Abstract
Background:
Although Clostridioides difficile surveillance often identifies emerging strains, clinical outcome evaluations are rarely performed. Ribotype (RT) 106 is a commonly isolated C. difficile strain worldwide; however, studies investigating RT 106 clinical outcomes are limited. The purpose of this study was to investigate clinical outcomes of RT 106 infections compared with two other endemic strains of varying virulence.
Methods:
This multicenter study evaluated adults hospitalized with C. difficile infection (CDI). C. difficile samples underwent PCR ribotyping and patients infected with RT 106 were compared to patients infected with a known hypervirulent strain (RT 027) and a strain associated with less virulence (RT 014-020). Electronic medical records were reviewed by blinded investigators to assess the primary outcome of poor clinical outcome (composite of initial clinical failure, discharge to a higher level of care, 90-day CDI recurrence, and CDI-contributable mortality).
Results:
A total of 396 patients with CDI were identified (RT 106, 32.3%; RT 027, 29.3%; RT 014-020, 38.3%). Patients infected with RT 014-020 less often experienced a poor clinical outcome (40%) compared with RT 106 (56%) and RT 027 (65%) infection (P<0.0001). After controlling for covariates and using RT 014-020 as a comparator, patients infected with RT 106 (OR, 2.25; 95% CI, 1.36-3.73) or RT 027 (OR, 2.56; 95% CI, 1.52-4.31) had higher odds of poor clinical outcome. Using RT 027 as the comparator, only RT 014-020 was associated with lower odds of poor clinical outcome (OR, 0.42; 95% CI, 0.27-0.65).
Conclusion:
This study demonstrated that the emergent C. difficile RT 106 was associated with increased rates of poor clinical outcomes compared to RT 014-020 and comparable poor clinical outcomes to RT 027. These findings can help to better understand the clinical significance of this and future emerging ribotypes.
Keywords: Clostridium difficile, ribotype 027, ribotype 014-020, surveillance, typing
Background
Clostridioides difficile infection (CDI) is a major public health threat in the United States (US) and internationally [1]. CDI occurs following a disruption of the host microbiome, which allows for C. difficile spore germination and production of toxins A and B. These toxins mediate disease symptoms ranging from mild diarrhea to death [2, 3].
Although CDI is more likely to be caused by a strain endemic in the local environment, outbreaks with epidemic strains are often associated with increased disease incidence and severity [4]. Multiple C. difficile strain typing techniques are used to distinguish these strains, including polymerase chain reaction (PCR) ribotyping [5]. PCR ribotyping was used to identify ribotype (RT) 027 strain which caused an epidemic in the mid 2000’s [6]. Notably, infection with RT 027 was associated with worse outcomes including higher rates of clinical failure, disease recurrence, and CDI-contributable mortality, and a higher likelihood of hospital discharge to a higher level of care [7, 8].
Rates of RT 027 infection have decreased considerably since the peak of the epidemic but overall CDI incidence rates have remained higher than prior to the RT 027 epidemic [9]. A novel strain, RT 106, has increased in prevalence in the US to become the most prevalent cause of community-associated (CA) CDI and the second most prevalent cause of healthcare-associated (HA) CDI [10]. Although initially identified in the United Kingdom, RT 106 has disseminated internationally and is the third most common ribotype (11.6%) in Texas, with RT 027 (17.5%) and RT 014-020 (16.1%) being the other two of the three most common [11, 12].
Although in vitro data exists describing RT 106, clinical outcomes associated with this ribotype are largely unknown. Previously we demonstrated that RT 014-020 infection was associated with significantly better outcomes than RT 027 [13]. This provided the background to develop the framework utilized here to assess clinical outcomes associated with emerging C. difficile ribotypes. Using RT 014-020 and RT 027 as strains associated with better and worse outcomes, respectively, we collected clinical data using blinded evaluators to determine differences in clinical outcomes between these three ribotypes. The objective of this study was to evaluate the clinical outcomes of RT 106 infection compared to infection with RT 014-020 and RT 027. We hypothesized that CDI caused by RT 106 would be associated with more favorable clinical outcomes than RT 027 and worse than RT 014-020.
