Treatment with Ceftriaxone in Complicated Diverticulitis Increases the Incidence of Intra-Abdominal Enterococcus faecium Detection

Abstract

Background:

Complicated diverticulitis of the sigmoid colon typically is treated by resection after initial antibiotic treatment. Third-generation cephalosporins are the drugs of choice but are not effective against enterococci and can induce colonic colonization by Enterococcus faecium within hours. Infections caused by enterococci, especially E. faecium, are difficult to treat but should be considered in the strategic treatment planning of hospital-acquired peritonitis (e.g., anastomotic leakage), especially in immunocompromised patients.

Methods:

To determine whether the duration of pre-operative ceftriaxone treatment in complicated diverticulitis increases the incidence of intra-abdominal E. faecium detection, we analyzed all patients operated on for diverticulitis of the sigmoid colon in our department between 2008 and 2016.

Results:

Analyzing 516 resections performed for complicated diverticulitis, we found that E. faecium generally was detected intra-abdominally more often in the group that underwent longer pre-operative ceftriaxone treatment (≥ 4 days). During primary resection, E. faecium was detected in 2.7%, 11.1%, and 37.0% cases in the group undergoing immediate operation, 1–3 days of antibiotic treatment, and ≥4 days of antibiotic treatment, respectively. Enterococcus faecium was detected in 0, 25.0%, and 70.6% of surgical revisions and 28.6%, 14.3%, and 56.0%, respectively, of incisional surgical site infections with identified pathogens. A multivariable analysis discovered anastomotic leakage and antibiotic treatment lasting ≥4 days to be independent risk factors for intra-abdominal isolation of E. faecium.

Conclusion:

A ceftriaxone treatment ≥4 days led to a higher incidence of intra-abdominal E. faecium. Our data further suggested that empiric coverage of E. faecium in the treatment of hospital-acquired peritonitis could be beneficial after a long duration of ceftriaxone treatment.

Diverticulitis of the sigmoid colon is one of the most common causes of abdominal infection in industrialized countries, with a rising incidence over the last decades [1]. The proportion of disease-affected adults rises with increasing age, even though only 10%–25% of the patients with diverticular disease develop an acute episode of diverticulitis [2]. Hospital admissions in Europe because of complicated diverticulitis have been on the rise over the last few decades [3].

Complicated diverticulitis is characterized by diverticulum-related bowel perforation with fecal or generalized peritonitis and the presence of an abscess or localized para-colic abscess [4,5]. Although the modern treatment of acute diverticulitis has intended to be conservative, complicated diverticulitis often leads to sigmoid resection to avoid complications [6 -10]. In perforated diverticulitis with generalized peritonitis (Hansen/Stock [HS] IIc), immediate surgery is necessary [11]. In perforated diverticulitis with localized abscess (HS IIb), initial antibiotic treatment followed by surgery after an interval is preferred [12].

Typically, third-generation cephalosporins accompanied by metronidazole are recommended to treat acute diverticulitis [13]. However, these drugs have no efficacy against enterococci, a commensal in the human bowel, and can have different side effects such as the early induction of colonization by E. faecium (within hours) [14 -16]. After proliferation in the bowel followed by translocation into deep tissue and the blood stream outside the intestinal tract, enterococci have high pathogenic potential [17,18]. It has been shown that enterococci play a relevant role in peritonitis, and coverage is advised in special circumstances [13,19,20]. For example, it is recommended to consider them in calculating treatment for hospital-acquired peritonitis (e.g., anastomotic leakage), especially in immunocompromised patients [13].

Enterococcal infections are difficult to treat because of their intrinsic and frequent extrinsic resistance to several antibiotics. Enterococcus faecalis and E. faecium typically comprise >90% of all enterococcus species found. Clinical isolates of E. faecium often are resistant to beta-lactam agents, making therapeutic coverage more difficult [21]. Furthermore, the increasing rates of vancomycin-resistant enterococci cause a problem for designing an adequate antimicrobial regimen [22].

Antibiotic therapy itself can cause bowel diseases with relevant morbidity that requires treatment, such as Clostridioides difficile infections [23]. Furthermore, it can result in a change of the intestinal microbiome and also can lead to colonization by multi-resistant pathogens and influence the spectrum of pathogens in secondary peritonitis. Increased use of third-generation cephalosporins has led to a significantly higher incidence of extended-spectrum beta-lactamases [24, 25]. However, their role in the rise of intra-abdominal E. faecium colonization has not yet been tested.

