Abstract
Purpose
This study aimed to evaluate the clinical significance of toxin positivity and toxin gene load, and the relation between them in the broad spectrum of Clostridium difficile infection (CDI) including colonization, significant diarrhea, and severe disease.
Methods
We included 2671 fecal samples submitted for CDI diagnosis and 180 samples from healthy individuals. The clinical spectrum was categorized as category I (toxigenic C. difficile positive without clinical CDI criteria), category II (mild CDI), and category III (severe CDI). Clinical parameters were compared based on toxin EIA and tcdB C t values. C t values of tcdB PCR for predicting toxin EIA positivity were assessed using receiver-operating characteristic (ROC) curves.
Results
The median C t values of tcdB PCR and toxin positivity were not significantly correlated with clinical spectrum of CDI (27.5, 28.2, and 26.1 for tcdB C t and 55.0, 56.6, and 60.9% for toxin EIA positivity in category I, II, and III, respectively, P > 0.05). There were significant differences in the tcdB C t values between toxin EIA-positive and -negative groups (P < 0.001). Optimal cutoff for the tcdB C t value for estimating toxin EIA positivity was 26.3 with 79.3% sensitivity and 83.6% specificity with good area under the curves (AUC, 0.848).
Conclusions
The C t values successfully predicted toxin EIA positivity and could be used as a surrogate for toxin EIA positivity in the diagnostic algorithm and routine analysis. Further studies are needed to validate the clinical significance of tcdB PCR C t value in toxigenic C. difficile colonization and infection.
Similar content being viewed by others
Avoid common mistakes on your manuscript.
Introduction
Clostridium difficile is a leading cause of antibiotic-associated colitis and imposes a great burden on the healthcare system globally [1, 2]. The clinical manifestations of C. difficile infection (CDI) have a broad spectrum ranging from mild diarrhea to toxic megacolon leading to colonic perforation and death. Moreover, they cannot be easily distinguished from other causes including other bacterial or viral infections, use of laxatives, or surgery, which makes an accurate laboratory diagnosis essential [3]. Due to insufficient sensitivity of the toxin enzyme immunoassay (EIA), many laboratories have switched to nucleic acid amplification testing (NAAT) with higher performance [4, 5]. However, use of NAAT as a single test is still questionable in terms of clinical significance and cost-effectiveness [3, 4, 6]. Importantly, C. difficile assays cannot distinguish asymptomatic colonization from symptomatic infection, whereas NAAT with its higher sensitivity detects more patients with asymptomatic colonization. Moreover, several studies show that a significant portion of patients with NAAT positive for C. difficile did not have clinically significant diarrhea [7, 8]. The experts and guidelines on this issue are still in conflict. The latest European guidelines still recommend the inclusion of toxin EIA in the diagnostic algorithm, whereas other guidelines recommend an NAAT-only approach [9,10,11]. The rationale of including toxin EIA is based on its correlation with clinical outcomes and CDI-related complications [3, 5]. This study aimed to evaluate the clinical significance of toxin positivity and toxin gene load, and the relation between them in the broad spectrum of CDI including colonization, significant CDI diarrhea, and severe CDI.
Methods
Study design and clinical samples
This study was approved by the Institutional Review Board of the Konkuk University Medical Center, Seoul, Korea (a tertiary referral hospital with 900 beds). This study included 2671 patients whose fecal samples were submitted to the clinical microbiology laboratory for C. difficile testing from March 2015 to October 2016. As a routine practice of C. difficile testing at our hospital, we performed the toxin EIA, and toxigenic culture (TC) and/or tcdB PCR. In addition, we included 180 residual stool samples after routine health care examination from asymptomatic healthy individuals and tested them with toxin EIA and tcdB PCR. This study required neither study-specific nor any other interventions. Therefore, written informed consent from enrolled patients was exempted.
