A 5-Year Retrospective Analysis of Raoultella planticola Bacteriuria

Introduction

Infections related to rare pathogens are very challenging to treat due to lack of enough data in the medical literature.^1,2 Raoultella planticola is a ubiquitous, non-motile, aerobic gram-negative bacteria. It belongs to the genus Raoultella under the Enterobacteriaceae family. Two species are clinically important: [1] Raoultella planticola and [2] Raoultella ornithinolytica. The microbiological identification of R. planticola remains a challenge even after 30 years of its identification.^3,4 This is because the conventional phenotypic assessments often misidentify it as Klebsiella spp., which is its remarkably close relative in the Enterobacteriaceae family. This might be one of the many reasons that the incidence of infections caused by R. planticola is underreported.

The gastrointestinal tract and the upper respiratory tract are the typical reservoirs of R. planticola.⁵ It commonly causes pneumonia, biliary tract infections, and bacteremia.^6–9 Only a handful of cases of urinary tract infection (UTI) caused by R. planticola in adults have thus far been published in the literature.^10–16 Most of the UTIs have been reported in immunocompromised adults, in patients with urinary tract procedures or instrumentation.^11–13,15

The recent emergence of extended-spectrum β-lactamase (ESBL) producing and carbapenem-resistant nosocomial R. planticola infections is a matter of concern. These strains can act as a reservoir for these resistance genes.¹⁷

Except for a few case reports/series, there has been no published comprehensive review or original study on R. planticola-associated UTIs. We conducted a retrospective study to better understand the characteristics of the organism, risk factors for acquisition of infection and to analyze the outcomes associated with UTI caused by R. planticola.

Materials and Methods

Search Strategy

This was a retrospective study conducted at a community-based teaching hospital in Massachusetts, United States. The study included data from January 2015 to September 2020 and included all urine cultures positive for R. planticola at our institution. The study was approved by our institutional review board (IRB #2020 −142).

Selection and Inclusion Criteria

All adult patients (>18 years) who had a urine culture growing R. planticola were included in the study. Primary identification and sensitivities of colonies were done with VITEK^® 2 compact automated system and confirmation was done with API 20-E, both from BioMérieux.

Data Extraction

Records of selected patients were reviewed and information including the demographics, clinical symptoms, physical examination findings, laboratory data, radiological parameters and microbiological data were extracted. Further details on hospitalization including the severity of the illness, antibiotics administered, hospital course and outcomes were also obtained. The data extracted was verified by two independent researchers. The data was then entered into a spreadsheet.

Data Analysis

All categorical data were expressed as counts, whereas continuous data were expressed as a mean. The data was analyzed using Google sheets^®. No statistical analysis was performed given the small sample size.

Results

Demographic Details

A total of 37 patients had a positive urine culture for R. planticola. The demographics of the patients are listed in Table 1. The mean age of our cohort was 77 years (47 years - 91 years). There were 28 females and 9 males in total. Out of our 37 patients, 9 were residents of nursing homes or group homes, and 28 presented from home. The body mass index (BMI) was measured to be <19 kg/m² in 3 patients, between 19kg/m² to 25 kg/m² in 15 patients, between 25 kg/m² to 30 kg/m² in 10 patients and more than 30 kg/m² in 9 patients. The mean BMI was 26.64 kg/m².

Table 1 Patient Demographics and Predisposing Factors

Risk Factors

We studied the systemic and genitourinary factors in our population that were associated and/or have prognostic significance to our patients with R. planticola-associated UTIs (Table 1).

Systemic Characteristics

Alcohol consumption was reported by 5 patients while 32 patients denied it. None of the patients was active smokers with 17 endorsing prior tobacco use. Amongst the 37 patients, 16 patients had a history of diabetes mellitus. Liver dysfunction was present in 7 patients. Chronic corticosteroid and immunosuppressant use were reported by 1 patient each, and chemotherapy by 2 patients.

Genitourinary Characteristics

A history of chronic kidney disease was present in 18 patients. Out of these 18 patients, 5 had an acute on chronic kidney injury. There was a history of a urological procedure or genitourinary instrumentation in 4 patients, with 2 having the procedure within a month of their UTI. An underlying structural abnormality of the genitourinary tract was present in 9 patients as depicted in Figure 1. A history of renal transplantation was present in 1 patient.

Clinical Features

As shown in Table 2, most of the patients were asymptomatic at the time of presentation. The symptomatic patients presented with the following symptoms altered mental status (6 patients), generalized fatigue (5 patients), fever (4 patients), dysuria (2 patients), flank pain (1 patient), the combination of fever, suprapubic pain, and increased frequency (1 patient), flank pain and turbid urine (1 patient), and altered mental status and increased urinary frequency (1 patient).

Table 2 Clinical Features of Patients with R. planticola-Associated UTIs

Four patients were bradycardic, and 8 patients were tachycardic at the time of presentation. Only 8 patients were febrile (>100.4^O F) at the time of presentation. Hypotension was present in 1 patient and the rest of the patients had systolic blood pressure greater than 90 mm Hg.