Therefore, we investigated whether ceftriaxone treatment leads to an increased prevalence of enterococci in the context of peritonitis, because generally, enterococci also increase the appearance of surgical site infections (SSIs) [26]. To answer this question, we analyzed the effects of the duration of initial antibiotic treatment in patients operated on for complicated diverticulitis.

Patients and Methods

Study population

All patients who underwent resection for complicated or frequently recurrent diverticulitis of the sigmoid colon between January 2008 and September 2016 at the secondary referral colorectal center of Marienhospital Stuttgart (Vinzenz von Paul Kliniken gGmbH, Germany) were included in a prospectively maintained database (n = 814). In this retrospective study, we evaluated all patients who were admitted for complicated sigmoid diverticulitis and operated on during the same hospital stay (n = 516). Patients with uncomplicated diverticulitis (HS I or IIa) usually were treated conservatively, which was not the subject of this study. This study was approved by our Institutional Review Board and conducted in accordance with the 1964 Declaration of Helsinki. Patients with colorectal cancer and those undergoing resection for diverticulitis of the right side were excluded.

Treatment algorithm

All patients suspected of having acute sigmoid diverticulitis underwent computed tomography scans after admission. In cases of HS IIc disease, surgery was performed immediately (“none” group), whereas initial antibiotic treatment was started, and early elective resection was planned, for HS IIb cases. In case of improvement of the symptoms and decreased inflammatory signs, operations were performed after approximately 7–10 days. In case of worsening and increase in inflammatory signs, operations were scheduled earlier. Pre-operative antibiotic treatment for <3 days and at least 4 days was categorized as short and long, respectively.

Design

The primary endpoint was the prevalence of E. faecium intra-abdominally during surgery secondary to complicated sigmoid diverticulitis with respect to the duration of the pre-operative treatment with ceftriaxone and metronidazole. Secondary endpoints were the prevalence of E. faecium in superficial or deep incisional SSI as well as the incidence of post-operative complications, both procedure related and general.

Procedures

Patient characteristics and intra-operative data were extracted from a prospective database. The patients were assessed for post-operative complications on a daily basis until discharge and subsequently underwent routine out-patient follow-up. Existing diagnosis and complications were documented by members of the Medical Documentation Department. Determination of wound classification and surveillance for SSIs and infectious complications were performed by a member of the Department of Hygiene and reported to the German National Reference Centre for Surveillance of Nosocomial Infections. For SSIs, the treating team was urged to obtain a subcutaneous swab on the day of diagnosis (BD ESwab™, Copan Italia SpA, Brescia, Italy). During primary resections with existing peritonitis or during revision surgery by re-laparotomy or re-laparoscopy, an intra-abdominal swab was taken routinely during the procedure. Swabs from drains were not taken because they can cause misleading results. Also, blood cultures were not collected routinely [27].

After anaerobic and aerobic overnight cultures, bacterial isolates were identified using the MALDI TOF Biotyper (Bruker Daltonics, Bremen, Germany). Antimicrobial susceptibility was determined by the broth microdilution method, and the results were interpreted according to the minimum inhibitory concentration interpretive breakpoints recommended by the European Committee on Antimicrobial Susceptibility Testing using WA96 (Beckman Coulter, Krefeld, Germany). All detected pathogens were analyzed and assigned according to their intra-abdominal or subcutaneous origin and added to the database by a member of the Department of Laboratory Medicine.

Peri-operative care

During the study period, the bowel was prepared with a small enema (Freka Clyss^®, Fresenius Kabi Deutschland GmbH, Bad Homburg, Germany) on the evening before and the morning of the day of surgery. Oral antibiotics were not administered at that time. Resections were performed using standardized procedures by nine Board-certified surgeons. Sigmoid resections typically were performed with double-stapled transanal anastomosis in the upper rectum. Closed-system intra-abdominal drainage was used routinely, inserted through a trocar incision away from the closed and covered extraction site (Blake™ Silicone Drain, Ethicon, Somerville, New Jersey, USA).

Antibiotics

Patients treated for acute complicated diverticulitis were given 2 g of ceftriaxone once and 500 mg of metronidazole thrice a day intravenously from the day of admission. The treatment was continued until 48–72 hours after the procedure.

Definition of complications

Complications were graded according to the Clavien-Dindo system [28]. Surgical site infections were diagnosed on the basis of the Centers for Disease Control and Prevention criteria [29]. Pneumonia was diagnosed by the presence of pulmonary infiltrates on a chest radiograph, together with elevated C-reactive protein (CRP) or white blood cell (WBC) count and clinical signs. Urinary tract infection was defined as a leukocyte- and/or nitrite-positive urine test combined with clinical signs, elevated WBC, fever, or some combination thereof.

