Advertisement

Enterococcal infective endocarditis is a marker of current occult or future incident colorectal neoplasia

Published:October 09, 2020DOI:https://doi.org/10.1016/j.ejim.2020.10.006

      Highlights

      • Patients with enterococcal endocarditis often have a history of prior cancer.
      • As S. gallolyticus, enterococcal endocarditis is associated with colorectal neoplasia.
      • Colonoscopy is generally safe in enterococcal endocarditis patients.
      • Enterococcal endocarditis may signal occult or future incident colorectal neoplasia.

      Abstract

      Background

      : Few studies suggest an association between Enterococcal infective endocarditis (EIE) and colorectal disease, including colorectal neoplasia (CRN) and colorectal cancer (CRC). In this study, we analyze differences in prevalence, risk factors and outcome of CRN and CRC between EIE and Streptococcus gallolyticus infective endocarditis (SGIE).

      Methods

      : Single center, observational study of 166 patients with definite EIE or SGIE. Clinical data were collected prospectively in a standardized IE protocol. Colonoscopy data were collected retrospectively on 90 patients.

      Results

      : 85 patients had EIE, 81 SGIE. EIE patients had a higher rate of prior cancer (20% vs 6%) and health-care associated infection (12% vs 1%), but similar mortality than SGIE. Colonoscopy performed in 90 patients showed intestinal diseases in 30 of 42 (71%) EIE patients vs. 40 of 48 (83%) SGIE patients (p = 0.174), with a predominance of CRN. Among 78 patients who underwent colonoscopy after IE diagnosis, no difference between EIE and SGIE was observed in the rate of non-neoplastic lesions (48% vs 47%), benign (32% vs 40%) or malignant (13% vs 15%) neoplastic lesions. Adverse events during colonoscopy were uncommon, although a careful handling of anticoagulation was required.

      Conclusions

      : EIE seems to be associated with colorectal disease, including colorectal neoplasia and colorectal cancer, to the same extent as SGIE. EIE should be considered a marker of colorectal neoplasia, even in patients with a clear health-care related acquisition. Colonoscopy is generally safe in EIE patients, and should be considered to early diagnose and treat colorectal disease.

      Keywords

      1. Introduction

      Streptococcus gallolyticus and Enterococcus species are members of the gut microbiota and a major cause of infective endocarditis (IE), mainly affecting elderly subjects with digestive or urinary sources of bacteremia [
      • Durante-Mangoni E.
      • Bradley S.
      • Selton-Suty C.
      • Tripodi M.F.
      • Barsic B.
      • Bouza E.
      • et al.
      Current Features of Infective Endocarditis in Elderly Patients Results of the International Collaboration on Endocarditis Prospective Cohort Study.
      ,
      • Murdoch D.R.
      • Corey G.R.
      • Hoen B.
      • Miró J.M.
      • Fowler Jr, V.G.
      • Bayer A.S
      Clinical presentation, etiology, and outcome of infective endocarditis in the 21st century: the International Collaboration on Endocarditis-Prospective Cohort Study.
      ,
      • Chirouze C.
      • Athan E.
      • Alla F.
      • Chu V.H.
      • Corey G.R.
      • Selton-Suty C.
      • et al.
      Enterococcal endocarditis in the beginning of the 21st century: analysis from the International Collaboration on Endocarditis-Prospective Cohort Study.
      ,
      • Anderson D.J.
      • Olaison L.
      • McDonald J.R.
      • Miro J.M.
      • Hoen B.
      • Selton-Suty C.
      • et al.
      Enterococcal prosthetic valve infective endocarditis: report of 45 episodes from the International Collaboration on Endocarditis-merged database.
      ]. Whilst S. gallolyticus causes spontaneous, community-acquired IE, often associated with colorectal disease [
      • Tripodi M.
      • Adinolfi L.
      • Ragone E.
      • Durante-Mangoni E.
      • Fortunato R.
      • Iarussi D.
      • et al.
      Clinical features of Streptococcus gallolyticus endocarditis and its association with chronic liver disease: an underestimated risk factor.
      ,
      • Klein R.S.
      • Recco R.A.
      • Catalano M.T.
      • Edberg S.C.
      • Casey J.I.
      • Steigbigel N.H
      Association of Streptococcus bovis with carcinoma of the colon.
      ], Enterococcal IE more often involves prosthetic valves and associates to health care procedures [
      • Chirouze C.
      • Athan E.
      • Alla F.
      • Chu V.H.
      • Corey G.R.
      • Selton-Suty C.
      • et al.
      Enterococcal endocarditis in the beginning of the 21st century: analysis from the International Collaboration on Endocarditis-Prospective Cohort Study.
      ,
      • Fernandez-Guerrero M.L.
      • Goyenechea A.
      • Verdejo C.
      • Roblas R.F.
      • de Górgolas M
      Enterococcal endocarditis on native and prosthetic valves. a review of clinical and prognostic factors with emphasis on hospital-acquired infections as a major determinant of outcome.
      ]. Whether enterococcal IE is also tied with colorectal diseases, especially colorectal neoplasms (CRN), remains unclear.
      A recent study assessed in a comparative fashion the incidence of CRN in patients with IE due to S. gallolyticus (SGIE) or Enterococcus spp. (EIE) [
      • Corredoira M.J.
      • García-País M.J.
      • Coira A.
      • Rabuñal R.
      • García-Garrote F.
      • Pita J.
      • et al.
      Differences between endocarditis caused by Streptococcus gallolyticus and Enterococcus spp. and their association with colorectal cancer.
      ]. The former was significantly more commonly associated with CRN than the latter, during both the acute phase and the subsequent long term follow-up of IE [
      • Corredoira M.J.
      • García-País M.J.
      • Coira A.
      • Rabuñal R.
      • García-Garrote F.
      • Pita J.
      • et al.
      Differences between endocarditis caused by Streptococcus gallolyticus and Enterococcus spp. and their association with colorectal cancer.
      ]. Based on these data, Corredoira et al. recommended a screening colonoscopy in all patients with prior or current SGIE but not EIE.
      The association between CRN and EIE has been recently re-analysed in more detail [
      • Pericàs J.M.
      • Corredoira J.
      • Moreno A.
      • García-País M.J.
      • Falces C.
      • Rabuñal R.
      • et al.
      Relationship Between Enterococcus faecalis Infective endocarditis and colorectal neoplasm: preliminary results from a cohort of 154 patients.
      ]. Among 154 cases, 67 underwent at least once a colonoscopy and the overall rate of CRN, especially CR cancer (CRC), was higher than expected. Interestingly, this was particularly true in those cases of Enterococcus faecalis IE with no obvious source of infection. Similar findings emerged from an Australian study on a smaller number of patients [
      • Athan E.
      • Cabiltes I.
      • Coghill S.
      • Bowe S.J
      Silent but deadly: patients with enterococcal bacteraemia should be assessed for colorectal neoplasia.
      ].
      Several questions, however, remain unanswered: i) should all or only a subset of EIE patients undergo a colonoscopy? ii) does the gut play any role in the disease onset when the apparent source of infection is genitourinary or nosocomial (procedure- or catheter-related)? iii) how feasible is colonoscopy in the IE setting, including management of anticoagulation or the timing of cardiac surgery?
      In this study, we tried to answer these questions analyzing our center experience with EIE with a specific focus on endoscopy-proven intestinal disease.

