Impact of subdivision of pathological stage I colorectal cancer

Abstract Aim Stage II‐IV colorectal cancers are subdivided according to TNM categories. However, stage I cases are a single category, despite the inclusion of both T1 and T2 cases, which may have different outcomes. The aim of this study was to evaluate the usefulness of subdividing stage I colorectal cancers by T category. Methods From 1984 to 2015, 844 patients with stage I colorectal cancer (T1: 446, T2: 398) underwent colorectal resection with lymph node dissection at three hospitals. The long‐term survival and recurrence rates were compared between T1 and T2. A Cox regression analysis was used to evaluate the risk factors associated with cancer recurrence. Results A comparison of the T1 and T2 groups revealed significant differences in 5‐year overall (95.9% vs 91.4%, P = .008), recurrence‐free (94.8% vs 87.1%, P = .0007), and cancer‐specific survival (97.6% vs 93.6%, P = .004), and in the overall (2.5% vs 6.8%, P = .003), local (0.2% vs 1.5%, P = .04), and lymph node recurrence rates (0.2% vs 1.5%, P = .04). All local and lymph node recurrences were associated with lower rectal cancer, and this difference was significant. The Cox multivariate analysis identified male sex (P = .01, hazard ratio: 4.00, 95% confidence interval: 1.38‐11.55), T2 (P = .02, hazard ratio: 2.98, 95% confidence interval: 1.17‐7.60), and venous invasion (P = .03, hazard ratio: 2.38, 95% confidence interval: 1.12‐5.10) as risk factors for recurrence. Conclusions The subdivision of stage I colorectal cancer according to T category clearly reflected the long‐term outcomes.


| INTRODUC TI ON
Colorectal cancers are staged according to the tumor-node-metastasis (TNM) system that was developed and updated by the International Union for Cancer Control (UICC). Each revision of this system has included subdivisions of stages II, III, and IV by TNM categories, and all three stages were divided into three classes (A, B, and C) in the 8th edition of the TNM classification. 1 For colorectal cancers, the subdivisions of stages II and III clearly reflect the prognoses associated with the corresponding T and N categories. 2 According to the 7th edition of the Cancer Staging Manual, the TNM classification of cancers of the colon and rectum provides more detail than other staging systems. In that edition, stage T4 was subdivided into T4a and T4b, while Stage II was subdivided into three subclasses rather than the binary subdivision used in previous editions. 2 In the 8th edition, the M category was subdivided into three categories, rather than two, because peritoneal metastasis was subdivided as category M1c, whereas the T and N categories were unchanged. 1 This latest subdivision clearly demonstrates the different prognoses associated with lesions of each stage.
In contrast, stage I was not subdivided even in this latest edition of the TNM classification. Although the Surveillance, Epidemiology and End Results Program (SEER) reported 5-year relative survival rates of ≥90% for both T1N0 and T2N0 colorectal cancer cases, a difference of 4.5% in the 5-year relative survival of rectal cancer was observed between T1 and T2 cases. 2 The Japanese Society for Cancer of the Colon and Rectum (JSCCR) also reported a significant difference in the recurrence rates between pT1 and pT2 cases (4.0% vs 7.3%, P = .0076). 3 Stage I colorectal cancers in the T1 category can be cured radically even by endoscopic treatment, whereas most cases in the T2 category require bowel resection with lymph node dissection. Moreover, T1 and T2 cases differ significantly in terms of the recurrence rate, according to data from the JSCCR. 3 These T-category-related differences in long-term outcomes might affect the surveillance schedule after curative resection. These data suggest that accurate prognostic predictions require the subdivision of pStage I cases. With this study, we aimed to clarify the usefulness of subdividing stage I colorectal cancers by examining the long-term outcomes of patients according to the T category.

| Patients
This was a retrospective cohort study of integrated data collected at Koga Community Hospital, Yokohama City University Gastroenterological Center, and Teikyo University. All colorectal cancer patients who underwent curative surgery at these three hospitals between 1984 and 2015 were initially enrolled. The study inclusion criteria were: (i) a histological diagnosis of stage I colorectal cancer according to the TNM system; and (ii) treatment via bowel resection with lymph node dissection. The exclusion criteria were: (i) classification as another stage (0, II, III, or IV); (ii) treatment via irradiation therapy or local excision; (iii) double cancer; (iv) a short follow-up (<36 months); and (v) incomplete data. The patients were divided into two groups depending on whether they were categorized as pathological (p) T1 or pT2 according to the following definitions: pT1, tumor invasion of the submucosa; pT2, tumor invasion of the muscularis propria. The classification of tumor site was according to the 9th edition of the Japanese Classification of Colorectal Appendiceal and Anal Carcinoma. 4 The range of rectum was defined from upper edge of puborectal muscle to promontory of sacrum. Upper side of the peritoneal reflection was defined as the upper rectum, and lower side was the lower rectum.
This study protocol was approved by the ethics committees of Koga Community Hospital, Yokohama City University Gastroenterological Center, and Teikyo University. Consent for this study was not received from all patients, because this was a retrospective observational study. However, the information on this study was disclosed on the homepage (http://www.sunko hkai.or.jp/) of our facility. If the patients and their family had an objection to this study, the corresponding data were deleted.

