Local Excision vs Radical Resection of Colorectal Carcinoma
Local Excision vs Radical Resection of Colorectal Carcinoma
This study included 7378 (18.0%) local excisions and 36,116 (83.0%) major resections for 30,232 colon and 13,262 patients with rectal cancer. There were 3553 (8.2%) patients with carcinoma in situ and 39,941 (91.8%) with a clinical stage I cancer (20,480 with T1 and 19,461 with T2 cancer).
Basic patient- and tumor-related demographics split by local excision versus major resection are shown in Table 1. In the local excision group, 25.7% of patients were aged 18 to 59 years, compared with 22.2% in the major resection group (Fig. 1). There was a lower rate of use of neoadjuvant radiotherapy in the local excision versus major resection group (0.9% vs 3.8%), but a higher rate of use of adjuvant therapy (8.4% vs 1.3%, respectively). There was a higher rate of patients with rectal cancer in the local excision group (50.4%) compared with the major resection group (26.4%) and a lower rate of stage I cancer (80.3% vs 94.2%, respectively, for local excision and major resection).
(Enlarge Image)
Figure 1.
Age distributions across T-stage groups.
Patient variables split by T-stage (Tis, T1, and T2) are shown in Table 2. In patients with Tis (carcinoma in situ), 30.6% undergoing local excision were aged 18 to 59 years. In stage I cancer, this rate was 24.5% (25.5% and 19.8%, respectively, for T1 and T2 cancer). Higher T stage was associated with increasing age, female proportion, and rate of radiotherapy use. The proportion of rectal cancer per group was similar between T stages, ranging from 26.8% to 31.3%. The predominance of moderately differentiated tumors seen in T1/2 cancer was not seen in Tis because of an increased rate of well-differentiated tumors. There were slightly more poorly differentiated tumors in the T2 versus T1 group (10.6% vs 7.2%, respectively). The rate of local excision was highest for Tis (40.8%) and lowest for T2 (4.3%).
The univariable and multivariable influence of patient demographics (age, gender, year of diagnosis, radiotherapy, differentiation, stage) on survival is shown for colon cancer in Table 3 and rectal cancer in Table 4. Unadjusted proportions of patients surviving to 5 years (CSS and OS, estimated by the Kaplan-Meier method) are shown in Table 5.
After adjusting for confounding factors (shown in Table 3 and Table 4), local tumor excision (compared to major resection) for carcinoma in situ was associated with HRs for 5-year CSS of 1.06 (P = 0.814) and 0.78 (P = 0.494) for colon and rectum lesions, respectively. There were also no OS differences for carcinoma in situ (colon OS HR: 1.04, P = 0.688, and rectum OS HR 1.05, P = 0.777, Table 5, Fig. 2).
(Enlarge Image)
Figure 2.
Adjusted HRs for OS and CSS. Point estimates above the horizontal line indicate worse survival for local excision compared to major resection and below indicate improved survival. Vertical lines indicate 95% CIs; if these cross the horizontal line (ie, 1.0), there is survival equivalence of the 2 techniques at P < 0.05.
Local excision of T1 and T2 colon cancer were associated with reduced CSS (HR: 1.31, P = 0.020, and HR: 2.89, P < 0.001, respectively) and reduced OS (HR: 1.56, P < 0.001, and HR: 2.41, P < 0.001). Local excision of T1 rectal cancer did not affect CSS (HR: 1.16, P = 0.236) but was associated with reduced OS (HR: 1.29, P < 0.001). However, local excision of T2 rectal cancer was associated with significantly reduced CSS (HR: 1.71, P < 0.001) and OS (HR: 1.71, P < 0.001, Fig. 2).
In patients with rectal cancer undergoing neoadjuvant radiotherapy (n = 1335), local excision, compared with major resection, was associated with oncological equivalence for T1 cancer (CSS adjusted HR: 1.21, 95% CI: 0.27–5.51, P = 0.802) and T2 cancer (adjusted HR: 1.23, 95% CI: 0.44–3.45, P = 0.701). This is compared to the oncological outcome of local excision from patients not undergoing radiotherapy: an oncological equivalence with T1 cancer (CSS adjusted HR 1.18, 95% CI 0.91–1.54, P = 0.220) and oncological disadvantage with T2 cancer (CSS adjusted HR: 1.83, 95% CI: 1.32–2.54, P < 0.001). There were insufficient patients to test for the same relationship in T0 cancer after neoadjuvant radiotherapy.
When split by site and stage, there were no significant changes in survival for local excision techniques over time. For major resections, an initial survival improvement over time for colon cancer became nonsignificant in the most recent years; there were no changes for rectal cancer (supplemental Table 1 available at http://links.lww.com/SLA/A438). All models were repeated to include patients with missing data, who accounted for less than 5% of the total cohort. These patients did not change the significance or magnitude of effect size for local excision in any model.
