Abstract
Purpose
For stage II and III rectal cancer, two commonly used treatment strategies are preoperative radiotherapy (RT) combined with surgery, and surgery combined with postoperative RT. This study aims to systematically compare the efficacy of preoperative RT versus postoperative RT for rectal cancer based on the Surveillance, Epidemiology, and End Results (SEER) database.
Materials and methods
Cases of rectal cancer were retrieved using SEER ∗ Stat software, including patients who underwent preoperative RT combined with surgery (PreORT group) and those received surgery combined with postoperative RT (PostORT group). Propensity score matching (PSM) was performed via the R software package (version 4.3.1) to balance intergroup differences. Kaplan–Meier curves for overall survival (OS) and cause-specific survival (CSS) were plotted for the cohort before and after PSM using IBM SPSS 23.0.
Results
A total of 22,079 rectal cancer cases were included, with 16,719 in the PreORT group and 5,360 in the PostORT group. Compared with the PostORT group, the PreORT group showed significantly higher 10-year OS (54.5% Vs. 58.9%, P < 0.001) and CSS (64.5% Vs. 67.9%, P < 0.001). However, in the matched cohort, the PostORT group had higher 10-year OS (55.7% Vs. 57.4%, P = 0.038) and CSS (64.8% Vs. 67.5%, P = 0.040).
Conclusion
While the PreORT group exhibited higher OS and CSS in the entire cohort, the PostORT group showed favorable survival outcomes in the matched cohort. However, given that PreORT remains the standard of care and SEER data lack critical clinical variables influencing treatment selection, these findings should be interpreted as hypothesis-generating rather than indicative of a change in clinical practice. Future randomized, multicenter, prospective studies with large sample sizes are therefore warranted to validate these results and identify subgroups most likely to benefit from each approach.
Supplementary Information
The online version contains supplementary material available at 10.1007/s12672-026-04701-1.
Keywords: Rectal cancer, SEER database, Preoperative radiotherapy, Postoperative radiotherapy
Introduction
Rectal cancer is one of the common malignant tumors of the digestive system. Globally, in 2020, there were an estimated 732,000 new cases of rectal cancer and 339,000 rectal cancer-related deaths [1]. In recent years, surgical techniques in rectal cancer treatment have advanced continuously. For instance, total mesorectal excision (TME) has become an important principle in the surgical treatment of rectal cancer, reducing the local recurrence rate to a certain extent [2, 3]. However, there is still room for improvement in the clinical outcomes of rectal cancer with surgery alone, especially for patients with stage II and stage III disease. Therefore, combining surgery with other treatment modalities, such as radiotherapy and chemotherapy, to improve the local control rate and survival rate of rectal cancer remains the main strategy for current treatment [4, 5].
In the radiotherapy strategies for rectal cancer, preoperative radiotherapy (PreORT) and postoperative radiotherapy (PostORT) are two important options. Sauer et al.'s randomized controlled study compared PreORT and PostORT for rectal cancer. The results showed that the preoperative group had a significantly lower local recurrence rate, a higher sphincter preservation rate in low rectal cancer, and lower grade 3–4 acute toxicity, with no significant difference in 5-year overall survival. This trial established preoperative chemoradiotherapy as the standard treatment [6]. Currently, both the NCCN and ESMO guidelines recommend preoperative chemoradiotherapy as the preferred treatment for stage II/III rectal cancer [5, 7].
However, PreORT may hinder the smooth progress of surgery due to side effects such as radiation-induced intestinal damage and bone marrow suppression [8]. At the same time, in terms of overall survival, PostORT may not be inferior to PreORT [6, 9]. In addition, comparing these two regimens using the Surveillance, Epidemiology, and End Results (SEER) database still holds significant clinical value. First, the SEER database has a large sample size and covers long-term follow-up data, which enables us to further verify the differences between the two regimens in terms of long-term clinical outcomes. Second, in real-world practice, there are still some patients who are not suitable for PreORT, and PostORT remains an important treatment option for this subgroup. This study can provide real-world evidence to support treatment decisions for such patients. Third, most of the existing mainstream conclusions are based on randomized controlled trials (RCTs). However, RCTs have strict inclusion criteria and exclude a large number of patients. The real-world data from the SEER database can better reflect the efficacy differences between the two regimens in a broader population, making up for the limitations of RCT results in clinical promotion.
The purpose of this study is to systematically compare the efficacy of preoperative RT versus postoperative RT for rectal cancer based on the SEER database.
Materials and methods
This study did not require approval from the institutional review board and was exempted from obtaining patient consent, as the relevant data used in the research were already publicly available and had been anonymized. Rectal cancer cases, including patients who underwent PreORT combined with surgery (PreORT group) and surgery combined with PostORT (PostORT group) were retrieved from SEER database using SEER ∗ Stat software 8.4.2 (SEER Program (www.seer.cancer.gov) SEER*Stat Database: Incidence—SEER Research Data, 17 Registries, Nov 2023 Sub (2000–2021)—Linked To County Attributes—Time Dependent (1990–2022) Income/Rurality, 1969–2022 Counties, National Cancer Institute, DCCPS, Surveillance Research Program, released April 2024, based on the November 2023 submission.).
Inclusion criteria
Patients with rectal cancer (either pathologically or clinically diagnosed) retrieved from SEER database had no other malignant tumors;
Diagnosis year of 2000–2021;
Age at diagnosis of 20–79.
