Abstract
Objectives:
The prognosis of patients with colorectal diverticular perforations requiring surgery is poor, and the efficacy of polymyxin B-immobilized fiber column direct hemoperfusion (PMX-DHP) in these patients has not yet been established. In the present study, we evaluated the prognostic factors and the efficacy of PMX-DHP in surgically treated patients with perforated colorectal diverticula.
Methods:
Of the 116 patients with colorectal perforations who underwent emergency surgery at our hospital between April 2018 and May 2023, we retrospectively reviewed 46 patients with perforated colorectal diverticula. Preoperative, surgical, and postoperative factors were compared between the survival and mortality groups. Subgroup analysis was performed to evaluate the efficacy of PMX-DHP in severe cases.
Results:
The postoperative mortality group included 7 patients (15.2%), and PMX-DHP was performed in 14 (30.4%). PMX-DHP was performed significantly more often in the mortality group (P<0.01). In multivariate analysis, antithrombotic drug administration (hazard ratio, 16.600; 95% confidence interval, 1.32-209; P<0.05) and higher lactate levels (≥ 3.0 mmol/L) (hazard ratio, 42.300; 95% confidence interval, 2.69-667, P<0.01) were independent risk factors for postoperative mortality. PMX-DHP was performed in severe cases (patients with higher lactate levels, APACHE II scores, ventilator management frequencies, and noradrenalin use). PMX-DHP was not effective in improving prognosis in severe cases.
Conclusions:
Administration of antithrombotic drugs and higher lactate levels (≥ 3.0 mmol/L) are independent prognostic factors in colorectal diverticular perforations. PMX-DHP may not be effective in patients with severe preoperative conditions.
Keywords: colorectal diverticulum, colorectal perforation, prognostic factor, polymyxin B-immobilized fiber column direct hemoperfusion, antithrombotic drugs, lactate levels
Introduction
Colorectal perforation is one of the most common abdominal emergencies with a high mortality rate and poor prognosis. Colorectal diverticula and cancer are the most common causes of colorectal perforations. In Japan, the prevalence of colorectal diverticula in adults was 23.9% between 2001 and 2010[1]. In the United States, the number of patients with colorectal diverticula is higher, and in one analysis of patients with an average age of 55 years, the prevalence of colorectal diverticula was 60%[2]. Previous reports in Japan showed an increasing trend in the number of patients with colorectal diverticula in recent years[1], and the prevalence of colorectal diverticula has been correlated with age, male sex, smoking history, severe weight increase in adulthood, HbA1c, alcohol intake history, and serum triglyceride levels[3]. Although the prevalence of colorectal diverticula in Japan is lower than in the United States, it is expected to continue to increase in the future as the population ages and diets become more Westernized. In principle, diverticular perforation leading to panperitonitis should be treated surgically. Although there are several studies on postoperative mortality risk factors for overall colorectal perforation[4-10], there are insufficient data regarding the mortality risk factors for colorectal diverticular perforation[11-13]. Polymyxin B-immobilized fiber column direct hemoperfusion (PMX-DHP), a blood purification therapy, is often performed after surgery for colorectal perforation. In recent years, there have been many negative opinions about its efficacy in improving mortality rates; however, there have been some reports on its efficacy under certain conditions[14]. In this study, we retrospectively evaluated the risk factors for postoperative mortality and the efficacy of PMX-DHP therapy in patients who underwent emergency surgery for colorectal diverticular perforations.
Methods
Patient selection
Of the 116 surgeries performed for colorectal perforations at our hospital between April 2018 and May 2023, 46 (39.7%) patients with colorectal perforations due to colonic diverticula were evaluated retrospectively (Figure 1). We excluded patients in whom the perforations were caused by lesions other than colonic diverticula, such as colorectal cancer (n=31, 26.7%), idiopathic causes (n=14, 12.1%), fecal impaction (n=7, 6.0%), postoperative suture failure (n=7, 6.0%), iatrogenic causes (n=6, 5.2%), volvulus (n=4, 3.4%), or traumatic injury (n=1, 0.9%). Idiopathic cases were defined as those in which malignant diseases were ruled out by histopathological examination of the resected specimen; however, the cause of the perforation could not be identified.
Figure 1.
Flow diagram of the study selection process.
We retrospectively investigated 46 patients with perforated colonic diverticula (n=46).
Study design
The clinical factors were compared between the survival and postoperative mortality groups during hospitalization. The following factors were evaluated: age, sex, body mass index (BMI), smoking history, past medical history (cardiac disease, cerebrovascular disease, diabetes mellitus), history of maintenance dialysis or administration of antithrombotic drugs or steroid use, Hinchey classification[15], preoperative Shock-Index (SI: pulse rate/systolic blood pressure), laboratory data (i.e., white blood cell [WBC] count, C-reactive protein [CRP], serum albumin, and serum lactate levels), APACHE II score[16], time from onset to surgery, surgical procedure (resection of the perforation site, one-stage anastomosis, stoma creation), operation time, blood loss, postoperative ventilator management, postoperative PMX-DHP therapy, and postoperative noradrenalin use. There are no strict guidelines for the use of PMX-DHP in our hospital; thus, the decision to introduce PMX-DHP was left to the discretion of the attending physician. Therefore, the characteristics of patients who underwent PMX-DHP therapy were analyzed retrospectively. Based on these findings, a subgroup analysis was conducted to evaluate the efficacy of PMX-DHP in severe cases.
