3.1 Sociodemographic and clinical characteristics of CRC patients
A total of 729 patients who underwent surgery for primary CRC in our hospital were enrolled in this study, including 58 patients in COPD cohort and 671 patients in non-COPD cohort. There were 48 (82.76%) males in COPD cohort and 376 (56.04%) males in non-COPD cohort, with statistically significant differences in the gender distribution between the two cohorts (P < 0.0001). In addition, we found obvious statistical differences in the age, BMI, concomitant coronary heart disease, and smoking history between COPD and non-COPD cohorts (P < 0.0001). There were no significant differences in concomitant diabetes mellitus, hypertension, alcohol consumption history, tumor location, TNM tumor stage, differentiated degree, histological type, maximum tumor diameter, neoadjuvant therapy and surgical time (P > 0.05). (Table 1)
Table 1
Clinical characteristics of CRC patients
Characteristics | | COPD cohort (n = 58) | Non-COPD cohort (n = 671) | P |
Gender | Male | 48(82.76%) | 376(56.04%) | < 0.0001 |
Age(year) | | 73.5(69.25, 78) | 62(54, 70) | < 0.0001 |
BMI | | 19.92(18.54, 21.86) | 22.67(20.57, 24.67) | < 0.0001 |
Diabetes | | 6(10.34%) | 86(12.82%) | 0.587 |
Hypertension | | 19(32.76%) | 199(29.66%) | 0.621 |
Coronary heart disease | | 16(27.59%) | 59(8.79%) | < 0.0001 |
Smoking history | | 25(43.10%) | 119(17.73%) | < 0.0001 |
Alcohol consumption history | | 5(8.62%) | 38(5.66%) | 0.531 |
Tumor Location | Rectum | 15(25.86%) | 195(29.06%) | 0.606 |
| Colon | 43(74.14%) | 476(70.94%) | |
TNM tumor stage | 0 stage | 3(5.17%) | 19(2.83%) | 0.354 |
| I stage | 9(15.52%) | 92(13.71%) | |
| II stage | 25(43.10%) | 227(33.83%) | |
| III stage | 15(25.86%) | 239(35.62%) | |
| IV stage | 6(10.34%) | 94(14.01%) | |
Histological type | Tubular adenocarcinoma | 54(93.10%) | 608(90.61%) | 0.193 |
| Mucinous adenocarcinoma + Signet-ring cell carcinoma | 1(1.72%) | 45(6.71%) | |
| Intramucosal carcinoma | 3(5.17%) | 18(2.68%) | |
Differentiated degree | High | 1(1.72%) | 21(3.13%) | 0.858 |
| Moderate | 48(82.76%) | 528(78.69%) | |
| Poor | 5(8.62%) | 61(9.09%) | |
| Other | 4(6.90%) | 61(9.09%) | |
Maximum tumor diameter | <5cm | 29(50.00%) | 405(60.36%) | 0.123 |
| ≥ 5cm | 29(50.00%) | 266(40.64%) | |
Neoadjuvant therapy | | 6(10.34%) | 87(12.97%) | 0.566 |
Surgical time | 2014–2019 | 29(50.00%) | 295(43.96%) | 0.375 |
| 2020–2022 | 29(50.00%) | 376(56.04%) | |
To accurately assess the difference in postoperative recovery and prognosis of CRC patients with and without COPD, we performed a propensity score for gender, age, BMI, TNM stage, and tumor location in both COPD and non-COPD cohorts to reduce the bias effect of confounders on the analysis result. After matching the two cohorts in a 1:1 ratio, 56 CRC patients with COPD and 56 CRC patients without COPD were included, including 46 men and 10 women in the COPD cohort and the same sex ratio in the paired control cohort. The median age was 73 (68.75, 78) years in the COPD cohort and 76 (66.75, 83.25) years in the non-COPD cohort. No obvious differences were observed between the paired COPD and non-COPD cohorts in the gender, age, BMI, concomitant diabetes mellitus, hypertension, coronary heart disease, bronchial asthma, smoking or drinking history, preoperative level of GI tumor indicators (CEA, AFR, CA199, CA724), tumor location, TNM stage, T stage, N stage, M stage, degree of differentiation, histological type, maximum tumor diameter, neoadjuvant therapy, operation year, operation method, operation timing, ASA score and postoperative adjuvant therapy (P > 0.05). (Table 2)
Table 2
Clinical characteristics of COPD cohort and non-COPD cohort after propensity matching analysis
Characteristics | | COPD cohort (n = 56) | Non-COPD cohort (n = 56) | P |
Gender | Male | 46(82.14%) | 46(82.14%) | 1 |
Age (year) | | 73(68.75, 78) | 76(66.75, 83.25) | 0.331 |
BMI | | 20(18.85, 21.94) | 20.83(19.02, 22.05) | 0.386 |
Diabetes | | 5(8.93%) | 7(12.50%) | 0.541 |
Hypertension | | 18(32.14%) | 21(37.50%) | 0.552 |
Coronary heart disease | | 15(26.79%) | 15(26.79%) | 1 |
Bronchial asthma | | 6(10.71%) | 1(1.79%) | 0.118 |
Smoking history | | 24(42.86%) | 24(42.86%) | 1 |
Alcohol consumption history | | 5(8.93%) | 9(16.07%) | 0.253 |
Tumor location | Rectum | 14(25.00%) | 14(25.00%) | 1 |
| Colon | 42(75.00%) | 42(75.00%) | |
TNM stage | 0 stage | 2(3.57%) | 3(5.36%) | 0.777 |
| I stage | 9(16.07%) | 6(10.71%) | |
| II stage | 25(44.64%) | 26(46.43%) | |
| III stage | 14(25.00%) | 15(26.79%) | |
| IV stage | 6(10.71%) | 6(10.71%) | |
T stage | 0 | 2(3.57%) | 3(5.36%) | 0.611 |
| 1 | 4(7.14%) | 1(1.79%) | |
| 2 | 3(5.36%) | 5(8.93%) | |
| 3 | 33(58.93%) | 30(53.57%) | |
| 4 | 14(25.00%) | 17(30.36%) | |
N stage | 0 | 36(64.29%) | 33(58.93%) | 0.520 |
| 1 | 15(26.79%) | 14(25.00%) | |
| 2 | 5(8.93%) | 9(16.07%) | |
M stage | 0 | 50(89.29%) | 50(89.29%) | 1 |
| 1 | 6(10.71%) | 6(10.71%) | |
Histological type | Tubular adenocarcinoma | 53(94.64%) | 48(85.71%) | 0.211 |
| Mucinous adenocarcinoma + Signet-ring cell carcinoma | 1(1.79%) | 5(8.93%) | |
| Intramucosal carcinoma | 2(3.57%) | 3(5.36%) | |
Differentiated degree | High | 1(1.79%) | 1(1.79%) | 0.464 |
| Moderate | 47(83.93%) | 43(76.79%) | |
| Poor | 5(8.93%) | 4(7.14%) | |
| Other | 3(5.36%) | 8(14.29%) | |
Maximum tumor diameter | <5cm | 27(48.21%) | 33(58.93%) | 0.256 |
| ≥ 5cm | 29(51.79%) | 23(41.07%) | |
Surgical time | 2014–2019 | 28(50.00%) | 31(55.36%) | 1.000 |
| 2020–2022 | 28(50.00%) | 25(44.64%) | |
Surgical approach | Laparoscopic | 45(80.36%) | 42(75.00%) | 0.496 |
| Open | 11(19.64%) | 14(25.00%) | |
Intervention type | Selevtive | 55(98.21%) | 53(94.64%) | 0.611 |
| Acute | 1(1.79%) | 3(5.36%) | |
ASA grade | I grade | 0(0.00%) | 1(1.79%) | 0.120 |
| II grade | 26(46.43%) | 27(48.21%) | |
| III grade | 23(41.07%) | 27(48.21%) | |
| IV grade | 7(12.50%) | 1(1.79%) | |
Neoadjuvant therapy | | 6(10.71%) | 6(10.71%) | 1 |
