Associated factors for postoperative morbidity and mortality following surgery of obstructive colon tumors

Background

Intestinal obstruction is the most frequent emergency during the course of colon cancer, occurring in 75% to 90% of cases.^1–4 It worsens oncological prognosis, negatively impacts overall survival, and can be life-threatening. In fact, mortality doubles from 3.4% to 8.5% in the case of non-elective surgery,⁵ implying a hazard ratio of 2.251.⁶ This applies irrespective of the T or N stage of the disease, and several studies have shown that for the same stage of the disease, occlusion worsens prognosis and reduces overall and disease-free survival.^7,8

Severe postoperative complications are more frequent in colon cancer surgery with occlusion, and this may be explained by the patient’s unfavorable preoperative conditions and by a doubling of the rate of specific complications, such as anastomotic fistula in the case of occlusion.^1,9 Indeed, the occurrence of postoperative complications is responsible for the decline in overall survival and recurrence-free survival, as shown by a meta-analysis of randomized phase III trials involving 5,530 patients.¹⁰

Prevention and management of postoperative complications are essential to improve the outcome of occluded colon cancer, as the occurrence of postoperative complications increases mortality.^11,12

Traditional predictive features have limitations in identifying at-risk patients sufficiently early to effectively intervene.

Our main objective was to determine factors associatedwith postoperative morbidity and mortality. By targeting these factors, we can reduce postoperative morbidity with the aim of reducing hospital stays, minimizing costs, allowing early initiation of adjuvant chemotherapy, and subsequently reducing mortality. Furthermore, we aimed to shed light on the current treatment landscape with different procedures and provide a comparative study of different surgical methods. In addition, we studied the restoration of bowel continuity after stoma creation to assess the encountered difficulties and challenges impeding this second procedure.

Methods

Results

Sixty-two cases were included in the study. The age distribution of patients in our study ranged from 28 to 100 years, with a median age of 62 years. Almost half of our patients were with no significant medical history (n=29, 45.8%). The most reported co-morbidity was of cardiovascular disease (n=22, 35.5%) (Table 1). The symptom duration was approximately 4,6 days on average. The obstruction’s tetrad was complete in more than half of cases (n=32; 51.6%). Seven of our patients had already diagnosed with a colon cancer (11.3%). The initial diagnosis was made 12,5 weeks after the intestinal obstruction. Upon examination, four patients had dehydration traits. Hemodynamic instability was observed in 3 cases. And three had clinical T4 tumors (Table 2).

Most of patients had elevated inflammatory markers (n=50, 82%). Anemia was noted in 34 patients (58.6%), while ionic disorders were observed in 26 patients (41.9%) (Table 3).

The Left localization was predominant (n=51, 82.3%). On abdominal computed tomography (CT-scan, synchronous tumor) was noted in three patients. Four patients had features of colonic ischemia (pneumoperitoneum, aeroporty, bowel wall hypoenhancement, and submucosal edema). The mean diameter of digestive tract was at 8.3 cm (3.7 – 13). An associated abscess was displaced in five cases. Metastases were observed in 11 patients (18%), with the majority in liver (n=9; 14.5%), followed by peritoneum (n=4; 6.6%) (Table 4). Only two patients had polymetastatic tumors.

Based on these findings, patients were operated on accordingly, taking into consideration the presence or absence of clinical, biological, and radiological signs, the tumor site and its number, and the patient’s general well-being. diverting ostomy was the most proposed procedure (n=22; 35.5%), followed by segmental resection (n=15; 24.2%). Subtotal colectomy was the least done procedure (n=2; 3.2%), reserved for synchronous tumor, coecal, or distant ischemic features. Other procedures included Hartmann’s procedure (n=5, 8.1%), segmental resection followed by double ostomy (n=11, 17.7%), and subtotal colectomy followed by ostomy (n=7, 11.3%). There were no demographic differences among the different surgical procedures (Table 1). However, patients who had a diverting ostomy more often had a complete tetrad of intestinal obstruction and had more complications from obstruction (Table 2). Whenever the obstruction ceased spontaneously, the patients were scheduled forsegmental resection, as in elective conditions. As mentioned above, patients who underwentsubtotal colectomy had more intestinal complications; therefore, forthey had the highest incidence of colonic ischemia. If the diameter of the intestinal lumen was large, anastomosis was rejected. STC was also proposed preoperatively in three patients with synchronous tumors (4.81%). Laparotomy was the procedure of choice; laparoscopy was the surgical route in two patients in the second stage of surgery after diverting the ostomy. Intraoperative exploration revealed synchronous tumor in six cases (10.3%). Three synchronous tumors were missed in the radiological assessment. Digestive perforation was observed in nine patients (15.3%). Peritoneal metastases were underdiagnosed on CT, as 15% of patients had carcinomatosis on surgical exploration. Ischemic features were seen in eight patients (13.6%), with pre-perforative coecum being the most reported sign (Table 5).

