Portal de Periódicos da CAPES

Impact of Lesion Location on Recurrence After Resection of T1 Colorectal Cancer: Post Hoc Analysis of a Nationwide Multicenter Cohort Study

2023; Elsevier BV; Volume: 166; Issue: 1 Linguagem: Inglês

10.1053/j.gastro.2023.09.038

1528-0012

Tatsunori Minamide, Hiroaki Ikematsu, Yoshiki Kajiwara, Shiro Oka, Yoichi Ajioka, Hideki Ueno, Kenichi Sugihara, Shôichi Saito, Yusuke Fukunaga, Hiroshi Kawachi, Manabu Takamatsu, Motohiro Kojima, Kinichi Hotta, Masayoshi Yamada, Yukihide Kanemitsu, Yutaka Saito, Shigeki Sekine, Shinji Tanaka, Shinji Nagata, Takahiro Nakamura, Kazutaka Yamada, Maki Konno, Soichiro Ishihara, Yusuke Saitoh, Kenji Matsuda, Kazutomo Togashi, Koji Komori, Megumi Ishiguro, Toshio Kuwai, Takashi Okuyama, Akihiro Ohuchi, Shinobu Ohnuma, Kazuhiro Sakamoto, Tamotsu Sugai, Kenji Katsumata, Hiro‐o Matsushita, Hiro‐o Yamano, Keisuke Nakai, Toshio Uraoka, Naohiko Akimoto, Hirotoshi Kobayashi,

Gastric Cancer Management and Outcomes

T1 colorectal cancer (CRC) can be selectively treated with endoscopic local resection (LR) or surgical resection (SR) by pathologic evaluation of lymph node metastasis (LNM) risk.1Nakadoi K. et al.J Gastroenterol Hepatol. 2012; 27: 1057-1062Crossref PubMed Scopus (142) Google Scholar, 2Suh J.H. et al.Endoscopy. 2012; 44: 590-595Crossref PubMed Scopus (90) Google Scholar, 3Ricciardi R. et al.Clin Gastroenterol Hepatol. 2006; 4: 1522-1527Abstract Full Text Full Text PDF PubMed Scopus (82) Google Scholar, 4Nascimbeni R. et al.Dis Colon Rectum. 2002; 45: 200-206Crossref PubMed Scopus (589) Google Scholar, 5Hashiguchi Y. et al.Int J Clin Oncol. 2020; 25: 1-42Crossref PubMed Scopus (1003) Google Scholar In previous studies, LNM frequency was 10%–15% in patients with high-risk lesions.6Ikematsu H. et al.Gastroenterology. 2013; 144: 551-559Abstract Full Text Full Text PDF PubMed Scopus (202) Google Scholar,7Belderbos T.D. et al.Clin Gastroenterol Hepatol. 2017; 15: 403-411.e1Abstract Full Text Full Text PDF PubMed Scopus (54) Google Scholar Consequently, other risk factors require clarification for refined stratification. Regarding lesion location, patients with high-risk T1 rectal cancer have been found to have higher recurrence risk than those with colon lesions in other locations after LR alone.6Ikematsu H. et al.Gastroenterology. 2013; 144: 551-559Abstract Full Text Full Text PDF PubMed Scopus (202) Google Scholar This study aimed to elucidate the impact of lesion location on T1 CRC recurrence, comparing LR to SR. We reviewed the clinical records of patients who underwent T1 CRC resection. Endpoints including recurrence and mortality rates were evaluated. The methods are available in the Supplementary Materials. This study included 3789 patients with complete resection of T1 CRC (LR, n =1044; SR, n = 2745) (Supplementary Figure 1A). The overall recurrence and mortality rates were 1.2% (95% confidence interval [CI], 0.7–2.1) vs 1.8% (95% CI, 1.4–2.4) (P = .256) and 8.5% (95% CI, 6.9–10.4) vs 4.8% (95% CI, 4.0–5.6) (P < .001) in the LR vs SR groups, respectively (Supplementary Figure 1B and C). Of these, 689, 355, and 2745 patients underwent LR for low-risk (group A), LR for high-risk (group B), and SR for high-risk (group C) lesions, respectively. We excluded 1372 patients with overall recurrence and mortality rates of 3.0% (95% CI, 2.1–4.2) and 6.6% (95% CI, 5.2–8.1), respectively, not significantly different from those of the included individuals. Rectal cancers were resected at a significantly younger age and larger size than colon cancers (Figure 1A). LNM rates did not differ between patients with colon and rectal cancer in group C (9.4% vs 10.9%, P = .275). In group A, no significant differences were observed in the overall recurrence and mortality rates between patients with colon and rectal cancer (0.5% [95% CI, 0.1–1.5] vs 2.1% [95% CI, 0.3–7.4] and P = .142; 6.6% [95% CI, 4.7–8.9, respectively] vs 7.4% [95% CI, 3.0–14.6], P = .824) (Figure 1B). The cumulative recurrence (P = .094, 0.6% [95% CI, 0.2–1.7] vs 3.2% [95% CI, 0.6–9.9] at 5 years) and recurrence-free survival (P = .543, 92.5% [95% CI, 89.6–94.6] vs 