Exclusion criteria
Patients did not receive surgical treatment;
Patients received intraoperative radiotherapy;
Patients received radiotherapy both before and after surgery;
Surgery was performed both before and after radiotherapy;
Patient received both surgery and radiotherapy, but the sequence of these treatments is unknown;
Patients are not in stage II/III according to the American Joint Committee on Cancer (AJCC) staging system;
Patients with incomplete data.
The following variables were extracted from the SEER database: sex, age, year of diagnosis, race, marital status at diagnosis, histologic type, median household income, TNM stage, chemotherapy, radiation recodes, sequence of surgery and radiation, tumor size, cause of death to site recode, survival months, and SEER cause-specific death classification. Tumor stage for different year at diagnosis was classified according to different versions of AJCC staging system.
Statistical analysis was performed using IBM SPSS, version 23.0 (IBM, Armonk, NY) and R software package (version 4.3.1, http://www.r-project.org). A p-value < 0.05 was considered significant. Overall survival (OS) (event = death from any cause) curves and cause-specific survival (CSS) (event = death from rectal cancer) curves were plotted by the Kaplan–Meier method.
Given that the median survival times for CSS in both groups were not reached, mean survival time was used to facilitate the understanding of differences in survival outcomes between the two groups, and the same statistical variable was consistently applied for OS for uniformity. The mean survival times were calculated using the Kaplan–Meier survival curve, and the survival time differences between different groups were compared using the log-rank test. Univariate Cox regression analyses were used to calculate the hazard ratio (HR) with 95% confidence intervals (95% CI). Then, a Cox proportional regression model was established for multivariate analysis of variables with statistically significant differences in univariate analyses, and the HR with 95% CI values of factors were calculated.
Propensity score matching (PSM) was used to adjust for differences between PreORT group and PostORT group using R software package as previously described [10]. Matching was conducted with a caliper of 0.02 and without replacement, using the nearest neighbor method witha1:1ratio and exact matching for all covariates. Standard mean difference (SMD) was used to examine the balance of covariate between the PreORT group and PostORT group for the data before and after PSM. For the cohort after PSM, the log-rank test was used to compare the differences in OS and CSS between PreORT group and PostORT group. To further explore the differences in OS and CSS between the two groups, a stratified comparative Log Rank test was conducted based on various parameters.
Results
Based on the inclusion criteria, 53,272 rectal cancer patients were retrieved from the SEER database. After case screening according to the exclusion criteria, a total of 22,079 rectal cancer patients were included, 16,719 in the PreORT group and 5,360 in the PostORT group (Fig. 1).
Fig. 1.
The flowchart of study population selection
With a median follow-up of 62 months (range, 0–263 months), the mean OS for PreORT group and PostORT group were 159.6 months (95% CI, 157.4–161.9 months) and 150.7 months (95% CI, 147.6–153.7 months), respectively (Fig. 2A). The mean CSS for PreORT group and PostORT group were 188.6 months (95% CI, 186.5–190.7 months) and 181.2 months (95% CI, 178.1–184.3 months), respectively (Fig. 2B).
Fig. 2.
Kaplan–Meier survival curves of rectal cancer screened from SEER database. a overall survival. b cause-specific survival
Before PSM, the log-rank test showed that sex, age, year of diagnosis, marital status, histologic type, median household income, T stage, N stage, treatment modality, chemotherapy, and AJCC stage were all factors affecting the mean survival time of OS and CSS (Tables 1 and 2). Since all variables were significant in the univariate analyses, they were all included in the multivariable analysis (Tables 1 and 2).
Table 1.
Univariate and multivariate Cox regression analysis of factors associated with overall survival
