Advertisement

Multiplex PCR Based Urinary Tract Infection (UTI) Analysis Compared to Traditional Urine Culture in Identifying Significant Pathogens in Symptomatic Patients

Open AccessPublished:November 09, 2019DOI:https://doi.org/10.1016/j.urology.2019.10.018

      Abstract

      Objective

      To evaluate whether multiplex PCR-based molecular testing is noninferior to urine culture for detection of bacterial infections in symptomatic patients.

      Methods

      Retrospective record review of 582 consecutive elderly patients presenting with symptoms of lower urinary tract infection (UTI) was conducted. All patients had traditional urine cultures and PCR molecular testing run in parallel.

      Results

      A total of 582 patients (mean age 77; range 60-95) with symptoms of lower UTI had both urine cultures and diagnostic PCR between March and July 2018. PCR detected uropathogens in 326 patients (56%, 326/582), while urine culture detected pathogens in 217 patients (37%, 217/582). PCR and culture agreed in 74% of cases (431/582): both were positive in 34% of cases (196/582) and both were negative in 40% of cases (235/582). However, PCR and culture disagreed in 26% of cases (151/582): PCR was positive while culture was negative in 22% of cases (130/582), and culture was positive while PCR was negative in 4% of cases (21/582). Polymicrobial infections were reported in 175 patients (30%, 175/582), with PCR reporting 166 and culture reporting 39. Further, polymicrobial infections were identified in 67 patients (12%, 67/582) in which culture results were negative. Agreement between PCR and urine culture for positive cultures was 90%, exceeding the noninferiority threshold of 85% (95% conflict of interest 85.7%-93.6%).

      Conclusion

      Multiplex PCR is noninferior to urine culture for detection and identification of bacteria. Further investigation may show that the accuracy and speed of PCR to diagnose UTI can significantly improve patient outcomes.
      Traditional urine culture is commonly regarded as the gold standard for detection and identification of pathogens. However, evidence has been accumulating to support use of molecular methods such as PCR. With antimicrobial resistance becoming both more common and complex, effective treatment of (urinary tract infection) UTIs is even more dependent on the accurate identification of pathogens. Some organisms can be fastidious, and therefore difficult to grow in culture. Further, PCR results can be obtained in a day or less, while culture can require 2 or more days. Previous studies have reported PCR to have both superior sensitivity and specificity, and have recommended PCR for rapid identification of pathogens in sepsis,
      • Chang S-S
      • Hsieh W-H
      • Liu T-S
      • et al.
      Multiplex PCR system for rapid detection of pathogens in patients with presumed sepsis – A systemic review and meta-analysis.
      • Scerbo MH
      • Kaplan HB
      • Dua A
      • et al.
      Beyond blood culture and Gram stain analysis: a review of molecular techniques for the early detection of bacteremia in surgical patients.
      • Florio W
      • Morici P
      • Ghelardi E
      • Barnini S
      • Lupetti A
      Recent advances in the microbiological diagnosis of bloodstream infections.
      and for diagnosis of genital infections and sexually transmitted diseases,
      • Munson E
      Molecular diagnostics update for the emerging (if not already widespread) sexually transmitted infection agent mycoplasma genitalium: just about ready for prime time.
      • Sahi S-V
      • Rogozińska E
      • Sobhy S
      • Khan KS
      Accuracy of tests used to detect infection with Chlamydia trachomatis in asymptomatic pregnant women: a systematic review.
      • Gayet-Ageron A
      • Combescure C
      • Lautenschlager S
      • Ninet B
      • Perneger TV
      Comparison of diagnostic accuracy of PCR targeting the 47-Kilodalton protein membrane gene of treponema pallidum and PCR targeting the DNA polymerase I Gene: systematic review and meta-analysis.
      parasitic infections,
      • Berry A
      • Benoit-Vical F
      • Fabre R
      • Cassaing S
      • Magnaval JF
      PCR-based methods to the diagnosis of imported malaria.
      tuberculosis,
      • Haas CT
      • Roe JK
      • Pollara G
      • Mehta M
      • Noursadeghi M
      Diagnostic ‘omics’ for active tuberculosis.
      and gastrointestinal infections.
      • Zboromyrska Y
      • Vila J
      Advanced PCR-based molecular diagnosis of gastrointestinal infections: challenges and opportunities.
      Few studies have compared multiplex PCR with urine culture for diagnosis of UTIs and acute cystitis. Although several studies have compared performance of PCR with urine culture for detection of a single pathogen, only 4 have tested multiplex PCR: one against 15 bacteria,
      • Lehmann LE
      • Hauser S
      • Malinka T
      • Klaschik S
      • Stüber F
      • Book M
      Real-time polymerase chain-reaction detection of pathogens is feasible to supplement the diagnostic sequence for urinary tract infections.
      a second against 14 bacteria together with 6 fungi,
      • Lehmann LE
      • Hauser S
      • Malinka T
      • et al.
      Rapid qualitative urinary tract infection pathogen identification by SeptiFast® Real-Time PCR.
      a third against 20,
      • Cybulski Z
      • Schmidt K
      • Grabiec A
      • et al.
      Usability application of multiplex polymerase chain reaction in the diagnosis of microorganisms isolated from urine of patients treated in cancer hospital.
      and the fourth against 9 bacteria.
      • van der Zee A
      • Roorda L
      • Bosman G
      • Ossewaarde JM
      Molecular diagnosis of urinary tract infections by semi-quantitative detection of uropathogens in a routine clinical hospital setting.
      Polymicrobial infections may occur in as many of 39% of UTIs
      • Price TK
      • Dune T
      • Hilt EE
      • et al.
      The clinical urine culture: enhanced techniques improve detection of clinically relevant microorganisms.
      ,
      • Hilt EE
      • McKinley K
      • Pearce MM
      • et al.
      Urine is not sterile: use of enhanced urine culture techniques to detect resident bacterial flora in the adult female bladder.
      and can display enhanced virulence and increased antibiotic resistance.
      • Croxall G
      • Weston V
      • Joseph S
      • Manning G
      • Cheetham P
      • McNally A
      Increased human pathogenic potential of Escherichia coli from polymicrobial urinary tract infections in comparison to isolates from monomicrobial culture samples.
      Simultaneous detection of a larger number of pathogens may confer benefits for outcome of UTIs. The long-term objective is to determine whether the speed and accuracy of multiplex PCR improves patient care and potentially reduces costs. This study is the first step to confirm that PCR is noninferior to traditional urine culture in detecting bacteria in symptomatic patients. In this study, we compared a multiplex panel of 31 bacteria against urine culture for diagnosis of patients presenting with symptoms of UTI. Though this PCR also reports resistance for each organism, these data are not reported here. Although it may sound intuitive that the resistance of a single organism is different from the susceptibility results of the polymicrobial “soup”, the data are being analyzed to better understand the patterns of the results.

