Systematic Review on the Correlation of Quantitative PCR Cycle Threshold Values of Gastrointestinal Pathogens With Patient Clinical Presentation and Outcomes

Background: Quantitative (q) polymerase chain reaction (PCR) cycle threshold (Ct) values represent the number of amplification cycles required for a positive PCR result and are a proxy of pathogen quantity in the tested sample. The clinical utility of Ct values remains unclear for gastrointestinal infections. Objectives: This systematic review assesses the global medical literature for associations between Ct values of gastrointestinal pathogens and patient presentation and clinical outcomes. Data Sources: MEDLINE, EMBASE, Cochrane library databases: searched January 14–17, 2020. Study Eligibility Criteria: Studies reporting on the presence or absence of an association between Ct values and clinical outcomes in adult and pediatric populations were included. Animal studies, reviews, meta-analyses, and non-English language studies were excluded. Participants: Humans infected with gastrointestinal pathogens, detected with qPCR. Interventions: Diagnostics assessing Ct values. Extracted data were reported narratively. Results: Thirty-three eligible studies were identified; the most commonly studied pathogens were Clostridioides difficile (n = 15), norovirus (n = 10), and rotavirus (n = 9). Statistically significant associations between low C. difficile Ct values and increased symptom severity or poor outcome were reported in 4/8 (50%) studies, and increased risk of death in 1/2 (50%) studies; no significant associations were found between Ct value and duration of symptoms or length of hospital stay. Among studies of norovirus, 5/7 (71%), mainly genogroup II, reported symptomatic cases with significantly lower median Ct values than controls. Significantly lower rotavirus Ct values were also observed in symptomatic cases vs. controls in 3/7 (43%) studies, and associated with more severe symptoms in 2/2 studies. Contradictory associations were identified for non-C. difficile bacterial and parasitic pathogens. Conclusions: In conclusion, some studies reported clinically useful associations between Ct values and patient or healthcare outcomes; additional, well-designed, large-scale trials are warranted based on these findings. Systematic Review Registration: [PROSPERO], identifier [CRD42020167239].


INTRODUCTION
Gastrointestinal infections contribute significantly to the burden of illness from infectious diseases worldwide (1,2). Rotavirus is the principal cause of diarrhea mortality, responsible for a high attributable fraction among all age groups (13.9%) (3). Shigella, the second most common cause of diarrhea mortality, is a key contributor to diarrheal death among children younger than 5 years (14.3%), mainly in low income countries (3).
Quantitative (q) polymerase chain reaction (PCR) is a robust and increasingly common methodology for rapid syndromic testing due to its sensitivity and specificity for identification of pathogens. In infectious diseases, qPCR cycle threshold (Ct) values represent the number of amplification cycles required for the fluorescent signal to exceed the basal threshold level. Ct values are inversely related to the number of copies of the target gene in a sample, meaning that lower Ct values correlate with higher pathogen loads. In infectious diseases, qPCR Ct values have potential utility in providing clinicians with information regarding genomic load that may help guide clinical and infection-control decisions. In addition, Ct values may help to clarify diagnostic uncertainty in cases where there is difficulty interpreting binary results, for example when distinguishing between causative infectious pathogen and asymptomatic carriage/colonization (4)(5)(6), particularly as identification of multiple pathogens is common (7,8).
Notably, unprecedented challenges from the COVID-19 pandemic have raised the interest in clinical and diagnostic utility of Ct values (9,10). However, in a recent systematic review of the utility of Ct values in respiratory infections (parallel to this study), no universal conclusions could be reached [In press: J Antimicrob Chemother 2021]. This systematic review assesses the global medical literature for associations between Ct values of gastrointestinal pathogens and patient or healthcare outcomes.

METHODS
This systematic review was undertaken according to the principles outlined in the Cochrane handbook and guidance published by the Center for Reviews and Dissemination. The original protocol was published in the PROSPERO database (CRD42020167239) and included broad search terms unrestricted by pathogen or disease type. This review focuses on gastrointestinal pathogens.