      2. Materials and methods

      2.1 Study design

      This is a retrospective analysis of observational data collected in the context of an ongoing prospective study at the Unit of Infectious & Transplant Medicine, Monaldi Hospital, University of Campania “Luigi Vanvitelli” in Naples, Italy. Patients hospitalized for definite IE due to Enterococcus spp or S. gallolyticus according to the existing criteria [
      • Li J.S.
      • Sexton D.J.
      • Mick N.
      • Nettles R.
      • Fowler Jr, V.G.
      • Ryan T.
      • et al.
      Proposed modifications to the Duke criteria for the diagnosis of infective endocarditis.
      ,
      • Habib G.
      • Lancellotti P.
      • Antunes M.J.
      • Bongiorni M.G.
      • Casalta J.P.
      • Del Zotti F.
      • et al.
      ESC Guidelines for the management of infective endocarditis: the Task Force for the Management of Infective Endocarditis of the European Society of Cardiology (ESC).
      ] between January 2000 and November 2019 were included in this study. Data were collected by means of dedicated case report forms and entered at the end of each patient hospitalization in an electronic database. Data on abdominal symptoms, clinical history of gut disease and colonoscopy investigations were collected retrospectively from patient charts and IE outpatient clinic reports for the purposes of this analysis.
      IE data collection was approved by the Ethics Committee of the University of Campania “Luigi Vanvitelli”. All patients gave their informed consent to participation in the study.

      2.2 Patients

      From a cohort of 746 patients admitted to our center for IE during the study period, we included in this analysis 166 patients (22.25%) with IE due to either Enterococcus species (EIE, case group), including E. faecalis, E. faecium and E. durans, or S. gallolyticus (SGIE, control group).
      Patient data were collected as part of a standardized protocol of IE evaluation in use at our Unit, including baseline clinical data, particularly clinical history, physical examination, chest X-ray, abdominal ultrasound scan and laboratory analyses. A trans-thoracic echocardiogram (TTE) was performed within 72 h of admission, followed by transesophageal echocardiogram (TEE) when needed. Detailed information about valve characteristics (native, prosthetic, type, etc.) and endocardial vegetations (number, size, length and position) were collected. Embolic events and their characteristics (location, extension, complications) were also recorded.
      Major comorbidities of patients, including kidney and liver disease, diabetes mellitus and prior neoplasms were always collected.

      2.3 Definitions

      IE was deemed to be: community-acquired (CA), when symptoms ensued outside of the hospital in a patient without prior health care contact in the preceding 6 months; hospital-acquired (HA), when symptoms occurred during hospitalization; health care associated (HCA), when disease symptoms ensued in patients who were on renal replacement therapy, or who had been hospitalized for any reason or who had undergone any invasive procedure or intravenous treatment during the prior 6 months. IE portal of entry was presumed to be gastrointestinal in all EIE patients who did not have a genito-urinary source and had no HA infection. Long term mortality was defined as death at the latest follow up after discharge.

      2.4 Colonoscopy evaluation

      Between 2000 and 2014, all data regarding intestinal disease evaluation and possible colonoscopy performance were collected retrospectively. Instead, since 2014, all patients with SGIE and EIE were actively assessed for the need of colonoscopy, during or after IE hospitalization. Patients were generally offered to undergo the exam, during either the acute phase of IE or the subsequent follow up. Exceptions were those with a colonoscopy done within six months of IE onset or before admission to our center for IE. Colonoscopy was not offered during the acute phase to patients who were hemodynamically unstable or when kidney disease or a coagulopathy or other severe complications were deemed to hamper the safety of the procedure. Data derived from pre-admission colonoscopy were included in the present analysis. No patient underwent virtual CT colonoscopy.
      We considered any endoscopy-proven intestinal lesion, classified in non-neoplastic disease (unspecific colitis/ulcerative colitis, diverticulosis/diverticulitis, polyposis) and colorectal neoplasia (adenoma and adenocarcinoma).
      Antimicrobial prophylaxis for colonoscopy was always administered when the exam was performed after IE diagnosis. Only sparse data on colonoscopy prophylaxis were available when the exam was performed before IE diagnosis.