Postoperative complications occurred in 37.5%, with the majority being surgical; 14 patients (22.6%) developed surgical postoperative complications. Specific surgical complications were the common (n=11; 18%), and they mostly occurring in diverting ostomy patients.

Seven deaths were numbered (11%) (Table 6). Mortality was attributed to 3 hemodynamic instability, acute renal injury with pulmonary edema, anastomotic leakage, pulmonary infection, and peritonitis secondary to tumor perforation after DS.

In the bivariate analysis, many factors were associated with postoperative mortality, such as neurological morbidity (p = 0.031), history of colorectal cancer (p = 0.005), complete obstructive syndrome (p = 0.043), hemodynamic instability (p=0.011), anemia (p=0.042), elevated CRP (p=0.006), peritoneal metastasis on CT (p=0.045), stercoral peritoneal effusion (p=0.014), per operative catecholamine usage (p=0.002), and postoperative catecholamine usage (p=0.004) (Table 7). On the Linear regression analysis, neurological comorbidity (p=0.027, OR=55%), hemodynamic instability (p≤0.000, OR=63%), peritoneal carcinosis (p<0.000, OR=60%), and catecholamine usage preoperatively (p=0.007, OR=57%) were associated with postoperative mortality. In contrast, anemia (p=0.03), elevated CA19-9 (p=0.016), and peritoneal metastasis on CT (p=0.02) were associated with postoperative morbidity (Table 8). None of the identified determinants of postoperative morbidity were retained in the linear regression.

Concerning specific surgical complications, the only statistically significant factors in the bivariate analysis were delayed consulting time (5 vs. 3 days, p=0.037), elevated CA 19-9 levels (p=0.021), peritoneal carcinomatosis on CT scan (p=0.02) or during surgical exploration (p=0.013), and longer anesthetic time (315 vs. 210 min, p=0.008) (Table 9). However, these factors were not confirmed in multivariate analysis (Table 10).

The second stage surgery was performed in only 21 patients from 42 patients (50%), Hartmann’s reversal in 1 patient only from 5 patients (20%), stoma closure following SR-DS or STC-DS in 5 patients from 19 patients (26.3%), and resection following diverting ostomy in 15 cases from 18 patients. Among patients not undergo the second surgical procedure (21, 50%), 11 patients were lost to follow-up after the first surgical procedure, seven patients were referred for palliative care due to polymetastatic disease (5 patients developed diffuse peritoneal carcinosis on postoperative CT, and 2 patients had multiple metastases from the preoperative period), three patients died following the first surgical procedure.

There was no statistical difference between the different surgical procedures when analyzing the postoperative morbidity or mortality in second-stage surgery (Table 11). Table 12 showcases cumulative postoperative morbidity and mortality after first and second surgeries.

Discussion

Intestinal obstructions are frequent and dreadful. Colonic tumors are the leading cause of death in developed countries. Its treatment is controversial because of the plethora of available procedures. An adequate assessment of the patient’s status and the results of each procedure will help in the choice of therapeutic measures. Our study provides a detailed treatment path for patients with obstructive colon tumors, with emphasis on the current treatment landscape. Postoperative complications occurred in 37.5% with the majority being surgical (n=14, 22.6%). Seven deaths were numbered (11%). Neurological comorbidity (p=0.027, OR=55%), hemodynamic instability (p≤0.000, OR=63%), peritoneal carcinosis (p<0.000, OR=60%), and catecholamine usage peroperatively (p=0.007, OR=57%) were associated with postoperative mortality.

