Abstract
Background/Aim: The purpose of this study was to investigate the clinical, pathological, and prognostic differences between adenocarcinoma (ADC) and mucinous adenocarcinoma (MUC) in colorectal cancer (CRC). Patients and Methods: This was a retrospective study of a Japanese high-volume cancer Center over a 10-year period. From April 2007 to December 2016, a total of 3,296 patients with primary CRC were included in the study. The clinical characteristics of MUC and ADC were compared. Then, propensity score matching was performed according to a 1:2 ratio. Multivariate analysis was used for independent risk factors related to prognosis. The overall survival (OS) and disease-free survival (DFS) of 126 cases of MUC and 256 cases of ADC were studied, as well as the survival rate of each stage. Results: MUC accounts for 3.82% of the total CRC. Compared to ADC, MUC is more common in female patients (47.62% vs. 38.77%; p=0.045), with higher carcinoembryonic antigen levels (56.35% vs. 34.95%; p<0.001), more ulcerative and infiltrative types (82.54% vs. 72.93%; p=0.016), higher incidence of perineural infiltration (51.59% vs. 41.04%; p=0.018), deeper infiltration (T3-T4: 90.48% vs. 65.84%; p<0.001), and more advanced cancer (stage III-IV: 59.52% vs. 44.79%; p=0.001). MUC is also more likely to recur (24.6% vs. 14.32%; p=0.001). Regarding the long-term survival rate, the OS (p<0.001) and DFS (p=0.05) is consequently worse. After propensity score matching, multivariate analysis showed that MUC was a common independent risk factor for DFS [odds ratio (OR)=4.277; 95% confidence interval (CI), 0.327-0.97; p=0.039], and also for OS (OR= 6.836; 95% CI, 0.274-0.831; p=0.009). In MUC, OS and DFS were still relatively worse (OS: p=0.017; DFS: p=0.038). However, only significant statistical differences were shown in stage II (OS: p=0.003; DFS: p=0.007). No significant differences were noted in the stages I, III, or IV. Conclusion: MUC is a high-risk factor for stage II CRC. Adjuvant chemotherapy should be routinely recommended for patients with MUC stage II, and special attention should be paid during their follow-up.
According to European and American guidelines, patients with a high risk of recurrence after radical surgery for stage II colorectal cancer (CRC) should receive adjuvant chemotherapy (1, 2). However, the definition of “high risk” varies. Mucinous adenocarcinoma (MUC) is defined as a high-risk factor in United States guidelines, but not in European guidelines. In Japan, clinical trials of conventional postoperative adjuvant chemotherapy for stage II CRC have not been found to be successful (3). In 2019, there are few Japanese guidelines for the treatment of stage II rectal cancer using adjuvant chemotherapy (4). There are also different opinions on whether or not adjuvant chemotherapy is required for patients with postoperative stage II MUC.
MUC is a rare tumor classification of CRC. This type of tumor contains neoplastic cells that produce a large amount of extracellular mucin. Histological markers require more than 50% of the tumor volume to consist of a mucin component (5). Some studies report that MUC has worse postoperative results than adenocarcinoma (ADC) (5, 6), while others report the postoperative results of MUC and ADC are the same (7). Some even report that the prognosis of MUC is better (8). This variation in opinion is most likely due to the inconsistent background characteristics of patients with an MUC or ADC history.
Recently, propensity score matching (PSM) has been widely used in clinical studies to minimize the potential for confounding between study groups (9, 10). The purpose of this study was to investigate the clinical, pathological, and prognostic differences between ADC and MUC in CRC using PSM.
Patients and Methods
This was a retrospective cohort study of a Japanese high-volume cancer Center over a 10-year period. The case selection was performed on 4,348 patients with CRC who underwent surgical resection at the International Medical Center of the Saitama Medical University from April 2007 to December 2016. Signet ring cell carcinoma, recurrent CRC, inflammatory colitis combined with malignant tumors, and multiple organ malignancies that were diagnosed simultaneously were excluded. With the approval of the ethics committee of our hospital, all patients submitted their written informed consent. The Ethics Committee of the Saitama Medical University International Medical Center approved the study.