Statistical analysis

The JMP 13.1 software program (SAS Institute Inc., Cary, NC) was used for statistical analysis. The baseline data and occurrence of endpoints were analyzed using the non-parametric U-test (Mann-Whitney-Wilcoxon) for continuous variables. The χ² test or Fisher exact test were used for categorical variables as appropriate. The albumin, CRP, and WBC data are presented in the tables in the original units of measurement.

Univariable analysis of the groups was performed using binary logistic regression with generalized models. Multivariable analysis incorporating all variables shown in Tables 1 and 2 (both significant and insignificant in univariate analysis) was performed for the intra-abdominal detection of E. faecium during primary resection or surgical revision in the short and long treatment groups. It consisted of multiple regression analysis with backward elimination. The procedure was rerun until no explanatory variable was left that could be removed without markedly worsening the prediction of the endpoint.

Table 1.

Baseline Data of All Patients

Group	None	Short (1–3 d Abx)	Long (≥ 4 d Abx)	p
Pre-treatment (%)	53 (10.5)	177 (34.3)	286 (55.4)
Age (y) (range)	63 (42–98)	63 (22–89)	65 (30–99)	065
Female (%)	31 (58.5)	120 (67.8)	166 (58.0)	0.10
Established disease (%)
Pulmonary	16 (30.2)	47 (26.6)	65 (22.7)	0.41
Cardiovascular (incl. hypertension)	30 (56.6)	94 (53.1)	153 (53.5)	0.90
Diabetes	4 (7.6)	17 (9.6)	32 (11.2)	0.68
Hepatic	2 (3.8)	12 (6.7)	18 (6.3)	0.73
Impaired renal function	10 (18.9)	13 (7.3)	21 (7.3)	0.02
Anemia	10 (18.9)	24 (13.6)	36 (12.6)	0.47
Smoking (%)	7 (13.2)	23 (13.0)	26 (9.1)	0.36
Steroid use (%)
Body Mass Index (kg/m²)	25.2 (18.6–33.1)	24.3 (16.8–31.3)	25.8 (17.7–32.8)	0.72
Albumin (g/dL)	3.8 (3.1–4.6)	4.1 (4–4.2)	3.9 (2.3–4.6)	0.26
Maximum CRP at admission (mg/L)	168.5 (0.6–429.0)	31.9 (0.4–360.0)	64.1 (0.2–363.0)	<0.001
Maximum WBC at admission ( × 10⁹)	11.7 (1.3–31.3)	10.3 (1.2–25.9)	10.6 (0.8–24.3)	0.32
Maximum CRP at day of surgery	168.5 (0.6–429.0)	33.2 (0.7–374.0)	10.7 (0.3–477.0)	<0.001
Maximum WBC at day of surgery	11.7 (1.3–31.3)	13.6 (1.0–25.5)	13.0 (5.1–28.8)	0.49
Pre-operative colonoscopy	0	1 (0.6)	0	<0.001

Data are reported as absolute numbers (percentage) or median (min-max); p values are calculated using the U test (Mann-Whitney-Wilcoxon) for continuous data, Pearson's χ² test for categorical data.

Abx = antibiotic treatment before surgery; CRP = C-reactive protein (normal value [nv]: <10.0 mg/L); WBC = white blood cell count/leukocytes (nv: 3.5–9.8 × 10⁹/L).

Table 2.

Intra- and Post-Operative Features (%)

Group	None	Short (1–3 d Abx)	Long (≥ 4 d Abx)	p
Pre-treatment	53 (10.5)	177 (34.3)	286 (55.4)
Approach
Laparoscopic	3 (5.7)	113 (63.8)	209 (73.1)
Open	46 (86.8)	57 (32.2)	61 (21.3)
Conversion	4 (7.6)	7 (4.0)	16 (5.6)	<0.001
Perforated sigmoid	53 (100.0)	90 (50.9)	153 (53.5)	<0.001
Diverticulitis stadium
IIA	0	87 (49.2)	131 (45.8)
IIB	0	90 (50.9)	155 (54.2)
IIC	53 (100.0)	0	0
III	0	0	0	<0.001
Cholecystectomy simultaneous	2 (3.8)	4 (2.3)	8 (2.8)	0.83
Cholecystectomy up to POD 30	0	1 (0.6)	1 (0.4)	0.83
Creation of
Colostomy	21 (39.6)	11 (6.2)	9 (3.2)
Ileostomy	3 (5.7)	3 (1.7)	5 (1.8)	<0.001
Wound contamination [38]
II	0	8 (4.5)	23 (8.0)
III	11 (25.6)	84 (47.5)	149 (52.1)
IV	40 (75.5)	76 (42.9)	112 (39.2)
Unknown	2 (3.8)	9 (5.1)	2 (0.7)	<0.001
Length of operation (min)	161.5 (88–235)	165 (82–297)	177 (56–342)	0.58
Post-operative length of stay (d)	13 (7–53)	8 (4–58)	9 (5–164)	<0.001

