Non-Fermentative Gram-negative Bloodstream Infection in Northern Italy: a Multicenter Cohort Study


 Background: The management of non-fermentative gram-negative bloodstream infection (NFGN-BSI) offers numerous challenges. In this study the aim is to analyse a large cohort of patients with NFGN-BSI recruited in the northern Italy to describe epidemiology, etiological and susceptibility pattern, therapeutic management and outcome.Methods: Multicentre retrospective cohort study of patients hospitalised at three large teaching hospitals in northern Italy in a fourth year period.Results: 355 BSI episodes were analyzed, due to P. aeruginosa (72.7%), A. baumannii (16.6%), and Stenotrophomonas maltophilia (10.7%). Overall, 21.4% of isolates were defined as DTR, highest rate among A. baumannii (64.4%). All-cause 30-day mortality rate was 17.5%. Rates of XDR or DTR A. baumannii isolation were significantly higher in non-surviving patients. Independent risk factors for 30-day mortality were: age (HR 1.03, 95%CI 1.00-1.04, p=0.003), septic shock (HR 2.84, 95%CI 1.67-4.82, p<0.001) and BSI due to Acinetobacter baumannii (HR 2.23, 95%CI 1.27-3.94, p=0.005). Conclusion: The overall prevalence of DTR was high in the NFGN BSI cohort analyzied, mainly among Acinetobacter baumannii episodes (64.4%). Acinetobacter baumannii is showed to be an independent predictor of mortality. These evidences marked the urgent need of new therapeutic options against this pathogen.Trial registration number: 79/2017/O/OssN. Approved: March14th, 2017.


Background
The management of non-fermentative gram-negative bloodstream infection (NFGN-BSI) offers numerous challenges. Indeed, there are several clinical and microbiological issues that may contribute to its high morbidity and mortality. NFGN-BSI is usually diagnosed in people with severe underlying conditions, critically ill and/or immunocompromised patients [1][2][3] . Isolates are generally resistant, or prone to acquire resistance, to rst-line antibiotics resulting in a high rate of initial inappropriate therapy and/or in the use of less effective and more toxic drugs. To counteract these ndings, antibiotic combination regimens are frequently employed with controversial results in terms of e cacy, toxicity and collateral environmental damage [3,4].
The knowledge of local epidemiology, etiological distribution in terms of causative agents and antibiotic resistance, therapeutic approach and factors associated with poor outcome are useful to guide infection control and antimicrobial stewardship policies and to inform clinicians regarding the best treatment approach [4].
With this premise, we analysed a large cohort of patients with NFGN-BSI recruited in three regions of northern Italy to describe the current epidemiology, etiological and susceptibility pattern distribution, Clinical charts and hospital records were reviewed, until 90 days after the index blood cultures (BCs), to gather study variables using a pre-established case report form. A local senior investigator systematically checked data for accuracy before they were recorded into a database. In addition, the numbers of patient days per year were recorded to assess the incidence of NFGN-BSI in the participating hospitals during the study period.
Ethic Committees of each centre approved the study, waiving of informed consent was obtained due to the retrospective non-interventional study design. Data were collected anonymously.

Participants
Inclusion criteria were all adult (≥18 years) patients diagnosed with NFGN-BSI, de ned as one or more positive BCs obtained from a patient suspected of having infection. Patients were considered only once at the time of rst episode (index BCs).

Variables and de nitions
The primary outcome was all-cause mortality within 30 days after index BCs [5].
BSI was classi ed according to the site of acquisition into nosocomial, healthcare-associated and community acquired using Friedman's criteria [7]. Clinical severity at infection onset was assessed according to SOFA score and new septic shock criteria [8]. BSI sources were established according to Centers for Disease Control and Prevention (CDC) criteria [9]. In the absence of a recognized source, BSI was considered as primary. BSI was de ned as complicated when the infection source was not fully removable.
The susceptibility pattern of isolates was classi ed according to Magiorakos et al. criteria [10] as non-MDR, MDR, XDR or PDR.
Finally the new de nition of "di cult to treat resistance" (DTR) was also assessed as reported elsewhere [11,12] Empirical therapy was de ned as antibiotics administered before the susceptibility report was available. It was considered appropriate when at least one in vitro active drug (according to the susceptibility pattern of the isolate) was administered within 24 h after drawing index BCs. Delayed or no active antibiotic administration within this period was considered as inappropriate empirical therapy. De nitive antibiotic therapy was de ned as antibiotic treatment administered according to susceptibility results. Antibiotic regimens including more than one anti-gram-negative agents, irrespective of their in vitro activity against the BSI isolate, during more than 50% of treatment duration were de ned as "combination regimen". Antibiotic therapy including at least two drugs showing in vitro activity against the BSI isolate was labelled as "2-in vitro active combination regimen". Duration of antibiotic treatment was de ned as the number of consecutive days during which the patient received an appropriate antibiotic regimen. Source control was de ned as the removal of the infection source within 7 days of index BCs, including the performance of non-surgical or surgical procedures to treat an obstructive focus or abscess at any site including, among others, the urinary tract, biliary tract and surgical site, and the removal of any device deemed as the source of BSI.

