Multi-organism gastrointestinal polymerase chain reaction positivity among pediatric transplant vs non-transplant populations: A single-center experience

Abstract

Background:

Diarrhea is a common problem in the pediatric post-solid organ transplant and post-hematopoietic stem cell transplant populations. Infectious etiology incidences are poorly defined, and the possibility of multi-organism positivity is often uninvestigated. The aim of this study is to utilize stool multiplex GIP assays to compare the PTP and NTP regarding the incidence and profiles of single-organism and multi-organism infectious diarrhea.

Methods:

A single-center retrospective review was conducted, investigating stool multiplex GIP panel results over a more than 3-year period, for pediatric patients. Assays test for 23 viral, bacterial, and protozoal organisms.

Results:

Positive assays in the PTP and NTP were 70/101 (69.3%) and 962/1716 (56.1%), respectively (P = .009). Thirty-two percent (32/101) of assays within the PTP were multi-organism positive, significantly more than 14.8% (254/1716) in the NTP (P < .00001). There was no significant difference in the incidence of single-organism positives, 37.6% (38/101) in PTP and 41.3% (708/1716) in the NTP. The PTP demonstrated a statistically significantly higher incidence of the following organisms within multi-agent positive GIPs (P < .05 for each): Clostridioides difficile, Cryptosporidium, EPEC, norovirus, and sapovirus.

Conclusions:

The pediatric PTP demonstrates higher incidence of positive GIPs, higher rate of multi-organism positivity, and unique infectious organism incidence profiles. These data can provide a framework for understanding organism-specific pathogenicity factors, assessing the clinical impact of enteric co-infection, and understanding the utility of this testing modality in this unique population.

1 |. INTRODUCTION

Etiologic identification and management of diarrhea in the pediatric SOT and HSCT populations are a challenging clinical scenario. Estimated all-cause post-SOT prevalence in adult counterparts reaches 40%−54% with far-reaching implications on nutrition, hydration, serum creatinine, serum levels of immunosuppressive medication, quality of life, and transplant viability.^1–8 Adult studies aimed at identification of post-transplant diarrheal etiology found a cause in 22% to 83% of cases with infectious attribution ranging from 28% to 77%.^9–11 These studies implemented conventional diagnostic modalities in a non-standardized or stepwise fashion, generally ceasing further testing after a single possible source was identified.

Multiplex GIP assays allow for investigation of several infectious causes in a rapid, sensitive, and specific fashion.¹² When compared to physician-directed standard laboratory methods, BioFire GIP testing demonstrates higher diagnostic yield (65% vs 46%, P < .001) and higher co-detection of multiple diarrheagenic organisms (20% vs 2%, P < .001).¹³ Underestimating the incidence of multi-agent positivity limits further investigation into potential contributors and understanding of clinical impact. This is especially important in vulnerable populations where failure to identify concomitant pathogens may preclude or delay organism-specific management, contributing to persistent diarrhea and untoward clinical sequelae. Multiple factors would suggest a possible predisposition to multi-agent enteric pathogen positivity in the pediatric PTP (see Discussion below for enumeration and consideration of these factors). To date, this has not been investigated. The objective of this study is to utilize the inherited qualities of GIP panels (higher positivity rates and higher detection of enteric co-infection), to assess incidence and organism profiles among single-agent and multi-agent positive assays, comparing the pediatric NTP and PTP.

2 |. MATERIALS AND METHODS

A single-center retrospective review of stool multiplex GIP panel (Film Array GI Panel, BioFire Diagnostics) results and associated demographics was conducted at a free-standing pediatric institution for a time period of 1/1/2016 to 5/22/2019, patients’ age 0–17.9 years. Institution laboratory guidelines require liquid consistency stool for performance of this test, such that samples can be aspirated into the analyzer without dilution. Additional clinical indications for performing GIP assays were at the discretion of the overseeing physician and included stool frequency, volume, and presence of blood. This assay includes the following organisms: adenovirus F 40/41, astrovirus, Campylobacter, C (Clostridioides) difficile, Cryptosporidium, Cyclospora cayetanensis, EAEC, Entamoeba histolytica, EPEC, ETEC, Giardia lamblia, norovirus GI/GII, Plesiomonas shigelloides, rotavirus A, STEC, EIEC, Salmonella, sapovirus, Vibrio, Vibrio cholera, and Yersinia enterocolitica. No exclusion criteria were applied to the NTP regarding past medical history. Chi-squared tests were used to compare cumulative incidence of positive assays, as well as single- and multi-organism positive assays, between the pediatric PTP and NTP. Given smaller numeric values, Fisher’s exact test was used to assess incidence of individual organisms as a part of a single- or multi-agent positive assay, comparing the two populations of interest.

