this study. For patients with TNM Stage 1, the difference between surgery and RT was not statistically significant. Adding CT or RT based on surgery has no survival benefit. Adding CT to RT does not benefit patients. Compared with monotherapy, combination therapy is more likely to reduce tolerance and increase adverse reactions in patients. In TNM Stage 2, the curative effect of CRT is superior to surgery. It is important to note that surgery combined CRT did not increase survival benefits. In TNM Stage 3, CT or RT alone did not increase survival benefits compared with no treatment. The curative effect of CRT is superior to that of surgery combined with CRT. In TNM Stage 4, The efficacy of CRT is superior to CT alone. Schizas D and D. C. Steinemann et al. suggest that surgery is the standard treatment regimens.(19, 25) In our study, surgery was significant, but only for stage 1 patients. In stage 2 and 3 patients, the survival benefit from surgery was not as good as that from CRT. Additional surgery on top of chemoradiotherapy did not increase survival benefits, consistent with previous studies.(26) The survival benefit of surgery in stage 4 patients was not analyzed due to the limited sample size. Considering the poor quality of life of surgery patients, surgery is not recommended as a standard treatment. We recommend RT-based multimodality treatment as paradigm for patients with rSCC, which is consistent with the view of previous studies.(17, 18, 27, 28) surgery can be used as salvage treatment after the failure of CRT.(16) But salvage surgery does not seem to increase survival benefits.(26) Immunotherapy has achieved remarkable efficacy in rAC and aSCC patients with MSI-H/MSS.(29–31) Neoadjuvant immunotherapy has achieved delightful results in rSCC. (32)Patients may experience prolonged remission. And can achieve surgical exemption, significantly improve the quality of life of patients. In the future, multi-center phase 3 clinical trials are worth looking forward to.
In this study, univariate and multivariate COX regression analysis were performed, and results showed that age, marital status, N, M, PIN, size, RT, CT, and surgery were independent prognostic factors for CSS. This is similar to the independent risk factor for OS.(33) The 1 -, 3 -, and 5-year AUC values of the model were 0.869, 0.777, 0.759, respectively, which indicated that the model had excellent discrimination. The calibration curve showed that the model had successful calibration. DCA showed that the model had outstanding clinical utility. The model constructed by the independent prognostic factors can successfully predict the 1-, 3- and 5-year CSS of patients with rSCC. In our study, age ≥ 70 years was an independent risk factor, which is consistent with previous studies.(34, 35) This may be because patients are more likely to not receive the full recommended treatment as they age.(36) Wang et al suggested that RC patients with partners have better survival prognosis(37).This is consistent with our study. Distant metastasis is an important factor affecting patients' CSS. In our study, the median CSS of stage 4 patients was only 16 months, far less than that of non-stage 4 patients.This study found that like aSCC, tumor size was an independent prognostic factor in patients with rSCC. Our results agree with the study by P. Goffredo et al., which may be explained by the fact that the larger the tumor, the more advanced the patient stage.(28) However, there is no consensus on the cut-off value of tumor size.
The study has several strengths. Firstly, the data were obtained from the SEER database and are therefore highly reproducible. Secondly, the treatment options of patients with different TNM stages were explored for the first time using a large sample size, which provided great guiding value for clinical treatment. Thirdly, the established prediction model has strong predictive performance and can accurately predict the survival mode of patients. Predictors are common clinical variables, which makes the model more broadly applicable.
There are also certain limitations in that study. Firstly, this study was based on retrospective information from the SEER database, which may have led to an inherent selection bias. Secondly, the clinicopathological variables included in the study were limited, such as the inability to analyze the effect of specific treatment regimens on patient cancer-specific survival, and there was also a lack of information on immunotherapy. At the time of prognostic analysis, common prognostic factors such as gene expression, microsatellite status, vascular invasion, and tumor deposition are lacking, which makes the prognostic model less comprehensive. Thirdly, this study only included information from online databases and lacked prospective data to verify the findings. It should be noted that due to the rarity of rectal squamous cell carcinoma, later analyses are more likely to be based on retrospective public databases. At the same time, the joint research of multiple countries and medical centers is very important for rSCC.