Postoperative adjuvant therapy | 18(32.14%) | 16(28.57%) | 0.681 | |
Pre-operative CEA level(ng/ml) | | 5.7(3.31, 11.71) | 4.69(2.86, 10.65) | 0.468 |
Pre-operative AFP level(ng/ml) | | 2.88(2.02, 3.71) | 2.76(2.14, 3.63) | 0.895 |
Pre-operative CA199 level(U/ml) | | 16.92(11.12, 30.65) | 18.22(8.63, 26.25) | 0.527 |
Pre-operative CA724 level(U/ml) | | 1.79(1.15, 3.13) | 1.85(1.17, 5.47) | 0.480 |
3.2 Comparison of intraoperative and postoperative outcomes between the two cohorts
After propensity matching, in COPD cohort, the operative duration was 4.17 (2.80, 5.18) hours, the anesthetic duration was 5.67 (4.31, 6.93) hours, the total length of hospital stay was 22 (17, 28.75) days, the duration of postoperative hospital stay was 11.5 (9, 16.75) days and the amount of intraoperative blood loss was 50 (30, 112.5) ml. In the non-COPD cohort, the operative duration was 4.42 (3.15, 5.36) hours, the anesthetic duration was 5.50 (4.58, 6.71) hours, the total length of hospital stay was 19 (16.75, 25) days, the duration of postoperative hospital stay was 10 (8, 14.5) days, and the intraoperative blood loss was 50 (45, 100) ml. There were no significant differences in these variables between the two cohorts (all P > 0.05). These results showed that COPD did not cause prolonged operation or anesthesia time, increased intraoperative blood loss, extended the length of hospital stay or postoperative hospital stay in CRC patients with COPD. (Table 3)
Table 3
Comparison of surgical and postoperative outcomes between the two cohorts
Characteristics | COPD cohort (n = 56) | Non-COPD cohort (n = 56) | P |
Operative duration(h) | 4.17 (2.80, 5.18) | 4.42(3.15, 5.36) | 0.256 |
Anaesthetics duration(h) | 5.67 (4.31, 6.93) | 5.50(4.58, 6.71) | 0.993 |
Total length of hospital stay(d) | 22(17, 28.75) | 19(16.75, 25) | 0.195 |
Duration of postoperative hospital stay(d) | 11.5(9, 16.75) | 10(8, 14.5) | 0.145 |
Intraoperative blood loss(ml) | 50(30, 112.5) | 50(45, 100) | 0.960 |
The time of postoperative abdominal drainage tube removal(d) | 8(6, 10) | 7(4.75, 9) | 0.210 |
The time of postoperative urinary catheter removal(d) | 5.5(4, 11) | 4(3, 7) | 0.039 |
The time of first postoperative defecation(d) | 6.5(4.75, 9) | 4(2.75, 6.25) | < 0.001 |
ICU admission stay (d) | | | < 0.001 |
0 | 17(30.36%) | 43(76.79%) | |
1 ~ 2 | 16(28.57%) | 7(12.50%) | |
3 ~ 5 | 12(21.43%) | 5(8.93%) | |
6 or more | 11(19.64%) | 1(1.79%) | |
Antibiotic use | | | 0.035 |
First and second generation cephalosporins | 27(48.2%) | 38(67.9%) | |
Carbapenems, third generation cephalosporins | 29(51.8%) | 18(32.1%) | |
Postoperative transfusion | 11(19.64%) | 9(16.07%) | 0.622 |
glucocorticoid | 5(8.93%) | 0(0.00%) | 0.067 |
secondary operation | 2(3.57%) | 0(0.00%) | 0.476 |
Non-invasive assisted ventilation | 11(19.64%) | 0(0.00%) | < 0.001 |
Invasive assisted ventilation | 17(30.36%) | 4(7.14%) | 0.002 |
Fiberoptic bronchoscopy | 6(10.71%) | 0(0.00%) | 0.036 |
Hemodialysis | 1(1.79%) | 0(0.00%) | 1.000 |