In this study, the term “antithrombotic drugs” encompasses both antiplatelet agents and anticoagulants. The time of onset was based on interviews with the patients. Lactate levels and APACHE II scores were treated as missing data for two patients who did not undergo blood gas testing. This study was a retrospective observational study carried out using the opt-out method, and the study design was approved by the Fukuoka Tokushukai Hospital Clinical Research Ethics Review Committee (approval no. 231004).
Statistical analysis
Statistical analyses were performed using EZR software with the Mann-Whitney U and Fisher's exact tests, and a P-value < 0.05 was considered significant. Logistic regression analysis was used for the multivariate analysis. Factors with insufficient sample sizes were described as not-estimated (N.E.).
Results
Patient characteristics
Patient characteristics are presented in Table 1. The median age was 69 years, and 26 patients (56.5%) were male. Thirteen patients (28.3%) had previous cardiac disease, and seven (15.2%) had previous cerebrovascular disease. Of the patients studied, 13 (28.3%) were receiving antithrombotic drugs for cerebrovascular or cardiac conditions, with none taking these medications for any other diseases. All patients had Hinchey classifications of III (67.4%) or IV (32.6%). Twelve patients (26.1%) had preoperative SIs ≥1.0, the median preoperative lactate level was 1.9 mmol/L, and the median APACHE II score was 14 points. The surgical factors are listed in Table 2. Median time from onset to surgery was 24 hours. The perforation site was the cecum, ascending colon, descending colon, sigmoid colon, and rectum in 2 (4.3%), 1 (2.1%), 1 (2.1%), 38 (82.6%), and 4 (8.7%) patients, respectively, with the sigmoid colon accounting for the majority. Regarding the surgical procedure, 9 patients (19.6%) underwent ileostomy creation only, and 29 (63.0%) underwent Hartmann's procedures. Forty-four patients (95.7%) underwent stoma creation. The median intraoperative blood loss for all patients was 50 ml. Blood loss was significantly higher in patients taking antithrombotic drugs (n=13, median 125 ml, range 5-1150 ml), compared to those not on antithrombotic drugs (n=33, median 25 ml, range 1-550 ml) (P<0.05). There were no deaths during surgery. Postoperative ventilator management was required in 16 patients (34.8%), PMX-DHP was performed in 14 (30.4%), and noradrenaline administration was required in 15 (32.6%). In terms of postoperative complications (Table 3), 19 patients (41.3%) had 23 complications of Clavien-Dindo classification ≥ grade III[17]. Four patients underwent echo-guided puncture drainage for intra-abdominal abscesses, two patients underwent chest drainage for pleural effusions, and three patients underwent surgical tracheostomies for respiratory failure. Grade V (mortality) Clavien-Dindo complications occurred in seven patients (15.2%), including five with sepsis, one with exacerbation of chronic heart failure, and one with a cerebral infarction.
Table 1.
Patients Characteristics.
| Variables | n=46 |
|---|---|
| Age (years) ※ | 69 (24-88) |
| Sex (Male/Female) | 26/20 |
| Body Mass Index※ | 21.7 (14.3-29.9) |
| ASA-PS (I/II/III) | 3/22/21 |
| Smoking history | 17 (37.0%) |
| Cardiac disease | 13 (28.3%) |
| Cerebrovascular disease | 7 (15.2%) |
| Diabetes mellitus | 6 (13.0%) |
| Maintenance dialysis | 3 (6.5%) |
| Antithrombotic drug | 13 (28.3%) |
| Steroid use | 5 (10.9%) |
| Hinchey classification (III/IV) | 31/15 |
| Shock-Index (≥1.0) | 12 (26.1%) |
| WBC (/μL) ※ | 9335 (360-30700) |
| CRP (mg/dL) ※ | 11.2 (0.02-46.9) |
| Albumin (g/dL) ※ | 3.7 (1.6-5.2) |
| Lactate (mmol/L) ※ | 1.9 (0.9-8.1) |
| APACHE II score※ | 14 (6-25) |
※Median (Range); ASA-PS, American Society of Anesthesiologists Physical Status; WBC, White Blood Cell; CRP, C-reactive protein
Table 2.