Data are reported as absolute numbers (percentage); p-values are calculated using the U test (Mann-Whitney-Wilcoxon) for continuous data, Pearson's chi-squared test for categorical data.

Abx = antibiotic treatment before surgery; POD = postoperative day.

Results

Included in this study were 516 patients who were operated on for acute, complicated diverticulitis. In the same period, 298 patients were operated on because of uncomplicated, recurrent diverticulitis and were excluded. Because of free perforation of the sigmoid colon and acute abdomen, 53 patients (10.3%) were operated on during the same day they were admitted (“none” group). After administering antibiotic treatment for 1–3 days, 177 patients (34.3%) were operated on (short treatment group). Antibiotic treatment was administered for at least 4 days to 286 patients (55.4%) before their operation (long treatment group). The age, sex distribution, and established diseases did not differ between the groups (Table 1). The “none” group showed impaired renal function more commonly than the short group (18.9% versus 7.3%).

The laboratory tests for inflammatory markers differed between groups. The CRP values were significantly higher in the none group, the WBC counts did not differ among the groups, and the CRP values did not differ between the short and long treatment groups.

Commonly, an open approach was used in the none group, whereas the laparoscopic approach was used for resection in the other groups (Table 2). The differences were statistically significant between all groups, as well as between the short and long groups (p = 0.03). In the none group, all patients were operated on for free perforation, whereas the differences in the stage of diverticulitis did not differ significantly between the short and long treatment groups. The same applied for colostomy and ileostomy, as well as the grading of site contamination.

All specific post-operative complications are listed in Table 3. Surgical site infections occurred more often in the none group, but the values did not differ significantly between the other two groups (p = 0.38). The same applied for the occurrence of post-operative pneumonia (p = 0.90).

Table 3.

Complications

Group	None	Short (1–3 d Abx)	Long (≥ 4 d Abx)	p
Pretreatment (%)	53 (10.5)	177 (34.3)	286 (55.4)
Post-operative UTI	3 (5.7)	8 (4.5)	22 (7.7)	0.39
Post-operative ileus	8 (15.1)	34 (19.2)	46 (16.1)	0.63
Pneumonia	5 (9.4)	4 (2.3)	7 (2.5)	0.02
Surgical site infection	20 (37.7)	17 (9.6)	35 (12.3)	<0.001
Death	2 (3.8)	1 (0.6)	3 (1.1)	0.16
Readmission	3 (5.7)	8 (4.5)	14 (4.9)	0.94
Anastomotic leakage	1 (1.9)	3 (1.7)	7 (2.5)	0.85
Surgical revision	8 (15.1)	11 (6.2)	22 (7.7)	0.11
Clostridial infection	2 (3.8)	3 (1.7)	9 (3.2)	0.57

UTI = Urinary tract infection.

During primary resection, bacterial pathogens could be detected in 100 patients (19.4%), and more pathogens were isolated intra-abdominally in the none group (Table 4). The rate of isolated pathogens differed between the short and long treatment groups (20.3% versus 9.4%; p < 0.001). We detected E. faecium in 15 patients (15.0%) comprising 2.7%, 11.1%, and 37% of the three groups, respectively.

Table 4.

Proportion of E. faecium Isolates (%) among Pathogens Recovered

Group	None	Short (1-3 d Abx)	Long (≥4 d Abx)	p
Primary resection	53	177	286
Pathogens recovered	37 (69.8)	36 (20.3)	27 (9.4)	<0.001
E. faecium isolates	1 (2.7)	4 (11.1)	10 (37.0)	0.001
Surgical revision	8 (15.1)	11 (6.2)	22 (7.7)	0.11
Pathogens recovered	5 (62.5)	8 (72.7)	17 (77.2)	0.92
E. faecium isolates	0	2 (25.0)	12 (70.6)	<0.01
Incisional surgical site infection	20 (37.7)	17 (9.6)	35 (12.2)	<0.001
Pathogens recovered	14 (70.0)	14 (82.4)	25 (71.4)	0.89
E. faecium isolates	4 (28.6)	2 (14.3)	14 (56.0)	0.02