Microbiology
BCs were incubated using the BACTEC FX Automated Blood Culture System (Becton Dickinson, Franklin Lakes, NJ). All positive BCs were processed with Maldi Biotyper MALDI-TOF system (Bruker Daltonics, Bremen, Germany) for rapid and reliable species identi cation of microorganisms. Antimicrobial susceptibility testing was performed using the Vitek 2 automated system (bioMerieux, Marcy l'Etoile, France) in two hospitals (Bologna and Genova) and the MicroScan system in the remaining hospital (Torino). The minimum inhibitory concentrations (MICs) were interpreted using EUCAST clinical breakpoints for all tested antibiotics.

Statistical analysis
For the descriptive analysis, categorical variables were presented as absolute numbers and their relative frequencies. Continuous variables were presented as the mean and standard deviation if normally distributed or as the median and interquartile range (IQR) if non-normally distributed.
Risk factors for all-cause 30-day mortality were analysed by univariate and multivariate analysis.
Categorical variables were compared using χ 2 or Fisher exact test when appropriate. Continuous variables were compared using the Mann-Whitney U-test. Signi cant and clinically relevant covariates identi ed in univariate analysis were introduced into a multivariable Cox regression survival model after verifying for proportional hazards and collinearity. Signi cance was considered for p<0.05. All the analysis were performed using SPSS software.

Results
Over the study period, 527 patients were diagnosed with NFGN-BSI. Of them, 172 were excluded: 146 patients had polymicrobial bacteremia, and for the remaining 26 patients clinical data were not available. Etiological distribution and susceptibility patterns are shown in Table 1. Most episodes were due to P. aeruginosa (n=258, 72.7%), followed by A. baumannii (n=59, 16.6%), and Stenotrophomonas maltophilia (n=38, 10.7%). Overall, 21.4% of isolates were de ned as DTR with highest rate among A. baumannii (64.4%). Susceptibility rates to individual antibiotic categories are shown in Figure 3. Both P. aeruginosa and A. baumannii maintained high rate susceptibility to colistin.
The general characteristics of study population are shown in Table 2. Overall, 65.6% of patients were male, the median age was 67 (IQR 55-79) years, the median Charlson index was 5.7 (IQR 3.6-7.4), and 24.5% were immunocompromised. Most patients were hospitalized at a medical ward at BSI onset (59.2%), and the majority of episodes were hospital acquired (70.1%). Infection source was not identi ed in 183 (51.5%) cases, in the remaining cases the most common sources of NFGN-BSI were CVC (n=102) and lower respiratory tract (n=49).
As shown in Figure 4, data on empirical and de nitive antibiotic therapy were available for 266 (75%) and 333 (94%) patients, respectively. Active therapy was administered in 35.2% and 73.8% of empiric and de nitive cohort patients, respectively. Empiric combination regimens were used in 55 (15.5%) patients with a 2-in vitro active drugs in 16 (4.5%). Combination regimens were used in 32.4% of patients in the de nitive cohort, 16.3% with 2-in vitro active drugs. Empiric and de nitive antibiotic regimens according to isolates are shown in Supplemental Tables 1 and 2.
All-cause 30-day mortality rate was 17.5%. Relapse at 90 days was observed in 7 patients (2%) within a median of 30 days (IQR 17-50) after index BCs. Compared to patients who were alive at day 30 (Table 2), non-surviving patients exhibited higher age, Charlson index and SOFA score, and higher rate of septic shock. Rates of A. baumannii, XDR or DTR isolation were signi cantly higher in non-surviving patients, while the rates of P. aeruginosa isolation and empirical active therapy were higher among survivors. For de nitive therapy, the use of any combination was associated with higher mortality while that of 2-in vitro active combination did not. At multivariate analysis adjusted for DTR, active empiric therapy, active targeted therapy and source control, the independent risk factors for all-cause 30-day mortality were: age (HR 1.03, 95%CI 1.00-1.04, p=0.003), septic shock (HR 2.84, 95%CI 1.67-4.82, p<0.001) and BSI due to Acinetobacter baumannii (HR 2.23, 95%CI 1.27-3.94, p=0.005).