Surgical Factors.
| Variables | n=46 |
|---|---|
| Time from onset to surgery (hour) | 24 (2-720) |
| Site of perforation | |
| cecum/ascending colon/descending colon/sigmoid colon/rectum | 2/1/1/38/4 |
| Surgical procedure | |
| Ileostomy creation only | 9 |
| Hartmann’s procedure | 29 |
| Laparoscopic drainage | 1 |
| Ileoceculectomy | 1 |
| Left hemicolectomy and ileostomy creation | 1 |
| Sigmoid colon resection and ileostomy creation | 1 |
| Sigmoid colon resection and transverse colostomy with double orifices creation | 1 |
| Sigmoid colostomy with double orifices creation | 1 |
| Rectum perforation site closure and ileostomy creation | 1 |
| Rectum perforation site closure and sigmoid colostomy with double orifices creation | 1 |
| Resection of the perforation site | 32 (70.0%) |
| One-stage anastomosis | 3 (6.5%) |
| Ostomy creation | 44 (95.7%) |
| Operation time (min) ※ | 169 (58-429) |
| Blood loss (ml) ※ | 50 (1-1150) |
| Postoperative ventilator management | 16 (34.8%) |
| Postoperative PMX-DHP therapy | 14 (30.4%) |
| Postoperative noradrenalin use | 15 (32.6%) |
※Median (Range); PMX-DHP, polymyxin B-immobilized fiber column direct hemoperfusion
Table 3.
Postoperative Complications ≥ Grade III (Clavien-Dindo Classification).
| Grade | Complication | n | Treatment |
|---|---|---|---|
| IIIa | Intra-abdominal abscess | 4 | Echo-guided puncture |
| Pleural effusion | 2 | Chest drainage | |
| IIIb | Respiratory failure | 3 | Surgical tracheostomy |
| Enterocutaneous fistula | 1 | Partial resection of small bowel | |
| Ostomy necrosis | 1 | Ostomy recreation | |
| Intra-abdominal abscess | 1 | Drainage under general anesthesia | |
| Colobladder fistula | 1 | Partial resection of the bladder | |
| Failure of suture at the colon end | 1 | Drainage under general anesthesia | |
| Non-occlusive mesenteric ischemia | 1 | Partial resection of small bowel | |
| Surgical Site Infection | 1 | Skin grafting | |
| V | Sepsis due to colon perforation | 5 | |
| Chronic heart failure | 1 | ||
| Cerebral infarction | 1 | ||
| Total | 23※ |
※Duplicates in 4 cases
Risk factors for postoperative mortality
Preoperative, surgical, and postoperative factors were compared between the survival (n=39) and postoperative mortality groups (n=7) (Table 4). Among the preoperative factors, the mortality group was significantly older (P<0.05), had more patients on antithrombotic drugs (P<0.05), had a higher incidence of Hinchey IV conditions (P<0.05), lower WBC counts (P<0.001), higher serum lactate levels (P<0.001) and higher APACHE II scores (P<0.001). No significant differences were found in the surgical factors. Among the postoperative factors, the mortality group had more postoperative ventilator management (P<0.01), more PMX-DHP was performed (P<0.01), and required more noradrenaline administration (P<0.01). Univariate analysis of risk factors for postoperative mortality (Table 5) showed that age (≥75 years) (P<0.05), history of cerebrovascular disease (P<0.05), administration of antithrombotic drugs (P<0.05), Hinchey IV (P<0.05), anormal WBC counts (P<0.05), higher lactate levels (≥3.0 mmol/L) (P<0.01), APACHE II scores (≥18) (P<0.01), postoperative ventilator management (P<0.05), induction of PMX-DHP therapy (P<0.01), and noradrenaline use (P<0.01) were significant. In multivariate analysis, administration of antithrombotic drugs (hazard ratio [HR], 16.6; 95% confidence interval [CI], 1.32-209; P<0.05) and higher lactate levels (≥ 3.0 mmol/L) (HR, 42.3; 95% CI, 2.69-667; P<0.01) were independent risk factors for postoperative mortality.
Table 4.
Clinical Characteristics in Patients with Perforated Diverticulitis.