      Materials and Methods

      Participants

      This was a single-site (Comprehensive Urology, Royal Oak, MI) retrospective study. IRB approval was obtained prior to commencing the study (IRB protocol number: 20171870). All patients meeting the inclusion criteria presenting to clinic between March and July 2018 (n = 582) were included in the study. Inclusion criteria: ≥60 years of age; symptoms of acute cystitis or UTI; sufficient urine sample volume for urinalysis, traditional culture, and PCR; all samples were shipped FedEx priority overnight. All samples except 4 were received by the laboratory the day after collecting the samples. Four samples were received 2 days after collection.
      This study focused on patients ≥60 years of age. UTI is common in this age group and can be more difficult to diagnose. Localized urogenital symptoms may not be present in this population and differentiation between UTI and asymptomatic bacteriuria can be difficult.
      • Rowe TA
      • Juthani-Mehta M
      Diagnosis and management of urinary tract infection in older adults.
      Patients in this age group could benefit significantly from better identification of UTI pathogens.

      Urine Culture

      Study participants provided a urine sample obtained either by self-administered clean catch or by catheterization. The urine was mixed and a sterile plastic loop (1 μL) used to inoculate blood agar plates. A sterile plastic loop (1 μL) was also used to inoculate colistin and nalidixic acid agar/MacConkey agar (CNA/MAC) plates, one loop-full of urine on the CNA side of the plate and another full loop-full on the MAC side of the plate. All plates were incubated at 35C in 5% CO2 for ≥18 hours and then examined for evidence of growth. Plates with <104 CFU/mL were reported as normal urogenital flora. For plates with growth (≥104 CFU/mL), the quantity and morphology of each organism was recorded. The maximum readable colony count using the 1 μL loop is >105 CFU/mL. Colony counts were performed on the blood agar plates. Species identification and colony counts were performed on CNA/MAC plates. For plates with ≤2 pathogens, species identification and colony counts were reported for each pathogen with ≥104 CFU/mL. If ≥3 pathogens were present, and one or 2 were predominant, species identification, and colony counts were reported. If ≥3 pathogens were present without predominant species, a mixed morphotype was reported.
      Samples for Gram stain were prepared by applying a thin, even smear on microscope slides, allowed to air dry, and then fixed with methanol. The slides were covered in crystal violet solution for one minute, rinsed with water, covered with iodine for 1 minute, and then rinsed with water again. Staphylococcus aureus 29213 was used as a positive control, and E. coli 35218 was used as a negative control.
      Pathogen identification was conducted using the VITEK 2 Compact System (bioMerieux, Durham, NC) in accordance with standard operating procedures. Briefly, a sterile swab was used to transfer morphologically similar colonies from positive blood agar plates to prepared polystyrene test tubes containing 3.0 mL of sterile saline. The sample was adjusted for density (equivalent to McFarland No. 0.50 to 0.63). The sample tube and an appropriate identification card were placed into the cassette and inserted into the VITEK 2 instrument. A GN card was used for Gram negative bacteria, and a GP card used for Gram positive bacteria. An YST card was used for yeast. Pathogen identification was then read from the VITEK 2 instrument.