| Surveillance
Postoperative surveillance was performed according to the JSCCR guidelines for the treatment of colorectal cancer. The patients were

| Analyzed parameters
The overall (OS), recurrence-free (RFS) and cancer-specific survival (CSS) rates, recurrence rates, and timing were compared between the pT1 and pT2 groups. Furthermore, the survival and recurrence rates were compared in the colon only and the rectal cancer only, respectively, between the pT1 and pT2 groups. The recurrence rates according to the recurrence site were compared between the pT1 and pT2 groups. The causes of death were analyzed between the pT1 and pT2 groups.

| Statistical analysis
The data are presented as medians for continuous variables and as frequencies and percentages (%) for categorical variables. The Mann-Whitney U test and Student's t-test were used to evaluate the significance of differences in continuous variables. The chi-squared test was used to evaluate the significance of differences in proportions. The survival rates were compared using a Kaplan-Meier analysis, and significant differences were determined using the log-rank test. A Cox multivariate regression analysis was used to identify significant risk factors for disease recurrence. In all statistical analyses, a P-value of <0.05 was considered statistically significant. IBM SPSS software for Windows, version 21 (SPSS Inc, Chicago, IL, USA) was used for the statistical analyses.

| Patients' characteristics
Some differences were observed in the backgrounds of patients in the pT1 and pT2 groups. Particularly, the pT2 group included a lot of rectal cancers, as well as a higher serum CEA concentration than that in the pT1 group. Regarding the histological findings, the pT2 group had more frequent lymphatic and venous invasion and larger tumor diameters relative to the pT1 group. In the surgical procedure, the rate of laparoscopic surgery in the pT1 group was higher than that in the pT2 group. And, the degree of lymph node dissection in the pT2 group was higher than that in the pT1 group. Regarding the surgical outcomes, patients in the pT2 group had a longer operative time, larger blood loss volume, higher incidence of postoperative complications, and greater number of dissected lymph nodes compared to the pT1 group. There were no inter-group differences in the surveillance duration or the incidence of surveillance duration ≥5 years (Table 1).

| Survival outcomes
The 5-year OS rates were 95.9% and 91.4% in the pT1 and pT2 groups, respectively, and this difference was significant (P = .008; Figure 2 5-year RFS rates were 94.8% and 87.1% in the pT1 and pT2 groups, respectively, and this difference was significant (P = .0007; Figure 3).
The 5-year CSS rates were 97.6% and 93.6% in the pT1 and pT2 groups, respectively, and this difference was significant (P = .004; Figure 4).
The survival outcomes of the pT1 and pT2 groups were also compared with respect to the tumor site. There were significant in-
An analysis according to the recurrence site revealed significant differences in local recurrence (pT1: 0.2% vs pT2: 1.5%, P = .04) and lymph node recurrence (0.2% vs 1.5%, P = .04; Table 2). All local and lymph node recurrences occurred in patients with lower rectal cancer. All cases of local recurrence occurred in male patients. Four of the seven patients with local recurrence underwent resection of recurrent lesion. Of the seven patients with lymph node recurrence, the inguinal lymph node, pelvic side wall lymph node, and para-aortic lymph node were affected in three patients, three patients, and one patient, respectively. Three patients with the inguinal lymph node recurrence and one patient with pelvic side wall lymph node recurrence underwent resection of recurrent lesion. The latter patient had been diagnosed with a pT1 lesion and also developed a hepatic recurrence. There were many hepatic metastases in the recurrence of colon cancer (Table 3).

| D ISCUSS I ON
In our study, we observed differences between the pT1 and T2  After 3 years, the tumor marker examination interval is extended to 6 months, and CT scans are recommended at the same interval until 5 years after surgery. 3

TA B L E 4 Risk factor for recurrence
In Japan, T1 depth of less than 1000 micrometers was classified to pT1a, and 1000 micrometers or deeper depth of pT1 was classified to pT1b from 2013. 19 The measuring method of a detailed invasion distance had been described in the Japanese Classification of Colorectal Carcinoma. 4 However, our pT1 dataset lacks invasion distance about half cases. And, uniform consensus about invasion distance of pT1 tumors has not yet been established internationally.
Future studies that include analyses of budding, perineural invasion, international uniformed invasion distance of pT1, and/or other new markers will refine the determination of risk factors associated with pStage I disease.
In conclusion, the subdivision of pStage I according to pT category appeared to provide a good reflection of the long-term outcomes. We believe that this subdivision into two pStage I classes would be useful for predicting the prognosis of patients and providing effective postoperative surveillance.

ACK N OWLED G EM ENTS
The authors thank Takaaki