Methods
Study design
A retrospective study was conducted including patients infected with C. difficile RTs 106, 027, and 014-020 from two hospital systems (20 hospitals) in Houston, TX between 2016 to 2019. CDI was defined as the presence of liquid stool and a stool test positive for C. difficile through PCR testing or toxin A/B through enzyme immunoassay (EIA). Discard patient stool samples were collected and ribotyped as part of an ongoing surveillance study [12]. All patients infected with RTs 106, 027, and 014-020 in the three-year study period were included. Of note, our strain typing technique does not distinguish between RT 014 and 020 and so patients infected with either of those ribotypes were grouped. Clinical data present at the time of stool collection was retrieved from electronic medical records by investigators blinded to C. difficile RT. This study was approved by the Committee for the Protection of Human Subjects at the University of Houston (CPHS000128).
Definitions
CDI severity and epidemiologic category were classified in accordance with the 2017 Infectious Diseases Society of America (IDSA)/Society for Healthcare Epidemiology (SHEA) C. difficile guidelines [2]. Severe and fulminant disease were grouped as one category for the purposes of our analysis. Epidemiologic definitions, including community-onset (CO) CDI and hospital-onset (HO) CDI were defined in accordance with the CDC multidrug-resistant organism and CDI module [14].
Poor clinical outcome was evaluated as a composite endpoint that included initial clinical failure, discharge to a higher level of care, 90-day CDI recurrence, and CDI-contributable mortality. Initial clinical failure was defined as having continued CDI symptoms on day 6 or later of treatment or documented treatment failure anytime prior to day 6 [15]. CDI recurrence was assessed 90 days after a positive toxin test as previously defined [16]. CDI-contributable mortality was defined as patient mortality within 30 days of CDI diagnosis and with at least one of the following clinical parameters present at time of death: diarrhea, fever, leukocytosis, and/or radiographic evidence of colitis [17]. Discharge to a higher level of care was determined relative to the residence of each patient prior to admission and assessed at hospital discharge as previously described [8].
Statistical analysis
First, likelihood of poor clinical outcomes was assessed between the three ribotypes using the Mantel Haenszel Chi square statistic. Independent predictors of poor prognosis were identified using the following covariates: age (in 1-year increments), sex, race/ethnicity, Charlson Comorbidity Index (CCI) score (in 1-point increments), C. difficile epidemiologic classification, lifetime CDI episode number (1, 2, or ≥3), CDI severity, and use of proton pump inhibitors (PPI). Those variables with a p<0.2 in univariate analyses were included in multivariable analysis to identify independent predictors associated with CDI poor clinical outcomes. All variables with a p<0.05 were included in the final model.
Next, two multivariable models were built using the covariates identified in the univariate/multivariable analyses to determine poor clinical outcomes associated with RT 106. The two models used RT 014-020 or RT 027 as the baseline ribotype comparators to be able to obtain a lower (RT 014-020) and upper boundary (RT 027) of poor clinical outcomes associated with ribotypes. Odds ratios (OR) and 95% confidence intervals (CI) were estimated. Those variables with p <0.05 were defined as statistically significant. All analyses were performed using SPSS software version 26 (IBM Corp., Armonk, NY) or SAS version 9.4 (SAS Institute, Cary NC).
Results
A total of 396 patients with CDI aged 66±16 (mean±SD) years were included (60% female, 71% Caucasian, 52% with PPI use) (Table 1). Of these, 128 (32%) patients were infected with RT 106, 116 (29%) with RT 027, and 152 (38%) with RT 014-020. Seventy two (18%) patients experienced initial clinical failure, 54 (14%) experienced 90-day CDI recurrence, 133 (34%) were discharged to a higher level of care, and 26 (7%) experienced CDI-contributable mortality. In total, 208 (53%) patients experienced a poor clinical outcome.
Table 1.