Eligibility Criteria
Literature searches of MEDLINE, EMBASE, and the Cochrane Library using search tools at ncbi.nlm.nih.gov/pubmed, embase.com and cochranelibrary.com were undertaken to identify studies reporting on the presence or absence of an association between qPCR Ct values and patient or healthcare outcomes (see Supplementary Table 1 for the PubMed search strategy). The search strategy comprised three concepts: (real-time [rt]-PCR OR qPCR) AND Ct values AND pathogen. Randomized-controlled, single-arm, nonrandomized comparative and observational (retrospective or prospective) studies were included. Animal studies, systematic reviews, non-systematic reviews and meta-analyses were excluded; however, additional publications were identified by manual citation searching of appropriate reviews. Searches were limited to English language studies, for reasons of feasibility.

Study Selection and Data Extraction
Titles and abstracts were screened, based on eligibility criteria, for inclusion by two independent reviewers who then assessed the full texts of relevant studies; a third reviewer resolved conflicts. Key data from all included studies were captured by one reviewer, and subsequently verified by another reviewer. Outcomes were broadly divided into the following categories: mortality, symptomatic vs. asymptomatic, severity of symptoms, duration of symptoms, intensive care unit (ICU) admission, hospitalization and length of stay (LOS).
The quality and risk of bias of each study was assessed using a tool relevant for each study design (Newcastle Ottawa Scale for cross-sectional, cohort, and case-control studies (11).

Overview of Studies Included
Literature searches, conducted January 14-17, 2020, identified 1,029 unique records. Application of distinct screening and restriction parameters specific to gastrointestinal infections identified 33 eligible studies. Most studies reported Ct value association for more than one pathogen; the most commonly studied pathogens were Clostridioides difficile (n = 15), norovirus (n = 10), and rotavirus (n = 9) (Figure 1). All studies identified gastrointestinal pathogens from stool samples. In studies of C. difficile, the majority used genes encoding toxin A or B as PCR targets. The majority of outcomes reported were related to symptoms, including symptom severity, symptomatic vs. asymptomatic and duration of symptoms. Mortality was assessed by three studies. No studies investigated associations between Ct values and hospitalization and/or ICU admission. The majority (84.8%; 28/33) of studies did not report normalized Ct values. Some (66.7%; 22/33) studies presented Ct value distributions.

Quality and Bias
Using Newcastle-Ottawa scales, all cross-sectional studies, cohort studies and case-control studies were classed as being of poor quality (Supplementary Tables 2-4, respectively). This was generally due to a lack of comparability between groups, insufficient or unjustified sample sizes, the use of nonrepresentative samples (often hospitalized patients or agespecific populations) and a lack of detail regarding patient follow-up or non-response; ascertainment of exposure and outcome was usually appropriate. In an assessment of qPCR methodology, 14/29 (48%) full-length articles were considered to have some or many gaps in the reported methodology (Supplementary Table 5).

Clostridioides difficile
C. difficile was the most commonly reported gastrointestinal pathogen with respect to articles describing associations between Ct values and patient or healthcare outcomes ( Table 1). Two studies investigated the association of Ct value with mortality, of which one (N = 1,013) reported no significant associations (12 (14). Reigadas et al. (n = 299) showed that Ct value was independently associated with poor outcome (p < 0.001) and classified patients into risk categories accordingly; high risk of poor outcome (median Ct <23.5); medium risk of poor outcome (median Ct 23.5-27.9); and low risk of poor outcome (median Ct ≥28.0) (17). A further three studies with numbers of PCRpositive patients ranging from 62 to 219, reported lower Ct values in patients with poorer outcomes or more severe disease; however, differences did not reach statistical significance (18)(19)(20).
Four studies investigated differences in Ct values in case vs. control subjects. In Crobach et al (22).
Two studies (N ≤ 120) investigated association of Ct value with duration of symptoms; no significant associations were reported in either study (23,24). One large study (N = 1,281) of diarrheal patients in the UK investigated Ct value and LOS; however, no significant associations were reported, except for patients with PCR-ribotype 027, where LOS was significantly increased in those with low vs. high Ct value (32.5 vs. 28 days; p = 0.018) (13).

Gastrointestinal Viruses
Associations between patient or healthcare outcomes and the Ct value of gastrointestinal viruses were investigated in 14 studies, with the most commonly studied viruses being norovirus and rotavirus (n = 10 and n = 9, respectively) ( Table 2). The majority of studies (n = 10) investigated the difference in Ct values (or viral load derived from Ct values) between cases and controls (symptomatic and asymptomatic, or patients with or without diarrhea).
In general, norovirus, particularly norovirus genogroup II (GII), infections were found to have significantly lower median Ct values in infections vs. controls.        Three studies investigated gastrointestinal viruses other than norovirus and rotavirus. In one study that investigated pathogen quantity and diarrhea in children <5 years old, associations between Ct value and diarrhea were reported for cases of adenovirus and astroviruses (29). No other associations between Ct values and cases vs. controls were identified (27,30).

