INHALATORY AND INTRAVENOUS COLISTIN IN TREATING VENTILATOR-ASSOCIATED PNEUMONIA DUE TO ACINETOBACTER SPECIES : SHOULD WE COMBINE THEM ?

Background/Aim. Acinetobacter is one of the most common causes of nosocomial infections, especially ventilatorassociated pneumonia (VAP). Considering the increased presence of multidrug-resistant microorganisms and the lack of novel antibiotics, colistin merged as the last-resort antibiotic for life threatening nosocomial infections. Intravenous use of antibiotics is accepted as a gold standard for the treatment of pneumonia, but additional administration of inhaled antibiotics in the treatment of VAP has shown to be advantageous in some clinical trials. The aim of this study was to investigate the effect of inhalatory colistin as an adjunct to intravenous colistin on the survival of patients with VAP caused by Acinetobacter species. Methods. We conducted a retrospective study to evaluate the efficacy of combination of inhalatory and intravenous colistin vs. intravenous colistin alone in 69 patients in the Intensive Care Units (ICU) with VAP caused by Acinetobacter baumannii. The patients were treated in the ICU at the Institute for Pulmonary Diseases of Vojvodina, Sremska Kamenica (Serbia) in the period from January, 2013 to March, 2018. Baseline demographic data, severity of the disease, comorbidities, colistin regimen and length of the treatment were collected. The primary outcome was 28-day mortality. Results. Twenty seven of total 69 (39.1%) patients received combined intravenous and inhalatory colistin. Forty two (60.9%) patients received only intravenous colistin. Compared to the combined use of the drug (intravenous and inhalatory colistin), patients receiving intravenous colistin alone had a significantly increased risk of death during 28 days [25.9% vs. 61.9%, respectively; odds ratio (OR) 4.464, 95% confidence interval (CI) 1.539?2.925; p = 0.006]. Length of colistin use (> 7 days) was also associated with reduced survival (OR 0.22; 95% CI 0.080?0.606; p = 0.003). After adjusting for baseline severity of the illness (APACHE score) and length of colistin treatment, patients receiving only intravenous colistin had greater 28-day mortality rate compared to patients receiving both intravenous and inhalatory colistin (OR 6.305; 95% CI 1.795?22.153; p = 0.004). Conclusion. Our results suggest that adding inhalatory to intravenous colistin might be beneficial in the treatment of VAP caused by Acinetobacter species.


Introduction
According to the Cochrane database review ventilator associated pneumonia(VAP) occurs in 10% of mechanically ventilated patients (1).Earlier studies reported that depending on the underlying conditions and the pathogenicity of the infecting organisms,the mortality rates varied from 10% to 70% (2)(3)(4).As stated in guidelines of the Infectious Diseases Society of America (IDSA) and the American Thoracic Society (ATS), the empirical treatment of VAP is based on the risk assessment of multidrug resistant infection.Inadequate initial therapy is associated with higher mortality and prolonged length of stay in intensive care unit (ICU LOS) (5).Early application of adequate antibiotic therapy is of crucial importance in the treatment of VAP.Postponement of antibiotic application as well as inadequate antibiotic therapy, even when later changed according to microbiological cultures, lead to higher mortality (6).The choice of therapy should be based on the initial microbiological map, minding theside effects, as well as the previousantibiotic therapy in the last two weeks (5,7).Due to its high virulence and increased antimicrobial resistance, Acinetobacter is one of the most common causes of nosocomial infections, especially VAP.Imipenem was recommended as the first line treatment of pneumonia caused by Acinetobacter baumannii, until its resistance occurred to most antibiotics including aminoglycosides, carbapenems and fluoroquinolones (8)(9)(10).In the 1950s antibiotics polymyxin B and E (also known as colistin) were introduced for the treatment of infections caused by Gram-negative bacilli, but even though they were highly effective, they fell out of favor in human medicine due to nephrotoxicity (11,12).Considering the increased presence of multidrug-resistant microorganisms (Acinetobacter baumannii, Klebsiella pneumoniae, Pseudomonas aeruginosa), and the lack of novel antibiotics, polymyxins emerged as the last-resort antibiotics for life threatening nosocomial infections in the 21 st century (13,14).Intravenous use of antibiotics is accepted as a gold standard for the treatment of pneumonia, but additional administration of inhaled antibiotics with systemic use in the treatment of VAP has shown to be advantageous in some clinical studies (15)(16)(17)(18).Even though the idea to enhance the antibiotic concentration in the lungs by inhalation is rational, there is not enough published reports to elucidate the benefits of such a root of administration (19)(20)(21).The studies related to this subject are scarce and have conflicting results.Despite the emerging colistin use, the recommendations for dosing regiments vary andthe beneficial effects of inhalatorytreatment remains insufficiently investigated (22,23).The aim of this study was to investigate the effect of inhalatory colistin as an adjunct to intravenous colistin on the survival of patients with VAP caused by Acinetobacter species.