      2.5 Statistical analysis

      Numerical data are presented as either mean ± standard deviation (SD) or median with interquartile range (IQR), whilst categorical/nominal data as number and percentage. Comparisons were made through the Mann-Whitney U test for numerical variables and either the Chi-squared or Fisher's exact test for nominal variables. A p-value <0.05 was considered statistically significant.
      Analyses were performed using the statistical software for Windows SPSS 23 (SPSS, Inc., Chicago, Illinois, USA).

      3. Results

      3.1 Clinical characteristics of the two study subgroups

      Of the 166 patients enrolled, 85 had EIE (82 E. faecalis, 2 E. durans, 1 E. faecium) and 81 had SGIE. The major clinical features of these two groups are summarized in Table 1. All patients were Caucasian and most were males, the median age being >60 years in both groups. EIE patients had more often a prosthetic valve IE and associated complications, including dehiscence and periprosthetic leaks. EIE were also more likely than SGIE patients to have a health care associated infection and a prior history of IE (Table 1). The presumed sources of infection at initial evaluation are also shown. Five patients (3%) presented two episodes of IE in the study period. Specifically, 2 patients had 2 distinct episodes of EIE and 1 patient suffered from 2 different episodes of SGIE. Also, 2 patients had 2 sequential episodes of IE related to E. faecalis and S. gallolyticus, respectively.
      Table 1Clinical features of the two study subgroups.
      ParameterEnterococcus IE (n = 85)S. gallolyticus IE (n = 81)P value
      Age (years), median [IQR]65 [59 - 72.75]65 [56 – 72]0.898
      Sex, n (%)0.179
      Males/Females56 (65.9)/29 (34.1)62 (76.5)/19 (23.5)
      IE Location0.083
      Aortic, n (%)42 (49.4)34 (42)
      Mitral, n (%)27 (31.8)23 (28.4)
      Tricuspid, n (%)4 (4.7)4 (4.9)
      PMK/AICD, n (%)3 (3.5)
      Multiple sites/valves, n (%)9 (10.6)20 (24.7)
      Valve type, n (%)0.000
      Native47 (55.3)69 (85.2)
      Prosthetic35 (41.2)11 (13.6)
      PMK/AICD3 (3.5)1 (1.2)
      Type of acquisition, n (%)0.009
      Community75 (88.2)80 (98.8)
      Nosocomial/HCA10 (11.8)1 (1.2)
      Previous IE, n (%)14 (16.5)5 (6.2)0.037
      Congenital heart disease, n (%)7 (8.2)6 (7.4)0.841
      Native valve predisposition, n (%)14 (16.5)18 (22.2)0.348
      Echocardiography, n (%)0.155
      TEE52 (61.2)58 (71.6)
      Regurgitant valve, n (%)0.502
      Aortic23 (27.1)27 (33.3)
      Mitral16 (18.8)19 (23.5)
      Tricuspid6 (7.1)2 (2.5)
      Multiple25 (29.4)19 (23.5)
      None15 (17.6)14 (17.2)
      Vegetation location, n (%)NA
      Aortic35 (41.2)32 (39.5)
      Mitral19 (22.3)21 (26)
      Tricuspid4 (4.7)3 (3.7)
      Catheter3 (3.5)3 (3.7)
      Wall2 (2.4)
      Multiple9 (10.6)13 (16)
      None13 (15.3)9 (11.1)
      Vegetation size (mm), median [IQR]13 [9 – 19]10 [8 – 16]0.530
      Valve perforation, n (%)14 (16.5)14 (17.3)0.887
      Abscess, n (%)6 (7)1 (1.2)0.117
      Fistula, n (%)1 (1.2)1
      Dehiscence, n (%)14 (16.5)1 (1.2)0.000
      Periprosthetic regurgitation, n (%)10 (11.8)1 (1.2)0.009
      Embolic events21 (24.7)38 (46.9)0.003
      Presumed source of infection
      Gastrointestinal70 (82.4)79 (97.5)
      Genito-urinary5 (5.8)1 (1.25)
      Other8 (9.4)1 (1.25)
      Unknown2 (2.4)0
      IE = infective endocarditis; PMK= pace-maker; AICD= automatic implantable electronic device; HCA=health care associated; TEE=transesophageal echocardiography; NA= not applicable; IQR = Interquartile range;.