      DNA Extraction and Analysis

      DNA was extracted from urine samples using the KingFisher/MagMAX Automated DNA Extraction instrument and the MagMAX DNA Multi-Sample Ultra Kit (ThermoFisher, Carlsbad, CA). Briefly, 400 μL of urine was transferred to wells in 96-well deep well plates, sealed, and centrifuged to concentrate the samples, after which supernatant was removed. Enzyme Lysis Mix (220 μL/well) was added and incubated for 20 minutes at 65C. Proteinase K Mix (PK Mix) was added (50 μL/well) and incubated for 30 minutes at 65C. Lysis buffer (125 μL/well) and DNA Binding Bead Mix (40 μL/well) were added and the samples shaken for a minimum of 5 minutes. The 96-well plate was then loaded into the KingFisher/MagMAX Automated DNA Extraction instrument, which was operated in accordance with standard operating procedures.
      DNA samples were analyzed using the Pathnostics Guidance UTI Test. Samples were mixed with universal PCR master mix and amplified using TaqMan technology on a Life Technologies 12K Flex OpenArray System. DNA samples were spotted in duplicate on 112-format OpenArray chips. Plasmids for each organism being tested for were used as positive controls. Candida tropicalis was used as an inhibition control. A data analysis tool developed by Pathnostics was used to sort data, assess the quality of data, summarize control sample data, identify positive assays, calculate concentrations, and generate draft reports. Probes and primers were used for the following pathogens:
      Bacteria: Acinetobacter baumannii, Actinobaculum schaalii, Aerococcus urinae, Alloscardovia omnicolens, Citrobacter freundii, Citrobacter koseri, Corynebacterium riegelii, Enterobacter aerogenes, Enterococcus faecalis, Escherichia coli, Klebsiella oxytoca, Klebsiella pneumoniae, Morganella morganii, Mycobacterium tuberculosis, Mycoplasma genitalium, Mycoplasma hominis, Pantoea agglomerans, Proteus mirabilis, Providencia stuartii, Pseudomonas aeruginosa, Serratia marcescens, Staphylococcus aureus, Streptococcus agalactiae, Ureaplasma urealyticum
      Bacterial Groups: Coagulase negative staphylococci (Staphylococcus epidermidis, Staphylococcus haemolyticus, Staphylococcus lugdunesis, Staphylococcus saprophyticus); Viridans group streptococci (Streptococcus anginosus, Streptococcus oralis, Streptococcus pasteuranus)

      Statistical Analysis

      Demographics and symptoms were compared for male and female patients with 2-sample t tests or Fisher's exact tests, as appropriate. Noninferiority of Guidance UTI to traditional culture in terms of detecting bacterial infections was assessed by comparing the lower 95% Wilson-score confidence interval (CI) for the positive percentage agreement to the noninferiority threshold of 85%. Incidences of bacterial infections in males and females (according to PCR and culture) were compared with Fisher's exact tests.
      The required sample size to yield 90% power to conclude noninferiority of Guidance UTI relative to traditional urine C&S was calculated assuming a Guidance UTI sensitivity (relative to culture) of 91% and assuming 60% of patients would test positive by culture (based on results in PGX-031 UTI TaqMan QuantStudio 12K Flex Data Analysis). This sample size was calculated in the Tests for One Proportion Procedure of NCSS Power Analysis Statistical Software, Version 14. A sample size of 310 patients with positive culture tests yields 90% power to conclude noninferiority. Assuming 60% of patients test positive for culture, a total of 517 patients should be enrolled in the study.