Intravenous pressor/vasoconstrictor | 6(10.71%) | 2(3.57%) | 0.271 |
Wound Infection | 2(3.57%) | 0(0.00%) | 0.476 |
Incision dehiscence | 2(3.57%) | 2(3.57%) | 1.000 |
Pneumonia | 17(30.36%) | 5(8.93%) | 0.004 |
Abdominal infection | 8(14.29%) | 3(5.36%) | 0.112 |
Catheter-related infection | 1(1.79%) | 0(0.00%) | 1 |
Urinary tract infection | 0(0.00%) | 1(1.79%) | 1.000 |
Septic shock | 6(10.71%) | 0(0.00%) | 0.036 |
Hypovolemic shock | 2(3.57%) | 0(0.00%) | 0.476 |
Respiratory failure | 4(7.14%) | 1(1.79%) | 0.360 |
Heart failure | 1(1.79%) | 0(0.00%) | 1 |
intestinal obstruction | 4(7.14%) | 0(0.00%) | 0.127 |
Anastomotic leakage | 0(0.00%) | 2(3.57%) | 0.476 |
Anastomotic hemorrhage | 3(5.36%) | 1(1.79%) | 0.611 |
Acute kidney injury | 2(3.57%) | 0(0.00%) | 0.476 |
After CRC resection, in COPD cohort, the time of abdominal drainage tube removal was 8 (6,10) days, the time of urinary catheter removal was 5.5 (4,11) days and the time of first defecation was 6.5 (4.75,9) days. In the paired non-COPD cohort, the time of abdominal drainage tube removal was 7 (4.75, 9) days, the time of urinary catheter removal was 4 (3, 7) days, and the time of first defecation was 4 (2.75, 6.25) days. The results demonstrated that the time of postoperative urinary catheter removal and the time of first defecation after surgery in CRC patients with COPD were longer than that in CRC patients without COPD (all P < 0.05). However, there was no statistical differences in the time of postoperative abdominal drainage tube removal between the two cohorts. (Table 3)
Furthermore, ICU admission rate after surgery was 69.64% in COPD cohort, whereas ICU admission rate was only 23.21% in non-COPD cohort, with a statistically significant difference (P < 0.001). In the COPD cohort, 11 patients (19.64%) stayed in ICU for 6 days or more, 12 patients (21.43%) stayed in ICU for 3 to 5 days, and 16 patients (28.57%) stayed in ICU for 1 to 2 days. In the non-COPD cohort, only 1 (1.79%) patient stayed in ICU for more than 6 days, 5 patients stayed in ICU for 3 to 5 days, and 7 patients stayed in ICU for 1 to 2 days. Moreover, non-invasive assisted ventilation rates were 19.64% and 0%, invasive assisted ventilation rates were 30.36% and 7.14%, and fiberoptic bronchoscopy rates were 10.71% and 0% in the COPD cohort and in the non-COPD cohort, respectiely. However, no statistical differences were observed in postoperative blood transfusion rate, glucocorticoid utilization rate, secondary operation rate, hemodialysis rate, and the utilization rate of intravenous pressor or vasoconstrictor between 2 cohorts (P > 0.05). In the use of antibiotics, we found that the usage of postoperative high-grade antibiotics such as carbapenems and third-generation cephalosporin in COPD cohort was more more common than that in non-COPD cohort (Table 3). The results indicate that CRC patients with COPD need more intensive care, more assisted ventilation and high-grade antibiotics after surgery.