      3.2 Comorbidities

      Charlson comorbidity index was significantly higher in EIE than SGIE patients (Table 2). EIE patients had an overall higher rate of comorbidities, including vascular and chronic kidney disease, and a higher rate of prior cancer (20% vs 6% in SGIE; p = 0.011) (Table 2). In particular, a prior known CRN was more prevalent among EIE than SGIE patients (9.4% vs 1.2%; p = 0.034). Also, a numerically higher prevalence of prior/current genitourinary cancer was observed in EIE patients (5.9% vs 1.2% in SGIE; p = 0.211). Interestingly, these five EIE patients with genitourinary cancers (4 prostate, 1 uterus) had been diagnosed with cancer a median of 10.5 months before IE onset.
      Table 2Main comorbidities observed in the two study groups.
      ComorbidityEnterococcus IE (n = 85)S. gallolyticus IE (n = 81)P value
      Charlson comorbidity index, median (range)4 (0–10)3 (0–12)0.021
      Chronic obstructive pulmonary disease, n (%)22 (25.9)16 (19.7)0.348
      Congestive heart failure, n (%)19 (22.3)13 (16)0.303
      Ischemic heart disease, n (%)12 (14.1)4 (4.9)0.064
      Peripheral arterial disease, n (%)10 (11.8)3 (3.7)0.080
      Cerebrovascular disease, n (%)10 (11.8)3 (3.7)0.080
      Diabetes mellitus, n (%)17 (20)16 (19.7)1
      Chronic kidney disease, n (%)14 (16.5)3 (3.7)0.009
      Dialysis, n (%)3 (3.5)0.245
      Drug abuse, n (%)5 (5.9)0.059
      Liver disease (non-cirrhotic), n (%)1 (1.2)3 (3.7)0.358
      Liver cirrhosis, n (%)4 (4.7)10 (12.3)0.096
      Peptic ulcer, n (%)3 (3.5)1 (1.2)0.620
      Rheumatic disease, n (%)4 (4.7)2 (2.5)0.682
      Hemiplegia, n (%)2 (2.3)0.497
      Dementia, n (%)2 (2.3)1 (1.2)1
      Neoplasia (known before IE diagnosis) *17 (20)5 (6.2)0.011
      Colorectal (adenoma and adenocarcinoma)8 (47)1 (20)0.034
      Genitourinary5 (29.4)1 (20)0.211
      Hepatic1 (5.9)
      Leukemia / lymphoma2 (11.8)1 (20)
      Breast1 (5.9)2 (40)
      * percentages below refer to the subgroup with prior known neoplasia (i.e. 17 in EIE and 5 in SGIE).

      3.3 Outcomes

      As shown in Table 1, embolic events were twice as incident among SGIE than EIE patients (p = 0.003). Cardiac surgery was indicated and actually performed in an equal proportion of subjects in the 2 groups: 43 (50.5%) in EIE and 37 (45.6%) in SGIE. In-hospital mortality was 10.5% in the EIE group vs 6.1% in the SGIE group (p = 0.305), whilst long term mortality was 12% and 13.6%, respectively (p = 0.729). Of note, the median follow up after discharge was 38 months (IQR 11–74).

      3.4 Colorectal disease evaluation

      Ninety patients (54.2%) went through at least one colonoscopy. The remaining subjects did not undergo the exam due to either IE-related mortality, presence of severe comorbidities, presence of a clear genitourinary source or refusal. Notably, only 8 patients overall (8.9%) were symptomatic for large bowel disease at the time of colonoscopy. The rate of colonoscopy was similar in EIE and SGIE groups (49% vs 59%; p = 0.203). Median Charlson comorbidity index was 4 (range 0–9) and 4 (range 0–10) in patients who did or did not undergo colonoscopy (p = 0.730).
      Table 3 summarizes data on colorectal disease observed. Colonoscopy was performed less often in the acute phase of IE in EIE (18 procedures, 43%) compared to SGIE (36 procedures, 75%) (p = 0.001). Among the 42 EIE patients and the 48 SGIE patients who underwent a colonoscopy, respectively 30 (71.4%) and 40 (83.3%) had an intestinal endoscopy-proven lesion (p = 0.174). In both groups, CRN emerged as the most frequent intestinal disease associated with IE, being adenomas more common than carcinomas (Table 3). No difference in the prevalence of CRN was observed between EIE and SGIE (52.4% vs 56.2%; p = 0.708). In contrast, inflammatory bowel diseases were seldom observed. Twelve patients had a positive colonoscopy for more than one bowel lesion (e.g. adenoma plus diverticulosis, polyposis plus diverticulosis) (Table 3). In the subgroup of patient who did colonoscopy after IE diagnosis, non-neoplastic lesions occurred in 15 of 31 EIE patients (48%) and in 22 of 47 SGIE patients (47%), whereas benign neoplastic disease (adenoma) occurred in 10 of 31 EIE patients (32%) and in 19 of 47 SGIE patients (40%). Malignant colonic lesions were found in 4 of 31 EIE patients (13%) and in 7 of 47 SGIE patients (15%) (Fig. 1). These data are also shown in supplementary Table 1.
      Table 3Endoscopy findings in EIE and SGIE patients.
      Enterococcus IE (n = 85)S. gallolyticus IE (n = 81)p value
      Underwent colonoscopy (any time)0.203
      Yes42 (49.4)48 (59.2)
      No43 (50.6)33 (40.8)
      Timing of colonoscopy<0.001
      During the acute phase of IE18 (42.9)37 (75)
      After IE13 (30.9)10 (18.7)
      Before IE11 (26.2)1 (6.3)
      Symptomatic patients
      - diarrhea10
      - constipation20
      - rectal bleeding or positive occult blood stool test21
      - iron deficiency anemia11
      Reason for not undergoing colonoscopy
      - early mortality94
      - overt genitourinary source51
      - patient refusal113
      - hemodynamic instability93
      - coagulopathy78
      - poor patient compliance214
      Endoscopy-proven intestinal disease
      all percentages refer to patients who underwent a colonoscopy.
      30 (71.4)40 (83.3)0.174
      Non-neoplastic disease
      more than one endoscopy-proven disease could be present (see text).
      15 (35.7)22 (45.8)
      - Unspecific Colitis, Ulcerative colitis2 (4.7)1 (2.1)
      - Diverticulosis/Diverticulitis7 (16.6)11 (22.9)
      - Non-neoplastic polyposis6 (14.3)10 (20.8)
      Colorectal neoplasia
      more than one endoscopy-proven disease could be present (see text).
      22 (52.4)27 (56.2)
      - Adenoma15 (35.7)20 (41.6)
      - Adenocarcinoma7 (16.6)7 (14.6)
      No intestinal disease (negative colonoscopy)12 (28.6)8 (16.6)
      low asterisk all percentages refer to patients who underwent a colonoscopy.
      low asterisklow asterisk more than one endoscopy-proven disease could be present (see text).
      Fig. 1
      Fig. 1Colonoscopy findings in patients without known colorectal disease at time of IE diagnosis.
      Two EIE and 2 SGIE patients who did not perform colonoscopy in the acute phase of IE developed signs/symptoms of large bowel disease (including diarrhea, constipation, iron-deficiency anemia) leading to colonoscopy a median of 5 months after being cured from IE. Of them, 2 were found to have colon cancer (adenocarcinoma). Overall, as shown in Table 3, the rate of large bowel disease, including CRN, was not significantly different in EIE compared to SGIE patients.
      The rate of colonoscopy was strictly related to the presumed source of infection in both study subgroups: it was 60.7% among SGIE and 54.3% among EIE patients with a presumed intestinal source. In contrast, the rate of colonoscopy was 26.6% in those EIE patients whose portal of entry was deemed to be other than gastrointestinal.