RT or surgery is recommended for patients with rSCC at Stage 1, and CRT is recommended for patients with rSCC at Stage 2, Stage 3, and Stage 4. Age, marital status, N, M, PIN, size, RT, CT, and surgery are independent risk factors for CSS in patients with rSCC. The prediction model composed of the above independent risk factors has excellent prediction efficiency.
Rectal squamous cell carcinoma
Surveillance, Epidemiology, and End Results
cancer specific survival
overall survival
receiver operating characteristic
decision curve analysis
Kaplan-Meier
radiotherapy
chemotherapy
Chemoradiotherapy:NO:no-treatment
Colorectal cancer
rectal cancer
adenocarcinoma
carcinoembryonic antigen
perineural invasion
area under the curve.
Ethics Approval and Consent to Participate
The SEER database used in this study is a public database, so there is no need for an ethical approval.
Consent for Publication
Not applicable
Availability of Data and Materials
Publicly available datasets were analyzed in this study. This data can be found here: Surveillance, Epidemiology, and End Results (SEER) database (https://seer.cancer.gov/).
Competing Interests
The authors declare that they have no competing interests.
Funding
There is no funding for this research.
Authors' Contributions
LR designed the study, conducted statistical analysis of the data, and explained the results, and wrote the manuscript. ZJH made graphs and tables. LR revised and improved the manuscript. The final manuscript was approved by all authors.
Acknowledgements
Not applicable.
Authors' Information
Rui Liu: [email protected]
Jiahui Zhang: [email protected]
Table 1
|
Characteristics (n) |
Female (331) |
Male (132) |
P value |
|
Time (months), median (IQR) |
59 (22.5, 84) |
54 (20, 84.25) |
0.572 |
|
Event, n (%) |
|
|
0.037 |
|
Live |
243 (73.4%) |
84 (63.6%) |
|
|
Dead |
88 (26.6%) |
48 (36.4%) |
|
|
Age, n (%) |
|
|
0.775 |
|
＜70 |
267 (80.7%) |
108 (81.8%) |
|
|
≥70 |
64 (19.3%) |
24 (18.2%) |
|
|
Race, n (%) |
|
|
0.523 |
|
White |
299 (90.3%) |
115 (87.1%) |
|
|
Black |
28 (8.5%) |
14 (10.6%) |
|
|
Other |
4 (1.2%) |
3 (2.3%) |
|
|
Marital status, n (%) |
|
|
0.666 |
|
Have partner |
138 (41.7%) |
61 (46.2%) |
|
|
Unknown |
23 (6.9%) |
9 (6.8%) |
|
|
NO partner |
170 (51.4%) |
62 (47%) |
|
|
Grade, n (%) |
|
|
0.535 |
|
Well differentiated |
20 (6%) |
12 (9.1%) |
|
|
Moderately differentiated |
147 (44.4%) |
54 (40.9%) |
|
|
Poorly differentiated |
156 (47.1%) |
61 (46.2%) |
|
|
Undifferentiated |
8 (2.4%) |
5 (3.8%) |