According to the World Cancer Research Fund, colorectal cancer (CRC) is the third most common cancer worldwide, regardless of sex, with 1.9 million new cases in 2020.¹⁵ In women, it is the second most common cancer, while in men, it is the third most common.¹⁵ On a smaller scale, the American Cancer Society estimated 106970 new cases of CRC in the United States.¹⁶ It is also the second leading cause of cancer-related deaths, with an estimated 52550 deaths by 2023.¹⁶

In Tunisia, a recent publication by the Global Cancer Observatory highlighted that colorectal cancer ranks 5th among newly diagnosed cancer cases, representing 9.6% of 988 new cases in 2020, and is the 5th leading cause of death, representing 4.6% of 542 deaths in 2020.¹⁷ A retrospective study, conducted between 1994 and 2009, and including patients registered in Tunisia’s northern cancer registry, with the aim of predicting regional colorectal cancer incidence in 2024, showed that there is an upward trend in cumulative risk estimated at + 3.9% [2.8% - 5.1%].¹⁸ Without screening, the age-standardized rate will reach 39.3/100,000 for men and 22.9/100,000 for women in 2024, that is double the figure announced at the end of the data collection.¹⁸

In literature reviews, bowel obstruction is the revealing manifestation of 8 to 40% of colonic tumors.^19–21 Moreover, occlusive syndrome is the most frequent emergency during the course of colonic cancers, accounting for 75% to 85% of cases.^1,2,22,23

The postoperative morbidity and mortality of colonic tumors in occlusion are higher than those in elective surgery, regardless of the type of surgery performed.²⁴ Mortality is observed in 9% of patients operated surgery for a colonic tumor obstruction, and the occurrence of postoperative complications affects half of all patients operated on, as shown by the French Association of Surgery.²⁵ This rate is threefold that of elective surgery for colonic tumors.²⁶

Intestinal obstruction impairs oncological prognosis and is responsible for a decline in 5-year-cancer-specific-survival from 70.9% to 52.9%.²⁷ For this reason, the National Cancer Control Network (2014) and the European Society for Medical Oncology (2012) consider occlusion, complicating colonic tumors, to be a poor prognostic factor in the development of adjuvant treatment modalities.^28,29

Therefore, a better understanding of the determinants of postoperative complications can improve therapeutic management and prognosis.

A total of 62 patients were included in the study, representing a heterogeneous series, including right and left localization, of colon cancer cases presenting with intestinal obstruction. Notably, we observed a wide range of surgical procedures in our study population, reflecting the diversity of patients with colon cancer requiring surgical intervention. Their ages ranged from 28 to 100 years old. Among these, 72.6% had left sided colon cancer and the remaining 27.4% were diagnosed with right sided colon cancer. This result is in accordance with those of other authors,^30–33 where distal tumors exhibitedobstructive behavior. This distribution underscores the varied sieges of obstructive colon malignancies encountered in clinical practice and emphasizes the importance of considering different therapeutic plans. Among our patients, 13.6% had ischemic features when operated upon. This high proportion was similarly observed (12%) by a Japanese team.³¹ Guo M Gao et al. noted a high incidence of peritonitis (19.7%) and colonic perforation (9%) in patients exceeding 70 years old.³⁴ Our study investigated patients admitted for colonic tumoral obstruction with an emphasis on their surgical outcometo define the factors associatedwith morbidity and mortality. We noted that postoperative complications occurred in a third of our patients (37.5%) and were essentially of surgical nature (22.6% of our patients). This incidence is similar to Shunji Morita et al. results (34%)³¹ but lower than what was observed in a large cohort of French Association of Surgery, which scored 51% of global morbidity.³⁵ In older patients, complications occur more frequently as highlighted by Guo M Gao et al. (62.2%).³⁴ With regard to the nature of postoperative complications, our results were lower than those observed by Bouare et al.³⁶ The highest proportion of surgical complications occurred in a large cohort study by the French Association of Surgery (90%).³⁵ Whereas ower rates (11,9%) were reported by Ker Kan Tan et al.³⁰

In our series, postoperative complications led to 11% of mortality, comparable to what was observed by Ker Kan Tan et al. (11.9%).³⁰ A lower mortality rate was reportedin a Dutch series (6.9%),³⁷ and higher rates were observed in senile patients, where 24.5% of patients died, as reported by Ming Gao Guo.³⁴ The selection of these at-risk patients to develop postoperative complications would lead to better perioperative management.