88.5% [95% CI, 78.3–94.1] at 5 years) did not differ between the patient groups (Figures 1C and D). In group B, no significant differences were found in the overall recurrence and mortality rates between patients with colon and rectal cancer (1.6% [95% CI, 0.4–4.1] vs 3.8% [95% CI, 1.0–9.4], P = .246 and 12.0% [95% CI, 8.3–16.8] vs 12.3% [95% CI, 6.7–20.1], P > .999, respectively) (Figure 1B). The cumulative recurrence (P = .279, 2.1% [95% CI, 0.7–5.0] vs 4.6% [95% CI, 1.5–10.5] at 5 years) and recurrence-free survival (P = .573, 84.0% [95% CI, 77.6–88.7] vs 88.1% [95% CI, 78.9–93.4] at 5 years) did not differ between the patient groups (Figure 1C and D). In group C, significant differences were found in the overall and distant recurrence rates between patients with colon and rectal cancer (1.2% [95% CI, 0.8–1.8] vs 3.5% [95% CI, 2.2–5.1], P < .001, 1.1% [95% CI, 0.7–1.7] vs 3.2% [95% CI, 2.0–4.7], P < .001) (Figure 1B). However, the overall mortality rates did not differ between the patient groups (5.1% [95% CI, 4.2–6.1] vs 3.9% [95% CI, 2.6–5.5], P = .222). The cumulative recurrence was significantly more frequent in the rectal cancer group than in the colon cancer group (P < .001, 1.3% [95% CI, 0.8–1.9] vs 3.7% [95% CI, 2.5–5.3] at 5 years), whereas the recurrence-free survival did not differ between them (P = .925, 94.9% [95% CI, 93.8–95.8] vs 94.3% [95% CI, 92.2–95.8] at 5 years) (Figure 1C and D). The independent risk factors for cumulative recurrence in group C were rectal location (hazard ratio [HR], 2.457; 95% CI, 1.387–4.350), tumor size of ≥20 mm (HR, 1.799; 95% CI, 1.005–3.221), and high budding grade (HR, 1.949; 95% CI, 1.040–3.654) (Figure 1E). The cumulative recurrence in this group did not differ between patients with high-risk rectal cancer with low budding grade and tumor size of <20 mm (2.2% [95% CI, 0.8–1.9] at 5 years) and those with colon cancer (P = .259), which may justify studying the impact of lessened surveillance in this subgroup. A posteriori multivariate analysis revealed that lymphovascular invasion and high budding grade were risk factors for LNM in the colon and rectal groups; meanwhile, submucosal invasion depth of ≥1000 μm was only associated in the colon (Supplementary Figure 1D). The overall recurrence rates after SR significantly differed between patients with high-risk T1 colon and rectal cancer, demonstrating the impact of rectal location on recurrence. However, no significant difference was found after LR for high-risk lesions. Particularly, the recurrence rate of rectal lesions was significantly lower than that previously reported.6Ikematsu H. et al.Gastroenterology. 2013; 144: 551-559Abstract Full Text Full Text PDF PubMed Scopus (202) Google Scholar,8Arthursson V. et al.Endoscopy. 2022; 54: 1071-1077Crossref PubMed Scopus (9) Google Scholar This result might be because of the different definition of high-risk lesion—that is, tumor budding was additionally assessed in our population—and patients' advanced age, increasing other disease-related deaths and reducing the opportunity for recurrence. Moreover, although the LNM rate was approximately 10% in patients with high-risk lesions, the recurrence rate after LR for high-risk lesions was much lower, and the recurrence-free survival rate was more than 80%, the reason for which is unknown. The outcomes after LR of low-risk lesions did not differ between patients with colon and rectal cancer, with significantly low recurrence rates and no CRC-related deaths. Therefore, low-risk rectal lesions can be treated using LR alone without stoma formation, which deteriorates quality of life.9Pommergaard H.C. et al.Colorectal Dis. 2014; 16: 662-671Crossref PubMed Scopus (165) Google Scholar,10Kakodkar R. et al.Colorectal Dis. 2006; 8: 650-656Crossref PubMed Scopus (82) Google Scholar This study has limitations: selection bias due to disparate patient demographic characteristics, insufficient information about surgical procedures and the number of lymph nodes harvested, pathologic diagnosis not independently adjudicated or standardized across sites regarding immunostaining, and a small number of