| Variables (n = 22,079) | Log Rank test | Univariate analysis | Multivariate analysis | ||||
|---|---|---|---|---|---|---|---|
| Mean survival time (95% CI) | χ2 | P value | HR (95% CI) | P value | HR (95% CI) | P value | |
| Sex | 36.109 | < 0.001 | < 0.001 | < 0.001 | |||
| Female | 163.2 (160.4–166.1) | Reference | Reference | ||||
| Male | 153.4 (151.2–155.6) | 1.155 (1.102–1.210) | < 0.001 | 1.252 (1.193–1.313) | < 0.001 | ||
| Age (y) | 688.546 | < 0.001 | < 0.001 | < 0.001 | |||
| 20–39 | 187.0 (180.1- | Reference | Reference | ||||
| 40–59 | 193.9) | 1.059 (0.943–1.189) | 0.334 | 1.142 (1.016–1.283) | 0.026 | ||
| 60–79 | 180.3 (177.9–182.7) | 1.991 (1.776–2.231) | < 0.001 | 2.206 (1.965–2.478) | < 0.001 | ||
| Year of diagnosis | 129.4 (127.0–131.8) | 179.196 | < 0.001 | < 0.001 | < 0.001 | ||
| 2000–2005 | 142.8 (139.6–145.9) | Reference | Reference | ||||
| 2006–2010 | 126.6 (124.6–128.6) | 0.858 (0.809–0.909) | < 0.001 | 0.847 (0.797–0.900) | < 0.001 | ||
| 2011–2015 | 100.3 (99.2–101.4) | 0.743 (0.698–0.790) | < 0.001 | 0.701 (0.656–0.749) | 0.001 | ||
| 2016–2021 | 62.9 (62.4–63.4) | 0.607 (0.558–0.660) | < 0.001 | 0.579 (0.530–0.632) | < 0.001 | ||
| Race | 1.794 | 0.180 | < 0.001 | < 0.001 | |||
| White | 157.3 (155.4–159.2) | Reference | Reference | ||||
| Black | 139.8 (133.7–145.9) | 1.314 (1.217–1.420) | < 0.001 | 1.287 (1.191–1.392) | < 0.001 | ||
| Other | 168.3 (163.0–173.6) | 0.867 (0.803–0.935) | < 0.001 | 0.934 (0.865–1.009) | 0.084 | ||
| Marital status | 47.353 | < 0.001 | < 0.001 | < 0.001 | |||
| Single | 150.0 (145.5–154.5) | Reference | Reference | ||||
| Married | 165.8 (163.7–168.0) | 0.803 (0.756–0.854) | < 0.001 | 0.735 (0.691–0.782) | < 0.001 | ||
| Other | 134.5 (130.7–138.3) | 1.224 (1.141–1.313) | < 0.001 | 1.038 (0.965–1.116) | 0.313 | ||
| Histologic type | 61.898 | < 0.00 | < 0.001 | < 0.001 | |||
| Adenocarcinoma | 158.6 (156.8–160.4) | 1 | Reference | 1 | Reference | ||
| Mucous tumor | 135.0 (128.2–141.9) | 1.396 (1.280–1.523) | < 0.001 | 1.183 (1.084–1.291) | < 0.001 | ||
| Other | 131.8 (115.7–148.0) | 1.434 (1.156–1.778) | 0.001 | 1.463 (1.179–1.816) | 0.001 | ||
| Median household income ($) | 89.981 | < 0.001 | Reference | Reference | |||
| 0–599,999 | 143.0 (138.8–147.1) | 0.859 | < 0.001 | 0.858 | < 0.001 | ||
| 600,000–799,999 | 155.4 (152.7–158.1) | (0.808–0.913) | < 0.001 | (0.807–0.912) | < 0.001 | ||
| 800,000–999,999 | 161.0 (157.8–164.2) | 0.792 (0.741–0.846) | < 0.001 | 0.807 (0.754–0.863) | < 0.001 | ||
| > 1,000,000 | 170.8 (166.2–175.4) | 0.686 (0.632–0.745) | < 0.001 | 0.721 (0.663–0.784) | < 0.001 | ||
| T stage | 323.906 | < 0.001 | < 0.001 | < 0.001 | |||
| T0 | 79.4 (45.5–113.3) | Reference | Reference | ||||
| T1 | 189.1 (179.0–199.1) | 0.417 (0.103–1.689) | 0.221 | 0.264 (0.065–1.070) | 0.062 | ||
| T2 | 182.6 (176.5–188.6) | 0.492 (0.123–1.973) | 0.317 | 0.296 (0.074–1.188) | 0.086 | ||
| T3 | 158.1 (156.2–160.1) | 0.718 (0.180–2.871) | 0.639 | 0.475 (0.119–1.903) | 0.293 | ||
| T4 | 122.8 (117.5–128.1) | 1.226 (0.306–4.907) | 0.774 | 0.816 (0.204–3.269) | 0.774 | ||
| N stage | 225.516 | < 0.001 | < 0.001 | < 0.001 | |||
| N0 | 163.2 (160.5–165.8) | Reference | Reference | ||||
| N1 | 163.6 (160.9–166.4) | 1.002 (0.952–1.054) | 0.947 | 1.235 (1.171–1.303) | < 0.001 | ||
| N2 | 128.9 (124.7–133.1) | 1.691 (1.593–1.796) | < 0.001 | 2.040 (1.918–2.170) | < 0.001 | ||
| Treatment modality | 32.517 | < 0.001 | < 0.001 | < 0.001 | |||
| PreORT group | 159.5 (157.3–161.7) | Reference | Reference | ||||
| PostORT group | 150.3 (147.3–153.3) | 1.146 (1.091–1.202) | < 0.001 | 0.996 (0.944–1.050) | 0.871 | ||
| Chemotherapy | 63.775 | < 0.001 | < 0.001 | < 0.001 | |||
| No | 126.4 (118.3–134.5) | Reference | Reference | ||||
| Yes | 158.3 (156.5–160.1) | 0.657 (0.593–0.729) | < 0.001 | 0.748 (0.673–0.832) | < 0.001 | ||
| AJCC stage | 48.324 | < 0.001 | < 0.001 | * | * | ||
| II | 163.2 (160.5–165.8) | Reference | |||||
| III | 153.6 (151.3–155.9) | 1.173 (1.120–1.229) | < 0.001 | ||||
*Degree of freedom reduced because of constant or linearly dependent covariates
Bold indicates that P < 0.05.
CI, confidence interval; HR, Hazard ratio; PreORT, preoperative radiotherapy; PostORT, postoperative radiotherapy; AJCC, American Joint Committee on Cancer
Table 2.