Surgical revision was needed in 41 of the 516 patients (7.9%). During these procedures, we isolated pathogens in 30 patients (73.2%), and the percentages between the groups did not differ. Enterococcus faecium was detected in 14 patients (46.7%), with no cases reported in the none group, and 25.0% and 70.6% in the short and long treatment groups, respectively. Post-operative SSI occurred in 72 patients (14.0%), and pathogens were isolated in 53 patients (73.6%). We identified E. faecium in 20 patients (27.8%), significantly more in the long group than in the other groups.

In 43 of the 177 patients (24.3%) of the short treatment group, the pathogens could be isolated from intra-abdominal swabs during either primary resection or surgical revision. Six of those swabs (14.0%) contained E. faecium. In the long treatment group, isolated pathogens were less frequent (43 patients; 15.0%), although the incidence of E. faecium increased (22 patients; 51.2%; p < 0.001).

To determine the effects of intra-abdominal detection of E. faecium, we compared the groups with and without detection during primary resection or surgical revision for the two groups with prior antibiotic treatment (short and long) (Table 5). The patients' conditions differed only for diabetes mellitus, with no significant difference observed for other diseases, as listed in Table 1. The stage of diverticulitis and laboratory tests for inflammatory signs did not differ significantly. There was a significantly worse outcome, with a 11.1% post-operative mortality rate, in the group with and nil in the group without intra-abdominal E. faecium detection (p < 0.01). Furthermore, there was a higher incidence of incisional SSI in the E. faecium-positive group. Subcutaneous swabs in the case of incisional SSI found E. faecium significantly more frequently in this group (69.2% versus 0; p < 0.001). The hospital stay was longer in the E. faecium-positive group.

Table 5.

Groups with Recovered Pathogens in Primary Resection or Surgical Revision after Antibiotic Treatment before Resection (“Short” and “Long” Groups)

E. faecium Detection	Yes (%)	No (%)	p
	27 (31.4)	59 (68.6)
Pre-condition
Age (y)	74 (33–89)	68 (29–90)	0.53
Diabetes mellitus	7 (25.9)	5 (8.5)	0.03
Diverticulitis stadium
IIA	6 (22.2)	6 (10.2)
IIB	21 (77.8)	53 (89.8)	0.13
Maximum laboratory values (range)
C-reactive protein^a at admission day	43.1 (0.9–292.0)	122.5 (1.8–363.0)	0.17
White blood cell count^b at admission day	11.9 (5.3–25.9)	11.7 (1.1–24.9)	0.92
C-reactive protein at day of surgery	87.3 (4.5–234.0)	87.9 (2.5–352.0)	0.49
White blood cell count at day of surgery	14.2 (5.1–25.5)	13.7 (2.6–26.8)	0.85
Length of operation (min)	160 (138-241)	171 (67-297)	1.0
Outcome
Incisional surgical site infection	13 (48.2)	12 (20.3)	<0.01
Death	3 (11.1)	0	<0.01
Readmission	4 (6.8)	1 (3.7)	0.57
Length of hospital stay (d) (range)	17 (6–164)	11 (6–58)	0.01

Patients had least one isolate of E. faecium recovered from intra-abdominal swabs during primary resection or surgical revision relative to the total number of patients from which bacterial species were isolated. For all other pre-existing diseases shown in Table 1, no significant differences exist.

Normal value <10.0 mg/L.

Nornal value 3.5–9.8 × 10⁹/L.

To identify the risk factors for detecting E. faecium intra-abdominally, we performed further analyses for all patients of the short and long treatment groups. A univariable analysis detected treatment in the long treatment group (p < 0.001), anastomotic leakage (p = 0.001), and diabetes mellitus (p = 0.02) as significant risk factors for intra-abdominal E. faecium.

Multivariable analysis kept long treatment, anastomotic leakage, high WBC count at admission (p = 0.03), laparoscopic procedure, hepatic and pulmonary diseases, creation of an ostomy, and CRP values on the day of operation in the final model (determination coefficient [r²] 0.42; p < 0.001). As we considered a p value <0.05 to be statistically significant, membership in the long treatment group (p < 0.001) and anastomotic leakage (p < 0.001) were proved to be independent risk factors for the occurrence of E. faecium.