recurrences in groups A and B, which might have limited statistical power. It would be also necessary to confirm these results in non-Japanese, non-Asian populations. In conclusion, this study revealed that rectal location affects recurrence after SR of high-risk T1 CRC, suggesting the need for careful surveillance, whereas LR may be an optimal approach for low-risk lesions given its low recurrence rate. We thank all the other members of the project of the International Collaboration Research for Lymph Node Metastasis of T1 Colorectal Carcinoma from the Japanese Society for Cancer of the Colon and Rectum for their assistance: Kenichi Sugihara,1 Shoichi Saito,2 Yusuke Fukunaga,3 Hiroshi Kawachi,4 Manabu Takamatsu,4 Motohiro Kojima,5 Kinichi Hotta,6 Masayoshi Yamada,7 Yukihide Kanemitsu,8 Yutaka Saito,7 Shigeki Sekine,9 Shinji Tanaka,10 Shinji Nagata,11 Takahiro Nakamura,12 Kazutaka Yamada,13 Maki Konno,14 Soichiro Ishihara,15 Yusuke Saitoh,16 Kenji Matsuda,17 Kazutomo Togashi,18 Koji Komori,19 Megumi Ishiguro,20 Toshio Kuwai,21 Takashi Okuyama,22 Akihiro Ohuchi,23 Shinobu Ohnuma,24 Kazuhiro Sakamoto,25 Tamotsu Sugai,26 Kenji Katsumata,27 Hiro-o Matsushita,28 Hiro-o Yamano,29 Keisuke Nakai,30 Toshio Uraoka,31 Naohiko Akimoto,32 and Hirotoshi Kobayashi33; from 1Tokyo Medical and Dental University, Tokyo, Japan; 2Department of Lower Gastrointestinal Medicine, Cancer Institute Hospital, Japanese Foundation for Cancer Research, Tokyo, Japan; 3Department of Colorectal Surgery, Cancer Institute Hospital, Japanese Foundation for Cancer Research, Tokyo, Japan; 4Department of Pathology, Cancer Institute Hospital, Japanese Foundation for Cancer Research, Tokyo, Japan; 5Division of Pathology, Exploratory Oncology Research and Clinical Trial Center, National Cancer Center, Kashiwa, Japan; 6Division of Endoscopy, Shizuoka Cancer Center, Sunto, Japan; 7Endoscopy Division, National Cancer Center Hospital, Tokyo, Japan; 8Department of Colorectal Surgery, National Cancer Center Hospital, Tokyo, Japan; 9Pathology and Clinical Laboratory Division, National Cancer Center Hospital, Tokyo, Japan; 10Department of Endoscopy, Hiroshima University Hospital, Hiroshima, Japan; 11Department of Gastroenterology, Hiroshima City Asa Citizens Hospital, Hiroshima, Japan; 12Department of Surgery, National Defense Medical College, Tokorozawa, Japan; 13Coloproctology Center Takano Hospital, Kumamoto, Japan; 14Coloproctology Center Takano Hospital, Kumamoto, Japan; 15Department of Surgical Oncology, The University of Tokyo, Tokyo, Japan; 16Digestive Disease Center, Asahikawa City Hospital, Hokkaido, Japan; 17Second Department of Surgery, School of Medicine, Wakayama Medical University, Wakayama, Japan; 18Department of Coloproctology, Aizu Medical Center, Fukushima Medical University, Fukushima, Japan; 19Department of Gastroenterological Surgery, Aichi Cancer Center Hospital, Nagoya, Japan; 20Medical Innovation Promotion Center, Tokyo Medical and Dental University, Tokyo, Japan; 21Department of Gastroenterology, National Hospital Organization Kure Medical Center and Chugoku Cancer Center, Kure, Japan; 22Department of Surgery, Dokkyo Medical University Saitama Medical Center, Saitama, Japan; 23Department of Gastroenterology, School of Medicine, Kurume University, Fukuoka, Japan; 24Department of Surgery, Tohoku University Graduate School of Medicine, Sendai, Japan; 25Department of Coloproctological Surgery, Juntendo University Faculty of Medicine, Tokyo, Japan; 26Department of Molecular Diagnostic Pathology, School of Medicine, Iwate Medical University, Iwate, Japan; 27Department of Gastrointestinal and Pediatric Surgery, Tokyo Medical University, Tokyo, Japan; 28Digestive Disease Center, Akita Red Cross Hospital, Akita, Japan; 29Department of Gastroenterology and Hepatology, Sapporo Medical University School of Medicine, Sapporo, Japan; 30Department of Internal Medicine, Hyogo College of Medicine, Hyogo, Japan; 31Department of Gastroenterology, National Hospital Organization Tokyo Medical Center, Tokyo, Japan; 32Department of Gastroenterology, Nippon Medical School, Graduate School of Medicine, Tokyo, Japan; and 33Department