      Results

      A total of 582 patients, mean age 77 (range 60-95) with symptoms of lower UTI had both urine culture and diagnostic PCR between March and July 2018 (Table 1). Sixty percent (60%, 347/582) were male and 40% (235/582) were female. Clinical symptoms included dysuria (38%, 221/582), incontinence (33%, 192/582), urine that was cloudy or had an odor (23%, 133/582), and pain or discomfort (7%, 40/582).
      Table 1Study participant demographics and symptoms
      ParameterTotalMaleFemalep Value
      p values are from Fisher's exact test or a t-test, as appropriate.
      Age (years)
       n582347235
       Mean (SD)77.2 (7.9)76.9 (7.9)77.7 (7.9).01
       Median777679
       Min, Max60, 9560, 9560, 95
      Symptoms – n (%)
       Dysuria219 (37.6%)110 (31.7%)109 (46.4%).0005
       Urinary incontinence196 (33.7%)145 (41.8%)51 (21.7%)<.0001
       Cloudy or Strong-smelling urine132 (22.7%)75 (21.6%)57 (24.3%).48
       Pain18 (3.09%)7 (2.02%)11 (4.68%).09
        Abdominal7 (1.20%)3 (0.86%)4 (1.70%).45
        Flank5 (0.86%)1 (0.29%)4 (1.70%).16
        Lower back4 (0.69%)1 (0.29%)3 (1.28%).31
        Penis/scrotum2 (0.34%)2 (0.58%)0 (0.00%)1.00
       Pelvic discomfort24 (4.12%)6 (1.73%)18 (7.66%).001
       Lower grade fever1 (0.17%)1 (0.29%)0 (0.00%)1.00
       Agitation2 (0.34%)2 (0.58%)0 (0.00%).52
       Frequency10 (1.72%)4 (1.15%)6 (2.55%).21
       Nocturia1 (0.17%)0 (0.00%)1 (0.43%).40
       Patient-reported hematuria1 (0.17%)0 (0.00%)1 (0.43%).40
      Dipstick results – n (%)
       Hematuria246 (42.7%)138 (40.4%)108 (46.2%).17
      Atypical symptoms – n (%)
       Increased falls/Tripping/Tired/Feeling Ill/Decline in ADLs7(1.2%)6 (1.73%)0 (0.00%)1.00
      Antibiotic usage – n (%)
       Antibiotic treatment in last 3 weeks89 (15.3%)30 (8.65%)59 (25.1%)<.0001
      low asterisk p values are from Fisher's exact test or a t-test, as appropriate.
      Sixty percent of patients (60%, 347/582) had positive results by PCR, urine culture, or both. PCR detected bacteria in 56% of patients (326/582), while urine culture detected pathogens in 37% of patients (217/582) (Table 2). PCR and culture agreed in 74% of cases (431/582): both PCR and culture were positive in 34% of patients (196/582), and both were negative in 40% (235/582). There was disagreement between PCR and culture in 26% of cases (151/582): PCR was positive while culture was negative in 22% of patients (130/582), and PCR was negative while culture was positive in 4% (21/582). The agreement between PCR and urine culture for positive cultures was 196/217 (90%), exceeding the noninferiority threshold of 85% (95% CI: 85.7%-93.6%).
      Table 2Agreement of PCR and traditional urine culture in patients with clinical symptoms of UTI
      PCR
      PositiveNegativeTotalAgreement
      PolymicrobialMonomicrobialTotal Positive
      CulturePositivePolymicrobial30 (5.2%)3 (0.5%)33 (5.7%)6 (1.0%)39 (6.7%)30/39 (76.9%)
      Monomicrobial69 (11.9%)94 (16.2%)163 (28.0%)15 (2.6%)178 (30.6%)94/178 (52.8%)
      Total Positive99 (17.0%)97 (16.7%)196 (33.7%)21 (3.6%)217 (37.3%)196/217 (90.3%)
      Negative67 (11.5%)63 (10.8%)130 (22.3%)235 (40.4%)365 (62.7%)235/365 (64.4%)
      Total166 (28.5%)160 (27.5%)326 (56.0%)256 (44.0%)582(196+235)/582 (74.1%)
      Agreement30/166 (18.1%)94/160 (58.8%)196/326 (60.1%)235/256 (91.8%)(196+235)/582 (74.1%)
      The agreement between PCR and urine culture for positive cultures was 196/217 (90%), exceeding the noninferiority threshold. Polymicrobial infections are those having ≥2 organisms infections. PCR was significantly more sensitive than urine culture for detecting polymicrobial infections.
      Polymicrobial percentage agreement (%, 95% CI) 76.9% (61.7, 87.4%).
      Monomicrobial percentage agreement (%, 95% CI) 52.8% (45.5, 60.0%).
      Negative percentage agreement (%, 95% CI) 64.4% (59.3, 69.1%).
      Overall percentage agreement (%, 95% CI) 61.7% (57.7, 65.5%).
      The multiplex panel used in this study tested for 31 bacteria, and PCR and urine culture together identified 29 different bacterial pathogens. PCR detected 24 bacteria, while culture detected 21 different bacteria.
      The most common organisms detected by PCR were E. coli (29% of PCR positives, 93/326), Actinobaculum schaali (27% of PCR positives, 89/326) and Viridans group Streptococci (27% of PCR positives, 89/326) (Fig. 1). Traditional culture failed to detect Actinobaculum schaali (n = 0), and Viridans group Streptococci was isolated in culture only rarely (6% of culture positives, 14/217). E. coli was also the most common bacterium detected by culture (34% of culture positives, 74/217) and Enterococcus faecalis was the second most common (21% of culture positives, 46/217).
      Figure 1
      Figure 1Frequency of detection of bacteria by PCR and urine culture, ordered in decreasing frequency of detection by urine culture.
      There were 8 bacteria that were identified only by PCR and 5 bacteria that were only detected by culture (Table 3). These 5 bacteria identified only by culture were not included on the multiplex PCR panel. These 5 bacteria were only detected in 8 patients by culture, accounting for 4% of the total positive culture samples (8/217): Enterobacter cloacae as monomicrobial infection in 2 patients (1%, 2/217), Enterococcus faecium alone in 1 patient (0.5%, 1/217), Enterobacter cloacae and faecium in 1 patient (0.5%, 1/217), Enterobacter cloacae and E. coli in 1 patient (0.5%, 1/217), Proteus mirabilis and Streptococcus gallolyticus in 1 patient (0.5%, 1/217), and Kocuria rosea and Kocuria kristinae in 1 patient each (0.5%, 1/217), (Table 3). Three of these 8 patients tested positive to other bacteria by PCR, 1 tested positive to Candida albicans, and 1 tested positive to JC virus by PCR. By contrast, there were 108 patients who tested positive for at least one of the 8 bacteria that were detected by PCR but not by culture. Only 45 of the 108 (42%) tested positive to other bacteria by culture.
      Table 3Bacteria detected only by PCR and only by culture
      Pathogens Detected Only by PCRPathogens Detected Only by Culture
      Actinobaculum schaaliiEnterobacter cloacae
      Alloscardovia omnicolensEnterococcus faecium
      Corynebacterium riegeliiKocuria kristinae
      Mycobacterium tuberculosisKocuria rosea
      Mycoplasma hominisStreptococcus gallolyticus
      Pantoea agglomerans
      Providencia stuartii
      Ureaplasma urealyticum
      The bacteria detected only by culture did not have PCR probes in the multiplex PCR panel used for this study.
      In 88 patients (15%, 88/582), both PCR and urine culture identified the same pathogens and the same number of bacteria. In 85/88 cases (97%) both culture and PCR reported a single pathogen. The most common bacteria with total agreement were E. coli (35%, 30/85), Enterococcus faecalis (19%, 16/85), and Klebsiella pneumoniae (14%, 12/85). Together, these 3 bacteria accounted for 58 of 85 cases (68%) of single bacteria total agreement, but only 29% (194/661) of the total bacteria reported by PCR.
      There were 175 cases of polymicrobial infection, defined as ≥2 bacteria. PCR detected 166 and culture detected 39 (Table 2). There were 30 cases in which both PCR and culture reported polymicrobial infections. There were 3 cases in which culture reported a polymicrobial infection while PCR reported a monomicrobial infection, and an additional 6 cases in which culture reported a polymicrobial infection and PCR was negative. In 1 of these 6 cases, both bacteria detected by culture were not included on the PCR panel; in another case, one of the bacteria detected by culture was not included on the PCR panel. Although it occurs rarely, molecular inhibition can cause PCR tests to be negative. This may have happened with the other 4 cases, which amount to less than 1% (4/582) of patients tested.
      In 67 polymicrobial cases (12% of patients, 67/582; 38% of polymicrobial infections, 67/175) culture was negative. Of the 30 cases in which PCR and culture detected polymicrobial infections, 7 were labelled as “Mixed” by culture so there could be no agreement on the organisms detected, 23 had at least one of the culture detected organisms detected by PCR and 17 had 2 of the culture detected organisms detected by PCR. There were 3 cases in which 2 pathogens were reported and there was total agreement between PCR and culture. In 2 cases, the pathogens were E. coli and Enterococcus faecalis; in 1 case, the pathogens were Proteus mirabilis and Enterococcus faecalis. Although there were 92 cases of at least 3 pathogens, there were none in which PCR and culture agreed on the organisms detected, since culture did not identify the organisms in most of these cases, but just labelled them as “Mixed”. Which bacteria cause infection and inflammation in polymicrobial situations is impossible to determine.