We compared the postoperative complications between COPD cohort and non-COPD cohort and found that a greater proportion of patients developed pneumonia and septic shock after CRC resection in COPD cohort than non-COPD cohort. However, there were no statistically significant differences in the incidence of wound infection, incision dehiscence, abdominal infection, catheter associated infection, urinary tract infection, hypovolemic shock, respiratory failure, heart failure, intestinal obstruction, anastomotic leakage, anastomotic hemorrhage, and acute kidney injury after surgery between the two cohorts (all P > 0.05). (Table 3)
3.3 Postoperative overall survival was worse in COPD cohort
To accurately assess the impact of COPD on the prognosis of CRC patients after surgery, we conducted survival correlation analysis in the 112 CRC patients. Our results showed that postoperative overall survival of patients was worse in COPD cohort than that in non-COPD cohort (P = 0.043) (Fig. 1A). However, we found no obvious differences in DFS (Fig. 1B) and CSS (Fig. 1C) between COPD and non-COPD cohorts.
A) Overall postoperative survival curve of patients in COPD group and non-COPD group;B) Postoperative disease-free survival curve of patients in COPD group and non-COPD group; C) Postoperative cancer-specific survival curves of patients in COPD group and non-COPD group;
To explore the effect of neoplastic and other prognostic outcomes on the prognosis of CRC patients, we defined other death causes as non-neoplastic causes, such as severe pneumonia, other comorbidity. A competition risk model was established to determine whether there was a difference in cancer-specific death risk of patients between COPD and non-COPD cohorts. The results confirmed that no distinct difference in the risk of cancer-related death between COPD and non-COPD cohorts after controlling for competing risks of non-neoplastic causes (Fig. 2).
CIF: Cumulative incidence function. Red : Cumulative incidence (curve) of non-tumor causes of death in COPD group; Blue: Cumulative incidence (curve) of non-tumor causes of death in non-COPD group; Black : Cumulative incidence (curve) of death from CRC in COPD group; Green : Cumulative incidence (curve) of death from CRC in non-COPD group.
3.4 COPD was an independent factor for overall survival of CRC patients
Univariate COX regression analysis showed that COPD, TNM stages, N stage, M stage, nerve invasion, surgical approach, ASA grade, preoperative CEA level, preoperative CA199 level, preoperative CA724 level and postoperative adjuvant therapy were correlated with overall survival of CRC patients (P < 0.05). However, age, sex, BMI, maximum tumor diameter, T-stage, tumor location, differentiated degree, histological type, surgical time, surgical year, diabetes, hypertension, coronary heart disease, bronchial asthma, smoking and alcohol consumption history, and neoadjuvant therapy were not associated with overall survival (P > 0.05). (Table 4)
Table 4
Univariate analysis of overall survival in patients with CRC
Characteristics | HR | 95%CI | P |