of Surgery, Tokyo Metropolitan Hiroo Hospital, Tokyo, Japan. We also thank Editage (www.editage.com) for English language editing. Tatsunori Minamide, MD (Conceptualization: Equal; Data curation: Supporting; Formal analysis: Lead; Methodology: Lead; Writing – original draft: Lead; Writing – review & editing: Equal) Hiroaki Ikematsu, MD, PhD (Conceptualization: Lead; Data curation: Supporting; Formal analysis: Equal; Supervision: Lead; Writing – original draft: Equal; Writing – review & editing: Lead) Yoshiki Kajiwara, MD, PhD (Conceptualization: Equal; Data curation: Lead; Project administration: Lead; Writing – review & editing: Equal) Shiro Oka, MD, PhD (Conceptualization: Equal; Data curation: Lead; Project administration: Lead; Writing – review & editing: Equal) Yoichi Ajioka, MD, PhD (Conceptualization: Equal; Supervision: Equal; Writing – review & editing: Equal) Hideki Ueno, MD, PhD (Conceptualization: Equal; Data curation: Supporting; Supervision: Equal; Writing – review & editing: Equal) This was a post hoc analysis of a multicenter retrospective cohort study. We reviewed the clinical records of patients who underwent T1 CRC resection at 27 institutions participating in the Japanese Society for Cancer of the Colon and Rectum (JSCCR) between July 2009 and December 2016. Among them, we included patients who had complete resection (R0 resection), defined as resection with histopathologic tumor-free margins. The exclusion criteria were colitic cancer from inflammatory bowel disease, familial adenomatous polyposis, synchronous advanced CRC (within 5 years before T1 CRC diagnosis), synchronous distant metastasis, synchronous cancer in any other organs, and unavailable information. Additionally, patients were excluded if they underwent SR for low-risk lesions or additional treatment other than SR for high-risk lesions. The study protocol was approved by the ethics committee of the JSCCR (approval date: March 13, 2018) and the institutional review board of the National Defense Medical College (approval number: 4479) and was conducted following the ethical principles outlined in the Declaration of Helsinki. Informed consent was obtained through opt-out methods. We obtained the following data from the medical records: age, sex, lesion characteristics, resection method, and follow-up data. Lesion locations were classified into colon (cecum, ascending colon, transverse colon, descending colon, sigmoid colon, or rectosigmoid) and rectum (upper or lower rectum). Resection methods of LR included polypectomy, endoscopic mucosal resection, endoscopic piecemeal mucosal resection, endoscopic submucosal dissection, transanal local excision, and transanal endoscopic microsurgery. Transanal local excision or transanal endoscopic microsurgery was included as LR because these were procedures without lymph node dissection. Follow-up data included the date of the last visit and survival outcome. Data on the recurrence date and location were also obtained. Resected specimens were fixed in formalin, dissected, and stained with hematoxylin and eosin. Subsequently, gastrointestinal pathologists in each institution conducted the pathologic diagnosis according to the Japanese Classification of Colorectal Carcinoma.1Nakadoi K. et al.J Gastroenterol Hepatol. 2012; 27: 1057-1062Crossref PubMed Scopus (142) Google Scholar Depth of submucosal invasion, lymphovascular invasion (lymphatic and/or venous invasion), histologic type, and budding grade were evaluated. When the muscularis mucosae were identifiable, submucosal invasion depth was measured from its lower border. However, the invasion depth was measured from the lesion surface if the muscularis mucosae could not be identified. Tumor budding was counted on the invasion front at 200× magnification. Immunostaining was used to identify the muscularis mucosae and lymphovascular invasion at the pathologists' discretion. The study period started from the time of introduction of the JSCCR guidelines for T1 CRC,2Suh J.H. et al.Endoscopy. 