Clinical Features and Outcomes of Raoultella Terrigena Infections. A Single-center Experience from Karachi, Pakistan: A Retrospective Study

Raoultella terrigena (formerly Klebsiella terrigena) is an environmental gram-negative rod. It may cause infections in humans, especially in the elderly and immunosuppressed patients. Moreover, this organism tends to be multi-drug resistant, limiting treatment options. Evidence on clinical presentation and outcomes of this infection is limited. We conducted this study to see its various clinical features, available treatment options, and associated clinical impacts of the R.terrigena infections. We conducted a cross-sectional study on all adult patients with clinical specimens positive for Raoultella terrigena


Background
Raoultella Terrigena, initially known as "Klebsiella Terrigenna," was discovered in 1981 and is a rare gramnegative organism, mostly found in soil and water 1 . It was separated from Klebsiella species in 2001 based on molecular analysis 2 . These organisms are oxidase negative, capsulated, immotile, facultatively anaerobic, and aerobic bacilli 2 . However, many microbiology laboratories continue to identify this organism as a "Klebsiella species," making the true incidence and pathogenesis uncertain 3 . The rst reported case as a human pathogen was in 2007 in a middle-aged post-liver transplant patient with endocarditis 4 . Another case report was published on sepsis secondary to R. terrigena in 2011 5 . A literature search reveals that most of the reported cases caused by genus Raoultella consist of R. orinithinolytica and R.planticola 6,7 , and are associated with diseases of the biliary tract and surgical interventions 6,8 as well as necrotizing fasciitis 9 ,cellulitis 10 , cystitis 11 ,cholangitis 12 , cholecystitis 13 , and pancreatitis 14 . However, despite reviewing the literature, we could not nd a sizeable descriptive study on clinical presentations and outcomes on Raoultella terrigena.
Ethical review: The study started after receiving approval for exemption from the Aga Khan University Ethics review committee. (ERC Reference No:2019-1232-3287). Patient con dentiality was maintained by not taking personal identi ers such as names or medical record numbers. As this was a retrospective study, the committee waived the requirement of informed consent.

Methodology
We conducted a cross-sectional study on all adults above 18 years of age, admitted to Aga Khan University Hospital (AKUH) between (January 2013 to December 2018), who had Raoultella terrigena in cultures. We excluded patients with recurrent Raoultella infections and colonizers. Patients who ful lled the eligibility criteria were consecutively included in the study. Data on patient-related variables such as age, sex, co-morbid, prior history of hospitalizations, invasive procedures, and use of antibiotics, etc., were collected from the Health care information and management system (HIMS) on a pre-tested structured proforma. Microbiological culture specimens related data extracted from Integrated Laboratory Management Services records.

Laboratory Identi cation
The identi cation process was performed in the microbiology laboratory at Aga Khan University Hospital Karachi. This involved culture inoculation on MacConkey, chocolate, and 5% sheep blood agar at 37 degrees Celsius for up to 48 hours. The blood agar was made by phlebotomizing sheep in the animal house at AKUH. Upon growth of lactose fermenter mucoid colonies on MacConkey agar, further biochemical tests involve the utilization of citrate, hydrogen sul de, urease, indole, and motility tests negative within acidic/acidic reaction on Tripe Sugar Test.
Further identi cation was carried out by API 20 E, which detects 91% to 95% of cases correctly. Antibiotic susceptibility testing was initially carried out by Mueller-Hinton agar with Kirby-Bauer disk diffusion test or Vitek-2 MSautomatedc system. Colistin minimum inhibitory concentrations were con rmed by colistin broth microdilution, which is currently the most reliable method. Results were interpreted as per Clinical and Laboratory Standards Institute antibiotic guidelines.

Statistical Methods
Descriptive analysis was performed for all patient-related variables with frequencies and proportions reported for categorical variables like sex, comorbid, clinical features, etc. And median with interquartile range reported for continuous variables like age, hospital stay. Chi-square test or Fisher exact test were used as appropriate to determine the association between two categorical variables, e.g., chronic kidney disease and death. SPSS software (version 25.0) was used for data analysis. A P-value of less than 0.05 is supposed to be statistically signi cant.

Results
There were 58 patients with R.terrigena isolated from different culture specimens. Out of those, 12 patients were identi ed as colonizers, remaining included in the study (see Figure-1). The median age was 61.5 years (IQR=28), with more males than females (60.9 % vs. 39.1%.) The most frequent co-morbid conditions were diabetes mellitus in n=18 (39.1%), followed by chronic kidney disease in n=16 (34.7%) and malignancy in n=11 (23.9) (see table-1). Most patients n= 34 (73.9%) had a previous history of hospitalization (within the past six months), and in the majority, n=24 (52.17%) had at least one hospital admission. 78% patients (n=36) reported antibiotic use in last 6 months. The most commonly used antibiotics were carbapenems in n=26 (56.5%). Also beta-lactam, beta-lactamase inhibitors in n=23 (50%) and glycopeptides in n=20 (43.5).
Out of 46 cases, eight patients lost to follow up, and treatment information was not available. The remaining 38 patients, n=31 patients (81.6%) received combination therapy, and n=6 (15.8%) received monotherapy. One patient died before starting treatment. The most frequent combination was carbapenem and colistemethate in n=11 (28.9%), followed by a combination of carbapenem with colistemethate and tigecycline in n=8 (21.1%) ( Table-3). In 71.1% cases, the treatment was inappropriate for R-terrigena while the remaining 28.9%, given appropriately. Mortality association with monotherapy was (p= 0.672), and with combination therapy (p= 0.70). Out of n=14 (36.84%) cases in whom repeat cultures for clearance needed, bacteriological clearance was achieved in n=9 (64.2%) cases. There were 23 cases that needed source control of underlying infection. It was achieved in n=12 (52.1%). Two cases (5.3%) relapsed.
The average hospital stay was a median of 11.50 days (IQR=17), with a median of 3 days before positive culture. Approximately n=23 (60.52%) of patients were seriously ill, requiring intensive care unit care with a median ICU stay of 6 days (IQR=7). In-hospital mortality recorded in n=17 (44.7%) patients.
In the subgroup analysis of factors associated with death in R.terrigena infections, it was found that chronic kidney disease (CKD) ((p=.029) and septic shock (p=.005) were signi cantly associated with mortality. Also, persons with a high (greater or equal to three) Charlson-comorbidity index had increased mortality (p=0.002).