      Discussion

      This study showed that multiplex PCR is not inferior to traditional urine culture, and in fact detected bacteria in 36% of symptomatic patients who had a negative urine culture. In addition, multiplex PCR detected more polymicrobial infections than urine culture, 28% of patients, compared to 7% of patients. In addition to higher detection rates, PCR can provide results in as little as 6 hours compared to traditional culture which takes 48 or more hours. The rapid, accurate identification offered by PCR can facilitate more appropriate and efficacious treatment and may improve clinical care and outcomes.
      UTI is a significant health concern in the United States, causing approximately 7 million visits to a doctor's office, 1 million emergency department visits, and over 100,000 hospitalizations annually, with the cost exceeding $2.6 billion each year.
      • Suskind AM
      • Saigal CS
      • Hanley JM
      • Lai J
      • Setodji CM
      • Clemens JQ
      Incidence and management of uncomplicated recurrent urinary tract infections in a national sample of women in the United States.
      ,

      Simmering JE, Tang F, Cavanaugh JE, Polgreen LA, Polgreen PM. The increase in hospitalizations for urinary tract infections and the associated costs in the United States, 1998-2011. Open Forum Infect Dis. 2017;4:ofw281. https://doi.org/10.1093/ofid/ofw281.

      Although the patients in this study were outpatients, all of them were at least 60 years old. About 40% of men and 28% of women 70-79 years of age will have nonspecific lower urinary tract symptoms (LUTS) that can be clinical difficult to differentiate from UTI.
      • Coyne KS
      • Sexton CC
      • Thompson CL
      • et al.
      The prevalence of lower urinary tract symptoms (LUTS) in the USA, the UK and Sweden: results from the epidemiology of LUTS (EpiLUTS) study.
      ,

      Irwin DE, Milsom I, Hunskaar S, et al. Population-based survey of urinary incontinence, overactive bladder, and other lower urinary tract symptoms in five countries: results of the EPIC study. Eur Urol. 2006;50:1306-1314; discussion 1314-1315. https://doi.org/10.1016/j.eururo.2006.09.019.