Univariate and multivariate Cox regression analysis of factors associated with cause-specific survival
| Variables (n = 22,079) | Log Rank test | Univariate analysis | Multivariate analysis | ||||
|---|---|---|---|---|---|---|---|
| Mean survival time (95% CI) | χ2 | P value | HR (95% CI) | P value | HR (95% CI) | P value | |
| Sex | 15.655 | < 0.001 | < 0.001 | < 0.001 | |||
| Female | 190.5 (187.8–193.3) | Reference | Reference | ||||
| Male | 184.0 (181.8–186.2) | 1.120 (1.059–1.184) | < 0.001 | 1.198 (1.132–1.267) | < 0.001 | ||
| Age (y) | 137.779 | < 0.001 | < 0.001 | < 0.001 | |||
| 20–39 | 195.5 (188.9–202.2) | Reference | Reference | ||||
| 40–59 | 196.0 (193.7–198.3) | 0.962 (0.850–1.088) | 0.535 | 1.060 (0.936–1.200) | 0.361 | ||
| 60–79 | 174.0 (171.3–176.7) | 1.368 (1.210–1.547) | < 0.001 | 1.575 (1.389–1.784) | < 0.001 | ||
| Year of diagnosis | 186.065 | < 0.001 | < 0.001 | < 0.001 | |||
| 2000–2005 | 171.2 (167.7–174.6) | Reference | Reference | ||||
| 2006–2010 | 140.5 (138.5–142.5) | 0.846 (0.789–0.907) | < 0.001 | 0.815 (0.758–0.876) | < 0.001 | ||
| 2011–2015 | 106.6 (105.6–107.7) | 0.712 (0.662–0.764) | < 0.001 | 0.642 (0.594–0.693) | < 0.001 | ||
| 2016–2021 | 64.7 (64.2–65.1) | 0.563 (0.512–0.619) | < 0.001 | 0.505 (0.456–0.558) | < 0.001 | ||
| Race | 0.554 | 0.457 | < 0.001 | < 0.001 | |||
| White | 187.5 (185.6–189.4) | Reference | Reference | ||||
| Black | 167.5 (161.2–173.8) | 1.390 (1.271–1.521) | < 0.001 | 1.354 (1.237–1.482) | < 0.001 | ||
| Other | 192.1 (187.1–197.2) | 0.935 (0.856–1.020) | 0.131 | 0.989 (0.904–1.081) | 0.804 | ||
| Marital status | 14.339 | < 0.001 | < 0.001 | < 0.001 | |||
| Single | 178.5 (174.1–183.0) | Reference | Reference | ||||
| Married | 193.2 (191.2–195.3) | 0.793 (0.739–0.852) | < 0.001 | 0.767 (0.713–0.825) | < 0.001 | ||
| Other | 169.9 (165.7–174.0) | 1.140 (1.049–1.238) | 0.002 | 1.040 (0.954–1.133) | 0.375 | ||
| Histologic type | 76.038 | < 0.001 | < 0.001 | < 0.001 | |||
| Adenocarcinoma | 188.1 (186.4–189.9) | Reference | Reference | ||||
| Mucous tumor | 161.6 (154.3–169.0) | 1.521 (1.376–1.681) | < 0.001 | 1.234 (1.115–1.366) | < 0.001 | ||
| Other | 156.9 (140.4–173.5) | 1.590 (1.245–2.031) | < 0.001 | 1.657 (1.296–2.118) | < 0.001 | ||
| Median household income ($) | 43.851 | < 0.001 | < 0.001 | < 0.001 | |||
| 0–599,999 | 176.6 (172.4–180.9) | Reference | Reference | ||||
| 600,000–799,999 | 185.3 (182.6–188.0) | 0.889 (0.827–0.956) | 0.002 | 0.880 (0.818–0.947) | 0.001 | ||
| 800,000–999,999 | 188.3 (185.1–191.5) | 0.838 (0.775–0.907) | < 0.001 | 0.844 (0.778–0.914) | < 0.001 | ||
| > 1,000,000 | 196.2 (191.8–200.7) | 0.726 (0.659–0.801) | < 0.001 | 0.754 (0.682–0.833) | < 0.001 | ||
| T stage | 345.478 | < 0.001 | < 0.001 | < 0.001 | |||
| T0 | 79.4 (45.5–113.3) | Reference | Reference | ||||
| T1 | 212.8 (204.0–221.5) | 0.293 (0.072–1.197) | 0.087 | 0.189 (0.046–0.772) | 0.020 | ||
| T2 | 212.4 (207.0–217.8) | 0.344 (0.086–1.382) | 0.133 | 0.215 (0.053–0.865) | 0.030 | ||
| T3 | 188.2 (186.3–190.1) | 0.552 (0.138–2.206) | 0.400 | 0.402 (0.100–1.610) | 0.198 | ||
| T4 | 146.9 (141.1–152.6) | 1.055 (0.264–4.226) | 0.939 | 0.765(0.191–3.066) | 0.705 | ||
| N stage | 421.691 | < 0.001 | < 0.001 | < 0.001 | |||
| N0 | 198.8 (196.2–201.3) | Reference | Reference | ||||
| N1 | 191.1 (188.5–193.7) | 1.159 (1.088–1.233) | < 0.001 | 1.439 (1.348–1.535) | < 0.001 | ||
| N2 | 148.9 (144.4–153.3) | 2.199 (2.050–2.358) | < 0.001 | 2.593 (2.413–2.786) | < 0.001 | ||
| Treatment modality | 21.458 | < 0.001 | < 0.001 | < 0.001 | |||
| PreORT group | 188.6 (186.5–190.7) | Reference | Reference | ||||
| PostORT group | 181.2 (178.1–184.3) | 1.146 (1.082–1.214) | < 0.001 | 0.984 (0.924–1.048) | 0.620 | ||
| Chemotherapy | 17.153 | < 0.001 | < 0.001 | < 0.001 | |||
| No | 169.1 (159.8–178.4) | Reference | Reference | ||||
| Yes | 187.3 (185.6–189.0) | 0.756 (0.662–0.864) | < 0.001 | 0.832 (0.726–0.952) | 0.008 | ||
| AJCC stage | 147.703 | < 0.001 | < 0.001 | * | * | ||
| II | 198.8 (196.2–201.3) | Reference | |||||
| III | 179.2 (176.9–181.5) | 1.420 (1.341–1.503) | < 0.001 | ||||
*Degree of freedom reduced because of constant or linearly dependent covariates
Bold indicates that P < 0.05
CI, confidence interval; HR, Hazard ratio; PreORT, preoperative radiotherapy; PostORT, postoperative radiotherapy; AJCC, American Joint Committee on Cancer
There were significant differences in the distributions of sex, age, year of diagnosis, histologic type, median household income, T stage, N stage, chemotherapy and AJCC stage (Table 3). 1693 cases in PostORT didn’t get a match due to the exact matching strategy in PSM. Finally, 3667 cases in PostORT group were matched with cases in PreORT group. SMDs for the data before and after matching show that PSM perfectly balances all covariates (Table 3). In the matched cohort, the PostORT group had higher 10-year OS (55.7% Vs. 57.4%, P = 0.038) and CSS (64.8% Vs. 67.5%, P = 0.040) (Fig. 3).