Non-C. difficile Bacterial and Parasitic Pathogens
Associations between patient clinical outcomes and the Ct value of non-C. difficile bacterial and parasitic pathogens were investigated in nine studies ( Table 3).
Among bacterial studies, the majority investigated associations between quantitative PCR-derived bacterial loads and cases vs. controls (symptomatic vs. asymptomatic, or patients with vs. without diarrhea), and most studies found significant associations. Among five studies reporting differences in Ct values between cases vs. controls, significantly lower median Ct values were reported in cases of enterotoxigenic Escherichia coli (ETEC), enteropathogenic E. coli (EPEC), Campylobacter spp., enteroinvasive E. coli (EIEC)/Shigella spp., and Salmonella spp. (22,27,29,30,37). However, associations were not consistent across studies, including two reports (n = 9 and n = 46) of no significant difference in cases vs. controls for Salmonella spp. (27,30). In one study, associations were notably weaker for Campylobacter spp. and typical EPEC (29). In a study (n = 143) of patients with EPEC, a 29% increase in risk of diarrhea was observed for each log 10 unit increase (calculated by Ct value) in bacterial load (OR 1.29; 95% CI 1.08-1.53) (37).
One study (n = 143) investigated the effect of EPEC bacterial load on the duration of symptoms; however, no significant association was observed (37).
All studies of parasites investigated associations between Ct values (or Cq) and cases vs. controls (symptomatic vs. asymptomatic, or patients with vs. without diarrhea). In studies including Cryptosporidium spp., two reported significantly lower Ct values in cases vs. controls, including Elfving et al. (n = 67; median Ct 32.1 vs. 36.8; p = 0.0009) (22,30). One further study also reported lower Ct values in cases vs. controls (n = 23), but did not reach statistical significance (27). Furthermore, and contrary to expected results, Haque et al. reported higher mean Ct values in Cryptosporidium parvum and Cryptosporidium hominus cases than controls, although the differences were not significant (p = 0.127 and 0.098) (39). In a study in children <5 years (n = N/A), strong pathogen quantity-dependent associations  • For Shigella/EIEC, Cq values <33.1 were defined as "diarrhea-associated" and <27.9 were defined as "highly diarrhea-associated" (ROC cut-off 26.1; Youden index 0.18) • For heat-stable ETEC, Cq values <26.2 were defined as "diarrhea-associated" and <22.8 were defined as "highly diarrhea-associated" (ROC cut-off 25.4; Youden index 0.25) • Campylobacter jejuni or C coli and EPEC were only moderately associated with diarrhea • For Campylobacter spp., Cq values <19.7 were defined as "diarrhea-associated" and <15.4 were defined as "highly diarrhea-associated" (ROC cut-off 25.8; Youden index 0.08) • For EPEC, Cq values <19.5 were defined as "diarrhea-associated" and <16.0 were defined as "highly diarrhea-associated" (ROC cut-off 19.9; Youden index 0.07) • Vibrio cholerae and Salmonella spp. showed associations with diarrhea • For Vibrio cholerae, Cq values <34.9 were defined as "diarrhea-associated" and <33.8 were defined as "highly diarrhea-associated" (ROC cut-off 29.3; Youden index 0.55) • For Salmonella spp, Cq values <32.4 were defined as "diarrhea-associated" and <30.7 were defined as "highly diarrhea-associated" (ROC cut-off 29.7; Youden index 0.29) • EAEC and Aeromonas spp were associated with diarrhea only in specific study sites or age strata   with diarrhea were reported in cases of Cryptosporidium spp. (29). Among three studies including Giardia spp. (n = 118, n = 375 and n = 92), none reported statistically significant lower Ct values in cases vs. controls (22,39,40). Notably, Haque et al. reported that Giardia lamblia parasite load, as measured by Ct values, was inversely related to diarrhea, which the authors suggest could be related to the primary role played by the immune system in diarrheal illness that results from these infections (39).