3 |. RESULTS

A total of 1817 GIPs were resulted within the given period, 1032 positive and 785 negative. The population included 831 females and 986 males. Age ranged from <0.1 to 17.9 years, median 5.2 years (IQR, 2.2–11.4). Sixty-four percent of samples were from Caucasian/non-Hispanic, 18% others, 17% African American, and <1% Asian, American Indian/Alaskan Native, and Native Hawaiian. There were no significant differences in age, gender, or ethnic distribution between the PTP and NTP. PTP stool samples were reviewed from 43 heart, 27 liver, 17 kidney, and 14 hematopoietic stem cell transplant recipients (Table 1). The NTP demonstrated a large degree of clinical heterogeneity: previously healthy infants with acute gastroenteritis, children with inflammatory bowel disease, and adolescents with oncologic diagnosis on chemotherapy.

TABLE 1.

Breakdown of single- and multi-organism positivity in post-transplant cohort by organ received

Post-transplant GIP positive assays
Transplant	Total assays	Positive assays (%)	Single organism (% of positives)	Multiple organisms (% of positives)
HSCT	14	5 (36)	4 (80)	1 (20)
Heart	43	35 (81)	13 (37)	22 (63)
Kidney	17	11 (65)	8 (73)	3 (27)
Liver	27	19 (70)	13 (68)	6 (32)

The PTP demonstrated a higher incidence of positive assays (70/101, 69.3%) compared to the NTP (962/1716, 56.1%) (P = .009). Of 1716 assays performed in the NTP, 708 (41.3%) were positive for a single organism and 254 (14.8%) were positive for multiple organisms. Of the 101 GIPs performed in the PTP, 38 (37.6%) were positive for a single organism and 32 (31.7%) were positive for multiple organisms. This represented a statistically significantly higher incidence of multi-organism positivity in the PTP (P < .00001). Comparing the incidences of single-organism positivity, Salmonella was significantly higher in the NTP (119/1716, 6.9% vs 0/101; P < .05) and C difficile was significantly higher in the PTP (20/101, 19.8% vs 147/1716, 8.6%; P < .05). The NTP also demonstrated a significantly higher incidence of Salmonella within multi-agent positive GIPs (74/1716, 4.3% vs 0/101; P < .05). The PTP demonstrated a significantly higher incidence of the following organisms within multi-agent positive GIPs (P < .05 for each): C difficile (15/101, 14.9% vs 88/1716, 5.1%), Cryptosporidium (2/101, 2% vs 4/1716, 0.2%), EPEC (17/101, 16.8% vs 101/1716, 6.3%), norovirus (13/101, 12.9% vs 65/1716 3.8%), and sapovirus (7/101, 6.9% vs 39/1716, 2.3%) (Figure 1). The most common co-occurring organisms in the NTP were EPEC and Salmonella (30 assays). The most common co-occurring organisms in the PTP were EPEC and C difficile/EPEC and norovirus (five assays each). Samples from heart transplant recipients demonstrated the highest percent of positive assays (35/43, 81%; P < .05) and the highest percent of multi-organism positive assays (22/43, 51%; P < .01). Interestingly, despite a greater proportion of multi-organism positivity in transplant patients, there were no assays with more than three organisms detected. The NTP demonstrated 10 assays with >3 organisms, the greatest being positive for five organisms. This rate did not reach statistical significance.

FIGURE 1.