      Thus, the diagnosis of UTI on the basis of clinical criteria alone has been reported to have an error rate of approximately 33%.
      • Schmiemann G
      • Kniehl E
      • Gebhardt K
      • Matejczyk MM
      • Hummers-Pradier E
      The diagnosis of urinary tract infection.
      The high numbers of men in this study may reflect that they were enrolled from a busy clinical urology practice that may see more men patients in general. Though women typically have more UTI's than men, the sex differences in this study should not affect the accuracy of PCR vs standard urine culture.
      Polymicrobial infections may be observed in about 39% of UTIs, a potential difficulty in determining appropriate treatment.
      • Price TK
      • Dune T
      • Hilt EE
      • et al.
      The clinical urine culture: enhanced techniques improve detection of clinically relevant microorganisms.
      ,
      • Hilt EE
      • McKinley K
      • Pearce MM
      • et al.
      Urine is not sterile: use of enhanced urine culture techniques to detect resident bacterial flora in the adult female bladder.
      Khasriya showed that urine sediment cultures in patients with LUTS were significantly different than voided urine cultures, and were more commonly polymicrobial, arguing that intracellular bacterial communities and bacteria adherent to the epithelial cells are detected by PCR but not by culture.
      • Khasriya R
      • Sathiananthamoorthy S
      • Ismail S
      • et al.
      Spectrum of bacterial colonization associated with urothelial cells from patients with chronic lower urinary tract symptoms.
      This may be clinically important, since each pathogen carries a unique pattern of antimicrobial resistance, and one bacterial species can confer antibiotic resistance on other bacterial species in a polymicrobial environment. This emphasizes an important limitation of traditional urine culture: the identification of polymicrobial UTIs is poor, whereas the accuracy of PCR for polymicrobial infections was clearly better. PCR can detect the fastidious organisms that are part of the microbiome; some are clearly pathogens while the role of others is under further investigation.
      While other studies have reported higher detection rates using PCR compared to urine culture, most tested against single pathogens, which is clearly not sufficiently comprehensive for clinical use.
      • Preiser W
      • Bräuninger S
      • Schwerdtfeger R
      • et al.
      Evaluation of diagnostic methods for the detection of cytomegalovirus in recipients of allogeneic stem cell transplants.
      • Gaydos CA
      • Theodore M
      • Dalesio N
      • Wood BJ
      • Quinn TC
      Comparison of three nucleic acid amplification tests for detection of chlamydia trachomatis in urine specimens.
      • Pillay A
      • Radebe F
      • Fehler G
      • Htun Y
      • Ballard RC
      Comparison of a TaqMan‐based real‐time polymerase chain reaction with conventional tests for the detection of Trichomonas vaginalis.
      A small number of studies has examined the performance of multiplex PCR, testing for between 9 and 20 pathogens.
      • Lehmann LE
      • Hauser S
      • Malinka T
      • Klaschik S
      • Stüber F
      • Book M
      Real-time polymerase chain-reaction detection of pathogens is feasible to supplement the diagnostic sequence for urinary tract infections.
      • Lehmann LE
      • Hauser S
      • Malinka T
      • et al.
      Rapid qualitative urinary tract infection pathogen identification by SeptiFast® Real-Time PCR.
      • Cybulski Z
      • Schmidt K
      • Grabiec A
      • et al.
      Usability application of multiplex polymerase chain reaction in the diagnosis of microorganisms isolated from urine of patients treated in cancer hospital.
      • van der Zee A
      • Roorda L
      • Bosman G
      • Ossewaarde JM
      Molecular diagnosis of urinary tract infections by semi-quantitative detection of uropathogens in a routine clinical hospital setting.
      In this study, our multiplex PCR tested for 31 bacteria and detected 24 different bacteria; the number of bacteria detected exceeded the size of panels used in previous studies.
      The importance of multiplexing at this level is underscored by detection rates for polymicrobial infections, defined as 2 or more pathogens. Multiplex PCR was better able to detect polymicrobial infections than culture and was also better able to identify the pathogens. Culture was rarely able to detect cases of >3 pathogens and was unable to identify the bacteria in these cases. In this study culture was unable to identify the pathogens in 85% of polymicrobial infections. Such cases may prove to be clinically significant, inasmuch as mutualism may increase resistance and complicate treatment.
      There were a small number of bacteria that were detected by culture but not by PCR. Cases in which culture results were positive and PCR was negative include on 3.6% (21/582) of patients involving 9 bacteria. In 8 of these 21 patients, PCR was negative because the multiplex panel did not include probes for the 5 bacteria infecting those patients. There are limits to multiplexing; it is not currently feasible to test for all known pathogens by PCR. However, judicious selection of PCR targets allows detection and identification of clinically significant bacteria.
      On the other hand, PCR detected 661 instances of 24 different bacteria in 56% of patients with about 50% of those being polymicrobial infections. Another advantage of PCR is that since no probe is used to detect lactobacillus, for example, this bacterium, commonly agreed to be a contaminant, is not detected and reported.
      E coli was the most common bacteria detected for both PCR and culture. However, the rest of the 5 most frequently detected bacteria were very different between PCR and culture. The rest of the top 5 for PCR were Actinobaculum schaallii, Viridans group Streptococci, Aerrococcus urinae, and Coagulase negative Staphylococcus. Actinobaculum schaalii was not detected at all by culture. The rest of the top 5 for culture were Enterococcus faecalis, Klebsiella pneumoniae, Viridans group Streptococci, and Pseudomonas aeruginosa, all of which were detected by PCR at higher rates than in culture. Others have grown these pathogens from the urine with enhanced urine culture techniques, confirming their importance in the urinary microbiome. Their failure to grow in traditional culture reflects their fastidious nature and further demonstrates the limitations of the traditional culture.
      The long-term objective is to determine whether the speed and accuracy of multiplex PCR improves patient care and potentially reduces costs. This study is the necessary first step to confirm that PCR is noninferior to traditional urine culture in identifying bacteria in symptomatic patients. The impact on patient outcomes of this more rapid and accurate organism detection technology will be required to validate the clinical utility before widespread use. Importantly, this multiplex PCR also reports antibiotic resistance for each organism and the susceptibility of the polymicrobial “soup”. These data are still being analyzed and will be reported in a separate manuscript. Finally, the clinical impact of the rapid identification of the organism and the resistance are unknown and are the subject an ongoing prospective trial.
      A potential weakness of this study is ascertaining the clinical significance of the pathogens identified. The urinary microbiome is still being identified and evaluated; some organisms may be pathogens, some protective and some interdependent. Although many bacteria detected in this study are known pathogens for UTI (eg, E. coli, Klebsiella pneumoniae, Enterococcus faecalis),
      • McLellan LK
      • Hunstad DA
      Urinary tract infection: pathogenesis and outlook.
      their contribution to pathogenesis may be less definite in cases of polymicrobial infection. That is, UTI symptoms in polymicrobial infections may be caused by a subset of the bacteria, or some combination. Evidence suggests that polymicrobial infections can form cooperative networks that enhance antibiotic resistance, wherein one bacterial species can confer antibiotic resistance on other bacterial species.
      • de Vos MGJ
      • Zagorski M
      • McNally A
      • Bollenbach T
      Interaction networks, ecological stability, and collective antibiotic tolerance in polymicrobial infections.
      A strength of this study is the inclusion criteria for enrolling patients that experienced urologists felt were symptomatic of UTI and required urine culture with or without empiric treatment. This should make the results of this study generalizable to the clinician who is tasked with evaluating and managing the symptomatic patient with clinical UTI symptoms. Another strength is the use of multiplex PCR for detection and identification of polymicrobial infection, allowing for the detection of a larger number and broader range of different bacteria than traditional culture. These data are being further evaluated to examine bacterial interdependence and patterns of bacterial combinations to identify new paradigms of pathogenesis and potential treatment.