Few studies have focused on determining factors associatedwith postoperative morbidity. Therefore, our study contributes to enriching the present data. In fact, diabetes mellitus is another factor to be considered, as it increases the occurrence of postoperative complications (p=0.004) and redo surgery (p=0.01) and is associated with longer stay in intensive care units (p=0.02).³⁸ Our univariate analysis linked a longer anesthetic time to the occurrence of specific surgical complications (315 vs. 210 min, p=0.008). Our result is consistent with a retrospective Romanian series involving 449 patients, published in 2022, and concluded that a threshold value for operating time exceeding 2 h 30 minwas associated with additional morbidity (p < 0.05) and an area under the curve of 0.83.³⁹ However, mastering the anesthesia time did not reduce postoperative mortality. A large multicenter French series 1983 patients retrospectively showed that operating time does not influence postoperative mortality, even if the operating time lasts more than 150 min (p=0.46).⁴⁰

There is a dearth of data regarding the risk factors for postoperative mortality in patients with obstructive colon tumors. Our study adds to the existing literature on postoperative mortality and underscores the importance of individualized surgical planning strategies based on accurate preoperative and peroperative assessments of patients. In fact, when considering that neurological comorbidity (p=0.027, OR=55%), hemodynamic instability (p≤0.000, OR=63%), peritoneal carcinosis (p<0.000, OR=60%), and catecholamine usage preoperatively (p=0.007, OR=57%) were associated with postoperative mortality, surgeons could modify their approach accordingly to minimize surgery’s impact and ramp up postoperative surveillance to detect subtle complications. Our results are consistent with those ofot previous studies. Patients with left obstructive colon tumorswho had neurological comorbidities were found to be prone to postoperative mortality (p=0.049; OR=1.89).⁴¹ The same study linked pulmonary comorbidities to postoperative mortality (p=0.04; OR=2).⁴¹ Coronary insufficiency (p=0.001; OR=5.83) and chronic renal disease (p<0.001; OR=23.54) negatively influenced the postoperative course in a prospective Egyptian study.⁴² To our knowledge, our results showing peritoneal carcinosis as an associated factor for postoperative mortality is the first to highlight such a finding. Mroua et al. found this interrelation only in their univariate analysis.⁴³ However, in 2018, the French Surgical Association did not confirm this statistical association in their report of 1220 patients operated surgery for left colonic tumors in occlusion (p=0.3).⁴¹ However, local studies do not share our findings. The Tunisian Surgical Association, analyzing 635 patients who underwent surgery for a left colonic tumor in occlusion at 21 hospital centers between 2003 and 2013, showed that intraoperative presence of peritoneal carcinosis did not increase mortality or specific surgical morbidity (p=1 and p=0.11, respectively).⁴⁴

Limitations: It is noteworthy that our study had some limitations, including its retrospective nature and small sample size. While our broad inclusion allowed for a comprehensive assessment of the course of patients with obstructive colonic tumors, it also contained heterogeneity in the study population. As different surgical procedures may exhibit variations in the postoperative course, the generalizability of our findings to patients with colon cancer with intestinal obstruction may be influenced by this heterogeneity.

Conclusion

Colonic cancers are among the most deadliest neoplasms. This can be explained by dreadful complications. Intestinal obstruction was the most frequently reported complication. This presents a therapeutic dilemma because of the multiplicity of procedures and frail patients. Treatment should be tailored to the patient’s medical history and repercussions of the affection to avoid the occurrence of postoperative complications. We determined that neurological comorbidity, hemodynamic instability, peritoneal carcinosis, and catecholamine use preoperativelywere associated with postoperative mortality. Although factors associated with postoperative morbidity were not determined in our study, we encourage further studies to resolve this question. Considering these factors, surgeons can identify high-risk patients, strengthen preoperative preparations, adjust surgical strategies and techniques, detect potential complications early, and intervene in a timely manner.

Declarations

Clinical trial number

Not applicable.