Discussion
In our large cohort of NFGN-BSI we have found high rates of DTR and carbapenem resistance, especially among A. baumannii. Empirical active therapy administration was signi cantly higher in surviving patients at univariate analysis, however it was not con rmed at multivariate analysis. Combination therapy, also with 2 active drug, was not associated to improving surviving at multivariate analysis. Of note, A. baumannii isolation resulted as independent risk factors for mortality at multivariate analysis. In our cohort nosocomial infections accounted for a large majority of cases (70%), mainly CVC related and pneumonia, according to literature [13,14]. Therefore, promote and improve infection control programs would play a critical role in reducing the rates of this kind of nosocomial infections.
The increasing importance of the NFGN bacteria is also related to their complex antimicrobial resistance pro le. In our cohort carbapenem resistance showed high prevalence, with 31.8% and 64.4% rates for P. aeruginosa and A. baumannii respectively. We have also analysed the prevalence of the new de nition DTR. This de nition re ects the use of second-line agents with poorer therapeutic index, resulting in a better prediction of poor outcome. In our cohort, the overall prevalence of DTR was 21.1%. It varied across species being highest among A. baumannii BSI with rates of 64.4%. This value was much higher than data present in literature [11][15]. Similarly, DTR prevalence for P.aeruginosa BSI, accounted for 13.6%, higher than the rates showed previously [11][15]. As expected, in A. baumannii strain, CR and DTR rates were comparable.
In our study active therapy did not result statistically associated with improved outocome as previously reported elsewhere. This nding deserves further investigation. Indeed, the classical way to de ne if an antimicrobial agent is useful to treat an infection is the MIC determination of strains. However, MIC determination have some concerns: i) clinical laboratories cannot determine MICs with su cient accuracy owing to the assay variation in the MIC test especially when automated or semi-automated methods are used, ii) the MIC does not represent a concentration directly compared with in vivo concentration found during treatment; iii) bacterial growth conditions in vitro could be different from those in vivo [16]. Also, the in vitro activity of antimicrobial often does not re ect the clinical feasibility due to the speci c pharmacokinetic/toxicodynamic pro le of the drugs and the source of infection [17].
These considerations could explain why in our cohort active therapy seems not associated to improving surviving. Similar experiences were previously reported elsewhere [18].
In our study, A. baumannii was an independent predictor of mortality. This is line with the characteristics of this pathogen that is commonly responsible for severe opportunistic nosocomial infections mainly in hospitalized immunocompromised patients [1][2][3]. Additionally, the complex antimicrobial resistance pro le and the limited therapeutic arsenal for this strain may explain this result. In this scenario, polymyxins remains in vitro the most active agent. However, the in vitro activity of polymyxins not re ect the clinical feasibility due to the suboptimal pharmacokinetic/toxicodynamic pro le of this class [17].
Our study has a number of limitations. Although we have analysed a large cohort of patients in three different centres, the results could be in uenced by the epidemiology of a restricted area of our country. Also, our cohort is from all large tertiary teaching hospital re ecting the complexity and epidemiology of patients managed in similar institutions. The retrospective collection of patient and microbiological data could have limited integrity and accuracy. However, a senior investigator and three young investigators revised all CRFs, and reconciled data reports and missing data with the medical records before including information in the database.
To conclude, the overall prevalence of DTR was high in our NFGN BSI cohort, mainly among Acinetobacter baumannii episodes. Furthermore, Acinetobacter baumannii is showed to be an independent predictor of mortality. These evidences marked the urgent need of new therapeutic options against this pathogen. Availability of data and materials The original data and materials from this study are available from the corresponding author on reasonable request.

Competing interests
All authors no reported competing interests.

Funding
No external funding was received for the present study.
Authors' contributions RP: data analysis and drafting the manuscript, SC: data collection, LB: data collection, LP: data collection, DRG: data collection,SA: support in data collection, TL: data collection, CC: support in data collection, AM: support in data collection, CV: support in study design, FGDR: support in study design,MB: manuscript revision ,MG: study design, data analysis and manuscript revision,PV: study design and manuscript revision  TMP/SMXR Trimethoprim/sulfamethoxazole resistance, XDR extensively drug-resistance. Figure 1 Flow diagram of study population.

Figure 2
Overall incidence of P. aeruginosa and A. baumannii BSI per 1000 patient days between 2013-2016 years.

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download. TablesAdditional.docx