Clinical variables associated with poor clinical outcome in hospitalized patient with CDI
| Poor Clinical Outcome* | |||
|---|---|---|---|
| No (n=188) | Yes (n=208) | P value | |
| Age (yr), mean (SD) | 63.2 (±16.88) | 68.7 (±14.04) | <0.001 |
| Female sex (n=239), no. (%) | 118 (63%) | 121 (58%) | 0.25 |
| Race, no. (%) | 0.43 | ||
| White (n=278) | 127 (68%) | 151 (73%) | |
| Black (n=83) | 42 (23%) | 41 (20%) | |
| Other (n=35) | 19 (10%) | 16 (8%) | |
| Ethnicity, no. (%) | 0.33 | ||
| Hispanic (n=64) | 34 (18%) | 30 (14%) | |
| Not Hispanic (n=332) | 154 (82%) | 178 (86%) | |
| CCI score, mean (SD) | 2.52 (±2.14) | 3.22 (±2.20) | <0.001 |
| PPI use (n=205), no (%) | 91 (48.4%) | 114 (54.8%) | 0.20 |
| Residence prior to admission, no (%) | 0.06 | ||
| Home (n=310) | 157 (8%) | 153 (7%) | |
| LTCF or SNF (n=62) | 22 (12%) | 40 (19%) | |
| Transfer from another healthcare facility (n=24) | 9 (5%) | 15 (7%) | |
| CDI episode no. (%) | 0.09 | ||
| 1 (n=292) | 147 (78%) | 145 (70%) | |
| 2 (n=64) | 28 (15%) | 36 (17%) | |
| ≥3 (n=40) | 13 (7%) | 27 (13%) | |
| Hospital-onset CDI (n=147), no. (%) | 57 (30%) | 90 (43%) | 0.01 |
| Severe/fulminant CDI (n=207), no. (%) | 87 (46%) | 120 (58%) | 0.02 |
Poor clinical outcome defined as a composite of initial clinical failure, discharge to a higher level of care, 90-day CDI recurrence, and CDI-contributable mortality.
Abbv: CCI, Charlson Comorbitiy Index score; PPI, proton pump inhibitor use at time of diagnosis; CDI, C. difficile infection
Poor clinical outcome occurred least often in patients infected with RT 014-020 strains (40%) followed by RT 106 (56%) and RT 027 (65%) (p<0.0001). The frequency of the composite endpoint and sub-components per ribotype are shown in Figure 1. Univariate analyses of other variables predictive of poor clinical outcome are shown in Table 1.
Figure 1.
Poor clinical outcome by composite endpoint and sub-components for ribotype 106 vs. comparators. Poor clinical outcome occurred least often in patients infected with RT 014-020 strains (40%) followed by RT 106 (56%) and RT 027 (65%) (p<0.0001).
In addition to ribotype, other variables included in the multivariable model included age, CCI score, PPI use, residence prior to admission, lifetime CDI episode number, HO-CDI, and severe/fulminant CDI. Using a forward selection process, age (OR, 1.02; 95% CI, 1.01-1.04), CCI (OR, 1.14; 95% CI, 1.03-1.26), CDI episode number (OR, 1.63; 95% CI, 1.15-2.31), and HO-CDI (OR, 1.94; 95% CI, 1.24-3.03) remained in the model. After controlling for these variables and using RT 014-020 as a comparator, patients infected with both RT 106 (OR, 2.25; 95% CI, 1.36-3.73) or RT 027 (OR, 2.56; 95% CI, 1.52-4.31) had significantly higher odds of poor clinical outcome than patients infected with RT 014-020. In the second model using RT 027 as the comparator, RT 014-020 was associated with significantly lower odds of poor clinical outcome (OR, 0.42; 95% CI, 0.27-0.65). Compared with RT 014-020, infection with RT 106 was no different than infection RT 027 in predicting poor clinical outcome and was not included in the final model.
Discussion
The recent RT 027 epidemic has heightened the awareness and importance of monitoring for new, novel strains of C. difficile as they appear in the community. In the USA, routine surveillance is done using the CDC Emerging Infections Program to assess for emerging strain types [9]. In Texas, we have performed state-wide monitoring for several years using PCR-ribotyping to type C. difficile strains using a surveillance system we have named CaCTUS [12]. Surveillance systems are generally limited in that they can detect emerging strains, but the clinical significance of these strains is generally unknown. We used our previous experience comparing RT 027 to other circulating strains to develop methodology to assess the clinical virulence of RT 106, an emergent ribotype in both the CDC and CaCTUS surveillance networks [13]. Using this methodology, we were able to determine that patients infected with RT 106 had higher rates of poor clinical outcome than those infected with RT 014-020 and comparable clinical outcomes to RT 027 infection. This study also validates our previous findings demonstrating RT 014-020 has decreased virulence compared to RT 027 and justifies the use of these two ribotypes to assess virulence of emerging ribotypes.