| Variables | Survival group (n=39) |
Mortality group (n=7) |
P-value |
|---|---|---|---|
| Preoperative factors | |||
| Age (years) ※ | 65 (24-88) | 80 (71-86) | <0.05 |
| Sex | 0.446 | ||
| Male | 21 | 5 | |
| Female | 18 | 2 | |
| Body Mass Index※ | 21.7 (14.3-29.9) | 20.5 (19.3-27.0) | 0.842 |
| Smoking history | 0.234 | ||
| Yes | 16 | 1 | |
| No | 23 | 6 | |
| Cardiac disease | 0.350 | ||
| Yes | 10 | 3 | |
| No | 29 | 4 | |
| Cerebrovascular disease | 0.060 | ||
| Yes | 4 | 3 | |
| No | 35 | 4 | |
| Diabetes mellitus | 0.570 | ||
| Yes | 6 | 0 | |
| No | 33 | 7 | |
| Maintenance dialysis | 1.000 | ||
| Yes | 3 | 0 | |
| No | 36 | 7 | |
| Antithrombotic drug | <0.05 | ||
| Yes | 8 | 5 | |
| No | 31 | 2 | |
| Steroid use | 1.000 | ||
| Yes | 4 | 1 | |
| No | 35 | 6 | |
| Hinchey classification | <0.05 | ||
| III | 29 | 2 | |
| IV | 10 | 5 | |
| Shock-Index | 0.064 | ||
| ≥1.0 | 8 | 4 | |
| 1.0> | 31 | 3 | |
| WBC (/μL) ※ | 11000 (1000-30700) | 2420 (360-7660) | <0.001 |
| CRP (mg/dL) ※ | 11.3 (0.02-46.9) | 1.17 (0.1-29.9) | 0.482 |
| Albumin (g/dL) ※ | 3.7 (2.5-5.2) | 3.0 (1.6-4.1) | 0.095 |
| Lactate (mmol/L) ※ | 2.4 (1.7-6.5) | 5.6 (2.8-8.1) | <0.001 |
| APACHE II score※ | 13 (6-23) | 21 (17-25) | <0.001 |
| Surgical factors | |||
| Time from onset to surgery (hour) | 48 (2-720) | 10 (5-48) | 0.102 |
| Resection of the perforation site | 0.413 | ||
| Yes | 26 | 6 | |
| No | 13 | 1 | |
| Ostomy creation | 1.000 | ||
| Yes | 37 | 7 | |
| No | 2 | 0 | |
| Operation time (min) ※ | 177 (58-429) | 156 (97-310) | 0.878 |
| Blood loss (ml) ※ | 50 (1-1150) | 100 (20-1000) | 0.317 |
| Postoperative factors | |||
| Ventilator management | <0.01 | ||
| Yes | 10 | 6 | |
| No | 29 | 1 | |
| PMX-DHP | <0.01 | ||
| Yes | 8 | 6 | |
| No | 31 | 1 | |
| Noradrenalin use | <0.01 | ||
| Yes | 9 | 6 | |
| No | 30 | 1 |
※Median (Range); WBC, White Blood Cell; CRP, C-reactive protein; PMX-DHP, Polymyxin B-immobilized fiber column direct hemoperfusion
Table 5.
Risk Factors for Postoperative Mortality.
| Variables | Object | Control | Univariate analysis P-value |
Multivariate analysis | |
|---|---|---|---|---|---|
| Hazard ratio (95%CI) | P-value | ||||
| Preoperative factors | |||||
| Age (years) | ≥75 | 75> | <0.05 | ||
| Sex | Male | Female | 0.395 | ||
| Body Mass Index | ≥23 | 23> | 0.538 | ||
| Smoking history | Yes | No | 0.205 | ||
| Cardiac disease | Yes | No | 0.359 | ||
| Cerebrovascular disease | Yes | No | <0.05 | ||
| Diabetes mellitus | Yes | No | N.E | ||
| Maintenance dialysis | Yes | No | N.E | ||
| Antithrombotic drug | Yes | No | <0.05 | 16.600 (1.320-209.000) | <0.05 |
| Steroid use | Yes | No | 0.754 | ||
| Hinchey classification | IV | III | <0.05 | ||
| Shock-Index | ≥1.0 | 1.0> | 0.056 | ||
| WBC (/μL) | ≥15000/3000> | 3000～14999 | <0.05 | ||
| CRP (mg/dL) | ≥5.0 | 5.0> | 0.190 | ||
| Albumin (g/dL) | 3.0> | ≥3.0 | 0.519 | ||
| Lactate (mmol/L) | ≥3.0 | 3.0> | <0.01 | 42.300 (2.690-667.000) | <0.01 |
| APACHE II score | ≥18 | 17> | <0.01 | ||
| Surgical factors | |||||
| Time from onset to surgery (hour) | ≥24 | 24> | 0.102 | ||
| Resection of the perforation site | No | Yes | 0.332 | ||
| Ostomy creation | No | Yes | N.E | ||
| Operation time (min) | ≥180 | 180> | 0.538 | ||
| Blood loss (ml) | ≥100 | 100> | 0.362 | ||
| Postoperative factors | |||||
| Ventilator management | Yes | No | <0.05 | ||
| PMX-DHP | Yes | No | <0.01 | ||
| Noradrenalin use | Yes | No | <0.01 | ||
※N.E; Not-entimated; WBC, White Blood Cell; CRP, C-reactive protein; PMX-DHP, Polymyxin B-immobilized fiber column direct hemoperfusion
Characteristics of patients who underwent PMX-DHP therapy
Patient characteristics according to the induction of PMX-DHP therapy were investigated (Table 6). In terms of preoperative factors, PMX-DHP therapy was performed more frequently in older patients (P<0.01), those on antithrombotic drugs (P<0.05), those classified as Hinchey IV (P<0.01), those with lower WBC counts (P<0.001), lower CRP levels (P<0.05), higher lactate levels (P<0.001), and higher APACHE II scores (P<0.01). Concerning surgical and postoperative factors, the time from onset to surgery was significantly shorter (P<0.05), and interventions such as ventilator management (P<0.0001) and noradrenaline use (P<0.0001) were more frequently in the PMX-DHP group. The median APACHE II score in the PMX-DHP group was 19.5 (range, 10-25). A subgroup analysis was performed to demonstrate the efficacy of PMX-DHP in severe cases with poor general condition. This analysis explored the correlation between PMX-DHP therapy and prognosis across different patient groups: those taking antithrombotic drugs (n=13), those with higher lactate levels (≥ 3.0 mmol/L) (n=13), those with both conditions (n=6) and those with neither condition (n=26) (Table 7). The group with neither antithrombotic drug use nor high lactate levels reported no deaths. However, PMX-DHP did not improve prognosis in the respective groups.