Laboratory Data

The laboratory results of the patients are depicted in Table 3. The mean white blood cell count was 8.96 x 10⁹/L (2.3 × 10⁹/L - 20.5 × 10⁹/L), platelet count was 246 x10⁹/L (52 × 10⁹/L - 599 × 10⁹/L), glucose was 152 mg/dL (49 mg/dl - 684 mg/dL), creatinine was 1.31 mg/dL (0.53 mg/dl - 6.3 mg/dL), and GFR was 56.87 mL/min (5.7 mL/min - 111.6 mL/min). A glycated hemoglobin level was measured only in 12 patients and the mean was 8.39% (5.3%- 13.7%). An elevated lactate level was present in 3 patients.

Table 3 Laboratory Results

Culture, Coexisting Organism, and Antibiotic Sensitivity

The microbiological data are detailed in Table 4. Resistance to ampicillin was found in 35 isolates, which seems to be an intrinsic character of the Raoultella spp., and 2 isolates were multidrug-resistant, but still susceptible to ciprofloxacin. Concomitant growth of other organisms was found in 9 urine culture samples; namely, Escherichia coli (5 cultures), carbapenem-resistant Pseudomonas aeruginosa (1 culture), Enterococcus sp. (1 culture), Group B Streptococcus (1 culture), and mixed gram-positive organisms (1 culture). None of the patients had positive blood cultures.

Table 4 Microbiological Characteristics

Clinical Course, Hospital Management, and Outcome

The severity of illness varied widely ranging from asymptomatic bacteriuria to septic shock as depicted in Table 5. Antibiotics were administered to 24 patients while 13 patients did not receive any antibiotics. The mean duration of antibiotic administration was 7.54 days. Ciprofloxacin was the most commonly used antibiotic and was administered to 9 patients. Ceftazidime/avibactam was used in the UTI caused by drug-resistant isolates. All patients responded to treatment. The mean length of stay was 3.18 days.

Table 5 The Clinical Course of the Patients with R. planticola-Associated UTIs

Discussion

Urinary tract infections can range from simple asymptomatic bacteriuria to complicated ascending tract infections leading to bacteremia and sepsis.¹⁸ The risk of complications is more pronounced in diabetics, elderly people with indwelling catheters, and immunocompromised individuals.^19–21 Raoultella planticola has been rarely reported to be significantly associated with UTI. To the best of our knowledge, this is the first review of the clinical characteristics of R. planticola-associated UTIs.

R. planticola UTI seems to occur approximately three times more common in females in our study population, which mirrors the gender distribution for UTIs with other organisms.²²

Advanced age, immunocompromised condition like cancer, diabetes mellitus and impaired renal function were noted to be significant risk factors for developing R. planticola UTI, similar to the risk factors in UTIs due to other organisms.^23,42 The mean age of our study population was 77 years. A diagnosis of diabetes mellitus was present in 43.24% (16/37) of the patients and 21% (8/37) had a glycated hemoglobin (HbA1C) >6%.

A BMI of greater than 25 kg/m² was present in 51.35% (19/37) of patients and the mean BMI of this cohort was 26.64 kg/m². Our study was underpowered to determine if obesity was an independent risk factor.

A large study of UTI in diabetics, done in Germany, has suggested that a GFR of less than 60 mL/min increases the risk of UTI.²³ However, it is unclear if chronic renal insufficiency is a risk independent of age and diabetes mellitus.²⁴ Twenty-three patients (including one patient who was post-renal transplant) in our study population had chronic kidney disease (GFR < 60 mL/min). In the subset of patients with chronic renal insufficiency, 52% (12/23) did not have a diagnosis of diabetes mellitus. Therefore, it is likely that impaired renal function increases the risk of R. planticola-associated UTIs. However, our study was not powered to make that determination.

In our study, the most common presentation was asymptomatic bacteriuria (16/37, 43.24%). However, 4 patients (10.81%) in the study population fulfilled the SIRS criteria for sepsis. Of those, 2 patients (5.4%) went into septic shock requiring aggressive intravenous fluid therapy and/or pressor support. Levy et al found that approximately 9–31% of all cases of sepsis can be attributed to urinary tract infection (depending on the geographical region).²⁵ Based on studies, Klebsiella spp. is responsible for 15% of all cases of urosepsis.^26,27 Due to its close similarity to Klebsiella spp., it is possible that many of those could have been due to R. planticola and may have been misdiagnosed in the past. There have been a few case reports of Raoultella bacteremia.^6–9,28,29 No cases of bacteremia secondary to UTI were observed in our study. With the increasing use of techniques such as Matrix-Assisted Desorption Ionization–Time of Flight Mass Spectrometry (MALDI-TOF MS), R. planticola is being detected more frequently than in the past.^28,29

Multiple studies have demonstrated that UTIs present differently in the elderly.^30,31 A study performed by D’Agata et al in-nursing home residents showed that altered mental status is by far the commonest presentation (approximately 40%). The classical UTI symptoms of dysuria (3.8%), costovertebral tenderness (2.3%), urinary frequency (1.5%), urgency (0%) and suprapubic pain (0%) are much less common in adults greater than 65 years.³⁰ Our study demonstrated similar findings.