Two Japanese pathologists reviewed all postoperative CRC samples. The pathologists examined the tumor tissue sections and counted the percentage of mucin-producing tumor cells. Tumors with more than 50% mucinous cells were labelled MUC (Figure 1), and those with less than 50% were labelled MUC components. MUC components were not included in the study. The depth of tumor invasion, lymph node metastasis, and clinical information were used to determine the stage of tumor (TNM classification) based on the Cancer Staging Manual by the American Joint Committee on Cancer.
Propensity score weights were used to balance basic variable logistic regression models using polynomials for comparative analysis. Pathological classification, sex, tumor location, tumor morphology type, depth of invasion, lymphatic invasion, perineural infiltration, vascular invasion, preoperative carcinoembryonic antigen (CEA) levels, and lymph node metastasis were incorporated into a multivariate model for PSM. Logistic regression analysis was performed to determine the differences in overall survival (OS) and disease-free survival rate (DFS) between MUC and ADC. All statistical analyses were performed using the SPSS software package (SPSS version 22; IBM, Tokyo, Japan). For differences in categorical variables, chi-square analysis and Fisher exact test were performed. The Kaplan-Meier method was used to estimate OS and a p-value <0.05 indicated statistical significance.
Results
A total of 3,296 patients were included in the study, of whom 3,170 (96.17%) were diagnosed with ADC. MUC was diagnosed in 126 cases (3.82%) (Figure 2). First, 126 patients with MUC and 3,170 patients with ADC were analyzed by single-factor analysis. Compared to ADC, MUC was more common in female patients (47.62% vs. 38.77%; p=0.045), with higher CEA levels (56.35% vs. 34.95%; p<0.001), more ulcerative and infiltrative types (82.54% vs. 72.93%; p=0.016), higher incidence of perineural infiltration (51.59% vs. 41.04%; p=0.018), deep infiltration (T3-T4: 90.48% vs. 65.84%; p<0.001) and advanced cancer (stage III-IV: 59.52% vs. 44.79%; p=0.001). MUC is also more likely to recur (24.6% vs. 14.32%; p=0.001). However, there were no differences in age [mean (y)±SD, 67.0±0.11 vs. 67.3±0.19], lymphatic infiltration (34.92% vs. 31.32%; p=0.39), or lymph node dissection (p=0.169) (Table I).
Then, we analyzed the long-term prognosis of the 2 groups and found that the OS in the MUC group was significantly worse (p<0.001), and the DFS was relatively poor (p=0.05) (Figure 3A and B).
The MUC and ADC groups were then matched according to the propensity score with a ratio of 1:2. A single-factor comparative analysis was performed in 126 patients with MUC and 252 patients with ADC. After matching, there was no statistical difference in the background characteristics of all patients, including age, sex, preoperative tumor marker CEA levels, tumor location, gross tumor classification, tumor invasion depth, vascular invasion, nerve invasion, lymph node metastasis, postoperative staging, and postoperative recurrence (Table II).
After multivariate analysis of the prognostic factors of matched patients, male sex [odds ratio (OR)=4.265; 95% confidence interval (CI)=1.029-3.014; p=0.039], perineural invasion (OR=4.151; 95% CI=0.19-0.968; p=0.042), and vascular invasion (OR=11.371; 95% CI=0.168-0.623; p=0.001) were identified as high independent risk factors for OS. In DFS, infiltrating and ulcerative types were independent risk factors (OR=4.159; 95% CI=0.096-0.955; p=0.041). Mucinous cell carcinoma was a strong independent risk factor in both OS (OR=6.836; 95% CI, 0.274-0.831; p=0.009) and DFS (OR=4.277; 95% CI, 0.327-0.97; p=0.039) (Table III).
After PSM, a re-evaluation of OS and DFS found that the prognosis of the MUC group was still significantly worse than that of the ADC group (OS: p=0.017; DFS: p=0.038) (Figure 4A and B). Then, the prognosis of each stage was analyzed. Eight patients with stage 1 MUC and 16 patients with stage I ADC all survived. One patient with MUC relapsed 7 years post surgery, while there were no recurrences in the ADC group. There was also no difference in the DFS of stage I (data not shown). Interestingly, significant statistical differences were only noted in stage II (OS: p=0.003; DFS: p=0.007) (Figure 5A and B). In stage III, there was no difference between MUC and ADC (OS: p=0.116; DFS: p=0.253) (Figure 6A and B). In stage IV, there was no difference in OS (p=0.358) and cancer-specific survival rate (p=0.819) (Figure 7A and B).