DISCUSSION
The objective of this systematic review was to assess the global medical literature for any correlation between Ct values and clincal outcomes of patients with gastrointestinal infections.
Lower Ct values correspond with greater quantities of detectible target gene and therefore a higher pathogen load, which may correspond with less favorable clinical outcomes. Here we report outcomes from studies identified that report on gastrointestinal pathogens only. This review gathers data from 33 studies, with the largest number of studies for C. difficile (n = 15). The most common outcomes reported were related to symptoms, including case vs. control, with vs. without diarrhea, and severity of symptoms. Evidence in this review suggests associations between Ct values and symptomatic C. difficile infections. Four out of eight studies reporting the association between lower Ct values and increased disease severity found the association to be significant, including two studies that reported lower Ct values as a predictor of poor outcome (14,17). Furthermore, 2/4 case vs. control studies reported significantly lower Ct values in symtomatic cases. Most of the C. difficile studies reported genes encoding toxin A/B as the target for PCR diagnostics, which when detected by other methods, is generally inferred as marker of disease severity (5).
All studies of norovirus and rotavirus reported lower Ct values in cases vs. controls; the majority for norovirus GII and ∼50% for rotavirus reported significant differences. Furthermore, two studies of rotavirus infections reported significant associatons between lower Ct values and severity of symptoms, including vomiting, severe dehydration and administering intravenous fluids (27,34). Notably, the association of Ct values and symptom severity was more pronounced for norovirus GII than norovirus GI (25,27,28). One possible explanation for this is the increased virulence observed with GII infection compared with other norovirus genogroups (41), although more investigation is necessary to draw firm conclusions.
This review found less evidence for the clinical utility of Ct values in non-C. difficile bacterial and parasitic infections compared with C. difficile and gastrointestinal viruses. Multiple studies reported significant associations between bacterial loads and symptomatic cases, particularly for Shigella (29,30). Two studies reported Shigella association with symptom severity (27,38). Inconsistencies were found in studies of parasitic infections; some studies indicated an association between low Ct values and symptomatic infection in patients with Cryptosporidium spp., however, evidence is limited (22,29,30). There is insufficient evidence to draw conclusions for other parasitic infections.
Among the studies included in this review, evidence suggests that Ct values may have utility in defining symptomatic causality, particularly in cases of polymicrobial infection. In one study of norovirus-positive samples, coinfection with rotavirus was observed in 3.7 and 7.4% of asymptomatic and diarrheal samples, respectively; probable etiology was determined based on relative Ct values, highlighting their utility for defining causitive organisms in this setting (28). Ct values may also aid causative diagnosis in patients with C. difficille infection, where asymptomatic colonization (5,6), and coinfections have been reported (42). C. difficile fecal load is already considered to be of diagnostic utility in distinguishing between infection and colonization (43,44). However, it is essential to consider Ct values within the context of clinical presentation rather than utilize Ct values as an independent marker of disease.
Despite multiple studies reporting significant associations between high genomic load (low Ct values) and symptomatic infections, particularly for C. difficile, norovirus and rotavirus, statistically significant evidence was inconsistent across studies despite similar trends. A possible explanation for this is the diversity of populations investigated across each study (e.g., hospital vs. community setting, pediatric vs. adult populations); adjusting for similar settings may uncover stronger trends toward Ct value and patient outcomes. Further assessments of associations between Ct values and LOS, hospital/ICU admission, for example, could also aid in understanding the utility of Ct values in the diagnosis of gastrointestinal infections.
When interpreting the studies in this systematic review, consideration must be given to the settings and populations in which they were conducted. Studies for some pathogens, such as norovirus, were conducted primarily in pediatric populations and as such their conclusions may not apply to adult populations. All but one of the studies investigating non-C. difficile bacterial pathogens and parasites were performed in non-industrialized countries; therefore, the clinical impact of Ct values for these pathogens in industrialized countries remains to be determined. Of the seven studies that detected parasites, five investigated a large list of GI pathogens and multiple pathogens were detected for 8-72% patients (22,27,29,30). These studies highlight the utility of syndromic testing in gastrointestinal infection, where multiplex testing is able to detect more pathogens and coinfections than conventional methods (42). It should also be noted that multiplex PCR for GI pathogens does not currently provide a picture of the microbiome, whereas culture-based techniques are able to provide an understanding of dysbiosis resulting from GI infections.