Results
Demographics
This study included 7378 (18.0%) local excisions and 36,116 (83.0%) major resections for 30,232 colon and 13,262 patients with rectal cancer. There were 3553 (8.2%) patients with carcinoma in situ and 39,941 (91.8%) with a clinical stage I cancer (20,480 with T1 and 19,461 with T2 cancer).
Basic patient- and tumor-related demographics split by local excision versus major resection are shown in Table 1. In the local excision group, 25.7% of patients were aged 18 to 59 years, compared with 22.2% in the major resection group (Fig. 1). There was a lower rate of use of neoadjuvant radiotherapy in the local excision versus major resection group (0.9% vs 3.8%), but a higher rate of use of adjuvant therapy (8.4% vs 1.3%, respectively). There was a higher rate of patients with rectal cancer in the local excision group (50.4%) compared with the major resection group (26.4%) and a lower rate of stage I cancer (80.3% vs 94.2%, respectively, for local excision and major resection).
(Enlarge Image)
Figure 1.
Age distributions across T-stage groups.
Patient variables split by T-stage (Tis, T1, and T2) are shown in Table 2. In patients with Tis (carcinoma in situ), 30.6% undergoing local excision were aged 18 to 59 years. In stage I cancer, this rate was 24.5% (25.5% and 19.8%, respectively, for T1 and T2 cancer). Higher T stage was associated with increasing age, female proportion, and rate of radiotherapy use. The proportion of rectal cancer per group was similar between T stages, ranging from 26.8% to 31.3%. The predominance of moderately differentiated tumors seen in T1/2 cancer was not seen in Tis because of an increased rate of well-differentiated tumors. There were slightly more poorly differentiated tumors in the T2 versus T1 group (10.6% vs 7.2%, respectively). The rate of local excision was highest for Tis (40.8%) and lowest for T2 (4.3%).
Survival by Demographics
The univariable and multivariable influence of patient demographics (age, gender, year of diagnosis, radiotherapy, differentiation, stage) on survival is shown for colon cancer in Table 3 and rectal cancer in Table 4. Unadjusted proportions of patients surviving to 5 years (CSS and OS, estimated by the Kaplan-Meier method) are shown in Table 5.
Multivariable Survival Analysis by Site and Stage
After adjusting for confounding factors (shown in Table 3 and Table 4), local tumor excision (compared to major resection) for carcinoma in situ was associated with HRs for 5-year CSS of 1.06 (P = 0.814) and 0.78 (P = 0.494) for colon and rectum lesions, respectively. There were also no OS differences for carcinoma in situ (colon OS HR: 1.04, P = 0.688, and rectum OS HR 1.05, P = 0.777, Table 5, Fig. 2).
(Enlarge Image)
Figure 2.
Adjusted HRs for OS and CSS. Point estimates above the horizontal line indicate worse survival for local excision compared to major resection and below indicate improved survival. Vertical lines indicate 95% CIs; if these cross the horizontal line (ie, 1.0), there is survival equivalence of the 2 techniques at P < 0.05.
Local excision of T1 and T2 colon cancer were associated with reduced CSS (HR: 1.31, P = 0.020, and HR: 2.89, P < 0.001, respectively) and reduced OS (HR: 1.56, P < 0.001, and HR: 2.41, P < 0.001). Local excision of T1 rectal cancer did not affect CSS (HR: 1.16, P = 0.236) but was associated with reduced OS (HR: 1.29, P < 0.001). However, local excision of T2 rectal cancer was associated with significantly reduced CSS (HR: 1.71, P < 0.001) and OS (HR: 1.71, P < 0.001, Fig. 2).
In patients with rectal cancer undergoing neoadjuvant radiotherapy (n = 1335), local excision, compared with major resection, was associated with oncological equivalence for T1 cancer (CSS adjusted HR: 1.21, 95% CI: 0.27–5.51, P = 0.802) and T2 cancer (adjusted HR: 1.23, 95% CI: 0.44–3.45, P = 0.701). This is compared to the oncological outcome of local excision from patients not undergoing radiotherapy: an oncological equivalence with T1 cancer (CSS adjusted HR 1.18, 95% CI 0.91–1.54, P = 0.220) and oncological disadvantage with T2 cancer (CSS adjusted HR: 1.83, 95% CI: 1.32–2.54, P < 0.001). There were insufficient patients to test for the same relationship in T0 cancer after neoadjuvant radiotherapy.
When split by site and stage, there were no significant changes in survival for local excision techniques over time. For major resections, an initial survival improvement over time for colon cancer became nonsignificant in the most recent years; there were no changes for rectal cancer (supplemental Table 1 available at http://links.lww.com/SLA/A438). All models were repeated to include patients with missing data, who accounted for less than 5% of the total cohort. These patients did not change the significance or magnitude of effect size for local excision in any model.
Source...