Laboratory tests
For TC, alcohol-treated stool samples were inoculated onto a chromogenic agar plate (chromID CD agar, bioMérieux, Marcy-l’Etoile, France), and incubated at 37 °C under anaerobic conditions (Forma Anaerobic System; Thermo Fisher Scientific, Waltham, MA, USA). Isolates with typical morphology suspicious for C. difficile were confirmed by Gram-staining and finally identified by matrix-assisted laser desorption/ionization time-of-flight (MALDI-TOF) using the VITEK MS system (bioMérieux). Toxin production was confirmed by EIA or in-house PCR which was performed as described previously [12]. Toxin EIA was performed using VIDAS C. difficile toxin A&B according to the manufacturer’s instructions. The assay principle combined a two-step enzyme immunoassay with fluorescence detection (enzyme-linked fluorescence immunoassay, ELFA) and the results were reported as negative, equivocal, or positive with cutoffs of 0.13 and 0.37 test value (TV) according to the manufacturer’s instructions. Equivocal results of toxin EIA were considered negative in this study. The tcdB PCR was performed using the Xpert C. difficile system (Cepheid, Sunnyvale, CA, USA). This assay detects toxin B (tcdB), binary toxin (cdt), and a point mutation associated with PCR ribotype 027. Using a maximum valid cycle threshold (C t) setting of 37 for tcdB, the limit-of-detection point estimate for toxigenic C. difficile was 1,657 CFUs/swab according to the manufacturer’s instructions. We collected the C t data in all positive results.
Clinical data collection
Among 2671 samples, 366 samples (13.7%) were positive for TC. Among 884 samples tested with tcdB PCR, 161 (18.2%) were positive; 434 samples (16.2%) were positive for TC and/or tcdB PCR. After excluding 152 duplicated samples in the same episode, we reviewed the data of 282 patients in detail (Fig. 1). For these patients, we collected clinical data through chart review including demographic data and laboratory data including white cell count, serum creatinine, and albumin concentrations tested within 3 days of fecal sample collection. We obtained the baseline serum creatinine concentrations from tests performed more than 6 months before study entry.
The underlying diseases of these patients included malignancy (33.0%), respiratory disease (16.7%), acute infectious diseases (11.0%), cerebral- and cardio-vascular disease (11.0%), surgical disorder (5.3%), and neurologic disease (5.3%). Of the patients, 67.7% had undergone antibiotic therapy and most (95.7%) were treated with metronidazole for CDI. The clinical spectrum was categorized as category I (toxigenic C. difficile positive without clinical CDI criteria, < 3 unformed stools in 24 h), mild CDI (≥ 3 unformed stools in 24 h and WBC count less than 15 × 109 cells/L, and serum creatinine less than 1.5 times the premorbid level, category II), and severe CDI (≥ 3 unformed stools in 24 h and WBC count of at least 15 × 109 cells/L or serum creatinine levels at least 1.5 times the premorbid level, category III). CDI-related complications included radiologically confirmed toxic megacolon status or hypovolemic shock with low blood pressure [2, 4]. Patients with laxative use in 48 h before sample collection were excluded from category II and III [13]. The duration of hospitalization and the 30-day all-cause mortalities were assessed as the clinical outcomes. The tcdB PCR-positive patients were divided as high C t and low C t based on the median C t value (C t, 27.1). Clinical characteristics were compared between these groups.
Statistical analysis
The difference between the continuous variables was analyzed using the Student’s t test or Mann–Whitney U test, and that between categorical variables was analyzed using the Chi-square test, Fisher’s exact test, McNemar test, or the Cochran–Armitage test for trends. The Kruskal–Wallis test and one-way analysis of variance (ANOVA) followed by the Games–Howel’s post hoc test was used to assess the differences between groups. The C t values of tcdB PCR for estimating toxin EIA positivity were assessed using receiver-operating characteristic (ROC) curves and the area under the curves (AUC). Statistical analysis was performed using MedCalc Statistical Software (version 15.8, MedCalc Software, Mariakerke, Belgium) and IBM SPSS Statistics 22.0 (IBM Corporation, Armonk, NY, USA). P values less than 0.05 were considered statistically significant.