Table 6.
Clinical Characteristics in Patients with Perforated Diverticulitis According to PMX-DHP Therapy.
| Variables | PMX-DHP group (n=14) |
Control group (n=32) |
P-value |
|---|---|---|---|
| Preoperative factors | |||
| Age (years) ※ | 81 (48-88) | 64 (24-87) | <0.01 |
| Sex | 1.000 | ||
| Male | 8 | 18 | |
| Female | 6 | 14 | |
| Body Mass Index※ | 20.3 (14.3-27.0) | 21.9 (14.6-29.9) | 0.252 |
| Smoking history | 0.195 | ||
| Yes | 3 | 14 | |
| No | 11 | 18 | |
| Cardiac disease | 0.171 | ||
| Yes | 6 | 7 | |
| No | 8 | 25 | |
| Cerebrovascular disease | 0.176 | ||
| Yes | 4 | 3 | |
| No | 10 | 29 | |
| Diabetes mellitus | 0.651 | ||
| Yes | 1 | 5 | |
| No | 13 | 27 | |
| Maintenance dialysis | 1.000 | ||
| Yes | 1 | 2 | |
| No | 13 | 30 | |
| Antithrombotic drug | <0.05 | ||
| Yes | 7 | 6 | |
| No | 7 | 26 | |
| Steroid use | 0.157 | ||
| Yes | 3 | 2 | |
| No | 11 | 30 | |
| Hinchey classification | <0.01 | ||
| III | 5 | 26 | |
| IV | 9 | 6 | |
| Shock-index | 0.143 | ||
| ≥1.0 | 6 | 6 | |
| 1.0> | 8 | 26 | |
| WBC (/μL) ※ | 3245 (360-17430) | 11965 (2800-30700) | <0.001 |
| CRP (mg/dL) ※ | 0.81 (0.03-45.1) | 13.5 (0.02-46.9) | <0.05 |
| Albumin (g/dL) ※ | 3.3 (1.6-4.8) | 3.8 (2.5-5.2) | 0.148 |
| Lactate (mmol/L) ※ | 4.1 (1.4-8.1) | 1.7 (0.9-6.5) | <0.001 |
| APACHE II score※ | 19.5 (10-25) | 12.5 (6-23) | <0.01 |
| Surgical factors | |||
| Time from onset to surgery (hour) | 10 (5-96) | 48 (2-720) | <0.05 |
| Resection of the perforation site | 0.169 | ||
| Yes | 12 | 20 | |
| No | 2 | 12 | |
| Ostomy creation | 1.000 | ||
| Yes | 14 | 30 | |
| No | 0 | 2 | |
| Operation time (min) ※ | 194 (97-310) | 163 (58-429) | 0.136 |
| Blood loss (ml) ※ | 80 (20-1150) | 25 (1-550) | 0.063 |
| Postoperative factors | |||
| Ventilator management | <0.0001 | ||
| Yes | 11 | 5 | |
| No | 3 | 27 | |
| Noradrenalin use | <0.0001 | ||
| Yes | 11 | 4 | |
| No | 3 | 28 |
※Median (Range); WBC, White Blood Cell; CRP, C-reactive protein
Table 7.
Subgroup Analysis of Postoperative Mortality According to PMX-DHP.
| Survival group |
Mortality group |
P-value | |
|---|---|---|---|
| Antithrombotic drug (n=13) | 0.266 | ||
| PMX-DHP group (n=7) | 3 | 4 | |
| Control group (n=6) | 5 | 1 | |
| Lactate (≥3.0 mmol/L) (n=13) | 0.192 | ||
| PMX-DHP group (n=10) | 4 | 6 | |
| Control group (n=3) | 3 | 0 | |
| Antithrombotic drug and Lactate (≥3.0 mmol/L) (n=6) | 0.333 | ||
| PMX-DHP group (n=5) | 1 | 4 | |
| Control group (n=1) | 1 | 0 | |
| Antithrombotic drug nor Lactate (≥3.0 mmol/L) (n=26) | 1.000 | ||
| PMX-DHP group (n=2) | 2 | 0 | |
| Control group (n=24) | 24 | 0 |
PMX-DHP, Polymyxin B-immobilized fiber column direct hemoperfusion
Discussion
In this retrospective observational study, the risk factors for postoperative mortality in patients with colorectal diverticular perforations were found to be the administration of antithrombotic drugs and preoperative lactate levels. In a subgroup analysis, PMX-DHP may not be effective in improving prognosis in patients with poor general conditions.