|
|
T, n (%) |
|
|
0.774 |
|
T1 |
112 (33.8%) |
47 (35.6%) |
|
|
T2 |
64 (19.3%) |
21 (15.9%) |
|
|
T3 |
105 (31.7%) |
46 (34.8%) |
|
|
T4 |
50 (15.1%) |
18 (13.6%) |
|
|
N, n (%) |
|
|
0.047 |
|
N0 |
219 (66.2%) |
76 (57.6%) |
|
|
N1 |
99 (29.9%) |
44 (33.3%) |
|
|
N2 |
13 (3.9%) |
12 (9.1%) |
|
|
M, n (%) |
|
|
0.118 |
|
M0 |
308 (93.1%) |
117 (88.6%) |
|
|
M1 |
23 (6.9%) |
15 (11.4%) |
|
|
CEA, n (%) |
|
|
0.493 |
|
Normal |
79 (23.9%) |
37 (28%) |
|
|
Elevated |
37 (11.2%) |
11 (8.3%) |
|
|
Unknown |
215 (65%) |
84 (63.6%) |
|
|
PIN, n (%) |
|
|
0.793 |
|
Negative |
134 (40.5%) |
58 (43.9%) |
|
|
Positive |
8 (2.4%) |
3 (2.3%) |
|
|
Unknown |
189 (57.1%) |
71 (53.8%) |
|
|
Size, n (%) |
|
|
0.315 |
|
＜73 |
220 (66.5%) |
81 (61.4%) |
|
|
≥73 |
24 (7.3%) |
15 (11.4%) |
|
|
Unknow |
87 (26.3%) |
36 (27.3%) |
|
|
Radiation, n (%) |
|
|
0.806 |
|
Performed |
274 (82.8%) |
108 (81.8%) |
|
|
None/Unknown |
57 (17.2%) |
24 (18.2%) |
|
|
Chemotherapy, n (%) |
|
|
0.720 |
|
Performed |
273 (82.5%) |
107 (81.1%) |
|
|
No/Unknown |
58 (17.5%) |
25 (18.9%) |
|
|
Surgery, n (%) |
|
|
0.188 |
|
Performed |
90 (27.2%) |
44 (33.3%) |
|
|
No |
241 (72.8%) |
88 (66.7%) |
|
Table 2
|
Characteristics(n) |
Stage 1(214) |
Stage 2(91) |
Stage 3(129) |
Stage 4(29) |
|
Event, n (%) |
|
|
|
|
|
Dead |
43 (20.1%) |
27 (29.7%) |
46 (35.7%) |
20 (69%) |
|
Live |
171 (79.9%) |
64 (70.3%) |
83 (64.3%) |
9 (31%) |
|
Time (months), median (IQR) |
63 (28.5, 86) |
58 (25.5, 90) |
56 (21, 80) |
16 (6, 61) |
|
Paradigm, n (%) |
|
|
|
|
|
surgery |
26 (12.1%) |
4 (4.4%) |
0 (0%) |
1 (3.4%) |
|
chemotherapy |
4 (1.9%) |
0 (0%) |
5 (3.9%) |
4 (13.8%) |
|
radiotherapy |
7 (3.3%) |
3 (3.3%) |
4 (3.1%) |
2 (6.9%) |
|
chemoradiotherapy |
101 (47.2%) |
64 (70.3%) |
88 (68.2%) |
15 (51.7%) |
|
surgery and chemotherapy |
4 (1.9%) |
0 (0%) |
1 (0.8%) |
0 (0%) |
|
surgery and radiotherapy |
3 (1.4%) |
1 (1.1%) |
0 (0%) |
0 (0%) |
|
surgery and chemoradiotherapy |
53 (24.8%) |
14 (15.4%) |
25 (19.4%) |
2 (6.9%) |
|
No |
16 (7.5%) |
5 (5.5%) |
6 (4.7%) |
5 (17.2%) |
Table 3
|
Characteristics |
Total(N) |
Univariate analysis |
Multivariate analysis |
||
|
Hazard ratio (95% CI) |
P value |
Hazard ratio (95% CI) |
P value |
||
|
Age |
463 |
|
< 0.001 |
|
|
|
≥70 |
88 |
Reference |
|
Reference |
|
|
＜70 |
375 |
0.441 (0.306 - 0.635) |
< 0.001 |
0.621 (0.420 - 0.918) |
0.017 |
|
Sex |
463 |
|
0.064 |
|
|
|
Male |
132 |
Reference |
|
|
|
|
Female |
331 |
0.713 (0.502 - 1.014) |
0.059 |
|
|
|
Race |
463 |
|