Results
Among 282 patients positive for TC and/or tcdB PCR, 119, 99, and 64 patients were category I, II, and III, respectively. The results of toxin EIA and tcdB PCR in each clinical category are described in supplemental Table 1. The median C t values of tcdB PCR and toxin positivity were not significantly correlated with clinical spectrum of CDI (27.5, 28.2, and 26.1 for tcdB PCR and 55.0, 56.6, and 60.9% for toxin EIA positivity in category I, II, and III, respectively, P > 0.05). There were significant differences in age between category II and III (61.5 vs. 67.9 years old, P = 0.0186). Comparison of clinical characteristics between groups based on results of the toxin EIA and tcdB PCR in patients with TC and/or tcdB PCR positivity is described in supplemental Table 2. Age and WBC count were significantly different between groups based on toxin EIA results (mean age, 60.7 vs. 68.0 years old, P = 0.0016 and mean WBC count, 8.7 vs. 11.3 109/L, P < 0.0001, respectively).
The distribution of tcdB PCR C t value between groups divided by toxin EIA results is presented in Fig. 2. There were significant differences in the tcdB C t values between toxin EIA-positive and -negative groups (P < 0.001). The optimal cutoff for the tcdB C t value for predicting toxin EIA positivity was established by the ROC curve analysis. The optimal tcdB C t cutoff was 26.3 with 79.3% (95% CI 65.9–89.2) sensitivity, and 83.6% (95% CI 71.2–92.2) specificity with high AUC (0.848, 95% CI 0.767–0.910) (Fig. 3). Among the 180 healthy individuals tested, 8 (4.4%) were positive for toxin EIA and/or tcdB PCR. The median C t value of the tcdB PCR-positive healthy group was 32.9 (IQR 30.1–35.2), which was significantly higher than category I, II, and III (C t, 27.5, 28.2, and 26.1; P, 0.0032, 0.0171, and 0.0004, respectively).
Discussion
In this study, we divided the toxigenic C. difficile-positive patients to category I, II, and III adopting the classical clinical criteria. The median C t values of tcdB PCR and rates of toxin positivity were not significantly correlated with clinical spectrum of CDI (27.5, 28.2, and 26.1 for tcdB PCR and 55.0, 56.6, and 60.9% for toxin EIA positivity in category I, II, and III, respectively, P > 0.05). Clinical criteria were based on laboratory test and the number of diarrhoeal stools in the medical chart. It can be influenced by the accuracy of medical chart records and there are many factors including drugs, other infections, or host factors that make correct categorization difficult. No reliable laboratory tests accurately distinguish CDI from C. difficile colonization and there is also no definite clinical difference between true CDI and non-CDI-related symptoms [9, 14, 15]. Moreover, a very high proportion of laxative therapy in patients diagnosed with CDI has been reported, making clinical criteria based on diarrhoeal stools impractical [7, 13, 16]. We can expect considerable overlap between category I and category II. However, the median C t value of the tcdB PCR in toxigenic C. difficile-positive healthy individuals was 32.9, which was significantly higher than those of category I, II, and III. These data suggest significant lower tcdB gene load in asymptomatic colonization. We included only 8 toxigenic C. difficile-positive individuals among 180 screened subjects. Further study including sufficient number of toxigenic C. difficile-positive healthy individuals would be valuable to assess C t value for differentiation between colonization and significant CDI.
There were significant differences in age between the mild and severe form of CDI (II vs. III) (P = 0.0186) (supplemental Table 1). Similarly, there were significant differences for age and WBC count between toxin EIA-negative and -positive groups (P = 0.0016 and P < 0.0001) (supplemental Table 2). The WBC count might be related to older age in toxin positive groups, suggesting more severe clinical findings in older groups. Regarding clinical impact of toxin positivity, there are different results between studies and more controlled study is needed to elucidate this issue [3, 8, 17].