Discussion
Although mucinous has been considered a specific type of CRC, several separate clinical features have been identified in this study. MUC accounted for 3.82%, which was significantly less than the 10-15% reported in Western countries (11). Some studies report that MUC is more common in young people, but we have not found any age difference with ADC in our patient cohort. MUCs are more common in women and are associated with more advanced cancers that are prone to vascular invasion, recurrence, and metastasis after surgery. Therefore, MUC is recognized as having a poor prognosis (6, 12). However, the mechanisms responsible for the worse prognosis are not clearly established. There are reports in the literature that MUC is more prone to peritoneal metastasis; this may be because mucus is produced under pressure, which allows cancer to separate from the tissue layer of the intestinal wall and enter the peritoneal cavity. However, this is only speculation (11, 13, 14). In our multivariable analysis of DFS and OS, we found that MUC is an independent risk factor for both. This is consistent with previous reports (5, 15, 16). Basic research on the mechanism of metastasis of MUC needs to be further evaluated in the future.
The postoperative recurrence rate of MUC is 24.6%, which is nearly 10% higher than the postoperative recurrence rate of ADC. We need to investigate why this difference in recurrence rate exists between the two cancers. In this study, we found that no patients with either MUC or ADC died in stage I. This shows that if the MUC is diagnosed early, a good prognosis can be expected. However, in 8 cases of stage I MUC, 1 case recurred 7 years after surgery. Sixteen patients with ADC did not experience relapses. This suggest that close postoperative follow-up should be recommended even for early MUC. In recent years, it has been reported that magnifying endoscopy can improve the screening of early CRC (17), and the development of endoscopic technology is expected to improve the early diagnosis rate.
In this study, stage III and IV CRC were routinely treated with adjuvant chemotherapy in accordance with the Japanese Colorectal Cancer Treatment Guidelines (4). There were no statistical differences in OS and DFS between MUC and ADC after matching, which is consistent with many other reported results (16, 18). This shows that even with worse biological characteristics, treatment for MUC can achieve as good a prognosis as ADC through the use of chemotherapy.
Since there are no significant differences between the tumors in stages I, III, and IV, the question remains: at what stage are MUC and ADC most different? The guidelines have different definitions of high-risk factors in CRCs, leading to incomplete implementation of adjuvant chemotherapy in patients with stage II CRC across the world. Our study found that the prognosis of OS and DFS in the MUC group was significantly worse than that in the ADC group in stage II CRC. Therefore, conventional adjuvant chemotherapy should be considered for patients with stage II MUC.
It is important to recognize the limitations of this study. The chemotherapy effect for CRC is closely related to the mutation of the KRAS and BRAF genes, or microsatellite instability (MSI) (19-21). In future analyses of MUC, research on the genetic component needs to be strengthened. Due to the different chemotherapy regimens and individual patient differences, we have not evaluated the chemotherapy regimens for patients in stages III and IV, which is a limitation of this study.
Finally, MUC is an independent risk factor associated with the prognosis of CRCs. Compared with ADC, the biggest difference in long-term effect is in stage II. This suggests the utility in routinely screen BRAF, KRAS mutant, VEGF expression and MSI status for stage II MUC, before making an individualized decision on adjuvant chemotherapy to improve the overall prognosis of patients with MUC in stage II (22, 23). Because the clinical proportion of stage II MUC is relatively lower, even though PSM could adjust for known confounding factors, a certain degree of selection bias cannot be ruled out. Prospective clinical trials involving multiple countries and multiple centers are needed in the future.
Conclusion
MUC is a high-risk factor for stage II CRC. Adjuvant chemotherapy should be routinely recommended for patients with stage II MUC, and special attention should be paid during follow-up.
Acknowledgements
Our thanks are extended to BioMed Proofreading, LLC for English copyediting.
Footnotes
Authors' Contributions
LMW drafted the manuscript. YH, GH, TI, HK, KH, NO, MA, TK and SY reviewed its content. All Authors have read and approved the final version of the article.
Conflicts of Interest
The Authors have no competing interests to declare with respect to this study.
Consent for Publication
All patients have agreed to use their personal medical data for research and publication.
- Received January 24, 2020.
- Revision received February 1, 2020.
- Accepted February 3, 2020.
- Copyright© 2020, International Institute of Anticancer Research (Dr. George J. Delinasios), All rights reserved