In this study, the common causes of colorectal perforation in the preceding five years were colorectal diverticula (39.7%) and colorectal cancer (26.7%). The number of colorectal diverticular perforation cases is expected to continue to increase owing to the Westernization of diets. Furthermore, colorectal perforations can progress from peritonitis to sepsis, with a mortality rate ranging from 12.3% to 21.7%[18]. The mortality rate of colorectal diverticular perforation in our hospital was 15.2%, which is similar to those of previous studies.
In this study, the sigmoid colon was the predominant site of perforation, consistent with previous literature[1]. In 46 emergency surgery cases, Hartmann surgery (63.0%) and ileostomy creation only (19.6%) were more frequently performed, because of its shorter operative time and the risk of suture failure under septic conditions. As for the anastomosis, only three patients (6.5%) underwent one-stage anastomoses after resection of the perforated site. A previous study reported that a comparison between the Hartmann's procedure group and the one-stage anastomosis group (including ileostomy creation) for perforated colorectal diverticulitis (Hinchey III/IV) showed no significant difference in postoperative mortality, while stoma-free survival at 1 year was significantly higher in the one-stage anastomosis group[19]. Considering the patients' quality of life, increasing the use of one-stage anastomoses, including ileostomy creations, is a future challenge in our hospital.
In previous studies of colorectal diverticular perforation, a history of steroid use, American Society of Anesthesiologists Physical Status (ASA-PS) score ≥ 3, abnormal creatinine levels, Hinchey classification IV, low albumin levels, low BMI ≤ 21 kg/m2, and Mannheim Peritonitis Index score ≥ 27[20] were reported as risk factors for postoperative mortality[11-13]. In our study, seven preoperative factors, including abnormal WBC counts, were associated with mortality. In general, WBC counts are often high in patients with infections but may be low due to severe sepsis associated with colorectal perforations[21]. In our study, administration of antithrombotic drugs and higher lactate levels (≥ 3.0 mmol/L) were independent risk factors.
The relationship between antithrombotic drug use and sepsis mortality remains uncertain. Some studies suggest that pre-admission oral aspirin may reduce sepsis mortality[22], while others find no beneficial effect on prognosis[23]. In this study, the group taking antithrombotic drugs exhibited higher mortality, yet there was no significant difference in the amount of surgical blood loss between the survival and postoperative mortality groups. This suggests that the underlying conditions necessitating antithrombotic therapy may be more closely related to perioperative mortality than bleeding due to anticoagulant effects. It is conceivable that patients with comorbidities requiring oral antithrombotic drugs could have higher mortality rates if they develop sepsis. Conversely, mortality might not increase if the underlying condition is well-managed with these drugs. Given the small sample size of this study, further research is needed to explore these associations more thoroughly.
In addition, serum lactate levels are considered indicators of peripheral circulatory failure[24]. Peripheral vasodilation due to sepsis causes tissue hypoperfusion, which increases the anaerobic metabolism and enhances lactate production. Lactate is metabolized by the liver and kidneys in the Cori cycle[25]. Thus, the administration of antithrombotic drugs was considered a patient factor related to medical history and lifestyle, whereas the lactate level was considered a factor reflecting the severity of the sepsis. These two factors have not been reported in previous literature, and the findings in this study are important for the future treatment of colorectal diverticular perforations.
Randomized controlled trials, such as the ABDOMIX study and EUPHRATES trial, which investigated the efficacy of PMX-DHP therapy for sepsis with peritonitis or high endotoxin-activity assay levels, showed its effect on blood pressure elevation but not on prognostic improvements[26,27]. There is a report suggesting that PMX-DHP is less effective in severe cases, such as in patients with colorectal perforations with APACHE II scores of 17 or higher[28]. The 2020 Japanese Guidelines for the Treatment of Sepsis state that PMX-DHP is “weakly recommended not to do” for the treatment of sepsis[29]. However, there are also reports on the efficacy of PMX-DHP in groups treated with continuous hemodiafiltration dialysis for acute renal failure associated with sepsis[14], suggesting that PMX-DHP therapy may be effective under certain conditions. In our study, PMX-DHP was frequently performed in critically ill patients with a high risk of postoperative mortality. Interestingly, the CRP levels in the PMX-DHP group were significantly lower than those in the control group. Typically, it takes approximately 6-8 hours for blood CRP levels to increase following the onset of inflammation[30]. The significantly lower CRP levels in the PMX-DHP group may be attributed to the timing of the preoperative blood tests, which were likely conducted soon after the onset of inflammation. This is supported by the shorter average time from onset to surgery in the PMX-DHP group (mean 10 hours) compared to the control group (mean 48 hours).