Currently, NAAT has become the predominant method for CDI diagnosis and some experts prefer an NAAT-only approach [10, 11, 18]. Nevertheless, recent studies have shown the clinical importance of toxin positivity [3, 5] and toxin EIA is still recommended in updated guidelines as the second step, a highly specific test in a two-step algorithm [9]. Positive results of toxin EIA have two meanings including positive free toxin in feces and organism burden above the threshold of toxin EIA [9]. As the sensitivity of toxin EIA is lower than that of NAAT, the clinical correlation of positive toxin EIA could reflect the impact of higher organism burden compared to that of positive NAAT. The C t value of real-time PCR or DNA copies can also reflect the organism burden and has shown a correlation with toxin EIA [5, 19]. Some recent studies evaluated the clinical impact of C t value in CDI [20, 21]. In this study, there was a significant difference in the C t value between toxin EIA-negative and -positive groups (median 29.5 vs 24.2, P < 0.001). Using the ROC curve analysis of C t value for detecting toxin EIA positivity, a good AUC value of 0.848 was obtained with the optimal C t cutoff (26.3). Using an algorithm with additional toxin, EIA steps could require additional time, cost for the additional step, and result in diagnostic delay. Based on our results, it could be replaced by using C t values with no additional cost. The optimal cutoff from ROC analysis is close to median C t value of all positive results (C t, 26.3 vs. 27.1), and could be applied to each laboratory. This cutoff is also very close to that of recent study on this issue (26.35), indicating easy adaptation to other laboratory [22]. All C t values of tcdB PCR-positive results in healthy individuals have shown an above optimal cutoff (26.3).
This study has its strengths and limitations. We analyzed the clinical values of toxin EIA positivity and tcdB gene load PCR in the broad spectrum of CDI ranging from healthy carriers to severe CDI. In addition, based on our results and published studies, it could be possible to adopt the C t value instead of the toxin EIA into the algorithm and routine analysis. The limitation of this study is the difference in the number of tested patients between toxin EIA and tcdB PCR. All patients were tested with TC and toxin EIA, but only 33% were tested with tcdB PCR. The lower number of tcdB PCR positivity compared to toxin EIA positivity can affect the statistical significance of the parameters between the low and high C t groups. Further studies are needed to validate the values of C t of tcdB PCR.
In conclusion, the C t values successfully predicted toxin EIA positivity, and could be used as a surrogate for toxin EIA positivity in the diagnostic algorithm and routine analysis. Further studies are needed to validate the clinical significance of tcdB PCR C t value in toxigenic C. difficile colonization and infection.
References
Martin JS, Monaghan TM, Wilcox MH. Clostridium difficile infection: epidemiology, diagnosis and understanding transmission. Nat Rev Gastroenterol Hepatol. 2016;13:206–16.
Cohen SH, Gerding DN, Johnson S, Kelly CP, Loo VG, McDonald LC, et al. Clinical practice guidelines for Clostridium difficile infection in adults: 2010 update by the society for healthcare epidemiology of America (SHEA) and the infectious diseases society of America (IDSA). Infect Control Hosp Epidemiol. 2010;31:431–55.
Planche TD, Davies KA, Coen PG, Finney JM, Monahan IM, Morris KA, et al. Differences in outcome according to Clostridium difficile testing method: a prospective multicentre diagnostic validation study of C. difficile infection. Lancet Infect Dis. 2013;13:936–45.
Bagdasarian N, Rao K, Malani PN. Diagnosis and treatment of Clostridium difficile in adults: a systematic review. JAMA. 2015;313:398–408.
Polage CR, Gyorke CE, Kennedy MA, Leslie JL, Chin DL, Wang S, et al. Overdiagnosis of Clostridium difficile infection in the molecular test era. JAMA Intern Med. 2015;175:1792–801.
Debast SB, Bauer MP, Kuijper EJ. European Society of Clinical Microbiology and Infectious Diseases: update of the treatment guidance document for Clostridium difficile infection. Clin Microbiol Infect. 2014;20:1–26.
Banaei N, Anikst V, Schroeder LF. Burden of Clostridium difficile infection in the United States. N Engl J Med. 2015;372:2368–9.