Table 3.
Standard mean difference for the data before and after propensity score matching
| Variables | Before propensity score matching (n (%)) | After propensity score matching (n (%)) | ||||||
|---|---|---|---|---|---|---|---|---|
| PreORT group (n = 16,719) | PostORT group (n = 5360) | P | SMD | PreORT group (n = 16,719) | PostORT group (n = 5360) | P | SMD | |
| Sex | < 0.001 | 0.061 | 1.000 | < 0.001 | ||||
| Female | 6217.0 (37.2) | 2151.0 (40.1) | 1321.0 (36.0) | 1321.0 (36.0) | ||||
| Male | 10,502.0 (62.8) | 3209.0 (59.9) | 2346.0 (64.0) | 2346.0 (64.0) | ||||
| Age (y) | < 0.001 | 0.143 | 1.000 | < 0.001 | ||||
| 20–39 | 1061.0 (6.3) | 240.0 (4.5) | 115.0 (3.1) | 115.0 (3.1) | ||||
| 40–59 | 8387.0 (50.2) | 2435.0 (45.4) | 1780.0 (48.5) | 1780.0 (48.5) | ||||
| 60–79 | 7271.0 (43.5) | 2685.0 (50.1) | 1772.0 (48.3) | 1772.0 (48.3) | ||||
| Year of diagnosis | < 0.001 | 0.761 | 1.000 | < 0.001 | ||||
| 2000–2005 | 1762.0 (10.5) | 2014.0 (37.6) | 1097.0 (29.9) | 1097.0 (29.9) | ||||
| 2006–2010 | 3472.0 (20.8) | 1325.0 (24.7) | 959.0 (26.2) | 959.0 (26.2) | ||||
| 2011–2015 | 4967.0 (29.7) | 1059.0 (19.8) | 832.0 (22.7) | 832.0 (22.7) | ||||
| 2016–2021 | 6518.0 (39.0) | 962.0 (17.9) | 779.0 (21.2) | 779.0 (21.2) | ||||
| Race | 0.277 | 0.025 | 1.000 | < 0.001 | ||||
| White | 13,480.0 (80.6) | 4321.0 (80.6) | 3128.0 (85.3) | 3128.0 (85.3) | ||||
| Black | 1284.0 (7.7) | 441.0 (8.2) | 208.0 (5.7) | 208.0 (5.7) | ||||
| Other | 1955.0 (11.7) | 598.0 (11.2) | 331.0 (9.0) | 331.0 (9.0) | ||||
| Marital status | < 0.001 | 0.120 | 1.000 | < 0.001 | ||||
| Single | 3269.0 (19.6) | 808.0 (15.1) | 508.0 (13.9) | 508.0 (13.9) | ||||
| Married | 10,380.0 (62.1) | 3475.0 (64.8) | 2514.0 (68.6) | 2514.0 (68.6) | ||||
| Other | 3070.0 (18.4) | 1077.0 (20.1) | 645.0 (17.6) | 645.0 (17.6) | ||||
| Histologic type | < 0.001 | 0.067 | 1.000 | < 0.001 | ||||
| Adenocarcinoma | 15,790.0 (94.4) | 4991.0 (93.1) | 3542.0 (96.6) | 3542.0 (96.6) | ||||
| Mucous tumor | 804.0 (4.8) | 298.0 (5.6) | 116.0 (3.2) | 116.0 (3.2) | ||||
| Other | 125 (0.8) | 71 (1.3) | 9.0 (0.2) | 9.0 (0.2) | ||||
| Household income ($) | < 0.001 | 0.103 | ||||||
| 0–599,999 | 2883.0 (17.2) | 946.0 (17.6) | 628.0 (17.1) | 628.0 (17.1) | ||||
| 600,000–799,999 | 6155.0 (36.8) | 2211.0 (41.2) | 1521.0 (41.5) | 1521.0 (41.5) | ||||
| 800,000–999,999 | 5054.0 (30.2) | 1458.0 (27.2) | 1036.0 (28.3) | 1036.0 (28.3) | ||||
| > 1,000,000 | 2627.0 (15.7) | 745.0 (13.9) | 482.0 (13.1) | 482.0 (13.1) | ||||
| T stage | < 0.001 | 0.312 | NAN | < 0.001 | ||||
| T0 | 5.0 (0.0) | 1.0 (0.0) | 0.0 (0.0) | 0.0 (0.0) | ||||
| T1 | 216.0 (1.3) | 260.0 (4.9) | 82.0 (2.2) | 82.0 (2.2) | ||||
| T2 | 981.0 (5.9) | 639.0 (11.9) | 320.0 (8.7) | 320.0 (8.7) | ||||
| T3 | 13,444.0 (80.4) | 3936.0 (73.4) | 2993.0 (81.6) | 2993.0 (81.6) | ||||
| T4 | 2073.0 (12.4) | 524.0 (9.8) | 272.0 (7.4) | 272.0 (7.4) | ||||
| N stage | < 0.001 | 0.268 | 1.000 | < 0.001 | ||||
| T0 | 5964.0 (35.7) | 2001.0 (37.3) | 1501.0 (40.9) | 1501.0 (40.9) | ||||
| T1 | 8447.0 (50.5) | 2136.0 (39.9) | 1520.0 (41.5) | 1520.0 (41.5) | ||||
| T2 | 2308.0 (13.8) | 1223.0 (22.8) | 646.0 (17.6) | 646.0 (17.6) | ||||
| Chemotherapy | < 0.001 | 0.304 | ||||||
| No | 253.0 (1.5) | 422.0 (7.9) | 57.0 (1.6) | 57.0 (1.6) | ||||
| Yes | 16,466.0 (98.5) | 4938.0 (92.1) | 3610.0 (98.4) | 3610.0 (98.4) | ||||
| AJCC Stage | 0.029 | 0.034 | 1.000 | < 0.001 | ||||
| II | 5964.0 (35.7) | 2001.0 (37.3) | 1501.0 (40.9) | 1501.0 (40.9) | ||||
| III | 10,755.0 (64.3) | 3359.0 (62.7) | 2166.0 (59.1) | 2166.0 (59.1) | ||||
SMD, Standard mean difference; NAN, not a number; AJCC, American Joint Committee on Cancer
Bold indicates that P < 0.05
Fig. 3.