Strengths of this study included the large sample size and associated clinical data from each of the three ribotypes used in this study. The clinical data collected for this study is readily available through most electronic medical records. Standardized ribotyping methods are also largely available, allowing this framework to have widespread applicability for other emerging strains. Blinding the investigator to ribotype during the clinical data collection helped reduce any bias in ascertaining clinical outcomes. Finally, although the majority of cases were diagnosed utilizing PCR testing, our relatively high rates of severe disease, recurrence, and CDI-contributable mortality imply we included a low rate of asymptomatic colonization.
Previous clinical studies evaluating CDI patients infected with RT 106 are limited but support our findings [18, 19]. A study in the United Kingdom compared RT 106 and RT 027 as part of a larger study to assess clinical outcomes in patients with CDI compared to uninfected controls [18]. Although underpowered to assess clinical outcomes by ribotype, reported CDI severity rates were lower for RT 106 (8%) compared to RT 027 (16%). Reported 60-day CDI recurrence rates were also lower for RT 106 (8%) compared to RT 027 (11%). The second study assessed outcomes from CDI patients based on North American pulsed-field gel electrophoresis type (NAP) 1 (which includes RT 027) isolates compared to a group of non-NAP1 strains that included NAP11 (RT 106) and NAP4 (RTs 014, 020) [19]. Similar to our findings, NAP1 infection was associated with increased odds of severe disease, severe outcomes, and death within 14 days compared with non-NAP1 strains after adjusting for other factors. We feel both of these studies complement with the results of our study well and also help justify our use of RT 014-020 and RT 027 as comparator ribotypes to assess and compare emerging ribotypes.
Our study has several limitations. The clinical data were multicenter involving over 20 hospitals but was collected in a distinct geographic location. Validation of this methodology in other centers will be required. Patients were primarily diagnosed using a PCR test detecting toxin B so identification of asymptomatic colonization may have occurred [20]. However, in addition to the high rates of poor CDI-attributable outcomes, all patients grew toxigenic C. difficile of known virulent ribotypes making identification of asymptomatic patients less likely. Since ribotypes 014 and 020 were indistinguishable and combined, distinctions between clinical outcomes of these two ribotypes will require further study. A 30-day time period is often used for clinical drug trials; however, we chose a 90-day time period for CDI recurrence evaluation in order to capture any recurrences in that defined time period. Due to our collection method of leftover stool samples, recurrence samples were not routinely available for typing. For these models, we chose to include only variables present at the time of the stool collection date. Thus, antibiotics were not included in the model. Future models will incorporate treatment choices to assess if antibiotic therapy affects strain types differently. Finally, this was an observational study and all findings should be regarded as associated and not causal.
In conclusion, this study demonstrated that infection with C. difficile RT 106 was associated with an increased rate of poor clinical outcome compared to RT 014-020 infection and comparable to RT 027 infection. These findings help to better understand the clinical significance of this and future emerging ribotypes.
Highlights.
Clostridioides difficile ribotype (RT) 106 is an emergent strain in the United States with unknown clinical virulence
We developed a novel methodology to assess clinical outcomes across common ribotypes
Patients with RT 106 experienced poor clinical outcomes comparable to RT 027, a known hypervirulent strain
Financial support:
This study was funded by the National Institutes of Health (NIH) National Institute of Allergy and Infectious Diseases (NIAID) U01AI124290-01.