Discussion

The potential involvement of E. faecium in hospital-acquired peritonitis can complicate the antibiotic treatment, especially if calculated regimens are necessary. Our study assessed the influence of pre-operative antibiotic therapy with third-generation cephalosporins on complicated sigmoid diverticulitis in the appearance of E. faecium intra-abdominally and in incisional SSI. We found that pre-operative antibiotic treatment of ≥4 days significantly increased the detection rates of E. faecium, which confirms the conclusion of a recent review on the increase of intestinal enterococcal colonization after use of ceftriaxone [30].

Cephalosporins have no efficacy against enterococci, and Siesing et al. have shown that the increase in the doses of cefuroxime in an orthopedic clinic led to more detection of enterococci in tissue samples [31]. It has been shown that this is particularly true for ceftriaxone, but not for cephalosporins in general [32]. Furthermore, vancomycin and third-generation cephalosporins are associated most commonly with the spread of vancomycin-resistant enterococci [33]. Dahms et al. could show this effect in abdominal surgery, especially transplant surgery, as well [34]. The more common detection could be the result of a selection of the intestinal microbiome with disruption of colonization and proliferation of enterococci. This was shown in the colon of mice after ceftriaxone application, leading to a reduction of the mucus-associated microbiota layer and segregation of E. faecium from the intestinal wall [35,36], subsequently leading to more frequent detection after perforation or anastomotic leakage.

During primary surgery, E. faecium was detectable in only about 3% of all patients who were operated on. However, the proportion of patients with detectable intra-abdominal bacteria was significantly higher (15%). The absolute increase in risk between the short and long treatment groups was significant, but not pronounced. In the case of revision surgery, however, E. faecium was detectable in 34% of all cases operated on, again a significant and this time clinically relevant absolute increase in risk. In the long treatment group, 70.6% of patients with detectable intra-abdominal bacteria yielded E. faecium. This effect may be favored by the antibiotic therapy, which usually was given between primary resection and revision. Unfortunately, no evaluable data were available for this period in our study population. The question is if, at all, under what conditions E. faecium has to be included in designing strategic antibiotic regimens. Post-operative enterococcal infections are associated with increased mortality rates [37]. In our study population, an increased mortality rate was observed in antibiotically pretreated patients with intra-abdominal E. faecium. The basic conditions such as age and previous diseases of both groups differed only with regard to established diabetes mellitus and were comparable otherwise. This supports the former findings of the harmful role of the presence of enterococci in peritoneal fluid in immunologically suppressed patients [38]. Therefore, some guidelines and reviews recommend using empiric therapy with anti-enterococcal efficacy in patients considered at risk for enterococcal infection [13,39]. Post-operative infection, substantial recent exposure to broad-spectrum antimicrobial therapy, and severe sepsis without improvement after initial therapy were named as risk factors [13].

On the other hand, we found a significantly higher rate of incisional SSI in the group with intra-abdominal E. faecium detection. As expected, subcutaneous swabs also very often detected E. faecium in the superficial abdominal site in incisional SSI. This indicates that antibiotic prophylaxis is ineffective in these patients who may require a change of the antibiotic agent [40]. The generous use of broad-spectrum antibiotics, on the other hand, can promote post-operative pulmonary infections [41].

Limitations

Our study was limited by its retrospective and single-center design. Nonetheless, the study population represents the complete spectrum observed at a large-volume colorectal center. A complete consecutive series without exclusion of any patient groups was evaluated. Hence, the results are applicable to everyday practice. Furthermore, the incidence of isolated pathogens was low during primary resection for complicated diverticulitis after antibiotic treatment. This could be attributable to the antimicrobial efficacy of the antibiotic serum concentrations, preventing bacterial growth in vitro and lowering the detection rate (false-negative swabs). Microbiologic surveys have to deal with this problem.

In addition, there are no data on the duration and medication of any pre-clinical antibiotic treatment. Of course, both can influence the prevalence of E. faecium. As well, post-operative blood cultures were not taken regularly; therefore, no statement can be made about the correlation between intra-abdominal detection of E. faecium and the development of positive blood cultures. The incidence of vancomycin resistance was low during the study period but increased afterward. In our study, we therefore could not make any reference to the effects of antibiotic therapy on the development of resistance. Finally, there is no comparative cohort treated with another or without any antibiotic agent. The increased detection of E. faecium could be the consequence of the colonic inflammation rather than of the antibiotic application. As discussed before, a shift in the colon microbiota in mice to E. faecium was induced hours after ceftriaxone application, which makes it more likely to be a consequence of the latter [16]. This should be addressed further in future studies.