COPD | | | |
No | 1 | | |
Yes | 1.832 | 1.012–3.311 | 0.046 |
Age(year) | 1.021 | 0.990–1.054 | 0.189 |
Gender | | | |
Male | 1 | | |
Female | 0.814 | 0.377–1.757 | 0.6 |
BMI(kg/m2) | 0.896 | 0.794–1.011 | 0.075 |
Maximum tumor diameter(cm) | | | |
<5 | 1 | | |
≥ 5 | 1.459 | 0.806–2.643 | 0.212 |
TNM tumor stage | | | |
IV | 1 | | |
III | 0.356 | 0.165–0.767 | 0.008 |
II | 0.107 | 0.046–0.249 | < 0.001 |
0-I | 0.101 | 0.031–0.329 | < 0.001 |
T stage | | | |
3–4 | 1 | | |
1–2 | 0.416 | 0.101–1.724 | 0.227 |
0 | 0.633 | 0.087–4.624 | 0.652 |
N stage | | | |
0 | 1 | | |
1 | 2.965 | 1.506–5.838 | 0.002 |
2 | 3.996 | 1.852–8.620 | < 0.001 |
M stage | | | |
0 | 1 | | |
1 | 5.771 | 2.854–11.667 | < 0.001 |
Tumor location | | | |
Rectum | 1 | | |
Colon | 0.832 | 0.411–1.686 | 0.61 |
Differentiated degree | | | |
Moderate + High | 1 | | |
Poor | 1.217 | 0.465–3.185 | 0.689 |
other | 1.215 | 0.509-2.900 | 0.661 |
Histological type | | | |
Tubular adenocarcinoma | 1 | | |
Mucinous adenocarcinoma + Signet-ring cell carcinoma | 1.255 | 0.387–4.073 | 0.705 |
Intramucosal carcinoma | 1.508 | 0.361–6.296 | 0.573 |
Nerve invasion | | | |
No | 1 | | |
Yes | 2.241 | 1.151–4.361 | 0.018 |
Vascular infiltration | | | |
No | 1 | | |
Yes | 1.74 | 0.953–3.175 | 0.071 |
Surgical approach | | | |
Laparoscopic | 1 | | |
Open | 2.049 | 1.097–3.827 | 0.024 |
Intervention type | | | |
Selevtive | 1 | | |
Acute | 1.649 | 0.397–6.854 | 0.491 |
Surgical time | | | |
2014–2019 | 1 | | |
2020–2022 | 1.735 | 0.819–3.676 | 0.151 |
ASA grade | | | |
I-II | 1 | | |
III-IV | 2.031 | 1.109–3.722 | 0.022 |
Diabetes | | | |
No | 1 | | |
Yes | 0.561 | 0.199–1.587 | 0.276 |
Hypertension | | | |
No | 1 | | |
Yes | 1.041 | 0.557–1.943 | 0.9 |
Coronary heart disease | | | |
No | 1 | | |
Yes | 1.615 | 0.852–3.060 | 0.142 |
Bronchial asthma | | | |
No | 1 | | |
Yes | 0.761 | 0.235–2.470 | 0.65 |
Smoking history | | | |
No | 1 | | |
Yes | 0.989 | 0.549–1.783 | 0.971 |
Alcohol consumption history | | | |
No | 1 | | |
Yes | 0.598 | 0.234–1.528 | 0.283 |
Pre-operative CEA level(ng/ml) | | | |
<5 | 1 | | |
≥ 5 | 2.074 | 1.098–3.915 | 0.025 |
Pre-operative CA199 level(U/ml) | | | |
<39 | 1 | | |
≥ 39 | 2.556 | 1.300-5.026 | 0.007 |
Pre-operative CA724 level(U/ml) | | | |
<6.9 | 1 | | |
≥ 6.9 | 2.237 | 1.120–4.469 | 0.023 |
Neoadjuvant therapy | | | |
No | 1 | | |
Yes | 0.528 | 0.128–2.185 | 0.378 |
Postoperative adjuvant therapy | | | |
No | 1 | | |
Yes | 0.406 | 0.195–0.846 | 0.016 |
Since Pearson's correlation test suggested obvious collinearity between TNM stages and N stage (correlation coefficient = 0.704, P < 0.001) and M stage (correlation coefficient = 0.669, P < 0.001), only TNM stages was included in the multivariate analysis.
Multivariate COX regression analysis showed that COPD (HR = 3.232, 95%CI: 1.537–6.795, P = 0.002) and TNM stages (Stage III vs. stage IV, HR = 0.261, 95%CI: 0.099–0.687, P = 0.007; Stage II vs. Stage IV, HR = 0.129, 95%CI: 0.046–0.360, P < 0.001; Stage 0-I vs. stage IV, HR = 0.104, 95%CI: 0.025–0.430, P = 0.002) were independent prognostic factors in CRC patients (Fig. 3).