Kaplan–Meier survival curves of rectal cancer screened from SEER database after propensity score matching. a overall survival. b cause-specific survival
In the matched cohort, stratified comparative Log Rank test was shown in Tables S1 and S2. In the subgroups of male patients, diagnosed between 2000 and 2005, married, with adenocarcinoma, median household income of $600,000–799,999, and receiving chemotherapy, OS of patients in the PostORT group was better than that in the PreORT group; in the subgroups of patients aged 40–59 years, diagnosed between 2000 and 2005, Black, married, with median household income of $600,000–799,999, T3 stage, and receiving chemotherapy, the CSS of patients in the PostORT group was better than that in the PreORT group.
Discussion
As one of the largest cancer registry databases globally, the SEER database enables real-world data (RWD) analysis that play an irreplaceable role in evaluating long-term outcomes of complex treatment strategies, particularly for large cohorts of rare subgroups where randomized trials are impractical [11]. This study systematically compares PreORT and PostORT efficacy for stage II/III rectal cancer using the SEER database. By applying PSM to a large initial cohort (n = 22,079; 3667 matched pairs), we minimized baseline differences, generating high-quality comparative effectiveness evidence that addresses RCT limitations in sample size and extended follow-up duration.
After PSM, the PostORT group demonstrated significantly higher 10-year OS (55.7% vs. 57.4%, P = 0.038) and CSS (64.8% vs. 67.5%, P = 0.040) compared to the PreORT group. These findings challenge the established guideline preference for preoperative chemoradiotherapy as the standard for stage II and III rectal cancer. While PreORT showed better survival in the unmatched cohort, the significant long-term survival advantage for PostORT in the balanced cohort underscores the need for personalized treatment strategies based on individual patient characteristics (e.g., age, comorbidities, tumor location, anticipated surgical margins). PostORT represents a potentially better choice for patients with concerns regarding PreORT tolerance.
Our results differ from the landmark CAO/ARO/AIO-94 trial, which demonstrated PreORT’s superiority in reducing local recurrence (6% vs. 13%) and improving sphincter preservation, but found no significant OS difference [6]. This divergence may be attributed to key methodological and contextual factors: study design (RCT vs. retrospective cohort), our substantially larger matched sample size (3667 pairs vs. 421 in CAO/ARO/AIO-94), longer median follow-up (62 months vs. 46 months), and significant advancements in radiotherapy techniques (e.g., IMRT, image-guidance improving normal tissue sparing) [12–15] and systemic therapy (e.g., adoption of oxaliplatin-based regimens) [16–18] occurring during our study period (2000–2021). These factors likely enhance the generalizability of our survival findings to contemporary practice.
While current NCCN and ESMO guidelines strongly favor PreORT based on its established local control and sphincter preservation benefits [5, 7], our survival findings contest the perception of universal PreORT superiority. This necessitates greater emphasis on individualized radiotherapy timing selection. The observed long-term survival advantage for PostORT in balanced patients supports the notion that it may be non-inferior, or potentially superior, to PreORT in OS [6, 9]. Mechanistically, PostORT allows precise targeting of residual disease based on surgical pathology and may facilitate delivery of full-dose, multi-agent adjuvant chemotherapy [19]. Conversely, PreORT-induced bowel toxicity can compromise postoperative recovery and tolerance to subsequent chemotherapy [13].
We acknowledge the well-documented value of PreORT: it effectively reduces tumor volume, facilitating surgery [4, 6, 20, 21], and significantly improves sphincter preservation rates, thereby enhancing quality of life [6, 22]. Therefore, PreORT remains the preferred approach for patients where sphincter preservation is challenging or tumor downstaging is critical for resectability. However, our results indicate that for patients where sphincter preservation is achievable without PreORT or who have contraindications to PreORT, PostORT represents a viable alternative associated with superior long-term survival in this analysis.