Organisms that demonstrated statistically significant difference of incidence as a part of a multi-organism positive gastrointestinal PCR

4 |. DISCUSSION

There is inherent difficulty in identifying an etiology for post-transplant diarrhea in the pediatric population. Accurate diagnosis and appropriate management are essential to avoid associated morbidity and mortality. Nevertheless, the cumulative incidence, causative organism(s), and risk factors of pediatric post-transplant infectious diarrhea remain largely uninvestigated. Published studies in the adult transplant population employ an average of 15 methods to investigate infectious causes, implemented in a non-standardized fashion (retrospective review of testing directed by overseeing physician) or stepwise approach (diagnostic algorithm ceasing further testing after one possible source is identified).^9–11 This may contribute to underestimating the incidence and clinical impact of multi-organism enteric infection. Stockmann et al¹³ demonstrated that physician-directed standard laboratory testing in pediatric diarrhea consisted of obtaining a median of three assays (range 1–10) and when directly compared to BioFire GIP testing generated lower diagnostic yield (46% vs 65%, P < .001) and lower co-detection of multiple diarrheagenic organisms (2% vs 20%, P < .001). The accessibility, sensitivity, and broad diagnostic net cast by multiplex GIP assays may provide a groundwork for addressing these questions.

Our finding of higher incidence of C difficile positivity in PTP single- and multi-agent assays is likely multifactorial. A retrospective study of 22 children’s hospitals demonstrated nearly double the incidence of inpatient CDI from 2001 to 2006.¹⁴ Among children with CDI, there is an exceptionally high prevalence of comorbidities.^14,15 Nylund et al¹⁶ demonstrated CDI adjusted odds ratios of 4.53 (95% CI 3.92–5.24) and 3.31 (95% CI 2.87–3.82) for conditions of SOT and HSCT, respectively. Pant et al¹⁷ found a 0.6% and 3.6% rate of CDI in pediatric non-transplant and SOT cohorts, respectively, yielding an odds ratio of 6.6 (95% CI 6.0–7.3; P < .05). Beyond initial infection, the pediatric PTP bears many risk factors associated with recurrent CDI: malignancy, recent hospitalization, recent surgery, antibiotic use, acid blocker use, immunosuppressant use, and hospital-acquired disease.¹⁵ Colonization in the infant population must be considered as a contributor to positive assays. The median age for C difficile positivity in the PTP and NTP was 6.1 and 5.8 years, respectively, with the PTP having a lower percentage of positive tests in patients <2 years old. Beyond infancy, higher rates of colonization and prolonged shedding in stool have been demonstrated in similar populations and are also likely contributing factors.^18–23

CDC surveillance reports a nearly threefold increase in cryptosporidiosis in the United States from 2004 to 2008, with a large portion of reported cases of age 0–9 years.²⁴ Pediatric PTP populations have increased risk for severe and prolonged infection.²⁵ Our observation of higher Cryptosporidium rates in PTP is found elsewhere in the literature; however, the 2% (2/101) attribution is generally lower than corresponding studies.^10,26,27 Bandin et al²⁸ and Krause et al²⁹ identified Cryptosporidium as responsible for 11% (7/64 and 6/57, respectively) of all-cause pediatric post-SOT diarrhea in separate cohorts. Alternatively, other sources report relative rarity within immunocompromised pediatric populations.^9,25,30 Disparate incidences are likely attributable to differences in populations, design, assays utilized, geographic endemicity, and seasonal effects.

Diarrheagenic E coli may be an underappreciated cause of diarrhea in children.³¹ Chronic EPEC infections are reported in pediatric³² and immunocompromised³³ populations. Our findings of 19.8% (20/101) EPEC positivity in the pediatric PTP exceed corresponding estimates. Gu et al³⁰ and Chao et al³⁴ utilize the BioFire GIP to observe a 2% (4/199) and 11.3% (35/311) EPEC positivity in pediatric and adult oncology cohorts, respectively. Choa et al³⁴ postulate that high EPEC GIP positivity could represent novel recognition of transmission in high-risk populations or false positives for various reasons (colonization, assay cross-reactivity with non-pathogenic organisms).