As mentioned above, due to phylogenetic similarities, it is difficult to differentiate between Klebsiella spp. and Raoultella spp. by microscopy. The biochemical tests to differentiate Raoultella spp. from Klebsiella spp. such as ornithine decarboxylase activity, histamine or D-melezitose utilization, is not routinely available in commercial test kits.^26,27 MALDI-TOF MS has emerged, in recent years, as a faster technique to identify Raoultella spp. and its reliability has been proven by multiple studies.^5,32–34 It essentially involves taking a sample from a bacterial colony, absorbing the lysed bacterial proteins onto a matrix, followed by ionizing and desorbing it with a laser. The resultant plume is analyzed through mass spectrometry to detect the signature pattern that identifies the bacteria. An additional advantage is the rapid detection of resistant strains. This technique is very effective in gram-negative bacteria, even with microcolonies.³⁵ One limitation is the need for pure colonies; mixed colonies can lead to erroneous results.

In our laboratory, we used the VITEK^® 2 compact automated system microbial identification (ID) and antibiotic susceptibility testing (AST) card for identification of organisms. The VITEK^® 2 compact automated system uses 47 biochemical tests on a compact card specifically designed to identify gram-negative bacilli and their sensitivity. This technique usually identifies the organisms within 10 hours.³⁶ A frequently encountered problem with the system is the misidentification of Raoultella spp. as Klebsiella spp.³² To avoid such errors, Analytical Profile Index-20E (API-20E) was used to differentiate between members of the Enterobacteriaceae family in our laboratory. This technique uses a miniaturized version of 20 biochemical tests. The positive and negative results are compiled to obtain a unique profile code used to identify the organism. The sensitivity for identification of Raoultella spp. is 93.3% in VITEK^® 2 compact automated system, and 97.4% in the MALDI-TOF MS method. Though MALDI-TOF is better than VITEK^® 2 compact automated system in terms of sensitivity, both systems detect Raoultella spp. with a high degree of accuracy and the high initial cost of the MALDI-TOF system has restricted its use mostly to higher centres.

R. planticola is intrinsically resistant to ampicillin due to the over-expression of chromosomally encoded class-A β-lactamase.^31,32 The other two major groups of resistance genes seen in Raoultella spp. are extended-spectrum β-lactamase (ESBL) and carbapenemase genes. Due to the ubiquitous nature of Raoultella spp, there is a very high risk of them acting as environmental reservoirs for resistance genes, such as carbapenemase.^37,38 An analysis of sewage water from a tertiary centre in Spain revealed the presence of multiple carbapenemases producing Raoultella spp.³⁹ These resistance genes can then be carried on mobile genetic elements like transposons and plasmids which are capable of transforming naive bacteria. In our study population, 2 patients had multidrug-resistant R. planticola-associated UTI. Both of these strains were resistant to ceftazidime which is routinely used in our institution to cover urinary gram-negative bacteria as per our hospital antibiogram. Though not reported, we can assume that this isolate probably had ESBL resistance gene/genes. Out of these two patients, one had a hospital-acquired catheter-associated UTI with a coexisting carbapenem-resistant Pseudomonas aeruginosa. The other patient had a coexisting Escherichia coli infection.

In addition to this, another 9 patients had a concomitant growth of a second organism found on urine culture. This is particularly concerning as mobile genetic elements, especially plasmids (chiefly IncF, IncI, IncA/C, IncL, IncN, and IncH plasmids) can transmit resistance genes between Enterobacteriaceae.^40,41

Limitations of the Study

Due to the low prevalence of R. planticola-associated UTIs in general, the sample size was small in our study. We were not able to establish the statistical significance of our findings as a result of the small sample size. As this was a retrospective study, we encountered missing data. For example, not all patients had glycated hemoglobin levels analyzed during their hospitalization. Information on whether a urinary catheter was placed during the current hospitalization or whether it was chronic was not documented in many patient’s charts. This left us unable to determine whether some of the UTIs were true infections or colonization that was discovered incidentally. Similarly, details on the structural abnormalities of the genitourinary tract and outpatient urological procedures were missing sometimes. The samples were analyzed by VITEK^® 2 compact automated system microbial identification (ID) and antibiotic susceptibility testing (AST) card which is less accurate than a MALDI-TOF MS.

Conclusion

Our study revealed that patients with R. planticola UTI had higher proportion of diabetes mellitus, renal failure compared to the general population. Unlike prior case reports, most of the infections occurred in immunocompetent patients. Our study also confirms the intrinsic resistance to ampicillin of R. species, which has been documented previously in the literature. Surprisingly, Escherichia coli seems to coexist with Raoultella spp. in a significant number of cultures. A worrying finding was the presence of multidrug-resistant isolates, one of which was associated with multidrug-resistant Pseudomonas aeruginosa, which raises concern for the transmission of resistance genes. This raises concern for the transmission of resistance genes. As our study was limited by relatively small sample size and the retrospective nature of the analysis, larger studies would help us further define the observations noted in this study.

Figure 1 Underlying structural abnormality of the genitourinary tract.