0.127 |
|
|
|
White |
414 |
Reference |
|
|
|
|
Black |
42 |
1.644 (1.001 - 2.700) |
0.050 |
|
|
|
Other |
7 |
0.501 (0.070 - 3.590) |
0.492 |
|
|
|
Marital status |
463 |
|
0.001 |
|
|
|
NO partner |
232 |
Reference |
|
Reference |
|
|
Have partner |
199 |
0.518 (0.358 - 0.749) |
< 0.001 |
0.710 (0.480 - 1.048) |
0.085 |
|
Unknown |
32 |
0.627 (0.304 - 1.294) |
0.207 |
0.435 (0.198 - 0.958) |
0.039 |
|
Grade |
463 |
|
0.657 |
|
|
|
Well |
32 |
0.915 (0.457 - 1.832) |
0.802 |
|
|
|
Moderately |
201 |
0.817 (0.571 - 1.169) |
0.269 |
|
|
|
Poorly |
217 |
Reference |
|
|
|
|
Undifferentiated |
13 |
1.223 (0.493 - 3.031) |
0.664 |
|
|
|
T |
463 |
|
< 0.001 |
|
|
|
T1 |
159 |
0.972 (0.551 - 1.715) |
0.922 |
0.892 (0.483 - 1.647) |
0.715 |
|
T2 |
85 |
Reference |
|
Reference |
|
|
T3 |
151 |
1.352 (0.790 - 2.316) |
0.272 |
0.897 (0.508 - 1.585) |
0.709 |
|
T4 |
68 |
3.768 (2.185 - 6.500) |
< 0.001 |
1.727 (0.932 - 3.201) |
0.083 |
|
N |
463 |
|
< 0.001 |
|
|
|
N0 |
295 |
Reference |
|
Reference |
|
|
N1 |
143 |
1.983 (1.403 - 2.804) |
< 0.001 |
1.653 (1.097 - 2.493) |
0.016 |
|
N2 |
25 |
1.146 (0.527 - 2.494) |
0.731 |
0.928 (0.401 - 2.146) |
0.861 |
|
M |
463 |
|
< 0.001 |
|
|
|
M0 |
425 |
Reference |
|
Reference |
|
|
M1 |
38 |
4.422 (2.877 - 6.797) |
< 0.001 |
2.161 (1.316 - 3.548) |
0.002 |
|
CEA |
463 |
|
< 0.001 |
|
|
|
Elevated |
48 |
Reference |
|
Reference |
|
|
Normal |
116 |
0.504 (0.310 - 0.820) |
0.006 |
0.683 (0.403 - 1.158) |
0.157 |
|
Unknown |
299 |
0.335 (0.216 - 0.520) |
< 0.001 |
0.628 (0.376 - 1.048) |
0.075 |
|
PIN |
463 |
|
< 0.001 |
|
|
|
Negative |
192 |
Reference |
|
Reference |
|
|
Positive |
11 |
3.999 (1.796 - 8.906) |
< 0.001 |
3.767 (1.491 - 9.517) |
0.005 |
|
Unknown |
260 |
1.698 (1.177 - 2.450) |
0.005 |
1.213 (0.815 - 1.806) |
0.342 |
|
Size |
463 |
|
< 0.001 |
|
|
|
＜73 |
301 |
Reference |
|
Reference |
|
|
≥73 |
39 |
4.731 (3.021 - 7.411) |
< 0.001 |
2.423 (1.417 - 4.145) |
0.001 |
|
Unknow |
123 |
1.723 (1.171 - 2.535) |
0.006 |
1.399 (0.918 - 2.132) |
0.119 |
|
Radiation |
463 |
|
< 0.001 |
|
|
|
Yes |
382 |
0.455 (0.311 - 0.667) |
< 0.001 |
0.522 (0.318 - 0.857) |
0.010 |
|
No/Unknown |
81 |
Reference |
|
Reference |
|
|
Chemotherapy |
463 |
|
< 0.001 |
|
|
|
Yes |
380 |
0.497 (0.338 - 0.730) |
< 0.001 |
0.493 (0.290 - 0.836) |
0.009 |
|
No/Unknown |
83 |
Reference |
|
Reference |
|
|
Surgery |
463 |
|
0.002 |
|
|
|
Yes |
134 |
0.523 (0.339 - 0.807) |
0.003 |
0.545 (0.323 - 0.919) |
0.023 |
|
No |
329 |
Reference |
|
Reference |
|