Subgroup analysis was performed to evaluate the efficacy of PMX-DHP in severe cases. PMX-DHP did not improve prognosis in groups identified with poor independent prognostic factors: those using antithrombotic drugs (n=13), those with higher lactate levels (n=13), and those with both conditions. Despite PMX-DHP being administered to patients in severe condition, its efficacy appears limited in these specific groups. It may be beneficial for a very narrow subset of patients requiring high-dose medications but do not exhibit several poor prognostic factors. Given the high cost of PMX-DHP therapy, future research should focus on identifying appropriate candidates and optimizing its effectiveness in real-world practice.
This study had some limitations. This was a single-center retrospective study, and the sample size was insufficient. In addition, although missing values may have caused bias, we considered them to be relatively small. Furthermore, PMX-DHP introduction criteria, the number of times PMX-DHP was performed, the duration of PMX-DHP therapy, and the blood flow rate were not standardized. Therefore, it is possible that there were differences in the efficacy of PMX-DHP from case to case.
Despite these limitations, we believe that our study, which used real-world data from a regional central hospital in Japan, has important clinical implications. The administration of antithrombotic drugs and higher preoperative lactate levels (≥ 3.0 mmol/L) were independent risk factors for postoperative mortality in patients with colorectal diverticular perforations, and PMX-DHP therapy may not be effective in patients with high mortality risks. After the diagnosis of colorectal perforation due to diverticulitis, it is essential to confirm the patient's history, including medications, and to improve sepsis and circulatory status through early surgical intervention.
Conflicts of Interest
There are no conflicts of interest.
Author Contributions
Nakagawa: Project development, data collection and analysis, and manuscript writing and editing
Ito: Project development, manuscript review and editing
Nakahara: Data collection
Sakamoto: Data collection
Hosoda: Data collection
Hayashi: Data collection
Nakamura: Data collection
Komiya: Data collection
Ishii: Data collection
Morimoto: Data collection
Yoshida: Data collection
Noritomi: Project development and manuscript editing
Approval by Institutional Review Board (IRB)
Fukuoka Tokushukai Hospital Clinical Research Ethics Review Committee
(approval number: 231004)
References
- 1.Yamamichi N, Shimamoto T, Takahashi Y, et al. Trend and risk factors of diverticulosis in Japan: age, gender, and lifestyle/metabolic-related factors may cooperatively affect on the colorectal diverticula formation. PLoS One. 2015 Apr;10(4):e0123688. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Peery AF, Barrett PR, Park D, et al. A high-fiber diet does not protect against asymptomatic diverticulosis. Gastroenterology. 2012 Feb;142(2):266-72.e1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Nagata N, Niikura R, Aoki T, et al. Increase in colonic diverticulosis and diverticular hemorrhage in an aging society: lessons from a 9-year colonoscopic study of 28,192 patients in Japan. Int J Colorectal Dis. 2014 Mar;29(3):379-85. [DOI] [PubMed] [Google Scholar]
- 4.Hsu CW, Wang JH, Kung YH, et al. What is the predictor of surgical mortality in adult colorectal perforation? The clinical characteristics and results of a multivariate logistic regression analysis. Surg Today. 2017 Jun;47(6):683-9. [DOI] [PubMed] [Google Scholar]
- 5.Ishikawa D, Takehara Y, Takata A, et al. Combination of dirty mass volume and APACHE II score predicts mortality in patients with colorectal perforation. World J Emerg Surg. 2021 Mar;16(1):17. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Shinkawa H, Yasuhara H, Naka S, et al. Factors affecting the early mortality of patients with nontraumatic colorectal perforation. Surg Today. 2003;33(1):13-7. [DOI] [PubMed] [Google Scholar]
- 7.Matsuoka T, Yamamoto R, Matsumura K, et al. Perioperative clinical parameters associated with short-term mortality after colorectal perforation. Eur J Trauma Emerg Surg. 2022 Aug;48(4):3017-24. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Harries RL, Twine CP, Kugathasan G, et al. Prognostic factors for survival following emergency Hartmann's procedure. Postgrad Med J. 2012 Apr;88(1038):205-9. [DOI] [PubMed] [Google Scholar]
- 9.Shimizu T, Hanasawa K, Sato K, et al. The clinical significance of serum procalcitonin levels following direct hemoperfusion with polymyxin B-immobilized fiber column in septic patients with colorectal perforation. Eur Surg Res. 2009;42(2):109-17. [DOI] [PubMed] [Google Scholar]
- 10.Tan KK, Hong CC, Zhang J, et al. Predictors of outcome following surgery in colonic perforation: an institution's experience over 6 years. J Gastrointest Surg. 2011 Feb;15(2):277-84. [DOI] [PubMed] [Google Scholar]