Differences in study methodology and qPCR workflow are likely to impact Ct values, including: specimen source, collection method, transport media type and volume, stability, quality of the sample, time of sampling vs. onset of infection, master mix components, type and concentration of passive reference dye, reaction efficiency, inter-and intravariability in assay platforms, and whether they were single or multiplex systems. Methodologies varied widely between studies and many (39%) had some or many gaps in reporting defined standardized methodologies. Therefore, within-study variability may have limited the ability to detect associations. The majority (84.8%; 28/33) of studies did not report normalized Ct values, which would have provided more accurate estimations of genomic load for each sample. Although outside the scope of our review, we noted not all (66.7%; 22/33) studies presented Ct value distributions. Further studies to understand the distribution of Ct values in relation to patient outcomes across the populations would be necessary if Ct values are to be utilized in clinical decision-making. After data analysis had been completed, we became aware of the Minimum Information for Publication of Quantitative rt-PCR Experiments (MIQE) guidelines (45), which should be applied to laboratory-developed tests. Some of the studies utilized in this review use commercially available assays and, therefore, when implemented in clinical diagnostic routines, applicable validation, and verification using external controls are necessary. Due to the late discovery it was not possible to re-assess the studies using laboratory-developed assays with the MIQE guidelines in mind; however, we believe that assessment of study methodology using these guidelines would not significantly alter the findings of this systematic review.
There were a number of limitations to this systematic review. The protocol restricted articles referring to Ct values as a measure of genomic load, therefore studies which reported genomic load in measures other than Ct value were not picked up in the database searches or excluded from during screening. Furthermore, articles describing Ct values but with no mention of Ct values in the title, abstract or keywords, were not retrieved based on the search parameters used in the database searches. In addition, late in the review we became aware of alternative wording for Ct values, including Cq and "crossing point" [discussed in detail in (45)]; while we have added articles with these terms manually, it is possible that some may have been missed. Another limitation to this review was the assessment of all included studies as poor quality for bias by the Newcastle-Ottawa scale. This is due to the majority of studies reporting Ct values as secondary outputs, as opposed to seeking to compare clinical outcomes against Ct values. Consequently, the studies did not fully align with the risk and bias assessment. There was considerable variability between studies. Given the high heterogeneity between studies, it was not possible to conduct aggregated/meta-analyses, a key limitation in the scope of this review. A number of studies only made comparative analysis between symptomatic and asymptomatic cases, which limits the clinical utility of these studies in defining Ct values as a measure of disease severity. However, Ct values of asymptomatic patients still hold clinical value in order to discriminate between infection and colonization, an observation reported in multiple studies (4-6, 25, 27). A single reviewer conducted the data extraction and a second reviewer checked all the data points. Whilst an acceptable approach, the methodology could have been optimized by double independent reviewer data extraction with a third reviewer for discrepancy resolution. Due to the large number of studies identified as potential data sources for this review, the single-reviewer extraction method ensured that the review remained feasible. Despite these limitations, we believe this review provides insights into the potential clinical utility of gastrointestinal pathogen Ct values. In summary, there is evidence to support relationships between Ct values and clinical outcomes in gastrointestinal infections. Considered alongside clinical presentation, Ct values could help to guide treatment decisions, particularly in cases of C. difficile, where treatment is guided by severity of disease and asymptomatic colonization has been observed (5,6,46). This review did not uncover sufficient evidence to draw conclusions on the clinical utility of Ct values for non-C. difficile bacterial and parasitic infections. This systematic review is the first to assess the relationship between Ct values and clinical outcomes in gastrointestinal infections, large-scale clinical trials with endpoints centered on Ct values are warranted to draw definitive evidence.

DATA AVAILABILITY STATEMENT
The original contributions presented in the study are included in the article/Supplementary Material, further inquiries can be directed to the corresponding author/s.

AUTHOR CONTRIBUTIONS
BV, DB, JP, SR, DM, GH, and JV were involved in conception and design of the study. All authors contributed to interpretation of the data, manuscript drafting and revision, and approved the submitted version.

FUNDING
Medical writing support for the development of this manuscript, under the direction of the authors, was provided by Isabella Talbot, BSc, of Ashfield MedComms, an Ashfield Health company, and funded by Qiagen Manchester Ltd. This study received funding from QIAGEN Manchester Ltd. The funder had the following involvement with the study: article processing fees and provision of medical writing support.