Methods
A retrospective analysis was conducted in the period from January 2013 to March 2018.All ethical procedures were done in accordance with requirements of Institute for pulmonary diseases of Vojvodina (IPDV).The study included a total of 69 patients who were treated in the Intensive Care Unit (ICU) of the IPDV.Those 69 patients received colistin for the treatment of VAP caused by Acinetobacter.Colistin was administered in two ways, only intravenously or in combination, both inhalatory and intravenously.The experimental group consisted of 27 patients who received both intravenous and inhalatory colistin, while the control group consisted of 42 patients who received only intravenous colistin.The criteria for diagnosing VAP were based on recommendations for hospital-acquired pneumonia (HAP) and VAP from 2016 (5).The patients were mechanically ventilated for a minimum of 48 hours, with a new infiltration on the chest X-ray or a progression of already existing infiltration with two of the following three criteria: fever over 38.5°C or hypothermia below 35.5°C, leukocytosis > 10000 / μl or leukopenia < 4000 / μl and purulent endotracheal aspiration.Non-invasive sampling and semi-quantitative determination were performed to determine the microbiological cause.The significant noninvasive quantitative sampling value was ≥10 5 CFU / ml.If the sampling was invasive, with the quantitative determination of the causative agent, the threshold for the diagnosis of VAP was≥ 10 4 CFU / ml forbronchoalveolar lavage (5).
Baselinedemographic data and severity of illness (Acute Physiology and Chronic Health Evaluation (APACHE II)) (24), and Sequential Organ Failure Assessment(SOFA) scores (25), presence of acute respiratory distress syndrome (ARDS) (26), septic shock (27) and acute renal failure (defined by KDIGO -The Kidney Disease: Improving Global Outcomesdefinition) (28), comorbidities, colistinregimen (intravenous versus intravenous and inhalatory) and length of treatment were recorded.The primary outcome was 28 day mortality.For statistical analysis, continuous variables were presented as mean and standard deviations (SD), while categorical variables wereexpressed as whole numbers and percentages.The influence of different colistin protocols on 28 day mortality was investigated using binary logistic regression analysis.All predictors that were statistically significant in univariate analysis were entered into multivariate model.The final model included APACHE score, length of treatment and colistin regimen.Statistical significance for all variables was set on p value 0.05.All statistical tests were performed using SPSS version 21.

Discussion
The results of this study indicated thatintravenous treatment with colistin is associated with 6 fold increase in 28 days mortality compared to combined intravenous and inhalation colistin regimen (61,9% vs. 25,9% respectively, OR 6,305, 95% CI 1,795-22,153).The combined treatment resulted in prolonged length of hospital stay that was not statistical significant (35 vs. 27 days, p=0.07) .Literature search revealed a small quantity of published studies that investigated the relation of the inhalatory colistin addition to the monotherapy and their correlation with the 28 days mortality rate.Nevertheless, results from previous studies examining effects of the inhalatory colistin addition to the monotherapy treatment are conflicting (21,29,30).These discrepancies between published studies were explained in the conclusion of study by Tumbarello et al.where it is stated that their investigation was conducted on a substantially larger population (being the largest study so far with208 patients) and significant improvement of clinical cure rates were observed (30)(31)(32).These findings are in direct correlation with our investigation elucidating the substantial decrease in risk of ICU and 28 day mortality when a combined treatment was carried out.Moreover, Tumbarello et al. emphasized that an important role in further investigation should be to optimize the colistin use in order to enhance the efficacy without increasing the adverse renal effects.Additionally, it was stressed out that randomized controlled trials are a necessity for further clarification of the benefits and risks of the combined treatment (30).Earlier review studies indicated that major adverse effect of colistin use could be nephrotoxicity, but results were inconclusive and could not allow for a more significant conclusion concerning the correlation of nephrotoxicity to colistin use (33).These concerns have also been raised in recent publications for both intravenous and inhalatoryroute of administration,where no increase in nephrotoxicity was reportedwith inhaled colistin as adjunctive therapy, which is also in accordance with our findings (21,(34)(35)(36).The overall conclusion of these studies was that the inhaled colistin seem to be beneficial in the VAP therapy and can be considered as safe,even though limitations and drawbacks were observed, mainly as inconsistent and limited data.A more detailed investigation of colistin nephrotoxicity and neurotoxicity was recently reported in the study of Abdellatif et al.where renal safety was underlined as one of several benefits of aerosolized colistin regimen versus intravenous (37).It should be noted that the significant benefits of the colistin inhalotory enrollment in the combined therapy was recognized in the latest HAP and VAP guidelines of IDSA and ATS suggesting both inhaled and systemic antibiotics for patients with VAP, but with very low quality evidence (5).Therefore, the results from our study could contributeto stronger evidence essential for future guidelines as well as to the ongoing investigation of this therapeutic approach.Two studies out of nine that were cited in the mentioned guidelines directly concentrated their researchon the beneficial effects of the inhaled colistin combined with colistin monotherapy (36,38).Korbilaet al. concluded that the application of the inhaled colistin was an independent predictor of cure of VAP, but no difference in all-cause in-hospital mortality and all-cause ICU mortality was detected (36).Three years later Doshi et al.published their results obtained from three tertiary-care academic medical centers stating that the addition of aerosolized colistin to IV colistin may improve clinical cure and mortality for patients with MDR-GN pneumonia (38).These findings are in accordance with our results elucidating the hypothesis of our research.As previously mentioned, results obtained in our study showed that patients receiving only intravenous colistin had greater ICU mortality compared to group of patients who received combined intravenous and inhalatory colistin (24/42, 57.1% vs. 13/27, 48.1%, p=0.465).These results are in correlation with other studies comparing these two regimens of colistin administration, where collected data showed ICU mortality of 35.9-52.9% vs. 24-43.3%(21,30,36,38).The present study has some limitations that are very similar to the limitations stated in almost all previous investigation published on this subject.The limitations of our study are retrospective single-center nature, slight variations in the administration of the inhalatorycolistin as well as dosing variations.

Conclusion
Our study demonstrated that adjunct of inhalatory colistin to intravenous colistin may significantly decrease28 dayand ICU mortality in the treatment of VAP caused by Acinetobacter.Therefore, we suggest the use of the mentioned treatment approach.High quality randomized controlled multicenter trials are urgently needed to validate the additional benefits of inhaled colistin in this setting.