This finding aligns with prior SEER analysis by Peng et al. in T3N0 rectal cancer (1998–2008), which also reported significantly reduced cancer mortality with PostORT (HR = 0.688; 95% CI, 0.578–0.819; P < 0.001) but no significant survival benefit for PreORT (HR = 0.863; 95% CI, 0.715–1.043; P = 0.127) [23]. Our study extends this observation to the broader stage II/III population, utilizing more contemporary data (2000–2021) reflecting modern treatment paradigms.
However, our findings require cautious interpretation due to potential inherent selection bias in treatment assignment—particularly given that PreORT had been established as the standard of care for locally advanced rectal cancer throughout the study period (2000–2001). Patients who received PostORT likely represented a highly selected subgroup: many may have had clinical under-staging before surgery (e.g., initially classified as clinical stage cT1-2N0) and were only reclassified as pathological stage pStage II/III after postoperative pathological examination. In fact, we found that in the data before PSM, the proportions of T1 (1.3% vs. 4.9%) and T2 (5.9% vs. 11.9%) stages in the PostORT group were both higher than those in the PreORT group. This indicates that the initial tumor burden of the PostORT cohort may be lower compared with the PreORT group. Even after PSM, this "tumor burden bias" remains difficult to fully eliminate, as PSM cannot account for unmeasured differences in baseline tumor characteristics (e.g., microscopic tumor extension) that may influence prognosis. Therefore, the observed survival advantage of the PostORT group in the matched cohort—and the reversal of survival curves compared to the entire cohort—might partially stem from this hidden selection bias (i.e., inherently better prognosis due to smaller initial tumor volume) rather than the therapeutic efficacy of PostORT itself. This limitation underscores the need for extra caution when attributing survival differences solely to the timing of radiotherapy, as baseline tumor burden could be a confounding factor that independently affects treatment outcomes.
This study has several limitations inherent to the SEER database. Lack of detailed radiotherapy parameters (dose, fields), specific chemotherapy regimens and administration timing, performance status, surgical quality metrics (e.g., TME completeness), and data on local recurrence outcomes or treatment-related toxicity hinders comprehensive assessment and mechanism analysis of efficacy comparison. Despite PSM, unmeasured confounders (e.g., performance status, exact comorbidity burden, postoperative complications) may persist. Furthermore, another major limitation of our study is that pathological stage in the PreORT group represents post-treatment status, whereas in the PostORT group it reflects untreated surgical pathology, and this inherent difference cannot be resolved by propensity score matching or other statistical methods. Notably, patient-reported outcomes and key quality-of-life indicators, such as anorectal and genitourinary function—central considerations in the PreORT/PostORT debate—are unavailable [24]. When formulating treatment decisions in clinical practice, full consideration should be given to the unmeasured confounders in the SEER database. This avoids absolutizing the survival difference of PostORT as a treatment advantage, as the absolute difference between the two groups was small (1.7% for 10 years OS and 2.7% for 10 years CSS) in this study.
Conclusions
While the PreORT group exhibited higher OS and CSS in the entire cohort, the PostORT group showed favorable survival outcomes in the propensity score-matched cohort. However, given that PreORT remains the standard of care and SEER data lack critical clinical variables influencing treatment selection, these findings should be interpreted as hypothesis-generating rather than indicative of a change in clinical practice. Future randomized, multicenter, prospective studies with large sample sizes are therefore warranted to validate these results and identify subgroups most likely to benefit from each approach.
Supplementary Information
Below is the link to the electronic supplementary material.
Acknowledgements
Not applicable.
Author contributions
Mingyuan He: Data curation, Formal analysis, Software, Validation, Writing—original draft, Writing—review and editing. Jinbao Wang: Data curation, Formal analysis, Software, Validation, Writing—original draft, Writing—review and editing. Ning Zhang: Data curation, Formal analysis, Software, Validation, Writing—original draft, Writing—review and editing. Hongfu Zhao: Data curation, Software, Writing—original draft. Guanghui Cheng: Conceptualization, Methodology, Quality control of data and algorithms, Writing—review and editing, review.
Funding
This work was partially supported by grants from the National Natural Science Foundation of China (82073331); Project of Science and Technology Department of Jilin Province (20230203082SF); Horizontal Project of Jilin University (2023YX0056) and the Natural Science Foundation of Jilin Province, China (212558JC010284904), Huilan Public Welfare Fund (Z610719157137141057).
Data availability
All data, models, or code generated or used during the study are available from the corresponding author by request.
Declarations
Ethics approval and consent to participate
Not applicable.
Consent for publication
Not applicable.
Competing interests
The authors declare that they have no competing interests.
Footnotes
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Mingyuan He, Jinbao Wang and Ning Zhang contributed equally to this work.