2012; 44: 590-595Crossref PubMed Scopus (90) Google Scholar which established high-risk factors of LNM: deep submucosal invasion depth of ≥1000 μm; lymphovascular invasion; poorly differentiated adenocarcinoma, signet-ring cell carcinoma, or mucinous carcinoma; and BD 2/3. T1 CRCs without these factors and with ≥1 factor were classified as "low-risk" and "high-risk" lesions, respectively. If the pathologic findings after LR confirmed low risk, watchful waiting without additional SR was considered acceptable. If the pathologic findings after LR confirmed high risk, SR with complete pericolic/perirectal and intermediate lymph node dissection (D2) was considered an additional treatment. Although LNM risk was high, the additional SR was not performed considering the patients' will and general conditions or at the physicians' discretion. In this study, the participants were categorized into 3 groups: patients with low-risk T1 CRC treated with LR alone (group A), those with high-risk T1 CRC treated with LR alone (group B), and those with high-risk T1 CRC treated with SR (group C). Surveillance after T1 CRC resection was performed according to the JSCCR guidelines.2Suh J.H. et al.Endoscopy. 2012; 44: 590-595Crossref PubMed Scopus (90) Google Scholar, 3Ricciardi R. et al.Clin Gastroenterol Hepatol. 2006; 4: 1522-1527Abstract Full Text Full Text PDF PubMed Scopus (82) Google Scholar, 4Nascimbeni R. et al.Dis Colon Rectum. 2002; 45: 200-206Crossref PubMed Scopus (589) Google Scholar For colon cancer, physical examination and blood tests, including serum tumor markers, were performed every 3 months for 3 years and, subsequently, every 6 months; abdominal computed tomography (CT) or ultrasonography and chest CT or radiology every 6 months for 3 years and, subsequently, yearly; and total colonoscopy at 1 and 3 years. For rectal cancer, physical examinations including digital rectal examination and blood tests including serum tumor markers were performed every 3 months for 3 years and every 6 months subsequently; abdominopelvic CT or ultrasonography and chest CT or radiology every 6 months for 3 years and then yearly; and total colonoscopy once a year for 3 years. This study's primary endpoint was the overall recurrence rates of T1 colon or rectal cancer by groups. The secondary endpoints were local or distant recurrence rates, mortality rates, cumulative recurrence, recurrence-free survival, and risk factors for recurrence. Recurrence-free survival was defined as the period between the initial T1 CRC resection date and the date of recurrence or mortality from any cause. The period was censored at the date of the last follow-up if any of these events were not observed. Local recurrence was defined as recurrence found within the surgical field and pelvis for colon and rectal lesions, respectively. Distant recurrence was defined as newly diagnosed metastasis in the liver, lung, peritoneum, bone, brain, or nonregional lymph nodes. Descriptive statistics were used to summarize the study population. Categorical variables were expressed as frequency (percentage) and compared using the Fisher exact test. Continuous variables were expressed as medians with interquartile ranges and analyzed using the Mann-Whitney U test. Cumulative recurrence was estimated using the cumulative incidence curves and compared using the Gray test, which considers mortality from causes other than CRC as a competing risk. Recurrence-free survival was calculated and compared using the Kaplan-Meier curves and log-rank test, respectively. A Fine-Gray subdistribution hazard model was used to assess the HR for cumulative recurrence after adjustment for potential confounders as follows: age, sex, initial resection method, lesion location, size, morphology, submucosal invasion depth, lymphovascular invasion, histologic type, and BD. Multivariate logistic regression analysis was used to assess the odds ratio for LNM after adjustment for potential confounders as follows: age, sex, size, morphology, submucosal invasion depth, lymphovascular invasion, histologic type, and BD. All tests were 2 sided, and statistical significance was considered at P < .05. Statistical tests were performed using EZR (Saitama Medical Center, Jichi Medical University).