Discussion
In our study, most patients had hospital-acquired respiratory tract infections, followed by urinary tract infections. Clinical symptoms identi ed in the vast majority were respiratory failure, septic shock, and altered mental status. The grown isolates were highly resistant to beta-lactams, carbapenems, and colistemethate.
Its clinical impact remained mostly unknown until 2007 when a middle-aged liver transplant person suffered infective endocarditis with R-terrigena (Table-4). Before his surgery for liver transplantation, he was found to be a colonizer with ESBL R-terrigena and was given piperacillin-tazobactam as a prophylactic agent. The organism identi ed after the death of the patient in tissue (aortic valve) 4 .
Previously stated that infections caused by the genus Raoultella are common in older-aged groups who were suffering from malignancy or underwent surgical interventions. Many of them developed infections of the biliary tract and had different mortality rates. 15 In our study group, most patients were male in the middle and older aged group, who were diabetics in the vast majority, followed by CKD and cancer patients. Around two-thirds of them had recent surgical interventions. We had only one patient with biliary tract involvement, and we found a higher rate of nosocomial pneumonia and urinary tract infections in our patients. (table-1).
Another case published in 2011 was about a monomicrobial R.terrigena infection. It was causing postlaparotomy intrabdominal sepsis with uid encasement 5 (Table-4). The organism identi ed in blood and drain cultures was sensitive to amoxicillin-clavulanic acid, cephalosporins, cipro oxacin, and imipenem.
The patient was treated successfully with imipenem piperacillin-tazobactam. In 2016, a case of subungual abscess caused by R.terrigena reported, resulting from working in a eld. It was pan sensitive and treated with imipenim 16 (Table-4). However, in most of our patient population, the grown isolates were highly resistant to many drugs, including carbapenems and colistemethate. These ndings also present in a recent small case series of three patients published from Pakistan in 2019 (Table-4). In that all three patients were female, two of them had diabetes and one had CKD. All patients had a complicated hospital course; they were already on treatment for other hospital-acquired infections. Two patients had complicated urinary tract infections; they grew highly drug-resistant R.terrigena in urine and blood cultures. One died and the other was critically ill and transferred to another hospital. The remaining one patient suffered nosocomial respiratory tract infection by R.terrigena sensitive to colistemethate and tigecycline in both respiratory and blood samples. This patient also died despite appropriate treatment 17 .
In comparison, we found a higher frequency of nosocomial pneumonia in our patients. Similarly, we also found increased mortality among our infected cases.
Although our study has a limited sample size and is a single-center experience, it is the largest cohort of patients reported of this rare gram-negative infection to date. In this era of hard to treat hospital-acquired infections and associated antibiotic resistance, we face challenges treating these kinds of pathogens.
These emerging opportunistic pathogens and associated drug resistance leave us with a limited choice of antibiotics. Interestingly, we found no outcome difference in treating patients with monotherapy against combination therapy. Patients with many comorbid conditions and immunosuppression can lead to infections with opportunistic organisms, including R.terrigena, leading to increased mortality if not identi ed and treated correctly.

Limitations
Because of the scarcity of available resources, we could not detect R.terrigena by molecular analysis (standard method).

Conclusion
Infections caused by R.terrigena are highly drug-resistant, mostly causing hospital-acquired respiratory tract infections that are di cult to treat, leading to prolonged hospital and ICU stays with a high mortality rate. Patients with underlying renal dysfunction, those on vasopressor support, and a high CCI score are at greater risk of death due to this infection.

Declarations
Authors' Contributions: Competing Interest I declare that the authors have no competing interests (Financial or non-nancial) as de ned by BMC, or other interests that might be perceived to in uence the results and/or discussion reported in this paper.
Ethical approval and consent to participate: Our study was reviewed and approved as exemption by Aga Khan University (AKU) ethics review committee (Refrence number: 2019-1232-3287). The study was performed in accordance with the Declaration of Helsinki and guidelines and regulations of Aga Khan University ethics review committee. The requirement for written informed consent was waived as the data was collected retrospectively from medical records and was de-identi ed, which had been approved by the the Aga Khan University (AKU) ethics reveiw committee.

Consent for publication:
Not applicable Animal Ethics: We con rm that all experimental protocols involving animals were approved by the Aga Khan University licensing committee and all methods were carried out in accordance with ARRIVE guidelines.
Availability of data and material: The datasets generated and/or analyzed during the current study are not publicly available due to institutional policy but are available from the corresponding author on reasonable request.

Funding:
Not applicable