      Conclusions

      This study shows that multiplex PCR is noninferior to traditional urine culture for detection and identification of bacteria in patients with clinical symptoms of UTI. PCR exhibited greater accuracy than culture for pathogens detected and identified bacteria in 36% of patients who had a negative traditional urine culture. PCR was much more sensitive in detecting polymicrobial infections than urine culture. The accuracy and speed of PCR testing over traditional urine culture, and the potential identification of polymicrobial infections with complicated resistance sharing mechanisms is exciting but requires further study to determine the clinical importance.

      References

        • Chang S-S
        • Hsieh W-H
        • Liu T-S
        • et al.
        Multiplex PCR system for rapid detection of pathogens in patients with presumed sepsis – A systemic review and meta-analysis.
        PLoS One. 2013; 8https://doi.org/10.1371/journal.pone.0062323
        • Scerbo MH
        • Kaplan HB
        • Dua A
        • et al.
        Beyond blood culture and Gram stain analysis: a review of molecular techniques for the early detection of bacteremia in surgical patients.
        Surg Infect (Larchmt). 2016; 17: 294-302https://doi.org/10.1089/sur.2015.099
        • Florio W
        • Morici P
        • Ghelardi E
        • Barnini S
        • Lupetti A
        Recent advances in the microbiological diagnosis of bloodstream infections.
        Crit Rev Microbiol. 2018; 44: 351-370https://doi.org/10.1080/1040841X.2017.1407745
        • Munson E
        Molecular diagnostics update for the emerging (if not already widespread) sexually transmitted infection agent mycoplasma genitalium: just about ready for prime time.
        J Clin Microbiol. 2017; 55: 2894-2902https://doi.org/10.1128/JCM.00818-17
        • Sahi S-V
        • Rogozińska E
        • Sobhy S
        • Khan KS
        Accuracy of tests used to detect infection with Chlamydia trachomatis in asymptomatic pregnant women: a systematic review.
        Curr Opin Obstet Gynecol. 2017; 29: 375-382https://doi.org/10.1097/GCO.0000000000000411
        • Gayet-Ageron A
        • Combescure C
        • Lautenschlager S
        • Ninet B
        • Perneger TV
        Comparison of diagnostic accuracy of PCR targeting the 47-Kilodalton protein membrane gene of treponema pallidum and PCR targeting the DNA polymerase I Gene: systematic review and meta-analysis.
        J Clin Microbiol. 2015; 53: 3522-3529https://doi.org/10.1128/JCM.01619-15
        • Berry A
        • Benoit-Vical F
        • Fabre R
        • Cassaing S
        • Magnaval JF
        PCR-based methods to the diagnosis of imported malaria.
        Parasite. 2008; 15: 484-488https://doi.org/10.1051/parasite/2008153484
        • Haas CT
        • Roe JK
        • Pollara G
        • Mehta M
        • Noursadeghi M
        Diagnostic ‘omics’ for active tuberculosis.
        BMC Med. 2016; 14https://doi.org/10.1186/s12916-016-0583-9
        • Zboromyrska Y
        • Vila J
        Advanced PCR-based molecular diagnosis of gastrointestinal infections: challenges and opportunities.
        Expert Rev Mol Diagn. 2016; 16: 631-640https://doi.org/10.1586/14737159.2016.1167599
        • Lehmann LE
        • Hauser S
        • Malinka T
        • Klaschik S
        • Stüber F
        • Book M
        Real-time polymerase chain-reaction detection of pathogens is feasible to supplement the diagnostic sequence for urinary tract infections.
        BJU Int. 2010; 106: 114-120https://doi.org/10.1111/j.1464-410X.2009.09017.x
        • Lehmann LE
        • Hauser S
        • Malinka T
        • et al.
        Rapid qualitative urinary tract infection pathogen identification by SeptiFast® Real-Time PCR.
        PLoS One. 2011; 6https://doi.org/10.1371/journal.pone.0017146
        • Cybulski Z
        • Schmidt K
        • Grabiec A
        • et al.
        Usability application of multiplex polymerase chain reaction in the diagnosis of microorganisms isolated from urine of patients treated in cancer hospital.
        Radiol Oncol. 2013; 47: 296-303https://doi.org/10.2478/raon-2013-0044
        • van der Zee A
        • Roorda L
        • Bosman G
        • Ossewaarde JM
        Molecular diagnosis of urinary tract infections by semi-quantitative detection of uropathogens in a routine clinical hospital setting.
        PLoS One. 2016; 11https://doi.org/10.1371/journal.pone.0150755
        • Price TK
        • Dune T
        • Hilt EE
        • et al.
        The clinical urine culture: enhanced techniques improve detection of clinically relevant microorganisms.
        J Clin Microbiol. 2016; 54: 1216-1222https://doi.org/10.1128/JCM.00044-16
        • Hilt EE
        • McKinley K
        • Pearce MM
        • et al.
        Urine is not sterile: use of enhanced urine culture techniques to detect resident bacterial flora in the adult female bladder.
        J Clin Microbiol. 2014; 52: 871-876https://doi.org/10.1128/JCM.02876-13
        • Croxall G
        • Weston V
        • Joseph S
        • Manning G
        • Cheetham P
        • McNally A
        Increased human pathogenic potential of Escherichia coli from polymicrobial urinary tract infections in comparison to isolates from monomicrobial culture samples.
        J Med Microbiol. 2011; 60: 102-109https://doi.org/10.1099/jmm.0.020602-0
        • Rowe TA
        • Juthani-Mehta M
        Diagnosis and management of urinary tract infection in older adults.
        Infect Dis Clin North Am. 2014; 28: 75-89https://doi.org/10.1016/j.idc.2013.10.004
        • Suskind AM
        • Saigal CS
        • Hanley JM
        • Lai J
        • Setodji CM
        • Clemens JQ
        Incidence and management of uncomplicated recurrent urinary tract infections in a national sample of women in the United States.
        Urology. 2016; 90: 50-55https://doi.org/10.1016/j.urology.2015.11.051
      1. Simmering JE, Tang F, Cavanaugh JE, Polgreen LA, Polgreen PM. The increase in hospitalizations for urinary tract infections and the associated costs in the United States, 1998-2011. Open Forum Infect Dis. 2017;4:ofw281. https://doi.org/10.1093/ofid/ofw281.