      3.5 Epidemiological correlates of colonoscopy in the study subjects

      The relation between colonoscopy-proven intestinal disease and the epidemiology of IE was also investigated. Of the 42 EIE patients who performed a colonoscopy, 36 (85.7%) were deemed to have CA and only 4 (9.5%) HCA/HA IE. However, endoscopy-proven lesions were observed in both EIE subgroups (25 of 36 [69.4%] CA vs 3 of 4 [75%] HCA/HA cases; p = 1). On the other hand, 10 of 12 EIE patients with a negative colonoscopy had a CA infection. Unfortunately, we were unable to describe gut diseases among EIE cases with a possible/probable genitourinary source of IE, as none of these patients underwent colonoscopy. Almost all SGIE patients had CA infection (Table 1) and only 1 patient had signs of a genitourinary source of infection.
      We analysed whether an association existed between performance of a colonoscopy and IE patient outcome. Hospital mortality was 2.2% in those who did and 15.7% in those who did not undergo colonoscopy (p<0.01). Indeed, the exam was more likely performed in those with a better clinical condition. In fact, long term survival did not differ, being 72.7% in those who did and 62.5% in those who did not undergo colonoscopy (p = 0.180).
      Among IE patients with CRN, long term survival was not significantly reduced compared to patients without CRN (75.5% vs 65.8%; p = 0.314).

      3.6 Outcome of colorectal cancer in EIE and SGIE

      Table 4 details features of the 14 CRC cases observed at any time among EIE and SGIE patients studied. The median age in this patient subgroup was 65 years, not different from the overall study group. Most of these 14 patients had a positive IE outcome with only one EIE patient who died. 9 patients were diagnosed with cancer at an early stage (4 in EIE and 5 in SGIE groups), and 1 EIE patient had a metastatic disease. Most CRC patients (10 of 14) were surgically treated, mainly with hemicolectomy, the remaining undergoing polypectomy. The outcome of CRC was worse in EIE as 4 of 7 patients died during follow up compared to only 1 of 7 among SGIE (p = 0.265), as detailed in Table 4.
      Table 4Features of the 14 colorectal cancer cases observed among EIE and SGIE patients.
      SexAgeEtiologyOutcome of IECRC stage (Astler-Coller / TNM)Timing of CRC diagnosis *TreatmentOutcome (timing of death [months after IE])
      F64EIECuredC / stage IIIAChemotherapy + radiotherapyAlive
      M71EIECuredB2 / stage IICLocal radiotherapy + rectal anterior resectionDeceased (156)
      F77EIECuredB1 / stage ICRectal anterior resectionAlive
      F62EIECuredB2 / stage IIAHemicolectomyDeceased (36)
      F80EIECuredD / stage IVCPalliative colostomyDeceased (72)
      M59EIECuredB1 / stage IBLeft hemicolectomyAlive
      F67EIEDeceasedA / stage IAPolypectomyDeceased (1)
      M63SGIECuredA / stage IBPolypectomyAlive
      M66SGIECuredB2 / stage IIABLeft hemicolectomyAlive
      M58SGIECuredB2 / stage IIACLeft hemicolectomy + colostomyAlive
      M69SGIECuredA / stage IBRight hemicolectomyAlive
      M58SGIECuredB2 / stage IIBLeft hemicolectomyDeceased (10)
      M69SGIECuredIn situ / stage 0CPolypectomyAlive
      M56SGIECuredA / stage ICPolypectomyAlive
      A: before IE onset; B: during IE hospitalization; C: after IE cure.

      3.7 Safety of colonoscopy in EIE patients

      Overall, 42 EIE patients underwent colonoscopy. Specifically, 12 did so before valve surgery, 5 after surgery and 25 without surgical indication. An ad hoc prophylaxis was done in all patients who underwent colonoscopy after IE cure. This consisted in the intravenous administration of ampicillin-sulbactam (2 gr + 1 gr) and gentamicin 160 mg, given in the context of day care hospitalization or inpatient admission. No specific prophylaxis was administered in addition to IE treatment when colonoscopy was performed during the acute phase of IE.
      The handling of anti-thrombotic medications during colonoscopy is shown in Table 5. Most patients temporarily withdrew vitamin K antagonists (VKA) and received bridging with low molecular weight heparin. Three patients stopped aspirin 5 days before the exam. None was on direct oral anticoagulant drugs.
      Table 5Management of anticoagulation in EIE patients undergoing colonoscopy.
      Colonoscopy in the active phase of IEColonoscopy during IE follow up
      Before valve surgeryAfter valve surgeryNonsurgical cases
      VKA withdrawal + LMWH bridging2074
      ASA withdrawal1143
      DOAC0000
      Complications of colonoscopy occurred in 2 patients (4.7%): bleeding after polypectomy requiring prolongation of hospital stay in 1 patient; cardiac arrest due to ventricular fibrillation, eventually leading to death, 5 h after colonoscopy completion in 1 patient with dilated cardiomyopathy; no evidence of electrolyte imbalances secondary to colonoscopy-related fluid loss was detected in this patient. In the SGIE subgroup, 1 patient, shortly after colonoscopy, developed an acute abdomen due to massive spleen embolism with impending capsule rupture, requiring emergency surgery.