Footnotes
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Potential competing interests: None to declare
All authors declare no conflict of interest
References:
- 1.Magill SS, O’Leary E, Janelle SJ, et al. Changes in Prevalence of Health Care-Associated Infections in U.S. Hospitals. N Engl J Med 2018; 379(18): 1732–44. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.McDonald LC, Gerding DN, Johnson S, et al. Clinical Practice Guidelines for Clostridium difficile Infection in Adults and Children: 2017 Update by the Infectious Diseases Society of America (IDSA) and Society for Healthcare Epidemiology of America (SHEA). Clin Infect Dis 2018; 66(7): e1–e48. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Vedantam G, Clark A, Chu M, McQuade R, Mallozzi M, Viswanathan VK. Clostridium difficile infection: toxins and non-toxin virulence factors, and their contributions to disease establishment and host response. Gut Microbes 2012; 3(2): 121–34. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Bartlett JG. Clostridium difficile infection: historic review. Anaerobe 2009; 15(6): 227–9. [DOI] [PubMed] [Google Scholar]
- 5.Martinson JN, Broadaway S, Lohman E, et al. Evaluation of portability and cost of a fluorescent PCR ribotyping protocol for Clostridium difficile epidemiology. J Clin Microbiol 2015; 53(4): 1192–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.McDonald LC, Killgore GE, Thompson A, et al. An epidemic, toxin gene-variant strain of Clostridium difficile. N Engl J Med 2005; 353(23): 2433–41. [DOI] [PubMed] [Google Scholar]
- 7.Loo VG, Poirier L, Miller MA, et al. A predominantly clonal multi-institutional outbreak of Clostridium difficile-associated diarrhea with high morbidity and mortality. N Engl J Med 2005; 353(23): 2442–9. [DOI] [PubMed] [Google Scholar]
- 8.Reveles KR, Dotson KM, Gonzales-Luna A, et al. Clostridioides (Formerly Clostridium) difficile Infection During Hospitalization Increases the Likelihood of Nonhome Patient Discharge. Clin Infect Dis 2019; 68(11): 1887–93. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Guh AY, Mu Y, Winston LG, et al. Trends in U.S. Burden of Clostridioides difficile Infection and Outcomes. N Engl J Med 2020; 382(14): 1320–30. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Clostridioides difficile infection (CDI) tracking. Centers for Disease Control and Prevention. Available at: https://www.cdc.gov/hai/eip/Annual-CDI-Report-2017.html. [Google Scholar]
- 11.Carlson TJ, Blasingame D, Gonzales-Luna AJ, Alnezary F, Garey KW. Clostridioides difficile ribotype 106: A systematic review of the antimicrobial susceptibility, genetics, and clinical outcomes of this common worldwide strain. Anaerobe 2020; 62: 102142. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Gonzales-Luna AJ, Carlson TJ, Dotson KM, et al. PCR ribotypes of Clostridioides difficile across Texas from 2011 to 2018 including emergence of ribotype 255. Emerg Microbes Infect 2020; 9(1): 341–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Aitken SL, Alam MJ, Khaleduzzaman M, et al. In the Endemic Setting, Clostridium difficile Ribotype 027 Is Virulent But Not Hypervirulent. Infect Control Hosp Epidemiol 2015; 36(11): 1318–23. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Centers for Disease Control and Prevention. Clostridioides difficile Infection (CDI) Tracking. Available at: https://www.cdc.gov/hai/eip/cdiff-tracking.html#reports. Accessed June 3.
- 15.Zar FA, Bakkanagari SR, Moorthi KMLST, Davis MB. A Comparison of Vancomycin and Metronidazole for the Treatment of Clostridium difficile-Associated Diarrhea, Stratified by Disease Severity. Clinical Infectious Diseases 2007; 45(3): 302–7. [DOI] [PubMed] [Google Scholar]
- 16.Shah DN, Aitken SL, Barragan LF, et al. Economic burden of primary compared with recurrent Clostridium difficile infection in hospitalized patients: a prospective cohort study. J Hosp Infect 2016; 93(3): 286–9. [DOI] [PubMed] [Google Scholar]
- 17.Lin Q, Pollock NR, Banz A, et al. Toxin A-Predominant Pathogenic Clostridioides difficile: A Novel Clinical Phenotype. Clin Infect Dis 2020; 70(12): 2628–33. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Sundram F, Guyot A, Carboo I, Green S, Lilaonitkul M, Scourfield A. Clostridium difficile ribotypes 027 and 106: clinical outcomes and risk factors. J Hosp Infect 2009; 72(2): 111–8. [DOI] [PubMed] [Google Scholar]
- 19.See I, Mu Y, Cohen J, et al. NAP1 strain type predicts outcomes from Clostridium difficile infection. Clin Infect Dis 2014; 58(10): 1394–400. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Koo HL, Van JN, Zhao M, et al. Real-time polymerase chain reaction detection of asymptomatic Clostridium difficile colonization and rising C. difficile-associated disease rates. Infect Control Hosp Epidemiol 2014; 35(6): 667–73. [DOI] [PubMed] [Google Scholar]