Anikst VE, Gaur RL, Schroeder LF, Banaei N. Organism burden, toxin concentration, and lactoferrin concentration do not distinguish between clinically significant and nonsignificant diarrhea in patients with Clostridium difficile. Diagn Microbiol Infect Dis. 2016;84:343–6.
Crobach MJ, Planche T, Eckert C, Barbut F, Terveer EM, Dekkers OM, et al. European Society of Clinical Microbiology and Infectious Diseases: update of the diagnostic guidance document for Clostridium difficile infection. Clin Microbiol Infect. 2016;22:S63–81.
Surawicz CM, Brandt LJ, Binion DG, Ananthakrishnan AN, Curry SR, Gilligan PH, et al. Guidelines for diagnosis, treatment, and prevention of Clostridium difficile infections. Am J Gastroenterol. 2013;108:478–98.
Fang FC, Polage CR, Wilcox MH. Point-counterpoint: what is the optimal approach for detection of Clostridium difficile infection? J Clin Microbiol. 2017;55:670–80.
Terhes G, Urban E, Soki J, Hamid KA, Nagy E. Community-acquired Clostridium difficile diarrhea caused by binary toxin, toxin A, and toxin B gene-positive isolates in Hungary. J Clin Microbiol. 2004;42:4316–8.
Truong CY, Gombar S, Wilson R, Sundararajan G, Tekic N, Holubar M, et al. Real-time electronic tracking of diarrheal episodes and laxative therapy enables verification of Clostridium difficile clinical testing criteria and reduction of Clostridium difficile infection rates. J Clin Microbiol. 2017;55:1276–84.
Burnham CA, Carroll KC. Diagnosis of Clostridium difficile infection: an ongoing conundrum for clinicians and for clinical laboratories. Clin Microbiol Rev. 2013;26:604–30.
Burnham CA, Dubberke ER, Kociolek LK, Polage CR, Riley TV. Clostridium difficile-diagnostic and clinical challenges. Clin Chem. 2016;62:310–4.
Polage CR, Solnick JV, Cohen SH. Nosocomial diarrhea: evaluation and treatment of causes other than Clostridium difficile. Clin Infect Dis. 2012;55:982–9.
Banaei N, Schroeder LF. Toxin Immunoassays and Clostridium difficile Infection. JAMA Intern Med. 2016;176:413.
Wilcox MH, Planche T, Fang FC, Gilligan P. What is the current role of algorithmic approaches for diagnosis of Clostridium difficile infection? J Clin Microbiol. 2010;48:4347–53.
Dionne LL, Raymond F, Corbeil J, Longtin J, Gervais P, Longtin Y. Correlation between Clostridium difficile bacterial load, commercial real-time PCR cycle thresholds, and results of diagnostic tests based on enzyme immunoassay and cell culture cytotoxicity assay. J Clin Microbiol. 2013;51:3624–30.
Jazmati N, Hellmich M, Licanin B, Plum G, Kaasch AJ. PCR cycle threshold value predicts the course of Clostridium difficile infection. Clin Microbiol Infect. 2016;22:e7–8.
Truong C, Schroeder LF, Gaur R, Anikst VE, Komo I, Watters C, et al. Clostridium difficile rates in asymptomatic and symptomatic hospitalized patients using nucleic acid testing. Diagn Microbiol Infect Dis. 2017;87:365–70.
Senchyna F, Gaur RL, Gombar S, Truong CY, Schroeder LF, Banaei N. Clostridium difficile PCR cycle threshold predicts free toxin. J Clin Microbiol. 2017;55:2651–60.
Acknowledgements
This work was supported by the Konkuk University Medical Center Research Grant 2017.
Author information
Authors and Affiliations
Corresponding author
Ethics declarations
Conflict of interest
The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.
Electronic supplementary material
Below is the link to the electronic supplementary material.
Rights and permissions
About this article
Cite this article
Kim, H.N., Kim, H., Moon, HW. et al. Toxin positivity and tcdB gene load in broad-spectrum Clostridium difficile infection. Infection 46, 113–117 (2018). https://doi.org/10.1007/s15010-017-1108-y
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s15010-017-1108-y