- 11.Broersen LHA, Horváth-Puhó E, Pereira AM, et al. Corticosteroid use and mortality risk in patients with perforated colonic diverticular disease: a population-based cohort study. BMJ Open Gastroenterol. 2017 Apr;4(1):e000136. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Ince M, Stocchi L, Khomvilai S, et al. Morbidity and mortality of the Hartmann procedure for diverticular disease over 18 years in a single institution. Colorectal Dis. 2012 Aug;14(8):e492-8. [DOI] [PubMed] [Google Scholar]
- 13.Mäkelä JT, Kiviniemi H, Laitinen S. Prognostic factors of perforated sigmoid diverticulitis in the elderly. Dig Surg. 2005;22(1-2):100-6. [DOI] [PubMed] [Google Scholar]
- 14.Iwagami M, Yasunaga H, Noiri E, et al. Potential survival benefit of polymyxin B hemoperfusion in septic shock patients on continuous renal replacement therapy: A propensity-matched analysis. Blood Purif. 2016;42(1):9-17. [DOI] [PubMed] [Google Scholar]
- 15.Nagata N, Ishii N, Manabe N, et al. Guidelines for Colonic Diverticular Bleeding and Colonic Diverticulitis: Japan Gastroenterological Association. Digestion. 2019;99 Suppl 1:1-26. [DOI] [PubMed] [Google Scholar]
- 16.Knaus WA, Draper EA, Wagner DP, et al. APACHE II: a severity of disease classification system. Crit Care Med. 1985 Oct;13(10):818-29. [PubMed] [Google Scholar]
- 17.Dindo D, Demartines N, Clavien PA. Classification of surgical complications: a new proposal with evaluation in a cohort of 6336 patients and results of a survey. Ann Surg. 2004 Aug;240(2):205-13. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Bielecki K, Kamiński P, Klukowski M. Large bowel perforation: morbidity and mortality. Tech Coloproctol. 2002 Dec;6(3):177-82. [DOI] [PubMed] [Google Scholar]
- 19.Lambrichts DPV, Vennix S, Musters GD, et al. Hartmann's procedure versus sigmoidectomy with primary anastomosis for perforated diverticulitis with purulent or faecal peritonitis (LADIES): a multicentre, parallel-group, randomised, open-label, superiority trial. Lancet Gastroenterol Hepatol. 2019 Aug;4(8):599-610. [DOI] [PubMed] [Google Scholar]
- 20.Gueiros LDS, Fonseca CMD, Duarte NMDM, et al. Mannheim's peritonitis index in the prediction of postoperative outcome of peritonitis. Rev Col Bras Cir. 2022 Sep;49:e20222991. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Solomkin JS, Bauman MP, Nelson RD, et al. Neutrophils dysfunction during the course of intra-abdominal infection. Ann Surg. 1981 Jul;194(1):9-17. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Hsu WT, Porta L, Chang IJ, et al. Association Between Aspirin Use and Sepsis Outcomes: A National Cohort Study. Anesth Analg. 2022 Jul;135(1):110-7. [DOI] [PubMed] [Google Scholar]
- 23.Eisen DP, Leder K, Woods RL, et al. Effect of aspirin on deaths associated with sepsis in healthy older people (ANTISEPSIS): a randomised, double-blind, placebo-controlled primary prevention trial. Lancet Respir Med. 2021 Feb;9(2):186-95. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Hernández G, Ospina-Tascón GA, Damiani LP, et al. Effect of a resuscitation strategy targeting peripheral perfusion status vs serum lactate levels on 28-day mortality among patients with septic shock: The ANDROMEDA-SHOCK Randomized Clinical Trial. JAMA. 2019 Feb;321(7):654-64. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Suhara T, Hishiki T, Kasahara M, et al. Inhibition of the oxygen sensor PHD2 in the liver improves survival in lactic acidosis by activating the Cori cycle. Proc Natl Acad Sci USA. 2015 Sep;112(37):11642-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Payen DM, Guilhot J, Launey Y, et al. Early use of polymyxin B hemoperfusion in patients with septic shock due to peritonitis: a multicenter randomized control trial. Intensive Care Med. 2015 Jun;41(6):975-84. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Dellinger RP, Bagshaw SM, Antonelli M, et al. Effect of targeted polymyxin B hemoperfusion on 28-day mortality in patients with septic shock and elevated endotoxin level: The EUPHRATES Randomized Clinical Trial. JAMA. 2018 Oct;320(14):1455-63. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Sugimoto K, Sato K, Maekawa H, et al. Analysis of the efficacy of direct hemoperfusion with polymyxin B-immobilized fiber (PMX-DHP) according to the prognostic factors in patients with colorectal perforation. Surg Today. 2013 Sep;43(9):1031-8. [DOI] [PubMed] [Google Scholar]
- 29.Egi M, Ogura H, Yatabe T, et al. The Japanese Clinical Practice Guidelines for Management of Sepsis and Septic Shock 2020 (J-SSCG 2020). J Intensive Care. 2021 Aug;9(1):53. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Marnell L, Mold C, Du Clos TW. C-reactive protein: ligands, receptors and role in inflammation. Clin Immunol. 2005 Nov;117(2):104-11. [DOI] [PubMed] [Google Scholar]