        • Coyne KS
        • Sexton CC
        • Thompson CL
        • et al.
        The prevalence of lower urinary tract symptoms (LUTS) in the USA, the UK and Sweden: results from the epidemiology of LUTS (EpiLUTS) study.
        BJU Int. 2009; 104: 352-360https://doi.org/10.1111/j.1464-410X.2009.08427.x
      2. Irwin DE, Milsom I, Hunskaar S, et al. Population-based survey of urinary incontinence, overactive bladder, and other lower urinary tract symptoms in five countries: results of the EPIC study. Eur Urol. 2006;50:1306-1314; discussion 1314-1315. https://doi.org/10.1016/j.eururo.2006.09.019.

        • Schmiemann G
        • Kniehl E
        • Gebhardt K
        • Matejczyk MM
        • Hummers-Pradier E
        The diagnosis of urinary tract infection.
        Dtsch Arztebl Int. 2010; 107: 361-367https://doi.org/10.3238/arztebl.2010.0361
        • Khasriya R
        • Sathiananthamoorthy S
        • Ismail S
        • et al.
        Spectrum of bacterial colonization associated with urothelial cells from patients with chronic lower urinary tract symptoms.
        J Clin Microbiol. 2013; 51: 2054-2062https://doi.org/10.1128/JCM.03314-12
        • Preiser W
        • Bräuninger S
        • Schwerdtfeger R
        • et al.
        Evaluation of diagnostic methods for the detection of cytomegalovirus in recipients of allogeneic stem cell transplants.
        J Clin Virol. 2001; 20: 59-70
        • Gaydos CA
        • Theodore M
        • Dalesio N
        • Wood BJ
        • Quinn TC
        Comparison of three nucleic acid amplification tests for detection of chlamydia trachomatis in urine specimens.
        J Clin Microbiol. 2004; 42: 3041-3045https://doi.org/10.1128/JCM.42.7.3041-3045.2004
        • Pillay A
        • Radebe F
        • Fehler G
        • Htun Y
        • Ballard RC
        Comparison of a TaqMan‐based real‐time polymerase chain reaction with conventional tests for the detection of Trichomonas vaginalis.
        Sex Transm Infect. 2007; 83: 126-129https://doi.org/10.1136/sti.2006.022376
        • McLellan LK
        • Hunstad DA
        Urinary tract infection: pathogenesis and outlook.
        Trends Mol Med. 2016; 22: 946-957https://doi.org/10.1016/j.molmed.2016.09.003
        • de Vos MGJ
        • Zagorski M
        • McNally A
        • Bollenbach T
        Interaction networks, ecological stability, and collective antibiotic tolerance in polymicrobial infections.
        Proc Natl Acad Sci USA. 2017; 114: 10666-10671https://doi.org/10.1073/pnas.1713372114