Conclusions

Our results indicate that ceftriaxone treatment administered for >3 days can lead to a higher incidence of detectable E. faecium intra-abdominally in patients with primary resection or with revision operations for sigmoid diverticulitis as well as in post-operative incisional SSI. Enterococcus faecium was the most frequently isolated pathogen in as many as 70% of patients undergoing revision procedures, especially with anastomotic leakage. Our data further suggest that empiric coverage of E. faecium while treating hospital-acquired peritonitis could be beneficial after a long duration of ceftriaxone treatment.

Footnotes

Author Disclosure Statement

None of the authors has any commercial associations to disclose that might create conflicts of interest, whether actual or potential.

References

Rafferty

, Shellito

, Hyman

, Buie

, Standards Committee of American Society of Colon and Rectal

Surgeons

. Practice parameters for sigmoid diverticulitis. Dis Colon Rectum, 2006; 49:939–944.

Martel

, Raskin

. History, incidence, and epidemiology of diverticulosis. J Clin Gastroenterol, 2008; 42:1125–1127.

Hupfeld

, Pommergaard

, Burcharth

, Rosenberg

. Emergency admissions for complicated colonic diverticulitis are increasing: A nationwide register-based cohort study. Int J Colorectal Dis, 2018; 33:879–886.

Jacobs

. Clinical practice. Diverticulitis. N Engl J Med, 2007; 357:2057–2066.

Hinchey

, Schaal

, Richards

. Treatment of perforated diverticular disease of the colon. Adv Surg, 1978; 12:85–109.

Siddiqui

, Zahid

, Hong

, Young

. Colorectal surgeon consensus with diverticulitis clinical practice guidelines. World J Gastrointest Surg, 2017; 9:224–232.

Scarpa

, Buchs

, Poncet

, et al. Short-term intravenous antibiotic treatment in uncomplicated diverticulitis does not increase the risk of recurrence compared to long-term treatment. Ann Coloproctol, 2015; 31:52–56.

Nally

, Kavanagh

. Current controversies in the management of diverticulitis: A review. Dig Surg, 2019; 36:195–205.

Devaraj

, Liu

, Tatum

, et al. Medically treated diverticular abscess associated with high risk of recurrence and disease complications. Dis Colon Rectum, 2016; 59:208–215.

10.

Boostrom

, Wolff

, Cima

, et al. Uncomplicated diverticulitis; more complicated than we thought. J Gastrointest Surg, 2012; 16:1744–1749.

11.

Hansen

, Graupe

, Stock

. [Prognostic factors in perforating diverticulitis of the large intestine] (Ger). Chirurg, 1998; 69:443–449.

12.

Kruis

, Germer

, Leifeld

, German Society for

Gastroenterology

, Digestive and Metabolic Disease and the German Society for General and Visceral

Surgery

. Diverticular disease: Guidelines of the German Society for Gastroenterology, Digestive and Metabolic Diseases and the German Society for General and Visceral Surgery. Digestion, 2014; 90:190–207.

13.

Mazuski

, Tessier

, May

, et al. The Surgical Infection Society Revised Guidelines on the Management of Intra-Abdominal Infection. Surg Infect, 2017; 18:1–76.

14.

Choi

, Jung

, Choi

, et al. Gallbladder pseudolithiasis caused by ceftriaxone in young adult. J Korean Surg Soc, 2011; 81:423–426.

15.

Saner

, Canbay

, Gerken

, Broelsch

. [Collateral damage of cephalosporins and quinolones and possibilities for control] (Ger). Med Klin (Munich), 2009; 104:114–118.

16.

Hendrickx

APA

, van de Kamer

, Willems

RJL

. Primary murine mucosal response during cephalosporin-induced intestinal colonization by Enterococcus faecium. Microbiology Open, 2018; 7:e00602.

17.

Nagpal

, Yadav

. Bacterial translocation from the gut to the distant organs: An overview. Ann Nutr Metab, 2017; 71(Suppl 1):11–16.

18.

Schietroma

, Pessia

, Colozzi

, et al. Does bacterial translocation influence the postoperative infections in splenectomized patients after abdominal trauma?. Surgeon, 2018; 16:94–100.

19.

Seguin

, Brianchon

, Launey

, et al. Are enterococci playing a role in postoperative peritonitis in critically ill patients?. Eur J Clin Microbiol Infect Dis, 2012; 31:1479–1485.

20.

Harbarth

, Uckay

. Are there patients with peritonitis who require empiric therapy for enterococcus?. Eur J Clin Microbiol Infect Dis, 2004; 23:73–77.