      4. Discussion

      The results of this study suggest that EIE is associated with colorectal disease, including CRN and CRC, to the same extent as SGIE, and corroborate recent findings both in enterococcal bacteremia [
      • Cabiltes I.
      • Coghill S.
      • Bowe S.J.
      • Athan E
      Enterococcal bacteraemia "silent but deadly": a population-based cohort study.
      ] and IE [
      • Pericàs J.M.
      • Corredoira J.
      • Moreno A.
      • García-País M.J.
      • Falces C.
      • Rabuñal R.
      • et al.
      Relationship Between Enterococcus faecalis Infective endocarditis and colorectal neoplasm: preliminary results from a cohort of 154 patients.
      ,
      • Athan E.
      • Cabiltes I.
      • Coghill S.
      • Bowe S.J
      Silent but deadly: patients with enterococcal bacteraemia should be assessed for colorectal neoplasia.
      ,
      • Escolà-Vergé L.
      • Peghin M.
      • Givone F.
      • Pérez-Rodríguez M.T.
      • Suárez-Varela M.
      • Meije Y.
      • et al.
      Prevalence of colorectal disease in Enterococcus faecalis infective endocarditis: results of an observational multicenter study.
      ]. Being part of the normal gut microbiota, enterococci cause bacteremia and IE exploiting intestinal barrier breaching disorders through which translocation to the bloodstream occurs. Based on our as well as other study results [
      • Escolà-Vergé L.
      • Peghin M.
      • Givone F.
      • Pérez-Rodríguez M.T.
      • Suárez-Varela M.
      • Meije Y.
      • et al.
      Prevalence of colorectal disease in Enterococcus faecalis infective endocarditis: results of an observational multicenter study.
      ], this may occur also in well-defined nosocomial EIE cases. The accumulating evidence on the association between EIE and CRN suggests that a significant burden of disease might be prevented by performing systematic colonoscopy in EIE patients to potentially diagnose and treat CRN earlier.
      We tried to understand whether all or only a subset of EIE patients should undergo a colonoscopy. Excluding those with a recent exam and those with severe organ dysfunction, our data suggest that all EIE patients should undergo a colonoscopy evaluation whenever feasible [
      • Escolà-Vergé L.
      • Peghin M.
      • Givone F.
      • Pérez-Rodríguez M.T.
      • Suárez-Varela M.
      • Meije Y.
      • et al.
      Prevalence of colorectal disease in Enterococcus faecalis infective endocarditis: results of an observational multicenter study.
      ]. However, we lack data to recommend a specific strategy in different EIE subgroups.
      The optimal timing of colonoscopy is not easy to determine. We believe that when cardiorenal function is stable, colonoscopy should better be performed during the active phase of EIE and before nonurgent cardiac surgery. This allows to carry out the exam on antimicrobial therapy and without anti-thrombotic medications, which may be subsequently complex to handle. If colonoscopy is not feasible before cardiac surgery, we currently aim at performing the exam 3 to 6 months after valve surgery. In this setting, we bridge VKA with LMWH and administer a single dose antibiotic prophylaxis just before the exam. Case by case evaluation is always mandatory.
      At variance with other studies [
      • Pericàs J.M.
      • Corredoira J.
      • Moreno A.
      • García-País M.J.
      • Falces C.
      • Rabuñal R.
      • et al.
      Relationship Between Enterococcus faecalis Infective endocarditis and colorectal neoplasm: preliminary results from a cohort of 154 patients.
      ,
      • Athan E.
      • Cabiltes I.
      • Coghill S.
      • Bowe S.J
      Silent but deadly: patients with enterococcal bacteraemia should be assessed for colorectal neoplasia.
      ,
      • Escolà-Vergé L.
      • Peghin M.
      • Givone F.
      • Pérez-Rodríguez M.T.
      • Suárez-Varela M.
      • Meije Y.
      • et al.
      Prevalence of colorectal disease in Enterococcus faecalis infective endocarditis: results of an observational multicenter study.
      ], we analysed the association of EIE with colorectal disorders comparing EIE with SGIE as a control group. SGIE is a recognized condition where colonoscopy is mandatory due to the high likelihood of a positive finding [
      • Tripodi M.
      • Adinolfi L.
      • Ragone E.
      • Durante-Mangoni E.
      • Fortunato R.
      • Iarussi D.
      • et al.
      Clinical features of Streptococcus gallolyticus endocarditis and its association with chronic liver disease: an underestimated risk factor.
      ,
      • Habib G.
      • Lancellotti P.