Norovirus is one of the leading causes of community-acquired diarrhea in the immunocompetent population, estimated to account for 19–21 million cases annually in the United States.³⁵ Post-transplant incidence estimates for combined noro- and sapovirus reach 18.4% in adults, demonstrating prolonged diarrheal course and higher symptom recurrence after initial resolution.³⁶ Roos-Weil et al³⁶ observed that 94% of noro- and sapovirus cases in their renal transplant cohort demonstrated a chronic diarrheal course with a mean duration of 8.7 months. Other case series also demonstrate prolonged asymptomatic shedding of norovirus, thought to further viral evolution and immune resistance.^37,38

In summary, multiple studies have contributed to understanding the differences between infectious diarrhea in the NTP and PTP. Detailed characterizations of individual diarrheagenic organisms demonstrate the potential for different behaviors between these cohorts. Age and immune status are strong determinants of organism pathogenicity. These findings would indicate a predisposition to multi-organism enteric positivity within the pediatric transplant population. To date, this has not been investigated. Implementation of GIP panels results in higher positivity rates and higher detection of enteric co-infection than conventional methods.^13,39,40 Our findings utilize this characteristic of the assay to demonstrate significant differences between the pediatric PTP and NTP regarding overall incidence of positive GIPs, incidence of multi-organism positive GIPs, and specific organisms involved in multi-agent positive GIPs.

Justification for these findings is complex as it involves multiple factors for each organism. Available literature in transplant cohorts demonstrates that certain pathogens exhibit higher risk of incipient infection, prolonged infection duration, prolonged organism shedding in stool, higher risk of recurrent infection, and prevalence of colonization. Our findings agree with these single-organism characterizations by identifying a novel pattern in which the pathogens that demonstrate these factors are those more commonly involved in multi-agent GIP positivity. This outlines the above factors as likely contributors to our findings. It simultaneously illustrates the complexity of interpreting infectious diarrheal testing in this population, as a positive assay (especially in the setting of multi-organism positivity) may represent early infection, periodic symptoms of prolonged initial infection, separate incidence of recurrent infection, shedding of organism in stool after infection resolution, asymptomatic colonization, or other false positives. The retrospective nature of our investigation precludes differentiation, and in many cases, this determination is beyond the reach of modern investigation. The importance of interpreting case-by-case GIP results in light of the clinical scenario, laboratory findings, and epidemiologic considerations cannot be overstated.

Study design also precludes confirmation of GIP results with conventional methods; therefore, validation of this modality in the pediatric PTP is beyond study scope. Additional study limitations include single-center experience limiting generalization of results, variability in physician practices (clinical indications used for obtaining GIP), heterogeneous nature of PTP (SOT and HSCT included), and small sample size of PTP compared to NTP.

In conclusion, our findings demonstrate higher overall incidence of positive GIPs, higher rate of multi-organism positive GIPs, and unique infectious organism incidence profiles in the pediatric PTP. Areas where further research is required are numerous and include the following: further defining incidence of common and severe etiologies contributing to pediatric post-transplant diarrhea, understanding organism-specific spectrum of clinical severity and risk factors for severe/prolonged course, organism-specific duration of stool shedding in this unique population, prevalence of organism-specific colonization in pediatric PTP, comparing clinical course and sequelae of single-agent vs multi-agent enteric infection, and existence/pathophysiology of commensal relationship between enteric infectious organisms. A better understanding of these facets can lead to more confident test interpretation and patient management in a population where diarrhea markedly impacts quality of life, morbidity, and mortality.

Abbreviations:

C difficile

Clostridioides difficile

CDC

Centers for Disease Control and Prevention

CDI

Clostridioides difficile infection

EAEC

enteroaggregative Escherichia coli

EIEC

enteroinvasive/Shigella Escherichia coli

EPEC

enteropathogenic Escherichia coli

ETEC

enterotoxigenic Escherichia coli

GIP

gastrointestinal polymerase chain reaction

HSCT

hematopoietic stem cell transplant

IQR

interquartile range

NTP

non-transplant population

PTP

post-transplant population

SOT

solid organ transplant

STEC

Shiga-like toxin-producing Escherichia coli