References
- 1.Siegel RL, Miller KD, Wagle NS, et al. Cancer statistics, 2023. CA Cancer J Clin. 2023;73:17–48. [DOI] [PubMed] [Google Scholar]
- 2.Zeng Z, Luo S, Zhang H, et al. Transanal vs Laparoscopic Total Mesorectal Excision and 3-Year Disease-Free Survival in Rectal Cancer: the TaLaR Randomized Clinical Trial. JAMA. 2025;333:774–83. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Rutgers ML, Burghgraef TA, Hol JC, et al. Total mesorectal excision in MRI-defined low rectal cancer: multicentre study comparing oncological outcomes of robotic, laparoscopic and transanal total mesorectal excision in high-volume centres. BJS Open. 2024;8:zrae029. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.van Gijn W, Marijnen CA, Nagtegaal ID, et al. Preoperative radiotherapy combined with total mesorectal excision for resectable rectal cancer: 12-year follow-up of the multicentre, randomised controlled TME trial. Lancet Oncol. 2011;12:575–82. [DOI] [PubMed] [Google Scholar]
- 5.Glynne-Jones R, Wyrwicz L, Tiret E, et al. Rectal cancer: ESMO clinical practice guidelines for diagnosis, treatment and follow-up. Ann Oncol. 2017;28:iv22–40. [DOI] [PubMed] [Google Scholar]
- 6.Sauer R, Becker H, Hohenberger W, et al. Preoperative versus postoperative chemoradiotherapy for rectal cancer. N Engl J Med. 2004;351:1731–40. [DOI] [PubMed] [Google Scholar]
- 7.NCCN Clinical Practice Guidelines in Oncology: Rectal Cancer. Available at: 2025. https://www.nccn.org/professionals/physician_gls/pdf/rectal.pdf (accessed March 31 2025).
- 8.Zhou ZR, Liu SX, Zhang TS, et al. Short-course preoperative radiotherapy with immediate surgery versus long-course chemoradiation with delayed surgery in the treatment of rectal cancer: a systematic review and meta-analysis. Surg Oncol. 2014;23:211–21. [DOI] [PubMed] [Google Scholar]
- 9.Duzova M, Basaran H, Inan G, et al. Preoperative versus postoperative chemoradiotherapy for locally advanced rectal cancer: outcomes of survival, toxicity, sphincter preserving and prognostic factors. Transpl Immunol. 2021;69:101489. [DOI] [PubMed] [Google Scholar]
- 10.Zhang Z, Zhao H, Cheng G. Long term clinical outcomes of cervical cancer patients who were recommended surgery but did not undergo it: a SEER database study. Eur J Surg Oncol. 2024;50:108572. [DOI] [PubMed] [Google Scholar]
- 11.Che WQ, Li YJ, Tsang CK, et al. How to use the Surveillance, Epidemiology, and End Results (SEER) data: research design and methodology. Mil Med Res. 2023;10:50. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Thong DW, Chakraborty P, Rajan R, et al. Adjuvant radiotherapy in incidental positive nodal disease in rectal cancer-a systemic review. J Surg Oncol. 2025. 10.1002/jso.70025. [DOI] [PubMed] [Google Scholar]
- 13.Pierrard J, Donnay L, Collard A, et al. Radiotherapy boost to the primary tumour in locally advanced rectal cancer: systematic review of practices and meta-analysis. Clin Transl Radiat Oncol. 2025;54:101014. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Chiloiro G, Gani C, Boldrini L. Rectal cancer MRI guided radiotherapy: a practical review for the physician. Semin Radiat Oncol. 2024;34:64–8. [DOI] [PubMed] [Google Scholar]
- 15.Dröge LH, Weber HE, Guhlich M, et al. Reduced toxicity in the treatment of locally advanced rectal cancer: a comparison of volumetric modulated arc therapy and 3D conformal radiotherapy. BMC Cancer. 2015;15:750. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Petersen SH, Harling H, Kirkeby LT, et al. Postoperative adjuvant chemotherapy in rectal cancer operated for cure. Cochrane Database Syst Rev. 2012;2012:CD004078. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Neugut AI, Fleischauer AT, Sundararajan V, et al. Use of adjuvant chemotherapy and radiation therapy for rectal cancer among the elderly: a population-based study. J Clin Oncol. 2002;20:2643–50. [DOI] [PubMed] [Google Scholar]
- 18.André T, Boni C, Mounedji-Boudiaf L, et al. Oxaliplatin, fluorouracil, and leucovorin as adjuvant treatment for colon cancer. N Engl J Med. 2004;350:2343–51. [DOI] [PubMed] [Google Scholar]
- 19.Sauer R, Liersch T, Merkel S, et al. Preoperative versus postoperative chemoradiotherapy for locally advanced rectal cancer: results of the German CAO/ARO/AIO-94 randomized phase III trial after a median follow-up of 11 years. J Clin Oncol. 2012;30:1926–33. [DOI] [PubMed] [Google Scholar]
- 20.Deng Y, Chi P, Lan P, et al. Modified FOLFOX6 with or without radiation versus fluorouracil and leucovorin with radiation in neoadjuvant treatment of locally advanced rectal cancer: initial results of the Chinese FOWARC multicenter, open-label, randomized three-arm phase III trial. J Clin Oncol. 2016;34:3300–7. [DOI] [PubMed] [Google Scholar]
- 21.Garcia-Aguilar J, Patil S, Gollub MJ, et al. Organ preservation in patients with rectal adenocarcinoma treated with total neoadjuvant therapy. J Clin Oncol. 2022;40:2546–56. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Roh MS, Colangelo LH, O’Connell MJ, et al. Preoperative multimodality therapy improves disease-free survival in patients with carcinoma of the rectum: NSABP R-03. J Clin Oncol. 2009;27:5124–30. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Peng LC, Milsom J, Garrett K, et al. Surveillance, epidemiology, and end results-based analysis of the impact of preoperative or postoperative radiotherapy on survival outcomes for T3N0 rectal cancer. Cancer Epidemiol. 2014;38:73–8. [DOI] [PubMed] [Google Scholar]
- 24.Loos M, Quentmeier P, Schuster T, et al. Effect of preoperative radio(chemo)therapy on long-term functional outcome in rectal cancer patients: a systematic review and meta-analysis. Ann Surg Oncol. 2013;20:1816–28. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
All data, models, or code generated or used during the study are available from the corresponding author by request.