      • Antunes M.J.
      • Bongiorni M.G.
      • Casalta J.P.
      • Del Zotti F.
      • et al.
      ESC Guidelines for the management of infective endocarditis: the Task Force for the Management of Infective Endocarditis of the European Society of Cardiology (ESC).
      ]. Reflecting the current recommendations, our SGIE patients underwent colonoscopy more often during the IE-related hospitalization compared with EIE. We believe that a similar strategy should now be considered also for EIE, in order to avoid late presentations of CRC, often associated with a more advanced stage.
      It was interesting to observe that, similar to SGIE, EIE was also associated mostly with CRN, being adenomas more common than carcinomas. In contrast, other intestinal disorders, including inflammatory bowel diseases, were uncommon. Once again, this could allude to a role for these IE gut pathogens in intestinal carcinogenesis. An in vitro study suggested a role for E. faecalis in colonic cell mutagenesis and consequent malignant transformation [
      • Wang X.
      • Allen T.D.
      • May R.J.
      • Lightfoot S.
      • Houchen C.W.
      • Huycke M.M
      Enterococcus faecalis induces aneuploidy and tetraploidy in colonic epithelial cells through a bystander effect.
      ]. Bacterial and host factors can contribute to CRN development and we believe further studies are warranted to clarify these mechanisms.
      Beyond CRN, we observed an association between EIE and other cancers, which was not evident for SGIE. This finding is consistent with previous studies [
      • Fernández-Cruz A.
      • Muñoz P.
      • Sandoval C.
      • Fariñas C.
      • Gutiérrez-Cuadra M.
      • Pericás Pulido J.M.
      • et al.
      Infective endocarditis in patients with cancer: a consequence of invasive procedures or a harbinger of neoplasm? A prospective, multicenter cohort.
      ,
      • Sun L.M.
      • Wu J.N.
      • Lin C.L.
      • Day J.D.
      • Liang J.A.
      • Liou L.R.
      • et al.
      Infective Endocarditis and cancer risk: a population-based cohort study.
      ] and suggests that when IE is diagnosed in a cancer survivor, enterococcal etiology be specifically considered. In some of these cases, EIE may actually emerge as a marker of occult neoplasia. In a large, Danish population-based cohort study, 997 cancers were identified among 8445 IE patients with a median follow-up of 3.5 years. The risk of abdominal and hematological cancers was high soon after IE diagnosis (within the first 3 months) and remained higher than expected in the long-term follow-up (12 months) for abdominal cancer [
      • Thomsen R.W.
      • Farkas D.K.
      • Friis S.
      • Svaerke C.
      • Ording A.G.
      • Norgaard M.
      • et al.
      Endocarditis and risk of cancer: a Danish nationwide cohort study.
      ].
      Beyond feasibility, safety of colonoscopy in the IE setting, including management of anticoagulation or the timing of cardiac surgery, is a concern. In our experience, colonoscopy was safe in most patients, although a few complications were observed. A careful evaluation should be carried out for each patient, balancing benefits and risks of the exam and obtaining a clear informed consent.
      Our and others’ results [
      • Pericàs J.M.
      • Corredoira J.
      • Moreno A.
      • García-País M.J.
      • Falces C.
      • Rabuñal R.
      • et al.
      Relationship Between Enterococcus faecalis Infective endocarditis and colorectal neoplasm: preliminary results from a cohort of 154 patients.
      ] suggest that, when possible, colonoscopy should be considered in all patients with EIE in which no specific genitourinary diseases are diagnosed. Unfortunately, our study could not clarify if patients with genitourinary conditions should also undergo colonoscopy, as none of these patients did so in our hands. However, if patient conditions allow, offering colonoscopy to these patients could also be beneficial to detect asymptomatic pre-cancerous lesion or an early stage CRC.
      This study has some limitations: it was a retrospective study and colonoscopy was not systematically performed, so endoscopic findings were not available for all patients; a selection bias towards patients who actually received colonoscopy cannot be excluded; also, the study group was relatively small. Notwithstanding, we add evidence to the relationship between EIE and CRN.
      In conclusion, EIE seems to be associated to CRN as much as SGIE, and should be considered a marker of CRN. Colonoscopy is generally safe in IE patients, and can be performed after careful case by case evaluation.