21.

Jia

, Li

, Wang

. Prevalence and antimicrobial resistance of Enterococcus species: A hospital-based study in China. Int J Environ Res Public Health, 2014; 11:3424–3442.

22.

Gastmeier

, Schroder

, Behnke

, et al. Dramatic increase in vancomycin-resistant enterococci in Germany. J Antimicrob Chemother, 2014; 69:1660–1664.

23.

Paterson

. “Collateral damage” from cephalosporin or quinolone antibiotic therapy. Clin Infect Dis, 2004; 38(Suppl 4):S341–S345.

24.

Urbanek

, Kolar

, Loveckova

, et al. Influence of third-generation cephalosporin utilization on the occurrence of ESBL-positive Klebsiella pneumoniae strains. J Clin Pharm Ther, 2007; 32:403–408.

25.

Marchaim

, Lazarovitch

, Efrati

, et al. Serious consequences to the use of cephalosporins as the first line of antimicrobial therapy administered in hemodialysis units. Nephron Clin Pract, 2005; 101:c58–c64.

26.

Pochhammer

, Kramer

, Schaffer

. [Enterococci and surgical site infections: Causal agent or harmless commensals?] (Ger). Chirurg, 2017; 88:377–384.

27.

Everts

, Heneghan

, Adholla

, Reller

. Validity of cultures of fluid collected through drainage catheters versus those obtained by direct aspiration. J Clin Microbiol, 2001; 39:66–68.

28.

Dindo

, Demartines

, Clavien

. Classification of surgical complications: A new proposal with evaluation in a cohort of 6336 patients and results of a survey. Ann Surg, 2004; 240:205–213.

29.

Horan

, Gaynes

, Martone

, et al. CDC definitions of nosocomial surgical site infections, 1992: A modification of CDC definitions of surgical wound infections. Infect Control Hosp Epidemiol, 1992; 13:606–608.

30.

Bhalodi

, van Engelen

TSR

, Virk

, Wiersinga

. Impact of antimicrobial therapy on the gut microbiome. J Antimicrob Chemother, 2019; 74(Supplement 1):i6–i15.

31.

Siesing

, Alva-Jorgensen

, Brodersen

, et al. Rising incidence of Enterococcus species in microbiological specimens from orthopedic patients correlates to increased use of cefuroxime: A study concentrating on tissue samples. Acta Orthop, 2013; 84:319–322.

32.

McKinnell

, Kunz

, Chamot

, et al. Association between vancomycin-resistant enterococci bacteremia and ceftriaxone usage. Infect Control Hosp Epidemiol, 2012; 33:718–724.

33.

Mazuski

. Vancomycin-resistant enterococcus: Risk factors, surveillance, infections, and treatment. Surg Infect, 2008; 9:567–571.

34.

Dahms

, Johnson

, Statz

, et al. Third-generation cephalosporins and vancomycin as risk factors for postoperative vancomycin-resistant enterococcus infection. Arch Surg, 1998; 133:1343–1346.

35.

Chakraborty

, Lam

, Kommineni

, et al. Ceftriaxone administration disrupts intestinal homeostasis, mediating noninflammatory proliferation and dissemination of commensal enterococci. Infect Immun, 2018; 86:e00674-18.

36.

Hendrickx

, Top

, Bayjanov

, et al. Antibiotic-driven dysbiosis mediates intraluminal agglutination and alternative segregation of Enterococcus faecium from the intestinal epithelium. MBio, 2015; 6:e01346-15.

37.

Sitges-Serra

, Lopez

, Girvent

, et al. Postoperative enterococcal infection after treatment of complicated intra-abdominal sepsis. Br J Surg, 2002; 89:361–367.

38.

Riche

, Dray

, Laisne

, et al. Factors associated with septic shock and mortality in generalized peritonitis: Comparison between community-acquired and postoperative peritonitis. Crit Care, 2009; 13:R99.

39.

Sartelli

. A focus on intra-abdominal infections. World J Emerg Surg, 2010; 5:9.

40.

Tanaka

, Nakamura

, Imai

, et al. The use of broad-spectrum antibiotics reduces the incidence of surgical site infection after pancreatoduodenectomy. Surg Today, 2018; 48:825–834.

41.

Saleh

, Khanna

, Chagin

, et al. Glycopeptides versus beta-lactams for the prevention of surgical site infections in cardiovascular and orthopedic surgery: A meta-analysis. Ann Surg, 2015; 261:72–80.