      Funding sources

      This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

      Author contributions

      All authors contributed to the study conception and design. Data collection and analysis were performed by Maria Paola Ursi, Lorenzo Bertolino, Roberto Andini, Fabiana D'Amico, Domenico Iossa, Arta Karruli, Eugenio D'Avenia, Sabrina Manduca, Mariano Bernardo. The first draft of the manuscript was written by Maria Paola Ursi, Lorenzo Bertolino, Rosa Zampino and Emanuele Durante Mangoni, and all authors commented on previous versions of the manuscript. All authors read and approved the final manuscript.

      Declarations of competing interest

      None

      Appendix. Supplementary materials

      References

        • Durante-Mangoni E.
        • Bradley S.
        • Selton-Suty C.
        • Tripodi M.F.
        • Barsic B.
        • Bouza E.
        • et al.
        Current Features of Infective Endocarditis in Elderly Patients Results of the International Collaboration on Endocarditis Prospective Cohort Study.
        Arch Intern Med. 2008; 168: 2095-2103https://doi.org/10.1001/archinte.168.19.2095
        • Murdoch D.R.
        • Corey G.R.
        • Hoen B.
        • Miró J.M.
        • Fowler Jr, V.G.
        • Bayer A.S
        Clinical presentation, etiology, and outcome of infective endocarditis in the 21st century: the International Collaboration on Endocarditis-Prospective Cohort Study.
        Arch Intern Med. 2009; 169: 463-473https://doi.org/10.1001/archinternmed.2008.603
        • Chirouze C.
        • Athan E.
        • Alla F.
        • Chu V.H.
        • Corey G.R.
        • Selton-Suty C.
        • et al.
        Enterococcal endocarditis in the beginning of the 21st century: analysis from the International Collaboration on Endocarditis-Prospective Cohort Study.
        Clin Microbiol Infect. 2013; 19: 1140-1147https://doi.org/10.1111/1469-0691.12166
        • Anderson D.J.
        • Olaison L.
        • McDonald J.R.
        • Miro J.M.
        • Hoen B.
        • Selton-Suty C.
        • et al.
        Enterococcal prosthetic valve infective endocarditis: report of 45 episodes from the International Collaboration on Endocarditis-merged database.
        Eur J Clin Microbiol Infect Dis. 2005; 24: 665-670https://doi.org/10.1007/s10096-005-0007-9
        • Tripodi M.
        • Adinolfi L.
        • Ragone E.
        • Durante-Mangoni E.
        • Fortunato R.
        • Iarussi D.
        • et al.
        Clinical features of Streptococcus gallolyticus endocarditis and its association with chronic liver disease: an underestimated risk factor.
        Clin Infect Dis. 2004; 38: 1394-1400https://doi.org/10.1086/392503
        • Klein R.S.
        • Recco R.A.
        • Catalano M.T.
        • Edberg S.C.
        • Casey J.I.
        • Steigbigel N.H
        Association of Streptococcus bovis with carcinoma of the colon.
        N Engl J Med. 1977; 297: 800-802https://doi.org/10.1056/NEJM197710132971503
        • Fernandez-Guerrero M.L.
        • Goyenechea A.
        • Verdejo C.
        • Roblas R.F.
        • de Górgolas M
        Enterococcal endocarditis on native and prosthetic valves. a review of clinical and prognostic factors with emphasis on hospital-acquired infections as a major determinant of outcome.
        Medicine (Baltimore). 2007; 86: 363-377https://doi.org/10.1097/MD.0b013e31815d5386
        • Corredoira M.J.
        • García-País M.J.
        • Coira A.
        • Rabuñal R.
        • García-Garrote F.
        • Pita J.
        • et al.
        Differences between endocarditis caused by Streptococcus gallolyticus and Enterococcus spp. and their association with colorectal cancer.
        Eur J Clin Microbiol Infect Dis. 2015; 34: 1657-1665https://doi.org/10.1086/392503
        • Pericàs J.M.
        • Corredoira J.
        • Moreno A.
        • García-País M.J.
        • Falces C.
        • Rabuñal R.
        • et al.
        Relationship Between Enterococcus faecalis Infective endocarditis and colorectal neoplasm: preliminary results from a cohort of 154 patients.
        Rev Esp Cardiol (Engl Ed). 2017; 70: 451-458https://doi.org/10.1016/j.rec.2016.10.013
        • Athan E.
        • Cabiltes I.
        • Coghill S.
        • Bowe S.J
        Silent but deadly: patients with enterococcal bacteraemia should be assessed for colorectal neoplasia.
        Med J Aust. 2019; 210: 86https://doi.org/10.5694/mja2.12027
        • Li J.S.
        • Sexton D.J.
        • Mick N.
        • Nettles R.
        • Fowler Jr, V.G.
        • Ryan T.
        • et al.
        Proposed modifications to the Duke criteria for the diagnosis of infective endocarditis.
        Clin Infect Dis. 2000; 30: 633-638https://doi.org/10.1086/313753
        • Cabiltes I.
        • Coghill S.
        • Bowe S.J.
        • Athan E
        Enterococcal bacteraemia "silent but deadly": a population-based cohort study.
        Intern Med J. 2019; https://doi.org/10.1111/imj.14396
        • Escolà-Vergé L.
        • Peghin M.
        • Givone F.
        • Pérez-Rodríguez M.T.
        • Suárez-Varela M.
        • Meije Y.
        • et al.
        Prevalence of colorectal disease in Enterococcus faecalis infective endocarditis: results of an observational multicenter study.
        Rev Esp Cardiol (Engl Ed). 2019; (pii:S1885-5857(19)30222-1)https://doi.org/10.1016/j.rec.2019.07.007
        • Habib G.
        • Lancellotti P.
        • Antunes M.J.
        • Bongiorni M.G.
        • Casalta J.P.
        • Del Zotti F.
        • et al.
        ESC Guidelines for the management of infective endocarditis: the Task Force for the Management of Infective Endocarditis of the European Society of Cardiology (ESC).
        Eur Heart J. 2015; 36: 3075-3128https://doi.org/10.1093/eurheartj/ehv319
        • Wang X.
        • Allen T.D.
        • May R.J.
        • Lightfoot S.
        • Houchen C.W.
        • Huycke M.M
        Enterococcus faecalis induces aneuploidy and tetraploidy in colonic epithelial cells through a bystander effect.
        Cancer Res. 2008; 68: 9909-9917https://doi.org/10.1158/0008-5472.CAN-08-1551
        • Fernández-Cruz A.
        • Muñoz P.
        • Sandoval C.
        • Fariñas C.
        • Gutiérrez-Cuadra M.
        • Pericás Pulido J.M.
        • et al.
        Infective endocarditis in patients with cancer: a consequence of invasive procedures or a harbinger of neoplasm? A prospective, multicenter cohort.
        Medicine (Baltimore). 2017; 96: e7913https://doi.org/10.1097/MD.0000000000007913
        • Sun L.M.
        • Wu J.N.
        • Lin C.L.
        • Day J.D.
        • Liang J.A.
        • Liou L.R.
        • et al.
        Infective Endocarditis and cancer risk: a population-based cohort study.
        Medicine (Baltimore). 2016; 95: e3198https://doi.org/10.1097/MD.0000000000003198
        • Thomsen R.W.
        • Farkas D.K.
        • Friis S.
        • Svaerke C.
        • Ording A.G.
        • Norgaard M.
        • et al.
        Endocarditis and risk of cancer: a Danish nationwide cohort study.
        Am J Med. 2013; 126: 58-67https://doi.org/10.1016/j.amjmed.2012.07.026